commit 0fa22ebb7a58c5c7db2fdad32a24d0e920e9974e Author: ModelHub XC Date: Sat Jun 13 09:03:16 2026 +0800 初始化项目，由ModelHub XC社区提供模型 Model: alirizaercan/qwen25_05b_base_full_ft_lunarlander_a4000 Source: Original Platform diff --git a/.gitattributes b/.gitattributes new file mode 100644 index 0000000..716d4a7 --- /dev/null +++ b/.gitattributes @@ -0,0 +1,39 @@ +*.7z filter=lfs diff=lfs merge=lfs -text +*.arrow filter=lfs diff=lfs merge=lfs -text +*.bin filter=lfs diff=lfs merge=lfs -text +*.bz2 filter=lfs diff=lfs merge=lfs -text +*.ckpt filter=lfs diff=lfs merge=lfs -text +*.ftz filter=lfs diff=lfs merge=lfs -text +*.gz filter=lfs diff=lfs merge=lfs -text +*.h5 filter=lfs diff=lfs merge=lfs -text +*.joblib filter=lfs diff=lfs merge=lfs -text +*.lfs.* filter=lfs diff=lfs merge=lfs -text +*.mlmodel filter=lfs diff=lfs merge=lfs -text +*.model filter=lfs diff=lfs merge=lfs -text +*.msgpack filter=lfs diff=lfs merge=lfs -text +*.npy filter=lfs diff=lfs merge=lfs -text +*.npz filter=lfs diff=lfs merge=lfs -text +*.onnx filter=lfs diff=lfs merge=lfs -text +*.ot filter=lfs diff=lfs merge=lfs -text +*.parquet filter=lfs diff=lfs merge=lfs -text +*.pb filter=lfs diff=lfs merge=lfs -text +*.pickle filter=lfs diff=lfs merge=lfs -text +*.pkl filter=lfs diff=lfs merge=lfs -text +*.pt filter=lfs diff=lfs merge=lfs -text +*.pth filter=lfs diff=lfs merge=lfs -text +*.rar filter=lfs diff=lfs merge=lfs -text +*.safetensors filter=lfs diff=lfs merge=lfs -text +saved_model/**/* filter=lfs diff=lfs merge=lfs -text +*.tar.* filter=lfs diff=lfs merge=lfs -text +*.tar filter=lfs diff=lfs merge=lfs -text +*.tflite filter=lfs diff=lfs merge=lfs -text +*.tgz filter=lfs diff=lfs merge=lfs -text +*.wasm filter=lfs diff=lfs merge=lfs -text +*.xz filter=lfs diff=lfs merge=lfs -text +*.zip filter=lfs diff=lfs merge=lfs -text +*.zst filter=lfs diff=lfs merge=lfs -text +*tfevents* filter=lfs diff=lfs merge=lfs -text +checkpoint-5800/tokenizer.json filter=lfs diff=lfs merge=lfs -text +checkpoint-6000/tokenizer.json filter=lfs diff=lfs merge=lfs -text +checkpoint-6125/tokenizer.json filter=lfs diff=lfs merge=lfs -text +tokenizer.json filter=lfs diff=lfs merge=lfs -text diff --git a/README.md b/README.md new file mode 100644 index 0000000..abd4ce3 --- /dev/null +++ b/README.md @@ -0,0 +1,97 @@ +--- +library_name: transformers +license: other +base_model: Qwen/Qwen2.5-0.5B +tags: +- llama-factory +- full +- generated_from_trainer +metrics: +- accuracy +model-index: +- name: qwen25_05b_base_full_ft_lunarlander_a4000 + results: [] +--- + + + +# qwen25_05b_base_full_ft_lunarlander_a4000 + +This model is a fine-tuned version of [Qwen/Qwen2.5-0.5B](https://huggingface.co/Qwen/Qwen2.5-0.5B) on the lunar_lander_270_reward_train dataset. +It achieves the following results on the evaluation set: +- Loss: 0.0253 +- Accuracy: 0.9905 + +## 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: 5e-06 +- train_batch_size: 1 +- eval_batch_size: 1 +- seed: 42 +- distributed_type: multi-GPU +- gradient_accumulation_steps: 32 +- total_train_batch_size: 32 +- 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: 100.0 +- num_epochs: 1.0 +- mixed_precision_training: Native AMP + +### Training results + +| Training Loss | Epoch | Step | Validation Loss | Accuracy | +|:-------------:|:------:|:----:|:---------------:|:--------:| +| 0.1271 | 0.0327 | 200 | 0.1269 | 0.9408 | +| 0.1134 | 0.0653 | 400 | 0.1141 | 0.9496 | +| 0.0951 | 0.0980 | 600 | 0.0991 | 0.9562 | +| 0.0866 | 0.1306 | 800 | 0.0930 | 0.9564 | +| 0.0909 | 0.1633 | 1000 | 0.0883 | 0.9608 | +| 0.0980 | 0.1959 | 1200 | 0.0856 | 0.963 | +| 0.0836 | 0.2286 | 1400 | 0.0839 | 0.9611 | +| 0.0709 | 0.2612 | 1600 | 0.0811 | 0.9644 | +| 0.0757 | 0.2939 | 1800 | 0.0765 | 0.9665 | +| 0.0701 | 0.3265 | 2000 | 0.0791 | 0.9655 | +| 0.0717 | 0.3592 | 2200 | 0.0681 | 0.9706 | +| 0.0637 | 0.3918 | 2400 | 0.0704 | 0.9686 | +| 0.0619 | 0.4245 | 2600 | 0.0618 | 0.9747 | +| 0.0572 | 0.4571 | 2800 | 0.0562 | 0.9774 | +| 0.0605 | 0.4898 | 3000 | 0.0561 | 0.9767 | +| 0.0648 | 0.5224 | 3200 | 0.0584 | 0.9745 | +| 0.0659 | 0.5551 | 3400 | 0.0540 | 0.9771 | +| 0.0564 | 0.5878 | 3600 | 0.0479 | 0.9799 | +| 0.0545 | 0.6204 | 3800 | 0.0432 | 0.9819 | +| 0.0468 | 0.6531 | 4000 | 0.0449 | 0.9817 | +| 0.0388 | 0.6857 | 4200 | 0.0395 | 0.9840 | +| 0.0328 | 0.7184 | 4400 | 0.0397 | 0.9831 | +| 0.0363 | 0.7510 | 4600 | 0.0357 | 0.9856 | +| 0.0314 | 0.7837 | 4800 | 0.0362 | 0.985 | +| 0.0280 | 0.8163 | 5000 | 0.0336 | 0.9861 | +| 0.0308 | 0.8490 | 5200 | 0.0292 | 0.9882 | +| 0.0252 | 0.8816 | 5400 | 0.0268 | 0.9894 | +| 0.0202 | 0.9143 | 5600 | 0.0258 | 0.9901 | +| 0.0252 | 0.9469 | 5800 | 0.0255 | 0.9902 | +| 0.0201 | 0.9796 | 6000 | 0.0253 | 0.9904 | + + +### Framework versions + +- Transformers 5.2.0 +- Pytorch 2.11.0+cu130 +- Datasets 4.0.0 +- Tokenizers 0.22.2 diff --git a/all_results.json b/all_results.json new file mode 100644 index 0000000..13a3b1e --- /dev/null +++ b/all_results.json @@ -0,0 +1,13 @@ +{ + "epoch": 1.0, + "eval_accuracy": 0.9905, + "eval_loss": 0.025262661278247833, + "eval_runtime": 156.0067, + "eval_samples_per_second": 25.64, + "eval_steps_per_second": 25.64, + "total_flos": 5.72407910760448e+16, + "train_loss": 0.07194560643361539, + "train_runtime": 34224.4204, + "train_samples_per_second": 5.727, + "train_steps_per_second": 0.179 +} \ No newline at end of file diff --git a/chat_template.jinja b/chat_template.jinja new file mode 100644 index 0000000..ce566c4 --- /dev/null +++ b/chat_template.jinja @@ -0,0 +1,9 @@ +{% set system_message = 'Below is an instruction that describes a task. Write a response that appropriately completes the request. + +' %}{% if messages[0]['role'] == 'system' %}{% set loop_messages = messages[1:] %}{% set system_message = messages[0]['content'] %}{% else %}{% set loop_messages = messages %}{% endif %}{% if system_message is defined %}{{ system_message }}{% endif %}{% for message in loop_messages %}{% set content = message['content'] %}{% if message['role'] == 'user' %}{{ '### Instruction: +' + content + ' + +### Response: +' }}{% elif message['role'] == 'assistant' %}{{ content + '<|endoftext|>' + ' + +' }}{% endif %}{% endfor %} \ No newline at end of file diff --git a/checkpoint-5800/chat_template.jinja b/checkpoint-5800/chat_template.jinja new file mode 100644 index 0000000..ce566c4 --- /dev/null +++ b/checkpoint-5800/chat_template.jinja @@ -0,0 +1,9 @@ +{% set system_message = 'Below is an instruction that describes a task. Write a response that appropriately completes the request. + +' %}{% if messages[0]['role'] == 'system' %}{% set loop_messages = messages[1:] %}{% set system_message = messages[0]['content'] %}{% else %}{% set loop_messages = messages %}{% endif %}{% if system_message is defined %}{{ system_message }}{% endif %}{% for message in loop_messages %}{% set content = message['content'] %}{% if message['role'] == 'user' %}{{ '### Instruction: +' + content + ' + +### Response: +' }}{% elif message['role'] == 'assistant' %}{{ content + '<|endoftext|>' + ' + +' }}{% endif %}{% endfor %} \ No newline at end of file diff --git a/checkpoint-5800/config.json b/checkpoint-5800/config.json new file mode 100644 index 0000000..642be54 --- /dev/null +++ b/checkpoint-5800/config.json @@ -0,0 +1,58 @@ +{ + "architectures": [ + "Qwen2ForCausalLM" + ], + "attention_dropout": 0.0, + "bos_token_id": null, + "dtype": "float32", + "eos_token_id": 151643, + "hidden_act": "silu", + "hidden_size": 896, + "initializer_range": 0.02, + "intermediate_size": 4864, + "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" + ], + "max_position_embeddings": 32768, + "max_window_layers": 24, + "model_type": "qwen2", + "num_attention_heads": 14, + "num_hidden_layers": 24, + "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_mrope": false, + "use_sliding_window": false, + "vocab_size": 151936 +} diff --git a/checkpoint-5800/generation_config.json b/checkpoint-5800/generation_config.json new file mode 100644 index 0000000..795a75e --- /dev/null +++ b/checkpoint-5800/generation_config.json @@ -0,0 +1,9 @@ +{ + "do_sample": false, + "eos_token_id": [ + 151643 + ], + "max_new_tokens": 2048, + "pad_token_id": 151643, + "transformers_version": "5.2.0" +} diff --git a/checkpoint-5800/model.safetensors b/checkpoint-5800/model.safetensors new file mode 100644 index 0000000..6042495 --- /dev/null +++ b/checkpoint-5800/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:272f1dcaf0c796e2aae409931633f49a52cc819396bc4a945cceb579847cd088 +size 1976163472 diff --git a/checkpoint-5800/tokenizer.json b/checkpoint-5800/tokenizer.json new file mode 100644 index 0000000..34510ff --- /dev/null +++ b/checkpoint-5800/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3fd169731d2cbde95e10bf356d66d5997fd885dd8dbb6fb4684da3f23b2585d8 +size 11421892 diff --git a/checkpoint-5800/tokenizer_config.json b/checkpoint-5800/tokenizer_config.json new file mode 100644 index 0000000..392ded1 --- /dev/null +++ b/checkpoint-5800/tokenizer_config.json @@ -0,0 +1,30 @@ +{ + "add_prefix_space": false, + "backend": "tokenizers", + "bos_token": null, + "clean_up_tokenization_spaces": false, + "eos_token": "<|endoftext|>", + "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/checkpoint-5800/trainer_state.json b/checkpoint-5800/trainer_state.json new file mode 100644 index 0000000..90ff41f --- /dev/null +++ b/checkpoint-5800/trainer_state.json @@ -0,0 +1,4355 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.9469387755102041, + "eval_steps": 200, + "global_step": 5800, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0016326530612244899, + "grad_norm": 7.955283164978027, + "learning_rate": 4.5000000000000003e-07, + "loss": 4.391213607788086, + "step": 10 + }, + { + "epoch": 0.0032653061224489797, + "grad_norm": 2.4868686199188232, + "learning_rate": 9.500000000000001e-07, + "loss": 1.9876020431518555, + "step": 20 + }, + { + "epoch": 0.004897959183673469, + "grad_norm": 0.25438952445983887, + "learning_rate": 1.45e-06, + "loss": 0.34127225875854494, + "step": 30 + }, + { + "epoch": 0.006530612244897959, + "grad_norm": 0.13044093549251556, + "learning_rate": 1.9500000000000004e-06, + "loss": 0.23643033504486083, + "step": 40 + }, + { + "epoch": 0.00816326530612245, + "grad_norm": 0.17984068393707275, + "learning_rate": 2.4500000000000003e-06, + "loss": 0.23313720226287843, + "step": 50 + }, + { + "epoch": 0.009795918367346938, + "grad_norm": 0.15249308943748474, + "learning_rate": 2.95e-06, + "loss": 0.22351114749908446, + "step": 60 + }, + { + "epoch": 0.011428571428571429, + "grad_norm": 0.07123949378728867, + "learning_rate": 3.45e-06, + "loss": 0.21615636348724365, + "step": 70 + }, + { + "epoch": 0.013061224489795919, + "grad_norm": 0.21456502377986908, + "learning_rate": 3.95e-06, + "loss": 0.21677422523498535, + "step": 80 + }, + { + "epoch": 0.014693877551020407, + "grad_norm": 0.22391553223133087, + "learning_rate": 4.450000000000001e-06, + "loss": 0.21422204971313477, + "step": 90 + }, + { + "epoch": 0.0163265306122449, + "grad_norm": 0.1745220273733139, + "learning_rate": 4.95e-06, + "loss": 0.19350455999374389, + "step": 100 + }, + { + "epoch": 0.017959183673469388, + "grad_norm": 0.18820920586585999, + "learning_rate": 4.9999724716692435e-06, + "loss": 0.1981007933616638, + "step": 110 + }, + { + "epoch": 0.019591836734693877, + "grad_norm": 0.21733109652996063, + "learning_rate": 4.9998773127856985e-06, + "loss": 0.16994532346725463, + "step": 120 + }, + { + "epoch": 0.02122448979591837, + "grad_norm": 0.40475547313690186, + "learning_rate": 4.999714186080037e-06, + "loss": 0.1539763927459717, + "step": 130 + }, + { + "epoch": 0.022857142857142857, + "grad_norm": 0.3676598370075226, + "learning_rate": 4.999483095987423e-06, + "loss": 0.16321685314178466, + "step": 140 + }, + { + "epoch": 0.024489795918367346, + "grad_norm": 0.19484871625900269, + "learning_rate": 4.9991840487908426e-06, + "loss": 0.16523454189300538, + "step": 150 + }, + { + "epoch": 0.026122448979591838, + "grad_norm": 0.21928830444812775, + "learning_rate": 4.998817052620928e-06, + "loss": 0.14424897432327272, + "step": 160 + }, + { + "epoch": 0.027755102040816326, + "grad_norm": 0.2790975868701935, + "learning_rate": 4.9983821174557415e-06, + "loss": 0.13810645341873168, + "step": 170 + }, + { + "epoch": 0.029387755102040815, + "grad_norm": 0.21702228486537933, + "learning_rate": 4.997879255120501e-06, + "loss": 0.1318126440048218, + "step": 180 + }, + { + "epoch": 0.031020408163265307, + "grad_norm": 0.16967810690402985, + "learning_rate": 4.997308479287263e-06, + "loss": 0.14249314069747926, + "step": 190 + }, + { + "epoch": 0.0326530612244898, + "grad_norm": 0.14007994532585144, + "learning_rate": 4.996669805474542e-06, + "loss": 0.12708702087402343, + "step": 200 + }, + { + "epoch": 0.0326530612244898, + "eval_accuracy": 0.9408333333333335, + "eval_loss": 0.12691152095794678, + "eval_runtime": 156.68, + "eval_samples_per_second": 25.53, + "eval_steps_per_second": 25.53, + "step": 200 + }, + { + "epoch": 0.03428571428571429, + "grad_norm": 0.2674005627632141, + "learning_rate": 4.9959632510469e-06, + "loss": 0.11019121408462525, + "step": 210 + }, + { + "epoch": 0.035918367346938776, + "grad_norm": 0.12073638290166855, + "learning_rate": 4.995188835214465e-06, + "loss": 0.12032746076583863, + "step": 220 + }, + { + "epoch": 0.037551020408163265, + "grad_norm": 0.11186162382364273, + "learning_rate": 4.994346579032413e-06, + "loss": 0.12122929096221924, + "step": 230 + }, + { + "epoch": 0.03918367346938775, + "grad_norm": 0.09840800613164902, + "learning_rate": 4.993436505400396e-06, + "loss": 0.11416752338409424, + "step": 240 + }, + { + "epoch": 0.04081632653061224, + "grad_norm": 0.19076374173164368, + "learning_rate": 4.992458639061916e-06, + "loss": 0.11376930475234985, + "step": 250 + }, + { + "epoch": 0.04244897959183674, + "grad_norm": 0.14405065774917603, + "learning_rate": 4.991413006603657e-06, + "loss": 0.10997685194015502, + "step": 260 + }, + { + "epoch": 0.044081632653061226, + "grad_norm": 0.23844055831432343, + "learning_rate": 4.990299636454759e-06, + "loss": 0.1000083327293396, + "step": 270 + }, + { + "epoch": 0.045714285714285714, + "grad_norm": 0.23128052055835724, + "learning_rate": 4.989118558886043e-06, + "loss": 0.1243711233139038, + "step": 280 + }, + { + "epoch": 0.0473469387755102, + "grad_norm": 0.19849570095539093, + "learning_rate": 4.9878698060091955e-06, + "loss": 0.11728224754333497, + "step": 290 + }, + { + "epoch": 0.04897959183673469, + "grad_norm": 0.13039851188659668, + "learning_rate": 4.986553411775886e-06, + "loss": 0.09353847503662109, + "step": 300 + }, + { + "epoch": 0.05061224489795919, + "grad_norm": 0.19887588918209076, + "learning_rate": 4.985169411976852e-06, + "loss": 0.11291944980621338, + "step": 310 + }, + { + "epoch": 0.052244897959183675, + "grad_norm": 0.25038304924964905, + "learning_rate": 4.98371784424092e-06, + "loss": 0.11149277687072753, + "step": 320 + }, + { + "epoch": 0.053877551020408164, + "grad_norm": 0.238360196352005, + "learning_rate": 4.982198748033986e-06, + "loss": 0.11344953775405883, + "step": 330 + }, + { + "epoch": 0.05551020408163265, + "grad_norm": 0.13686658442020416, + "learning_rate": 4.980612164657943e-06, + "loss": 0.09844974875450134, + "step": 340 + }, + { + "epoch": 0.05714285714285714, + "grad_norm": 0.0829666405916214, + "learning_rate": 4.978958137249551e-06, + "loss": 0.10898438692092896, + "step": 350 + }, + { + "epoch": 0.05877551020408163, + "grad_norm": 0.10389993339776993, + "learning_rate": 4.977236710779277e-06, + "loss": 0.09371156096458436, + "step": 360 + }, + { + "epoch": 0.060408163265306125, + "grad_norm": 0.18182680010795593, + "learning_rate": 4.9754479320500584e-06, + "loss": 0.1049533724784851, + "step": 370 + }, + { + "epoch": 0.062040816326530614, + "grad_norm": 0.14488115906715393, + "learning_rate": 4.973591849696041e-06, + "loss": 0.10510741472244263, + "step": 380 + }, + { + "epoch": 0.0636734693877551, + "grad_norm": 0.14816738665103912, + "learning_rate": 4.971668514181252e-06, + "loss": 0.10956356525421143, + "step": 390 + }, + { + "epoch": 0.0653061224489796, + "grad_norm": 0.20584072172641754, + "learning_rate": 4.969677977798225e-06, + "loss": 0.11344989538192748, + "step": 400 + }, + { + "epoch": 0.0653061224489796, + "eval_accuracy": 0.9496250000000002, + "eval_loss": 0.11409253627061844, + "eval_runtime": 165.7682, + "eval_samples_per_second": 24.13, + "eval_steps_per_second": 24.13, + "step": 400 + }, + { + "epoch": 0.06693877551020408, + "grad_norm": 0.1342787891626358, + "learning_rate": 4.967620294666589e-06, + "loss": 0.09949527978897095, + "step": 410 + }, + { + "epoch": 0.06857142857142857, + "grad_norm": 0.17336596548557281, + "learning_rate": 4.965495520731583e-06, + "loss": 0.10606559514999389, + "step": 420 + }, + { + "epoch": 0.07020408163265306, + "grad_norm": 0.19437003135681152, + "learning_rate": 4.963303713762547e-06, + "loss": 0.11677042245864869, + "step": 430 + }, + { + "epoch": 0.07183673469387755, + "grad_norm": 0.12076073884963989, + "learning_rate": 4.961044933351342e-06, + "loss": 0.1019243597984314, + "step": 440 + }, + { + "epoch": 0.07346938775510205, + "grad_norm": 0.10272067785263062, + "learning_rate": 4.958719240910737e-06, + "loss": 0.10125980377197266, + "step": 450 + }, + { + "epoch": 0.07510204081632653, + "grad_norm": 0.1080622524023056, + "learning_rate": 4.956326699672736e-06, + "loss": 0.09888594150543213, + "step": 460 + }, + { + "epoch": 0.07673469387755102, + "grad_norm": 0.1307847797870636, + "learning_rate": 4.953867374686856e-06, + "loss": 0.10753540992736817, + "step": 470 + }, + { + "epoch": 0.0783673469387755, + "grad_norm": 0.2128073275089264, + "learning_rate": 4.951341332818365e-06, + "loss": 0.11452326774597169, + "step": 480 + }, + { + "epoch": 0.08, + "grad_norm": 0.21006478369235992, + "learning_rate": 4.9487486427464594e-06, + "loss": 0.09555010795593262, + "step": 490 + }, + { + "epoch": 0.08163265306122448, + "grad_norm": 0.24889807403087616, + "learning_rate": 4.946089374962396e-06, + "loss": 0.09771155118942261, + "step": 500 + }, + { + "epoch": 0.08326530612244898, + "grad_norm": 0.13042858242988586, + "learning_rate": 4.943363601767582e-06, + "loss": 0.10659778118133545, + "step": 510 + }, + { + "epoch": 0.08489795918367347, + "grad_norm": 0.2210501879453659, + "learning_rate": 4.940571397271598e-06, + "loss": 0.10161921977996827, + "step": 520 + }, + { + "epoch": 0.08653061224489796, + "grad_norm": 0.11963305622339249, + "learning_rate": 4.937712837390194e-06, + "loss": 0.09155921936035157, + "step": 530 + }, + { + "epoch": 0.08816326530612245, + "grad_norm": 0.07435746490955353, + "learning_rate": 4.93478799984322e-06, + "loss": 0.11164016723632812, + "step": 540 + }, + { + "epoch": 0.08979591836734693, + "grad_norm": 0.16348940134048462, + "learning_rate": 4.9317969641525154e-06, + "loss": 0.10738091468811035, + "step": 550 + }, + { + "epoch": 0.09142857142857143, + "grad_norm": 0.12221129983663559, + "learning_rate": 4.928739811639741e-06, + "loss": 0.09491395950317383, + "step": 560 + }, + { + "epoch": 0.09306122448979592, + "grad_norm": 0.1107218861579895, + "learning_rate": 4.9256166254241775e-06, + "loss": 0.08408294320106506, + "step": 570 + }, + { + "epoch": 0.0946938775510204, + "grad_norm": 0.207329660654068, + "learning_rate": 4.922427490420457e-06, + "loss": 0.10272791385650634, + "step": 580 + }, + { + "epoch": 0.0963265306122449, + "grad_norm": 0.14269684255123138, + "learning_rate": 4.919172493336259e-06, + "loss": 0.10189608335494996, + "step": 590 + }, + { + "epoch": 0.09795918367346938, + "grad_norm": 0.10970286279916763, + "learning_rate": 4.915851722669952e-06, + "loss": 0.09506216645240784, + "step": 600 + }, + { + "epoch": 0.09795918367346938, + "eval_accuracy": 0.9561666666666667, + "eval_loss": 0.09913250058889389, + "eval_runtime": 158.3659, + "eval_samples_per_second": 25.258, + "eval_steps_per_second": 25.258, + "step": 600 + }, + { + "epoch": 0.09959183673469388, + "grad_norm": 0.11382903158664703, + "learning_rate": 4.912465268708186e-06, + "loss": 0.09665942788124085, + "step": 610 + }, + { + "epoch": 0.10122448979591837, + "grad_norm": 0.24785904586315155, + "learning_rate": 4.90901322352344e-06, + "loss": 0.10539349317550659, + "step": 620 + }, + { + "epoch": 0.10285714285714286, + "grad_norm": 0.18585072457790375, + "learning_rate": 4.905495680971518e-06, + "loss": 0.10907092094421386, + "step": 630 + }, + { + "epoch": 0.10448979591836735, + "grad_norm": 0.16385921835899353, + "learning_rate": 4.901912736688995e-06, + "loss": 0.09973281621932983, + "step": 640 + }, + { + "epoch": 0.10612244897959183, + "grad_norm": 0.10820458829402924, + "learning_rate": 4.898264488090619e-06, + "loss": 0.09869267344474793, + "step": 650 + }, + { + "epoch": 0.10775510204081633, + "grad_norm": 0.18020115792751312, + "learning_rate": 4.894551034366665e-06, + "loss": 0.10499387979507446, + "step": 660 + }, + { + "epoch": 0.10938775510204081, + "grad_norm": 0.10356947779655457, + "learning_rate": 4.890772476480229e-06, + "loss": 0.10153095722198487, + "step": 670 + }, + { + "epoch": 0.1110204081632653, + "grad_norm": 0.08888879418373108, + "learning_rate": 4.886928917164496e-06, + "loss": 0.08613221645355225, + "step": 680 + }, + { + "epoch": 0.1126530612244898, + "grad_norm": 0.11978956311941147, + "learning_rate": 4.8830204609199335e-06, + "loss": 0.09545036554336547, + "step": 690 + }, + { + "epoch": 0.11428571428571428, + "grad_norm": 0.11924679577350616, + "learning_rate": 4.8790472140114635e-06, + "loss": 0.0950542688369751, + "step": 700 + }, + { + "epoch": 0.11591836734693878, + "grad_norm": 0.0637463852763176, + "learning_rate": 4.875009284465561e-06, + "loss": 0.10127921104431152, + "step": 710 + }, + { + "epoch": 0.11755102040816326, + "grad_norm": 0.0894392654299736, + "learning_rate": 4.8709067820673235e-06, + "loss": 0.09835844039916992, + "step": 720 + }, + { + "epoch": 0.11918367346938775, + "grad_norm": 0.06312194466590881, + "learning_rate": 4.866739818357489e-06, + "loss": 0.08583117127418519, + "step": 730 + }, + { + "epoch": 0.12081632653061225, + "grad_norm": 0.1375141739845276, + "learning_rate": 4.862508506629394e-06, + "loss": 0.08735222220420838, + "step": 740 + }, + { + "epoch": 0.12244897959183673, + "grad_norm": 0.12715794146060944, + "learning_rate": 4.858212961925902e-06, + "loss": 0.09324318170547485, + "step": 750 + }, + { + "epoch": 0.12408163265306123, + "grad_norm": 0.14359286427497864, + "learning_rate": 4.8538533010362735e-06, + "loss": 0.09482367634773255, + "step": 760 + }, + { + "epoch": 0.12571428571428572, + "grad_norm": 0.1228424683213234, + "learning_rate": 4.849429642492985e-06, + "loss": 0.09595230221748352, + "step": 770 + }, + { + "epoch": 0.1273469387755102, + "grad_norm": 0.08908221125602722, + "learning_rate": 4.844942106568517e-06, + "loss": 0.08924583196640015, + "step": 780 + }, + { + "epoch": 0.12897959183673469, + "grad_norm": 0.15417414903640747, + "learning_rate": 4.84039081527207e-06, + "loss": 0.09454164505004883, + "step": 790 + }, + { + "epoch": 0.1306122448979592, + "grad_norm": 0.1259608417749405, + "learning_rate": 4.835775892346264e-06, + "loss": 0.08659166097640991, + "step": 800 + }, + { + "epoch": 0.1306122448979592, + "eval_accuracy": 0.9564166666666667, + "eval_loss": 0.09297937154769897, + "eval_runtime": 157.5491, + "eval_samples_per_second": 25.389, + "eval_steps_per_second": 25.389, + "step": 800 + }, + { + "epoch": 0.13224489795918368, + "grad_norm": 0.15664513409137726, + "learning_rate": 4.8310974632637584e-06, + "loss": 0.09094155430793763, + "step": 810 + }, + { + "epoch": 0.13387755102040816, + "grad_norm": 0.12366823107004166, + "learning_rate": 4.82635565522385e-06, + "loss": 0.09823386669158936, + "step": 820 + }, + { + "epoch": 0.13551020408163264, + "grad_norm": 0.09367755055427551, + "learning_rate": 4.821550597149012e-06, + "loss": 0.09080367088317871, + "step": 830 + }, + { + "epoch": 0.13714285714285715, + "grad_norm": 0.15230220556259155, + "learning_rate": 4.816682419681389e-06, + "loss": 0.09046205282211303, + "step": 840 + }, + { + "epoch": 0.13877551020408163, + "grad_norm": 0.09214162826538086, + "learning_rate": 4.811751255179244e-06, + "loss": 0.09502801299095154, + "step": 850 + }, + { + "epoch": 0.1404081632653061, + "grad_norm": 0.09514431655406952, + "learning_rate": 4.806757237713361e-06, + "loss": 0.09261927604675294, + "step": 860 + }, + { + "epoch": 0.14204081632653062, + "grad_norm": 0.13052736222743988, + "learning_rate": 4.8017005030634e-06, + "loss": 0.08243635892868043, + "step": 870 + }, + { + "epoch": 0.1436734693877551, + "grad_norm": 0.05511025711894035, + "learning_rate": 4.796581188714205e-06, + "loss": 0.10203326940536499, + "step": 880 + }, + { + "epoch": 0.14530612244897959, + "grad_norm": 0.09896261245012283, + "learning_rate": 4.7913994338520655e-06, + "loss": 0.08544076681137085, + "step": 890 + }, + { + "epoch": 0.1469387755102041, + "grad_norm": 0.18726249039173126, + "learning_rate": 4.786155379360932e-06, + "loss": 0.0944068193435669, + "step": 900 + }, + { + "epoch": 0.14857142857142858, + "grad_norm": 0.09349548071622849, + "learning_rate": 4.7808491678185855e-06, + "loss": 0.0877715528011322, + "step": 910 + }, + { + "epoch": 0.15020408163265306, + "grad_norm": 0.13616853952407837, + "learning_rate": 4.775480943492764e-06, + "loss": 0.09196346998214722, + "step": 920 + }, + { + "epoch": 0.15183673469387754, + "grad_norm": 0.061041463166475296, + "learning_rate": 4.770050852337236e-06, + "loss": 0.09890784621238709, + "step": 930 + }, + { + "epoch": 0.15346938775510205, + "grad_norm": 0.09928455948829651, + "learning_rate": 4.764559041987836e-06, + "loss": 0.09260086417198181, + "step": 940 + }, + { + "epoch": 0.15510204081632653, + "grad_norm": 0.10996410250663757, + "learning_rate": 4.759005661758446e-06, + "loss": 0.09412651658058166, + "step": 950 + }, + { + "epoch": 0.156734693877551, + "grad_norm": 0.09731809049844742, + "learning_rate": 4.753390862636939e-06, + "loss": 0.09459404945373535, + "step": 960 + }, + { + "epoch": 0.15836734693877552, + "grad_norm": 0.12398040294647217, + "learning_rate": 4.747714797281076e-06, + "loss": 0.09960640072822571, + "step": 970 + }, + { + "epoch": 0.16, + "grad_norm": 0.14135490357875824, + "learning_rate": 4.741977620014347e-06, + "loss": 0.09338371157646179, + "step": 980 + }, + { + "epoch": 0.16163265306122448, + "grad_norm": 0.09384343028068542, + "learning_rate": 4.736179486821788e-06, + "loss": 0.08515415191650391, + "step": 990 + }, + { + "epoch": 0.16326530612244897, + "grad_norm": 0.10298261791467667, + "learning_rate": 4.730320555345728e-06, + "loss": 0.09090787768363953, + "step": 1000 + }, + { + "epoch": 0.16326530612244897, + "eval_accuracy": 0.9607916666666667, + "eval_loss": 0.08832322806119919, + "eval_runtime": 166.8275, + "eval_samples_per_second": 23.977, + "eval_steps_per_second": 23.977, + "step": 1000 + }, + { + "epoch": 0.16489795918367348, + "grad_norm": 0.05896759778261185, + "learning_rate": 4.724400984881508e-06, + "loss": 0.08005568385124207, + "step": 1010 + }, + { + "epoch": 0.16653061224489796, + "grad_norm": 0.07907038927078247, + "learning_rate": 4.7184209363731546e-06, + "loss": 0.08133251667022705, + "step": 1020 + }, + { + "epoch": 0.16816326530612244, + "grad_norm": 0.07444760203361511, + "learning_rate": 4.7123805724089935e-06, + "loss": 0.08690541982650757, + "step": 1030 + }, + { + "epoch": 0.16979591836734695, + "grad_norm": 0.16517433524131775, + "learning_rate": 4.706280057217239e-06, + "loss": 