commit 179635f48e3990efd86f58df0dba9d2149892975 Author: ModelHub XC Date: Mon Sep 7 09:24:12 2026 +0800 初始化项目,由ModelHub XC社区提供模型 Model: laion/glm46-swesmith-maxeps-131k-fixthink Source: Original Platform diff --git a/.gitattributes b/.gitattributes new file mode 100644 index 0000000..b1699fd --- /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 + +training_args.bin filter=lfs diff=lfs merge=lfs -text +model-00002-of-00004.safetensors 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 +merges.txt filter=lfs diff=lfs merge=lfs -text +model-00004-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..dcbbc90 --- /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: glm46-swesmith-maxeps-131k-fixthink + results: [] +--- + + + +# glm46-swesmith-maxeps-131k-fixthink + +This model is a fine-tuned version of [Qwen/Qwen3-8B](https://huggingface.co/Qwen/Qwen3-8B) on the /data/cat/ws/befe330h-befe330h-otagent/huggingface/hub/datasets--penfever--glm46-swesmith-maxeps-131k/snapshots/4d4c2d4a9d21f73870ed31c7bc6028035b3b6ca7_thinking_preprocessed 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 4.57.6 +- Pytorch 2.9.0+cu128 +- Datasets 4.4.1 +- Tokenizers 0.22.2 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..209e331 --- /dev/null +++ b/all_results.json @@ -0,0 +1,16 @@ +{ + "achieved_tflops_per_gpu": 6.044664775062429, + "achieved_tflops_per_gpu_theoretical": 203.27427376939005, + "epoch": 7.0, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1423269808292389, + "mfu_percent": 0.6111895626958977, + "mfu_percent_theoretical": 20.55351605352781, + "total_flos": 2.6890471642266337e+18, + "train_loss": 0.1819671289975788, + "train_runtime": 55607.8638, + "train_samples_per_second": 0.959, + "train_steps_per_second": 0.06, + "valid_targets_mean": 5815.2, + "valid_targets_min": 1781 +} \ 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..59b4193 --- /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": "4.57.6", + "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..9adbb28 --- /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": "4.57.6" +} 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..b1cdb86 --- /dev/null +++ b/model-00001-of-00004.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d78554cd6a8387184f7acd4648eeefc7ef13576f39c50ca9b11d1b0c097333b1 +size 4902257696 diff --git a/model-00002-of-00004.safetensors b/model-00002-of-00004.safetensors new file mode 100644 index 0000000..86cd071 --- /dev/null +++ b/model-00002-of-00004.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6d99bb411de10393a99af38d0e7027f0836fe42ff2b1e04f341b908378573874 +size 4915960368 diff --git a/model-00003-of-00004.safetensors b/model-00003-of-00004.safetensors new file mode 100644 index 0000000..0ac2230 --- /dev/null +++ b/model-00003-of-00004.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d5561d90f3dce480734f13b6ea18c0b75b21a1c945ce511e47cafa1b4353fea2 +size 4983068496 diff --git a/model-00004-of-00004.safetensors b/model-00004-of-00004.safetensors new file mode 100644 index 0000000..94a26b6 --- /dev/null +++ b/model-00004-of-00004.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2b65fa64ecee457f78b29d0b1766fa9af6afb6b56b7aa2f46faffd357c812278 +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..608269e --- /dev/null +++ b/run_summary.json @@ -0,0 +1,12 @@ +{ + "agent_name": "4d4c2d4a9d21f73870ed31c7bc6028035b3b6ca7_thinking_preprocessed", + "training_start": null, + "training_end": null, + "created_by": "DCAgent", + "base_model_name": "Qwen/Qwen3-8B", + "dataset_name": "/data/cat/ws/befe330h-befe330h-otagent/huggingface/hub/datasets--penfever--glm46-swesmith-maxeps-131k/snapshots/4d4c2d4a9d21f73870ed31c7bc6028035b3b6ca7_thinking_preprocessed", + "training_type": "SFT", + "training_parameters": "https://huggingface.co/laion/glm46-swesmith-maxeps-131k-fixthink/blob/main/config.json", + "wandb_link": "https://wandb.ai/dogml/OpenThoughts-Agent/runs/sft_glm46-swesmith-maxeps-131k_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..209e331 --- /dev/null +++ b/train_results.json @@ -0,0 +1,16 @@ +{ + "achieved_tflops_per_gpu": 6.044664775062429, + "achieved_tflops_per_gpu_theoretical": 203.27427376939005, + "epoch": 7.0, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1423269808292389, + "mfu_percent": 0.6111895626958977, + "mfu_percent_theoretical": 20.55351605352781, + "total_flos": 2.6890471642266337e+18, + "train_loss": 0.1819671289975788, + "train_runtime": 55607.8638, + "train_samples_per_second": 0.959, + "train_steps_per_second": 0.06, + "valid_targets_mean": 5815.2, + "valid_targets_min": 1781 +} \ No newline at end of file diff --git a/trainer_log.jsonl b/trainer_log.jsonl new file mode 100644 index 0000000..e2dfcff --- /dev/null +++ b/trainer_log.jsonl @@ -0,0 +1,668 @@ +{"current_steps": 5, "total_steps": 3339, "loss": 0.5478, "lr": 4.790419161676648e-07, "epoch": 0.01049317943336831, "percentage": 0.15, "elapsed_time": "0:01:32", "remaining_time": "17:07:48"} +{"current_steps": 10, "total_steps": 3339, "loss": 0.5587, "lr": 1.0778443113772456e-06, "epoch": 0.02098635886673662, "percentage": 0.3, "elapsed_time": "0:02:57", "remaining_time": "16:24:48"} +{"current_steps": 15, "total_steps": 3339, "loss": 0.5301, "lr": 1.6766467065868263e-06, "epoch": 0.03147953830010493, "percentage": 0.45, "elapsed_time": "0:04:22", "remaining_time": "16:11:09"} +{"current_steps": 20, "total_steps": 3339, "loss": 0.5158, "lr": 2.2754491017964075e-06, "epoch": 0.04197271773347324, "percentage": 0.6, "elapsed_time": "0:05:48", "remaining_time": "16:05:16"} +{"current_steps": 25, "total_steps": 3339, "loss": 0.4672, "lr": 2.874251497005988e-06, "epoch": 0.05246589716684155, "percentage": 0.75, "elapsed_time": "0:07:16", "remaining_time": "16:04:32"} +{"current_steps": 30, "total_steps": 3339, "loss": 0.4327, "lr": 3.4730538922155693e-06, "epoch": 0.06295907660020986, "percentage": 0.9, "elapsed_time": "0:08:43", "remaining_time": "16:02:14"} +{"current_steps": 35, "total_steps": 3339, "loss": 0.4349, "lr": 4.07185628742515e-06, "epoch": 0.07345225603357818, "percentage": 1.05, "elapsed_time": "0:10:09", "remaining_time": "15:59:43"} +{"current_steps": 40, "total_steps": 3339, "loss": 0.385, "lr": 4.670658682634731e-06, "epoch": 0.08394543546694648, "percentage": 1.2, "elapsed_time": "0:11:36", "remaining_time": "15:57:36"} +{"current_steps": 45, "total_steps": 3339, "loss": 0.3846, "lr": 5.2694610778443125e-06, "epoch": 0.0944386149003148, "percentage": 1.35, "elapsed_time": "0:13:03", "remaining_time": "15:56:22"} +{"current_steps": 50, "total_steps": 3339, "loss": 0.3689, "lr": 5.868263473053893e-06, "epoch": 0.1049317943336831, "percentage": 1.5, "elapsed_time": "0:14:31", "remaining_time": "15:55:33"} +{"current_steps": 55, "total_steps": 3339, "loss": 0.364, "lr": 6.4670658682634736e-06, "epoch": 0.11542497376705142, "percentage": 1.65, "elapsed_time": "0:15:57", "remaining_time": "15:53:17"} +{"current_steps": 60, "total_steps": 3339, "loss": 0.3559, "lr": 7.0658682634730545e-06, "epoch": 0.1259181532004197, "percentage": 1.8, "elapsed_time": "0:17:23", "remaining_time": "15:50:42"} +{"current_steps": 65, "total_steps": 3339, "loss": 0.3527, "lr": 7.664670658682636e-06, "epoch": 0.13641133263378805, "percentage": 1.95, "elapsed_time": "0:18:50", "remaining_time": "15:48:42"} +{"current_steps": 70, "total_steps": 3339, "loss": 0.3472, "lr": 8.263473053892216e-06, "epoch": 0.14690451206715635, "percentage": 2.1, "elapsed_time": "0:20:15", "remaining_time": "15:45:46"} +{"current_steps": 75, "total_steps": 3339, "loss": 0.3256, "lr": 8.862275449101796e-06, "epoch": 0.15739769150052466, "percentage": 2.25, "elapsed_time": "0:21:39", "remaining_time": "15:42:29"} +{"current_steps": 80, "total_steps": 3339, "loss": 0.3088, "lr": 9.461077844311378e-06, "epoch": 0.16789087093389296, "percentage": 2.4, "elapsed_time": "0:23:05", "remaining_time": "15:40:55"} +{"current_steps": 85, "total_steps": 3339, "loss": 0.3132, "lr": 1.0059880239520958e-05, "epoch": 0.1783840503672613, "percentage": 2.55, "elapsed_time": "0:24:33", "remaining_time": "15:40:01"} +{"current_steps": 90, "total_steps": 3339, "loss": 0.3157, "lr": 1.065868263473054e-05, "epoch": 0.1888772298006296, "percentage": 2.7, "elapsed_time": "0:25:58", "remaining_time": "15:37:57"} +{"current_steps": 95, "total_steps": 3339, "loss": 0.3021, "lr": 1.125748502994012e-05, "epoch": 0.1993704092339979, "percentage": 2.85, "elapsed_time": "0:27:24", "remaining_time": "15:36:03"} +{"current_steps": 100, "total_steps": 3339, "loss": 0.2913, "lr": 1.1856287425149702e-05, "epoch": 0.2098635886673662, "percentage": 2.99, "elapsed_time": "0:28:50", "remaining_time": "15:34:03"} +{"current_steps": 105, "total_steps": 3339, "loss": 0.299, "lr": 1.2455089820359284e-05, "epoch": 0.2203567681007345, "percentage": 3.14, "elapsed_time": "0:30:15", "remaining_time": "15:31:48"} +{"current_steps": 110, "total_steps": 3339, "loss": 0.2922, "lr": 1.3053892215568864e-05, "epoch": 0.23084994753410285, "percentage": 3.29, "elapsed_time": "0:31:39", "remaining_time": "15:29:24"} +{"current_steps": 115, "total_steps": 3339, "loss": 0.2916, "lr": 1.3652694610778446e-05, "epoch": 0.24134312696747115, "percentage": 3.44, "elapsed_time": "0:33:05", "remaining_time": "15:27:38"} +{"current_steps": 120, "total_steps": 3339, "loss": 0.2873, "lr": 1.4251497005988024e-05, "epoch": 0.2518363064008394, "percentage": 3.59, "elapsed_time": "0:34:30", "remaining_time": "15:25:44"} +{"current_steps": 125, "total_steps": 3339, "loss": 0.2819, "lr": 1.4850299401197606e-05, "epoch": 0.2623294858342078, "percentage": 3.74, "elapsed_time": "0:35:56", "remaining_time": "15:24:16"} +{"current_steps": 130, "total_steps": 3339, "loss": 0.2735, "lr": 1.5449101796407188e-05, "epoch": 0.2728226652675761, "percentage": 3.89, "elapsed_time": "0:37:21", "remaining_time": "15:22:21"} +{"current_steps": 135, "total_steps": 3339, "loss": 0.2725, "lr": 1.604790419161677e-05, "epoch": 0.2833158447009444, "percentage": 4.04, "elapsed_time": "0:38:48", "remaining_time": "15:20:52"} +{"current_steps": 140, "total_steps": 3339, "loss": 0.2745, "lr": 1.6646706586826348e-05, "epoch": 0.2938090241343127, "percentage": 4.19, "elapsed_time": "0:40:14", "remaining_time": "15:19:32"} +{"current_steps": 145, "total_steps": 3339, "loss": 0.2756, "lr": 1.724550898203593e-05, "epoch": 0.304302203567681, "percentage": 4.34, "elapsed_time": "0:41:38", "remaining_time": "15:17:25"} +{"current_steps": 150, "total_steps": 3339, "loss": 0.262, "lr": 1.784431137724551e-05, "epoch": 0.3147953830010493, "percentage": 4.49, "elapsed_time": "0:43:05", "remaining_time": "15:15:57"} +{"current_steps": 155, "total_steps": 3339, "loss": 0.2752, "lr": 1.8443113772455093e-05, "epoch": 0.3252885624344176, "percentage": 4.64, "elapsed_time": "0:44:31", "remaining_time": "15:14:36"} +{"current_steps": 160, "total_steps": 3339, "loss": 0.2699, "lr": 1.904191616766467e-05, "epoch": 0.3357817418677859, "percentage": 4.79, "elapsed_time": "0:45:55", "remaining_time": "15:12:35"} +{"current_steps": 165, "total_steps": 3339, "loss": 0.2647, "lr": 1.9640718562874253e-05, "epoch": 0.3462749213011542, "percentage": 4.94, "elapsed_time": "0:47:19", "remaining_time": "15:10:20"} +{"current_steps": 170, "total_steps": 3339, "loss": 0.269, "lr": 2.0239520958083835e-05, "epoch": 0.3567681007345226, "percentage": 5.09, "elapsed_time": "0:48:46", "remaining_time": "15:09:05"} +{"current_steps": 175, "total_steps": 3339, "loss": 0.2593, "lr": 2.0838323353293413e-05, "epoch": 0.3672612801678909, "percentage": 5.24, "elapsed_time": "0:50:12", "remaining_time": "15:07:50"} +{"current_steps": 180, "total_steps": 3339, "loss": 0.2679, "lr": 2.1437125748502995e-05, "epoch": 0.3777544596012592, "percentage": 5.39, "elapsed_time": "0:51:37", "remaining_time": "15:06:02"} +{"current_steps": 185, "total_steps": 3339, "loss": 0.2668, "lr": 2.2035928143712577e-05, "epoch": 0.3882476390346275, "percentage": 5.54, "elapsed_time": "0:53:01", "remaining_time": "15:04:02"} +{"current_steps": 190, "total_steps": 3339, "loss": 0.2611, "lr": 2.2634730538922155e-05, "epoch": 0.3987408184679958, "percentage": 5.69, "elapsed_time": "0:54:26", "remaining_time": "15:02:24"} +{"current_steps": 195, "total_steps": 3339, "loss": 0.2534, "lr": 2.323353293413174e-05, "epoch": 0.4092339979013641, "percentage": 5.84, "elapsed_time": "0:55:49", "remaining_time": "15:00:02"} +{"current_steps": 200, "total_steps": 3339, "loss": 0.257, "lr": 2.383233532934132e-05, "epoch": 0.4197271773347324, "percentage": 5.99, "elapsed_time": "0:57:14", "remaining_time": "14:58:29"} +{"current_steps": 205, "total_steps": 3339, "loss": 0.2471, "lr": 2.44311377245509e-05, "epoch": 0.4302203567681007, "percentage": 6.14, "elapsed_time": "0:58:38", "remaining_time": "14:56:32"} +{"current_steps": 210, "total_steps": 3339, "loss": 0.2462, "lr": 2.502994011976048e-05, "epoch": 0.440713536201469, "percentage": 6.29, "elapsed_time": "1:00:02", "remaining_time": "14:54:34"} +{"current_steps": 215, "total_steps": 3339, "loss": 0.2499, "lr": 2.5628742514970064e-05, "epoch": 0.45120671563483733, "percentage": 6.44, "elapsed_time": "1:01:27", "remaining_time": "14:53:05"} +{"current_steps": 220, "total_steps": 3339, "loss": 0.2541, "lr": 2.6227544910179643e-05, "epoch": 0.4616998950682057, "percentage": 6.59, "elapsed_time": "1:02:53", "remaining_time": "14:51:31"} +{"current_steps": 225, "total_steps": 3339, "loss": 0.2519, "lr": 2.682634730538922e-05, "epoch": 0.472193074501574, "percentage": 6.74, "elapsed_time": "1:04:17", "remaining_time": "14:49:46"} +{"current_steps": 230, "total_steps": 3339, "loss": 0.2439, "lr": 2.7425149700598806e-05, "epoch": 0.4826862539349423, "percentage": 6.89, "elapsed_time": "1:05:42", "remaining_time": "14:48:08"} +{"current_steps": 235, "total_steps": 3339, "loss": 0.2436, "lr": 2.8023952095808385e-05, "epoch": 0.4931794333683106, "percentage": 7.04, "elapsed_time": "1:07:06", "remaining_time": "14:46:25"} +{"current_steps": 240, "total_steps": 3339, "loss": 0.2359, "lr": 2.8622754491017966e-05, "epoch": 0.5036726128016789, "percentage": 7.19, "elapsed_time": "1:08:31", "remaining_time": "14:44:47"} +{"current_steps": 245, "total_steps": 3339, "loss": 0.2377, "lr": 2.9221556886227548e-05, "epoch": 0.5141657922350472, "percentage": 7.34, "elapsed_time": "1:09:57", "remaining_time": "14:43:32"} +{"current_steps": 250, "total_steps": 3339, "loss": 0.2408, "lr": 2.982035928143713e-05, "epoch": 0.5246589716684156, "percentage": 7.49, "elapsed_time": "1:11:23", "remaining_time": "14:42:07"} +{"current_steps": 255, "total_steps": 3339, "loss": 0.24, "lr": 3.041916167664671e-05, "epoch": 0.5351521511017838, "percentage": 7.64, "elapsed_time": "1:12:49", "remaining_time": "14:40:40"} +{"current_steps": 260, "total_steps": 3339, "loss": 0.2371, "lr": 3.101796407185629e-05, "epoch": 0.5456453305351522, "percentage": 7.79, "elapsed_time": "1:14:13", "remaining_time": "14:38:55"} +{"current_steps": 265, "total_steps": 3339, "loss": 0.2438, "lr": 3.161676646706587e-05, "epoch": 0.5561385099685204, "percentage": 7.94, "elapsed_time": "1:15:36", "remaining_time": "14:37:07"} +{"current_steps": 270, "total_steps": 3339, "loss": 0.2372, "lr": 3.221556886227545e-05, "epoch": 0.5666316894018888, "percentage": 8.09, "elapsed_time": "1:17:01", "remaining_time": "14:35:28"} +{"current_steps": 275, "total_steps": 3339, "loss": 0.2386, "lr": 3.2814371257485035e-05, "epoch": 0.577124868835257, "percentage": 8.24, "elapsed_time": "1:18:27", "remaining_time": "14:34:08"} +{"current_steps": 280, "total_steps": 3339, "loss": 0.2417, "lr": 3.3413173652694614e-05, "epoch": 0.5876180482686254, "percentage": 8.39, "elapsed_time": "1:19:52", "remaining_time": "14:32:42"} +{"current_steps": 285, "total_steps": 3339, "loss": 0.2361, "lr": 3.40119760479042e-05, "epoch": 0.5981112277019937, "percentage": 8.54, "elapsed_time": "1:21:16", "remaining_time": "14:31:00"} +{"current_steps": 290, "total_steps": 3339, "loss": 0.2317, "lr": 3.461077844311378e-05, "epoch": 0.608604407135362, "percentage": 8.69, "elapsed_time": "1:22:42", "remaining_time": "14:29:30"} +{"current_steps": 295, "total_steps": 3339, "loss": 0.2427, "lr": 3.5209580838323356e-05, "epoch": 0.6190975865687304, "percentage": 8.83, "elapsed_time": "1:24:07", "remaining_time": "14:28:02"} +{"current_steps": 300, "total_steps": 3339, "loss": 0.2351, "lr": 3.5808383233532934e-05, "epoch": 0.6295907660020986, "percentage": 8.98, "elapsed_time": "1:25:32", "remaining_time": "14:26:28"} +{"current_steps": 305, "total_steps": 3339, "loss": 0.2384, "lr": 3.640718562874252e-05, "epoch": 0.640083945435467, "percentage": 9.13, "elapsed_time": "1:26:57", "remaining_time": "14:25:03"} +{"current_steps": 310, "total_steps": 3339, "loss": 0.2351, "lr": 3.70059880239521e-05, "epoch": 0.6505771248688352, "percentage": 9.28, "elapsed_time": "1:28:23", "remaining_time": "14:23:36"} +{"current_steps": 315, "total_steps": 3339, "loss": 0.2393, "lr": 3.7604790419161676e-05, "epoch": 0.6610703043022036, "percentage": 9.43, "elapsed_time": "1:29:47", "remaining_time": "14:22:00"} +{"current_steps": 320, "total_steps": 3339, "loss": 0.2268, "lr": 3.820359281437126e-05, "epoch": 0.6715634837355718, "percentage": 9.58, "elapsed_time": "1:31:10", "remaining_time": "14:20:08"} +{"current_steps": 325, "total_steps": 3339, "loss": 0.2302, "lr": 3.880239520958084e-05, "epoch": 0.6820566631689402, "percentage": 9.73, "elapsed_time": "1:32:34", "remaining_time": "14:18:32"} +{"current_steps": 330, "total_steps": 3339, "loss": 0.2331, "lr": 3.9401197604790425e-05, "epoch": 0.6925498426023085, "percentage": 9.88, "elapsed_time": "1:33:57", "remaining_time": "14:16:45"} +{"current_steps": 335, "total_steps": 3339, "loss": 0.2285, "lr": 4e-05, "epoch": 0.7030430220356768, "percentage": 10.03, "elapsed_time": "1:35:23", "remaining_time": "14:15:20"} +{"current_steps": 340, "total_steps": 3339, "loss": 0.2214, "lr": 3.9999726756517076e-05, "epoch": 0.7135362014690452, "percentage": 10.18, "elapsed_time": "1:36:47", "remaining_time": "14:13:49"} +{"current_steps": 345, "total_steps": 3339, "loss": 0.2242, "lr": 3.9998907033534495e-05, "epoch": 0.7240293809024134, "percentage": 10.33, "elapsed_time": "1:38:11", "remaining_time": "14:12:12"} +{"current_steps": 350, "total_steps": 3339, "loss": 0.2279, "lr": 3.999754085345065e-05, "epoch": 0.7345225603357818, "percentage": 10.48, "elapsed_time": "1:39:35", "remaining_time": "14:10:28"} +{"current_steps": 355, "total_steps": 3339, "loss": 0.2298, "lr": 3.9995628253595536e-05, "epoch": 0.74501573976915, "percentage": 10.63, "elapsed_time": "1:40:58", "remaining_time": "14:08:45"} +{"current_steps": 360, "total_steps": 3339, "loss": 0.2265, "lr": 3.999316928622968e-05, "epoch": 0.7555089192025184, "percentage": 10.78, "elapsed_time": "1:42:22", "remaining_time": "14:07:07"} +{"current_steps": 365, "total_steps": 3339, "loss": 0.2319, "lr": 3.999016401854278e-05, "epoch": 0.7660020986358866, "percentage": 10.93, "elapsed_time": "1:43:45", "remaining_time": "14:05:28"} +{"current_steps": 370, "total_steps": 3339, "loss": 0.2244, "lr": 3.9986612532651804e-05, "epoch": 0.776495278069255, "percentage": 11.08, "elapsed_time": "1:45:09", "remaining_time": "14:03:49"} +{"current_steps": 375, "total_steps": 3339, "loss": 0.2335, "lr": 3.9982514925598794e-05, "epoch": 0.7869884575026233, "percentage": 11.23, "elapsed_time": "1:46:34", "remaining_time": "14:02:22"} +{"current_steps": 380, "total_steps": 3339, "loss": 0.227, "lr": 3.997787130934819e-05, "epoch": 0.7974816369359916, "percentage": 11.38, "elapsed_time": "1:47:59", "remaining_time": "14:00:53"} +{"current_steps": 385, "total_steps": 3339, "loss": 0.2321, "lr": 3.997268181078379e-05, "epoch": 0.8079748163693599, "percentage": 11.53, "elapsed_time": "1:49:23", "remaining_time": "13:59:17"} +{"current_steps": 390, "total_steps": 3339, "loss": 0.2248, "lr": 3.9966946571705244e-05, "epoch": 0.8184679958027282, "percentage": 11.68, "elapsed_time": "1:50:47", "remaining_time": "13:57:48"} +{"current_steps": 395, "total_steps": 3339, "loss": 0.2195, "lr": 3.996066574882424e-05, "epoch": 0.8289611752360966, "percentage": 11.83, "elapsed_time": "1:52:11", "remaining_time": "13:56:07"} +{"current_steps": 400, "total_steps": 3339, "loss": 0.2246, "lr": 3.9953839513760154e-05, "epoch": 0.8394543546694648, "percentage": 11.98, "elapsed_time": "1:53:35", "remaining_time": "13:54:39"} +{"current_steps": 405, "total_steps": 3339, "loss": 0.2237, "lr": 3.9946468053035417e-05, "epoch": 0.8499475341028332, "percentage": 12.13, "elapsed_time": "1:55:00", "remaining_time": "13:53:10"} +{"current_steps": 410, "total_steps": 3339, "loss": 0.2254, "lr": 3.993855156807039e-05, "epoch": 0.8604407135362014, "percentage": 12.28, "elapsed_time": "1:56:23", "remaining_time": "13:51:30"} +{"current_steps": 415, "total_steps": 3339, "loss": 0.2205, "lr": 3.993009027517786e-05, "epoch": 0.8709338929695698, "percentage": 12.43, "elapsed_time": "1:57:47", "remaining_time": "13:49:58"} +{"current_steps": 420, "total_steps": 3339, "loss": 0.2191, "lr": 3.992108440555715e-05, "epoch": 0.881427072402938, "percentage": 12.58, "elapsed_time": "1:59:11", "remaining_time": "13:48:25"} +{"current_steps": 425, "total_steps": 3339, "loss": 0.2177, "lr": 3.991153420528777e-05, "epoch": 0.8919202518363064, "percentage": 12.73, "elapsed_time": "2:00:36", "remaining_time": "13:46:54"} +{"current_steps": 430, "total_steps": 3339, "loss": 0.225, "lr": 3.9901439935322726e-05, "epoch": 0.9024134312696747, "percentage": 12.88, "elapsed_time": "2:02:00", "remaining_time": "13:45:26"} +{"current_steps": 435, "total_steps": 3339, "loss": 0.2247, "lr": 3.989080187148136e-05, "epoch": 0.912906610703043, "percentage": 13.03, "elapsed_time": "2:03:25", "remaining_time": "13:43:55"} +{"current_steps": 440, "total_steps": 3339, "loss": 0.2196, "lr": 3.9879620304441834e-05, "epoch": 0.9233997901364114, "percentage": 13.18, "elapsed_time": "2:04:50", "remaining_time": "13:42:29"} +{"current_steps": 445, "total_steps": 3339, "loss": 0.2314, "lr": 3.986789553973319e-05, "epoch": 0.9338929695697796, "percentage": 13.33, "elapsed_time": "2:06:13", "remaining_time": "13:40:54"} +{"current_steps": 450, "total_steps": 3339, "loss": 0.2164, "lr": 3.985562789772697e-05, "epoch": 0.944386149003148, "percentage": 13.48, "elapsed_time": "2:07:37", "remaining_time": "13:39:20"} +{"current_steps": 455, "total_steps": 3339, "loss": 0.2144, "lr": 3.9842817713628494e-05, "epoch": 0.9548793284365162, "percentage": 13.63, "elapsed_time": "2:09:00", "remaining_time": "13:37:45"} +{"current_steps": 460, "total_steps": 3339, "loss": 0.2282, "lr": 3.982946533746771e-05, "epoch": 0.9653725078698846, "percentage": 13.78, "elapsed_time": "2:10:24", "remaining_time": "13:36:09"} +{"current_steps": 465, "total_steps": 3339, "loss": 0.2142, "lr": 3.9815571134089584e-05, "epoch": 0.9758656873032528, "percentage": 13.93, "elapsed_time": "2:11:47", "remaining_time": "13:34:33"} +{"current_steps": 470, "total_steps": 3339, "loss": 0.2206, "lr": 3.9801135483144174e-05, "epoch": 0.9863588667366212, "percentage": 14.08, "elapsed_time": "2:13:12", "remaining_time": "13:33:05"} +{"current_steps": 475, "total_steps": 3339, "loss": 0.2224, "lr": 3.9786158779076226e-05, "epoch": 0.9968520461699895, "percentage": 14.23, "elapsed_time": "2:14:36", "remaining_time": "13:31:39"} +{"current_steps": 480, "total_steps": 3339, "loss": 0.208, "lr": 3.977064143111443e-05, "epoch": 1.006295907660021, "percentage": 14.38, "elapsed_time": "2:15:51", "remaining_time": "13:29:09"} +{"current_steps": 485, "total_steps": 3339, "loss": 0.202, "lr": 3.9754583863260195e-05, "epoch": 1.0167890870933893, "percentage": 14.53, "elapsed_time": "2:17:15", "remaining_time": "13:27:41"} +{"current_steps": 490, "total_steps": 3339, "loss": 0.2074, "lr": 3.97379865142761e-05, "epoch": 1.0272822665267576, "percentage": 14.68, "elapsed_time": "2:18:40", "remaining_time": "13:26:14"} +{"current_steps": 495, "total_steps": 3339, "loss": 0.212, "lr": 3.97208498376739e-05, "epoch": 1.0377754459601258, "percentage": 14.82, "elapsed_time": "2:20:04", "remaining_time": "13:24:47"} +{"current_steps": 500, "total_steps": 3339, "loss": 0.2093, "lr": 3.97031743017021e-05, "epoch": 1.0482686253934943, "percentage": 14.97, "elapsed_time": "2:21:28", "remaining_time": "13:23:18"} +{"current_steps": 505, "total_steps": 3339, "loss": 0.2052, "lr": 3.9684960389333206e-05, "epoch": 1.0587618048268626, "percentage": 15.12, "elapsed_time": "2:22:52", "remaining_time": "13:21:45"} +{"current_steps": 510, "total_steps": 3339, "loss": 0.2151, "lr": 3.966620859825051e-05, "epoch": 1.0692549842602308, "percentage": 15.27, "elapsed_time": "2:24:14", "remaining_time": "13:20:09"} +{"current_steps": 515, "total_steps": 3339, "loss": 0.21, "lr": 3.964691944083448e-05, "epoch": 1.079748163693599, "percentage": 15.42, "elapsed_time": "2:25:38", "remaining_time": "13:18:39"} +{"current_steps": 520, "total_steps": 3339, "loss": 0.2127, "lr": 3.9627093444148766e-05, "epoch": 1.0902413431269675, "percentage": 15.57, "elapsed_time": "2:27:01", "remaining_time": "13:17:02"} +{"current_steps": 525, "total_steps": 3339, "loss": 0.2079, "lr": 3.960673114992581e-05, "epoch": 1.1007345225603358, "percentage": 15.72, "elapsed_time": "2:28:25", "remaining_time": "13:15:31"} +{"current_steps": 530, "total_steps": 3339, "loss": 0.2083, "lr": 3.958583311455203e-05, "epoch": 1.111227701993704, "percentage": 15.87, "elapsed_time": "2:29:49", "remaining_time": "13:14:02"} +{"current_steps": 535, "total_steps": 3339, "loss": 0.2104, "lr": 3.9564399909052626e-05, "epoch": 1.1217208814270725, "percentage": 16.02, "elapsed_time": "2:31:12", "remaining_time": "13:12:31"} +{"current_steps": 540, "total_steps": 3339, "loss": 0.2077, "lr": 3.954243211907596e-05, "epoch": 1.1322140608604407, "percentage": 16.17, "elapsed_time": "2:32:37", "remaining_time": "13:11:04"} +{"current_steps": 545, "total_steps": 3339, "loss": 0.2124, "lr": 3.9519930344877606e-05, "epoch": 1.142707240293809, "percentage": 16.32, "elapsed_time": "2:34:00", "remaining_time": "13:09:33"} +{"current_steps": 550, "total_steps": 3339, "loss": 0.2154, "lr": 3.949689520130384e-05, "epoch": 1.1532004197271772, "percentage": 16.47, "elapsed_time": "2:35:24", "remaining_time": "13:08:05"} +{"current_steps": 555, "total_steps": 3339, "loss": 0.2073, "lr": 3.947332731777498e-05, "epoch": 1.1636935991605457, "percentage": 16.62, "elapsed_time": "2:36:48", "remaining_time": "13:06:33"} +{"current_steps": 560, "total_steps": 3339, "loss": 0.2032, "lr": 3.944922733826806e-05, "epoch": 1.174186778593914, "percentage": 16.77, "elapsed_time": "2:38:11", "remaining_time": "13:05:02"} +{"current_steps": 565, "total_steps": 3339, "loss": 0.2117, "lr": 3.942459592129933e-05, "epoch": 1.1846799580272822, "percentage": 16.92, "elapsed_time": "2:39:34", "remaining_time": "13:03:29"} +{"current_steps": 570, "total_steps": 3339, "loss": 0.2063, "lr": 3.9399433739906195e-05, "epoch": 1.1951731374606505, "percentage": 17.07, "elapsed_time": "2:40:58", "remaining_time": "13:02:01"} +{"current_steps": 575, "total_steps": 3339, "loss": 0.1963, "lr": 3.937374148162888e-05, "epoch": 1.205666316894019, "percentage": 17.22, "elapsed_time": "2:42:21", "remaining_time": "13:00:27"} +{"current_steps": 580, "total_steps": 3339, "loss": 0.2086, "lr": 3.934751984849158e-05, "epoch": 1.2161594963273872, "percentage": 17.37, "elapsed_time": "2:43:44", "remaining_time": "12:58:53"} +{"current_steps": 585, "total_steps": 3339, "loss": 0.2049, "lr": 3.932076955698334e-05, "epoch": 1.2266526757607554, "percentage": 17.52, "elapsed_time": "2:45:07", "remaining_time": "12:57:21"} +{"current_steps": 590, "total_steps": 3339, "loss": 0.2054, "lr": 3.929349133803844e-05, "epoch": 1.237145855194124, "percentage": 17.67, "elapsed_time": "2:46:30", "remaining_time": "12:55:50"} +{"current_steps": 595, "total_steps": 3339, "loss": 0.1986, "lr": 3.926568593701644e-05, "epoch": 1.2476390346274921, "percentage": 17.82, "elapsed_time": "2:47:54", "remaining_time": "12:54:20"} +{"current_steps": 600, "total_steps": 3339, "loss": 0.2041, "lr": 3.9237354113681795e-05, "epoch": 1.2581322140608604, "percentage": 17.97, "elapsed_time": "2:49:17", "remaining_time": "12:52:49"} +{"current_steps": 605, "total_steps": 3339, "loss": 0.2043, "lr": 3.920849664218312e-05, "epoch": 1.2686253934942289, "percentage": 18.12, "elapsed_time": "2:50:41", "remaining_time": "12:51:22"} +{"current_steps": 610, "total_steps": 3339, "loss": 0.2038, "lr": 3.917911431103201e-05, "epoch": 1.2791185729275971, "percentage": 18.27, "elapsed_time": "2:52:05", "remaining_time": "12:49:52"} +{"current_steps": 615, "total_steps": 3339, "loss": 0.2087, "lr": 3.914920792308153e-05, "epoch": 1.2896117523609654, "percentage": 18.42, "elapsed_time": "2:53:29", "remaining_time": "12:48:27"} +{"current_steps": 620, "total_steps": 3339, "loss": 0.2088, "lr": 3.9118778295504224e-05, "epoch": 1.3001049317943336, "percentage": 18.57, "elapsed_time": "2:54:52", "remaining_time": "12:46:55"} +{"current_steps": 625, "total_steps": 3339, "loss": 0.2122, "lr": 3.908782625976984e-05, "epoch": 1.3105981112277019, "percentage": 18.72, "elapsed_time": "2:56:16", "remaining_time": "12:45:27"} +{"current_steps": 630, "total_steps": 3339, "loss": 0.2086, "lr": 3.905635266162258e-05, "epoch": 1.3210912906610703, "percentage": 18.87, "elapsed_time": "2:57:39", "remaining_time": "12:43:57"} +{"current_steps": 635, "total_steps": 3339, "loss": 0.206, "lr": 3.9024358361058e-05, "epoch": 1.3315844700944386, "percentage": 19.02, "elapsed_time": "2:59:03", "remaining_time": "12:42:26"} +{"current_steps": 640, "total_steps": 3339, "loss": 0.2027, "lr": 3.8991844232299526e-05, "epoch": 1.3420776495278068, "percentage": 19.17, "elapsed_time": "3:00:27", "remaining_time": "12:41:00"} +{"current_steps": 645, "total_steps": 3339, "loss": 0.2089, "lr": 3.8958811163774526e-05, "epoch": 1.3525708289611753, "percentage": 19.32, "elapsed_time": "3:01:51", "remaining_time": "12:39:34"} +{"current_steps": 650, "total_steps": 3339, "loss": 0.2011, "lr": 3.892526005809007e-05, "epoch": 1.3630640083945436, "percentage": 19.47, "elapsed_time": "3:03:14", "remaining_time": "12:38:03"} +{"current_steps": 655, "total_steps": 3339, "loss": 0.2038, "lr": 3.8891191832008256e-05, "epoch": 1.3735571878279118, "percentage": 19.62, "elapsed_time": "3:04:38", "remaining_time": "12:36:37"} +{"current_steps": 660, "total_steps": 3339, "loss": 0.209, "lr": 3.885660741642116e-05, "epoch": 1.3840503672612803, "percentage": 19.77, "elapsed_time": "3:06:01", "remaining_time": "12:35:06"} +{"current_steps": 665, "total_steps": 3339, "loss": 0.2093, "lr": 3.882150775632539e-05, "epoch": 1.3945435466946485, "percentage": 19.92, "elapsed_time": "3:07:24", "remaining_time": "12:33:35"} +{"current_steps": 670, "total_steps": 3339, "loss": 0.2036, "lr": 3.87858938107963e-05, "epoch": 1.4050367261280168, "percentage": 20.07, "elapsed_time": "3:08:48", "remaining_time": "12:32:06"} +{"current_steps": 675, "total_steps": 3339, "loss": 0.1979, "lr": 3.8749766552961736e-05, "epoch": 1.415529905561385, "percentage": 20.22, "elapsed_time": "3:10:11", "remaining_time": "12:30:35"} +{"current_steps": 680, "total_steps": 3339, "loss": 0.2041, "lr": 3.871312696997547e-05, "epoch": 1.4260230849947533, "percentage": 20.37, "elapsed_time": "3:11:34", "remaining_time": "12:29:05"} +{"current_steps": 685, "total_steps": 3339, "loss": 0.2037, "lr": 3.867597606299023e-05, "epoch": 1.4365162644281217, "percentage": 20.52, "elapsed_time": "3:12:55", "remaining_time": "12:27:30"} +{"current_steps": 690, "total_steps": 3339, "loss": 0.1972, "lr": 3.863831484713033e-05, "epoch": 1.44700944386149, "percentage": 20.66, "elapsed_time": "3:14:18", "remaining_time": "12:25:59"} +{"current_steps": 695, "total_steps": 3339, "loss": 0.2015, "lr": 3.860014435146397e-05, "epoch": 1.4575026232948582, "percentage": 20.81, "elapsed_time": "3:15:42", "remaining_time": "12:24:32"} +{"current_steps": 700, "total_steps": 3339, "loss": 0.1951, "lr": 3.856146561897505e-05, "epoch": 1.4679958027282267, "percentage": 20.96, "elapsed_time": "3:17:06", "remaining_time": "12:23:07"} +{"current_steps": 705, "total_steps": 3339, "loss": 0.2097, "lr": 3.852227970653473e-05, "epoch": 1.478488982161595, "percentage": 21.11, "elapsed_time": "3:18:29", "remaining_time": "12:21:37"} +{"current_steps": 710, "total_steps": 3339, "loss": 0.1968, "lr": 3.8482587684872536e-05, "epoch": 1.4889821615949632, "percentage": 21.26, "elapsed_time": "3:19:53", "remaining_time": "12:20:09"} +{"current_steps": 715, "total_steps": 3339, "loss": 0.2097, "lr": 3.844239063854709e-05, "epoch": 1.4994753410283317, "percentage": 21.41, "elapsed_time": "3:21:16", "remaining_time": "12:18:39"} +{"current_steps": 720, "total_steps": 3339, "loss": 0.1965, "lr": 3.8401689665916485e-05, "epoch": 1.5099685204617, "percentage": 21.56, "elapsed_time": "3:22:40", "remaining_time": "12:17:13"} +{"current_steps": 725, "total_steps": 3339, "loss": 0.2111, "lr": 3.836048587910828e-05, "epoch": 1.5204616998950682, "percentage": 21.71, "elapsed_time": "3:24:04", "remaining_time": "12:15:47"} +{"current_steps": 730, "total_steps": 3339, "loss": 0.201, "lr": 3.831878040398908e-05, "epoch": 1.5309548793284367, "percentage": 21.86, "elapsed_time": "3:25:29", "remaining_time": "12:14:24"} +{"current_steps": 735, "total_steps": 3339, "loss": 0.2025, "lr": 3.8276574380133826e-05, "epoch": 1.5414480587618047, "percentage": 22.01, "elapsed_time": "3:26:52", "remaining_time": "12:12:56"} +{"current_steps": 740, "total_steps": 3339, "loss": 0.2008, "lr": 3.823386896079462e-05, "epoch": 1.5519412381951732, "percentage": 22.16, "elapsed_time": "3:28:15", "remaining_time": "12:11:27"} +{"current_steps": 745, "total_steps": 3339, "loss": 0.2024, "lr": 3.8190665312869194e-05, "epoch": 1.5624344176285414, "percentage": 22.31, "elapsed_time": "3:29:39", "remaining_time": "12:10:00"} +{"current_steps": 750, "total_steps": 3339, "loss": 0.2052, "lr": 3.814696461686909e-05, "epoch": 1.5729275970619097, "percentage": 22.46, "elapsed_time": "3:31:04", "remaining_time": "12:08:39"} +{"current_steps": 755, "total_steps": 3339, "loss": 0.2014, "lr": 3.8102768066887335e-05, "epoch": 1.5834207764952781, "percentage": 22.61, "elapsed_time": "3:32:29", "remaining_time": "12:07:16"} +{"current_steps": 760, "total_steps": 3339, "loss": 0.1936, "lr": 3.805807687056586e-05, "epoch": 1.5939139559286464, "percentage": 22.76, "elapsed_time": "3:33:52", "remaining_time": "12:05:46"} +{"current_steps": 765, "total_steps": 3339, "loss": 0.2022, "lr": 3.801289224906248e-05, "epoch": 1.6044071353620146, "percentage": 22.91, "elapsed_time": "3:35:16", "remaining_time": "12:04:19"} +{"current_steps": 770, "total_steps": 3339, "loss": 0.1975, "lr": 3.7967215437017514e-05, "epoch": 1.614900314795383, "percentage": 23.06, "elapsed_time": "3:36:41", "remaining_time": "12:02:57"} +{"current_steps": 775, "total_steps": 3339, "loss": 0.2026, "lr": 3.792104768252011e-05, "epoch": 1.6253934942287513, "percentage": 23.21, "elapsed_time": "3:38:05", "remaining_time": "12:01:31"} +{"current_steps": 780, "total_steps": 3339, "loss": 0.1994, "lr": 3.787439024707405e-05, "epoch": 1.6358866736621196, "percentage": 23.36, "elapsed_time": "3:39:29", "remaining_time": "12:00:07"} +{"current_steps": 785, "total_steps": 3339, "loss": 0.2034, "lr": 3.782724440556336e-05, "epoch": 1.646379853095488, "percentage": 23.51, "elapsed_time": "3:40:54", "remaining_time": "11:58:42"} +{"current_steps": 790, "total_steps": 3339, "loss": 0.1927, "lr": 3.7779611446217434e-05, "epoch": 1.656873032528856, "percentage": 23.66, "elapsed_time": "3:42:17", "remaining_time": "11:57:12"} +{"current_steps": 795, "total_steps": 3339, "loss": 0.2028, "lr": 3.7731492670575846e-05, "epoch": 1.6673662119622246, "percentage": 23.81, "elapsed_time": "3:43:40", "remaining_time": "11:55:46"} +{"current_steps": 800, "total_steps": 3339, "loss": 0.2084, "lr": 3.7682889393452775e-05, "epoch": 1.6778593913955928, "percentage": 23.96, "elapsed_time": "3:45:05", "remaining_time": "11:54:22"} +{"current_steps": 805, "total_steps": 3339, "loss": 0.2002, "lr": 3.763380294290109e-05, "epoch": 1.688352570828961, "percentage": 24.11, "elapsed_time": "3:46:29", "remaining_time": "11:52:57"} +{"current_steps": 810, "total_steps": 3339, "loss": 0.1965, "lr": 3.7584234660176074e-05, "epoch": 1.6988457502623295, "percentage": 24.26, "elapsed_time": "3:47:53", "remaining_time": "11:51:31"} +{"current_steps": 815, "total_steps": 3339, "loss": 0.2015, "lr": 3.753418589969874e-05, "epoch": 1.7093389296956978, "percentage": 24.41, "elapsed_time": "3:49:17", "remaining_time": "11:50:06"} +{"current_steps": 820, "total_steps": 3339, "loss": 0.1976, "lr": 3.7483658029018854e-05, "epoch": 1.719832109129066, "percentage": 24.56, "elapsed_time": "3:50:40", "remaining_time": "11:48:38"} +{"current_steps": 825, "total_steps": 3339, "loss": 0.1963, "lr": 3.743265242877756e-05, "epoch": 1.7303252885624345, "percentage": 24.71, "elapsed_time": "3:52:03", "remaining_time": "11:47:08"} +{"current_steps": 830, "total_steps": 3339, "loss": 0.2063, "lr": 3.7381170492669624e-05, "epoch": 1.7408184679958028, "percentage": 24.86, "elapsed_time": "3:53:25", "remaining_time": "11:45:38"} +{"current_steps": 835, "total_steps": 3339, "loss": 0.2027, "lr": 3.732921362740541e-05, "epoch": 1.751311647429171, "percentage": 25.01, "elapsed_time": "3:54:48", "remaining_time": "11:44:09"} +{"current_steps": 840, "total_steps": 3339, "loss": 0.2016, "lr": 3.727678325267241e-05, "epoch": 1.7618048268625395, "percentage": 25.16, "elapsed_time": "3:56:10", "remaining_time": "11:42:37"} +{"current_steps": 845, "total_steps": 3339, "loss": 0.2047, "lr": 3.722388080109643e-05, "epoch": 1.7722980062959075, "percentage": 25.31, "elapsed_time": "3:57:32", "remaining_time": "11:41:06"} +{"current_steps": 850, "total_steps": 3339, "loss": 0.194, "lr": 3.7170507718202494e-05, "epoch": 1.782791185729276, "percentage": 25.46, "elapsed_time": "3:58:55", "remaining_time": "11:39:38"} +{"current_steps": 855, "total_steps": 3339, "loss": 0.2033, "lr": 3.71166654623753e-05, "epoch": 1.7932843651626444, "percentage": 25.61, "elapsed_time": "4:00:19", "remaining_time": "11:38:12"} +{"current_steps": 860, "total_steps": 3339, "loss": 0.197, "lr": 3.70623555048194e-05, "epoch": 1.8037775445960125, "percentage": 25.76, "elapsed_time": "4:01:42", "remaining_time": "11:36:43"} +{"current_steps": 865, "total_steps": 3339, "loss": 0.1897, "lr": 3.700757932951899e-05, "epoch": 1.814270724029381, "percentage": 25.91, "elapsed_time": "4:03:05", "remaining_time": "11:35:15"} +{"current_steps": 870, "total_steps": 3339, "loss": 0.1914, "lr": 3.695233843319737e-05, "epoch": 1.8247639034627492, "percentage": 26.06, "elapsed_time": "4:04:27", "remaining_time": "11:33:44"} +{"current_steps": 875, "total_steps": 3339, "loss": 0.204, "lr": 3.689663432527602e-05, "epoch": 1.8352570828961174, "percentage": 26.21, "elapsed_time": "4:05:51", "remaining_time": "11:32:19"} +{"current_steps": 880, "total_steps": 3339, "loss": 0.1942, "lr": 3.684046852783339e-05, "epoch": 1.845750262329486, "percentage": 26.36, "elapsed_time": "4:07:14", "remaining_time": "11:30:52"} +{"current_steps": 885, "total_steps": 3339, "loss": 0.1938, "lr": 3.67838425755633e-05, "epoch": 1.8562434417628542, "percentage": 26.5, "elapsed_time": "4:08:39", "remaining_time": "11:29:29"} +{"current_steps": 890, "total_steps": 3339, "loss": 0.191, "lr": 3.672675801573298e-05, "epoch": 1.8667366211962224, "percentage": 26.65, "elapsed_time": "4:10:02", "remaining_time": "11:28:03"} +{"current_steps": 895, "total_steps": 3339, "loss": 0.1867, "lr": 3.6669216408140835e-05, "epoch": 1.8772298006295909, "percentage": 26.8, "elapsed_time": "4:11:26", "remaining_time": "11:26:37"} +{"current_steps": 900, "total_steps": 3339, "loss": 0.2065, "lr": 3.661121932507378e-05, "epoch": 1.887722980062959, "percentage": 26.95, "elapsed_time": "4:12:50", "remaining_time": "11:25:11"} +{"current_steps": 905, "total_steps": 3339, "loss": 0.197, "lr": 3.655276835126432e-05, "epoch": 1.8982161594963274, "percentage": 27.1, "elapsed_time": "4:14:14", "remaining_time": "11:23:47"} +{"current_steps": 910, "total_steps": 3339, "loss": 0.1985, "lr": 3.6493865083847226e-05, "epoch": 1.9087093389296959, "percentage": 27.25, "elapsed_time": "4:15:33", "remaining_time": "11:22:09"} +{"current_steps": 915, "total_steps": 3339, "loss": 0.1969, "lr": 3.6434511132315886e-05, "epoch": 