diff --git a/.gitattributes b/.gitattributes index a6344aac8c09253b3b630fb776ae94478aa0275b..bcf655a15cd231da13035dfa734da946e54d510c 100644 --- a/.gitattributes +++ b/.gitattributes @@ -33,3 +33,11 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text *.zip filter=lfs diff=lfs merge=lfs -text *.zst filter=lfs diff=lfs merge=lfs -text *tfevents* filter=lfs diff=lfs merge=lfs -text +qwen2.5-1.5b-instruct/checkpoint-142/tokenizer.json filter=lfs diff=lfs merge=lfs -text +qwen2.5-1.5b-instruct/checkpoint-213/tokenizer.json filter=lfs diff=lfs merge=lfs -text +qwen2.5-1.5b-instruct/checkpoint-71/tokenizer.json filter=lfs diff=lfs merge=lfs -text +qwen2.5-1.5b-instruct/tokenizer.json filter=lfs diff=lfs merge=lfs -text +qwen2.5-3b-instruct/checkpoint-142/tokenizer.json filter=lfs diff=lfs merge=lfs -text +qwen2.5-3b-instruct/checkpoint-213/tokenizer.json filter=lfs diff=lfs merge=lfs -text +qwen2.5-3b-instruct/checkpoint-71/tokenizer.json filter=lfs diff=lfs merge=lfs -text +qwen2.5-3b-instruct/tokenizer.json filter=lfs diff=lfs merge=lfs -text diff --git a/qwen2.5-1.5b-instruct/README.md b/qwen2.5-1.5b-instruct/README.md new file mode 100644 index 0000000000000000000000000000000000000000..242cc5bbf973ac13e8abbc608ae69a51c6b03b8f --- /dev/null +++ b/qwen2.5-1.5b-instruct/README.md @@ -0,0 +1,69 @@ +--- +library_name: transformers +license: other +base_model: Qwen/Qwen2.5-1.5B-Instruct +tags: +- llama-factory +- full +- generated_from_trainer +model-index: +- name: qwen2.5-1.5b-instruct_max_grad_norm_5.0 + results: [] +--- + + + +# qwen2.5-1.5b-instruct_max_grad_norm_5.0 + +This model is a fine-tuned version of [Qwen/Qwen2.5-1.5B-Instruct](https://huggingface.co/Qwen/Qwen2.5-1.5B-Instruct) on the healthcare_chatdoctor_train_20k_trace_generation_valid dataset. +It achieves the following results on the evaluation set: +- Loss: 1.1803 + +## 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: 1e-05 +- train_batch_size: 1 +- eval_batch_size: 1 +- seed: 42 +- distributed_type: multi-GPU +- num_devices: 2 +- gradient_accumulation_steps: 128 +- total_train_batch_size: 256 +- total_eval_batch_size: 2 +- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments +- lr_scheduler_type: cosine +- lr_scheduler_warmup_steps: 0.1 +- num_epochs: 3 + +### Training results + +| Training Loss | Epoch | Step | Validation Loss | +|:-------------:|:------:|:----:|:---------------:| +| 1.2639 | 0.7111 | 50 | 1.2486 | +| 1.1955 | 1.4124 | 100 | 1.1989 | +| 1.1315 | 2.1138 | 150 | 1.1811 | +| 1.1373 | 2.8249 | 200 | 1.1768 | + + +### Framework versions + +- Transformers 5.2.0 +- Pytorch 2.11.0+cu128 +- Datasets 4.0.0 +- Tokenizers 0.22.2 diff --git a/qwen2.5-1.5b-instruct/all_results.json b/qwen2.5-1.5b-instruct/all_results.json new file mode 100644 index 0000000000000000000000000000000000000000..db5e8d742c7b4e775056f78b92a0a8c41da79511 --- /dev/null +++ b/qwen2.5-1.5b-instruct/all_results.json @@ -0,0 +1,12 @@ +{ + "epoch": 3.0, + "eval_loss": 1.1803252696990967, + "eval_runtime": 126.8683, + "eval_samples_per_second": 1.576, + "eval_steps_per_second": 0.788, + "total_flos": 4.5153968281196954e+17, + "train_loss": 0.0, + "train_runtime": 4.5863, + "train_samples_per_second": 11774.243, + "train_steps_per_second": 46.443 +} \ No newline at end of file diff --git a/qwen2.5-1.5b-instruct/chat_template.jinja b/qwen2.5-1.5b-instruct/chat_template.jinja new file mode 100644 index 0000000000000000000000000000000000000000..bdf7919a96cfe43d50914a007b9c0877bd0ec27e --- /dev/null +++ b/qwen2.5-1.5b-instruct/chat_template.jinja @@ -0,0 +1,54 @@ +{%- if tools %} + {{- '<|im_start|>system\n' }} + {%- if messages[0]['role'] == 'system' %} + {{- messages[0]['content'] }} + {%- else %} + {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }} + {%- endif %} + {{- "\n\n# 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' }} + {%- else %} + {{- '<|im_start|>system\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\n' }} + {%- endif %} +{%- endif %} +{%- for message in messages %} + {%- if (message.role == "user") or (message.role == "system" and not loop.first) or (message.role == "assistant" and not message.tool_calls) %} + {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }} + {%- elif message.role == "assistant" %} + {{- '<|im_start|>' + message.role }} + {%- if message.content %} + {{- '\n' + message.content }} + {%- endif %} + {%- for tool_call in message.tool_calls %} + {%- if tool_call.function is defined %} + {%- set tool_call = tool_call.function %} + {%- endif %} + {{- '\n\n{"name": "' }} + {{- tool_call.name }} + {{- '", "arguments": ' }} + {{- tool_call.arguments | tojson }} + {{- '}\n' }} + {%- endfor %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "tool" %} + {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != "tool") %} + {{- '<|im_start|>user' }} + {%- endif %} + {{- '\n\n' }} + {{- message.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' }} +{%- endif %} diff --git a/qwen2.5-1.5b-instruct/checkpoint-142/chat_template.jinja b/qwen2.5-1.5b-instruct/checkpoint-142/chat_template.jinja new file mode 100644 index 0000000000000000000000000000000000000000..bdf7919a96cfe43d50914a007b9c0877bd0ec27e --- /dev/null +++ b/qwen2.5-1.5b-instruct/checkpoint-142/chat_template.jinja @@ -0,0 +1,54 @@ +{%- if tools %} + {{- '<|im_start|>system\n' }} + {%- if messages[0]['role'] == 'system' %} + {{- messages[0]['content'] }} + {%- else %} + {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }} + {%- endif %} + {{- "\n\n# 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' }} + {%- else %} + {{- '<|im_start|>system\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\n' }} + {%- endif %} +{%- endif %} +{%- for message in messages %} + {%- if (message.role == "user") or (message.role == "system" and not loop.first) or (message.role == "assistant" and not message.tool_calls) %} + {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }} + {%- elif message.role == "assistant" %} + {{- '<|im_start|>' + message.role }} + {%- if message.content %} + {{- '\n' + message.content }} + {%- endif %} + {%- for tool_call in message.tool_calls %} + {%- if tool_call.function is defined %} + {%- set tool_call = tool_call.function %} + {%- endif %} + {{- '\n\n{"name": "' }} + {{- tool_call.name }} + {{- '", "arguments": ' }} + {{- tool_call.arguments | tojson }} + {{- '}\n' }} + {%- endfor %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "tool" %} + {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != "tool") %} + {{- '<|im_start|>user' }} + {%- endif %} + {{- '\n\n' }} + {{- message.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' }} +{%- endif %} diff --git a/qwen2.5-1.5b-instruct/checkpoint-142/config.json b/qwen2.5-1.5b-instruct/checkpoint-142/config.json new file mode 100644 index 0000000000000000000000000000000000000000..499065fd744a64fc8ad2884da26e85a6e99d468a --- /dev/null +++ b/qwen2.5-1.5b-instruct/checkpoint-142/config.json @@ -0,0 +1,61 @@ +{ + "architectures": [ + "Qwen2ForCausalLM" + ], + "attention_dropout": 0.0, + "bos_token_id": null, + "dtype": "float32", + "eos_token_id": 151645, + "hidden_act": "silu", + "hidden_size": 1536, + "initializer_range": 0.02, + "intermediate_size": 8960, + "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" + ], + "max_position_embeddings": 32768, + "max_window_layers": 21, + "model_type": "qwen2", + "num_attention_heads": 12, + "num_hidden_layers": 28, + "num_key_value_heads": 2, + "pad_token_id": 151643, + "rms_norm_eps": 1e-06, + "rope_parameters": { + "rope_theta": 1000000.0, + "rope_type": "default" + }, + "sliding_window": null, + "tie_word_embeddings": true, + "transformers_version": "5.2.0", + "use_cache": false, + "use_sliding_window": false, + "vocab_size": 151936 +} diff --git a/qwen2.5-1.5b-instruct/checkpoint-142/generation_config.json b/qwen2.5-1.5b-instruct/checkpoint-142/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..c5b4e11411fed3e7a5e6557165f888a7b0b1248b --- /dev/null +++ b/qwen2.5-1.5b-instruct/checkpoint-142/generation_config.json @@ -0,0 +1,13 @@ +{ + "do_sample": true, + "eos_token_id": [ + 151645, + 151643 + ], + "pad_token_id": 151643, + "repetition_penalty": 1.1, + "temperature": 0.7, + "top_k": 20, + "top_p": 0.8, + "transformers_version": "5.2.0" +} diff --git a/qwen2.5-1.5b-instruct/checkpoint-142/model.safetensors b/qwen2.5-1.5b-instruct/checkpoint-142/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..ff4cbbcc4e48c7d8130a0b3afb346b8f1a09c5a4 --- /dev/null +++ b/qwen2.5-1.5b-instruct/checkpoint-142/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d8f0868330bdeff640a003ac39e4267049a3966f47ebdc5491d4b2fc5c0ee90f +size 6174895536 diff --git a/qwen2.5-1.5b-instruct/checkpoint-142/optimizer.pt b/qwen2.5-1.5b-instruct/checkpoint-142/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..57488053ab0a74fba1aa96a17ce8b2d3b1efcee1 --- /dev/null +++ b/qwen2.5-1.5b-instruct/checkpoint-142/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0c2809c124c8a741d0d5b59283ac6a912c7a9fad540e6b5d2a6f5bd65df63dc6 +size 12350013801 diff --git a/qwen2.5-1.5b-instruct/checkpoint-142/rng_state_0.pth b/qwen2.5-1.5b-instruct/checkpoint-142/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..870021e3fa5ac35c2f711adf0c93a556ab4842da --- /dev/null +++ b/qwen2.5-1.5b-instruct/checkpoint-142/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:95e5fc2074c0df31522a514f862c86cb00d71c946a7f15cc9ec0e53a69fb28a7 +size 14917 diff --git a/qwen2.5-1.5b-instruct/checkpoint-142/rng_state_1.pth b/qwen2.5-1.5b-instruct/checkpoint-142/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..21f20da5eb1da017f08aaa88bd19cf24d40e3fbf --- /dev/null +++ b/qwen2.5-1.5b-instruct/checkpoint-142/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0e7153eae67b6c9232a41bc996a2bf5b83229b8c7230d61911ac0fd40e64154e +size 14917 diff --git a/qwen2.5-1.5b-instruct/checkpoint-142/scheduler.pt b/qwen2.5-1.5b-instruct/checkpoint-142/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..2967e7243507a115f9424b1b386bf3657f047c75 --- /dev/null +++ b/qwen2.5-1.5b-instruct/checkpoint-142/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:922deb71175797a96c4bddc5e1cef1c097df38c0dfa52334e97c89e0f573f2bb +size 1465 diff --git a/qwen2.5-1.5b-instruct/checkpoint-142/tokenizer.json b/qwen2.5-1.5b-instruct/checkpoint-142/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..34510ff0037cd50428af467a17ead5a96140a32c --- /dev/null +++ b/qwen2.5-1.5b-instruct/checkpoint-142/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3fd169731d2cbde95e10bf356d66d5997fd885dd8dbb6fb4684da3f23b2585d8 +size 11421892 diff --git a/qwen2.5-1.5b-instruct/checkpoint-142/tokenizer_config.json b/qwen2.5-1.5b-instruct/checkpoint-142/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..145e2c7a68b109ae0253dda29fc2e66cb01b4b85 --- /dev/null +++ b/qwen2.5-1.5b-instruct/checkpoint-142/tokenizer_config.json @@ -0,0 +1,30 @@ +{ + "add_prefix_space": false, + "backend": "tokenizers", + "bos_token": null, + "clean_up_tokenization_spaces": false, + "eos_token": "<|im_end|>", + "errors": "replace", + "extra_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|>" + ], + "is_local": false, + "model_max_length": 131072, + "pad_token": "<|endoftext|>", + "padding_side": "right", + "split_special_tokens": false, + "tokenizer_class": "Qwen2Tokenizer", + "unk_token": null +} diff --git a/qwen2.5-1.5b-instruct/checkpoint-142/trainer_state.json b/qwen2.5-1.5b-instruct/checkpoint-142/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..ca0061650a924dfecfde3c2065bc14ebdcd47cc4 --- /dev/null +++ b/qwen2.5-1.5b-instruct/checkpoint-142/trainer_state.json @@ -0,0 +1,1044 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 2.0, + "eval_steps": 50, + "global_step": 142, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.014222222222222223, + "grad_norm": 0.036850932985544205, + "learning_rate": 0.0, + "loss": 1.6560986042022705, + "step": 1 + }, + { + "epoch": 0.028444444444444446, + "grad_norm": 0.037402428686618805, + "learning_rate": 4.5454545454545457e-07, + "loss": 1.681471347808838, + "step": 2 + }, + { + "epoch": 0.042666666666666665, + "grad_norm": 0.03194897621870041, + "learning_rate": 9.090909090909091e-07, + "loss": 1.6596678495407104, + "step": 3 + }, + { + "epoch": 0.05688888888888889, + "grad_norm": 0.02700614556670189, + "learning_rate": 1.3636363636363636e-06, + "loss": 1.634795904159546, + "step": 4 + }, + { + "epoch": 0.07111111111111111, + "grad_norm": 0.021732285618782043, + "learning_rate": 1.8181818181818183e-06, + "loss": 1.606718897819519, + "step": 5 + }, + { + "epoch": 0.08533333333333333, + "grad_norm": 0.018985262140631676, + "learning_rate": 2.2727272727272728e-06, + "loss": 1.5807676315307617, + "step": 6 + }, + { + "epoch": 0.09955555555555555, + "grad_norm": 0.018287895247340202, + "learning_rate": 2.7272727272727272e-06, + "loss": 1.5490249395370483, + "step": 7 + }, + { + "epoch": 0.11377777777777778, + "grad_norm": 0.017051130533218384, + "learning_rate": 3.181818181818182e-06, + "loss": 1.5317950248718262, + "step": 8 + }, + { + "epoch": 0.128, + "grad_norm": 0.015596743673086166, + "learning_rate": 3.6363636363636366e-06, + "loss": 1.5076513290405273, + "step": 9 + }, + { + "epoch": 0.14222222222222222, + "grad_norm": 0.014579818584024906, + "learning_rate": 4.0909090909090915e-06, + "loss": 1.4786231517791748, + "step": 10 + }, + { + "epoch": 0.15644444444444444, + "grad_norm": 0.012900307774543762, + "learning_rate": 4.5454545454545455e-06, + "loss": 1.469682216644287, + "step": 11 + }, + { + "epoch": 0.17066666666666666, + "grad_norm": 0.011142965406179428, + "learning_rate": 5e-06, + "loss": 1.4563651084899902, + "step": 12 + }, + { + "epoch": 0.18488888888888888, + "grad_norm": 0.011200510896742344, + "learning_rate": 5.4545454545454545e-06, + "loss": 1.4510159492492676, + "step": 13 + }, + { + "epoch": 0.1991111111111111, + "grad_norm": 0.01117027923464775, + "learning_rate": 5.90909090909091e-06, + "loss": 1.440371036529541, + "step": 14 + }, + { + "epoch": 0.21333333333333335, + "grad_norm": 0.010404860600829124, + "learning_rate": 6.363636363636364e-06, + "loss": 1.4149150848388672, + "step": 15 + }, + { + "epoch": 0.22755555555555557, + "grad_norm": 0.009069642052054405, + "learning_rate": 6.818181818181818e-06, + "loss": 1.3951767683029175, + "step": 16 + }, + { + "epoch": 0.24177777777777779, + "grad_norm": 0.008563642390072346, + "learning_rate": 7.272727272727273e-06, + "loss": 1.3866009712219238, + "step": 17 + }, + { + "epoch": 0.256, + "grad_norm": 0.008158586919307709, + "learning_rate": 7.727272727272727e-06, + "loss": 1.3851287364959717, + "step": 18 + }, + { + "epoch": 0.2702222222222222, + "grad_norm": 0.008019132539629936, + "learning_rate": 8.181818181818183e-06, + "loss": 1.3660919666290283, + "step": 19 + }, + { + "epoch": 0.28444444444444444, + "grad_norm": 0.007505647838115692, + "learning_rate": 8.636363636363637e-06, + "loss": 1.3991025686264038, + "step": 20 + }, + { + "epoch": 0.2986666666666667, + "grad_norm": 0.0068238647654652596, + "learning_rate": 9.090909090909091e-06, + "loss": 1.3563741445541382, + "step": 21 + }, + { + "epoch": 0.3128888888888889, + "grad_norm": 0.00681255292147398, + "learning_rate": 9.545454545454547e-06, + "loss": 1.3432800769805908, + "step": 22 + }, + { + "epoch": 0.32711111111111113, + "grad_norm": 0.0064102653414011, + "learning_rate": 1e-05, + "loss": 1.3540418148040771, + "step": 23 + }, + { + "epoch": 0.3413333333333333, + "grad_norm": 0.006286031566560268, + "learning_rate": 9.999323662872998e-06, + "loss": 1.3571672439575195, + "step": 24 + }, + { + "epoch": 0.35555555555555557, + "grad_norm": 0.006193141452968121, + "learning_rate": 9.99729483446475e-06, + "loss": 1.3181281089782715, + "step": 25 + }, + { + "epoch": 0.36977777777777776, + "grad_norm": 0.005919129587709904, + "learning_rate": 9.993914063644053e-06, + "loss": 1.3311502933502197, + "step": 26 + }, + { + "epoch": 0.384, + "grad_norm": 0.005563116166740656, + "learning_rate": 9.989182265027232e-06, + "loss": 1.3282787799835205, + "step": 27 + }, + { + "epoch": 0.3982222222222222, + "grad_norm": 0.005611867178231478, + "learning_rate": 9.98310071873072e-06, + "loss": 1.325502872467041, + "step": 28 + }, + { + "epoch": 0.41244444444444445, + "grad_norm": 0.0053426348604261875, + "learning_rate": 9.975671070024741e-06, + "loss": 1.3453574180603027, + "step": 29 + }, + { + "epoch": 0.4266666666666667, + "grad_norm": 0.0058601596392691135, + "learning_rate": 9.966895328888195e-06, + "loss": 1.3257673978805542, + "step": 30 + }, + { + "epoch": 0.4408888888888889, + "grad_norm": 0.005306159611791372, + "learning_rate": 9.956775869464901e-06, + "loss": 1.317690372467041, + "step": 31 + }, + { + "epoch": 0.45511111111111113, + "grad_norm": 0.0048169931396842, + "learning_rate": 9.945315429421307e-06, + "loss": 1.3243460655212402, + "step": 32 + }, + { + "epoch": 0.4693333333333333, + "grad_norm": 0.0048377541825175285, + "learning_rate": 9.932517109205849e-06, + "loss": 1.3194491863250732, + "step": 33 + }, + { + "epoch": 0.48355555555555557, + "grad_norm": 0.004818843677639961, + "learning_rate": 9.918384371210178e-06, + "loss": 1.3047914505004883, + "step": 34 + }, + { + "epoch": 0.49777777777777776, + "grad_norm": 0.005093662533909082, + "learning_rate": 9.902921038832456e-06, + "loss": 1.2955162525177002, + "step": 35 + }, + { + "epoch": 0.512, + "grad_norm": 0.005024498328566551, + "learning_rate": 9.886131295443003e-06, + "loss": 1.3118610382080078, + "step": 36 + }, + { + "epoch": 0.5262222222222223, + "grad_norm": 0.004711818415671587, + "learning_rate": 9.868019683252543e-06, + "loss": 1.302067518234253, + "step": 37 + }, + { + "epoch": 0.5404444444444444, + "grad_norm": 0.0047568087466061115, + "learning_rate": 9.848591102083375e-06, + "loss": 1.2789689302444458, + "step": 38 + }, + { + "epoch": 0.5546666666666666, + "grad_norm": 0.004525096621364355, + "learning_rate": 9.82785080804381e-06, + "loss": 1.2827372550964355, + "step": 39 + }, + { + "epoch": 0.5688888888888889, + "grad_norm": 0.0044178408570587635, + "learning_rate": 9.805804412106197e-06, + "loss": 1.2863564491271973, + "step": 40 + }, + { + "epoch": 0.5831111111111111, + "grad_norm": 0.00429871492087841, + "learning_rate": 9.782457878588977e-06, + "loss": 1.2768702507019043, + "step": 41 + }, + { + "epoch": 0.5973333333333334, + "grad_norm": 0.004442654084414244, + "learning_rate": 9.75781752354311e-06, + "loss": 1.2826354503631592, + "step": 42 + }, + { + "epoch": 0.6115555555555555, + "grad_norm": 0.004453128203749657, + "learning_rate": 9.731890013043367e-06, + "loss": 1.275444746017456, + "step": 43 + }, + { + "epoch": 0.6257777777777778, + "grad_norm": 0.004429741762578487, + "learning_rate": 9.704682361384941e-06, + "loss": 1.2660303115844727, + "step": 44 + }, + { + "epoch": 0.64, + "grad_norm": 0.004699238575994968, + "learning_rate": 9.676201929185809e-06, + "loss": 1.2627251148223877, + "step": 45 + }, + { + "epoch": 0.6542222222222223, + "grad_norm": 0.004564967937767506, + "learning_rate": 9.646456421395447e-06, + "loss": 1.2888470888137817, + "step": 46 + }, + { + "epoch": 0.6684444444444444, + "grad_norm": 0.004141585901379585, + "learning_rate": 9.615453885210368e-06, + "loss": 1.2826390266418457, + "step": 47 + }, + { + "epoch": 0.6826666666666666, + "grad_norm": 0.004241476766765118, + "learning_rate": 9.583202707897075e-06, + "loss": 1.2571580410003662, + "step": 48 + }, + { + "epoch": 0.6968888888888889, + "grad_norm": 0.004200824070721865, + "learning_rate": 9.549711614523007e-06, + "loss": 1.2616808414459229, + "step": 49 + }, + { + "epoch": 0.7111111111111111, + "grad_norm": 0.004399280063807964, + "learning_rate": 9.514989665596114e-06, + "loss": 1.263854742050171, + "step": 50 + }, + { + "epoch": 0.7111111111111111, + "eval_loss": 1.248613715171814, + "eval_runtime": 54.7705, + "eval_samples_per_second": 3.652, + "eval_steps_per_second": 1.826, + "step": 50 + }, + { + "epoch": 0.7253333333333334, + "grad_norm": 0.004282557871192694, + "learning_rate": 9.479046254613673e-06, + "loss": 1.279876947402954, + "step": 51 + }, + { + "epoch": 0.7395555555555555, + "grad_norm": 0.004135944880545139, + "learning_rate": 9.441891105521005e-06, + "loss": 1.2600014209747314, + "step": 52 + }, + { + "epoch": 0.7537777777777778, + "grad_norm": 0.0039484258741140366, + "learning_rate": 9.40353427008083e-06, + "loss": 1.2702335119247437, + "step": 53 + }, + { + "epoch": 0.768, + "grad_norm": 0.004160050302743912, + "learning_rate": 9.3639861251539e-06, + "loss": 1.2441630363464355, + "step": 54 + }, + { + "epoch": 0.7822222222222223, + "grad_norm": 0.004111913964152336, + "learning_rate": 9.323257369891702e-06, + "loss": 1.2441895008087158, + "step": 55 + }, + { + "epoch": 0.7964444444444444, + "grad_norm": 0.004063262604176998, + "learning_rate": 9.281359022841966e-06, + "loss": 1.255817174911499, + "step": 56 + }, + { + "epoch": 0.8106666666666666, + "grad_norm": 0.003984581213444471, + "learning_rate": 9.238302418967757e-06, + "loss": 1.2536640167236328, + "step": 57 + }, + { + "epoch": 0.8248888888888889, + "grad_norm": 0.003996702842414379, + "learning_rate": 9.194099206580981e-06, + "loss": 1.2499892711639404, + "step": 58 + }, + { + "epoch": 0.8391111111111111, + "grad_norm": 0.0044861058704555035, + "learning_rate": 9.14876134419111e-06, + "loss": 1.248606562614441, + "step": 59 + }, + { + "epoch": 0.8533333333333334, + "grad_norm": 0.004101422615349293, + "learning_rate": 9.102301097269974e-06, + "loss": 1.247856616973877, + "step": 60 + }, + { + "epoch": 0.8675555555555555, + "grad_norm": 0.004009375814348459, + "learning_rate": 9.05473103493355e-06, + "loss": 1.257920742034912, + "step": 61 + }, + { + "epoch": 0.8817777777777778, + "grad_norm": 0.0039364611729979515, + "learning_rate": 9.006064026541549e-06, + "loss": 1.2761049270629883, + "step": 62 + }, + { + "epoch": 0.896, + "grad_norm": 0.004427333828061819, + "learning_rate": 8.956313238215824e-06, + "loss": 1.245169997215271, + "step": 63 + }, + { + "epoch": 0.9102222222222223, + "grad_norm": 0.004076611250638962, + "learning_rate": 8.905492129278478e-06, + "loss": 1.234323263168335, + "step": 64 + }, + { + "epoch": 0.9244444444444444, + "grad_norm": 0.004032611381262541, + "learning_rate": 8.85361444861063e-06, + "loss": 1.2250256538391113, + "step": 65 + }, + { + "epoch": 0.9386666666666666, + "grad_norm": 0.004036127123981714, + "learning_rate": 8.800694230932885e-06, + "loss": 1.2339348793029785, + "step": 66 + }, + { + "epoch": 0.9528888888888889, + "grad_norm": 0.004113379865884781, + "learning_rate": 8.74674579300843e-06, + "loss": 1.252026081085205, + "step": 67 + }, + { + "epoch": 0.9671111111111111, + "grad_norm": 0.004064187407493591, + "learning_rate": 8.691783729769874e-06, + "loss": 1.2473349571228027, + "step": 68 + }, + { + "epoch": 0.9813333333333333, + "grad_norm": 0.0040993476286530495, + "learning_rate": 8.635822910370793e-06, + "loss": 1.2227911949157715, + "step": 69 + }, + { + "epoch": 0.9955555555555555, + "grad_norm": 0.003903420874848962, + "learning_rate": 8.578878474163115e-06, + "loss": 1.2343428134918213, + "step": 70 + }, + { + "epoch": 1.0, + "grad_norm": 0.006894194986671209, + "learning_rate": 8.520965826601394e-06, + "loss": 1.2428948879241943, + "step": 71 + }, + { + "epoch": 1.0142222222222221, + "grad_norm": 0.004313764162361622, + "learning_rate": 8.462100635075097e-06, + "loss": 1.202027440071106, + "step": 72 + }, + { + "epoch": 1.0284444444444445, + "grad_norm": 0.004148328676819801, + "learning_rate": 8.40229882467003e-06, + "loss": 1.2224020957946777, + "step": 73 + }, + { + "epoch": 1.0426666666666666, + "grad_norm": 0.00453830836340785, + "learning_rate": 8.341576573860049e-06, + "loss": 1.2196844816207886, + "step": 74 + }, + { + "epoch": 1.056888888888889, + "grad_norm": 0.004094650503247976, + "learning_rate": 8.279950310130218e-06, + "loss": 1.2222049236297607, + "step": 75 + }, + { + "epoch": 1.0711111111111111, + "grad_norm": 0.004169773776084185, + "learning_rate": 8.2174367055326e-06, + "loss": 1.2127679586410522, + "step": 76 + }, + { + "epoch": 1.0853333333333333, + "grad_norm": 0.004176456015557051, + "learning_rate": 8.154052672175888e-06, + "loss": 1.1962798833847046, + "step": 77 + }, + { + "epoch": 1.0995555555555556, + "grad_norm": 0.004215880297124386, + "learning_rate": 8.08981535765009e-06, + "loss": 1.192876935005188, + "step": 78 + }, + { + "epoch": 1.1137777777777778, + "grad_norm": 0.004112801048904657, + "learning_rate": 8.024742140387506e-06, + "loss": 1.204209566116333, + "step": 79 + }, + { + "epoch": 1.1280000000000001, + "grad_norm": 0.00416502496227622, + "learning_rate": 7.95885062496126e-06, + "loss": 1.2061021327972412, + "step": 80 + }, + { + "epoch": 1.1422222222222222, + "grad_norm": 0.004163650795817375, + "learning_rate": 7.892158637322647e-06, + "loss": 1.2074682712554932, + "step": 81 + }, + { + "epoch": 1.1564444444444444, + "grad_norm": 0.004129140637814999, + "learning_rate": 7.824684219978591e-06, + "loss": 1.1797451972961426, + "step": 82 + }, + { + "epoch": 1.1706666666666667, + "grad_norm": 0.004320825915783644, + "learning_rate": 7.756445627110523e-06, + "loss": 1.2151604890823364, + "step": 83 + }, + { + "epoch": 1.1848888888888889, + "grad_norm": 0.004057695157825947, + "learning_rate": 7.68746131963598e-06, + "loss": 1.1917520761489868, + "step": 84 + }, + { + "epoch": 1.199111111111111, + "grad_norm": 0.004045524168759584, + "learning_rate": 7.6177499602143e-06, + "loss": 1.1966304779052734, + "step": 85 + }, + { + "epoch": 1.2133333333333334, + "grad_norm": 0.004168134648352861, + "learning_rate": 7.547330408197695e-06, + "loss": 1.2256059646606445, + "step": 86 + }, + { + "epoch": 1.2275555555555555, + "grad_norm": 0.0040689390152692795, + "learning_rate": 7.476221714529167e-06, + "loss": 1.1984763145446777, + "step": 87 + }, + { + "epoch": 1.2417777777777779, + "grad_norm": 0.004095707554370165, + "learning_rate": 7.404443116588548e-06, + "loss": 1.1815118789672852, + "step": 88 + }, + { + "epoch": 1.256, + "grad_norm": 0.004128715489059687, + "learning_rate": 7.332014032988123e-06, + "loss": 1.1813948154449463, + "step": 89 + }, + { + "epoch": 1.2702222222222221, + "grad_norm": 0.004037311766296625, + "learning_rate": 7.2589540583192165e-06, + "loss": 1.2083604335784912, + "step": 90 + }, + { + "epoch": 1.2844444444444445, + "grad_norm": 0.004094353877007961, + "learning_rate": 7.185282957851175e-06, + "loss": 1.2008360624313354, + "step": 91 + }, + { + "epoch": 1.2986666666666666, + "grad_norm": 0.004135950934141874, + "learning_rate": 7.111020662184174e-06, + "loss": 1.1902484893798828, + "step": 92 + }, + { + "epoch": 1.3128888888888888, + "grad_norm": 0.003980172332376242, + "learning_rate": 7.036187261857289e-06, + "loss": 1.1865317821502686, + "step": 93 + }, + { + "epoch": 1.3271111111111111, + "grad_norm": 0.003914009779691696, + "learning_rate": 6.960803001913315e-06, + "loss": 1.1860058307647705, + "step": 94 + }, + { + "epoch": 1.3413333333333333, + "grad_norm": 0.003945492208003998, + "learning_rate": 6.884888276421766e-06, + "loss": 1.1842578649520874, + "step": 95 + }, + { + "epoch": 1.3555555555555556, + "grad_norm": 0.0038985933642834425, + "learning_rate": 6.8084636229615786e-06, + "loss": 1.1814080476760864, + "step": 96 + }, + { + "epoch": 1.3697777777777778, + "grad_norm": 0.003922275733202696, + "learning_rate": 6.731549717064975e-06, + "loss": 1.193676233291626, + "step": 97 + }, + { + "epoch": 1.384, + "grad_norm": 0.004165121354162693, + "learning_rate": 6.654167366624009e-06, + "loss": 1.2141278982162476, + "step": 98 + }, + { + "epoch": 1.3982222222222223, + "grad_norm": 0.004156369715929031, + "learning_rate": 6.576337506261314e-06, + "loss": 1.1865811347961426, + "step": 99 + }, + { + "epoch": 1.4124444444444444, + "grad_norm": 0.00419853487983346, + "learning_rate": 6.498081191666549e-06, + "loss": 1.1954832077026367, + "step": 100 + }, + { + "epoch": 1.4124444444444444, + "eval_loss": 1.1988927125930786, + "eval_runtime": 11.4257, + "eval_samples_per_second": 17.504, + "eval_steps_per_second": 8.752, + "step": 100 + }, + { + "epoch": 1.4266666666666667, + "grad_norm": 0.003968748729676008, + "learning_rate": 6.419419593900109e-06, + "loss": 1.1804523468017578, + "step": 101 + }, + { + "epoch": 1.4408888888888889, + "grad_norm": 0.003949783742427826, + "learning_rate": 6.340373993665607e-06, + "loss": 1.2068874835968018, + "step": 102 + }, + { + "epoch": 1.455111111111111, + "grad_norm": 0.004127713851630688, + "learning_rate": 6.2609657755527135e-06, + "loss": 1.1804969310760498, + "step": 103 + }, + { + "epoch": 1.4693333333333334, + "grad_norm": 0.004182465374469757, + "learning_rate": 6.1812164222518626e-06, + "loss": 1.1953659057617188, + "step": 104 + }, + { + "epoch": 1.4835555555555555, + "grad_norm": 0.004224156495183706, + "learning_rate": 6.101147508742456e-06, + "loss": 1.1933740377426147, + "step": 105 + }, + { + "epoch": 1.4977777777777779, + "grad_norm": 0.0041472818702459335, + "learning_rate": 6.020780696456059e-06, + "loss": 1.201953411102295, + "step": 106 + }, + { + "epoch": 1.512, + "grad_norm": 0.0038781315088272095, + "learning_rate": 5.940137727416247e-06, + "loss": 1.1897786855697632, + "step": 107 + }, + { + "epoch": 1.5262222222222221, + "grad_norm": 0.004033711273223162, + "learning_rate": 5.859240418356614e-06, + "loss": 1.1798899173736572, + "step": 108 + }, + { + "epoch": 1.5404444444444443, + "grad_norm": 0.00406253058463335, + "learning_rate": 5.778110654818602e-06, + "loss": 1.1853947639465332, + "step": 109 + }, + { + "epoch": 1.5546666666666666, + "grad_norm": 0.004025817848742008, + "learning_rate": 5.696770385230679e-06, + "loss": 1.1760761737823486, + "step": 110 + }, + { + "epoch": 1.568888888888889, + "grad_norm": 0.004029016941785812, + "learning_rate": 5.615241614970546e-06, + "loss": 1.1980795860290527, + "step": 111 + }, + { + "epoch": 1.5831111111111111, + "grad_norm": 0.004079029895365238, + "learning_rate": 5.533546400411899e-06, + "loss": 1.1878252029418945, + "step": 112 + }, + { + "epoch": 1.5973333333333333, + "grad_norm": 0.004004180897027254, + "learning_rate": 5.4517068429574215e-06, + "loss": 1.1891889572143555, + "step": 113 + }, + { + "epoch": 1.6115555555555554, + "grad_norm": 0.004044163506478071, + "learning_rate": 5.3697450830595775e-06, + "loss": 1.1935696601867676, + "step": 114 + }, + { + "epoch": 1.6257777777777778, + "grad_norm": 0.0040819537825882435, + "learning_rate": 5.287683294230855e-06, + "loss": 1.1648902893066406, + "step": 115 + }, + { + "epoch": 1.6400000000000001, + "grad_norm": 0.0042525348253548145, + "learning_rate": 5.20554367704505e-06, + "loss": 1.1749862432479858, + "step": 116 + }, + { + "epoch": 1.6542222222222223, + "grad_norm": 0.004129970911890268, + "learning_rate": 5.123348453131242e-06, + "loss": 1.200908899307251, + "step": 117 + }, + { + "epoch": 1.6684444444444444, + "grad_norm": 0.004072704818099737, + "learning_rate": 5.041119859162068e-06, + "loss": 1.189719796180725, + "step": 118 + }, + { + "epoch": 1.6826666666666665, + "grad_norm": 0.004113580100238323, + "learning_rate": 4.958880140837934e-06, + "loss": 1.1798235177993774, + "step": 119 + }, + { + "epoch": 1.696888888888889, + "grad_norm": 0.003983311355113983, + "learning_rate": 4.876651546868759e-06, + "loss": 1.166198492050171, + "step": 120 + }, + { + "epoch": 1.7111111111111112, + "grad_norm": 0.003966879565268755, + "learning_rate": 4.794456322954953e-06, + "loss": 1.1802079677581787, + "step": 121 + }, + { + "epoch": 1.7253333333333334, + "grad_norm": 0.004070781171321869, + "learning_rate": 4.7123167057691446e-06, + "loss": 1.1807469129562378, + "step": 122 + }, + { + "epoch": 1.7395555555555555, + "grad_norm": 0.0039869495667517185, + "learning_rate": 4.630254916940424e-06, + "loss": 1.1979055404663086, + "step": 123 + }, + { + "epoch": 1.7537777777777777, + "grad_norm": 0.004057880956679583, + "learning_rate": 4.548293157042581e-06, + "loss": 1.192582130432129, + "step": 124 + }, + { + "epoch": 1.768, + "grad_norm": 0.004071912728250027, + "learning_rate": 4.466453599588103e-06, + "loss": 1.1831550598144531, + "step": 125 + }, + { + "epoch": 1.7822222222222224, + "grad_norm": 0.004058274906128645, + "learning_rate": 4.384758385029457e-06, + "loss": 1.1674884557724, + "step": 126 + }, + { + "epoch": 1.7964444444444445, + "grad_norm": 0.0038475256878882647, + "learning_rate": 4.3032296147693225e-06, + "loss": 1.181553602218628, + "step": 127 + }, + { + "epoch": 1.8106666666666666, + "grad_norm": 0.0039198137819767, + "learning_rate": 4.2218893451814005e-06, + "loss": 1.153355360031128, + "step": 128 + }, + { + "epoch": 1.8248888888888888, + "grad_norm": 0.003993947524577379, + "learning_rate": 4.140759581643386e-06, + "loss": 1.1762878894805908, + "step": 129 + }, + { + "epoch": 1.8391111111111111, + "grad_norm": 0.003767550690099597, + "learning_rate": 4.059862272583755e-06, + "loss": 1.1792393922805786, + "step": 130 + }, + { + "epoch": 1.8533333333333335, + "grad_norm": 0.004042964428663254, + "learning_rate": 3.979219303543942e-06, + "loss": 1.1686198711395264, + "step": 131 + }, + { + "epoch": 1.8675555555555556, + "grad_norm": 0.004027931950986385, + "learning_rate": 3.898852491257547e-06, + "loss": 1.179172396659851, + "step": 132 + }, + { + "epoch": 1.8817777777777778, + "grad_norm": 0.003951320890337229, + "learning_rate": 3.818783577748138e-06, + "loss": 1.1667160987854004, + "step": 133 + }, + { + "epoch": 1.896, + "grad_norm": 0.003887354861944914, + "learning_rate": 3.7390342244472886e-06, + "loss": 1.1844608783721924, + "step": 134 + }, + { + "epoch": 1.9102222222222223, + "grad_norm": 0.003996608778834343, + "learning_rate": 3.659626006334395e-06, + "loss": 1.1989758014678955, + "step": 135 + }, + { + "epoch": 1.9244444444444444, + "grad_norm": 0.003960869275033474, + "learning_rate": 3.5805804060998926e-06, + "loss": 1.1868371963500977, + "step": 136 + }, + { + "epoch": 1.9386666666666668, + "grad_norm": 0.003877371782436967, + "learning_rate": 3.501918808333453e-06, + "loss": 1.166593313217163, + "step": 137 + }, + { + "epoch": 1.952888888888889, + "grad_norm": 0.004059459548443556, + "learning_rate": 3.4236624937386874e-06, + "loss": 1.1770471334457397, + "step": 138 + }, + { + "epoch": 1.967111111111111, + "grad_norm": 0.0038016922771930695, + "learning_rate": 3.3458326333759927e-06, + "loss": 1.1796729564666748, + "step": 139 + }, + { + "epoch": 1.9813333333333332, + "grad_norm": 0.00393063947558403, + "learning_rate": 3.268450282935026e-06, + "loss": 1.1624293327331543, + "step": 140 + }, + { + "epoch": 1.9955555555555555, + "grad_norm": 0.0037788248155266047, + "learning_rate": 3.1915363770384223e-06, + "loss": 1.1687278747558594, + "step": 141 + }, + { + "epoch": 2.0, + "grad_norm": 0.006625148933380842, + "learning_rate": 3.1151117235782346e-06, + "loss": 1.1975963115692139, + "step": 142 + } + ], + "logging_steps": 1, + "max_steps": 213, + "num_input_tokens_seen": 0, + "num_train_epochs": 3, + "save_steps": 500, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 3.010264556330025e+17, + "train_batch_size": 1, + "trial_name": null, + "trial_params": null +} diff --git a/qwen2.5-1.5b-instruct/checkpoint-142/training_args.bin b/qwen2.5-1.5b-instruct/checkpoint-142/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..9fe3caacd835eb5bd003416a09e775b3c6032cff --- /dev/null +++ b/qwen2.5-1.5b-instruct/checkpoint-142/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:439cfa655d37de4fb57f512de1d989fa4287c7c18d86c9d7c320a2bd5256fe19 +size 5649 diff --git a/qwen2.5-1.5b-instruct/checkpoint-213/chat_template.jinja b/qwen2.5-1.5b-instruct/checkpoint-213/chat_template.jinja new file mode 100644 index 0000000000000000000000000000000000000000..bdf7919a96cfe43d50914a007b9c0877bd0ec27e --- /dev/null +++ b/qwen2.5-1.5b-instruct/checkpoint-213/chat_template.jinja @@ -0,0 +1,54 @@ +{%- if tools %} + {{- '<|im_start|>system\n' }} + {%- if messages[0]['role'] == 'system' %} + {{- messages[0]['content'] }} + {%- else %} + {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }} + {%- endif %} + {{- "\n\n# 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' }} + {%- else %} + {{- '<|im_start|>system\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\n' }} + {%- endif %} +{%- endif %} +{%- for message in messages %} + {%- if (message.role == "user") or (message.role == "system" and not loop.first) or (message.role == "assistant" and not message.tool_calls) %} + {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }} + {%- elif message.role == "assistant" %} + {{- '<|im_start|>' + message.role }} + {%- if message.content %} + {{- '\n' + message.content }} + {%- endif %} + {%- for tool_call in message.tool_calls %} + {%- if tool_call.function is defined %} + {%- set tool_call = tool_call.function %} + {%- endif %} + {{- '\n\n{"name": "' }} + {{- tool_call.name }} + {{- '", "arguments": ' }} + {{- tool_call.arguments | tojson }} + {{- '}\n' }} + {%- endfor %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "tool" %} + {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != "tool") %} + {{- '<|im_start|>user' }} + {%- endif %} + {{- '\n\n' }} + {{- message.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' }} +{%- endif %} diff --git a/qwen2.5-1.5b-instruct/checkpoint-213/config.json b/qwen2.5-1.5b-instruct/checkpoint-213/config.json new file mode 100644 index 0000000000000000000000000000000000000000..499065fd744a64fc8ad2884da26e85a6e99d468a --- /dev/null +++ b/qwen2.5-1.5b-instruct/checkpoint-213/config.json @@ -0,0 +1,61 @@ +{ + "architectures": [ + "Qwen2ForCausalLM" + ], + "attention_dropout": 0.0, + "bos_token_id": null, + "dtype": "float32", + "eos_token_id": 151645, + "hidden_act": "silu", + "hidden_size": 1536, + "initializer_range": 0.02, + "intermediate_size": 8960, + "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" + ], + "max_position_embeddings": 32768, + "max_window_layers": 21, + "model_type": "qwen2", + "num_attention_heads": 12, + "num_hidden_layers": 28, + "num_key_value_heads": 2, + "pad_token_id": 151643, + "rms_norm_eps": 1e-06, + "rope_parameters": { + "rope_theta": 1000000.0, + "rope_type": "default" + }, + "sliding_window": null, + "tie_word_embeddings": true, + "transformers_version": "5.2.0", + "use_cache": false, + "use_sliding_window": false, + "vocab_size": 151936 +} diff --git a/qwen2.5-1.5b-instruct/checkpoint-213/generation_config.json b/qwen2.5-1.5b-instruct/checkpoint-213/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..c5b4e11411fed3e7a5e6557165f888a7b0b1248b --- /dev/null +++ b/qwen2.5-1.5b-instruct/checkpoint-213/generation_config.json @@ -0,0 +1,13 @@ +{ + "do_sample": true, + "eos_token_id": [ + 151645, + 151643 + ], + "pad_token_id": 151643, + "repetition_penalty": 1.1, + "temperature": 0.7, + "top_k": 20, + "top_p": 0.8, + "transformers_version": "5.2.0" +} diff --git a/qwen2.5-1.5b-instruct/checkpoint-213/model.safetensors b/qwen2.5-1.5b-instruct/checkpoint-213/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..7c9b6652e3790d6f8e2afa2bd0c78cb51bf27f05 --- /dev/null +++ b/qwen2.5-1.5b-instruct/checkpoint-213/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:670fc682f9aec84643a6aac7c761dc5caa0aae09dac649f0bed07488527fd569 +size 6174895536 diff --git a/qwen2.5-1.5b-instruct/checkpoint-213/optimizer.pt b/qwen2.5-1.5b-instruct/checkpoint-213/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..017b367405314f9aa32a01459c2c15dcae2330d6 --- /dev/null +++ b/qwen2.5-1.5b-instruct/checkpoint-213/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f6447eddee0bc2a3f237cdcb0209e907d2be045f995bc2f5a05b99907bf9d204 +size 12350013801 diff --git a/qwen2.5-1.5b-instruct/checkpoint-213/rng_state_0.pth b/qwen2.5-1.5b-instruct/checkpoint-213/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..870021e3fa5ac35c2f711adf0c93a556ab4842da --- /dev/null +++ b/qwen2.5-1.5b-instruct/checkpoint-213/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:95e5fc2074c0df31522a514f862c86cb00d71c946a7f15cc9ec0e53a69fb28a7 +size 14917 diff --git a/qwen2.5-1.5b-instruct/checkpoint-213/rng_state_1.pth b/qwen2.5-1.5b-instruct/checkpoint-213/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..21f20da5eb1da017f08aaa88bd19cf24d40e3fbf --- /dev/null +++ b/qwen2.5-1.5b-instruct/checkpoint-213/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0e7153eae67b6c9232a41bc996a2bf5b83229b8c7230d61911ac0fd40e64154e +size 14917 diff --git a/qwen2.5-1.5b-instruct/checkpoint-213/scheduler.pt b/qwen2.5-1.5b-instruct/checkpoint-213/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..39cf87775c3a71d10e7311e0cab769e112f357ba --- /dev/null +++ b/qwen2.5-1.5b-instruct/checkpoint-213/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c38a67b98c17be81d4c6d00b60029cdedd43e9aba35ca28ec28ef946c400f268 +size 1465 diff --git a/qwen2.5-1.5b-instruct/checkpoint-213/tokenizer.json b/qwen2.5-1.5b-instruct/checkpoint-213/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..34510ff0037cd50428af467a17ead5a96140a32c --- /dev/null +++ b/qwen2.5-1.5b-instruct/checkpoint-213/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3fd169731d2cbde95e10bf356d66d5997fd885dd8dbb6fb4684da3f23b2585d8 +size 11421892 diff --git a/qwen2.5-1.5b-instruct/checkpoint-213/tokenizer_config.json b/qwen2.5-1.5b-instruct/checkpoint-213/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..145e2c7a68b109ae0253dda29fc2e66cb01b4b85 --- /dev/null +++ b/qwen2.5-1.5b-instruct/checkpoint-213/tokenizer_config.json @@ -0,0 +1,30 @@ +{ + "add_prefix_space": false, + "backend": "tokenizers", + "bos_token": null, + "clean_up_tokenization_spaces": false, + "eos_token": "<|im_end|>", + "errors": "replace", + "extra_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|>" + ], + "is_local": false, + "model_max_length": 131072, + "pad_token": "<|endoftext|>", + "padding_side": "right", + "split_special_tokens": false, + "tokenizer_class": "Qwen2Tokenizer", + "unk_token": null +} diff --git a/qwen2.5-1.5b-instruct/checkpoint-213/trainer_state.json b/qwen2.5-1.5b-instruct/checkpoint-213/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..3c8f0af3feb999d359fd27af65c1a062f62ac150 --- /dev/null +++ b/qwen2.5-1.5b-instruct/checkpoint-213/trainer_state.json @@ -0,0 +1,1557 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 3.0, + "eval_steps": 50, + "global_step": 213, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.014222222222222223, + "grad_norm": 0.036850932985544205, + "learning_rate": 0.0, + "loss": 1.6560986042022705, + "step": 1 + }, + { + "epoch": 0.028444444444444446, + "grad_norm": 0.037402428686618805, + "learning_rate": 4.5454545454545457e-07, + "loss": 1.681471347808838, + "step": 2 + }, + { + "epoch": 0.042666666666666665, + "grad_norm": 0.03194897621870041, + "learning_rate": 9.090909090909091e-07, + "loss": 1.6596678495407104, + "step": 3 + }, + { + "epoch": 0.05688888888888889, + "grad_norm": 0.02700614556670189, + "learning_rate": 1.3636363636363636e-06, + "loss": 1.634795904159546, + "step": 4 + }, + { + "epoch": 0.07111111111111111, + "grad_norm": 0.021732285618782043, + "learning_rate": 1.8181818181818183e-06, + "loss": 1.606718897819519, + "step": 5 + }, + { + "epoch": 0.08533333333333333, + "grad_norm": 0.018985262140631676, + "learning_rate": 2.2727272727272728e-06, + "loss": 1.5807676315307617, + "step": 6 + }, + { + "epoch": 0.09955555555555555, + "grad_norm": 0.018287895247340202, + "learning_rate": 2.7272727272727272e-06, + "loss": 1.5490249395370483, + "step": 7 + }, + { + "epoch": 0.11377777777777778, + "grad_norm": 0.017051130533218384, + "learning_rate": 3.181818181818182e-06, + "loss": 1.5317950248718262, + "step": 8 + }, + { + "epoch": 0.128, + "grad_norm": 0.015596743673086166, + "learning_rate": 3.6363636363636366e-06, + "loss": 1.5076513290405273, + "step": 9 + }, + { + "epoch": 0.14222222222222222, + "grad_norm": 0.014579818584024906, + "learning_rate": 4.0909090909090915e-06, + "loss": 1.4786231517791748, + "step": 10 + }, + { + "epoch": 0.15644444444444444, + "grad_norm": 0.012900307774543762, + "learning_rate": 4.5454545454545455e-06, + "loss": 1.469682216644287, + "step": 11 + }, + { + "epoch": 0.17066666666666666, + "grad_norm": 0.011142965406179428, + "learning_rate": 5e-06, + "loss": 1.4563651084899902, + "step": 12 + }, + { + "epoch": 0.18488888888888888, + "grad_norm": 0.011200510896742344, + "learning_rate": 5.4545454545454545e-06, + "loss": 1.4510159492492676, + "step": 13 + }, + { + "epoch": 0.1991111111111111, + "grad_norm": 0.01117027923464775, + "learning_rate": 5.90909090909091e-06, + "loss": 1.440371036529541, + "step": 14 + }, + { + "epoch": 0.21333333333333335, + "grad_norm": 0.010404860600829124, + "learning_rate": 6.363636363636364e-06, + "loss": 1.4149150848388672, + "step": 15 + }, + { + "epoch": 0.22755555555555557, + "grad_norm": 0.009069642052054405, + "learning_rate": 6.818181818181818e-06, + "loss": 1.3951767683029175, + "step": 16 + }, + { + "epoch": 0.24177777777777779, + "grad_norm": 0.008563642390072346, + "learning_rate": 7.272727272727273e-06, + "loss": 1.3866009712219238, + "step": 17 + }, + { + "epoch": 0.256, + "grad_norm": 0.008158586919307709, + "learning_rate": 7.727272727272727e-06, + "loss": 1.3851287364959717, + "step": 18 + }, + { + "epoch": 0.2702222222222222, + "grad_norm": 0.008019132539629936, + "learning_rate": 8.181818181818183e-06, + "loss": 1.3660919666290283, + "step": 19 + }, + { + "epoch": 0.28444444444444444, + "grad_norm": 