0.09946454763412475, + "step": 1040 + }, + { + "epoch": 0.17142857142857143, + "grad_norm": 0.18107500672340393, + "learning_rate": 4.700119556661521e-06, + "loss": 0.08886862993240356, + "step": 1050 + }, + { + "epoch": 0.1730612244897959, + "grad_norm": 0.15386350452899933, + "learning_rate": 4.693899238236383e-06, + "loss": 0.09117826819419861, + "step": 1060 + }, + { + "epoch": 0.17469387755102042, + "grad_norm": 0.1098460704088211, + "learning_rate": 4.687619271062725e-06, + "loss": 0.09286551475524903, + "step": 1070 + }, + { + "epoch": 0.1763265306122449, + "grad_norm": 0.09872590750455856, + "learning_rate": 4.681279825883201e-06, + "loss": 0.09102073907852173, + "step": 1080 + }, + { + "epoch": 0.17795918367346938, + "grad_norm": 0.12001907080411911, + "learning_rate": 4.674881075057581e-06, + "loss": 0.09481253027915955, + "step": 1090 + }, + { + "epoch": 0.17959183673469387, + "grad_norm": 0.13000014424324036, + "learning_rate": 4.668423192558069e-06, + "loss": 0.08999797105789184, + "step": 1100 + }, + { + "epoch": 0.18122448979591838, + "grad_norm": 0.07537361979484558, + "learning_rate": 4.661906353964563e-06, + "loss": 0.07345216870307922, + "step": 1110 + }, + { + "epoch": 0.18285714285714286, + "grad_norm": 0.10346294194459915, + "learning_rate": 4.65533073645989e-06, + "loss": 0.08096543550491334, + "step": 1120 + }, + { + "epoch": 0.18448979591836734, + "grad_norm": 0.12576258182525635, + "learning_rate": 4.6486965188249825e-06, + "loss": 0.08680164813995361, + "step": 1130 + }, + { + "epoch": 0.18612244897959185, + "grad_norm": 0.17314419150352478, + "learning_rate": 4.642003881434024e-06, + "loss": 0.08470291495323182, + "step": 1140 + }, + { + "epoch": 0.18775510204081633, + "grad_norm": 0.1576225906610489, + "learning_rate": 4.635253006249536e-06, + "loss": 0.10852066278457642, + "step": 1150 + }, + { + "epoch": 0.1893877551020408, + "grad_norm": 0.15372557938098907, + "learning_rate": 4.628444076817444e-06, + "loss": 0.1022921085357666, + "step": 1160 + }, + { + "epoch": 0.1910204081632653, + "grad_norm": 0.15021173655986786, + "learning_rate": 4.621577278262072e-06, + "loss": 0.09645735621452331, + "step": 1170 + }, + { + "epoch": 0.1926530612244898, + "grad_norm": 0.07714739441871643, + "learning_rate": 4.614652797281118e-06, + "loss": 0.08636590242385864, + "step": 1180 + }, + { + "epoch": 0.19428571428571428, + "grad_norm": 0.17956240475177765, + "learning_rate": 4.6076708221405806e-06, + "loss": 0.09293715357780456, + "step": 1190 + }, + { + "epoch": 0.19591836734693877, + "grad_norm": 0.13412988185882568, + "learning_rate": 4.600631542669633e-06, + "loss": 0.09799563884735107, + "step": 1200 + }, + { + "epoch": 0.19591836734693877, + "eval_accuracy": 0.963, + "eval_loss": 0.08563050627708435, + "eval_runtime": 171.1854, + "eval_samples_per_second": 23.366, + "eval_steps_per_second": 23.366, + "step": 1200 + }, + { + "epoch": 0.19755102040816327, + "grad_norm": 0.1682359278202057, + "learning_rate": 4.593535150255463e-06, + "loss": 0.08345703482627868, + "step": 1210 + }, + { + "epoch": 0.19918367346938776, + "grad_norm": 0.09281504899263382, + "learning_rate": 4.586381837838076e-06, + "loss": 0.08114275336265564, + "step": 1220 + }, + { + "epoch": 0.20081632653061224, + "grad_norm": 0.12355420738458633, + "learning_rate": 4.579171799905044e-06, + "loss": 0.0780484676361084, + "step": 1230 + }, + { + "epoch": 0.20244897959183675, + "grad_norm": 0.12121636420488358, + "learning_rate": 4.571905232486214e-06, + "loss": 0.07816853523254394, + "step": 1240 + }, + { + "epoch": 0.20408163265306123, + "grad_norm": 0.09326840937137604, + "learning_rate": 4.56458233314839e-06, + "loss": 0.08001847267150879, + "step": 1250 + }, + { + "epoch": 0.2057142857142857, + "grad_norm": 0.11423090845346451, + "learning_rate": 4.557203300989948e-06, + "loss": 0.08720387816429138, + "step": 1260 + }, + { + "epoch": 0.2073469387755102, + "grad_norm": 0.124808169901371, + "learning_rate": 4.549768336635432e-06, + "loss": 0.08812607526779175, + "step": 1270 + }, + { + "epoch": 0.2089795918367347, + "grad_norm": 0.08311110734939575, + "learning_rate": 4.542277642230096e-06, + "loss": 0.0880801498889923, + "step": 1280 + }, + { + "epoch": 0.21061224489795918, + "grad_norm": 0.1179559975862503, + "learning_rate": 4.53473142143441e-06, + "loss": 0.09044411182403564, + "step": 1290 + }, + { + "epoch": 0.21224489795918366, + "grad_norm": 0.13122901320457458, + "learning_rate": 4.527129879418519e-06, + "loss": 0.06792758107185363, + "step": 1300 + }, + { + "epoch": 0.21387755102040817, + "grad_norm": 0.10990814864635468, + "learning_rate": 4.519473222856667e-06, + "loss": 0.07861824035644531, + "step": 1310 + }, + { + "epoch": 0.21551020408163266, + "grad_norm": 0.14613105356693268, + "learning_rate": 4.51176165992158e-06, + "loss": 0.08266589641571045, + "step": 1320 + }, + { + "epoch": 0.21714285714285714, + "grad_norm": 0.07217947393655777, + "learning_rate": 4.503995400278801e-06, + "loss": 0.08272708654403686, + "step": 1330 + }, + { + "epoch": 0.21877551020408162, + "grad_norm": 0.12690335512161255, + "learning_rate": 4.4961746550809945e-06, + "loss": 0.08644292950630188, + "step": 1340 + }, + { + "epoch": 0.22040816326530613, + "grad_norm": 0.11895330995321274, + "learning_rate": 4.488299636962203e-06, + "loss": 0.08812489509582519, + "step": 1350 + }, + { + "epoch": 0.2220408163265306, + "grad_norm": 0.08727742731571198, + "learning_rate": 4.4803705600320655e-06, + "loss": 0.08752686977386474, + "step": 1360 + }, + { + "epoch": 0.2236734693877551, + "grad_norm": 0.1101253554224968, + "learning_rate": 4.472387639869998e-06, + "loss": 0.08365487456321716, + "step": 1370 + }, + { + "epoch": 0.2253061224489796, + "grad_norm": 0.0881471335887909, + "learning_rate": 4.464351093519332e-06, + "loss": 0.08101181387901306, + "step": 1380 + }, + { + "epoch": 0.22693877551020408, + "grad_norm": 0.1676328480243683, + "learning_rate": 4.45626113948141e-06, + "loss": 0.08873040080070496, + "step": 1390 + }, + { + "epoch": 0.22857142857142856, + "grad_norm": 0.13493730127811432, + "learning_rate": 4.44811799770965e-06, + "loss": 0.08360225558280945, + "step": 1400 + }, + { + "epoch": 0.22857142857142856, + "eval_accuracy": 0.961125, + "eval_loss": 0.0838993564248085, + "eval_runtime": 159.3888, + "eval_samples_per_second": 25.096, + "eval_steps_per_second": 25.096, + "step": 1400 + }, + { + "epoch": 0.23020408163265307, + "grad_norm": 0.18087805807590485, + "learning_rate": 4.4399218896035615e-06, + "loss": 0.07942168116569519, + "step": 1410 + }, + { + "epoch": 0.23183673469387756, + "grad_norm": 0.15980401635169983, + "learning_rate": 4.431673038002726e-06, + "loss": 0.07959137558937072, + "step": 1420 + }, + { + "epoch": 0.23346938775510204, + "grad_norm": 0.09756626188755035, + "learning_rate": 4.423371667180743e-06, + "loss": 0.08047830462455749, + "step": 1430 + }, + { + "epoch": 0.23510204081632652, + "grad_norm": 0.11922340840101242, + "learning_rate": 4.415018002839127e-06, + "loss": 0.07832802534103393, + "step": 1440 + }, + { + "epoch": 0.23673469387755103, + "grad_norm": 0.09443783015012741, + "learning_rate": 4.40661227210117e-06, + "loss": 0.07132846713066102, + "step": 1450 + }, + { + "epoch": 0.2383673469387755, + "grad_norm": 0.15397194027900696, + "learning_rate": 4.398154703505776e-06, + "loss": 0.09222683906555176, + "step": 1460 + }, + { + "epoch": 0.24, + "grad_norm": 0.06159692630171776, + "learning_rate": 4.389645527001234e-06, + "loss": 0.07937920093536377, + "step": 1470 + }, + { + "epoch": 0.2416326530612245, + "grad_norm": 0.11855911463499069, + "learning_rate": 4.381084973938976e-06, + "loss": 0.07650941014289855, + "step": 1480 + }, + { + "epoch": 0.24326530612244898, + "grad_norm": 0.1305486559867859, + "learning_rate": 4.372473277067285e-06, + "loss": 0.07730157375335693, + "step": 1490 + }, + { + "epoch": 0.24489795918367346, + "grad_norm": 0.09036920964717865, + "learning_rate": 4.363810670524962e-06, + "loss": 0.07192737460136414, + "step": 1500 + }, + { + "epoch": 0.24653061224489795, + "grad_norm": 0.0936962217092514, + "learning_rate": 4.355097389834966e-06, + "loss": 0.09040454030036926, + "step": 1510 + }, + { + "epoch": 0.24816326530612245, + "grad_norm": 0.0778910219669342, + "learning_rate": 4.346333671898007e-06, + "loss": 0.06948431730270385, + "step": 1520 + }, + { + "epoch": 0.24979591836734694, + "grad_norm": 0.0979679748415947, + "learning_rate": 4.337519754986108e-06, + "loss": 0.07317689657211304, + "step": 1530 + }, + { + "epoch": 0.25142857142857145, + "grad_norm": 0.15620048344135284, + "learning_rate": 4.328655878736123e-06, + "loss": 0.09126524925231934, + "step": 1540 + }, + { + "epoch": 0.2530612244897959, + "grad_norm": 0.09080824255943298, + "learning_rate": 4.319742284143226e-06, + "loss": 0.09065211415290833, + "step": 1550 + }, + { + "epoch": 0.2546938775510204, + "grad_norm": 0.046968940645456314, + "learning_rate": 4.310779213554353e-06, + "loss": 0.0718291699886322, + "step": 1560 + }, + { + "epoch": 0.2563265306122449, + "grad_norm": 0.11479081958532333, + "learning_rate": 4.301766910661619e-06, + "loss": 0.07913097143173217, + "step": 1570 + }, + { + "epoch": 0.25795918367346937, + "grad_norm": 0.1042279377579689, + "learning_rate": 4.292705620495689e-06, + "loss": 0.06828063130378723, + "step": 1580 + }, + { + "epoch": 0.2595918367346939, + "grad_norm": 0.17096509039402008, + "learning_rate": 4.283595589419115e-06, + "loss": 0.07042163014411926, + "step": 1590 + }, + { + "epoch": 0.2612244897959184, + "grad_norm": 0.10892639309167862, + "learning_rate": 4.274437065119643e-06, + "loss": 0.07090482115745544, + "step": 1600 + }, + { + "epoch": 0.2612244897959184, + "eval_accuracy": 0.964375, + "eval_loss": 0.08112655580043793, + "eval_runtime": 158.0164, + "eval_samples_per_second": 25.314, + "eval_steps_per_second": 25.314, + "step": 1600 + }, + { + "epoch": 0.26285714285714284, + "grad_norm": 0.09731441736221313, + "learning_rate": 4.265230296603471e-06, + "loss": 0.08803416490554809, + "step": 1610 + }, + { + "epoch": 0.26448979591836735, + "grad_norm": 0.0956069752573967, + "learning_rate": 4.255975534188488e-06, + "loss": 0.0734250545501709, + "step": 1620 + }, + { + "epoch": 0.26612244897959186, + "grad_norm": 0.14361435174942017, + "learning_rate": 4.246673029497458e-06, + "loss": 0.07703657150268554, + "step": 1630 + }, + { + "epoch": 0.2677551020408163, + "grad_norm": 0.08011971414089203, + "learning_rate": 4.237323035451188e-06, + "loss": 0.08410758376121522, + "step": 1640 + }, + { + "epoch": 0.2693877551020408, + "grad_norm": 0.1244175136089325, + "learning_rate": 4.2279258062616455e-06, + "loss": 0.0808764636516571, + "step": 1650 + }, + { + "epoch": 0.2710204081632653, + "grad_norm": 0.11162611097097397, + "learning_rate": 4.21848159742505e-06, + "loss": 0.07636536359786987, + "step": 1660 + }, + { + "epoch": 0.2726530612244898, + "grad_norm": 0.16819386184215546, + "learning_rate": 4.208990665714927e-06, + "loss": 0.07789689898490906, + "step": 1670 + }, + { + "epoch": 0.2742857142857143, + "grad_norm": 0.07522564381361008, + "learning_rate": 4.19945326917512e-06, + "loss": 0.07442367672920228, + "step": 1680 + }, + { + "epoch": 0.27591836734693875, + "grad_norm": 0.08352494984865189, + "learning_rate": 4.1898696671127855e-06, + "loss": 0.07646024227142334, + "step": 1690 + }, + { + "epoch": 0.27755102040816326, + "grad_norm": 0.12580762803554535, + "learning_rate": 4.180240120091333e-06, + "loss": 0.07484800815582275, + "step": 1700 + }, + { + "epoch": 0.27918367346938777, + "grad_norm": 0.14690619707107544, + "learning_rate": 4.170564889923348e-06, + "loss": 0.07452926635742188, + "step": 1710 + }, + { + "epoch": 0.2808163265306122, + "grad_norm": 0.08760543167591095, + "learning_rate": 4.1608442396634675e-06, + "loss": 0.07579211592674255, + "step": 1720 + }, + { + "epoch": 0.28244897959183674, + "grad_norm": 0.05263754725456238, + "learning_rate": 4.151078433601234e-06, + "loss": 0.08277950882911682, + "step": 1730 + }, + { + "epoch": 0.28408163265306124, + "grad_norm": 0.12929993867874146, + "learning_rate": 4.141267737253905e-06, + "loss": 0.07354526519775391, + "step": 1740 + }, + { + "epoch": 0.2857142857142857, + "grad_norm": 0.1444639414548874, + "learning_rate": 4.1314124173592356e-06, + "loss": 0.07377716898918152, + "step": 1750 + }, + { + "epoch": 0.2873469387755102, + "grad_norm": 0.15250442922115326, + "learning_rate": 4.121512741868229e-06, + "loss": 0.0802634596824646, + "step": 1760 + }, + { + "epoch": 0.2889795918367347, + "grad_norm": 0.14867356419563293, + "learning_rate": 4.111568979937845e-06, + "loss": 0.08141970038414001, + "step": 1770 + }, + { + "epoch": 0.29061224489795917, + "grad_norm": 0.1100233644247055, + "learning_rate": 4.10158140192369e-06, + "loss": 0.07370007634162903, + "step": 1780 + }, + { + "epoch": 0.2922448979591837, + "grad_norm": 0.10663364082574844, + "learning_rate": 4.091550279372659e-06, + "loss": 0.0754891276359558, + "step": 1790 + }, + { + "epoch": 0.2938775510204082, + "grad_norm": 0.07148909568786621, + "learning_rate": 4.081475885015558e-06, + "loss": 0.07567381858825684, + "step": 1800 + }, + { + "epoch": 0.2938775510204082, + "eval_accuracy": 0.9665416666666667, + "eval_loss": 0.07648575305938721, + "eval_runtime": 176.7261, + "eval_samples_per_second": 22.634, + "eval_steps_per_second": 22.634, + "step": 1800 + }, + { + "epoch": 0.29551020408163264, + "grad_norm": 0.07446923106908798, + "learning_rate": 4.071358492759686e-06, + "loss": 0.06706757545471191, + "step": 1810 + }, + { + "epoch": 0.29714285714285715, + "grad_norm": 0.144475519657135, + "learning_rate": 4.061198377681391e-06, + "loss": 0.07407833337783813, + "step": 1820 + }, + { + "epoch": 0.2987755102040816, + "grad_norm": 0.16540028154850006, + "learning_rate": 4.050995816018585e-06, + "loss": 0.07593105435371399, + "step": 1830 + }, + { + "epoch": 0.3004081632653061, + "grad_norm": 0.09237829595804214, + "learning_rate": 4.040751085163239e-06, + "loss": 0.08375502228736878, + "step": 1840 + }, + { + "epoch": 0.3020408163265306, + "grad_norm": 0.11202957481145859, + "learning_rate": 4.0304644636538435e-06, + "loss": 0.0692280650138855, + "step": 1850 + }, + { + "epoch": 0.3036734693877551, + "grad_norm": 0.16075952351093292, + "learning_rate": 4.020136231167824e-06, + "loss": 0.07824563384056091, + "step": 1860 + }, + { + "epoch": 0.3053061224489796, + "grad_norm": 0.12567447125911713, + "learning_rate": 4.0097666685139516e-06, + "loss": 0.07255304455757142, + "step": 1870 + }, + { + "epoch": 0.3069387755102041, + "grad_norm": 0.08912669867277145, + "learning_rate": 3.999356057624696e-06, + "loss": 0.06304961442947388, + "step": 1880 + }, + { + "epoch": 0.30857142857142855, + "grad_norm": 0.15736177563667297, + "learning_rate": 3.988904681548568e-06, + "loss": 0.07315694689750671, + "step": 1890 + }, + { + "epoch": 0.31020408163265306, + "grad_norm": 0.13224704563617706, + "learning_rate": 3.978412824442421e-06, + "loss": 0.07607222199440003, + "step": 1900 + }, + { + "epoch": 0.31183673469387757, + "grad_norm": 0.0942244678735733, + "learning_rate": 3.9678807715637245e-06, + "loss": 0.06320942640304565, + "step": 1910 + }, + { + "epoch": 0.313469387755102, + "grad_norm": 0.1153193786740303, + "learning_rate": 3.95730880926281e-06, + "loss": 0.07607839703559875, + "step": 1920 + }, + { + "epoch": 0.31510204081632653, + "grad_norm": 0.24464164674282074, + "learning_rate": 3.946697224975083e-06, + "loss": 0.07458690404891968, + "step": 1930 + }, + { + "epoch": 0.31673469387755104, + "grad_norm": 0.14584548771381378, + "learning_rate": 3.936046307213215e-06, + "loss": 0.07876120805740357, + "step": 1940 + }, + { + "epoch": 0.3183673469387755, + "grad_norm": 0.0880698636174202, + "learning_rate": 3.925356345559288e-06, + "loss": 0.07082821726799012, + "step": 1950 + }, + { + "epoch": 0.32, + "grad_norm": 0.08619305491447449, + "learning_rate": 3.914627630656931e-06, + "loss": 0.06648402214050293, + "step": 1960 + }, + { + "epoch": 0.3216326530612245, + "grad_norm": 0.1341329663991928, + "learning_rate": 3.903860454203413e-06, + "loss": 0.07000230550765991, + "step": 1970 + }, + { + "epoch": 0.32326530612244897, + "grad_norm": 0.1504107564687729, + "learning_rate": 3.8930551089417145e-06, + "loss": 0.06887786984443664, + "step": 1980 + }, + { + "epoch": 0.3248979591836735, + "grad_norm": 0.1262267827987671, + "learning_rate": 3.882211888652566e-06, + "loss": 0.08991954326629639, + "step": 1990 + }, + { + "epoch": 0.32653061224489793, + "grad_norm": 0.2052707076072693, + "learning_rate": 3.871331088146463e-06, + "loss": 0.07011507153511047, + "step": 2000 + }, + { + "epoch": 0.32653061224489793, + "eval_accuracy": 0.9654583333333334, + "eval_loss": 0.07913381606340408, + "eval_runtime": 161.4221, + "eval_samples_per_second": 24.78, + "eval_steps_per_second": 24.78, + "step": 2000 + }, + { + "epoch": 0.32816326530612244, + "grad_norm": 0.11643362790346146, + "learning_rate": 3.860413003255648e-06, + "loss": 0.08266008496284485, + "step": 2010 + }, + { + "epoch": 0.32979591836734695, + "grad_norm": 0.1472826600074768, + "learning_rate": 3.8494579308260695e-06, + "loss": 0.07043538689613342, + "step": 2020 + }, + { + "epoch": 0.3314285714285714, + "grad_norm": 0.10538945347070694, + "learning_rate": 3.838466168709312e-06, + "loss": 0.06780247688293457, + "step": 2030 + }, + { + "epoch": 0.3330612244897959, + "grad_norm": 0.14366035163402557, + "learning_rate": 3.827438015754491e-06, + "loss": 0.06872215867042542, + "step": 2040 + }, + { + "epoch": 0.3346938775510204, + "grad_norm": 0.11022552847862244, + "learning_rate": 3.81637377180014e-06, + "loss": 0.06861716508865356, + "step": 2050 + }, + { + "epoch": 0.3363265306122449, + "grad_norm": 0.07990925014019012, + "learning_rate": 3.8052737376660446e-06, + "loss": 0.07096996307373046, + "step": 2060 + }, + { + "epoch": 0.3379591836734694, + "grad_norm": 0.11924420297145844, + "learning_rate": 3.7941382151450756e-06, + "loss": 0.07092551589012146, + "step": 2070 + }, + { + "epoch": 0.3395918367346939, + "grad_norm": 0.08096475899219513, + "learning_rate": 3.782967506994975e-06, + "loss": 0.06780472993850709, + "step": 2080 + }, + { + "epoch": 0.34122448979591835, + "grad_norm": 0.14714907109737396, + "learning_rate": 3.7717619169301302e-06, + "loss": 0.07661795020103454, + "step": 2090 + }, + { + "epoch": 0.34285714285714286, + "grad_norm": 0.14839571714401245, + "learning_rate": 3.7605217496133127e-06, + "loss": 0.07939797639846802, + "step": 2100 + }, + { + "epoch": 0.34448979591836737, + "grad_norm": 0.15081271529197693, + "learning_rate": 3.749247310647398e-06, + "loss": 0.07869908213615417, + "step": 2110 + }, + { + "epoch": 0.3461224489795918, + "grad_norm": 0.13346275687217712, + "learning_rate": 3.737938906567053e-06, + "loss": 0.07526676058769226, + "step": 2120 + }, + { + "epoch": 0.34775510204081633, + "grad_norm": 0.112598717212677, + "learning_rate": 3.7265968448304057e-06, + "loss": 0.07366414070129394, + "step": 2130 + }, + { + "epoch": 0.34938775510204084, + "grad_norm": 0.12423531711101532, + "learning_rate": 3.7152214338106827e-06, + "loss": 0.06439908146858216, + "step": 2140 + }, + { + "epoch": 0.3510204081632653, + "grad_norm": 0.21758921444416046, + "learning_rate": 3.703812982787828e-06, + "loss": 0.08098108768463134, + "step": 2150 + }, + { + "epoch": 0.3526530612244898, + "grad_norm": 0.1298121213912964, + "learning_rate": 3.6923718019400906e-06, + "loss": 0.08134381771087647, + "step": 2160 + }, + { + "epoch": 0.35428571428571426, + "grad_norm": 0.09018377214670181, + "learning_rate": 3.6808982023355934e-06, + "loss": 0.07070590257644653, + "step": 2170 + }, + { + "epoch": 0.35591836734693877, + "grad_norm": 0.1329157054424286, + "learning_rate": 3.6693924959238774e-06, + "loss": 0.06200396418571472, + "step": 2180 + }, + { + "epoch": 0.3575510204081633, + "grad_norm": 0.08813945204019547, + "learning_rate": 3.657854995527416e-06, + "loss": 0.06886230111122131, + "step": 2190 + }, + { + "epoch": 0.35918367346938773, + "grad_norm": 0.13459829986095428, + "learning_rate": 3.646286014833114e-06, + "loss": 0.07170915603637695, + "step": 2200 + }, + { + "epoch": 0.35918367346938773, + "eval_accuracy": 0.970625, + "eval_loss": 0.06810569018125534, + "eval_runtime": 157.4227, + "eval_samples_per_second": 25.409, + "eval_steps_per_second": 25.409, + "step": 2200 + }, + { + "epoch": 0.36081632653061224, + "grad_norm": 0.13405098021030426, + "learning_rate": 3.634685868383776e-06, + "loss": 0.07438129186630249, + "step": 2210 + }, + { + "epoch": 0.36244897959183675, + "grad_norm": 0.1342102289199829, + "learning_rate": 3.6230548715695555e-06, + "loss": 0.07625046968460084, + "step": 2220 + }, + { + "epoch": 0.3640816326530612, + "grad_norm": 0.1532502919435501, + "learning_rate": 3.611393340619381e-06, + "loss": 0.06719872951507569, + "step": 2230 + }, + { + "epoch": 0.3657142857142857, + "grad_norm": 0.11218152195215225, + "learning_rate": 3.5997015925923563e-06, + "loss": 0.07131683230400085, + "step": 2240 + }, + { + "epoch": 0.3673469387755102, + "grad_norm": 0.09200239181518555, + "learning_rate": 3.587979945369141e-06, + "loss": 0.06948224902153015, + "step": 2250 + }, + { + "epoch": 0.3689795918367347, + "grad_norm": 0.08099328726530075, + "learning_rate": 3.5762287176433097e-06, + "loss": 0.07263018488883972, + "step": 2260 + }, + { + "epoch": 0.3706122448979592, + "grad_norm": 0.09097494184970856, + "learning_rate": 3.564448228912682e-06, + "loss": 0.06066258549690247, + "step": 2270 + }, + { + "epoch": 0.3722448979591837, + "grad_norm": 0.07033897191286087, + "learning_rate": 3.5526387994706428e-06, + "loss": 0.07041062712669373, + "step": 2280 + }, + { + "epoch": 0.37387755102040815, + "grad_norm": 0.07659374922513962, + "learning_rate": 3.5408007503974283e-06, + "loss": 0.07272242307662964, + "step": 2290 + }, + { + "epoch": 0.37551020408163266, + "grad_norm": 0.0836353451013565, + "learning_rate": 3.5289344035513996e-06, + "loss": 0.07268807291984558, + "step": 2300 + }, + { + "epoch": 0.37714285714285717, + "grad_norm": 0.1426168978214264, + "learning_rate": 3.5170400815602894e-06, + "loss": 0.05777261257171631, + "step": 2310 + }, + { + "epoch": 0.3787755102040816, + "grad_norm": 0.07479120045900345, + "learning_rate": 3.5051181078124336e-06, + "loss": 0.06174240708351135, + "step": 2320 + }, + { + "epoch": 0.38040816326530613, + "grad_norm": 0.14598114788532257, + "learning_rate": 3.4931688064479745e-06, + "loss": 0.06616324782371522, + "step": 2330 + }, + { + "epoch": 0.3820408163265306, + "grad_norm": 0.22399863600730896, + "learning_rate": 3.4811925023500524e-06, + "loss": 0.06692227125167846, + "step": 2340 + }, + { + "epoch": 0.3836734693877551, + "grad_norm": 0.11600285023450851, + "learning_rate": 3.4691895211359693e-06, + "loss": 0.0634126603603363, + "step": 2350 + }, + { + "epoch": 0.3853061224489796, + "grad_norm": 0.0577060841023922, + "learning_rate": 3.457160189148337e-06, + "loss": 0.06884999275207519, + "step": 2360 + }, + { + "epoch": 0.38693877551020406, + "grad_norm": 0.09160040318965912, + "learning_rate": 3.445104833446205e-06, + "loss": 0.05986430644989014, + "step": 2370 + }, + { + "epoch": 0.38857142857142857, + "grad_norm": 0.1284518539905548, + "learning_rate": 3.433023781796167e-06, + "loss": 0.06759394407272339, + "step": 2380 + }, + { + "epoch": 0.3902040816326531, + "grad_norm": 0.08572858572006226, + "learning_rate": 3.4209173626634494e-06, + "loss": 0.06813116669654846, + "step": 2390 + }, + { + "epoch": 0.39183673469387753, + "grad_norm": 0.11550725251436234, + "learning_rate": 3.4087859052029837e-06, + "loss": 0.06371633410453796, + "step": 2400 + }, + { + "epoch": 0.39183673469387753, + "eval_accuracy": 0.9685833333333334, + "eval_loss": 0.07035119831562042, + "eval_runtime": 156.9729, + "eval_samples_per_second": 25.482, + "eval_steps_per_second": 25.482, + "step": 2400 + }, + { + "epoch": 0.39346938775510204, + "grad_norm": 0.22328943014144897, + "learning_rate": 3.3966297392504535e-06, + "loss": 0.06804299950599671, + "step": 2410 + }, + { + "epoch": 0.39510204081632655, + "grad_norm": 0.18518126010894775, + "learning_rate": 3.3844491953133272e-06, + "loss": 0.07678125500679016, + "step": 2420 + }, + { + "epoch": 0.396734693877551, + "grad_norm": 0.09818598628044128, + "learning_rate": 3.372244604561875e-06, + "loss": 0.06537411212921143, + "step": 2430 + }, + { + "epoch": 0.3983673469387755, + "grad_norm": 0.08255010843276978, + "learning_rate": 3.360016298820163e-06, + "loss": 0.05702770948410034, + "step": 2440 + }, + { + "epoch": 0.4, + "grad_norm": 0.10319415479898453, + "learning_rate": 3.3477646105570293e-06, + "loss": 0.06334798336029053, + "step": 2450 + }, + { + "epoch": 0.4016326530612245, + "grad_norm": 0.06523872911930084, + "learning_rate": 3.3354898728770503e-06, + "loss": 0.051438117027282716, + "step": 2460 + }, + { + "epoch": 0.403265306122449, + "grad_norm": 0.15883173048496246, + "learning_rate": 3.3231924195114774e-06, + "loss": 0.06427289247512817, + "step": 2470 + }, + { + "epoch": 0.4048979591836735, + "grad_norm": 0.1734374463558197, + "learning_rate": 3.3108725848091685e-06, + "loss": 0.06652372479438781, + "step": 2480 + }, + { + "epoch": 0.40653061224489795, + "grad_norm": 0.10557752102613449, + "learning_rate": 3.2985307037274937e-06, + "loss": 0.07374808788299561, + "step": 2490 + }, + { + "epoch": 0.40816326530612246, + "grad_norm": 0.12942200899124146, + "learning_rate": 3.286167111823234e-06, + "loss": 0.06349012851715088, + "step": 2500 + }, + { + "epoch": 0.4097959183673469, + "grad_norm": 0.1858324259519577, + "learning_rate": 3.2737821452434502e-06, + "loss": 0.06953533291816712, + "step": 2510 + }, + { + "epoch": 0.4114285714285714, + "grad_norm": 0.10682573169469833, + "learning_rate": 3.26137614071635e-06, + "loss": 0.05826004147529602, + "step": 2520 + }, + { + "epoch": 0.41306122448979593, + "grad_norm": 0.12533792853355408, + "learning_rate": 3.2489494355421293e-06, + "loss": 0.05837242007255554, + "step": 2530 + }, + { + "epoch": 0.4146938775510204, + "grad_norm": 0.10952792316675186, + "learning_rate": 3.2365023675838027e-06, + "loss": 0.06359312534332276, + "step": 2540 + }, + { + "epoch": 0.4163265306122449, + "grad_norm": 0.08605755865573883, + "learning_rate": 3.22403527525802e-06, + "loss": 0.07361164093017578, + "step": 2550 + }, + { + "epoch": 0.4179591836734694, + "grad_norm": 0.15942610800266266, + "learning_rate": 3.211548497525861e-06, + "loss": 0.05989238619804382, + "step": 2560 + }, + { + "epoch": 0.41959183673469386, + "grad_norm": 0.16340239346027374, + "learning_rate": 3.1990423738836206e-06, + "loss": 0.07593320608139038, + "step": 2570 + }, + { + "epoch": 0.42122448979591837, + "grad_norm": 0.11011330783367157, + "learning_rate": 3.1865172443535797e-06, + "loss": 0.06850495338439941, + "step": 2580 + }, + { + "epoch": 0.4228571428571429, + "grad_norm": 0.10529608279466629, + "learning_rate": 3.173973449474762e-06, + "loss": 0.056969082355499266, + "step": 2590 + }, + { + "epoch": 0.42448979591836733, + "grad_norm": 0.12615667283535004, + "learning_rate": 3.1614113302936716e-06, + "loss": 0.06189748048782349, + "step": 2600 + }, + { + "epoch": 0.42448979591836733, + "eval_accuracy": 0.9746666666666668, + "eval_loss": 0.0618240088224411, + "eval_runtime": 157.0858, + "eval_samples_per_second": 25.464, + "eval_steps_per_second": 25.464, + "step": 2600 + }, + { + "epoch": 0.42612244897959184, + "grad_norm": 0.1343226134777069, + "learning_rate": 3.1488312283550216e-06, + "loss": 0.061647194623947146, + "step": 2610 + }, + { + "epoch": 