1.9192025183630639, "percentage": 27.4, "elapsed_time": "4:16:57", "remaining_time": "11:20:43"} +{"current_steps": 920, "total_steps": 3339, "loss": 0.1957, "lr": 3.637470811847834e-05, "epoch": 1.9296956977964324, "percentage": 27.55, "elapsed_time": "4:18:21", "remaining_time": "11:19:17"} +{"current_steps": 925, "total_steps": 3339, "loss": 0.1945, "lr": 3.631445767641298e-05, "epoch": 1.9401888772298006, "percentage": 27.7, "elapsed_time": "4:19:43", "remaining_time": "11:17:49"} +{"current_steps": 930, "total_steps": 3339, "loss": 0.1995, "lr": 3.625376145242386e-05, "epoch": 1.9506820566631689, "percentage": 27.85, "elapsed_time": "4:21:08", "remaining_time": "11:16:27"} +{"current_steps": 935, "total_steps": 3339, "loss": 0.2035, "lr": 3.619262110499575e-05, "epoch": 1.9611752360965373, "percentage": 28.0, "elapsed_time": "4:22:31", "remaining_time": "11:14:58"} +{"current_steps": 940, "total_steps": 3339, "loss": 0.1964, "lr": 3.6131038304748794e-05, "epoch": 1.9716684155299056, "percentage": 28.15, "elapsed_time": "4:23:54", "remaining_time": "11:13:30"} +{"current_steps": 945, "total_steps": 3339, "loss": 0.205, "lr": 3.6069014734392874e-05, "epoch": 1.9821615949632738, "percentage": 28.3, "elapsed_time": "4:25:17", "remaining_time": "11:12:04"} +{"current_steps": 950, "total_steps": 3339, "loss": 0.1964, "lr": 3.600655208868163e-05, "epoch": 1.9926547743966423, "percentage": 28.45, "elapsed_time": "4:26:38", "remaining_time": "11:10:33"} +{"current_steps": 955, "total_steps": 3339, "loss": 0.1905, "lr": 3.594365207436615e-05, "epoch": 2.0020986358866737, "percentage": 28.6, "elapsed_time": "4:27:54", "remaining_time": "11:08:47"} +{"current_steps": 960, "total_steps": 3339, "loss": 0.1841, "lr": 3.588031641014833e-05, "epoch": 2.012591815320042, "percentage": 28.75, "elapsed_time": "4:29:17", "remaining_time": "11:07:21"} +{"current_steps": 965, "total_steps": 3339, "loss": 0.1863, "lr": 3.5816546826633925e-05, "epoch": 2.02308499475341, "percentage": 28.9, "elapsed_time": "4:30:41", "remaining_time": "11:05:56"} +{"current_steps": 970, "total_steps": 3339, "loss": 0.1797, "lr": 3.575234506628524e-05, "epoch": 2.0335781741867787, "percentage": 29.05, "elapsed_time": "4:32:06", "remaining_time": "11:04:32"} +{"current_steps": 975, "total_steps": 3339, "loss": 0.1827, "lr": 3.568771288337353e-05, "epoch": 2.0440713536201467, "percentage": 29.2, "elapsed_time": "4:33:29", "remaining_time": "11:03:06"} +{"current_steps": 980, "total_steps": 3339, "loss": 0.1853, "lr": 3.562265204393108e-05, "epoch": 2.054564533053515, "percentage": 29.35, "elapsed_time": "4:34:54", "remaining_time": "11:01:45"} +{"current_steps": 985, "total_steps": 3339, "loss": 0.1818, "lr": 3.555716432570293e-05, "epoch": 2.0650577124868836, "percentage": 29.5, "elapsed_time": "4:36:18", "remaining_time": "11:00:19"} +{"current_steps": 990, "total_steps": 3339, "loss": 0.1886, "lr": 3.5491251518098296e-05, "epoch": 2.0755508919202517, "percentage": 29.65, "elapsed_time": "4:37:41", "remaining_time": "10:58:53"} +{"current_steps": 995, "total_steps": 3339, "loss": 0.1843, "lr": 3.542491542214169e-05, "epoch": 2.08604407135362, "percentage": 29.8, "elapsed_time": "4:39:04", "remaining_time": "10:57:25"} +{"current_steps": 1000, "total_steps": 3339, "loss": 0.1822, "lr": 3.535815785042371e-05, "epoch": 2.0965372507869886, "percentage": 29.95, "elapsed_time": "4:40:28", "remaining_time": "10:56:00"} +{"current_steps": 1005, "total_steps": 3339, "loss": 0.1872, "lr": 3.529098062705148e-05, "epoch": 2.1070304302203566, "percentage": 30.1, "elapsed_time": "4:41:50", "remaining_time": "10:54:33"} +{"current_steps": 1010, "total_steps": 3339, "loss": 0.1852, "lr": 3.522338558759886e-05, "epoch": 2.117523609653725, "percentage": 30.25, "elapsed_time": "4:43:15", "remaining_time": "10:53:10"} +{"current_steps": 1015, "total_steps": 3339, "loss": 0.1792, "lr": 3.5155374579056254e-05, "epoch": 2.1280167890870936, "percentage": 30.4, "elapsed_time": "4:44:38", "remaining_time": "10:51:42"} +{"current_steps": 1020, "total_steps": 3339, "loss": 0.181, "lr": 3.508694945978015e-05, "epoch": 2.1385099685204616, "percentage": 30.55, "elapsed_time": "4:46:01", "remaining_time": "10:50:16"} +{"current_steps": 1025, "total_steps": 3339, "loss": 0.179, "lr": 3.501811209944232e-05, "epoch": 2.14900314795383, "percentage": 30.7, "elapsed_time": "4:47:25", "remaining_time": "10:48:51"} +{"current_steps": 1030, "total_steps": 3339, "loss": 0.1888, "lr": 3.494886437897879e-05, "epoch": 2.159496327387198, "percentage": 30.85, "elapsed_time": "4:48:48", "remaining_time": "10:47:26"} +{"current_steps": 1035, "total_steps": 3339, "loss": 0.1923, "lr": 3.4879208190538386e-05, "epoch": 2.1699895068205666, "percentage": 31.0, "elapsed_time": "4:50:12", "remaining_time": "10:46:02"} +{"current_steps": 1040, "total_steps": 3339, "loss": 0.186, "lr": 3.4809145437431065e-05, "epoch": 2.180482686253935, "percentage": 31.15, "elapsed_time": "4:51:36", "remaining_time": "10:44:36"} +{"current_steps": 1045, "total_steps": 3339, "loss": 0.1873, "lr": 3.473867803407589e-05, "epoch": 2.190975865687303, "percentage": 31.3, "elapsed_time": "4:53:05", "remaining_time": "10:43:24"} +{"current_steps": 1050, "total_steps": 3339, "loss": 0.1824, "lr": 3.466780790594875e-05, "epoch": 2.2014690451206715, "percentage": 31.45, "elapsed_time": "4:54:30", "remaining_time": "10:42:00"} +{"current_steps": 1055, "total_steps": 3339, "loss": 0.181, "lr": 3.4596536989529685e-05, "epoch": 2.21196222455404, "percentage": 31.6, "elapsed_time": "4:55:53", "remaining_time": "10:40:34"} +{"current_steps": 1060, "total_steps": 3339, "loss": 0.1819, "lr": 3.452486723225005e-05, "epoch": 2.222455403987408, "percentage": 31.75, "elapsed_time": "4:57:16", "remaining_time": "10:39:09"} +{"current_steps": 1065, "total_steps": 3339, "loss": 0.1906, "lr": 3.445280059243926e-05, "epoch": 2.2329485834207765, "percentage": 31.9, "elapsed_time": "4:58:39", "remaining_time": "10:37:42"} +{"current_steps": 1070, "total_steps": 3339, "loss": 0.1887, "lr": 3.438033903927128e-05, "epoch": 2.243441762854145, "percentage": 32.05, "elapsed_time": "5:00:03", "remaining_time": "10:36:17"} +{"current_steps": 1075, "total_steps": 3339, "loss": 0.1863, "lr": 3.430748455271082e-05, "epoch": 2.253934942287513, "percentage": 32.2, "elapsed_time": "5:01:26", "remaining_time": "10:34:50"} +{"current_steps": 1080, "total_steps": 3339, "loss": 0.1873, "lr": 3.423423912345925e-05, "epoch": 2.2644281217208815, "percentage": 32.35, "elapsed_time": "5:02:50", "remaining_time": "10:33:26"} +{"current_steps": 1085, "total_steps": 3339, "loss": 0.186, "lr": 3.416060475290019e-05, "epoch": 2.27492130115425, "percentage": 32.49, "elapsed_time": "5:04:15", "remaining_time": "10:32:03"} +{"current_steps": 1090, "total_steps": 3339, "loss": 0.1848, "lr": 3.408658345304482e-05, "epoch": 2.285414480587618, "percentage": 32.64, "elapsed_time": "5:05:38", "remaining_time": "10:30:37"} +{"current_steps": 1095, "total_steps": 3339, "loss": 0.1785, "lr": 3.401217724647693e-05, "epoch": 2.2959076600209865, "percentage": 32.79, "elapsed_time": "5:07:01", "remaining_time": "10:29:10"} +{"current_steps": 1100, "total_steps": 3339, "loss": 0.1868, "lr": 3.393738816629761e-05, "epoch": 2.3064008394543545, "percentage": 32.94, "elapsed_time": "5:08:24", "remaining_time": "10:27:45"} +{"current_steps": 1105, "total_steps": 3339, "loss": 0.1809, "lr": 3.386221825606975e-05, "epoch": 2.316894018887723, "percentage": 33.09, "elapsed_time": "5:09:47", "remaining_time": "10:26:18"} +{"current_steps": 1110, "total_steps": 3339, "loss": 0.1852, "lr": 3.378666956976215e-05, "epoch": 2.3273871983210914, "percentage": 33.24, "elapsed_time": "5:11:10", "remaining_time": "10:24:52"} +{"current_steps": 1115, "total_steps": 3339, "loss": 0.1842, "lr": 3.371074417169343e-05, "epoch": 2.3378803777544594, "percentage": 33.39, "elapsed_time": "5:12:34", "remaining_time": "10:23:28"} +{"current_steps": 1120, "total_steps": 3339, "loss": 0.1769, "lr": 3.363444413647561e-05, "epoch": 2.348373557187828, "percentage": 33.54, "elapsed_time": "5:13:58", "remaining_time": "10:22:02"} +{"current_steps": 1125, "total_steps": 3339, "loss": 0.1861, "lr": 3.3557771548957425e-05, "epoch": 2.3588667366211964, "percentage": 33.69, "elapsed_time": "5:15:21", "remaining_time": "10:20:38"} +{"current_steps": 1130, "total_steps": 3339, "loss": 0.182, "lr": 3.348072850416737e-05, "epoch": 2.3693599160545644, "percentage": 33.84, "elapsed_time": "5:16:46", "remaining_time": "10:19:14"} +{"current_steps": 1135, "total_steps": 3339, "loss": 0.1909, "lr": 3.340331710725641e-05, "epoch": 2.379853095487933, "percentage": 33.99, "elapsed_time": "5:18:10", "remaining_time": "10:17:50"} +{"current_steps": 1140, "total_steps": 3339, "loss": 0.1765, "lr": 3.3325539473440546e-05, "epoch": 2.390346274921301, "percentage": 34.14, "elapsed_time": "5:19:32", "remaining_time": "10:16:22"} +{"current_steps": 1145, "total_steps": 3339, "loss": 0.1886, "lr": 3.324739772794292e-05, "epoch": 2.4008394543546694, "percentage": 34.29, "elapsed_time": "5:20:57", "remaining_time": "10:14:59"} +{"current_steps": 1150, "total_steps": 3339, "loss": 0.1783, "lr": 3.316889400593581e-05, "epoch": 2.411332633788038, "percentage": 34.44, "elapsed_time": "5:22:20", "remaining_time": "10:13:34"} +{"current_steps": 1155, "total_steps": 3339, "loss": 0.1764, "lr": 3.309003045248224e-05, "epoch": 2.421825813221406, "percentage": 34.59, "elapsed_time": "5:23:44", "remaining_time": "10:12:09"} +{"current_steps": 1160, "total_steps": 3339, "loss": 0.1782, "lr": 3.301080922247743e-05, "epoch": 2.4323189926547744, "percentage": 34.74, "elapsed_time": "5:25:06", "remaining_time": "10:10:42"} +{"current_steps": 1165, "total_steps": 3339, "loss": 0.1666, "lr": 3.2931232480589853e-05, "epoch": 2.442812172088143, "percentage": 34.89, "elapsed_time": "5:26:30", "remaining_time": "10:09:18"} +{"current_steps": 1170, "total_steps": 3339, "loss": 0.1764, "lr": 3.285130240120212e-05, "epoch": 2.453305351521511, "percentage": 35.04, "elapsed_time": "5:27:55", "remaining_time": "10:07:55"} +{"current_steps": 1175, "total_steps": 3339, "loss": 0.1775, "lr": 3.277102116835157e-05, "epoch": 2.4637985309548793, "percentage": 35.19, "elapsed_time": "5:29:18", "remaining_time": "10:06:30"} +{"current_steps": 1180, "total_steps": 3339, "loss": 0.1837, "lr": 3.2690390975670554e-05, "epoch": 2.474291710388248, "percentage": 35.34, "elapsed_time": "5:30:42", "remaining_time": "10:05:04"} +{"current_steps": 1185, "total_steps": 3339, "loss": 0.1806, "lr": 3.260941402632655e-05, "epoch": 2.484784889821616, "percentage": 35.49, "elapsed_time": "5:32:05", "remaining_time": "10:03:39"} +{"current_steps": 1190, "total_steps": 3339, "loss": 0.1811, "lr": 3.252809253296192e-05, "epoch": 2.4952780692549843, "percentage": 35.64, "elapsed_time": "5:33:30", "remaining_time": "10:02:16"} +{"current_steps": 1195, "total_steps": 3339, "loss": 0.18, "lr": 3.244642871763348e-05, "epoch": 2.5057712486883528, "percentage": 35.79, "elapsed_time": "5:34:54", "remaining_time": "10:00:52"} +{"current_steps": 1200, "total_steps": 3339, "loss": 0.1862, "lr": 3.236442481175176e-05, "epoch": 2.516264428121721, "percentage": 35.94, "elapsed_time": "5:36:18", "remaining_time": "9:59:28"} +{"current_steps": 1205, "total_steps": 3339, "loss": 0.1762, "lr": 3.228208305602004e-05, "epoch": 2.5267576075550893, "percentage": 36.09, "elapsed_time": "5:37:41", "remaining_time": "9:58:02"} +{"current_steps": 1210, "total_steps": 3339, "loss": 0.1819, "lr": 3.219940570037314e-05, "epoch": 2.5372507869884577, "percentage": 36.24, "elapsed_time": "5:39:04", "remaining_time": "9:56:36"} +{"current_steps": 1215, "total_steps": 3339, "loss": 0.1841, "lr": 3.211639500391592e-05, "epoch": 2.5477439664218258, "percentage": 36.39, "elapsed_time": "5:40:27", "remaining_time": "9:55:10"} +{"current_steps": 1220, "total_steps": 3339, "loss": 0.1805, "lr": 3.2033053234861557e-05, "epoch": 2.5582371458551942, "percentage": 36.54, "elapsed_time": "5:41:50", "remaining_time": "9:53:43"} +{"current_steps": 1225, "total_steps": 3339, "loss": 0.1832, "lr": 3.194938267046958e-05, "epoch": 2.5687303252885627, "percentage": 36.69, "elapsed_time": "5:43:14", "remaining_time": "9:52:20"} +{"current_steps": 1230, "total_steps": 3339, "loss": 0.1894, "lr": 3.186538559698364e-05, "epoch": 2.5792235047219307, "percentage": 36.84, "elapsed_time": "5:44:37", "remaining_time": "9:50:53"} +{"current_steps": 1235, "total_steps": 3339, "loss": 0.177, "lr": 3.1781064309569015e-05, "epoch": 2.589716684155299, "percentage": 36.99, "elapsed_time": "5:46:00", "remaining_time": "9:49:29"} +{"current_steps": 1240, "total_steps": 3339, "loss": 0.1898, "lr": 3.1696421112249925e-05, "epoch": 2.6002098635886672, "percentage": 37.14, "elapsed_time": "5:47:24", "remaining_time": "9:48:04"} +{"current_steps": 1245, "total_steps": 3339, "loss": 0.1839, "lr": 3.16114583178466e-05, "epoch": 2.6107030430220357, "percentage": 37.29, "elapsed_time": "5:48:48", "remaining_time": "9:46:39"} +{"current_steps": 1250, "total_steps": 3339, "loss": 0.1818, "lr": 3.1526178247912005e-05, "epoch": 2.6211962224554037, "percentage": 37.44, "elapsed_time": "5:50:12", "remaining_time": "9:45:16"} +{"current_steps": 1255, "total_steps": 3339, "loss": 0.1913, "lr": 3.1440583232668484e-05, "epoch": 2.631689401888772, "percentage": 37.59, "elapsed_time": "5:51:35", "remaining_time": "9:43:51"} +{"current_steps": 1260, "total_steps": 3339, "loss": 0.1876, "lr": 3.1354675610944046e-05, "epoch": 2.6421825813221407, "percentage": 37.74, "elapsed_time": "5:52:57", "remaining_time": "9:42:23"} +{"current_steps": 1265, "total_steps": 3339, "loss": 0.1853, "lr": 3.1268457730108454e-05, "epoch": 2.6526757607555087, "percentage": 37.89, "elapsed_time": "5:54:21", "remaining_time": "9:40:58"} +{"current_steps": 1270, "total_steps": 3339, "loss": 0.1818, "lr": 3.1181931946009125e-05, "epoch": 2.663168940188877, "percentage": 38.04, "elapsed_time": "5:55:44", "remaining_time": "9:39:33"} +{"current_steps": 1275, "total_steps": 3339, "loss": 0.1756, "lr": 3.1095100622906715e-05, "epoch": 2.6736621196222456, "percentage": 38.19, "elapsed_time": "5:57:08", "remaining_time": "9:38:08"} +{"current_steps": 1280, "total_steps": 3339, "loss": 0.1793, "lr": 3.100796613341055e-05, "epoch": 2.6841552990556137, "percentage": 38.33, "elapsed_time": "5:58:32", "remaining_time": "9:36:44"} +{"current_steps": 1285, "total_steps": 3339, "loss": 0.183, "lr": 3.092053085841375e-05, "epoch": 2.694648478488982, "percentage": 38.48, "elapsed_time": "5:59:57", "remaining_time": "9:35:21"} +{"current_steps": 1290, "total_steps": 3339, "loss": 0.1815, "lr": 3.083279718702824e-05, "epoch": 2.7051416579223506, "percentage": 38.63, "elapsed_time": "6:01:19", "remaining_time": "9:33:55"} +{"current_steps": 1295, "total_steps": 3339, "loss": 0.1758, "lr": 3.074476751651941e-05, "epoch": 2.7156348373557186, "percentage": 38.78, "elapsed_time": "6:02:44", "remaining_time": "9:32:32"} +{"current_steps": 1300, "total_steps": 3339, "loss": 0.1792, "lr": 3.065644425224063e-05, "epoch": 2.726128016789087, "percentage": 38.93, "elapsed_time": "6:04:07", "remaining_time": "9:31:07"} +{"current_steps": 1305, "total_steps": 3339, "loss": 0.1811, "lr": 3.056782980756754e-05, "epoch": 2.7366211962224556, "percentage": 39.08, "elapsed_time": "6:05:31", "remaining_time": "9:29:43"} +{"current_steps": 1310, "total_steps": 3339, "loss": 0.1766, "lr": 3.047892660383209e-05, "epoch": 2.7471143756558236, "percentage": 39.23, "elapsed_time": "6:06:56", "remaining_time": "9:28:19"} +{"current_steps": 1315, "total_steps": 3339, "loss": 0.1818, "lr": 3.0389737070256388e-05, "epoch": 2.757607555089192, "percentage": 39.38, "elapsed_time": "6:08:18", "remaining_time": "9:26:53"} +{"current_steps": 1320, "total_steps": 3339, "loss": 0.1792, "lr": 3.0300263643886302e-05, "epoch": 2.7681007345225606, "percentage": 39.53, "elapsed_time": "6:09:43", "remaining_time": "9:25:30"} +{"current_steps": 1325, "total_steps": 3339, "loss": 0.1786, "lr": 3.0210508769524906e-05, "epoch": 2.7785939139559286, "percentage": 39.68, "elapsed_time": "6:11:06", "remaining_time": "9:24:05"} +{"current_steps": 1330, "total_steps": 3339, "loss": 0.1837, "lr": 3.0120474899665643e-05, "epoch": 2.789087093389297, "percentage": 39.83, "elapsed_time": "6:12:30", "remaining_time": "9:22:40"} +{"current_steps": 1335, "total_steps": 3339, "loss": 0.1775, "lr": 3.0030164494425333e-05, "epoch": 2.7995802728226655, "percentage": 39.98, "elapsed_time": "6:13:54", "remaining_time": "9:21:16"} +{"current_steps": 1340, "total_steps": 3339, "loss": 0.1767, "lr": 2.993958002147695e-05, "epoch": 2.8100734522560336, "percentage": 40.13, "elapsed_time": "6:15:18", "remaining_time": "9:19:52"} +{"current_steps": 1345, "total_steps": 3339, "loss": 0.1848, "lr": 2.9848723955982168e-05, "epoch": 2.820566631689402, "percentage": 40.28, "elapsed_time": "6:16:42", "remaining_time": "9:18:29"} +{"current_steps": 1350, "total_steps": 3339, "loss": 0.1834, "lr": 2.9757598780523784e-05, "epoch": 2.83105981112277, "percentage": 40.43, "elapsed_time": "6:18:05", "remaining_time": "9:17:03"} +{"current_steps": 1355, "total_steps": 3339, "loss": 0.1778, "lr": 2.966620698503781e-05, "epoch": 2.8415529905561385, "percentage": 40.58, "elapsed_time": "6:19:26", "remaining_time": "9:15:35"} +{"current_steps": 1360, "total_steps": 3339, "loss": 0.1703, "lr": 2.9574551066745518e-05, "epoch": 2.8520461699895066, "percentage": 40.73, "elapsed_time": "6:20:48", "remaining_time": "9:14:08"} +{"current_steps": 1365, "total_steps": 3339, "loss": 0.1778, "lr": 2.9482633530085122e-05, "epoch": 2.862539349422875, "percentage": 40.88, "elapsed_time": "6:22:13", "remaining_time": "9:12:44"} +{"current_steps": 1370, "total_steps": 3339, "loss": 0.1872, "lr": 2.939045688664342e-05, "epoch": 2.8730325288562435, "percentage": 41.03, "elapsed_time": "6:23:36", "remaining_time": "9:11:20"} +{"current_steps": 1375, "total_steps": 3339, "loss": 0.1806, "lr": 2.929802365508712e-05, "epoch": 2.8835257082896115, "percentage": 41.18, "elapsed_time": "6:25:02", "remaining_time": "9:09:58"} +{"current_steps": 1380, "total_steps": 3339, "loss": 0.1799, "lr": 2.920533636109404e-05, "epoch": 2.89401888772298, "percentage": 41.33, "elapsed_time": "6:26:28", "remaining_time": "9:08:36"} +{"current_steps": 1385, "total_steps": 3339, "loss": 0.179, "lr": 2.911239753728407e-05, "epoch": 2.9045120671563485, "percentage": 41.48, "elapsed_time": "6:27:51", "remaining_time": "9:07:12"} +{"current_steps": 1390, "total_steps": 3339, "loss": 0.1816, "lr": 2.9019209723150008e-05, "epoch": 2.9150052465897165, "percentage": 41.63, "elapsed_time": "6:29:16", "remaining_time": "9:05:49"} +{"current_steps": 1395, "total_steps": 3339, "loss": 0.1846, "lr": 2.8925775464988154e-05, "epoch": 2.925498426023085, "percentage": 41.78, "elapsed_time": "6:30:40", "remaining_time": "9:04:26"} +{"current_steps": 1400, "total_steps": 3339, "loss": 0.1715, "lr": 2.8832097315828706e-05, "epoch": 2.9359916054564534, "percentage": 41.93, "elapsed_time": "6:32:05", "remaining_time": "9:03:03"} +{"current_steps": 1405, "total_steps": 3339, "loss": 0.1801, "lr": 2.873817783536605e-05, "epoch": 2.9464847848898215, "percentage": 42.08, "elapsed_time": "6:33:29", "remaining_time": "9:01:39"} +{"current_steps": 1410, "total_steps": 3339, "loss": 0.183, "lr": 2.8644019589888768e-05, "epoch": 2.95697796432319, "percentage": 42.23, "elapsed_time": "6:34:53", "remaining_time": "9:00:14"} +{"current_steps": 1415, "total_steps": 3339, "loss": 0.18, "lr": 2.854962515220957e-05, "epoch": 2.9674711437565584, "percentage": 42.38, "elapsed_time": "6:36:17", "remaining_time": "8:58:50"} +{"current_steps": 1420, "total_steps": 3339, "loss": 0.1795, "lr": 2.845499710159494e-05, "epoch": 2.9779643231899264, "percentage": 42.53, "elapsed_time": "6:37:38", "remaining_time": "8:57:22"} +{"current_steps": 1425, "total_steps": 3339, "loss": 0.1848, "lr": 2.8360138023694693e-05, "epoch": 2.988457502623295, "percentage": 42.68, "elapsed_time": "6:39:03", "remaining_time": "8:55:59"} +{"current_steps": 1430, "total_steps": 3339, "loss": 0.1833, "lr": 2.826505051047131e-05, "epoch": 2.9989506820566634, "percentage": 42.83, "elapsed_time": "6:40:27", "remaining_time": "8:54:35"} +{"current_steps": 1435, "total_steps": 3339, "loss": 0.1621, "lr": 2.816973716012912e-05, "epoch": 3.0083945435466948, "percentage": 42.98, "elapsed_time": "6:41:43", "remaining_time": "8:53:00"} +{"current_steps": 1440, "total_steps": 3339, "loss": 0.1656, "lr": 2.807420057704331e-05, "epoch": 3.018887722980063, "percentage": 43.13, "elapsed_time": "6:43:06", "remaining_time": "8:51:36"} +{"current_steps": 1445, "total_steps": 3339, "loss": 0.1681, "lr": 2.797844337168874e-05, "epoch": 3.0293809024134313, "percentage": 43.28, "elapsed_time": "6:44:30", "remaining_time": "8:50:11"} +{"current_steps": 1450, "total_steps": 3339, "loss": 0.1675, "lr": 2.7882468160568653e-05, "epoch": 3.0398740818467997, "percentage": 43.43, "elapsed_time": "6:45:54", "remaining_time": "8:48:47"} +{"current_steps": 1455, "total_steps": 3339, "loss": 0.1639, "lr": 2.778627756614314e-05, "epoch": 3.0503672612801678, "percentage": 43.58, "elapsed_time": "6:47:17", "remaining_time": "8:47:23"} +{"current_steps": 1460, "total_steps": 3339, "loss": 0.1672, "lr": 2.768987421675751e-05, "epoch": 3.0608604407135362, "percentage": 43.73, "elapsed_time": "6:48:40", "remaining_time": "8:45:57"} +{"current_steps": 1465, "total_steps": 3339, "loss": 0.1732, "lr": 2.7593260746570448e-05, "epoch": 3.0713536201469047, "percentage": 43.88, "elapsed_time": "6:50:04", "remaining_time": "8:44:33"} +{"current_steps": 1470, "total_steps": 3339, "loss": 0.165, "lr": 2.749643979548207e-05, "epoch": 3.0818467995802727, "percentage": 44.03, "elapsed_time": "6:51:27", "remaining_time": "8:43:07"} +{"current_steps": 1475, "total_steps": 3339, "loss": 0.1706, "lr": 2.7399414009061767e-05, "epoch": 3.092339979013641, "percentage": 44.17, "elapsed_time": "6:52:49", "remaining_time": "8:41:42"} +{"current_steps": 1480, "total_steps": 3339, "loss": 0.1648, "lr": 2.7302186038475915e-05, "epoch": 3.1028331584470092, "percentage": 44.32, "elapsed_time": "6:54:13", "remaining_time": "8:40:18"} +{"current_steps": 1485, "total_steps": 3339, "loss": 0.167, "lr": 2.7204758540415456e-05, "epoch": 3.1133263378803777, "percentage": 44.47, "elapsed_time": "6:55:37", "remaining_time": "8:38:54"} +{"current_steps": 1490, "total_steps": 3339, "loss": 0.1748, "lr": 2.710713417702327e-05, "epoch": 3.123819517313746, "percentage": 44.62, "elapsed_time": "6:57:01", "remaining_time": "8:37:29"} +{"current_steps": 1495, "total_steps": 3339, "loss": 0.1648, "lr": 2.700931561582147e-05, "epoch": 3.134312696747114, "percentage": 44.77, "elapsed_time": "6:58:24", "remaining_time": "8:36:04"} +{"current_steps": 1500, "total_steps": 3339, "loss": 0.1662, "lr": 2.6911305529638482e-05, "epoch": 3.1448058761804827, "percentage": 44.92, "elapsed_time": "6:59:48", "remaining_time": "8:34:40"} +{"current_steps": 1505, "total_steps": 3339, "loss": 0.1712, "lr": 2.681310659653605e-05, "epoch": 3.155299055613851, "percentage": 45.07, "elapsed_time": "7:01:48", "remaining_time": "8:34:01"} +{"current_steps": 1510, "total_steps": 3339, "loss": 0.169, "lr": 2.671472149973602e-05, "epoch": 3.165792235047219, "percentage": 45.22, "elapsed_time": "7:03:11", "remaining_time": "8:32:35"} +{"current_steps": 1515, "total_steps": 3339, "loss": 0.1672, "lr": 2.6616152927547037e-05, "epoch": 3.1762854144805877, "percentage": 45.37, "elapsed_time": "7:04:34", "remaining_time": "8:31:10"} +{"current_steps": 1520, "total_steps": 3339, "loss": 0.1681, "lr": 2.6517403573291112e-05, "epoch": 3.186778593913956, "percentage": 45.52, "elapsed_time": "7:05:57", "remaining_time": "8:29:44"} +{"current_steps": 1525, "total_steps": 3339, "loss": 0.1698, "lr": 2.6418476135229978e-05, "epoch": 3.197271773347324, "percentage": 45.67, "elapsed_time": "7:07:20", "remaining_time": "8:28:20"} +{"current_steps": 1530, "total_steps": 3339, "loss": 0.1689, "lr": 2.6319373316491425e-05, "epoch": 3.2077649527806926, "percentage": 45.82, "elapsed_time": "7:08:43", "remaining_time": "8:26:54"} +{"current_steps": 1535, "total_steps": 3339, "loss": 0.1656, "lr": 2.622009782499537e-05, "epoch": 3.2182581322140607, "percentage": 45.97, "elapsed_time": "7:10:07", "remaining_time": "8:25:30"} +{"current_steps": 1540, "total_steps": 3339, "loss": 0.1798, "lr": 2.6120652373379944e-05, "epoch": 3.228751311647429, "percentage": 46.12, "elapsed_time": "7:11:31", "remaining_time": "8:24:05"} +{"current_steps": 1545, "total_steps": 3339, "loss": 0.1756, "lr": 2.6021039678927275e-05, "epoch": 3.2392444910807976, "percentage": 46.27, "elapsed_time": "7:12:54", "remaining_time": "8:22:40"} +{"current_steps": 1550, "total_steps": 3339, "loss": 0.1644, "lr": 2.592126246348934e-05, "epoch": 3.2497376705141656, "percentage": 46.42, "elapsed_time": "7:14:18", "remaining_time": "8:21:16"} +{"current_steps": 1555, "total_steps": 3339, "loss": 0.1717, "lr": 2.5821323453413517e-05, "epoch": 3.260230849947534, "percentage": 46.57, "elapsed_time": "7:15:43", "remaining_time": "8:19:53"} +{"current_steps": 1560, "total_steps": 3339, "loss": 0.1675, "lr": 2.5721225379468133e-05, "epoch": 3.2707240293809026, "percentage": 46.72, "elapsed_time": "7:17:06", "remaining_time": "8:18:28"} +{"current_steps": 1565, "total_steps": 3339, "loss": 0.1681, "lr": 2.5620970976767816e-05, "epoch": 3.2812172088142706, "percentage": 46.87, "elapsed_time": "7:18:28", "remaining_time": "8:17:02"} +{"current_steps": 1570, "total_steps": 3339, "loss": 0.1721, "lr": 2.5520562984698787e-05, "epoch": 3.291710388247639, "percentage": 47.02, "elapsed_time": "7:19:52", "remaining_time": "8:15:38"} +{"current_steps": 1575, "total_steps": 3339, "loss": 0.1673, "lr": 2.542000414684399e-05, "epoch": 3.3022035676810075, "percentage": 47.17, "elapsed_time": "7:21:17", "remaining_time": "8:14:14"} +{"current_steps": 1580, "total_steps": 3339, "loss": 0.1672, "lr": 2.5319297210908143e-05, "epoch": 3.3126967471143756, "percentage": 47.32, "elapsed_time": "7:22:40", "remaining_time": "8:12:49"} +{"current_steps": 1585, "total_steps": 3339, "loss": 0.1705, "lr": 2.5218444928642633e-05, "epoch": 3.323189926547744, "percentage": 47.47, "elapsed_time": "7:24:03", "remaining_time": "8:11:24"} +{"current_steps": 1590, "total_steps": 3339, "loss": 0.1724, "lr": 2.5117450055770334e-05, "epoch": 3.3336831059811125, "percentage": 47.62, "elapsed_time": "7:25:27", "remaining_time": "8:10:00"} +{"current_steps": 1595, "total_steps": 3339, "loss": 0.1671, "lr": 2.5016315351910355e-05, "epoch": 3.3441762854144805, "percentage": 47.77, "elapsed_time": "7:26:50", "remaining_time": "8:08:34"} +{"current_steps": 1600, "total_steps": 3339, "loss": 0.1695, "lr": 2.4915043580502543e-05, "epoch": 3.354669464847849, "percentage": 47.92, "elapsed_time": "7:28:13", "remaining_time": "8:07:09"} +{"current_steps": 1605, "total_steps": 3339, "loss": 0.1695, "lr": 2.4813637508732068e-05, "epoch": 3.365162644281217, "percentage": 48.07, "elapsed_time": "7:29:36", "remaining_time": "8:05:44"} +{"current_steps": 1610, "total_steps": 3339, "loss": 0.165, "lr": 2.4712099907453756e-05, "epoch": 3.3756558237145855, "percentage": 48.22, "elapsed_time": "7:30:59", "remaining_time": "8:04:19"} +{"current_steps": 1615, "total_steps": 3339, "loss": 0.1601, "lr": 2.4610433551116374e-05, "epoch": 3.386149003147954, "percentage": 48.37, "elapsed_time": "7:32:23", "remaining_time": "8:02:55"} +{"current_steps": 1620, "total_steps": 3339, "loss": 0.1728, "lr": 2.4508641217686868e-05, "epoch": 3.396642182581322, "percentage": 48.52, "elapsed_time": "7:33:47", "remaining_time": "8:01:30"} +{"current_steps": 1625, "total_steps": 3339, "loss": 0.1663, "lr": 2.4406725688574394e-05, "epoch": 3.4071353620146905, "percentage": 48.67, "elapsed_time": "7:35:09", "remaining_time": "8:00:05"} +{"current_steps": 1630, "total_steps": 3339, "loss": 0.1669, "lr": 2.430468974855438e-05, "epoch": 3.417628541448059, "percentage": 48.82, "elapsed_time": "7:36:31", "remaining_time": "7:58:39"} +{"current_steps": 1635, "total_steps": 3339, "loss": 0.1676, "lr": 2.4202536185692386e-05, "epoch": 3.428121720881427, "percentage": 48.97, "elapsed_time": "7:37:55", "remaining_time": "7:57:14"} +{"current_steps": 1640, "total_steps": 3339, "loss": 0.167, "lr": 2.410026779126794e-05, "epoch": 3.4386149003147954, "percentage": 49.12, "elapsed_time": "7:39:18", "remaining_time": "7:55:49"} +{"current_steps": 1645, "total_steps": 3339, "loss": 0.1714, "lr": 2.399788735969827e-05, "epoch": 3.4491080797481635, "percentage": 49.27, "elapsed_time": "7:40:41", "remaining_time": "7:54:25"} +{"current_steps": 1650, "total_steps": 3339, "loss": 0.1674, "lr": 2.3895397688461946e-05, "epoch": 3.459601259181532, "percentage": 49.42, "elapsed_time": "7:42:04", "remaining_time": "7:52:59"} +{"current_steps": 1655, "total_steps": 3339, "loss": 0.165, "lr": 2.379280157802245e-05, "epoch": 3.4700944386149004, "percentage": 49.57, "elapsed_time": "7:43:26", "remaining_time": "7:51:33"} +{"current_steps": 1660, "total_steps": 3339, "loss": 0.1673, "lr": 2.3690101831751627e-05, "epoch": 3.4805876180482684, "percentage": 49.72, "elapsed_time": "7:44:49", "remaining_time": "7:50:09"} +{"current_steps": 1665, "total_steps": 3339, "loss": 0.1671, "lr": 2.3587301255853122e-05, "epoch": 3.491080797481637, "percentage": 49.87, "elapsed_time": "7:46:12", "remaining_time": "7:48:44"} +{"current_steps": 1670, "total_steps": 3339, "loss": 0.1715, "lr": 2.3484402659285666e-05, "epoch": 3.5015739769150054, "percentage": 50.01, "elapsed_time": "7:47:35", "remaining_time": "7:47:18"} +{"current_steps": 1675, "total_steps": 3339, "loss": 0.1677, "lr": 2.3381408853686362e-05, "epoch": 3.5120671563483734, "percentage": 50.16, "elapsed_time": "7:48:58", "remaining_time": "7:45:53"} +{"current_steps": 1680, "total_steps": 3339, "loss": 0.1674, "lr": 2.327832265329381e-05, "epoch": 3.522560335781742, "percentage": 50.31, "elapsed_time": "7:50:21", "remaining_time": "7:44:28"} +{"current_steps": 1685, "total_steps": 3339, "loss": 0.1672, "lr": 2.3175146874871265e-05, "epoch": 3.5330535152151104, "percentage": 50.46, "elapsed_time": "7:51:43", "remaining_time": "7:43:03"} +{"current_steps": 1690, "total_steps": 3339, "loss": 0.1718, "lr": 2.3071884337629638e-05, "epoch": 3.5435466946484784, "percentage": 50.61, "elapsed_time": "7:53:07", "remaining_time": "7:41:38"} +{"current_steps": 1695, "total_steps": 3339, "loss": 0.1635, "lr": 2.296853786315045e-05, "epoch": 3.554039874081847, "percentage": 50.76, "elapsed_time": "7:54:29", "remaining_time": "7:40:12"} +{"current_steps": 1700, "total_steps": 3339, "loss": 0.1701, "lr": 2.2865110275308774e-05, "epoch": 3.5645330535152153, "percentage": 50.91, "elapsed_time": "7:55:52", "remaining_time": "7:38:48"} +{"current_steps": 1705, "total_steps": 3339, "loss": 0.1738, "lr": 2.276160440019603e-05, "epoch": 3.5750262329485833, "percentage": 51.06, "elapsed_time": "7:57:15", "remaining_time": "7:37:23"} +{"current_steps": 1710, "total_steps": 3339, "loss": 0.1658, "lr": 2.2658023066042815e-05, "epoch": 3.585519412381952, "percentage": 51.21, "elapsed_time": "7:58:39", "remaining_time": "7:35:59"} +{"current_steps": 1715, "total_steps": 3339, "loss": 0.1722, "lr": 2.2554369103141566e-05, "epoch": 3.5960125918153203, "percentage": 51.36, "elapsed_time": "8:00:03", "remaining_time": "7:34:34"} +{"current_steps": 1720, "total_steps": 3339, "loss": 0.1661, "lr": 2.2450645343769285e-05, "epoch": 3.6065057712486883, "percentage": 51.51, "elapsed_time": "8:01:27", "remaining_time": "7:33:10"} +{"current_steps": 1725, "total_steps": 3339, "loss": 0.1638, "lr": 2.2346854622110082e-05, "epoch": 3.616998950682057, "percentage": 51.66, "elapsed_time": "8:02:49", "remaining_time": "7:31:45"} +{"current_steps": 1730, "total_steps": 3339, "loss": 0.168, "lr": 2.224299977417779e-05, "epoch": 3.627492130115425, "percentage": 51.81, "elapsed_time": "8:04:11", "remaining_time": "7:30:19"} +{"current_steps": 1735, "total_steps": 3339, "loss": 0.1633, "lr": 2.2139083637738455e-05, "epoch": 3.6379853095487933, "percentage": 51.96, "elapsed_time": "8:05:34", "remaining_time": "7:28:54"} +{"current_steps": 1740, "total_steps": 3339, "loss": 0.1647, "lr": 2.2035109052232768e-05, "epoch": 3.6484784889821618, "percentage": 52.11, "elapsed_time": "8:06:56", "remaining_time": "7:27:29"} +{"current_steps": 1745, "total_steps": 3339, "loss": 0.1679, "lr": 2.1931078858698526e-05, "epoch": 3.65897166841553, "percentage": 52.26, "elapsed_time": "8:08:19", "remaining_time": "7:26:03"} +{"current_steps": 1750, "total_steps": 3339, "loss": 0.1571, "lr": 2.1826995899692968e-05, "epoch": 3.6694648478488983, "percentage": 52.41, "elapsed_time": "8:09:43", "remaining_time": "7:24:40"} +{"current_steps": 1755, "total_steps": 3339, "loss": 0.1715, "lr": 2.1722863019215122e-05, "epoch": 3.6799580272822663, "percentage": 52.56, "elapsed_time": "8:11:07", "remaining_time": "7:23:16"} +{"current_steps": 1760, "total_steps": 3339, "loss": 0.1704, "lr": 2.1618683062628068e-05, "epoch": 3.6904512067156348, "percentage": 52.71, "elapsed_time": "8:12:32", "remaining_time": "7:21:53"} +{"current_steps": 1765, "total_steps": 3339, "loss": 0.1643, "lr": 2.151445887658125e-05, "epoch": 3.7009443861490032, "percentage": 52.86, "elapsed_time": "8:13:55", "remaining_time": "7:20:28"} +{"current_steps": 1770, "total_steps": 3339, "loss": 0.169, "lr": 2.1410193308932602e-05, "epoch": 3.7114375655823713, "percentage": 53.01, "elapsed_time": "8:15:19", "remaining_time": "7:19:04"} +{"current_steps": 1775, "total_steps": 3339, "loss": 0.1715, "lr": 2.1305889208670827e-05, "epoch": 3.7219307450157397, "percentage": 53.16, "elapsed_time": "8:16:41", "remaining_time": "7:17:38"} +{"current_steps": 1780, "total_steps": 3339, "loss": 0.1595, "lr": 2.120154942583748e-05, "epoch": 3.732423924449108, "percentage": 53.31, "elapsed_time": "8:18:04", "remaining_time": "7:16:13"} +{"current_steps": 1785, "total_steps": 3339, "loss": 0.1606, "lr": 2.1097176811449127e-05, "epoch": 3.7429171038824762, "percentage": 53.46, "elapsed_time": "8:19:27", "remaining_time": "7:14:49"} +{"current_steps": 1790, "total_steps": 3339, "loss": 0.1681, "lr": 2.0992774217419445e-05, "epoch": 3.7534102833158447, "percentage": 53.61, "elapsed_time": "8:20:50", "remaining_time": "7:13:24"} +{"current_steps": 1795, "total_steps": 3339, "loss": 0.1633, "lr": 2.088834449648127e-05, "epoch": 3.763903462749213, "percentage": 53.76, "elapsed_time": "8:22:13", "remaining_time": "7:11:59"} +{"current_steps": 1800, "total_steps": 3339, "loss": 0.1695, "lr": 2.0783890502108665e-05, "epoch": 3.774396642182581, "percentage": 53.91, "elapsed_time": "8:23:37", "remaining_time": "7:10:35"} +{"current_steps": 1805, "total_steps": 3339, "loss": 0.1656, "lr": 2.067941508843896e-05, "epoch": 3.7848898216159497, "percentage": 54.06, "elapsed_time": "8:25:01", "remaining_time": "7:09:11"} +{"current_steps": 1810, "total_steps": 3339, "loss": 0.1656, "lr": 2.057492111019474e-05, "epoch": 3.795383001049318, "percentage": 54.21, "elapsed_time": "8:26:24", "remaining_time": "7:07:47"} +{"current_steps": 1815, "total_steps": 3339, "loss": 0.1543, "lr": 2.0470411422605866e-05, "epoch": 3.805876180482686, "percentage": 54.36, "elapsed_time": "8:27:48", "remaining_time": "7:06:23"} +{"current_steps": 1820, "total_steps": 3339, "loss": 0.1643, "lr": 2.0365888881331437e-05, "epoch": 3.8163693599160546, "percentage": 54.51, "elapsed_time": "8:29:11", "remaining_time": "7:04:58"} +{"current_steps": 1825, "total_steps": 3339, "loss": 0.1624, "lr": 2.0261356342381785e-05, "epoch": 3.826862539349423, "percentage": 54.66, "elapsed_time": "8:30:34", "remaining_time": "7:03:34"} +{"current_steps": 1830, "total_steps": 3339, "loss": 0.1631, "lr": 2.01568166620404e-05, "epoch": 3.837355718782791, "percentage": 54.81, "elapsed_time": "8:31:57", "remaining_time": "7:02:09"} +{"current_steps": 1835, "total_steps": 3339, "loss": 0.1643, "lr": 2.0052272696785926e-05, "epoch": 3.8478488982161596, "percentage": 54.96, "elapsed_time": "8:33:21", "remaining_time": "7:00:45"} +{"current_steps": 1840, "total_steps": 3339, "loss": 0.1636, "lr": 1.9947727303214077e-05, "epoch": 3.858342077649528, "percentage": 55.11, "elapsed_time": "8:34:45", "remaining_time": "6:59:21"} +{"current_steps": 1845, "total_steps": 3339, "loss": 0.1662, "lr": 1.9843183337959607e-05, "epoch": 3.868835257082896, "percentage": 55.26, "elapsed_time": "8:36:08", "remaining_time": "6:57:57"} +{"current_steps": 1850, "total_steps": 3339, "loss": 0.1631, "lr": 1.973864365761822e-05, "epoch": 3.8793284365162646, "percentage": 55.41, "elapsed_time": "8:37:32", "remaining_time": "6:56:33"} +{"current_steps": 1855, "total_steps": 3339, "loss": 0.1714, "lr": 1.9634111118668566e-05, "epoch": 3.8898216159496326, "percentage": 55.56, "elapsed_time": "8:38:56", "remaining_time": "6:55:09"} +{"current_steps": 1860, "total_steps": 3339, "loss": 0.1618, "lr": 1.9529588577394138e-05, "epoch": 3.900314795383001, "percentage": 55.71, "elapsed_time": "8:40:20", "remaining_time": "6:53:45"} +{"current_steps": 1865, "total_steps": 3339, "loss": 0.1661, "lr": 1.9425078889805263e-05, "epoch": 3.910807974816369, "percentage": 55.86, "elapsed_time": "8:41:42", "remaining_time": "6:52:19"} +{"current_steps": 1870, "total_steps": 3339, "loss": 0.171, "lr": 1.9320584911561048e-05, "epoch": 3.9213011542497376, "percentage": 56.0, "elapsed_time": "8:43:04", "remaining_time": "6:50:54"} +{"current_steps": 1875, "total_steps": 3339, "loss": 0.1607, "lr": 1.921610949789134e-05, "epoch": 3.931794333683106, "percentage": 56.15, "elapsed_time": "8:44:27", "remaining_time": "6:49:29"} +{"current_steps": 1880, "total_steps": 3339, "loss": 0.1686, "lr": 1.911165550351874e-05, "epoch": 3.942287513116474, "percentage": 56.3, "elapsed_time": "8:45:49", "remaining_time": "6:48:04"} +{"current_steps": 1885, "total_steps": 3339, "loss": 0.1666, "lr": 1.9007225782580562e-05, "epoch": 3.9527806925498425, "percentage": 56.45, "elapsed_time": "8:47:10", "remaining_time": "6:46:38"} +{"current_steps": 1890, "total_steps": 3339, "loss": 0.1613, "lr": 1.8902823188550877e-05, "epoch": 3.963273871983211, "percentage": 56.6, "elapsed_time": "8:48:32", "remaining_time": "6:45:13"} +{"current_steps": 1895, "total_steps": 3339, "loss": 0.1686, "lr": 1.8798450574162526e-05, "epoch": 3.973767051416579, "percentage": 56.75, "elapsed_time": "8:49:54", "remaining_time": "6:43:47"} +{"current_steps": 1900, "total_steps": 3339, "loss": 0.1664, "lr": 1.869411079132918e-05, "epoch": 3.9842602308499475, "percentage": 56.9, "elapsed_time": "8:51:17", "remaining_time": "6:42:23"} +{"current_steps": 1905, "total_steps": 3339, "loss": 0.1651, "lr": 1.85898066910674e-05, "epoch": 3.994753410283316, "percentage": 57.05, "elapsed_time": "8:52:40", "remaining_time": "6:40:58"} +{"current_steps": 1910, "total_steps": 3339, "loss": 0.1653, "lr": 1.848554112341876e-05, "epoch": 4.004197271773347, "percentage": 57.2, "elapsed_time": "8:53:55", "remaining_time": "6:39:27"} +{"current_steps": 1915, "total_steps": 3339, "loss": 0.1595, "lr": 1.8381316937371935e-05, "epoch": 4.014690451206715, "percentage": 57.35, "elapsed_time": "8:55:18", "remaining_time": "6:38:03"} +{"current_steps": 1920, "total_steps": 3339, "loss": 0.1521, "lr": 1.8277136980784888e-05, "epoch": 4.025183630640084, "percentage": 57.5, "elapsed_time": "8:56:41", "remaining_time": "6:36:38"} +{"current_steps": 1925, "total_steps": 3339, "loss": 0.1588, "lr": 1.817300410030704e-05, "epoch": 4.035676810073452, "percentage": 57.65, "elapsed_time": "8:58:04", "remaining_time": "6:35:14"} +{"current_steps": 1930, "total_steps": 3339, "loss": 0.156, "lr": 1.806892114130148e-05, "epoch": 4.04616998950682, "percentage": 57.8, "elapsed_time": "8:59:27", "remaining_time": "6:33:49"} +{"current_steps": 1935, "total_steps": 3339, "loss": 0.1587, "lr": 1.7964890947767242e-05, "epoch": 4.056663168940189, "percentage": 57.95, "elapsed_time": "9:00:51", "remaining_time": "6:32:25"} +{"current_steps": 1940, "total_steps": 3339, "loss": 0.1563, "lr": 1.786091636226156e-05, "epoch": 4.067156348373557, "percentage": 58.1, "elapsed_time": "9:02:13", "remaining_time": "6:31:00"} +{"current_steps": 1945, "total_steps": 3339, "loss": 0.156, "lr": 1.7757000225822216e-05, "epoch": 4.077649527806925, "percentage": 