0.007505647838115692, + "learning_rate": 8.636363636363637e-06, + "loss": 1.3991025686264038, + "step": 20 + }, + { + "epoch": 0.2986666666666667, + "grad_norm": 0.0068238647654652596, + "learning_rate": 9.090909090909091e-06, + "loss": 1.3563741445541382, + "step": 21 + }, + { + "epoch": 0.3128888888888889, + "grad_norm": 0.00681255292147398, + "learning_rate": 9.545454545454547e-06, + "loss": 1.3432800769805908, + "step": 22 + }, + { + "epoch": 0.32711111111111113, + "grad_norm": 0.0064102653414011, + "learning_rate": 1e-05, + "loss": 1.3540418148040771, + "step": 23 + }, + { + "epoch": 0.3413333333333333, + "grad_norm": 0.006286031566560268, + "learning_rate": 9.999323662872998e-06, + "loss": 1.3571672439575195, + "step": 24 + }, + { + "epoch": 0.35555555555555557, + "grad_norm": 0.006193141452968121, + "learning_rate": 9.99729483446475e-06, + "loss": 1.3181281089782715, + "step": 25 + }, + { + "epoch": 0.36977777777777776, + "grad_norm": 0.005919129587709904, + "learning_rate": 9.993914063644053e-06, + "loss": 1.3311502933502197, + "step": 26 + }, + { + "epoch": 0.384, + "grad_norm": 0.005563116166740656, + "learning_rate": 9.989182265027232e-06, + "loss": 1.3282787799835205, + "step": 27 + }, + { + "epoch": 0.3982222222222222, + "grad_norm": 0.005611867178231478, + "learning_rate": 9.98310071873072e-06, + "loss": 1.325502872467041, + "step": 28 + }, + { + "epoch": 0.41244444444444445, + "grad_norm": 0.0053426348604261875, + "learning_rate": 9.975671070024741e-06, + "loss": 1.3453574180603027, + "step": 29 + }, + { + "epoch": 0.4266666666666667, + "grad_norm": 0.0058601596392691135, + "learning_rate": 9.966895328888195e-06, + "loss": 1.3257673978805542, + "step": 30 + }, + { + "epoch": 0.4408888888888889, + "grad_norm": 0.005306159611791372, + "learning_rate": 9.956775869464901e-06, + "loss": 1.317690372467041, + "step": 31 + }, + { + "epoch": 0.45511111111111113, + "grad_norm": 0.0048169931396842, + "learning_rate": 9.945315429421307e-06, + "loss": 1.3243460655212402, + "step": 32 + }, + { + "epoch": 0.4693333333333333, + "grad_norm": 0.0048377541825175285, + "learning_rate": 9.932517109205849e-06, + "loss": 1.3194491863250732, + "step": 33 + }, + { + "epoch": 0.48355555555555557, + "grad_norm": 0.004818843677639961, + "learning_rate": 9.918384371210178e-06, + "loss": 1.3047914505004883, + "step": 34 + }, + { + "epoch": 0.49777777777777776, + "grad_norm": 0.005093662533909082, + "learning_rate": 9.902921038832456e-06, + "loss": 1.2955162525177002, + "step": 35 + }, + { + "epoch": 0.512, + "grad_norm": 0.005024498328566551, + "learning_rate": 9.886131295443003e-06, + "loss": 1.3118610382080078, + "step": 36 + }, + { + "epoch": 0.5262222222222223, + "grad_norm": 0.004711818415671587, + "learning_rate": 9.868019683252543e-06, + "loss": 1.302067518234253, + "step": 37 + }, + { + "epoch": 0.5404444444444444, + "grad_norm": 0.0047568087466061115, + "learning_rate": 9.848591102083375e-06, + "loss": 1.2789689302444458, + "step": 38 + }, + { + "epoch": 0.5546666666666666, + "grad_norm": 0.004525096621364355, + "learning_rate": 9.82785080804381e-06, + "loss": 1.2827372550964355, + "step": 39 + }, + { + "epoch": 0.5688888888888889, + "grad_norm": 0.0044178408570587635, + "learning_rate": 9.805804412106197e-06, + "loss": 1.2863564491271973, + "step": 40 + }, + { + "epoch": 0.5831111111111111, + "grad_norm": 0.00429871492087841, + "learning_rate": 9.782457878588977e-06, + "loss": 1.2768702507019043, + "step": 41 + }, + { + "epoch": 0.5973333333333334, + "grad_norm": 0.004442654084414244, + "learning_rate": 9.75781752354311e-06, + "loss": 1.2826354503631592, + "step": 42 + }, + { + "epoch": 0.6115555555555555, + "grad_norm": 0.004453128203749657, + "learning_rate": 9.731890013043367e-06, + "loss": 1.275444746017456, + "step": 43 + }, + { + "epoch": 0.6257777777777778, + "grad_norm": 0.004429741762578487, + "learning_rate": 9.704682361384941e-06, + "loss": 1.2660303115844727, + "step": 44 + }, + { + "epoch": 0.64, + "grad_norm": 0.004699238575994968, + "learning_rate": 9.676201929185809e-06, + "loss": 1.2627251148223877, + "step": 45 + }, + { + "epoch": 0.6542222222222223, + "grad_norm": 0.004564967937767506, + "learning_rate": 9.646456421395447e-06, + "loss": 1.2888470888137817, + "step": 46 + }, + { + "epoch": 0.6684444444444444, + "grad_norm": 0.004141585901379585, + "learning_rate": 9.615453885210368e-06, + "loss": 1.2826390266418457, + "step": 47 + }, + { + "epoch": 0.6826666666666666, + "grad_norm": 0.004241476766765118, + "learning_rate": 9.583202707897075e-06, + "loss": 1.2571580410003662, + "step": 48 + }, + { + "epoch": 0.6968888888888889, + "grad_norm": 0.004200824070721865, + "learning_rate": 9.549711614523007e-06, + "loss": 1.2616808414459229, + "step": 49 + }, + { + "epoch": 0.7111111111111111, + "grad_norm": 0.004399280063807964, + "learning_rate": 9.514989665596114e-06, + "loss": 1.263854742050171, + "step": 50 + }, + { + "epoch": 0.7111111111111111, + "eval_loss": 1.248613715171814, + "eval_runtime": 54.7705, + "eval_samples_per_second": 3.652, + "eval_steps_per_second": 1.826, + "step": 50 + }, + { + "epoch": 0.7253333333333334, + "grad_norm": 0.004282557871192694, + "learning_rate": 9.479046254613673e-06, + "loss": 1.279876947402954, + "step": 51 + }, + { + "epoch": 0.7395555555555555, + "grad_norm": 0.004135944880545139, + "learning_rate": 9.441891105521005e-06, + "loss": 1.2600014209747314, + "step": 52 + }, + { + "epoch": 0.7537777777777778, + "grad_norm": 0.0039484258741140366, + "learning_rate": 9.40353427008083e-06, + "loss": 1.2702335119247437, + "step": 53 + }, + { + "epoch": 0.768, + "grad_norm": 0.004160050302743912, + "learning_rate": 9.3639861251539e-06, + "loss": 1.2441630363464355, + "step": 54 + }, + { + "epoch": 0.7822222222222223, + "grad_norm": 0.004111913964152336, + "learning_rate": 9.323257369891702e-06, + "loss": 1.2441895008087158, + "step": 55 + }, + { + "epoch": 0.7964444444444444, + "grad_norm": 0.004063262604176998, + "learning_rate": 9.281359022841966e-06, + "loss": 1.255817174911499, + "step": 56 + }, + { + "epoch": 0.8106666666666666, + "grad_norm": 0.003984581213444471, + "learning_rate": 9.238302418967757e-06, + "loss": 1.2536640167236328, + "step": 57 + }, + { + "epoch": 0.8248888888888889, + "grad_norm": 0.003996702842414379, + "learning_rate": 9.194099206580981e-06, + "loss": 1.2499892711639404, + "step": 58 + }, + { + "epoch": 0.8391111111111111, + "grad_norm": 0.0044861058704555035, + "learning_rate": 9.14876134419111e-06, + "loss": 1.248606562614441, + "step": 59 + }, + { + "epoch": 0.8533333333333334, + "grad_norm": 0.004101422615349293, + "learning_rate": 9.102301097269974e-06, + "loss": 1.247856616973877, + "step": 60 + }, + { + "epoch": 0.8675555555555555, + "grad_norm": 0.004009375814348459, + "learning_rate": 9.05473103493355e-06, + "loss": 1.257920742034912, + "step": 61 + }, + { + "epoch": 0.8817777777777778, + "grad_norm": 0.0039364611729979515, + "learning_rate": 9.006064026541549e-06, + "loss": 1.2761049270629883, + "step": 62 + }, + { + "epoch": 0.896, + "grad_norm": 0.004427333828061819, + "learning_rate": 8.956313238215824e-06, + "loss": 1.245169997215271, + "step": 63 + }, + { + "epoch": 0.9102222222222223, + "grad_norm": 0.004076611250638962, + "learning_rate": 8.905492129278478e-06, + "loss": 1.234323263168335, + "step": 64 + }, + { + "epoch": 0.9244444444444444, + "grad_norm": 0.004032611381262541, + "learning_rate": 8.85361444861063e-06, + "loss": 1.2250256538391113, + "step": 65 + }, + { + "epoch": 0.9386666666666666, + "grad_norm": 0.004036127123981714, + "learning_rate": 8.800694230932885e-06, + "loss": 1.2339348793029785, + "step": 66 + }, + { + "epoch": 0.9528888888888889, + "grad_norm": 0.004113379865884781, + "learning_rate": 8.74674579300843e-06, + "loss": 1.252026081085205, + "step": 67 + }, + { + "epoch": 0.9671111111111111, + "grad_norm": 0.004064187407493591, + "learning_rate": 8.691783729769874e-06, + "loss": 1.2473349571228027, + "step": 68 + }, + { + "epoch": 0.9813333333333333, + "grad_norm": 0.0040993476286530495, + "learning_rate": 8.635822910370793e-06, + "loss": 1.2227911949157715, + "step": 69 + }, + { + "epoch": 0.9955555555555555, + "grad_norm": 0.003903420874848962, + "learning_rate": 8.578878474163115e-06, + "loss": 1.2343428134918213, + "step": 70 + }, + { + "epoch": 1.0, + "grad_norm": 0.006894194986671209, + "learning_rate": 8.520965826601394e-06, + "loss": 1.2428948879241943, + "step": 71 + }, + { + "epoch": 1.0142222222222221, + "grad_norm": 0.004313764162361622, + "learning_rate": 8.462100635075097e-06, + "loss": 1.202027440071106, + "step": 72 + }, + { + "epoch": 1.0284444444444445, + "grad_norm": 0.004148328676819801, + "learning_rate": 8.40229882467003e-06, + "loss": 1.2224020957946777, + "step": 73 + }, + { + "epoch": 1.0426666666666666, + "grad_norm": 0.00453830836340785, + "learning_rate": 8.341576573860049e-06, + "loss": 1.2196844816207886, + "step": 74 + }, + { + "epoch": 1.056888888888889, + "grad_norm": 0.004094650503247976, + "learning_rate": 8.279950310130218e-06, + "loss": 1.2222049236297607, + "step": 75 + }, + { + "epoch": 1.0711111111111111, + "grad_norm": 0.004169773776084185, + "learning_rate": 8.2174367055326e-06, + "loss": 1.2127679586410522, + "step": 76 + }, + { + "epoch": 1.0853333333333333, + "grad_norm": 0.004176456015557051, + "learning_rate": 8.154052672175888e-06, + "loss": 1.1962798833847046, + "step": 77 + }, + { + "epoch": 1.0995555555555556, + "grad_norm": 0.004215880297124386, + "learning_rate": 8.08981535765009e-06, + "loss": 1.192876935005188, + "step": 78 + }, + { + "epoch": 1.1137777777777778, + "grad_norm": 0.004112801048904657, + "learning_rate": 8.024742140387506e-06, + "loss": 1.204209566116333, + "step": 79 + }, + { + "epoch": 1.1280000000000001, + "grad_norm": 0.00416502496227622, + "learning_rate": 7.95885062496126e-06, + "loss": 1.2061021327972412, + "step": 80 + }, + { + "epoch": 1.1422222222222222, + "grad_norm": 0.004163650795817375, + "learning_rate": 7.892158637322647e-06, + "loss": 1.2074682712554932, + "step": 81 + }, + { + "epoch": 1.1564444444444444, + "grad_norm": 0.004129140637814999, + "learning_rate": 7.824684219978591e-06, + "loss": 1.1797451972961426, + "step": 82 + }, + { + "epoch": 1.1706666666666667, + "grad_norm": 0.004320825915783644, + "learning_rate": 7.756445627110523e-06, + "loss": 1.2151604890823364, + "step": 83 + }, + { + "epoch": 1.1848888888888889, + "grad_norm": 0.004057695157825947, + "learning_rate": 7.68746131963598e-06, + "loss": 1.1917520761489868, + "step": 84 + }, + { + "epoch": 1.199111111111111, + "grad_norm": 0.004045524168759584, + "learning_rate": 7.6177499602143e-06, + "loss": 1.1966304779052734, + "step": 85 + }, + { + "epoch": 1.2133333333333334, + "grad_norm": 0.004168134648352861, + "learning_rate": 7.547330408197695e-06, + "loss": 1.2256059646606445, + "step": 86 + }, + { + "epoch": 1.2275555555555555, + "grad_norm": 0.0040689390152692795, + "learning_rate": 7.476221714529167e-06, + "loss": 1.1984763145446777, + "step": 87 + }, + { + "epoch": 1.2417777777777779, + "grad_norm": 0.004095707554370165, + "learning_rate": 7.404443116588548e-06, + "loss": 1.1815118789672852, + "step": 88 + }, + { + "epoch": 1.256, + "grad_norm": 0.004128715489059687, + "learning_rate": 7.332014032988123e-06, + "loss": 1.1813948154449463, + "step": 89 + }, + { + "epoch": 1.2702222222222221, + "grad_norm": 0.004037311766296625, + "learning_rate": 7.2589540583192165e-06, + "loss": 1.2083604335784912, + "step": 90 + }, + { + "epoch": 1.2844444444444445, + "grad_norm": 0.004094353877007961, + "learning_rate": 7.185282957851175e-06, + "loss": 1.2008360624313354, + "step": 91 + }, + { + "epoch": 1.2986666666666666, + "grad_norm": 0.004135950934141874, + "learning_rate": 7.111020662184174e-06, + "loss": 1.1902484893798828, + "step": 92 + }, + { + "epoch": 1.3128888888888888, + "grad_norm": 0.003980172332376242, + "learning_rate": 7.036187261857289e-06, + "loss": 1.1865317821502686, + "step": 93 + }, + { + "epoch": 1.3271111111111111, + "grad_norm": 0.003914009779691696, + "learning_rate": 6.960803001913315e-06, + "loss": 1.1860058307647705, + "step": 94 + }, + { + "epoch": 1.3413333333333333, + "grad_norm": 0.003945492208003998, + "learning_rate": 6.884888276421766e-06, + "loss": 1.1842578649520874, + "step": 95 + }, + { + "epoch": 1.3555555555555556, + "grad_norm": 0.0038985933642834425, + "learning_rate": 6.8084636229615786e-06, + "loss": 1.1814080476760864, + "step": 96 + }, + { + "epoch": 1.3697777777777778, + "grad_norm": 0.003922275733202696, + "learning_rate": 6.731549717064975e-06, + "loss": 1.193676233291626, + "step": 97 + }, + { + "epoch": 1.384, + "grad_norm": 0.004165121354162693, + "learning_rate": 6.654167366624009e-06, + "loss": 1.2141278982162476, + "step": 98 + }, + { + "epoch": 1.3982222222222223, + "grad_norm": 0.004156369715929031, + "learning_rate": 6.576337506261314e-06, + "loss": 1.1865811347961426, + "step": 99 + }, + { + "epoch": 1.4124444444444444, + "grad_norm": 0.00419853487983346, + "learning_rate": 6.498081191666549e-06, + "loss": 1.1954832077026367, + "step": 100 + }, + { + "epoch": 1.4124444444444444, + "eval_loss": 1.1988927125930786, + "eval_runtime": 11.4257, + "eval_samples_per_second": 17.504, + "eval_steps_per_second": 8.752, + "step": 100 + }, + { + "epoch": 1.4266666666666667, + "grad_norm": 0.003968748729676008, + "learning_rate": 6.419419593900109e-06, + "loss": 1.1804523468017578, + "step": 101 + }, + { + "epoch": 1.4408888888888889, + "grad_norm": 0.003949783742427826, + "learning_rate": 6.340373993665607e-06, + "loss": 1.2068874835968018, + "step": 102 + }, + { + "epoch": 1.455111111111111, + "grad_norm": 0.004127713851630688, + "learning_rate": 6.2609657755527135e-06, + "loss": 1.1804969310760498, + "step": 103 + }, + { + "epoch": 1.4693333333333334, + "grad_norm": 0.004182465374469757, + "learning_rate": 6.1812164222518626e-06, + "loss": 1.1953659057617188, + "step": 104 + }, + { + "epoch": 1.4835555555555555, + "grad_norm": 0.004224156495183706, + "learning_rate": 6.101147508742456e-06, + "loss": 1.1933740377426147, + "step": 105 + }, + { + "epoch": 1.4977777777777779, + "grad_norm": 0.0041472818702459335, + "learning_rate": 6.020780696456059e-06, + "loss": 1.201953411102295, + "step": 106 + }, + { + "epoch": 1.512, + "grad_norm": 0.0038781315088272095, + "learning_rate": 5.940137727416247e-06, + "loss": 1.1897786855697632, + "step": 107 + }, + { + "epoch": 1.5262222222222221, + "grad_norm": 0.004033711273223162, + "learning_rate": 5.859240418356614e-06, + "loss": 1.1798899173736572, + "step": 108 + }, + { + "epoch": 1.5404444444444443, + "grad_norm": 0.00406253058463335, + "learning_rate": 5.778110654818602e-06, + "loss": 1.1853947639465332, + "step": 109 + }, + { + "epoch": 1.5546666666666666, + "grad_norm": 0.004025817848742008, + "learning_rate": 5.696770385230679e-06, + "loss": 1.1760761737823486, + "step": 110 + }, + { + "epoch": 1.568888888888889, + "grad_norm": 0.004029016941785812, + "learning_rate": 5.615241614970546e-06, + "loss": 1.1980795860290527, + "step": 111 + }, + { + "epoch": 1.5831111111111111, + "grad_norm": 0.004079029895365238, + "learning_rate": 5.533546400411899e-06, + "loss": 1.1878252029418945, + "step": 112 + }, + { + "epoch": 1.5973333333333333, + "grad_norm": 0.004004180897027254, + "learning_rate": 5.4517068429574215e-06, + "loss": 1.1891889572143555, + "step": 113 + }, + { + "epoch": 1.6115555555555554, + "grad_norm": 0.004044163506478071, + "learning_rate": 5.3697450830595775e-06, + "loss": 1.1935696601867676, + "step": 114 + }, + { + "epoch": 1.6257777777777778, + "grad_norm": 0.0040819537825882435, + "learning_rate": 5.287683294230855e-06, + "loss": 1.1648902893066406, + "step": 115 + }, + { + "epoch": 1.6400000000000001, + "grad_norm": 0.0042525348253548145, + "learning_rate": 5.20554367704505e-06, + "loss": 1.1749862432479858, + "step": 116 + }, + { + "epoch": 1.6542222222222223, + "grad_norm": 0.004129970911890268, + "learning_rate": 5.123348453131242e-06, + "loss": 1.200908899307251, + "step": 117 + }, + { + "epoch": 1.6684444444444444, + "grad_norm": 0.004072704818099737, + "learning_rate": 5.041119859162068e-06, + "loss": 1.189719796180725, + "step": 118 + }, + { + "epoch": 1.6826666666666665, + "grad_norm": 0.004113580100238323, + "learning_rate": 4.958880140837934e-06, + "loss": 1.1798235177993774, + "step": 119 + }, + { + "epoch": 1.696888888888889, + "grad_norm": 0.003983311355113983, + "learning_rate": 4.876651546868759e-06, + "loss": 1.166198492050171, + "step": 120 + }, + { + "epoch": 1.7111111111111112, + "grad_norm": 0.003966879565268755, + "learning_rate": 4.794456322954953e-06, + "loss": 1.1802079677581787, + "step": 121 + }, + { + "epoch": 1.7253333333333334, + "grad_norm": 0.004070781171321869, + "learning_rate": 4.7123167057691446e-06, + "loss": 1.1807469129562378, + "step": 122 + }, + { + "epoch": 1.7395555555555555, + "grad_norm": 0.0039869495667517185, + "learning_rate": 4.630254916940424e-06, + "loss": 1.1979055404663086, + "step": 123 + }, + { + "epoch": 1.7537777777777777, + "grad_norm": 0.004057880956679583, + "learning_rate": 4.548293157042581e-06, + "loss": 1.192582130432129, + "step": 124 + }, + { + "epoch": 1.768, + "grad_norm": 0.004071912728250027, + "learning_rate": 4.466453599588103e-06, + "loss": 1.1831550598144531, + "step": 125 + }, + { + "epoch": 1.7822222222222224, + "grad_norm": 0.004058274906128645, + "learning_rate": 4.384758385029457e-06, + "loss": 1.1674884557724, + "step": 126 + }, + { + "epoch": 1.7964444444444445, + "grad_norm": 0.0038475256878882647, + "learning_rate": 4.3032296147693225e-06, + "loss": 1.181553602218628, + "step": 127 + }, + { + "epoch": 1.8106666666666666, + "grad_norm": 0.0039198137819767, + "learning_rate": 4.2218893451814005e-06, + "loss": 1.153355360031128, + "step": 128 + }, + { + "epoch": 1.8248888888888888, + "grad_norm": 0.003993947524577379, + "learning_rate": 4.140759581643386e-06, + "loss": 1.1762878894805908, + "step": 129 + }, + { + "epoch": 1.8391111111111111, + "grad_norm": 0.003767550690099597, + "learning_rate": 4.059862272583755e-06, + "loss": 1.1792393922805786, + "step": 130 + }, + { + "epoch": 1.8533333333333335, + "grad_norm": 0.004042964428663254, + "learning_rate": 3.979219303543942e-06, + "loss": 1.1686198711395264, + "step": 131 + }, + { + "epoch": 1.8675555555555556, + "grad_norm": 0.004027931950986385, + "learning_rate": 3.898852491257547e-06, + "loss": 1.179172396659851, + "step": 132 + }, + { + "epoch": 1.8817777777777778, + "grad_norm": 0.003951320890337229, + "learning_rate": 3.818783577748138e-06, + "loss": 1.1667160987854004, + "step": 133 + }, + { + "epoch": 1.896, + "grad_norm": 0.003887354861944914, + "learning_rate": 3.7390342244472886e-06, + "loss": 1.1844608783721924, + "step": 134 + }, + { + "epoch": 1.9102222222222223, + "grad_norm": 0.003996608778834343, + "learning_rate": 3.659626006334395e-06, + "loss": 1.1989758014678955, + "step": 135 + }, + { + "epoch": 1.9244444444444444, + "grad_norm": 0.003960869275033474, + "learning_rate": 3.5805804060998926e-06, + "loss": 1.1868371963500977, + "step": 136 + }, + { + "epoch": 1.9386666666666668, + "grad_norm": 0.003877371782436967, + "learning_rate": 3.501918808333453e-06, + "loss": 1.166593313217163, + "step": 137 + }, + { + "epoch": 1.952888888888889, + "grad_norm": 0.004059459548443556, + "learning_rate": 3.4236624937386874e-06, + "loss": 1.1770471334457397, + "step": 138 + }, + { + "epoch": 1.967111111111111, + "grad_norm": 0.0038016922771930695, + "learning_rate": 3.3458326333759927e-06, + "loss": 1.1796729564666748, + "step": 139 + }, + { + "epoch": 1.9813333333333332, + "grad_norm": 0.00393063947558403, + "learning_rate": 3.268450282935026e-06, + "loss": 1.1624293327331543, + "step": 140 + }, + { + "epoch": 1.9955555555555555, + "grad_norm": 0.0037788248155266047, + "learning_rate": 3.1915363770384223e-06, + "loss": 1.1687278747558594, + "step": 141 + }, + { + "epoch": 2.0, + "grad_norm": 0.006625148933380842, + "learning_rate": 3.1151117235782346e-06, + "loss": 1.1975963115692139, + "step": 142 + }, + { + "epoch": 2.014222222222222, + "grad_norm": 0.003876346629112959, + "learning_rate": 3.0391969980866874e-06, + "loss": 1.1713850498199463, + "step": 143 + }, + { + "epoch": 2.0284444444444443, + "grad_norm": 0.003996741026639938, + "learning_rate": 2.963812738142713e-06, + "loss": 1.1817915439605713, + "step": 144 + }, + { + "epoch": 2.042666666666667, + "grad_norm": 0.003873717039823532, + "learning_rate": 2.8889793378158284e-06, + "loss": 1.1671693325042725, + "step": 145 + }, + { + "epoch": 2.056888888888889, + "grad_norm": 0.003939088433980942, + "learning_rate": 2.814717042148827e-06, + "loss": 1.1553094387054443, + "step": 146 + }, + { + "epoch": 2.071111111111111, + "grad_norm": 0.003916384186595678, + "learning_rate": 2.7410459416807856e-06, + "loss": 1.1611030101776123, + "step": 147 + }, + { + "epoch": 2.0853333333333333, + "grad_norm": 0.0037560381460934877, + "learning_rate": 2.6679859670118785e-06, + "loss": 1.169109582901001, + "step": 148 + }, + { + "epoch": 2.0995555555555554, + "grad_norm": 0.0038683414459228516, + "learning_rate": 2.5955568834114523e-06, + "loss": 1.1539714336395264, + "step": 149 + }, + { + "epoch": 2.113777777777778, + "grad_norm": 0.003892498556524515, + "learning_rate": 2.523778285470835e-06, + "loss": 1.1314873695373535, + "step": 150 + }, + { + "epoch": 2.113777777777778, + "eval_loss": 1.1810542345046997, + "eval_runtime": 77.0472, + "eval_samples_per_second": 2.596, + "eval_steps_per_second": 1.298, + "step": 150 + }, + { + "epoch": 2.128, + "grad_norm": 0.003935188986361027, + "learning_rate": 2.452669591802307e-06, + "loss": 1.1695539951324463, + "step": 151 + }, + { + "epoch": 2.1422222222222222, + "grad_norm": 0.0038220868445932865, + "learning_rate": 2.3822500397857016e-06, + "loss": 1.1494641304016113, + "step": 152 + }, + { + "epoch": 2.1564444444444444, + "grad_norm": 0.0038105787243694067, + "learning_rate": 2.3125386803640188e-06, + "loss": 1.1655616760253906, + "step": 153 + }, + { + "epoch": 2.1706666666666665, + "grad_norm": 0.003957564476877451, + "learning_rate": 2.243554372889479e-06, + "loss": 1.1519088745117188, + "step": 154 + }, + { + "epoch": 2.1848888888888887, + "grad_norm": 0.003938976209610701, + "learning_rate": 2.1753157800214107e-06, + "loss": 1.1479674577713013, + "step": 155 + }, + { + "epoch": 2.1991111111111112, + "grad_norm": 0.0037570318672806025, + "learning_rate": 2.1078413626773547e-06, + "loss": 1.1504448652267456, + "step": 156 + }, + { + "epoch": 2.2133333333333334, + "grad_norm": 0.003837036434561014, + "learning_rate": 2.0411493750387423e-06, + "loss": 1.142012119293213, + "step": 157 + }, + { + "epoch": 2.2275555555555555, + "grad_norm": 0.0037807337939739227, + "learning_rate": 1.9752578596124955e-06, + "loss": 1.1382803916931152, + "step": 158 + }, + { + "epoch": 2.2417777777777776, + "grad_norm": 0.0038557867519557476, + "learning_rate": 1.9101846423499113e-06, + "loss": 1.1602444648742676, + "step": 159 + }, + { + "epoch": 2.2560000000000002, + "grad_norm": 0.003806622698903084, + "learning_rate": 1.8459473278241125e-06, + "loss": 1.146360158920288, + "step": 160 + }, + { + "epoch": 2.2702222222222224, + "grad_norm": 0.003777023172006011, + "learning_rate": 1.7825632944674016e-06, + "loss": 1.1420152187347412, + "step": 161 + }, + { + "epoch": 2.2844444444444445, + "grad_norm": 0.0038270214572548866, + "learning_rate": 1.7200496898697832e-06, + "loss": 1.1756829023361206, + "step": 162 + }, + { + "epoch": 2.2986666666666666, + "grad_norm": 0.003796679200604558, + "learning_rate": 1.6584234261399535e-06, + "loss": 1.1596589088439941, + "step": 163 + }, + { + "epoch": 2.3128888888888888, + "grad_norm": 0.003775167977437377, + "learning_rate": 1.5977011753299726e-06, + "loss": 1.1395504474639893, + "step": 164 + }, + { + "epoch": 2.327111111111111, + "grad_norm": 0.0038062240928411484, + "learning_rate": 1.5378993649249053e-06, + "loss": 1.1680355072021484, + "step": 165 + }, + { + "epoch": 2.3413333333333335, + "grad_norm": 0.0038756304420530796, + "learning_rate": 1.4790341733986085e-06, + "loss": 1.1566658020019531, + "step": 166 + }, + { + "epoch": 2.3555555555555556, + "grad_norm": 0.00390993244946003, + "learning_rate": 1.4211215258368866e-06, + "loss": 1.1633788347244263, + "step": 167 + }, + { + "epoch": 2.3697777777777778, + "grad_norm": 0.0038066902197897434, + "learning_rate": 1.3641770896292083e-06, + "loss": 1.1514759063720703, + "step": 168 + }, + { + "epoch": 2.384, + "grad_norm": 0.003804982639849186, + "learning_rate": 1.3082162702301276e-06, + "loss": 1.1515188217163086, + "step": 169 + }, + { + "epoch": 2.398222222222222, + "grad_norm": 0.003867222461849451, + "learning_rate": 1.2532542069915722e-06, + "loss": 1.1591559648513794, + "step": 170 + }, + { + "epoch": 2.4124444444444446, + "grad_norm": 0.0037913229316473007, + "learning_rate": 1.1993057690671174e-06, + "loss": 1.1669801473617554, + "step": 171 + }, + { + "epoch": 2.4266666666666667, + "grad_norm": 0.0037280265241861343, + "learning_rate": 1.1463855513893695e-06, + "loss": 1.155794382095337, + "step": 172 + }, + { + "epoch": 2.440888888888889, + "grad_norm": 0.0040002260357141495, + "learning_rate": 1.0945078707215224e-06, + "loss": 1.1434683799743652, + "step": 173 + }, + { + "epoch": 2.455111111111111, + "grad_norm": 0.0037701099645346403, + "learning_rate": 1.0436867617841768e-06, + "loss": 1.1586582660675049, + "step": 174 + }, + { + "epoch": 2.469333333333333, + "grad_norm": 0.0038694143295288086, + "learning_rate": 9.939359734584552e-07, + "loss": 1.1643059253692627, + "step": 175 + }, + { + "epoch": 2.4835555555555557, + "grad_norm": 0.0036939869169145823, + "learning_rate": 9.452689650664515e-07, + "loss": 1.137399673461914, + "step": 176 + }, + { + "epoch": 2.497777777777778, + "grad_norm": 0.0037478527519851923, + "learning_rate": 8.976989027300265e-07, + "loss": 1.150512456893921, + "step": 177 + }, + { + "epoch": 2.512, + "grad_norm": 0.003883050289005041, + "learning_rate": 8.512386558088919e-07, + "loss": 1.1556127071380615, + "step": 178 + }, + { + "epoch": 2.526222222222222, + "grad_norm": 0.003667640034109354, + "learning_rate": 8.059007934190194e-07, + "loss": 1.1506540775299072, + "step": 179 + }, + { + "epoch": 2.5404444444444443, + "grad_norm": 0.003665938274934888, + "learning_rate": 7.61697581032243e-07, + "loss": 1.1606539487838745, + "step": 180 + }, + { + "epoch": 2.554666666666667, + "grad_norm": 0.0037777498364448547, + "learning_rate": 7.186409771580355e-07, + "loss": 1.145754098892212, + "step": 181 + }, + { + "epoch": 2.568888888888889, + "grad_norm": 0.0037169207353144884, + "learning_rate": 6.76742630108298e-07, + "loss": 1.1506030559539795, + "step": 182 + }, + { + "epoch": 2.583111111111111, + "grad_norm": 0.0036668803077191114, + "learning_rate": 6.360138748461015e-07, + "loss": 1.1366521120071411, + "step": 183 + }, + { + "epoch": 2.5973333333333333, + "grad_norm": 0.0037295217625796795, + "learning_rate": 5.964657299191712e-07, + "loss": 1.1723542213439941, + "step": 184 + }, + { + "epoch": 2.6115555555555554, + "grad_norm": 0.003689161501824856, + "learning_rate": 5.581088944789953e-07, + "loss": 1.1543989181518555, + "step": 185 + }, + { + "epoch": 2.6257777777777775, + "grad_norm": 0.0038293329998850822, + "learning_rate": 5.20953745386329e-07, + "loss": 1.1481022834777832, + "step": 186 + }, + { + "epoch": 2.64, + "grad_norm": 0.0037391053047031164, + "learning_rate": 4.850103344038853e-07, + "loss": 1.1510019302368164, + "step": 187 + }, + { + "epoch": 2.6542222222222223, + "grad_norm": 0.0036369990557432175, + "learning_rate": 4.502883854769935e-07, + "loss": 1.1454750299453735, + "step": 188 + }, + { + "epoch": 2.6684444444444444, + "grad_norm": 0.003734483616426587, + "learning_rate": 4.167972921029262e-07, + "loss": 1.1555531024932861, + "step": 189 + }, + { + "epoch": 2.6826666666666665, + "grad_norm": 0.0037341697607189417, + "learning_rate": 3.8454611478963235e-07, + "loss": 1.141042709350586, + "step": 190 + }, + { + "epoch": 2.696888888888889, + "grad_norm": 0.0036616334691643715, + "learning_rate": 3.535435786045538e-07, + "loss": 1.1597721576690674, + "step": 191 + }, + { + "epoch": 2.7111111111111112, + "grad_norm": 0.0037469498347491026, + "learning_rate": 3.237980708141919e-07, + "loss": 1.1370229721069336, + "step": 192 + }, + { + "epoch": 2.7253333333333334, + "grad_norm": 0.003746448317542672, + "learning_rate": 2.9531763861505967e-07, + "loss": 1.1386423110961914, + "step": 193 + }, + { + "epoch": 2.7395555555555555, + "grad_norm": 0.0038595127407461405, + "learning_rate": 2.681099869566328e-07, + "loss": 1.141430139541626, + "step": 194 + }, + { + "epoch": 2.7537777777777777, + "grad_norm": 0.0037963776849210262, + "learning_rate": 2.4218247645689306e-07, + "loss": 1.1269652843475342, + "step": 195 + }, + { + "epoch": 2.768, + "grad_norm": 0.0037523838691413403, + "learning_rate": 2.175421214110235e-07, + "loss": 1.1652940511703491, + "step": 196 + }, + { + "epoch": 2.7822222222222224, + "grad_norm": 0.0037649553269147873, + "learning_rate": 1.941955878938029e-07, + "loss": 1.1484644412994385, + "step": 197 + }, + { + "epoch": 2.7964444444444445, + "grad_norm": 0.0036591957323253155, + "learning_rate": 1.721491919561913e-07, + "loss": 1.1459333896636963, + "step": 198 + }, + { + "epoch": 2.8106666666666666, + "grad_norm": 0.003836029442027211, + "learning_rate": 1.514088979166256e-07, + "loss": 1.157778024673462, + "step": 199 + }, + { + "epoch": 2.824888888888889, + "grad_norm": 0.003767541144043207, + "learning_rate": 1.3198031674745814e-07, + "loss": 1.1373133659362793, + "step": 200 + }, + { + "epoch": 2.824888888888889, + "eval_loss": 1.1767549514770508, + "eval_runtime": 17.7324, + "eval_samples_per_second": 11.279, + "eval_steps_per_second": 5.639, + "step": 200 + }, + { + "epoch": 2.8391111111111114, + "grad_norm": 0.0036713366862386465, + "learning_rate": 1.138687045569975e-07, + "loss": 1.1472009420394897, + "step": 201 + }, + { + "epoch": 2.8533333333333335, + "grad_norm": 0.003775710705667734, + "learning_rate": 9.70789611675449e-08, + "loss": 1.1417003870010376, + "step": 202 + }, + { + "epoch": 2.8675555555555556, + "grad_norm": 0.003818412311375141, + "learning_rate": 8.161562878982399e-08, + "loss": 1.1214667558670044, + "step": 203 + }, + { + "epoch": 2.8817777777777778, + "grad_norm": 0.0038233508821576834, + "learning_rate": 6.74828907941516e-08, + "loss": 1.138871431350708, + "step": 204 + }, + { + "epoch": 2.896, + "grad_norm": 0.0036753350868821144, + "learning_rate": 5.468457057869358e-08, + "loss": 1.1596763134002686, + "step": 205 + }, + { + "epoch": 2.910222222222222, + "grad_norm": 0.0036339866928756237, + "learning_rate": 4.322413053509944e-08, + "loss": 1.1474013328552246, + "step": 206 + }, + { + "epoch": 2.924444444444444, + "grad_norm": 0.003777341218665242, + "learning_rate": 3.31046711118066e-08, + "loss": 1.1420172452926636, + "step": 207 + }, + { + "epoch": 2.9386666666666668, + "grad_norm": 0.0037329578772187233, + "learning_rate": 2.4328929975260262e-08, + "loss": 1.149836778640747, + "step": 208 + }, + { + "epoch": 2.952888888888889, + "grad_norm": 0.00386771815828979, + "learning_rate": 1.6899281269279756e-08, + "loss": 1.142824649810791, + "step": 209 + }, + { + "epoch": 2.967111111111111, + "grad_norm": 0.0036802676040679216, + "learning_rate": 1.0817734972768946e-08, + "loss": 1.1392970085144043, + "step": 210 + }, + { + "epoch": 2.981333333333333, + "grad_norm": 0.003702987916767597, + "learning_rate": 6.085936355947897e-09, + "loss": 1.1546673774719238, + "step": 211 + }, + { + "epoch": 2.9955555555555557, + "grad_norm": 0.0036437278613448143, + "learning_rate": 2.7051655352494654e-09, + "loss": 1.1519522666931152, + "step": 212 + }, + { + "epoch": 3.0, + "grad_norm": 0.006544829346239567, + "learning_rate": 6.763371270035457e-10, + "loss": 1.1223903894424438, + "step": 213 + } + ], + "logging_steps": 1, + "max_steps": 213, + "num_input_tokens_seen": 0, + "num_train_epochs": 3, + "save_steps": 500, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 4.5153968281196954e+17, + "train_batch_size": 1, + "trial_name": null, + "trial_params": null +} diff --git a/qwen2.5-1.5b-instruct/checkpoint-213/training_args.bin b/qwen2.5-1.5b-instruct/checkpoint-213/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..9fe3caacd835eb5bd003416a09e775b3c6032cff --- /dev/null +++ b/qwen2.5-1.5b-instruct/checkpoint-213/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:439cfa655d37de4fb57f512de1d989fa4287c7c18d86c9d7c320a2bd5256fe19 +size 5649 diff --git a/qwen2.5-1.5b-instruct/checkpoint-71/chat_template.jinja b/qwen2.5-1.5b-instruct/checkpoint-71/chat_template.jinja new file mode 100644 index 0000000000000000000000000000000000000000..bdf7919a96cfe43d50914a007b9c0877bd0ec27e --- /dev/null +++ b/qwen2.5-1.5b-instruct/checkpoint-71/chat_template.jinja @@ -0,0 +1,54 @@ +{%- if tools %} + {{- '<|im_start|>system\n' }} + {%- if messages[0]['role'] == 'system' %} + {{- messages[0]['content'] }} + {%- else %} + {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }} + {%- endif %} + {{- "\n\n# 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' }} + {%- else %} + {{- '<|im_start|>system\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\n' }} + {%- endif %} +{%- endif %} +{%- for message in messages %} + {%- if (message.role == "user") or (message.role == "system" and not loop.first) or (message.role == "assistant" and not message.tool_calls) %} + {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }} + {%- elif message.role == "assistant" %} + {{- '<|im_start|>' + message.role }} + {%- if message.content %} + {{- '\n' + message.content }} + {%- endif %} + {%- for tool_call in message.tool_calls %} + {%- if tool_call.function is defined %} + {%- set tool_call = tool_call.function %} + {%- endif %} + {{- '\n\n{"name": "' }} + {{- tool_call.name }} + {{- '", "arguments": ' }} + {{- tool_call.arguments | tojson }} + {{- '}\n' }} + {%- endfor %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "tool" %} + {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != "tool") %} + {{- '<|im_start|>user' }} + {%- endif %} + {{- '\n\n' }} + {{- message.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' }} +{%- endif %} diff --git a/qwen2.5-1.5b-instruct/checkpoint-71/config.json b/qwen2.5-1.5b-instruct/checkpoint-71/config.json new file mode 100644 index 0000000000000000000000000000000000000000..499065fd744a64fc8ad2884da26e85a6e99d468a --- /dev/null +++ b/qwen2.5-1.5b-instruct/checkpoint-71/config.json @@ -0,0 +1,61 @@ +{ + "architectures": [ + "Qwen2ForCausalLM" + ], + "attention_dropout": 0.0, + "bos_token_id": null, + "dtype": "float32", + "eos_token_id": 151645, + "hidden_act": "silu", + "hidden_size": 1536, + "initializer_range": 0.02, + "intermediate_size": 8960, + "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" + ], + "max_position_embeddings": 32768, + "max_window_layers": 21, + "model_type": "qwen2", + "num_attention_heads": 12, + "num_hidden_layers": 28, + "num_key_value_heads": 2, + "pad_token_id": 151643, + "rms_norm_eps": 1e-06, + "rope_parameters": { + "rope_theta": 1000000.0, + "rope_type": "default" + }, + "sliding_window": null, + "tie_word_embeddings": true, + "transformers_version": "5.2.0", + "use_cache": false, + "use_sliding_window": false, + "vocab_size": 151936 +} diff --git a/qwen2.5-1.5b-instruct/checkpoint-71/generation_config.json b/qwen2.5-1.5b-instruct/checkpoint-71/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..c5b4e11411fed3e7a5e6557165f888a7b0b1248b --- /dev/null +++ b/qwen2.5-1.5b-instruct/checkpoint-71/generation_config.json @@ -0,0 +1,13 @@ +{ + "do_sample": true, + "eos_token_id": [ + 151645, + 151643 + ], + "pad_token_id": 151643, + "repetition_penalty": 1.1, + "temperature": 0.7, + "top_k": 20, + "top_p": 0.8, + "transformers_version": "5.2.0" +} diff --git a/qwen2.5-1.5b-instruct/checkpoint-71/model.safetensors b/qwen2.5-1.5b-instruct/checkpoint-71/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..0823ada9d981217fe21e3a7c9626b83d75b9d004 --- /dev/null +++ b/qwen2.5-1.5b-instruct/checkpoint-71/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:49287d78b0916c490ddf0955a3f46c4109ba8705b5dd8d79cfc7a3023a4ce2c8 +size 6174895536 diff --git a/qwen2.5-1.5b-instruct/checkpoint-71/optimizer.pt b/qwen2.5-1.5b-instruct/checkpoint-71/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..65ece90ae23ac6276255fb81afcedd6c0cb5547f --- /dev/null +++ b/qwen2.5-1.5b-instruct/checkpoint-71/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:173b616688d0890693b2ed96cb88c91d2c3ebe788e38f42f1bc70d71463101a7 +size 12350013801 diff --git a/qwen2.5-1.5b-instruct/checkpoint-71/rng_state_0.pth b/qwen2.5-1.5b-instruct/checkpoint-71/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..870021e3fa5ac35c2f711adf0c93a556ab4842da --- /dev/null +++ b/qwen2.5-1.5b-instruct/checkpoint-71/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:95e5fc2074c0df31522a514f862c86cb00d71c946a7f15cc9ec0e53a69fb28a7 +size 14917 diff --git a/qwen2.5-1.5b-instruct/checkpoint-71/rng_state_1.pth b/qwen2.5-1.5b-instruct/checkpoint-71/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..21f20da5eb1da017f08aaa88bd19cf24d40e3fbf --- /dev/null +++ b/qwen2.5-1.5b-instruct/checkpoint-71/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0e7153eae67b6c9232a41bc996a2bf5b83229b8c7230d61911ac0fd40e64154e +size 14917 diff --git a/qwen2.5-1.5b-instruct/checkpoint-71/scheduler.pt b/qwen2.5-1.5b-instruct/checkpoint-71/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..c3be1d29bde40f225713ff5084d3d9d69858fcfd --- /dev/null +++ b/qwen2.5-1.5b-instruct/checkpoint-71/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9c96c28f12671e2f5db493417daefb3ff0b1bc36602843e0d7ef6df3649c27b3 +size 1465 diff --git a/qwen2.5-1.5b-instruct/checkpoint-71/tokenizer.json b/qwen2.5-1.5b-instruct/checkpoint-71/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..34510ff0037cd50428af467a17ead5a96140a32c --- /dev/null +++ b/qwen2.5-1.5b-instruct/checkpoint-71/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3fd169731d2cbde95e10bf356d66d5997fd885dd8dbb6fb4684da3f23b2585d8 +size 11421892 diff --git a/qwen2.5-1.5b-instruct/checkpoint-71/tokenizer_config.json b/qwen2.5-1.5b-instruct/checkpoint-71/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..145e2c7a68b109ae0253dda29fc2e66cb01b4b85 --- /dev/null +++ b/qwen2.5-1.5b-instruct/checkpoint-71/tokenizer_config.json @@ -0,0 +1,30 @@ +{ + "add_prefix_space": false, + "backend": "tokenizers", + "bos_token": null, + "clean_up_tokenization_spaces": false, + "eos_token": "<|im_end|>", + "errors": "replace", + "extra_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|>" + ], + "is_local": false, + "model_max_length": 131072, + "pad_token": "<|endoftext|>", + "padding_side": "right", + "split_special_tokens": false, + "tokenizer_class": "Qwen2Tokenizer", + "unk_token": null +} diff --git a/qwen2.5-1.5b-instruct/checkpoint-71/trainer_state.json b/qwen2.5-1.5b-instruct/checkpoint-71/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..036df38211f947218ce1b705f1ce2eec85423e58 --- /dev/null +++ b/qwen2.5-1.5b-instruct/checkpoint-71/trainer_state.json @@ -0,0 +1,539 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.0, + "eval_steps": 50, + "global_step": 71, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.014222222222222223, + "grad_norm": 0.036850932985544205, + "learning_rate": 0.0, + "loss": 1.6560986042022705, + "step": 1 + }, + { + "epoch": 0.028444444444444446, + "grad_norm": 0.037402428686618805, + "learning_rate": 4.5454545454545457e-07, + "loss": 1.681471347808838, + "step": 2 + }, + { + "epoch": 0.042666666666666665, + "grad_norm": 0.03194897621870041, + "learning_rate": 9.090909090909091e-07, + "loss": 1.6596678495407104, + "step": 3 + }, + { + "epoch": 0.05688888888888889, + "grad_norm": 0.02700614556670189, + "learning_rate": 1.3636363636363636e-06, + "loss": 1.634795904159546, + "step": 4 + }, + { + "epoch": 0.07111111111111111, + "grad_norm": 0.021732285618782043, + "learning_rate": 1.8181818181818183e-06, + "loss": 1.606718897819519, + "step": 5 + }, + { + "epoch": 0.08533333333333333, + "grad_norm": 0.018985262140631676, + "learning_rate": 2.2727272727272728e-06, + "loss": 1.5807676315307617, + "step": 6 + }, + { + "epoch": 0.09955555555555555, + "grad_norm": 0.018287895247340202, + "learning_rate": 2.7272727272727272e-06, + "loss": 1.5490249395370483, + "step": 7 + }, + { + "epoch": 0.11377777777777778, + "grad_norm": 0.017051130533218384, + "learning_rate": 3.181818181818182e-06, + "loss": 1.5317950248718262, + "step": 8 + }, + { + "epoch": 0.128, + "grad_norm": 0.015596743673086166, + "learning_rate": 3.6363636363636366e-06, + "loss": 1.5076513290405273, + "step": 9 + }, + { + "epoch": 0.14222222222222222, + "grad_norm": 0.014579818584024906, + "learning_rate": 4.0909090909090915e-06, + "loss": 1.4786231517791748, + "step": 10 + }, + { + "epoch": 0.15644444444444444, + "grad_norm": 0.012900307774543762, + "learning_rate": 4.5454545454545455e-06, + "loss": 1.469682216644287, + "step": 11 + }, + { + "epoch": 0.17066666666666666, + "grad_norm": 0.011142965406179428, + "learning_rate": 5e-06, + "loss": 1.4563651084899902, + "step": 12 + }, + { + "epoch": 0.18488888888888888, + "grad_norm": 0.011200510896742344, + "learning_rate": 5.4545454545454545e-06, + "loss": 1.4510159492492676, + "step": 13 + }, + { + "epoch": 0.1991111111111111, + "grad_norm": 0.01117027923464775, + "learning_rate": 5.90909090909091e-06, + "loss": 1.440371036529541, + "step": 14 + }, + { + "epoch": 0.21333333333333335, + "grad_norm": 0.010404860600829124, + "learning_rate": 6.363636363636364e-06, + "loss": 1.4149150848388672, + "step": 15 + }, + { + "epoch": 0.22755555555555557, + "grad_norm": 0.009069642052054405, + "learning_rate": 6.818181818181818e-06, + "loss": 1.3951767683029175, + "step": 16 + }, + { + "epoch": 0.24177777777777779, + "grad_norm": 0.008563642390072346, + "learning_rate": 7.272727272727273e-06, + "loss": 1.3866009712219238, + "step": 17 + }, + { + "epoch": 0.256, + "grad_norm": 0.008158586919307709, + "learning_rate": 7.727272727272727e-06, + "loss": 1.3851287364959717, + "step": 18 + }, + { + "epoch": 0.2702222222222222, + "grad_norm": 0.008019132539629936, + "learning_rate": 8.181818181818183e-06, + "loss": 1.3660919666290283, + "step": 19 + }, + { + "epoch": 0.28444444444444444, + "grad_norm": 0.007505647838115692, + "learning_rate": 8.636363636363637e-06, + "loss": 1.3991025686264038, + "step": 20 + }, + { + "epoch": 0.2986666666666667, + "grad_norm": 0.0068238647654652596, + "learning_rate": 9.090909090909091e-06, + "loss": 1.3563741445541382, + "step": 21 + }, + { + "epoch": 0.3128888888888889, + "grad_norm": 0.00681255292147398, + "learning_rate": 9.545454545454547e-06, + "loss": 1.3432800769805908, + "step": 22 + }, + { + "epoch": 0.32711111111111113, + "grad_norm": 0.0064102653414011, + "learning_rate": 1e-05, + "loss": 1.3540418148040771, + "step": 23 + }, + { + "epoch": 0.3413333333333333, + "grad_norm": 0.006286031566560268, + "learning_rate": 9.999323662872998e-06, + "loss": 1.3571672439575195, + "step": 24 + }, + { + "epoch": 0.35555555555555557, + "grad_norm": 0.006193141452968121, + "learning_rate": 9.99729483446475e-06, + "loss": 1.3181281089782715, + "step": 25 + }, + { + "epoch": 0.36977777777777776, + "grad_norm": 0.005919129587709904, + "learning_rate": 9.993914063644053e-06, + "loss": 1.3311502933502197, + "step": 26 + }, + { + "epoch": 0.384, + "grad_norm": 0.005563116166740656, + "learning_rate": 9.989182265027232e-06, + "loss": 1.3282787799835205, + "step": 27 + }, + { + "epoch": 0.3982222222222222, + "grad_norm": 0.005611867178231478, + "learning_rate": 9.98310071873072e-06, + "loss": 1.325502872467041, + "step": 28 + }, + { + "epoch": 0.41244444444444445, + "grad_norm": 0.0053426348604261875, + "learning_rate": 9.975671070024741e-06, + "loss": 1.3453574180603027, + "step": 29 + }, + { + "epoch": 0.4266666666666667, + "grad_norm": 0.0058601596392691135, + "learning_rate": 9.966895328888195e-06, + "loss": 1.3257673978805542, + "step": 30 + }, + { + "epoch": 0.4408888888888889, + "grad_norm": 0.005306159611791372, + "learning_rate": 9.956775869464901e-06, + "loss": 1.317690372467041, + "step": 31 + }, + { + "epoch": 0.45511111111111113, + "grad_norm": 0.0048169931396842, + "learning_rate": 9.945315429421307e-06, + "loss": 1.3243460655212402, + "step": 32 + }, + { + "epoch": 0.4693333333333333, + "grad_norm": 0.0048377541825175285, + "learning_rate": 9.932517109205849e-06, + "loss": 1.3194491863250732, + "step": 33 + }, + { + "epoch": 0.48355555555555557, + "grad_norm": 0.004818843677639961, + "learning_rate": 9.918384371210178e-06, + "loss": 1.3047914505004883, + "step": 34 + }, + { + "epoch": 0.49777777777777776, + "grad_norm": 0.005093662533909082, + "learning_rate": 9.902921038832456e-06, + "loss": 1.2955162525177002, + "step": 35 + }, + { + "epoch": 0.512, + "grad_norm": 0.005024498328566551, + "learning_rate": 9.886131295443003e-06, + "loss": 1.3118610382080078, + "step": 36 + }, + { + "epoch": 0.5262222222222223, + "grad_norm": 0.004711818415671587, + "learning_rate": 9.868019683252543e-06, + "loss": 1.302067518234253, + "step": 37 + }, + { + "epoch": 0.5404444444444444, + "grad_norm": 0.0047568087466061115, + "learning_rate": 9.848591102083375e-06, + "loss": 1.2789689302444458, + "step": 38 + }, + { + "epoch": 