0.42775510204081635, + "grad_norm": 0.11777225136756897, + "learning_rate": 3.136233485692451e-06, + "loss": 0.05916886925697327, + "step": 2620 + }, + { + "epoch": 0.4293877551020408, + "grad_norm": 0.14721977710723877, + "learning_rate": 3.12361844481922e-06, + "loss": 0.059158307313919065, + "step": 2630 + }, + { + "epoch": 0.4310204081632653, + "grad_norm": 0.12912684679031372, + "learning_rate": 3.110986448718902e-06, + "loss": 0.05914759635925293, + "step": 2640 + }, + { + "epoch": 0.4326530612244898, + "grad_norm": 0.13003695011138916, + "learning_rate": 3.0983378408360597e-06, + "loss": 0.059078723192214966, + "step": 2650 + }, + { + "epoch": 0.4342857142857143, + "grad_norm": 0.12714017927646637, + "learning_rate": 3.085672965066899e-06, + "loss": 0.05707222819328308, + "step": 2660 + }, + { + "epoch": 0.4359183673469388, + "grad_norm": 0.13708427548408508, + "learning_rate": 3.0729921657499294e-06, + "loss": 0.05138048529624939, + "step": 2670 + }, + { + "epoch": 0.43755102040816324, + "grad_norm": 0.0949573740363121, + "learning_rate": 3.0602957876565943e-06, + "loss": 0.050283265113830564, + "step": 2680 + }, + { + "epoch": 0.43918367346938775, + "grad_norm": 0.08723263442516327, + "learning_rate": 3.0475841759819023e-06, + "loss": 0.06342648863792419, + "step": 2690 + }, + { + "epoch": 0.44081632653061226, + "grad_norm": 0.09944969415664673, + "learning_rate": 3.034857676335038e-06, + "loss": 0.06127093434333801, + "step": 2700 + }, + { + "epoch": 0.4424489795918367, + "grad_norm": 0.10576151311397552, + "learning_rate": 3.0221166347299673e-06, + "loss": 0.05412284731864929, + "step": 2710 + }, + { + "epoch": 0.4440816326530612, + "grad_norm": 0.1598891168832779, + "learning_rate": 3.009361397576031e-06, + "loss": 0.05579071640968323, + "step": 2720 + }, + { + "epoch": 0.44571428571428573, + "grad_norm": 0.16953758895397186, + "learning_rate": 2.9965923116685224e-06, + "loss": 0.06472888588905334, + "step": 2730 + }, + { + "epoch": 0.4473469387755102, + "grad_norm": 0.10883133113384247, + "learning_rate": 2.9838097241792646e-06, + "loss": 0.06263615489006043, + "step": 2740 + }, + { + "epoch": 0.4489795918367347, + "grad_norm": 0.08914396911859512, + "learning_rate": 2.9710139826471646e-06, + "loss": 0.0624904453754425, + "step": 2750 + }, + { + "epoch": 0.4506122448979592, + "grad_norm": 0.1059069111943245, + "learning_rate": 2.9582054349687706e-06, + "loss": 0.05488716959953308, + "step": 2760 + }, + { + "epoch": 0.45224489795918366, + "grad_norm": 0.09883642196655273, + "learning_rate": 2.9453844293888067e-06, + "loss": 0.06263139843940735, + "step": 2770 + }, + { + "epoch": 0.45387755102040817, + "grad_norm": 0.11156519502401352, + "learning_rate": 2.932551314490712e-06, + "loss": 0.06324111819267272, + "step": 2780 + }, + { + "epoch": 0.4555102040816327, + "grad_norm": 0.08826332539319992, + "learning_rate": 2.9197064391871565e-06, + "loss": 0.06384775042533875, + "step": 2790 + }, + { + "epoch": 0.45714285714285713, + "grad_norm": 0.0698089748620987, + "learning_rate": 2.9068501527105595e-06, + "loss": 0.05722553730010986, + "step": 2800 + }, + { + "epoch": 0.45714285714285713, + "eval_accuracy": 0.977375, + "eval_loss": 0.05622806400060654, + "eval_runtime": 170.7758, + "eval_samples_per_second": 23.423, + "eval_steps_per_second": 23.423, + "step": 2800 + }, + { + "epoch": 0.45877551020408164, + "grad_norm": 0.07031046599149704, + "learning_rate": 2.893982804603592e-06, + "loss": 0.060957270860672, + "step": 2810 + }, + { + "epoch": 0.46040816326530615, + "grad_norm": 0.08772005140781403, + "learning_rate": 2.8811047447096714e-06, + "loss": 0.05667365789413452, + "step": 2820 + }, + { + "epoch": 0.4620408163265306, + "grad_norm": 0.10329847782850266, + "learning_rate": 2.8682163231634557e-06, + "loss": 0.06144403219223023, + "step": 2830 + }, + { + "epoch": 0.4636734693877551, + "grad_norm": 0.06306952983140945, + "learning_rate": 2.8553178903813193e-06, + "loss": 0.04665087461471558, + "step": 2840 + }, + { + "epoch": 0.46530612244897956, + "grad_norm": 0.11146903783082962, + "learning_rate": 2.842409797051825e-06, + "loss": 0.06363347172737122, + "step": 2850 + }, + { + "epoch": 0.4669387755102041, + "grad_norm": 0.04504350945353508, + "learning_rate": 2.829492394126192e-06, + "loss": 0.05332474112510681, + "step": 2860 + }, + { + "epoch": 0.4685714285714286, + "grad_norm": 0.06039620563387871, + "learning_rate": 2.8165660328087546e-06, + "loss": 0.07094461917877197, + "step": 2870 + }, + { + "epoch": 0.47020408163265304, + "grad_norm": 0.08084474503993988, + "learning_rate": 2.8036310645474096e-06, + "loss": 0.058167701959609984, + "step": 2880 + }, + { + "epoch": 0.47183673469387755, + "grad_norm": 0.05352522060275078, + "learning_rate": 2.7906878410240655e-06, + "loss": 0.0486674964427948, + "step": 2890 + }, + { + "epoch": 0.47346938775510206, + "grad_norm": 0.09431054443120956, + "learning_rate": 2.777736714145077e-06, + "loss": 0.062025409936904904, + "step": 2900 + }, + { + "epoch": 0.4751020408163265, + "grad_norm": 0.09863542020320892, + "learning_rate": 2.7647780360316787e-06, + "loss": 0.061118441820144656, + "step": 2910 + }, + { + "epoch": 0.476734693877551, + "grad_norm": 0.082881860435009, + "learning_rate": 2.7518121590104145e-06, + "loss": 0.061808472871780394, + "step": 2920 + }, + { + "epoch": 0.47836734693877553, + "grad_norm": 0.06983684748411179, + "learning_rate": 2.7388394356035524e-06, + "loss": 0.05849882960319519, + "step": 2930 + }, + { + "epoch": 0.48, + "grad_norm": 0.10832090675830841, + "learning_rate": 2.7258602185195043e-06, + "loss": 0.06612529158592224, + "step": 2940 + }, + { + "epoch": 0.4816326530612245, + "grad_norm": 0.1516750454902649, + "learning_rate": 2.7128748606432353e-06, + "loss": 0.058327901363372806, + "step": 2950 + }, + { + "epoch": 0.483265306122449, + "grad_norm": 0.14055661857128143, + "learning_rate": 2.699883715026669e-06, + "loss": 0.055737996101379396, + "step": 2960 + }, + { + "epoch": 0.48489795918367345, + "grad_norm": 0.0793638750910759, + "learning_rate": 2.686887134879089e-06, + "loss": 0.051851028203964235, + "step": 2970 + }, + { + "epoch": 0.48653061224489796, + "grad_norm": 0.07475428283214569, + "learning_rate": 2.6738854735575358e-06, + "loss": 0.06357839703559875, + "step": 2980 + }, + { + "epoch": 0.4881632653061225, + "grad_norm": 0.13335047662258148, + "learning_rate": 2.660879084557198e-06, + "loss": 0.06160399913787842, + "step": 2990 + }, + { + "epoch": 0.4897959183673469, + "grad_norm": 0.12583893537521362, + "learning_rate": 2.6478683215018025e-06, + "loss": 0.060502588748931885, + "step": 3000 + }, + { + "epoch": 0.4897959183673469, + "eval_accuracy": 0.9767083333333334, + "eval_loss": 0.0561000294983387, + "eval_runtime": 156.4202, + "eval_samples_per_second": 25.572, + "eval_steps_per_second": 25.572, + "step": 3000 + }, + { + "epoch": 0.49142857142857144, + "grad_norm": 0.1411438137292862, + "learning_rate": 2.6348535381340025e-06, + "loss": 0.060459482669830325, + "step": 3010 + }, + { + "epoch": 0.4930612244897959, + "grad_norm": 0.12584605813026428, + "learning_rate": 2.6218350883057546e-06, + "loss": 0.0632048487663269, + "step": 3020 + }, + { + "epoch": 0.4946938775510204, + "grad_norm": 0.10522248595952988, + "learning_rate": 2.608813325968702e-06, + "loss": 0.04655081033706665, + "step": 3030 + }, + { + "epoch": 0.4963265306122449, + "grad_norm": 0.11343304812908173, + "learning_rate": 2.59578860516455e-06, + "loss": 0.05672039389610291, + "step": 3040 + }, + { + "epoch": 0.49795918367346936, + "grad_norm": 0.12666337192058563, + "learning_rate": 2.5827612800154382e-06, + "loss": 0.047120341658592226, + "step": 3050 + }, + { + "epoch": 0.4995918367346939, + "grad_norm": 0.08930598944425583, + "learning_rate": 2.5697317047143177e-06, + "loss": 0.06063341498374939, + "step": 3060 + }, + { + "epoch": 0.5012244897959184, + "grad_norm": 0.07818852365016937, + "learning_rate": 2.556700233515316e-06, + "loss": 0.05240688323974609, + "step": 3070 + }, + { + "epoch": 0.5028571428571429, + "grad_norm": 0.1425711065530777, + "learning_rate": 2.5436672207241064e-06, + "loss": 0.053180736303329465, + "step": 3080 + }, + { + "epoch": 0.5044897959183674, + "grad_norm": 0.1457373946905136, + "learning_rate": 2.5306330206882764e-06, + "loss": 0.06092001795768738, + "step": 3090 + }, + { + "epoch": 0.5061224489795918, + "grad_norm": 0.12738314270973206, + "learning_rate": 2.5175979877876935e-06, + "loss": 0.06613186597824097, + "step": 3100 + }, + { + "epoch": 0.5077551020408163, + "grad_norm": 0.09169036149978638, + "learning_rate": 2.5045624764248687e-06, + "loss": 0.06115126609802246, + "step": 3110 + }, + { + "epoch": 0.5093877551020408, + "grad_norm": 0.09637737274169922, + "learning_rate": 2.491526841015321e-06, + "loss": 0.044120225310325625, + "step": 3120 + }, + { + "epoch": 0.5110204081632653, + "grad_norm": 0.15029804408550262, + "learning_rate": 2.478491435977945e-06, + "loss": 0.07098687291145325, + "step": 3130 + }, + { + "epoch": 0.5126530612244898, + "grad_norm": 0.1228741705417633, + "learning_rate": 2.4654566157253683e-06, + "loss": 0.05738852024078369, + "step": 3140 + }, + { + "epoch": 0.5142857142857142, + "grad_norm": 0.09150952100753784, + "learning_rate": 2.452422734654321e-06, + "loss": 0.053479206562042234, + "step": 3150 + }, + { + "epoch": 0.5159183673469387, + "grad_norm": 0.14547918736934662, + "learning_rate": 2.4393901471359977e-06, + "loss": 0.056794995069503786, + "step": 3160 + }, + { + "epoch": 0.5175510204081633, + "grad_norm": 0.10099287331104279, + "learning_rate": 2.4263592075064238e-06, + "loss": 0.05802353620529175, + "step": 3170 + }, + { + "epoch": 0.5191836734693878, + "grad_norm": 0.10859636962413788, + "learning_rate": 2.41333027005682e-06, + "loss": 0.05282890200614929, + "step": 3180 + }, + { + "epoch": 0.5208163265306123, + "grad_norm": 0.12420635670423508, + "learning_rate": 2.400303689023972e-06, + "loss": 0.05190218687057495, + "step": 3190 + }, + { + "epoch": 0.5224489795918368, + "grad_norm": 0.1902378350496292, + "learning_rate": 2.3872798185805986e-06, + "loss": 0.06482383608818054, + "step": 3200 + }, + { + "epoch": 0.5224489795918368, + "eval_accuracy": 0.9744583333333335, + "eval_loss": 0.058392543345689774, + "eval_runtime": 156.8655, + "eval_samples_per_second": 25.5, + "eval_steps_per_second": 25.5, + "step": 3200 + }, + { + "epoch": 0.5240816326530612, + "grad_norm": 0.15229448676109314, + "learning_rate": 2.37425901282572e-06, + "loss": 0.06062687039375305, + "step": 3210 + }, + { + "epoch": 0.5257142857142857, + "grad_norm": 0.08411400765180588, + "learning_rate": 2.3612416257750335e-06, + "loss": 0.06494970321655273, + "step": 3220 + }, + { + "epoch": 0.5273469387755102, + "grad_norm": 0.114021435379982, + "learning_rate": 2.348228011351287e-06, + "loss": 0.05140092372894287, + "step": 3230 + }, + { + "epoch": 0.5289795918367347, + "grad_norm": 0.08826510608196259, + "learning_rate": 2.335218523374656e-06, + "loss": 0.041680914163589475, + "step": 3240 + }, + { + "epoch": 0.5306122448979592, + "grad_norm": 0.10536732524633408, + "learning_rate": 2.3222135155531248e-06, + "loss": 0.04369679093360901, + "step": 3250 + }, + { + "epoch": 0.5322448979591837, + "grad_norm": 0.16883279383182526, + "learning_rate": 2.309213341472868e-06, + "loss": 0.050543594360351565, + "step": 3260 + }, + { + "epoch": 0.5338775510204081, + "grad_norm": 0.15277455747127533, + "learning_rate": 2.29621835458864e-06, + "loss": 0.05258342027664185, + "step": 3270 + }, + { + "epoch": 0.5355102040816326, + "grad_norm": 0.07162481546401978, + "learning_rate": 2.283228908214163e-06, + "loss": 0.055678969621658324, + "step": 3280 + }, + { + "epoch": 0.5371428571428571, + "grad_norm": 0.11578263342380524, + "learning_rate": 2.2702453555125182e-06, + "loss": 0.048620349168777464, + "step": 3290 + }, + { + "epoch": 0.5387755102040817, + "grad_norm": 0.21185705065727234, + "learning_rate": 2.2572680494865527e-06, + "loss": 0.046547812223434445, + "step": 3300 + }, + { + "epoch": 0.5404081632653062, + "grad_norm": 0.19949115812778473, + "learning_rate": 2.2442973429692705e-06, + "loss": 0.04967870116233826, + "step": 3310 + }, + { + "epoch": 0.5420408163265306, + "grad_norm": 0.10855370759963989, + "learning_rate": 2.231333588614248e-06, + "loss": 0.0508593499660492, + "step": 3320 + }, + { + "epoch": 0.5436734693877551, + "grad_norm": 0.11827342212200165, + "learning_rate": 2.2183771388860427e-06, + "loss": 0.04854269027709961, + "step": 3330 + }, + { + "epoch": 0.5453061224489796, + "grad_norm": 0.1081443578004837, + "learning_rate": 2.2054283460506096e-06, + "loss": 0.05934192538261414, + "step": 3340 + }, + { + "epoch": 0.5469387755102041, + "grad_norm": 0.14441704750061035, + "learning_rate": 2.1924875621657256e-06, + "loss": 0.05332346558570862, + "step": 3350 + }, + { + "epoch": 0.5485714285714286, + "grad_norm": 0.11531820893287659, + "learning_rate": 2.179555139071415e-06, + "loss": 0.04938814043998718, + "step": 3360 + }, + { + "epoch": 0.5502040816326531, + "grad_norm": 0.08365523815155029, + "learning_rate": 2.166631428380387e-06, + "loss": 0.05734975934028626, + "step": 3370 + }, + { + "epoch": 0.5518367346938775, + "grad_norm": 0.09165070950984955, + "learning_rate": 2.1537167814684728e-06, + "loss": 0.047475004196166994, + "step": 3380 + }, + { + "epoch": 0.553469387755102, + "grad_norm": 0.13237863779067993, + "learning_rate": 2.140811549465074e-06, + "loss": 0.062284767627716064, + "step": 3390 + }, + { + "epoch": 0.5551020408163265, + "grad_norm": 0.1698858141899109, + "learning_rate": 2.127916083243615e-06, + "loss": 0.06585791110992431, + "step": 3400 + }, + { + "epoch": 0.5551020408163265, + "eval_accuracy": 0.9770833333333334, + "eval_loss": 0.05396538972854614, + "eval_runtime": 156.044, + "eval_samples_per_second": 25.634, + "eval_steps_per_second": 25.634, + "step": 3400 + }, + { + "epoch": 0.556734693877551, + "grad_norm": 0.1104230135679245, + "learning_rate": 2.1150307334120033e-06, + "loss": 0.04635375440120697, + "step": 3410 + }, + { + "epoch": 0.5583673469387755, + "grad_norm": 0.21372850239276886, + "learning_rate": 2.1021558503030973e-06, + "loss": 0.05940789580345154, + "step": 3420 + }, + { + "epoch": 0.56, + "grad_norm": 0.2013176679611206, + "learning_rate": 2.0892917839651818e-06, + "loss": 0.047165653109550475, + "step": 3430 + }, + { + "epoch": 0.5616326530612245, + "grad_norm": 0.09189072996377945, + "learning_rate": 2.076438884152449e-06, + "loss": 0.04550817906856537, + "step": 3440 + }, + { + "epoch": 0.563265306122449, + "grad_norm": 0.10521821677684784, + "learning_rate": 2.0635975003154934e-06, + "loss": 0.045033201575279236, + "step": 3450 + }, + { + "epoch": 0.5648979591836735, + "grad_norm": 0.10311131924390793, + "learning_rate": 2.0507679815918043e-06, + "loss": 0.04502847194671631, + "step": 3460 + }, + { + "epoch": 0.566530612244898, + "grad_norm": 0.07487176358699799, + "learning_rate": 2.0379506767962788e-06, + "loss": 0.04697479605674744, + "step": 3470 + }, + { + "epoch": 0.5681632653061225, + "grad_norm": 0.12441752105951309, + "learning_rate": 2.0251459344117345e-06, + "loss": 0.052119261026382445, + "step": 3480 + }, + { + "epoch": 0.5697959183673469, + "grad_norm": 0.12823422253131866, + "learning_rate": 2.0123541025794397e-06, + "loss": 0.05213663578033447, + "step": 3490 + }, + { + "epoch": 0.5714285714285714, + "grad_norm": 0.11340193450450897, + "learning_rate": 1.9995755290896414e-06, + "loss": 0.05319364666938782, + "step": 3500 + }, + { + "epoch": 0.5730612244897959, + "grad_norm": 0.10469839721918106, + "learning_rate": 1.986810561372113e-06, + "loss": 0.046560952067375184, + "step": 3510 + }, + { + "epoch": 0.5746938775510204, + "grad_norm": 0.07797456532716751, + "learning_rate": 1.9740595464867092e-06, + "loss": 0.05037369728088379, + "step": 3520 + }, + { + "epoch": 0.5763265306122449, + "grad_norm": 0.13010293245315552, + "learning_rate": 1.961322831113927e-06, + "loss": 0.0543134331703186, + "step": 3530 + }, + { + "epoch": 0.5779591836734694, + "grad_norm": 0.13981281220912933, + "learning_rate": 1.9486007615454837e-06, + "loss": 0.052860957384109494, + "step": 3540 + }, + { + "epoch": 0.5795918367346938, + "grad_norm": 0.14817261695861816, + "learning_rate": 1.9358936836748972e-06, + "loss": 0.05510561466217041, + "step": 3550 + }, + { + "epoch": 0.5812244897959183, + "grad_norm": 0.15079492330551147, + "learning_rate": 1.9232019429880874e-06, + "loss": 0.06125714182853699, + "step": 3560 + }, + { + "epoch": 0.5828571428571429, + "grad_norm": 0.10553149878978729, + "learning_rate": 1.910525884553976e-06, + "loss": 0.051503753662109374, + "step": 3570 + }, + { + "epoch": 0.5844897959183674, + "grad_norm": 0.09094442427158356, + "learning_rate": 1.8978658530151113e-06, + "loss": 0.05377998948097229, + "step": 3580 + }, + { + "epoch": 0.5861224489795919, + "grad_norm": 0.09810740500688553, + "learning_rate": 1.8852221925782935e-06, + "loss": 0.05506981611251831, + "step": 3590 + }, + { + "epoch": 0.5877551020408164, + "grad_norm": 0.13397352397441864, + "learning_rate": 1.872595247005218e-06, + "loss": 0.056353437900543216, + "step": 3600 + }, + { + "epoch": 0.5877551020408164, + "eval_accuracy": 0.979875, + "eval_loss": 0.04787343740463257, + "eval_runtime": 157.8457, + "eval_samples_per_second": 25.341, + "eval_steps_per_second": 25.341, + "step": 3600 + }, + { + "epoch": 0.5893877551020408, + "grad_norm": 0.07899880409240723, + "learning_rate": 1.8599853596031289e-06, + "loss": 0.05329249501228332, + "step": 3610 + }, + { + "epoch": 0.5910204081632653, + "grad_norm": 0.10414254665374756, + "learning_rate": 1.8473928732154847e-06, + "loss": 0.04905630648136139, + "step": 3620 + }, + { + "epoch": 0.5926530612244898, + "grad_norm": 0.1037079319357872, + "learning_rate": 1.8348181302126367e-06, + "loss": 0.04863021373748779, + "step": 3630 + }, + { + "epoch": 0.5942857142857143, + "grad_norm": 0.06960397213697433, + "learning_rate": 1.8222614724825222e-06, + "loss": 0.04666702449321747, + "step": 3640 + }, + { + "epoch": 0.5959183673469388, + "grad_norm": 0.11159648001194, + "learning_rate": 1.8097232414213662e-06, + "loss": 0.05730679631233215, + "step": 3650 + }, + { + "epoch": 0.5975510204081632, + "grad_norm": 0.10231669992208481, + "learning_rate": 1.7972037779244e-06, + "loss": 0.04433938264846802, + "step": 3660 + }, + { + "epoch": 0.5991836734693877, + "grad_norm": 0.1256829798221588, + "learning_rate": 1.784703422376598e-06, + "loss": 0.04502029418945312, + "step": 3670 + }, + { + "epoch": 0.6008163265306122, + "grad_norm": 0.1006830632686615, + "learning_rate": 1.7722225146434146e-06, + "loss": 0.04091633260250092, + "step": 3680 + }, + { + "epoch": 0.6024489795918367, + "grad_norm": 0.12239495664834976, + "learning_rate": 1.7597613940615492e-06, + "loss": 0.0417140930891037, + "step": 3690 + }, + { + "epoch": 0.6040816326530613, + "grad_norm": 0.09386949241161346, + "learning_rate": 1.7473203994297197e-06, + "loss": 0.045112240314483645, + "step": 3700 + }, + { + "epoch": 0.6057142857142858, + "grad_norm": 0.1845707893371582, + "learning_rate": 1.7348998689994492e-06, + "loss": 0.04915813505649567, + "step": 3710 + }, + { + "epoch": 0.6073469387755102, + "grad_norm": 0.1345887929201126, + "learning_rate": 1.7225001404658714e-06, + "loss": 0.05195118188858032, + "step": 3720 + }, + { + "epoch": 0.6089795918367347, + "grad_norm": 0.16226522624492645, + "learning_rate": 1.7101215509585482e-06, + "loss": 0.047248026728630065, + "step": 3730 + }, + { + "epoch": 0.6106122448979592, + "grad_norm": 0.12012234330177307, + "learning_rate": 1.6977644370323032e-06, + "loss": 0.05229860544204712, + "step": 3740 + }, + { + "epoch": 0.6122448979591837, + "grad_norm": 0.16154533624649048, + "learning_rate": 1.6854291346580737e-06, + "loss": 0.048448923230171206, + "step": 3750 + }, + { + "epoch": 0.6138775510204082, + "grad_norm": 0.08616314828395844, + "learning_rate": 1.673115979213773e-06, + "loss": 0.04779277443885803, + "step": 3760 + }, + { + "epoch": 0.6155102040816327, + "grad_norm": 0.20649370551109314, + "learning_rate": 1.6608253054751739e-06, + "loss": 0.04001966416835785, + "step": 3770 + }, + { + "epoch": 0.6171428571428571, + "grad_norm": 0.22419355809688568, + "learning_rate": 1.6485574476068056e-06, + "loss": 0.04915740489959717, + "step": 3780 + }, + { + "epoch": 0.6187755102040816, + "grad_norm": 0.11812295019626617, + "learning_rate": 1.6363127391528694e-06, + "loss": 0.055313241481781, + "step": 3790 + }, + { + "epoch": 0.6204081632653061, + "grad_norm": 0.12400855869054794, + "learning_rate": 1.6240915130281696e-06, + "loss": 0.05452631115913391, + "step": 3800 + }, + { + "epoch": 0.6204081632653061, + "eval_accuracy": 0.981875, + "eval_loss": 0.04318515956401825, + "eval_runtime": 158.5838, + "eval_samples_per_second": 25.223, + "eval_steps_per_second": 25.223, + "step": 3800 + }, + { + "epoch": 0.6220408163265306, + "grad_norm": 0.094053253531456, + "learning_rate": 1.6118941015090614e-06, + "loss": 0.047327810525894166, + "step": 3810 + }, + { + "epoch": 0.6236734693877551, + "grad_norm": 0.11286808550357819, + "learning_rate": 1.599720836224419e-06, + "loss": 0.04089818894863129, + "step": 3820 + }, + { + "epoch": 0.6253061224489795, + "grad_norm": 0.0909293070435524, + "learning_rate": 1.5875720481466162e-06, + "loss": 0.04378651976585388, + "step": 3830 + }, + { + "epoch": 0.626938775510204, + "grad_norm": 0.1207573264837265, + "learning_rate": 1.5754480675825295e-06, + "loss": 0.055971544981002805, + "step": 3840 + }, + { + "epoch": 0.6285714285714286, + "grad_norm": 0.09171696752309799, + "learning_rate": 1.563349224164556e-06, + "loss": 0.04364039599895477, + "step": 3850 + }, + { + "epoch": 0.6302040816326531, + "grad_norm": 0.1100732609629631, + "learning_rate": 1.5512758468416561e-06, + "loss": 0.04227602481842041, + "step": 3860 + }, + { + "epoch": 0.6318367346938776, + "grad_norm": 0.08587154746055603, + "learning_rate": 1.5392282638704032e-06, + "loss": 0.03987776041030884, + "step": 3870 + }, + { + "epoch": 0.6334693877551021, + "grad_norm": 0.14273083209991455, + "learning_rate": 1.5272068028060616e-06, + "loss": 0.05040489435195923, + "step": 3880 + }, + { + "epoch": 0.6351020408163265, + "grad_norm": 0.12788064777851105, + "learning_rate": 1.5152117904936827e-06, + "loss": 0.04129888415336609, + "step": 3890 + }, + { + "epoch": 0.636734693877551, + "grad_norm": 0.14224086701869965, + "learning_rate": 1.5032435530592165e-06, + "loss": 0.04145504236221313, + "step": 3900 + }, + { + "epoch": 0.6383673469387755, + "grad_norm": 0.07414636015892029, + "learning_rate": 1.4913024159006444e-06, + "loss": 0.03938417434692383, + "step": 3910 + }, + { + "epoch": 0.64, + "grad_norm": 0.1443425565958023, + "learning_rate": 1.4793887036791338e-06, + "loss": 0.047359606623649596, + "step": 3920 + }, + { + "epoch": 0.6416326530612245, + "grad_norm": 0.1462288498878479, + "learning_rate": 1.4675027403102107e-06, + "loss": 0.047661393880844116, + "step": 3930 + }, + { + "epoch": 0.643265306122449, + "grad_norm": 0.15937893092632294, + "learning_rate": 1.4556448489549519e-06, + "loss": 0.04829471111297608, + "step": 3940 + }, + { + "epoch": 0.6448979591836734, + "grad_norm": 0.10469378530979156, + "learning_rate": 1.4438153520111978e-06, + "loss": 0.047904521226882935, + "step": 3950 + }, + { + "epoch": 0.6465306122448979, + "grad_norm": 0.10817313939332962, + "learning_rate": 1.4320145711047894e-06, + "loss": 0.04648992121219635, + "step": 3960 + }, + { + "epoch": 0.6481632653061224, + "grad_norm": 0.12307990342378616, + "learning_rate": 1.4202428270808256e-06, + "loss": 0.044715133309364316, + "step": 3970 + }, + { + "epoch": 0.649795918367347, + "grad_norm": 0.14425775408744812, + "learning_rate": 1.408500439994933e-06, + "loss": 0.042470496892929074, + "step": 3980 + }, + { + "epoch": 0.6514285714285715, + "grad_norm": 0.11694067716598511, + "learning_rate": 1.3967877291045706e-06, + "loss": 0.03910423815250397, + "step": 3990 + }, + { + "epoch": 0.6530612244897959, + "grad_norm": 0.13050039112567902, + "learning_rate": 1.3851050128603454e-06, + "loss": 0.046834564208984374, + "step": 4000 + }, + { + "epoch": 0.6530612244897959, + "eval_accuracy": 0.9817083333333334, + "eval_loss": 0.04485878720879555, + "eval_runtime": 155.2113, + "eval_samples_per_second": 25.771, + "eval_steps_per_second": 25.771, + "step": 4000 + }, + { + "epoch": 0.6546938775510204, + "grad_norm": 0.16441842913627625, + "learning_rate": 1.3734526088973576e-06, + "loss": 0.048332995176315306, + "step": 4010 + }, + { + "epoch": 0.6563265306122449, + "grad_norm": 0.16489046812057495, + "learning_rate": 1.3618308340265628e-06, + "loss": 0.05632186532020569, + "step": 4020 + }, + { + "epoch": 0.6579591836734694, + "grad_norm": 0.0776519775390625, + "learning_rate": 1.3502400042261583e-06, + "loss": 0.0413711816072464, + "step": 4030 + }, + { + "epoch": 0.6595918367346939, + "grad_norm": 0.07065974175930023, + "learning_rate": 1.3386804346329913e-06, + "loss": 0.04443484246730804, + "step": 4040 + }, + { + "epoch": 0.6612244897959184, + "grad_norm": 0.058040838688611984, + "learning_rate": 1.3271524395339958e-06, + "loss": 0.03936767876148224, + "step": 4050 + }, + { + "epoch": 0.6628571428571428, + "grad_norm": 0.12576180696487427, + "learning_rate": 1.31565633235764e-06, + "loss": 0.03893322050571442, + "step": 4060 + }, + { + "epoch": 0.6644897959183673, + "grad_norm": 0.1022111177444458, + "learning_rate": 1.30419242566541e-06, + "loss": 0.05221283435821533, + "step": 4070 + }, + { + "epoch": 0.6661224489795918, + "grad_norm": 0.07897438853979111, + "learning_rate": 1.2927610311433098e-06, + "loss": 0.04261190891265869, + "step": 4080 + }, + { + "epoch": 0.6677551020408163, + "grad_norm": 0.16535215079784393, + "learning_rate": 1.2813624595933885e-06, + "loss": 0.04804243743419647, + "step": 4090 + }, + { + "epoch": 0.6693877551020408, + "grad_norm": 0.10754700005054474, + "learning_rate": 1.2699970209252877e-06, + "loss": 0.03677414059638977, + "step": 4100 + }, + { + "epoch": 0.6710204081632654, + "grad_norm": 0.13121150434017181, + "learning_rate": 1.2586650241478173e-06, + "loss": 0.04266786575317383, + "step": 4110 + }, + { + "epoch": 0.6726530612244898, + "grad_norm": 0.11523906886577606, + "learning_rate": 1.2473667773605542e-06, + "loss": 0.045382776856422426, + "step": 4120 + }, + { + "epoch": 0.6742857142857143, + "grad_norm": 0.16150203347206116, + "learning_rate": 1.236102587745464e-06, + "loss": 0.0483401358127594, + "step": 4130 + }, + { + "epoch": 0.6759183673469388, + "grad_norm": 0.2037627100944519, + "learning_rate": 1.2248727615585501e-06, + "loss": 0.0531754195690155, + "step": 4140 + }, + { + "epoch": 0.6775510204081633, + "grad_norm": 0.08756938576698303, + "learning_rate": 1.2136776041215293e-06, + "loss": 0.046069496870040895, + "step": 4150 + }, + { + "epoch": 0.6791836734693878, + "grad_norm": 0.10196162015199661, + "learning_rate": 1.2025174198135255e-06, + "loss": 0.03938464820384979, + "step": 4160 + }, + { + "epoch": 0.6808163265306122, + "grad_norm": 0.10243439674377441, + "learning_rate": 1.1913925120627987e-06, + "loss": 0.04410702586174011, + "step": 4170 + }, + { + "epoch": 0.6824489795918367, + "grad_norm": 0.16230984032154083, + "learning_rate": 1.1803031833384923e-06, + "loss": 0.03496443629264832, + "step": 4180 + }, + { + "epoch": 0.6840816326530612, + "grad_norm": 0.1036972925066948, + "learning_rate": 1.1692497351424117e-06, + "loss": 0.04019207060337067, + "step": 4190 + }, + { + "epoch": 0.6857142857142857, + "grad_norm": 0.19966299831867218, + "learning_rate": 