58.25, "elapsed_time": "9:03:36", "remaining_time": "6:29:36"} +{"current_steps": 1950, "total_steps": 3339, "loss": 0.1571, "lr": 1.765314537788993e-05, "epoch": 4.088142707240293, "percentage": 58.4, "elapsed_time": "9:05:00", "remaining_time": "6:28:13"} +{"current_steps": 1955, "total_steps": 3339, "loss": 0.1565, "lr": 1.7549354656230725e-05, "epoch": 4.098635886673662, "percentage": 58.55, "elapsed_time": "9:06:24", "remaining_time": "6:26:48"} +{"current_steps": 1960, "total_steps": 3339, "loss": 0.1597, "lr": 1.744563089685844e-05, "epoch": 4.10912906610703, "percentage": 58.7, "elapsed_time": "9:07:47", "remaining_time": "6:25:24"} +{"current_steps": 1965, "total_steps": 3339, "loss": 0.148, "lr": 1.7341976933957195e-05, "epoch": 4.119622245540398, "percentage": 58.85, "elapsed_time": "9:09:10", "remaining_time": "6:24:00"} +{"current_steps": 1970, "total_steps": 3339, "loss": 0.1556, "lr": 1.723839559980398e-05, "epoch": 4.130115424973767, "percentage": 59.0, "elapsed_time": "9:10:32", "remaining_time": "6:22:35"} +{"current_steps": 1975, "total_steps": 3339, "loss": 0.1526, "lr": 1.713488972469124e-05, "epoch": 4.140608604407135, "percentage": 59.15, "elapsed_time": "9:11:55", "remaining_time": "6:21:10"} +{"current_steps": 1980, "total_steps": 3339, "loss": 0.1557, "lr": 1.7031462136849557e-05, "epoch": 4.151101783840503, "percentage": 59.3, "elapsed_time": "9:13:18", "remaining_time": "6:19:46"} +{"current_steps": 1985, "total_steps": 3339, "loss": 0.1522, "lr": 1.6928115662370372e-05, "epoch": 4.161594963273872, "percentage": 59.45, "elapsed_time": "9:14:41", "remaining_time": "6:18:21"} +{"current_steps": 1990, "total_steps": 3339, "loss": 0.1568, "lr": 1.682485312512874e-05, "epoch": 4.17208814270724, "percentage": 59.6, "elapsed_time": "9:16:04", "remaining_time": "6:16:57"} +{"current_steps": 1995, "total_steps": 3339, "loss": 0.1564, "lr": 1.67216773467062e-05, "epoch": 4.182581322140608, "percentage": 59.75, "elapsed_time": "9:17:26", "remaining_time": "6:15:32"} +{"current_steps": 2000, "total_steps": 3339, "loss": 0.1611, "lr": 1.661859114631365e-05, "epoch": 4.193074501573977, "percentage": 59.9, "elapsed_time": "9:18:50", "remaining_time": "6:14:08"} +{"current_steps": 2005, "total_steps": 3339, "loss": 0.1534, "lr": 1.6515597340714348e-05, "epoch": 4.203567681007345, "percentage": 60.05, "elapsed_time": "9:20:12", "remaining_time": "6:12:43"} +{"current_steps": 2010, "total_steps": 3339, "loss": 0.1534, "lr": 1.641269874414689e-05, "epoch": 4.214060860440713, "percentage": 60.2, "elapsed_time": "9:21:34", "remaining_time": "6:11:18"} +{"current_steps": 2015, "total_steps": 3339, "loss": 0.1623, "lr": 1.6309898168248383e-05, "epoch": 4.224554039874082, "percentage": 60.35, "elapsed_time": "9:22:57", "remaining_time": "6:09:54"} +{"current_steps": 2020, "total_steps": 3339, "loss": 0.1579, "lr": 1.6207198421977558e-05, "epoch": 4.23504721930745, "percentage": 60.5, "elapsed_time": "9:24:21", "remaining_time": "6:08:30"} +{"current_steps": 2025, "total_steps": 3339, "loss": 0.1501, "lr": 1.6104602311538058e-05, "epoch": 4.245540398740818, "percentage": 60.65, "elapsed_time": "9:25:45", "remaining_time": "6:07:06"} +{"current_steps": 2030, "total_steps": 3339, "loss": 0.1597, "lr": 1.6002112640301732e-05, "epoch": 4.256033578174187, "percentage": 60.8, "elapsed_time": "9:27:09", "remaining_time": "6:05:42"} +{"current_steps": 2035, "total_steps": 3339, "loss": 0.1527, "lr": 1.5899732208732066e-05, "epoch": 4.266526757607555, "percentage": 60.95, "elapsed_time": "9:28:32", "remaining_time": "6:04:18"} +{"current_steps": 2040, "total_steps": 3339, "loss": 0.1599, "lr": 1.5797463814307617e-05, "epoch": 4.277019937040923, "percentage": 61.1, "elapsed_time": "9:29:55", "remaining_time": "6:02:54"} +{"current_steps": 2045, "total_steps": 3339, "loss": 0.1584, "lr": 1.5695310251445623e-05, "epoch": 4.287513116474292, "percentage": 61.25, "elapsed_time": "9:31:17", "remaining_time": "6:01:29"} +{"current_steps": 2050, "total_steps": 3339, "loss": 0.1544, "lr": 1.5593274311425606e-05, "epoch": 4.29800629590766, "percentage": 61.4, "elapsed_time": "9:32:40", "remaining_time": "6:00:05"} +{"current_steps": 2055, "total_steps": 3339, "loss": 0.1538, "lr": 1.5491358782313136e-05, "epoch": 4.308499475341028, "percentage": 61.55, "elapsed_time": "9:34:04", "remaining_time": "5:58:41"} +{"current_steps": 2060, "total_steps": 3339, "loss": 0.1548, "lr": 1.5389566448883626e-05, "epoch": 4.318992654774396, "percentage": 61.7, "elapsed_time": "9:35:27", "remaining_time": "5:57:17"} +{"current_steps": 2065, "total_steps": 3339, "loss": 0.1567, "lr": 1.5287900092546247e-05, "epoch": 4.329485834207765, "percentage": 61.84, "elapsed_time": "9:36:48", "remaining_time": "5:55:51"} +{"current_steps": 2070, "total_steps": 3339, "loss": 0.1577, "lr": 1.518636249126793e-05, "epoch": 4.339979013641133, "percentage": 61.99, "elapsed_time": "9:38:12", "remaining_time": "5:54:28"} +{"current_steps": 2075, "total_steps": 3339, "loss": 0.1581, "lr": 1.5084956419497455e-05, "epoch": 4.350472193074501, "percentage": 62.14, "elapsed_time": "9:39:33", "remaining_time": "5:53:02"} +{"current_steps": 2080, "total_steps": 3339, "loss": 0.1524, "lr": 1.498368464808965e-05, "epoch": 4.36096537250787, "percentage": 62.29, "elapsed_time": "9:40:56", "remaining_time": "5:51:38"} +{"current_steps": 2085, "total_steps": 3339, "loss": 0.154, "lr": 1.4882549944229662e-05, "epoch": 4.371458551941238, "percentage": 62.44, "elapsed_time": "9:42:17", "remaining_time": "5:50:13"} +{"current_steps": 2090, "total_steps": 3339, "loss": 0.1577, "lr": 1.4781555071357372e-05, "epoch": 4.381951731374606, "percentage": 62.59, "elapsed_time": "9:43:40", "remaining_time": "5:48:48"} +{"current_steps": 2095, "total_steps": 3339, "loss": 0.1525, "lr": 1.468070278909186e-05, "epoch": 4.392444910807975, "percentage": 62.74, "elapsed_time": "9:45:03", "remaining_time": "5:47:24"} +{"current_steps": 2100, "total_steps": 3339, "loss": 0.1569, "lr": 1.4579995853156008e-05, "epoch": 4.402938090241343, "percentage": 62.89, "elapsed_time": "9:46:26", "remaining_time": "5:46:00"} +{"current_steps": 2105, "total_steps": 3339, "loss": 0.1562, "lr": 1.4479437015301217e-05, "epoch": 4.413431269674711, "percentage": 63.04, "elapsed_time": "9:47:50", "remaining_time": "5:44:36"} +{"current_steps": 2110, "total_steps": 3339, "loss": 0.1569, "lr": 1.4379029023232188e-05, "epoch": 4.42392444910808, "percentage": 63.19, "elapsed_time": "9:49:13", "remaining_time": "5:43:12"} +{"current_steps": 2115, "total_steps": 3339, "loss": 0.1661, "lr": 1.4278774620531875e-05, "epoch": 4.434417628541448, "percentage": 63.34, "elapsed_time": "9:50:35", "remaining_time": "5:41:47"} +{"current_steps": 2120, "total_steps": 3339, "loss": 0.1595, "lr": 1.4178676546586484e-05, "epoch": 4.444910807974816, "percentage": 63.49, "elapsed_time": "9:51:58", "remaining_time": "5:40:23"} +{"current_steps": 2125, "total_steps": 3339, "loss": 0.1564, "lr": 1.407873753651067e-05, "epoch": 4.455403987408185, "percentage": 63.64, "elapsed_time": "9:53:21", "remaining_time": "5:38:59"} +{"current_steps": 2130, "total_steps": 3339, "loss": 0.1533, "lr": 1.3978960321072733e-05, "epoch": 4.465897166841553, "percentage": 63.79, "elapsed_time": "9:54:44", "remaining_time": "5:37:34"} +{"current_steps": 2135, "total_steps": 3339, "loss": 0.1551, "lr": 1.3879347626620066e-05, "epoch": 4.476390346274921, "percentage": 63.94, "elapsed_time": "9:56:07", "remaining_time": "5:36:10"} +{"current_steps": 2140, "total_steps": 3339, "loss": 0.151, "lr": 1.3779902175004634e-05, "epoch": 4.48688352570829, "percentage": 64.09, "elapsed_time": "9:57:30", "remaining_time": "5:34:46"} +{"current_steps": 2145, "total_steps": 3339, "loss": 0.1556, "lr": 1.3680626683508582e-05, "epoch": 4.497376705141658, "percentage": 64.24, "elapsed_time": "9:58:53", "remaining_time": "5:33:22"} +{"current_steps": 2150, "total_steps": 3339, "loss": 0.1518, "lr": 1.3581523864770027e-05, "epoch": 4.507869884575026, "percentage": 64.39, "elapsed_time": "10:00:17", "remaining_time": "5:31:58"} +{"current_steps": 2155, "total_steps": 3339, "loss": 0.1589, "lr": 1.3482596426708896e-05, "epoch": 4.518363064008394, "percentage": 64.54, "elapsed_time": "10:01:38", "remaining_time": "5:30:33"} +{"current_steps": 2160, "total_steps": 3339, "loss": 0.1539, "lr": 1.3383847072452966e-05, "epoch": 4.528856243441763, "percentage": 64.69, "elapsed_time": "10:03:02", "remaining_time": "5:29:09"} +{"current_steps": 2165, "total_steps": 3339, "loss": 0.1518, "lr": 1.3285278500263987e-05, "epoch": 4.539349422875131, "percentage": 64.84, "elapsed_time": "10:04:24", "remaining_time": "5:27:45"} +{"current_steps": 2170, "total_steps": 3339, "loss": 0.1563, "lr": 1.3186893403463954e-05, "epoch": 4.5498426023085, "percentage": 64.99, "elapsed_time": "10:05:47", "remaining_time": "5:26:20"} +{"current_steps": 2175, "total_steps": 3339, "loss": 0.1572, "lr": 1.3088694470361523e-05, "epoch": 4.560335781741868, "percentage": 65.14, "elapsed_time": "10:07:09", "remaining_time": "5:24:56"} +{"current_steps": 2180, "total_steps": 3339, "loss": 0.1595, "lr": 1.2990684384178538e-05, "epoch": 4.570828961175236, "percentage": 65.29, "elapsed_time": "10:08:32", "remaining_time": "5:23:32"} +{"current_steps": 2185, "total_steps": 3339, "loss": 0.16, "lr": 1.289286582297674e-05, "epoch": 4.581322140608604, "percentage": 65.44, "elapsed_time": "10:09:56", "remaining_time": "5:22:08"} +{"current_steps": 2190, "total_steps": 3339, "loss": 0.1583, "lr": 1.279524145958455e-05, "epoch": 4.591815320041973, "percentage": 65.59, "elapsed_time": "10:11:20", "remaining_time": "5:20:44"} +{"current_steps": 2195, "total_steps": 3339, "loss": 0.1604, "lr": 1.2697813961524088e-05, "epoch": 4.602308499475341, "percentage": 65.74, "elapsed_time": "10:12:43", "remaining_time": "5:19:20"} +{"current_steps": 2200, "total_steps": 3339, "loss": 0.1551, "lr": 1.260058599093824e-05, "epoch": 4.612801678908709, "percentage": 65.89, "elapsed_time": "10:14:05", "remaining_time": "5:17:56"} +{"current_steps": 2205, "total_steps": 3339, "loss": 0.1563, "lr": 1.2503560204517937e-05, "epoch": 4.623294858342078, "percentage": 66.04, "elapsed_time": "10:15:28", "remaining_time": "5:16:31"} +{"current_steps": 2210, "total_steps": 3339, "loss": 0.1557, "lr": 1.2406739253429558e-05, "epoch": 4.633788037775446, "percentage": 66.19, "elapsed_time": "10:16:50", "remaining_time": "5:15:07"} +{"current_steps": 2215, "total_steps": 3339, "loss": 0.1485, "lr": 1.2310125783242497e-05, "epoch": 4.644281217208814, "percentage": 66.34, "elapsed_time": "10:18:12", "remaining_time": "5:13:42"} +{"current_steps": 2220, "total_steps": 3339, "loss": 0.156, "lr": 1.2213722433856865e-05, "epoch": 4.654774396642183, "percentage": 66.49, "elapsed_time": "10:19:34", "remaining_time": "5:12:17"} +{"current_steps": 2225, "total_steps": 3339, "loss": 0.1608, "lr": 1.2117531839431349e-05, "epoch": 4.665267576075551, "percentage": 66.64, "elapsed_time": "10:20:57", "remaining_time": "5:10:53"} +{"current_steps": 2230, "total_steps": 3339, "loss": 0.1591, "lr": 1.2021556628311266e-05, "epoch": 4.675760755508919, "percentage": 66.79, "elapsed_time": "10:22:20", "remaining_time": "5:09:30"} +{"current_steps": 2235, "total_steps": 3339, "loss": 0.1543, "lr": 1.19257994229567e-05, "epoch": 4.686253934942288, "percentage": 66.94, "elapsed_time": "10:23:43", "remaining_time": "5:08:05"} +{"current_steps": 2240, "total_steps": 3339, "loss": 0.1554, "lr": 1.183026283987089e-05, "epoch": 4.696747114375656, "percentage": 67.09, "elapsed_time": "10:25:05", "remaining_time": "5:06:41"} +{"current_steps": 2245, "total_steps": 3339, "loss": 0.1562, "lr": 1.17349494895287e-05, "epoch": 4.707240293809024, "percentage": 67.24, "elapsed_time": "10:26:28", "remaining_time": "5:05:17"} +{"current_steps": 2250, "total_steps": 3339, "loss": 0.1585, "lr": 1.1639861976305315e-05, "epoch": 4.717733473242393, "percentage": 67.39, "elapsed_time": "10:27:50", "remaining_time": "5:03:52"} +{"current_steps": 2255, "total_steps": 3339, "loss": 0.1546, "lr": 1.1545002898405064e-05, "epoch": 4.728226652675761, "percentage": 67.54, "elapsed_time": "10:29:13", "remaining_time": "5:02:28"} +{"current_steps": 2260, "total_steps": 3339, "loss": 0.1549, "lr": 1.1450374847790434e-05, "epoch": 4.738719832109129, "percentage": 67.68, "elapsed_time": "10:30:37", "remaining_time": "5:01:04"} +{"current_steps": 2265, "total_steps": 3339, "loss": 0.1557, "lr": 1.1355980410111232e-05, "epoch": 4.749213011542498, "percentage": 67.83, "elapsed_time": "10:32:00", "remaining_time": "4:59:40"} +{"current_steps": 2270, "total_steps": 3339, "loss": 0.1567, "lr": 1.1261822164633962e-05, "epoch": 4.759706190975866, "percentage": 67.98, "elapsed_time": "10:33:22", "remaining_time": "4:58:16"} +{"current_steps": 2275, "total_steps": 3339, "loss": 0.1546, "lr": 1.1167902684171299e-05, "epoch": 4.770199370409234, "percentage": 68.13, "elapsed_time": "10:34:45", "remaining_time": "4:56:52"} +{"current_steps": 2280, "total_steps": 3339, "loss": 0.1538, "lr": 1.1074224535011853e-05, "epoch": 4.780692549842602, "percentage": 68.28, "elapsed_time": "10:36:07", "remaining_time": "4:55:28"} +{"current_steps": 2285, "total_steps": 3339, "loss": 0.1527, "lr": 1.098079027684999e-05, "epoch": 4.791185729275971, "percentage": 68.43, "elapsed_time": "10:37:30", "remaining_time": "4:54:03"} +{"current_steps": 2290, "total_steps": 3339, "loss": 0.1563, "lr": 1.0887602462715931e-05, "epoch": 4.801678908709339, "percentage": 68.58, "elapsed_time": "10:38:53", "remaining_time": "4:52:39"} +{"current_steps": 2295, "total_steps": 3339, "loss": 0.1517, "lr": 1.0794663638905974e-05, "epoch": 4.812172088142708, "percentage": 68.73, "elapsed_time": "10:40:15", "remaining_time": "4:51:15"} +{"current_steps": 2300, "total_steps": 3339, "loss": 0.1528, "lr": 1.0701976344912886e-05, "epoch": 4.822665267576076, "percentage": 68.88, "elapsed_time": "10:41:36", "remaining_time": "4:49:50"} +{"current_steps": 2305, "total_steps": 3339, "loss": 0.1568, "lr": 1.0609543113356585e-05, "epoch": 4.833158447009444, "percentage": 69.03, "elapsed_time": "10:42:59", "remaining_time": "4:48:26"} +{"current_steps": 2310, "total_steps": 3339, "loss": 0.1538, "lr": 1.0517366469914883e-05, "epoch": 4.843651626442812, "percentage": 69.18, "elapsed_time": "10:44:20", "remaining_time": "4:47:01"} +{"current_steps": 2315, "total_steps": 3339, "loss": 0.154, "lr": 1.0425448933254496e-05, "epoch": 4.854144805876181, "percentage": 69.33, "elapsed_time": "10:45:40", "remaining_time": "4:45:36"} +{"current_steps": 2320, "total_steps": 3339, "loss": 0.1488, "lr": 1.0333793014962197e-05, "epoch": 4.864637985309549, "percentage": 69.48, "elapsed_time": "10:47:02", "remaining_time": "4:44:11"} +{"current_steps": 2325, "total_steps": 3339, "loss": 0.1537, "lr": 1.0242401219476226e-05, "epoch": 4.875131164742917, "percentage": 69.63, "elapsed_time": "10:48:24", "remaining_time": "4:42:47"} +{"current_steps": 2330, "total_steps": 3339, "loss": 0.1656, "lr": 1.0151276044017836e-05, "epoch": 4.885624344176286, "percentage": 69.78, "elapsed_time": "10:49:46", "remaining_time": "4:41:23"} +{"current_steps": 2335, "total_steps": 3339, "loss": 0.156, "lr": 1.0060419978523055e-05, "epoch": 4.896117523609654, "percentage": 69.93, "elapsed_time": "10:51:09", "remaining_time": "4:39:58"} +{"current_steps": 2340, "total_steps": 3339, "loss": 0.1558, "lr": 9.969835505574677e-06, "epoch": 4.906610703043022, "percentage": 70.08, "elapsed_time": "10:52:31", "remaining_time": "4:38:34"} +{"current_steps": 2345, "total_steps": 3339, "loss": 0.1487, "lr": 9.879525100334365e-06, "epoch": 4.917103882476391, "percentage": 70.23, "elapsed_time": "10:53:53", "remaining_time": "4:37:10"} +{"current_steps": 2350, "total_steps": 3339, "loss": 0.1568, "lr": 9.789491230475104e-06, "epoch": 4.927597061909759, "percentage": 70.38, "elapsed_time": "10:55:15", "remaining_time": "4:35:46"} +{"current_steps": 2355, "total_steps": 3339, "loss": 0.1622, "lr": 9.699736356113705e-06, "epoch": 4.938090241343127, "percentage": 70.53, "elapsed_time": "10:56:35", "remaining_time": "4:34:20"} +{"current_steps": 2360, "total_steps": 3339, "loss": 0.1539, "lr": 9.610262929743625e-06, "epoch": 4.948583420776496, "percentage": 70.68, "elapsed_time": "10:57:57", "remaining_time": "4:32:56"} +{"current_steps": 2365, "total_steps": 3339, "loss": 0.1489, "lr": 9.521073396167916e-06, "epoch": 4.959076600209864, "percentage": 70.83, "elapsed_time": "10:59:20", "remaining_time": "4:31:32"} +{"current_steps": 2370, "total_steps": 3339, "loss": 0.1617, "lr": 9.432170192432466e-06, "epoch": 4.969569779643232, "percentage": 70.98, "elapsed_time": "11:00:43", "remaining_time": "4:30:08"} +{"current_steps": 2375, "total_steps": 3339, "loss": 0.1512, "lr": 9.343555747759376e-06, "epoch": 4.980062959076601, "percentage": 71.13, "elapsed_time": "11:02:04", "remaining_time": "4:28:43"} +{"current_steps": 2380, "total_steps": 3339, "loss": 0.1511, "lr": 9.255232483480594e-06, "epoch": 4.990556138509969, "percentage": 71.28, "elapsed_time": "11:03:26", "remaining_time": "4:27:19"} +{"current_steps": 2385, "total_steps": 3339, "loss": 0.1436, "lr": 9.167202812971767e-06, "epoch": 5.0, "percentage": 71.43, "elapsed_time": "11:04:41", "remaining_time": "4:25:52"} +{"current_steps": 2390, "total_steps": 3339, "loss": 0.1508, "lr": 9.079469141586255e-06, "epoch": 5.010493179433368, "percentage": 71.58, "elapsed_time": "11:06:04", "remaining_time": "4:24:28"} +{"current_steps": 2395, "total_steps": 3339, "loss": 0.1448, "lr": 8.99203386658946e-06, "epoch": 5.020986358866737, "percentage": 71.73, "elapsed_time": "11:07:26", "remaining_time": "4:23:04"} +{"current_steps": 2400, "total_steps": 3339, "loss": 0.1531, "lr": 8.904899377093289e-06, "epoch": 5.031479538300105, "percentage": 71.88, "elapsed_time": "11:08:47", "remaining_time": "4:21:39"} +{"current_steps": 2405, "total_steps": 3339, "loss": 0.1478, "lr": 8.81806805399088e-06, "epoch": 5.041972717733473, "percentage": 72.03, "elapsed_time": "11:10:08", "remaining_time": "4:20:15"} +{"current_steps": 2410, "total_steps": 3339, "loss": 0.1474, "lr": 8.73154226989155e-06, "epoch": 5.052465897166842, "percentage": 72.18, "elapsed_time": "11:11:28", "remaining_time": "4:18:50"} +{"current_steps": 2415, "total_steps": 3339, "loss": 0.1449, "lr": 8.64532438905596e-06, "epoch": 5.06295907660021, "percentage": 72.33, "elapsed_time": "11:12:50", "remaining_time": "4:17:26"} +{"current_steps": 2420, "total_steps": 3339, "loss": 0.1482, "lr": 8.559416767331514e-06, "epoch": 5.073452256033578, "percentage": 72.48, "elapsed_time": "11:14:13", "remaining_time": "4:16:02"} +{"current_steps": 2425, "total_steps": 3339, "loss": 0.1495, "lr": 8.473821752087998e-06, "epoch": 5.083945435466947, "percentage": 72.63, "elapsed_time": "11:15:35", "remaining_time": "4:14:37"} +{"current_steps": 2430, "total_steps": 3339, "loss": 0.1521, "lr": 8.388541682153403e-06, "epoch": 5.094438614900315, "percentage": 72.78, "elapsed_time": "11:16:56", "remaining_time": "4:13:13"} +{"current_steps": 2435, "total_steps": 3339, "loss": 0.1459, "lr": 8.303578887750073e-06, "epoch": 5.104931794333683, "percentage": 72.93, "elapsed_time": "11:18:18", "remaining_time": "4:11:49"} +{"current_steps": 2440, "total_steps": 3339, "loss": 0.1489, "lr": 8.218935690430991e-06, "epoch": 5.115424973767052, "percentage": 73.08, "elapsed_time": "11:19:39", "remaining_time": "4:10:25"} +{"current_steps": 2445, "total_steps": 3339, "loss": 0.1467, "lr": 8.134614403016364e-06, "epoch": 5.12591815320042, "percentage": 73.23, "elapsed_time": "11:21:01", "remaining_time": "4:09:00"} +{"current_steps": 2450, "total_steps": 3339, "loss": 0.156, "lr": 8.050617329530421e-06, "epoch": 5.136411332633788, "percentage": 73.38, "elapsed_time": "11:22:23", "remaining_time": "4:07:36"} +{"current_steps": 2455, "total_steps": 3339, "loss": 0.1454, "lr": 7.966946765138445e-06, "epoch": 5.146904512067156, "percentage": 73.53, "elapsed_time": "11:23:44", "remaining_time": "4:06:12"} +{"current_steps": 2460, "total_steps": 3339, "loss": 0.1484, "lr": 7.883604996084083e-06, "epoch": 5.157397691500525, "percentage": 73.67, "elapsed_time": "11:25:05", "remaining_time": "4:04:47"} +{"current_steps": 2465, "total_steps": 3339, "loss": 0.1508, "lr": 7.800594299626858e-06, "epoch": 5.167890870933893, "percentage": 73.82, "elapsed_time": "11:26:27", "remaining_time": "4:03:23"} +{"current_steps": 2470, "total_steps": 3339, "loss": 0.1496, "lr": 7.717916943979964e-06, "epoch": 5.178384050367261, "percentage": 73.97, "elapsed_time": "11:27:49", "remaining_time": "4:01:59"} +{"current_steps": 2475, "total_steps": 3339, "loss": 0.1498, "lr": 7.635575188248245e-06, "epoch": 5.18887722980063, "percentage": 74.12, "elapsed_time": "11:29:09", "remaining_time": "4:00:34"} +{"current_steps": 2480, "total_steps": 3339, "loss": 0.1522, "lr": 7.553571282366521e-06, "epoch": 5.199370409233998, "percentage": 74.27, "elapsed_time": "11:30:30", "remaining_time": "3:59:10"} +{"current_steps": 2485, "total_steps": 3339, "loss": 0.146, "lr": 7.47190746703808e-06, "epoch": 5.209863588667366, "percentage": 74.42, "elapsed_time": "11:31:50", "remaining_time": "3:57:45"} +{"current_steps": 2490, "total_steps": 3339, "loss": 0.1433, "lr": 7.390585973673452e-06, "epoch": 5.220356768100735, "percentage": 74.57, "elapsed_time": "11:33:12", "remaining_time": "3:56:21"} +{"current_steps": 2495, "total_steps": 3339, "loss": 0.1477, "lr": 7.309609024329452e-06, "epoch": 5.230849947534103, "percentage": 74.72, "elapsed_time": "11:34:33", "remaining_time": "3:54:57"} +{"current_steps": 2500, "total_steps": 3339, "loss": 0.1523, "lr": 7.2289788316484365e-06, "epoch": 5.241343126967471, "percentage": 74.87, "elapsed_time": "11:35:55", "remaining_time": "3:53:33"} +{"current_steps": 2505, "total_steps": 3339, "loss": 0.1504, "lr": 7.14869759879788e-06, "epoch": 5.25183630640084, "percentage": 75.02, "elapsed_time": "11:37:17", "remaining_time": "3:52:09"} +{"current_steps": 2510, "total_steps": 3339, "loss": 0.1495, "lr": 7.068767519410149e-06, "epoch": 5.262329485834208, "percentage": 75.17, "elapsed_time": "11:38:39", "remaining_time": "3:50:45"} +{"current_steps": 2515, "total_steps": 3339, "loss": 0.1525, "lr": 6.989190777522577e-06, "epoch": 5.272822665267576, "percentage": 75.32, "elapsed_time": "11:40:00", "remaining_time": "3:49:20"} +{"current_steps": 2520, "total_steps": 3339, "loss": 0.1442, "lr": 6.9099695475177655e-06, "epoch": 5.283315844700945, "percentage": 75.47, "elapsed_time": "11:41:21", "remaining_time": "3:47:56"} +{"current_steps": 2525, "total_steps": 3339, "loss": 0.1463, "lr": 6.8311059940641975e-06, "epoch": 5.293809024134313, "percentage": 75.62, "elapsed_time": "11:42:42", "remaining_time": "3:46:32"} +{"current_steps": 2530, "total_steps": 3339, "loss": 0.1554, "lr": 6.752602272057081e-06, "epoch": 5.304302203567681, "percentage": 75.77, "elapsed_time": "11:44:03", "remaining_time": "3:45:08"} +{"current_steps": 2535, "total_steps": 3339, "loss": 0.1458, "lr": 6.674460526559452e-06, "epoch": 5.31479538300105, "percentage": 75.92, "elapsed_time": "11:45:25", "remaining_time": "3:43:44"} +{"current_steps": 2540, "total_steps": 3339, "loss": 0.1401, "lr": 6.596682892743593e-06, "epoch": 5.325288562434418, "percentage": 76.07, "elapsed_time": "11:46:47", "remaining_time": "3:42:20"} +{"current_steps": 2545, "total_steps": 3339, "loss": 0.148, "lr": 6.519271495832642e-06, "epoch": 5.335781741867786, "percentage": 76.22, "elapsed_time": "11:48:09", "remaining_time": "3:40:55"} +{"current_steps": 2550, "total_steps": 3339, "loss": 0.1515, "lr": 6.44222845104258e-06, "epoch": 5.346274921301154, "percentage": 76.37, "elapsed_time": "11:49:31", "remaining_time": "3:39:32"} +{"current_steps": 2555, "total_steps": 3339, "loss": 0.1439, "lr": 6.365555863524395e-06, "epoch": 5.356768100734523, "percentage": 76.52, "elapsed_time": "11:50:52", "remaining_time": "3:38:07"} +{"current_steps": 2560, "total_steps": 3339, "loss": 0.1443, "lr": 6.289255828306573e-06, "epoch": 5.367261280167891, "percentage": 76.67, "elapsed_time": "11:52:13", "remaining_time": "3:36:43"} +{"current_steps": 2565, "total_steps": 3339, "loss": 0.1471, "lr": 6.213330430237856e-06, "epoch": 5.37775445960126, "percentage": 76.82, "elapsed_time": "11:53:35", "remaining_time": "3:35:19"} +{"current_steps": 2570, "total_steps": 3339, "loss": 0.1501, "lr": 6.137781743930256e-06, "epoch": 5.388247639034628, "percentage": 76.97, "elapsed_time": "11:54:56", "remaining_time": "3:33:55"} +{"current_steps": 2575, "total_steps": 3339, "loss": 0.1506, "lr": 6.062611833702392e-06, "epoch": 5.398740818467996, "percentage": 77.12, "elapsed_time": "11:56:18", "remaining_time": "3:32:31"} +{"current_steps": 2580, "total_steps": 3339, "loss": 0.1509, "lr": 5.987822753523076e-06, "epoch": 5.409233997901364, "percentage": 77.27, "elapsed_time": "11:57:40", "remaining_time": "3:31:07"} +{"current_steps": 2585, "total_steps": 3339, "loss": 0.1484, "lr": 5.9134165469551886e-06, "epoch": 5.419727177334733, "percentage": 77.42, "elapsed_time": "11:59:02", "remaining_time": "3:29:43"} +{"current_steps": 2590, "total_steps": 3339, "loss": 0.148, "lr": 5.839395247099819e-06, "epoch": 5.430220356768101, "percentage": 77.57, "elapsed_time": "12:00:24", "remaining_time": "3:28:20"} +{"current_steps": 2595, "total_steps": 3339, "loss": 0.1537, "lr": 5.765760876540756e-06, "epoch": 5.440713536201469, "percentage": 77.72, "elapsed_time": "12:01:46", "remaining_time": "3:26:56"} +{"current_steps": 2600, "total_steps": 3339, "loss": 0.1458, "lr": 5.692515447289184e-06, "epoch": 5.451206715634838, "percentage": 77.87, "elapsed_time": "12:03:09", "remaining_time": "3:25:32"} +{"current_steps": 2605, "total_steps": 3339, "loss": 0.1437, "lr": 5.6196609607287215e-06, "epoch": 5.461699895068206, "percentage": 78.02, "elapsed_time": "12:04:30", "remaining_time": "3:24:08"} +{"current_steps": 2610, "total_steps": 3339, "loss": 0.1493, "lr": 5.5471994075607435e-06, "epoch": 5.472193074501574, "percentage": 78.17, "elapsed_time": "12:05:51", "remaining_time": "3:22:44"} +{"current_steps": 2615, "total_steps": 3339, "loss": 0.1488, "lr": 5.475132767749951e-06, "epoch": 5.482686253934943, "percentage": 78.32, "elapsed_time": "12:07:13", "remaining_time": "3:21:20"} +{"current_steps": 2620, "total_steps": 3339, "loss": 0.1462, "lr": 5.4034630104703204e-06, "epoch": 5.493179433368311, "percentage": 78.47, "elapsed_time": "12:08:35", "remaining_time": "3:19:56"} +{"current_steps": 2625, "total_steps": 3339, "loss": 0.1466, "lr": 5.332192094051256e-06, "epoch": 5.503672612801679, "percentage": 78.62, "elapsed_time": "12:09:56", "remaining_time": "3:18:32"} +{"current_steps": 2630, "total_steps": 3339, "loss": 0.1527, "lr": 5.261321965924114e-06, "epoch": 5.5141657922350475, "percentage": 78.77, "elapsed_time": "12:11:18", "remaining_time": "3:17:08"} +{"current_steps": 2635, "total_steps": 3339, "loss": 0.1432, "lr": 5.190854562568941e-06, "epoch": 5.524658971668416, "percentage": 78.92, "elapsed_time": "12:12:39", "remaining_time": "3:15:44"} +{"current_steps": 2640, "total_steps": 3339, "loss": 0.1457, "lr": 5.120791809461621e-06, "epoch": 5.535152151101784, "percentage": 79.07, "elapsed_time": "12:14:01", "remaining_time": "3:14:20"} +{"current_steps": 2645, "total_steps": 3339, "loss": 0.148, "lr": 5.051135621021217e-06, "epoch": 5.5456453305351525, "percentage": 79.22, "elapsed_time": "12:15:23", "remaining_time": "3:12:57"} +{"current_steps": 2650, "total_steps": 3339, "loss": 0.141, "lr": 4.981887900557685e-06, "epoch": 5.5561385099685205, "percentage": 79.37, "elapsed_time": "12:16:45", "remaining_time": "3:11:33"} +{"current_steps": 2655, "total_steps": 3339, "loss": 0.1449, "lr": 4.913050540219864e-06, "epoch": 5.566631689401889, "percentage": 79.51, "elapsed_time": "12:18:07", "remaining_time": "3:10:09"} +{"current_steps": 2660, "total_steps": 3339, "loss": 0.1397, "lr": 4.844625420943752e-06, "epoch": 5.5771248688352575, "percentage": 79.66, "elapsed_time": "12:19:26", "remaining_time": "3:08:45"} +{"current_steps": 2665, "total_steps": 3339, "loss": 0.1466, "lr": 4.776614412401144e-06, "epoch": 5.5876180482686255, "percentage": 79.81, "elapsed_time": "12:20:46", "remaining_time": "3:07:20"} +{"current_steps": 2670, "total_steps": 3339, "loss": 0.1482, "lr": 4.709019372948524e-06, "epoch": 5.5981112277019935, "percentage": 79.96, "elapsed_time": "12:22:07", "remaining_time": "3:05:56"} +{"current_steps": 2675, "total_steps": 3339, "loss": 0.148, "lr": 4.641842149576303e-06, "epoch": 5.608604407135362, "percentage": 80.11, "elapsed_time": "12:23:29", "remaining_time": "3:04:33"} +{"current_steps": 2680, "total_steps": 3339, "loss": 0.1487, "lr": 4.575084577858318e-06, "epoch": 5.6190975865687305, "percentage": 80.26, "elapsed_time": "12:24:50", "remaining_time": "3:03:09"} +{"current_steps": 2685, "total_steps": 3339, "loss": 0.1455, "lr": 4.50874848190171e-06, "epoch": 5.6295907660020985, "percentage": 80.41, "elapsed_time": "12:26:12", "remaining_time": "3:01:45"} +{"current_steps": 2690, "total_steps": 3339, "loss": 0.1544, "lr": 4.4428356742970744e-06, "epoch": 5.640083945435467, "percentage": 80.56, "elapsed_time": "12:27:34", "remaining_time": "3:00:21"} +{"current_steps": 2695, "total_steps": 3339, "loss": 0.1515, "lr": 4.37734795606892e-06, "epoch": 5.6505771248688355, "percentage": 80.71, "elapsed_time": "12:28:56", "remaining_time": "2:58:58"} +{"current_steps": 2700, "total_steps": 3339, "loss": 0.1445, "lr": 4.312287116626479e-06, "epoch": 5.6610703043022035, "percentage": 80.86, "elapsed_time": "12:30:18", "remaining_time": "2:57:34"} +{"current_steps": 2705, "total_steps": 3339, "loss": 0.1446, "lr": 4.247654933714771e-06, "epoch": 5.6715634837355715, "percentage": 81.01, "elapsed_time": "12:31:40", "remaining_time": "2:56:10"} +{"current_steps": 2710, "total_steps": 3339, "loss": 0.1557, "lr": 4.1834531733660835e-06, "epoch": 5.68205666316894, "percentage": 81.16, "elapsed_time": "12:33:02", "remaining_time": "2:54:47"} +{"current_steps": 2715, "total_steps": 3339, "loss": 0.1509, "lr": 4.119683589851673e-06, "epoch": 5.6925498426023085, "percentage": 81.31, "elapsed_time": "12:34:23", "remaining_time": "2:53:23"} +{"current_steps": 2720, "total_steps": 3339, "loss": 0.1521, "lr": 4.056347925633855e-06, "epoch": 5.7030430220356765, "percentage": 81.46, "elapsed_time": "12:35:45", "remaining_time": "2:51:59"} +{"current_steps": 2725, "total_steps": 3339, "loss": 0.1484, "lr": 3.993447911318378e-06, "epoch": 5.713536201469045, "percentage": 81.61, "elapsed_time": "12:37:07", "remaining_time": "2:50:35"} +{"current_steps": 2730, "total_steps": 3339, "loss": 0.1517, "lr": 3.930985265607135e-06, "epoch": 5.724029380902413, "percentage": 81.76, "elapsed_time": "12:38:29", "remaining_time": "2:49:12"} +{"current_steps": 2735, "total_steps": 3339, "loss": 0.1478, "lr": 3.868961695251212e-06, "epoch": 5.7345225603357814, "percentage": 81.91, "elapsed_time": "12:39:51", "remaining_time": "2:47:48"} +{"current_steps": 2740, "total_steps": 3339, "loss": 0.1493, "lr": 3.8073788950042546e-06, "epoch": 5.74501573976915, "percentage": 82.06, "elapsed_time": "12:41:12", "remaining_time": "2:46:24"} +{"current_steps": 2745, "total_steps": 3339, "loss": 0.1475, "lr": 3.7462385475761463e-06, "epoch": 5.755508919202518, "percentage": 82.21, "elapsed_time": "12:42:34", "remaining_time": "2:45:00"} +{"current_steps": 2750, "total_steps": 3339, "loss": 0.1489, "lr": 3.6855423235870256e-06, "epoch": 5.766002098635886, "percentage": 82.36, "elapsed_time": "12:43:56", "remaining_time": "2:43:37"} +{"current_steps": 2755, "total_steps": 3339, "loss": 0.1511, "lr": 3.625291881521662e-06, "epoch": 5.776495278069255, "percentage": 82.51, "elapsed_time": "12:45:17", "remaining_time": "2:42:13"} +{"current_steps": 2760, "total_steps": 3339, "loss": 0.1464, "lr": 3.5654888676841214e-06, "epoch": 5.786988457502623, "percentage": 82.66, "elapsed_time": "12:46:38", "remaining_time": "2:40:49"} +{"current_steps": 2765, "total_steps": 3339, "loss": 0.1431, "lr": 3.506134916152779e-06, "epoch": 5.797481636935991, "percentage": 82.81, "elapsed_time": "12:48:00", "remaining_time": "2:39:26"} +{"current_steps": 2770, "total_steps": 3339, "loss": 0.152, "lr": 3.4472316487356873e-06, "epoch": 5.807974816369359, "percentage": 82.96, "elapsed_time": "12:49:22", "remaining_time": "2:38:02"} +{"current_steps": 2775, "total_steps": 3339, "loss": 0.1462, "lr": 3.3887806749262285e-06, "epoch": 5.818467995802728, "percentage": 83.11, "elapsed_time": "12:50:44", "remaining_time": "2:36:38"} +{"current_steps": 2780, "total_steps": 3339, "loss": 0.1505, "lr": 3.3307835918591703e-06, "epoch": 5.828961175236096, "percentage": 83.26, "elapsed_time": "12:52:05", "remaining_time": "2:35:15"} +{"current_steps": 2785, "total_steps": 3339, "loss": 0.1518, "lr": 3.273241984267026e-06, "epoch": 5.839454354669465, "percentage": 83.41, "elapsed_time": "12:53:25", "remaining_time": "2:33:51"} +{"current_steps": 2790, "total_steps": 3339, "loss": 0.1432, "lr": 3.2161574244367056e-06, "epoch": 5.849947534102833, "percentage": 83.56, "elapsed_time": "12:54:47", "remaining_time": "2:32:27"} +{"current_steps": 2795, "total_steps": 3339, "loss": 0.1523, "lr": 3.1595314721666103e-06, "epoch": 5.860440713536201, "percentage": 83.71, "elapsed_time": "12:56:08", "remaining_time": "2:31:03"} +{"current_steps": 2800, "total_steps": 3339, "loss": 0.1457, "lr": 3.1033656747239836e-06, "epoch": 5.870933892969569, "percentage": 83.86, "elapsed_time": "12:57:30", "remaining_time": "2:29:40"} +{"current_steps": 2805, "total_steps": 3339, "loss": 0.1434, "lr": 3.0476615668026334e-06, "epoch": 5.881427072402938, "percentage": 84.01, "elapsed_time": "12:58:51", "remaining_time": "2:28:16"} +{"current_steps": 2810, "total_steps": 3339, "loss": 0.1455, "lr": 2.9924206704810153e-06, "epoch": 5.891920251836306, "percentage": 84.16, "elapsed_time": "13:00:12", "remaining_time": "2:26:52"} +{"current_steps": 2815, "total_steps": 3339, "loss": 0.1369, "lr": 2.937644495180607e-06, "epoch": 5.902413431269674, "percentage": 84.31, "elapsed_time": "13:01:34", "remaining_time": "2:25:29"} +{"current_steps": 2820, "total_steps": 3339, "loss": 0.1484, "lr": 2.883334537624707e-06, "epoch": 5.912906610703043, "percentage": 84.46, "elapsed_time": "13:02:54", "remaining_time": "2:24:05"} +{"current_steps": 2825, "total_steps": 3339, "loss": 0.1507, "lr": 2.8294922817975125e-06, "epoch": 5.923399790136411, "percentage": 84.61, "elapsed_time": "13:04:16", "remaining_time": "2:22:41"} +{"current_steps": 2830, "total_steps": 3339, "loss": 0.1451, "lr": 2.7761191989035753e-06, "epoch": 5.933892969569779, "percentage": 84.76, "elapsed_time": "13:05:38", "remaining_time": "2:21:18"} +{"current_steps": 2835, "total_steps": 3339, "loss": 0.1496, "lr": 2.723216747327597e-06, "epoch": 5.944386149003148, "percentage": 84.91, "elapsed_time": "13:07:00", "remaining_time": "2:19:54"} +{"current_steps": 2840, "total_steps": 3339, "loss": 0.1487, "lr": 2.6707863725945913e-06, "epoch": 5.954879328436516, "percentage": 85.06, "elapsed_time": "13:08:23", "remaining_time": "2:18:31"} +{"current_steps": 2845, "total_steps": 3339, "loss": 0.1413, "lr": 2.6188295073303825e-06, "epoch": 5.965372507869884, "percentage": 85.21, "elapsed_time": "13:09:43", "remaining_time": "2:17:07"} +{"current_steps": 2850, "total_steps": 3339, "loss": 0.1487, "lr": 2.5673475712224493e-06, "epoch": 5.975865687303253, "percentage": 85.35, "elapsed_time": "13:11:05", "remaining_time": "2:15:44"} +{"current_steps": 2855, "total_steps": 3339, "loss": 0.1456, "lr": 2.5163419709811487e-06, "epoch": 5.986358866736621, "percentage": 85.5, "elapsed_time": "13:12:27", "remaining_time": "2:14:20"} +{"current_steps": 2860, "total_steps": 3339, "loss": 0.1424, "lr": 2.4658141003012626e-06, "epoch": 5.996852046169989, "percentage": 85.65, "elapsed_time": "13:13:49", "remaining_time": "2:12:57"} +{"current_steps": 2865, "total_steps": 3339, "loss": 0.1425, "lr": 2.415765339823928e-06, "epoch": 6.006295907660021, "percentage": 85.8, "elapsed_time": "13:15:02", "remaining_time": "2:11:32"} +{"current_steps": 2870, "total_steps": 3339, "loss": 0.1441, "lr": 2.366197057098909e-06, "epoch": 6.0167890870933896, "percentage": 85.95, "elapsed_time": "13:16:24", "remaining_time": "2:10:08"} +{"current_steps": 2875, "total_steps": 3339, "loss": 0.1452, "lr": 2.317110606547228e-06, "epoch": 6.027282266526758, "percentage": 86.1, "elapsed_time": "13:17:46", "remaining_time": "2:08:45"} +{"current_steps": 2880, "total_steps": 3339, "loss": 0.1433, "lr": 2.2685073294241587e-06, "epoch": 6.037775445960126, "percentage": 86.25, "elapsed_time": "13:19:08", "remaining_time": "2:07:21"} +{"current_steps": 2885, "total_steps": 3339, "loss": 0.1432, "lr": 2.2203885537825664e-06, "epoch": 6.0482686253934945, "percentage": 86.4, "elapsed_time": "13:20:31", "remaining_time": "2:05:58"} +{"current_steps": 2890, "total_steps": 3339, "loss": 0.1487, "lr": 2.172755594436642e-06, "epoch": 6.0587618048268626, "percentage": 86.55, "elapsed_time": "13:21:52", "remaining_time": "2:04:34"} +{"current_steps": 2895, "total_steps": 3339, "loss": 0.1378, "lr": 2.125609752925952e-06, "epoch": 6.069254984260231, "percentage": 86.7, "elapsed_time": "13:23:13", "remaining_time": "2:03:11"} +{"current_steps": 2900, "total_steps": 3339, "loss": 0.1416, "lr": 2.078952317479899e-06, "epoch": 6.0797481636935995, "percentage": 86.85, "elapsed_time": "13:24:36", "remaining_time": "2:01:48"} +{"current_steps": 2905, "total_steps": 3339, "loss": 0.146, "lr": 2.0327845629824883e-06, "epoch": 6.0902413431269675, "percentage": 87.0, "elapsed_time": "13:25:57", "remaining_time": "2:00:24"} +{"current_steps": 2910, "total_steps": 3339, "loss": 0.1414, "lr": 1.9871077509375316e-06, "epoch": 6.1007345225603355, "percentage": 87.15, "elapsed_time": "13:27:19", "remaining_time": "1:59:01"} +{"current_steps": 2915, "total_steps": 3339, "loss": 0.1424, "lr": 1.9419231294341467e-06, "epoch": 6.1112277019937045, "percentage": 87.3, "elapsed_time": "13:28:40", "remaining_time": "1:57:37"} +{"current_steps": 2920, "total_steps": 3339, "loss": 0.1453, "lr": 1.8972319331126686e-06, "epoch": 6.1217208814270725, "percentage": 87.45, "elapsed_time": "13:30:02", "remaining_time": "1:56:14"} +{"current_steps": 2925, "total_steps": 3339, "loss": 0.1489, "lr": 1.853035383130919e-06, "epoch": 6.1322140608604405, "percentage": 87.6, "elapsed_time": "13:31:25", "remaining_time": "1:54:50"} +{"current_steps": 2930, "total_steps": 3339, "loss": 0.1488, "lr": 1.8093346871308104e-06, "epoch": 6.142707240293809, "percentage": 87.75, "elapsed_time": "13:32:47", "remaining_time": "1:53:27"} +{"current_steps": 2935, "total_steps": 3339, "loss": 0.1451, "lr": 1.7661310392053877e-06, "epoch": 6.1532004197271775, "percentage": 87.9, "elapsed_time": "13:34:09", "remaining_time": "1:52:04"} +{"current_steps": 2940, "total_steps": 3339, "loss": 0.1442, "lr": 1.7234256198661748e-06, "epoch": 6.1636935991605455, "percentage": 88.05, "elapsed_time": "13:35:31", "remaining_time": "1:50:40"} +{"current_steps": 2945, "total_steps": 3339, "loss": 0.1537, "lr": 1.6812195960109257e-06, "epoch": 6.1741867785939135, "percentage": 88.2, "elapsed_time": "13:36:52", "remaining_time": "1:49:17"} +{"current_steps": 2950, "total_steps": 3339, "loss": 0.1447, "lr": 1.6395141208917297e-06, "epoch": 6.184679958027282, "percentage": 88.35, "elapsed_time": "13:38:13", "remaining_time": "1:47:53"} +{"current_steps": 2955, "total_steps": 3339, "loss": 0.141, "lr": 1.5983103340835194e-06, "epoch": 6.1951731374606505, "percentage": 88.5, "elapsed_time": "13:39:37", "remaining_time": "1:46:30"} +{"current_steps": 2960, "total_steps": 3339, "loss": 0.1487, "lr": 1.5576093614529141e-06, "epoch": 6.2056663168940185, "percentage": 88.65, "elapsed_time": "13:41:03", "remaining_time": "1:45:07"} +{"current_steps": 2965, "total_steps": 3339, "loss": 0.1424, "lr": 1.5174123151274668e-06, "epoch": 6.216159496327387, "percentage": 88.8, "elapsed_time": "13:42:33", "remaining_time": "1:43:45"} +{"current_steps": 2970, "total_steps": 3339, "loss": 0.1447, "lr": 1.4777202934652746e-06, "epoch": 6.226652675760755, "percentage": 88.95, "elapsed_time": "13:43:59", "remaining_time": "1:42:22"} +{"current_steps": 2975, "total_steps": 3339, "loss": 0.1432, "lr": 1.4385343810249586e-06, "epoch": 6.2371458551941235, "percentage": 89.1, "elapsed_time": "13:45:26", "remaining_time": "1:40:59"} +{"current_steps": 2980, "total_steps": 