0.5546666666666666, + "grad_norm": 0.004525096621364355, + "learning_rate": 9.82785080804381e-06, + "loss": 1.2827372550964355, + "step": 39 + }, + { + "epoch": 0.5688888888888889, + "grad_norm": 0.0044178408570587635, + "learning_rate": 9.805804412106197e-06, + "loss": 1.2863564491271973, + "step": 40 + }, + { + "epoch": 0.5831111111111111, + "grad_norm": 0.00429871492087841, + "learning_rate": 9.782457878588977e-06, + "loss": 1.2768702507019043, + "step": 41 + }, + { + "epoch": 0.5973333333333334, + "grad_norm": 0.004442654084414244, + "learning_rate": 9.75781752354311e-06, + "loss": 1.2826354503631592, + "step": 42 + }, + { + "epoch": 0.6115555555555555, + "grad_norm": 0.004453128203749657, + "learning_rate": 9.731890013043367e-06, + "loss": 1.275444746017456, + "step": 43 + }, + { + "epoch": 0.6257777777777778, + "grad_norm": 0.004429741762578487, + "learning_rate": 9.704682361384941e-06, + "loss": 1.2660303115844727, + "step": 44 + }, + { + "epoch": 0.64, + "grad_norm": 0.004699238575994968, + "learning_rate": 9.676201929185809e-06, + "loss": 1.2627251148223877, + "step": 45 + }, + { + "epoch": 0.6542222222222223, + "grad_norm": 0.004564967937767506, + "learning_rate": 9.646456421395447e-06, + "loss": 1.2888470888137817, + "step": 46 + }, + { + "epoch": 0.6684444444444444, + "grad_norm": 0.004141585901379585, + "learning_rate": 9.615453885210368e-06, + "loss": 1.2826390266418457, + "step": 47 + }, + { + "epoch": 0.6826666666666666, + "grad_norm": 0.004241476766765118, + "learning_rate": 9.583202707897075e-06, + "loss": 1.2571580410003662, + "step": 48 + }, + { + "epoch": 0.6968888888888889, + "grad_norm": 0.004200824070721865, + "learning_rate": 9.549711614523007e-06, + "loss": 1.2616808414459229, + "step": 49 + }, + { + "epoch": 0.7111111111111111, + "grad_norm": 0.004399280063807964, + "learning_rate": 9.514989665596114e-06, + "loss": 1.263854742050171, + "step": 50 + }, + { + "epoch": 0.7111111111111111, + "eval_loss": 1.248613715171814, + "eval_runtime": 54.7705, + "eval_samples_per_second": 3.652, + "eval_steps_per_second": 1.826, + "step": 50 + }, + { + "epoch": 0.7253333333333334, + "grad_norm": 0.004282557871192694, + "learning_rate": 9.479046254613673e-06, + "loss": 1.279876947402954, + "step": 51 + }, + { + "epoch": 0.7395555555555555, + "grad_norm": 0.004135944880545139, + "learning_rate": 9.441891105521005e-06, + "loss": 1.2600014209747314, + "step": 52 + }, + { + "epoch": 0.7537777777777778, + "grad_norm": 0.0039484258741140366, + "learning_rate": 9.40353427008083e-06, + "loss": 1.2702335119247437, + "step": 53 + }, + { + "epoch": 0.768, + "grad_norm": 0.004160050302743912, + "learning_rate": 9.3639861251539e-06, + "loss": 1.2441630363464355, + "step": 54 + }, + { + "epoch": 0.7822222222222223, + "grad_norm": 0.004111913964152336, + "learning_rate": 9.323257369891702e-06, + "loss": 1.2441895008087158, + "step": 55 + }, + { + "epoch": 0.7964444444444444, + "grad_norm": 0.004063262604176998, + "learning_rate": 9.281359022841966e-06, + "loss": 1.255817174911499, + "step": 56 + }, + { + "epoch": 0.8106666666666666, + "grad_norm": 0.003984581213444471, + "learning_rate": 9.238302418967757e-06, + "loss": 1.2536640167236328, + "step": 57 + }, + { + "epoch": 0.8248888888888889, + "grad_norm": 0.003996702842414379, + "learning_rate": 9.194099206580981e-06, + "loss": 1.2499892711639404, + "step": 58 + }, + { + "epoch": 0.8391111111111111, + "grad_norm": 0.0044861058704555035, + "learning_rate": 9.14876134419111e-06, + "loss": 1.248606562614441, + "step": 59 + }, + { + "epoch": 0.8533333333333334, + "grad_norm": 0.004101422615349293, + "learning_rate": 9.102301097269974e-06, + "loss": 1.247856616973877, + "step": 60 + }, + { + "epoch": 0.8675555555555555, + "grad_norm": 0.004009375814348459, + "learning_rate": 9.05473103493355e-06, + "loss": 1.257920742034912, + "step": 61 + }, + { + "epoch": 0.8817777777777778, + "grad_norm": 0.0039364611729979515, + "learning_rate": 9.006064026541549e-06, + "loss": 1.2761049270629883, + "step": 62 + }, + { + "epoch": 0.896, + "grad_norm": 0.004427333828061819, + "learning_rate": 8.956313238215824e-06, + "loss": 1.245169997215271, + "step": 63 + }, + { + "epoch": 0.9102222222222223, + "grad_norm": 0.004076611250638962, + "learning_rate": 8.905492129278478e-06, + "loss": 1.234323263168335, + "step": 64 + }, + { + "epoch": 0.9244444444444444, + "grad_norm": 0.004032611381262541, + "learning_rate": 8.85361444861063e-06, + "loss": 1.2250256538391113, + "step": 65 + }, + { + "epoch": 0.9386666666666666, + "grad_norm": 0.004036127123981714, + "learning_rate": 8.800694230932885e-06, + "loss": 1.2339348793029785, + "step": 66 + }, + { + "epoch": 0.9528888888888889, + "grad_norm": 0.004113379865884781, + "learning_rate": 8.74674579300843e-06, + "loss": 1.252026081085205, + "step": 67 + }, + { + "epoch": 0.9671111111111111, + "grad_norm": 0.004064187407493591, + "learning_rate": 8.691783729769874e-06, + "loss": 1.2473349571228027, + "step": 68 + }, + { + "epoch": 0.9813333333333333, + "grad_norm": 0.0040993476286530495, + "learning_rate": 8.635822910370793e-06, + "loss": 1.2227911949157715, + "step": 69 + }, + { + "epoch": 0.9955555555555555, + "grad_norm": 0.003903420874848962, + "learning_rate": 8.578878474163115e-06, + "loss": 1.2343428134918213, + "step": 70 + }, + { + "epoch": 1.0, + "grad_norm": 0.006894194986671209, + "learning_rate": 8.520965826601394e-06, + "loss": 1.2428948879241943, + "step": 71 + } + ], + "logging_steps": 1, + "max_steps": 213, + "num_input_tokens_seen": 0, + "num_train_epochs": 3, + "save_steps": 500, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 1.5051322805138227e+17, + "train_batch_size": 1, + "trial_name": null, + "trial_params": null +} diff --git a/qwen2.5-1.5b-instruct/checkpoint-71/training_args.bin b/qwen2.5-1.5b-instruct/checkpoint-71/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..9fe3caacd835eb5bd003416a09e775b3c6032cff --- /dev/null +++ b/qwen2.5-1.5b-instruct/checkpoint-71/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:439cfa655d37de4fb57f512de1d989fa4287c7c18d86c9d7c320a2bd5256fe19 +size 5649 diff --git a/qwen2.5-1.5b-instruct/config.json b/qwen2.5-1.5b-instruct/config.json new file mode 100644 index 0000000000000000000000000000000000000000..499065fd744a64fc8ad2884da26e85a6e99d468a --- /dev/null +++ b/qwen2.5-1.5b-instruct/config.json @@ -0,0 +1,61 @@ +{ + "architectures": [ + "Qwen2ForCausalLM" + ], + "attention_dropout": 0.0, + "bos_token_id": null, + "dtype": "float32", + "eos_token_id": 151645, + "hidden_act": "silu", + "hidden_size": 1536, + "initializer_range": 0.02, + "intermediate_size": 8960, + "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" + ], + "max_position_embeddings": 32768, + "max_window_layers": 21, + "model_type": "qwen2", + "num_attention_heads": 12, + "num_hidden_layers": 28, + "num_key_value_heads": 2, + "pad_token_id": 151643, + "rms_norm_eps": 1e-06, + "rope_parameters": { + "rope_theta": 1000000.0, + "rope_type": "default" + }, + "sliding_window": null, + "tie_word_embeddings": true, + "transformers_version": "5.2.0", + "use_cache": false, + "use_sliding_window": false, + "vocab_size": 151936 +} diff --git a/qwen2.5-1.5b-instruct/eval_results.json b/qwen2.5-1.5b-instruct/eval_results.json new file mode 100644 index 0000000000000000000000000000000000000000..0e81088543b8b758d83a792f6c676d9f5775fb08 --- /dev/null +++ b/qwen2.5-1.5b-instruct/eval_results.json @@ -0,0 +1,7 @@ +{ + "epoch": 3.0, + "eval_loss": 1.1803252696990967, + "eval_runtime": 126.8683, + "eval_samples_per_second": 1.576, + "eval_steps_per_second": 0.788 +} \ No newline at end of file diff --git a/qwen2.5-1.5b-instruct/generation_config.json b/qwen2.5-1.5b-instruct/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..c5b4e11411fed3e7a5e6557165f888a7b0b1248b --- /dev/null +++ b/qwen2.5-1.5b-instruct/generation_config.json @@ -0,0 +1,13 @@ +{ + "do_sample": true, + "eos_token_id": [ + 151645, + 151643 + ], + "pad_token_id": 151643, + "repetition_penalty": 1.1, + "temperature": 0.7, + "top_k": 20, + "top_p": 0.8, + "transformers_version": "5.2.0" +} diff --git a/qwen2.5-1.5b-instruct/model.safetensors b/qwen2.5-1.5b-instruct/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..7c9b6652e3790d6f8e2afa2bd0c78cb51bf27f05 --- /dev/null +++ b/qwen2.5-1.5b-instruct/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:670fc682f9aec84643a6aac7c761dc5caa0aae09dac649f0bed07488527fd569 +size 6174895536 diff --git a/qwen2.5-1.5b-instruct/tokenizer.json b/qwen2.5-1.5b-instruct/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..34510ff0037cd50428af467a17ead5a96140a32c --- /dev/null +++ b/qwen2.5-1.5b-instruct/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3fd169731d2cbde95e10bf356d66d5997fd885dd8dbb6fb4684da3f23b2585d8 +size 11421892 diff --git a/qwen2.5-1.5b-instruct/tokenizer_config.json b/qwen2.5-1.5b-instruct/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..145e2c7a68b109ae0253dda29fc2e66cb01b4b85 --- /dev/null +++ b/qwen2.5-1.5b-instruct/tokenizer_config.json @@ -0,0 +1,30 @@ +{ + "add_prefix_space": false, + "backend": "tokenizers", + "bos_token": null, + "clean_up_tokenization_spaces": false, + "eos_token": "<|im_end|>", + "errors": "replace", + "extra_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|>" + ], + "is_local": false, + "model_max_length": 131072, + "pad_token": "<|endoftext|>", + "padding_side": "right", + "split_special_tokens": false, + "tokenizer_class": "Qwen2Tokenizer", + "unk_token": null +} diff --git a/qwen2.5-1.5b-instruct/train_results.json b/qwen2.5-1.5b-instruct/train_results.json new file mode 100644 index 0000000000000000000000000000000000000000..5a54526dc0c979a99cf4c22251663e8f026d0706 --- /dev/null +++ b/qwen2.5-1.5b-instruct/train_results.json @@ -0,0 +1,8 @@ +{ + "epoch": 3.0, + "total_flos": 4.5153968281196954e+17, + "train_loss": 0.0, + "train_runtime": 4.5863, + "train_samples_per_second": 11774.243, + "train_steps_per_second": 46.443 +} \ No newline at end of file diff --git a/qwen2.5-1.5b-instruct/trainer_log.jsonl b/qwen2.5-1.5b-instruct/trainer_log.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..bd68ef8fa1b30298da98de79a09993924353d44e --- /dev/null +++ b/qwen2.5-1.5b-instruct/trainer_log.jsonl @@ -0,0 +1,219 @@ +{"current_steps": 1, "total_steps": 213, "loss": 1.6560986042022705, "lr": 0.0, "epoch": 0.014222222222222223, "percentage": 0.47, "elapsed_time": "0:00:47", "remaining_time": "2:48:24"} +{"current_steps": 2, "total_steps": 213, "loss": 1.681471347808838, "lr": 4.5454545454545457e-07, "epoch": 0.028444444444444446, "percentage": 0.94, "elapsed_time": "0:01:35", "remaining_time": "2:47:26"} +{"current_steps": 3, "total_steps": 213, "loss": 1.6596678495407104, "lr": 9.090909090909091e-07, "epoch": 0.042666666666666665, "percentage": 1.41, "elapsed_time": "0:02:23", "remaining_time": "2:46:52"} +{"current_steps": 4, "total_steps": 213, "loss": 1.634795904159546, "lr": 1.3636363636363636e-06, "epoch": 0.05688888888888889, "percentage": 1.88, "elapsed_time": "0:03:10", "remaining_time": "2:46:06"} +{"current_steps": 5, "total_steps": 213, "loss": 1.606718897819519, "lr": 1.8181818181818183e-06, "epoch": 0.07111111111111111, "percentage": 2.35, "elapsed_time": "0:03:58", "remaining_time": "2:45:26"} +{"current_steps": 6, "total_steps": 213, "loss": 1.5807676315307617, "lr": 2.2727272727272728e-06, "epoch": 0.08533333333333333, "percentage": 2.82, "elapsed_time": "0:04:46", "remaining_time": "2:44:40"} +{"current_steps": 7, "total_steps": 213, "loss": 1.5490249395370483, "lr": 2.7272727272727272e-06, "epoch": 0.09955555555555555, "percentage": 3.29, "elapsed_time": "0:05:32", "remaining_time": "2:43:17"} +{"current_steps": 8, "total_steps": 213, "loss": 1.5317950248718262, "lr": 3.181818181818182e-06, "epoch": 0.11377777777777778, "percentage": 3.76, "elapsed_time": "0:06:20", "remaining_time": "2:42:29"} +{"current_steps": 9, "total_steps": 213, "loss": 1.5076513290405273, "lr": 3.6363636363636366e-06, "epoch": 0.128, "percentage": 4.23, "elapsed_time": "0:07:08", "remaining_time": "2:41:41"} +{"current_steps": 10, "total_steps": 213, "loss": 1.4786231517791748, "lr": 4.0909090909090915e-06, "epoch": 0.14222222222222222, "percentage": 4.69, "elapsed_time": "0:07:55", "remaining_time": "2:40:57"} +{"current_steps": 11, "total_steps": 213, "loss": 1.469682216644287, "lr": 4.5454545454545455e-06, "epoch": 0.15644444444444444, "percentage": 5.16, "elapsed_time": "0:08:43", "remaining_time": "2:40:09"} +{"current_steps": 12, "total_steps": 213, "loss": 1.4563651084899902, "lr": 5e-06, "epoch": 0.17066666666666666, "percentage": 5.63, "elapsed_time": "0:09:30", "remaining_time": "2:39:13"} +{"current_steps": 13, "total_steps": 213, "loss": 1.4510159492492676, "lr": 5.4545454545454545e-06, "epoch": 0.18488888888888888, "percentage": 6.1, "elapsed_time": "0:10:17", "remaining_time": "2:38:22"} +{"current_steps": 14, "total_steps": 213, "loss": 1.440371036529541, "lr": 5.90909090909091e-06, "epoch": 0.1991111111111111, "percentage": 6.57, "elapsed_time": "0:11:04", "remaining_time": "2:37:32"} +{"current_steps": 15, "total_steps": 213, "loss": 1.4149150848388672, "lr": 6.363636363636364e-06, "epoch": 0.21333333333333335, "percentage": 7.04, "elapsed_time": "0:11:52", "remaining_time": "2:36:50"} +{"current_steps": 16, "total_steps": 213, "loss": 1.3951767683029175, "lr": 6.818181818181818e-06, "epoch": 0.22755555555555557, "percentage": 7.51, "elapsed_time": "0:12:40", "remaining_time": "2:36:02"} +{"current_steps": 17, "total_steps": 213, "loss": 1.3866009712219238, "lr": 7.272727272727273e-06, "epoch": 0.24177777777777779, "percentage": 7.98, "elapsed_time": "0:13:27", "remaining_time": "2:35:13"} +{"current_steps": 18, "total_steps": 213, "loss": 1.3851287364959717, "lr": 7.727272727272727e-06, "epoch": 0.256, "percentage": 8.45, "elapsed_time": "0:14:15", "remaining_time": "2:34:26"} +{"current_steps": 19, "total_steps": 213, "loss": 1.3660919666290283, "lr": 8.181818181818183e-06, "epoch": 0.2702222222222222, "percentage": 8.92, "elapsed_time": "0:15:02", "remaining_time": "2:33:34"} +{"current_steps": 20, "total_steps": 213, "loss": 1.3991025686264038, "lr": 8.636363636363637e-06, "epoch": 0.28444444444444444, "percentage": 9.39, "elapsed_time": "0:15:50", "remaining_time": "2:32:50"} +{"current_steps": 21, "total_steps": 213, "loss": 1.3563741445541382, "lr": 9.090909090909091e-06, "epoch": 0.2986666666666667, "percentage": 9.86, "elapsed_time": "0:16:37", "remaining_time": "2:32:03"} +{"current_steps": 22, "total_steps": 213, "loss": 1.3432800769805908, "lr": 9.545454545454547e-06, "epoch": 0.3128888888888889, "percentage": 10.33, "elapsed_time": "0:17:24", "remaining_time": "2:31:09"} +{"current_steps": 23, "total_steps": 213, "loss": 1.3540418148040771, "lr": 1e-05, "epoch": 0.32711111111111113, "percentage": 10.8, "elapsed_time": "0:18:12", "remaining_time": "2:30:21"} +{"current_steps": 24, "total_steps": 213, "loss": 1.3571672439575195, "lr": 9.999323662872998e-06, "epoch": 0.3413333333333333, "percentage": 11.27, "elapsed_time": "0:18:59", "remaining_time": "2:29:34"} +{"current_steps": 25, "total_steps": 213, "loss": 1.3181281089782715, "lr": 9.99729483446475e-06, "epoch": 0.35555555555555557, "percentage": 11.74, "elapsed_time": "0:19:47", "remaining_time": "2:28:46"} +{"current_steps": 26, "total_steps": 213, "loss": 1.3311502933502197, "lr": 9.993914063644053e-06, "epoch": 0.36977777777777776, "percentage": 12.21, "elapsed_time": "0:20:34", "remaining_time": "2:28:02"} +{"current_steps": 27, "total_steps": 213, "loss": 1.3282787799835205, "lr": 9.989182265027232e-06, "epoch": 0.384, "percentage": 12.68, "elapsed_time": "0:21:23", "remaining_time": "2:27:19"} +{"current_steps": 28, "total_steps": 213, "loss": 1.325502872467041, "lr": 9.98310071873072e-06, "epoch": 0.3982222222222222, "percentage": 13.15, "elapsed_time": "0:22:09", "remaining_time": "2:26:27"} +{"current_steps": 29, "total_steps": 213, "loss": 1.3453574180603027, "lr": 9.975671070024741e-06, "epoch": 0.41244444444444445, "percentage": 13.62, "elapsed_time": "0:22:58", "remaining_time": "2:25:44"} +{"current_steps": 30, "total_steps": 213, "loss": 1.3257673978805542, "lr": 9.966895328888195e-06, "epoch": 0.4266666666666667, "percentage": 14.08, "elapsed_time": "0:23:45", "remaining_time": "2:24:56"} +{"current_steps": 31, "total_steps": 213, "loss": 1.317690372467041, "lr": 9.956775869464901e-06, "epoch": 0.4408888888888889, "percentage": 14.55, "elapsed_time": "0:24:33", "remaining_time": "2:24:11"} +{"current_steps": 32, "total_steps": 213, "loss": 1.3243460655212402, "lr": 9.945315429421307e-06, "epoch": 0.45511111111111113, "percentage": 15.02, "elapsed_time": "0:25:20", "remaining_time": "2:23:20"} +{"current_steps": 33, "total_steps": 213, "loss": 1.3194491863250732, "lr": 9.932517109205849e-06, "epoch": 0.4693333333333333, "percentage": 15.49, "elapsed_time": "0:26:08", "remaining_time": "2:22:34"} +{"current_steps": 34, "total_steps": 213, "loss": 1.3047914505004883, "lr": 9.918384371210178e-06, "epoch": 0.48355555555555557, "percentage": 15.96, "elapsed_time": "0:26:55", "remaining_time": "2:21:46"} +{"current_steps": 35, "total_steps": 213, "loss": 1.2955162525177002, "lr": 9.902921038832456e-06, "epoch": 0.49777777777777776, "percentage": 16.43, "elapsed_time": "0:27:43", "remaining_time": "2:20:59"} +{"current_steps": 36, "total_steps": 213, "loss": 1.3118610382080078, "lr": 9.886131295443003e-06, "epoch": 0.512, "percentage": 16.9, "elapsed_time": "0:28:31", "remaining_time": "2:20:13"} +{"current_steps": 37, "total_steps": 213, "loss": 1.302067518234253, "lr": 9.868019683252543e-06, "epoch": 0.5262222222222223, "percentage": 17.37, "elapsed_time": "0:29:18", "remaining_time": "2:19:26"} +{"current_steps": 38, "total_steps": 213, "loss": 1.2789689302444458, "lr": 9.848591102083375e-06, "epoch": 0.5404444444444444, "percentage": 17.84, "elapsed_time": "0:30:06", "remaining_time": "2:18:38"} +{"current_steps": 39, "total_steps": 213, "loss": 1.2827372550964355, "lr": 9.82785080804381e-06, "epoch": 0.5546666666666666, "percentage": 18.31, "elapsed_time": "0:30:54", "remaining_time": "2:17:54"} +{"current_steps": 40, "total_steps": 213, "loss": 1.2863564491271973, "lr": 9.805804412106197e-06, "epoch": 0.5688888888888889, "percentage": 18.78, "elapsed_time": "0:31:42", "remaining_time": "2:17:07"} +{"current_steps": 41, "total_steps": 213, "loss": 1.2768702507019043, "lr": 9.782457878588977e-06, "epoch": 0.5831111111111111, "percentage": 19.25, "elapsed_time": "0:32:30", "remaining_time": "2:16:22"} +{"current_steps": 42, "total_steps": 213, "loss": 1.2826354503631592, "lr": 9.75781752354311e-06, "epoch": 0.5973333333333334, "percentage": 19.72, "elapsed_time": "0:33:18", "remaining_time": "2:15:37"} +{"current_steps": 43, "total_steps": 213, "loss": 1.275444746017456, "lr": 9.731890013043367e-06, "epoch": 0.6115555555555555, "percentage": 20.19, "elapsed_time": "0:34:05", "remaining_time": "2:14:48"} +{"current_steps": 44, "total_steps": 213, "loss": 1.2660303115844727, "lr": 9.704682361384941e-06, "epoch": 0.6257777777777778, "percentage": 20.66, "elapsed_time": "0:34:52", "remaining_time": "2:13:58"} +{"current_steps": 45, "total_steps": 213, "loss": 1.2627251148223877, "lr": 9.676201929185809e-06, "epoch": 0.64, "percentage": 21.13, "elapsed_time": "0:35:40", "remaining_time": "2:13:10"} +{"current_steps": 46, "total_steps": 213, "loss": 1.2888470888137817, "lr": 9.646456421395447e-06, "epoch": 0.6542222222222223, "percentage": 21.6, "elapsed_time": "0:36:28", "remaining_time": "2:12:24"} +{"current_steps": 47, "total_steps": 213, "loss": 1.2826390266418457, "lr": 9.615453885210368e-06, "epoch": 0.6684444444444444, "percentage": 22.07, "elapsed_time": "0:37:16", "remaining_time": "2:11:38"} +{"current_steps": 48, "total_steps": 213, "loss": 1.2571580410003662, "lr": 9.583202707897075e-06, "epoch": 0.6826666666666666, "percentage": 22.54, "elapsed_time": "0:38:03", "remaining_time": "2:10:49"} +{"current_steps": 49, "total_steps": 213, "loss": 1.2616808414459229, "lr": 9.549711614523007e-06, "epoch": 0.6968888888888889, "percentage": 23.0, "elapsed_time": "0:38:50", "remaining_time": "2:10:00"} +{"current_steps": 50, "total_steps": 213, "loss": 1.263854742050171, "lr": 9.514989665596114e-06, "epoch": 0.7111111111111111, "percentage": 23.47, "elapsed_time": "0:39:38", "remaining_time": "2:09:14"} +{"current_steps": 50, "total_steps": 213, "eval_loss": 1.248613715171814, "epoch": 0.7111111111111111, "percentage": 23.47, "elapsed_time": "0:40:34", "remaining_time": "2:12:15"} +{"current_steps": 51, "total_steps": 213, "loss": 1.279876947402954, "lr": 9.479046254613673e-06, "epoch": 0.7253333333333334, "percentage": 23.94, "elapsed_time": "0:41:22", "remaining_time": "2:11:25"} +{"current_steps": 52, "total_steps": 213, "loss": 1.2600014209747314, "lr": 9.441891105521005e-06, "epoch": 0.7395555555555555, "percentage": 24.41, "elapsed_time": "0:42:10", "remaining_time": "2:10:34"} +{"current_steps": 53, "total_steps": 213, "loss": 1.2702335119247437, "lr": 9.40353427008083e-06, "epoch": 0.7537777777777778, "percentage": 24.88, "elapsed_time": "0:42:58", "remaining_time": "2:09:43"} +{"current_steps": 54, "total_steps": 213, "loss": 1.2441630363464355, "lr": 9.3639861251539e-06, "epoch": 0.768, "percentage": 25.35, "elapsed_time": "0:43:46", "remaining_time": "2:08:52"} +{"current_steps": 55, "total_steps": 213, "loss": 1.2441895008087158, "lr": 9.323257369891702e-06, "epoch": 0.7822222222222223, "percentage": 25.82, "elapsed_time": "0:44:33", "remaining_time": "2:08:00"} +{"current_steps": 56, "total_steps": 213, "loss": 1.255817174911499, "lr": 9.281359022841966e-06, "epoch": 0.7964444444444444, "percentage": 26.29, "elapsed_time": "0:45:21", "remaining_time": "2:07:11"} +{"current_steps": 57, "total_steps": 213, "loss": 1.2536640167236328, "lr": 9.238302418967757e-06, "epoch": 0.8106666666666666, "percentage": 26.76, "elapsed_time": "0:46:09", "remaining_time": "2:06:18"} +{"current_steps": 58, "total_steps": 213, "loss": 1.2499892711639404, "lr": 9.194099206580981e-06, "epoch": 0.8248888888888889, "percentage": 27.23, "elapsed_time": "0:46:56", "remaining_time": "2:05:26"} +{"current_steps": 59, "total_steps": 213, "loss": 1.248606562614441, "lr": 9.14876134419111e-06, "epoch": 0.8391111111111111, "percentage": 27.7, "elapsed_time": "0:47:43", "remaining_time": "2:04:34"} +{"current_steps": 60, "total_steps": 213, "loss": 1.247856616973877, "lr": 9.102301097269974e-06, "epoch": 0.8533333333333334, "percentage": 28.17, "elapsed_time": "0:48:30", "remaining_time": "2:03:42"} +{"current_steps": 61, "total_steps": 213, "loss": 1.257920742034912, "lr": 9.05473103493355e-06, "epoch": 0.8675555555555555, "percentage": 28.64, "elapsed_time": "0:49:18", "remaining_time": "2:02:51"} +{"current_steps": 62, "total_steps": 213, "loss": 1.2761049270629883, "lr": 9.006064026541549e-06, "epoch": 0.8817777777777778, "percentage": 29.11, "elapsed_time": "0:50:05", "remaining_time": "2:02:00"} +{"current_steps": 63, "total_steps": 213, "loss": 1.245169997215271, "lr": 8.956313238215824e-06, "epoch": 0.896, "percentage": 29.58, "elapsed_time": "0:50:53", "remaining_time": "2:01:11"} +{"current_steps": 64, "total_steps": 213, "loss": 1.234323263168335, "lr": 8.905492129278478e-06, "epoch": 0.9102222222222223, "percentage": 30.05, "elapsed_time": "0:51:41", "remaining_time": "2:00:20"} +{"current_steps": 65, "total_steps": 213, "loss": 1.2250256538391113, "lr": 8.85361444861063e-06, "epoch": 0.9244444444444444, "percentage": 30.52, "elapsed_time": "0:52:28", "remaining_time": "1:59:29"} +{"current_steps": 66, "total_steps": 213, "loss": 1.2339348793029785, "lr": 8.800694230932885e-06, "epoch": 0.9386666666666666, "percentage": 30.99, "elapsed_time": "0:53:16", "remaining_time": "1:58:39"} +{"current_steps": 67, "total_steps": 213, "loss": 1.252026081085205, "lr": 8.74674579300843e-06, "epoch": 0.9528888888888889, "percentage": 31.46, "elapsed_time": "0:54:03", "remaining_time": "1:57:48"} +{"current_steps": 68, "total_steps": 213, "loss": 1.2473349571228027, "lr": 8.691783729769874e-06, "epoch": 0.9671111111111111, "percentage": 31.92, "elapsed_time": "0:54:50", "remaining_time": "1:56:56"} +{"current_steps": 69, "total_steps": 213, "loss": 1.2227911949157715, "lr": 8.635822910370793e-06, "epoch": 0.9813333333333333, "percentage": 32.39, "elapsed_time": "0:55:38", "remaining_time": "1:56:08"} +{"current_steps": 70, "total_steps": 213, "loss": 1.2343428134918213, "lr": 8.578878474163115e-06, "epoch": 0.9955555555555555, "percentage": 32.86, "elapsed_time": "0:56:26", "remaining_time": "1:55:18"} +{"current_steps": 71, "total_steps": 213, "loss": 1.2428948879241943, "lr": 8.520965826601394e-06, "epoch": 1.0, "percentage": 33.33, "elapsed_time": "0:56:41", "remaining_time": "1:53:22"} +{"current_steps": 72, "total_steps": 213, "loss": 1.202027440071106, "lr": 8.462100635075097e-06, "epoch": 1.0142222222222221, "percentage": 33.8, "elapsed_time": "0:58:21", "remaining_time": "1:54:17"} +{"current_steps": 73, "total_steps": 213, "loss": 1.2224020957946777, "lr": 8.40229882467003e-06, "epoch": 1.0284444444444445, "percentage": 34.27, "elapsed_time": "0:59:08", "remaining_time": "1:53:26"} +{"current_steps": 74, "total_steps": 213, "loss": 1.2196844816207886, "lr": 8.341576573860049e-06, "epoch": 1.0426666666666666, "percentage": 34.74, "elapsed_time": "0:59:56", "remaining_time": "1:52:35"} +{"current_steps": 75, "total_steps": 213, "loss": 1.2222049236297607, "lr": 8.279950310130218e-06, "epoch": 1.056888888888889, "percentage": 35.21, "elapsed_time": "1:00:43", "remaining_time": "1:51:44"} +{"current_steps": 76, "total_steps": 213, "loss": 1.2127679586410522, "lr": 8.2174367055326e-06, "epoch": 1.0711111111111111, "percentage": 35.68, "elapsed_time": "1:01:31", "remaining_time": "1:50:53"} +{"current_steps": 77, "total_steps": 213, "loss": 1.1962798833847046, "lr": 8.154052672175888e-06, "epoch": 1.0853333333333333, "percentage": 36.15, "elapsed_time": "1:02:18", "remaining_time": "1:50:03"} +{"current_steps": 78, "total_steps": 213, "loss": 1.192876935005188, "lr": 8.08981535765009e-06, "epoch": 1.0995555555555556, "percentage": 36.62, "elapsed_time": "1:03:06", "remaining_time": "1:49:13"} +{"current_steps": 79, "total_steps": 213, "loss": 1.204209566116333, "lr": 8.024742140387506e-06, "epoch": 1.1137777777777778, "percentage": 37.09, "elapsed_time": "1:03:53", "remaining_time": "1:48:22"} +{"current_steps": 80, "total_steps": 213, "loss": 1.2061021327972412, "lr": 7.95885062496126e-06, "epoch": 1.1280000000000001, "percentage": 37.56, "elapsed_time": "1:04:41", "remaining_time": "1:47:32"} +{"current_steps": 81, "total_steps": 213, "loss": 1.2074682712554932, "lr": 7.892158637322647e-06, "epoch": 1.1422222222222222, "percentage": 38.03, "elapsed_time": "1:05:28", "remaining_time": "1:46:42"} +{"current_steps": 82, "total_steps": 213, "loss": 1.1797451972961426, "lr": 7.824684219978591e-06, "epoch": 1.1564444444444444, "percentage": 38.5, "elapsed_time": "1:06:15", "remaining_time": "1:45:50"} +{"current_steps": 83, "total_steps": 213, "loss": 1.2151604890823364, "lr": 7.756445627110523e-06, "epoch": 1.1706666666666667, "percentage": 38.97, "elapsed_time": "1:07:02", "remaining_time": "1:45:00"} +{"current_steps": 84, "total_steps": 213, "loss": 1.1917520761489868, "lr": 7.68746131963598e-06, "epoch": 1.1848888888888889, "percentage": 39.44, "elapsed_time": "1:07:49", "remaining_time": "1:44:10"} +{"current_steps": 85, "total_steps": 213, "loss": 1.1966304779052734, "lr": 7.6177499602143e-06, "epoch": 1.199111111111111, "percentage": 39.91, "elapsed_time": "1:08:36", "remaining_time": "1:43:19"} +{"current_steps": 86, "total_steps": 213, "loss": 1.2256059646606445, "lr": 7.547330408197695e-06, "epoch": 1.2133333333333334, "percentage": 40.38, "elapsed_time": "1:09:24", "remaining_time": "1:42:30"} +{"current_steps": 87, "total_steps": 213, "loss": 1.1984763145446777, "lr": 7.476221714529167e-06, "epoch": 1.2275555555555555, "percentage": 40.85, "elapsed_time": "1:10:12", "remaining_time": "1:41:40"} +{"current_steps": 88, "total_steps": 213, "loss": 1.1815118789672852, "lr": 7.404443116588548e-06, "epoch": 1.2417777777777779, "percentage": 41.31, "elapsed_time": "1:10:59", "remaining_time": "1:40:50"} +{"current_steps": 89, "total_steps": 213, "loss": 1.1813948154449463, "lr": 7.332014032988123e-06, "epoch": 1.256, "percentage": 41.78, "elapsed_time": "1:11:47", "remaining_time": "1:40:01"} +{"current_steps": 90, "total_steps": 213, "loss": 1.2083604335784912, "lr": 7.2589540583192165e-06, "epoch": 1.2702222222222221, "percentage": 42.25, "elapsed_time": "1:12:34", "remaining_time": "1:39:11"} +{"current_steps": 91, "total_steps": 213, "loss": 1.2008360624313354, "lr": 7.185282957851175e-06, "epoch": 1.2844444444444445, "percentage": 42.72, "elapsed_time": "1:13:21", "remaining_time": "1:38:21"} +{"current_steps": 92, "total_steps": 213, "loss": 1.1902484893798828, "lr": 7.111020662184174e-06, "epoch": 1.2986666666666666, "percentage": 43.19, "elapsed_time": "1:14:09", "remaining_time": "1:37:31"} +{"current_steps": 93, "total_steps": 213, "loss": 1.1865317821502686, "lr": 7.036187261857289e-06, "epoch": 1.3128888888888888, "percentage": 43.66, "elapsed_time": "1:14:56", "remaining_time": "1:36:41"} +{"current_steps": 94, "total_steps": 213, "loss": 1.1860058307647705, "lr": 6.960803001913315e-06, "epoch": 1.3271111111111111, "percentage": 44.13, "elapsed_time": "1:15:44", "remaining_time": "1:35:52"} +{"current_steps": 95, "total_steps": 213, "loss": 1.1842578649520874, "lr": 6.884888276421766e-06, "epoch": 1.3413333333333333, "percentage": 44.6, "elapsed_time": "1:16:31", "remaining_time": "1:35:03"} +{"current_steps": 96, "total_steps": 213, "loss": 1.1814080476760864, "lr": 6.8084636229615786e-06, "epoch": 1.3555555555555556, "percentage": 45.07, "elapsed_time": "1:17:20", "remaining_time": "1:34:15"} +{"current_steps": 97, "total_steps": 213, "loss": 1.193676233291626, "lr": 6.731549717064975e-06, "epoch": 1.3697777777777778, "percentage": 45.54, "elapsed_time": "1:18:07", "remaining_time": "1:33:25"} +{"current_steps": 98, "total_steps": 213, "loss": 1.2141278982162476, "lr": 6.654167366624009e-06, "epoch": 1.384, "percentage": 46.01, "elapsed_time": "1:18:55", "remaining_time": "1:32:36"} +{"current_steps": 99, "total_steps": 213, "loss": 1.1865811347961426, "lr": 6.576337506261314e-06, "epoch": 1.3982222222222223, "percentage": 46.48, "elapsed_time": "1:19:43", "remaining_time": "1:31:47"} +{"current_steps": 100, "total_steps": 213, "loss": 1.1954832077026367, "lr": 6.498081191666549e-06, "epoch": 1.4124444444444444, "percentage": 46.95, "elapsed_time": "1:20:31", "remaining_time": "1:30:59"} +{"current_steps": 100, "total_steps": 213, "eval_loss": 1.1988927125930786, "epoch": 1.4124444444444444, "percentage": 46.95, "elapsed_time": "1:20:42", "remaining_time": "1:31:12"} +{"current_steps": 101, "total_steps": 213, "loss": 1.1804523468017578, "lr": 6.419419593900109e-06, "epoch": 1.4266666666666667, "percentage": 47.42, "elapsed_time": "1:21:30", "remaining_time": "1:30:22"} +{"current_steps": 102, "total_steps": 213, "loss": 1.2068874835968018, "lr": 6.340373993665607e-06, "epoch": 1.4408888888888889, "percentage": 47.89, "elapsed_time": "1:22:17", "remaining_time": "1:29:33"} +{"current_steps": 103, "total_steps": 213, "loss": 1.1804969310760498, "lr": 6.2609657755527135e-06, "epoch": 1.455111111111111, "percentage": 48.36, "elapsed_time": "1:23:05", "remaining_time": "1:28:43"} +{"current_steps": 104, "total_steps": 213, "loss": 1.1953659057617188, "lr": 6.1812164222518626e-06, "epoch": 1.4693333333333334, "percentage": 48.83, "elapsed_time": "1:23:53", "remaining_time": "1:27:55"} +{"current_steps": 105, "total_steps": 213, "loss": 1.1933740377426147, "lr": 6.101147508742456e-06, "epoch": 1.4835555555555555, "percentage": 49.3, "elapsed_time": "1:24:41", "remaining_time": "1:27:06"} +{"current_steps": 106, "total_steps": 213, "loss": 1.201953411102295, "lr": 6.020780696456059e-06, "epoch": 1.4977777777777779, "percentage": 49.77, "elapsed_time": "1:25:28", "remaining_time": "1:26:16"} +{"current_steps": 107, "total_steps": 213, "loss": 1.1897786855697632, "lr": 5.940137727416247e-06, "epoch": 1.512, "percentage": 50.23, "elapsed_time": "1:26:16", "remaining_time": "1:25:27"} +{"current_steps": 108, "total_steps": 213, "loss": 1.1798899173736572, "lr": 5.859240418356614e-06, "epoch": 1.5262222222222221, "percentage": 50.7, "elapsed_time": "1:27:03", "remaining_time": "1:24:38"} +{"current_steps": 109, "total_steps": 213, "loss": 1.1853947639465332, "lr": 5.778110654818602e-06, "epoch": 1.5404444444444443, "percentage": 51.17, "elapsed_time": "1:27:52", "remaining_time": "1:23:50"} +{"current_steps": 110, "total_steps": 213, "loss": 1.1760761737823486, "lr": 5.696770385230679e-06, "epoch": 1.5546666666666666, "percentage": 51.64, "elapsed_time": "1:28:39", "remaining_time": "1:23:00"} +{"current_steps": 111, "total_steps": 213, "loss": 1.1980795860290527, "lr": 5.615241614970546e-06, "epoch": 1.568888888888889, "percentage": 52.11, "elapsed_time": "1:29:26", "remaining_time": "1:22:11"} +{"current_steps": 112, "total_steps": 213, "loss": 1.1878252029418945, "lr": 5.533546400411899e-06, "epoch": 1.5831111111111111, "percentage": 52.58, "elapsed_time": "1:30:14", "remaining_time": "1:21:22"} +{"current_steps": 113, "total_steps": 213, "loss": 1.1891889572143555, "lr": 5.4517068429574215e-06, "epoch": 1.5973333333333333, "percentage": 53.05, "elapsed_time": "1:31:02", "remaining_time": "1:20:33"} +{"current_steps": 114, "total_steps": 213, "loss": 1.1935696601867676, "lr": 5.3697450830595775e-06, "epoch": 1.6115555555555554, "percentage": 53.52, "elapsed_time": "1:31:49", "remaining_time": "1:19:44"} +{"current_steps": 115, "total_steps": 213, "loss": 1.1648902893066406, "lr": 5.287683294230855e-06, "epoch": 1.6257777777777778, "percentage": 53.99, "elapsed_time": "1:32:36", "remaining_time": "1:18:54"} +{"current_steps": 116, "total_steps": 213, "loss": 1.1749862432479858, "lr": 5.20554367704505e-06, "epoch": 1.6400000000000001, "percentage": 54.46, "elapsed_time": "1:33:23", "remaining_time": "1:18:05"} +{"current_steps": 117, "total_steps": 213, "loss": 1.200908899307251, "lr": 5.123348453131242e-06, "epoch": 1.6542222222222223, "percentage": 54.93, "elapsed_time": "1:34:12", "remaining_time": "1:17:17"} +{"current_steps": 118, "total_steps": 213, "loss": 1.189719796180725, "lr": 5.041119859162068e-06, "epoch": 1.6684444444444444, "percentage": 55.4, "elapsed_time": "1:34:59", "remaining_time": "1:16:28"} +{"current_steps": 119, "total_steps": 213, "loss": 1.1798235177993774, "lr": 4.958880140837934e-06, "epoch": 1.6826666666666665, "percentage": 55.87, "elapsed_time": "1:35:47", "remaining_time": "1:15:40"} +{"current_steps": 120, "total_steps": 213, "loss": 1.166198492050171, "lr": 4.876651546868759e-06, "epoch": 1.696888888888889, "percentage": 56.34, "elapsed_time": "1:36:34", "remaining_time": "1:14:50"} +{"current_steps": 121, "total_steps": 213, "loss": 1.1802079677581787, "lr": 4.794456322954953e-06, "epoch": 1.7111111111111112, "percentage": 56.81, "elapsed_time": "1:37:22", "remaining_time": "1:14:02"} +{"current_steps": 122, "total_steps": 213, "loss": 1.1807469129562378, "lr": 4.7123167057691446e-06, "epoch": 1.7253333333333334, "percentage": 57.28, "elapsed_time": "1:38:10", "remaining_time": "1:13:13"} +{"current_steps": 123, "total_steps": 213, "loss": 1.1979055404663086, "lr": 4.630254916940424e-06, "epoch": 1.7395555555555555, "percentage": 57.75, "elapsed_time": "1:38:57", "remaining_time": "1:12:24"} +{"current_steps": 124, "total_steps": 213, "loss": 1.192582130432129, "lr": 4.548293157042581e-06, "epoch": 1.7537777777777777, "percentage": 58.22, "elapsed_time": "1:39:45", "remaining_time": "1:11:36"} +{"current_steps": 125, "total_steps": 213, "loss": 1.1831550598144531, "lr": 4.466453599588103e-06, "epoch": 1.768, "percentage": 58.69, "elapsed_time": "1:40:33", "remaining_time": "1:10:47"} +{"current_steps": 126, "total_steps": 213, "loss": 1.1674884557724, "lr": 4.384758385029457e-06, "epoch": 1.7822222222222224, "percentage": 59.15, "elapsed_time": "1:41:20", "remaining_time": "1:09:58"} +{"current_steps": 127, "total_steps": 213, "loss": 1.181553602218628, "lr": 4.3032296147693225e-06, "epoch": 1.7964444444444445, "percentage": 59.62, "elapsed_time": "1:42:08", "remaining_time": "1:09:10"} +{"current_steps": 128, "total_steps": 213, "loss": 1.153355360031128, "lr": 4.2218893451814005e-06, "epoch": 1.8106666666666666, "percentage": 60.09, "elapsed_time": "1:42:55", "remaining_time": "1:08:20"} +{"current_steps": 129, "total_steps": 213, "loss": 1.1762878894805908, "lr": 4.140759581643386e-06, "epoch": 1.8248888888888888, "percentage": 60.56, "elapsed_time": "1:43:42", "remaining_time": "1:07:32"} +{"current_steps": 130, "total_steps": 213, "loss": 1.1792393922805786, "lr": 4.059862272583755e-06, "epoch": 1.8391111111111111, "percentage": 61.03, "elapsed_time": "1:44:30", "remaining_time": "1:06:43"} +{"current_steps": 131, "total_steps": 213, "loss": 1.1686198711395264, "lr": 3.979219303543942e-06, "epoch": 1.8533333333333335, "percentage": 61.5, "elapsed_time": "1:45:18", "remaining_time": "1:05:54"} +{"current_steps": 132, "total_steps": 213, "loss": 1.179172396659851, "lr": 3.898852491257547e-06, "epoch": 1.8675555555555556, "percentage": 61.97, "elapsed_time": "1:46:06", "remaining_time": "1:05:06"} +{"current_steps": 133, "total_steps": 213, "loss": 1.1667160987854004, "lr": 3.818783577748138e-06, "epoch": 1.8817777777777778, "percentage": 62.44, "elapsed_time": "1:46:53", "remaining_time": "1:04:17"} +{"current_steps": 134, "total_steps": 213, "loss": 1.1844608783721924, "lr": 3.7390342244472886e-06, "epoch": 1.896, "percentage": 62.91, "elapsed_time": "1:47:41", "remaining_time": "1:03:29"} +{"current_steps": 135, "total_steps": 213, "loss": 1.1989758014678955, "lr": 3.659626006334395e-06, "epoch": 1.9102222222222223, "percentage": 63.38, "elapsed_time": "1:48:28", "remaining_time": "1:02:40"} +{"current_steps": 136, "total_steps": 213, "loss": 1.1868371963500977, "lr": 3.5805804060998926e-06, "epoch": 1.9244444444444444, "percentage": 63.85, "elapsed_time": "1:49:16", "remaining_time": "1:01:52"} +{"current_steps": 137, "total_steps": 213, "loss": 1.166593313217163, "lr": 3.501918808333453e-06, "epoch": 1.9386666666666668, "percentage": 64.32, "elapsed_time": "1:50:04", "remaining_time": "1:01:03"} +{"current_steps": 138, "total_steps": 213, "loss": 1.1770471334457397, "lr": 3.4236624937386874e-06, "epoch": 1.952888888888889, "percentage": 64.79, "elapsed_time": "1:50:51", "remaining_time": "1:00:15"} +{"current_steps": 139, "total_steps": 213, "loss": 1.1796729564666748, "lr": 3.3458326333759927e-06, "epoch": 1.967111111111111, "percentage": 65.26, "elapsed_time": "1:51:39", "remaining_time": "0:59:26"} +{"current_steps": 140, "total_steps": 213, "loss": 1.1624293327331543, "lr": 3.268450282935026e-06, "epoch": 1.9813333333333332, "percentage": 65.73, "elapsed_time": "1:52:25", "remaining_time": "0:58:37"} +{"current_steps": 141, "total_steps": 213, "loss": 1.1687278747558594, "lr": 3.1915363770384223e-06, "epoch": 1.9955555555555555, "percentage": 66.2, "elapsed_time": "1:53:13", "remaining_time": "0:57:49"} +{"current_steps": 142, "total_steps": 213, "loss": 1.1975963115692139, "lr": 3.1151117235782346e-06, "epoch": 2.0, "percentage": 66.67, "elapsed_time": "1:53:28", "remaining_time": "0:56:44"} +{"current_steps": 143, "total_steps": 213, "loss": 1.1713850498199463, "lr": 3.0391969980866874e-06, "epoch": 2.014222222222222, "percentage": 67.14, "elapsed_time": "1:55:07", "remaining_time": "0:56:21"} +{"current_steps": 144, "total_steps": 213, "loss": 1.1817915439605713, "lr": 2.963812738142713e-06, "epoch": 2.0284444444444443, "percentage": 67.61, "elapsed_time": "1:55:54", "remaining_time": "0:55:32"} +{"current_steps": 145, "total_steps": 213, "loss": 1.1671693325042725, "lr": 2.8889793378158284e-06, "epoch": 2.042666666666667, "percentage": 68.08, "elapsed_time": "1:56:43", "remaining_time": "0:54:44"} +{"current_steps": 146, "total_steps": 213, "loss": 1.1553094387054443, "lr": 2.814717042148827e-06, "epoch": 2.056888888888889, "percentage": 68.54, "elapsed_time": "1:57:30", "remaining_time": "0:53:55"} +{"current_steps": 147, "total_steps": 213, "loss": 1.1611030101776123, "lr": 2.7410459416807856e-06, "epoch": 2.071111111111111, "percentage": 69.01, "elapsed_time": "1:58:18", "remaining_time": "0:53:06"} +{"current_steps": 148, "total_steps": 213, "loss": 1.169109582901001, "lr": 2.6679859670118785e-06, "epoch": 2.0853333333333333, "percentage": 69.48, "elapsed_time": "1:59:04", "remaining_time": "0:52:17"} +{"current_steps": 149, "total_steps": 213, "loss": 1.1539714336395264, "lr": 2.5955568834114523e-06, "epoch": 2.0995555555555554, "percentage": 69.95, "elapsed_time": "1:59:52", "remaining_time": "0:51:29"} +{"current_steps": 150, "total_steps": 213, "loss": 1.1314873695373535, "lr": 2.523778285470835e-06, "epoch": 2.113777777777778, "percentage": 70.42, "elapsed_time": "2:00:38", "remaining_time": "0:50:40"} +{"current_steps": 150, "total_steps": 213, "eval_loss": 1.1810542345046997, "epoch": 2.113777777777778, "percentage": 70.42, "elapsed_time": "2:01:55", "remaining_time": "0:51:12"} +{"current_steps": 151, "total_steps": 213, "loss": 1.1695539951324463, "lr": 2.452669591802307e-06, "epoch": 2.128, "percentage": 70.89, "elapsed_time": "2:02:43", "remaining_time": "0:50:23"} +{"current_steps": 152, "total_steps": 213, "loss": 1.1494641304016113, "lr": 2.3822500397857016e-06, "epoch": 2.1422222222222222, "percentage": 71.36, "elapsed_time": "2:03:31", "remaining_time": "0:49:34"} +{"current_steps": 153, "total_steps": 213, "loss": 1.1655616760253906, "lr": 2.3125386803640188e-06, "epoch": 2.1564444444444444, "percentage": 71.83, "elapsed_time": "2:04:19", "remaining_time": "0:48:45"} +{"current_steps": 154, "total_steps": 213, "loss": 1.1519088745117188, "lr": 2.243554372889479e-06, "epoch": 2.1706666666666665, "percentage": 72.3, "elapsed_time": "2:05:06", "remaining_time": "0:47:55"} +{"current_steps": 155, "total_steps": 213, "loss": 1.1479674577713013, "lr": 2.1753157800214107e-06, "epoch": 2.1848888888888887, "percentage": 72.77, "elapsed_time": "2:05:53", "remaining_time": "0:47:06"} +{"current_steps": 156, "total_steps": 213, "loss": 1.1504448652267456, "lr": 2.1078413626773547e-06, "epoch": 2.1991111111111112, "percentage": 73.24, "elapsed_time": "2:06:41", "remaining_time": "0:46:17"} +{"current_steps": 157, "total_steps": 213, "loss": 1.142012119293213, "lr": 2.0411493750387423e-06, "epoch": 2.2133333333333334, "percentage": 73.71, "elapsed_time": "2:07:28", "remaining_time": "0:45:28"} +{"current_steps": 158, "total_steps": 213, "loss": 1.1382803916931152, "lr": 1.9752578596124955e-06, "epoch": 2.2275555555555555, "percentage": 74.18, "elapsed_time": "2:08:15", "remaining_time": "0:44:38"} +{"current_steps": 159, "total_steps": 213, "loss": 1.1602444648742676, "lr": 1.9101846423499113e-06, "epoch": 2.2417777777777776, "percentage": 74.65, "elapsed_time": "2:09:03", "remaining_time": "0:43:49"} +{"current_steps": 160, "total_steps": 213, "loss": 1.146360158920288, "lr": 1.8459473278241125e-06, "epoch": 2.2560000000000002, "percentage": 75.12, "elapsed_time": "2:09:51", "remaining_time": "0:43:00"} +{"current_steps": 161, "total_steps": 213, "loss": 1.1420152187347412, "lr": 1.7825632944674016e-06, "epoch": 2.2702222222222224, "percentage": 75.59, "elapsed_time": "2:10:38", "remaining_time": "0:42:11"} +{"current_steps": 162, "total_steps": 213, "loss": 1.1756829023361206, "lr": 1.7200496898697832e-06, "epoch": 2.2844444444444445, "percentage": 76.06, "elapsed_time": "2:11:26", "remaining_time": "0:41:22"} +{"current_steps": 163, "total_steps": 213, "loss": 1.1596589088439941, "lr": 1.6584234261399535e-06, "epoch": 2.2986666666666666, "percentage": 76.53, "elapsed_time": "2:12:14", "remaining_time": "0:40:33"} +{"current_steps": 164, "total_steps": 213, "loss": 1.1395504474639893, "lr": 1.5977011753299726e-06, "epoch": 2.3128888888888888, "percentage": 77.0, "elapsed_time": "2:13:02", "remaining_time": "0:39:44"} +{"current_steps": 165, "total_steps": 213, "loss": 1.1680355072021484, "lr": 1.5378993649249053e-06, "epoch": 2.327111111111111, "percentage": 77.46, "elapsed_time": "2:13:50", "remaining_time": "0:38:56"} +{"current_steps": 166, "total_steps": 213, "loss": 1.1566658020019531, "lr": 1.4790341733986085e-06, "epoch": 2.3413333333333335, "percentage": 77.93, "elapsed_time": "2:14:38", "remaining_time": "0:38:07"} +{"current_steps": 167, "total_steps": 213, "loss": 1.1633788347244263, "lr": 1.4211215258368866e-06, "epoch": 2.3555555555555556, "percentage": 78.4, "elapsed_time": "2:15:25", "remaining_time": "0:37:18"} +{"current_steps": 168, "total_steps": 213, "loss": 1.1514759063720703, "lr": 1.3641770896292083e-06, "epoch": 2.3697777777777778, "percentage": 78.87, "elapsed_time": "2:16:11", "remaining_time": "0:36:28"} +{"current_steps": 169, "total_steps": 213, "loss": 1.1515188217163086, "lr": 1.3082162702301276e-06, "epoch": 2.384, "percentage": 79.34, "elapsed_time": "2:16:59", "remaining_time": "0:35:39"} +{"current_steps": 170, "total_steps": 213, "loss": 1.1591559648513794, "lr": 1.2532542069915722e-06, "epoch": 2.398222222222222, "percentage": 