1.1582324680008252e-06, + "loss": 0.03875584900379181, + "step": 4200 + }, + { + "epoch": 0.6857142857142857, + "eval_accuracy": 0.9839583333333334, + "eval_loss": 0.0395464189350605, + "eval_runtime": 153.8599, + "eval_samples_per_second": 25.998, + "eval_steps_per_second": 25.998, + "step": 4200 + }, + { + "epoch": 0.6873469387755102, + "grad_norm": 0.15349307656288147, + "learning_rate": 1.1472516814562945e-06, + "loss": 0.04495754837989807, + "step": 4210 + }, + { + "epoch": 0.6889795918367347, + "grad_norm": 0.16430068016052246, + "learning_rate": 1.1363076740595294e-06, + "loss": 0.0422965407371521, + "step": 4220 + }, + { + "epoch": 0.6906122448979591, + "grad_norm": 0.1903417408466339, + "learning_rate": 1.1254007433612705e-06, + "loss": 0.04501302838325501, + "step": 4230 + }, + { + "epoch": 0.6922448979591836, + "grad_norm": 0.05951374024152756, + "learning_rate": 1.1145311859042026e-06, + "loss": 0.03593409359455109, + "step": 4240 + }, + { + "epoch": 0.6938775510204082, + "grad_norm": 0.05894222855567932, + "learning_rate": 1.1036992972148865e-06, + "loss": 0.034275355935096743, + "step": 4250 + }, + { + "epoch": 0.6955102040816327, + "grad_norm": 0.09788545221090317, + "learning_rate": 1.092905371795729e-06, + "loss": 0.039912134408950806, + "step": 4260 + }, + { + "epoch": 0.6971428571428572, + "grad_norm": 0.2608252763748169, + "learning_rate": 1.0821497031169729e-06, + "loss": 0.03563873171806335, + "step": 4270 + }, + { + "epoch": 0.6987755102040817, + "grad_norm": 0.10350881516933441, + "learning_rate": 1.0714325836087191e-06, + "loss": 0.036586591601371767, + "step": 4280 + }, + { + "epoch": 0.7004081632653061, + "grad_norm": 0.10214894264936447, + "learning_rate": 1.0607543046529763e-06, + "loss": 0.03775157630443573, + "step": 4290 + }, + { + "epoch": 0.7020408163265306, + "grad_norm": 0.1326843649148941, + "learning_rate": 1.050115156575738e-06, + "loss": 0.03171868622303009, + "step": 4300 + }, + { + "epoch": 0.7036734693877551, + "grad_norm": 0.12274391204118729, + "learning_rate": 1.0395154286390896e-06, + "loss": 0.028169712424278258, + "step": 4310 + }, + { + "epoch": 0.7053061224489796, + "grad_norm": 0.12580160796642303, + "learning_rate": 1.028955409033342e-06, + "loss": 0.03762204349040985, + "step": 4320 + }, + { + "epoch": 0.7069387755102041, + "grad_norm": 0.26383766531944275, + "learning_rate": 1.0184353848692005e-06, + "loss": 0.04035975933074951, + "step": 4330 + }, + { + "epoch": 0.7085714285714285, + "grad_norm": 0.101371631026268, + "learning_rate": 1.0079556421699533e-06, + "loss": 0.04394418001174927, + "step": 4340 + }, + { + "epoch": 0.710204081632653, + "grad_norm": 0.12708403170108795, + "learning_rate": 9.97516465863698e-07, + "loss": 0.04101270735263825, + "step": 4350 + }, + { + "epoch": 0.7118367346938775, + "grad_norm": 0.08838815987110138, + "learning_rate": 9.87118139775595e-07, + "loss": 0.04232141673564911, + "step": 4360 + }, + { + "epoch": 0.713469387755102, + "grad_norm": 0.13831272721290588, + "learning_rate": 9.767609466201494e-07, + "loss": 0.03110981285572052, + "step": 4370 + }, + { + "epoch": 0.7151020408163266, + "grad_norm": 0.14169402420520782, + "learning_rate": 9.664451679935252e-07, + "loss": 0.036100244522094725, + "step": 4380 + }, + { + "epoch": 0.7167346938775511, + "grad_norm": 0.09364918619394302, + "learning_rate": 9.561710843658892e-07, + "loss": 0.028025591373443605, + "step": 4390 + }, + { + "epoch": 0.7183673469387755, + "grad_norm": 0.08178938925266266, + "learning_rate": 9.459389750737854e-07, + "loss": 0.03276517987251282, + "step": 4400 + }, + { + "epoch": 0.7183673469387755, + "eval_accuracy": 0.9830833333333334, + "eval_loss": 0.03968822583556175, + "eval_runtime": 170.9863, + "eval_samples_per_second": 23.394, + "eval_steps_per_second": 23.394, + "step": 4400 + }, + { + "epoch": 0.72, + "grad_norm": 0.21406538784503937, + "learning_rate": 9.357491183125398e-07, + "loss": 0.04240003228187561, + "step": 4410 + }, + { + "epoch": 0.7216326530612245, + "grad_norm": 0.2528787851333618, + "learning_rate": 9.256017911286958e-07, + "loss": 0.03953196704387665, + "step": 4420 + }, + { + "epoch": 0.723265306122449, + "grad_norm": 0.1817523092031479, + "learning_rate": 9.154972694124866e-07, + "loss": 0.03630510270595551, + "step": 4430 + }, + { + "epoch": 0.7248979591836735, + "grad_norm": 0.19527313113212585, + "learning_rate": 9.054358278903272e-07, + "loss": 0.04581010937690735, + "step": 4440 + }, + { + "epoch": 0.726530612244898, + "grad_norm": 0.3551980257034302, + "learning_rate": 8.954177401173491e-07, + "loss": 0.04633161723613739, + "step": 4450 + }, + { + "epoch": 0.7281632653061224, + "grad_norm": 0.1003197506070137, + "learning_rate": 8.854432784699624e-07, + "loss": 0.03686977326869965, + "step": 4460 + }, + { + "epoch": 0.7297959183673469, + "grad_norm": 0.1250677853822708, + "learning_rate": 8.755127141384495e-07, + "loss": 0.03494169116020203, + "step": 4470 + }, + { + "epoch": 0.7314285714285714, + "grad_norm": 0.23732195794582367, + "learning_rate": 8.656263171195931e-07, + "loss": 0.04174305498600006, + "step": 4480 + }, + { + "epoch": 0.7330612244897959, + "grad_norm": 0.2553883194923401, + "learning_rate": 8.557843562093329e-07, + "loss": 0.04263312518596649, + "step": 4490 + }, + { + "epoch": 0.7346938775510204, + "grad_norm": 0.1312720775604248, + "learning_rate": 8.459870989954611e-07, + "loss": 0.03155354261398315, + "step": 4500 + }, + { + "epoch": 0.7363265306122448, + "grad_norm": 0.2632088363170624, + "learning_rate": 8.362348118503424e-07, + "loss": 0.05083600878715515, + "step": 4510 + }, + { + "epoch": 0.7379591836734694, + "grad_norm": 0.1293715387582779, + "learning_rate": 8.265277599236785e-07, + "loss": 0.031102049350738525, + "step": 4520 + }, + { + "epoch": 0.7395918367346939, + "grad_norm": 0.15807363390922546, + "learning_rate": 8.168662071352918e-07, + "loss": 0.033837145566940306, + "step": 4530 + }, + { + "epoch": 0.7412244897959184, + "grad_norm": 0.08791180700063705, + "learning_rate": 8.072504161679534e-07, + "loss": 0.0337544322013855, + "step": 4540 + }, + { + "epoch": 0.7428571428571429, + "grad_norm": 0.10295934230089188, + "learning_rate": 7.976806484602417e-07, + "loss": 0.03291972577571869, + "step": 4550 + }, + { + "epoch": 0.7444897959183674, + "grad_norm": 0.11660359799861908, + "learning_rate": 7.881571641994331e-07, + "loss": 0.032550466060638425, + "step": 4560 + }, + { + "epoch": 0.7461224489795918, + "grad_norm": 0.1896156519651413, + "learning_rate": 7.786802223144286e-07, + "loss": 0.030972787737846376, + "step": 4570 + }, + { + "epoch": 0.7477551020408163, + "grad_norm": 0.2084398865699768, + "learning_rate": 7.69250080468712e-07, + "loss": 0.0415867954492569, + "step": 4580 + }, + { + "epoch": 0.7493877551020408, + "grad_norm": 0.12618854641914368, + "learning_rate": 7.59866995053348e-07, + "loss": 0.0360225111246109, + "step": 4590 + }, + { + "epoch": 0.7510204081632653, + "grad_norm": 0.12291006743907928, + "learning_rate": 7.505312211800081e-07, + "loss": 0.03627174496650696, + "step": 4600 + }, + { + "epoch": 0.7510204081632653, + "eval_accuracy": 0.9855833333333334, + "eval_loss": 0.03568050265312195, + "eval_runtime": 156.9074, + "eval_samples_per_second": 25.493, + "eval_steps_per_second": 25.493, + "step": 4600 + }, + { + "epoch": 0.7526530612244898, + "grad_norm": 0.1491072177886963, + "learning_rate": 7.41243012674035e-07, + "loss": 0.033873635530471805, + "step": 4610 + }, + { + "epoch": 0.7542857142857143, + "grad_norm": 0.14268337190151215, + "learning_rate": 7.320026220675447e-07, + "loss": 0.03949807286262512, + "step": 4620 + }, + { + "epoch": 0.7559183673469387, + "grad_norm": 0.15548810362815857, + "learning_rate": 7.228103005925557e-07, + "loss": 0.03478931188583374, + "step": 4630 + }, + { + "epoch": 0.7575510204081632, + "grad_norm": 0.14298392832279205, + "learning_rate": 7.136662981741613e-07, + "loss": 0.030326524376869203, + "step": 4640 + }, + { + "epoch": 0.7591836734693878, + "grad_norm": 0.11854545772075653, + "learning_rate": 7.045708634237341e-07, + "loss": 0.034136393666267396, + "step": 4650 + }, + { + "epoch": 0.7608163265306123, + "grad_norm": 0.1346571296453476, + "learning_rate": 6.955242436321663e-07, + "loss": 0.040317881107330325, + "step": 4660 + }, + { + "epoch": 0.7624489795918368, + "grad_norm": 0.13217571377754211, + "learning_rate": 6.865266847631463e-07, + "loss": 0.02895120084285736, + "step": 4670 + }, + { + "epoch": 0.7640816326530612, + "grad_norm": 0.0761382132768631, + "learning_rate": 6.775784314464717e-07, + "loss": 0.029726022481918336, + "step": 4680 + }, + { + "epoch": 0.7657142857142857, + "grad_norm": 0.17583715915679932, + "learning_rate": 6.686797269713971e-07, + "loss": 0.02383600324392319, + "step": 4690 + }, + { + "epoch": 0.7673469387755102, + "grad_norm": 0.13434812426567078, + "learning_rate": 6.598308132800221e-07, + "loss": 0.03801000118255615, + "step": 4700 + }, + { + "epoch": 0.7689795918367347, + "grad_norm": 0.0841183140873909, + "learning_rate": 6.510319309607099e-07, + "loss": 0.030053699016571046, + "step": 4710 + }, + { + "epoch": 0.7706122448979592, + "grad_norm": 0.10911499708890915, + "learning_rate": 6.422833192415476e-07, + "loss": 0.03296325206756592, + "step": 4720 + }, + { + "epoch": 0.7722448979591837, + "grad_norm": 0.08562543988227844, + "learning_rate": 6.335852159838423e-07, + "loss": 0.029071077704429626, + "step": 4730 + }, + { + "epoch": 0.7738775510204081, + "grad_norm": 0.24464724957942963, + "learning_rate": 6.249378576756538e-07, + "loss": 0.03307548463344574, + "step": 4740 + }, + { + "epoch": 0.7755102040816326, + "grad_norm": 0.17294660210609436, + "learning_rate": 6.163414794253641e-07, + "loss": 0.03507235944271088, + "step": 4750 + }, + { + "epoch": 0.7771428571428571, + "grad_norm": 0.18331816792488098, + "learning_rate": 6.077963149552868e-07, + "loss": 0.038254004716873166, + "step": 4760 + }, + { + "epoch": 0.7787755102040816, + "grad_norm": 0.10756120830774307, + "learning_rate": 5.993025965953103e-07, + "loss": 0.03428310751914978, + "step": 4770 + }, + { + "epoch": 0.7804081632653062, + "grad_norm": 0.16071061789989471, + "learning_rate": 5.908605552765833e-07, + "loss": 0.032243117690086365, + "step": 4780 + }, + { + "epoch": 0.7820408163265307, + "grad_norm": 0.14897944033145905, + "learning_rate": 5.824704205252343e-07, + "loss": 0.03812848925590515, + "step": 4790 + }, + { + "epoch": 0.7836734693877551, + "grad_norm": 0.1536232829093933, + "learning_rate": 5.741324204561335e-07, + "loss": 0.031389710307121274, + "step": 4800 + }, + { + "epoch": 0.7836734693877551, + "eval_accuracy": 0.985, + "eval_loss": 0.03615976870059967, + "eval_runtime": 153.2713, + "eval_samples_per_second": 26.098, + "eval_steps_per_second": 26.098, + "step": 4800 + }, + { + "epoch": 0.7853061224489796, + "grad_norm": 0.15687023103237152, + "learning_rate": 5.65846781766688e-07, + "loss": 0.027873021364212037, + "step": 4810 + }, + { + "epoch": 0.7869387755102041, + "grad_norm": 0.14541009068489075, + "learning_rate": 5.576137297306794e-07, + "loss": 0.0407156765460968, + "step": 4820 + }, + { + "epoch": 0.7885714285714286, + "grad_norm": 0.16398967802524567, + "learning_rate": 5.494334881921393e-07, + "loss": 0.03936888575553894, + "step": 4830 + }, + { + "epoch": 0.7902040816326531, + "grad_norm": 0.2177809625864029, + "learning_rate": 5.413062795592633e-07, + "loss": 0.03059300184249878, + "step": 4840 + }, + { + "epoch": 0.7918367346938775, + "grad_norm": 0.1487113982439041, + "learning_rate": 5.332323247983632e-07, + "loss": 0.038730043172836306, + "step": 4850 + }, + { + "epoch": 0.793469387755102, + "grad_norm": 0.1111542284488678, + "learning_rate": 5.252118434278605e-07, + "loss": 0.033898717164993285, + "step": 4860 + }, + { + "epoch": 0.7951020408163265, + "grad_norm": 0.10296906530857086, + "learning_rate": 5.172450535123166e-07, + "loss": 0.033659690618515016, + "step": 4870 + }, + { + "epoch": 0.796734693877551, + "grad_norm": 0.11844902485609055, + "learning_rate": 5.093321716565064e-07, + "loss": 0.0293354332447052, + "step": 4880 + }, + { + "epoch": 0.7983673469387755, + "grad_norm": 0.08401476591825485, + "learning_rate": 5.014734129995258e-07, + "loss": 0.027682244777679443, + "step": 4890 + }, + { + "epoch": 0.8, + "grad_norm": 0.15870651602745056, + "learning_rate": 4.936689912089443e-07, + "loss": 0.029118484258651732, + "step": 4900 + }, + { + "epoch": 0.8016326530612244, + "grad_norm": 0.10739132016897202, + "learning_rate": 4.859191184749968e-07, + "loss": 0.031624987721443176, + "step": 4910 + }, + { + "epoch": 0.803265306122449, + "grad_norm": 0.1523323655128479, + "learning_rate": 4.782240055048113e-07, + "loss": 0.03415923118591309, + "step": 4920 + }, + { + "epoch": 0.8048979591836735, + "grad_norm": 0.31860172748565674, + "learning_rate": 4.705838615166836e-07, + "loss": 0.035749971866607666, + "step": 4930 + }, + { + "epoch": 0.806530612244898, + "grad_norm": 0.15043528378009796, + "learning_rate": 4.629988942343863e-07, + "loss": 0.029294773936271667, + "step": 4940 + }, + { + "epoch": 0.8081632653061225, + "grad_norm": 0.1007913127541542, + "learning_rate": 4.55469309881523e-07, + "loss": 0.026727235317230223, + "step": 4950 + }, + { + "epoch": 0.809795918367347, + "grad_norm": 0.10851600021123886, + "learning_rate": 4.479953131759199e-07, + "loss": 0.029011189937591553, + "step": 4960 + }, + { + "epoch": 0.8114285714285714, + "grad_norm": 0.19257904589176178, + "learning_rate": 4.4057710732406113e-07, + "loss": 0.040424668788909913, + "step": 4970 + }, + { + "epoch": 0.8130612244897959, + "grad_norm": 0.08985571563243866, + "learning_rate": 4.332148940155617e-07, + "loss": 0.031291928887367246, + "step": 4980 + }, + { + "epoch": 0.8146938775510204, + "grad_norm": 0.09874805808067322, + "learning_rate": 4.259088734176883e-07, + "loss": 0.028314629197120668, + "step": 4990 + }, + { + "epoch": 0.8163265306122449, + "grad_norm": 0.2056759148836136, + "learning_rate": 4.186592441699119e-07, + "loss": 0.028049173951148986, + "step": 5000 + }, + { + "epoch": 0.8163265306122449, + "eval_accuracy": 0.986125, + "eval_loss": 0.033585112541913986, + "eval_runtime": 153.2119, + "eval_samples_per_second": 26.108, + "eval_steps_per_second": 26.108, + "step": 5000 + }, + { + "epoch": 0.8179591836734694, + "grad_norm": 0.10125680267810822, + "learning_rate": 4.1146620337850965e-07, + "loss": 0.03612466752529144, + "step": 5010 + }, + { + "epoch": 0.8195918367346938, + "grad_norm": 0.10107982903718948, + "learning_rate": 4.043299466112058e-07, + "loss": 0.026968371868133546, + "step": 5020 + }, + { + "epoch": 0.8212244897959183, + "grad_norm": 0.13236716389656067, + "learning_rate": 3.9725066789185446e-07, + "loss": 0.030766445398330688, + "step": 5030 + }, + { + "epoch": 0.8228571428571428, + "grad_norm": 0.1182790994644165, + "learning_rate": 3.9022855969516384e-07, + "loss": 0.028759539127349854, + "step": 5040 + }, + { + "epoch": 0.8244897959183674, + "grad_norm": 0.1430918574333191, + "learning_rate": 3.832638129414637e-07, + "loss": 0.03126598298549652, + "step": 5050 + }, + { + "epoch": 0.8261224489795919, + "grad_norm": 0.24944688379764557, + "learning_rate": 3.7635661699151366e-07, + "loss": 0.030412834882736207, + "step": 5060 + }, + { + "epoch": 0.8277551020408164, + "grad_norm": 0.24371539056301117, + "learning_rate": 3.695071596413574e-07, + "loss": 0.03904336988925934, + "step": 5070 + }, + { + "epoch": 0.8293877551020408, + "grad_norm": 0.25066161155700684, + "learning_rate": 3.627156271172125e-07, + "loss": 0.032776132225990295, + "step": 5080 + }, + { + "epoch": 0.8310204081632653, + "grad_norm": 0.1635182797908783, + "learning_rate": 3.559822040704108e-07, + "loss": 0.027044805884361266, + "step": 5090 + }, + { + "epoch": 0.8326530612244898, + "grad_norm": 0.16698378324508667, + "learning_rate": 3.4930707357237666e-07, + "loss": 0.03271254003047943, + "step": 5100 + }, + { + "epoch": 0.8342857142857143, + "grad_norm": 0.18812412023544312, + "learning_rate": 3.4269041710964903e-07, + "loss": 0.030745366215705873, + "step": 5110 + }, + { + "epoch": 0.8359183673469388, + "grad_norm": 0.18739669024944305, + "learning_rate": 3.36132414578948e-07, + "loss": 0.032180759310722354, + "step": 5120 + }, + { + "epoch": 0.8375510204081633, + "grad_norm": 0.13260763883590698, + "learning_rate": 3.296332442822839e-07, + "loss": 0.026755687594413758, + "step": 5130 + }, + { + "epoch": 0.8391836734693877, + "grad_norm": 0.06771053373813629, + "learning_rate": 3.231930829221083e-07, + "loss": 0.02895594835281372, + "step": 5140 + }, + { + "epoch": 0.8408163265306122, + "grad_norm": 0.09441749006509781, + "learning_rate": 3.168121055965112e-07, + "loss": 0.033178862929344174, + "step": 5150 + }, + { + "epoch": 0.8424489795918367, + "grad_norm": 0.0863656997680664, + "learning_rate": 3.104904857944591e-07, + "loss": 0.02276728004217148, + "step": 5160 + }, + { + "epoch": 0.8440816326530612, + "grad_norm": 0.2047131359577179, + "learning_rate": 3.0422839539107794e-07, + "loss": 0.028258222341537475, + "step": 5170 + }, + { + "epoch": 0.8457142857142858, + "grad_norm": 0.2389051467180252, + "learning_rate": 2.980260046429834e-07, + "loss": 0.02725991904735565, + "step": 5180 + }, + { + "epoch": 0.8473469387755102, + "grad_norm": 0.1733924299478531, + "learning_rate": 2.9188348218364635e-07, + "loss": 0.0327379584312439, + "step": 5190 + }, + { + "epoch": 0.8489795918367347, + "grad_norm": 0.19450697302818298, + "learning_rate": 2.8580099501881285e-07, + "loss": 0.030778995156288146, + "step": 5200 + }, + { + "epoch": 0.8489795918367347, + "eval_accuracy": 0.9881666666666667, + "eval_loss": 0.02920014224946499, + "eval_runtime": 155.6487, + "eval_samples_per_second": 25.699, + "eval_steps_per_second": 25.699, + "step": 5200 + }, + { + "epoch": 0.8506122448979592, + "grad_norm": 0.14990773797035217, + "learning_rate": 2.7977870852195997e-07, + "loss": 0.03692179024219513, + "step": 5210 + }, + { + "epoch": 0.8522448979591837, + "grad_norm": 0.16051384806632996, + "learning_rate": 2.7381678642980303e-07, + "loss": 0.030528882145881654, + "step": 5220 + }, + { + "epoch": 0.8538775510204082, + "grad_norm": 0.1660623997449875, + "learning_rate": 2.6791539083784014e-07, + "loss": 0.025125756859779358, + "step": 5230 + }, + { + "epoch": 0.8555102040816327, + "grad_norm": 0.07568248361349106, + "learning_rate": 2.620746821959475e-07, + "loss": 0.02720560133457184, + "step": 5240 + }, + { + "epoch": 0.8571428571428571, + "grad_norm": 0.2634601593017578, + "learning_rate": 2.562948193040174e-07, + "loss": 0.033940815925598146, + "step": 5250 + }, + { + "epoch": 0.8587755102040816, + "grad_norm": 0.15926218032836914, + "learning_rate": 2.5057595930763873e-07, + "loss": 0.028036150336265563, + "step": 5260 + }, + { + "epoch": 0.8604081632653061, + "grad_norm": 0.19775289297103882, + "learning_rate": 2.4491825769382526e-07, + "loss": 0.02731768786907196, + "step": 5270 + }, + { + "epoch": 0.8620408163265306, + "grad_norm": 0.07538789510726929, + "learning_rate": 2.39321868286789e-07, + "loss": 0.0313712328672409, + "step": 5280 + }, + { + "epoch": 0.8636734693877551, + "grad_norm": 0.10647427290678024, + "learning_rate": 2.337869432437567e-07, + "loss": 0.02781860828399658, + "step": 5290 + }, + { + "epoch": 0.8653061224489796, + "grad_norm": 0.20678064227104187, + "learning_rate": 2.283136330508337e-07, + "loss": 0.028782907128334045, + "step": 5300 + }, + { + "epoch": 0.866938775510204, + "grad_norm": 0.18386024236679077, + "learning_rate": 2.2290208651891244e-07, + "loss": 0.02515394687652588, + "step": 5310 + }, + { + "epoch": 0.8685714285714285, + "grad_norm": 0.13139928877353668, + "learning_rate": 2.1755245077962618e-07, + "loss": 0.02002216875553131, + "step": 5320 + }, + { + "epoch": 0.8702040816326531, + "grad_norm": 0.10402868688106537, + "learning_rate": 2.1226487128134854e-07, + "loss": 0.03221684396266937, + "step": 5330 + }, + { + "epoch": 0.8718367346938776, + "grad_norm": 0.056439850479364395, + "learning_rate": 2.070394917852403e-07, + "loss": 0.023763810098171235, + "step": 5340 + }, + { + "epoch": 0.8734693877551021, + "grad_norm": 0.1421876698732376, + "learning_rate": 2.0187645436133913e-07, + "loss": 0.03192594051361084, + "step": 5350 + }, + { + "epoch": 0.8751020408163265, + "grad_norm": 0.20487430691719055, + "learning_rate": 1.967758993846966e-07, + "loss": 0.025955894589424135, + "step": 5360 + }, + { + "epoch": 0.876734693877551, + "grad_norm": 0.2183534801006317, + "learning_rate": 1.9173796553156487e-07, + "loss": 0.027435585856437683, + "step": 5370 + }, + { + "epoch": 0.8783673469387755, + "grad_norm": 0.19724565744400024, + "learning_rate": 1.8676278977562219e-07, + "loss": 0.028912261128425598, + "step": 5380 + }, + { + "epoch": 0.88, + "grad_norm": 0.21412917971611023, + "learning_rate": 1.818505073842508e-07, + "loss": 0.029015693068504333, + "step": 5390 + }, + { + "epoch": 0.8816326530612245, + "grad_norm": 0.08429288119077682, + "learning_rate": 1.770012519148598e-07, + "loss": 0.025249993801116942, + "step": 5400 + }, + { + "epoch": 0.8816326530612245, + "eval_accuracy": 0.989375, + "eval_loss": 0.02678457647562027, + "eval_runtime": 156.2763, + "eval_samples_per_second": 25.596, + "eval_steps_per_second": 25.596, + "step": 5400 + }, + { + "epoch": 0.883265306122449, + "grad_norm": 0.15489192306995392, + "learning_rate": 1.722151552112525e-07, + "loss": 0.03486272096633911, + "step": 5410 + }, + { + "epoch": 0.8848979591836734, + "grad_norm": 0.1594313681125641, + "learning_rate": 1.6749234740004227e-07, + "loss": 0.024751077592372894, + "step": 5420 + }, + { + "epoch": 0.8865306122448979, + "grad_norm": 0.12505307793617249, + "learning_rate": 1.6283295688711564e-07, + "loss": 0.025553828477859496, + "step": 5430 + }, + { + "epoch": 0.8881632653061224, + "grad_norm": 0.08852038532495499, + "learning_rate": 1.582371103541394e-07, + "loss": 0.025743791460990907, + "step": 5440 + }, + { + "epoch": 0.889795918367347, + "grad_norm": 0.08450407534837723, + "learning_rate": 1.5370493275511767e-07, + "loss": 0.02687264084815979, + "step": 5450 + }, + { + "epoch": 0.8914285714285715, + "grad_norm": 0.054211657494306564, + "learning_rate": 1.4923654731299363e-07, + "loss": 0.023073682188987733, + "step": 5460 + }, + { + "epoch": 0.893061224489796, + "grad_norm": 0.23101477324962616, + "learning_rate": 1.4483207551630045e-07, + "loss": 0.0280561625957489, + "step": 5470 + }, + { + "epoch": 0.8946938775510204, + "grad_norm": 0.19923920929431915, + "learning_rate": 1.4049163711585684e-07, + "loss": 0.02133965492248535, + "step": 5480 + }, + { + "epoch": 0.8963265306122449, + "grad_norm": 0.13227398693561554, + "learning_rate": 1.3621535012151244e-07, + "loss": 0.017637649178504945, + "step": 5490 + }, + { + "epoch": 0.8979591836734694, + "grad_norm": 0.11182940751314163, + "learning_rate": 1.3200333079893823e-07, + "loss": 0.02847398817539215, + "step": 5500 + }, + { + "epoch": 0.8995918367346939, + "grad_norm": 0.06762958317995071, + "learning_rate": 1.2785569366646628e-07, + "loss": 0.020397891104221345, + "step": 5510 + }, + { + "epoch": 0.9012244897959184, + "grad_norm": 0.06652984023094177, + "learning_rate": 1.2377255149197576e-07, + "loss": 0.03015478849411011, + "step": 5520 + }, + { + "epoch": 0.9028571428571428, + "grad_norm": 0.16087517142295837, + "learning_rate": 1.1975401528982718e-07, + "loss": 0.02614833414554596, + "step": 5530 + }, + { + "epoch": 0.9044897959183673, + "grad_norm": 0.0873415544629097, + "learning_rate": 1.1580019431784384e-07, + "loss": 0.023097904026508333, + "step": 5540 + }, + { + "epoch": 0.9061224489795918, + "grad_norm": 0.26114869117736816, + "learning_rate": 1.1191119607434076e-07, + "loss": 0.028036117553710938, + "step": 5550 + }, + { + "epoch": 0.9077551020408163, + "grad_norm": 0.08690405637025833, + "learning_rate": 1.0808712629520435e-07, + "loss": 0.027321887016296387, + "step": 5560 + }, + { + "epoch": 0.9093877551020408, + "grad_norm": 0.19979159533977509, + "learning_rate": 1.0432808895101426e-07, + "loss": 0.03543014228343964, + "step": 5570 + }, + { + "epoch": 0.9110204081632653, + "grad_norm": 0.09599421173334122, + "learning_rate": 1.0063418624421895e-07, + "loss": 0.02044544517993927, + "step": 5580 + }, + { + "epoch": 0.9126530612244897, + "grad_norm": 0.22291070222854614, + "learning_rate": 9.700551860635538e-08, + "loss": 0.027166444063186645, + "step": 5590 + }, + { + "epoch": 0.9142857142857143, + "grad_norm": 0.10956079512834549, + "learning_rate": 9.344218469532012e-08, + "loss": 0.02024113982915878, + "step": 5600 + }, + { + "epoch": 0.9142857142857143, + "eval_accuracy": 0.9900833333333334, + "eval_loss": 0.025807492434978485, + "eval_runtime": 154.9378, + "eval_samples_per_second": 25.817, + "eval_steps_per_second": 25.817, + "step": 5600 + }, + { + "epoch": 0.9159183673469388, + "grad_norm": 0.14949144423007965, + "learning_rate": 8.99442813926854e-08, + "loss": 0.024004274606704713, + "step": 5610 + }, + { + "epoch": 0.9175510204081633, + "grad_norm": 0.2315555214881897, + "learning_rate": 8.651190380106673e-08, + "loss": 0.024219857156276704, + "step": 5620 + }, + { + "epoch": 0.9191836734693878, + "grad_norm": 0.06734409183263779, + "learning_rate": 8.314514524153522e-08, + "loss": 0.02802177369594574, + "step": 5630 + }, + { + "epoch": 0.9208163265306123, + "grad_norm": 0.21427671611309052, + "learning_rate": 7.984409725108222e-08, + "loss": 0.021109704673290253, + "step": 5640 + }, + { + "epoch": 0.9224489795918367, + "grad_norm": 0.18013973534107208, + "learning_rate": 7.660884958012899e-08, + "loss": 0.020703735947608947, + "step": 5650 + }, + { + "epoch": 0.9240816326530612, + "grad_norm": 0.2695794701576233, + "learning_rate": 7.343949019008756e-08, + "loss": 0.03247661292552948, + "step": 5660 + }, + { + "epoch": 0.9257142857142857, + "grad_norm": 0.15175238251686096, + "learning_rate": 7.03361052509688e-08, + "loss": 0.021647430956363678, + "step": 5670 + }, + { + "epoch": 0.9273469387755102, + "grad_norm": 0.18325908482074738, + "learning_rate": 6.729877913903976e-08, + "loss": 0.028400209546089173, + "step": 5680 + }, + { + "epoch": 0.9289795918367347, + "grad_norm": 0.15859369933605194, + "learning_rate": 6.432759443452924e-08, + "loss": 0.014232246577739716, + "step": 5690 + }, + { + "epoch": 0.9306122448979591, + "grad_norm": 0.13438236713409424, + "learning_rate": 6.14226319193828e-08, + "loss": 0.02244553864002228, + "step": 5700 + }, + { + "epoch": 0.9322448979591836, + "grad_norm": 0.27041247487068176, + "learning_rate": 5.858397057506709e-08, + "loss": 0.029011306166648865, + "step": 5710 + }, + { + "epoch": 0.9338775510204081, + "grad_norm": 0.12215204536914825, + "learning_rate": 5.581168758042127e-08, + "loss": 0.028310227394104003, + "step": 5720 + }, + { + "epoch": 0.9355102040816327, + "grad_norm": 0.2103251814842224, + "learning_rate": 5.310585830955922e-08, + "loss": 0.021468037366867067, + "step": 5730 + }, + { + "epoch": 0.9371428571428572, + "grad_norm": 0.17152205109596252, + "learning_rate": 5.0466556329819835e-08, + "loss": 0.024149742722511292, + "step": 5740 + }, + { + "epoch": 0.9387755102040817, + "grad_norm": 0.08712399005889893, + "learning_rate": 4.789385339976832e-08, + "loss": 0.01809695065021515, + "step": 5750 + }, + { + "epoch": 0.9404081632653061, + "grad_norm": 0.15006859600543976, + "learning_rate": 4.53878194672433e-08, + "loss": 0.022831857204437256, + "step": 5760 + }, + { + "epoch": 0.9420408163265306, + "grad_norm": 0.06355110555887222, + "learning_rate": 4.2948522667455864e-08, + "loss": 