3339, "loss": 0.1476, "lr": 1.3998556485360372e-06, "epoch": 6.247639034627492, "percentage": 89.25, "elapsed_time": "13:46:50", "remaining_time": "1:39:36"} +{"current_steps": 2985, "total_steps": 3339, "loss": 0.1428, "lr": 1.3616851528696694e-06, "epoch": 6.25813221406086, "percentage": 89.4, "elapsed_time": "13:48:13", "remaining_time": "1:38:13"} +{"current_steps": 2990, "total_steps": 3339, "loss": 0.1434, "lr": 1.3240239370097752e-06, "epoch": 6.268625393494228, "percentage": 89.55, "elapsed_time": "13:49:35", "remaining_time": "1:36:49"} +{"current_steps": 2995, "total_steps": 3339, "loss": 0.143, "lr": 1.2868730300245357e-06, "epoch": 6.279118572927597, "percentage": 89.7, "elapsed_time": "13:50:59", "remaining_time": "1:35:26"} +{"current_steps": 3000, "total_steps": 3339, "loss": 0.1394, "lr": 1.250233447038267e-06, "epoch": 6.289611752360965, "percentage": 89.85, "elapsed_time": "13:52:26", "remaining_time": "1:34:03"} +{"current_steps": 3005, "total_steps": 3339, "loss": 0.1482, "lr": 1.2141061892037009e-06, "epoch": 6.300104931794333, "percentage": 90.0, "elapsed_time": "13:54:30", "remaining_time": "1:32:45"} +{"current_steps": 3010, "total_steps": 3339, "loss": 0.1499, "lr": 1.1784922436746094e-06, "epoch": 6.310598111227702, "percentage": 90.15, "elapsed_time": "13:55:54", "remaining_time": "1:31:21"} +{"current_steps": 3015, "total_steps": 3339, "loss": 0.1425, "lr": 1.1433925835788484e-06, "epoch": 6.32109129066107, "percentage": 90.3, "elapsed_time": "13:57:18", "remaining_time": "1:29:58"} +{"current_steps": 3020, "total_steps": 3339, "loss": 0.1456, "lr": 1.1088081679917506e-06, "epoch": 6.331584470094438, "percentage": 90.45, "elapsed_time": "13:58:42", "remaining_time": "1:28:35"} +{"current_steps": 3025, "total_steps": 3339, "loss": 0.1448, "lr": 1.0747399419099347e-06, "epoch": 6.342077649527807, "percentage": 90.6, "elapsed_time": "14:00:06", "remaining_time": "1:27:12"} +{"current_steps": 3030, "total_steps": 3339, "loss": 0.144, "lr": 1.0411888362254775e-06, "epoch": 6.352570828961175, "percentage": 90.75, "elapsed_time": "14:01:30", "remaining_time": "1:25:48"} +{"current_steps": 3035, "total_steps": 3339, "loss": 0.1455, "lr": 1.008155767700476e-06, "epoch": 6.363064008394543, "percentage": 90.9, "elapsed_time": "14:02:54", "remaining_time": "1:24:25"} +{"current_steps": 3040, "total_steps": 3339, "loss": 0.1452, "lr": 9.75641638942002e-07, "epoch": 6.373557187827912, "percentage": 91.05, "elapsed_time": "14:04:35", "remaining_time": "1:23:04"} +{"current_steps": 3045, "total_steps": 3339, "loss": 0.1437, "lr": 9.436473383774269e-07, "epoch": 6.38405036726128, "percentage": 91.19, "elapsed_time": "14:06:19", "remaining_time": "1:21:42"} +{"current_steps": 3050, "total_steps": 3339, "loss": 0.1426, "lr": 9.121737402301689e-07, "epoch": 6.394543546694648, "percentage": 91.34, "elapsed_time": "14:07:44", "remaining_time": "1:20:19"} +{"current_steps": 3055, "total_steps": 3339, "loss": 0.1423, "lr": 8.812217044957827e-07, "epoch": 6.405036726128017, "percentage": 91.49, "elapsed_time": "14:09:08", "remaining_time": "1:18:56"} +{"current_steps": 3060, "total_steps": 3339, "loss": 0.1472, "lr": 8.507920769184763e-07, "epoch": 6.415529905561385, "percentage": 91.64, "elapsed_time": "14:10:30", "remaining_time": "1:17:32"} +{"current_steps": 3065, "total_steps": 3339, "loss": 0.1468, "lr": 8.208856889679895e-07, "epoch": 6.426023084994753, "percentage": 91.79, "elapsed_time": "14:11:51", "remaining_time": "1:16:09"} +{"current_steps": 3070, "total_steps": 3339, "loss": 0.1447, "lr": 7.915033578168829e-07, "epoch": 6.436516264428121, "percentage": 91.94, "elapsed_time": "14:13:14", "remaining_time": "1:14:45"} +{"current_steps": 3075, "total_steps": 3339, "loss": 0.1414, "lr": 7.626458863182051e-07, "epoch": 6.44700944386149, "percentage": 92.09, "elapsed_time": "14:14:35", "remaining_time": "1:13:22"} +{"current_steps": 3080, "total_steps": 3339, "loss": 0.1444, "lr": 7.343140629835632e-07, "epoch": 6.457502623294858, "percentage": 92.24, "elapsed_time": "14:15:56", "remaining_time": "1:11:58"} +{"current_steps": 3085, "total_steps": 3339, "loss": 0.1504, "lr": 7.065086619615624e-07, "epoch": 6.467995802728226, "percentage": 92.39, "elapsed_time": "14:17:18", "remaining_time": "1:10:35"} +{"current_steps": 3090, "total_steps": 3339, "loss": 0.1505, "lr": 6.792304430166652e-07, "epoch": 6.478488982161595, "percentage": 92.54, "elapsed_time": "14:18:39", "remaining_time": "1:09:11"} +{"current_steps": 3095, "total_steps": 3339, "loss": 0.1414, "lr": 6.524801515084256e-07, "epoch": 6.488982161594963, "percentage": 92.69, "elapsed_time": "14:20:00", "remaining_time": "1:07:48"} +{"current_steps": 3100, "total_steps": 3339, "loss": 0.145, "lr": 6.262585183711256e-07, "epoch": 6.499475341028331, "percentage": 92.84, "elapsed_time": "14:21:21", "remaining_time": "1:06:24"} +{"current_steps": 3105, "total_steps": 3339, "loss": 0.1449, "lr": 6.005662600938067e-07, "epoch": 6.5099685204617, "percentage": 92.99, "elapsed_time": "14:22:42", "remaining_time": "1:05:00"} +{"current_steps": 3110, "total_steps": 3339, "loss": 0.1393, "lr": 5.754040787006765e-07, "epoch": 6.520461699895068, "percentage": 93.14, "elapsed_time": "14:24:05", "remaining_time": "1:03:37"} +{"current_steps": 3115, "total_steps": 3339, "loss": 0.1337, "lr": 5.507726617319442e-07, "epoch": 6.530954879328436, "percentage": 93.29, "elapsed_time": "14:25:26", "remaining_time": "1:02:14"} +{"current_steps": 3120, "total_steps": 3339, "loss": 0.1405, "lr": 5.266726822250268e-07, "epoch": 6.541448058761805, "percentage": 93.44, "elapsed_time": "14:26:48", "remaining_time": "1:00:50"} +{"current_steps": 3125, "total_steps": 3339, "loss": 0.1424, "lr": 5.031047986961613e-07, "epoch": 6.551941238195173, "percentage": 93.59, "elapsed_time": "14:28:08", "remaining_time": "0:59:27"} +{"current_steps": 3130, "total_steps": 3339, "loss": 0.1376, "lr": 4.800696551224038e-07, "epoch": 6.562434417628541, "percentage": 93.74, "elapsed_time": "14:29:29", "remaining_time": "0:58:03"} +{"current_steps": 3135, "total_steps": 3339, "loss": 0.1396, "lr": 4.575678809240369e-07, "epoch": 6.57292759706191, "percentage": 93.89, "elapsed_time": "14:30:49", "remaining_time": "0:56:39"} +{"current_steps": 3140, "total_steps": 3339, "loss": 0.1465, "lr": 4.35600090947379e-07, "epoch": 6.583420776495278, "percentage": 94.04, "elapsed_time": "14:32:11", "remaining_time": "0:55:16"} +{"current_steps": 3145, "total_steps": 3339, "loss": 0.1404, "lr": 4.141668854479752e-07, "epoch": 6.593913955928646, "percentage": 94.19, "elapsed_time": "14:33:32", "remaining_time": "0:53:53"} +{"current_steps": 3150, "total_steps": 3339, "loss": 0.1419, "lr": 3.9326885007419546e-07, "epoch": 6.604407135362015, "percentage": 94.34, "elapsed_time": "14:34:53", "remaining_time": "0:52:29"} +{"current_steps": 3155, "total_steps": 3339, "loss": 0.1501, "lr": 3.729065558512401e-07, "epoch": 6.614900314795383, "percentage": 94.49, "elapsed_time": "14:36:14", "remaining_time": "0:51:06"} +{"current_steps": 3160, "total_steps": 3339, "loss": 0.1446, "lr": 3.5308055916552395e-07, "epoch": 6.625393494228751, "percentage": 94.64, "elapsed_time": "14:37:36", "remaining_time": "0:49:42"} +{"current_steps": 3165, "total_steps": 3339, "loss": 0.1444, "lr": 3.337914017494903e-07, "epoch": 6.635886673662119, "percentage": 94.79, "elapsed_time": "14:38:57", "remaining_time": "0:48:19"} +{"current_steps": 3170, "total_steps": 3339, "loss": 0.1471, "lr": 3.1503961066679414e-07, "epoch": 6.646379853095488, "percentage": 94.94, "elapsed_time": "14:40:19", "remaining_time": "0:46:55"} +{"current_steps": 3175, "total_steps": 3339, "loss": 0.1415, "lr": 2.968256982979023e-07, "epoch": 6.656873032528856, "percentage": 95.09, "elapsed_time": "14:41:40", "remaining_time": "0:45:32"} +{"current_steps": 3180, "total_steps": 3339, "loss": 0.1426, "lr": 2.7915016232610505e-07, "epoch": 6.667366211962225, "percentage": 95.24, "elapsed_time": "14:43:01", "remaining_time": "0:44:09"} +{"current_steps": 3185, "total_steps": 3339, "loss": 0.1487, "lr": 2.6201348572389764e-07, "epoch": 6.677859391395593, "percentage": 95.39, "elapsed_time": "14:44:23", "remaining_time": "0:42:45"} +{"current_steps": 3190, "total_steps": 3339, "loss": 0.1433, "lr": 2.4541613673980667e-07, "epoch": 6.688352570828961, "percentage": 95.54, "elapsed_time": "14:45:45", "remaining_time": "0:41:22"} +{"current_steps": 3195, "total_steps": 3339, "loss": 0.1424, "lr": 2.2935856888557372e-07, "epoch": 6.698845750262329, "percentage": 95.69, "elapsed_time": "14:47:07", "remaining_time": "0:39:58"} +{"current_steps": 3200, "total_steps": 3339, "loss": 0.1435, "lr": 2.1384122092377391e-07, "epoch": 6.709338929695698, "percentage": 95.84, "elapsed_time": "14:48:29", "remaining_time": "0:38:35"} +{"current_steps": 3205, "total_steps": 3339, "loss": 0.1387, "lr": 1.9886451685582787e-07, "epoch": 6.719832109129066, "percentage": 95.99, "elapsed_time": "14:49:51", "remaining_time": "0:37:12"} +{"current_steps": 3210, "total_steps": 3339, "loss": 0.1449, "lr": 1.8442886591041542e-07, "epoch": 6.730325288562434, "percentage": 96.14, "elapsed_time": "14:51:12", "remaining_time": "0:35:48"} +{"current_steps": 3215, "total_steps": 3339, "loss": 0.1476, "lr": 1.7053466253229123e-07, "epoch": 6.740818467995803, "percentage": 96.29, "elapsed_time": "14:52:34", "remaining_time": "0:34:25"} +{"current_steps": 3220, "total_steps": 3339, "loss": 0.1441, "lr": 1.5718228637150668e-07, "epoch": 6.751311647429171, "percentage": 96.44, "elapsed_time": "14:53:56", "remaining_time": "0:33:02"} +{"current_steps": 3225, "total_steps": 3339, "loss": 0.1389, "lr": 1.44372102273036e-07, "epoch": 6.761804826862539, "percentage": 96.59, "elapsed_time": "14:55:18", "remaining_time": "0:31:38"} +{"current_steps": 3230, "total_steps": 3339, "loss": 0.1505, "lr": 1.3210446026681533e-07, "epoch": 6.772298006295908, "percentage": 96.74, "elapsed_time": "14:56:40", "remaining_time": "0:30:15"} +{"current_steps": 3235, "total_steps": 3339, "loss": 0.1394, "lr": 1.2037969555816597e-07, "epoch": 6.782791185729276, "percentage": 96.89, "elapsed_time": "14:58:01", "remaining_time": "0:28:52"} +{"current_steps": 3240, "total_steps": 3339, "loss": 0.1439, "lr": 1.0919812851864387e-07, "epoch": 6.793284365162644, "percentage": 97.04, "elapsed_time": "14:59:22", "remaining_time": "0:27:28"} +{"current_steps": 3245, "total_steps": 3339, "loss": 0.139, "lr": 9.856006467727775e-08, "epoch": 6.803777544596013, "percentage": 97.18, "elapsed_time": "15:00:44", "remaining_time": "0:26:05"} +{"current_steps": 3250, "total_steps": 3339, "loss": 0.1375, "lr": 8.84657947122336e-08, "epoch": 6.814270724029381, "percentage": 97.33, "elapsed_time": "15:02:06", "remaining_time": "0:24:42"} +{"current_steps": 3255, "total_steps": 3339, "loss": 0.143, "lr": 7.891559444285435e-08, "epoch": 6.824763903462749, "percentage": 97.48, "elapsed_time": "15:03:27", "remaining_time": "0:23:18"} +{"current_steps": 3260, "total_steps": 3339, "loss": 0.1425, "lr": 6.990972482214364e-08, "epoch": 6.835257082896118, "percentage": 97.63, "elapsed_time": "15:04:49", "remaining_time": "0:21:55"} +{"current_steps": 3265, "total_steps": 3339, "loss": 0.1433, "lr": 6.144843192961825e-08, "epoch": 6.845750262329486, "percentage": 97.78, "elapsed_time": "15:06:09", "remaining_time": "0:20:32"} +{"current_steps": 3270, "total_steps": 3339, "loss": 0.1492, "lr": 5.3531946964588966e-08, "epoch": 6.856243441762854, "percentage": 97.93, "elapsed_time": "15:07:31", "remaining_time": "0:19:08"} +{"current_steps": 3275, "total_steps": 3339, "loss": 0.1413, "lr": 4.616048623985236e-08, "epoch": 6.866736621196223, "percentage": 98.08, "elapsed_time": "15:08:50", "remaining_time": "0:17:45"} +{"current_steps": 3280, "total_steps": 3339, "loss": 0.1411, "lr": 3.9334251175764396e-08, "epoch": 6.877229800629591, "percentage": 98.23, "elapsed_time": "15:10:11", "remaining_time": "0:16:22"} +{"current_steps": 3285, "total_steps": 3339, "loss": 0.141, "lr": 3.305342829475589e-08, "epoch": 6.887722980062959, "percentage": 98.38, "elapsed_time": "15:11:32", "remaining_time": "0:14:59"} +{"current_steps": 3290, "total_steps": 3339, "loss": 0.1428, "lr": 2.7318189216212208e-08, "epoch": 6.898216159496327, "percentage": 98.53, "elapsed_time": "15:12:53", "remaining_time": "0:13:35"} +{"current_steps": 3295, "total_steps": 3339, "loss": 0.1419, "lr": 2.2128690651808115e-08, "epoch": 6.908709338929696, "percentage": 98.68, "elapsed_time": "15:14:16", "remaining_time": "0:12:12"} +{"current_steps": 3300, "total_steps": 3339, "loss": 0.1415, "lr": 1.7485074401206725e-08, "epoch": 6.919202518363064, "percentage": 98.83, "elapsed_time": "15:15:37", "remaining_time": "0:10:49"} +{"current_steps": 3305, "total_steps": 3339, "loss": 0.1427, "lr": 1.3387467348195959e-08, "epoch": 6.929695697796433, "percentage": 98.98, "elapsed_time": "15:16:59", "remaining_time": "0:09:26"} +{"current_steps": 3310, "total_steps": 3339, "loss": 0.1452, "lr": 9.835981457222421e-09, "epoch": 6.940188877229801, "percentage": 99.13, "elapsed_time": "15:18:21", "remaining_time": "0:08:02"} +{"current_steps": 3315, "total_steps": 3339, "loss": 0.1412, "lr": 6.830713770318298e-09, "epoch": 6.950682056663169, "percentage": 99.28, "elapsed_time": "15:19:43", "remaining_time": "0:06:39"} +{"current_steps": 3320, "total_steps": 3339, "loss": 0.1427, "lr": 4.371746404467914e-09, "epoch": 6.961175236096537, "percentage": 99.43, "elapsed_time": "15:21:04", "remaining_time": "0:05:16"} +{"current_steps": 3325, "total_steps": 3339, "loss": 0.1407, "lr": 2.4591465493495336e-09, "epoch": 6.971668415529906, "percentage": 99.58, "elapsed_time": "15:22:25", "remaining_time": "0:03:53"} +{"current_steps": 3330, "total_steps": 3339, "loss": 0.1453, "lr": 1.092966465510159e-09, "epoch": 6.982161594963274, "percentage": 99.73, "elapsed_time": "15:23:46", "remaining_time": "0:02:29"} +{"current_steps": 3335, "total_steps": 3339, "loss": 0.1466, "lr": 2.732434829266772e-10, "epoch": 6.992654774396642, "percentage": 99.88, "elapsed_time": "15:25:07", "remaining_time": "0:01:06"} +{"current_steps": 3339, "total_steps": 3339, "epoch": 7.0, "percentage": 100.0, "elapsed_time": "15:26:44", "remaining_time": "0:00:00"} diff --git a/trainer_state.json b/trainer_state.json new file mode 100644 index 0000000..4546e4a --- /dev/null +++ b/trainer_state.json @@ -0,0 +1,7384 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 7.0, + "eval_steps": 500, + "global_step": 3339, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.01049317943336831, + "grad_norm": 10.442070712794635, + "learning_rate": 4.790419161676648e-07, + "loss": 0.5478, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.2598276138305664, + "step": 5, + "valid_targets_mean": 5905.0, + "valid_targets_min": 2570 + }, + { + "epoch": 0.02098635886673662, + "grad_norm": 10.092963683152188, + "learning_rate": 1.0778443113772456e-06, + "loss": 0.5587, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.2579394578933716, + "step": 10, + "valid_targets_mean": 5802.2, + "valid_targets_min": 250 + }, + { + "epoch": 0.03147953830010493, + "grad_norm": 7.0791755036316, + "learning_rate": 1.6766467065868263e-06, + "loss": 0.5301, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.24459198117256165, + "step": 15, + "valid_targets_mean": 6244.0, + "valid_targets_min": 3736 + }, + { + "epoch": 0.04197271773347324, + "grad_norm": 4.90305077968075, + "learning_rate": 2.2754491017964075e-06, + "loss": 0.5158, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.26417461037635803, + "step": 20, + "valid_targets_mean": 7787.6, + "valid_targets_min": 5618 + }, + { + "epoch": 0.05246589716684155, + "grad_norm": 3.539133111950705, + "learning_rate": 2.874251497005988e-06, + "loss": 0.4672, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.23148763179779053, + "step": 25, + "valid_targets_mean": 7682.0, + "valid_targets_min": 3874 + }, + { + "epoch": 0.06295907660020986, + "grad_norm": 1.9513961592534688, + "learning_rate": 3.4730538922155693e-06, + "loss": 0.4327, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1966111958026886, + "step": 30, + "valid_targets_mean": 5997.2, + "valid_targets_min": 1142 + }, + { + "epoch": 0.07345225603357818, + "grad_norm": 1.4806215791740838, + "learning_rate": 4.07185628742515e-06, + "loss": 0.4349, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.18876728415489197, + "step": 35, + "valid_targets_mean": 6697.6, + "valid_targets_min": 3596 + }, + { + "epoch": 0.08394543546694648, + "grad_norm": 1.0513852110854274, + "learning_rate": 4.670658682634731e-06, + "loss": 0.385, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.17462614178657532, + "step": 40, + "valid_targets_mean": 7213.8, + "valid_targets_min": 3468 + }, + { + "epoch": 0.0944386149003148, + "grad_norm": 0.8448792964085587, + "learning_rate": 5.2694610778443125e-06, + "loss": 0.3846, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.17532816529273987, + "step": 45, + "valid_targets_mean": 5703.9, + "valid_targets_min": 366 + }, + { + "epoch": 0.1049317943336831, + "grad_norm": 0.6611181441663576, + "learning_rate": 5.868263473053893e-06, + "loss": 0.3689, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.17552967369556427, + "step": 50, + "valid_targets_mean": 6948.4, + "valid_targets_min": 3677 + }, + { + "epoch": 0.11542497376705142, + "grad_norm": 0.5763133002863221, + "learning_rate": 6.4670658682634736e-06, + "loss": 0.364, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.16234669089317322, + "step": 55, + "valid_targets_mean": 7262.8, + "valid_targets_min": 5121 + }, + { + "epoch": 0.1259181532004197, + "grad_norm": 0.5797092201424425, + "learning_rate": 7.0658682634730545e-06, + "loss": 0.3559, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.19157901406288147, + "step": 60, + "valid_targets_mean": 8252.1, + "valid_targets_min": 4470 + }, + { + "epoch": 0.13641133263378805, + "grad_norm": 0.5570227928674149, + "learning_rate": 7.664670658682636e-06, + "loss": 0.3527, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.18043449521064758, + "step": 65, + "valid_targets_mean": 6852.2, + "valid_targets_min": 4327 + }, + { + "epoch": 0.14690451206715635, + "grad_norm": 0.514648674471187, + "learning_rate": 8.263473053892216e-06, + "loss": 0.3472, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.18866193294525146, + "step": 70, + "valid_targets_mean": 8363.0, + "valid_targets_min": 6778 + }, + { + "epoch": 0.15739769150052466, + "grad_norm": 0.48117680106981386, + "learning_rate": 8.862275449101796e-06, + "loss": 0.3256, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.16611936688423157, + "step": 75, + "valid_targets_mean": 7270.4, + "valid_targets_min": 4190 + }, + { + "epoch": 0.16789087093389296, + "grad_norm": 0.5090714973625744, + "learning_rate": 9.461077844311378e-06, + "loss": 0.3088, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.156036838889122, + "step": 80, + "valid_targets_mean": 5659.9, + "valid_targets_min": 2290 + }, + { + "epoch": 0.1783840503672613, + "grad_norm": 0.495155198419227, + "learning_rate": 1.0059880239520958e-05, + "loss": 0.3132, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1740761399269104, + "step": 85, + "valid_targets_mean": 7120.0, + "valid_targets_min": 4711 + }, + { + "epoch": 0.1888772298006296, + "grad_norm": 0.48660107390632623, + "learning_rate": 1.065868263473054e-05, + "loss": 0.3157, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1523314267396927, + "step": 90, + "valid_targets_mean": 7334.6, + "valid_targets_min": 5440 + }, + { + "epoch": 0.1993704092339979, + "grad_norm": 0.5151799197587337, + "learning_rate": 1.125748502994012e-05, + "loss": 0.3021, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.14953424036502838, + "step": 95, + "valid_targets_mean": 6514.2, + "valid_targets_min": 3367 + }, + { + "epoch": 0.2098635886673662, + "grad_norm": 0.4444602570908872, + "learning_rate": 1.1856287425149702e-05, + "loss": 0.2913, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.15660414099693298, + "step": 100, + "valid_targets_mean": 7354.2, + "valid_targets_min": 3000 + }, + { + "epoch": 0.2203567681007345, + "grad_norm": 0.4690864277135588, + "learning_rate": 1.2455089820359284e-05, + "loss": 0.299, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.13274423778057098, + "step": 105, + "valid_targets_mean": 6205.4, + "valid_targets_min": 4049 + }, + { + "epoch": 0.23084994753410285, + "grad_norm": 0.4396706992964899, + "learning_rate": 1.3053892215568864e-05, + "loss": 0.2922, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.146137073636055, + "step": 110, + "valid_targets_mean": 6435.5, + "valid_targets_min": 4433 + }, + { + "epoch": 0.24134312696747115, + "grad_norm": 0.444445172882143, + "learning_rate": 1.3652694610778446e-05, + "loss": 0.2916, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.13934937119483948, + "step": 115, + "valid_targets_mean": 6306.5, + "valid_targets_min": 4760 + }, + { + "epoch": 0.2518363064008394, + "grad_norm": 0.4790757081055015, + "learning_rate": 1.4251497005988024e-05, + "loss": 0.2873, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.13322457671165466, + "step": 120, + "valid_targets_mean": 6649.1, + "valid_targets_min": 4272 + }, + { + "epoch": 0.2623294858342078, + "grad_norm": 0.4355874273213463, + "learning_rate": 1.4850299401197606e-05, + "loss": 0.2819, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.14607881009578705, + "step": 125, + "valid_targets_mean": 7614.6, + "valid_targets_min": 5871 + }, + { + "epoch": 0.2728226652675761, + "grad_norm": 0.44449645209510946, + "learning_rate": 1.5449101796407188e-05, + "loss": 0.2735, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.15480495989322662, + "step": 130, + "valid_targets_mean": 7294.1, + "valid_targets_min": 5012 + }, + { + "epoch": 0.2833158447009444, + "grad_norm": 0.5042014058167168, + "learning_rate": 1.604790419161677e-05, + "loss": 0.2725, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.14767739176750183, + "step": 135, + "valid_targets_mean": 6540.5, + "valid_targets_min": 5008 + }, + { + "epoch": 0.2938090241343127, + "grad_norm": 0.6131099638004378, + "learning_rate": 1.6646706586826348e-05, + "loss": 0.2745, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.14182569086551666, + "step": 140, + "valid_targets_mean": 6346.6, + "valid_targets_min": 4130 + }, + { + "epoch": 0.304302203567681, + "grad_norm": 0.4561058399466216, + "learning_rate": 1.724550898203593e-05, + "loss": 0.2756, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11705484986305237, + "step": 145, + "valid_targets_mean": 7584.0, + "valid_targets_min": 5008 + }, + { + "epoch": 0.3147953830010493, + "grad_norm": 0.48600693993994903, + "learning_rate": 1.784431137724551e-05, + "loss": 0.262, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.12611989676952362, + "step": 150, + "valid_targets_mean": 6218.5, + "valid_targets_min": 764 + }, + { + "epoch": 0.3252885624344176, + "grad_norm": 0.45109432940005745, + "learning_rate": 1.8443113772455093e-05, + "loss": 0.2752, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.13387374579906464, + "step": 155, + "valid_targets_mean": 6948.8, + "valid_targets_min": 5189 + }, + { + "epoch": 0.3357817418677859, + "grad_norm": 0.5288217361833911, + "learning_rate": 1.904191616766467e-05, + "loss": 0.2699, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11767670512199402, + "step": 160, + "valid_targets_mean": 4625.1, + "valid_targets_min": 349 + }, + { + "epoch": 0.3462749213011542, + "grad_norm": 0.49693883800017535, + "learning_rate": 1.9640718562874253e-05, + "loss": 0.2647, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1478138417005539, + "step": 165, + "valid_targets_mean": 7615.9, + "valid_targets_min": 5205 + }, + { + "epoch": 0.3567681007345226, + "grad_norm": 0.44781816365597465, + "learning_rate": 2.0239520958083835e-05, + "loss": 0.269, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1448097676038742, + "step": 170, + "valid_targets_mean": 7566.9, + "valid_targets_min": 4967 + }, + { + "epoch": 0.3672612801678909, + "grad_norm": 0.4377414441672674, + "learning_rate": 2.0838323353293413e-05, + "loss": 0.2593, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.13878971338272095, + "step": 175, + "valid_targets_mean": 7618.2, + "valid_targets_min": 5329 + }, + { + "epoch": 0.3777544596012592, + "grad_norm": 0.5315516549401317, + "learning_rate": 2.1437125748502995e-05, + "loss": 0.2679, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.128665953874588, + "step": 180, + "valid_targets_mean": 5878.2, + "valid_targets_min": 1914 + }, + { + "epoch": 0.3882476390346275, + "grad_norm": 0.5074662604874222, + "learning_rate": 2.2035928143712577e-05, + "loss": 0.2668, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.13005158305168152, + "step": 185, + "valid_targets_mean": 6618.8, + "valid_targets_min": 2840 + }, + { + "epoch": 0.3987408184679958, + "grad_norm": 0.7615341272684868, + "learning_rate": 2.2634730538922155e-05, + "loss": 0.2611, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.12154854834079742, + "step": 190, + "valid_targets_mean": 6694.0, + "valid_targets_min": 4284 + }, + { + "epoch": 0.4092339979013641, + "grad_norm": 0.4773516440422621, + "learning_rate": 2.323353293413174e-05, + "loss": 0.2534, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1193661242723465, + "step": 195, + "valid_targets_mean": 6876.0, + "valid_targets_min": 4220 + }, + { + "epoch": 0.4197271773347324, + "grad_norm": 0.5288002322801338, + "learning_rate": 2.383233532934132e-05, + "loss": 0.257, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1273195743560791, + "step": 200, + "valid_targets_mean": 6374.9, + "valid_targets_min": 3949 + }, + { + "epoch": 0.4302203567681007, + "grad_norm": 0.48232466996680506, + "learning_rate": 2.44311377245509e-05, + "loss": 0.2471, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11985370516777039, + "step": 205, + "valid_targets_mean": 7847.9, + "valid_targets_min": 4092 + }, + { + "epoch": 0.440713536201469, + "grad_norm": 0.44817416869209087, + "learning_rate": 2.502994011976048e-05, + "loss": 0.2462, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.13444887101650238, + "step": 210, + "valid_targets_mean": 7243.0, + "valid_targets_min": 4423 + }, + { + "epoch": 0.45120671563483733, + "grad_norm": 0.5049073442307994, + "learning_rate": 2.5628742514970064e-05, + "loss": 0.2499, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.14578238129615784, + "step": 215, + "valid_targets_mean": 7838.8, + "valid_targets_min": 5069 + }, + { + "epoch": 0.4616998950682057, + "grad_norm": 0.542799924475802, + "learning_rate": 2.6227544910179643e-05, + "loss": 0.2541, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.12193570286035538, + "step": 220, + "valid_targets_mean": 6121.5, + "valid_targets_min": 1003 + }, + { + "epoch": 0.472193074501574, + "grad_norm": 0.4317893281003648, + "learning_rate": 2.682634730538922e-05, + "loss": 0.2519, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1273801326751709, + "step": 225, + "valid_targets_mean": 8576.8, + "valid_targets_min": 4025 + }, + { + "epoch": 0.4826862539349423, + "grad_norm": 0.46444653807630826, + "learning_rate": 2.7425149700598806e-05, + "loss": 0.2439, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.12182192504405975, + "step": 230, + "valid_targets_mean": 7348.0, + "valid_targets_min": 700 + }, + { + "epoch": 0.4931794333683106, + "grad_norm": 0.4514835069168026, + "learning_rate": 2.8023952095808385e-05, + "loss": 0.2436, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10713709890842438, + "step": 235, + "valid_targets_mean": 6862.8, + "valid_targets_min": 5758 + }, + { + "epoch": 0.5036726128016789, + "grad_norm": 0.4515082052592365, + "learning_rate": 2.8622754491017966e-05, + "loss": 0.2359, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11493110656738281, + "step": 240, + "valid_targets_mean": 6710.5, + "valid_targets_min": 3918 + }, + { + "epoch": 0.5141657922350472, + "grad_norm": 0.4866253654951501, + "learning_rate": 2.9221556886227548e-05, + "loss": 0.2377, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.13693702220916748, + "step": 245, + "valid_targets_mean": 7029.1, + "valid_targets_min": 3887 + }, + { + "epoch": 0.5246589716684156, + "grad_norm": 0.4670954273732565, + "learning_rate": 2.982035928143713e-05, + "loss": 0.2408, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.13592128455638885, + "step": 250, + "valid_targets_mean": 8156.4, + "valid_targets_min": 5231 + }, + { + "epoch": 0.5351521511017838, + "grad_norm": 0.5116091256707135, + "learning_rate": 3.041916167664671e-05, + "loss": 0.24, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.14138075709342957, + "step": 255, + "valid_targets_mean": 7603.2, + "valid_targets_min": 3515 + }, + { + "epoch": 0.5456453305351522, + "grad_norm": 0.49381176948684474, + "learning_rate": 3.101796407185629e-05, + "loss": 0.2371, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10804753005504608, + "step": 260, + "valid_targets_mean": 5739.0, + "valid_targets_min": 854 + }, + { + "epoch": 0.5561385099685204, + "grad_norm": 0.5057048420375377, + "learning_rate": 3.161676646706587e-05, + "loss": 0.2438, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11901626735925674, + "step": 265, + "valid_targets_mean": 5782.8, + "valid_targets_min": 4021 + }, + { + "epoch": 0.5666316894018888, + "grad_norm": 0.4856327454584469, + "learning_rate": 3.221556886227545e-05, + "loss": 0.2372, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.13004529476165771, + "step": 270, + "valid_targets_mean": 6982.0, + "valid_targets_min": 3945 + }, + { + "epoch": 0.577124868835257, + "grad_norm": 0.5776994400850526, + "learning_rate": 3.2814371257485035e-05, + "loss": 0.2386, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1195303201675415, + "step": 275, + "valid_targets_mean": 6730.1, + "valid_targets_min": 3743 + }, + { + "epoch": 0.5876180482686254, + "grad_norm": 0.4961135164843524, + "learning_rate": 3.3413173652694614e-05, + "loss": 0.2417, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10341168195009232, + "step": 280, + "valid_targets_mean": 6805.9, + "valid_targets_min": 1689 + }, + { + "epoch": 0.5981112277019937, + "grad_norm": 0.43158715919426693, + "learning_rate": 3.40119760479042e-05, + "loss": 0.2361, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.12573376297950745, + "step": 285, + "valid_targets_mean": 7673.2, + "valid_targets_min": 2930 + }, + { + "epoch": 0.608604407135362, + "grad_norm": 0.44658707252402824, + "learning_rate": 3.461077844311378e-05, + "loss": 0.2317, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.12525373697280884, + "step": 290, + "valid_targets_mean": 7587.9, + "valid_targets_min": 5964 + }, + { + "epoch": 0.6190975865687304, + "grad_norm": 0.505529888519472, + "learning_rate": 3.5209580838323356e-05, + "loss": 0.2427, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11592252552509308, + "step": 295, + "valid_targets_mean": 6551.8, + "valid_targets_min": 3820 + }, + { + "epoch": 0.6295907660020986, + "grad_norm": 0.47045291182652393, + "learning_rate": 3.5808383233532934e-05, + "loss": 0.2351, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09940166026353836, + "step": 300, + "valid_targets_mean": 6256.0, + "valid_targets_min": 4308 + }, + { + "epoch": 0.640083945435467, + "grad_norm": 0.5819742984902145, + "learning_rate": 3.640718562874252e-05, + "loss": 0.2384, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.12186205387115479, + "step": 305, + "valid_targets_mean": 5801.5, + "valid_targets_min": 3290 + }, + { + "epoch": 0.6505771248688352, + "grad_norm": 0.5363473683573847, + "learning_rate": 3.70059880239521e-05, + "loss": 0.2351, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10016492009162903, + "step": 310, + "valid_targets_mean": 5232.6, + "valid_targets_min": 300 + }, + { + "epoch": 0.6610703043022036, + "grad_norm": 0.4789689396787, + "learning_rate": 3.7604790419161676e-05, + "loss": 0.2393, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11646223068237305, + "step": 315, + "valid_targets_mean": 6174.9, + "valid_targets_min": 1718 + }, + { + "epoch": 0.6715634837355718, + "grad_norm": 0.4398019580612534, + "learning_rate": 3.820359281437126e-05, + "loss": 0.2268, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.13057488203048706, + "step": 320, + "valid_targets_mean": 7400.0, + "valid_targets_min": 4923 + }, + { + "epoch": 0.6820566631689402, + "grad_norm": 0.4733313785929033, + "learning_rate": 3.880239520958084e-05, + "loss": 0.2302, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10850711911916733, + "step": 325, + "valid_targets_mean": 6663.8, + "valid_targets_min": 1167 + }, + { + "epoch": 0.6925498426023085, + "grad_norm": 0.47434509221532284, + "learning_rate": 3.9401197604790425e-05, + "loss": 0.2331, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1422533094882965, + "step": 330, + "valid_targets_mean": 7683.2, + "valid_targets_min": 4582 + }, + { + "epoch": 0.7030430220356768, + "grad_norm": 0.4581420855921195, + "learning_rate": 4e-05, + "loss": 0.2285, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.12746475636959076, + "step": 335, + "valid_targets_mean": 6639.6, + "valid_targets_min": 3621 + }, + { + "epoch": 0.7135362014690452, + "grad_norm": 0.9121402094767356, + "learning_rate": 3.9999726756517076e-05, + "loss": 0.2214, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.12314047664403915, + "step": 340, + "valid_targets_mean": 6983.0, + "valid_targets_min": 3756 + }, + { + "epoch": 0.7240293809024134, + "grad_norm": 0.6199468683819878, + "learning_rate": 3.9998907033534495e-05, + "loss": 0.2242, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09141039103269577, + "step": 345, + "valid_targets_mean": 4505.4, + "valid_targets_min": 594 + }, + { + "epoch": 0.7345225603357818, + "grad_norm": 0.47183916650282887, + "learning_rate": 3.999754085345065e-05, + "loss": 0.2279, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11257795244455338, + "step": 350, + "valid_targets_mean": 6230.9, + "valid_targets_min": 1162 + }, + { + "epoch": 0.74501573976915, + "grad_norm": 0.44132517764849544, + "learning_rate": 3.9995628253595536e-05, + "loss": 0.2298, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09789767116308212, + "step": 355, + "valid_targets_mean": 6479.1, + "valid_targets_min": 666 + }, + { + "epoch": 0.7555089192025184, + "grad_norm": 0.44013247404632355, + "learning_rate": 3.999316928622968e-05, + "loss": 0.2265, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09201842546463013, + "step": 360, + "valid_targets_mean": 6625.6, + "valid_targets_min": 630 + }, + { + "epoch": 0.7660020986358866, + "grad_norm": 0.45026997589480183, + "learning_rate": 3.999016401854278e-05, + "loss": 0.2319, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.13638240098953247, + "step": 365, + "valid_targets_mean": 7620.2, + "valid_targets_min": 3090 + }, + { + "epoch": 0.776495278069255, + "grad_norm": 0.49138122387375666, + "learning_rate": 3.9986612532651804e-05, + "loss": 0.2244, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11600857973098755, + "step": 370, + "valid_targets_mean": 7170.2, + "valid_targets_min": 4445 + }, + { + "epoch": 0.7869884575026233, + "grad_norm": 0.4622724679630308, + "learning_rate": 3.9982514925598794e-05, + "loss": 0.2335, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10520058125257492, + "step": 375, + "valid_targets_mean": 6474.5, + "valid_targets_min": 2723 + }, + { + "epoch": 0.7974816369359916, + "grad_norm": 0.4759975969806025, + "learning_rate": 3.997787130934819e-05, + "loss": 0.227, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11512656509876251, + "step": 380, + "valid_targets_mean": 6460.6, + "valid_targets_min": 4597 + }, + { + "epoch": 0.8079748163693599, + "grad_norm": 0.46249109508603226, + "learning_rate": 3.997268181078379e-05, + "loss": 0.2321, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11230267584323883, + "step": 385, + "valid_targets_mean": 6711.8, + "valid_targets_min": 4850 + }, + { + "epoch": 0.8184679958027282, + "grad_norm": 0.5006831543857353, + "learning_rate": 3.9966946571705244e-05, + "loss": 0.2248, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1262255311012268, + "step": 390, + "valid_targets_mean": 6133.6, + "valid_targets_min": 3674 + }, + { + "epoch": 0.8289611752360966, + "grad_norm": 0.4303159020764481, + "learning_rate": 3.996066574882424e-05, + "loss": 0.2195, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09490875899791718, + "step": 395, + "valid_targets_mean": 6189.0, + "valid_targets_min": 1429 + }, + { + "epoch": 0.8394543546694648, + "grad_norm": 0.45205335798505436, + "learning_rate": 3.9953839513760154e-05, + "loss": 0.2246, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.14661428332328796, + "step": 400, + "valid_targets_mean": 7913.9, + "valid_targets_min": 3322 + }, + { + "epoch": 0.8499475341028332, + "grad_norm": 0.4635074333313668, + "learning_rate": 3.9946468053035417e-05, + "loss": 0.2237, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10226143896579742, + "step": 405, + "valid_targets_mean": 6755.9, + "valid_targets_min": 3629 + }, + { + "epoch": 0.8604407135362014, + "grad_norm": 0.45410343452389834, + "learning_rate": 3.993855156807039e-05, + "loss": 0.2254, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11254068464040756, + "step": 410, + "valid_targets_mean": 6444.2, + "valid_targets_min": 4531 + }, + { + "epoch": 0.8709338929695698, + "grad_norm": 0.45650594668745553, + "learning_rate": 3.993009027517786e-05, + "loss": 0.2205, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.13951584696769714, + "step": 415, + "valid_targets_mean": 8224.6, + "valid_targets_min": 5268 + }, + { + "epoch": 0.881427072402938, + "grad_norm": 0.46915644741084966, + "learning_rate": 3.992108440555715e-05, + "loss": 0.2191, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10114838182926178, + "step": 420, + "valid_targets_mean": 5602.1, + "valid_targets_min": 648 + }, + { + "epoch": 0.8919202518363064, + "grad_norm": 0.42441467712770836, + "learning_rate": 3.991153420528777e-05, + "loss": 0.2177, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10630941390991211, + "step": 425, + "valid_targets_mean": 7337.6, + "valid_targets_min": 4878 + }, + { + "epoch": 0.9024134312696747, + "grad_norm": 0.3857461321513521, + "learning_rate": 3.9901439935322726e-05, + "loss": 0.225, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10085228085517883, + "step": 430, + "valid_targets_mean": 9627.6, + "valid_targets_min": 4733 + }, + { + "epoch": 0.912906610703043, + "grad_norm": 0.43473037371673884, + "learning_rate": 3.989080187148136e-05, + "loss": 0.2247, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09486043453216553, + "step": 435, + "valid_targets_mean": 6688.5, + "valid_targets_min": 4730 + }, + { + "epoch": 0.9233997901364114, + "grad_norm": 0.42766307509826645, + "learning_rate": 3.9879620304441834e-05, + "loss": 0.2196, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1074979156255722, + "step": 440, + "valid_targets_mean": 6539.4, + "valid_targets_min": 615 + }, + { + "epoch": 0.9338929695697796, + "grad_norm": 0.4407848515744525, + "learning_rate": 3.986789553973319e-05, + "loss": 0.2314, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09944035112857819, + "step": 445, + "valid_targets_mean": 6428.0, + "valid_targets_min": 3529 + }, + { + "epoch": 0.944386149003148, + "grad_norm": 0.41655497392082325, + "learning_rate": 3.985562789772697e-05, + "loss": 0.2164, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09469069540500641, + "step": 450, + "valid_targets_mean": 5372.9, + "valid_targets_min": 3737 + }, + { + "epoch": 0.9548793284365162, + "grad_norm": 0.44887841881078855, + "learning_rate": 3.9842817713628494e-05, + "loss": 0.2144, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11075478047132492, + "step": 455, + "valid_targets_mean": 5717.0, + "valid_targets_min": 483 + }, + { + "epoch": 0.9653725078698846, + "grad_norm": 0.4695273240032865, + "learning_rate": 3.982946533746771e-05, + "loss": 0.2282, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1085926964879036, + "step": 460, + "valid_targets_mean": 6351.1, + "valid_targets_min": 4575 + }, + { + "epoch": 0.9758656873032528, + "grad_norm": 0.4674193387804589, + "learning_rate": 3.9815571134089584e-05, + "loss": 0.2142, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.102556973695755, + "step": 