79.81, "elapsed_time": "2:17:46", "remaining_time": "0:34:51"} +{"current_steps": 171, "total_steps": 213, "loss": 1.1669801473617554, "lr": 1.1993057690671174e-06, "epoch": 2.4124444444444446, "percentage": 80.28, "elapsed_time": "2:18:34", "remaining_time": "0:34:02"} +{"current_steps": 172, "total_steps": 213, "loss": 1.155794382095337, "lr": 1.1463855513893695e-06, "epoch": 2.4266666666666667, "percentage": 80.75, "elapsed_time": "2:19:21", "remaining_time": "0:33:13"} +{"current_steps": 173, "total_steps": 213, "loss": 1.1434683799743652, "lr": 1.0945078707215224e-06, "epoch": 2.440888888888889, "percentage": 81.22, "elapsed_time": "2:20:08", "remaining_time": "0:32:24"} +{"current_steps": 174, "total_steps": 213, "loss": 1.1586582660675049, "lr": 1.0436867617841768e-06, "epoch": 2.455111111111111, "percentage": 81.69, "elapsed_time": "2:20:56", "remaining_time": "0:31:35"} +{"current_steps": 175, "total_steps": 213, "loss": 1.1643059253692627, "lr": 9.939359734584552e-07, "epoch": 2.469333333333333, "percentage": 82.16, "elapsed_time": "2:21:44", "remaining_time": "0:30:46"} +{"current_steps": 176, "total_steps": 213, "loss": 1.137399673461914, "lr": 9.452689650664515e-07, "epoch": 2.4835555555555557, "percentage": 82.63, "elapsed_time": "2:22:31", "remaining_time": "0:29:57"} +{"current_steps": 177, "total_steps": 213, "loss": 1.150512456893921, "lr": 8.976989027300265e-07, "epoch": 2.497777777777778, "percentage": 83.1, "elapsed_time": "2:23:19", "remaining_time": "0:29:09"} +{"current_steps": 178, "total_steps": 213, "loss": 1.1556127071380615, "lr": 8.512386558088919e-07, "epoch": 2.512, "percentage": 83.57, "elapsed_time": "2:24:06", "remaining_time": "0:28:20"} +{"current_steps": 179, "total_steps": 213, "loss": 1.1506540775299072, "lr": 8.059007934190194e-07, "epoch": 2.526222222222222, "percentage": 84.04, "elapsed_time": "2:24:53", "remaining_time": "0:27:31"} +{"current_steps": 180, "total_steps": 213, "loss": 1.1606539487838745, "lr": 7.61697581032243e-07, "epoch": 2.5404444444444443, "percentage": 84.51, "elapsed_time": "2:25:41", "remaining_time": "0:26:42"} +{"current_steps": 181, "total_steps": 213, "loss": 1.145754098892212, "lr": 7.186409771580355e-07, "epoch": 2.554666666666667, "percentage": 84.98, "elapsed_time": "2:26:28", "remaining_time": "0:25:53"} +{"current_steps": 182, "total_steps": 213, "loss": 1.1506030559539795, "lr": 6.76742630108298e-07, "epoch": 2.568888888888889, "percentage": 85.45, "elapsed_time": "2:27:15", "remaining_time": "0:25:04"} +{"current_steps": 183, "total_steps": 213, "loss": 1.1366521120071411, "lr": 6.360138748461015e-07, "epoch": 2.583111111111111, "percentage": 85.92, "elapsed_time": "2:28:02", "remaining_time": "0:24:16"} +{"current_steps": 184, "total_steps": 213, "loss": 1.1723542213439941, "lr": 5.964657299191712e-07, "epoch": 2.5973333333333333, "percentage": 86.38, "elapsed_time": "2:28:51", "remaining_time": "0:23:27"} +{"current_steps": 185, "total_steps": 213, "loss": 1.1543989181518555, "lr": 5.581088944789953e-07, "epoch": 2.6115555555555554, "percentage": 86.85, "elapsed_time": "2:29:39", "remaining_time": "0:22:39"} +{"current_steps": 186, "total_steps": 213, "loss": 1.1481022834777832, "lr": 5.20953745386329e-07, "epoch": 2.6257777777777775, "percentage": 87.32, "elapsed_time": "2:30:27", "remaining_time": "0:21:50"} +{"current_steps": 187, "total_steps": 213, "loss": 1.1510019302368164, "lr": 4.850103344038853e-07, "epoch": 2.64, "percentage": 87.79, "elapsed_time": "2:31:15", "remaining_time": "0:21:01"} +{"current_steps": 188, "total_steps": 213, "loss": 1.1454750299453735, "lr": 4.502883854769935e-07, "epoch": 2.6542222222222223, "percentage": 88.26, "elapsed_time": "2:32:02", "remaining_time": "0:20:13"} +{"current_steps": 189, "total_steps": 213, "loss": 1.1555531024932861, "lr": 4.167972921029262e-07, "epoch": 2.6684444444444444, "percentage": 88.73, "elapsed_time": "2:32:50", "remaining_time": "0:19:24"} +{"current_steps": 190, "total_steps": 213, "loss": 1.141042709350586, "lr": 3.8454611478963235e-07, "epoch": 2.6826666666666665, "percentage": 89.2, "elapsed_time": "2:33:38", "remaining_time": "0:18:35"} +{"current_steps": 191, "total_steps": 213, "loss": 1.1597721576690674, "lr": 3.535435786045538e-07, "epoch": 2.696888888888889, "percentage": 89.67, "elapsed_time": "2:34:25", "remaining_time": "0:17:47"} +{"current_steps": 192, "total_steps": 213, "loss": 1.1370229721069336, "lr": 3.237980708141919e-07, "epoch": 2.7111111111111112, "percentage": 90.14, "elapsed_time": "2:35:12", "remaining_time": "0:16:58"} +{"current_steps": 193, "total_steps": 213, "loss": 1.1386423110961914, "lr": 2.9531763861505967e-07, "epoch": 2.7253333333333334, "percentage": 90.61, "elapsed_time": "2:36:00", "remaining_time": "0:16:10"} +{"current_steps": 194, "total_steps": 213, "loss": 1.141430139541626, "lr": 2.681099869566328e-07, "epoch": 2.7395555555555555, "percentage": 91.08, "elapsed_time": "2:36:48", "remaining_time": "0:15:21"} +{"current_steps": 195, "total_steps": 213, "loss": 1.1269652843475342, "lr": 2.4218247645689306e-07, "epoch": 2.7537777777777777, "percentage": 91.55, "elapsed_time": "2:37:34", "remaining_time": "0:14:32"} +{"current_steps": 196, "total_steps": 213, "loss": 1.1652940511703491, "lr": 2.175421214110235e-07, "epoch": 2.768, "percentage": 92.02, "elapsed_time": "2:38:22", "remaining_time": "0:13:44"} +{"current_steps": 197, "total_steps": 213, "loss": 1.1484644412994385, "lr": 1.941955878938029e-07, "epoch": 2.7822222222222224, "percentage": 92.49, "elapsed_time": "2:39:09", "remaining_time": "0:12:55"} +{"current_steps": 198, "total_steps": 213, "loss": 1.1459333896636963, "lr": 1.721491919561913e-07, "epoch": 2.7964444444444445, "percentage": 92.96, "elapsed_time": "2:39:56", "remaining_time": "0:12:07"} +{"current_steps": 199, "total_steps": 213, "loss": 1.157778024673462, "lr": 1.514088979166256e-07, "epoch": 2.8106666666666666, "percentage": 93.43, "elapsed_time": "2:40:43", "remaining_time": "0:11:18"} +{"current_steps": 200, "total_steps": 213, "loss": 1.1373133659362793, "lr": 1.3198031674745814e-07, "epoch": 2.824888888888889, "percentage": 93.9, "elapsed_time": "2:41:30", "remaining_time": "0:10:29"} +{"current_steps": 200, "total_steps": 213, "eval_loss": 1.1767549514770508, "epoch": 2.824888888888889, "percentage": 93.9, "elapsed_time": "2:41:48", "remaining_time": "0:10:31"} +{"current_steps": 201, "total_steps": 213, "loss": 1.1472009420394897, "lr": 1.138687045569975e-07, "epoch": 2.8391111111111114, "percentage": 94.37, "elapsed_time": "2:42:36", "remaining_time": "0:09:42"} +{"current_steps": 202, "total_steps": 213, "loss": 1.1417003870010376, "lr": 9.70789611675449e-08, "epoch": 2.8533333333333335, "percentage": 94.84, "elapsed_time": "2:43:23", "remaining_time": "0:08:53"} +{"current_steps": 203, "total_steps": 213, "loss": 1.1214667558670044, "lr": 8.161562878982399e-08, "epoch": 2.8675555555555556, "percentage": 95.31, "elapsed_time": "2:44:10", "remaining_time": "0:08:05"} +{"current_steps": 204, "total_steps": 213, "loss": 1.138871431350708, "lr": 6.74828907941516e-08, "epoch": 2.8817777777777778, "percentage": 95.77, "elapsed_time": "2:44:58", "remaining_time": "0:07:16"} +{"current_steps": 205, "total_steps": 213, "loss": 1.1596763134002686, "lr": 5.468457057869358e-08, "epoch": 2.896, "percentage": 96.24, "elapsed_time": "2:45:46", "remaining_time": "0:06:28"} +{"current_steps": 206, "total_steps": 213, "loss": 1.1474013328552246, "lr": 4.322413053509944e-08, "epoch": 2.910222222222222, "percentage": 96.71, "elapsed_time": "2:46:34", "remaining_time": "0:05:39"} +{"current_steps": 207, "total_steps": 213, "loss": 1.1420172452926636, "lr": 3.31046711118066e-08, "epoch": 2.924444444444444, "percentage": 97.18, "elapsed_time": "2:47:21", "remaining_time": "0:04:51"} +{"current_steps": 208, "total_steps": 213, "loss": 1.149836778640747, "lr": 2.4328929975260262e-08, "epoch": 2.9386666666666668, "percentage": 97.65, "elapsed_time": "2:48:08", "remaining_time": "0:04:02"} +{"current_steps": 209, "total_steps": 213, "loss": 1.142824649810791, "lr": 1.6899281269279756e-08, "epoch": 2.952888888888889, "percentage": 98.12, "elapsed_time": "2:48:55", "remaining_time": "0:03:13"} +{"current_steps": 210, "total_steps": 213, "loss": 1.1392970085144043, "lr": 1.0817734972768946e-08, "epoch": 2.967111111111111, "percentage": 98.59, "elapsed_time": "2:49:42", "remaining_time": "0:02:25"} +{"current_steps": 211, "total_steps": 213, "loss": 1.1546673774719238, "lr": 6.085936355947897e-09, "epoch": 2.981333333333333, "percentage": 99.06, "elapsed_time": "2:50:30", "remaining_time": "0:01:36"} +{"current_steps": 212, "total_steps": 213, "loss": 1.1519522666931152, "lr": 2.7051655352494654e-09, "epoch": 2.9955555555555557, "percentage": 99.53, "elapsed_time": "2:51:17", "remaining_time": "0:00:48"} +{"current_steps": 213, "total_steps": 213, "loss": 1.1223903894424438, "lr": 6.763371270035457e-10, "epoch": 3.0, "percentage": 100.0, "elapsed_time": "2:51:32", "remaining_time": "0:00:00"} +{"current_steps": 213, "total_steps": 213, "epoch": 3.0, "percentage": 100.0, "elapsed_time": "2:52:22", "remaining_time": "0:00:00"} +{"current_steps": 213, "total_steps": 213, "epoch": 3.0, "percentage": 100.0, "elapsed_time": "0:00:00", "remaining_time": "0:00:00"} diff --git a/qwen2.5-1.5b-instruct/trainer_state.json b/qwen2.5-1.5b-instruct/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..0b00a33d020b277e6912a69e713efeafad0cea4c --- /dev/null +++ b/qwen2.5-1.5b-instruct/trainer_state.json @@ -0,0 +1,1566 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 3.0, + "eval_steps": 50, + "global_step": 213, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.014222222222222223, + "grad_norm": 0.036850932985544205, + "learning_rate": 0.0, + "loss": 1.6560986042022705, + "step": 1 + }, + { + "epoch": 0.028444444444444446, + "grad_norm": 0.037402428686618805, + "learning_rate": 4.5454545454545457e-07, + "loss": 1.681471347808838, + "step": 2 + }, + { + "epoch": 0.042666666666666665, + "grad_norm": 0.03194897621870041, + "learning_rate": 9.090909090909091e-07, + "loss": 1.6596678495407104, + "step": 3 + }, + { + "epoch": 0.05688888888888889, + "grad_norm": 0.02700614556670189, + "learning_rate": 1.3636363636363636e-06, + "loss": 1.634795904159546, + "step": 4 + }, + { + "epoch": 0.07111111111111111, + "grad_norm": 0.021732285618782043, + "learning_rate": 1.8181818181818183e-06, + "loss": 1.606718897819519, + "step": 5 + }, + { + "epoch": 0.08533333333333333, + "grad_norm": 0.018985262140631676, + "learning_rate": 2.2727272727272728e-06, + "loss": 1.5807676315307617, + "step": 6 + }, + { + "epoch": 0.09955555555555555, + "grad_norm": 0.018287895247340202, + "learning_rate": 2.7272727272727272e-06, + "loss": 1.5490249395370483, + "step": 7 + }, + { + "epoch": 0.11377777777777778, + "grad_norm": 0.017051130533218384, + "learning_rate": 3.181818181818182e-06, + "loss": 1.5317950248718262, + "step": 8 + }, + { + "epoch": 0.128, + "grad_norm": 0.015596743673086166, + "learning_rate": 3.6363636363636366e-06, + "loss": 1.5076513290405273, + "step": 9 + }, + { + "epoch": 0.14222222222222222, + "grad_norm": 0.014579818584024906, + "learning_rate": 4.0909090909090915e-06, + "loss": 1.4786231517791748, + "step": 10 + }, + { + "epoch": 0.15644444444444444, + "grad_norm": 0.012900307774543762, + "learning_rate": 4.5454545454545455e-06, + "loss": 1.469682216644287, + "step": 11 + }, + { + "epoch": 0.17066666666666666, + "grad_norm": 0.011142965406179428, + "learning_rate": 5e-06, + "loss": 1.4563651084899902, + "step": 12 + }, + { + "epoch": 0.18488888888888888, + "grad_norm": 0.011200510896742344, + "learning_rate": 5.4545454545454545e-06, + "loss": 1.4510159492492676, + "step": 13 + }, + { + "epoch": 0.1991111111111111, + "grad_norm": 0.01117027923464775, + "learning_rate": 5.90909090909091e-06, + "loss": 1.440371036529541, + "step": 14 + }, + { + "epoch": 0.21333333333333335, + "grad_norm": 0.010404860600829124, + "learning_rate": 6.363636363636364e-06, + "loss": 1.4149150848388672, + "step": 15 + }, + { + "epoch": 0.22755555555555557, + "grad_norm": 0.009069642052054405, + "learning_rate": 6.818181818181818e-06, + "loss": 1.3951767683029175, + "step": 16 + }, + { + "epoch": 0.24177777777777779, + "grad_norm": 0.008563642390072346, + "learning_rate": 7.272727272727273e-06, + "loss": 1.3866009712219238, + "step": 17 + }, + { + "epoch": 0.256, + "grad_norm": 0.008158586919307709, + "learning_rate": 7.727272727272727e-06, + "loss": 1.3851287364959717, + "step": 18 + }, + { + "epoch": 0.2702222222222222, + "grad_norm": 0.008019132539629936, + "learning_rate": 8.181818181818183e-06, + "loss": 1.3660919666290283, + "step": 19 + }, + { + "epoch": 0.28444444444444444, + "grad_norm": 0.007505647838115692, + "learning_rate": 8.636363636363637e-06, + "loss": 1.3991025686264038, + "step": 20 + }, + { + "epoch": 0.2986666666666667, + "grad_norm": 0.0068238647654652596, + "learning_rate": 9.090909090909091e-06, + "loss": 1.3563741445541382, + "step": 21 + }, + { + "epoch": 0.3128888888888889, + "grad_norm": 0.00681255292147398, + "learning_rate": 9.545454545454547e-06, + "loss": 1.3432800769805908, + "step": 22 + }, + { + "epoch": 0.32711111111111113, + "grad_norm": 0.0064102653414011, + "learning_rate": 1e-05, + "loss": 1.3540418148040771, + "step": 23 + }, + { + "epoch": 0.3413333333333333, + "grad_norm": 0.006286031566560268, + "learning_rate": 9.999323662872998e-06, + "loss": 1.3571672439575195, + "step": 24 + }, + { + "epoch": 0.35555555555555557, + "grad_norm": 0.006193141452968121, + "learning_rate": 9.99729483446475e-06, + "loss": 1.3181281089782715, + "step": 25 + }, + { + "epoch": 0.36977777777777776, + "grad_norm": 0.005919129587709904, + "learning_rate": 9.993914063644053e-06, + "loss": 1.3311502933502197, + "step": 26 + }, + { + "epoch": 0.384, + "grad_norm": 0.005563116166740656, + "learning_rate": 9.989182265027232e-06, + "loss": 1.3282787799835205, + "step": 27 + }, + { + "epoch": 0.3982222222222222, + "grad_norm": 0.005611867178231478, + "learning_rate": 9.98310071873072e-06, + "loss": 1.325502872467041, + "step": 28 + }, + { + "epoch": 0.41244444444444445, + "grad_norm": 0.0053426348604261875, + "learning_rate": 9.975671070024741e-06, + "loss": 1.3453574180603027, + "step": 29 + }, + { + "epoch": 0.4266666666666667, + "grad_norm": 0.0058601596392691135, + "learning_rate": 9.966895328888195e-06, + "loss": 1.3257673978805542, + "step": 30 + }, + { + "epoch": 0.4408888888888889, + "grad_norm": 0.005306159611791372, + "learning_rate": 9.956775869464901e-06, + "loss": 1.317690372467041, + "step": 31 + }, + { + "epoch": 0.45511111111111113, + "grad_norm": 0.0048169931396842, + "learning_rate": 9.945315429421307e-06, + "loss": 1.3243460655212402, + "step": 32 + }, + { + "epoch": 0.4693333333333333, + "grad_norm": 0.0048377541825175285, + "learning_rate": 9.932517109205849e-06, + "loss": 1.3194491863250732, + "step": 33 + }, + { + "epoch": 0.48355555555555557, + "grad_norm": 0.004818843677639961, + "learning_rate": 9.918384371210178e-06, + "loss": 1.3047914505004883, + "step": 34 + }, + { + "epoch": 0.49777777777777776, + "grad_norm": 0.005093662533909082, + "learning_rate": 9.902921038832456e-06, + "loss": 1.2955162525177002, + "step": 35 + }, + { + "epoch": 0.512, + "grad_norm": 0.005024498328566551, + "learning_rate": 9.886131295443003e-06, + "loss": 1.3118610382080078, + "step": 36 + }, + { + "epoch": 0.5262222222222223, + "grad_norm": 0.004711818415671587, + "learning_rate": 9.868019683252543e-06, + "loss": 1.302067518234253, + "step": 37 + }, + { + "epoch": 0.5404444444444444, + "grad_norm": 0.0047568087466061115, + "learning_rate": 9.848591102083375e-06, + "loss": 1.2789689302444458, + "step": 38 + }, + { + "epoch": 0.5546666666666666, + "grad_norm": 0.004525096621364355, + "learning_rate": 9.82785080804381e-06, + "loss": 1.2827372550964355, + "step": 39 + }, + { + "epoch": 0.5688888888888889, + "grad_norm": 0.0044178408570587635, + "learning_rate": 9.805804412106197e-06, + "loss": 1.2863564491271973, + "step": 40 + }, + { + "epoch": 0.5831111111111111, + "grad_norm": 0.00429871492087841, + "learning_rate": 9.782457878588977e-06, + "loss": 1.2768702507019043, + "step": 41 + }, + { + "epoch": 0.5973333333333334, + "grad_norm": 0.004442654084414244, + "learning_rate": 9.75781752354311e-06, + "loss": 1.2826354503631592, + "step": 42 + }, + { + "epoch": 0.6115555555555555, + "grad_norm": 0.004453128203749657, + "learning_rate": 9.731890013043367e-06, + "loss": 1.275444746017456, + "step": 43 + }, + { + "epoch": 0.6257777777777778, + "grad_norm": 0.004429741762578487, + "learning_rate": 9.704682361384941e-06, + "loss": 1.2660303115844727, + "step": 44 + }, + { + "epoch": 0.64, + "grad_norm": 0.004699238575994968, + "learning_rate": 9.676201929185809e-06, + "loss": 1.2627251148223877, + "step": 45 + }, + { + "epoch": 0.6542222222222223, + "grad_norm": 0.004564967937767506, + "learning_rate": 9.646456421395447e-06, + "loss": 1.2888470888137817, + "step": 46 + }, + { + "epoch": 0.6684444444444444, + "grad_norm": 0.004141585901379585, + "learning_rate": 9.615453885210368e-06, + "loss": 1.2826390266418457, + "step": 47 + }, + { + "epoch": 0.6826666666666666, + "grad_norm": 0.004241476766765118, + "learning_rate": 9.583202707897075e-06, + "loss": 1.2571580410003662, + "step": 48 + }, + { + "epoch": 0.6968888888888889, + "grad_norm": 0.004200824070721865, + "learning_rate": 9.549711614523007e-06, + "loss": 1.2616808414459229, + "step": 49 + }, + { + "epoch": 0.7111111111111111, + "grad_norm": 0.004399280063807964, + "learning_rate": 9.514989665596114e-06, + "loss": 1.263854742050171, + "step": 50 + }, + { + "epoch": 0.7111111111111111, + "eval_loss": 1.248613715171814, + "eval_runtime": 54.7705, + "eval_samples_per_second": 3.652, + "eval_steps_per_second": 1.826, + "step": 50 + }, + { + "epoch": 0.7253333333333334, + "grad_norm": 0.004282557871192694, + "learning_rate": 9.479046254613673e-06, + "loss": 1.279876947402954, + "step": 51 + }, + { + "epoch": 0.7395555555555555, + "grad_norm": 0.004135944880545139, + "learning_rate": 9.441891105521005e-06, + "loss": 1.2600014209747314, + "step": 52 + }, + { + "epoch": 0.7537777777777778, + "grad_norm": 0.0039484258741140366, + "learning_rate": 9.40353427008083e-06, + "loss": 1.2702335119247437, + "step": 53 + }, + { + "epoch": 0.768, + "grad_norm": 0.004160050302743912, + "learning_rate": 9.3639861251539e-06, + "loss": 1.2441630363464355, + "step": 54 + }, + { + "epoch": 0.7822222222222223, + "grad_norm": 0.004111913964152336, + "learning_rate": 9.323257369891702e-06, + "loss": 1.2441895008087158, + "step": 55 + }, + { + "epoch": 0.7964444444444444, + "grad_norm": 0.004063262604176998, + "learning_rate": 9.281359022841966e-06, + "loss": 1.255817174911499, + "step": 56 + }, + { + "epoch": 0.8106666666666666, + "grad_norm": 0.003984581213444471, + "learning_rate": 9.238302418967757e-06, + "loss": 1.2536640167236328, + "step": 57 + }, + { + "epoch": 0.8248888888888889, + "grad_norm": 0.003996702842414379, + "learning_rate": 9.194099206580981e-06, + "loss": 1.2499892711639404, + "step": 58 + }, + { + "epoch": 0.8391111111111111, + "grad_norm": 0.0044861058704555035, + "learning_rate": 9.14876134419111e-06, + "loss": 1.248606562614441, + "step": 59 + }, + { + "epoch": 0.8533333333333334, + "grad_norm": 0.004101422615349293, + "learning_rate": 9.102301097269974e-06, + "loss": 1.247856616973877, + "step": 60 + }, + { + "epoch": 0.8675555555555555, + "grad_norm": 0.004009375814348459, + "learning_rate": 9.05473103493355e-06, + "loss": 1.257920742034912, + "step": 61 + }, + { + "epoch": 0.8817777777777778, + "grad_norm": 0.0039364611729979515, + "learning_rate": 9.006064026541549e-06, + "loss": 1.2761049270629883, + "step": 62 + }, + { + "epoch": 0.896, + "grad_norm": 0.004427333828061819, + "learning_rate": 8.956313238215824e-06, + "loss": 1.245169997215271, + "step": 63 + }, + { + "epoch": 0.9102222222222223, + "grad_norm": 0.004076611250638962, + "learning_rate": 8.905492129278478e-06, + "loss": 1.234323263168335, + "step": 64 + }, + { + "epoch": 0.9244444444444444, + "grad_norm": 0.004032611381262541, + "learning_rate": 8.85361444861063e-06, + "loss": 1.2250256538391113, + "step": 65 + }, + { + "epoch": 0.9386666666666666, + "grad_norm": 0.004036127123981714, + "learning_rate": 8.800694230932885e-06, + "loss": 1.2339348793029785, + "step": 66 + }, + { + "epoch": 0.9528888888888889, + "grad_norm": 0.004113379865884781, + "learning_rate": 8.74674579300843e-06, + "loss": 1.252026081085205, + "step": 67 + }, + { + "epoch": 0.9671111111111111, + "grad_norm": 0.004064187407493591, + "learning_rate": 8.691783729769874e-06, + "loss": 1.2473349571228027, + "step": 68 + }, + { + "epoch": 0.9813333333333333, + "grad_norm": 0.0040993476286530495, + "learning_rate": 8.635822910370793e-06, + "loss": 1.2227911949157715, + "step": 69 + }, + { + "epoch": 0.9955555555555555, + "grad_norm": 0.003903420874848962, + "learning_rate": 8.578878474163115e-06, + "loss": 1.2343428134918213, + "step": 70 + }, + { + "epoch": 1.0, + "grad_norm": 0.006894194986671209, + "learning_rate": 8.520965826601394e-06, + "loss": 1.2428948879241943, + "step": 71 + }, + { + "epoch": 1.0142222222222221, + "grad_norm": 0.004313764162361622, + "learning_rate": 8.462100635075097e-06, + "loss": 1.202027440071106, + "step": 72 + }, + { + "epoch": 1.0284444444444445, + "grad_norm": 0.004148328676819801, + "learning_rate": 8.40229882467003e-06, + "loss": 1.2224020957946777, + "step": 73 + }, + { + "epoch": 1.0426666666666666, + "grad_norm": 0.00453830836340785, + "learning_rate": 8.341576573860049e-06, + "loss": 1.2196844816207886, + "step": 74 + }, + { + "epoch": 1.056888888888889, + "grad_norm": 0.004094650503247976, + "learning_rate": 8.279950310130218e-06, + "loss": 1.2222049236297607, + "step": 75 + }, + { + "epoch": 1.0711111111111111, + "grad_norm": 0.004169773776084185, + "learning_rate": 8.2174367055326e-06, + "loss": 1.2127679586410522, + "step": 76 + }, + { + "epoch": 1.0853333333333333, + "grad_norm": 0.004176456015557051, + "learning_rate": 8.154052672175888e-06, + "loss": 1.1962798833847046, + "step": 77 + }, + { + "epoch": 1.0995555555555556, + "grad_norm": 0.004215880297124386, + "learning_rate": 8.08981535765009e-06, + "loss": 1.192876935005188, + "step": 78 + }, + { + "epoch": 1.1137777777777778, + "grad_norm": 0.004112801048904657, + "learning_rate": 8.024742140387506e-06, + "loss": 1.204209566116333, + "step": 79 + }, + { + "epoch": 1.1280000000000001, + "grad_norm": 0.00416502496227622, + "learning_rate": 7.95885062496126e-06, + "loss": 1.2061021327972412, + "step": 80 + }, + { + "epoch": 1.1422222222222222, + "grad_norm": 0.004163650795817375, + "learning_rate": 7.892158637322647e-06, + "loss": 1.2074682712554932, + "step": 81 + }, + { + "epoch": 1.1564444444444444, + "grad_norm": 0.004129140637814999, + "learning_rate": 7.824684219978591e-06, + "loss": 1.1797451972961426, + "step": 82 + }, + { + "epoch": 1.1706666666666667, + "grad_norm": 0.004320825915783644, + "learning_rate": 7.756445627110523e-06, + "loss": 1.2151604890823364, + "step": 83 + }, + { + "epoch": 1.1848888888888889, + "grad_norm": 0.004057695157825947, + "learning_rate": 7.68746131963598e-06, + "loss": 1.1917520761489868, + "step": 84 + }, + { + "epoch": 1.199111111111111, + "grad_norm": 0.004045524168759584, + "learning_rate": 7.6177499602143e-06, + "loss": 1.1966304779052734, + "step": 85 + }, + { + "epoch": 1.2133333333333334, + "grad_norm": 0.004168134648352861, + "learning_rate": 7.547330408197695e-06, + "loss": 1.2256059646606445, + "step": 86 + }, + { + "epoch": 1.2275555555555555, + "grad_norm": 0.0040689390152692795, + "learning_rate": 7.476221714529167e-06, + "loss": 1.1984763145446777, + "step": 87 + }, + { + "epoch": 1.2417777777777779, + "grad_norm": 0.004095707554370165, + "learning_rate": 7.404443116588548e-06, + "loss": 1.1815118789672852, + "step": 88 + }, + { + "epoch": 1.256, + "grad_norm": 0.004128715489059687, + "learning_rate": 7.332014032988123e-06, + "loss": 1.1813948154449463, + "step": 89 + }, + { + "epoch": 1.2702222222222221, + "grad_norm": 0.004037311766296625, + "learning_rate": 7.2589540583192165e-06, + "loss": 1.2083604335784912, + "step": 90 + }, + { + "epoch": 1.2844444444444445, + "grad_norm": 0.004094353877007961, + "learning_rate": 7.185282957851175e-06, + "loss": 1.2008360624313354, + "step": 91 + }, + { + "epoch": 1.2986666666666666, + "grad_norm": 0.004135950934141874, + "learning_rate": 7.111020662184174e-06, + "loss": 1.1902484893798828, + "step": 92 + }, + { + "epoch": 1.3128888888888888, + "grad_norm": 0.003980172332376242, + "learning_rate": 7.036187261857289e-06, + "loss": 1.1865317821502686, + "step": 93 + }, + { + "epoch": 1.3271111111111111, + "grad_norm": 0.003914009779691696, + "learning_rate": 6.960803001913315e-06, + "loss": 1.1860058307647705, + "step": 94 + }, + { + "epoch": 1.3413333333333333, + "grad_norm": 0.003945492208003998, + "learning_rate": 6.884888276421766e-06, + "loss": 1.1842578649520874, + "step": 95 + }, + { + "epoch": 1.3555555555555556, + "grad_norm": 0.0038985933642834425, + "learning_rate": 6.8084636229615786e-06, + "loss": 1.1814080476760864, + "step": 96 + }, + { + "epoch": 1.3697777777777778, + "grad_norm": 0.003922275733202696, + "learning_rate": 6.731549717064975e-06, + "loss": 1.193676233291626, + "step": 97 + }, + { + "epoch": 1.384, + "grad_norm": 0.004165121354162693, + "learning_rate": 6.654167366624009e-06, + "loss": 1.2141278982162476, + "step": 98 + }, + { + "epoch": 1.3982222222222223, + "grad_norm": 0.004156369715929031, + "learning_rate": 6.576337506261314e-06, + "loss": 1.1865811347961426, + "step": 99 + }, + { + "epoch": 1.4124444444444444, + "grad_norm": 0.00419853487983346, + "learning_rate": 6.498081191666549e-06, + "loss": 1.1954832077026367, + "step": 100 + }, + { + "epoch": 1.4124444444444444, + "eval_loss": 1.1988927125930786, + "eval_runtime": 11.4257, + "eval_samples_per_second": 17.504, + "eval_steps_per_second": 8.752, + "step": 100 + }, + { + "epoch": 1.4266666666666667, + "grad_norm": 0.003968748729676008, + "learning_rate": 6.419419593900109e-06, + "loss": 1.1804523468017578, + "step": 101 + }, + { + "epoch": 1.4408888888888889, + "grad_norm": 0.003949783742427826, + "learning_rate": 6.340373993665607e-06, + "loss": 1.2068874835968018, + "step": 102 + }, + { + "epoch": 1.455111111111111, + "grad_norm": 0.004127713851630688, + "learning_rate": 6.2609657755527135e-06, + "loss": 1.1804969310760498, + "step": 103 + }, + { + "epoch": 1.4693333333333334, + "grad_norm": 0.004182465374469757, + "learning_rate": 6.1812164222518626e-06, + "loss": 1.1953659057617188, + "step": 104 + }, + { + "epoch": 1.4835555555555555, + "grad_norm": 0.004224156495183706, + "learning_rate": 6.101147508742456e-06, + "loss": 1.1933740377426147, + "step": 105 + }, + { + "epoch": 1.4977777777777779, + "grad_norm": 0.0041472818702459335, + "learning_rate": 6.020780696456059e-06, + "loss": 1.201953411102295, + "step": 106 + }, + { + "epoch": 1.512, + "grad_norm": 0.0038781315088272095, + "learning_rate": 5.940137727416247e-06, + "loss": 1.1897786855697632, + "step": 107 + }, + { + "epoch": 1.5262222222222221, + "grad_norm": 0.004033711273223162, + "learning_rate": 5.859240418356614e-06, + "loss": 1.1798899173736572, + "step": 108 + }, + { + "epoch": 1.5404444444444443, + "grad_norm": 0.00406253058463335, + "learning_rate": 5.778110654818602e-06, + "loss": 1.1853947639465332, + "step": 109 + }, + { + "epoch": 1.5546666666666666, + "grad_norm": 0.004025817848742008, + "learning_rate": 5.696770385230679e-06, + "loss": 1.1760761737823486, + "step": 110 + }, + { + "epoch": 1.568888888888889, + "grad_norm": 0.004029016941785812, + "learning_rate": 5.615241614970546e-06, + "loss": 1.1980795860290527, + "step": 111 + }, + { + "epoch": 1.5831111111111111, + "grad_norm": 0.004079029895365238, + "learning_rate": 5.533546400411899e-06, + "loss": 1.1878252029418945, + "step": 112 + }, + { + "epoch": 1.5973333333333333, + "grad_norm": 0.004004180897027254, + "learning_rate": 5.4517068429574215e-06, + "loss": 1.1891889572143555, + "step": 113 + }, + { + "epoch": 1.6115555555555554, + "grad_norm": 0.004044163506478071, + "learning_rate": 5.3697450830595775e-06, + "loss": 1.1935696601867676, + "step": 114 + }, + { + "epoch": 1.6257777777777778, + "grad_norm": 0.0040819537825882435, + "learning_rate": 5.287683294230855e-06, + "loss": 1.1648902893066406, + "step": 115 + }, + { + "epoch": 1.6400000000000001, + "grad_norm": 0.0042525348253548145, + "learning_rate": 5.20554367704505e-06, + "loss": 1.1749862432479858, + "step": 116 + }, + { + "epoch": 1.6542222222222223, + "grad_norm": 0.004129970911890268, + "learning_rate": 5.123348453131242e-06, + "loss": 1.200908899307251, + "step": 117 + }, + { + "epoch": 1.6684444444444444, + "grad_norm": 0.004072704818099737, + "learning_rate": 5.041119859162068e-06, + "loss": 1.189719796180725, + "step": 118 + }, + { + "epoch": 1.6826666666666665, + "grad_norm": 0.004113580100238323, + "learning_rate": 4.958880140837934e-06, + "loss": 1.1798235177993774, + "step": 119 + }, + { + "epoch": 1.696888888888889, + "grad_norm": 0.003983311355113983, + "learning_rate": 4.876651546868759e-06, + "loss": 1.166198492050171, + "step": 120 + }, + { + "epoch": 1.7111111111111112, + "grad_norm": 0.003966879565268755, + "learning_rate": 4.794456322954953e-06, + "loss": 1.1802079677581787, + "step": 121 + }, + { + "epoch": 1.7253333333333334, + "grad_norm": 0.004070781171321869, + "learning_rate": 4.7123167057691446e-06, + "loss": 1.1807469129562378, + "step": 122 + }, + { + "epoch": 1.7395555555555555, + "grad_norm": 0.0039869495667517185, + "learning_rate": 4.630254916940424e-06, + "loss": 1.1979055404663086, + "step": 123 + }, + { + "epoch": 1.7537777777777777, + "grad_norm": 0.004057880956679583, + "learning_rate": 4.548293157042581e-06, + "loss": 1.192582130432129, + "step": 124 + }, + { + "epoch": 1.768, + "grad_norm": 0.004071912728250027, + "learning_rate": 4.466453599588103e-06, + "loss": 1.1831550598144531, + "step": 125 + }, + { + "epoch": 1.7822222222222224, + "grad_norm": 0.004058274906128645, + "learning_rate": 4.384758385029457e-06, + "loss": 1.1674884557724, + "step": 126 + }, + { + "epoch": 1.7964444444444445, + "grad_norm": 0.0038475256878882647, + "learning_rate": 4.3032296147693225e-06, + "loss": 1.181553602218628, + "step": 127 + }, + { + "epoch": 1.8106666666666666, + "grad_norm": 0.0039198137819767, + "learning_rate": 4.2218893451814005e-06, + "loss": 1.153355360031128, + "step": 128 + }, + { + "epoch": 1.8248888888888888, + "grad_norm": 0.003993947524577379, + "learning_rate": 4.140759581643386e-06, + "loss": 1.1762878894805908, + "step": 129 + }, + { + "epoch": 1.8391111111111111, + "grad_norm": 0.003767550690099597, + "learning_rate": 4.059862272583755e-06, + "loss": 1.1792393922805786, + "step": 130 + }, + { + "epoch": 1.8533333333333335, + "grad_norm": 0.004042964428663254, + "learning_rate": 3.979219303543942e-06, + "loss": 1.1686198711395264, + "step": 131 + }, + { + "epoch": 1.8675555555555556, + "grad_norm": 0.004027931950986385, + "learning_rate": 3.898852491257547e-06, + "loss": 1.179172396659851, + "step": 132 + }, + { + "epoch": 1.8817777777777778, + "grad_norm": 0.003951320890337229, + "learning_rate": 3.818783577748138e-06, + "loss": 1.1667160987854004, + "step": 133 + }, + { + "epoch": 1.896, + "grad_norm": 0.003887354861944914, + "learning_rate": 3.7390342244472886e-06, + "loss": 1.1844608783721924, + "step": 134 + }, + { + "epoch": 1.9102222222222223, + "grad_norm": 0.003996608778834343, + "learning_rate": 3.659626006334395e-06, + "loss": 1.1989758014678955, + "step": 135 + }, + { + "epoch": 1.9244444444444444, + "grad_norm": 0.003960869275033474, + "learning_rate": 3.5805804060998926e-06, + "loss": 1.1868371963500977, + "step": 136 + }, + { + "epoch": 1.9386666666666668, + "grad_norm": 0.003877371782436967, + "learning_rate": 3.501918808333453e-06, + "loss": 1.166593313217163, + "step": 137 + }, + { + "epoch": 1.952888888888889, + "grad_norm": 0.004059459548443556, + "learning_rate": 3.4236624937386874e-06, + "loss": 1.1770471334457397, + "step": 138 + }, + { + "epoch": 1.967111111111111, + "grad_norm": 0.0038016922771930695, + "learning_rate": 3.3458326333759927e-06, + "loss": 1.1796729564666748, + "step": 139 + }, + { + "epoch": 1.9813333333333332, + "grad_norm": 0.00393063947558403, + "learning_rate": 3.268450282935026e-06, + "loss": 1.1624293327331543, + "step": 140 + }, + { + "epoch": 1.9955555555555555, + "grad_norm": 0.0037788248155266047, + "learning_rate": 3.1915363770384223e-06, + "loss": 1.1687278747558594, + "step": 141 + }, + { + "epoch": 2.0, + "grad_norm": 0.006625148933380842, + "learning_rate": 3.1151117235782346e-06, + "loss": 1.1975963115692139, + "step": 142 + }, + { + "epoch": 2.014222222222222, + "grad_norm": 0.003876346629112959, + "learning_rate": 3.0391969980866874e-06, + "loss": 1.1713850498199463, + "step": 143 + }, + { + "epoch": 2.0284444444444443, + "grad_norm": 0.003996741026639938, + "learning_rate": 2.963812738142713e-06, + "loss": 1.1817915439605713, + "step": 144 + }, + { + "epoch": 2.042666666666667, + "grad_norm": 0.003873717039823532, + "learning_rate": 2.8889793378158284e-06, + "loss": 1.1671693325042725, + "step": 145 + }, + { + "epoch": 2.056888888888889, + "grad_norm": 0.003939088433980942, + "learning_rate": 2.814717042148827e-06, + "loss": 1.1553094387054443, + "step": 146 + }, + { + "epoch": 2.071111111111111, + "grad_norm": 0.003916384186595678, + "learning_rate": 2.7410459416807856e-06, + "loss": 1.1611030101776123, + "step": 147 + }, + { + "epoch": 2.0853333333333333, + "grad_norm": 0.0037560381460934877, + "learning_rate": 2.6679859670118785e-06, + "loss": 1.169109582901001, + "step": 148 + }, + { + "epoch": 2.0995555555555554, + "grad_norm": 0.0038683414459228516, + "learning_rate": 2.5955568834114523e-06, + "loss": 1.1539714336395264, + "step": 149 + }, + { + "epoch": 2.113777777777778, + "grad_norm": 0.003892498556524515, + "learning_rate": 2.523778285470835e-06, + "loss": 1.1314873695373535, + "step": 150 + }, + { + "epoch": 2.113777777777778, + "eval_loss": 1.1810542345046997, + "eval_runtime": 77.0472, + "eval_samples_per_second": 2.596, + "eval_steps_per_second": 1.298, + "step": 150 + }, + { + "epoch": 2.128, + "grad_norm": 0.003935188986361027, + "learning_rate": 2.452669591802307e-06, + "loss": 1.1695539951324463, + "step": 151 + }, + { + "epoch": 2.1422222222222222, + "grad_norm": 0.0038220868445932865, + "learning_rate": 2.3822500397857016e-06, + "loss": 1.1494641304016113, + "step": 152 + }, + { + "epoch": 2.1564444444444444, + "grad_norm": 0.0038105787243694067, + "learning_rate": 2.3125386803640188e-06, + "loss": 1.1655616760253906, + "step": 153 + }, + { + "epoch": 2.1706666666666665, + "grad_norm": 0.003957564476877451, + "learning_rate": 2.243554372889479e-06, + "loss": 1.1519088745117188, + "step": 154 + }, + { + "epoch": 2.1848888888888887, + "grad_norm": 0.003938976209610701, + "learning_rate": 2.1753157800214107e-06, + "loss": 1.1479674577713013, + "step": 155 + }, + { + "epoch": 2.1991111111111112, + "grad_norm": 0.0037570318672806025, + "learning_rate": 2.1078413626773547e-06, + "loss": 1.1504448652267456, + "step": 156 + }, + { + "epoch": 2.2133333333333334, + "grad_norm": 0.003837036434561014, + "learning_rate": 2.0411493750387423e-06, + "loss": 1.142012119293213, + "step": 157 + }, + { + "epoch": 2.2275555555555555, + "grad_norm": 0.0037807337939739227, + "learning_rate": 1.9752578596124955e-06, + "loss": 1.1382803916931152, + "step": 158 + }, + { + "epoch": 2.2417777777777776, + "grad_norm": 0.0038557867519557476, + "learning_rate": 1.9101846423499113e-06, + "loss": 1.1602444648742676, + "step": 159 + }, + { + "epoch": 2.2560000000000002, + "grad_norm": 0.003806622698903084, + "learning_rate": 1.8459473278241125e-06, + "loss": 1.146360158920288, + "step": 160 + }, + { + "epoch": 2.2702222222222224, + "grad_norm": 0.003777023172006011, + "learning_rate": 1.7825632944674016e-06, + "loss": 1.1420152187347412, + "step": 161 + }, + { + "epoch": 2.2844444444444445, + "grad_norm": 0.0038270214572548866, + "learning_rate": 1.7200496898697832e-06, + "loss": 1.1756829023361206, + "step": 162 + }, + { + "epoch": 2.2986666666666666, + "grad_norm": 0.003796679200604558, + "learning_rate": 1.6584234261399535e-06, + "loss": 1.1596589088439941, + "step": 163 + }, + { + "epoch": 2.3128888888888888, + "grad_norm": 0.003775167977437377, + "learning_rate": 1.5977011753299726e-06, + "loss": 1.1395504474639893, + "step": 164 + }, + { + "epoch": 2.327111111111111, + "grad_norm": 0.0038062240928411484, + "learning_rate": 1.5378993649249053e-06, + "loss": 1.1680355072021484, + "step": 165 + }, + { + "epoch": 2.3413333333333335, + "grad_norm": 0.0038756304420530796, + "learning_rate": 1.4790341733986085e-06, + "loss": 1.1566658020019531, + "step": 166 + }, + { + "epoch": 2.3555555555555556, + "grad_norm": 0.00390993244946003, + "learning_rate": 1.4211215258368866e-06, + "loss": 1.1633788347244263, + "step": 167 + }, + { + "epoch": 2.3697777777777778, + "grad_norm": 0.0038066902197897434, + "learning_rate": 1.3641770896292083e-06, + "loss": 1.1514759063720703, + "step": 168 + }, + { + "epoch": 2.384, + "grad_norm": 0.003804982639849186, + "learning_rate": 1.3082162702301276e-06, + "loss": 1.1515188217163086, + "step": 169 + }, + { + "epoch": 2.398222222222222, + "grad_norm": 0.003867222461849451, + "learning_rate": 1.2532542069915722e-06, + "loss": 1.1591559648513794, + "step": 170 + }, + { + "epoch": 2.4124444444444446, + "grad_norm": 0.0037913229316473007, + "learning_rate": 1.1993057690671174e-06, + "loss": 1.1669801473617554, + "step": 171 + }, + { + "epoch": 2.4266666666666667, + "grad_norm": 0.0037280265241861343, + "learning_rate": 1.1463855513893695e-06, + "loss": 1.155794382095337, + "step": 172 + }, + { + "epoch": 2.440888888888889, + "grad_norm": 0.0040002260357141495, + "learning_rate": 1.0945078707215224e-06, + "loss": 1.1434683799743652, + "step": 173 + }, + { + "epoch": 2.455111111111111, + "grad_norm": 0.0037701099645346403, + "learning_rate": 1.0436867617841768e-06, + "loss": 1.1586582660675049, + "step": 174 + }, + { + "epoch": 2.469333333333333, + "grad_norm": 0.0038694143295288086, + "learning_rate": 9.939359734584552e-07, + "loss": 1.1643059253692627, + "step": 175 + }, + { + "epoch": 2.4835555555555557, + "grad_norm": 0.0036939869169145823, + "learning_rate": 9.452689650664515e-07, + "loss": 1.137399673461914, + "step": 176 + }, + { + "epoch": 2.497777777777778, + "grad_norm": 0.0037478527519851923, + "learning_rate": 8.976989027300265e-07, + "loss": 1.150512456893921, + "step": 177 + }, + { + "epoch": 2.512, + "grad_norm": 0.003883050289005041, + "learning_rate": 8.512386558088919e-07, + "loss": 1.1556127071380615, + "step": 178 + }, + { + "epoch": 2.526222222222222, + "grad_norm": 0.003667640034109354, + "learning_rate": 8.059007934190194e-07, + "loss": 1.1506540775299072, + "step": 179 + }, + { + "epoch": 2.5404444444444443, + "grad_norm": 0.003665938274934888, + "learning_rate": 7.61697581032243e-07, + "loss": 1.1606539487838745, + "step": 180 + }, + { + "epoch": 2.554666666666667, + "grad_norm": 0.0037777498364448547, + "learning_rate": 7.186409771580355e-07, + "loss": 1.145754098892212, + "step": 181 + }, + { + "epoch": 2.568888888888889, + "grad_norm": 0.0037169207353144884, + "learning_rate": 6.76742630108298e-07, + "loss": 1.1506030559539795, + "step": 182 + }, + { + "epoch": 2.583111111111111, + "grad_norm": 0.0036668803077191114, + "learning_rate": 6.360138748461015e-07, + "loss": 1.1366521120071411, + "step": 183 + }, + { + "epoch": 2.5973333333333333, + "grad_norm": 0.0037295217625796795, + "learning_rate": 5.964657299191712e-07, + "loss": 1.1723542213439941, + "step": 184 + }, + { + "epoch": 2.6115555555555554, + "grad_norm": 0.003689161501824856, + "learning_rate": 5.581088944789953e-07, + "loss": 1.1543989181518555, + "step": 185 + }, + { + "epoch": 2.6257777777777775, + "grad_norm": 0.0038293329998850822, + "learning_rate": 5.20953745386329e-07, + "loss": 1.1481022834777832, + "step": 186 + }, + { + "epoch": 2.64, + "grad_norm": 0.0037391053047031164, + "learning_rate": 4.850103344038853e-07, + "loss": 1.1510019302368164, + "step": 187 + }, + { + "epoch": 2.6542222222222223, + "grad_norm": 0.0036369990557432175, + "learning_rate": 4.502883854769935e-07, + "loss": 1.1454750299453735, + "step": 188 + }, + { + "epoch": 2.6684444444444444, + "grad_norm": 0.003734483616426587, + "learning_rate": 4.167972921029262e-07, + "loss": 1.1555531024932861, + "step": 189 + }, + { + "epoch": 2.6826666666666665, + "grad_norm": 0.0037341697607189417, + "learning_rate": 3.8454611478963235e-07, + "loss": 1.141042709350586, + "step": 190 + }, + { + "epoch": 2.696888888888889, + "grad_norm": 0.0036616334691643715, + "learning_rate": 3.535435786045538e-07, + "loss": 1.1597721576690674, + "step": 191 + }, + { + "epoch": 2.7111111111111112, + "grad_norm": 0.0037469498347491026, + "learning_rate": 3.237980708141919e-07, + "loss": 1.1370229721069336, + "step": 192 + }, + { + "epoch": 2.7253333333333334, + "grad_norm": 0.003746448317542672, + "learning_rate": 2.9531763861505967e-07, + "loss": 1.1386423110961914, + "step": 193 + }, + { + "epoch": 2.7395555555555555, + "grad_norm": 0.0038595127407461405, + "learning_rate": 2.681099869566328e-07, + "loss": 1.141430139541626, + "step": 194 + }, + { + "epoch": 2.7537777777777777, + "grad_norm": 0.0037963776849210262, + "learning_rate": 2.4218247645689306e-07, + "loss": 1.1269652843475342, + "step": 195 + }, + { + "epoch": 2.768, + "grad_norm": 0.0037523838691413403, + "learning_rate": 2.175421214110235e-07, + "loss": 1.1652940511703491, + "step": 196 + }, + { + "epoch": 2.7822222222222224, + "grad_norm": 0.0037649553269147873, + "learning_rate": 1.941955878938029e-07, + "loss": 1.1484644412994385, + "step": 197 + }, + { + "epoch": 2.7964444444444445, + "grad_norm": 0.0036591957323253155, + "learning_rate": 1.721491919561913e-07, + "loss": 1.1459333896636963, + "step": 198 + }, + { + "epoch": 2.8106666666666666, + "grad_norm": 0.003836029442027211, + "learning_rate": 1.514088979166256e-07, + "loss": 1.157778024673462, + "step": 199 + }, + { + "epoch": 2.824888888888889, + "grad_norm": 0.003767541144043207, + "learning_rate": 1.3198031674745814e-07, + "loss": 1.1373133659362793, + "step": 200 + }, + { + "epoch": 2.824888888888889, + "eval_loss": 1.1767549514770508, + "eval_runtime": 17.7324, + "eval_samples_per_second": 11.279, + "eval_steps_per_second": 5.639, + "step": 200 + }, + { + "epoch": 2.8391111111111114, + "grad_norm": 0.0036713366862386465, + "learning_rate": 1.138687045569975e-07, + "loss": 1.1472009420394897, + "step": 201 + }, + { + "epoch": 2.8533333333333335, + "grad_norm": 0.003775710705667734, + "learning_rate": 9.70789611675449e-08, + "loss": 1.1417003870010376, + "step": 202 + }, + { + "epoch": 2.8675555555555556, + "grad_norm": 0.003818412311375141, + "learning_rate": 8.161562878982399e-08, + "loss": 1.1214667558670044, + "step": 203 + }, + { + "epoch": 2.8817777777777778, + "grad_norm": 0.0038233508821576834, + "learning_rate": 6.74828907941516e-08, + "loss": 1.138871431350708, + "step": 204 + }, + { + "epoch": 2.896, + "grad_norm": 0.0036753350868821144, + "learning_rate": 5.468457057869358e-08, + "loss": 1.1596763134002686, + "step": 205 + }, + { + "epoch": 2.910222222222222, + "grad_norm": 0.0036339866928756237, + "learning_rate": 4.322413053509944e-08, + "loss": 1.1474013328552246, + "step": 206 + }, + { + "epoch": 2.924444444444444, + "grad_norm": 0.003777341218665242, + "learning_rate": 3.31046711118066e-08, + "loss": 1.1420172452926636, + "step": 207 + }, + { + "epoch": 2.9386666666666668, + "grad_norm": 0.0037329578772187233, + "learning_rate": 2.4328929975260262e-08, + "loss": 1.149836778640747, + "step": 208 + }, + { + "epoch": 2.952888888888889, + "grad_norm": 0.00386771815828979, + "learning_rate": 1.6899281269279756e-08, + "loss": 1.142824649810791, + "step": 209 + }, + { + "epoch": 2.967111111111111, + "grad_norm": 0.0036802676040679216, + "learning_rate": 1.0817734972768946e-08, + "loss": 1.1392970085144043, + "step": 210 + }, + { + "epoch": 2.981333333333333, + "grad_norm": 0.003702987916767597, + "learning_rate": 6.085936355947897e-09, + "loss": 1.1546673774719238, + "step": 211 + }, + { + "epoch": 2.9955555555555557, + "grad_norm": 0.0036437278613448143, + "learning_rate": 2.7051655352494654e-09, + "loss": 1.1519522666931152, + "step": 212 + }, + { + "epoch": 3.0, + "grad_norm": 0.006544829346239567, + "learning_rate": 6.763371270035457e-10, + "loss": 1.1223903894424438, + "step": 213 + }, + { + "epoch": 3.0, + "step": 213, + "total_flos": 4.5153968281196954e+17, + "train_loss": 0.0, + "train_runtime": 4.5863, + "train_samples_per_second": 11774.243, + "train_steps_per_second": 46.443 + } + ], + "logging_steps": 1, + "max_steps": 213, + "num_input_tokens_seen": 0, + "num_train_epochs": 3, + "save_steps": 500, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 4.5153968281196954e+17, + "train_batch_size": 1, + "trial_name": null, + "trial_params": null +} diff --git a/qwen2.5-1.5b-instruct/training_args.bin b/qwen2.5-1.5b-instruct/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..cf39d7c3d673c085a29e8f5066faa91e72181179 --- /dev/null +++ b/qwen2.5-1.5b-instruct/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:191a35d716d5e6eb5183c0f6b71c5c48a64b55747675582ade81b04d1bb4a020 +size 5713 diff --git a/qwen2.5-1.5b-instruct/training_eval_loss.png b/qwen2.5-1.5b-instruct/training_eval_loss.png new file mode 100644 index 0000000000000000000000000000000000000000..6b92798ae8a213ee40b3a7360042dbc057541f19 Binary files /dev/null and b/qwen2.5-1.5b-instruct/training_eval_loss.png differ diff --git a/qwen2.5-1.5b-instruct/training_loss.png b/qwen2.5-1.5b-instruct/training_loss.png new file mode 100644 index 0000000000000000000000000000000000000000..de0c7d50253e5718ac3106ce4d4630ebbc27ac27 Binary files /dev/null and b/qwen2.5-1.5b-instruct/training_loss.png differ diff --git a/qwen2.5-3b-instruct/README.md b/qwen2.5-3b-instruct/README.md new file mode 100644 index 0000000000000000000000000000000000000000..e96a9549afc0d9b9ecb30c9e4904220ba1307aa1 --- /dev/null +++ b/qwen2.5-3b-instruct/README.md @@ -0,0 +1,69 @@ +--- +library_name: transformers +license: other +base_model: Qwen/Qwen2.5-3B-Instruct +tags: +- llama-factory +- full +- generated_from_trainer +model-index: +- name: qwen2.5-3b-instruct + results: [] +--- + + + +# qwen2.5-3b-instruct + +This model is a fine-tuned version of [Qwen/Qwen2.5-3B-Instruct](https://huggingface.co/Qwen/Qwen2.5-3B-Instruct) on the healthcare_chatdoctor_train_20k_trace_generation_valid dataset. +It achieves the following results on the evaluation set: +- Loss: 1.0748 + +## 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: 1e-05 +- train_batch_size: 1 +- eval_batch_size: 1 +- seed: 42 +- distributed_type: multi-GPU +- num_devices: 2 +- gradient_accumulation_steps: 128 +- total_train_batch_size: 256 +- total_eval_batch_size: 2 +- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments +- lr_scheduler_type: cosine +- lr_scheduler_warmup_steps: 0.1 +- num_epochs: 3 + +### Training results + +| Training Loss | Epoch | Step | Validation Loss | +|:-------------:|:------:|:----:|:---------------:| +| 1.1558 | 0.7111 | 50 | 1.1429 | +| 1.0880 | 1.4124 | 100 | 1.0955 | +| 1.0246 | 2.1138 | 150 | 1.0790 | +| 1.0300 | 2.8249 | 200 | 1.0748 | + + +### Framework versions + +- Transformers 5.2.0 +- Pytorch 2.11.0+cu128 +- Datasets 4.0.0 +- Tokenizers 0.22.2 diff --git a/qwen2.5-3b-instruct/all_results.json b/qwen2.5-3b-instruct/all_results.json new file mode 100644 index 0000000000000000000000000000000000000000..8c5648ce9d1f4b737b0536534128b7b5761683de --- /dev/null +++ b/qwen2.5-3b-instruct/all_results.json @@ -0,0 +1,12 @@ +{ + "epoch": 3.0, + "eval_loss": 1.0747698545455933, + "eval_runtime": 42.9121, + "eval_samples_per_second": 4.661, + "eval_steps_per_second": 2.33, + "total_flos": 9.561792243333857e+17, + "train_loss": 1.1234366983315194, + "train_runtime": 19384.6824, + "train_samples_per_second": 2.786, + "train_steps_per_second": 0.011 +} \ No newline at end of file diff --git a/qwen2.5-3b-instruct/chat_template.jinja b/qwen2.5-3b-instruct/chat_template.jinja new file mode 100644 index 0000000000000000000000000000000000000000..bdf7919a96cfe43d50914a007b9c0877bd0ec27e --- /dev/null +++ b/qwen2.5-3b-instruct/chat_template.jinja @@ -0,0 +1,54 @@ +{%- if tools %} + {{- '<|im_start|>system\n' }} + {%- if messages[0]['role'] == 'system' %} + {{- messages[0]['content'] }} + {%- else %} + {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }} + {%- endif %} + {{- "\n\n# 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' }} + {%- else %} + {{- '<|im_start|>system\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\n' }} + {%- endif %} +{%- endif %} +{%- for message in messages %} + {%- if (message.role == "user") or (message.role == "system" and not loop.first) or (message.role == "assistant" and not message.tool_calls) %} + {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }} + {%- elif message.role == "assistant" %} + {{- '<|im_start|>' + message.role }} + {%- if message.content %} + {{- '\n' + message.content }} + {%- endif %} + {%- for tool_call in message.tool_calls %} + {%- if tool_call.function is defined %} + {%- set tool_call = tool_call.function %} + {%- endif %} + {{- '\n\n{"name": "' }} + {{- tool_call.name }} + {{- '", "arguments": ' }} + {{- tool_call.arguments | tojson }} + {{- '}\n' }} + {%- endfor %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "tool" %} + {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != "tool") %} + {{- '<|im_start|>user' }} + {%- endif %} + {{- '\n\n' }} + {{- message.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' }} +{%- endif %} diff --git a/qwen2.5-3b-instruct/checkpoint-142/chat_template.jinja b/qwen2.5-3b-instruct/checkpoint-142/chat_template.jinja new file mode 100644 index 0000000000000000000000000000000000000000..bdf7919a96cfe43d50914a007b9c0877bd0ec27e --- /dev/null +++ b/qwen2.5-3b-instruct/checkpoint-142/chat_template.jinja @@ -0,0 +1,54 @@ +{%- if tools %} + {{- '<|im_start|>system\n' }} + {%- if messages[0]['role'] == 'system' %} + {{- messages[0]['content'] }} + {%- else %} + {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }} + {%- endif %} + {{- "\n\n# 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' }} + {%- else %} + {{- '<|im_start|>system\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\n' }} + {%- endif %} +{%- endif %} +{%- for message in messages %} + {%- if (message.role == "user") or (message.role == "system" and not loop.first) or (message.role == "assistant" and not message.tool_calls) %} + {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }} + {%- elif message.role == "assistant" %} + {{- '<|im_start|>' + message.role }} + {%- if message.content %} + {{- '\n' + message.content }} + {%- endif %} + {%- for tool_call in message.tool_calls %} + {%- if tool_call.function is defined %} + {%- set tool_call = tool_call.function %} + {%- endif %} + {{- '\n\n{"name": "' }} + {{- tool_call.name }} + {{- '", "arguments": ' }} + {{- tool_call.arguments | tojson }} + {{- '}\n' }} + {%- endfor %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "tool" %} + {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != "tool") %} + {{- '<|im_start|>user' }} + {%- endif %} + {{- '\n\n' }} + {{- message.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' }} +{%- endif %} diff --git a/qwen2.5-3b-instruct/checkpoint-142/config.json b/qwen2.5-3b-instruct/checkpoint-142/config.json new file mode 100644 index 0000000000000000000000000000000000000000..4814292863d1e5aa2f4fda1db18f4082cd61aa93 --- /dev/null +++ b/qwen2.5-3b-instruct/checkpoint-142/config.json @@ -0,0 +1,69 @@ +{ + "architectures": [ + "Qwen2ForCausalLM" + ], + "attention_dropout": 0.0, + "bos_token_id": null, + "dtype": "float32", + "eos_token_id": 151645, + "hidden_act": "silu", + "hidden_size": 2048, + "initializer_range": 0.02, + "intermediate_size": 11008, + "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": 32768, + "max_window_layers": 70, + "model_type": "qwen2", + "num_attention_heads": 16, + "num_hidden_layers": 36, + "num_key_value_heads": 2, + "pad_token_id": 151643, + "rms_norm_eps": 1e-06, + "rope_parameters": { + "rope_theta": 1000000.0, + "rope_type": "default" + }, + "sliding_window": null, + "tie_word_embeddings": true, + "transformers_version": "5.2.0", + "use_cache": false, + "use_sliding_window": false, + "vocab_size": 151936 +} diff --git a/qwen2.5-3b-instruct/checkpoint-142/generation_config.json b/qwen2.5-3b-instruct/checkpoint-142/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..47205fbf3386a215e7c2b9d62df6576f49b77e7b --- /dev/null +++ b/qwen2.5-3b-instruct/checkpoint-142/generation_config.json @@ -0,0 +1,13 @@ +{ + "do_sample": true, + "eos_token_id": [ + 151645, + 151643 + ], + "pad_token_id": 151643, + "repetition_penalty": 1.05, + "temperature": 0.7, + "top_k": 20, + "top_p": 0.8, + "transformers_version": "5.2.0" +} diff --git a/qwen2.5-3b-instruct/checkpoint-142/model.safetensors b/qwen2.5-3b-instruct/checkpoint-142/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..ecf27266e20240fad9f90aa224f2d4c58029e09c --- /dev/null +++ b/qwen2.5-3b-instruct/checkpoint-142/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3b27bc5c7d25ba65336c6d79f927762d7c8661265d2c236f7a21d66f6e36a01e +size 12343804296 diff --git a/qwen2.5-3b-instruct/checkpoint-142/optimizer.pt b/qwen2.5-3b-instruct/checkpoint-142/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..d860d1b76354b5a7a35a15a5c8668cc03f1850da --- /dev/null +++ b/qwen2.5-3b-instruct/checkpoint-142/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d4c66bc22528b4f90c03bb4c1ce4268e81499cb23d79bb0521fb791562240379 +size 24687895753 diff --git a/qwen2.5-3b-instruct/checkpoint-142/rng_state_0.pth b/qwen2.5-3b-instruct/checkpoint-142/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..870021e3fa5ac35c2f711adf0c93a556ab4842da --- /dev/null +++ b/qwen2.5-3b-instruct/checkpoint-142/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:95e5fc2074c0df31522a514f862c86cb00d71c946a7f15cc9ec0e53a69fb28a7 +size 14917 diff --git a/qwen2.5-3b-instruct/checkpoint-142/rng_state_1.pth b/qwen2.5-3b-instruct/checkpoint-142/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..21f20da5eb1da017f08aaa88bd19cf24d40e3fbf --- /dev/null +++ b/qwen2.5-3b-instruct/checkpoint-142/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0e7153eae67b6c9232a41bc996a2bf5b83229b8c7230d61911ac0fd40e64154e +size 14917 diff --git a/qwen2.5-3b-instruct/checkpoint-142/scheduler.pt b/qwen2.5-3b-instruct/checkpoint-142/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..2967e7243507a115f9424b1b386bf3657f047c75 --- /dev/null +++ b/qwen2.5-3b-instruct/checkpoint-142/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:922deb71175797a96c4bddc5e1cef1c097df38c0dfa52334e97c89e0f573f2bb +size 1465 diff --git a/qwen2.5-3b-instruct/checkpoint-142/tokenizer.json b/qwen2.5-3b-instruct/checkpoint-142/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..34510ff0037cd50428af467a17ead5a96140a32c --- /dev/null +++ b/qwen2.5-3b-instruct/checkpoint-142/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3fd169731d2cbde95e10bf356d66d5997fd885dd8dbb6fb4684da3f23b2585d8 +size 11421892 diff --git a/qwen2.5-3b-instruct/checkpoint-142/tokenizer_config.json b/qwen2.5-3b-instruct/checkpoint-142/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..145e2c7a68b109ae0253dda29fc2e66cb01b4b85 --- /dev/null +++ b/qwen2.5-3b-instruct/checkpoint-142/tokenizer_config.json @@ -0,0 +1,30 @@ +{ + "add_prefix_space": false, + "backend": "tokenizers", + "bos_token": null, + "clean_up_tokenization_spaces": false, + "eos_token": "<|im_end|>", + "errors": "replace", + "extra_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|>" + ], + "is_local": false, + "model_max_length": 131072, + "pad_token": "<|endoftext|>", + "padding_side": "right", + "split_special_tokens": false, + "tokenizer_class": "Qwen2Tokenizer", + "unk_token": null +} diff --git a/qwen2.5-3b-instruct/checkpoint-142/trainer_state.json b/qwen2.5-3b-instruct/checkpoint-142/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..099b390e82cf9e0d0fa9244e3e0982bcf5f07b61 --- /dev/null +++ b/qwen2.5-3b-instruct/checkpoint-142/trainer_state.json @@ -0,0 +1,1044 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 2.0, + "eval_steps": 50, + "global_step": 142, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.014222222222222223, + "grad_norm": 0.07561660557985306, + "learning_rate": 0.0, + "loss": 1.6462128162384033, + "step": 1 + }, + { + "epoch": 0.028444444444444446, + "grad_norm": 0.07576845586299896, + "learning_rate": 4.5454545454545457e-07, + "loss": 1.6739654541015625, + "step": 2 + }, + { + "epoch": 0.042666666666666665, + "grad_norm": 0.060671016573905945, + "learning_rate": 9.090909090909091e-07, + "loss": 1.6208744049072266, + "step": 3 + }, + { + "epoch": 0.05688888888888889, + "grad_norm": 0.04539854824542999, + "learning_rate": 1.3636363636363636e-06, + "loss": 1.5681328773498535, + "step": 4 + }, + { + "epoch": 0.07111111111111111, + "grad_norm": 0.030098970979452133, + "learning_rate": 1.8181818181818183e-06, + "loss": 1.5172258615493774, + "step": 5 + }, + { + "epoch": 0.08533333333333333, + "grad_norm": 0.02540394477546215, + "learning_rate": 2.2727272727272728e-06, + "loss": 1.4882113933563232, + "step": 6 + }, + { + "epoch": 0.09955555555555555, + "grad_norm": 0.026057474315166473, + "learning_rate": 2.7272727272727272e-06, + "loss": 1.453606128692627, + "step": 7 + }, + { + "epoch": 0.11377777777777778, + "grad_norm": 0.024274546653032303, + "learning_rate": 3.181818181818182e-06, + "loss": 1.433072805404663, + "step": 8 + }, + { + "epoch": 0.128, + "grad_norm": 0.02080286666750908, + "learning_rate": 3.6363636363636366e-06, + "loss": 1.4036792516708374, + "step": 9 + }, + { + "epoch": 0.14222222222222222, + "grad_norm": 0.021423645317554474, + "learning_rate": 4.0909090909090915e-06, + "loss": 1.3751729726791382, + "step": 10 + }, + { + "epoch": 0.15644444444444444, + "grad_norm": 0.018808694556355476, + "learning_rate": 4.5454545454545455e-06, + "loss": 1.3682663440704346, + "step": 11 + }, + { + "epoch": 0.17066666666666666, + "grad_norm": 0.013645165599882603, + "learning_rate": 5e-06, + "loss": 1.3491919040679932, + "step": 12 + }, + { + "epoch": 0.18488888888888888, + "grad_norm": 0.012575564906001091, + "learning_rate": 5.4545454545454545e-06, + "loss": 1.3444881439208984, + "step": 13 + }, + { + "epoch": 0.1991111111111111, + "grad_norm": 0.011944238096475601, + "learning_rate": 5.90909090909091e-06, + "loss": 1.3321316242218018, + "step": 14 + }, + { + "epoch": 0.21333333333333335, + "grad_norm": 0.012251932173967361, + "learning_rate": 6.363636363636364e-06, + "loss": 1.3083908557891846, + "step": 15 + }, + { + "epoch": 0.22755555555555557, + "grad_norm": 0.011485234834253788, + "learning_rate": 6.818181818181818e-06, + "loss": 1.287674903869629, + "step": 16 + }, + { + "epoch": 0.24177777777777779, + "grad_norm": 0.010052796453237534, + "learning_rate": 7.272727272727273e-06, + "loss": 1.2796708345413208, + "step": 17 + }, + { + "epoch": 0.256, + "grad_norm": 0.009297606535255909, + "learning_rate": 7.727272727272727e-06, + "loss": 1.2778702974319458, + "step": 18 + }, + { + "epoch": 0.2702222222222222, + "grad_norm": 0.009387261234223843, + "learning_rate": 8.181818181818183e-06, + "loss": 1.258772373199463, + "step": 19 + }, + { + "epoch": 0.28444444444444444, + "grad_norm": 0.009283723309636116, + "learning_rate": 8.636363636363637e-06, + "loss": 1.2937917709350586, + "step": 20 + }, + { + "epoch": 0.2986666666666667, + "grad_norm": 0.008504457771778107, + "learning_rate": 9.090909090909091e-06, + "loss": 1.247847318649292, + "step": 21 + }, + { + "epoch": 0.3128888888888889, + "grad_norm": 0.007998796179890633, + "learning_rate": 9.545454545454547e-06, + "loss": 1.235945701599121, + "step": 22 + }, + { + "epoch": 0.32711111111111113, + "grad_norm": 0.007681620307266712, + "learning_rate": 1e-05, + "loss": 1.2462598085403442, + "step": 23 + }, + { + "epoch": 0.3413333333333333, + "grad_norm": 0.007152756676077843, + "learning_rate": 9.999323662872998e-06, + "loss": 1.2484906911849976, + "step": 24 + }, + { + "epoch": 0.35555555555555557, + "grad_norm": 0.00654159439727664, + "learning_rate": 9.99729483446475e-06, + "loss": 1.2089664936065674, + "step": 25 + }, + { + "epoch": 0.36977777777777776, + "grad_norm": 0.0063508739694952965, + "learning_rate": 9.993914063644053e-06, + "loss": 1.2245509624481201, + "step": 26 + }, + { + "epoch": 0.384, + "grad_norm": 0.006452301051467657, + "learning_rate": 9.989182265027232e-06, + "loss": 1.2176213264465332, + "step": 27 + }, + { + "epoch": 0.3982222222222222, + "grad_norm": 0.006349734030663967, + "learning_rate": 9.98310071873072e-06, + "loss": 1.2157793045043945, + "step": 28 + }, + { + "epoch": 0.41244444444444445, + "grad_norm": 0.006191881373524666, + "learning_rate": 9.975671070024741e-06, + "loss": 1.237949013710022, + "step": 29 + }, + { + "epoch": 0.4266666666666667, + "grad_norm": 0.005793207325041294, + "learning_rate": 9.966895328888195e-06, + "loss": 1.21578049659729, + "step": 30 + }, + { + "epoch": 0.4408888888888889, + "grad_norm": 0.005483992397785187, + "learning_rate": 9.956775869464901e-06, + "loss": 1.2089630365371704, + "step": 31 + }, + { + "epoch": 0.45511111111111113, + "grad_norm": 0.005602920427918434, + "learning_rate": 9.945315429421307e-06, + "loss": 1.2175853252410889, + "step": 32 + }, + { + "epoch": 0.4693333333333333, + "grad_norm": 0.005212367512285709, + "learning_rate": 9.932517109205849e-06, + "loss": 1.2096197605133057, + "step": 33 + }, + { + "epoch": 0.48355555555555557, + "grad_norm": 0.005095571745187044, + "learning_rate": 9.918384371210178e-06, + "loss": 1.1974353790283203, + "step": 34 + }, + { + "epoch": 0.49777777777777776, + "grad_norm": 0.005125134252011776, + "learning_rate": 9.902921038832456e-06, + "loss": 1.1847140789031982, + "step": 35 + }, + { + "epoch": 0.512, + "grad_norm": 0.004882226698100567, + "learning_rate": 9.886131295443003e-06, + "loss": 1.2025952339172363, + "step": 36 + }, + { + "epoch": 0.5262222222222223, + "grad_norm": 0.004733556881546974, + "learning_rate": 9.868019683252543e-06, + "loss": 1.193751573562622, + "step": 37 + }, + { + "epoch": 0.5404444444444444, + "grad_norm": 0.004847981501370668, + "learning_rate": 9.848591102083375e-06, + "loss": 1.1725599765777588, + "step": 38 + }, + { + "epoch": 0.5546666666666666, + "grad_norm": 0.004716528579592705, + "learning_rate": 9.82785080804381e-06, + "loss": 1.1763365268707275, + "step": 39 + }, + { + "epoch": 0.5688888888888889, + "grad_norm": 0.004721782635897398, + "learning_rate": 9.805804412106197e-06, + "loss": 1.180063009262085, + "step": 40 + }, + { + "epoch": 0.5831111111111111, + "grad_norm": 0.004470364656299353, + "learning_rate": 9.782457878588977e-06, + "loss": 1.169750690460205, + "step": 41 + }, + { + "epoch": 0.5973333333333334, + "grad_norm": 0.0046170237474143505, + "learning_rate": 9.75781752354311e-06, + "loss": 1.174955129623413, + "step": 42 + }, + { + "epoch": 0.6115555555555555, + "grad_norm": 0.004407198633998632, + "learning_rate": 9.731890013043367e-06, + "loss": 1.1694097518920898, + "step": 43 + }, + { + "epoch": 0.6257777777777778, + "grad_norm": 0.004441778641194105, + "learning_rate": 9.704682361384941e-06, + "loss": 1.1582658290863037, + "step": 44 + }, + { + "epoch": 0.64, + "grad_norm": 0.004318015184253454, + "learning_rate": 9.676201929185809e-06, + "loss": 1.1544437408447266, + "step": 45 + }, + { + "epoch": 0.6542222222222223, + "grad_norm": 0.004480153322219849, + "learning_rate": 9.646456421395447e-06, + "loss": 1.1822636127471924, + "step": 46 + }, + { + "epoch": 0.6684444444444444, + "grad_norm": 0.004228127654641867, + "learning_rate": 9.615453885210368e-06, + "loss": 1.1748132705688477, + "step": 47 + }, + { + "epoch": 0.6826666666666666, + "grad_norm": 0.004379024263471365, + "learning_rate": 9.583202707897075e-06, + "loss": 1.1558542251586914, + "step": 48 + }, + { + "epoch": 0.6968888888888889, + "grad_norm": 0.004313027486205101, + "learning_rate": 9.549711614523007e-06, + "loss": 1.1546211242675781, + "step": 49 + }, + { + "epoch": 0.7111111111111111, + "grad_norm": 0.004527685698121786, + "learning_rate": 9.514989665596114e-06, + "loss": 1.1557996273040771, + "step": 50 + }, + { + "epoch": 0.7111111111111111, + "eval_loss": 1.1428771018981934, + "eval_runtime": 47.9864, + "eval_samples_per_second": 4.168, + "eval_steps_per_second": 2.084, + "step": 50 + }, + { + "epoch": 0.7253333333333334, + "grad_norm": 0.0040644253604114056, + "learning_rate": 9.479046254613673e-06, + "loss": 1.172366976737976, + "step": 51 + }, + { + "epoch": 0.7395555555555555, + "grad_norm": 0.004353945609182119, + "learning_rate": 9.441891105521005e-06, + "loss": 1.1539039611816406, + "step": 52 + }, + { + "epoch": 0.7537777777777778, + "grad_norm": 0.004351011943072081, + "learning_rate": 9.40353427008083e-06, + "loss": 1.159185767173767, + "step": 53 + }, + { + "epoch": 0.768, + "grad_norm": 0.004263842478394508, + "learning_rate": 9.3639861251539e-06, + "loss": 1.1381170749664307, + "step": 54 + }, + { + "epoch": 0.7822222222222223, + "grad_norm": 0.0042404914274811745, + "learning_rate": 9.323257369891702e-06, + "loss": 1.1402206420898438, + "step": 55 + }, + { + "epoch": 0.7964444444444444, + "grad_norm": 0.004360774997621775, + "learning_rate": 9.281359022841966e-06, + "loss": 1.1499946117401123, + "step": 56 + }, + { + "epoch": 0.8106666666666666, + "grad_norm": 0.004249541088938713, + "learning_rate": 9.238302418967757e-06, + "loss": 1.1507560014724731, + "step": 57 + }, + { + "epoch": 0.8248888888888889, + "grad_norm": 0.004316243343055248, + "learning_rate": 9.194099206580981e-06, + "loss": 1.14259934425354, + "step": 58 + }, + { + "epoch": 0.8391111111111111, + "grad_norm": 0.004926053807139397, + "learning_rate": 9.14876134419111e-06, + "loss": 1.1445270776748657, + "step": 59 + }, + { + "epoch": 0.8533333333333334, + "grad_norm": 0.004064859822392464, + "learning_rate": 9.102301097269974e-06, + "loss": 1.1424087285995483, + "step": 60 + }, + { + "epoch": 0.8675555555555555, + "grad_norm": 0.004263672977685928, + "learning_rate": 9.05473103493355e-06, + "loss": 1.1524295806884766, + "step": 61 + }, + { + "epoch": 0.8817777777777778, + "grad_norm": 0.004161916207522154, + "learning_rate": 9.006064026541549e-06, + "loss": 1.1665949821472168, + "step": 62 + }, + { + "epoch": 0.896, + "grad_norm": 0.00454193539917469, + "learning_rate": 8.956313238215824e-06, + "loss": 1.1387211084365845, + "step": 63 + }, + { + "epoch": 0.9102222222222223, + "grad_norm": 0.004263832233846188, + "learning_rate": 8.905492129278478e-06, + "loss": 1.1322274208068848, + "step": 64 + }, + { + "epoch": 0.9244444444444444, + "grad_norm": 0.004334372468292713, + "learning_rate": 8.85361444861063e-06, + "loss": 1.1179461479187012, + "step": 65 + }, + { + "epoch": 0.9386666666666666, + "grad_norm": 0.0041003478690981865, + "learning_rate": 8.800694230932885e-06, + "loss": 1.1307452917099, + "step": 66 + }, + { + "epoch": 0.9528888888888889, + "grad_norm": 0.004313213285058737, + "learning_rate": 8.74674579300843e-06, + "loss": 1.1472362279891968, + "step": 67 + }, + { + "epoch": 0.9671111111111111, + "grad_norm": 0.004027621820569038, + "learning_rate": 8.691783729769874e-06, + "loss": 1.140931487083435, + "step": 68 + }, + { + "epoch": 0.9813333333333333, + "grad_norm": 0.004112871363759041, + "learning_rate": 8.635822910370793e-06, + "loss": 1.115431547164917, + "step": 69 + }, + { + "epoch": 0.9955555555555555, + "grad_norm": 0.004120040684938431, + "learning_rate": 8.578878474163115e-06, + "loss": 1.1306861639022827, + "step": 70 + }, + { + "epoch": 1.0, + "grad_norm": 0.007299893535673618, + "learning_rate": 8.520965826601394e-06, + "loss": 1.1355249881744385, + "step": 71 + }, + { + "epoch": 1.0142222222222221, + "grad_norm": 0.004595829173922539, + "learning_rate": 8.462100635075097e-06, + "loss": 1.098466157913208, + "step": 72 + }, + { + "epoch": 1.0284444444444445, + "grad_norm": 0.004596133250743151, + "learning_rate": 8.40229882467003e-06, + "loss": 1.1134223937988281, + "step": 73 + }, + { + "epoch": 1.0426666666666666, + "grad_norm": 0.0047966958954930305, + "learning_rate": 8.341576573860049e-06, + "loss": 1.1099815368652344, + "step": 74 + }, + { + "epoch": 1.056888888888889, + "grad_norm": 0.0042020766995847225, + "learning_rate": 8.279950310130218e-06, + "loss": 1.1130825281143188, + "step": 75 + }, + { + "epoch": 1.0711111111111111, + "grad_norm": 0.004611601121723652, + "learning_rate": 8.2174367055326e-06, + "loss": 1.1052048206329346, + "step": 76 + }, + { + "epoch": 1.0853333333333333, + "grad_norm": 0.004341403488069773, + "learning_rate": 8.154052672175888e-06, + "loss": 1.087442398071289, + "step": 77 + }, + { + "epoch": 1.0995555555555556, + "grad_norm": 0.004340282175689936, + "learning_rate": 8.08981535765009e-06, + "loss": 1.0848444700241089, + "step": 78 + }, + { + "epoch": 1.1137777777777778, + "grad_norm": 0.004358356352895498, + "learning_rate": 8.024742140387506e-06, + "loss": 1.1008961200714111, + "step": 79 + }, + { + "epoch": 1.1280000000000001, + "grad_norm": 0.004328620154410601, + "learning_rate": 7.95885062496126e-06, + "loss": 1.098643183708191, + "step": 80 + }, + { + "epoch": 1.1422222222222222, + "grad_norm": 0.004357457626610994, + "learning_rate": 7.892158637322647e-06, + "loss": 1.098900318145752, + "step": 81 + }, + { + "epoch": 1.1564444444444444, + "grad_norm": 0.004075393080711365, + "learning_rate": 7.824684219978591e-06, + "loss": 1.0768826007843018, + "step": 82 + }, + { + "epoch": 1.1706666666666667, + "grad_norm": 0.004524864722043276, + "learning_rate": 7.756445627110523e-06, + "loss": 1.1066226959228516, + "step": 83 + }, + { + "epoch": 1.1848888888888889, + "grad_norm": 0.004255601204931736, + "learning_rate": 7.68746131963598e-06, + "loss": 1.085633397102356, + "step": 84 + }, + { + "epoch": 1.199111111111111, + "grad_norm": 0.004353133961558342, + "learning_rate": 7.6177499602143e-06, + "loss": 1.0916088819503784, + "step": 85 + }, + { + "epoch": 1.2133333333333334, + "grad_norm": 0.004467015620321035, + "learning_rate": 7.547330408197695e-06, + "loss": 1.1167348623275757, + "step": 86 + }, + { + "epoch": 1.2275555555555555, + "grad_norm": 0.004218449350446463, + "learning_rate": 7.476221714529167e-06, + "loss": 1.093672513961792, + "step": 87 + }, + { + "epoch": 1.2417777777777779, + "grad_norm": 0.0046919044107198715, + "learning_rate": 7.404443116588548e-06, + "loss": 1.0770323276519775, + "step": 88 + }, + { + "epoch": 1.256, + "grad_norm": 0.004413353744894266, + "learning_rate": 7.332014032988123e-06, + "loss": 1.07706618309021, + "step": 89 + }, + { + "epoch": 1.2702222222222221, + "grad_norm": 0.004050100222229958, + "learning_rate": 7.2589540583192165e-06, + "loss": 1.1019604206085205, + "step": 90 + }, + { + "epoch": 1.2844444444444445, + "grad_norm": 0.004370351787656546, + "learning_rate": 7.185282957851175e-06, + "loss": 1.0928822755813599, + "step": 91 + }, + { + "epoch": 1.2986666666666666, + "grad_norm": 0.004322490654885769, + "learning_rate": 7.111020662184174e-06, + "loss": 1.083510160446167, + "step": 92 + }, + { + "epoch": 1.3128888888888888, + "grad_norm": 0.004432035610079765, + "learning_rate": 7.036187261857289e-06, + "loss": 1.0836291313171387, + "step": 93 + }, + { + "epoch": 1.3271111111111111, + "grad_norm": 0.004410918802022934, + "learning_rate": 6.960803001913315e-06, + "loss": 1.0802290439605713, + "step": 94 + }, + { + "epoch": 1.3413333333333333, + "grad_norm": 0.004076072946190834, + "learning_rate": 6.884888276421766e-06, + "loss": 1.0798417329788208, + "step": 95 + }, + { + "epoch": 1.3555555555555556, + "grad_norm": 0.004036934580653906, + "learning_rate": 6.8084636229615786e-06, + "loss": 1.0784028768539429, + "step": 96 + }, + { + "epoch": 1.3697777777777778, + "grad_norm": 0.004370865412056446, + "learning_rate": 6.731549717064975e-06, + "loss": 1.0908350944519043, + "step": 97 + }, + { + "epoch": 1.384, + "grad_norm": 0.004513819236308336, + "learning_rate": 6.654167366624009e-06, + "loss": 1.1067761182785034, + "step": 98 + }, + { + "epoch": 1.3982222222222223, + "grad_norm": 0.004452561028301716, + "learning_rate": 6.576337506261314e-06, + "loss": 1.0768399238586426, + "step": 99 + }, + { + "epoch": 1.4124444444444444, + "grad_norm": 0.00441731046885252, + "learning_rate": 6.498081191666549e-06, + "loss": 1.087970495223999, + "step": 100 + }, + { + "epoch": 1.4124444444444444, + "eval_loss": 1.0954939126968384, + "eval_runtime": 48.1789, + "eval_samples_per_second": 4.151, + "eval_steps_per_second": 2.076, + "step": 100 + }, + { + "epoch": 1.4266666666666667, + "grad_norm": 0.004392492119222879, + "learning_rate": 6.419419593900109e-06, + "loss": 1.0731995105743408, + "step": 101 + }, + { + "epoch": 1.4408888888888889, + "grad_norm": 0.004050276707857847, + "learning_rate": 6.340373993665607e-06, + "loss": 1.0977025032043457, + "step": 102 + }, + { + "epoch": 1.455111111111111, + "grad_norm": 0.004420506302267313, + "learning_rate": 6.2609657755527135e-06, + "loss": 1.0753436088562012, + "step": 103 + }, + { + "epoch": 1.4693333333333334, + "grad_norm": 0.004381162114441395, + "learning_rate": 6.1812164222518626e-06, + "loss": 1.0888622999191284, + "step": 104 + }, + { + "epoch": 1.4835555555555555, + "grad_norm": 0.004510694183409214, + "learning_rate": 6.101147508742456e-06, + "loss": 1.0883121490478516, + "step": 105 + }, + { + "epoch": 1.4977777777777779, + "grad_norm": 0.004262086004018784, + "learning_rate": 6.020780696456059e-06, + "loss": 1.0957694053649902, + "step": 106 + }, + { + "epoch": 1.512, + "grad_norm": 0.004126217216253281, + "learning_rate": 5.940137727416247e-06, + "loss": 1.0853359699249268, + "step": 107 + }, + { + "epoch": 1.5262222222222221, + "grad_norm": 0.004340359941124916, + "learning_rate": 5.859240418356614e-06, + "loss": 1.072435736656189, + "step": 108 + }, + { + "epoch": 1.5404444444444443, + "grad_norm": 0.0040999967604875565, + "learning_rate": 5.778110654818602e-06, + "loss": 1.0823040008544922, + "step": 109 + }, + { + "epoch": 1.5546666666666666, + "grad_norm": 0.004626137670129538, + "learning_rate": 5.696770385230679e-06, + "loss": 1.0717535018920898, + "step": 110 + }, + { + "epoch": 1.568888888888889, + "grad_norm": 0.00404947716742754, + "learning_rate": 5.615241614970546e-06, + "loss": 1.091806411743164, + "step": 111 + }, + { + "epoch": 1.5831111111111111, + "grad_norm": 0.004189760889858007, + "learning_rate": 5.533546400411899e-06, + "loss": 1.0810648202896118, + "step": 112 + }, + { + "epoch": 1.5973333333333333, + "grad_norm": 0.004373203497380018, + "learning_rate": 5.4517068429574215e-06, + "loss": 1.082428216934204, + "step": 113 + }, + { + "epoch": 1.6115555555555554, + "grad_norm": 0.0044046551920473576, + "learning_rate": 5.3697450830595775e-06, + "loss": 1.087543249130249, + "step": 114 + }, + { + "epoch": 1.6257777777777778, + "grad_norm": 0.004288993310183287, + "learning_rate": 5.287683294230855e-06, + "loss": 1.0576050281524658, + "step": 115 + }, + { + "epoch": 1.6400000000000001, + "grad_norm": 0.004459703341126442, + "learning_rate": 5.20554367704505e-06, + "loss": 1.0710945129394531, + "step": 116 + }, + { + "epoch": 1.6542222222222223, + "grad_norm": 0.004256462212651968, + "learning_rate": 5.123348453131242e-06, + "loss": 1.0972342491149902, + "step": 117 + }, + { + "epoch": 1.6684444444444444, + "grad_norm": 0.004406606312841177, + "learning_rate": 5.041119859162068e-06, + "loss": 1.0853114128112793, + "step": 118 + }, + { + "epoch": 1.6826666666666665, + "grad_norm": 0.00414910726249218, + "learning_rate": 4.958880140837934e-06, + "loss": 1.0747385025024414, + "step": 119 + }, + { + "epoch": 1.696888888888889, + "grad_norm": 0.004288721829652786, + "learning_rate": 4.876651546868759e-06, + "loss": 1.062084674835205, + "step": 120 + }, + { + "epoch": 1.7111111111111112, + "grad_norm": 0.004232042469084263, + "learning_rate": 4.794456322954953e-06, + "loss": 1.0754766464233398, + "step": 121 + }, + { + "epoch": 1.7253333333333334, + "grad_norm": 0.004211317747831345, + "learning_rate": 4.7123167057691446e-06, + "loss": 1.073303461074829, + "step": 122 + }, + { + "epoch": 1.7395555555555555, + "grad_norm": 0.004253690596669912, + "learning_rate": 4.630254916940424e-06, + "loss": 1.0916731357574463, + "step": 123 + }, + { + "epoch": 1.7537777777777777, + "grad_norm": 0.004212565720081329, + "learning_rate": 4.548293157042581e-06, + "loss": 1.0850458145141602, + "step": 124 + }, + { + "epoch": 1.768, + "grad_norm": 0.004267064854502678, + "learning_rate": 4.466453599588103e-06, + "loss": 1.0751748085021973, + "step": 125 + }, + { + "epoch": 1.7822222222222224, + "grad_norm": 0.004149065352976322, + "learning_rate": 4.384758385029457e-06, + "loss": 1.0639723539352417, + "step": 126 + }, + { + "epoch": 1.7964444444444445, + "grad_norm": 0.004006257280707359, + "learning_rate": 4.3032296147693225e-06, + "loss": 1.0787277221679688, + "step": 127 + }, + { + "epoch": 1.8106666666666666, + "grad_norm": 0.004197373986244202, + "learning_rate": 4.2218893451814005e-06, + "loss": 1.0513901710510254, + "step": 128 + }, + { + "epoch": 1.8248888888888888, + "grad_norm": 0.004353092983365059, + "learning_rate": 4.140759581643386e-06, + "loss": 1.0729951858520508, + "step": 129 + }, + { + "epoch": 1.8391111111111111, + "grad_norm": 0.003996321931481361, + "learning_rate": 4.059862272583755e-06, + "loss": 1.07584810256958, + "step": 130 + }, + { + "epoch": 1.8533333333333335, + "grad_norm": 0.0042655752040445805, + "learning_rate": 3.979219303543942e-06, + "loss": 1.0640318393707275, + "step": 131 + }, + { + "epoch": 1.8675555555555556, + "grad_norm": 0.004138551652431488, + "learning_rate": 3.898852491257547e-06, + "loss": 1.074927568435669, + "step": 132 + }, + { + "epoch": 1.8817777777777778, + "grad_norm": 0.0041953204199671745, + "learning_rate": 3.818783577748138e-06, + "loss": 1.0660316944122314, + "step": 133 + }, + { + "epoch": 1.896, + "grad_norm": 0.004029198549687862, + "learning_rate": 3.7390342244472886e-06, + "loss": 1.0795130729675293, + "step": 134 + }, + { + "epoch": 1.9102222222222223, + "grad_norm": 0.004120152443647385, + "learning_rate": 3.659626006334395e-06, + "loss": 1.094170093536377, + "step": 135 + }, + { + "epoch": 1.9244444444444444, + "grad_norm": 0.0042063528671860695, + "learning_rate": 3.5805804060998926e-06, + "loss": 1.0817457437515259, + "step": 136 + }, + { + "epoch": 1.9386666666666668, + "grad_norm": 0.004157442133873701, + "learning_rate": 3.501918808333453e-06, + "loss": 1.0608152151107788, + "step": 137 + }, + { + "epoch": 1.952888888888889, + "grad_norm": 0.004061015788465738, + "learning_rate": 3.4236624937386874e-06, + "loss": 1.0709357261657715, + "step": 138 + }, + { + "epoch": 1.967111111111111, + "grad_norm": 0.004056522622704506, + "learning_rate": 3.3458326333759927e-06, + "loss": 1.0752217769622803, + "step": 139 + }, + { + "epoch": 1.9813333333333332, + "grad_norm": 0.004041316919028759, + "learning_rate": 3.268450282935026e-06, + "loss": 1.0565910339355469, + "step": 140 + }, + { + "epoch": 1.9955555555555555, + "grad_norm": 0.00407154718413949, + "learning_rate": 3.1915363770384223e-06, + "loss": 1.064276099205017, + "step": 141 + }, + { + "epoch": 2.0, + "grad_norm": 0.006891065277159214, + "learning_rate": 3.1151117235782346e-06, + "loss": 1.0892033576965332, + "step": 142 + } + ], + "logging_steps": 1, + "max_steps": 213, + "num_input_tokens_seen": 0, + "num_train_epochs": 3, + "save_steps": 500, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 6.374528159538217e+17, + "train_batch_size": 1, + "trial_name": null, + "trial_params": null +} diff --git a/qwen2.5-3b-instruct/checkpoint-142/training_args.bin b/qwen2.5-3b-instruct/checkpoint-142/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..745b9d6d614f3064b70369b5149a4865ef67205e --- /dev/null +++ b/qwen2.5-3b-instruct/checkpoint-142/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:84503fc1ab703b8b0cb47dbab1adb9be59585b090c011d1af556d442423fd3c6 +size 5585 diff --git a/qwen2.5-3b-instruct/checkpoint-213/chat_template.jinja b/qwen2.5-3b-instruct/checkpoint-213/chat_template.jinja new file mode 100644 index 0000000000000000000000000000000000000000..bdf7919a96cfe43d50914a007b9c0877bd0ec27e --- /dev/null +++ b/qwen2.5-3b-instruct/checkpoint-213/chat_template.jinja @@ -0,0 +1,54 @@ +{%- if tools %} + {{- '<|im_start|>system\n' }} + {%- if messages[0]['role'] == 'system' %} + {{- messages[0]['content'] }} + {%- else %} + {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }} + {%- endif %} + {{- "\n\n# 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' }} + {%- else %} + {{- '<|im_start|>system\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\n' }} + {%- endif %} +{%- endif %} +{%- for message in messages %} + {%- if (message.role == "user") or (message.role == "system" and not loop.first) or (message.role == "assistant" and not message.tool_calls) %} + {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }} + {%- elif message.role == "assistant" %} + {{- '<|im_start|>' + message.role }} + {%- if message.content %} + {{- '\n' + message.content }} + {%- endif %} + {%- for tool_call in message.tool_calls %} + {%- if tool_call.function is defined %} + {%- set tool_call = tool_call.function %} + {%- endif %} + {{- '\n\n{"name": "' }} + {{- tool_call.name }} + {{- '", "arguments": ' }} + {{- tool_call.arguments | tojson }} + {{- '}\n' }} + {%- endfor %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "tool" %} + {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != "tool") %} + {{- '<|im_start|>user' }} + {%- endif %} + {{- '\n\n' }} + {{- message.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' }} +{%- endif %} diff --git a/qwen2.5-3b-instruct/checkpoint-213/config.json b/qwen2.5-3b-instruct/checkpoint-213/config.json new file mode 100644 index 0000000000000000000000000000000000000000..4814292863d1e5aa2f4fda1db18f4082cd61aa93 --- /dev/null +++ b/qwen2.5-3b-instruct/checkpoint-213/config.json @@ -0,0 +1,69 @@ +{ + "architectures": [ + "Qwen2ForCausalLM" + ], + "attention_dropout": 0.0, + "bos_token_id": null, + "dtype": "float32", + "eos_token_id": 151645, + "hidden_act": "silu", + "hidden_size": 2048, + "initializer_range": 0.02, + "intermediate_size": 11008, + "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": 32768, + "max_window_layers": 70, + "model_type": "qwen2", + "num_attention_heads": 16, + "num_hidden_layers": 36, + "num_key_value_heads": 2, + "pad_token_id": 151643, + "rms_norm_eps": 1e-06, + "rope_parameters": { + "rope_theta": 1000000.0, + "rope_type": "default" + }, + "sliding_window": null, + "tie_word_embeddings": true, + "transformers_version": "5.2.0", + "use_cache": false, + "use_sliding_window": false, + "vocab_size": 151936 +} diff --git a/qwen2.5-3b-instruct/checkpoint-213/generation_config.json b/qwen2.5-3b-instruct/checkpoint-213/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..47205fbf3386a215e7c2b9d62df6576f49b77e7b --- /dev/null +++ b/qwen2.5-3b-instruct/checkpoint-213/generation_config.json @@ -0,0 +1,13 @@ +{ + "do_sample": true, + "eos_token_id": [ + 151645, + 151643 + ], + "pad_token_id": 151643, + "repetition_penalty": 1.05, + "temperature": 0.7, + "top_k": 20, + "top_p": 0.8, + "transformers_version": "5.2.0" +} diff --git a/qwen2.5-3b-instruct/checkpoint-213/model.safetensors b/qwen2.5-3b-instruct/checkpoint-213/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..74524f99eff11d5088665628147ba894149b6000 --- /dev/null +++ b/qwen2.5-3b-instruct/checkpoint-213/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d0e8497d095a6f44dad6178b028fb64ddba2353611464e5056ab51f620ec21c7 +size 12343804296 diff --git a/qwen2.5-3b-instruct/checkpoint-213/optimizer.pt b/qwen2.5-3b-instruct/checkpoint-213/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..869db51d233d1b0c1910d12e7278d5b22a22d623 --- /dev/null +++ b/qwen2.5-3b-instruct/checkpoint-213/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1aa3b67cdf2298a099454f83f43421be0800899bccb77ae5d047a075a4cabde0 +size 24687895753 diff --git a/qwen2.5-3b-instruct/checkpoint-213/rng_state_0.pth b/qwen2.5-3b-instruct/checkpoint-213/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..870021e3fa5ac35c2f711adf0c93a556ab4842da --- /dev/null +++ b/qwen2.5-3b-instruct/checkpoint-213/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:95e5fc2074c0df31522a514f862c86cb00d71c946a7f15cc9ec0e53a69fb28a7 +size 14917 diff --git a/qwen2.5-3b-instruct/checkpoint-213/rng_state_1.pth b/qwen2.5-3b-instruct/checkpoint-213/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..21f20da5eb1da017f08aaa88bd19cf24d40e3fbf --- /dev/null +++ b/qwen2.5-3b-instruct/checkpoint-213/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0e7153eae67b6c9232a41bc996a2bf5b83229b8c7230d61911ac0fd40e64154e +size 14917 diff --git a/qwen2.5-3b-instruct/checkpoint-213/scheduler.pt b/qwen2.5-3b-instruct/checkpoint-213/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..39cf87775c3a71d10e7311e0cab769e112f357ba --- /dev/null +++ b/qwen2.5-3b-instruct/checkpoint-213/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c38a67b98c17be81d4c6d00b60029cdedd43e9aba35ca28ec28ef946c400f268 +size 1465 diff --git a/qwen2.5-3b-instruct/checkpoint-213/tokenizer.json b/qwen2.5-3b-instruct/checkpoint-213/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..34510ff0037cd50428af467a17ead5a96140a32c --- /dev/null +++ b/qwen2.5-3b-instruct/checkpoint-213/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3fd169731d2cbde95e10bf356d66d5997fd885dd8dbb6fb4684da3f23b2585d8 +size 11421892 diff --git a/qwen2.5-3b-instruct/checkpoint-213/tokenizer_config.json b/qwen2.5-3b-instruct/checkpoint-213/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..145e2c7a68b109ae0253dda29fc2e66cb01b4b85 --- /dev/null +++ b/qwen2.5-3b-instruct/checkpoint-213/tokenizer_config.json @@ -0,0 +1,30 @@ +{ + "add_prefix_space": false, + "backend": "tokenizers", + "bos_token": null, + "clean_up_tokenization_spaces": false, + "eos_token": "<|im_end|>", + "errors": "replace", + "extra_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|>" + ], + "is_local": false, + "model_max_length": 131072, + "pad_token": "<|endoftext|>", + "padding_side": "right", + "split_special_tokens": false, + "tokenizer_class": "Qwen2Tokenizer", + "unk_token": null +} diff --git a/qwen2.5-3b-instruct/checkpoint-213/trainer_state.json b/qwen2.5-3b-instruct/checkpoint-213/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..abecd7c22c6918d3abd79fea3ec3e3bce97a4ece --- /dev/null +++ b/qwen2.5-3b-instruct/checkpoint-213/trainer_state.json @@ -0,0 +1,1557 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 3.0, + "eval_steps": 50, + "global_step": 213, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.014222222222222223, + "grad_norm": 0.07561660557985306, + "learning_rate": 0.0, + "loss": 1.6462128162384033, + "step": 1 + }, + { + "epoch": 0.028444444444444446, + "grad_norm": 0.07576845586299896, + "learning_rate": 4.5454545454545457e-07, + "loss": 1.6739654541015625, + "step": 2 + }, + { + "epoch": 0.042666666666666665, + "grad_norm": 0.060671016573905945, + "learning_rate": 9.090909090909091e-07, + "loss": 1.6208744049072266, + "step": 3 + }, + { + "epoch": 0.05688888888888889, + "grad_norm": 0.04539854824542999, + "learning_rate": 1.3636363636363636e-06, + "loss": 1.5681328773498535, + "step": 4 + }, + { + "epoch": 0.07111111111111111, + "grad_norm": 0.030098970979452133, + "learning_rate": 1.8181818181818183e-06, + "loss": 1.5172258615493774, + "step": 5 + }, + { + "epoch": 0.08533333333333333, + "grad_norm": 0.02540394477546215, + "learning_rate": 2.2727272727272728e-06, + "loss": 1.4882113933563232, + "step": 6 + }, + { + "epoch": 0.09955555555555555, + "grad_norm": 0.026057474315166473, + "learning_rate": 2.7272727272727272e-06, + "loss": 1.453606128692627, + "step": 7 + }, + { + "epoch": 0.11377777777777778, + "grad_norm": 0.024274546653032303, + "learning_rate": 3.181818181818182e-06, + "loss": 1.433072805404663, + "step": 8 + }, + { + "epoch": 0.128, + "grad_norm": 0.02080286666750908, + "learning_rate": 3.6363636363636366e-06, + "loss": 1.4036792516708374, + "step": 9 + }, + { + "epoch": 0.14222222222222222, + "grad_norm": 0.021423645317554474, + "learning_rate": 4.0909090909090915e-06, + "loss": 1.3751729726791382, + "step": 10 + }, + { + "epoch": 0.15644444444444444, + "grad_norm": 0.018808694556355476, + "learning_rate": 4.5454545454545455e-06, + "loss": 1.3682663440704346, + "step": 11 + }, + { + "epoch": 0.17066666666666666, + "grad_norm": 0.013645165599882603, + "learning_rate": 5e-06, + "loss": 1.3491919040679932, + "step": 12 + }, + { + "epoch": 0.18488888888888888, + "grad_norm": 0.012575564906001091, + "learning_rate": 5.4545454545454545e-06, + "loss": 1.3444881439208984, + "step": 13 + }, + { + "epoch": 0.1991111111111111, + "grad_norm": 0.011944238096475601, + "learning_rate": 5.90909090909091e-06, + "loss": 1.3321316242218018, + "step": 14 + }, + { + "epoch": 0.21333333333333335, + "grad_norm": 0.012251932173967361, + "learning_rate": 6.363636363636364e-06, + "loss": 1.3083908557891846, + "step": 15 + }, + { + "epoch": 0.22755555555555557, + "grad_norm": 0.011485234834253788, + "learning_rate": 6.818181818181818e-06, + "loss": 1.287674903869629, + "step": 16 + }, + { + "epoch": 0.24177777777777779, + "grad_norm": 0.010052796453237534, + "learning_rate": 7.272727272727273e-06, + "loss": 1.2796708345413208, + "step": 17 + }, + { + "epoch": 0.256, + "grad_norm": 0.009297606535255909, + "learning_rate": 7.727272727272727e-06, + "loss": 1.2778702974319458, + "step": 18 + }, + { + "epoch": 0.2702222222222222, + "grad_norm": 0.009387261234223843, + "learning_rate": 8.181818181818183e-06, + "loss": 1.258772373199463, + "step": 19 + }, + { + "epoch": 0.28444444444444444, + "grad_norm": 0.009283723309636116, + "learning_rate": 8.636363636363637e-06, + "loss": 1.2937917709350586, + "step": 20 + }, + { + "epoch": 0.2986666666666667, + "grad_norm": 0.008504457771778107, + "learning_rate": 9.090909090909091e-06, + "loss": 1.247847318649292, + "step": 21 + }, + { + "epoch": 0.3128888888888889, + "grad_norm": 0.007998796179890633, + "learning_rate": 9.545454545454547e-06, + "loss": 1.235945701599121, + "step": 22 + }, + { + "epoch": 0.32711111111111113, + "grad_norm": 0.007681620307266712, + "learning_rate": 1e-05, + "loss": 1.2462598085403442, + "step": 23 + }, + { + "epoch": 0.3413333333333333, + "grad_norm": 0.007152756676077843, + "learning_rate": 9.999323662872998e-06, + "loss": 1.2484906911849976, + "step": 