0.015157052874565124, + "step": 5770 + }, + { + "epoch": 0.9436734693877551, + "grad_norm": 0.14330919086933136, + "learning_rate": 4.057602932113658e-08, + "loss": 0.02486245185136795, + "step": 5780 + }, + { + "epoch": 0.9453061224489796, + "grad_norm": 0.12216096371412277, + "learning_rate": 3.82704039327339e-08, + "loss": 0.024203602969646455, + "step": 5790 + }, + { + "epoch": 0.9469387755102041, + "grad_norm": 0.19883109629154205, + "learning_rate": 3.6031709188658336e-08, + "loss": 0.025162199139595033, + "step": 5800 + }, + { + "epoch": 0.9469387755102041, + "eval_accuracy": 0.99025, + "eval_loss": 0.025506427511572838, + "eval_runtime": 154.1326, + "eval_samples_per_second": 25.952, + "eval_steps_per_second": 25.952, + "step": 5800 + } + ], + "logging_steps": 10, + "max_steps": 6125, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 200, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 5.420352461078528e+16, + "train_batch_size": 1, + "trial_name": null, + "trial_params": null +} diff --git a/checkpoint-5800/training_args.bin b/checkpoint-5800/training_args.bin new file mode 100644 index 0000000..a9304c0 --- /dev/null +++ b/checkpoint-5800/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:10a2085a9c347d21774d35a2558a0376b0ffbb33ca9517fafbd4cd5d12552b02 +size 5585 diff --git a/checkpoint-6000/chat_template.jinja b/checkpoint-6000/chat_template.jinja new file mode 100644 index 0000000..ce566c4 --- /dev/null +++ b/checkpoint-6000/chat_template.jinja @@ -0,0 +1,9 @@ +{% set system_message = 'Below is an instruction that describes a task. Write a response that appropriately completes the request. + +' %}{% if messages[0]['role'] == 'system' %}{% set loop_messages = messages[1:] %}{% set system_message = messages[0]['content'] %}{% else %}{% set loop_messages = messages %}{% endif %}{% if system_message is defined %}{{ system_message }}{% endif %}{% for message in loop_messages %}{% set content = message['content'] %}{% if message['role'] == 'user' %}{{ '### Instruction: +' + content + ' + +### Response: +' }}{% elif message['role'] == 'assistant' %}{{ content + '<|endoftext|>' + ' + +' }}{% endif %}{% endfor %} \ No newline at end of file diff --git a/checkpoint-6000/config.json b/checkpoint-6000/config.json new file mode 100644 index 0000000..642be54 --- /dev/null +++ b/checkpoint-6000/config.json @@ -0,0 +1,58 @@ +{ + "architectures": [ + "Qwen2ForCausalLM" + ], + "attention_dropout": 0.0, + "bos_token_id": null, + "dtype": "float32", + "eos_token_id": 151643, + "hidden_act": "silu", + "hidden_size": 896, + "initializer_range": 0.02, + "intermediate_size": 4864, + "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" + ], + "max_position_embeddings": 32768, + "max_window_layers": 24, + "model_type": "qwen2", + "num_attention_heads": 14, + "num_hidden_layers": 24, + "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_mrope": false, + "use_sliding_window": false, + "vocab_size": 151936 +} diff --git a/checkpoint-6000/generation_config.json b/checkpoint-6000/generation_config.json new file mode 100644 index 0000000..795a75e --- /dev/null +++ b/checkpoint-6000/generation_config.json @@ -0,0 +1,9 @@ +{ + "do_sample": false, + "eos_token_id": [ + 151643 + ], + "max_new_tokens": 2048, + "pad_token_id": 151643, + "transformers_version": "5.2.0" +} diff --git a/checkpoint-6000/model.safetensors b/checkpoint-6000/model.safetensors new file mode 100644 index 0000000..f161b17 --- /dev/null +++ b/checkpoint-6000/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d29c7bfd6402cc8e1a34af00183efcd1a3b4bbdb17368612d067d710771c6301 +size 1976163472 diff --git a/checkpoint-6000/tokenizer.json b/checkpoint-6000/tokenizer.json new file mode 100644 index 0000000..34510ff --- /dev/null +++ b/checkpoint-6000/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3fd169731d2cbde95e10bf356d66d5997fd885dd8dbb6fb4684da3f23b2585d8 +size 11421892 diff --git a/checkpoint-6000/tokenizer_config.json b/checkpoint-6000/tokenizer_config.json new file mode 100644 index 0000000..392ded1 --- /dev/null +++ b/checkpoint-6000/tokenizer_config.json @@ -0,0 +1,30 @@ +{ + "add_prefix_space": false, + "backend": "tokenizers", + "bos_token": null, + "clean_up_tokenization_spaces": false, + "eos_token": "<|endoftext|>", + "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/checkpoint-6000/trainer_state.json b/checkpoint-6000/trainer_state.json new file mode 100644 index 0000000..e96e4dd --- /dev/null +++ b/checkpoint-6000/trainer_state.json @@ -0,0 +1,4504 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.9795918367346939, + "eval_steps": 200, + "global_step": 6000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0016326530612244899, + "grad_norm": 7.955283164978027, + "learning_rate": 4.5000000000000003e-07, + "loss": 4.391213607788086, + "step": 10 + }, + { + "epoch": 0.0032653061224489797, + "grad_norm": 2.4868686199188232, + "learning_rate": 9.500000000000001e-07, + "loss": 1.9876020431518555, + "step": 20 + }, + { + "epoch": 0.004897959183673469, + "grad_norm": 0.25438952445983887, + "learning_rate": 1.45e-06, + "loss": 0.34127225875854494, + "step": 30 + }, + { + "epoch": 0.006530612244897959, + "grad_norm": 0.13044093549251556, + "learning_rate": 1.9500000000000004e-06, + "loss": 0.23643033504486083, + "step": 40 + }, + { + "epoch": 0.00816326530612245, + "grad_norm": 0.17984068393707275, + "learning_rate": 2.4500000000000003e-06, + "loss": 0.23313720226287843, + "step": 50 + }, + { + "epoch": 0.009795918367346938, + "grad_norm": 0.15249308943748474, + "learning_rate": 2.95e-06, + "loss": 0.22351114749908446, + "step": 60 + }, + { + "epoch": 0.011428571428571429, + "grad_norm": 0.07123949378728867, + "learning_rate": 3.45e-06, + "loss": 0.21615636348724365, + "step": 70 + }, + { + "epoch": 0.013061224489795919, + "grad_norm": 0.21456502377986908, + "learning_rate": 3.95e-06, + "loss": 0.21677422523498535, + "step": 80 + }, + { + "epoch": 0.014693877551020407, + "grad_norm": 0.22391553223133087, + "learning_rate": 4.450000000000001e-06, + "loss": 0.21422204971313477, + "step": 90 + }, + { + "epoch": 0.0163265306122449, + "grad_norm": 0.1745220273733139, + "learning_rate": 4.95e-06, + "loss": 0.19350455999374389, + "step": 100 + }, + { + "epoch": 0.017959183673469388, + "grad_norm": 0.18820920586585999, + "learning_rate": 4.9999724716692435e-06, + "loss": 0.1981007933616638, + "step": 110 + }, + { + "epoch": 0.019591836734693877, + "grad_norm": 0.21733109652996063, + "learning_rate": 4.9998773127856985e-06, + "loss": 0.16994532346725463, + "step": 120 + }, + { + "epoch": 0.02122448979591837, + "grad_norm": 0.40475547313690186, + "learning_rate": 4.999714186080037e-06, + "loss": 0.1539763927459717, + "step": 130 + }, + { + "epoch": 0.022857142857142857, + "grad_norm": 0.3676598370075226, + "learning_rate": 4.999483095987423e-06, + "loss": 0.16321685314178466, + "step": 140 + }, + { + "epoch": 0.024489795918367346, + "grad_norm": 0.19484871625900269, + "learning_rate": 4.9991840487908426e-06, + "loss": 0.16523454189300538, + "step": 150 + }, + { + "epoch": 0.026122448979591838, + "grad_norm": 0.21928830444812775, + "learning_rate": 4.998817052620928e-06, + "loss": 0.14424897432327272, + "step": 160 + }, + { + "epoch": 0.027755102040816326, + "grad_norm": 0.2790975868701935, + "learning_rate": 4.9983821174557415e-06, + "loss": 0.13810645341873168, + "step": 170 + }, + { + "epoch": 0.029387755102040815, + "grad_norm": 0.21702228486537933, + "learning_rate": 4.997879255120501e-06, + "loss": 0.1318126440048218, + "step": 180 + }, + { + "epoch": 0.031020408163265307, + "grad_norm": 0.16967810690402985, + "learning_rate": 4.997308479287263e-06, + "loss": 0.14249314069747926, + "step": 190 + }, + { + "epoch": 0.0326530612244898, + "grad_norm": 0.14007994532585144, + "learning_rate": 4.996669805474542e-06, + "loss": 0.12708702087402343, + "step": 200 + }, + { + "epoch": 0.0326530612244898, + "eval_accuracy": 0.9408333333333335, + "eval_loss": 0.12691152095794678, + "eval_runtime": 156.68, + "eval_samples_per_second": 25.53, + "eval_steps_per_second": 25.53, + "step": 200 + }, + { + "epoch": 0.03428571428571429, + "grad_norm": 0.2674005627632141, + "learning_rate": 4.9959632510469e-06, + "loss": 0.11019121408462525, + "step": 210 + }, + { + "epoch": 0.035918367346938776, + "grad_norm": 0.12073638290166855, + "learning_rate": 4.995188835214465e-06, + "loss": 0.12032746076583863, + "step": 220 + }, + { + "epoch": 0.037551020408163265, + "grad_norm": 0.11186162382364273, + "learning_rate": 4.994346579032413e-06, + "loss": 0.12122929096221924, + "step": 230 + }, + { + "epoch": 0.03918367346938775, + "grad_norm": 0.09840800613164902, + "learning_rate": 4.993436505400396e-06, + "loss": 0.11416752338409424, + "step": 240 + }, + { + "epoch": 0.04081632653061224, + "grad_norm": 0.19076374173164368, + "learning_rate": 4.992458639061916e-06, + "loss": 0.11376930475234985, + "step": 250 + }, + { + "epoch": 0.04244897959183674, + "grad_norm": 0.14405065774917603, + "learning_rate": 4.991413006603657e-06, + "loss": 0.10997685194015502, + "step": 260 + }, + { + "epoch": 0.044081632653061226, + "grad_norm": 0.23844055831432343, + "learning_rate": 4.990299636454759e-06, + "loss": 0.1000083327293396, + "step": 270 + }, + { + "epoch": 0.045714285714285714, + "grad_norm": 0.23128052055835724, + "learning_rate": 4.989118558886043e-06, + "loss": 0.1243711233139038, + "step": 280 + }, + { + "epoch": 0.0473469387755102, + "grad_norm": 0.19849570095539093, + "learning_rate": 4.9878698060091955e-06, + "loss": 0.11728224754333497, + "step": 290 + }, + { + "epoch": 0.04897959183673469, + "grad_norm": 0.13039851188659668, + "learning_rate": 4.986553411775886e-06, + "loss": 0.09353847503662109, + "step": 300 + }, + { + "epoch": 0.05061224489795919, + "grad_norm": 0.19887588918209076, + "learning_rate": 4.985169411976852e-06, + "loss": 0.11291944980621338, + "step": 310 + }, + { + "epoch": 0.052244897959183675, + "grad_norm": 0.25038304924964905, + "learning_rate": 4.98371784424092e-06, + "loss": 0.11149277687072753, + "step": 320 + }, + { + "epoch": 0.053877551020408164, + "grad_norm": 0.238360196352005, + "learning_rate": 4.982198748033986e-06, + "loss": 0.11344953775405883, + "step": 330 + }, + { + "epoch": 0.05551020408163265, + "grad_norm": 0.13686658442020416, + "learning_rate": 4.980612164657943e-06, + "loss": 0.09844974875450134, + "step": 340 + }, + { + "epoch": 0.05714285714285714, + "grad_norm": 0.0829666405916214, + "learning_rate": 4.978958137249551e-06, + "loss": 0.10898438692092896, + "step": 350 + }, + { + "epoch": 0.05877551020408163, + "grad_norm": 0.10389993339776993, + "learning_rate": 4.977236710779277e-06, + "loss": 0.09371156096458436, + "step": 360 + }, + { + "epoch": 0.060408163265306125, + "grad_norm": 0.18182680010795593, + "learning_rate": 4.9754479320500584e-06, + "loss": 0.1049533724784851, + "step": 370 + }, + { + "epoch": 0.062040816326530614, + "grad_norm": 0.14488115906715393, + "learning_rate": 4.973591849696041e-06, + "loss": 0.10510741472244263, + "step": 380 + }, + { + "epoch": 0.0636734693877551, + "grad_norm": 0.14816738665103912, + "learning_rate": 4.971668514181252e-06, + "loss": 0.10956356525421143, + "step": 390 + }, + { + "epoch": 0.0653061224489796, + "grad_norm": 0.20584072172641754, + "learning_rate": 4.969677977798225e-06, + "loss": 0.11344989538192748, + "step": 400 + }, + { + "epoch": 0.0653061224489796, + "eval_accuracy": 0.9496250000000002, + "eval_loss": 0.11409253627061844, + "eval_runtime": 165.7682, + "eval_samples_per_second": 24.13, + "eval_steps_per_second": 24.13, + "step": 400 + }, + { + "epoch": 0.06693877551020408, + "grad_norm": 0.1342787891626358, + "learning_rate": 4.967620294666589e-06, + "loss": 0.09949527978897095, + "step": 410 + }, + { + "epoch": 0.06857142857142857, + "grad_norm": 0.17336596548557281, + "learning_rate": 4.965495520731583e-06, + "loss": 0.10606559514999389, + "step": 420 + }, + { + "epoch": 0.07020408163265306, + "grad_norm": 0.19437003135681152, + "learning_rate": 4.963303713762547e-06, + "loss": 0.11677042245864869, + "step": 430 + }, + { + "epoch": 0.07183673469387755, + "grad_norm": 0.12076073884963989, + "learning_rate": 4.961044933351342e-06, + "loss": 0.1019243597984314, + "step": 440 + }, + { + "epoch": 0.07346938775510205, + "grad_norm": 0.10272067785263062, + "learning_rate": 4.958719240910737e-06, + "loss": 0.10125980377197266, + "step": 450 + }, + { + "epoch": 0.07510204081632653, + "grad_norm": 0.1080622524023056, + "learning_rate": 4.956326699672736e-06, + "loss": 0.09888594150543213, + "step": 460 + }, + { + "epoch": 0.07673469387755102, + "grad_norm": 0.1307847797870636, + "learning_rate": 4.953867374686856e-06, + "loss": 0.10753540992736817, + "step": 470 + }, + { + "epoch": 0.0783673469387755, + "grad_norm": 0.2128073275089264, + "learning_rate": 4.951341332818365e-06, + "loss": 0.11452326774597169, + "step": 480 + }, + { + "epoch": 0.08, + "grad_norm": 0.21006478369235992, + "learning_rate": 4.9487486427464594e-06, + "loss": 0.09555010795593262, + "step": 490 + }, + { + "epoch": 0.08163265306122448, + "grad_norm": 0.24889807403087616, + "learning_rate": 4.946089374962396e-06, + "loss": 0.09771155118942261, + "step": 500 + }, + { + "epoch": 0.08326530612244898, + "grad_norm": 0.13042858242988586, + "learning_rate": 4.943363601767582e-06, + "loss": 0.10659778118133545, + "step": 510 + }, + { + "epoch": 0.08489795918367347, + "grad_norm": 0.2210501879453659, + "learning_rate": 4.940571397271598e-06, + "loss": 0.10161921977996827, + "step": 520 + }, + { + "epoch": 0.08653061224489796, + "grad_norm": 0.11963305622339249, + "learning_rate": 4.937712837390194e-06, + "loss": 0.09155921936035157, + "step": 530 + }, + { + "epoch": 0.08816326530612245, + "grad_norm": 0.07435746490955353, + "learning_rate": 4.93478799984322e-06, + "loss": 0.11164016723632812, + "step": 540 + }, + { + "epoch": 0.08979591836734693, + "grad_norm": 0.16348940134048462, + "learning_rate": 4.9317969641525154e-06, + "loss": 0.10738091468811035, + "step": 550 + }, + { + "epoch": 0.09142857142857143, + "grad_norm": 0.12221129983663559, + "learning_rate": 4.928739811639741e-06, + "loss": 0.09491395950317383, + "step": 560 + }, + { + "epoch": 0.09306122448979592, + "grad_norm": 0.1107218861579895, + "learning_rate": 4.9256166254241775e-06, + "loss": 0.08408294320106506, + "step": 570 + }, + { + "epoch": 0.0946938775510204, + "grad_norm": 0.207329660654068, + "learning_rate": 4.922427490420457e-06, + "loss": 0.10272791385650634, + "step": 580 + }, + { + "epoch": 0.0963265306122449, + "grad_norm": 0.14269684255123138, + "learning_rate": 4.919172493336259e-06, + "loss": 0.10189608335494996, + "step": 590 + }, + { + "epoch": 0.09795918367346938, + "grad_norm": 0.10970286279916763, + "learning_rate": 4.915851722669952e-06, + "loss": 0.09506216645240784, + "step": 600 + }, + { + "epoch": 0.09795918367346938, + "eval_accuracy": 0.9561666666666667, + "eval_loss": 0.09913250058889389, + "eval_runtime": 158.3659, + "eval_samples_per_second": 25.258, + "eval_steps_per_second": 25.258, + "step": 600 + }, + { + "epoch": 0.09959183673469388, + "grad_norm": 0.11382903158664703, + "learning_rate": 4.912465268708186e-06, + "loss": 0.09665942788124085, + "step": 610 + }, + { + "epoch": 0.10122448979591837, + "grad_norm": 0.24785904586315155, + "learning_rate": 4.90901322352344e-06, + "loss": 0.10539349317550659, + "step": 620 + }, + { + "epoch": 0.10285714285714286, + "grad_norm": 0.18585072457790375, + "learning_rate": 4.905495680971518e-06, + "loss": 0.10907092094421386, + "step": 630 + }, + { + "epoch": 0.10448979591836735, + "grad_norm": 0.16385921835899353, + "learning_rate": 4.901912736688995e-06, + "loss": 0.09973281621932983, + "step": 640 + }, + { + "epoch": 0.10612244897959183, + "grad_norm": 0.10820458829402924, + "learning_rate": 4.898264488090619e-06, + "loss": 0.09869267344474793, + "step": 650 + }, + { + "epoch": 0.10775510204081633, + "grad_norm": 0.18020115792751312, + "learning_rate": 4.894551034366665e-06, + "loss": 0.10499387979507446, + "step": 660 + }, + { + "epoch": 0.10938775510204081, + "grad_norm": 0.10356947779655457, + "learning_rate": 4.890772476480229e-06, + "loss": 0.10153095722198487, + "step": 670 + }, + { + "epoch": 0.1110204081632653, + "grad_norm": 0.08888879418373108, + "learning_rate": 4.886928917164496e-06, + "loss": 0.08613221645355225, + "step": 680 + }, + { + "epoch": 0.1126530612244898, + "grad_norm": 0.11978956311941147, + "learning_rate": 4.8830204609199335e-06, + "loss": 0.09545036554336547, + "step": 690 + }, + { + "epoch": 0.11428571428571428, + "grad_norm": 0.11924679577350616, + "learning_rate": 4.8790472140114635e-06, + "loss": 0.0950542688369751, + "step": 700 + }, + { + "epoch": 0.11591836734693878, + "grad_norm": 0.0637463852763176, + "learning_rate": 4.875009284465561e-06, + "loss": 0.10127921104431152, + "step": 710 + }, + { + "epoch": 0.11755102040816326, + "grad_norm": 0.0894392654299736, + "learning_rate": 4.8709067820673235e-06, + "loss": 0.09835844039916992, + "step": 720 + }, + { + "epoch": 0.11918367346938775, + "grad_norm": 0.06312194466590881, + "learning_rate": 4.866739818357489e-06, + "loss": 0.08583117127418519, + "step": 730 + }, + { + "epoch": 0.12081632653061225, + "grad_norm": 0.1375141739845276, + "learning_rate": 4.862508506629394e-06, + "loss": 0.08735222220420838, + "step": 740 + }, + { + "epoch": 0.12244897959183673, + "grad_norm": 0.12715794146060944, + "learning_rate": 4.858212961925902e-06, + "loss": 0.09324318170547485, + "step": 750 + }, + { + "epoch": 0.12408163265306123, + "grad_norm": 0.14359286427497864, + "learning_rate": 4.8538533010362735e-06, + "loss": 0.09482367634773255, + "step": 760 + }, + { + "epoch": 0.12571428571428572, + "grad_norm": 0.1228424683213234, + "learning_rate": 4.849429642492985e-06, + "loss": 0.09595230221748352, + "step": 770 + }, + { + "epoch": 0.1273469387755102, + "grad_norm": 0.08908221125602722, + "learning_rate": 4.844942106568517e-06, + "loss": 0.08924583196640015, + "step": 780 + }, + { + "epoch": 0.12897959183673469, + "grad_norm": 0.15417414903640747, + "learning_rate": 4.84039081527207e-06, + "loss": 0.09454164505004883, + "step": 790 + }, + { + "epoch": 0.1306122448979592, + "grad_norm": 0.1259608417749405, + "learning_rate": 4.835775892346264e-06, + "loss": 0.08659166097640991, + "step": 800 + }, + { + "epoch": 0.1306122448979592, + "eval_accuracy": 0.9564166666666667, + "eval_loss": 0.09297937154769897, + "eval_runtime": 157.5491, + "eval_samples_per_second": 25.389, + "eval_steps_per_second": 25.389, + "step": 800 + }, + { + "epoch": 0.13224489795918368, + "grad_norm": 0.15664513409137726, + "learning_rate": 4.8310974632637584e-06, + "loss": 0.09094155430793763, + "step": 810 + }, + { + "epoch": 0.13387755102040816, + "grad_norm": 0.12366823107004166, + "learning_rate": 4.82635565522385e-06, + "loss": 0.09823386669158936, + "step": 820 + }, + { + "epoch": 0.13551020408163264, + "grad_norm": 0.09367755055427551, + "learning_rate": 4.821550597149012e-06, + "loss": 0.09080367088317871, + "step": 830 + }, + { + "epoch": 0.13714285714285715, + "grad_norm": 0.15230220556259155, + "learning_rate": 4.816682419681389e-06, + "loss": 0.09046205282211303, + "step": 840 + }, + { + "epoch": 0.13877551020408163, + "grad_norm": 0.09214162826538086, + "learning_rate": 4.811751255179244e-06, + "loss": 0.09502801299095154, + "step": 850 + }, + { + "epoch": 0.1404081632653061, + "grad_norm": 0.09514431655406952, + "learning_rate": 4.806757237713361e-06, + "loss": 0.09261927604675294, + "step": 860 + }, + { + "epoch": 0.14204081632653062, + "grad_norm": 0.13052736222743988, + "learning_rate": 4.8017005030634e-06, + "loss": 0.08243635892868043, + "step": 870 + }, + { + "epoch": 0.1436734693877551, + "grad_norm": 0.05511025711894035, + "learning_rate": 4.796581188714205e-06, + "loss": 0.10203326940536499, + "step": 880 + }, + { + "epoch": 0.14530612244897959, + "grad_norm": 0.09896261245012283, + "learning_rate": 4.7913994338520655e-06, + "loss": 0.08544076681137085, + "step": 890 + }, + { + "epoch": 0.1469387755102041, + "grad_norm": 0.18726249039173126, + "learning_rate": 4.786155379360932e-06, + "loss": 0.0944068193435669, + "step": 900 + }, + { + "epoch": 0.14857142857142858, + "grad_norm": 0.09349548071622849, + "learning_rate": 4.7808491678185855e-06, + "loss": 0.0877715528011322, + "step": 910 + }, + { + "epoch": 0.15020408163265306, + "grad_norm": 0.13616853952407837, + "learning_rate": 4.775480943492764e-06, + "loss": 0.09196346998214722, + "step": 920 + }, + { + "epoch": 0.15183673469387754, + "grad_norm": 0.061041463166475296, + "learning_rate": 4.770050852337236e-06, + "loss": 0.09890784621238709, + "step": 930 + }, + { + "epoch": 0.15346938775510205, + "grad_norm": 0.09928455948829651, + "learning_rate": 4.764559041987836e-06, + "loss": 0.09260086417198181, + "step": 940 + }, + { + "epoch": 0.15510204081632653, + "grad_norm": 0.10996410250663757, + "learning_rate": 4.759005661758446e-06, + "loss": 0.09412651658058166, + "step": 950 + }, + { + "epoch": 0.156734693877551, + "grad_norm": 0.09731809049844742, + "learning_rate": 4.753390862636939e-06, + "loss": 0.09459404945373535, + "step": 960 + }, + { + "epoch": 0.15836734693877552, + "grad_norm": 0.12398040294647217, + "learning_rate": 4.747714797281076e-06, + "loss": 0.09960640072822571, + "step": 970 + }, + { + "epoch": 0.16, + "grad_norm": 0.14135490357875824, + "learning_rate": 4.741977620014347e-06, + "loss": 0.09338371157646179, + "step": 980 + }, + { + "epoch": 0.16163265306122448, + "grad_norm": 0.09384343028068542, + "learning_rate": 4.736179486821788e-06, + "loss": 0.08515415191650391, + "step": 990 + }, + { + "epoch": 0.16326530612244897, + "grad_norm": 0.10298261791467667, + "learning_rate": 4.730320555345728e-06, + "loss": 0.09090787768363953, + "step": 1000 + }, + { + "epoch": 0.16326530612244897, + "eval_accuracy": 0.9607916666666667, + "eval_loss": 0.08832322806119919, + "eval_runtime": 166.8275, + "eval_samples_per_second": 23.977, + "eval_steps_per_second": 23.977, + "step": 1000 + }, + { + "epoch": 0.16489795918367348, + "grad_norm": 0.05896759778261185, + "learning_rate": 4.724400984881508e-06, + "loss": 0.08005568385124207, + "step": 1010 + }, + { + "epoch": 0.16653061224489796, + "grad_norm": 0.07907038927078247, + "learning_rate": 4.7184209363731546e-06, + "loss": 0.08133251667022705, + "step": 1020 + }, + { + "epoch": 0.16816326530612244, + "grad_norm": 0.07444760203361511, + "learning_rate": 4.7123805724089935e-06, + "loss": 0.08690541982650757, + "step": 1030 + }, + { + "epoch": 0.16979591836734695, + "grad_norm": 0.16517433524131775, + "learning_rate": 4.706280057217239e-06, + "loss": 0.09946454763412475, + "step": 1040 + }, + { + "epoch": 0.17142857142857143, + "grad_norm": 0.18107500672340393, + "learning_rate": 4.700119556661521e-06, + "loss": 0.08886862993240356, + "step": 1050 + }, + { + "epoch": 0.1730612244897959, + "grad_norm": 0.15386350452899933, + "learning_rate": 4.693899238236383e-06, + "loss": 0.09117826819419861, + "step": 1060 + }, + { + "epoch": 0.17469387755102042, + "grad_norm": 0.1098460704088211, + "learning_rate": 4.687619271062725e-06, + "loss": 0.09286551475524903, + "step": 1070 + }, + { + "epoch": 0.1763265306122449, + "grad_norm": 0.09872590750455856, + "learning_rate": 4.681279825883201e-06, + "loss": 0.09102073907852173, + "step": 1080 + }, + { + "epoch": 0.17795918367346938, + "grad_norm": 0.12001907080411911, + "learning_rate": 4.674881075057581e-06, + "loss": 0.09481253027915955, + "step": 1090 + }, + { + "epoch": 0.17959183673469387, + "grad_norm": 0.13000014424324036, + "learning_rate": 4.668423192558069e-06, + "loss": 0.08999797105789184, + "step": 1100 + }, + { + "epoch": 0.18122448979591838, + "grad_norm": 0.07537361979484558, + "learning_rate": 4.661906353964563e-06, + "loss": 0.07345216870307922, + "step": 1110 + }, + { + "epoch": 0.18285714285714286, + "grad_norm": 0.10346294194459915, + "learning_rate": 4.65533073645989e-06, + "loss": 0.08096543550491334, + "step": 1120 + }, + { + "epoch": 0.18448979591836734, + "grad_norm": 0.12576258182525635, + "learning_rate": 4.6486965188249825e-06, + "loss": 0.08680164813995361, + "step": 1130 + }, + { + "epoch": 0.18612244897959185, + "grad_norm": 0.17314419150352478, + "learning_rate": 4.642003881434024e-06, + "loss": 0.08470291495323182, + "step": 1140 + }, + { + "epoch": 0.18775510204081633, + "grad_norm": 0.1576225906610489, + "learning_rate": 4.635253006249536e-06, + "loss": 0.10852066278457642, + "step": 1150 + }, + { + "epoch": 0.1893877551020408, + "grad_norm": 0.15372557938098907, + "learning_rate": 4.628444076817444e-06, + "loss": 0.1022921085357666, + "step": 1160 + }, + { + "epoch": 0.1910204081632653, + "grad_norm": 0.15021173655986786, + "learning_rate": 4.621577278262072e-06, + "loss": 0.09645735621452331, + "step": 1170 + }, + { + "epoch": 0.1926530612244898, + "grad_norm": 0.07714739441871643, + "learning_rate": 4.614652797281118e-06, + "loss": 0.08636590242385864, + "step": 1180 + }, + { + "epoch": 0.19428571428571428, + "grad_norm": 0.17956240475177765, + "learning_rate": 4.6076708221405806e-06, + "loss": 0.09293715357780456, + "step": 1190 + }, + { + "epoch": 0.19591836734693877, + "grad_norm": 0.13412988185882568, + "learning_rate": 4.600631542669633e-06, + "loss": 0.09799563884735107, + "step": 1200 + }, + { + "epoch": 0.19591836734693877, + "eval_accuracy": 0.963, + "eval_loss": 0.08563050627708435, + "eval_runtime": 171.1854, + "eval_samples_per_second": 23.366, + "eval_steps_per_second": 23.366, + "step": 1200 + }, + { + "epoch": 0.19755102040816327, + "grad_norm": 0.1682359278202057, + "learning_rate": 4.593535150255463e-06, + "loss": 0.08345703482627868, + "step": 1210 + }, + { + "epoch": 0.19918367346938776, + "grad_norm": 0.09281504899263382, + "learning_rate": 4.586381837838076e-06, + "loss": 0.08114275336265564, + "step": 1220 + }, + { + "epoch": 0.20081632653061224, + "grad_norm": 0.12355420738458633, + "learning_rate": 4.579171799905044e-06, + "loss": 0.0780484676361084, + "step": 1230 + }, + { + "epoch": 0.20244897959183675, + "grad_norm": 0.12121636420488358, + "learning_rate": 4.571905232486214e-06, + "loss": 0.07816853523254394, + "step": 1240 + }, + { + "epoch": 0.20408163265306123, + "grad_norm": 0.09326840937137604, + "learning_rate": 4.56458233314839e-06, + "loss": 0.08001847267150879, + "step": 1250 + }, + { + "epoch": 0.2057142857142857, + "grad_norm": 0.11423090845346451, + "learning_rate": 4.557203300989948e-06, + "loss": 0.08720387816429138, + "step": 1260 + }, + { + "epoch": 0.2073469387755102, + "grad_norm": 0.124808169901371, + "learning_rate": 4.549768336635432e-06, + "loss": 0.08812607526779175, + "step": 1270 + }, + { + "epoch": 0.2089795918367347, + "grad_norm": 0.08311110734939575, + "learning_rate": 4.542277642230096e-06, + "loss": 0.0880801498889923, + "step": 1280 + }, + { + "epoch": 0.21061224489795918, + "grad_norm": 0.1179559975862503, + "learning_rate": 4.53473142143441e-06, + "loss": 0.09044411182403564, + "step": 1290 + }, + { + "epoch": 0.21224489795918366, + "grad_norm": 0.13122901320457458, + "learning_rate": 4.527129879418519e-06, + "loss": 0.06792758107185363, + "step": 1300 + }, + { + "epoch": 0.21387755102040817, + "grad_norm": 0.10990814864635468, + "learning_rate": 4.519473222856667e-06, + "loss": 0.07861824035644531, + "step": 1310 + }, + { + "epoch": 0.21551020408163266, + "grad_norm": 0.14613105356693268, + "learning_rate": 4.51176165992158e-06, + "loss": 0.08266589641571045, + "step": 1320 + }, + { + "epoch": 0.21714285714285714, + "grad_norm": 