465, + "valid_targets_mean": 6746.1, + "valid_targets_min": 4121 + }, + { + "epoch": 0.9863588667366212, + "grad_norm": 0.43027944579612265, + "learning_rate": 3.9801135483144174e-05, + "loss": 0.2206, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10300008952617645, + "step": 470, + "valid_targets_mean": 6275.4, + "valid_targets_min": 4440 + }, + { + "epoch": 0.9968520461699895, + "grad_norm": 0.4553669538356775, + "learning_rate": 3.9786158779076226e-05, + "loss": 0.2224, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09901054203510284, + "step": 475, + "valid_targets_mean": 6079.8, + "valid_targets_min": 3270 + }, + { + "epoch": 1.006295907660021, + "grad_norm": 0.5141367626612066, + "learning_rate": 3.977064143111443e-05, + "loss": 0.208, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11563676595687866, + "step": 480, + "valid_targets_mean": 6947.8, + "valid_targets_min": 5223 + }, + { + "epoch": 1.0167890870933893, + "grad_norm": 0.44768746605898224, + "learning_rate": 3.9754583863260195e-05, + "loss": 0.202, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10534417629241943, + "step": 485, + "valid_targets_mean": 7922.2, + "valid_targets_min": 3596 + }, + { + "epoch": 1.0272822665267576, + "grad_norm": 0.4320578176769016, + "learning_rate": 3.97379865142761e-05, + "loss": 0.2074, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09221038222312927, + "step": 490, + "valid_targets_mean": 6658.9, + "valid_targets_min": 3889 + }, + { + "epoch": 1.0377754459601258, + "grad_norm": 0.4663529918729838, + "learning_rate": 3.97208498376739e-05, + "loss": 0.212, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10989321768283844, + "step": 495, + "valid_targets_mean": 7902.9, + "valid_targets_min": 1168 + }, + { + "epoch": 1.0482686253934943, + "grad_norm": 0.44372155120467116, + "learning_rate": 3.97031743017021e-05, + "loss": 0.2093, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11609729379415512, + "step": 500, + "valid_targets_mean": 7281.1, + "valid_targets_min": 5114 + }, + { + "epoch": 1.0587618048268626, + "grad_norm": 0.6864781611865786, + "learning_rate": 3.9684960389333206e-05, + "loss": 0.2052, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09195898473262787, + "step": 505, + "valid_targets_mean": 6692.9, + "valid_targets_min": 266 + }, + { + "epoch": 1.0692549842602308, + "grad_norm": 0.5245622817534252, + "learning_rate": 3.966620859825051e-05, + "loss": 0.2151, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10989820957183838, + "step": 510, + "valid_targets_mean": 6483.5, + "valid_targets_min": 4790 + }, + { + "epoch": 1.079748163693599, + "grad_norm": 0.4708650614620763, + "learning_rate": 3.964691944083448e-05, + "loss": 0.21, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1011229082942009, + "step": 515, + "valid_targets_mean": 6219.8, + "valid_targets_min": 3720 + }, + { + "epoch": 1.0902413431269675, + "grad_norm": 0.4594897950402482, + "learning_rate": 3.9627093444148766e-05, + "loss": 0.2127, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10303577780723572, + "step": 520, + "valid_targets_mean": 6787.2, + "valid_targets_min": 4548 + }, + { + "epoch": 1.1007345225603358, + "grad_norm": 0.44122564056313573, + "learning_rate": 3.960673114992581e-05, + "loss": 0.2079, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09995207190513611, + "step": 525, + "valid_targets_mean": 5998.5, + "valid_targets_min": 816 + }, + { + "epoch": 1.111227701993704, + "grad_norm": 0.4884840600183993, + "learning_rate": 3.958583311455203e-05, + "loss": 0.2083, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.13188013434410095, + "step": 530, + "valid_targets_mean": 7562.2, + "valid_targets_min": 3922 + }, + { + "epoch": 1.1217208814270725, + "grad_norm": 0.45988581469672457, + "learning_rate": 3.9564399909052626e-05, + "loss": 0.2104, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0921846479177475, + "step": 535, + "valid_targets_mean": 6951.1, + "valid_targets_min": 1165 + }, + { + "epoch": 1.1322140608604407, + "grad_norm": 0.45266493691744136, + "learning_rate": 3.954243211907596e-05, + "loss": 0.2077, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09937339276075363, + "step": 540, + "valid_targets_mean": 7355.6, + "valid_targets_min": 4068 + }, + { + "epoch": 1.142707240293809, + "grad_norm": 0.4719115503349866, + "learning_rate": 3.9519930344877606e-05, + "loss": 0.2124, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09784740209579468, + "step": 545, + "valid_targets_mean": 5580.9, + "valid_targets_min": 500 + }, + { + "epoch": 1.1532004197271772, + "grad_norm": 0.41994490005109103, + "learning_rate": 3.949689520130384e-05, + "loss": 0.2154, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11583942919969559, + "step": 550, + "valid_targets_mean": 8311.1, + "valid_targets_min": 6525 + }, + { + "epoch": 1.1636935991605457, + "grad_norm": 0.4396076879076679, + "learning_rate": 3.947332731777498e-05, + "loss": 0.2073, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10304717719554901, + "step": 555, + "valid_targets_mean": 6858.4, + "valid_targets_min": 546 + }, + { + "epoch": 1.174186778593914, + "grad_norm": 0.4410137446094941, + "learning_rate": 3.944922733826806e-05, + "loss": 0.2032, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11203482002019882, + "step": 560, + "valid_targets_mean": 7976.4, + "valid_targets_min": 3852 + }, + { + "epoch": 1.1846799580272822, + "grad_norm": 0.45060698031653185, + "learning_rate": 3.942459592129933e-05, + "loss": 0.2117, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.12652936577796936, + "step": 565, + "valid_targets_mean": 8402.8, + "valid_targets_min": 6922 + }, + { + "epoch": 1.1951731374606505, + "grad_norm": 0.44960533339075365, + "learning_rate": 3.9399433739906195e-05, + "loss": 0.2063, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11032892763614655, + "step": 570, + "valid_targets_mean": 6879.6, + "valid_targets_min": 4463 + }, + { + "epoch": 1.205666316894019, + "grad_norm": 0.40891261039473853, + "learning_rate": 3.937374148162888e-05, + "loss": 0.1963, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10610020905733109, + "step": 575, + "valid_targets_mean": 8212.9, + "valid_targets_min": 5715 + }, + { + "epoch": 1.2161594963273872, + "grad_norm": 0.4943058934180262, + "learning_rate": 3.934751984849158e-05, + "loss": 0.2086, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10550324618816376, + "step": 580, + "valid_targets_mean": 6367.5, + "valid_targets_min": 3802 + }, + { + "epoch": 1.2266526757607554, + "grad_norm": 0.4703813332230257, + "learning_rate": 3.932076955698334e-05, + "loss": 0.2049, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0797218382358551, + "step": 585, + "valid_targets_mean": 4991.4, + "valid_targets_min": 970 + }, + { + "epoch": 1.237145855194124, + "grad_norm": 0.42104119869177054, + "learning_rate": 3.929349133803844e-05, + "loss": 0.2054, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07701364159584045, + "step": 590, + "valid_targets_mean": 5127.9, + "valid_targets_min": 324 + }, + { + "epoch": 1.2476390346274921, + "grad_norm": 0.41604478601692463, + "learning_rate": 3.926568593701644e-05, + "loss": 0.1986, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.12172184884548187, + "step": 595, + "valid_targets_mean": 8070.8, + "valid_targets_min": 5142 + }, + { + "epoch": 1.2581322140608604, + "grad_norm": 0.47393122507909924, + "learning_rate": 3.9237354113681795e-05, + "loss": 0.2041, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11034677922725677, + "step": 600, + "valid_targets_mean": 6446.1, + "valid_targets_min": 3682 + }, + { + "epoch": 1.2686253934942289, + "grad_norm": 0.4179462515196094, + "learning_rate": 3.920849664218312e-05, + "loss": 0.2043, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0843837708234787, + "step": 605, + "valid_targets_mean": 6985.2, + "valid_targets_min": 5105 + }, + { + "epoch": 1.2791185729275971, + "grad_norm": 0.40880913860347773, + "learning_rate": 3.917911431103201e-05, + "loss": 0.2038, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0910620167851448, + "step": 610, + "valid_targets_mean": 7530.9, + "valid_targets_min": 4496 + }, + { + "epoch": 1.2896117523609654, + "grad_norm": 0.4095319501364525, + "learning_rate": 3.914920792308153e-05, + "loss": 0.2087, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09878174960613251, + "step": 615, + "valid_targets_mean": 6751.8, + "valid_targets_min": 3991 + }, + { + "epoch": 1.3001049317943336, + "grad_norm": 0.4506689796690072, + "learning_rate": 3.9118778295504224e-05, + "loss": 0.2088, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10704606026411057, + "step": 620, + "valid_targets_mean": 6370.0, + "valid_targets_min": 477 + }, + { + "epoch": 1.3105981112277019, + "grad_norm": 0.46388489916153236, + "learning_rate": 3.908782625976984e-05, + "loss": 0.2122, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08457791805267334, + "step": 625, + "valid_targets_mean": 5144.4, + "valid_targets_min": 301 + }, + { + "epoch": 1.3210912906610703, + "grad_norm": 0.4193806296157824, + "learning_rate": 3.905635266162258e-05, + "loss": 0.2086, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10484787821769714, + "step": 630, + "valid_targets_mean": 7334.6, + "valid_targets_min": 1101 + }, + { + "epoch": 1.3315844700944386, + "grad_norm": 0.4270554384125975, + "learning_rate": 3.9024358361058e-05, + "loss": 0.206, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10570015013217926, + "step": 635, + "valid_targets_mean": 7018.4, + "valid_targets_min": 1812 + }, + { + "epoch": 1.3420776495278068, + "grad_norm": 0.46023429861282056, + "learning_rate": 3.8991844232299526e-05, + "loss": 0.2027, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1281335949897766, + "step": 640, + "valid_targets_mean": 7558.1, + "valid_targets_min": 5144 + }, + { + "epoch": 1.3525708289611753, + "grad_norm": 0.40400315590462454, + "learning_rate": 3.8958811163774526e-05, + "loss": 0.2089, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08377764374017715, + "step": 645, + "valid_targets_mean": 5568.4, + "valid_targets_min": 1547 + }, + { + "epoch": 1.3630640083945436, + "grad_norm": 0.4165881898530997, + "learning_rate": 3.892526005809007e-05, + "loss": 0.2011, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1130603700876236, + "step": 650, + "valid_targets_mean": 7709.5, + "valid_targets_min": 4068 + }, + { + "epoch": 1.3735571878279118, + "grad_norm": 0.44924096975403294, + "learning_rate": 3.8891191832008256e-05, + "loss": 0.2038, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10047546774148941, + "step": 655, + "valid_targets_mean": 6542.5, + "valid_targets_min": 4680 + }, + { + "epoch": 1.3840503672612803, + "grad_norm": 0.4222630020682009, + "learning_rate": 3.885660741642116e-05, + "loss": 0.209, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10015548765659332, + "step": 660, + "valid_targets_mean": 7341.5, + "valid_targets_min": 4984 + }, + { + "epoch": 1.3945435466946485, + "grad_norm": 0.4343816792421444, + "learning_rate": 3.882150775632539e-05, + "loss": 0.2093, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10170470178127289, + "step": 665, + "valid_targets_mean": 7560.9, + "valid_targets_min": 4740 + }, + { + "epoch": 1.4050367261280168, + "grad_norm": 0.43895827864606435, + "learning_rate": 3.87858938107963e-05, + "loss": 0.2036, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10412415862083435, + "step": 670, + "valid_targets_mean": 6737.1, + "valid_targets_min": 5160 + }, + { + "epoch": 1.415529905561385, + "grad_norm": 0.5016028159461988, + "learning_rate": 3.8749766552961736e-05, + "loss": 0.1979, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09616001695394516, + "step": 675, + "valid_targets_mean": 5862.6, + "valid_targets_min": 2828 + }, + { + "epoch": 1.4260230849947533, + "grad_norm": 0.406456166127644, + "learning_rate": 3.871312696997547e-05, + "loss": 0.2041, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09848299622535706, + "step": 680, + "valid_targets_mean": 7496.1, + "valid_targets_min": 4233 + }, + { + "epoch": 1.4365162644281217, + "grad_norm": 0.4243777498549692, + "learning_rate": 3.867597606299023e-05, + "loss": 0.2037, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10640639066696167, + "step": 685, + "valid_targets_mean": 8608.2, + "valid_targets_min": 6053 + }, + { + "epoch": 1.44700944386149, + "grad_norm": 0.4414623443643935, + "learning_rate": 3.863831484713033e-05, + "loss": 0.1972, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11067424714565277, + "step": 690, + "valid_targets_mean": 6397.6, + "valid_targets_min": 4089 + }, + { + "epoch": 1.4575026232948582, + "grad_norm": 0.3946603582520573, + "learning_rate": 3.860014435146397e-05, + "loss": 0.2015, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07396245002746582, + "step": 695, + "valid_targets_mean": 5697.6, + "valid_targets_min": 2186 + }, + { + "epoch": 1.4679958027282267, + "grad_norm": 0.4278635147030236, + "learning_rate": 3.856146561897505e-05, + "loss": 0.1951, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0998871922492981, + "step": 700, + "valid_targets_mean": 7134.9, + "valid_targets_min": 3930 + }, + { + "epoch": 1.478488982161595, + "grad_norm": 0.4408456344312936, + "learning_rate": 3.852227970653473e-05, + "loss": 0.2097, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11778020858764648, + "step": 705, + "valid_targets_mean": 6837.9, + "valid_targets_min": 4187 + }, + { + "epoch": 1.4889821615949632, + "grad_norm": 0.4382834292978791, + "learning_rate": 3.8482587684872536e-05, + "loss": 0.1968, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08697176724672318, + "step": 710, + "valid_targets_mean": 5439.9, + "valid_targets_min": 318 + }, + { + "epoch": 1.4994753410283317, + "grad_norm": 0.4111781891346404, + "learning_rate": 3.844239063854709e-05, + "loss": 0.2097, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11067388951778412, + "step": 715, + "valid_targets_mean": 8192.6, + "valid_targets_min": 5565 + }, + { + "epoch": 1.5099685204617, + "grad_norm": 0.44356410064904767, + "learning_rate": 3.8401689665916485e-05, + "loss": 0.1965, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0996721088886261, + "step": 720, + "valid_targets_mean": 7183.6, + "valid_targets_min": 2219 + }, + { + "epoch": 1.5204616998950682, + "grad_norm": 0.47697409772680943, + "learning_rate": 3.836048587910828e-05, + "loss": 0.2111, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.093361496925354, + "step": 725, + "valid_targets_mean": 5792.6, + "valid_targets_min": 3653 + }, + { + "epoch": 1.5309548793284367, + "grad_norm": 0.4078324727801479, + "learning_rate": 3.831878040398908e-05, + "loss": 0.201, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1209375262260437, + "step": 730, + "valid_targets_mean": 8004.8, + "valid_targets_min": 5088 + }, + { + "epoch": 1.5414480587618047, + "grad_norm": 0.47197867122584225, + "learning_rate": 3.8276574380133826e-05, + "loss": 0.2025, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09673216938972473, + "step": 735, + "valid_targets_mean": 6692.4, + "valid_targets_min": 1274 + }, + { + "epoch": 1.5519412381951732, + "grad_norm": 0.397058596804463, + "learning_rate": 3.823386896079462e-05, + "loss": 0.2008, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10966745764017105, + "step": 740, + "valid_targets_mean": 7848.8, + "valid_targets_min": 5927 + }, + { + "epoch": 1.5624344176285414, + "grad_norm": 0.38271046277171716, + "learning_rate": 3.8190665312869194e-05, + "loss": 0.2024, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10054107010364532, + "step": 745, + "valid_targets_mean": 7927.5, + "valid_targets_min": 5058 + }, + { + "epoch": 1.5729275970619097, + "grad_norm": 0.4219444101395016, + "learning_rate": 3.814696461686909e-05, + "loss": 0.2052, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1056118905544281, + "step": 750, + "valid_targets_mean": 7095.5, + "valid_targets_min": 4810 + }, + { + "epoch": 1.5834207764952781, + "grad_norm": 0.42783254365913237, + "learning_rate": 3.8102768066887335e-05, + "loss": 0.2014, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09082961827516556, + "step": 755, + "valid_targets_mean": 5680.6, + "valid_targets_min": 1358 + }, + { + "epoch": 1.5939139559286464, + "grad_norm": 0.41825847596382354, + "learning_rate": 3.805807687056586e-05, + "loss": 0.1936, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0992087721824646, + "step": 760, + "valid_targets_mean": 7267.9, + "valid_targets_min": 5131 + }, + { + "epoch": 1.6044071353620146, + "grad_norm": 0.41664988629768784, + "learning_rate": 3.801289224906248e-05, + "loss": 0.2022, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10876358300447464, + "step": 765, + "valid_targets_mean": 7545.2, + "valid_targets_min": 3949 + }, + { + "epoch": 1.614900314795383, + "grad_norm": 0.4137441641357081, + "learning_rate": 3.7967215437017514e-05, + "loss": 0.1975, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10791847109794617, + "step": 770, + "valid_targets_mean": 7212.2, + "valid_targets_min": 4768 + }, + { + "epoch": 1.6253934942287513, + "grad_norm": 0.45219540866514485, + "learning_rate": 3.792104768252011e-05, + "loss": 0.2026, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08391132950782776, + "step": 775, + "valid_targets_mean": 5368.8, + "valid_targets_min": 615 + }, + { + "epoch": 1.6358866736621196, + "grad_norm": 0.3987242030228965, + "learning_rate": 3.787439024707405e-05, + "loss": 0.1994, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09216152131557465, + "step": 780, + "valid_targets_mean": 6794.9, + "valid_targets_min": 483 + }, + { + "epoch": 1.646379853095488, + "grad_norm": 0.45195506699081656, + "learning_rate": 3.782724440556336e-05, + "loss": 0.2034, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.12410935759544373, + "step": 785, + "valid_targets_mean": 6614.8, + "valid_targets_min": 1924 + }, + { + "epoch": 1.656873032528856, + "grad_norm": 0.3964556004108242, + "learning_rate": 3.7779611446217434e-05, + "loss": 0.1927, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09435314685106277, + "step": 790, + "valid_targets_mean": 8387.0, + "valid_targets_min": 3873 + }, + { + "epoch": 1.6673662119622246, + "grad_norm": 0.4077439304977771, + "learning_rate": 3.7731492670575846e-05, + "loss": 0.2028, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08365722000598907, + "step": 795, + "valid_targets_mean": 6189.0, + "valid_targets_min": 332 + }, + { + "epoch": 1.6778593913955928, + "grad_norm": 0.40266792031294235, + "learning_rate": 3.7682889393452775e-05, + "loss": 0.2084, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09859654307365417, + "step": 800, + "valid_targets_mean": 6637.4, + "valid_targets_min": 2242 + }, + { + "epoch": 1.688352570828961, + "grad_norm": 0.408296677682496, + "learning_rate": 3.763380294290109e-05, + "loss": 0.2002, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10042356699705124, + "step": 805, + "valid_targets_mean": 6032.1, + "valid_targets_min": 499 + }, + { + "epoch": 1.6988457502623295, + "grad_norm": 0.41324819064978646, + "learning_rate": 3.7584234660176074e-05, + "loss": 0.1965, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09822611510753632, + "step": 810, + "valid_targets_mean": 7266.5, + "valid_targets_min": 4168 + }, + { + "epoch": 1.7093389296956978, + "grad_norm": 0.4235838127793138, + "learning_rate": 3.753418589969874e-05, + "loss": 0.2015, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10357680171728134, + "step": 815, + "valid_targets_mean": 7080.1, + "valid_targets_min": 3991 + }, + { + "epoch": 1.719832109129066, + "grad_norm": 0.4527516901022831, + "learning_rate": 3.7483658029018854e-05, + "loss": 0.1976, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0956818163394928, + "step": 820, + "valid_targets_mean": 6897.4, + "valid_targets_min": 3801 + }, + { + "epoch": 1.7303252885624345, + "grad_norm": 0.40909146454101564, + "learning_rate": 3.743265242877756e-05, + "loss": 0.1963, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0946725532412529, + "step": 825, + "valid_targets_mean": 7805.6, + "valid_targets_min": 5623 + }, + { + "epoch": 1.7408184679958028, + "grad_norm": 0.412706296069022, + "learning_rate": 3.7381170492669624e-05, + "loss": 0.2063, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10043913871049881, + "step": 830, + "valid_targets_mean": 6858.9, + "valid_targets_min": 3743 + }, + { + "epoch": 1.751311647429171, + "grad_norm": 0.39956247120422494, + "learning_rate": 3.732921362740541e-05, + "loss": 0.2027, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10793577879667282, + "step": 835, + "valid_targets_mean": 7797.1, + "valid_targets_min": 5890 + }, + { + "epoch": 1.7618048268625395, + "grad_norm": 0.41436525323067336, + "learning_rate": 3.727678325267241e-05, + "loss": 0.2016, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09502030909061432, + "step": 840, + "valid_targets_mean": 6251.0, + "valid_targets_min": 3968 + }, + { + "epoch": 1.7722980062959075, + "grad_norm": 0.4797881543462193, + "learning_rate": 3.722388080109643e-05, + "loss": 0.2047, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09240148961544037, + "step": 845, + "valid_targets_mean": 6233.5, + "valid_targets_min": 2263 + }, + { + "epoch": 1.782791185729276, + "grad_norm": 0.4170698240641954, + "learning_rate": 3.7170507718202494e-05, + "loss": 0.194, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07663221657276154, + "step": 850, + "valid_targets_mean": 6374.0, + "valid_targets_min": 4582 + }, + { + "epoch": 1.7932843651626444, + "grad_norm": 0.44755981452881793, + "learning_rate": 3.71166654623753e-05, + "loss": 0.2033, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10387059301137924, + "step": 855, + "valid_targets_mean": 6731.5, + "valid_targets_min": 2450 + }, + { + "epoch": 1.8037775445960125, + "grad_norm": 0.38968333636802666, + "learning_rate": 3.70623555048194e-05, + "loss": 0.197, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1014275923371315, + "step": 860, + "valid_targets_mean": 7916.1, + "valid_targets_min": 4449 + }, + { + "epoch": 1.814270724029381, + "grad_norm": 0.42881262320950814, + "learning_rate": 3.700757932951899e-05, + "loss": 0.1897, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11223836243152618, + "step": 865, + "valid_targets_mean": 7489.8, + "valid_targets_min": 4955 + }, + { + "epoch": 1.8247639034627492, + "grad_norm": 0.42363186988234797, + "learning_rate": 3.695233843319737e-05, + "loss": 0.1914, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09501855075359344, + "step": 870, + "valid_targets_mean": 7973.4, + "valid_targets_min": 5263 + }, + { + "epoch": 1.8352570828961174, + "grad_norm": 0.5182645247622273, + "learning_rate": 3.689663432527602e-05, + "loss": 0.204, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08103088289499283, + "step": 875, + "valid_targets_mean": 4704.4, + "valid_targets_min": 760 + }, + { + "epoch": 1.845750262329486, + "grad_norm": 0.4367148842488578, + "learning_rate": 3.684046852783339e-05, + "loss": 0.1942, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0984354019165039, + "step": 880, + "valid_targets_mean": 7248.1, + "valid_targets_min": 4539 + }, + { + "epoch": 1.8562434417628542, + "grad_norm": 0.38270693579706977, + "learning_rate": 3.67838425755633e-05, + "loss": 0.1938, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08837896585464478, + "step": 885, + "valid_targets_mean": 7056.6, + "valid_targets_min": 4588 + }, + { + "epoch": 1.8667366211962224, + "grad_norm": 0.40854562348221457, + "learning_rate": 3.672675801573298e-05, + "loss": 0.191, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09182108938694, + "step": 890, + "valid_targets_mean": 6378.6, + "valid_targets_min": 3933 + }, + { + "epoch": 1.8772298006295909, + "grad_norm": 0.39924998949276075, + "learning_rate": 3.6669216408140835e-05, + "loss": 0.1867, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08204784989356995, + "step": 895, + "valid_targets_mean": 7945.0, + "valid_targets_min": 4271 + }, + { + "epoch": 1.887722980062959, + "grad_norm": 0.452304158450851, + "learning_rate": 3.661121932507378e-05, + "loss": 0.2065, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11216147243976593, + "step": 900, + "valid_targets_mean": 6035.9, + "valid_targets_min": 1186 + }, + { + "epoch": 1.8982161594963274, + "grad_norm": 0.4404227384776657, + "learning_rate": 3.655276835126432e-05, + "loss": 0.197, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10932591557502747, + "step": 905, + "valid_targets_mean": 7168.1, + "valid_targets_min": 2162 + }, + { + "epoch": 1.9087093389296959, + "grad_norm": 0.4542137143794065, + "learning_rate": 3.6493865083847226e-05, + "loss": 0.1985, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10697949677705765, + "step": 910, + "valid_targets_mean": 6365.0, + "valid_targets_min": 3426 + }, + { + "epoch": 1.9192025183630639, + "grad_norm": 0.404954990032281, + "learning_rate": 3.6434511132315886e-05, + "loss": 0.1969, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10255680978298187, + "step": 915, + "valid_targets_mean": 8168.1, + "valid_targets_min": 5917 + }, + { + "epoch": 1.9296956977964324, + "grad_norm": 0.3897075840461183, + "learning_rate": 3.637470811847834e-05, + "loss": 0.1957, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09185780584812164, + "step": 920, + "valid_targets_mean": 6572.1, + "valid_targets_min": 3549 + }, + { + "epoch": 1.9401888772298006, + "grad_norm": 0.39651237229197694, + "learning_rate": 3.631445767641298e-05, + "loss": 0.1945, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08244992792606354, + "step": 925, + "valid_targets_mean": 6651.2, + "valid_targets_min": 616 + }, + { + "epoch": 1.9506820566631689, + "grad_norm": 0.4484085883874254, + "learning_rate": 3.625376145242386e-05, + "loss": 0.1995, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08280104398727417, + "step": 930, + "valid_targets_mean": 5030.5, + "valid_targets_min": 441 + }, + { + "epoch": 1.9611752360965373, + "grad_norm": 0.46615234233246866, + "learning_rate": 3.619262110499575e-05, + "loss": 0.2035, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10085698962211609, + "step": 935, + "valid_targets_mean": 6407.9, + "valid_targets_min": 3935 + }, + { + "epoch": 1.9716684155299056, + "grad_norm": 0.43598934684074614, + "learning_rate": 3.6131038304748794e-05, + "loss": 0.1964, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06743644177913666, + "step": 940, + "valid_targets_mean": 4689.6, + "valid_targets_min": 322 + }, + { + "epoch": 1.9821615949632738, + "grad_norm": 0.4458899296120857, + "learning_rate": 3.6069014734392874e-05, + "loss": 0.205, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10298512130975723, + "step": 945, + "valid_targets_mean": 7427.0, + "valid_targets_min": 5165 + }, + { + "epoch": 1.9926547743966423, + "grad_norm": 0.4437501410689644, + "learning_rate": 3.600655208868163e-05, + "loss": 0.1964, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.107498399913311, + "step": 950, + "valid_targets_mean": 6088.8, + "valid_targets_min": 3661 + }, + { + "epoch": 2.0020986358866737, + "grad_norm": 0.44017088666167625, + "learning_rate": 3.594365207436615e-05, + "loss": 0.1905, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08558373153209686, + "step": 955, + "valid_targets_mean": 5909.6, + "valid_targets_min": 4031 + }, + { + "epoch": 2.012591815320042, + "grad_norm": 0.5124575314629869, + "learning_rate": 3.588031641014833e-05, + "loss": 0.1841, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09741701185703278, + "step": 960, + "valid_targets_mean": 7122.0, + "valid_targets_min": 4984 + }, + { + "epoch": 2.02308499475341, + "grad_norm": 0.4672307456910968, + "learning_rate": 3.5816546826633925e-05, + "loss": 0.1863, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09654739499092102, + "step": 965, + "valid_targets_mean": 6102.2, + "valid_targets_min": 3175 + }, + { + "epoch": 2.0335781741867787, + "grad_norm": 0.3966122654644788, + "learning_rate": 3.575234506628524e-05, + "loss": 0.1797, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07941184192895889, + "step": 970, + "valid_targets_mean": 6045.1, + "valid_targets_min": 4480 + }, + { + "epoch": 2.0440713536201467, + "grad_norm": 0.4802533802340644, + "learning_rate": 3.568771288337353e-05, + "loss": 0.1827, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06682239472866058, + "step": 975, + "valid_targets_mean": 5073.0, + "valid_targets_min": 325 + }, + { + "epoch": 2.054564533053515, + "grad_norm": 0.46343938423686243, + "learning_rate": 3.562265204393108e-05, + "loss": 0.1853, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09843286871910095, + "step": 980, + "valid_targets_mean": 6371.8, + "valid_targets_min": 3524 + }, + { + "epoch": 2.0650577124868836, + "grad_norm": 0.42827771977065954, + "learning_rate": 3.555716432570293e-05, + "loss": 0.1818, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08701205253601074, + "step": 985, + "valid_targets_mean": 7376.1, + "valid_targets_min": 4391 + }, + { + "epoch": 2.0755508919202517, + "grad_norm": 0.4178123662722871, + "learning_rate": 3.5491251518098296e-05, + "loss": 0.1886, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07098545879125595, + "step": 990, + "valid_targets_mean": 6282.4, + "valid_targets_min": 2473 + }, + { + "epoch": 2.08604407135362, + "grad_norm": 0.4369245156721469, + "learning_rate": 3.542491542214169e-05, + "loss": 0.1843, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09313167631626129, + "step": 995, + "valid_targets_mean": 6320.6, + "valid_targets_min": 1656 + }, + { + "epoch": 2.0965372507869886, + "grad_norm": 0.44284801458408135, + "learning_rate": 3.535815785042371e-05, + "loss": 0.1822, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08810308575630188, + "step": 1000, + "valid_targets_mean": 7316.4, + "valid_targets_min": 3643 + }, + { + "epoch": 2.1070304302203566, + "grad_norm": 0.44247309384226824, + "learning_rate": 3.529098062705148e-05, + "loss": 0.1872, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.100520059466362, + "step": 1005, + "valid_targets_mean": 7869.4, + "valid_targets_min": 5358 + }, + { + "epoch": 2.117523609653725, + "grad_norm": 0.4423483630719316, + "learning_rate": 3.522338558759886e-05, + "loss": 0.1852, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08891274034976959, + "step": 1010, + "valid_targets_mean": 5825.1, + "valid_targets_min": 3607 + }, + { + "epoch": 2.1280167890870936, + "grad_norm": 0.409018786588809, + "learning_rate": 3.5155374579056254e-05, + "loss": 0.1792, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07287713885307312, + "step": 1015, + "valid_targets_mean": 5844.8, + "valid_targets_min": 4117 + }, + { + "epoch": 2.1385099685204616, + "grad_norm": 0.3884171047064284, + "learning_rate": 3.508694945978015e-05, + "loss": 0.181, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09158478677272797, + "step": 1020, + "valid_targets_mean": 6647.6, + "valid_targets_min": 3679 + }, + { + "epoch": 2.14900314795383, + "grad_norm": 0.42805033070866694, + "learning_rate": 3.501811209944232e-05, + "loss": 0.179, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09017354995012283, + "step": 1025, + "valid_targets_mean": 6048.2, + "valid_targets_min": 1404 + }, + { + "epoch": 2.159496327387198, + "grad_norm": 0.42409257025656566, + "learning_rate": 3.494886437897879e-05, + "loss": 0.1888, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11794978380203247, + "step": 1030, + "valid_targets_mean": 8129.6, + "valid_targets_min": 4234 + }, + { + "epoch": 2.1699895068205666, + "grad_norm": 0.4368637883307038, + "learning_rate": 3.4879208190538386e-05, + "loss": 0.1923, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09482675045728683, + "step": 1035, + "valid_targets_mean": 6817.6, + "valid_targets_min": 1165 + }, + { + "epoch": 2.180482686253935, + "grad_norm": 0.4103453370736601, + "learning_rate": 3.4809145437431065e-05, + "loss": 0.186, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10131502151489258, + "step": 1040, + "valid_targets_mean": 8052.9, + "valid_targets_min": 2162 + }, + { + "epoch": 2.190975865687303, + "grad_norm": 0.4357026429613, + "learning_rate": 3.473867803407589e-05, + "loss": 0.1873, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08020630478858948, + "step": 1045, + "valid_targets_mean": 4987.2, + "valid_targets_min": 689 + }, + { + "epoch": 2.2014690451206715, + "grad_norm": 0.45113814281483533, + "learning_rate": 3.466780790594875e-05, + "loss": 0.1824, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08971951901912689, + "step": 1050, + "valid_targets_mean": 6716.1, + "valid_targets_min": 4812 + }, + { + "epoch": 2.21196222455404, + "grad_norm": 0.42132690859795957, + "learning_rate": 3.4596536989529685e-05, + "loss": 0.181, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11115249246358871, + "step": 1055, + "valid_targets_mean": 7619.2, + "valid_targets_min": 5118 + }, + { + "epoch": 2.222455403987408, + "grad_norm": 0.4261634694667753, + "learning_rate": 3.452486723225005e-05, + "loss": 0.1819, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08706986904144287, + "step": 1060, + "valid_targets_mean": 6290.1, + "valid_targets_min": 4356 + }, + { + "epoch": 2.2329485834207765, + "grad_norm": 0.4270185460150402, + "learning_rate": 3.445280059243926e-05, + "loss": 0.1906, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0887756198644638, + "step": 1065, + "valid_targets_mean": 6846.2, + "valid_targets_min": 1784 + }, + { + "epoch": 2.243441762854145, + "grad_norm": 0.4393645555629806, + "learning_rate": 3.438033903927128e-05, + "loss": 0.1887, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10316319763660431, + "step": 1070, + "valid_targets_mean": 7630.8, + "valid_targets_min": 5329 + }, + { + "epoch": 2.253934942287513, + "grad_norm": 0.4074615753689165, + "learning_rate": 3.430748455271082e-05, + "loss": 0.1863, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09186771512031555, + "step": 1075, + "valid_targets_mean": 6324.9, + "valid_targets_min": 801 + }, + { + "epoch": 2.2644281217208815, + "grad_norm": 0.4551348547576463, + "learning_rate": 3.423423912345925e-05, + "loss": 0.1873, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09558658301830292, + "step": 1080, + "valid_targets_mean": 6825.9, + "valid_targets_min": 4330 + }, + { + "epoch": 2.27492130115425, + "grad_norm": 0.4063524767748916, + "learning_rate": 3.416060475290019e-05, + "loss": 0.186, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09127890318632126, + "step": 1085, + "valid_targets_mean": 7566.4, + "valid_targets_min": 5486 + }, + { + "epoch": 2.285414480587618, + "grad_norm": 0.42727438558638087, + "learning_rate": 3.408658345304482e-05, + "loss": 0.1848, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09472738206386566, + "step": 1090, + "valid_targets_mean": 7277.4, + "valid_targets_min": 4568 + }, + { + "epoch": 2.2959076600209865, + "grad_norm": 0.42540236336669357, + "learning_rate": 3.401217724647693e-05, + "loss": 0.1785, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08184987306594849, + "step": 1095, + "valid_targets_mean": 6295.1, + "valid_targets_min": 1003 + }, + { + "epoch": 2.3064008394543545, + "grad_norm": 0.48732574293736025, + "learning_rate": 3.393738816629761e-05, + "loss": 0.1868, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09464119374752045, + "step": 1100, + "valid_targets_mean": 6031.1, + "valid_targets_min": 970 + }, + { + "epoch": 2.316894018887723, + "grad_norm": 0.39703792599369886, + "learning_rate": 3.386221825606975e-05, + "loss": 0.1809, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0924624651670456, + "step": 1105, + "valid_targets_mean": 7210.1, + "valid_targets_min": 500 + }, + { + "epoch": 2.3273871983210914, + "grad_norm": 0.42365586554451284, + "learning_rate": 3.378666956976215e-05, + "loss": 0.1852, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09636393189430237, + "step": 1110, + "valid_targets_mean": 7290.6, + "valid_targets_min": 4543 + }, + { + "epoch": 2.3378803777544594, + "grad_norm": 0.4820250065269673, + "learning_rate": 3.371074417169343e-05, + "loss": 0.1842, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09943065047264099, + "step": 1115, + "valid_targets_mean": 7099.0, + "valid_targets_min": 3746 + }, + { + "epoch": 2.348373557187828, + "grad_norm": 0.35705638541542845, + "learning_rate": 3.363444413647561e-05, + "loss": 0.1769, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09401725232601166, + "step": 1120, + "valid_targets_mean": 7498.1, + "valid_targets_min": 5871 + }, + { + "epoch": 2.3588667366211964, + "grad_norm": 0.3911955004684999, + "learning_rate": 3.3557771548957425e-05, + "loss": 0.1861, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09180471301078796, + "step": 1125, + "valid_targets_mean": 7101.4, + "valid_targets_min": 4271 + }, + { + "epoch": 2.3693599160545644, + "grad_norm": 0.3964501833121612, + "learning_rate": 3.348072850416737e-05, + "loss": 0.182, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10606725513935089, + "step": 1130, + "valid_targets_mean": 8038.5, + "valid_targets_min": 5168 + }, + { + "epoch": 2.379853095487933, + "grad_norm": 0.455907895656185, + "learning_rate": 3.340331710725641e-05, + "loss": 0.1909, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08738633990287781, + "step": 1135, + "valid_targets_mean": 7328.1, + "valid_targets_min": 4889 + }, + { + "epoch": 2.390346274921301, + "grad_norm": 0.36903201730487273, + "learning_rate": 3.3325539473440546e-05, + "loss": 0.1765, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08207184076309204, + "step": 1140, + "valid_targets_mean": 8535.2, + "valid_targets_min": 4353 + }, + { + "epoch": 2.4008394543546694, + "grad_norm": 0.44307801071227315, + "learning_rate": 3.324739772794292e-05, + "loss": 0.1886, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09129495918750763, + "step": 1145, + "valid_targets_mean": 6202.0, + "valid_targets_min": 3426 + }, + { + "epoch": 2.411332633788038, + "grad_norm": 0.4087694969229485, + "learning_rate": 3.316889400593581e-05, + "loss": 0.1783, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10222087800502777, + "step": 1150, + "valid_targets_mean": 7357.5, + "valid_targets_min": 4601 + }, + { + "epoch": 2.421825813221406, + "grad_norm": 0.45778394915553267, + "learning_rate": 3.309003045248224e-05, + "loss": 0.1764, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08947615325450897, + "step": 1155, + "valid_targets_mean": 6312.9, + "valid_targets_min": 2011 + }, + { + "epoch": 2.4323189926547744, + "grad_norm": 0.4582642436473117, + "learning_rate": 3.301080922247743e-05, + "loss": 0.1782, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09737248718738556, + "step": 1160, + "valid_targets_mean": 6401.2, + "valid_targets_min": 854 + }, + { + "epoch": 2.442812172088143, + "grad_norm": 0.39629383689675235, + "learning_rate": 3.2931232480589853e-05, + "loss": 0.1666, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0754726454615593, + "step": 1165, + "valid_targets_mean": 6189.4, + "valid_targets_min": 200 + }, + { + "epoch": 2.453305351521511, + "grad_norm": 0.459614313277168, + "learning_rate": 3.285130240120212e-05, + "loss": 0.1764, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10912300646305084, + "step": 1170, + "valid_targets_mean": 7041.2, + "valid_targets_min": 4857 + }, + { + "epoch": 2.4637985309548793, + "grad_norm": 0.3952639860647928, + "learning_rate": 3.277102116835157e-05, + "loss": 0.1775, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0902990773320198, + "step": 1175, + "valid_targets_mean": 6684.1, + "valid_targets_min": 5614 + }, + { + "epoch": 2.474291710388248, + "grad_norm": 0.4179206732119482, + "learning_rate": 3.2690390975670554e-05, + "loss": 0.1837, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09472492337226868, + "step": 1180, + "valid_targets_mean": 7415.2, + "valid_targets_min": 4528 + }, + { + "epoch": 2.484784889821616, + "grad_norm": 0.43289452705337866, + "learning_rate": 3.260941402632655e-05, + "loss": 0.1806, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08920637518167496, + "step": 1185, + "valid_targets_mean": 6525.2, + "valid_targets_min": 4823 + }, + { + "epoch": 2.4952780692549843, + "grad_norm": 0.4086445458434871, + "learning_rate": 