24 + }, + { + "epoch": 0.35555555555555557, + "grad_norm": 0.00654159439727664, + "learning_rate": 9.99729483446475e-06, + "loss": 1.2089664936065674, + "step": 25 + }, + { + "epoch": 0.36977777777777776, + "grad_norm": 0.0063508739694952965, + "learning_rate": 9.993914063644053e-06, + "loss": 1.2245509624481201, + "step": 26 + }, + { + "epoch": 0.384, + "grad_norm": 0.006452301051467657, + "learning_rate": 9.989182265027232e-06, + "loss": 1.2176213264465332, + "step": 27 + }, + { + "epoch": 0.3982222222222222, + "grad_norm": 0.006349734030663967, + "learning_rate": 9.98310071873072e-06, + "loss": 1.2157793045043945, + "step": 28 + }, + { + "epoch": 0.41244444444444445, + "grad_norm": 0.006191881373524666, + "learning_rate": 9.975671070024741e-06, + "loss": 1.237949013710022, + "step": 29 + }, + { + "epoch": 0.4266666666666667, + "grad_norm": 0.005793207325041294, + "learning_rate": 9.966895328888195e-06, + "loss": 1.21578049659729, + "step": 30 + }, + { + "epoch": 0.4408888888888889, + "grad_norm": 0.005483992397785187, + "learning_rate": 9.956775869464901e-06, + "loss": 1.2089630365371704, + "step": 31 + }, + { + "epoch": 0.45511111111111113, + "grad_norm": 0.005602920427918434, + "learning_rate": 9.945315429421307e-06, + "loss": 1.2175853252410889, + "step": 32 + }, + { + "epoch": 0.4693333333333333, + "grad_norm": 0.005212367512285709, + "learning_rate": 9.932517109205849e-06, + "loss": 1.2096197605133057, + "step": 33 + }, + { + "epoch": 0.48355555555555557, + "grad_norm": 0.005095571745187044, + "learning_rate": 9.918384371210178e-06, + "loss": 1.1974353790283203, + "step": 34 + }, + { + "epoch": 0.49777777777777776, + "grad_norm": 0.005125134252011776, + "learning_rate": 9.902921038832456e-06, + "loss": 1.1847140789031982, + "step": 35 + }, + { + "epoch": 0.512, + "grad_norm": 0.004882226698100567, + "learning_rate": 9.886131295443003e-06, + "loss": 1.2025952339172363, + "step": 36 + }, + { + "epoch": 0.5262222222222223, + "grad_norm": 0.004733556881546974, + "learning_rate": 9.868019683252543e-06, + "loss": 1.193751573562622, + "step": 37 + }, + { + "epoch": 0.5404444444444444, + "grad_norm": 0.004847981501370668, + "learning_rate": 9.848591102083375e-06, + "loss": 1.1725599765777588, + "step": 38 + }, + { + "epoch": 0.5546666666666666, + "grad_norm": 0.004716528579592705, + "learning_rate": 9.82785080804381e-06, + "loss": 1.1763365268707275, + "step": 39 + }, + { + "epoch": 0.5688888888888889, + "grad_norm": 0.004721782635897398, + "learning_rate": 9.805804412106197e-06, + "loss": 1.180063009262085, + "step": 40 + }, + { + "epoch": 0.5831111111111111, + "grad_norm": 0.004470364656299353, + "learning_rate": 9.782457878588977e-06, + "loss": 1.169750690460205, + "step": 41 + }, + { + "epoch": 0.5973333333333334, + "grad_norm": 0.0046170237474143505, + "learning_rate": 9.75781752354311e-06, + "loss": 1.174955129623413, + "step": 42 + }, + { + "epoch": 0.6115555555555555, + "grad_norm": 0.004407198633998632, + "learning_rate": 9.731890013043367e-06, + "loss": 1.1694097518920898, + "step": 43 + }, + { + "epoch": 0.6257777777777778, + "grad_norm": 0.004441778641194105, + "learning_rate": 9.704682361384941e-06, + "loss": 1.1582658290863037, + "step": 44 + }, + { + "epoch": 0.64, + "grad_norm": 0.004318015184253454, + "learning_rate": 9.676201929185809e-06, + "loss": 1.1544437408447266, + "step": 45 + }, + { + "epoch": 0.6542222222222223, + "grad_norm": 0.004480153322219849, + "learning_rate": 9.646456421395447e-06, + "loss": 1.1822636127471924, + "step": 46 + }, + { + "epoch": 0.6684444444444444, + "grad_norm": 0.004228127654641867, + "learning_rate": 9.615453885210368e-06, + "loss": 1.1748132705688477, + "step": 47 + }, + { + "epoch": 0.6826666666666666, + "grad_norm": 0.004379024263471365, + "learning_rate": 9.583202707897075e-06, + "loss": 1.1558542251586914, + "step": 48 + }, + { + "epoch": 0.6968888888888889, + "grad_norm": 0.004313027486205101, + "learning_rate": 9.549711614523007e-06, + "loss": 1.1546211242675781, + "step": 49 + }, + { + "epoch": 0.7111111111111111, + "grad_norm": 0.004527685698121786, + "learning_rate": 9.514989665596114e-06, + "loss": 1.1557996273040771, + "step": 50 + }, + { + "epoch": 0.7111111111111111, + "eval_loss": 1.1428771018981934, + "eval_runtime": 47.9864, + "eval_samples_per_second": 4.168, + "eval_steps_per_second": 2.084, + "step": 50 + }, + { + "epoch": 0.7253333333333334, + "grad_norm": 0.0040644253604114056, + "learning_rate": 9.479046254613673e-06, + "loss": 1.172366976737976, + "step": 51 + }, + { + "epoch": 0.7395555555555555, + "grad_norm": 0.004353945609182119, + "learning_rate": 9.441891105521005e-06, + "loss": 1.1539039611816406, + "step": 52 + }, + { + "epoch": 0.7537777777777778, + "grad_norm": 0.004351011943072081, + "learning_rate": 9.40353427008083e-06, + "loss": 1.159185767173767, + "step": 53 + }, + { + "epoch": 0.768, + "grad_norm": 0.004263842478394508, + "learning_rate": 9.3639861251539e-06, + "loss": 1.1381170749664307, + "step": 54 + }, + { + "epoch": 0.7822222222222223, + "grad_norm": 0.0042404914274811745, + "learning_rate": 9.323257369891702e-06, + "loss": 1.1402206420898438, + "step": 55 + }, + { + "epoch": 0.7964444444444444, + "grad_norm": 0.004360774997621775, + "learning_rate": 9.281359022841966e-06, + "loss": 1.1499946117401123, + "step": 56 + }, + { + "epoch": 0.8106666666666666, + "grad_norm": 0.004249541088938713, + "learning_rate": 9.238302418967757e-06, + "loss": 1.1507560014724731, + "step": 57 + }, + { + "epoch": 0.8248888888888889, + "grad_norm": 0.004316243343055248, + "learning_rate": 9.194099206580981e-06, + "loss": 1.14259934425354, + "step": 58 + }, + { + "epoch": 0.8391111111111111, + "grad_norm": 0.004926053807139397, + "learning_rate": 9.14876134419111e-06, + "loss": 1.1445270776748657, + "step": 59 + }, + { + "epoch": 0.8533333333333334, + "grad_norm": 0.004064859822392464, + "learning_rate": 9.102301097269974e-06, + "loss": 1.1424087285995483, + "step": 60 + }, + { + "epoch": 0.8675555555555555, + "grad_norm": 0.004263672977685928, + "learning_rate": 9.05473103493355e-06, + "loss": 1.1524295806884766, + "step": 61 + }, + { + "epoch": 0.8817777777777778, + "grad_norm": 0.004161916207522154, + "learning_rate": 9.006064026541549e-06, + "loss": 1.1665949821472168, + "step": 62 + }, + { + "epoch": 0.896, + "grad_norm": 0.00454193539917469, + "learning_rate": 8.956313238215824e-06, + "loss": 1.1387211084365845, + "step": 63 + }, + { + "epoch": 0.9102222222222223, + "grad_norm": 0.004263832233846188, + "learning_rate": 8.905492129278478e-06, + "loss": 1.1322274208068848, + "step": 64 + }, + { + "epoch": 0.9244444444444444, + "grad_norm": 0.004334372468292713, + "learning_rate": 8.85361444861063e-06, + "loss": 1.1179461479187012, + "step": 65 + }, + { + "epoch": 0.9386666666666666, + "grad_norm": 0.0041003478690981865, + "learning_rate": 8.800694230932885e-06, + "loss": 1.1307452917099, + "step": 66 + }, + { + "epoch": 0.9528888888888889, + "grad_norm": 0.004313213285058737, + "learning_rate": 8.74674579300843e-06, + "loss": 1.1472362279891968, + "step": 67 + }, + { + "epoch": 0.9671111111111111, + "grad_norm": 0.004027621820569038, + "learning_rate": 8.691783729769874e-06, + "loss": 1.140931487083435, + "step": 68 + }, + { + "epoch": 0.9813333333333333, + "grad_norm": 0.004112871363759041, + "learning_rate": 8.635822910370793e-06, + "loss": 1.115431547164917, + "step": 69 + }, + { + "epoch": 0.9955555555555555, + "grad_norm": 0.004120040684938431, + "learning_rate": 8.578878474163115e-06, + "loss": 1.1306861639022827, + "step": 70 + }, + { + "epoch": 1.0, + "grad_norm": 0.007299893535673618, + "learning_rate": 8.520965826601394e-06, + "loss": 1.1355249881744385, + "step": 71 + }, + { + "epoch": 1.0142222222222221, + "grad_norm": 0.004595829173922539, + "learning_rate": 8.462100635075097e-06, + "loss": 1.098466157913208, + "step": 72 + }, + { + "epoch": 1.0284444444444445, + "grad_norm": 0.004596133250743151, + "learning_rate": 8.40229882467003e-06, + "loss": 1.1134223937988281, + "step": 73 + }, + { + "epoch": 1.0426666666666666, + "grad_norm": 0.0047966958954930305, + "learning_rate": 8.341576573860049e-06, + "loss": 1.1099815368652344, + "step": 74 + }, + { + "epoch": 1.056888888888889, + "grad_norm": 0.0042020766995847225, + "learning_rate": 8.279950310130218e-06, + "loss": 1.1130825281143188, + "step": 75 + }, + { + "epoch": 1.0711111111111111, + "grad_norm": 0.004611601121723652, + "learning_rate": 8.2174367055326e-06, + "loss": 1.1052048206329346, + "step": 76 + }, + { + "epoch": 1.0853333333333333, + "grad_norm": 0.004341403488069773, + "learning_rate": 8.154052672175888e-06, + "loss": 1.087442398071289, + "step": 77 + }, + { + "epoch": 1.0995555555555556, + "grad_norm": 0.004340282175689936, + "learning_rate": 8.08981535765009e-06, + "loss": 1.0848444700241089, + "step": 78 + }, + { + "epoch": 1.1137777777777778, + "grad_norm": 0.004358356352895498, + "learning_rate": 8.024742140387506e-06, + "loss": 1.1008961200714111, + "step": 79 + }, + { + "epoch": 1.1280000000000001, + "grad_norm": 0.004328620154410601, + "learning_rate": 7.95885062496126e-06, + "loss": 1.098643183708191, + "step": 80 + }, + { + "epoch": 1.1422222222222222, + "grad_norm": 0.004357457626610994, + "learning_rate": 7.892158637322647e-06, + "loss": 1.098900318145752, + "step": 81 + }, + { + "epoch": 1.1564444444444444, + "grad_norm": 0.004075393080711365, + "learning_rate": 7.824684219978591e-06, + "loss": 1.0768826007843018, + "step": 82 + }, + { + "epoch": 1.1706666666666667, + "grad_norm": 0.004524864722043276, + "learning_rate": 7.756445627110523e-06, + "loss": 1.1066226959228516, + "step": 83 + }, + { + "epoch": 1.1848888888888889, + "grad_norm": 0.004255601204931736, + "learning_rate": 7.68746131963598e-06, + "loss": 1.085633397102356, + "step": 84 + }, + { + "epoch": 1.199111111111111, + "grad_norm": 0.004353133961558342, + "learning_rate": 7.6177499602143e-06, + "loss": 1.0916088819503784, + "step": 85 + }, + { + "epoch": 1.2133333333333334, + "grad_norm": 0.004467015620321035, + "learning_rate": 7.547330408197695e-06, + "loss": 1.1167348623275757, + "step": 86 + }, + { + "epoch": 1.2275555555555555, + "grad_norm": 0.004218449350446463, + "learning_rate": 7.476221714529167e-06, + "loss": 1.093672513961792, + "step": 87 + }, + { + "epoch": 1.2417777777777779, + "grad_norm": 0.0046919044107198715, + "learning_rate": 7.404443116588548e-06, + "loss": 1.0770323276519775, + "step": 88 + }, + { + "epoch": 1.256, + "grad_norm": 0.004413353744894266, + "learning_rate": 7.332014032988123e-06, + "loss": 1.07706618309021, + "step": 89 + }, + { + "epoch": 1.2702222222222221, + "grad_norm": 0.004050100222229958, + "learning_rate": 7.2589540583192165e-06, + "loss": 1.1019604206085205, + "step": 90 + }, + { + "epoch": 1.2844444444444445, + "grad_norm": 0.004370351787656546, + "learning_rate": 7.185282957851175e-06, + "loss": 1.0928822755813599, + "step": 91 + }, + { + "epoch": 1.2986666666666666, + "grad_norm": 0.004322490654885769, + "learning_rate": 7.111020662184174e-06, + "loss": 1.083510160446167, + "step": 92 + }, + { + "epoch": 1.3128888888888888, + "grad_norm": 0.004432035610079765, + "learning_rate": 7.036187261857289e-06, + "loss": 1.0836291313171387, + "step": 93 + }, + { + "epoch": 1.3271111111111111, + "grad_norm": 0.004410918802022934, + "learning_rate": 6.960803001913315e-06, + "loss": 1.0802290439605713, + "step": 94 + }, + { + "epoch": 1.3413333333333333, + "grad_norm": 0.004076072946190834, + "learning_rate": 6.884888276421766e-06, + "loss": 1.0798417329788208, + "step": 95 + }, + { + "epoch": 1.3555555555555556, + "grad_norm": 0.004036934580653906, + "learning_rate": 6.8084636229615786e-06, + "loss": 1.0784028768539429, + "step": 96 + }, + { + "epoch": 1.3697777777777778, + "grad_norm": 0.004370865412056446, + "learning_rate": 6.731549717064975e-06, + "loss": 1.0908350944519043, + "step": 97 + }, + { + "epoch": 1.384, + "grad_norm": 0.004513819236308336, + "learning_rate": 6.654167366624009e-06, + "loss": 1.1067761182785034, + "step": 98 + }, + { + "epoch": 1.3982222222222223, + "grad_norm": 0.004452561028301716, + "learning_rate": 6.576337506261314e-06, + "loss": 1.0768399238586426, + "step": 99 + }, + { + "epoch": 1.4124444444444444, + "grad_norm": 0.00441731046885252, + "learning_rate": 6.498081191666549e-06, + "loss": 1.087970495223999, + "step": 100 + }, + { + "epoch": 1.4124444444444444, + "eval_loss": 1.0954939126968384, + "eval_runtime": 48.1789, + "eval_samples_per_second": 4.151, + "eval_steps_per_second": 2.076, + "step": 100 + }, + { + "epoch": 1.4266666666666667, + "grad_norm": 0.004392492119222879, + "learning_rate": 6.419419593900109e-06, + "loss": 1.0731995105743408, + "step": 101 + }, + { + "epoch": 1.4408888888888889, + "grad_norm": 0.004050276707857847, + "learning_rate": 6.340373993665607e-06, + "loss": 1.0977025032043457, + "step": 102 + }, + { + "epoch": 1.455111111111111, + "grad_norm": 0.004420506302267313, + "learning_rate": 6.2609657755527135e-06, + "loss": 1.0753436088562012, + "step": 103 + }, + { + "epoch": 1.4693333333333334, + "grad_norm": 0.004381162114441395, + "learning_rate": 6.1812164222518626e-06, + "loss": 1.0888622999191284, + "step": 104 + }, + { + "epoch": 1.4835555555555555, + "grad_norm": 0.004510694183409214, + "learning_rate": 6.101147508742456e-06, + "loss": 1.0883121490478516, + "step": 105 + }, + { + "epoch": 1.4977777777777779, + "grad_norm": 0.004262086004018784, + "learning_rate": 6.020780696456059e-06, + "loss": 1.0957694053649902, + "step": 106 + }, + { + "epoch": 1.512, + "grad_norm": 0.004126217216253281, + "learning_rate": 5.940137727416247e-06, + "loss": 1.0853359699249268, + "step": 107 + }, + { + "epoch": 1.5262222222222221, + "grad_norm": 0.004340359941124916, + "learning_rate": 5.859240418356614e-06, + "loss": 1.072435736656189, + "step": 108 + }, + { + "epoch": 1.5404444444444443, + "grad_norm": 0.0040999967604875565, + "learning_rate": 5.778110654818602e-06, + "loss": 1.0823040008544922, + "step": 109 + }, + { + "epoch": 1.5546666666666666, + "grad_norm": 0.004626137670129538, + "learning_rate": 5.696770385230679e-06, + "loss": 1.0717535018920898, + "step": 110 + }, + { + "epoch": 1.568888888888889, + "grad_norm": 0.00404947716742754, + "learning_rate": 5.615241614970546e-06, + "loss": 1.091806411743164, + "step": 111 + }, + { + "epoch": 1.5831111111111111, + "grad_norm": 0.004189760889858007, + "learning_rate": 5.533546400411899e-06, + "loss": 1.0810648202896118, + "step": 112 + }, + { + "epoch": 1.5973333333333333, + "grad_norm": 0.004373203497380018, + "learning_rate": 5.4517068429574215e-06, + "loss": 1.082428216934204, + "step": 113 + }, + { + "epoch": 1.6115555555555554, + "grad_norm": 0.0044046551920473576, + "learning_rate": 5.3697450830595775e-06, + "loss": 1.087543249130249, + "step": 114 + }, + { + "epoch": 1.6257777777777778, + "grad_norm": 0.004288993310183287, + "learning_rate": 5.287683294230855e-06, + "loss": 1.0576050281524658, + "step": 115 + }, + { + "epoch": 1.6400000000000001, + "grad_norm": 0.004459703341126442, + "learning_rate": 5.20554367704505e-06, + "loss": 1.0710945129394531, + "step": 116 + }, + { + "epoch": 1.6542222222222223, + "grad_norm": 0.004256462212651968, + "learning_rate": 5.123348453131242e-06, + "loss": 1.0972342491149902, + "step": 117 + }, + { + "epoch": 1.6684444444444444, + "grad_norm": 0.004406606312841177, + "learning_rate": 5.041119859162068e-06, + "loss": 1.0853114128112793, + "step": 118 + }, + { + "epoch": 1.6826666666666665, + "grad_norm": 0.00414910726249218, + "learning_rate": 4.958880140837934e-06, + "loss": 1.0747385025024414, + "step": 119 + }, + { + "epoch": 1.696888888888889, + "grad_norm": 0.004288721829652786, + "learning_rate": 4.876651546868759e-06, + "loss": 1.062084674835205, + "step": 120 + }, + { + "epoch": 1.7111111111111112, + "grad_norm": 0.004232042469084263, + "learning_rate": 4.794456322954953e-06, + "loss": 1.0754766464233398, + "step": 121 + }, + { + "epoch": 1.7253333333333334, + "grad_norm": 0.004211317747831345, + "learning_rate": 4.7123167057691446e-06, + "loss": 1.073303461074829, + "step": 122 + }, + { + "epoch": 1.7395555555555555, + "grad_norm": 0.004253690596669912, + "learning_rate": 4.630254916940424e-06, + "loss": 1.0916731357574463, + "step": 123 + }, + { + "epoch": 1.7537777777777777, + "grad_norm": 0.004212565720081329, + "learning_rate": 4.548293157042581e-06, + "loss": 1.0850458145141602, + "step": 124 + }, + { + "epoch": 1.768, + "grad_norm": 0.004267064854502678, + "learning_rate": 4.466453599588103e-06, + "loss": 1.0751748085021973, + "step": 125 + }, + { + "epoch": 1.7822222222222224, + "grad_norm": 0.004149065352976322, + "learning_rate": 4.384758385029457e-06, + "loss": 1.0639723539352417, + "step": 126 + }, + { + "epoch": 1.7964444444444445, + "grad_norm": 0.004006257280707359, + "learning_rate": 4.3032296147693225e-06, + "loss": 1.0787277221679688, + "step": 127 + }, + { + "epoch": 1.8106666666666666, + "grad_norm": 0.004197373986244202, + "learning_rate": 4.2218893451814005e-06, + "loss": 1.0513901710510254, + "step": 128 + }, + { + "epoch": 1.8248888888888888, + "grad_norm": 0.004353092983365059, + "learning_rate": 4.140759581643386e-06, + "loss": 1.0729951858520508, + "step": 129 + }, + { + "epoch": 1.8391111111111111, + "grad_norm": 0.003996321931481361, + "learning_rate": 4.059862272583755e-06, + "loss": 1.07584810256958, + "step": 130 + }, + { + "epoch": 1.8533333333333335, + "grad_norm": 0.0042655752040445805, + "learning_rate": 3.979219303543942e-06, + "loss": 1.0640318393707275, + "step": 131 + }, + { + "epoch": 1.8675555555555556, + "grad_norm": 0.004138551652431488, + "learning_rate": 3.898852491257547e-06, + "loss": 1.074927568435669, + "step": 132 + }, + { + "epoch": 1.8817777777777778, + "grad_norm": 0.0041953204199671745, + "learning_rate": 3.818783577748138e-06, + "loss": 1.0660316944122314, + "step": 133 + }, + { + "epoch": 1.896, + "grad_norm": 0.004029198549687862, + "learning_rate": 3.7390342244472886e-06, + "loss": 1.0795130729675293, + "step": 134 + }, + { + "epoch": 1.9102222222222223, + "grad_norm": 0.004120152443647385, + "learning_rate": 3.659626006334395e-06, + "loss": 1.094170093536377, + "step": 135 + }, + { + "epoch": 1.9244444444444444, + "grad_norm": 0.0042063528671860695, + "learning_rate": 3.5805804060998926e-06, + "loss": 1.0817457437515259, + "step": 136 + }, + { + "epoch": 1.9386666666666668, + "grad_norm": 0.004157442133873701, + "learning_rate": 3.501918808333453e-06, + "loss": 1.0608152151107788, + "step": 137 + }, + { + "epoch": 1.952888888888889, + "grad_norm": 0.004061015788465738, + "learning_rate": 3.4236624937386874e-06, + "loss": 1.0709357261657715, + "step": 138 + }, + { + "epoch": 1.967111111111111, + "grad_norm": 0.004056522622704506, + "learning_rate": 3.3458326333759927e-06, + "loss": 1.0752217769622803, + "step": 139 + }, + { + "epoch": 1.9813333333333332, + "grad_norm": 0.004041316919028759, + "learning_rate": 3.268450282935026e-06, + "loss": 1.0565910339355469, + "step": 140 + }, + { + "epoch": 1.9955555555555555, + "grad_norm": 0.00407154718413949, + "learning_rate": 3.1915363770384223e-06, + "loss": 1.064276099205017, + "step": 141 + }, + { + "epoch": 2.0, + "grad_norm": 0.006891065277159214, + "learning_rate": 3.1151117235782346e-06, + "loss": 1.0892033576965332, + "step": 142 + }, + { + "epoch": 2.014222222222222, + "grad_norm": 0.004137777257710695, + "learning_rate": 3.0391969980866874e-06, + "loss": 1.0656744241714478, + "step": 143 + }, + { + "epoch": 2.0284444444444443, + "grad_norm": 0.0042611584067344666, + "learning_rate": 2.963812738142713e-06, + "loss": 1.0707381963729858, + "step": 144 + }, + { + "epoch": 2.042666666666667, + "grad_norm": 0.004273556638509035, + "learning_rate": 2.8889793378158284e-06, + "loss": 1.059180736541748, + "step": 145 + }, + { + "epoch": 2.056888888888889, + "grad_norm": 0.004242253955453634, + "learning_rate": 2.814717042148827e-06, + "loss": 1.0461338758468628, + "step": 146 + }, + { + "epoch": 2.071111111111111, + "grad_norm": 0.004076985642313957, + "learning_rate": 2.7410459416807856e-06, + "loss": 1.0546996593475342, + "step": 147 + }, + { + "epoch": 2.0853333333333333, + "grad_norm": 0.003977393265813589, + "learning_rate": 2.6679859670118785e-06, + "loss": 1.061549186706543, + "step": 148 + }, + { + "epoch": 2.0995555555555554, + "grad_norm": 0.0039635817520320415, + "learning_rate": 2.5955568834114523e-06, + "loss": 1.0474296808242798, + "step": 149 + }, + { + "epoch": 2.113777777777778, + "grad_norm": 0.004132823087275028, + "learning_rate": 2.523778285470835e-06, + "loss": 1.024568796157837, + "step": 150 + }, + { + "epoch": 2.113777777777778, + "eval_loss": 1.079006552696228, + "eval_runtime": 47.4217, + "eval_samples_per_second": 4.217, + "eval_steps_per_second": 2.109, + "step": 150 + }, + { + "epoch": 2.128, + "grad_norm": 0.004181793890893459, + "learning_rate": 2.452669591802307e-06, + "loss": 1.0624573230743408, + "step": 151 + }, + { + "epoch": 2.1422222222222222, + "grad_norm": 0.0041247266344726086, + "learning_rate": 2.3822500397857016e-06, + "loss": 1.0404682159423828, + "step": 152 + }, + { + "epoch": 2.1564444444444444, + "grad_norm": 0.004161086864769459, + "learning_rate": 2.3125386803640188e-06, + "loss": 1.0580246448516846, + "step": 153 + }, + { + "epoch": 2.1706666666666665, + "grad_norm": 0.004000307526439428, + "learning_rate": 2.243554372889479e-06, + "loss": 1.0445423126220703, + "step": 154 + }, + { + "epoch": 2.1848888888888887, + "grad_norm": 0.004109199624508619, + "learning_rate": 2.1753157800214107e-06, + "loss": 1.0412821769714355, + "step": 155 + }, + { + "epoch": 2.1991111111111112, + "grad_norm": 0.003884031670168042, + "learning_rate": 2.1078413626773547e-06, + "loss": 1.0429916381835938, + "step": 156 + }, + { + "epoch": 2.2133333333333334, + "grad_norm": 0.004056555684655905, + "learning_rate": 2.0411493750387423e-06, + "loss": 1.0391228199005127, + "step": 157 + }, + { + "epoch": 2.2275555555555555, + "grad_norm": 0.004009357187896967, + "learning_rate": 1.9752578596124955e-06, + "loss": 1.031305193901062, + "step": 158 + }, + { + "epoch": 2.2417777777777776, + "grad_norm": 0.0040836031548678875, + "learning_rate": 1.9101846423499113e-06, + "loss": 1.055808424949646, + "step": 159 + }, + { + "epoch": 2.2560000000000002, + "grad_norm": 0.004020580090582371, + "learning_rate": 1.8459473278241125e-06, + "loss": 1.0406150817871094, + "step": 160 + }, + { + "epoch": 2.2702222222222224, + "grad_norm": 0.003893360961228609, + "learning_rate": 1.7825632944674016e-06, + "loss": 1.0353386402130127, + "step": 161 + }, + { + "epoch": 2.2844444444444445, + "grad_norm": 0.004030768759548664, + "learning_rate": 1.7200496898697832e-06, + "loss": 1.0691417455673218, + "step": 162 + }, + { + "epoch": 2.2986666666666666, + "grad_norm": 0.003913589753210545, + "learning_rate": 1.6584234261399535e-06, + "loss": 1.0500602722167969, + "step": 163 + }, + { + "epoch": 2.3128888888888888, + "grad_norm": 0.003951441030949354, + "learning_rate": 1.5977011753299726e-06, + "loss": 1.0313360691070557, + "step": 164 + }, + { + "epoch": 2.327111111111111, + "grad_norm": 0.004080148879438639, + "learning_rate": 1.5378993649249053e-06, + "loss": 1.0616838932037354, + "step": 165 + }, + { + "epoch": 2.3413333333333335, + "grad_norm": 0.004066621884703636, + "learning_rate": 1.4790341733986085e-06, + "loss": 1.0500047206878662, + "step": 166 + }, + { + "epoch": 2.3555555555555556, + "grad_norm": 0.0040935976430773735, + "learning_rate": 1.4211215258368866e-06, + "loss": 1.0583405494689941, + "step": 167 + }, + { + "epoch": 2.3697777777777778, + "grad_norm": 0.004103758372366428, + "learning_rate": 1.3641770896292083e-06, + "loss": 1.0435129404067993, + "step": 168 + }, + { + "epoch": 2.384, + "grad_norm": 0.004010037053376436, + "learning_rate": 1.3082162702301276e-06, + "loss": 1.0434541702270508, + "step": 169 + }, + { + "epoch": 2.398222222222222, + "grad_norm": 0.004051943309605122, + "learning_rate": 1.2532542069915722e-06, + "loss": 1.052001953125, + "step": 170 + }, + { + "epoch": 2.4124444444444446, + "grad_norm": 0.0039450060576200485, + "learning_rate": 1.1993057690671174e-06, + "loss": 1.055966854095459, + "step": 171 + }, + { + "epoch": 2.4266666666666667, + "grad_norm": 0.00400911271572113, + "learning_rate": 1.1463855513893695e-06, + "loss": 1.0502328872680664, + "step": 172 + }, + { + "epoch": 2.440888888888889, + "grad_norm": 0.004119160585105419, + "learning_rate": 1.0945078707215224e-06, + "loss": 1.0345993041992188, + "step": 173 + }, + { + "epoch": 2.455111111111111, + "grad_norm": 0.004055999685078859, + "learning_rate": 1.0436867617841768e-06, + "loss": 1.0489795207977295, + "step": 174 + }, + { + "epoch": 2.469333333333333, + "grad_norm": 0.003937677945941687, + "learning_rate": 9.939359734584552e-07, + "loss": 1.057159662246704, + "step": 175 + }, + { + "epoch": 2.4835555555555557, + "grad_norm": 0.003958112560212612, + "learning_rate": 9.452689650664515e-07, + "loss": 1.0306581258773804, + "step": 176 + }, + { + "epoch": 2.497777777777778, + "grad_norm": 0.004141568671911955, + "learning_rate": 8.976989027300265e-07, + "loss": 1.0416128635406494, + "step": 177 + }, + { + "epoch": 2.512, + "grad_norm": 0.004062578082084656, + "learning_rate": 8.512386558088919e-07, + "loss": 1.0480129718780518, + "step": 178 + }, + { + "epoch": 2.526222222222222, + "grad_norm": 0.003910359926521778, + "learning_rate": 8.059007934190194e-07, + "loss": 1.047067642211914, + "step": 179 + }, + { + "epoch": 2.5404444444444443, + "grad_norm": 0.004017462953925133, + "learning_rate": 7.61697581032243e-07, + "loss": 1.053797721862793, + "step": 180 + }, + { + "epoch": 2.554666666666667, + "grad_norm": 0.0039033186621963978, + "learning_rate": 7.186409771580355e-07, + "loss": 1.0373890399932861, + "step": 181 + }, + { + "epoch": 2.568888888888889, + "grad_norm": 0.004007395822554827, + "learning_rate": 6.76742630108298e-07, + "loss": 1.043588399887085, + "step": 182 + }, + { + "epoch": 2.583111111111111, + "grad_norm": 0.003964278846979141, + "learning_rate": 6.360138748461015e-07, + "loss": 1.03302001953125, + "step": 183 + }, + { + "epoch": 2.5973333333333333, + "grad_norm": 0.0039159939624369144, + "learning_rate": 5.964657299191712e-07, + "loss": 1.0640901327133179, + "step": 184 + }, + { + "epoch": 2.6115555555555554, + "grad_norm": 0.003952207043766975, + "learning_rate": 5.581088944789953e-07, + "loss": 1.050436019897461, + "step": 185 + }, + { + "epoch": 2.6257777777777775, + "grad_norm": 0.004017418250441551, + "learning_rate": 5.20953745386329e-07, + "loss": 1.0415809154510498, + "step": 186 + }, + { + "epoch": 2.64, + "grad_norm": 0.003962535876780748, + "learning_rate": 4.850103344038853e-07, + "loss": 1.0449340343475342, + "step": 187 + }, + { + "epoch": 2.6542222222222223, + "grad_norm": 0.00386899639852345, + "learning_rate": 4.502883854769935e-07, + "loss": 1.0389983654022217, + "step": 188 + }, + { + "epoch": 2.6684444444444444, + "grad_norm": 0.003973118029534817, + "learning_rate": 4.167972921029262e-07, + "loss": 1.0497729778289795, + "step": 189 + }, + { + "epoch": 2.6826666666666665, + "grad_norm": 0.003942268900573254, + "learning_rate": 3.8454611478963235e-07, + "loss": 1.036604642868042, + "step": 190 + }, + { + "epoch": 2.696888888888889, + "grad_norm": 0.0038068366702646017, + "learning_rate": 3.535435786045538e-07, + "loss": 1.0535285472869873, + "step": 191 + }, + { + "epoch": 2.7111111111111112, + "grad_norm": 0.0040411436930298805, + "learning_rate": 3.237980708141919e-07, + "loss": 1.0318130254745483, + "step": 192 + }, + { + "epoch": 2.7253333333333334, + "grad_norm": 0.003907525446265936, + "learning_rate": 2.9531763861505967e-07, + "loss": 1.0349407196044922, + "step": 193 + }, + { + "epoch": 2.7395555555555555, + "grad_norm": 0.004061816725879908, + "learning_rate": 2.681099869566328e-07, + "loss": 1.0367615222930908, + "step": 194 + }, + { + "epoch": 2.7537777777777777, + "grad_norm": 0.003970359917730093, + "learning_rate": 2.4218247645689306e-07, + "loss": 1.0251959562301636, + "step": 195 + }, + { + "epoch": 2.768, + "grad_norm": 0.003819686360657215, + "learning_rate": 2.175421214110235e-07, + "loss": 1.0571613311767578, + "step": 196 + }, + { + "epoch": 2.7822222222222224, + "grad_norm": 0.0039050441700965166, + "learning_rate": 1.941955878938029e-07, + "loss": 1.0457653999328613, + "step": 197 + }, + { + "epoch": 2.7964444444444445, + "grad_norm": 0.003968564793467522, + "learning_rate": 1.721491919561913e-07, + "loss": 1.0394766330718994, + "step": 198 + }, + { + "epoch": 2.8106666666666666, + "grad_norm": 0.003920818213373423, + "learning_rate": 1.514088979166256e-07, + "loss": 1.0483171939849854, + "step": 199 + }, + { + "epoch": 2.824888888888889, + "grad_norm": 0.003929824102669954, + "learning_rate": 1.3198031674745814e-07, + "loss": 1.0300005674362183, + "step": 200 + }, + { + "epoch": 2.824888888888889, + "eval_loss": 1.0748209953308105, + "eval_runtime": 46.195, + "eval_samples_per_second": 4.329, + "eval_steps_per_second": 2.165, + "step": 200 + }, + { + "epoch": 2.8391111111111114, + "grad_norm": 0.0038694061804562807, + "learning_rate": 1.138687045569975e-07, + "loss": 1.0445573329925537, + "step": 201 + }, + { + "epoch": 2.8533333333333335, + "grad_norm": 0.00398522661998868, + "learning_rate": 9.70789611675449e-08, + "loss": 1.0359022617340088, + "step": 202 + }, + { + "epoch": 2.8675555555555556, + "grad_norm": 0.004058049060404301, + "learning_rate": 8.161562878982399e-08, + "loss": 1.0159897804260254, + "step": 203 + }, + { + "epoch": 2.8817777777777778, + "grad_norm": 0.004108336754143238, + "learning_rate": 6.74828907941516e-08, + "loss": 1.0329415798187256, + "step": 204 + }, + { + "epoch": 2.896, + "grad_norm": 0.0038334683049470186, + "learning_rate": 5.468457057869358e-08, + "loss": 1.0553950071334839, + "step": 205 + }, + { + "epoch": 2.910222222222222, + "grad_norm": 0.0037339332047849894, + "learning_rate": 4.322413053509944e-08, + "loss": 1.0415141582489014, + "step": 206 + }, + { + "epoch": 2.924444444444444, + "grad_norm": 0.004045915324240923, + "learning_rate": 3.31046711118066e-08, + "loss": 1.0387455224990845, + "step": 207 + }, + { + "epoch": 2.9386666666666668, + "grad_norm": 0.0039448207244277, + "learning_rate": 2.4328929975260262e-08, + "loss": 1.041945457458496, + "step": 208 + }, + { + "epoch": 2.952888888888889, + "grad_norm": 0.004023151472210884, + "learning_rate": 1.6899281269279756e-08, + "loss": 1.0387485027313232, + "step": 209 + }, + { + "epoch": 2.967111111111111, + "grad_norm": 0.0038402979262173176, + "learning_rate": 1.0817734972768946e-08, + "loss": 1.0304040908813477, + "step": 210 + }, + { + "epoch": 2.981333333333333, + "grad_norm": 0.003873219946399331, + "learning_rate": 6.085936355947897e-09, + "loss": 1.0460115671157837, + "step": 211 + }, + { + "epoch": 2.9955555555555557, + "grad_norm": 0.0037621951196342707, + "learning_rate": 2.7051655352494654e-09, + "loss": 1.0432226657867432, + "step": 212 + }, + { + "epoch": 3.0, + "grad_norm": 0.006865890696644783, + "learning_rate": 6.763371270035457e-10, + "loss": 1.0157432556152344, + "step": 213 + } + ], + "logging_steps": 1, + "max_steps": 213, + "num_input_tokens_seen": 0, + "num_train_epochs": 3, + "save_steps": 500, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 9.561792243333857e+17, + "train_batch_size": 1, + "trial_name": null, + "trial_params": null +} diff --git a/qwen2.5-3b-instruct/checkpoint-213/training_args.bin b/qwen2.5-3b-instruct/checkpoint-213/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..745b9d6d614f3064b70369b5149a4865ef67205e --- /dev/null +++ b/qwen2.5-3b-instruct/checkpoint-213/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:84503fc1ab703b8b0cb47dbab1adb9be59585b090c011d1af556d442423fd3c6 +size 5585 diff --git a/qwen2.5-3b-instruct/checkpoint-71/chat_template.jinja b/qwen2.5-3b-instruct/checkpoint-71/chat_template.jinja new file mode 100644 index 0000000000000000000000000000000000000000..bdf7919a96cfe43d50914a007b9c0877bd0ec27e --- /dev/null +++ b/qwen2.5-3b-instruct/checkpoint-71/chat_template.jinja @@ -0,0 +1,54 @@ +{%- if tools %} + {{- '<|im_start|>system\n' }} + {%- if messages[0]['role'] == 'system' %} + {{- messages[0]['content'] }} + {%- else %} + {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }} + {%- endif %} + {{- "\n\n# 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' }} + {%- else %} + {{- '<|im_start|>system\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\n' }} + {%- endif %} +{%- endif %} +{%- for message in messages %} + {%- if (message.role == "user") or (message.role == "system" and not loop.first) or (message.role == "assistant" and not message.tool_calls) %} + {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }} + {%- elif message.role == "assistant" %} + {{- '<|im_start|>' + message.role }} + {%- if message.content %} + {{- '\n' + message.content }} + {%- endif %} + {%- for tool_call in message.tool_calls %} + {%- if tool_call.function is defined %} + {%- set tool_call = tool_call.function %} + {%- endif %} + {{- '\n\n{"name": "' }} + {{- tool_call.name }} + {{- '", "arguments": ' }} + {{- tool_call.arguments | tojson }} + {{- '}\n' }} + {%- endfor %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "tool" %} + {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != "tool") %} + {{- '<|im_start|>user' }} + {%- endif %} + {{- '\n\n' }} + {{- message.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' }} +{%- endif %} diff --git a/qwen2.5-3b-instruct/checkpoint-71/config.json b/qwen2.5-3b-instruct/checkpoint-71/config.json new file mode 100644 index 0000000000000000000000000000000000000000..4814292863d1e5aa2f4fda1db18f4082cd61aa93 --- /dev/null +++ b/qwen2.5-3b-instruct/checkpoint-71/config.json @@ -0,0 +1,69 @@ +{ + "architectures": [ + "Qwen2ForCausalLM" + ], + "attention_dropout": 0.0, + "bos_token_id": null, + "dtype": "float32", + "eos_token_id": 151645, + "hidden_act": "silu", + "hidden_size": 2048, + "initializer_range": 0.02, + "intermediate_size": 11008, + "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": 32768, + "max_window_layers": 70, + "model_type": "qwen2", + "num_attention_heads": 16, + "num_hidden_layers": 36, + "num_key_value_heads": 2, + "pad_token_id": 151643, + "rms_norm_eps": 1e-06, + "rope_parameters": { + "rope_theta": 1000000.0, + "rope_type": "default" + }, + "sliding_window": null, + "tie_word_embeddings": true, + "transformers_version": "5.2.0", + "use_cache": false, + "use_sliding_window": false, + "vocab_size": 151936 +} diff --git a/qwen2.5-3b-instruct/checkpoint-71/generation_config.json b/qwen2.5-3b-instruct/checkpoint-71/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..47205fbf3386a215e7c2b9d62df6576f49b77e7b --- /dev/null +++ b/qwen2.5-3b-instruct/checkpoint-71/generation_config.json @@ -0,0 +1,13 @@ +{ + "do_sample": true, + "eos_token_id": [ + 151645, + 151643 + ], + "pad_token_id": 151643, + "repetition_penalty": 1.05, + "temperature": 0.7, + "top_k": 20, + "top_p": 0.8, + "transformers_version": "5.2.0" +} diff --git a/qwen2.5-3b-instruct/checkpoint-71/model.safetensors b/qwen2.5-3b-instruct/checkpoint-71/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..f84d6c888f0a1969ecdc47164205fa9a7af2ca7e --- /dev/null +++ b/qwen2.5-3b-instruct/checkpoint-71/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2bee460b090abf3f92d454d91f500d3611bb9515ce0d249cf31e1e1a1eec1004 +size 12343804296 diff --git a/qwen2.5-3b-instruct/checkpoint-71/optimizer.pt b/qwen2.5-3b-instruct/checkpoint-71/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..9d6073c0b536585cd8ac09c9bd56c0e11ed78824 --- /dev/null +++ b/qwen2.5-3b-instruct/checkpoint-71/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9af48865a0e6d6b5853c7ad6b0024e2f362663a47983251524a93cfd5d214d11 +size 24687895753 diff --git a/qwen2.5-3b-instruct/checkpoint-71/rng_state_0.pth b/qwen2.5-3b-instruct/checkpoint-71/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..870021e3fa5ac35c2f711adf0c93a556ab4842da --- /dev/null +++ b/qwen2.5-3b-instruct/checkpoint-71/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:95e5fc2074c0df31522a514f862c86cb00d71c946a7f15cc9ec0e53a69fb28a7 +size 14917 diff --git a/qwen2.5-3b-instruct/checkpoint-71/rng_state_1.pth b/qwen2.5-3b-instruct/checkpoint-71/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..21f20da5eb1da017f08aaa88bd19cf24d40e3fbf --- /dev/null +++ b/qwen2.5-3b-instruct/checkpoint-71/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0e7153eae67b6c9232a41bc996a2bf5b83229b8c7230d61911ac0fd40e64154e +size 14917 diff --git a/qwen2.5-3b-instruct/checkpoint-71/scheduler.pt b/qwen2.5-3b-instruct/checkpoint-71/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..c3be1d29bde40f225713ff5084d3d9d69858fcfd --- /dev/null +++ b/qwen2.5-3b-instruct/checkpoint-71/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9c96c28f12671e2f5db493417daefb3ff0b1bc36602843e0d7ef6df3649c27b3 +size 1465 diff --git a/qwen2.5-3b-instruct/checkpoint-71/tokenizer.json b/qwen2.5-3b-instruct/checkpoint-71/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..34510ff0037cd50428af467a17ead5a96140a32c --- /dev/null +++ b/qwen2.5-3b-instruct/checkpoint-71/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3fd169731d2cbde95e10bf356d66d5997fd885dd8dbb6fb4684da3f23b2585d8 +size 11421892 diff --git a/qwen2.5-3b-instruct/checkpoint-71/tokenizer_config.json b/qwen2.5-3b-instruct/checkpoint-71/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..145e2c7a68b109ae0253dda29fc2e66cb01b4b85 --- /dev/null +++ b/qwen2.5-3b-instruct/checkpoint-71/tokenizer_config.json @@ -0,0 +1,30 @@ +{ + "add_prefix_space": false, + "backend": "tokenizers", + "bos_token": null, + "clean_up_tokenization_spaces": false, + "eos_token": "<|im_end|>", + "errors": "replace", + "extra_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|>" + ], + "is_local": false, + "model_max_length": 131072, + "pad_token": "<|endoftext|>", + "padding_side": "right", + "split_special_tokens": false, + "tokenizer_class": "Qwen2Tokenizer", + "unk_token": null +} diff --git a/qwen2.5-3b-instruct/checkpoint-71/trainer_state.json b/qwen2.5-3b-instruct/checkpoint-71/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..fad05944842c3102e064ec73033d381bbc417cb7 --- /dev/null +++ b/qwen2.5-3b-instruct/checkpoint-71/trainer_state.json @@ -0,0 +1,539 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.0, + "eval_steps": 50, + "global_step": 71, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.014222222222222223, + "grad_norm": 0.07561660557985306, + "learning_rate": 0.0, + "loss": 1.6462128162384033, + "step": 1 + }, + { + "epoch": 0.028444444444444446, + "grad_norm": 0.07576845586299896, + "learning_rate": 4.5454545454545457e-07, + "loss": 1.6739654541015625, + "step": 2 + }, + { + "epoch": 0.042666666666666665, + "grad_norm": 0.060671016573905945, + "learning_rate": 9.090909090909091e-07, + "loss": 1.6208744049072266, + "step": 3 + }, + { + "epoch": 0.05688888888888889, + "grad_norm": 0.04539854824542999, + "learning_rate": 1.3636363636363636e-06, + "loss": 1.5681328773498535, + "step": 4 + }, + { + "epoch": 0.07111111111111111, + "grad_norm": 0.030098970979452133, + "learning_rate": 1.8181818181818183e-06, + "loss": 1.5172258615493774, + "step": 5 + }, + { + "epoch": 0.08533333333333333, + "grad_norm": 0.02540394477546215, + "learning_rate": 2.2727272727272728e-06, + "loss": 1.4882113933563232, + "step": 6 + }, + { + "epoch": 0.09955555555555555, + "grad_norm": 0.026057474315166473, + "learning_rate": 2.7272727272727272e-06, + "loss": 1.453606128692627, + "step": 7 + }, + { + "epoch": 0.11377777777777778, + "grad_norm": 0.024274546653032303, + "learning_rate": 3.181818181818182e-06, + "loss": 1.433072805404663, + "step": 8 + }, + { + "epoch": 0.128, + "grad_norm": 0.02080286666750908, + "learning_rate": 3.6363636363636366e-06, + "loss": 1.4036792516708374, + "step": 9 + }, + { + "epoch": 0.14222222222222222, + "grad_norm": 0.021423645317554474, + "learning_rate": 4.0909090909090915e-06, + "loss": 1.3751729726791382, + "step": 10 + }, + { + "epoch": 0.15644444444444444, + "grad_norm": 0.018808694556355476, + "learning_rate": 4.5454545454545455e-06, + "loss": 1.3682663440704346, + "step": 11 + }, + { + "epoch": 0.17066666666666666, + "grad_norm": 0.013645165599882603, + "learning_rate": 5e-06, + "loss": 1.3491919040679932, + "step": 12 + }, + { + "epoch": 0.18488888888888888, + "grad_norm": 0.012575564906001091, + "learning_rate": 5.4545454545454545e-06, + "loss": 1.3444881439208984, + "step": 13 + }, + { + "epoch": 0.1991111111111111, + "grad_norm": 0.011944238096475601, + "learning_rate": 5.90909090909091e-06, + "loss": 1.3321316242218018, + "step": 14 + }, + { + "epoch": 0.21333333333333335, + "grad_norm": 0.012251932173967361, + "learning_rate": 6.363636363636364e-06, + "loss": 1.3083908557891846, + "step": 15 + }, + { + "epoch": 0.22755555555555557, + "grad_norm": 0.011485234834253788, + "learning_rate": 6.818181818181818e-06, + "loss": 1.287674903869629, + "step": 16 + }, + { + "epoch": 0.24177777777777779, + "grad_norm": 0.010052796453237534, + "learning_rate": 7.272727272727273e-06, + "loss": 1.2796708345413208, + "step": 17 + }, + { + "epoch": 0.256, + "grad_norm": 0.009297606535255909, + "learning_rate": 7.727272727272727e-06, + "loss": 1.2778702974319458, + "step": 18 + }, + { + "epoch": 0.2702222222222222, + "grad_norm": 0.009387261234223843, + "learning_rate": 8.181818181818183e-06, + "loss": 1.258772373199463, + "step": 19 + }, + { + "epoch": 0.28444444444444444, + "grad_norm": 0.009283723309636116, + "learning_rate": 8.636363636363637e-06, + "loss": 1.2937917709350586, + "step": 20 + }, + { + "epoch": 0.2986666666666667, + "grad_norm": 0.008504457771778107, + "learning_rate": 9.090909090909091e-06, + "loss": 1.247847318649292, + "step": 21 + }, + { + "epoch": 0.3128888888888889, + "grad_norm": 0.007998796179890633, + "learning_rate": 9.545454545454547e-06, + "loss": 1.235945701599121, + "step": 22 + }, + { + "epoch": 0.32711111111111113, + "grad_norm": 0.007681620307266712, + "learning_rate": 1e-05, + "loss": 1.2462598085403442, + "step": 23 + }, + { + "epoch": 0.3413333333333333, + "grad_norm": 0.007152756676077843, + "learning_rate": 9.999323662872998e-06, + "loss": 1.2484906911849976, + "step": 24 + }, + { + "epoch": 0.35555555555555557, + "grad_norm": 0.00654159439727664, + "learning_rate": 9.99729483446475e-06, + "loss": 1.2089664936065674, + "step": 25 + }, + { + "epoch": 0.36977777777777776, + "grad_norm": 0.0063508739694952965, + "learning_rate": 9.993914063644053e-06, + "loss": 1.2245509624481201, + "step": 26 + }, + { + "epoch": 0.384, + "grad_norm": 0.006452301051467657, + "learning_rate": 9.989182265027232e-06, + "loss": 1.2176213264465332, + "step": 27 + }, + { + "epoch": 0.3982222222222222, + "grad_norm": 0.006349734030663967, + "learning_rate": 9.98310071873072e-06, + "loss": 1.2157793045043945, + "step": 28 + }, + { + "epoch": 0.41244444444444445, + "grad_norm": 0.006191881373524666, + "learning_rate": 9.975671070024741e-06, + "loss": 1.237949013710022, + "step": 29 + }, + { + "epoch": 0.4266666666666667, + "grad_norm": 0.005793207325041294, + "learning_rate": 9.966895328888195e-06, + "loss": 1.21578049659729, + "step": 30 + }, + { + "epoch": 0.4408888888888889, + "grad_norm": 0.005483992397785187, + "learning_rate": 9.956775869464901e-06, + "loss": 1.2089630365371704, + "step": 31 + }, + { + "epoch": 0.45511111111111113, + "grad_norm": 0.005602920427918434, + "learning_rate": 9.945315429421307e-06, + "loss": 1.2175853252410889, + "step": 32 + }, + { + "epoch": 0.4693333333333333, + "grad_norm": 0.005212367512285709, + "learning_rate": 9.932517109205849e-06, + "loss": 1.2096197605133057, + "step": 33 + }, + { + "epoch": 0.48355555555555557, + "grad_norm": 0.005095571745187044, + "learning_rate": 9.918384371210178e-06, + "loss": 1.1974353790283203, + "step": 34 + }, + { + "epoch": 0.49777777777777776, + "grad_norm": 0.005125134252011776, + "learning_rate": 9.902921038832456e-06, + "loss": 1.1847140789031982, + "step": 35 + }, + { + "epoch": 0.512, + "grad_norm": 0.004882226698100567, + "learning_rate": 9.886131295443003e-06, + "loss": 1.2025952339172363, + "step": 36 + }, + { + "epoch": 0.5262222222222223, + "grad_norm": 0.004733556881546974, + "learning_rate": 9.868019683252543e-06, + "loss": 1.193751573562622, + "step": 37 + }, + { + "epoch": 0.5404444444444444, + "grad_norm": 0.004847981501370668, + "learning_rate": 9.848591102083375e-06, + "loss": 1.1725599765777588, + "step": 38 + }, + { + "epoch": 0.5546666666666666, + "grad_norm": 0.004716528579592705, + "learning_rate": 9.82785080804381e-06, + "loss": 1.1763365268707275, + "step": 39 + }, + { + "epoch": 0.5688888888888889, + "grad_norm": 0.004721782635897398, + "learning_rate": 9.805804412106197e-06, + "loss": 1.180063009262085, + "step": 40 + }, + { + "epoch": 0.5831111111111111, + "grad_norm": 0.004470364656299353, + "learning_rate": 9.782457878588977e-06, + "loss": 1.169750690460205, + "step": 41 + }, + { + "epoch": 0.5973333333333334, + "grad_norm": 0.0046170237474143505, + "learning_rate": 9.75781752354311e-06, + "loss": 1.174955129623413, + "step": 42 + }, + { + "epoch": 0.6115555555555555, + "grad_norm": 0.004407198633998632, + "learning_rate": 9.731890013043367e-06, + "loss": 1.1694097518920898, + "step": 43 + }, + { + "epoch": 0.6257777777777778, + "grad_norm": 0.004441778641194105, + "learning_rate": 9.704682361384941e-06, + "loss": 