0.07217947393655777, + "learning_rate": 4.503995400278801e-06, + "loss": 0.08272708654403686, + "step": 1330 + }, + { + "epoch": 0.21877551020408162, + "grad_norm": 0.12690335512161255, + "learning_rate": 4.4961746550809945e-06, + "loss": 0.08644292950630188, + "step": 1340 + }, + { + "epoch": 0.22040816326530613, + "grad_norm": 0.11895330995321274, + "learning_rate": 4.488299636962203e-06, + "loss": 0.08812489509582519, + "step": 1350 + }, + { + "epoch": 0.2220408163265306, + "grad_norm": 0.08727742731571198, + "learning_rate": 4.4803705600320655e-06, + "loss": 0.08752686977386474, + "step": 1360 + }, + { + "epoch": 0.2236734693877551, + "grad_norm": 0.1101253554224968, + "learning_rate": 4.472387639869998e-06, + "loss": 0.08365487456321716, + "step": 1370 + }, + { + "epoch": 0.2253061224489796, + "grad_norm": 0.0881471335887909, + "learning_rate": 4.464351093519332e-06, + "loss": 0.08101181387901306, + "step": 1380 + }, + { + "epoch": 0.22693877551020408, + "grad_norm": 0.1676328480243683, + "learning_rate": 4.45626113948141e-06, + "loss": 0.08873040080070496, + "step": 1390 + }, + { + "epoch": 0.22857142857142856, + "grad_norm": 0.13493730127811432, + "learning_rate": 4.44811799770965e-06, + "loss": 0.08360225558280945, + "step": 1400 + }, + { + "epoch": 0.22857142857142856, + "eval_accuracy": 0.961125, + "eval_loss": 0.0838993564248085, + "eval_runtime": 159.3888, + "eval_samples_per_second": 25.096, + "eval_steps_per_second": 25.096, + "step": 1400 + }, + { + "epoch": 0.23020408163265307, + "grad_norm": 0.18087805807590485, + "learning_rate": 4.4399218896035615e-06, + "loss": 0.07942168116569519, + "step": 1410 + }, + { + "epoch": 0.23183673469387756, + "grad_norm": 0.15980401635169983, + "learning_rate": 4.431673038002726e-06, + "loss": 0.07959137558937072, + "step": 1420 + }, + { + "epoch": 0.23346938775510204, + "grad_norm": 0.09756626188755035, + "learning_rate": 4.423371667180743e-06, + "loss": 0.08047830462455749, + "step": 1430 + }, + { + "epoch": 0.23510204081632652, + "grad_norm": 0.11922340840101242, + "learning_rate": 4.415018002839127e-06, + "loss": 0.07832802534103393, + "step": 1440 + }, + { + "epoch": 0.23673469387755103, + "grad_norm": 0.09443783015012741, + "learning_rate": 4.40661227210117e-06, + "loss": 0.07132846713066102, + "step": 1450 + }, + { + "epoch": 0.2383673469387755, + "grad_norm": 0.15397194027900696, + "learning_rate": 4.398154703505776e-06, + "loss": 0.09222683906555176, + "step": 1460 + }, + { + "epoch": 0.24, + "grad_norm": 0.06159692630171776, + "learning_rate": 4.389645527001234e-06, + "loss": 0.07937920093536377, + "step": 1470 + }, + { + "epoch": 0.2416326530612245, + "grad_norm": 0.11855911463499069, + "learning_rate": 4.381084973938976e-06, + "loss": 0.07650941014289855, + "step": 1480 + }, + { + "epoch": 0.24326530612244898, + "grad_norm": 0.1305486559867859, + "learning_rate": 4.372473277067285e-06, + "loss": 0.07730157375335693, + "step": 1490 + }, + { + "epoch": 0.24489795918367346, + "grad_norm": 0.09036920964717865, + "learning_rate": 4.363810670524962e-06, + "loss": 0.07192737460136414, + "step": 1500 + }, + { + "epoch": 0.24653061224489795, + "grad_norm": 0.0936962217092514, + "learning_rate": 4.355097389834966e-06, + "loss": 0.09040454030036926, + "step": 1510 + }, + { + "epoch": 0.24816326530612245, + "grad_norm": 0.0778910219669342, + "learning_rate": 4.346333671898007e-06, + "loss": 0.06948431730270385, + "step": 1520 + }, + { + "epoch": 0.24979591836734694, + "grad_norm": 0.0979679748415947, + "learning_rate": 4.337519754986108e-06, + "loss": 0.07317689657211304, + "step": 1530 + }, + { + "epoch": 0.25142857142857145, + "grad_norm": 0.15620048344135284, + "learning_rate": 4.328655878736123e-06, + "loss": 0.09126524925231934, + "step": 1540 + }, + { + "epoch": 0.2530612244897959, + "grad_norm": 0.09080824255943298, + "learning_rate": 4.319742284143226e-06, + "loss": 0.09065211415290833, + "step": 1550 + }, + { + "epoch": 0.2546938775510204, + "grad_norm": 0.046968940645456314, + "learning_rate": 4.310779213554353e-06, + "loss": 0.0718291699886322, + "step": 1560 + }, + { + "epoch": 0.2563265306122449, + "grad_norm": 0.11479081958532333, + "learning_rate": 4.301766910661619e-06, + "loss": 0.07913097143173217, + "step": 1570 + }, + { + "epoch": 0.25795918367346937, + "grad_norm": 0.1042279377579689, + "learning_rate": 4.292705620495689e-06, + "loss": 0.06828063130378723, + "step": 1580 + }, + { + "epoch": 0.2595918367346939, + "grad_norm": 0.17096509039402008, + "learning_rate": 4.283595589419115e-06, + "loss": 0.07042163014411926, + "step": 1590 + }, + { + "epoch": 0.2612244897959184, + "grad_norm": 0.10892639309167862, + "learning_rate": 4.274437065119643e-06, + "loss": 0.07090482115745544, + "step": 1600 + }, + { + "epoch": 0.2612244897959184, + "eval_accuracy": 0.964375, + "eval_loss": 0.08112655580043793, + "eval_runtime": 158.0164, + "eval_samples_per_second": 25.314, + "eval_steps_per_second": 25.314, + "step": 1600 + }, + { + "epoch": 0.26285714285714284, + "grad_norm": 0.09731441736221313, + "learning_rate": 4.265230296603471e-06, + "loss": 0.08803416490554809, + "step": 1610 + }, + { + "epoch": 0.26448979591836735, + "grad_norm": 0.0956069752573967, + "learning_rate": 4.255975534188488e-06, + "loss": 0.0734250545501709, + "step": 1620 + }, + { + "epoch": 0.26612244897959186, + "grad_norm": 0.14361435174942017, + "learning_rate": 4.246673029497458e-06, + "loss": 0.07703657150268554, + "step": 1630 + }, + { + "epoch": 0.2677551020408163, + "grad_norm": 0.08011971414089203, + "learning_rate": 4.237323035451188e-06, + "loss": 0.08410758376121522, + "step": 1640 + }, + { + "epoch": 0.2693877551020408, + "grad_norm": 0.1244175136089325, + "learning_rate": 4.2279258062616455e-06, + "loss": 0.0808764636516571, + "step": 1650 + }, + { + "epoch": 0.2710204081632653, + "grad_norm": 0.11162611097097397, + "learning_rate": 4.21848159742505e-06, + "loss": 0.07636536359786987, + "step": 1660 + }, + { + "epoch": 0.2726530612244898, + "grad_norm": 0.16819386184215546, + "learning_rate": 4.208990665714927e-06, + "loss": 0.07789689898490906, + "step": 1670 + }, + { + "epoch": 0.2742857142857143, + "grad_norm": 0.07522564381361008, + "learning_rate": 4.19945326917512e-06, + "loss": 0.07442367672920228, + "step": 1680 + }, + { + "epoch": 0.27591836734693875, + "grad_norm": 0.08352494984865189, + "learning_rate": 4.1898696671127855e-06, + "loss": 0.07646024227142334, + "step": 1690 + }, + { + "epoch": 0.27755102040816326, + "grad_norm": 0.12580762803554535, + "learning_rate": 4.180240120091333e-06, + "loss": 0.07484800815582275, + "step": 1700 + }, + { + "epoch": 0.27918367346938777, + "grad_norm": 0.14690619707107544, + "learning_rate": 4.170564889923348e-06, + "loss": 0.07452926635742188, + "step": 1710 + }, + { + "epoch": 0.2808163265306122, + "grad_norm": 0.08760543167591095, + "learning_rate": 4.1608442396634675e-06, + "loss": 0.07579211592674255, + "step": 1720 + }, + { + "epoch": 0.28244897959183674, + "grad_norm": 0.05263754725456238, + "learning_rate": 4.151078433601234e-06, + "loss": 0.08277950882911682, + "step": 1730 + }, + { + "epoch": 0.28408163265306124, + "grad_norm": 0.12929993867874146, + "learning_rate": 4.141267737253905e-06, + "loss": 0.07354526519775391, + "step": 1740 + }, + { + "epoch": 0.2857142857142857, + "grad_norm": 0.1444639414548874, + "learning_rate": 4.1314124173592356e-06, + "loss": 0.07377716898918152, + "step": 1750 + }, + { + "epoch": 0.2873469387755102, + "grad_norm": 0.15250442922115326, + "learning_rate": 4.121512741868229e-06, + "loss": 0.0802634596824646, + "step": 1760 + }, + { + "epoch": 0.2889795918367347, + "grad_norm": 0.14867356419563293, + "learning_rate": 4.111568979937845e-06, + "loss": 0.08141970038414001, + "step": 1770 + }, + { + "epoch": 0.29061224489795917, + "grad_norm": 0.1100233644247055, + "learning_rate": 4.10158140192369e-06, + "loss": 0.07370007634162903, + "step": 1780 + }, + { + "epoch": 0.2922448979591837, + "grad_norm": 0.10663364082574844, + "learning_rate": 4.091550279372659e-06, + "loss": 0.0754891276359558, + "step": 1790 + }, + { + "epoch": 0.2938775510204082, + "grad_norm": 0.07148909568786621, + "learning_rate": 4.081475885015558e-06, + "loss": 0.07567381858825684, + "step": 1800 + }, + { + "epoch": 0.2938775510204082, + "eval_accuracy": 0.9665416666666667, + "eval_loss": 0.07648575305938721, + "eval_runtime": 176.7261, + "eval_samples_per_second": 22.634, + "eval_steps_per_second": 22.634, + "step": 1800 + }, + { + "epoch": 0.29551020408163264, + "grad_norm": 0.07446923106908798, + "learning_rate": 4.071358492759686e-06, + "loss": 0.06706757545471191, + "step": 1810 + }, + { + "epoch": 0.29714285714285715, + "grad_norm": 0.144475519657135, + "learning_rate": 4.061198377681391e-06, + "loss": 0.07407833337783813, + "step": 1820 + }, + { + "epoch": 0.2987755102040816, + "grad_norm": 0.16540028154850006, + "learning_rate": 4.050995816018585e-06, + "loss": 0.07593105435371399, + "step": 1830 + }, + { + "epoch": 0.3004081632653061, + "grad_norm": 0.09237829595804214, + "learning_rate": 4.040751085163239e-06, + "loss": 0.08375502228736878, + "step": 1840 + }, + { + "epoch": 0.3020408163265306, + "grad_norm": 0.11202957481145859, + "learning_rate": 4.0304644636538435e-06, + "loss": 0.0692280650138855, + "step": 1850 + }, + { + "epoch": 0.3036734693877551, + "grad_norm": 0.16075952351093292, + "learning_rate": 4.020136231167824e-06, + "loss": 0.07824563384056091, + "step": 1860 + }, + { + "epoch": 0.3053061224489796, + "grad_norm": 0.12567447125911713, + "learning_rate": 4.0097666685139516e-06, + "loss": 0.07255304455757142, + "step": 1870 + }, + { + "epoch": 0.3069387755102041, + "grad_norm": 0.08912669867277145, + "learning_rate": 3.999356057624696e-06, + "loss": 0.06304961442947388, + "step": 1880 + }, + { + "epoch": 0.30857142857142855, + "grad_norm": 0.15736177563667297, + "learning_rate": 3.988904681548568e-06, + "loss": 0.07315694689750671, + "step": 1890 + }, + { + "epoch": 0.31020408163265306, + "grad_norm": 0.13224704563617706, + "learning_rate": 3.978412824442421e-06, + "loss": 0.07607222199440003, + "step": 1900 + }, + { + "epoch": 0.31183673469387757, + "grad_norm": 0.0942244678735733, + "learning_rate": 3.9678807715637245e-06, + "loss": 0.06320942640304565, + "step": 1910 + }, + { + "epoch": 0.313469387755102, + "grad_norm": 0.1153193786740303, + "learning_rate": 3.95730880926281e-06, + "loss": 0.07607839703559875, + "step": 1920 + }, + { + "epoch": 0.31510204081632653, + "grad_norm": 0.24464164674282074, + "learning_rate": 3.946697224975083e-06, + "loss": 0.07458690404891968, + "step": 1930 + }, + { + "epoch": 0.31673469387755104, + "grad_norm": 0.14584548771381378, + "learning_rate": 3.936046307213215e-06, + "loss": 0.07876120805740357, + "step": 1940 + }, + { + "epoch": 0.3183673469387755, + "grad_norm": 0.0880698636174202, + "learning_rate": 3.925356345559288e-06, + "loss": 0.07082821726799012, + "step": 1950 + }, + { + "epoch": 0.32, + "grad_norm": 0.08619305491447449, + "learning_rate": 3.914627630656931e-06, + "loss": 0.06648402214050293, + "step": 1960 + }, + { + "epoch": 0.3216326530612245, + "grad_norm": 0.1341329663991928, + "learning_rate": 3.903860454203413e-06, + "loss": 0.07000230550765991, + "step": 1970 + }, + { + "epoch": 0.32326530612244897, + "grad_norm": 0.1504107564687729, + "learning_rate": 3.8930551089417145e-06, + "loss": 0.06887786984443664, + "step": 1980 + }, + { + "epoch": 0.3248979591836735, + "grad_norm": 0.1262267827987671, + "learning_rate": 3.882211888652566e-06, + "loss": 0.08991954326629639, + "step": 1990 + }, + { + "epoch": 0.32653061224489793, + "grad_norm": 0.2052707076072693, + "learning_rate": 3.871331088146463e-06, + "loss": 0.07011507153511047, + "step": 2000 + }, + { + "epoch": 0.32653061224489793, + "eval_accuracy": 0.9654583333333334, + "eval_loss": 0.07913381606340408, + "eval_runtime": 161.4221, + "eval_samples_per_second": 24.78, + "eval_steps_per_second": 24.78, + "step": 2000 + }, + { + "epoch": 0.32816326530612244, + "grad_norm": 0.11643362790346146, + "learning_rate": 3.860413003255648e-06, + "loss": 0.08266008496284485, + "step": 2010 + }, + { + "epoch": 0.32979591836734695, + "grad_norm": 0.1472826600074768, + "learning_rate": 3.8494579308260695e-06, + "loss": 0.07043538689613342, + "step": 2020 + }, + { + "epoch": 0.3314285714285714, + "grad_norm": 0.10538945347070694, + "learning_rate": 3.838466168709312e-06, + "loss": 0.06780247688293457, + "step": 2030 + }, + { + "epoch": 0.3330612244897959, + "grad_norm": 0.14366035163402557, + "learning_rate": 3.827438015754491e-06, + "loss": 0.06872215867042542, + "step": 2040 + }, + { + "epoch": 0.3346938775510204, + "grad_norm": 0.11022552847862244, + "learning_rate": 3.81637377180014e-06, + "loss": 0.06861716508865356, + "step": 2050 + }, + { + "epoch": 0.3363265306122449, + "grad_norm": 0.07990925014019012, + "learning_rate": 3.8052737376660446e-06, + "loss": 0.07096996307373046, + "step": 2060 + }, + { + "epoch": 0.3379591836734694, + "grad_norm": 0.11924420297145844, + "learning_rate": 3.7941382151450756e-06, + "loss": 0.07092551589012146, + "step": 2070 + }, + { + "epoch": 0.3395918367346939, + "grad_norm": 0.08096475899219513, + "learning_rate": 3.782967506994975e-06, + "loss": 0.06780472993850709, + "step": 2080 + }, + { + "epoch": 0.34122448979591835, + "grad_norm": 0.14714907109737396, + "learning_rate": 3.7717619169301302e-06, + "loss": 0.07661795020103454, + "step": 2090 + }, + { + "epoch": 0.34285714285714286, + "grad_norm": 0.14839571714401245, + "learning_rate": 3.7605217496133127e-06, + "loss": 0.07939797639846802, + "step": 2100 + }, + { + "epoch": 0.34448979591836737, + "grad_norm": 0.15081271529197693, + "learning_rate": 3.749247310647398e-06, + "loss": 0.07869908213615417, + "step": 2110 + }, + { + "epoch": 0.3461224489795918, + "grad_norm": 0.13346275687217712, + "learning_rate": 3.737938906567053e-06, + "loss": 0.07526676058769226, + "step": 2120 + }, + { + "epoch": 0.34775510204081633, + "grad_norm": 0.112598717212677, + "learning_rate": 3.7265968448304057e-06, + "loss": 0.07366414070129394, + "step": 2130 + }, + { + "epoch": 0.34938775510204084, + "grad_norm": 0.12423531711101532, + "learning_rate": 3.7152214338106827e-06, + "loss": 0.06439908146858216, + "step": 2140 + }, + { + "epoch": 0.3510204081632653, + "grad_norm": 0.21758921444416046, + "learning_rate": 3.703812982787828e-06, + "loss": 0.08098108768463134, + "step": 2150 + }, + { + "epoch": 0.3526530612244898, + "grad_norm": 0.1298121213912964, + "learning_rate": 3.6923718019400906e-06, + "loss": 0.08134381771087647, + "step": 2160 + }, + { + "epoch": 0.35428571428571426, + "grad_norm": 0.09018377214670181, + "learning_rate": 3.6808982023355934e-06, + "loss": 0.07070590257644653, + "step": 2170 + }, + { + "epoch": 0.35591836734693877, + "grad_norm": 0.1329157054424286, + "learning_rate": 3.6693924959238774e-06, + "loss": 0.06200396418571472, + "step": 2180 + }, + { + "epoch": 0.3575510204081633, + "grad_norm": 0.08813945204019547, + "learning_rate": 3.657854995527416e-06, + "loss": 0.06886230111122131, + "step": 2190 + }, + { + "epoch": 0.35918367346938773, + "grad_norm": 0.13459829986095428, + "learning_rate": 3.646286014833114e-06, + "loss": 0.07170915603637695, + "step": 2200 + }, + { + "epoch": 0.35918367346938773, + "eval_accuracy": 0.970625, + "eval_loss": 0.06810569018125534, + "eval_runtime": 157.4227, + "eval_samples_per_second": 25.409, + "eval_steps_per_second": 25.409, + "step": 2200 + }, + { + "epoch": 0.36081632653061224, + "grad_norm": 0.13405098021030426, + "learning_rate": 3.634685868383776e-06, + "loss": 0.07438129186630249, + "step": 2210 + }, + { + "epoch": 0.36244897959183675, + "grad_norm": 0.1342102289199829, + "learning_rate": 3.6230548715695555e-06, + "loss": 0.07625046968460084, + "step": 2220 + }, + { + "epoch": 0.3640816326530612, + "grad_norm": 0.1532502919435501, + "learning_rate": 3.611393340619381e-06, + "loss": 0.06719872951507569, + "step": 2230 + }, + { + "epoch": 0.3657142857142857, + "grad_norm": 0.11218152195215225, + "learning_rate": 3.5997015925923563e-06, + "loss": 0.07131683230400085, + "step": 2240 + }, + { + "epoch": 0.3673469387755102, + "grad_norm": 0.09200239181518555, + "learning_rate": 3.587979945369141e-06, + "loss": 0.06948224902153015, + "step": 2250 + }, + { + "epoch": 0.3689795918367347, + "grad_norm": 0.08099328726530075, + "learning_rate": 3.5762287176433097e-06, + "loss": 0.07263018488883972, + "step": 2260 + }, + { + "epoch": 0.3706122448979592, + "grad_norm": 0.09097494184970856, + "learning_rate": 3.564448228912682e-06, + "loss": 0.06066258549690247, + "step": 2270 + }, + { + "epoch": 0.3722448979591837, + "grad_norm": 0.07033897191286087, + "learning_rate": 3.5526387994706428e-06, + "loss": 0.07041062712669373, + "step": 2280 + }, + { + "epoch": 0.37387755102040815, + "grad_norm": 0.07659374922513962, + "learning_rate": 3.5408007503974283e-06, + "loss": 0.07272242307662964, + "step": 2290 + }, + { + "epoch": 0.37551020408163266, + "grad_norm": 0.0836353451013565, + "learning_rate": 3.5289344035513996e-06, + "loss": 0.07268807291984558, + "step": 2300 + }, + { + "epoch": 0.37714285714285717, + "grad_norm": 0.1426168978214264, + "learning_rate": 3.5170400815602894e-06, + "loss": 0.05777261257171631, + "step": 2310 + }, + { + "epoch": 0.3787755102040816, + "grad_norm": 0.07479120045900345, + "learning_rate": 3.5051181078124336e-06, + "loss": 0.06174240708351135, + "step": 2320 + }, + { + "epoch": 0.38040816326530613, + "grad_norm": 0.14598114788532257, + "learning_rate": 3.4931688064479745e-06, + "loss": 0.06616324782371522, + "step": 2330 + }, + { + "epoch": 0.3820408163265306, + "grad_norm": 0.22399863600730896, + "learning_rate": 3.4811925023500524e-06, + "loss": 0.06692227125167846, + "step": 2340 + }, + { + "epoch": 0.3836734693877551, + "grad_norm": 0.11600285023450851, + "learning_rate": 3.4691895211359693e-06, + "loss": 0.0634126603603363, + "step": 2350 + }, + { + "epoch": 0.3853061224489796, + "grad_norm": 0.0577060841023922, + "learning_rate": 3.457160189148337e-06, + "loss": 0.06884999275207519, + "step": 2360 + }, + { + "epoch": 0.38693877551020406, + "grad_norm": 0.09160040318965912, + "learning_rate": 3.445104833446205e-06, + "loss": 0.05986430644989014, + "step": 2370 + }, + { + "epoch": 0.38857142857142857, + "grad_norm": 0.1284518539905548, + "learning_rate": 3.433023781796167e-06, + "loss": 0.06759394407272339, + "step": 2380 + }, + { + "epoch": 0.3902040816326531, + "grad_norm": 0.08572858572006226, + "learning_rate": 3.4209173626634494e-06, + "loss": 0.06813116669654846, + "step": 2390 + }, + { + "epoch": 0.39183673469387753, + "grad_norm": 0.11550725251436234, + "learning_rate": 3.4087859052029837e-06, + "loss": 0.06371633410453796, + "step": 2400 + }, + { + "epoch": 0.39183673469387753, + "eval_accuracy": 0.9685833333333334, + "eval_loss": 0.07035119831562042, + "eval_runtime": 156.9729, + "eval_samples_per_second": 25.482, + "eval_steps_per_second": 25.482, + "step": 2400 + }, + { + "epoch": 0.39346938775510204, + "grad_norm": 0.22328943014144897, + "learning_rate": 3.3966297392504535e-06, + "loss": 0.06804299950599671, + "step": 2410 + }, + { + "epoch": 0.39510204081632655, + "grad_norm": 0.18518126010894775, + "learning_rate": 3.3844491953133272e-06, + "loss": 0.07678125500679016, + "step": 2420 + }, + { + "epoch": 0.396734693877551, + "grad_norm": 0.09818598628044128, + "learning_rate": 3.372244604561875e-06, + "loss": 0.06537411212921143, + "step": 2430 + }, + { + "epoch": 0.3983673469387755, + "grad_norm": 0.08255010843276978, + "learning_rate": 3.360016298820163e-06, + "loss": 0.05702770948410034, + "step": 2440 + }, + { + "epoch": 0.4, + "grad_norm": 0.10319415479898453, + "learning_rate": 3.3477646105570293e-06, + "loss": 0.06334798336029053, + "step": 2450 + }, + { + "epoch": 0.4016326530612245, + "grad_norm": 0.06523872911930084, + "learning_rate": 3.3354898728770503e-06, + "loss": 0.051438117027282716, + "step": 2460 + }, + { + "epoch": 0.403265306122449, + "grad_norm": 0.15883173048496246, + "learning_rate": 3.3231924195114774e-06, + "loss": 0.06427289247512817, + "step": 2470 + }, + { + "epoch": 0.4048979591836735, + "grad_norm": 0.1734374463558197, + "learning_rate": 3.3108725848091685e-06, + "loss": 0.06652372479438781, + "step": 2480 + }, + { + "epoch": 0.40653061224489795, + "grad_norm": 0.10557752102613449, + "learning_rate": 3.2985307037274937e-06, + "loss": 0.07374808788299561, + "step": 2490 + }, + { + "epoch": 0.40816326530612246, + "grad_norm": 0.12942200899124146, + "learning_rate": 3.286167111823234e-06, + "loss": 0.06349012851715088, + "step": 2500 + }, + { + "epoch": 0.4097959183673469, + "grad_norm": 0.1858324259519577, + "learning_rate": 3.2737821452434502e-06, + "loss": 0.06953533291816712, + "step": 2510 + }, + { + "epoch": 0.4114285714285714, + "grad_norm": 0.10682573169469833, + "learning_rate": 3.26137614071635e-06, + "loss": 0.05826004147529602, + "step": 2520 + }, + { + "epoch": 0.41306122448979593, + "grad_norm": 0.12533792853355408, + "learning_rate": 3.2489494355421293e-06, + "loss": 0.05837242007255554, + "step": 2530 + }, + { + "epoch": 0.4146938775510204, + "grad_norm": 0.10952792316675186, + "learning_rate": 3.2365023675838027e-06, + "loss": 0.06359312534332276, + "step": 2540 + }, + { + "epoch": 0.4163265306122449, + "grad_norm": 0.08605755865573883, + "learning_rate": 3.22403527525802e-06, + "loss": 0.07361164093017578, + "step": 2550 + }, + { + "epoch": 0.4179591836734694, + "grad_norm": 0.15942610800266266, + "learning_rate": 3.211548497525861e-06, + "loss": 0.05989238619804382, + "step": 2560 + }, + { + "epoch": 0.41959183673469386, + "grad_norm": 0.16340239346027374, + "learning_rate": 3.1990423738836206e-06, + "loss": 0.07593320608139038, + "step": 2570 + }, + { + "epoch": 0.42122448979591837, + "grad_norm": 0.11011330783367157, + "learning_rate": 3.1865172443535797e-06, + "loss": 0.06850495338439941, + "step": 2580 + }, + { + "epoch": 0.4228571428571429, + "grad_norm": 0.10529608279466629, + "learning_rate": 3.173973449474762e-06, + "loss": 0.056969082355499266, + "step": 2590 + }, + { + "epoch": 0.42448979591836733, + "grad_norm": 0.12615667283535004, + "learning_rate": 3.1614113302936716e-06, + "loss": 0.06189748048782349, + "step": 2600 + }, + { + "epoch": 0.42448979591836733, + "eval_accuracy": 0.9746666666666668, + "eval_loss": 0.0618240088224411, + "eval_runtime": 157.0858, + "eval_samples_per_second": 25.464, + "eval_steps_per_second": 25.464, + "step": 2600 + }, + { + "epoch": 0.42612244897959184, + "grad_norm": 0.1343226134777069, + "learning_rate": 3.1488312283550216e-06, + "loss": 0.061647194623947146, + "step": 2610 + }, + { + "epoch": 0.42775510204081635, + "grad_norm": 0.11777225136756897, + "learning_rate": 3.136233485692451e-06, + "loss": 0.05916886925697327, + "step": 2620 + }, + { + "epoch": 0.4293877551020408, + "grad_norm": 0.14721977710723877, + "learning_rate": 3.12361844481922e-06, + "loss": 0.059158307313919065, + "step": 2630 + }, + { + "epoch": 0.4310204081632653, + "grad_norm": 0.12912684679031372, + "learning_rate": 3.110986448718902e-06, + "loss": 0.05914759635925293, + "step": 2640 + }, + { + "epoch": 0.4326530612244898, + "grad_norm": 0.13003695011138916, + "learning_rate": 3.0983378408360597e-06, + "loss": 0.059078723192214966, + "step": 2650 + }, + { + "epoch": 0.4342857142857143, + "grad_norm": 0.12714017927646637, + "learning_rate": 3.085672965066899e-06, + "loss": 0.05707222819328308, + "step": 2660 + }, + { + "epoch": 0.4359183673469388, + "grad_norm": 0.13708427548408508, + "learning_rate": 3.0729921657499294e-06, + "loss": 0.05138048529624939, + "step": 2670 + }, + { + "epoch": 0.43755102040816324, + "grad_norm": 0.0949573740363121, + "learning_rate": 3.0602957876565943e-06, + "loss": 0.050283265113830564, + "step": 2680 + }, + { + "epoch": 0.43918367346938775, + "grad_norm": 0.08723263442516327, + "learning_rate": 3.0475841759819023e-06, + "loss": 0.06342648863792419, + "step": 2690 + }, + { + "epoch": 0.44081632653061226, + "grad_norm": 0.09944969415664673, + "learning_rate": 3.034857676335038e-06, + "loss": 0.06127093434333801, + "step": 2700 + }, + { + "epoch": 0.4424489795918367, + "grad_norm": 0.10576151311397552, + "learning_rate": 3.0221166347299673e-06, + "loss": 0.05412284731864929, + "step": 2710 + }, + { + "epoch": 0.4440816326530612, + "grad_norm": 0.1598891168832779, + "learning_rate": 3.009361397576031e-06, + "loss": 0.05579071640968323, + "step": 2720 + }, + { + "epoch": 0.44571428571428573, + "grad_norm": 0.16953758895397186, + "learning_rate": 2.9965923116685224e-06, + "loss": 0.06472888588905334, + "step": 2730 + }, + { + "epoch": 0.4473469387755102, + "grad_norm": 0.10883133113384247, + "learning_rate": 2.9838097241792646e-06, + "loss": 0.06263615489006043, + "step": 2740 + }, + { + "epoch": 0.4489795918367347, + "grad_norm": 0.08914396911859512, + "learning_rate": 2.9710139826471646e-06, + "loss": 0.0624904453754425, + "step": 2750 + }, + { + "epoch": 0.4506122448979592, + "grad_norm": 0.1059069111943245, + "learning_rate": 2.9582054349687706e-06, + "loss": 0.05488716959953308, + "step": 2760 + }, + { + "epoch": 0.45224489795918366, + "grad_norm": 0.09883642196655273, + "learning_rate": 2.9453844293888067e-06, + "loss": 0.06263139843940735, + "step": 2770 + }, + { + "epoch": 0.45387755102040817, + "grad_norm": 0.11156519502401352, + "learning_rate": 2.932551314490712e-06, + "loss": 0.06324111819267272, + "step": 2780 + }, + { + "epoch": 0.4555102040816327, + "grad_norm": 0.08826332539319992, + "learning_rate": 2.9197064391871565e-06, + "loss": 0.06384775042533875, + "step": 2790 + }, + { + "epoch": 0.45714285714285713, + "grad_norm": 0.0698089748620987, + "learning_rate": 2.9068501527105595e-06, + "loss": 0.05722553730010986, + "step": 2800 + }, + { + "epoch": 0.45714285714285713, + "eval_accuracy": 0.977375, + "eval_loss": 0.05622806400060654, + "eval_runtime": 170.7758, + "eval_samples_per_second": 23.423, + "eval_steps_per_second": 23.423, + "step": 2800 + }, + { + "epoch": 0.45877551020408164, + "grad_norm": 0.07031046599149704, + "learning_rate": 2.893982804603592e-06, + "loss": 0.060957270860672, + "step": 2810 + }, + { + "epoch": 0.46040816326530615, + "grad_norm": 0.08772005140781403, + "learning_rate": 2.8811047447096714e-06, + "loss": 0.05667365789413452, + "step": 2820 + }, + { + "epoch": 0.4620408163265306, + "grad_norm": 0.10329847782850266, + "learning_rate": 2.8682163231634557e-06, + "loss": 0.06144403219223023, + "step": 2830 + }, + { + "epoch": 0.4636734693877551, + "grad_norm": 0.06306952983140945, + "learning_rate": 2.8553178903813193e-06, + "loss": 0.04665087461471558, + "step": 2840 + }, + { + "epoch": 0.46530612244897956, + "grad_norm": 0.11146903783082962, + "learning_rate": 2.842409797051825e-06, + "loss": 0.06363347172737122, + "step": 2850 + }, + { + "epoch": 0.4669387755102041, + "grad_norm": 0.04504350945353508, + "learning_rate": 2.829492394126192e-06, + "loss": 0.05332474112510681, + "step": 2860 + }, + { + "epoch": 0.4685714285714286, + "grad_norm": 0.06039620563387871, + "learning_rate": 2.8165660328087546e-06, + "loss": 0.07094461917877197, + "step": 2870 + }, + { + "epoch": 0.47020408163265304, + "grad_norm": 0.08084474503993988, + "learning_rate": 2.8036310645474096e-06, + "loss": 0.058167701959609984, + "step": 2880 + }, + { + "epoch": 0.47183673469387755, + "grad_norm": 0.05352522060275078, + "learning_rate": 2.7906878410240655e-06, + "loss": 0.0486674964427948, + "step": 2890 + }, + { + "epoch": 0.47346938775510206, + "grad_norm": 0.09431054443120956, + "learning_rate": 