3.252809253296192e-05, + "loss": 0.1811, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0858517736196518, + "step": 1190, + "valid_targets_mean": 6638.0, + "valid_targets_min": 477 + }, + { + "epoch": 2.5057712486883528, + "grad_norm": 0.41600280138486223, + "learning_rate": 3.244642871763348e-05, + "loss": 0.18, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0942123606801033, + "step": 1195, + "valid_targets_mean": 7091.9, + "valid_targets_min": 3829 + }, + { + "epoch": 2.516264428121721, + "grad_norm": 0.4538417196886871, + "learning_rate": 3.236442481175176e-05, + "loss": 0.1862, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08888477832078934, + "step": 1200, + "valid_targets_mean": 6174.1, + "valid_targets_min": 3753 + }, + { + "epoch": 2.5267576075550893, + "grad_norm": 0.38418266028614384, + "learning_rate": 3.228208305602004e-05, + "loss": 0.1762, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07455168664455414, + "step": 1205, + "valid_targets_mean": 6350.9, + "valid_targets_min": 3549 + }, + { + "epoch": 2.5372507869884577, + "grad_norm": 0.5054440973584818, + "learning_rate": 3.219940570037314e-05, + "loss": 0.1819, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09286289662122726, + "step": 1210, + "valid_targets_mean": 5103.2, + "valid_targets_min": 1973 + }, + { + "epoch": 2.5477439664218258, + "grad_norm": 0.40792500922036945, + "learning_rate": 3.211639500391592e-05, + "loss": 0.1841, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09160646796226501, + "step": 1215, + "valid_targets_mean": 7091.2, + "valid_targets_min": 5960 + }, + { + "epoch": 2.5582371458551942, + "grad_norm": 0.45707439908719016, + "learning_rate": 3.2033053234861557e-05, + "loss": 0.1805, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1148623377084732, + "step": 1220, + "valid_targets_mean": 5661.2, + "valid_targets_min": 2723 + }, + { + "epoch": 2.5687303252885627, + "grad_norm": 0.42073329032911105, + "learning_rate": 3.194938267046958e-05, + "loss": 0.1832, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07760588824748993, + "step": 1225, + "valid_targets_mean": 6177.0, + "valid_targets_min": 3413 + }, + { + "epoch": 2.5792235047219307, + "grad_norm": 0.4428605708045142, + "learning_rate": 3.186538559698364e-05, + "loss": 0.1894, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07915692031383514, + "step": 1230, + "valid_targets_mean": 5899.2, + "valid_targets_min": 1483 + }, + { + "epoch": 2.589716684155299, + "grad_norm": 0.45836056899440347, + "learning_rate": 3.1781064309569015e-05, + "loss": 0.177, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07306934893131256, + "step": 1235, + "valid_targets_mean": 5256.4, + "valid_targets_min": 2124 + }, + { + "epoch": 2.6002098635886672, + "grad_norm": 0.4020244934968488, + "learning_rate": 3.1696421112249925e-05, + "loss": 0.1898, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.11009399592876434, + "step": 1240, + "valid_targets_mean": 7515.2, + "valid_targets_min": 4308 + }, + { + "epoch": 2.6107030430220357, + "grad_norm": 0.4022462707955614, + "learning_rate": 3.16114583178466e-05, + "loss": 0.1839, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08671146631240845, + "step": 1245, + "valid_targets_mean": 6869.9, + "valid_targets_min": 1036 + }, + { + "epoch": 2.6211962224554037, + "grad_norm": 0.47498208512338697, + "learning_rate": 3.1526178247912005e-05, + "loss": 0.1818, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07975402474403381, + "step": 1250, + "valid_targets_mean": 6258.4, + "valid_targets_min": 2139 + }, + { + "epoch": 2.631689401888772, + "grad_norm": 0.41197440308021555, + "learning_rate": 3.1440583232668484e-05, + "loss": 0.1913, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10407009720802307, + "step": 1255, + "valid_targets_mean": 6670.4, + "valid_targets_min": 5530 + }, + { + "epoch": 2.6421825813221407, + "grad_norm": 0.4164427328588479, + "learning_rate": 3.1354675610944046e-05, + "loss": 0.1876, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10782954096794128, + "step": 1260, + "valid_targets_mean": 7447.9, + "valid_targets_min": 5403 + }, + { + "epoch": 2.6526757607555087, + "grad_norm": 0.41741776877299636, + "learning_rate": 3.1268457730108454e-05, + "loss": 0.1853, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10540814697742462, + "step": 1265, + "valid_targets_mean": 6496.2, + "valid_targets_min": 960 + }, + { + "epoch": 2.663168940188877, + "grad_norm": 0.42961826653694224, + "learning_rate": 3.1181931946009125e-05, + "loss": 0.1818, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10980040580034256, + "step": 1270, + "valid_targets_mean": 6583.2, + "valid_targets_min": 3258 + }, + { + "epoch": 2.6736621196222456, + "grad_norm": 0.3829942173928751, + "learning_rate": 3.1095100622906715e-05, + "loss": 0.1756, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09991838037967682, + "step": 1275, + "valid_targets_mean": 8077.5, + "valid_targets_min": 4044 + }, + { + "epoch": 2.6841552990556137, + "grad_norm": 0.43251339018074875, + "learning_rate": 3.100796613341055e-05, + "loss": 0.1793, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08717729896306992, + "step": 1280, + "valid_targets_mean": 6728.1, + "valid_targets_min": 4296 + }, + { + "epoch": 2.694648478488982, + "grad_norm": 0.42913123037571893, + "learning_rate": 3.092053085841375e-05, + "loss": 0.183, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0834718644618988, + "step": 1285, + "valid_targets_mean": 6064.0, + "valid_targets_min": 3509 + }, + { + "epoch": 2.7051416579223506, + "grad_norm": 0.37766342906026606, + "learning_rate": 3.083279718702824e-05, + "loss": 0.1815, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08643440902233124, + "step": 1290, + "valid_targets_mean": 7712.2, + "valid_targets_min": 5434 + }, + { + "epoch": 2.7156348373557186, + "grad_norm": 0.3689035325149897, + "learning_rate": 3.074476751651941e-05, + "loss": 0.1758, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08829987049102783, + "step": 1295, + "valid_targets_mean": 7431.0, + "valid_targets_min": 5088 + }, + { + "epoch": 2.726128016789087, + "grad_norm": 0.4223787524659475, + "learning_rate": 3.065644425224063e-05, + "loss": 0.1792, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07150721549987793, + "step": 1300, + "valid_targets_mean": 5736.1, + "valid_targets_min": 3613 + }, + { + "epoch": 2.7366211962224556, + "grad_norm": 0.4531507032318413, + "learning_rate": 3.056782980756754e-05, + "loss": 0.1811, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08609944581985474, + "step": 1305, + "valid_targets_mean": 6046.4, + "valid_targets_min": 5153 + }, + { + "epoch": 2.7471143756558236, + "grad_norm": 0.4090522165313531, + "learning_rate": 3.047892660383209e-05, + "loss": 0.1766, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09299299120903015, + "step": 1310, + "valid_targets_mean": 6313.6, + "valid_targets_min": 1945 + }, + { + "epoch": 2.757607555089192, + "grad_norm": 0.41103975255649333, + "learning_rate": 3.0389737070256388e-05, + "loss": 0.1818, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0955432802438736, + "step": 1315, + "valid_targets_mean": 6863.9, + "valid_targets_min": 3300 + }, + { + "epoch": 2.7681007345225606, + "grad_norm": 0.4426263135685565, + "learning_rate": 3.0300263643886302e-05, + "loss": 0.1792, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08449938893318176, + "step": 1320, + "valid_targets_mean": 5699.4, + "valid_targets_min": 3892 + }, + { + "epoch": 2.7785939139559286, + "grad_norm": 0.40123747981786717, + "learning_rate": 3.0210508769524906e-05, + "loss": 0.1786, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0938343033194542, + "step": 1325, + "valid_targets_mean": 7248.2, + "valid_targets_min": 3304 + }, + { + "epoch": 2.789087093389297, + "grad_norm": 0.4160444015132139, + "learning_rate": 3.0120474899665643e-05, + "loss": 0.1837, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08788036555051804, + "step": 1330, + "valid_targets_mean": 6417.8, + "valid_targets_min": 3648 + }, + { + "epoch": 2.7995802728226655, + "grad_norm": 0.37098940484651294, + "learning_rate": 3.0030164494425333e-05, + "loss": 0.1775, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10098370909690857, + "step": 1335, + "valid_targets_mean": 7696.4, + "valid_targets_min": 4310 + }, + { + "epoch": 2.8100734522560336, + "grad_norm": 0.3987879216445246, + "learning_rate": 2.993958002147695e-05, + "loss": 0.1767, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.081761434674263, + "step": 1340, + "valid_targets_mean": 5893.6, + "valid_targets_min": 324 + }, + { + "epoch": 2.820566631689402, + "grad_norm": 0.40020051809598217, + "learning_rate": 2.9848723955982168e-05, + "loss": 0.1848, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0817759558558464, + "step": 1345, + "valid_targets_mean": 7260.2, + "valid_targets_min": 5517 + }, + { + "epoch": 2.83105981112277, + "grad_norm": 0.39035620635372087, + "learning_rate": 2.9757598780523784e-05, + "loss": 0.1834, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07814863324165344, + "step": 1350, + "valid_targets_mean": 6776.4, + "valid_targets_min": 5021 + }, + { + "epoch": 2.8415529905561385, + "grad_norm": 0.4155116705285445, + "learning_rate": 2.966620698503781e-05, + "loss": 0.1778, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0795905739068985, + "step": 1355, + "valid_targets_mean": 5605.9, + "valid_targets_min": 867 + }, + { + "epoch": 2.8520461699895066, + "grad_norm": 0.3651857984148604, + "learning_rate": 2.9574551066745518e-05, + "loss": 0.1703, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07493982464075089, + "step": 1360, + "valid_targets_mean": 7037.2, + "valid_targets_min": 3814 + }, + { + "epoch": 2.862539349422875, + "grad_norm": 0.4968863677791486, + "learning_rate": 2.9482633530085122e-05, + "loss": 0.1778, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07866144180297852, + "step": 1365, + "valid_targets_mean": 6898.5, + "valid_targets_min": 4352 + }, + { + "epoch": 2.8730325288562435, + "grad_norm": 0.38513687465002544, + "learning_rate": 2.939045688664342e-05, + "loss": 0.1872, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08800777792930603, + "step": 1370, + "valid_targets_mean": 7048.1, + "valid_targets_min": 827 + }, + { + "epoch": 2.8835257082896115, + "grad_norm": 0.3916384621636145, + "learning_rate": 2.929802365508712e-05, + "loss": 0.1806, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06416353583335876, + "step": 1375, + "valid_targets_mean": 5427.6, + "valid_targets_min": 2208 + }, + { + "epoch": 2.89401888772298, + "grad_norm": 0.41249100527389265, + "learning_rate": 2.920533636109404e-05, + "loss": 0.1799, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10218751430511475, + "step": 1380, + "valid_targets_mean": 7861.0, + "valid_targets_min": 4433 + }, + { + "epoch": 2.9045120671563485, + "grad_norm": 0.42276159919001566, + "learning_rate": 2.911239753728407e-05, + "loss": 0.179, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07027000188827515, + "step": 1385, + "valid_targets_mean": 6150.2, + "valid_targets_min": 3833 + }, + { + "epoch": 2.9150052465897165, + "grad_norm": 0.40300943305559933, + "learning_rate": 2.9019209723150008e-05, + "loss": 0.1816, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08272185921669006, + "step": 1390, + "valid_targets_mean": 6389.5, + "valid_targets_min": 3653 + }, + { + "epoch": 2.925498426023085, + "grad_norm": 0.43316087735907716, + "learning_rate": 2.8925775464988154e-05, + "loss": 0.1846, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0895342081785202, + "step": 1395, + "valid_targets_mean": 7209.4, + "valid_targets_min": 5388 + }, + { + "epoch": 2.9359916054564534, + "grad_norm": 0.4064825781797068, + "learning_rate": 2.8832097315828706e-05, + "loss": 0.1715, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08410394191741943, + "step": 1400, + "valid_targets_mean": 7775.5, + "valid_targets_min": 4484 + }, + { + "epoch": 2.9464847848898215, + "grad_norm": 0.40703924065434194, + "learning_rate": 2.873817783536605e-05, + "loss": 0.1801, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1052619144320488, + "step": 1405, + "valid_targets_mean": 7129.4, + "valid_targets_min": 4160 + }, + { + "epoch": 2.95697796432319, + "grad_norm": 0.5452337380337632, + "learning_rate": 2.8644019589888768e-05, + "loss": 0.183, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09623530507087708, + "step": 1410, + "valid_targets_mean": 7220.1, + "valid_targets_min": 686 + }, + { + "epoch": 2.9674711437565584, + "grad_norm": 0.44040184620813055, + "learning_rate": 2.854962515220957e-05, + "loss": 0.18, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07843069732189178, + "step": 1415, + "valid_targets_mean": 6220.1, + "valid_targets_min": 4010 + }, + { + "epoch": 2.9779643231899264, + "grad_norm": 0.4344119880088962, + "learning_rate": 2.845499710159494e-05, + "loss": 0.1795, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08755984902381897, + "step": 1420, + "valid_targets_mean": 5477.9, + "valid_targets_min": 459 + }, + { + "epoch": 2.988457502623295, + "grad_norm": 0.37362054458345395, + "learning_rate": 2.8360138023694693e-05, + "loss": 0.1848, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0953589528799057, + "step": 1425, + "valid_targets_mean": 7391.5, + "valid_targets_min": 4296 + }, + { + "epoch": 2.9989506820566634, + "grad_norm": 0.39440895581693197, + "learning_rate": 2.826505051047131e-05, + "loss": 0.1833, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09483747184276581, + "step": 1430, + "valid_targets_mean": 6179.8, + "valid_targets_min": 533 + }, + { + "epoch": 3.0083945435466948, + "grad_norm": 0.4041240911588325, + "learning_rate": 2.816973716012912e-05, + "loss": 0.1621, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09408942610025406, + "step": 1435, + "valid_targets_mean": 7690.8, + "valid_targets_min": 5866 + }, + { + "epoch": 3.018887722980063, + "grad_norm": 0.42307810208526814, + "learning_rate": 2.807420057704331e-05, + "loss": 0.1656, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07292736321687698, + "step": 1440, + "valid_targets_mean": 6039.2, + "valid_targets_min": 3628 + }, + { + "epoch": 3.0293809024134313, + "grad_norm": 0.4101584152685082, + "learning_rate": 2.797844337168874e-05, + "loss": 0.1681, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08154639601707458, + "step": 1445, + "valid_targets_mean": 6886.5, + "valid_targets_min": 4324 + }, + { + "epoch": 3.0398740818467997, + "grad_norm": 0.5043063134998965, + "learning_rate": 2.7882468160568653e-05, + "loss": 0.1675, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0975562334060669, + "step": 1450, + "valid_targets_mean": 6367.8, + "valid_targets_min": 198 + }, + { + "epoch": 3.0503672612801678, + "grad_norm": 0.44588200134148054, + "learning_rate": 2.778627756614314e-05, + "loss": 0.1639, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08662178367376328, + "step": 1455, + "valid_targets_mean": 6065.1, + "valid_targets_min": 3582 + }, + { + "epoch": 3.0608604407135362, + "grad_norm": 0.40847902962750265, + "learning_rate": 2.768987421675751e-05, + "loss": 0.1672, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.086537204682827, + "step": 1460, + "valid_targets_mean": 6205.0, + "valid_targets_min": 594 + }, + { + "epoch": 3.0713536201469047, + "grad_norm": 0.4139291851349987, + "learning_rate": 2.7593260746570448e-05, + "loss": 0.1732, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07508321106433868, + "step": 1465, + "valid_targets_mean": 6099.2, + "valid_targets_min": 4512 + }, + { + "epoch": 3.0818467995802727, + "grad_norm": 0.4970832264114012, + "learning_rate": 2.749643979548207e-05, + "loss": 0.165, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0886954590678215, + "step": 1470, + "valid_targets_mean": 7735.8, + "valid_targets_min": 5072 + }, + { + "epoch": 3.092339979013641, + "grad_norm": 0.4209444210135973, + "learning_rate": 2.7399414009061767e-05, + "loss": 0.1706, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07565881311893463, + "step": 1475, + "valid_targets_mean": 6115.8, + "valid_targets_min": 4781 + }, + { + "epoch": 3.1028331584470092, + "grad_norm": 0.4078773311168684, + "learning_rate": 2.7302186038475915e-05, + "loss": 0.1648, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07618191093206406, + "step": 1480, + "valid_targets_mean": 5800.4, + "valid_targets_min": 200 + }, + { + "epoch": 3.1133263378803777, + "grad_norm": 0.45033910300912927, + "learning_rate": 2.7204758540415456e-05, + "loss": 0.167, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09146006405353546, + "step": 1485, + "valid_targets_mean": 7705.2, + "valid_targets_min": 5064 + }, + { + "epoch": 3.123819517313746, + "grad_norm": 0.43851175707436846, + "learning_rate": 2.710713417702327e-05, + "loss": 0.1748, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08708958327770233, + "step": 1490, + "valid_targets_mean": 6961.9, + "valid_targets_min": 2985 + }, + { + "epoch": 3.134312696747114, + "grad_norm": 0.3993240690243315, + "learning_rate": 2.700931561582147e-05, + "loss": 0.1648, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06856022775173187, + "step": 1495, + "valid_targets_mean": 6112.1, + "valid_targets_min": 4307 + }, + { + "epoch": 3.1448058761804827, + "grad_norm": 0.402674271094955, + "learning_rate": 2.6911305529638482e-05, + "loss": 0.1662, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08153347671031952, + "step": 1500, + "valid_targets_mean": 6892.5, + "valid_targets_min": 3414 + }, + { + "epoch": 3.155299055613851, + "grad_norm": 0.4801621987181525, + "learning_rate": 2.681310659653605e-05, + "loss": 0.1712, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09445063769817352, + "step": 1505, + "valid_targets_mean": 6658.4, + "valid_targets_min": 2874 + }, + { + "epoch": 3.165792235047219, + "grad_norm": 0.417576742594345, + "learning_rate": 2.671472149973602e-05, + "loss": 0.169, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07494966685771942, + "step": 1510, + "valid_targets_mean": 5719.5, + "valid_targets_min": 219 + }, + { + "epoch": 3.1762854144805877, + "grad_norm": 0.46567069961720287, + "learning_rate": 2.6616152927547037e-05, + "loss": 0.1672, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08902396261692047, + "step": 1515, + "valid_targets_mean": 6702.5, + "valid_targets_min": 4458 + }, + { + "epoch": 3.186778593913956, + "grad_norm": 0.466037224661499, + "learning_rate": 2.6517403573291112e-05, + "loss": 0.1681, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0927564799785614, + "step": 1520, + "valid_targets_mean": 6709.8, + "valid_targets_min": 5235 + }, + { + "epoch": 3.197271773347324, + "grad_norm": 0.3994725862205657, + "learning_rate": 2.6418476135229978e-05, + "loss": 0.1698, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07826617360115051, + "step": 1525, + "valid_targets_mean": 6350.1, + "valid_targets_min": 3938 + }, + { + "epoch": 3.2077649527806926, + "grad_norm": 0.38658746179736664, + "learning_rate": 2.6319373316491425e-05, + "loss": 0.1689, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08795974403619766, + "step": 1530, + "valid_targets_mean": 8101.9, + "valid_targets_min": 5368 + }, + { + "epoch": 3.2182581322140607, + "grad_norm": 0.3880560459982342, + "learning_rate": 2.622009782499537e-05, + "loss": 0.1656, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08751142024993896, + "step": 1535, + "valid_targets_mean": 8263.8, + "valid_targets_min": 5121 + }, + { + "epoch": 3.228751311647429, + "grad_norm": 0.4449467557459084, + "learning_rate": 2.6120652373379944e-05, + "loss": 0.1798, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08977397531270981, + "step": 1540, + "valid_targets_mean": 6350.8, + "valid_targets_min": 3487 + }, + { + "epoch": 3.2392444910807976, + "grad_norm": 0.4173952696657384, + "learning_rate": 2.6021039678927275e-05, + "loss": 0.1756, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09539172053337097, + "step": 1545, + "valid_targets_mean": 7225.2, + "valid_targets_min": 1429 + }, + { + "epoch": 3.2497376705141656, + "grad_norm": 0.4248128734297759, + "learning_rate": 2.592126246348934e-05, + "loss": 0.1644, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09295059740543365, + "step": 1550, + "valid_targets_mean": 6882.9, + "valid_targets_min": 1358 + }, + { + "epoch": 3.260230849947534, + "grad_norm": 0.41904240124939646, + "learning_rate": 2.5821323453413517e-05, + "loss": 0.1717, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06466634571552277, + "step": 1555, + "valid_targets_mean": 5784.2, + "valid_targets_min": 614 + }, + { + "epoch": 3.2707240293809026, + "grad_norm": 0.504583742170718, + "learning_rate": 2.5721225379468133e-05, + "loss": 0.1675, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07877649366855621, + "step": 1560, + "valid_targets_mean": 6239.6, + "valid_targets_min": 2709 + }, + { + "epoch": 3.2812172088142706, + "grad_norm": 0.40444247280749485, + "learning_rate": 2.5620970976767816e-05, + "loss": 0.1681, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07856304943561554, + "step": 1565, + "valid_targets_mean": 6097.6, + "valid_targets_min": 4027 + }, + { + "epoch": 3.291710388247639, + "grad_norm": 0.41156567255930954, + "learning_rate": 2.5520562984698787e-05, + "loss": 0.1721, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08800213783979416, + "step": 1570, + "valid_targets_mean": 6618.8, + "valid_targets_min": 4601 + }, + { + "epoch": 3.3022035676810075, + "grad_norm": 0.43118019165224486, + "learning_rate": 2.542000414684399e-05, + "loss": 0.1673, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09648343920707703, + "step": 1575, + "valid_targets_mean": 7660.2, + "valid_targets_min": 3833 + }, + { + "epoch": 3.3126967471143756, + "grad_norm": 0.40607862134841427, + "learning_rate": 2.5319297210908143e-05, + "loss": 0.1672, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07327889651060104, + "step": 1580, + "valid_targets_mean": 6962.8, + "valid_targets_min": 4476 + }, + { + "epoch": 3.323189926547744, + "grad_norm": 0.41004678014175855, + "learning_rate": 2.5218444928642633e-05, + "loss": 0.1705, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08325806260108948, + "step": 1585, + "valid_targets_mean": 7239.4, + "valid_targets_min": 5199 + }, + { + "epoch": 3.3336831059811125, + "grad_norm": 0.41435882843308675, + "learning_rate": 2.5117450055770334e-05, + "loss": 0.1724, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07585517317056656, + "step": 1590, + "valid_targets_mean": 6045.6, + "valid_targets_min": 1814 + }, + { + "epoch": 3.3441762854144805, + "grad_norm": 0.5125445608511127, + "learning_rate": 2.5016315351910355e-05, + "loss": 0.1671, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08650140464305878, + "step": 1595, + "valid_targets_mean": 5994.4, + "valid_targets_min": 1710 + }, + { + "epoch": 3.354669464847849, + "grad_norm": 0.4333139383208153, + "learning_rate": 2.4915043580502543e-05, + "loss": 0.1695, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08141575008630753, + "step": 1600, + "valid_targets_mean": 7064.8, + "valid_targets_min": 5358 + }, + { + "epoch": 3.365162644281217, + "grad_norm": 0.4306329822361818, + "learning_rate": 2.4813637508732068e-05, + "loss": 0.1695, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08281993120908737, + "step": 1605, + "valid_targets_mean": 5479.8, + "valid_targets_min": 4160 + }, + { + "epoch": 3.3756558237145855, + "grad_norm": 0.4587442567740161, + "learning_rate": 2.4712099907453756e-05, + "loss": 0.165, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10962776839733124, + "step": 1610, + "valid_targets_mean": 7223.1, + "valid_targets_min": 3685 + }, + { + "epoch": 3.386149003147954, + "grad_norm": 0.4990644006983264, + "learning_rate": 2.4610433551116374e-05, + "loss": 0.1601, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0753771960735321, + "step": 1615, + "valid_targets_mean": 5392.4, + "valid_targets_min": 630 + }, + { + "epoch": 3.396642182581322, + "grad_norm": 0.3877879093803534, + "learning_rate": 2.4508641217686868e-05, + "loss": 0.1728, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07319890707731247, + "step": 1620, + "valid_targets_mean": 6380.2, + "valid_targets_min": 3735 + }, + { + "epoch": 3.4071353620146905, + "grad_norm": 0.43457812934938034, + "learning_rate": 2.4406725688574394e-05, + "loss": 0.1663, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08114920556545258, + "step": 1625, + "valid_targets_mean": 5747.9, + "valid_targets_min": 332 + }, + { + "epoch": 3.417628541448059, + "grad_norm": 0.4146879970061769, + "learning_rate": 2.430468974855438e-05, + "loss": 0.1669, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10119670629501343, + "step": 1630, + "valid_targets_mean": 6672.0, + "valid_targets_min": 768 + }, + { + "epoch": 3.428121720881427, + "grad_norm": 0.41663529464095644, + "learning_rate": 2.4202536185692386e-05, + "loss": 0.1676, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09076064825057983, + "step": 1635, + "valid_targets_mean": 7449.0, + "valid_targets_min": 6018 + }, + { + "epoch": 3.4386149003147954, + "grad_norm": 0.3892902822253302, + "learning_rate": 2.410026779126794e-05, + "loss": 0.167, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07737547159194946, + "step": 1640, + "valid_targets_mean": 7792.4, + "valid_targets_min": 3941 + }, + { + "epoch": 3.4491080797481635, + "grad_norm": 0.379012381873314, + "learning_rate": 2.399788735969827e-05, + "loss": 0.1714, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.079659603536129, + "step": 1645, + "valid_targets_mean": 7189.1, + "valid_targets_min": 4689 + }, + { + "epoch": 3.459601259181532, + "grad_norm": 0.4589264699535379, + "learning_rate": 2.3895397688461946e-05, + "loss": 0.1674, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09099084138870239, + "step": 1650, + "valid_targets_mean": 7408.6, + "valid_targets_min": 3745 + }, + { + "epoch": 3.4700944386149004, + "grad_norm": 0.39875397124527184, + "learning_rate": 2.379280157802245e-05, + "loss": 0.165, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07003942131996155, + "step": 1655, + "valid_targets_mean": 5763.9, + "valid_targets_min": 1705 + }, + { + "epoch": 3.4805876180482684, + "grad_norm": 0.4102811878767304, + "learning_rate": 2.3690101831751627e-05, + "loss": 0.1673, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08362659811973572, + "step": 1660, + "valid_targets_mean": 6938.2, + "valid_targets_min": 5100 + }, + { + "epoch": 3.491080797481637, + "grad_norm": 0.3754963994529459, + "learning_rate": 2.3587301255853122e-05, + "loss": 0.1671, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08203849196434021, + "step": 1665, + "valid_targets_mean": 6376.9, + "valid_targets_min": 3209 + }, + { + "epoch": 3.5015739769150054, + "grad_norm": 0.4341372107892201, + "learning_rate": 2.3484402659285666e-05, + "loss": 0.1715, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07443380355834961, + "step": 1670, + "valid_targets_mean": 5842.4, + "valid_targets_min": 4031 + }, + { + "epoch": 3.5120671563483734, + "grad_norm": 0.39323300910634584, + "learning_rate": 2.3381408853686362e-05, + "loss": 0.1677, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09358431398868561, + "step": 1675, + "valid_targets_mean": 7877.4, + "valid_targets_min": 6369 + }, + { + "epoch": 3.522560335781742, + "grad_norm": 0.4527760664816108, + "learning_rate": 2.327832265329381e-05, + "loss": 0.1674, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08503014594316483, + "step": 1680, + "valid_targets_mean": 6846.4, + "valid_targets_min": 4914 + }, + { + "epoch": 3.5330535152151104, + "grad_norm": 0.4129749408602394, + "learning_rate": 2.3175146874871265e-05, + "loss": 0.1672, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08560128509998322, + "step": 1685, + "valid_targets_mean": 6881.5, + "valid_targets_min": 3857 + }, + { + "epoch": 3.5435466946484784, + "grad_norm": 0.4184934074251371, + "learning_rate": 2.3071884337629638e-05, + "loss": 0.1718, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06725214421749115, + "step": 1690, + "valid_targets_mean": 5971.5, + "valid_targets_min": 669 + }, + { + "epoch": 3.554039874081847, + "grad_norm": 0.4154428663277812, + "learning_rate": 2.296853786315045e-05, + "loss": 0.1635, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07005719840526581, + "step": 1695, + "valid_targets_mean": 6704.8, + "valid_targets_min": 3304 + }, + { + "epoch": 3.5645330535152153, + "grad_norm": 0.3992991439781619, + "learning_rate": 2.2865110275308774e-05, + "loss": 0.1701, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09991228580474854, + "step": 1700, + "valid_targets_mean": 8103.6, + "valid_targets_min": 4641 + }, + { + "epoch": 3.5750262329485833, + "grad_norm": 0.43006525202928947, + "learning_rate": 2.276160440019603e-05, + "loss": 0.1738, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09327185153961182, + "step": 1705, + "valid_targets_mean": 7360.6, + "valid_targets_min": 6295 + }, + { + "epoch": 3.585519412381952, + "grad_norm": 0.4217529857423982, + "learning_rate": 2.2658023066042815e-05, + "loss": 0.1658, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07933661341667175, + "step": 1710, + "valid_targets_mean": 5988.6, + "valid_targets_min": 4134 + }, + { + "epoch": 3.5960125918153203, + "grad_norm": 0.38201380244393573, + "learning_rate": 2.2554369103141566e-05, + "loss": 0.1722, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07573413103818893, + "step": 1715, + "valid_targets_mean": 6481.5, + "valid_targets_min": 3702 + }, + { + "epoch": 3.6065057712486883, + "grad_norm": 0.3926210592419528, + "learning_rate": 2.2450645343769285e-05, + "loss": 0.1661, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08732487261295319, + "step": 1720, + "valid_targets_mean": 6903.0, + "valid_targets_min": 1571 + }, + { + "epoch": 3.616998950682057, + "grad_norm": 0.41111662103005986, + "learning_rate": 2.2346854622110082e-05, + "loss": 0.1638, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0891902893781662, + "step": 1725, + "valid_targets_mean": 7023.1, + "valid_targets_min": 4203 + }, + { + "epoch": 3.627492130115425, + "grad_norm": 0.3878286095130208, + "learning_rate": 2.224299977417779e-05, + "loss": 0.168, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08115678280591965, + "step": 1730, + "valid_targets_mean": 7307.6, + "valid_targets_min": 3800 + }, + { + "epoch": 3.6379853095487933, + "grad_norm": 0.4053109928124687, + "learning_rate": 2.2139083637738455e-05, + "loss": 0.1633, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08877220749855042, + "step": 1735, + "valid_targets_mean": 7286.0, + "valid_targets_min": 5917 + }, + { + "epoch": 3.6484784889821618, + "grad_norm": 0.4026062943546612, + "learning_rate": 2.2035109052232768e-05, + "loss": 0.1647, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08257147669792175, + "step": 1740, + "valid_targets_mean": 6384.8, + "valid_targets_min": 3720 + }, + { + "epoch": 3.65897166841553, + "grad_norm": 0.4073735691254093, + "learning_rate": 2.1931078858698526e-05, + "loss": 0.1679, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09360416233539581, + "step": 1745, + "valid_targets_mean": 7319.1, + "valid_targets_min": 4178 + }, + { + "epoch": 3.6694648478488983, + "grad_norm": 0.3758234452327913, + "learning_rate": 2.1826995899692968e-05, + "loss": 0.1571, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07552816718816757, + "step": 1750, + "valid_targets_mean": 6783.5, + "valid_targets_min": 4587 + }, + { + "epoch": 3.6799580272822663, + "grad_norm": 0.3625689887881155, + "learning_rate": 2.1722863019215122e-05, + "loss": 0.1715, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06475590169429779, + "step": 1755, + "valid_targets_mean": 6420.6, + "valid_targets_min": 4370 + }, + { + "epoch": 3.6904512067156348, + "grad_norm": 0.3814821074057992, + "learning_rate": 2.1618683062628068e-05, + "loss": 0.1704, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07419635355472565, + "step": 1760, + "valid_targets_mean": 6360.2, + "valid_targets_min": 880 + }, + { + "epoch": 3.7009443861490032, + "grad_norm": 0.4224182295340642, + "learning_rate": 2.151445887658125e-05, + "loss": 0.1643, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07284858822822571, + "step": 1765, + "valid_targets_mean": 6183.0, + "valid_targets_min": 4336 + }, + { + "epoch": 3.7114375655823713, + "grad_norm": 0.39451651845722235, + "learning_rate": 2.1410193308932602e-05, + "loss": 0.169, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06923623383045197, + "step": 1770, + "valid_targets_mean": 5225.6, + "valid_targets_min": 684 + }, + { + "epoch": 3.7219307450157397, + "grad_norm": 0.38030426750431207, + "learning_rate": 2.1305889208670827e-05, + "loss": 0.1715, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07710374891757965, + "step": 1775, + "valid_targets_mean": 7815.2, + "valid_targets_min": 5443 + }, + { + "epoch": 3.732423924449108, + "grad_norm": 0.3980747472714809, + "learning_rate": 2.120154942583748e-05, + "loss": 0.1595, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.054851315915584564, + "step": 1780, + "valid_targets_mean": 4349.1, + "valid_targets_min": 1336 + }, + { + "epoch": 3.7429171038824762, + "grad_norm": 0.40078386595591353, + "learning_rate": 2.1097176811449127e-05, + "loss": 0.1606, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06544072180986404, + "step": 1785, + "valid_targets_mean": 5950.1, + "valid_targets_min": 2615 + }, + { + "epoch": 3.7534102833158447, + "grad_norm": 0.3846972910569676, + "learning_rate": 2.0992774217419445e-05, + "loss": 0.1681, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08092211931943893, + "step": 1790, + "valid_targets_mean": 6076.6, + "valid_targets_min": 2290 + }, + { + "epoch": 3.763903462749213, + "grad_norm": 0.420224973573116, + "learning_rate": 2.088834449648127e-05, + "loss": 0.1633, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08635681867599487, + "step": 1795, + "valid_targets_mean": 6068.2, + "valid_targets_min": 1075 + }, + { + "epoch": 3.774396642182581, + "grad_norm": 0.4671569721203918, + "learning_rate": 2.0783890502108665e-05, + "loss": 0.1695, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07879254221916199, + "step": 1800, + "valid_targets_mean": 5715.6, + "valid_targets_min": 2996 + }, + { + "epoch": 3.7848898216159497, + "grad_norm": 0.3907378517294046, + "learning_rate": 2.067941508843896e-05, + "loss": 0.1656, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08353912830352783, + "step": 1805, + "valid_targets_mean": 7182.6, + "valid_targets_min": 3367 + }, + { + "epoch": 3.795383001049318, + "grad_norm": 0.4049460668959915, + "learning_rate": 2.057492111019474e-05, + "loss": 0.1656, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0975283533334732, + "step": 1810, + "valid_targets_mean": 7144.6, + "valid_targets_min": 1914 + }, + { + "epoch": 3.805876180482686, + "grad_norm": 0.3883882723408151, + "learning_rate": 2.0470411422605866e-05, + "loss": 0.1543, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07384220510721207, + "step": 1815, + "valid_targets_mean": 6721.4, + "valid_targets_min": 4313 + }, + { + "epoch": 3.8163693599160546, + "grad_norm": 0.4282085359535585, + "learning_rate": 2.0365888881331437e-05, + "loss": 0.1643, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0948980301618576, + "step": 1820, + "valid_targets_mean": 6577.1, + "valid_targets_min": 1962 + }, + { + "epoch": 3.826862539349423, + "grad_norm": 0.38348459896882664, + "learning_rate": 2.0261356342381785e-05, + "loss": 0.1624, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06430776417255402, + "step": 1825, + "valid_targets_mean": 6165.0, + "valid_targets_min": 3682 + }, + { + "epoch": 3.837355718782791, + "grad_norm": 0.41418819636764725, + "learning_rate": 2.01568166620404e-05, + "loss": 0.1631, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06571393460035324, + "step": 1830, + "valid_targets_mean": 5519.6, + "valid_targets_min": 3957 + }, + { + "epoch": 3.8478488982161596, + "grad_norm": 0.5029360981906253, + "learning_rate": 2.0052272696785926e-05, + "loss": 0.1643, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10120285302400589, + "step": 1835, + "valid_targets_mean": 8763.6, + "valid_targets_min": 5460 + }, + { + "epoch": 3.858342077649528, + "grad_norm": 0.42210611894229516, + "learning_rate": 1.9947727303214077e-05, + "loss": 0.1636, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.083087258040905, + "step": 1840, + "valid_targets_mean": 6512.6, + "valid_targets_min": 3834 + }, + { + "epoch": 3.868835257082896, + "grad_norm": 0.39079013325089157, + "learning_rate": 1.9843183337959607e-05, + "loss": 0.1662, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08630498498678207, + "step": 1845, + "valid_targets_mean": 7534.5, + "valid_targets_min": 3235 + }, + { + "epoch": 3.8793284365162646, + "grad_norm": 0.38262125163711497, + "learning_rate": 1.973864365761822e-05, + "loss": 0.1631, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0831310823559761, + "step": 1850, + "valid_targets_mean": 7668.1, + "valid_targets_min": 4589 + }, + { + "epoch": 3.8898216159496326, + "grad_norm": 0.38445316596373974, + "learning_rate": 1.9634111118668566e-05, + "loss": 0.1714, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08890377730131149, + "step": 1855, + "valid_targets_mean": 7810.9, + "valid_targets_min": 4969 + }, + { + "epoch": 3.900314795383001, + "grad_norm": 0.40207321193677115, + "learning_rate": 1.9529588577394138e-05, + "loss": 0.1618, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07207867503166199, + "step": 1860, + "valid_targets_mean": 5959.2, + "valid_targets_min": 3426 + }, + { + "epoch": 3.910807974816369, + "grad_norm": 0.4016860319845282, + "learning_rate": 1.9425078889805263e-05, + "loss": 0.1661, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09732590615749359, + "step": 1865, + "valid_targets_mean": 8373.6, + "valid_targets_min": 4059 + }, + { + "epoch": 3.9213011542497376, + "grad_norm": 0.43373670485932747, + "learning_rate": 1.9320584911561048e-05, + "loss": 0.171, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0957561507821083, + "step": 1870, + "valid_targets_mean": 6685.6, + "valid_targets_min": 1358 + }, + { + "epoch": 3.931794333683106, + "grad_norm": 0.42146077002197013, + "learning_rate": 1.921610949789134e-05, + "loss": 0.1607, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06969386339187622, + "step": 1875, + "valid_targets_mean": 6524.0, + "valid_targets_min": 4300 + }, + { + "epoch": 3.942287513116474, + "grad_norm": 0.4400353031601922, + "learning_rate": 1.911165550351874e-05, + "loss": 0.1686, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08482322096824646, + "step": 1880, + "valid_targets_mean": 5664.2, + "valid_targets_min": 512 + }, + { + "epoch": 3.9527806925498425, + "grad_norm": 0.4064192167987498, + "learning_rate": 1.9007225782580562e-05, + "loss": 0.1666, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07865047454833984, + "step": 1885, + "valid_targets_mean": 6331.1, + "valid_targets_min": 3235 + }, + { + "epoch": 3.963273871983211, + "grad_norm": 0.7621316486026284, + "learning_rate": 1.8902823188550877e-05, + "loss": 0.1613, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07833646237850189, + "step": 1890, + "valid_targets_mean": 6456.4, + "valid_targets_min": 2231 + }, + { + "epoch": 3.973767051416579, + "grad_norm": 0.437370240086765, + "learning_rate": 1.8798450574162526e-05, + "loss": 0.1686, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10028866678476334, + "step": 1895, + "valid_targets_mean": 6577.6, + "valid_targets_min": 3902 + }, + { + "epoch": 3.9842602308499475, + "grad_norm": 0.39310572077044664, + "learning_rate": 1.869411079132918e-05, + "loss": 0.1664, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08249735087156296, + "step": 1900, + "valid_targets_mean": 7470.2, + "valid_targets_min": 4211 + }, + { + "epoch": 3.994753410283316, + "grad_norm": 0.36779840500167005, + "learning_rate": 1.85898066910674e-05, + "loss": 0.1651, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09547023475170135, + "step": 1905, + "valid_targets_mean": 8968.4, + "valid_targets_min": 5719 + }, + { + "epoch": 4.004197271773347, + "grad_norm": 0.39149366548919723, + "learning_rate": 