1.1582658290863037, + "step": 44 + }, + { + "epoch": 0.64, + "grad_norm": 0.004318015184253454, + "learning_rate": 9.676201929185809e-06, + "loss": 1.1544437408447266, + "step": 45 + }, + { + "epoch": 0.6542222222222223, + "grad_norm": 0.004480153322219849, + "learning_rate": 9.646456421395447e-06, + "loss": 1.1822636127471924, + "step": 46 + }, + { + "epoch": 0.6684444444444444, + "grad_norm": 0.004228127654641867, + "learning_rate": 9.615453885210368e-06, + "loss": 1.1748132705688477, + "step": 47 + }, + { + "epoch": 0.6826666666666666, + "grad_norm": 0.004379024263471365, + "learning_rate": 9.583202707897075e-06, + "loss": 1.1558542251586914, + "step": 48 + }, + { + "epoch": 0.6968888888888889, + "grad_norm": 0.004313027486205101, + "learning_rate": 9.549711614523007e-06, + "loss": 1.1546211242675781, + "step": 49 + }, + { + "epoch": 0.7111111111111111, + "grad_norm": 0.004527685698121786, + "learning_rate": 9.514989665596114e-06, + "loss": 1.1557996273040771, + "step": 50 + }, + { + "epoch": 0.7111111111111111, + "eval_loss": 1.1428771018981934, + "eval_runtime": 47.9864, + "eval_samples_per_second": 4.168, + "eval_steps_per_second": 2.084, + "step": 50 + }, + { + "epoch": 0.7253333333333334, + "grad_norm": 0.0040644253604114056, + "learning_rate": 9.479046254613673e-06, + "loss": 1.172366976737976, + "step": 51 + }, + { + "epoch": 0.7395555555555555, + "grad_norm": 0.004353945609182119, + "learning_rate": 9.441891105521005e-06, + "loss": 1.1539039611816406, + "step": 52 + }, + { + "epoch": 0.7537777777777778, + "grad_norm": 0.004351011943072081, + "learning_rate": 9.40353427008083e-06, + "loss": 1.159185767173767, + "step": 53 + }, + { + "epoch": 0.768, + "grad_norm": 0.004263842478394508, + "learning_rate": 9.3639861251539e-06, + "loss": 1.1381170749664307, + "step": 54 + }, + { + "epoch": 0.7822222222222223, + "grad_norm": 0.0042404914274811745, + "learning_rate": 9.323257369891702e-06, + "loss": 1.1402206420898438, + "step": 55 + }, + { + "epoch": 0.7964444444444444, + "grad_norm": 0.004360774997621775, + "learning_rate": 9.281359022841966e-06, + "loss": 1.1499946117401123, + "step": 56 + }, + { + "epoch": 0.8106666666666666, + "grad_norm": 0.004249541088938713, + "learning_rate": 9.238302418967757e-06, + "loss": 1.1507560014724731, + "step": 57 + }, + { + "epoch": 0.8248888888888889, + "grad_norm": 0.004316243343055248, + "learning_rate": 9.194099206580981e-06, + "loss": 1.14259934425354, + "step": 58 + }, + { + "epoch": 0.8391111111111111, + "grad_norm": 0.004926053807139397, + "learning_rate": 9.14876134419111e-06, + "loss": 1.1445270776748657, + "step": 59 + }, + { + "epoch": 0.8533333333333334, + "grad_norm": 0.004064859822392464, + "learning_rate": 9.102301097269974e-06, + "loss": 1.1424087285995483, + "step": 60 + }, + { + "epoch": 0.8675555555555555, + "grad_norm": 0.004263672977685928, + "learning_rate": 9.05473103493355e-06, + "loss": 1.1524295806884766, + "step": 61 + }, + { + "epoch": 0.8817777777777778, + "grad_norm": 0.004161916207522154, + "learning_rate": 9.006064026541549e-06, + "loss": 1.1665949821472168, + "step": 62 + }, + { + "epoch": 0.896, + "grad_norm": 0.00454193539917469, + "learning_rate": 8.956313238215824e-06, + "loss": 1.1387211084365845, + "step": 63 + }, + { + "epoch": 0.9102222222222223, + "grad_norm": 0.004263832233846188, + "learning_rate": 8.905492129278478e-06, + "loss": 1.1322274208068848, + "step": 64 + }, + { + "epoch": 0.9244444444444444, + "grad_norm": 0.004334372468292713, + "learning_rate": 8.85361444861063e-06, + "loss": 1.1179461479187012, + "step": 65 + }, + { + "epoch": 0.9386666666666666, + "grad_norm": 0.0041003478690981865, + "learning_rate": 8.800694230932885e-06, + "loss": 1.1307452917099, + "step": 66 + }, + { + "epoch": 0.9528888888888889, + "grad_norm": 0.004313213285058737, + "learning_rate": 8.74674579300843e-06, + "loss": 1.1472362279891968, + "step": 67 + }, + { + "epoch": 0.9671111111111111, + "grad_norm": 0.004027621820569038, + "learning_rate": 8.691783729769874e-06, + "loss": 1.140931487083435, + "step": 68 + }, + { + "epoch": 0.9813333333333333, + "grad_norm": 0.004112871363759041, + "learning_rate": 8.635822910370793e-06, + "loss": 1.115431547164917, + "step": 69 + }, + { + "epoch": 0.9955555555555555, + "grad_norm": 0.004120040684938431, + "learning_rate": 8.578878474163115e-06, + "loss": 1.1306861639022827, + "step": 70 + }, + { + "epoch": 1.0, + "grad_norm": 0.007299893535673618, + "learning_rate": 8.520965826601394e-06, + "loss": 1.1355249881744385, + "step": 71 + } + ], + "logging_steps": 1, + "max_steps": 213, + "num_input_tokens_seen": 0, + "num_train_epochs": 3, + "save_steps": 500, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 3.187264086479995e+17, + "train_batch_size": 1, + "trial_name": null, + "trial_params": null +} diff --git a/qwen2.5-3b-instruct/checkpoint-71/training_args.bin b/qwen2.5-3b-instruct/checkpoint-71/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..745b9d6d614f3064b70369b5149a4865ef67205e --- /dev/null +++ b/qwen2.5-3b-instruct/checkpoint-71/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:84503fc1ab703b8b0cb47dbab1adb9be59585b090c011d1af556d442423fd3c6 +size 5585 diff --git a/qwen2.5-3b-instruct/config.json b/qwen2.5-3b-instruct/config.json new file mode 100644 index 0000000000000000000000000000000000000000..4814292863d1e5aa2f4fda1db18f4082cd61aa93 --- /dev/null +++ b/qwen2.5-3b-instruct/config.json @@ -0,0 +1,69 @@ +{ + "architectures": [ + "Qwen2ForCausalLM" + ], + "attention_dropout": 0.0, + "bos_token_id": null, + "dtype": "float32", + "eos_token_id": 151645, + "hidden_act": "silu", + "hidden_size": 2048, + "initializer_range": 0.02, + "intermediate_size": 11008, + "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": 32768, + "max_window_layers": 70, + "model_type": "qwen2", + "num_attention_heads": 16, + "num_hidden_layers": 36, + "num_key_value_heads": 2, + "pad_token_id": 151643, + "rms_norm_eps": 1e-06, + "rope_parameters": { + "rope_theta": 1000000.0, + "rope_type": "default" + }, + "sliding_window": null, + "tie_word_embeddings": true, + "transformers_version": "5.2.0", + "use_cache": false, + "use_sliding_window": false, + "vocab_size": 151936 +} diff --git a/qwen2.5-3b-instruct/eval_results.json b/qwen2.5-3b-instruct/eval_results.json new file mode 100644 index 0000000000000000000000000000000000000000..6068c2c0c446a66b8dfea5d24403af27e9dfce69 --- /dev/null +++ b/qwen2.5-3b-instruct/eval_results.json @@ -0,0 +1,7 @@ +{ + "epoch": 3.0, + "eval_loss": 1.0747698545455933, + "eval_runtime": 42.9121, + "eval_samples_per_second": 4.661, + "eval_steps_per_second": 2.33 +} \ No newline at end of file diff --git a/qwen2.5-3b-instruct/generation_config.json b/qwen2.5-3b-instruct/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..47205fbf3386a215e7c2b9d62df6576f49b77e7b --- /dev/null +++ b/qwen2.5-3b-instruct/generation_config.json @@ -0,0 +1,13 @@ +{ + "do_sample": true, + "eos_token_id": [ + 151645, + 151643 + ], + "pad_token_id": 151643, + "repetition_penalty": 1.05, + "temperature": 0.7, + "top_k": 20, + "top_p": 0.8, + "transformers_version": "5.2.0" +} diff --git a/qwen2.5-3b-instruct/model.safetensors b/qwen2.5-3b-instruct/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..74524f99eff11d5088665628147ba894149b6000 --- /dev/null +++ b/qwen2.5-3b-instruct/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d0e8497d095a6f44dad6178b028fb64ddba2353611464e5056ab51f620ec21c7 +size 12343804296 diff --git a/qwen2.5-3b-instruct/tokenizer.json b/qwen2.5-3b-instruct/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..34510ff0037cd50428af467a17ead5a96140a32c --- /dev/null +++ b/qwen2.5-3b-instruct/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3fd169731d2cbde95e10bf356d66d5997fd885dd8dbb6fb4684da3f23b2585d8 +size 11421892 diff --git a/qwen2.5-3b-instruct/tokenizer_config.json b/qwen2.5-3b-instruct/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..145e2c7a68b109ae0253dda29fc2e66cb01b4b85 --- /dev/null +++ b/qwen2.5-3b-instruct/tokenizer_config.json @@ -0,0 +1,30 @@ +{ + "add_prefix_space": false, + "backend": "tokenizers", + "bos_token": null, + "clean_up_tokenization_spaces": false, + "eos_token": "<|im_end|>", + "errors": "replace", + "extra_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|>" + ], + "is_local": false, + "model_max_length": 131072, + "pad_token": "<|endoftext|>", + "padding_side": "right", + "split_special_tokens": false, + "tokenizer_class": "Qwen2Tokenizer", + "unk_token": null +} diff --git a/qwen2.5-3b-instruct/train_results.json b/qwen2.5-3b-instruct/train_results.json new file mode 100644 index 0000000000000000000000000000000000000000..98573f1f09fa79a6b20d764ea37cd478e9b02986 --- /dev/null +++ b/qwen2.5-3b-instruct/train_results.json @@ -0,0 +1,8 @@ +{ + "epoch": 3.0, + "total_flos": 9.561792243333857e+17, + "train_loss": 1.1234366983315194, + "train_runtime": 19384.6824, + "train_samples_per_second": 2.786, + "train_steps_per_second": 0.011 +} \ No newline at end of file diff --git a/qwen2.5-3b-instruct/trainer_log.jsonl b/qwen2.5-3b-instruct/trainer_log.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..6426e68941d5276b4d34391a70319eb3be36c8a0 --- /dev/null +++ b/qwen2.5-3b-instruct/trainer_log.jsonl @@ -0,0 +1,221 @@ +{"current_steps": 1, "total_steps": 213, "loss": 1.6462128162384033, "lr": 0.0, "epoch": 0.014222222222222223, "percentage": 0.47, "elapsed_time": "0:01:30", "remaining_time": "5:19:30"} +{"current_steps": 2, "total_steps": 213, "loss": 1.6739654541015625, "lr": 4.5454545454545457e-07, "epoch": 0.028444444444444446, "percentage": 0.94, "elapsed_time": "0:03:00", "remaining_time": "5:17:36"} +{"current_steps": 3, "total_steps": 213, "loss": 1.6208744049072266, "lr": 9.090909090909091e-07, "epoch": 0.042666666666666665, "percentage": 1.41, "elapsed_time": "0:04:31", "remaining_time": "5:16:10"} +{"current_steps": 1, "total_steps": 213, "loss": 1.6462128162384033, "lr": 0.0, "epoch": 0.014222222222222223, "percentage": 0.47, "elapsed_time": "0:01:32", "remaining_time": "5:25:25"} +{"current_steps": 2, "total_steps": 213, "loss": 1.6739654541015625, "lr": 4.5454545454545457e-07, "epoch": 0.028444444444444446, "percentage": 0.94, "elapsed_time": "0:03:02", "remaining_time": "5:21:13"} +{"current_steps": 3, "total_steps": 213, "loss": 1.6208744049072266, "lr": 9.090909090909091e-07, "epoch": 0.042666666666666665, "percentage": 1.41, "elapsed_time": "0:04:34", "remaining_time": "5:19:49"} +{"current_steps": 4, "total_steps": 213, "loss": 1.5681328773498535, "lr": 1.3636363636363636e-06, "epoch": 0.05688888888888889, "percentage": 1.88, "elapsed_time": "0:06:05", "remaining_time": "5:17:56"} +{"current_steps": 5, "total_steps": 213, "loss": 1.5172258615493774, "lr": 1.8181818181818183e-06, "epoch": 0.07111111111111111, "percentage": 2.35, "elapsed_time": "0:07:36", "remaining_time": "5:16:21"} +{"current_steps": 6, "total_steps": 213, "loss": 1.4882113933563232, "lr": 2.2727272727272728e-06, "epoch": 0.08533333333333333, "percentage": 2.82, "elapsed_time": "0:09:06", "remaining_time": "5:14:26"} +{"current_steps": 7, "total_steps": 213, "loss": 1.453606128692627, "lr": 2.7272727272727272e-06, "epoch": 0.09955555555555555, "percentage": 3.29, "elapsed_time": "0:10:34", "remaining_time": "5:11:21"} +{"current_steps": 8, "total_steps": 213, "loss": 1.433072805404663, "lr": 3.181818181818182e-06, "epoch": 0.11377777777777778, "percentage": 3.76, "elapsed_time": "0:12:05", "remaining_time": "5:09:38"} +{"current_steps": 9, "total_steps": 213, "loss": 1.4036792516708374, "lr": 3.6363636363636366e-06, "epoch": 0.128, "percentage": 4.23, "elapsed_time": "0:13:35", "remaining_time": "5:07:57"} +{"current_steps": 10, "total_steps": 213, "loss": 1.3751729726791382, "lr": 4.0909090909090915e-06, "epoch": 0.14222222222222222, "percentage": 4.69, "elapsed_time": "0:15:06", "remaining_time": "5:06:32"} +{"current_steps": 11, "total_steps": 213, "loss": 1.3682663440704346, "lr": 4.5454545454545455e-06, "epoch": 0.15644444444444444, "percentage": 5.16, "elapsed_time": "0:16:35", "remaining_time": "5:04:47"} +{"current_steps": 12, "total_steps": 213, "loss": 1.3491919040679932, "lr": 5e-06, "epoch": 0.17066666666666666, "percentage": 5.63, "elapsed_time": "0:18:05", "remaining_time": "5:03:03"} +{"current_steps": 13, "total_steps": 213, "loss": 1.3444881439208984, "lr": 5.4545454545454545e-06, "epoch": 0.18488888888888888, "percentage": 6.1, "elapsed_time": "0:19:35", "remaining_time": "5:01:24"} +{"current_steps": 14, "total_steps": 213, "loss": 1.3321316242218018, "lr": 5.90909090909091e-06, "epoch": 0.1991111111111111, "percentage": 6.57, "elapsed_time": "0:21:05", "remaining_time": "4:59:50"} +{"current_steps": 15, "total_steps": 213, "loss": 1.3083908557891846, "lr": 6.363636363636364e-06, "epoch": 0.21333333333333335, "percentage": 7.04, "elapsed_time": "0:22:37", "remaining_time": "4:58:35"} +{"current_steps": 16, "total_steps": 213, "loss": 1.287674903869629, "lr": 6.818181818181818e-06, "epoch": 0.22755555555555557, "percentage": 7.51, "elapsed_time": "0:24:07", "remaining_time": "4:57:00"} +{"current_steps": 17, "total_steps": 213, "loss": 1.2796708345413208, "lr": 7.272727272727273e-06, "epoch": 0.24177777777777779, "percentage": 7.98, "elapsed_time": "0:25:36", "remaining_time": "4:55:20"} +{"current_steps": 18, "total_steps": 213, "loss": 1.2778702974319458, "lr": 7.727272727272727e-06, "epoch": 0.256, "percentage": 8.45, "elapsed_time": "0:27:08", "remaining_time": "4:53:57"} +{"current_steps": 19, "total_steps": 213, "loss": 1.258772373199463, "lr": 8.181818181818183e-06, "epoch": 0.2702222222222222, "percentage": 8.92, "elapsed_time": "0:28:37", "remaining_time": "4:52:14"} +{"current_steps": 20, "total_steps": 213, "loss": 1.2937917709350586, "lr": 8.636363636363637e-06, "epoch": 0.28444444444444444, "percentage": 9.39, "elapsed_time": "0:30:06", "remaining_time": "4:50:37"} +{"current_steps": 21, "total_steps": 213, "loss": 1.247847318649292, "lr": 9.090909090909091e-06, "epoch": 0.2986666666666667, "percentage": 9.86, "elapsed_time": "0:31:37", "remaining_time": "4:49:04"} +{"current_steps": 22, "total_steps": 213, "loss": 1.235945701599121, "lr": 9.545454545454547e-06, "epoch": 0.3128888888888889, "percentage": 10.33, "elapsed_time": "0:33:05", "remaining_time": "4:47:13"} +{"current_steps": 23, "total_steps": 213, "loss": 1.2462598085403442, "lr": 1e-05, "epoch": 0.32711111111111113, "percentage": 10.8, "elapsed_time": "0:34:34", "remaining_time": "4:45:37"} +{"current_steps": 24, "total_steps": 213, "loss": 1.2484906911849976, "lr": 9.999323662872998e-06, "epoch": 0.3413333333333333, "percentage": 11.27, "elapsed_time": "0:36:03", "remaining_time": "4:43:59"} +{"current_steps": 25, "total_steps": 213, "loss": 1.2089664936065674, "lr": 9.99729483446475e-06, "epoch": 0.35555555555555557, "percentage": 11.74, "elapsed_time": "0:37:33", "remaining_time": "4:42:25"} +{"current_steps": 26, "total_steps": 213, "loss": 1.2245509624481201, "lr": 9.993914063644053e-06, "epoch": 0.36977777777777776, "percentage": 12.21, "elapsed_time": "0:39:03", "remaining_time": "4:40:55"} +{"current_steps": 27, "total_steps": 213, "loss": 1.2176213264465332, "lr": 9.989182265027232e-06, "epoch": 0.384, "percentage": 12.68, "elapsed_time": "0:40:34", "remaining_time": "4:39:29"} +{"current_steps": 28, "total_steps": 213, "loss": 1.2157793045043945, "lr": 9.98310071873072e-06, "epoch": 0.3982222222222222, "percentage": 13.15, "elapsed_time": "0:42:02", "remaining_time": "4:37:45"} +{"current_steps": 29, "total_steps": 213, "loss": 1.237949013710022, "lr": 9.975671070024741e-06, "epoch": 0.41244444444444445, "percentage": 13.62, "elapsed_time": "0:43:32", "remaining_time": "4:36:16"} +{"current_steps": 30, "total_steps": 213, "loss": 1.21578049659729, "lr": 9.966895328888195e-06, "epoch": 0.4266666666666667, "percentage": 14.08, "elapsed_time": "0:45:01", "remaining_time": "4:34:40"} +{"current_steps": 31, "total_steps": 213, "loss": 1.2089630365371704, "lr": 9.956775869464901e-06, "epoch": 0.4408888888888889, "percentage": 14.55, "elapsed_time": "0:46:32", "remaining_time": "4:33:11"} +{"current_steps": 32, "total_steps": 213, "loss": 1.2175853252410889, "lr": 9.945315429421307e-06, "epoch": 0.45511111111111113, "percentage": 15.02, "elapsed_time": "0:48:00", "remaining_time": "4:31:32"} +{"current_steps": 33, "total_steps": 213, "loss": 1.2096197605133057, "lr": 9.932517109205849e-06, "epoch": 0.4693333333333333, "percentage": 15.49, "elapsed_time": "0:49:30", "remaining_time": "4:30:01"} +{"current_steps": 34, "total_steps": 213, "loss": 1.1974353790283203, "lr": 9.918384371210178e-06, "epoch": 0.48355555555555557, "percentage": 15.96, "elapsed_time": "0:50:59", "remaining_time": "4:28:27"} +{"current_steps": 35, "total_steps": 213, "loss": 1.1847140789031982, "lr": 9.902921038832456e-06, "epoch": 0.49777777777777776, "percentage": 16.43, "elapsed_time": "0:52:29", "remaining_time": "4:26:55"} +{"current_steps": 36, "total_steps": 213, "loss": 1.2025952339172363, "lr": 9.886131295443003e-06, "epoch": 0.512, "percentage": 16.9, "elapsed_time": "0:53:59", "remaining_time": "4:25:27"} +{"current_steps": 37, "total_steps": 213, "loss": 1.193751573562622, "lr": 9.868019683252543e-06, "epoch": 0.5262222222222223, "percentage": 17.37, "elapsed_time": "0:55:28", "remaining_time": "4:23:54"} +{"current_steps": 38, "total_steps": 213, "loss": 1.1725599765777588, "lr": 9.848591102083375e-06, "epoch": 0.5404444444444444, "percentage": 17.84, "elapsed_time": "0:56:58", "remaining_time": "4:22:20"} +{"current_steps": 39, "total_steps": 213, "loss": 1.1763365268707275, "lr": 9.82785080804381e-06, "epoch": 0.5546666666666666, "percentage": 18.31, "elapsed_time": "0:58:28", "remaining_time": "4:20:55"} +{"current_steps": 40, "total_steps": 213, "loss": 1.180063009262085, "lr": 9.805804412106197e-06, "epoch": 0.5688888888888889, "percentage": 18.78, "elapsed_time": "0:59:58", "remaining_time": "4:19:25"} +{"current_steps": 41, "total_steps": 213, "loss": 1.169750690460205, "lr": 9.782457878588977e-06, "epoch": 0.5831111111111111, "percentage": 19.25, "elapsed_time": "1:01:30", "remaining_time": "4:18:00"} +{"current_steps": 42, "total_steps": 213, "loss": 1.174955129623413, "lr": 9.75781752354311e-06, "epoch": 0.5973333333333334, "percentage": 19.72, "elapsed_time": "1:03:01", "remaining_time": "4:16:34"} +{"current_steps": 43, "total_steps": 213, "loss": 1.1694097518920898, "lr": 9.731890013043367e-06, "epoch": 0.6115555555555555, "percentage": 20.19, "elapsed_time": "1:04:29", "remaining_time": "4:14:57"} +{"current_steps": 44, "total_steps": 213, "loss": 1.1582658290863037, "lr": 9.704682361384941e-06, "epoch": 0.6257777777777778, "percentage": 20.66, "elapsed_time": "1:05:58", "remaining_time": "4:13:23"} +{"current_steps": 45, "total_steps": 213, "loss": 1.1544437408447266, "lr": 9.676201929185809e-06, "epoch": 0.64, "percentage": 21.13, "elapsed_time": "1:07:27", "remaining_time": "4:11:51"} +{"current_steps": 46, "total_steps": 213, "loss": 1.1822636127471924, "lr": 9.646456421395447e-06, "epoch": 0.6542222222222223, "percentage": 21.6, "elapsed_time": "1:08:58", "remaining_time": "4:10:25"} +{"current_steps": 47, "total_steps": 213, "loss": 1.1748132705688477, "lr": 9.615453885210368e-06, "epoch": 0.6684444444444444, "percentage": 22.07, "elapsed_time": "1:10:29", "remaining_time": "4:08:56"} +{"current_steps": 48, "total_steps": 213, "loss": 1.1558542251586914, "lr": 9.583202707897075e-06, "epoch": 0.6826666666666666, "percentage": 22.54, "elapsed_time": "1:11:57", "remaining_time": "4:07:21"} +{"current_steps": 49, "total_steps": 213, "loss": 1.1546211242675781, "lr": 9.549711614523007e-06, "epoch": 0.6968888888888889, "percentage": 23.0, "elapsed_time": "1:13:26", "remaining_time": "4:05:47"} +{"current_steps": 50, "total_steps": 213, "loss": 1.1557996273040771, "lr": 9.514989665596114e-06, "epoch": 0.7111111111111111, "percentage": 23.47, "elapsed_time": "1:14:57", "remaining_time": "4:04:21"} +{"current_steps": 50, "total_steps": 213, "eval_loss": 1.1428771018981934, "epoch": 0.7111111111111111, "percentage": 23.47, "elapsed_time": "1:15:45", "remaining_time": "4:06:57"} +{"current_steps": 51, "total_steps": 213, "loss": 1.172366976737976, "lr": 9.479046254613673e-06, "epoch": 0.7253333333333334, "percentage": 23.94, "elapsed_time": "1:17:16", "remaining_time": "4:05:26"} +{"current_steps": 52, "total_steps": 213, "loss": 1.1539039611816406, "lr": 9.441891105521005e-06, "epoch": 0.7395555555555555, "percentage": 24.41, "elapsed_time": "1:18:46", "remaining_time": "4:03:52"} +{"current_steps": 53, "total_steps": 213, "loss": 1.159185767173767, "lr": 9.40353427008083e-06, "epoch": 0.7537777777777778, "percentage": 24.88, "elapsed_time": "1:20:16", "remaining_time": "4:02:19"} +{"current_steps": 54, "total_steps": 213, "loss": 1.1381170749664307, "lr": 9.3639861251539e-06, "epoch": 0.768, "percentage": 25.35, "elapsed_time": "1:21:46", "remaining_time": "4:00:47"} +{"current_steps": 55, "total_steps": 213, "loss": 1.1402206420898438, "lr": 9.323257369891702e-06, "epoch": 0.7822222222222223, "percentage": 25.82, "elapsed_time": "1:23:16", "remaining_time": "3:59:12"} +{"current_steps": 56, "total_steps": 213, "loss": 1.1499946117401123, "lr": 9.281359022841966e-06, "epoch": 0.7964444444444444, "percentage": 26.29, "elapsed_time": "1:24:46", "remaining_time": "3:57:41"} +{"current_steps": 57, "total_steps": 213, "loss": 1.1507560014724731, "lr": 9.238302418967757e-06, "epoch": 0.8106666666666666, "percentage": 26.76, "elapsed_time": "1:26:16", "remaining_time": "3:56:06"} +{"current_steps": 58, "total_steps": 213, "loss": 1.14259934425354, "lr": 9.194099206580981e-06, "epoch": 0.8248888888888889, "percentage": 27.23, "elapsed_time": "1:27:44", "remaining_time": "3:54:30"} +{"current_steps": 59, "total_steps": 213, "loss": 1.1445270776748657, "lr": 9.14876134419111e-06, "epoch": 0.8391111111111111, "percentage": 27.7, "elapsed_time": "1:29:13", "remaining_time": "3:52:53"} +{"current_steps": 60, "total_steps": 213, "loss": 1.1424087285995483, "lr": 9.102301097269974e-06, "epoch": 0.8533333333333334, "percentage": 28.17, "elapsed_time": "1:30:42", "remaining_time": "3:51:17"} +{"current_steps": 61, "total_steps": 213, "loss": 1.1524295806884766, "lr": 9.05473103493355e-06, "epoch": 0.8675555555555555, "percentage": 28.64, "elapsed_time": "1:32:11", "remaining_time": "3:49:44"} +{"current_steps": 62, "total_steps": 213, "loss": 1.1665949821472168, "lr": 9.006064026541549e-06, "epoch": 0.8817777777777778, "percentage": 29.11, "elapsed_time": "1:33:41", "remaining_time": "3:48:10"} +{"current_steps": 63, "total_steps": 213, "loss": 1.1387211084365845, "lr": 8.956313238215824e-06, "epoch": 0.896, "percentage": 29.58, "elapsed_time": "1:35:12", "remaining_time": "3:46:40"} +{"current_steps": 64, "total_steps": 213, "loss": 1.1322274208068848, "lr": 8.905492129278478e-06, "epoch": 0.9102222222222223, "percentage": 30.05, "elapsed_time": "1:36:41", "remaining_time": "3:45:05"} +{"current_steps": 65, "total_steps": 213, "loss": 1.1179461479187012, "lr": 8.85361444861063e-06, "epoch": 0.9244444444444444, "percentage": 30.52, "elapsed_time": "1:38:10", "remaining_time": "3:43:31"} +{"current_steps": 66, "total_steps": 213, "loss": 1.1307452917099, "lr": 8.800694230932885e-06, "epoch": 0.9386666666666666, "percentage": 30.99, "elapsed_time": "1:39:39", "remaining_time": "3:41:57"} +{"current_steps": 67, "total_steps": 213, "loss": 1.1472362279891968, "lr": 8.74674579300843e-06, "epoch": 0.9528888888888889, "percentage": 31.46, "elapsed_time": "1:41:08", "remaining_time": "3:40:22"} +{"current_steps": 68, "total_steps": 213, "loss": 1.140931487083435, "lr": 8.691783729769874e-06, "epoch": 0.9671111111111111, "percentage": 31.92, "elapsed_time": "1:42:36", "remaining_time": "3:38:48"} +{"current_steps": 69, "total_steps": 213, "loss": 1.115431547164917, "lr": 8.635822910370793e-06, "epoch": 0.9813333333333333, "percentage": 32.39, "elapsed_time": "1:44:07", "remaining_time": "3:37:17"} +{"current_steps": 70, "total_steps": 213, "loss": 1.1306861639022827, "lr": 8.578878474163115e-06, "epoch": 0.9955555555555555, "percentage": 32.86, "elapsed_time": "1:45:37", "remaining_time": "3:35:46"} +{"current_steps": 71, "total_steps": 213, "loss": 1.1355249881744385, "lr": 8.520965826601394e-06, "epoch": 1.0, "percentage": 33.33, "elapsed_time": "1:46:05", "remaining_time": "3:32:11"} +{"current_steps": 72, "total_steps": 213, "loss": 1.098466157913208, "lr": 8.462100635075097e-06, "epoch": 1.0142222222222221, "percentage": 33.8, "elapsed_time": "1:49:11", "remaining_time": "3:33:50"} +{"current_steps": 73, "total_steps": 213, "loss": 1.1134223937988281, "lr": 8.40229882467003e-06, "epoch": 1.0284444444444445, "percentage": 34.27, "elapsed_time": "1:50:40", "remaining_time": "3:32:16"} +{"current_steps": 74, "total_steps": 213, "loss": 1.1099815368652344, "lr": 8.341576573860049e-06, "epoch": 1.0426666666666666, "percentage": 34.74, "elapsed_time": "1:52:10", "remaining_time": "3:30:41"} +{"current_steps": 75, "total_steps": 213, "loss": 1.1130825281143188, "lr": 8.279950310130218e-06, "epoch": 1.056888888888889, "percentage": 35.21, "elapsed_time": "1:53:39", "remaining_time": "3:29:07"} +{"current_steps": 76, "total_steps": 213, "loss": 1.1052048206329346, "lr": 8.2174367055326e-06, "epoch": 1.0711111111111111, "percentage": 35.68, "elapsed_time": "1:55:08", "remaining_time": "3:27:34"} +{"current_steps": 77, "total_steps": 213, "loss": 1.087442398071289, "lr": 8.154052672175888e-06, "epoch": 1.0853333333333333, "percentage": 36.15, "elapsed_time": "1:56:38", "remaining_time": "3:26:00"} +{"current_steps": 78, "total_steps": 213, "loss": 1.0848444700241089, "lr": 8.08981535765009e-06, "epoch": 1.0995555555555556, "percentage": 36.62, "elapsed_time": "1:58:07", "remaining_time": "3:24:27"} +{"current_steps": 79, "total_steps": 213, "loss": 1.1008961200714111, "lr": 8.024742140387506e-06, "epoch": 1.1137777777777778, "percentage": 37.09, "elapsed_time": "1:59:36", "remaining_time": "3:22:52"} +{"current_steps": 80, "total_steps": 213, "loss": 1.098643183708191, "lr": 7.95885062496126e-06, "epoch": 1.1280000000000001, "percentage": 37.56, "elapsed_time": "2:01:06", "remaining_time": "3:21:20"} +{"current_steps": 81, "total_steps": 213, "loss": 1.098900318145752, "lr": 7.892158637322647e-06, "epoch": 1.1422222222222222, "percentage": 38.03, "elapsed_time": "2:02:35", "remaining_time": "3:19:46"} +{"current_steps": 82, "total_steps": 213, "loss": 1.0768826007843018, "lr": 7.824684219978591e-06, "epoch": 1.1564444444444444, "percentage": 38.5, "elapsed_time": "2:04:02", "remaining_time": "3:18:10"} +{"current_steps": 83, "total_steps": 213, "loss": 1.1066226959228516, "lr": 7.756445627110523e-06, "epoch": 1.1706666666666667, "percentage": 38.97, "elapsed_time": "2:05:32", "remaining_time": "3:16:38"} +{"current_steps": 84, "total_steps": 213, "loss": 1.085633397102356, "lr": 7.68746131963598e-06, "epoch": 1.1848888888888889, "percentage": 39.44, "elapsed_time": "2:07:01", "remaining_time": "3:15:04"} +{"current_steps": 85, "total_steps": 213, "loss": 1.0916088819503784, "lr": 7.6177499602143e-06, "epoch": 1.199111111111111, "percentage": 39.91, "elapsed_time": "2:08:29", "remaining_time": "3:13:30"} +{"current_steps": 86, "total_steps": 213, "loss": 1.1167348623275757, "lr": 7.547330408197695e-06, "epoch": 1.2133333333333334, "percentage": 40.38, "elapsed_time": "2:09:59", "remaining_time": "3:11:58"} +{"current_steps": 87, "total_steps": 213, "loss": 1.093672513961792, "lr": 7.476221714529167e-06, "epoch": 1.2275555555555555, "percentage": 40.85, "elapsed_time": "2:11:29", "remaining_time": "3:10:25"} +{"current_steps": 88, "total_steps": 213, "loss": 1.0770323276519775, "lr": 7.404443116588548e-06, "epoch": 1.2417777777777779, "percentage": 41.31, "elapsed_time": "2:12:57", "remaining_time": "3:08:51"} +{"current_steps": 89, "total_steps": 213, "loss": 1.07706618309021, "lr": 7.332014032988123e-06, "epoch": 1.256, "percentage": 41.78, "elapsed_time": "2:14:27", "remaining_time": "3:07:19"} +{"current_steps": 90, "total_steps": 213, "loss": 1.1019604206085205, "lr": 7.2589540583192165e-06, "epoch": 1.2702222222222221, "percentage": 42.25, "elapsed_time": "2:15:55", "remaining_time": "3:05:46"} +{"current_steps": 91, "total_steps": 213, "loss": 1.0928822755813599, "lr": 7.185282957851175e-06, "epoch": 1.2844444444444445, "percentage": 42.72, "elapsed_time": "2:17:24", "remaining_time": "3:04:12"} +{"current_steps": 92, "total_steps": 213, "loss": 1.083510160446167, "lr": 7.111020662184174e-06, "epoch": 1.2986666666666666, "percentage": 43.19, "elapsed_time": "2:18:54", "remaining_time": "3:02:41"} +{"current_steps": 93, "total_steps": 213, "loss": 1.0836291313171387, "lr": 7.036187261857289e-06, "epoch": 1.3128888888888888, "percentage": 43.66, "elapsed_time": "2:20:23", "remaining_time": "3:01:08"} +{"current_steps": 94, "total_steps": 213, "loss": 1.0802290439605713, "lr": 6.960803001913315e-06, "epoch": 1.3271111111111111, "percentage": 44.13, "elapsed_time": "2:21:52", "remaining_time": "2:59:36"} +{"current_steps": 95, "total_steps": 213, "loss": 1.0798417329788208, "lr": 6.884888276421766e-06, "epoch": 1.3413333333333333, "percentage": 44.6, "elapsed_time": "2:23:22", "remaining_time": "2:58:05"} +{"current_steps": 96, "total_steps": 213, "loss": 1.0784028768539429, "lr": 6.8084636229615786e-06, "epoch": 1.3555555555555556, "percentage": 45.07, "elapsed_time": "2:24:53", "remaining_time": "2:56:35"} +{"current_steps": 97, "total_steps": 213, "loss": 1.0908350944519043, "lr": 6.731549717064975e-06, "epoch": 1.3697777777777778, "percentage": 45.54, "elapsed_time": "2:26:23", "remaining_time": "2:55:03"} +{"current_steps": 98, "total_steps": 213, "loss": 1.1067761182785034, "lr": 6.654167366624009e-06, "epoch": 1.384, "percentage": 46.01, "elapsed_time": "2:27:53", "remaining_time": "2:53:32"} +{"current_steps": 99, "total_steps": 213, "loss": 1.0768399238586426, "lr": 6.576337506261314e-06, "epoch": 1.3982222222222223, "percentage": 46.48, "elapsed_time": "2:29:23", "remaining_time": "2:52:02"} +{"current_steps": 100, "total_steps": 213, "loss": 1.087970495223999, "lr": 6.498081191666549e-06, "epoch": 1.4124444444444444, "percentage": 46.95, "elapsed_time": "2:30:53", "remaining_time": "2:50:30"} +{"current_steps": 100, "total_steps": 213, "eval_loss": 1.0954939126968384, "epoch": 1.4124444444444444, "percentage": 46.95, "elapsed_time": "2:31:41", "remaining_time": "2:51:24"} +{"current_steps": 101, "total_steps": 213, "loss": 1.0731995105743408, "lr": 6.419419593900109e-06, "epoch": 1.4266666666666667, "percentage": 47.42, "elapsed_time": "2:33:11", "remaining_time": "2:49:52"} +{"current_steps": 102, "total_steps": 213, "loss": 1.0977025032043457, "lr": 6.340373993665607e-06, "epoch": 1.4408888888888889, "percentage": 47.89, "elapsed_time": "2:34:40", "remaining_time": "2:48:19"} +{"current_steps": 103, "total_steps": 213, "loss": 1.0753436088562012, "lr": 6.2609657755527135e-06, "epoch": 1.455111111111111, "percentage": 48.36, "elapsed_time": "2:36:10", "remaining_time": "2:46:46"} +{"current_steps": 104, "total_steps": 213, "loss": 1.0888622999191284, "lr": 6.1812164222518626e-06, "epoch": 1.4693333333333334, "percentage": 48.83, "elapsed_time": "2:37:40", "remaining_time": "2:45:15"} +{"current_steps": 105, "total_steps": 213, "loss": 1.0883121490478516, "lr": 6.101147508742456e-06, "epoch": 1.4835555555555555, "percentage": 49.3, "elapsed_time": "2:39:11", "remaining_time": "2:43:43"} +{"current_steps": 106, "total_steps": 213, "loss": 1.0957694053649902, "lr": 6.020780696456059e-06, "epoch": 1.4977777777777779, "percentage": 49.77, "elapsed_time": "2:40:39", "remaining_time": "2:42:10"} +{"current_steps": 107, "total_steps": 213, "loss": 1.0853359699249268, "lr": 5.940137727416247e-06, "epoch": 1.512, "percentage": 50.23, "elapsed_time": "2:42:09", "remaining_time": "2:40:38"} +{"current_steps": 108, "total_steps": 213, "loss": 1.072435736656189, "lr": 5.859240418356614e-06, "epoch": 1.5262222222222221, "percentage": 50.7, "elapsed_time": "2:43:39", "remaining_time": "2:39:06"} +{"current_steps": 109, "total_steps": 213, "loss": 1.0823040008544922, "lr": 5.778110654818602e-06, "epoch": 1.5404444444444443, "percentage": 51.17, "elapsed_time": "2:45:09", "remaining_time": "2:37:35"} +{"current_steps": 110, "total_steps": 213, "loss": 1.0717535018920898, "lr": 5.696770385230679e-06, "epoch": 1.5546666666666666, "percentage": 51.64, "elapsed_time": "2:46:39", "remaining_time": "2:36:02"} +{"current_steps": 111, "total_steps": 213, "loss": 1.091806411743164, "lr": 5.615241614970546e-06, "epoch": 1.568888888888889, "percentage": 52.11, "elapsed_time": "2:48:08", "remaining_time": "2:34:30"} +{"current_steps": 112, "total_steps": 213, "loss": 1.0810648202896118, "lr": 5.533546400411899e-06, "epoch": 1.5831111111111111, "percentage": 52.58, "elapsed_time": "2:49:37", "remaining_time": "2:32:58"} +{"current_steps": 113, "total_steps": 213, "loss": 1.082428216934204, "lr": 5.4517068429574215e-06, "epoch": 1.5973333333333333, "percentage": 53.05, "elapsed_time": "2:51:07", "remaining_time": "2:31:26"} +{"current_steps": 114, "total_steps": 213, "loss": 1.087543249130249, "lr": 5.3697450830595775e-06, "epoch": 1.6115555555555554, "percentage": 53.52, "elapsed_time": "2:52:36", "remaining_time": "2:29:53"} +{"current_steps": 115, "total_steps": 213, "loss": 1.0576050281524658, "lr": 5.287683294230855e-06, "epoch": 1.6257777777777778, "percentage": 53.99, "elapsed_time": "2:54:05", "remaining_time": "2:28:21"} +{"current_steps": 116, "total_steps": 213, "loss": 1.0710945129394531, "lr": 5.20554367704505e-06, "epoch": 1.6400000000000001, "percentage": 54.46, "elapsed_time": "2:55:34", "remaining_time": "2:26:49"} +{"current_steps": 117, "total_steps": 213, "loss": 1.0972342491149902, "lr": 5.123348453131242e-06, "epoch": 1.6542222222222223, "percentage": 54.93, "elapsed_time": "2:57:05", "remaining_time": "2:25:18"} +{"current_steps": 118, "total_steps": 213, "loss": 1.0853114128112793, "lr": 5.041119859162068e-06, "epoch": 1.6684444444444444, "percentage": 55.4, "elapsed_time": "2:58:34", "remaining_time": "2:23:46"} +{"current_steps": 119, "total_steps": 213, "loss": 1.0747385025024414, "lr": 4.958880140837934e-06, "epoch": 1.6826666666666665, "percentage": 55.87, "elapsed_time": "3:00:04", "remaining_time": "2:22:14"} +{"current_steps": 120, "total_steps": 213, "loss": 1.062084674835205, "lr": 4.876651546868759e-06, "epoch": 1.696888888888889, "percentage": 56.34, "elapsed_time": "3:01:33", "remaining_time": "2:20:42"} +{"current_steps": 121, "total_steps": 213, "loss": 1.0754766464233398, "lr": 4.794456322954953e-06, "epoch": 1.7111111111111112, "percentage": 56.81, "elapsed_time": "3:03:03", "remaining_time": "2:19:11"} +{"current_steps": 122, "total_steps": 213, "loss": 1.073303461074829, "lr": 4.7123167057691446e-06, "epoch": 1.7253333333333334, "percentage": 57.28, "elapsed_time": "3:04:33", "remaining_time": "2:17:39"} +{"current_steps": 123, "total_steps": 213, "loss": 1.0916731357574463, "lr": 4.630254916940424e-06, "epoch": 1.7395555555555555, "percentage": 57.75, "elapsed_time": "3:06:03", "remaining_time": "2:16:08"} +{"current_steps": 124, "total_steps": 213, "loss": 1.0850458145141602, "lr": 4.548293157042581e-06, "epoch": 1.7537777777777777, "percentage": 58.22, "elapsed_time": "3:07:32", "remaining_time": "2:14:36"} +{"current_steps": 125, "total_steps": 213, "loss": 1.0751748085021973, "lr": 4.466453599588103e-06, "epoch": 1.768, "percentage": 58.69, "elapsed_time": "3:09:02", "remaining_time": "2:13:05"} +{"current_steps": 126, "total_steps": 213, "loss": 1.0639723539352417, "lr": 4.384758385029457e-06, "epoch": 1.7822222222222224, "percentage": 59.15, "elapsed_time": "3:10:31", "remaining_time": "2:11:33"} +{"current_steps": 127, "total_steps": 213, "loss": 1.0787277221679688, "lr": 4.3032296147693225e-06, "epoch": 1.7964444444444445, "percentage": 59.62, "elapsed_time": "3:12:02", "remaining_time": "2:10:02"} +{"current_steps": 128, "total_steps": 213, "loss": 1.0513901710510254, "lr": 4.2218893451814005e-06, "epoch": 1.8106666666666666, "percentage": 60.09, "elapsed_time": "3:13:29", "remaining_time": "2:08:29"} +{"current_steps": 129, "total_steps": 213, "loss": 1.0729951858520508, "lr": 4.140759581643386e-06, "epoch": 1.8248888888888888, "percentage": 60.56, "elapsed_time": "3:15:00", "remaining_time": "2:06:58"} +{"current_steps": 130, "total_steps": 213, "loss": 1.07584810256958, "lr": 4.059862272583755e-06, "epoch": 1.8391111111111111, "percentage": 61.03, "elapsed_time": "3:16:29", "remaining_time": "2:05:27"} +{"current_steps": 131, "total_steps": 213, "loss": 1.0640318393707275, "lr": 3.979219303543942e-06, "epoch": 1.8533333333333335, "percentage": 61.5, "elapsed_time": "3:17:59", "remaining_time": "2:03:56"} +{"current_steps": 132, "total_steps": 213, "loss": 1.074927568435669, "lr": 3.898852491257547e-06, "epoch": 1.8675555555555556, "percentage": 61.97, "elapsed_time": "3:19:30", "remaining_time": "2:02:25"} +{"current_steps": 133, "total_steps": 213, "loss": 1.0660316944122314, "lr": 3.818783577748138e-06, "epoch": 1.8817777777777778, "percentage": 62.44, "elapsed_time": "3:21:00", "remaining_time": "2:00:54"} +{"current_steps": 134, "total_steps": 213, "loss": 1.0795130729675293, "lr": 3.7390342244472886e-06, "epoch": 1.896, "percentage": 62.91, "elapsed_time": "3:22:30", "remaining_time": "1:59:23"} +{"current_steps": 135, "total_steps": 213, "loss": 1.094170093536377, "lr": 3.659626006334395e-06, "epoch": 1.9102222222222223, "percentage": 63.38, "elapsed_time": "3:24:00", "remaining_time": "1:57:52"} +{"current_steps": 136, "total_steps": 213, "loss": 1.0817457437515259, "lr": 3.5805804060998926e-06, "epoch": 1.9244444444444444, "percentage": 63.85, "elapsed_time": "3:25:30", "remaining_time": "1:56:21"} +{"current_steps": 137, "total_steps": 213, "loss": 1.0608152151107788, "lr": 3.501918808333453e-06, "epoch": 1.9386666666666668, "percentage": 64.32, "elapsed_time": "3:27:00", "remaining_time": "1:54:50"} +{"current_steps": 138, "total_steps": 213, "loss": 1.0709357261657715, "lr": 3.4236624937386874e-06, "epoch": 1.952888888888889, "percentage": 64.79, "elapsed_time": "3:28:30", "remaining_time": "1:53:19"} +{"current_steps": 139, "total_steps": 213, "loss": 1.0752217769622803, "lr": 3.3458326333759927e-06, "epoch": 1.967111111111111, "percentage": 65.26, "elapsed_time": "3:29:59", "remaining_time": "1:51:47"} +{"current_steps": 140, "total_steps": 213, "loss": 1.0565910339355469, "lr": 3.268450282935026e-06, "epoch": 1.9813333333333332, "percentage": 65.73, "elapsed_time": "3:31:28", "remaining_time": "1:50:15"} +{"current_steps": 141, "total_steps": 213, "loss": 1.064276099205017, "lr": 3.1915363770384223e-06, "epoch": 1.9955555555555555, "percentage": 66.2, "elapsed_time": "3:32:58", "remaining_time": "1:48:44"} +{"current_steps": 142, "total_steps": 213, "loss": 1.0892033576965332, "lr": 3.1151117235782346e-06, "epoch": 2.0, "percentage": 66.67, "elapsed_time": "3:33:26", "remaining_time": "1:46:43"} +{"current_steps": 143, "total_steps": 213, "loss": 1.0656744241714478, "lr": 3.0391969980866874e-06, "epoch": 2.014222222222222, "percentage": 67.14, "elapsed_time": "3:36:32", "remaining_time": "1:45:59"} +{"current_steps": 144, "total_steps": 213, "loss": 1.0707381963729858, "lr": 2.963812738142713e-06, "epoch": 2.0284444444444443, "percentage": 67.61, "elapsed_time": "3:38:01", "remaining_time": "1:44:28"} +{"current_steps": 145, "total_steps": 213, "loss": 1.059180736541748, "lr": 2.8889793378158284e-06, "epoch": 2.042666666666667, "percentage": 68.08, "elapsed_time": "3:39:32", "remaining_time": "1:42:57"} +{"current_steps": 146, "total_steps": 213, "loss": 1.0461338758468628, "lr": 2.814717042148827e-06, "epoch": 2.056888888888889, "percentage": 68.54, "elapsed_time": "3:41:00", "remaining_time": "1:41:25"} +{"current_steps": 147, "total_steps": 213, "loss": 1.0546996593475342, "lr": 2.7410459416807856e-06, "epoch": 2.071111111111111, "percentage": 69.01, "elapsed_time": "3:42:31", "remaining_time": "1:39:54"} +{"current_steps": 148, "total_steps": 213, "loss": 1.061549186706543, "lr": 2.6679859670118785e-06, "epoch": 2.0853333333333333, "percentage": 69.48, "elapsed_time": "3:43:59", "remaining_time": "1:38:22"} +{"current_steps": 149, "total_steps": 213, "loss": 1.0474296808242798, "lr": 2.5955568834114523e-06, "epoch": 2.0995555555555554, "percentage": 69.95, "elapsed_time": "3:45:28", "remaining_time": "1:36:50"} +{"current_steps": 150, "total_steps": 213, "loss": 1.024568796157837, "lr": 2.523778285470835e-06, "epoch": 2.113777777777778, "percentage": 70.42, "elapsed_time": "3:46:55", "remaining_time": "1:35:18"} +{"current_steps": 150, "total_steps": 213, "eval_loss": 1.079006552696228, "epoch": 2.113777777777778, "percentage": 70.42, "elapsed_time": "3:47:43", "remaining_time": "1:35:38"} +{"current_steps": 151, "total_steps": 213, "loss": 1.0624573230743408, "lr": 2.452669591802307e-06, "epoch": 2.128, "percentage": 70.89, "elapsed_time": "3:49:13", "remaining_time": "1:34:07"} +{"current_steps": 152, "total_steps": 213, "loss": 1.0404682159423828, "lr": 2.3822500397857016e-06, "epoch": 2.1422222222222222, "percentage": 71.36, "elapsed_time": "3:50:43", "remaining_time": "1:32:35"} +{"current_steps": 153, "total_steps": 213, "loss": 1.0580246448516846, "lr": 2.3125386803640188e-06, "epoch": 2.1564444444444444, "percentage": 71.83, "elapsed_time": "3:52:13", "remaining_time": "1:31:04"} +{"current_steps": 154, "total_steps": 213, "loss": 1.0445423126220703, "lr": 2.243554372889479e-06, "epoch": 2.1706666666666665, "percentage": 72.3, "elapsed_time": "3:53:43", "remaining_time": "1:29:32"} +{"current_steps": 155, "total_steps": 213, "loss": 1.0412821769714355, "lr": 2.1753157800214107e-06, "epoch": 2.1848888888888887, "percentage": 72.77, "elapsed_time": "3:55:11", "remaining_time": "1:28:00"} +{"current_steps": 156, "total_steps": 213, "loss": 1.0429916381835938, "lr": 2.1078413626773547e-06, "epoch": 2.1991111111111112, "percentage": 73.24, "elapsed_time": "3:56:42", "remaining_time": "1:26:29"} +{"current_steps": 157, "total_steps": 213, "loss": 1.0391228199005127, "lr": 2.0411493750387423e-06, "epoch": 2.2133333333333334, "percentage": 73.71, "elapsed_time": "3:58:10", "remaining_time": "1:24:57"} +{"current_steps": 158, "total_steps": 213, "loss": 1.031305193901062, "lr": 1.9752578596124955e-06, "epoch": 2.2275555555555555, "percentage": 74.18, "elapsed_time": "3:59:39", "remaining_time": "1:23:25"} +{"current_steps": 159, "total_steps": 213, "loss": 1.055808424949646, "lr": 1.9101846423499113e-06, "epoch": 2.2417777777777776, "percentage": 74.65, "elapsed_time": "4:01:09", "remaining_time": "1:21:54"} +{"current_steps": 160, "total_steps": 213, "loss": 1.0406150817871094, "lr": 1.8459473278241125e-06, "epoch": 2.2560000000000002, "percentage": 75.12, "elapsed_time": "4:02:39", "remaining_time": "1:20:22"} +{"current_steps": 161, "total_steps": 213, "loss": 1.0353386402130127, "lr": 1.7825632944674016e-06, "epoch": 2.2702222222222224, "percentage": 75.59, "elapsed_time": "4:04:08", "remaining_time": "1:18:51"} +{"current_steps": 162, "total_steps": 213, "loss": 1.0691417455673218, "lr": 1.7200496898697832e-06, "epoch": 2.2844444444444445, "percentage": 76.06, "elapsed_time": "4:05:38", "remaining_time": "1:17:19"} +{"current_steps": 163, "total_steps": 213, "loss": 1.0500602722167969, "lr": 1.6584234261399535e-06, "epoch": 2.2986666666666666, "percentage": 76.53, "elapsed_time": "4:07:08", "remaining_time": "1:15:48"} +{"current_steps": 164, "total_steps": 213, "loss": 1.0313360691070557, "lr": 1.5977011753299726e-06, "epoch": 2.3128888888888888, "percentage": 77.0, "elapsed_time": "4:08:38", "remaining_time": "1:14:17"} +{"current_steps": 165, "total_steps": 213, "loss": 1.0616838932037354, "lr": 1.5378993649249053e-06, "epoch": 2.327111111111111, "percentage": 77.46, "elapsed_time": "4:10:09", "remaining_time": "1:12:46"} +{"current_steps": 166, "total_steps": 213, "loss": 1.0500047206878662, "lr": 1.4790341733986085e-06, "epoch": 2.3413333333333335, "percentage": 77.93, "elapsed_time": "4:11:40", "remaining_time": "1:11:15"} +{"current_steps": 167, "total_steps": 213, "loss": 1.0583405494689941, "lr": 1.4211215258368866e-06, "epoch": 2.3555555555555556, "percentage": 78.4, "elapsed_time": "4:13:08", "remaining_time": "1:09:43"} +{"current_steps": 168, "total_steps": 213, "loss": 1.0435129404067993, "lr": 1.3641770896292083e-06, "epoch": 2.3697777777777778, "percentage": 78.87, "elapsed_time": "4:14:35", "remaining_time": "1:08:11"} +{"current_steps": 169, "total_steps": 213, "loss": 1.0434541702270508, "lr": 1.3082162702301276e-06, "epoch": 2.384, "percentage": 79.34, "elapsed_time": "4:16:05", "remaining_time": "1:06:40"} +{"current_steps": 170, "total_steps": 213, "loss": 1.052001953125, "lr": 1.2532542069915722e-06, "epoch": 2.398222222222222, "percentage": 79.81, "elapsed_time": "4:17:34", "remaining_time": "1:05:09"} +{"current_steps": 171, "total_steps": 213, "loss": 1.055966854095459, "lr": 1.1993057690671174e-06, "epoch": 2.4124444444444446, "percentage": 80.28, "elapsed_time": "4:19:04", "remaining_time": "1:03:38"} +{"current_steps": 172, "total_steps": 213, "loss": 1.0502328872680664, "lr": 1.1463855513893695e-06, "epoch": 2.4266666666666667, "percentage": 