2.777736714145077e-06, + "loss": 0.062025409936904904, + "step": 2900 + }, + { + "epoch": 0.4751020408163265, + "grad_norm": 0.09863542020320892, + "learning_rate": 2.7647780360316787e-06, + "loss": 0.061118441820144656, + "step": 2910 + }, + { + "epoch": 0.476734693877551, + "grad_norm": 0.082881860435009, + "learning_rate": 2.7518121590104145e-06, + "loss": 0.061808472871780394, + "step": 2920 + }, + { + "epoch": 0.47836734693877553, + "grad_norm": 0.06983684748411179, + "learning_rate": 2.7388394356035524e-06, + "loss": 0.05849882960319519, + "step": 2930 + }, + { + "epoch": 0.48, + "grad_norm": 0.10832090675830841, + "learning_rate": 2.7258602185195043e-06, + "loss": 0.06612529158592224, + "step": 2940 + }, + { + "epoch": 0.4816326530612245, + "grad_norm": 0.1516750454902649, + "learning_rate": 2.7128748606432353e-06, + "loss": 0.058327901363372806, + "step": 2950 + }, + { + "epoch": 0.483265306122449, + "grad_norm": 0.14055661857128143, + "learning_rate": 2.699883715026669e-06, + "loss": 0.055737996101379396, + "step": 2960 + }, + { + "epoch": 0.48489795918367345, + "grad_norm": 0.0793638750910759, + "learning_rate": 2.686887134879089e-06, + "loss": 0.051851028203964235, + "step": 2970 + }, + { + "epoch": 0.48653061224489796, + "grad_norm": 0.07475428283214569, + "learning_rate": 2.6738854735575358e-06, + "loss": 0.06357839703559875, + "step": 2980 + }, + { + "epoch": 0.4881632653061225, + "grad_norm": 0.13335047662258148, + "learning_rate": 2.660879084557198e-06, + "loss": 0.06160399913787842, + "step": 2990 + }, + { + "epoch": 0.4897959183673469, + "grad_norm": 0.12583893537521362, + "learning_rate": 2.6478683215018025e-06, + "loss": 0.060502588748931885, + "step": 3000 + }, + { + "epoch": 0.4897959183673469, + "eval_accuracy": 0.9767083333333334, + "eval_loss": 0.0561000294983387, + "eval_runtime": 156.4202, + "eval_samples_per_second": 25.572, + "eval_steps_per_second": 25.572, + "step": 3000 + }, + { + "epoch": 0.49142857142857144, + "grad_norm": 0.1411438137292862, + "learning_rate": 2.6348535381340025e-06, + "loss": 0.060459482669830325, + "step": 3010 + }, + { + "epoch": 0.4930612244897959, + "grad_norm": 0.12584605813026428, + "learning_rate": 2.6218350883057546e-06, + "loss": 0.0632048487663269, + "step": 3020 + }, + { + "epoch": 0.4946938775510204, + "grad_norm": 0.10522248595952988, + "learning_rate": 2.608813325968702e-06, + "loss": 0.04655081033706665, + "step": 3030 + }, + { + "epoch": 0.4963265306122449, + "grad_norm": 0.11343304812908173, + "learning_rate": 2.59578860516455e-06, + "loss": 0.05672039389610291, + "step": 3040 + }, + { + "epoch": 0.49795918367346936, + "grad_norm": 0.12666337192058563, + "learning_rate": 2.5827612800154382e-06, + "loss": 0.047120341658592226, + "step": 3050 + }, + { + "epoch": 0.4995918367346939, + "grad_norm": 0.08930598944425583, + "learning_rate": 2.5697317047143177e-06, + "loss": 0.06063341498374939, + "step": 3060 + }, + { + "epoch": 0.5012244897959184, + "grad_norm": 0.07818852365016937, + "learning_rate": 2.556700233515316e-06, + "loss": 0.05240688323974609, + "step": 3070 + }, + { + "epoch": 0.5028571428571429, + "grad_norm": 0.1425711065530777, + "learning_rate": 2.5436672207241064e-06, + "loss": 0.053180736303329465, + "step": 3080 + }, + { + "epoch": 0.5044897959183674, + "grad_norm": 0.1457373946905136, + "learning_rate": 2.5306330206882764e-06, + "loss": 0.06092001795768738, + "step": 3090 + }, + { + "epoch": 0.5061224489795918, + "grad_norm": 0.12738314270973206, + "learning_rate": 2.5175979877876935e-06, + "loss": 0.06613186597824097, + "step": 3100 + }, + { + "epoch": 0.5077551020408163, + "grad_norm": 0.09169036149978638, + "learning_rate": 2.5045624764248687e-06, + "loss": 0.06115126609802246, + "step": 3110 + }, + { + "epoch": 0.5093877551020408, + "grad_norm": 0.09637737274169922, + "learning_rate": 2.491526841015321e-06, + "loss": 0.044120225310325625, + "step": 3120 + }, + { + "epoch": 0.5110204081632653, + "grad_norm": 0.15029804408550262, + "learning_rate": 2.478491435977945e-06, + "loss": 0.07098687291145325, + "step": 3130 + }, + { + "epoch": 0.5126530612244898, + "grad_norm": 0.1228741705417633, + "learning_rate": 2.4654566157253683e-06, + "loss": 0.05738852024078369, + "step": 3140 + }, + { + "epoch": 0.5142857142857142, + "grad_norm": 0.09150952100753784, + "learning_rate": 2.452422734654321e-06, + "loss": 0.053479206562042234, + "step": 3150 + }, + { + "epoch": 0.5159183673469387, + "grad_norm": 0.14547918736934662, + "learning_rate": 2.4393901471359977e-06, + "loss": 0.056794995069503786, + "step": 3160 + }, + { + "epoch": 0.5175510204081633, + "grad_norm": 0.10099287331104279, + "learning_rate": 2.4263592075064238e-06, + "loss": 0.05802353620529175, + "step": 3170 + }, + { + "epoch": 0.5191836734693878, + "grad_norm": 0.10859636962413788, + "learning_rate": 2.41333027005682e-06, + "loss": 0.05282890200614929, + "step": 3180 + }, + { + "epoch": 0.5208163265306123, + "grad_norm": 0.12420635670423508, + "learning_rate": 2.400303689023972e-06, + "loss": 0.05190218687057495, + "step": 3190 + }, + { + "epoch": 0.5224489795918368, + "grad_norm": 0.1902378350496292, + "learning_rate": 2.3872798185805986e-06, + "loss": 0.06482383608818054, + "step": 3200 + }, + { + "epoch": 0.5224489795918368, + "eval_accuracy": 0.9744583333333335, + "eval_loss": 0.058392543345689774, + "eval_runtime": 156.8655, + "eval_samples_per_second": 25.5, + "eval_steps_per_second": 25.5, + "step": 3200 + }, + { + "epoch": 0.5240816326530612, + "grad_norm": 0.15229448676109314, + "learning_rate": 2.37425901282572e-06, + "loss": 0.06062687039375305, + "step": 3210 + }, + { + "epoch": 0.5257142857142857, + "grad_norm": 0.08411400765180588, + "learning_rate": 2.3612416257750335e-06, + "loss": 0.06494970321655273, + "step": 3220 + }, + { + "epoch": 0.5273469387755102, + "grad_norm": 0.114021435379982, + "learning_rate": 2.348228011351287e-06, + "loss": 0.05140092372894287, + "step": 3230 + }, + { + "epoch": 0.5289795918367347, + "grad_norm": 0.08826510608196259, + "learning_rate": 2.335218523374656e-06, + "loss": 0.041680914163589475, + "step": 3240 + }, + { + "epoch": 0.5306122448979592, + "grad_norm": 0.10536732524633408, + "learning_rate": 2.3222135155531248e-06, + "loss": 0.04369679093360901, + "step": 3250 + }, + { + "epoch": 0.5322448979591837, + "grad_norm": 0.16883279383182526, + "learning_rate": 2.309213341472868e-06, + "loss": 0.050543594360351565, + "step": 3260 + }, + { + "epoch": 0.5338775510204081, + "grad_norm": 0.15277455747127533, + "learning_rate": 2.29621835458864e-06, + "loss": 0.05258342027664185, + "step": 3270 + }, + { + "epoch": 0.5355102040816326, + "grad_norm": 0.07162481546401978, + "learning_rate": 2.283228908214163e-06, + "loss": 0.055678969621658324, + "step": 3280 + }, + { + "epoch": 0.5371428571428571, + "grad_norm": 0.11578263342380524, + "learning_rate": 2.2702453555125182e-06, + "loss": 0.048620349168777464, + "step": 3290 + }, + { + "epoch": 0.5387755102040817, + "grad_norm": 0.21185705065727234, + "learning_rate": 2.2572680494865527e-06, + "loss": 0.046547812223434445, + "step": 3300 + }, + { + "epoch": 0.5404081632653062, + "grad_norm": 0.19949115812778473, + "learning_rate": 2.2442973429692705e-06, + "loss": 0.04967870116233826, + "step": 3310 + }, + { + "epoch": 0.5420408163265306, + "grad_norm": 0.10855370759963989, + "learning_rate": 2.231333588614248e-06, + "loss": 0.0508593499660492, + "step": 3320 + }, + { + "epoch": 0.5436734693877551, + "grad_norm": 0.11827342212200165, + "learning_rate": 2.2183771388860427e-06, + "loss": 0.04854269027709961, + "step": 3330 + }, + { + "epoch": 0.5453061224489796, + "grad_norm": 0.1081443578004837, + "learning_rate": 2.2054283460506096e-06, + "loss": 0.05934192538261414, + "step": 3340 + }, + { + "epoch": 0.5469387755102041, + "grad_norm": 0.14441704750061035, + "learning_rate": 2.1924875621657256e-06, + "loss": 0.05332346558570862, + "step": 3350 + }, + { + "epoch": 0.5485714285714286, + "grad_norm": 0.11531820893287659, + "learning_rate": 2.179555139071415e-06, + "loss": 0.04938814043998718, + "step": 3360 + }, + { + "epoch": 0.5502040816326531, + "grad_norm": 0.08365523815155029, + "learning_rate": 2.166631428380387e-06, + "loss": 0.05734975934028626, + "step": 3370 + }, + { + "epoch": 0.5518367346938775, + "grad_norm": 0.09165070950984955, + "learning_rate": 2.1537167814684728e-06, + "loss": 0.047475004196166994, + "step": 3380 + }, + { + "epoch": 0.553469387755102, + "grad_norm": 0.13237863779067993, + "learning_rate": 2.140811549465074e-06, + "loss": 0.062284767627716064, + "step": 3390 + }, + { + "epoch": 0.5551020408163265, + "grad_norm": 0.1698858141899109, + "learning_rate": 2.127916083243615e-06, + "loss": 0.06585791110992431, + "step": 3400 + }, + { + "epoch": 0.5551020408163265, + "eval_accuracy": 0.9770833333333334, + "eval_loss": 0.05396538972854614, + "eval_runtime": 156.044, + "eval_samples_per_second": 25.634, + "eval_steps_per_second": 25.634, + "step": 3400 + }, + { + "epoch": 0.556734693877551, + "grad_norm": 0.1104230135679245, + "learning_rate": 2.1150307334120033e-06, + "loss": 0.04635375440120697, + "step": 3410 + }, + { + "epoch": 0.5583673469387755, + "grad_norm": 0.21372850239276886, + "learning_rate": 2.1021558503030973e-06, + "loss": 0.05940789580345154, + "step": 3420 + }, + { + "epoch": 0.56, + "grad_norm": 0.2013176679611206, + "learning_rate": 2.0892917839651818e-06, + "loss": 0.047165653109550475, + "step": 3430 + }, + { + "epoch": 0.5616326530612245, + "grad_norm": 0.09189072996377945, + "learning_rate": 2.076438884152449e-06, + "loss": 0.04550817906856537, + "step": 3440 + }, + { + "epoch": 0.563265306122449, + "grad_norm": 0.10521821677684784, + "learning_rate": 2.0635975003154934e-06, + "loss": 0.045033201575279236, + "step": 3450 + }, + { + "epoch": 0.5648979591836735, + "grad_norm": 0.10311131924390793, + "learning_rate": 2.0507679815918043e-06, + "loss": 0.04502847194671631, + "step": 3460 + }, + { + "epoch": 0.566530612244898, + "grad_norm": 0.07487176358699799, + "learning_rate": 2.0379506767962788e-06, + "loss": 0.04697479605674744, + "step": 3470 + }, + { + "epoch": 0.5681632653061225, + "grad_norm": 0.12441752105951309, + "learning_rate": 2.0251459344117345e-06, + "loss": 0.052119261026382445, + "step": 3480 + }, + { + "epoch": 0.5697959183673469, + "grad_norm": 0.12823422253131866, + "learning_rate": 2.0123541025794397e-06, + "loss": 0.05213663578033447, + "step": 3490 + }, + { + "epoch": 0.5714285714285714, + "grad_norm": 0.11340193450450897, + "learning_rate": 1.9995755290896414e-06, + "loss": 0.05319364666938782, + "step": 3500 + }, + { + "epoch": 0.5730612244897959, + "grad_norm": 0.10469839721918106, + "learning_rate": 1.986810561372113e-06, + "loss": 0.046560952067375184, + "step": 3510 + }, + { + "epoch": 0.5746938775510204, + "grad_norm": 0.07797456532716751, + "learning_rate": 1.9740595464867092e-06, + "loss": 0.05037369728088379, + "step": 3520 + }, + { + "epoch": 0.5763265306122449, + "grad_norm": 0.13010293245315552, + "learning_rate": 1.961322831113927e-06, + "loss": 0.0543134331703186, + "step": 3530 + }, + { + "epoch": 0.5779591836734694, + "grad_norm": 0.13981281220912933, + "learning_rate": 1.9486007615454837e-06, + "loss": 0.052860957384109494, + "step": 3540 + }, + { + "epoch": 0.5795918367346938, + "grad_norm": 0.14817261695861816, + "learning_rate": 1.9358936836748972e-06, + "loss": 0.05510561466217041, + "step": 3550 + }, + { + "epoch": 0.5812244897959183, + "grad_norm": 0.15079492330551147, + "learning_rate": 1.9232019429880874e-06, + "loss": 0.06125714182853699, + "step": 3560 + }, + { + "epoch": 0.5828571428571429, + "grad_norm": 0.10553149878978729, + "learning_rate": 1.910525884553976e-06, + "loss": 0.051503753662109374, + "step": 3570 + }, + { + "epoch": 0.5844897959183674, + "grad_norm": 0.09094442427158356, + "learning_rate": 1.8978658530151113e-06, + "loss": 0.05377998948097229, + "step": 3580 + }, + { + "epoch": 0.5861224489795919, + "grad_norm": 0.09810740500688553, + "learning_rate": 1.8852221925782935e-06, + "loss": 0.05506981611251831, + "step": 3590 + }, + { + "epoch": 0.5877551020408164, + "grad_norm": 0.13397352397441864, + "learning_rate": 1.872595247005218e-06, + "loss": 0.056353437900543216, + "step": 3600 + }, + { + "epoch": 0.5877551020408164, + "eval_accuracy": 0.979875, + "eval_loss": 0.04787343740463257, + "eval_runtime": 157.8457, + "eval_samples_per_second": 25.341, + "eval_steps_per_second": 25.341, + "step": 3600 + }, + { + "epoch": 0.5893877551020408, + "grad_norm": 0.07899880409240723, + "learning_rate": 1.8599853596031289e-06, + "loss": 0.05329249501228332, + "step": 3610 + }, + { + "epoch": 0.5910204081632653, + "grad_norm": 0.10414254665374756, + "learning_rate": 1.8473928732154847e-06, + "loss": 0.04905630648136139, + "step": 3620 + }, + { + "epoch": 0.5926530612244898, + "grad_norm": 0.1037079319357872, + "learning_rate": 1.8348181302126367e-06, + "loss": 0.04863021373748779, + "step": 3630 + }, + { + "epoch": 0.5942857142857143, + "grad_norm": 0.06960397213697433, + "learning_rate": 1.8222614724825222e-06, + "loss": 0.04666702449321747, + "step": 3640 + }, + { + "epoch": 0.5959183673469388, + "grad_norm": 0.11159648001194, + "learning_rate": 1.8097232414213662e-06, + "loss": 0.05730679631233215, + "step": 3650 + }, + { + "epoch": 0.5975510204081632, + "grad_norm": 0.10231669992208481, + "learning_rate": 1.7972037779244e-06, + "loss": 0.04433938264846802, + "step": 3660 + }, + { + "epoch": 0.5991836734693877, + "grad_norm": 0.1256829798221588, + "learning_rate": 1.784703422376598e-06, + "loss": 0.04502029418945312, + "step": 3670 + }, + { + "epoch": 0.6008163265306122, + "grad_norm": 0.1006830632686615, + "learning_rate": 1.7722225146434146e-06, + "loss": 0.04091633260250092, + "step": 3680 + }, + { + "epoch": 0.6024489795918367, + "grad_norm": 0.12239495664834976, + "learning_rate": 1.7597613940615492e-06, + "loss": 0.0417140930891037, + "step": 3690 + }, + { + "epoch": 0.6040816326530613, + "grad_norm": 0.09386949241161346, + "learning_rate": 1.7473203994297197e-06, + "loss": 0.045112240314483645, + "step": 3700 + }, + { + "epoch": 0.6057142857142858, + "grad_norm": 0.1845707893371582, + "learning_rate": 1.7348998689994492e-06, + "loss": 0.04915813505649567, + "step": 3710 + }, + { + "epoch": 0.6073469387755102, + "grad_norm": 0.1345887929201126, + "learning_rate": 1.7225001404658714e-06, + "loss": 0.05195118188858032, + "step": 3720 + }, + { + "epoch": 0.6089795918367347, + "grad_norm": 0.16226522624492645, + "learning_rate": 1.7101215509585482e-06, + "loss": 0.047248026728630065, + "step": 3730 + }, + { + "epoch": 0.6106122448979592, + "grad_norm": 0.12012234330177307, + "learning_rate": 1.6977644370323032e-06, + "loss": 0.05229860544204712, + "step": 3740 + }, + { + "epoch": 0.6122448979591837, + "grad_norm": 0.16154533624649048, + "learning_rate": 1.6854291346580737e-06, + "loss": 0.048448923230171206, + "step": 3750 + }, + { + "epoch": 0.6138775510204082, + "grad_norm": 0.08616314828395844, + "learning_rate": 1.673115979213773e-06, + "loss": 0.04779277443885803, + "step": 3760 + }, + { + "epoch": 0.6155102040816327, + "grad_norm": 0.20649370551109314, + "learning_rate": 1.6608253054751739e-06, + "loss": 0.04001966416835785, + "step": 3770 + }, + { + "epoch": 0.6171428571428571, + "grad_norm": 0.22419355809688568, + "learning_rate": 1.6485574476068056e-06, + "loss": 0.04915740489959717, + "step": 3780 + }, + { + "epoch": 0.6187755102040816, + "grad_norm": 0.11812295019626617, + "learning_rate": 1.6363127391528694e-06, + "loss": 0.055313241481781, + "step": 3790 + }, + { + "epoch": 0.6204081632653061, + "grad_norm": 0.12400855869054794, + "learning_rate": 1.6240915130281696e-06, + "loss": 0.05452631115913391, + "step": 3800 + }, + { + "epoch": 0.6204081632653061, + "eval_accuracy": 0.981875, + "eval_loss": 0.04318515956401825, + "eval_runtime": 158.5838, + "eval_samples_per_second": 25.223, + "eval_steps_per_second": 25.223, + "step": 3800 + }, + { + "epoch": 0.6220408163265306, + "grad_norm": 0.094053253531456, + "learning_rate": 1.6118941015090614e-06, + "loss": 0.047327810525894166, + "step": 3810 + }, + { + "epoch": 0.6236734693877551, + "grad_norm": 0.11286808550357819, + "learning_rate": 1.599720836224419e-06, + "loss": 0.04089818894863129, + "step": 3820 + }, + { + "epoch": 0.6253061224489795, + "grad_norm": 0.0909293070435524, + "learning_rate": 1.5875720481466162e-06, + "loss": 0.04378651976585388, + "step": 3830 + }, + { + "epoch": 0.626938775510204, + "grad_norm": 0.1207573264837265, + "learning_rate": 1.5754480675825295e-06, + "loss": 0.055971544981002805, + "step": 3840 + }, + { + "epoch": 0.6285714285714286, + "grad_norm": 0.09171696752309799, + "learning_rate": 1.563349224164556e-06, + "loss": 0.04364039599895477, + "step": 3850 + }, + { + "epoch": 0.6302040816326531, + "grad_norm": 0.1100732609629631, + "learning_rate": 1.5512758468416561e-06, + "loss": 0.04227602481842041, + "step": 3860 + }, + { + "epoch": 0.6318367346938776, + "grad_norm": 0.08587154746055603, + "learning_rate": 1.5392282638704032e-06, + "loss": 0.03987776041030884, + "step": 3870 + }, + { + "epoch": 0.6334693877551021, + "grad_norm": 0.14273083209991455, + "learning_rate": 1.5272068028060616e-06, + "loss": 0.05040489435195923, + "step": 3880 + }, + { + "epoch": 0.6351020408163265, + "grad_norm": 0.12788064777851105, + "learning_rate": 1.5152117904936827e-06, + "loss": 0.04129888415336609, + "step": 3890 + }, + { + "epoch": 0.636734693877551, + "grad_norm": 0.14224086701869965, + "learning_rate": 1.5032435530592165e-06, + "loss": 0.04145504236221313, + "step": 3900 + }, + { + "epoch": 0.6383673469387755, + "grad_norm": 0.07414636015892029, + "learning_rate": 1.4913024159006444e-06, + "loss": 0.03938417434692383, + "step": 3910 + }, + { + "epoch": 0.64, + "grad_norm": 0.1443425565958023, + "learning_rate": 1.4793887036791338e-06, + "loss": 0.047359606623649596, + "step": 3920 + }, + { + "epoch": 0.6416326530612245, + "grad_norm": 0.1462288498878479, + "learning_rate": 1.4675027403102107e-06, + "loss": 0.047661393880844116, + "step": 3930 + }, + { + "epoch": 0.643265306122449, + "grad_norm": 0.15937893092632294, + "learning_rate": 1.4556448489549519e-06, + "loss": 0.04829471111297608, + "step": 3940 + }, + { + "epoch": 0.6448979591836734, + "grad_norm": 0.10469378530979156, + "learning_rate": 1.4438153520111978e-06, + "loss": 0.047904521226882935, + "step": 3950 + }, + { + "epoch": 0.6465306122448979, + "grad_norm": 0.10817313939332962, + "learning_rate": 1.4320145711047894e-06, + "loss": 0.04648992121219635, + "step": 3960 + }, + { + "epoch": 0.6481632653061224, + "grad_norm": 0.12307990342378616, + "learning_rate": 1.4202428270808256e-06, + "loss": 0.044715133309364316, + "step": 3970 + }, + { + "epoch": 0.649795918367347, + "grad_norm": 0.14425775408744812, + "learning_rate": 1.408500439994933e-06, + "loss": 0.042470496892929074, + "step": 3980 + }, + { + "epoch": 0.6514285714285715, + "grad_norm": 0.11694067716598511, + "learning_rate": 1.3967877291045706e-06, + "loss": 0.03910423815250397, + "step": 3990 + }, + { + "epoch": 0.6530612244897959, + "grad_norm": 0.13050039112567902, + "learning_rate": 1.3851050128603454e-06, + "loss": 0.046834564208984374, + "step": 4000 + }, + { + "epoch": 0.6530612244897959, + "eval_accuracy": 0.9817083333333334, + "eval_loss": 0.04485878720879555, + "eval_runtime": 155.2113, + "eval_samples_per_second": 25.771, + "eval_steps_per_second": 25.771, + "step": 4000 + }, + { + "epoch": 0.6546938775510204, + "grad_norm": 0.16441842913627625, + "learning_rate": 1.3734526088973576e-06, + "loss": 0.048332995176315306, + "step": 4010 + }, + { + "epoch": 0.6563265306122449, + "grad_norm": 0.16489046812057495, + "learning_rate": 1.3618308340265628e-06, + "loss": 0.05632186532020569, + "step": 4020 + }, + { + "epoch": 0.6579591836734694, + "grad_norm": 0.0776519775390625, + "learning_rate": 1.3502400042261583e-06, + "loss": 0.0413711816072464, + "step": 4030 + }, + { + "epoch": 0.6595918367346939, + "grad_norm": 0.07065974175930023, + "learning_rate": 1.3386804346329913e-06, + "loss": 0.04443484246730804, + "step": 4040 + }, + { + "epoch": 0.6612244897959184, + "grad_norm": 0.058040838688611984, + "learning_rate": 1.3271524395339958e-06, + "loss": 0.03936767876148224, + "step": 4050 + }, + { + "epoch": 0.6628571428571428, + "grad_norm": 0.12576180696487427, + "learning_rate": 1.31565633235764e-06, + "loss": 0.03893322050571442, + "step": 4060 + }, + { + "epoch": 0.6644897959183673, + "grad_norm": 0.1022111177444458, + "learning_rate": 1.30419242566541e-06, + "loss": 0.05221283435821533, + "step": 4070 + }, + { + "epoch": 0.6661224489795918, + "grad_norm": 0.07897438853979111, + "learning_rate": 1.2927610311433098e-06, + "loss": 0.04261190891265869, + "step": 4080 + }, + { + "epoch": 0.6677551020408163, + "grad_norm": 0.16535215079784393, + "learning_rate": 1.2813624595933885e-06, + "loss": 0.04804243743419647, + "step": 4090 + }, + { + "epoch": 0.6693877551020408, + "grad_norm": 0.10754700005054474, + "learning_rate": 1.2699970209252877e-06, + "loss": 0.03677414059638977, + "step": 4100 + }, + { + "epoch": 0.6710204081632654, + "grad_norm": 0.13121150434017181, + "learning_rate": 1.2586650241478173e-06, + "loss": 0.04266786575317383, + "step": 4110 + }, + { + "epoch": 0.6726530612244898, + "grad_norm": 0.11523906886577606, + "learning_rate": 1.2473667773605542e-06, + "loss": 0.045382776856422426, + "step": 4120 + }, + { + "epoch": 0.6742857142857143, + "grad_norm": 0.16150203347206116, + "learning_rate": 1.236102587745464e-06, + "loss": 0.0483401358127594, + "step": 4130 + }, + { + "epoch": 0.6759183673469388, + "grad_norm": 0.2037627100944519, + "learning_rate": 1.2248727615585501e-06, + "loss": 0.0531754195690155, + "step": 4140 + }, + { + "epoch": 0.6775510204081633, + "grad_norm": 0.08756938576698303, + "learning_rate": 1.2136776041215293e-06, + "loss": 0.046069496870040895, + "step": 4150 + }, + { + "epoch": 0.6791836734693878, + "grad_norm": 0.10196162015199661, + "learning_rate": 1.2025174198135255e-06, + "loss": 0.03938464820384979, + "step": 4160 + }, + { + "epoch": 0.6808163265306122, + "grad_norm": 0.10243439674377441, + "learning_rate": 1.1913925120627987e-06, + "loss": 0.04410702586174011, + "step": 4170 + }, + { + "epoch": 0.6824489795918367, + "grad_norm": 0.16230984032154083, + "learning_rate": 1.1803031833384923e-06, + "loss": 0.03496443629264832, + "step": 4180 + }, + { + "epoch": 0.6840816326530612, + "grad_norm": 0.1036972925066948, + "learning_rate": 1.1692497351424117e-06, + "loss": 0.04019207060337067, + "step": 4190 + }, + { + "epoch": 0.6857142857142857, + "grad_norm": 0.19966299831867218, + "learning_rate": 1.1582324680008252e-06, + "loss": 0.03875584900379181, + "step": 4200 + }, + { + "epoch": 0.6857142857142857, + "eval_accuracy": 0.9839583333333334, + "eval_loss": 0.0395464189350605, + "eval_runtime": 153.8599, + "eval_samples_per_second": 25.998, + "eval_steps_per_second": 25.998, + "step": 4200 + }, + { + "epoch": 0.6873469387755102, + "grad_norm": 0.15349307656288147, + "learning_rate": 1.1472516814562945e-06, + "loss": 0.04495754837989807, + "step": 4210 + }, + { + "epoch": 0.6889795918367347, + "grad_norm": 0.16430068016052246, + "learning_rate": 1.1363076740595294e-06, + "loss": 0.0422965407371521, + "step": 4220 + }, + { + "epoch": 0.6906122448979591, + "grad_norm": 0.1903417408466339, + "learning_rate": 1.1254007433612705e-06, + "loss": 0.04501302838325501, + "step": 4230 + }, + { + "epoch": 0.6922448979591836, + "grad_norm": 0.05951374024152756, + "learning_rate": 1.1145311859042026e-06, + "loss": 0.03593409359455109, + "step": 4240 + }, + { + "epoch": 0.6938775510204082, + "grad_norm": 0.05894222855567932, + "learning_rate": 1.1036992972148865e-06, + "loss": 0.034275355935096743, + "step": 4250 + }, + { + "epoch": 0.6955102040816327, + "grad_norm": 0.09788545221090317, + "learning_rate": 1.092905371795729e-06, + "loss": 0.039912134408950806, + "step": 4260 + }, + { + "epoch": 0.6971428571428572, + "grad_norm": 0.2608252763748169, + "learning_rate": 1.0821497031169729e-06, + "loss": 0.03563873171806335, + "step": 4270 + }, + { + "epoch": 0.6987755102040817, + "grad_norm": 0.10350881516933441, + "learning_rate": 1.0714325836087191e-06, + "loss": 0.036586591601371767, + "step": 4280 + }, + { + "epoch": 0.7004081632653061, + "grad_norm": 0.10214894264936447, + "learning_rate": 1.0607543046529763e-06, + "loss": 0.03775157630443573, + "step": 4290 + }, + { + "epoch": 0.7020408163265306, + "grad_norm": 0.1326843649148941, + "learning_rate": 1.050115156575738e-06, + "loss": 0.03171868622303009, + "step": 4300 + }, + { + "epoch": 0.7036734693877551, + "grad_norm": 0.12274391204118729, + "learning_rate": 1.0395154286390896e-06, + "loss": 0.028169712424278258, + "step": 4310 + }, + { + "epoch": 0.7053061224489796, + "grad_norm": 0.12580160796642303, + "learning_rate": 1.028955409033342e-06, + "loss": 0.03762204349040985, + "step": 4320 + }, + { + "epoch": 0.7069387755102041, + "grad_norm": 0.26383766531944275, + "learning_rate": 1.0184353848692005e-06, + "loss": 0.04035975933074951, + "step": 4330 + }, + { + "epoch": 0.7085714285714285, + "grad_norm": 0.101371631026268, + "learning_rate": 1.0079556421699533e-06, + "loss": 0.04394418001174927, + "step": 4340 + }, + { + "epoch": 0.710204081632653, + "grad_norm": 0.12708403170108795, + "learning_rate": 9.97516465863698e-07, + "loss": 0.04101270735263825, + "step": 4350 + }, + { + "epoch": 0.7118367346938775, + "grad_norm": 0.08838815987110138, + "learning_rate": 9.87118139775595e-07, + "loss": 0.04232141673564911, + "step": 4360 + }, + { + "epoch": 0.713469387755102, + "grad_norm": 0.13831272721290588, + "learning_rate": 9.767609466201494e-07, + "loss": 0.03110981285572052, + "step": 4370 + }, + { + "epoch": 0.7151020408163266, + "grad_norm": 0.14169402420520782, + "learning_rate": 9.664451679935252e-07, + "loss": 0.036100244522094725, + "step": 4380 + }, + { + "epoch": 0.7167346938775511, + "grad_norm": 0.09364918619394302, + "learning_rate": 9.561710843658892e-07, + "loss": 0.028025591373443605, + "step": 4390 + }, + { + "epoch": 0.7183673469387755, + "grad_norm": 0.08178938925266266, + "learning_rate": 9.459389750737854e-07, + "loss": 0.03276517987251282, + "step": 4400 + }, + { + "epoch": 0.7183673469387755, + "eval_accuracy": 0.9830833333333334, + "eval_loss": 0.03968822583556175, + "eval_runtime": 170.9863, + "eval_samples_per_second": 23.394, + "eval_steps_per_second": 23.394, + "step": 4400 + }, + { + "epoch": 0.72, + "grad_norm": 0.21406538784503937, + "learning_rate": 9.357491183125398e-07, + "loss": 0.04240003228187561, + "step": 4410 + }, + { + "epoch": 0.7216326530612245, + "grad_norm": 0.2528787851333618, + "learning_rate": 9.256017911286958e-07, + "loss": 0.03953196704387665, + "step": 4420 + }, + { + "epoch": 0.723265306122449, + "grad_norm": 0.1817523092031479, + "learning_rate": 9.154972694124866e-07, + "loss": 0.03630510270595551, + "step": 4430 + }, + { + "epoch": 0.7248979591836735, + "grad_norm": 0.19527313113212585, + "learning_rate": 9.054358278903272e-07, + "loss": 0.04581010937690735, + "step": 4440 + }, + { + "epoch": 0.726530612244898, + "grad_norm": 0.3551980257034302, + "learning_rate": 8.954177401173491e-07, + "loss": 0.04633161723613739, + "step": 4450 + }, + { + "epoch": 0.7281632653061224, + "grad_norm": 0.1003197506070137, + "learning_rate": 8.854432784699624e-07, + "loss": 0.03686977326869965, + "step": 4460 + }, + { + "epoch": 0.7297959183673469, + "grad_norm": 0.1250677853822708, + "learning_rate": 8.755127141384495e-07, + "loss": 