1.848554112341876e-05, + "loss": 0.1653, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07257845252752304, + "step": 1910, + "valid_targets_mean": 6606.9, + "valid_targets_min": 4689 + }, + { + "epoch": 4.014690451206715, + "grad_norm": 0.45263145023416623, + "learning_rate": 1.8381316937371935e-05, + "loss": 0.1595, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0793987587094307, + "step": 1915, + "valid_targets_mean": 6458.4, + "valid_targets_min": 1781 + }, + { + "epoch": 4.025183630640084, + "grad_norm": 0.4365833658424274, + "learning_rate": 1.8277136980784888e-05, + "loss": 0.1521, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0758817046880722, + "step": 1920, + "valid_targets_mean": 6823.9, + "valid_targets_min": 1914 + }, + { + "epoch": 4.035676810073452, + "grad_norm": 0.42095370295373735, + "learning_rate": 1.817300410030704e-05, + "loss": 0.1588, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07962150871753693, + "step": 1925, + "valid_targets_mean": 6994.0, + "valid_targets_min": 5131 + }, + { + "epoch": 4.04616998950682, + "grad_norm": 0.41896736271959933, + "learning_rate": 1.806892114130148e-05, + "loss": 0.156, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06516960263252258, + "step": 1930, + "valid_targets_mean": 5917.6, + "valid_targets_min": 3201 + }, + { + "epoch": 4.056663168940189, + "grad_norm": 0.4255485471635825, + "learning_rate": 1.7964890947767242e-05, + "loss": 0.1587, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07960453629493713, + "step": 1935, + "valid_targets_mean": 7058.5, + "valid_targets_min": 630 + }, + { + "epoch": 4.067156348373557, + "grad_norm": 1.0067728543725611, + "learning_rate": 1.786091636226156e-05, + "loss": 0.1563, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09122170507907867, + "step": 1940, + "valid_targets_mean": 7027.4, + "valid_targets_min": 4935 + }, + { + "epoch": 4.077649527806925, + "grad_norm": 0.6131165231428474, + "learning_rate": 1.7757000225822216e-05, + "loss": 0.156, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07546645402908325, + "step": 1945, + "valid_targets_mean": 5387.8, + "valid_targets_min": 474 + }, + { + "epoch": 4.088142707240293, + "grad_norm": 0.44326264246871583, + "learning_rate": 1.765314537788993e-05, + "loss": 0.1571, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08575525134801865, + "step": 1950, + "valid_targets_mean": 6552.4, + "valid_targets_min": 3711 + }, + { + "epoch": 4.098635886673662, + "grad_norm": 0.40925231537302365, + "learning_rate": 1.7549354656230725e-05, + "loss": 0.1565, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08443048596382141, + "step": 1955, + "valid_targets_mean": 7531.9, + "valid_targets_min": 4470 + }, + { + "epoch": 4.10912906610703, + "grad_norm": 0.3994415861798036, + "learning_rate": 1.744563089685844e-05, + "loss": 0.1597, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08722895383834839, + "step": 1960, + "valid_targets_mean": 7725.9, + "valid_targets_min": 4753 + }, + { + "epoch": 4.119622245540398, + "grad_norm": 0.4134717029040427, + "learning_rate": 1.7341976933957195e-05, + "loss": 0.148, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08465741574764252, + "step": 1965, + "valid_targets_mean": 6957.8, + "valid_targets_min": 5434 + }, + { + "epoch": 4.130115424973767, + "grad_norm": 0.42458255750458085, + "learning_rate": 1.723839559980398e-05, + "loss": 0.1556, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07274483144283295, + "step": 1970, + "valid_targets_mean": 6318.1, + "valid_targets_min": 3580 + }, + { + "epoch": 4.140608604407135, + "grad_norm": 0.4179318626296571, + "learning_rate": 1.713488972469124e-05, + "loss": 0.1526, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07811453938484192, + "step": 1975, + "valid_targets_mean": 6550.6, + "valid_targets_min": 4337 + }, + { + "epoch": 4.151101783840503, + "grad_norm": 0.4502510200308509, + "learning_rate": 1.7031462136849557e-05, + "loss": 0.1557, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.05958006903529167, + "step": 1980, + "valid_targets_mean": 5198.2, + "valid_targets_min": 322 + }, + { + "epoch": 4.161594963273872, + "grad_norm": 0.37211337265115246, + "learning_rate": 1.6928115662370372e-05, + "loss": 0.1522, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07018811255693436, + "step": 1985, + "valid_targets_mean": 7271.2, + "valid_targets_min": 5364 + }, + { + "epoch": 4.17208814270724, + "grad_norm": 0.4214163040654945, + "learning_rate": 1.682485312512874e-05, + "loss": 0.1568, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.05935029312968254, + "step": 1990, + "valid_targets_mean": 5156.1, + "valid_targets_min": 3793 + }, + { + "epoch": 4.182581322140608, + "grad_norm": 0.4696750920424235, + "learning_rate": 1.67216773467062e-05, + "loss": 0.1564, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07500176876783371, + "step": 1995, + "valid_targets_mean": 6490.1, + "valid_targets_min": 2699 + }, + { + "epoch": 4.193074501573977, + "grad_norm": 0.42560792488166155, + "learning_rate": 1.661859114631365e-05, + "loss": 0.1611, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07586848735809326, + "step": 2000, + "valid_targets_mean": 7146.2, + "valid_targets_min": 4677 + }, + { + "epoch": 4.203567681007345, + "grad_norm": 0.4133187394249617, + "learning_rate": 1.6515597340714348e-05, + "loss": 0.1534, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07094684988260269, + "step": 2005, + "valid_targets_mean": 6295.6, + "valid_targets_min": 4453 + }, + { + "epoch": 4.214060860440713, + "grad_norm": 0.4447588456309257, + "learning_rate": 1.641269874414689e-05, + "loss": 0.1534, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07924395799636841, + "step": 2010, + "valid_targets_mean": 5545.4, + "valid_targets_min": 2211 + }, + { + "epoch": 4.224554039874082, + "grad_norm": 0.47666844990601065, + "learning_rate": 1.6309898168248383e-05, + "loss": 0.1623, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0635794848203659, + "step": 2015, + "valid_targets_mean": 4879.9, + "valid_targets_min": 1133 + }, + { + "epoch": 4.23504721930745, + "grad_norm": 0.48557243432668556, + "learning_rate": 1.6207198421977558e-05, + "loss": 0.1579, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08019214868545532, + "step": 2020, + "valid_targets_mean": 6683.0, + "valid_targets_min": 3775 + }, + { + "epoch": 4.245540398740818, + "grad_norm": 0.41517469905464643, + "learning_rate": 1.6104602311538058e-05, + "loss": 0.1501, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07885520160198212, + "step": 2025, + "valid_targets_mean": 7664.9, + "valid_targets_min": 1973 + }, + { + "epoch": 4.256033578174187, + "grad_norm": 0.41735951082963135, + "learning_rate": 1.6002112640301732e-05, + "loss": 0.1597, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0588902086019516, + "step": 2030, + "valid_targets_mean": 5111.6, + "valid_targets_min": 3596 + }, + { + "epoch": 4.266526757607555, + "grad_norm": 0.4094259158569839, + "learning_rate": 1.5899732208732066e-05, + "loss": 0.1527, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08648552000522614, + "step": 2035, + "valid_targets_mean": 7056.1, + "valid_targets_min": 4468 + }, + { + "epoch": 4.277019937040923, + "grad_norm": 0.45564364688216985, + "learning_rate": 1.5797463814307617e-05, + "loss": 0.1599, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06630299240350723, + "step": 2040, + "valid_targets_mean": 5274.6, + "valid_targets_min": 3410 + }, + { + "epoch": 4.287513116474292, + "grad_norm": 0.4194056347596156, + "learning_rate": 1.5695310251445623e-05, + "loss": 0.1584, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08637500554323196, + "step": 2045, + "valid_targets_mean": 7606.8, + "valid_targets_min": 4751 + }, + { + "epoch": 4.29800629590766, + "grad_norm": 0.4061322477098895, + "learning_rate": 1.5593274311425606e-05, + "loss": 0.1544, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0862482562661171, + "step": 2050, + "valid_targets_mean": 6504.8, + "valid_targets_min": 3546 + }, + { + "epoch": 4.308499475341028, + "grad_norm": 0.3918891160682465, + "learning_rate": 1.5491358782313136e-05, + "loss": 0.1538, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08805958181619644, + "step": 2055, + "valid_targets_mean": 8237.2, + "valid_targets_min": 5070 + }, + { + "epoch": 4.318992654774396, + "grad_norm": 0.429931806730124, + "learning_rate": 1.5389566448883626e-05, + "loss": 0.1548, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.075901098549366, + "step": 2060, + "valid_targets_mean": 5939.5, + "valid_targets_min": 1092 + }, + { + "epoch": 4.329485834207765, + "grad_norm": 0.38984301257628173, + "learning_rate": 1.5287900092546247e-05, + "loss": 0.1567, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08842956274747849, + "step": 2065, + "valid_targets_mean": 8008.1, + "valid_targets_min": 5168 + }, + { + "epoch": 4.339979013641133, + "grad_norm": 0.418672786924916, + "learning_rate": 1.518636249126793e-05, + "loss": 0.1577, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07579734921455383, + "step": 2070, + "valid_targets_mean": 6765.6, + "valid_targets_min": 1924 + }, + { + "epoch": 4.350472193074501, + "grad_norm": 0.4176023081999143, + "learning_rate": 1.5084956419497455e-05, + "loss": 0.1581, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07979007065296173, + "step": 2075, + "valid_targets_mean": 6787.9, + "valid_targets_min": 4399 + }, + { + "epoch": 4.36096537250787, + "grad_norm": 0.4150017078760464, + "learning_rate": 1.498368464808965e-05, + "loss": 0.1524, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08629097044467926, + "step": 2080, + "valid_targets_mean": 7039.1, + "valid_targets_min": 5428 + }, + { + "epoch": 4.371458551941238, + "grad_norm": 0.41382165898509804, + "learning_rate": 1.4882549944229662e-05, + "loss": 0.154, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08121803402900696, + "step": 2085, + "valid_targets_mean": 6801.4, + "valid_targets_min": 4142 + }, + { + "epoch": 4.381951731374606, + "grad_norm": 0.43026821451292013, + "learning_rate": 1.4781555071357372e-05, + "loss": 0.1577, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07599809765815735, + "step": 2090, + "valid_targets_mean": 6773.5, + "valid_targets_min": 3222 + }, + { + "epoch": 4.392444910807975, + "grad_norm": 0.425262289838467, + "learning_rate": 1.468070278909186e-05, + "loss": 0.1525, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07883192598819733, + "step": 2095, + "valid_targets_mean": 7382.5, + "valid_targets_min": 4144 + }, + { + "epoch": 4.402938090241343, + "grad_norm": 0.3911519649345201, + "learning_rate": 1.4579995853156008e-05, + "loss": 0.1569, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06916022300720215, + "step": 2100, + "valid_targets_mean": 7324.8, + "valid_targets_min": 4876 + }, + { + "epoch": 4.413431269674711, + "grad_norm": 0.42191681431601175, + "learning_rate": 1.4479437015301217e-05, + "loss": 0.1562, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07117653638124466, + "step": 2105, + "valid_targets_mean": 6722.2, + "valid_targets_min": 5426 + }, + { + "epoch": 4.42392444910808, + "grad_norm": 0.4117425742863475, + "learning_rate": 1.4379029023232188e-05, + "loss": 0.1569, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08402608335018158, + "step": 2110, + "valid_targets_mean": 6980.0, + "valid_targets_min": 3753 + }, + { + "epoch": 4.434417628541448, + "grad_norm": 0.4217596176839472, + "learning_rate": 1.4278774620531875e-05, + "loss": 0.1661, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08579067140817642, + "step": 2115, + "valid_targets_mean": 6372.4, + "valid_targets_min": 3272 + }, + { + "epoch": 4.444910807974816, + "grad_norm": 0.42304737141621285, + "learning_rate": 1.4178676546586484e-05, + "loss": 0.1595, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08386508375406265, + "step": 2120, + "valid_targets_mean": 7414.4, + "valid_targets_min": 1424 + }, + { + "epoch": 4.455403987408185, + "grad_norm": 0.38426413931742126, + "learning_rate": 1.407873753651067e-05, + "loss": 0.1564, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0678657814860344, + "step": 2125, + "valid_targets_mean": 7154.8, + "valid_targets_min": 3999 + }, + { + "epoch": 4.465897166841553, + "grad_norm": 0.3928872123959849, + "learning_rate": 1.3978960321072733e-05, + "loss": 0.1533, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07085883617401123, + "step": 2130, + "valid_targets_mean": 6518.4, + "valid_targets_min": 4307 + }, + { + "epoch": 4.476390346274921, + "grad_norm": 0.372810764108029, + "learning_rate": 1.3879347626620066e-05, + "loss": 0.1551, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06526800245046616, + "step": 2135, + "valid_targets_mean": 7368.6, + "valid_targets_min": 5216 + }, + { + "epoch": 4.48688352570829, + "grad_norm": 0.4030823943696792, + "learning_rate": 1.3779902175004634e-05, + "loss": 0.151, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0782521516084671, + "step": 2140, + "valid_targets_mean": 6990.0, + "valid_targets_min": 4030 + }, + { + "epoch": 4.497376705141658, + "grad_norm": 0.36594345319473276, + "learning_rate": 1.3680626683508582e-05, + "loss": 0.1556, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07061035186052322, + "step": 2145, + "valid_targets_mean": 7575.2, + "valid_targets_min": 4969 + }, + { + "epoch": 4.507869884575026, + "grad_norm": 0.4030167809018935, + "learning_rate": 1.3581523864770027e-05, + "loss": 0.1518, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07609860599040985, + "step": 2150, + "valid_targets_mean": 7314.5, + "valid_targets_min": 2186 + }, + { + "epoch": 4.518363064008394, + "grad_norm": 0.35684505880189, + "learning_rate": 1.3482596426708896e-05, + "loss": 0.1589, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06325547397136688, + "step": 2155, + "valid_targets_mean": 6480.9, + "valid_targets_min": 2612 + }, + { + "epoch": 4.528856243441763, + "grad_norm": 0.3604699451460754, + "learning_rate": 1.3383847072452966e-05, + "loss": 0.1539, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07662935554981232, + "step": 2160, + "valid_targets_mean": 8428.0, + "valid_targets_min": 5494 + }, + { + "epoch": 4.539349422875131, + "grad_norm": 0.38300248487288385, + "learning_rate": 1.3285278500263987e-05, + "loss": 0.1518, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07316838204860687, + "step": 2165, + "valid_targets_mean": 6899.1, + "valid_targets_min": 3887 + }, + { + "epoch": 4.5498426023085, + "grad_norm": 0.4471326993175294, + "learning_rate": 1.3186893403463954e-05, + "loss": 0.1563, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07809764891862869, + "step": 2170, + "valid_targets_mean": 7787.6, + "valid_targets_min": 5820 + }, + { + "epoch": 4.560335781741868, + "grad_norm": 0.4363458225581669, + "learning_rate": 1.3088694470361523e-05, + "loss": 0.1572, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07364889979362488, + "step": 2175, + "valid_targets_mean": 6103.8, + "valid_targets_min": 3597 + }, + { + "epoch": 4.570828961175236, + "grad_norm": 0.39294827932171045, + "learning_rate": 1.2990684384178538e-05, + "loss": 0.1595, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06333635747432709, + "step": 2180, + "valid_targets_mean": 5953.6, + "valid_targets_min": 533 + }, + { + "epoch": 4.581322140608604, + "grad_norm": 0.4000662062658891, + "learning_rate": 1.289286582297674e-05, + "loss": 0.16, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08443352580070496, + "step": 2185, + "valid_targets_mean": 8523.2, + "valid_targets_min": 4977 + }, + { + "epoch": 4.591815320041973, + "grad_norm": 0.42455826844378164, + "learning_rate": 1.279524145958455e-05, + "loss": 0.1583, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07433786243200302, + "step": 2190, + "valid_targets_mean": 6190.4, + "valid_targets_min": 1888 + }, + { + "epoch": 4.602308499475341, + "grad_norm": 0.45334280486068473, + "learning_rate": 1.2697813961524088e-05, + "loss": 0.1604, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07792102545499802, + "step": 2195, + "valid_targets_mean": 5404.4, + "valid_targets_min": 278 + }, + { + "epoch": 4.612801678908709, + "grad_norm": 0.4023394849639411, + "learning_rate": 1.260058599093824e-05, + "loss": 0.1551, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0884859561920166, + "step": 2200, + "valid_targets_mean": 7846.1, + "valid_targets_min": 6307 + }, + { + "epoch": 4.623294858342078, + "grad_norm": 0.3824948239360034, + "learning_rate": 1.2503560204517937e-05, + "loss": 0.1563, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07234790921211243, + "step": 2205, + "valid_targets_mean": 7304.8, + "valid_targets_min": 4683 + }, + { + "epoch": 4.633788037775446, + "grad_norm": 0.4568253190089592, + "learning_rate": 1.2406739253429558e-05, + "loss": 0.1557, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0741955116391182, + "step": 2210, + "valid_targets_mean": 5904.6, + "valid_targets_min": 3991 + }, + { + "epoch": 4.644281217208814, + "grad_norm": 0.4175466770704802, + "learning_rate": 1.2310125783242497e-05, + "loss": 0.1485, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08565805852413177, + "step": 2215, + "valid_targets_mean": 6703.1, + "valid_targets_min": 3367 + }, + { + "epoch": 4.654774396642183, + "grad_norm": 0.46439072024818684, + "learning_rate": 1.2213722433856865e-05, + "loss": 0.156, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08224919438362122, + "step": 2220, + "valid_targets_mean": 6804.8, + "valid_targets_min": 4889 + }, + { + "epoch": 4.665267576075551, + "grad_norm": 0.6335320908451477, + "learning_rate": 1.2117531839431349e-05, + "loss": 0.1608, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09963464736938477, + "step": 2225, + "valid_targets_mean": 7838.5, + "valid_targets_min": 5114 + }, + { + "epoch": 4.675760755508919, + "grad_norm": 0.390916402664337, + "learning_rate": 1.2021556628311266e-05, + "loss": 0.1591, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06823018938302994, + "step": 2230, + "valid_targets_mean": 7283.5, + "valid_targets_min": 5060 + }, + { + "epoch": 4.686253934942288, + "grad_norm": 0.39209144452249445, + "learning_rate": 1.19257994229567e-05, + "loss": 0.1543, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08256030827760696, + "step": 2235, + "valid_targets_mean": 7879.4, + "valid_targets_min": 5561 + }, + { + "epoch": 4.696747114375656, + "grad_norm": 0.4063964287117399, + "learning_rate": 1.183026283987089e-05, + "loss": 0.1554, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0840786024928093, + "step": 2240, + "valid_targets_mean": 6486.2, + "valid_targets_min": 4246 + }, + { + "epoch": 4.707240293809024, + "grad_norm": 0.3927177351854844, + "learning_rate": 1.17349494895287e-05, + "loss": 0.1562, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07412371784448624, + "step": 2245, + "valid_targets_mean": 7209.9, + "valid_targets_min": 3392 + }, + { + "epoch": 4.717733473242393, + "grad_norm": 0.42690749987062604, + "learning_rate": 1.1639861976305315e-05, + "loss": 0.1585, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08397528529167175, + "step": 2250, + "valid_targets_mean": 7209.9, + "valid_targets_min": 4104 + }, + { + "epoch": 4.728226652675761, + "grad_norm": 0.4239548924179682, + "learning_rate": 1.1545002898405064e-05, + "loss": 0.1546, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07825271040201187, + "step": 2255, + "valid_targets_mean": 6185.2, + "valid_targets_min": 1042 + }, + { + "epoch": 4.738719832109129, + "grad_norm": 0.41896934312757894, + "learning_rate": 1.1450374847790434e-05, + "loss": 0.1549, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08589155226945877, + "step": 2260, + "valid_targets_mean": 7447.4, + "valid_targets_min": 4713 + }, + { + "epoch": 4.749213011542498, + "grad_norm": 0.38591393562981063, + "learning_rate": 1.1355980410111232e-05, + "loss": 0.1557, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07966472208499908, + "step": 2265, + "valid_targets_mean": 7346.4, + "valid_targets_min": 2754 + }, + { + "epoch": 4.759706190975866, + "grad_norm": 0.433203012628369, + "learning_rate": 1.1261822164633962e-05, + "loss": 0.1567, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06243942305445671, + "step": 2270, + "valid_targets_mean": 5323.8, + "valid_targets_min": 2071 + }, + { + "epoch": 4.770199370409234, + "grad_norm": 0.40847213588420767, + "learning_rate": 1.1167902684171299e-05, + "loss": 0.1546, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07794161140918732, + "step": 2275, + "valid_targets_mean": 7951.9, + "valid_targets_min": 1966 + }, + { + "epoch": 4.780692549842602, + "grad_norm": 0.40464624487396406, + "learning_rate": 1.1074224535011853e-05, + "loss": 0.1538, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06902956962585449, + "step": 2280, + "valid_targets_mean": 5876.9, + "valid_targets_min": 3621 + }, + { + "epoch": 4.791185729275971, + "grad_norm": 0.3468512463698264, + "learning_rate": 1.098079027684999e-05, + "loss": 0.1527, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06479433923959732, + "step": 2285, + "valid_targets_mean": 8605.9, + "valid_targets_min": 5899 + }, + { + "epoch": 4.801678908709339, + "grad_norm": 0.43223261834776333, + "learning_rate": 1.0887602462715931e-05, + "loss": 0.1563, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07753467559814453, + "step": 2290, + "valid_targets_mean": 6797.6, + "valid_targets_min": 4157 + }, + { + "epoch": 4.812172088142708, + "grad_norm": 0.42542333185629694, + "learning_rate": 1.0794663638905974e-05, + "loss": 0.1517, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0873844251036644, + "step": 2295, + "valid_targets_mean": 7011.0, + "valid_targets_min": 5064 + }, + { + "epoch": 4.822665267576076, + "grad_norm": 0.40786720892283435, + "learning_rate": 1.0701976344912886e-05, + "loss": 0.1528, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08020438253879547, + "step": 2300, + "valid_targets_mean": 6572.8, + "valid_targets_min": 3745 + }, + { + "epoch": 4.833158447009444, + "grad_norm": 0.8421600789671039, + "learning_rate": 1.0609543113356585e-05, + "loss": 0.1568, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.081072136759758, + "step": 2305, + "valid_targets_mean": 5681.5, + "valid_targets_min": 1186 + }, + { + "epoch": 4.843651626442812, + "grad_norm": 0.4151228932189121, + "learning_rate": 1.0517366469914883e-05, + "loss": 0.1538, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07828246802091599, + "step": 2310, + "valid_targets_mean": 6720.6, + "valid_targets_min": 2974 + }, + { + "epoch": 4.854144805876181, + "grad_norm": 0.41040554148594127, + "learning_rate": 1.0425448933254496e-05, + "loss": 0.154, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06813888996839523, + "step": 2315, + "valid_targets_mean": 6291.6, + "valid_targets_min": 4436 + }, + { + "epoch": 4.864637985309549, + "grad_norm": 0.40592462502892096, + "learning_rate": 1.0333793014962197e-05, + "loss": 0.1488, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06569113582372665, + "step": 2320, + "valid_targets_mean": 5929.8, + "valid_targets_min": 2319 + }, + { + "epoch": 4.875131164742917, + "grad_norm": 0.472068953550755, + "learning_rate": 1.0242401219476226e-05, + "loss": 0.1537, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07884496450424194, + "step": 2325, + "valid_targets_mean": 6154.5, + "valid_targets_min": 1884 + }, + { + "epoch": 4.885624344176286, + "grad_norm": 0.4368900848207927, + "learning_rate": 1.0151276044017836e-05, + "loss": 0.1656, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.10132862627506256, + "step": 2330, + "valid_targets_mean": 7474.2, + "valid_targets_min": 441 + }, + { + "epoch": 4.896117523609654, + "grad_norm": 0.48069179270538637, + "learning_rate": 1.0060419978523055e-05, + "loss": 0.156, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07669992744922638, + "step": 2335, + "valid_targets_mean": 5272.9, + "valid_targets_min": 250 + }, + { + "epoch": 4.906610703043022, + "grad_norm": 0.4324966099072891, + "learning_rate": 9.969835505574677e-06, + "loss": 0.1558, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07218886911869049, + "step": 2340, + "valid_targets_mean": 5916.1, + "valid_targets_min": 5012 + }, + { + "epoch": 4.917103882476391, + "grad_norm": 0.3558719959865187, + "learning_rate": 9.879525100334365e-06, + "loss": 0.1487, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07323960214853287, + "step": 2345, + "valid_targets_mean": 7195.0, + "valid_targets_min": 4476 + }, + { + "epoch": 4.927597061909759, + "grad_norm": 0.4320909513047001, + "learning_rate": 9.789491230475104e-06, + "loss": 0.1568, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07984884083271027, + "step": 2350, + "valid_targets_mean": 6772.6, + "valid_targets_min": 3408 + }, + { + "epoch": 4.938090241343127, + "grad_norm": 0.41994476431344213, + "learning_rate": 9.699736356113705e-06, + "loss": 0.1622, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08414190262556076, + "step": 2355, + "valid_targets_mean": 6599.5, + "valid_targets_min": 3413 + }, + { + "epoch": 4.948583420776496, + "grad_norm": 0.3884949480159219, + "learning_rate": 9.610262929743625e-06, + "loss": 0.1539, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08472582697868347, + "step": 2360, + "valid_targets_mean": 7285.2, + "valid_targets_min": 3624 + }, + { + "epoch": 4.959076600209864, + "grad_norm": 0.37395378244555055, + "learning_rate": 9.521073396167916e-06, + "loss": 0.1489, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0813298374414444, + "step": 2365, + "valid_targets_mean": 7554.1, + "valid_targets_min": 5973 + }, + { + "epoch": 4.969569779643232, + "grad_norm": 0.47482567588277286, + "learning_rate": 9.432170192432466e-06, + "loss": 0.1617, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08917862176895142, + "step": 2370, + "valid_targets_mean": 7035.6, + "valid_targets_min": 1948 + }, + { + "epoch": 4.980062959076601, + "grad_norm": 0.7574178538162328, + "learning_rate": 9.343555747759376e-06, + "loss": 0.1512, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07299146801233292, + "step": 2375, + "valid_targets_mean": 6641.0, + "valid_targets_min": 3957 + }, + { + "epoch": 4.990556138509969, + "grad_norm": 0.4060234818918656, + "learning_rate": 9.255232483480594e-06, + "loss": 0.1511, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07931849360466003, + "step": 2380, + "valid_targets_mean": 7202.9, + "valid_targets_min": 4695 + }, + { + "epoch": 5.0, + "grad_norm": 0.4990145794174156, + "learning_rate": 9.167202812971767e-06, + "loss": 0.1436, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1183566004037857, + "step": 2385, + "valid_targets_mean": 6717.5, + "valid_targets_min": 3529 + }, + { + "epoch": 5.010493179433368, + "grad_norm": 0.4252274473663138, + "learning_rate": 9.079469141586255e-06, + "loss": 0.1508, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0829816609621048, + "step": 2390, + "valid_targets_mean": 6728.4, + "valid_targets_min": 3902 + }, + { + "epoch": 5.020986358866737, + "grad_norm": 0.41006210637321566, + "learning_rate": 8.99203386658946e-06, + "loss": 0.1448, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0762028619647026, + "step": 2395, + "valid_targets_mean": 7042.4, + "valid_targets_min": 4048 + }, + { + "epoch": 5.031479538300105, + "grad_norm": 0.4367520986727133, + "learning_rate": 8.904899377093289e-06, + "loss": 0.1531, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08191351592540741, + "step": 2400, + "valid_targets_mean": 6554.4, + "valid_targets_min": 3836 + }, + { + "epoch": 5.041972717733473, + "grad_norm": 0.4185830235603568, + "learning_rate": 8.81806805399088e-06, + "loss": 0.1478, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06936196982860565, + "step": 2405, + "valid_targets_mean": 6696.0, + "valid_targets_min": 3801 + }, + { + "epoch": 5.052465897166842, + "grad_norm": 0.42733835717866425, + "learning_rate": 8.73154226989155e-06, + "loss": 0.1474, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06674794107675552, + "step": 2410, + "valid_targets_mean": 6792.2, + "valid_targets_min": 4760 + }, + { + "epoch": 5.06295907660021, + "grad_norm": 0.4331518267811541, + "learning_rate": 8.64532438905596e-06, + "loss": 0.1449, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06603024899959564, + "step": 2415, + "valid_targets_mean": 5464.8, + "valid_targets_min": 2721 + }, + { + "epoch": 5.073452256033578, + "grad_norm": 0.44096615281191126, + "learning_rate": 8.559416767331514e-06, + "loss": 0.1482, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06579803675413132, + "step": 2420, + "valid_targets_mean": 5505.8, + "valid_targets_min": 2139 + }, + { + "epoch": 5.083945435466947, + "grad_norm": 0.3957543488994706, + "learning_rate": 8.473821752087998e-06, + "loss": 0.1495, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07741297781467438, + "step": 2425, + "valid_targets_mean": 7735.9, + "valid_targets_min": 4534 + }, + { + "epoch": 5.094438614900315, + "grad_norm": 0.4456038823158726, + "learning_rate": 8.388541682153403e-06, + "loss": 0.1521, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06927037239074707, + "step": 2430, + "valid_targets_mean": 6136.6, + "valid_targets_min": 2732 + }, + { + "epoch": 5.104931794333683, + "grad_norm": 0.4957862361803205, + "learning_rate": 8.303578887750073e-06, + "loss": 0.1459, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0768423080444336, + "step": 2435, + "valid_targets_mean": 6469.9, + "valid_targets_min": 4497 + }, + { + "epoch": 5.115424973767052, + "grad_norm": 0.46125428697798265, + "learning_rate": 8.218935690430991e-06, + "loss": 0.1489, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06259961426258087, + "step": 2440, + "valid_targets_mean": 5514.5, + "valid_targets_min": 4528 + }, + { + "epoch": 5.12591815320042, + "grad_norm": 0.4173138110643007, + "learning_rate": 8.134614403016364e-06, + "loss": 0.1467, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06864993274211884, + "step": 2445, + "valid_targets_mean": 7222.2, + "valid_targets_min": 3977 + }, + { + "epoch": 5.136411332633788, + "grad_norm": 0.4544283251795367, + "learning_rate": 8.050617329530421e-06, + "loss": 0.156, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06448602676391602, + "step": 2450, + "valid_targets_mean": 5016.2, + "valid_targets_min": 1884 + }, + { + "epoch": 5.146904512067156, + "grad_norm": 0.474458502120218, + "learning_rate": 7.966946765138445e-06, + "loss": 0.1454, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07070116698741913, + "step": 2455, + "valid_targets_mean": 5980.2, + "valid_targets_min": 4097 + }, + { + "epoch": 5.157397691500525, + "grad_norm": 0.40438045362599306, + "learning_rate": 7.883604996084083e-06, + "loss": 0.1484, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06624791026115417, + "step": 2460, + "valid_targets_mean": 6595.4, + "valid_targets_min": 324 + }, + { + "epoch": 5.167890870933893, + "grad_norm": 0.4528672630557165, + "learning_rate": 7.800594299626858e-06, + "loss": 0.1508, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0688638985157013, + "step": 2465, + "valid_targets_mean": 5398.9, + "valid_targets_min": 536 + }, + { + "epoch": 5.178384050367261, + "grad_norm": 0.40732443574012556, + "learning_rate": 7.717916943979964e-06, + "loss": 0.1496, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06999336183071136, + "step": 2470, + "valid_targets_mean": 6783.5, + "valid_targets_min": 5144 + }, + { + "epoch": 5.18887722980063, + "grad_norm": 0.7892658464876883, + "learning_rate": 7.635575188248245e-06, + "loss": 0.1498, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.061632752418518066, + "step": 2475, + "valid_targets_mean": 5858.1, + "valid_targets_min": 1249 + }, + { + "epoch": 5.199370409233998, + "grad_norm": 0.42595727302778713, + "learning_rate": 7.553571282366521e-06, + "loss": 0.1522, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08732441067695618, + "step": 2480, + "valid_targets_mean": 7687.6, + "valid_targets_min": 906 + }, + { + "epoch": 5.209863588667366, + "grad_norm": 0.3652515280893857, + "learning_rate": 7.47190746703808e-06, + "loss": 0.146, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06488369405269623, + "step": 2485, + "valid_targets_mean": 6861.8, + "valid_targets_min": 5043 + }, + { + "epoch": 5.220356768100735, + "grad_norm": 0.45160219072433433, + "learning_rate": 7.390585973673452e-06, + "loss": 0.1433, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07607679069042206, + "step": 2490, + "valid_targets_mean": 6759.8, + "valid_targets_min": 4330 + }, + { + "epoch": 5.230849947534103, + "grad_norm": 0.4454637274316162, + "learning_rate": 7.309609024329452e-06, + "loss": 0.1477, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0715937465429306, + "step": 2495, + "valid_targets_mean": 6435.5, + "valid_targets_min": 689 + }, + { + "epoch": 5.241343126967471, + "grad_norm": 0.46899614521449595, + "learning_rate": 7.2289788316484365e-06, + "loss": 0.1523, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08401565253734589, + "step": 2500, + "valid_targets_mean": 6436.8, + "valid_targets_min": 2807 + }, + { + "epoch": 5.25183630640084, + "grad_norm": 0.4335181996476465, + "learning_rate": 7.14869759879788e-06, + "loss": 0.1504, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08296830952167511, + "step": 2505, + "valid_targets_mean": 7016.9, + "valid_targets_min": 4130 + }, + { + "epoch": 5.262329485834208, + "grad_norm": 0.41481664562862386, + "learning_rate": 7.068767519410149e-06, + "loss": 0.1495, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06817290931940079, + "step": 2510, + "valid_targets_mean": 6680.9, + "valid_targets_min": 3035 + }, + { + "epoch": 5.272822665267576, + "grad_norm": 0.4195671396174188, + "learning_rate": 6.989190777522577e-06, + "loss": 0.1525, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07647854089736938, + "step": 2515, + "valid_targets_mean": 6277.1, + "valid_targets_min": 4030 + }, + { + "epoch": 5.283315844700945, + "grad_norm": 0.47686198024233745, + "learning_rate": 6.9099695475177655e-06, + "loss": 0.1442, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06897547841072083, + "step": 2520, + "valid_targets_mean": 6105.6, + "valid_targets_min": 960 + }, + { + "epoch": 5.293809024134313, + "grad_norm": 0.4229949303276319, + "learning_rate": 6.8311059940641975e-06, + "loss": 0.1463, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07781479507684708, + "step": 2525, + "valid_targets_mean": 6787.8, + "valid_targets_min": 3096 + }, + { + "epoch": 5.304302203567681, + "grad_norm": 0.38859079263316465, + "learning_rate": 6.752602272057081e-06, + "loss": 0.1554, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07891763001680374, + "step": 2530, + "valid_targets_mean": 7881.9, + "valid_targets_min": 5008 + }, + { + "epoch": 5.31479538300105, + "grad_norm": 0.47017762731596824, + "learning_rate": 6.674460526559452e-06, + "loss": 0.1458, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08782370388507843, + "step": 2535, + "valid_targets_mean": 6609.6, + "valid_targets_min": 5168 + }, + { + "epoch": 5.325288562434418, + "grad_norm": 0.36499619258803695, + "learning_rate": 6.596682892743593e-06, + "loss": 0.1401, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06861883401870728, + "step": 2540, + "valid_targets_mean": 9020.2, + "valid_targets_min": 5944 + }, + { + "epoch": 5.335781741867786, + "grad_norm": 0.43205542779884704, + "learning_rate": 6.519271495832642e-06, + "loss": 0.148, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07403698563575745, + "step": 2545, + "valid_targets_mean": 6897.8, + "valid_targets_min": 4598 + }, + { + "epoch": 5.346274921301154, + "grad_norm": 0.4083599914670493, + "learning_rate": 6.44222845104258e-06, + "loss": 0.1515, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06735797226428986, + "step": 2550, + "valid_targets_mean": 6999.2, + "valid_targets_min": 1718 + }, + { + "epoch": 5.356768100734523, + "grad_norm": 0.547169252089301, + "learning_rate": 6.365555863524395e-06, + "loss": 0.1439, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0876869410276413, + "step": 2555, + "valid_targets_mean": 6570.6, + "valid_targets_min": 2444 + }, + { + "epoch": 5.367261280167891, + "grad_norm": 0.43395094479438084, + "learning_rate": 6.289255828306573e-06, + "loss": 0.1443, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08128630369901657, + "step": 2560, + "valid_targets_mean": 7177.0, + "valid_targets_min": 4766 + }, + { + "epoch": 5.37775445960126, + "grad_norm": 0.4513131429020154, + "learning_rate": 6.213330430237856e-06, + "loss": 0.1471, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07187579572200775, + "step": 2565, + "valid_targets_mean": 6668.5, + "valid_targets_min": 1914 + }, + { + "epoch": 5.388247639034628, + "grad_norm": 0.42307586490369853, + "learning_rate": 6.137781743930256e-06, + "loss": 0.1501, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07994256913661957, + "step": 2570, + "valid_targets_mean": 6909.8, + "valid_targets_min": 5506 + }, + { + "epoch": 5.398740818467996, + "grad_norm": 0.43220979824321343, + "learning_rate": 6.062611833702392e-06, + "loss": 0.1506, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07739168405532837, + "step": 2575, + "valid_targets_mean": 6462.4, + "valid_targets_min": 614 + }, + { + "epoch": 5.409233997901364, + "grad_norm": 0.44216176944387225, + "learning_rate": 5.987822753523076e-06, + "loss": 0.1509, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07117493450641632, + "step": 2580, + "valid_targets_mean": 6391.1, + "valid_targets_min": 4288 + }, + { + "epoch": 5.419727177334733, + "grad_norm": 0.46752139351788846, + "learning_rate": 5.9134165469551886e-06, + "loss": 0.1484, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06774599850177765, + "step": 2585, + "valid_targets_mean": 5765.6, + "valid_targets_min": 348 + }, + { + "epoch": 5.430220356768101, + "grad_norm": 0.40911304765551415, + "learning_rate": 5.839395247099819e-06, + "loss": 0.148, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06528888642787933, + "step": 2590, + "valid_targets_mean": 7063.2, + "valid_targets_min": 3657 + }, + { + "epoch": 5.440713536201469, + "grad_norm": 0.40928793920370016, + "learning_rate": 5.765760876540756e-06, + "loss": 0.1537, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07219304144382477, + "step": 2595, + "valid_targets_mean": 6107.6, + "valid_targets_min": 2967 + }, + { + "epoch": 5.451206715634838, + "grad_norm": 0.3781749377835114, + "learning_rate": 5.692515447289184e-06, + "loss": 0.1458, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.05351231247186661, + "step": 2600, + "valid_targets_mean": 6460.2, + "valid_targets_min": 4264 + }, + { + "epoch": 5.461699895068206, + "grad_norm": 0.39237284429273434, + "learning_rate": 5.6196609607287215e-06, + "loss": 0.1437, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06967274844646454, + "step": 2605, + "valid_targets_mean": 7265.5, + "valid_targets_min": 4758 + }, + { + "epoch": 5.472193074501574, + "grad_norm": 0.3675261366398043, + "learning_rate": 5.5471994075607435e-06, + "loss": 0.1493, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06748726963996887, + "step": 2610, + "valid_targets_mean": 7475.9, + "valid_targets_min": 5897 + }, + { + "epoch": 5.482686253934943, + "grad_norm": 0.4363457031650436, + "learning_rate": 5.475132767749951e-06, + "loss": 0.1488, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08594280481338501, + "step": 2615, + "valid_targets_mean": 7848.8, + "valid_targets_min": 6173 + }, + { + "epoch": 5.493179433368311, + "grad_norm": 0.43585421625502074, + "learning_rate": 5.4034630104703204e-06, + "loss": 0.1462, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06766653805971146, + "step": 2620, + "valid_targets_mean": 6035.0, + "valid_targets_min": 648 + }, + { + "epoch": 5.503672612801679, + "grad_norm": 0.43808966640952685, + "learning_rate": 5.332192094051256e-06, + "loss": 0.1466, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07552677392959595, + "step": 2625, + "valid_targets_mean": 5618.8, + "valid_targets_min": 405 + }, + { + "epoch": 5.5141657922350475, + "grad_norm": 0.45626295298872316, + "learning_rate": 5.261321965924114e-06, + "loss": 0.1527, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07340364158153534, + "step": 2630, + "valid_targets_mean": 