80.75, "elapsed_time": "4:20:32", "remaining_time": "1:02:06"} +{"current_steps": 173, "total_steps": 213, "loss": 1.0345993041992188, "lr": 1.0945078707215224e-06, "epoch": 2.440888888888889, "percentage": 81.22, "elapsed_time": "4:22:00", "remaining_time": "1:00:34"} +{"current_steps": 174, "total_steps": 213, "loss": 1.0489795207977295, "lr": 1.0436867617841768e-06, "epoch": 2.455111111111111, "percentage": 81.69, "elapsed_time": "4:23:31", "remaining_time": "0:59:03"} +{"current_steps": 175, "total_steps": 213, "loss": 1.057159662246704, "lr": 9.939359734584552e-07, "epoch": 2.469333333333333, "percentage": 82.16, "elapsed_time": "4:25:00", "remaining_time": "0:57:32"} +{"current_steps": 176, "total_steps": 213, "loss": 1.0306581258773804, "lr": 9.452689650664515e-07, "epoch": 2.4835555555555557, "percentage": 82.63, "elapsed_time": "4:26:30", "remaining_time": "0:56:01"} +{"current_steps": 177, "total_steps": 213, "loss": 1.0416128635406494, "lr": 8.976989027300265e-07, "epoch": 2.497777777777778, "percentage": 83.1, "elapsed_time": "4:27:59", "remaining_time": "0:54:30"} +{"current_steps": 178, "total_steps": 213, "loss": 1.0480129718780518, "lr": 8.512386558088919e-07, "epoch": 2.512, "percentage": 83.57, "elapsed_time": "4:29:28", "remaining_time": "0:52:59"} +{"current_steps": 179, "total_steps": 213, "loss": 1.047067642211914, "lr": 8.059007934190194e-07, "epoch": 2.526222222222222, "percentage": 84.04, "elapsed_time": "4:30:57", "remaining_time": "0:51:28"} +{"current_steps": 180, "total_steps": 213, "loss": 1.053797721862793, "lr": 7.61697581032243e-07, "epoch": 2.5404444444444443, "percentage": 84.51, "elapsed_time": "4:32:27", "remaining_time": "0:49:57"} +{"current_steps": 181, "total_steps": 213, "loss": 1.0373890399932861, "lr": 7.186409771580355e-07, "epoch": 2.554666666666667, "percentage": 84.98, "elapsed_time": "4:33:56", "remaining_time": "0:48:25"} +{"current_steps": 182, "total_steps": 213, "loss": 1.043588399887085, "lr": 6.76742630108298e-07, "epoch": 2.568888888888889, "percentage": 85.45, "elapsed_time": "4:35:25", "remaining_time": "0:46:54"} +{"current_steps": 183, "total_steps": 213, "loss": 1.03302001953125, "lr": 6.360138748461015e-07, "epoch": 2.583111111111111, "percentage": 85.92, "elapsed_time": "4:36:54", "remaining_time": "0:45:23"} +{"current_steps": 184, "total_steps": 213, "loss": 1.0640901327133179, "lr": 5.964657299191712e-07, "epoch": 2.5973333333333333, "percentage": 86.38, "elapsed_time": "4:38:26", "remaining_time": "0:43:53"} +{"current_steps": 185, "total_steps": 213, "loss": 1.050436019897461, "lr": 5.581088944789953e-07, "epoch": 2.6115555555555554, "percentage": 86.85, "elapsed_time": "4:39:56", "remaining_time": "0:42:22"} +{"current_steps": 186, "total_steps": 213, "loss": 1.0415809154510498, "lr": 5.20953745386329e-07, "epoch": 2.6257777777777775, "percentage": 87.32, "elapsed_time": "4:41:26", "remaining_time": "0:40:51"} +{"current_steps": 187, "total_steps": 213, "loss": 1.0449340343475342, "lr": 4.850103344038853e-07, "epoch": 2.64, "percentage": 87.79, "elapsed_time": "4:42:56", "remaining_time": "0:39:20"} +{"current_steps": 188, "total_steps": 213, "loss": 1.0389983654022217, "lr": 4.502883854769935e-07, "epoch": 2.6542222222222223, "percentage": 88.26, "elapsed_time": "4:44:25", "remaining_time": "0:37:49"} +{"current_steps": 189, "total_steps": 213, "loss": 1.0497729778289795, "lr": 4.167972921029262e-07, "epoch": 2.6684444444444444, "percentage": 88.73, "elapsed_time": "4:45:54", "remaining_time": "0:36:18"} +{"current_steps": 190, "total_steps": 213, "loss": 1.036604642868042, "lr": 3.8454611478963235e-07, "epoch": 2.6826666666666665, "percentage": 89.2, "elapsed_time": "4:47:24", "remaining_time": "0:34:47"} +{"current_steps": 191, "total_steps": 213, "loss": 1.0535285472869873, "lr": 3.535435786045538e-07, "epoch": 2.696888888888889, "percentage": 89.67, "elapsed_time": "4:48:53", "remaining_time": "0:33:16"} +{"current_steps": 192, "total_steps": 213, "loss": 1.0318130254745483, "lr": 3.237980708141919e-07, "epoch": 2.7111111111111112, "percentage": 90.14, "elapsed_time": "4:50:23", "remaining_time": "0:31:45"} +{"current_steps": 193, "total_steps": 213, "loss": 1.0349407196044922, "lr": 2.9531763861505967e-07, "epoch": 2.7253333333333334, "percentage": 90.61, "elapsed_time": "4:51:53", "remaining_time": "0:30:14"} +{"current_steps": 194, "total_steps": 213, "loss": 1.0367615222930908, "lr": 2.681099869566328e-07, "epoch": 2.7395555555555555, "percentage": 91.08, "elapsed_time": "4:53:22", "remaining_time": "0:28:43"} +{"current_steps": 195, "total_steps": 213, "loss": 1.0251959562301636, "lr": 2.4218247645689306e-07, "epoch": 2.7537777777777777, "percentage": 91.55, "elapsed_time": "4:54:50", "remaining_time": "0:27:12"} +{"current_steps": 196, "total_steps": 213, "loss": 1.0571613311767578, "lr": 2.175421214110235e-07, "epoch": 2.768, "percentage": 92.02, "elapsed_time": "4:56:19", "remaining_time": "0:25:42"} +{"current_steps": 197, "total_steps": 213, "loss": 1.0457653999328613, "lr": 1.941955878938029e-07, "epoch": 2.7822222222222224, "percentage": 92.49, "elapsed_time": "4:57:48", "remaining_time": "0:24:11"} +{"current_steps": 198, "total_steps": 213, "loss": 1.0394766330718994, "lr": 1.721491919561913e-07, "epoch": 2.7964444444444445, "percentage": 92.96, "elapsed_time": "4:59:18", "remaining_time": "0:22:40"} +{"current_steps": 199, "total_steps": 213, "loss": 1.0483171939849854, "lr": 1.514088979166256e-07, "epoch": 2.8106666666666666, "percentage": 93.43, "elapsed_time": "5:00:47", "remaining_time": "0:21:09"} +{"current_steps": 200, "total_steps": 213, "loss": 1.0300005674362183, "lr": 1.3198031674745814e-07, "epoch": 2.824888888888889, "percentage": 93.9, "elapsed_time": "5:02:15", "remaining_time": "0:19:38"} +{"current_steps": 200, "total_steps": 213, "eval_loss": 1.0748209953308105, "epoch": 2.824888888888889, "percentage": 93.9, "elapsed_time": "5:03:01", "remaining_time": "0:19:41"} +{"current_steps": 201, "total_steps": 213, "loss": 1.0445573329925537, "lr": 1.138687045569975e-07, "epoch": 2.8391111111111114, "percentage": 94.37, "elapsed_time": "5:04:32", "remaining_time": "0:18:10"} +{"current_steps": 202, "total_steps": 213, "loss": 1.0359022617340088, "lr": 9.70789611675449e-08, "epoch": 2.8533333333333335, "percentage": 94.84, "elapsed_time": "5:06:01", "remaining_time": "0:16:39"} +{"current_steps": 203, "total_steps": 213, "loss": 1.0159897804260254, "lr": 8.161562878982399e-08, "epoch": 2.8675555555555556, "percentage": 95.31, "elapsed_time": "5:07:30", "remaining_time": "0:15:08"} +{"current_steps": 204, "total_steps": 213, "loss": 1.0329415798187256, "lr": 6.74828907941516e-08, "epoch": 2.8817777777777778, "percentage": 95.77, "elapsed_time": "5:08:59", "remaining_time": "0:13:37"} +{"current_steps": 205, "total_steps": 213, "loss": 1.0553950071334839, "lr": 5.468457057869358e-08, "epoch": 2.896, "percentage": 96.24, "elapsed_time": "5:10:31", "remaining_time": "0:12:07"} +{"current_steps": 206, "total_steps": 213, "loss": 1.0415141582489014, "lr": 4.322413053509944e-08, "epoch": 2.910222222222222, "percentage": 96.71, "elapsed_time": "5:12:02", "remaining_time": "0:10:36"} +{"current_steps": 207, "total_steps": 213, "loss": 1.0387455224990845, "lr": 3.31046711118066e-08, "epoch": 2.924444444444444, "percentage": 97.18, "elapsed_time": "5:13:31", "remaining_time": "0:09:05"} +{"current_steps": 208, "total_steps": 213, "loss": 1.041945457458496, "lr": 2.4328929975260262e-08, "epoch": 2.9386666666666668, "percentage": 97.65, "elapsed_time": "5:15:00", "remaining_time": "0:07:34"} +{"current_steps": 209, "total_steps": 213, "loss": 1.0387485027313232, "lr": 1.6899281269279756e-08, "epoch": 2.952888888888889, "percentage": 98.12, "elapsed_time": "5:16:28", "remaining_time": "0:06:03"} +{"current_steps": 210, "total_steps": 213, "loss": 1.0304040908813477, "lr": 1.0817734972768946e-08, "epoch": 2.967111111111111, "percentage": 98.59, "elapsed_time": "5:17:57", "remaining_time": "0:04:32"} +{"current_steps": 211, "total_steps": 213, "loss": 1.0460115671157837, "lr": 6.085936355947897e-09, "epoch": 2.981333333333333, "percentage": 99.06, "elapsed_time": "5:19:26", "remaining_time": "0:03:01"} +{"current_steps": 212, "total_steps": 213, "loss": 1.0432226657867432, "lr": 2.7051655352494654e-09, "epoch": 2.9955555555555557, "percentage": 99.53, "elapsed_time": "5:20:56", "remaining_time": "0:01:30"} +{"current_steps": 213, "total_steps": 213, "loss": 1.0157432556152344, "lr": 6.763371270035457e-10, "epoch": 3.0, "percentage": 100.0, "elapsed_time": "5:21:23", "remaining_time": "0:00:00"} +{"current_steps": 213, "total_steps": 213, "epoch": 3.0, "percentage": 100.0, "elapsed_time": "5:23:00", "remaining_time": "0:00:00"} diff --git a/qwen2.5-3b-instruct/trainer_state.json b/qwen2.5-3b-instruct/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..339ed49a2180ee11036fa41f002af3a966a339a1 --- /dev/null +++ b/qwen2.5-3b-instruct/trainer_state.json @@ -0,0 +1,1566 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 3.0, + "eval_steps": 50, + "global_step": 213, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.014222222222222223, + "grad_norm": 0.07561660557985306, + "learning_rate": 0.0, + "loss": 1.6462128162384033, + "step": 1 + }, + { + "epoch": 0.028444444444444446, + "grad_norm": 0.07576845586299896, + "learning_rate": 4.5454545454545457e-07, + "loss": 1.6739654541015625, + "step": 2 + }, + { + "epoch": 0.042666666666666665, + "grad_norm": 0.060671016573905945, + "learning_rate": 9.090909090909091e-07, + "loss": 1.6208744049072266, + "step": 3 + }, + { + "epoch": 0.05688888888888889, + "grad_norm": 0.04539854824542999, + "learning_rate": 1.3636363636363636e-06, + "loss": 1.5681328773498535, + "step": 4 + }, + { + "epoch": 0.07111111111111111, + "grad_norm": 0.030098970979452133, + "learning_rate": 1.8181818181818183e-06, + "loss": 1.5172258615493774, + "step": 5 + }, + { + "epoch": 0.08533333333333333, + "grad_norm": 0.02540394477546215, + "learning_rate": 2.2727272727272728e-06, + "loss": 1.4882113933563232, + "step": 6 + }, + { + "epoch": 0.09955555555555555, + "grad_norm": 0.026057474315166473, + "learning_rate": 2.7272727272727272e-06, + "loss": 1.453606128692627, + "step": 7 + }, + { + "epoch": 0.11377777777777778, + "grad_norm": 0.024274546653032303, + "learning_rate": 3.181818181818182e-06, + "loss": 1.433072805404663, + "step": 8 + }, + { + "epoch": 0.128, + "grad_norm": 0.02080286666750908, + "learning_rate": 3.6363636363636366e-06, + "loss": 1.4036792516708374, + "step": 9 + }, + { + "epoch": 0.14222222222222222, + "grad_norm": 0.021423645317554474, + "learning_rate": 4.0909090909090915e-06, + "loss": 1.3751729726791382, + "step": 10 + }, + { + "epoch": 0.15644444444444444, + "grad_norm": 0.018808694556355476, + "learning_rate": 4.5454545454545455e-06, + "loss": 1.3682663440704346, + "step": 11 + }, + { + "epoch": 0.17066666666666666, + "grad_norm": 0.013645165599882603, + "learning_rate": 5e-06, + "loss": 1.3491919040679932, + "step": 12 + }, + { + "epoch": 0.18488888888888888, + "grad_norm": 0.012575564906001091, + "learning_rate": 5.4545454545454545e-06, + "loss": 1.3444881439208984, + "step": 13 + }, + { + "epoch": 0.1991111111111111, + "grad_norm": 0.011944238096475601, + "learning_rate": 5.90909090909091e-06, + "loss": 1.3321316242218018, + "step": 14 + }, + { + "epoch": 0.21333333333333335, + "grad_norm": 0.012251932173967361, + "learning_rate": 6.363636363636364e-06, + "loss": 1.3083908557891846, + "step": 15 + }, + { + "epoch": 0.22755555555555557, + "grad_norm": 0.011485234834253788, + "learning_rate": 6.818181818181818e-06, + "loss": 1.287674903869629, + "step": 16 + }, + { + "epoch": 0.24177777777777779, + "grad_norm": 0.010052796453237534, + "learning_rate": 7.272727272727273e-06, + "loss": 1.2796708345413208, + "step": 17 + }, + { + "epoch": 0.256, + "grad_norm": 0.009297606535255909, + "learning_rate": 7.727272727272727e-06, + "loss": 1.2778702974319458, + "step": 18 + }, + { + "epoch": 0.2702222222222222, + "grad_norm": 0.009387261234223843, + "learning_rate": 8.181818181818183e-06, + "loss": 1.258772373199463, + "step": 19 + }, + { + "epoch": 0.28444444444444444, + "grad_norm": 0.009283723309636116, + "learning_rate": 8.636363636363637e-06, + "loss": 1.2937917709350586, + "step": 20 + }, + { + "epoch": 0.2986666666666667, + "grad_norm": 0.008504457771778107, + "learning_rate": 9.090909090909091e-06, + "loss": 1.247847318649292, + "step": 21 + }, + { + "epoch": 0.3128888888888889, + "grad_norm": 0.007998796179890633, + "learning_rate": 9.545454545454547e-06, + "loss": 1.235945701599121, + "step": 22 + }, + { + "epoch": 0.32711111111111113, + "grad_norm": 0.007681620307266712, + "learning_rate": 1e-05, + "loss": 1.2462598085403442, + "step": 23 + }, + { + "epoch": 0.3413333333333333, + "grad_norm": 0.007152756676077843, + "learning_rate": 9.999323662872998e-06, + "loss": 1.2484906911849976, + "step": 24 + }, + { + "epoch": 0.35555555555555557, + "grad_norm": 0.00654159439727664, + "learning_rate": 9.99729483446475e-06, + "loss": 1.2089664936065674, + "step": 25 + }, + { + "epoch": 0.36977777777777776, + "grad_norm": 0.0063508739694952965, + "learning_rate": 9.993914063644053e-06, + "loss": 1.2245509624481201, + "step": 26 + }, + { + "epoch": 0.384, + "grad_norm": 0.006452301051467657, + "learning_rate": 9.989182265027232e-06, + "loss": 1.2176213264465332, + "step": 27 + }, + { + "epoch": 0.3982222222222222, + "grad_norm": 0.006349734030663967, + "learning_rate": 9.98310071873072e-06, + "loss": 1.2157793045043945, + "step": 28 + }, + { + "epoch": 0.41244444444444445, + "grad_norm": 0.006191881373524666, + "learning_rate": 9.975671070024741e-06, + "loss": 1.237949013710022, + "step": 29 + }, + { + "epoch": 0.4266666666666667, + "grad_norm": 0.005793207325041294, + "learning_rate": 9.966895328888195e-06, + "loss": 1.21578049659729, + "step": 30 + }, + { + "epoch": 0.4408888888888889, + "grad_norm": 0.005483992397785187, + "learning_rate": 9.956775869464901e-06, + "loss": 1.2089630365371704, + "step": 31 + }, + { + "epoch": 0.45511111111111113, + "grad_norm": 0.005602920427918434, + "learning_rate": 9.945315429421307e-06, + "loss": 1.2175853252410889, + "step": 32 + }, + { + "epoch": 0.4693333333333333, + "grad_norm": 0.005212367512285709, + "learning_rate": 9.932517109205849e-06, + "loss": 1.2096197605133057, + "step": 33 + }, + { + "epoch": 0.48355555555555557, + "grad_norm": 0.005095571745187044, + "learning_rate": 9.918384371210178e-06, + "loss": 1.1974353790283203, + "step": 34 + }, + { + "epoch": 0.49777777777777776, + "grad_norm": 0.005125134252011776, + "learning_rate": 9.902921038832456e-06, + "loss": 1.1847140789031982, + "step": 35 + }, + { + "epoch": 0.512, + "grad_norm": 0.004882226698100567, + "learning_rate": 9.886131295443003e-06, + "loss": 1.2025952339172363, + "step": 36 + }, + { + "epoch": 0.5262222222222223, + "grad_norm": 0.004733556881546974, + "learning_rate": 9.868019683252543e-06, + "loss": 1.193751573562622, + "step": 37 + }, + { + "epoch": 0.5404444444444444, + "grad_norm": 0.004847981501370668, + "learning_rate": 9.848591102083375e-06, + "loss": 1.1725599765777588, + "step": 38 + }, + { + "epoch": 0.5546666666666666, + "grad_norm": 0.004716528579592705, + "learning_rate": 9.82785080804381e-06, + "loss": 1.1763365268707275, + "step": 39 + }, + { + "epoch": 0.5688888888888889, + "grad_norm": 0.004721782635897398, + "learning_rate": 9.805804412106197e-06, + "loss": 1.180063009262085, + "step": 40 + }, + { + "epoch": 0.5831111111111111, + "grad_norm": 0.004470364656299353, + "learning_rate": 9.782457878588977e-06, + "loss": 1.169750690460205, + "step": 41 + }, + { + "epoch": 0.5973333333333334, + "grad_norm": 0.0046170237474143505, + "learning_rate": 9.75781752354311e-06, + "loss": 1.174955129623413, + "step": 42 + }, + { + "epoch": 0.6115555555555555, + "grad_norm": 0.004407198633998632, + "learning_rate": 9.731890013043367e-06, + "loss": 1.1694097518920898, + "step": 43 + }, + { + "epoch": 0.6257777777777778, + "grad_norm": 0.004441778641194105, + "learning_rate": 9.704682361384941e-06, + "loss": 1.1582658290863037, + "step": 44 + }, + { + "epoch": 0.64, + "grad_norm": 0.004318015184253454, + "learning_rate": 9.676201929185809e-06, + "loss": 1.1544437408447266, + "step": 45 + }, + { + "epoch": 0.6542222222222223, + "grad_norm": 0.004480153322219849, + "learning_rate": 9.646456421395447e-06, + "loss": 1.1822636127471924, + "step": 46 + }, + { + "epoch": 0.6684444444444444, + "grad_norm": 0.004228127654641867, + "learning_rate": 9.615453885210368e-06, + "loss": 1.1748132705688477, + "step": 47 + }, + { + "epoch": 0.6826666666666666, + "grad_norm": 0.004379024263471365, + "learning_rate": 9.583202707897075e-06, + "loss": 1.1558542251586914, + "step": 48 + }, + { + "epoch": 0.6968888888888889, + "grad_norm": 0.004313027486205101, + "learning_rate": 9.549711614523007e-06, + "loss": 1.1546211242675781, + "step": 49 + }, + { + "epoch": 0.7111111111111111, + "grad_norm": 0.004527685698121786, + "learning_rate": 9.514989665596114e-06, + "loss": 1.1557996273040771, + "step": 50 + }, + { + "epoch": 0.7111111111111111, + "eval_loss": 1.1428771018981934, + "eval_runtime": 47.9864, + "eval_samples_per_second": 4.168, + "eval_steps_per_second": 2.084, + "step": 50 + }, + { + "epoch": 0.7253333333333334, + "grad_norm": 0.0040644253604114056, + "learning_rate": 9.479046254613673e-06, + "loss": 1.172366976737976, + "step": 51 + }, + { + "epoch": 0.7395555555555555, + "grad_norm": 0.004353945609182119, + "learning_rate": 9.441891105521005e-06, + "loss": 1.1539039611816406, + "step": 52 + }, + { + "epoch": 0.7537777777777778, + "grad_norm": 0.004351011943072081, + "learning_rate": 9.40353427008083e-06, + "loss": 1.159185767173767, + "step": 53 + }, + { + "epoch": 0.768, + "grad_norm": 0.004263842478394508, + "learning_rate": 9.3639861251539e-06, + "loss": 1.1381170749664307, + "step": 54 + }, + { + "epoch": 0.7822222222222223, + "grad_norm": 0.0042404914274811745, + "learning_rate": 9.323257369891702e-06, + "loss": 1.1402206420898438, + "step": 55 + }, + { + "epoch": 0.7964444444444444, + "grad_norm": 0.004360774997621775, + "learning_rate": 9.281359022841966e-06, + "loss": 1.1499946117401123, + "step": 56 + }, + { + "epoch": 0.8106666666666666, + "grad_norm": 0.004249541088938713, + "learning_rate": 9.238302418967757e-06, + "loss": 1.1507560014724731, + "step": 57 + }, + { + "epoch": 0.8248888888888889, + "grad_norm": 0.004316243343055248, + "learning_rate": 9.194099206580981e-06, + "loss": 1.14259934425354, + "step": 58 + }, + { + "epoch": 0.8391111111111111, + "grad_norm": 0.004926053807139397, + "learning_rate": 9.14876134419111e-06, + "loss": 1.1445270776748657, + "step": 59 + }, + { + "epoch": 0.8533333333333334, + "grad_norm": 0.004064859822392464, + "learning_rate": 9.102301097269974e-06, + "loss": 1.1424087285995483, + "step": 60 + }, + { + "epoch": 0.8675555555555555, + "grad_norm": 0.004263672977685928, + "learning_rate": 9.05473103493355e-06, + "loss": 1.1524295806884766, + "step": 61 + }, + { + "epoch": 0.8817777777777778, + "grad_norm": 0.004161916207522154, + "learning_rate": 9.006064026541549e-06, + "loss": 1.1665949821472168, + "step": 62 + }, + { + "epoch": 0.896, + "grad_norm": 0.00454193539917469, + "learning_rate": 8.956313238215824e-06, + "loss": 1.1387211084365845, + "step": 63 + }, + { + "epoch": 0.9102222222222223, + "grad_norm": 0.004263832233846188, + "learning_rate": 8.905492129278478e-06, + "loss": 1.1322274208068848, + "step": 64 + }, + { + "epoch": 0.9244444444444444, + "grad_norm": 0.004334372468292713, + "learning_rate": 8.85361444861063e-06, + "loss": 1.1179461479187012, + "step": 65 + }, + { + "epoch": 0.9386666666666666, + "grad_norm": 0.0041003478690981865, + "learning_rate": 8.800694230932885e-06, + "loss": 1.1307452917099, + "step": 66 + }, + { + "epoch": 0.9528888888888889, + "grad_norm": 0.004313213285058737, + "learning_rate": 8.74674579300843e-06, + "loss": 1.1472362279891968, + "step": 67 + }, + { + "epoch": 0.9671111111111111, + "grad_norm": 0.004027621820569038, + "learning_rate": 8.691783729769874e-06, + "loss": 1.140931487083435, + "step": 68 + }, + { + "epoch": 0.9813333333333333, + "grad_norm": 0.004112871363759041, + "learning_rate": 8.635822910370793e-06, + "loss": 1.115431547164917, + "step": 69 + }, + { + "epoch": 0.9955555555555555, + "grad_norm": 0.004120040684938431, + "learning_rate": 8.578878474163115e-06, + "loss": 1.1306861639022827, + "step": 70 + }, + { + "epoch": 1.0, + "grad_norm": 0.007299893535673618, + "learning_rate": 8.520965826601394e-06, + "loss": 1.1355249881744385, + "step": 71 + }, + { + "epoch": 1.0142222222222221, + "grad_norm": 0.004595829173922539, + "learning_rate": 8.462100635075097e-06, + "loss": 1.098466157913208, + "step": 72 + }, + { + "epoch": 1.0284444444444445, + "grad_norm": 0.004596133250743151, + "learning_rate": 8.40229882467003e-06, + "loss": 1.1134223937988281, + "step": 73 + }, + { + "epoch": 1.0426666666666666, + "grad_norm": 0.0047966958954930305, + "learning_rate": 8.341576573860049e-06, + "loss": 1.1099815368652344, + "step": 74 + }, + { + "epoch": 1.056888888888889, + "grad_norm": 0.0042020766995847225, + "learning_rate": 8.279950310130218e-06, + "loss": 1.1130825281143188, + "step": 75 + }, + { + "epoch": 1.0711111111111111, + "grad_norm": 0.004611601121723652, + "learning_rate": 8.2174367055326e-06, + "loss": 1.1052048206329346, + "step": 76 + }, + { + "epoch": 1.0853333333333333, + "grad_norm": 0.004341403488069773, + "learning_rate": 8.154052672175888e-06, + "loss": 1.087442398071289, + "step": 77 + }, + { + "epoch": 1.0995555555555556, + "grad_norm": 0.004340282175689936, + "learning_rate": 8.08981535765009e-06, + "loss": 1.0848444700241089, + "step": 78 + }, + { + "epoch": 1.1137777777777778, + "grad_norm": 0.004358356352895498, + "learning_rate": 8.024742140387506e-06, + "loss": 1.1008961200714111, + "step": 79 + }, + { + "epoch": 1.1280000000000001, + "grad_norm": 0.004328620154410601, + "learning_rate": 7.95885062496126e-06, + "loss": 1.098643183708191, + "step": 80 + }, + { + "epoch": 1.1422222222222222, + "grad_norm": 0.004357457626610994, + "learning_rate": 7.892158637322647e-06, + "loss": 1.098900318145752, + "step": 81 + }, + { + "epoch": 1.1564444444444444, + "grad_norm": 0.004075393080711365, + "learning_rate": 7.824684219978591e-06, + "loss": 1.0768826007843018, + "step": 82 + }, + { + "epoch": 1.1706666666666667, + "grad_norm": 0.004524864722043276, + "learning_rate": 7.756445627110523e-06, + "loss": 1.1066226959228516, + "step": 83 + }, + { + "epoch": 1.1848888888888889, + "grad_norm": 0.004255601204931736, + "learning_rate": 7.68746131963598e-06, + "loss": 1.085633397102356, + "step": 84 + }, + { + "epoch": 1.199111111111111, + "grad_norm": 0.004353133961558342, + "learning_rate": 7.6177499602143e-06, + "loss": 1.0916088819503784, + "step": 85 + }, + { + "epoch": 1.2133333333333334, + "grad_norm": 0.004467015620321035, + "learning_rate": 7.547330408197695e-06, + "loss": 1.1167348623275757, + "step": 86 + }, + { + "epoch": 1.2275555555555555, + "grad_norm": 0.004218449350446463, + "learning_rate": 7.476221714529167e-06, + "loss": 1.093672513961792, + "step": 87 + }, + { + "epoch": 1.2417777777777779, + "grad_norm": 0.0046919044107198715, + "learning_rate": 7.404443116588548e-06, + "loss": 1.0770323276519775, + "step": 88 + }, + { + "epoch": 1.256, + "grad_norm": 0.004413353744894266, + "learning_rate": 7.332014032988123e-06, + "loss": 1.07706618309021, + "step": 89 + }, + { + "epoch": 1.2702222222222221, + "grad_norm": 0.004050100222229958, + "learning_rate": 7.2589540583192165e-06, + "loss": 1.1019604206085205, + "step": 90 + }, + { + "epoch": 1.2844444444444445, + "grad_norm": 0.004370351787656546, + "learning_rate": 7.185282957851175e-06, + "loss": 1.0928822755813599, + "step": 91 + }, + { + "epoch": 1.2986666666666666, + "grad_norm": 0.004322490654885769, + "learning_rate": 7.111020662184174e-06, + "loss": 1.083510160446167, + "step": 92 + }, + { + "epoch": 1.3128888888888888, + "grad_norm": 0.004432035610079765, + "learning_rate": 7.036187261857289e-06, + "loss": 1.0836291313171387, + "step": 93 + }, + { + "epoch": 1.3271111111111111, + "grad_norm": 0.004410918802022934, + "learning_rate": 6.960803001913315e-06, + "loss": 1.0802290439605713, + "step": 94 + }, + { + "epoch": 1.3413333333333333, + "grad_norm": 0.004076072946190834, + "learning_rate": 6.884888276421766e-06, + "loss": 1.0798417329788208, + "step": 95 + }, + { + "epoch": 1.3555555555555556, + "grad_norm": 0.004036934580653906, + "learning_rate": 6.8084636229615786e-06, + "loss": 1.0784028768539429, + "step": 96 + }, + { + "epoch": 1.3697777777777778, + "grad_norm": 0.004370865412056446, + "learning_rate": 6.731549717064975e-06, + "loss": 1.0908350944519043, + "step": 97 + }, + { + "epoch": 1.384, + "grad_norm": 0.004513819236308336, + "learning_rate": 6.654167366624009e-06, + "loss": 1.1067761182785034, + "step": 98 + }, + { + "epoch": 1.3982222222222223, + "grad_norm": 0.004452561028301716, + "learning_rate": 6.576337506261314e-06, + "loss": 1.0768399238586426, + "step": 99 + }, + { + "epoch": 1.4124444444444444, + "grad_norm": 0.00441731046885252, + "learning_rate": 6.498081191666549e-06, + "loss": 1.087970495223999, + "step": 100 + }, + { + "epoch": 1.4124444444444444, + "eval_loss": 1.0954939126968384, + "eval_runtime": 48.1789, + "eval_samples_per_second": 4.151, + "eval_steps_per_second": 2.076, + "step": 100 + }, + { + "epoch": 1.4266666666666667, + "grad_norm": 0.004392492119222879, + "learning_rate": 6.419419593900109e-06, + "loss": 1.0731995105743408, + "step": 101 + }, + { + "epoch": 1.4408888888888889, + "grad_norm": 0.004050276707857847, + "learning_rate": 6.340373993665607e-06, + "loss": 1.0977025032043457, + "step": 102 + }, + { + "epoch": 1.455111111111111, + "grad_norm": 0.004420506302267313, + "learning_rate": 6.2609657755527135e-06, + "loss": 1.0753436088562012, + "step": 103 + }, + { + "epoch": 1.4693333333333334, + "grad_norm": 0.004381162114441395, + "learning_rate": 6.1812164222518626e-06, + "loss": 1.0888622999191284, + "step": 104 + }, + { + "epoch": 1.4835555555555555, + "grad_norm": 0.004510694183409214, + "learning_rate": 6.101147508742456e-06, + "loss": 1.0883121490478516, + "step": 105 + }, + { + "epoch": 1.4977777777777779, + "grad_norm": 0.004262086004018784, + "learning_rate": 6.020780696456059e-06, + "loss": 1.0957694053649902, + "step": 106 + }, + { + "epoch": 1.512, + "grad_norm": 0.004126217216253281, + "learning_rate": 5.940137727416247e-06, + "loss": 1.0853359699249268, + "step": 107 + }, + { + "epoch": 1.5262222222222221, + "grad_norm": 0.004340359941124916, + "learning_rate": 5.859240418356614e-06, + "loss": 1.072435736656189, + "step": 108 + }, + { + "epoch": 1.5404444444444443, + "grad_norm": 0.0040999967604875565, + "learning_rate": 5.778110654818602e-06, + "loss": 1.0823040008544922, + "step": 109 + }, + { + "epoch": 1.5546666666666666, + "grad_norm": 0.004626137670129538, + "learning_rate": 5.696770385230679e-06, + "loss": 1.0717535018920898, + "step": 110 + }, + { + "epoch": 1.568888888888889, + "grad_norm": 0.00404947716742754, + "learning_rate": 5.615241614970546e-06, + "loss": 1.091806411743164, + "step": 111 + }, + { + "epoch": 1.5831111111111111, + "grad_norm": 0.004189760889858007, + "learning_rate": 5.533546400411899e-06, + "loss": 1.0810648202896118, + "step": 112 + }, + { + "epoch": 1.5973333333333333, + "grad_norm": 0.004373203497380018, + "learning_rate": 5.4517068429574215e-06, + "loss": 1.082428216934204, + "step": 113 + }, + { + "epoch": 1.6115555555555554, + "grad_norm": 0.0044046551920473576, + "learning_rate": 5.3697450830595775e-06, + "loss": 1.087543249130249, + "step": 114 + }, + { + "epoch": 1.6257777777777778, + "grad_norm": 0.004288993310183287, + "learning_rate": 5.287683294230855e-06, + "loss": 1.0576050281524658, + "step": 115 + }, + { + "epoch": 1.6400000000000001, + "grad_norm": 0.004459703341126442, + "learning_rate": 5.20554367704505e-06, + "loss": 1.0710945129394531, + "step": 116 + }, + { + "epoch": 1.6542222222222223, + "grad_norm": 0.004256462212651968, + "learning_rate": 5.123348453131242e-06, + "loss": 1.0972342491149902, + "step": 117 + }, + { + "epoch": 1.6684444444444444, + "grad_norm": 0.004406606312841177, + "learning_rate": 5.041119859162068e-06, + "loss": 1.0853114128112793, + "step": 118 + }, + { + "epoch": 1.6826666666666665, + "grad_norm": 0.00414910726249218, + "learning_rate": 4.958880140837934e-06, + "loss": 1.0747385025024414, + "step": 119 + }, + { + "epoch": 1.696888888888889, + "grad_norm": 0.004288721829652786, + "learning_rate": 4.876651546868759e-06, + "loss": 1.062084674835205, + "step": 120 + }, + { + "epoch": 1.7111111111111112, + "grad_norm": 0.004232042469084263, + "learning_rate": 4.794456322954953e-06, + "loss": 1.0754766464233398, + "step": 121 + }, + { + "epoch": 1.7253333333333334, + "grad_norm": 0.004211317747831345, + "learning_rate": 4.7123167057691446e-06, + "loss": 1.073303461074829, + "step": 122 + }, + { + "epoch": 1.7395555555555555, + "grad_norm": 0.004253690596669912, + "learning_rate": 4.630254916940424e-06, + "loss": 1.0916731357574463, + "step": 123 + }, + { + "epoch": 1.7537777777777777, + "grad_norm": 0.004212565720081329, + "learning_rate": 4.548293157042581e-06, + "loss": 1.0850458145141602, + "step": 124 + }, + { + "epoch": 1.768, + "grad_norm": 0.004267064854502678, + "learning_rate": 4.466453599588103e-06, + "loss": 1.0751748085021973, + "step": 125 + }, + { + "epoch": 1.7822222222222224, + "grad_norm": 0.004149065352976322, + "learning_rate": 4.384758385029457e-06, + "loss": 1.0639723539352417, + "step": 126 + }, + { + "epoch": 1.7964444444444445, + "grad_norm": 0.004006257280707359, + "learning_rate": 4.3032296147693225e-06, + "loss": 1.0787277221679688, + "step": 127 + }, + { + "epoch": 1.8106666666666666, + "grad_norm": 0.004197373986244202, + "learning_rate": 4.2218893451814005e-06, + "loss": 1.0513901710510254, + "step": 128 + }, + { + "epoch": 1.8248888888888888, + "grad_norm": 0.004353092983365059, + "learning_rate": 4.140759581643386e-06, + "loss": 1.0729951858520508, + "step": 129 + }, + { + "epoch": 1.8391111111111111, + "grad_norm": 0.003996321931481361, + "learning_rate": 4.059862272583755e-06, + "loss": 1.07584810256958, + "step": 130 + }, + { + "epoch": 1.8533333333333335, + "grad_norm": 0.0042655752040445805, + "learning_rate": 3.979219303543942e-06, + "loss": 1.0640318393707275, + "step": 131 + }, + { + "epoch": 1.8675555555555556, + "grad_norm": 0.004138551652431488, + "learning_rate": 3.898852491257547e-06, + "loss": 1.074927568435669, + "step": 132 + }, + { + "epoch": 1.8817777777777778, + "grad_norm": 0.0041953204199671745, + "learning_rate": 3.818783577748138e-06, + "loss": 1.0660316944122314, + "step": 133 + }, + { + "epoch": 1.896, + "grad_norm": 0.004029198549687862, + "learning_rate": 3.7390342244472886e-06, + "loss": 1.0795130729675293, + "step": 134 + }, + { + "epoch": 1.9102222222222223, + "grad_norm": 0.004120152443647385, + "learning_rate": 3.659626006334395e-06, + "loss": 1.094170093536377, + "step": 135 + }, + { + "epoch": 1.9244444444444444, + "grad_norm": 0.0042063528671860695, + "learning_rate": 3.5805804060998926e-06, + "loss": 1.0817457437515259, + "step": 136 + }, + { + "epoch": 1.9386666666666668, + "grad_norm": 0.004157442133873701, + "learning_rate": 3.501918808333453e-06, + "loss": 1.0608152151107788, + "step": 137 + }, + { + "epoch": 1.952888888888889, + "grad_norm": 0.004061015788465738, + "learning_rate": 3.4236624937386874e-06, + "loss": 1.0709357261657715, + "step": 138 + }, + { + "epoch": 1.967111111111111, + "grad_norm": 0.004056522622704506, + "learning_rate": 3.3458326333759927e-06, + "loss": 1.0752217769622803, + "step": 139 + }, + { + "epoch": 1.9813333333333332, + "grad_norm": 0.004041316919028759, + "learning_rate": 3.268450282935026e-06, + "loss": 1.0565910339355469, + "step": 140 + }, + { + "epoch": 1.9955555555555555, + "grad_norm": 0.00407154718413949, + "learning_rate": 3.1915363770384223e-06, + "loss": 1.064276099205017, + "step": 141 + }, + { + "epoch": 2.0, + "grad_norm": 0.006891065277159214, + "learning_rate": 3.1151117235782346e-06, + "loss": 1.0892033576965332, + "step": 142 + }, + { + "epoch": 2.014222222222222, + "grad_norm": 0.004137777257710695, + "learning_rate": 3.0391969980866874e-06, + "loss": 1.0656744241714478, + "step": 143 + }, + { + "epoch": 2.0284444444444443, + "grad_norm": 0.0042611584067344666, + "learning_rate": 2.963812738142713e-06, + "loss": 1.0707381963729858, + "step": 144 + }, + { + "epoch": 2.042666666666667, + "grad_norm": 0.004273556638509035, + "learning_rate": 2.8889793378158284e-06, + "loss": 1.059180736541748, + "step": 145 + }, + { + "epoch": 2.056888888888889, + "grad_norm": 0.004242253955453634, + "learning_rate": 2.814717042148827e-06, + "loss": 1.0461338758468628, + "step": 146 + }, + { + "epoch": 2.071111111111111, + "grad_norm": 0.004076985642313957, + "learning_rate": 2.7410459416807856e-06, + "loss": 1.0546996593475342, + "step": 147 + }, + { + "epoch": 2.0853333333333333, + "grad_norm": 0.003977393265813589, + "learning_rate": 2.6679859670118785e-06, + "loss": 1.061549186706543, + "step": 148 + }, + { + "epoch": 2.0995555555555554, + "grad_norm": 0.0039635817520320415, + "learning_rate": 2.5955568834114523e-06, + "loss": 1.0474296808242798, + "step": 149 + }, + { + "epoch": 2.113777777777778, + "grad_norm": 0.004132823087275028, + "learning_rate": 2.523778285470835e-06, + "loss": 1.024568796157837, + "step": 150 + }, + { + "epoch": 2.113777777777778, + "eval_loss": 1.079006552696228, + "eval_runtime": 47.4217, + "eval_samples_per_second": 4.217, + "eval_steps_per_second": 2.109, + "step": 150 + }, + { + "epoch": 2.128, + "grad_norm": 0.004181793890893459, + "learning_rate": 2.452669591802307e-06, + "loss": 1.0624573230743408, + "step": 151 + }, + { + "epoch": 2.1422222222222222, + "grad_norm": 0.0041247266344726086, + "learning_rate": 2.3822500397857016e-06, + "loss": 1.0404682159423828, + "step": 152 + }, + { + "epoch": 2.1564444444444444, + "grad_norm": 0.004161086864769459, + "learning_rate": 2.3125386803640188e-06, + "loss": 1.0580246448516846, + "step": 153 + }, + { + "epoch": 2.1706666666666665, + "grad_norm": 0.004000307526439428, + "learning_rate": 2.243554372889479e-06, + "loss": 1.0445423126220703, + "step": 154 + }, + { + "epoch": 2.1848888888888887, + "grad_norm": 0.004109199624508619, + "learning_rate": 2.1753157800214107e-06, + "loss": 1.0412821769714355, + "step": 155 + }, + { + "epoch": 2.1991111111111112, + "grad_norm": 0.003884031670168042, + "learning_rate": 2.1078413626773547e-06, + "loss": 1.0429916381835938, + "step": 156 + }, + { + "epoch": 2.2133333333333334, + "grad_norm": 0.004056555684655905, + "learning_rate": 2.0411493750387423e-06, + "loss": 1.0391228199005127, + "step": 157 + }, + { + "epoch": 2.2275555555555555, + "grad_norm": 0.004009357187896967, + "learning_rate": 1.9752578596124955e-06, + "loss": 1.031305193901062, + "step": 158 + }, + { + "epoch": 2.2417777777777776, + "grad_norm": 0.0040836031548678875, + "learning_rate": 1.9101846423499113e-06, + "loss": 1.055808424949646, + "step": 159 + }, + { + "epoch": 2.2560000000000002, + "grad_norm": 0.004020580090582371, + "learning_rate": 1.8459473278241125e-06, + "loss": 1.0406150817871094, + "step": 160 + }, + { + "epoch": 2.2702222222222224, + "grad_norm": 0.003893360961228609, + "learning_rate": 1.7825632944674016e-06, + "loss": 1.0353386402130127, + "step": 161 + }, + { + "epoch": 2.2844444444444445, + "grad_norm": 0.004030768759548664, + "learning_rate": 1.7200496898697832e-06, + "loss": 1.0691417455673218, + "step": 162 + }, + { + "epoch": 2.2986666666666666, + "grad_norm": 0.003913589753210545, + "learning_rate": 1.6584234261399535e-06, + "loss": 1.0500602722167969, + "step": 163 + }, + { + "epoch": 2.3128888888888888, + "grad_norm": 0.003951441030949354, + "learning_rate": 1.5977011753299726e-06, + "loss": 1.0313360691070557, + "step": 164 + }, + { + "epoch": 2.327111111111111, + "grad_norm": 0.004080148879438639, + "learning_rate": 1.5378993649249053e-06, + "loss": 1.0616838932037354, + "step": 165 + }, + { + "epoch": 2.3413333333333335, + "grad_norm": 0.004066621884703636, + "learning_rate": 1.4790341733986085e-06, + "loss": 1.0500047206878662, + "step": 166 + }, + { + "epoch": 2.3555555555555556, + "grad_norm": 0.0040935976430773735, + "learning_rate": 1.4211215258368866e-06, + "loss": 1.0583405494689941, + "step": 167 + }, + { + "epoch": 2.3697777777777778, + "grad_norm": 0.004103758372366428, + "learning_rate": 1.3641770896292083e-06, + "loss": 1.0435129404067993, + "step": 168 + }, + { + "epoch": 2.384, + "grad_norm": 0.004010037053376436, + "learning_rate": 1.3082162702301276e-06, + "loss": 1.0434541702270508, + "step": 169 + }, + { + "epoch": 2.398222222222222, + "grad_norm": 0.004051943309605122, + "learning_rate": 1.2532542069915722e-06, + "loss": 1.052001953125, + "step": 170 + }, + { + "epoch": 2.4124444444444446, + "grad_norm": 0.0039450060576200485, + "learning_rate": 1.1993057690671174e-06, + "loss": 1.055966854095459, + "step": 171 + }, + { + "epoch": 2.4266666666666667, + "grad_norm": 0.00400911271572113, + "learning_rate": 1.1463855513893695e-06, + "loss": 1.0502328872680664, + "step": 172 + }, + { + "epoch": 2.440888888888889, + "grad_norm": 0.004119160585105419, + "learning_rate": 1.0945078707215224e-06, + "loss": 1.0345993041992188, + "step": 173 + }, + { + "epoch": 2.455111111111111, + "grad_norm": 0.004055999685078859, + "learning_rate": 1.0436867617841768e-06, + "loss": 1.0489795207977295, + "step": 174 + }, + { + "epoch": 2.469333333333333, + "grad_norm": 0.003937677945941687, + "learning_rate": 9.939359734584552e-07, + "loss": 1.057159662246704, + "step": 175 + }, + { + "epoch": 2.4835555555555557, + "grad_norm": 0.003958112560212612, + "learning_rate": 9.452689650664515e-07, + "loss": 1.0306581258773804, + "step": 176 + }, + { + "epoch": 2.497777777777778, + "grad_norm": 0.004141568671911955, + "learning_rate": 8.976989027300265e-07, + "loss": 1.0416128635406494, + "step": 177 + }, + { + "epoch": 2.512, + "grad_norm": 0.004062578082084656, + "learning_rate": 8.512386558088919e-07, + "loss": 1.0480129718780518, + "step": 178 + }, + { + "epoch": 2.526222222222222, + "grad_norm": 0.003910359926521778, + "learning_rate": 8.059007934190194e-07, + "loss": 1.047067642211914, + "step": 179 + }, + { + "epoch": 2.5404444444444443, + "grad_norm": 0.004017462953925133, + "learning_rate": 7.61697581032243e-07, + "loss": 1.053797721862793, + "step": 180 + }, + { + "epoch": 2.554666666666667, + "grad_norm": 0.0039033186621963978, + "learning_rate": 7.186409771580355e-07, + "loss": 1.0373890399932861, + "step": 181 + }, + { + "epoch": 2.568888888888889, + "grad_norm": 0.004007395822554827, + "learning_rate": 6.76742630108298e-07, + "loss": 1.043588399887085, + "step": 182 + }, + { + "epoch": 2.583111111111111, + "grad_norm": 0.003964278846979141, + "learning_rate": 6.360138748461015e-07, + "loss": 1.03302001953125, + "step": 183 + }, + { + "epoch": 2.5973333333333333, + "grad_norm": 0.0039159939624369144, + "learning_rate": 5.964657299191712e-07, + "loss": 1.0640901327133179, + "step": 184 + }, + { + "epoch": 2.6115555555555554, + "grad_norm": 0.003952207043766975, + "learning_rate": 5.581088944789953e-07, + "loss": 1.050436019897461, + "step": 185 + }, + { + "epoch": 2.6257777777777775, + "grad_norm": 0.004017418250441551, + "learning_rate": 5.20953745386329e-07, + "loss": 1.0415809154510498, + "step": 186 + }, + { + "epoch": 2.64, + "grad_norm": 0.003962535876780748, + "learning_rate": 4.850103344038853e-07, + "loss": 1.0449340343475342, + "step": 187 + }, + { + "epoch": 2.6542222222222223, + "grad_norm": 0.00386899639852345, + "learning_rate": 4.502883854769935e-07, + "loss": 1.0389983654022217, + "step": 188 + }, + { + "epoch": 2.6684444444444444, + "grad_norm": 0.003973118029534817, + "learning_rate": 4.167972921029262e-07, + "loss": 1.0497729778289795, + "step": 189 + }, + { + "epoch": 2.6826666666666665, + "grad_norm": 0.003942268900573254, + "learning_rate": 3.8454611478963235e-07, + "loss": 1.036604642868042, + "step": 190 + }, + { + "epoch": 2.696888888888889, + "grad_norm": 0.0038068366702646017, + "learning_rate": 3.535435786045538e-07, + "loss": 1.0535285472869873, + "step": 191 + }, + { + "epoch": 2.7111111111111112, + "grad_norm": 0.0040411436930298805, + "learning_rate": 3.237980708141919e-07, + "loss": 1.0318130254745483, + "step": 192 + }, + { + "epoch": 2.7253333333333334, + "grad_norm": 0.003907525446265936, + "learning_rate": 2.9531763861505967e-07, + "loss": 1.0349407196044922, + "step": 193 + }, + { + "epoch": 2.7395555555555555, + "grad_norm": 0.004061816725879908, + "learning_rate": 2.681099869566328e-07, + "loss": 1.0367615222930908, + "step": 194 + }, + { + "epoch": 2.7537777777777777, + "grad_norm": 0.003970359917730093, + "learning_rate": 2.4218247645689306e-07, + "loss": 1.0251959562301636, + "step": 195 + }, + { + "epoch": 2.768, + "grad_norm": 0.003819686360657215, + "learning_rate": 2.175421214110235e-07, + "loss": 1.0571613311767578, + "step": 196 + }, + { + "epoch": 2.7822222222222224, + "grad_norm": 0.0039050441700965166, + "learning_rate": 1.941955878938029e-07, + "loss": 1.0457653999328613, + "step": 197 + }, + { + "epoch": 2.7964444444444445, + "grad_norm": 0.003968564793467522, + "learning_rate": 1.721491919561913e-07, + "loss": 1.0394766330718994, + "step": 198 + }, + { + "epoch": 2.8106666666666666, + "grad_norm": 0.003920818213373423, + "learning_rate": 1.514088979166256e-07, + "loss": 1.0483171939849854, + "step": 199 + }, + { + "epoch": 2.824888888888889, + "grad_norm": 0.003929824102669954, + "learning_rate": 1.3198031674745814e-07, + "loss": 1.0300005674362183, + "step": 200 + }, + { + "epoch": 2.824888888888889, + "eval_loss": 1.0748209953308105, + "eval_runtime": 46.195, + "eval_samples_per_second": 4.329, + "eval_steps_per_second": 2.165, + "step": 200 + }, + { + "epoch": 2.8391111111111114, + "grad_norm": 0.0038694061804562807, + "learning_rate": 1.138687045569975e-07, + "loss": 1.0445573329925537, + "step": 201 + }, + { + "epoch": 2.8533333333333335, + "grad_norm": 0.00398522661998868, + "learning_rate": 9.70789611675449e-08, + "loss": 1.0359022617340088, + "step": 202 + }, + { + "epoch": 2.8675555555555556, + "grad_norm": 0.004058049060404301, + "learning_rate": 8.161562878982399e-08, + "loss": 1.0159897804260254, + "step": 203 + }, + { + "epoch": 2.8817777777777778, + "grad_norm": 0.004108336754143238, + "learning_rate": 6.74828907941516e-08, + "loss": 1.0329415798187256, + "step": 204 + }, + { + "epoch": 2.896, + "grad_norm": 0.0038334683049470186, + "learning_rate": 5.468457057869358e-08, + "loss": 1.0553950071334839, + "step": 205 + }, + { + "epoch": 2.910222222222222, + "grad_norm": 0.0037339332047849894, + "learning_rate": 4.322413053509944e-08, + "loss": 1.0415141582489014, + "step": 206 + }, + { + "epoch": 2.924444444444444, + "grad_norm": 0.004045915324240923, + "learning_rate": 3.31046711118066e-08, + "loss": 1.0387455224990845, + "step": 207 + }, + { + "epoch": 2.9386666666666668, + "grad_norm": 0.0039448207244277, + "learning_rate": 2.4328929975260262e-08, + "loss": 1.041945457458496, + "step": 208 + }, + { + "epoch": 2.952888888888889, + "grad_norm": 0.004023151472210884, + "learning_rate": 1.6899281269279756e-08, + "loss": 1.0387485027313232, + "step": 209 + }, + { + "epoch": 2.967111111111111, + "grad_norm": 0.0038402979262173176, + "learning_rate": 1.0817734972768946e-08, + "loss": 1.0304040908813477, + "step": 210 + }, + { + "epoch": 2.981333333333333, + "grad_norm": 0.003873219946399331, + "learning_rate": 6.085936355947897e-09, + "loss": 1.0460115671157837, + "step": 211 + }, + { + "epoch": 2.9955555555555557, + "grad_norm": 0.0037621951196342707, + "learning_rate": 2.7051655352494654e-09, + "loss": 1.0432226657867432, + "step": 212 + }, + { + "epoch": 3.0, + "grad_norm": 0.006865890696644783, + "learning_rate": 6.763371270035457e-10, + "loss": 1.0157432556152344, + "step": 213 + }, + { + "epoch": 3.0, + "step": 213, + "total_flos": 9.561792243333857e+17, + "train_loss": 1.1234366983315194, + "train_runtime": 19384.6824, + "train_samples_per_second": 2.786, + "train_steps_per_second": 0.011 + } + ], + "logging_steps": 1, + "max_steps": 213, + "num_input_tokens_seen": 0, + "num_train_epochs": 3, + "save_steps": 500, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 9.561792243333857e+17, + "train_batch_size": 1, + "trial_name": null, + "trial_params": null +} diff --git a/qwen2.5-3b-instruct/training_args.bin b/qwen2.5-3b-instruct/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..745b9d6d614f3064b70369b5149a4865ef67205e --- /dev/null +++ b/qwen2.5-3b-instruct/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:84503fc1ab703b8b0cb47dbab1adb9be59585b090c011d1af556d442423fd3c6 +size 5585 diff --git a/qwen2.5-3b-instruct/training_eval_loss.png b/qwen2.5-3b-instruct/training_eval_loss.png new file mode 100644 index 0000000000000000000000000000000000000000..471d60761327cb8a0c40a783f149938148934ead Binary files /dev/null and b/qwen2.5-3b-instruct/training_eval_loss.png differ diff --git a/qwen2.5-3b-instruct/training_loss.png b/qwen2.5-3b-instruct/training_loss.png new file mode 100644 index 0000000000000000000000000000000000000000..6a25839e232818cf702b4f4bb6c1c6a033d16957 Binary files /dev/null and b/qwen2.5-3b-instruct/training_loss.png differ