6599.6, + "valid_targets_min": 3839 + }, + { + "epoch": 5.524658971668416, + "grad_norm": 0.43645448650256147, + "learning_rate": 5.190854562568941e-06, + "loss": 0.1432, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06660012900829315, + "step": 2635, + "valid_targets_mean": 5888.1, + "valid_targets_min": 764 + }, + { + "epoch": 5.535152151101784, + "grad_norm": 0.4166178767903497, + "learning_rate": 5.120791809461621e-06, + "loss": 0.1457, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07351326197385788, + "step": 2640, + "valid_targets_mean": 6478.4, + "valid_targets_min": 4644 + }, + { + "epoch": 5.5456453305351525, + "grad_norm": 0.45659833508858655, + "learning_rate": 5.051135621021217e-06, + "loss": 0.148, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07900229096412659, + "step": 2645, + "valid_targets_mean": 7141.2, + "valid_targets_min": 5010 + }, + { + "epoch": 5.5561385099685205, + "grad_norm": 0.42559819078418587, + "learning_rate": 4.981887900557685e-06, + "loss": 0.141, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06843407452106476, + "step": 2650, + "valid_targets_mean": 6465.2, + "valid_targets_min": 547 + }, + { + "epoch": 5.566631689401889, + "grad_norm": 0.4245545261621145, + "learning_rate": 4.913050540219864e-06, + "loss": 0.1449, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07030477374792099, + "step": 2655, + "valid_targets_mean": 7377.6, + "valid_targets_min": 668 + }, + { + "epoch": 5.5771248688352575, + "grad_norm": 0.4376717496006258, + "learning_rate": 4.844625420943752e-06, + "loss": 0.1397, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07000290602445602, + "step": 2660, + "valid_targets_mean": 5965.5, + "valid_targets_min": 1547 + }, + { + "epoch": 5.5876180482686255, + "grad_norm": 0.43399947447598736, + "learning_rate": 4.776614412401144e-06, + "loss": 0.1466, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08628326654434204, + "step": 2665, + "valid_targets_mean": 7372.4, + "valid_targets_min": 2276 + }, + { + "epoch": 5.5981112277019935, + "grad_norm": 0.4179858762752855, + "learning_rate": 4.709019372948524e-06, + "loss": 0.1482, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06386978924274445, + "step": 2670, + "valid_targets_mean": 5711.0, + "valid_targets_min": 4049 + }, + { + "epoch": 5.608604407135362, + "grad_norm": 0.42844458690580944, + "learning_rate": 4.641842149576303e-06, + "loss": 0.148, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07220377027988434, + "step": 2675, + "valid_targets_mean": 7681.8, + "valid_targets_min": 5904 + }, + { + "epoch": 5.6190975865687305, + "grad_norm": 0.4281055075959464, + "learning_rate": 4.575084577858318e-06, + "loss": 0.1487, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08181042224168777, + "step": 2680, + "valid_targets_mean": 6801.0, + "valid_targets_min": 4406 + }, + { + "epoch": 5.6295907660020985, + "grad_norm": 0.3991086972443232, + "learning_rate": 4.50874848190171e-06, + "loss": 0.1455, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0653497725725174, + "step": 2685, + "valid_targets_mean": 6991.0, + "valid_targets_min": 4089 + }, + { + "epoch": 5.640083945435467, + "grad_norm": 0.41818544630195964, + "learning_rate": 4.4428356742970744e-06, + "loss": 0.1544, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07604652643203735, + "step": 2690, + "valid_targets_mean": 6563.8, + "valid_targets_min": 4816 + }, + { + "epoch": 5.6505771248688355, + "grad_norm": 0.39930401441589847, + "learning_rate": 4.37734795606892e-06, + "loss": 0.1515, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06771788001060486, + "step": 2695, + "valid_targets_mean": 6592.1, + "valid_targets_min": 4253 + }, + { + "epoch": 5.6610703043022035, + "grad_norm": 0.38044348903017194, + "learning_rate": 4.312287116626479e-06, + "loss": 0.1445, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07486169040203094, + "step": 2700, + "valid_targets_mean": 8361.0, + "valid_targets_min": 6841 + }, + { + "epoch": 5.6715634837355715, + "grad_norm": 0.4379406375634741, + "learning_rate": 4.247654933714771e-06, + "loss": 0.1446, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06736306846141815, + "step": 2705, + "valid_targets_mean": 6065.5, + "valid_targets_min": 1098 + }, + { + "epoch": 5.68205666316894, + "grad_norm": 0.42679685672000484, + "learning_rate": 4.1834531733660835e-06, + "loss": 0.1557, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06418700516223907, + "step": 2710, + "valid_targets_mean": 5853.2, + "valid_targets_min": 2580 + }, + { + "epoch": 5.6925498426023085, + "grad_norm": 0.39998357119716993, + "learning_rate": 4.119683589851673e-06, + "loss": 0.1509, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07677870988845825, + "step": 2715, + "valid_targets_mean": 6717.0, + "valid_targets_min": 4929 + }, + { + "epoch": 5.7030430220356765, + "grad_norm": 0.410590433158088, + "learning_rate": 4.056347925633855e-06, + "loss": 0.1521, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06188821420073509, + "step": 2720, + "valid_targets_mean": 6539.4, + "valid_targets_min": 3720 + }, + { + "epoch": 5.713536201469045, + "grad_norm": 0.41589590611443916, + "learning_rate": 3.993447911318378e-06, + "loss": 0.1484, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07664629071950912, + "step": 2725, + "valid_targets_mean": 6923.2, + "valid_targets_min": 3299 + }, + { + "epoch": 5.724029380902413, + "grad_norm": 0.5011737081463074, + "learning_rate": 3.930985265607135e-06, + "loss": 0.1517, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07968798279762268, + "step": 2730, + "valid_targets_mean": 4686.0, + "valid_targets_min": 1042 + }, + { + "epoch": 5.7345225603357814, + "grad_norm": 0.4158568540815471, + "learning_rate": 3.868961695251212e-06, + "loss": 0.1478, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07564239203929901, + "step": 2735, + "valid_targets_mean": 7301.1, + "valid_targets_min": 3746 + }, + { + "epoch": 5.74501573976915, + "grad_norm": 0.4217835213858697, + "learning_rate": 3.8073788950042546e-06, + "loss": 0.1493, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06772969663143158, + "step": 2740, + "valid_targets_mean": 6273.0, + "valid_targets_min": 3823 + }, + { + "epoch": 5.755508919202518, + "grad_norm": 0.38164029011515127, + "learning_rate": 3.7462385475761463e-06, + "loss": 0.1475, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07324448972940445, + "step": 2745, + "valid_targets_mean": 7362.5, + "valid_targets_min": 6030 + }, + { + "epoch": 5.766002098635886, + "grad_norm": 0.4185406006841416, + "learning_rate": 3.6855423235870256e-06, + "loss": 0.1489, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08540073037147522, + "step": 2750, + "valid_targets_mean": 7771.1, + "valid_targets_min": 3222 + }, + { + "epoch": 5.776495278069255, + "grad_norm": 0.4726871114570811, + "learning_rate": 3.625291881521662e-06, + "loss": 0.1511, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09558320045471191, + "step": 2755, + "valid_targets_mean": 7249.9, + "valid_targets_min": 3267 + }, + { + "epoch": 5.786988457502623, + "grad_norm": 0.5712290865771344, + "learning_rate": 3.5654888676841214e-06, + "loss": 0.1464, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06757961958646774, + "step": 2760, + "valid_targets_mean": 5331.9, + "valid_targets_min": 1424 + }, + { + "epoch": 5.797481636935991, + "grad_norm": 0.469600351312921, + "learning_rate": 3.506134916152779e-06, + "loss": 0.1431, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06742135435342789, + "step": 2765, + "valid_targets_mean": 6370.1, + "valid_targets_min": 4422 + }, + { + "epoch": 5.807974816369359, + "grad_norm": 0.5182371548309018, + "learning_rate": 3.4472316487356873e-06, + "loss": 0.152, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0650864839553833, + "step": 2770, + "valid_targets_mean": 6103.9, + "valid_targets_min": 429 + }, + { + "epoch": 5.818467995802728, + "grad_norm": 0.4334214453298438, + "learning_rate": 3.3887806749262285e-06, + "loss": 0.1462, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06549567729234695, + "step": 2775, + "valid_targets_mean": 6187.1, + "valid_targets_min": 3814 + }, + { + "epoch": 5.828961175236096, + "grad_norm": 0.4296434483439814, + "learning_rate": 3.3307835918591703e-06, + "loss": 0.1505, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08607367426156998, + "step": 2780, + "valid_targets_mean": 7481.6, + "valid_targets_min": 5090 + }, + { + "epoch": 5.839454354669465, + "grad_norm": 0.4092845211669697, + "learning_rate": 3.273241984267026e-06, + "loss": 0.1518, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0703398585319519, + "step": 2785, + "valid_targets_mean": 6800.1, + "valid_targets_min": 3413 + }, + { + "epoch": 5.849947534102833, + "grad_norm": 0.43715937651202774, + "learning_rate": 3.2161574244367056e-06, + "loss": 0.1432, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06645885109901428, + "step": 2790, + "valid_targets_mean": 6081.4, + "valid_targets_min": 332 + }, + { + "epoch": 5.860440713536201, + "grad_norm": 0.41748384386378085, + "learning_rate": 3.1595314721666103e-06, + "loss": 0.1523, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08072896301746368, + "step": 2795, + "valid_targets_mean": 6978.6, + "valid_targets_min": 5756 + }, + { + "epoch": 5.870933892969569, + "grad_norm": 0.42897525280854465, + "learning_rate": 3.1033656747239836e-06, + "loss": 0.1457, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06166823208332062, + "step": 2800, + "valid_targets_mean": 5959.6, + "valid_targets_min": 2186 + }, + { + "epoch": 5.881427072402938, + "grad_norm": 0.3943046052087005, + "learning_rate": 3.0476615668026334e-06, + "loss": 0.1434, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06551754474639893, + "step": 2805, + "valid_targets_mean": 6759.5, + "valid_targets_min": 4168 + }, + { + "epoch": 5.891920251836306, + "grad_norm": 0.4320299480191815, + "learning_rate": 2.9924206704810153e-06, + "loss": 0.1455, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08540228009223938, + "step": 2810, + "valid_targets_mean": 7075.1, + "valid_targets_min": 3526 + }, + { + "epoch": 5.902413431269674, + "grad_norm": 0.36447007850719293, + "learning_rate": 2.937644495180607e-06, + "loss": 0.1369, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.048238955438137054, + "step": 2815, + "valid_targets_mean": 7871.5, + "valid_targets_min": 1831 + }, + { + "epoch": 5.912906610703043, + "grad_norm": 0.39285857177498656, + "learning_rate": 2.883334537624707e-06, + "loss": 0.1484, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06805549561977386, + "step": 2820, + "valid_targets_mean": 6736.0, + "valid_targets_min": 4220 + }, + { + "epoch": 5.923399790136411, + "grad_norm": 0.4158701462159782, + "learning_rate": 2.8294922817975125e-06, + "loss": 0.1507, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06890353560447693, + "step": 2825, + "valid_targets_mean": 6997.6, + "valid_targets_min": 4716 + }, + { + "epoch": 5.933892969569779, + "grad_norm": 0.4414516645782903, + "learning_rate": 2.7761191989035753e-06, + "loss": 0.1451, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06870522350072861, + "step": 2830, + "valid_targets_mean": 6177.5, + "valid_targets_min": 217 + }, + { + "epoch": 5.944386149003148, + "grad_norm": 0.40983202777828187, + "learning_rate": 2.723216747327597e-06, + "loss": 0.1496, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06563228368759155, + "step": 2835, + "valid_targets_mean": 6595.6, + "valid_targets_min": 3133 + }, + { + "epoch": 5.954879328436516, + "grad_norm": 0.4246304551727905, + "learning_rate": 2.6707863725945913e-06, + "loss": 0.1487, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06634394824504852, + "step": 2840, + "valid_targets_mean": 5991.1, + "valid_targets_min": 1914 + }, + { + "epoch": 5.965372507869884, + "grad_norm": 0.3945330425962242, + "learning_rate": 2.6188295073303825e-06, + "loss": 0.1413, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07293495535850525, + "step": 2845, + "valid_targets_mean": 7528.6, + "valid_targets_min": 4590 + }, + { + "epoch": 5.975865687303253, + "grad_norm": 0.4390010580207722, + "learning_rate": 2.5673475712224493e-06, + "loss": 0.1487, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07403187453746796, + "step": 2850, + "valid_targets_mean": 6619.8, + "valid_targets_min": 3655 + }, + { + "epoch": 5.986358866736621, + "grad_norm": 0.411795975271099, + "learning_rate": 2.5163419709811487e-06, + "loss": 0.1456, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0659165158867836, + "step": 2855, + "valid_targets_mean": 5921.4, + "valid_targets_min": 3603 + }, + { + "epoch": 5.996852046169989, + "grad_norm": 0.4289574229732313, + "learning_rate": 2.4658141003012626e-06, + "loss": 0.1424, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06474661082029343, + "step": 2860, + "valid_targets_mean": 6295.4, + "valid_targets_min": 3992 + }, + { + "epoch": 6.006295907660021, + "grad_norm": 0.43495368850424004, + "learning_rate": 2.415765339823928e-06, + "loss": 0.1425, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06514771282672882, + "step": 2865, + "valid_targets_mean": 6017.6, + "valid_targets_min": 3674 + }, + { + "epoch": 6.0167890870933896, + "grad_norm": 0.40030858253272283, + "learning_rate": 2.366197057098909e-06, + "loss": 0.1441, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06733272969722748, + "step": 2870, + "valid_targets_mean": 6533.1, + "valid_targets_min": 2435 + }, + { + "epoch": 6.027282266526758, + "grad_norm": 0.4217102705930779, + "learning_rate": 2.317110606547228e-06, + "loss": 0.1452, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07546806335449219, + "step": 2875, + "valid_targets_mean": 6877.2, + "valid_targets_min": 3936 + }, + { + "epoch": 6.037775445960126, + "grad_norm": 0.43230954566902513, + "learning_rate": 2.2685073294241587e-06, + "loss": 0.1433, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06885918974876404, + "step": 2880, + "valid_targets_mean": 6396.2, + "valid_targets_min": 3735 + }, + { + "epoch": 6.0482686253934945, + "grad_norm": 0.414917324754456, + "learning_rate": 2.2203885537825664e-06, + "loss": 0.1432, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08821605145931244, + "step": 2885, + "valid_targets_mean": 7277.2, + "valid_targets_min": 3133 + }, + { + "epoch": 6.0587618048268626, + "grad_norm": 0.4417007657048786, + "learning_rate": 2.172755594436642e-06, + "loss": 0.1487, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07898689806461334, + "step": 2890, + "valid_targets_mean": 7009.6, + "valid_targets_min": 474 + }, + { + "epoch": 6.069254984260231, + "grad_norm": 0.40634795648961697, + "learning_rate": 2.125609752925952e-06, + "loss": 0.1378, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0624513253569603, + "step": 2895, + "valid_targets_mean": 7431.5, + "valid_targets_min": 5792 + }, + { + "epoch": 6.0797481636935995, + "grad_norm": 0.528028727747029, + "learning_rate": 2.078952317479899e-06, + "loss": 0.1416, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0651518702507019, + "step": 2900, + "valid_targets_mean": 7488.6, + "valid_targets_min": 6293 + }, + { + "epoch": 6.0902413431269675, + "grad_norm": 0.390927160736462, + "learning_rate": 2.0327845629824883e-06, + "loss": 0.146, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06875327229499817, + "step": 2905, + "valid_targets_mean": 6601.8, + "valid_targets_min": 4826 + }, + { + "epoch": 6.1007345225603355, + "grad_norm": 0.3788549155713362, + "learning_rate": 1.9871077509375316e-06, + "loss": 0.1414, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06497135758399963, + "step": 2910, + "valid_targets_mean": 7513.4, + "valid_targets_min": 4803 + }, + { + "epoch": 6.1112277019937045, + "grad_norm": 0.4647881988414928, + "learning_rate": 1.9419231294341467e-06, + "loss": 0.1424, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07887689024209976, + "step": 2915, + "valid_targets_mean": 5823.0, + "valid_targets_min": 730 + }, + { + "epoch": 6.1217208814270725, + "grad_norm": 0.4125321204598141, + "learning_rate": 1.8972319331126686e-06, + "loss": 0.1453, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07588649541139603, + "step": 2920, + "valid_targets_mean": 8039.0, + "valid_targets_min": 4456 + }, + { + "epoch": 6.1322140608604405, + "grad_norm": 0.43772759013626494, + "learning_rate": 1.853035383130919e-06, + "loss": 0.1489, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07713251560926437, + "step": 2925, + "valid_targets_mean": 7164.6, + "valid_targets_min": 5243 + }, + { + "epoch": 6.142707240293809, + "grad_norm": 0.40104851335229985, + "learning_rate": 1.8093346871308104e-06, + "loss": 0.1488, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07819400727748871, + "step": 2930, + "valid_targets_mean": 7948.4, + "valid_targets_min": 6004 + }, + { + "epoch": 6.1532004197271775, + "grad_norm": 0.3915333046425395, + "learning_rate": 1.7661310392053877e-06, + "loss": 0.1451, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06114370748400688, + "step": 2935, + "valid_targets_mean": 6627.1, + "valid_targets_min": 790 + }, + { + "epoch": 6.1636935991605455, + "grad_norm": 0.4322222049284193, + "learning_rate": 1.7234256198661748e-06, + "loss": 0.1442, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06770876795053482, + "step": 2940, + "valid_targets_mean": 6536.2, + "valid_targets_min": 320 + }, + { + "epoch": 6.1741867785939135, + "grad_norm": 0.5016528635002883, + "learning_rate": 1.6812195960109257e-06, + "loss": 0.1537, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07365211844444275, + "step": 2945, + "valid_targets_mean": 6059.4, + "valid_targets_min": 3041 + }, + { + "epoch": 6.184679958027282, + "grad_norm": 0.4321875363933191, + "learning_rate": 1.6395141208917297e-06, + "loss": 0.1447, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06925156712532043, + "step": 2950, + "valid_targets_mean": 6382.8, + "valid_targets_min": 2723 + }, + { + "epoch": 6.1951731374606505, + "grad_norm": 0.40614940431445695, + "learning_rate": 1.5983103340835194e-06, + "loss": 0.141, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.05599888414144516, + "step": 2955, + "valid_targets_mean": 6009.4, + "valid_targets_min": 3486 + }, + { + "epoch": 6.2056663168940185, + "grad_norm": 0.415332121975671, + "learning_rate": 1.5576093614529141e-06, + "loss": 0.1487, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08072960376739502, + "step": 2960, + "valid_targets_mean": 7450.6, + "valid_targets_min": 1924 + }, + { + "epoch": 6.216159496327387, + "grad_norm": 0.4461730384235302, + "learning_rate": 1.5174123151274668e-06, + "loss": 0.1424, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07739134877920151, + "step": 2965, + "valid_targets_mean": 7016.2, + "valid_targets_min": 3556 + }, + { + "epoch": 6.226652675760755, + "grad_norm": 0.4807072359918127, + "learning_rate": 1.4777202934652746e-06, + "loss": 0.1447, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06426989287137985, + "step": 2970, + "valid_targets_mean": 5483.0, + "valid_targets_min": 369 + }, + { + "epoch": 6.2371458551941235, + "grad_norm": 0.39446868321521805, + "learning_rate": 1.4385343810249586e-06, + "loss": 0.1432, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07102306932210922, + "step": 2975, + "valid_targets_mean": 8048.2, + "valid_targets_min": 4575 + }, + { + "epoch": 6.247639034627492, + "grad_norm": 0.9024916856028814, + "learning_rate": 1.3998556485360372e-06, + "loss": 0.1476, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07840659469366074, + "step": 2980, + "valid_targets_mean": 7506.6, + "valid_targets_min": 5756 + }, + { + "epoch": 6.25813221406086, + "grad_norm": 0.4232707769311399, + "learning_rate": 1.3616851528696694e-06, + "loss": 0.1428, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06752631813287735, + "step": 2985, + "valid_targets_mean": 6435.6, + "valid_targets_min": 738 + }, + { + "epoch": 6.268625393494228, + "grad_norm": 0.438115625465784, + "learning_rate": 1.3240239370097752e-06, + "loss": 0.1434, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09530520439147949, + "step": 2990, + "valid_targets_mean": 7564.8, + "valid_targets_min": 5332 + }, + { + "epoch": 6.279118572927597, + "grad_norm": 0.41953296625718955, + "learning_rate": 1.2868730300245357e-06, + "loss": 0.143, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06750574707984924, + "step": 2995, + "valid_targets_mean": 6020.0, + "valid_targets_min": 4445 + }, + { + "epoch": 6.289611752360965, + "grad_norm": 0.44332020811355083, + "learning_rate": 1.250233447038267e-06, + "loss": 0.1394, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07387255132198334, + "step": 3000, + "valid_targets_mean": 6266.6, + "valid_targets_min": 3836 + }, + { + "epoch": 6.300104931794333, + "grad_norm": 0.4406327553018184, + "learning_rate": 1.2141061892037009e-06, + "loss": 0.1482, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06112957000732422, + "step": 3005, + "valid_targets_mean": 5626.0, + "valid_targets_min": 871 + }, + { + "epoch": 6.310598111227702, + "grad_norm": 0.46914119959390144, + "learning_rate": 1.1784922436746094e-06, + "loss": 0.1499, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07894495129585266, + "step": 3010, + "valid_targets_mean": 5780.9, + "valid_targets_min": 816 + }, + { + "epoch": 6.32109129066107, + "grad_norm": 0.4229393881315381, + "learning_rate": 1.1433925835788484e-06, + "loss": 0.1425, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06454028189182281, + "step": 3015, + "valid_targets_mean": 6409.5, + "valid_targets_min": 3720 + }, + { + "epoch": 6.331584470094438, + "grad_norm": 0.48313386320671453, + "learning_rate": 1.1088081679917506e-06, + "loss": 0.1456, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0579468309879303, + "step": 3020, + "valid_targets_mean": 4705.8, + "valid_targets_min": 366 + }, + { + "epoch": 6.342077649527807, + "grad_norm": 0.4078934662103234, + "learning_rate": 1.0747399419099347e-06, + "loss": 0.1448, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06732167303562164, + "step": 3025, + "valid_targets_mean": 6300.5, + "valid_targets_min": 3000 + }, + { + "epoch": 6.352570828961175, + "grad_norm": 0.41669125432485954, + "learning_rate": 1.0411888362254775e-06, + "loss": 0.144, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07622279226779938, + "step": 3030, + "valid_targets_mean": 8540.2, + "valid_targets_min": 5870 + }, + { + "epoch": 6.363064008394543, + "grad_norm": 0.45928830101301443, + "learning_rate": 1.008155767700476e-06, + "loss": 0.1455, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07824111729860306, + "step": 3035, + "valid_targets_mean": 7116.4, + "valid_targets_min": 546 + }, + { + "epoch": 6.373557187827912, + "grad_norm": 0.49757185391479913, + "learning_rate": 9.75641638942002e-07, + "loss": 0.1452, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06644590944051743, + "step": 3040, + "valid_targets_mean": 6397.2, + "valid_targets_min": 3801 + }, + { + "epoch": 6.38405036726128, + "grad_norm": 0.4369871259494309, + "learning_rate": 9.436473383774269e-07, + "loss": 0.1437, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.059577181935310364, + "step": 3045, + "valid_targets_mean": 5425.9, + "valid_targets_min": 1001 + }, + { + "epoch": 6.394543546694648, + "grad_norm": 0.41972804842596223, + "learning_rate": 9.121737402301689e-07, + "loss": 0.1426, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0738314837217331, + "step": 3050, + "valid_targets_mean": 6497.6, + "valid_targets_min": 4325 + }, + { + "epoch": 6.405036726128017, + "grad_norm": 0.3919526277157298, + "learning_rate": 8.812217044957827e-07, + "loss": 0.1423, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07838955521583557, + "step": 3055, + "valid_targets_mean": 7497.6, + "valid_targets_min": 5308 + }, + { + "epoch": 6.415529905561385, + "grad_norm": 0.4244281006351969, + "learning_rate": 8.507920769184763e-07, + "loss": 0.1472, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09122315049171448, + "step": 3060, + "valid_targets_mean": 7805.6, + "valid_targets_min": 6840 + }, + { + "epoch": 6.426023084994753, + "grad_norm": 0.421873492937389, + "learning_rate": 8.208856889679895e-07, + "loss": 0.1468, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08135659992694855, + "step": 3065, + "valid_targets_mean": 7197.5, + "valid_targets_min": 4214 + }, + { + "epoch": 6.436516264428121, + "grad_norm": 0.4579684500315175, + "learning_rate": 7.915033578168829e-07, + "loss": 0.1447, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07099463790655136, + "step": 3070, + "valid_targets_mean": 5747.6, + "valid_targets_min": 2162 + }, + { + "epoch": 6.44700944386149, + "grad_norm": 0.4548759253008951, + "learning_rate": 7.626458863182051e-07, + "loss": 0.1414, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06701983511447906, + "step": 3075, + "valid_targets_mean": 5760.2, + "valid_targets_min": 3607 + }, + { + "epoch": 6.457502623294858, + "grad_norm": 0.4348519308171314, + "learning_rate": 7.343140629835632e-07, + "loss": 0.1444, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.062151968479156494, + "step": 3080, + "valid_targets_mean": 5968.0, + "valid_targets_min": 4453 + }, + { + "epoch": 6.467995802728226, + "grad_norm": 0.8065305094990384, + "learning_rate": 7.065086619615624e-07, + "loss": 0.1504, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06456480920314789, + "step": 3085, + "valid_targets_mean": 6700.5, + "valid_targets_min": 3187 + }, + { + "epoch": 6.478488982161595, + "grad_norm": 0.4409812009114144, + "learning_rate": 6.792304430166652e-07, + "loss": 0.1505, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07492567598819733, + "step": 3090, + "valid_targets_mean": 5884.1, + "valid_targets_min": 459 + }, + { + "epoch": 6.488982161594963, + "grad_norm": 0.486028171782539, + "learning_rate": 6.524801515084256e-07, + "loss": 0.1414, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0853385478258133, + "step": 3095, + "valid_targets_mean": 7553.2, + "valid_targets_min": 4761 + }, + { + "epoch": 6.499475341028331, + "grad_norm": 0.5095856838029553, + "learning_rate": 6.262585183711256e-07, + "loss": 0.145, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0921904668211937, + "step": 3100, + "valid_targets_mean": 8242.2, + "valid_targets_min": 6174 + }, + { + "epoch": 6.5099685204617, + "grad_norm": 0.42762008567933446, + "learning_rate": 6.005662600938067e-07, + "loss": 0.1449, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07059694826602936, + "step": 3105, + "valid_targets_mean": 6785.2, + "valid_targets_min": 3863 + }, + { + "epoch": 6.520461699895068, + "grad_norm": 0.39695033631297755, + "learning_rate": 5.754040787006765e-07, + "loss": 0.1393, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08176220208406448, + "step": 3110, + "valid_targets_mean": 8184.5, + "valid_targets_min": 6700 + }, + { + "epoch": 6.530954879328436, + "grad_norm": 0.39799240046919354, + "learning_rate": 5.507726617319442e-07, + "loss": 0.1337, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06416059285402298, + "step": 3115, + "valid_targets_mean": 7240.1, + "valid_targets_min": 4834 + }, + { + "epoch": 6.541448058761805, + "grad_norm": 0.4320783892467951, + "learning_rate": 5.266726822250268e-07, + "loss": 0.1405, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08427301794290543, + "step": 3120, + "valid_targets_mean": 7478.1, + "valid_targets_min": 5891 + }, + { + "epoch": 6.551941238195173, + "grad_norm": 0.4271761177702572, + "learning_rate": 5.031047986961613e-07, + "loss": 0.1424, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06461581587791443, + "step": 3125, + "valid_targets_mean": 6137.1, + "valid_targets_min": 4817 + }, + { + "epoch": 6.562434417628541, + "grad_norm": 0.4283640881002276, + "learning_rate": 4.800696551224038e-07, + "loss": 0.1376, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07027660310268402, + "step": 3130, + "valid_targets_mean": 9156.1, + "valid_targets_min": 5731 + }, + { + "epoch": 6.57292759706191, + "grad_norm": 0.42456445283546934, + "learning_rate": 4.575678809240369e-07, + "loss": 0.1396, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07222200930118561, + "step": 3135, + "valid_targets_mean": 6680.2, + "valid_targets_min": 5320 + }, + { + "epoch": 6.583420776495278, + "grad_norm": 0.39072576679551, + "learning_rate": 4.35600090947379e-07, + "loss": 0.1465, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06995102763175964, + "step": 3140, + "valid_targets_mean": 8072.5, + "valid_targets_min": 4531 + }, + { + "epoch": 6.593913955928646, + "grad_norm": 0.43028177549385166, + "learning_rate": 4.141668854479752e-07, + "loss": 0.1404, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0781894326210022, + "step": 3145, + "valid_targets_mean": 7461.8, + "valid_targets_min": 6054 + }, + { + "epoch": 6.604407135362015, + "grad_norm": 0.447662023713449, + "learning_rate": 3.9326885007419546e-07, + "loss": 0.1419, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0901523008942604, + "step": 3150, + "valid_targets_mean": 7322.8, + "valid_targets_min": 5137 + }, + { + "epoch": 6.614900314795383, + "grad_norm": 0.4614464648188136, + "learning_rate": 3.729065558512401e-07, + "loss": 0.1501, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06360362470149994, + "step": 3155, + "valid_targets_mean": 4600.5, + "valid_targets_min": 2071 + }, + { + "epoch": 6.625393494228751, + "grad_norm": 0.4312790884831707, + "learning_rate": 3.5308055916552395e-07, + "loss": 0.1446, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0776384025812149, + "step": 3160, + "valid_targets_mean": 7624.0, + "valid_targets_min": 5053 + }, + { + "epoch": 6.635886673662119, + "grad_norm": 0.4335013004968067, + "learning_rate": 3.337914017494903e-07, + "loss": 0.1444, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07908298075199127, + "step": 3165, + "valid_targets_mean": 7580.2, + "valid_targets_min": 3504 + }, + { + "epoch": 6.646379853095488, + "grad_norm": 0.43369268441305897, + "learning_rate": 3.1503961066679414e-07, + "loss": 0.1471, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07887262105941772, + "step": 3170, + "valid_targets_mean": 6178.0, + "valid_targets_min": 2852 + }, + { + "epoch": 6.656873032528856, + "grad_norm": 0.44104949668668153, + "learning_rate": 2.968256982979023e-07, + "loss": 0.1415, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06571593135595322, + "step": 3175, + "valid_targets_mean": 5838.2, + "valid_targets_min": 2612 + }, + { + "epoch": 6.667366211962225, + "grad_norm": 0.46934245922459233, + "learning_rate": 2.7915016232610505e-07, + "loss": 0.1426, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08003304898738861, + "step": 3180, + "valid_targets_mean": 6708.9, + "valid_targets_min": 3123 + }, + { + "epoch": 6.677859391395593, + "grad_norm": 0.4266321352323123, + "learning_rate": 2.6201348572389764e-07, + "loss": 0.1487, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07078324258327484, + "step": 3185, + "valid_targets_mean": 6869.1, + "valid_targets_min": 3645 + }, + { + "epoch": 6.688352570828961, + "grad_norm": 0.37731438715374194, + "learning_rate": 2.4541613673980667e-07, + "loss": 0.1433, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06519486755132675, + "step": 3190, + "valid_targets_mean": 7482.2, + "valid_targets_min": 5159 + }, + { + "epoch": 6.698845750262329, + "grad_norm": 0.36126087526973744, + "learning_rate": 2.2935856888557372e-07, + "loss": 0.1424, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0589570589363575, + "step": 3195, + "valid_targets_mean": 8831.9, + "valid_targets_min": 5456 + }, + { + "epoch": 6.709338929695698, + "grad_norm": 0.4382920406465457, + "learning_rate": 2.1384122092377391e-07, + "loss": 0.1435, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07247689366340637, + "step": 3200, + "valid_targets_mean": 6216.6, + "valid_targets_min": 4153 + }, + { + "epoch": 6.719832109129066, + "grad_norm": 0.3951749180119475, + "learning_rate": 1.9886451685582787e-07, + "loss": 0.1387, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.059506893157958984, + "step": 3205, + "valid_targets_mean": 6716.2, + "valid_targets_min": 3627 + }, + { + "epoch": 6.730325288562434, + "grad_norm": 0.4626707779131085, + "learning_rate": 1.8442886591041542e-07, + "loss": 0.1449, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07900651544332504, + "step": 3210, + "valid_targets_mean": 6243.0, + "valid_targets_min": 3938 + }, + { + "epoch": 6.740818467995803, + "grad_norm": 0.5036559734330851, + "learning_rate": 1.7053466253229123e-07, + "loss": 0.1476, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06927962601184845, + "step": 3215, + "valid_targets_mean": 6083.0, + "valid_targets_min": 2570 + }, + { + "epoch": 6.751311647429171, + "grad_norm": 0.40228617771958, + "learning_rate": 1.5718228637150668e-07, + "loss": 0.1441, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07740859687328339, + "step": 3220, + "valid_targets_mean": 7914.0, + "valid_targets_min": 5669 + }, + { + "epoch": 6.761804826862539, + "grad_norm": 0.39473750761015464, + "learning_rate": 1.44372102273036e-07, + "loss": 0.1389, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0655541718006134, + "step": 3225, + "valid_targets_mean": 6408.2, + "valid_targets_min": 554 + }, + { + "epoch": 6.772298006295908, + "grad_norm": 0.45340182016030783, + "learning_rate": 1.3210446026681533e-07, + "loss": 0.1505, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06713958084583282, + "step": 3230, + "valid_targets_mean": 5370.5, + "valid_targets_min": 2763 + }, + { + "epoch": 6.782791185729276, + "grad_norm": 0.4455180349574164, + "learning_rate": 1.2037969555816597e-07, + "loss": 0.1394, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06141781806945801, + "step": 3235, + "valid_targets_mean": 6221.1, + "valid_targets_min": 3181 + }, + { + "epoch": 6.793284365162644, + "grad_norm": 0.4678945238034796, + "learning_rate": 1.0919812851864387e-07, + "loss": 0.1439, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06151239573955536, + "step": 3240, + "valid_targets_mean": 4730.2, + "valid_targets_min": 1966 + }, + { + "epoch": 6.803777544596013, + "grad_norm": 0.4332032913580075, + "learning_rate": 9.856006467727775e-08, + "loss": 0.139, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06429820507764816, + "step": 3245, + "valid_targets_mean": 6662.9, + "valid_targets_min": 504 + }, + { + "epoch": 6.814270724029381, + "grad_norm": 0.46521227540696236, + "learning_rate": 8.84657947122336e-08, + "loss": 0.1375, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06694504618644714, + "step": 3250, + "valid_targets_mean": 6065.1, + "valid_targets_min": 2028 + }, + { + "epoch": 6.824763903462749, + "grad_norm": 0.4132197269531175, + "learning_rate": 7.891559444285435e-08, + "loss": 0.143, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06884525716304779, + "step": 3255, + "valid_targets_mean": 7244.1, + "valid_targets_min": 5319 + }, + { + "epoch": 6.835257082896118, + "grad_norm": 0.42055756340832384, + "learning_rate": 6.990972482214364e-08, + "loss": 0.1425, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06303881853818893, + "step": 3260, + "valid_targets_mean": 5586.9, + "valid_targets_min": 634 + }, + { + "epoch": 6.845750262329486, + "grad_norm": 0.41439812150191874, + "learning_rate": 6.144843192961825e-08, + "loss": 0.1433, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06820205599069595, + "step": 3265, + "valid_targets_mean": 7523.4, + "valid_targets_min": 3904 + }, + { + "epoch": 6.856243441762854, + "grad_norm": 0.4012850799436006, + "learning_rate": 5.3531946964588966e-08, + "loss": 0.1492, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08137405663728714, + "step": 3270, + "valid_targets_mean": 8135.6, + "valid_targets_min": 4717 + }, + { + "epoch": 6.866736621196223, + "grad_norm": 0.4223526810002007, + "learning_rate": 4.616048623985236e-08, + "loss": 0.1413, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07093162834644318, + "step": 3275, + "valid_targets_mean": 7110.4, + "valid_targets_min": 5026 + }, + { + "epoch": 6.877229800629591, + "grad_norm": 0.48289406377020566, + "learning_rate": 3.9334251175764396e-08, + "loss": 0.1411, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09231691807508469, + "step": 3280, + "valid_targets_mean": 7332.9, + "valid_targets_min": 6687 + }, + { + "epoch": 6.887722980062959, + "grad_norm": 0.4156290776101049, + "learning_rate": 3.305342829475589e-08, + "loss": 0.141, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06290020048618317, + "step": 3285, + "valid_targets_mean": 6575.0, + "valid_targets_min": 871 + }, + { + "epoch": 6.898216159496327, + "grad_norm": 0.41062369967843687, + "learning_rate": 2.7318189216212208e-08, + "loss": 0.1428, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0663066953420639, + "step": 3290, + "valid_targets_mean": 6362.2, + "valid_targets_min": 4179 + }, + { + "epoch": 6.908709338929696, + "grad_norm": 0.3952639477099376, + "learning_rate": 2.2128690651808115e-08, + "loss": 0.1419, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08482208102941513, + "step": 3295, + "valid_targets_mean": 7574.1, + "valid_targets_min": 5456 + }, + { + "epoch": 6.919202518363064, + "grad_norm": 0.4229287670153666, + "learning_rate": 1.7485074401206725e-08, + "loss": 0.1415, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06756614148616791, + "step": 3300, + "valid_targets_mean": 6032.6, + "valid_targets_min": 1358 + }, + { + "epoch": 6.929695697796433, + "grad_norm": 0.46558061059610945, + "learning_rate": 1.3387467348195959e-08, + "loss": 0.1427, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.061902180314064026, + "step": 3305, + "valid_targets_mean": 6283.4, + "valid_targets_min": 2580 + }, + { + "epoch": 6.940188877229801, + "grad_norm": 0.4162039582286657, + "learning_rate": 9.835981457222421e-09, + "loss": 0.1452, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.08325707167387009, + "step": 3310, + "valid_targets_mean": 6870.5, + "valid_targets_min": 4757 + }, + { + "epoch": 6.950682056663169, + "grad_norm": 0.38733372724939175, + "learning_rate": 6.830713770318298e-09, + "loss": 0.1412, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06899566948413849, + "step": 3315, + "valid_targets_mean": 7944.4, + "valid_targets_min": 3961 + }, + { + "epoch": 6.961175236096537, + "grad_norm": 0.3902285685010875, + "learning_rate": 4.371746404467914e-09, + "loss": 0.1427, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.06764066964387894, + "step": 3320, + "valid_targets_mean": 7789.8, + "valid_targets_min": 5617 + }, + { + "epoch": 6.971668415529906, + "grad_norm": 0.4545615339265071, + "learning_rate": 2.4591465493495336e-09, + "loss": 0.1407, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.09259332716464996, + "step": 3325, + "valid_targets_mean": 6868.5, + "valid_targets_min": 4701 + }, + { + "epoch": 6.982161594963274, + "grad_norm": 0.47336969792935046, + "learning_rate": 1.092966465510159e-09, + "loss": 0.1453, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.0696336179971695, + "step": 3330, + "valid_targets_mean": 7315.4, + "valid_targets_min": 4030 + }, + { + "epoch": 6.992654774396642, + "grad_norm": 0.44209208534841, + "learning_rate": 2.732434829266772e-10, + "loss": 0.1466, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.07687655836343765, + "step": 3335, + "valid_targets_mean": 6446.9, + "valid_targets_min": 4024 + }, + { + "epoch": 7.0, + "loss_nan_ranks": 0, + "loss_rank_avg": 0.1423269808292389, + "step": 3339, + "total_flos": 2.6890471642266337e+18, + "train_loss": 0.1819671289975788, + "train_runtime": 55607.8638, + "train_samples_per_second": 0.959, + "train_steps_per_second": 0.06, + "valid_targets_mean": 5815.2, + "valid_targets_min": 1781 + } + ], + "logging_steps": 5, + "max_steps": 3339, + "num_input_tokens_seen": 0, + "num_train_epochs": 7, + "save_steps": 1500, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 2.6890471642266337e+18, + "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..176b53f --- /dev/null +++ b/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bb612f056b634ae0e64fff52f03872fbf16575f18da6d9ffc056ab97ac256cd0 +size 8593 diff --git a/training_loss.png b/training_loss.png new file mode 100644 index 0000000..8829dd7 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