0.03494169116020203, + "step": 4470 + }, + { + "epoch": 0.7314285714285714, + "grad_norm": 0.23732195794582367, + "learning_rate": 8.656263171195931e-07, + "loss": 0.04174305498600006, + "step": 4480 + }, + { + "epoch": 0.7330612244897959, + "grad_norm": 0.2553883194923401, + "learning_rate": 8.557843562093329e-07, + "loss": 0.04263312518596649, + "step": 4490 + }, + { + "epoch": 0.7346938775510204, + "grad_norm": 0.1312720775604248, + "learning_rate": 8.459870989954611e-07, + "loss": 0.03155354261398315, + "step": 4500 + }, + { + "epoch": 0.7363265306122448, + "grad_norm": 0.2632088363170624, + "learning_rate": 8.362348118503424e-07, + "loss": 0.05083600878715515, + "step": 4510 + }, + { + "epoch": 0.7379591836734694, + "grad_norm": 0.1293715387582779, + "learning_rate": 8.265277599236785e-07, + "loss": 0.031102049350738525, + "step": 4520 + }, + { + "epoch": 0.7395918367346939, + "grad_norm": 0.15807363390922546, + "learning_rate": 8.168662071352918e-07, + "loss": 0.033837145566940306, + "step": 4530 + }, + { + "epoch": 0.7412244897959184, + "grad_norm": 0.08791180700063705, + "learning_rate": 8.072504161679534e-07, + "loss": 0.0337544322013855, + "step": 4540 + }, + { + "epoch": 0.7428571428571429, + "grad_norm": 0.10295934230089188, + "learning_rate": 7.976806484602417e-07, + "loss": 0.03291972577571869, + "step": 4550 + }, + { + "epoch": 0.7444897959183674, + "grad_norm": 0.11660359799861908, + "learning_rate": 7.881571641994331e-07, + "loss": 0.032550466060638425, + "step": 4560 + }, + { + "epoch": 0.7461224489795918, + "grad_norm": 0.1896156519651413, + "learning_rate": 7.786802223144286e-07, + "loss": 0.030972787737846376, + "step": 4570 + }, + { + "epoch": 0.7477551020408163, + "grad_norm": 0.2084398865699768, + "learning_rate": 7.69250080468712e-07, + "loss": 0.0415867954492569, + "step": 4580 + }, + { + "epoch": 0.7493877551020408, + "grad_norm": 0.12618854641914368, + "learning_rate": 7.59866995053348e-07, + "loss": 0.0360225111246109, + "step": 4590 + }, + { + "epoch": 0.7510204081632653, + "grad_norm": 0.12291006743907928, + "learning_rate": 7.505312211800081e-07, + "loss": 0.03627174496650696, + "step": 4600 + }, + { + "epoch": 0.7510204081632653, + "eval_accuracy": 0.9855833333333334, + "eval_loss": 0.03568050265312195, + "eval_runtime": 156.9074, + "eval_samples_per_second": 25.493, + "eval_steps_per_second": 25.493, + "step": 4600 + }, + { + "epoch": 0.7526530612244898, + "grad_norm": 0.1491072177886963, + "learning_rate": 7.41243012674035e-07, + "loss": 0.033873635530471805, + "step": 4610 + }, + { + "epoch": 0.7542857142857143, + "grad_norm": 0.14268337190151215, + "learning_rate": 7.320026220675447e-07, + "loss": 0.03949807286262512, + "step": 4620 + }, + { + "epoch": 0.7559183673469387, + "grad_norm": 0.15548810362815857, + "learning_rate": 7.228103005925557e-07, + "loss": 0.03478931188583374, + "step": 4630 + }, + { + "epoch": 0.7575510204081632, + "grad_norm": 0.14298392832279205, + "learning_rate": 7.136662981741613e-07, + "loss": 0.030326524376869203, + "step": 4640 + }, + { + "epoch": 0.7591836734693878, + "grad_norm": 0.11854545772075653, + "learning_rate": 7.045708634237341e-07, + "loss": 0.034136393666267396, + "step": 4650 + }, + { + "epoch": 0.7608163265306123, + "grad_norm": 0.1346571296453476, + "learning_rate": 6.955242436321663e-07, + "loss": 0.040317881107330325, + "step": 4660 + }, + { + "epoch": 0.7624489795918368, + "grad_norm": 0.13217571377754211, + "learning_rate": 6.865266847631463e-07, + "loss": 0.02895120084285736, + "step": 4670 + }, + { + "epoch": 0.7640816326530612, + "grad_norm": 0.0761382132768631, + "learning_rate": 6.775784314464717e-07, + "loss": 0.029726022481918336, + "step": 4680 + }, + { + "epoch": 0.7657142857142857, + "grad_norm": 0.17583715915679932, + "learning_rate": 6.686797269713971e-07, + "loss": 0.02383600324392319, + "step": 4690 + }, + { + "epoch": 0.7673469387755102, + "grad_norm": 0.13434812426567078, + "learning_rate": 6.598308132800221e-07, + "loss": 0.03801000118255615, + "step": 4700 + }, + { + "epoch": 0.7689795918367347, + "grad_norm": 0.0841183140873909, + "learning_rate": 6.510319309607099e-07, + "loss": 0.030053699016571046, + "step": 4710 + }, + { + "epoch": 0.7706122448979592, + "grad_norm": 0.10911499708890915, + "learning_rate": 6.422833192415476e-07, + "loss": 0.03296325206756592, + "step": 4720 + }, + { + "epoch": 0.7722448979591837, + "grad_norm": 0.08562543988227844, + "learning_rate": 6.335852159838423e-07, + "loss": 0.029071077704429626, + "step": 4730 + }, + { + "epoch": 0.7738775510204081, + "grad_norm": 0.24464724957942963, + "learning_rate": 6.249378576756538e-07, + "loss": 0.03307548463344574, + "step": 4740 + }, + { + "epoch": 0.7755102040816326, + "grad_norm": 0.17294660210609436, + "learning_rate": 6.163414794253641e-07, + "loss": 0.03507235944271088, + "step": 4750 + }, + { + "epoch": 0.7771428571428571, + "grad_norm": 0.18331816792488098, + "learning_rate": 6.077963149552868e-07, + "loss": 0.038254004716873166, + "step": 4760 + }, + { + "epoch": 0.7787755102040816, + "grad_norm": 0.10756120830774307, + "learning_rate": 5.993025965953103e-07, + "loss": 0.03428310751914978, + "step": 4770 + }, + { + "epoch": 0.7804081632653062, + "grad_norm": 0.16071061789989471, + "learning_rate": 5.908605552765833e-07, + "loss": 0.032243117690086365, + "step": 4780 + }, + { + "epoch": 0.7820408163265307, + "grad_norm": 0.14897944033145905, + "learning_rate": 5.824704205252343e-07, + "loss": 0.03812848925590515, + "step": 4790 + }, + { + "epoch": 0.7836734693877551, + "grad_norm": 0.1536232829093933, + "learning_rate": 5.741324204561335e-07, + "loss": 0.031389710307121274, + "step": 4800 + }, + { + "epoch": 0.7836734693877551, + "eval_accuracy": 0.985, + "eval_loss": 0.03615976870059967, + "eval_runtime": 153.2713, + "eval_samples_per_second": 26.098, + "eval_steps_per_second": 26.098, + "step": 4800 + }, + { + "epoch": 0.7853061224489796, + "grad_norm": 0.15687023103237152, + "learning_rate": 5.65846781766688e-07, + "loss": 0.027873021364212037, + "step": 4810 + }, + { + "epoch": 0.7869387755102041, + "grad_norm": 0.14541009068489075, + "learning_rate": 5.576137297306794e-07, + "loss": 0.0407156765460968, + "step": 4820 + }, + { + "epoch": 0.7885714285714286, + "grad_norm": 0.16398967802524567, + "learning_rate": 5.494334881921393e-07, + "loss": 0.03936888575553894, + "step": 4830 + }, + { + "epoch": 0.7902040816326531, + "grad_norm": 0.2177809625864029, + "learning_rate": 5.413062795592633e-07, + "loss": 0.03059300184249878, + "step": 4840 + }, + { + "epoch": 0.7918367346938775, + "grad_norm": 0.1487113982439041, + "learning_rate": 5.332323247983632e-07, + "loss": 0.038730043172836306, + "step": 4850 + }, + { + "epoch": 0.793469387755102, + "grad_norm": 0.1111542284488678, + "learning_rate": 5.252118434278605e-07, + "loss": 0.033898717164993285, + "step": 4860 + }, + { + "epoch": 0.7951020408163265, + "grad_norm": 0.10296906530857086, + "learning_rate": 5.172450535123166e-07, + "loss": 0.033659690618515016, + "step": 4870 + }, + { + "epoch": 0.796734693877551, + "grad_norm": 0.11844902485609055, + "learning_rate": 5.093321716565064e-07, + "loss": 0.0293354332447052, + "step": 4880 + }, + { + "epoch": 0.7983673469387755, + "grad_norm": 0.08401476591825485, + "learning_rate": 5.014734129995258e-07, + "loss": 0.027682244777679443, + "step": 4890 + }, + { + "epoch": 0.8, + "grad_norm": 0.15870651602745056, + "learning_rate": 4.936689912089443e-07, + "loss": 0.029118484258651732, + "step": 4900 + }, + { + "epoch": 0.8016326530612244, + "grad_norm": 0.10739132016897202, + "learning_rate": 4.859191184749968e-07, + "loss": 0.031624987721443176, + "step": 4910 + }, + { + "epoch": 0.803265306122449, + "grad_norm": 0.1523323655128479, + "learning_rate": 4.782240055048113e-07, + "loss": 0.03415923118591309, + "step": 4920 + }, + { + "epoch": 0.8048979591836735, + "grad_norm": 0.31860172748565674, + "learning_rate": 4.705838615166836e-07, + "loss": 0.035749971866607666, + "step": 4930 + }, + { + "epoch": 0.806530612244898, + "grad_norm": 0.15043528378009796, + "learning_rate": 4.629988942343863e-07, + "loss": 0.029294773936271667, + "step": 4940 + }, + { + "epoch": 0.8081632653061225, + "grad_norm": 0.1007913127541542, + "learning_rate": 4.55469309881523e-07, + "loss": 0.026727235317230223, + "step": 4950 + }, + { + "epoch": 0.809795918367347, + "grad_norm": 0.10851600021123886, + "learning_rate": 4.479953131759199e-07, + "loss": 0.029011189937591553, + "step": 4960 + }, + { + "epoch": 0.8114285714285714, + "grad_norm": 0.19257904589176178, + "learning_rate": 4.4057710732406113e-07, + "loss": 0.040424668788909913, + "step": 4970 + }, + { + "epoch": 0.8130612244897959, + "grad_norm": 0.08985571563243866, + "learning_rate": 4.332148940155617e-07, + "loss": 0.031291928887367246, + "step": 4980 + }, + { + "epoch": 0.8146938775510204, + "grad_norm": 0.09874805808067322, + "learning_rate": 4.259088734176883e-07, + "loss": 0.028314629197120668, + "step": 4990 + }, + { + "epoch": 0.8163265306122449, + "grad_norm": 0.2056759148836136, + "learning_rate": 4.186592441699119e-07, + "loss": 0.028049173951148986, + "step": 5000 + }, + { + "epoch": 0.8163265306122449, + "eval_accuracy": 0.986125, + "eval_loss": 0.033585112541913986, + "eval_runtime": 153.2119, + "eval_samples_per_second": 26.108, + "eval_steps_per_second": 26.108, + "step": 5000 + }, + { + "epoch": 0.8179591836734694, + "grad_norm": 0.10125680267810822, + "learning_rate": 4.1146620337850965e-07, + "loss": 0.03612466752529144, + "step": 5010 + }, + { + "epoch": 0.8195918367346938, + "grad_norm": 0.10107982903718948, + "learning_rate": 4.043299466112058e-07, + "loss": 0.026968371868133546, + "step": 5020 + }, + { + "epoch": 0.8212244897959183, + "grad_norm": 0.13236716389656067, + "learning_rate": 3.9725066789185446e-07, + "loss": 0.030766445398330688, + "step": 5030 + }, + { + "epoch": 0.8228571428571428, + "grad_norm": 0.1182790994644165, + "learning_rate": 3.9022855969516384e-07, + "loss": 0.028759539127349854, + "step": 5040 + }, + { + "epoch": 0.8244897959183674, + "grad_norm": 0.1430918574333191, + "learning_rate": 3.832638129414637e-07, + "loss": 0.03126598298549652, + "step": 5050 + }, + { + "epoch": 0.8261224489795919, + "grad_norm": 0.24944688379764557, + "learning_rate": 3.7635661699151366e-07, + "loss": 0.030412834882736207, + "step": 5060 + }, + { + "epoch": 0.8277551020408164, + "grad_norm": 0.24371539056301117, + "learning_rate": 3.695071596413574e-07, + "loss": 0.03904336988925934, + "step": 5070 + }, + { + "epoch": 0.8293877551020408, + "grad_norm": 0.25066161155700684, + "learning_rate": 3.627156271172125e-07, + "loss": 0.032776132225990295, + "step": 5080 + }, + { + "epoch": 0.8310204081632653, + "grad_norm": 0.1635182797908783, + "learning_rate": 3.559822040704108e-07, + "loss": 0.027044805884361266, + "step": 5090 + }, + { + "epoch": 0.8326530612244898, + "grad_norm": 0.16698378324508667, + "learning_rate": 3.4930707357237666e-07, + "loss": 0.03271254003047943, + "step": 5100 + }, + { + "epoch": 0.8342857142857143, + "grad_norm": 0.18812412023544312, + "learning_rate": 3.4269041710964903e-07, + "loss": 0.030745366215705873, + "step": 5110 + }, + { + "epoch": 0.8359183673469388, + "grad_norm": 0.18739669024944305, + "learning_rate": 3.36132414578948e-07, + "loss": 0.032180759310722354, + "step": 5120 + }, + { + "epoch": 0.8375510204081633, + "grad_norm": 0.13260763883590698, + "learning_rate": 3.296332442822839e-07, + "loss": 0.026755687594413758, + "step": 5130 + }, + { + "epoch": 0.8391836734693877, + "grad_norm": 0.06771053373813629, + "learning_rate": 3.231930829221083e-07, + "loss": 0.02895594835281372, + "step": 5140 + }, + { + "epoch": 0.8408163265306122, + "grad_norm": 0.09441749006509781, + "learning_rate": 3.168121055965112e-07, + "loss": 0.033178862929344174, + "step": 5150 + }, + { + "epoch": 0.8424489795918367, + "grad_norm": 0.0863656997680664, + "learning_rate": 3.104904857944591e-07, + "loss": 0.02276728004217148, + "step": 5160 + }, + { + "epoch": 0.8440816326530612, + "grad_norm": 0.2047131359577179, + "learning_rate": 3.0422839539107794e-07, + "loss": 0.028258222341537475, + "step": 5170 + }, + { + "epoch": 0.8457142857142858, + "grad_norm": 0.2389051467180252, + "learning_rate": 2.980260046429834e-07, + "loss": 0.02725991904735565, + "step": 5180 + }, + { + "epoch": 0.8473469387755102, + "grad_norm": 0.1733924299478531, + "learning_rate": 2.9188348218364635e-07, + "loss": 0.0327379584312439, + "step": 5190 + }, + { + "epoch": 0.8489795918367347, + "grad_norm": 0.19450697302818298, + "learning_rate": 2.8580099501881285e-07, + "loss": 0.030778995156288146, + "step": 5200 + }, + { + "epoch": 0.8489795918367347, + "eval_accuracy": 0.9881666666666667, + "eval_loss": 0.02920014224946499, + "eval_runtime": 155.6487, + "eval_samples_per_second": 25.699, + "eval_steps_per_second": 25.699, + "step": 5200 + }, + { + "epoch": 0.8506122448979592, + "grad_norm": 0.14990773797035217, + "learning_rate": 2.7977870852195997e-07, + "loss": 0.03692179024219513, + "step": 5210 + }, + { + "epoch": 0.8522448979591837, + "grad_norm": 0.16051384806632996, + "learning_rate": 2.7381678642980303e-07, + "loss": 0.030528882145881654, + "step": 5220 + }, + { + "epoch": 0.8538775510204082, + "grad_norm": 0.1660623997449875, + "learning_rate": 2.6791539083784014e-07, + "loss": 0.025125756859779358, + "step": 5230 + }, + { + "epoch": 0.8555102040816327, + "grad_norm": 0.07568248361349106, + "learning_rate": 2.620746821959475e-07, + "loss": 0.02720560133457184, + "step": 5240 + }, + { + "epoch": 0.8571428571428571, + "grad_norm": 0.2634601593017578, + "learning_rate": 2.562948193040174e-07, + "loss": 0.033940815925598146, + "step": 5250 + }, + { + "epoch": 0.8587755102040816, + "grad_norm": 0.15926218032836914, + "learning_rate": 2.5057595930763873e-07, + "loss": 0.028036150336265563, + "step": 5260 + }, + { + "epoch": 0.8604081632653061, + "grad_norm": 0.19775289297103882, + "learning_rate": 2.4491825769382526e-07, + "loss": 0.02731768786907196, + "step": 5270 + }, + { + "epoch": 0.8620408163265306, + "grad_norm": 0.07538789510726929, + "learning_rate": 2.39321868286789e-07, + "loss": 0.0313712328672409, + "step": 5280 + }, + { + "epoch": 0.8636734693877551, + "grad_norm": 0.10647427290678024, + "learning_rate": 2.337869432437567e-07, + "loss": 0.02781860828399658, + "step": 5290 + }, + { + "epoch": 0.8653061224489796, + "grad_norm": 0.20678064227104187, + "learning_rate": 2.283136330508337e-07, + "loss": 0.028782907128334045, + "step": 5300 + }, + { + "epoch": 0.866938775510204, + "grad_norm": 0.18386024236679077, + "learning_rate": 2.2290208651891244e-07, + "loss": 0.02515394687652588, + "step": 5310 + }, + { + "epoch": 0.8685714285714285, + "grad_norm": 0.13139928877353668, + "learning_rate": 2.1755245077962618e-07, + "loss": 0.02002216875553131, + "step": 5320 + }, + { + "epoch": 0.8702040816326531, + "grad_norm": 0.10402868688106537, + "learning_rate": 2.1226487128134854e-07, + "loss": 0.03221684396266937, + "step": 5330 + }, + { + "epoch": 0.8718367346938776, + "grad_norm": 0.056439850479364395, + "learning_rate": 2.070394917852403e-07, + "loss": 0.023763810098171235, + "step": 5340 + }, + { + "epoch": 0.8734693877551021, + "grad_norm": 0.1421876698732376, + "learning_rate": 2.0187645436133913e-07, + "loss": 0.03192594051361084, + "step": 5350 + }, + { + "epoch": 0.8751020408163265, + "grad_norm": 0.20487430691719055, + "learning_rate": 1.967758993846966e-07, + "loss": 0.025955894589424135, + "step": 5360 + }, + { + "epoch": 0.876734693877551, + "grad_norm": 0.2183534801006317, + "learning_rate": 1.9173796553156487e-07, + "loss": 0.027435585856437683, + "step": 5370 + }, + { + "epoch": 0.8783673469387755, + "grad_norm": 0.19724565744400024, + "learning_rate": 1.8676278977562219e-07, + "loss": 0.028912261128425598, + "step": 5380 + }, + { + "epoch": 0.88, + "grad_norm": 0.21412917971611023, + "learning_rate": 1.818505073842508e-07, + "loss": 0.029015693068504333, + "step": 5390 + }, + { + "epoch": 0.8816326530612245, + "grad_norm": 0.08429288119077682, + "learning_rate": 1.770012519148598e-07, + "loss": 0.025249993801116942, + "step": 5400 + }, + { + "epoch": 0.8816326530612245, + "eval_accuracy": 0.989375, + "eval_loss": 0.02678457647562027, + "eval_runtime": 156.2763, + "eval_samples_per_second": 25.596, + "eval_steps_per_second": 25.596, + "step": 5400 + }, + { + "epoch": 0.883265306122449, + "grad_norm": 0.15489192306995392, + "learning_rate": 1.722151552112525e-07, + "loss": 0.03486272096633911, + "step": 5410 + }, + { + "epoch": 0.8848979591836734, + "grad_norm": 0.1594313681125641, + "learning_rate": 1.6749234740004227e-07, + "loss": 0.024751077592372894, + "step": 5420 + }, + { + "epoch": 0.8865306122448979, + "grad_norm": 0.12505307793617249, + "learning_rate": 1.6283295688711564e-07, + "loss": 0.025553828477859496, + "step": 5430 + }, + { + "epoch": 0.8881632653061224, + "grad_norm": 0.08852038532495499, + "learning_rate": 1.582371103541394e-07, + "loss": 0.025743791460990907, + "step": 5440 + }, + { + "epoch": 0.889795918367347, + "grad_norm": 0.08450407534837723, + "learning_rate": 1.5370493275511767e-07, + "loss": 0.02687264084815979, + "step": 5450 + }, + { + "epoch": 0.8914285714285715, + "grad_norm": 0.054211657494306564, + "learning_rate": 1.4923654731299363e-07, + "loss": 0.023073682188987733, + "step": 5460 + }, + { + "epoch": 0.893061224489796, + "grad_norm": 0.23101477324962616, + "learning_rate": 1.4483207551630045e-07, + "loss": 0.0280561625957489, + "step": 5470 + }, + { + "epoch": 0.8946938775510204, + "grad_norm": 0.19923920929431915, + "learning_rate": 1.4049163711585684e-07, + "loss": 0.02133965492248535, + "step": 5480 + }, + { + "epoch": 0.8963265306122449, + "grad_norm": 0.13227398693561554, + "learning_rate": 1.3621535012151244e-07, + "loss": 0.017637649178504945, + "step": 5490 + }, + { + "epoch": 0.8979591836734694, + "grad_norm": 0.11182940751314163, + "learning_rate": 1.3200333079893823e-07, + "loss": 0.02847398817539215, + "step": 5500 + }, + { + "epoch": 0.8995918367346939, + "grad_norm": 0.06762958317995071, + "learning_rate": 1.2785569366646628e-07, + "loss": 0.020397891104221345, + "step": 5510 + }, + { + "epoch": 0.9012244897959184, + "grad_norm": 0.06652984023094177, + "learning_rate": 1.2377255149197576e-07, + "loss": 0.03015478849411011, + "step": 5520 + }, + { + "epoch": 0.9028571428571428, + "grad_norm": 0.16087517142295837, + "learning_rate": 1.1975401528982718e-07, + "loss": 0.02614833414554596, + "step": 5530 + }, + { + "epoch": 0.9044897959183673, + "grad_norm": 0.0873415544629097, + "learning_rate": 1.1580019431784384e-07, + "loss": 0.023097904026508333, + "step": 5540 + }, + { + "epoch": 0.9061224489795918, + "grad_norm": 0.26114869117736816, + "learning_rate": 1.1191119607434076e-07, + "loss": 0.028036117553710938, + "step": 5550 + }, + { + "epoch": 0.9077551020408163, + "grad_norm": 0.08690405637025833, + "learning_rate": 1.0808712629520435e-07, + "loss": 0.027321887016296387, + "step": 5560 + }, + { + "epoch": 0.9093877551020408, + "grad_norm": 0.19979159533977509, + "learning_rate": 1.0432808895101426e-07, + "loss": 0.03543014228343964, + "step": 5570 + }, + { + "epoch": 0.9110204081632653, + "grad_norm": 0.09599421173334122, + "learning_rate": 1.0063418624421895e-07, + "loss": 0.02044544517993927, + "step": 5580 + }, + { + "epoch": 0.9126530612244897, + "grad_norm": 0.22291070222854614, + "learning_rate": 9.700551860635538e-08, + "loss": 0.027166444063186645, + "step": 5590 + }, + { + "epoch": 0.9142857142857143, + "grad_norm": 0.10956079512834549, + "learning_rate": 9.344218469532012e-08, + "loss": 0.02024113982915878, + "step": 5600 + }, + { + "epoch": 0.9142857142857143, + "eval_accuracy": 0.9900833333333334, + "eval_loss": 0.025807492434978485, + "eval_runtime": 154.9378, + "eval_samples_per_second": 25.817, + "eval_steps_per_second": 25.817, + "step": 5600 + }, + { + "epoch": 0.9159183673469388, + "grad_norm": 0.14949144423007965, + "learning_rate": 8.99442813926854e-08, + "loss": 0.024004274606704713, + "step": 5610 + }, + { + "epoch": 0.9175510204081633, + "grad_norm": 0.2315555214881897, + "learning_rate": 8.651190380106673e-08, + "loss": 0.024219857156276704, + "step": 5620 + }, + { + "epoch": 0.9191836734693878, + "grad_norm": 0.06734409183263779, + "learning_rate": 8.314514524153522e-08, + "loss": 0.02802177369594574, + "step": 5630 + }, + { + "epoch": 0.9208163265306123, + "grad_norm": 0.21427671611309052, + "learning_rate": 7.984409725108222e-08, + "loss": 0.021109704673290253, + "step": 5640 + }, + { + "epoch": 0.9224489795918367, + "grad_norm": 0.18013973534107208, + "learning_rate": 7.660884958012899e-08, + "loss": 0.020703735947608947, + "step": 5650 + }, + { + "epoch": 0.9240816326530612, + "grad_norm": 0.2695794701576233, + "learning_rate": 7.343949019008756e-08, + "loss": 0.03247661292552948, + "step": 5660 + }, + { + "epoch": 0.9257142857142857, + "grad_norm": 0.15175238251686096, + "learning_rate": 7.03361052509688e-08, + "loss": 0.021647430956363678, + "step": 5670 + }, + { + "epoch": 0.9273469387755102, + "grad_norm": 0.18325908482074738, + "learning_rate": 6.729877913903976e-08, + "loss": 0.028400209546089173, + "step": 5680 + }, + { + "epoch": 0.9289795918367347, + "grad_norm": 0.15859369933605194, + "learning_rate": 6.432759443452924e-08, + "loss": 0.014232246577739716, + "step": 5690 + }, + { + "epoch": 0.9306122448979591, + "grad_norm": 0.13438236713409424, + "learning_rate": 6.14226319193828e-08, + "loss": 0.02244553864002228, + "step": 5700 + }, + { + "epoch": 0.9322448979591836, + "grad_norm": 0.27041247487068176, + "learning_rate": 5.858397057506709e-08, + "loss": 0.029011306166648865, + "step": 5710 + }, + { + "epoch": 0.9338775510204081, + "grad_norm": 0.12215204536914825, + "learning_rate": 5.581168758042127e-08, + "loss": 0.028310227394104003, + "step": 5720 + }, + { + "epoch": 0.9355102040816327, + "grad_norm": 0.2103251814842224, + "learning_rate": 5.310585830955922e-08, + "loss": 0.021468037366867067, + "step": 5730 + }, + { + "epoch": 0.9371428571428572, + "grad_norm": 0.17152205109596252, + "learning_rate": 5.0466556329819835e-08, + "loss": 0.024149742722511292, + "step": 5740 + }, + { + "epoch": 0.9387755102040817, + "grad_norm": 0.08712399005889893, + "learning_rate": 4.789385339976832e-08, + "loss": 0.01809695065021515, + "step": 5750 + }, + { + "epoch": 0.9404081632653061, + "grad_norm": 0.15006859600543976, + "learning_rate": 4.53878194672433e-08, + "loss": 0.022831857204437256, + "step": 5760 + }, + { + "epoch": 0.9420408163265306, + "grad_norm": 0.06355110555887222, + "learning_rate": 4.2948522667455864e-08, + "loss": 0.015157052874565124, + "step": 5770 + }, + { + "epoch": 0.9436734693877551, + "grad_norm": 0.14330919086933136, + "learning_rate": 4.057602932113658e-08, + "loss": 0.02486245185136795, + "step": 5780 + }, + { + "epoch": 0.9453061224489796, + "grad_norm": 0.12216096371412277, + "learning_rate": 3.82704039327339e-08, + "loss": 0.024203602969646455, + "step": 5790 + }, + { + "epoch": 0.9469387755102041, + "grad_norm": 0.19883109629154205, + "learning_rate": 3.6031709188658336e-08, + "loss": 0.025162199139595033, + "step": 5800 + }, + { + "epoch": 0.9469387755102041, + "eval_accuracy": 0.99025, + "eval_loss": 0.025506427511572838, + "eval_runtime": 154.1326, + "eval_samples_per_second": 25.952, + "eval_steps_per_second": 25.952, + "step": 5800 + }, + { + "epoch": 0.9485714285714286, + "grad_norm": 0.25931984186172485, + "learning_rate": 3.386000595557964e-08, + "loss": 0.027550938725471496, + "step": 5810 + }, + { + "epoch": 0.950204081632653, + "grad_norm": 0.07933743298053741, + "learning_rate": 3.1755353278770936e-08, + "loss": 0.024577364325523376, + "step": 5820 + }, + { + "epoch": 0.9518367346938775, + "grad_norm": 0.13839955627918243, + "learning_rate": 2.9717808380503866e-08, + "loss": 0.021038763225078583, + "step": 5830 + }, + { + "epoch": 0.953469387755102, + "grad_norm": 0.14696243405342102, + "learning_rate": 2.7747426658492893e-08, + "loss": 0.02484375685453415, + "step": 5840 + }, + { + "epoch": 0.9551020408163265, + "grad_norm": 0.1538810133934021, + "learning_rate": 2.5844261684388473e-08, + "loss": 0.023458963632583617, + "step": 5850 + }, + { + "epoch": 0.9567346938775511, + "grad_norm": 0.08468759804964066, + "learning_rate": 2.4008365202321525e-08, + "loss": 0.03132525682449341, + "step": 5860 + }, + { + "epoch": 0.9583673469387755, + "grad_norm": 0.1627626270055771, + "learning_rate": 2.2239787127495695e-08, + "loss": 0.03181464076042175, + "step": 5870 + }, + { + "epoch": 0.96, + "grad_norm": 0.17413999140262604, + "learning_rate": 2.0538575544831186e-08, + "loss": 0.019847501814365388, + "step": 5880 + }, + { + "epoch": 0.9616326530612245, + "grad_norm": 0.23201079666614532, + "learning_rate": 1.8904776707655837e-08, + "loss": 0.02205444723367691, + "step": 5890 + }, + { + "epoch": 0.963265306122449, + "grad_norm": 0.17272613942623138, + "learning_rate": 1.7338435036449164e-08, + "loss": 0.022021512687206268, + "step": 5900 + }, + { + "epoch": 0.9648979591836735, + "grad_norm": 0.1446181833744049, + "learning_rate": 1.5839593117634166e-08, + "loss": 0.02300398051738739, + "step": 5910 + }, + { + "epoch": 0.966530612244898, + "grad_norm": 0.21913999319076538, + "learning_rate": 1.4408291702418808e-08, + "loss": 0.024825331568717957, + "step": 5920 + }, + { + "epoch": 0.9681632653061224, + "grad_norm": 0.12140090018510818, + "learning_rate": 1.3044569705689125e-08, + "loss": 0.02171301543712616, + "step": 5930 + }, + { + "epoch": 0.9697959183673469, + "grad_norm": 0.08353517204523087, + "learning_rate": 1.1748464204950072e-08, + "loss": 0.02564157545566559, + "step": 5940 + }, + { + "epoch": 0.9714285714285714, + "grad_norm": 0.2671544551849365, + "learning_rate": 1.052001043931855e-08, + "loss": 0.02467280924320221, + "step": 5950 + }, + { + "epoch": 0.9730612244897959, + "grad_norm": 0.16360966861248016, + "learning_rate": 9.3592418085639e-09, + "loss": 0.02429400235414505, + "step": 5960 + }, + { + "epoch": 0.9746938775510204, + "grad_norm": 0.08739493787288666, + "learning_rate": 8.266189872201957e-09, + "loss": 0.02331821173429489, + "step": 5970 + }, + { + "epoch": 0.976326530612245, + "grad_norm": 0.23632948100566864, + "learning_rate": 7.24088434863407e-09, + "loss": 0.02990926504135132, + "step": 5980 + }, + { + "epoch": 0.9779591836734693, + "grad_norm": 0.2005041390657425, + "learning_rate": 6.283353114342749e-09, + "loss": 0.023541297018527984, + "step": 5990 + }, + { + "epoch": 0.9795918367346939, + "grad_norm": 0.16744142770767212, + "learning_rate": 5.393622203129778e-09, + "loss": 0.020145678520202638, + "step": 6000 + }, + { + "epoch": 0.9795918367346939, + "eval_accuracy": 0.9904166666666667, + "eval_loss": 0.025301460176706314, + "eval_runtime": 157.9993, + "eval_samples_per_second": 25.317, + "eval_steps_per_second": 25.317, + "step": 6000 + } + ], + "logging_steps": 10, + "max_steps": 6125, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 200, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 5.60726116663296e+16, + "train_batch_size": 1, + "trial_name": null, + "trial_params": null +} diff --git a/checkpoint-6000/training_args.bin b/checkpoint-6000/training_args.bin new file mode 100644 index 0000000..a9304c0 --- /dev/null +++ b/checkpoint-6000/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:10a2085a9c347d21774d35a2558a0376b0ffbb33ca9517fafbd4cd5d12552b02 +size 5585 diff --git a/checkpoint-6125/chat_template.jinja b/checkpoint-6125/chat_template.jinja new file mode 100644 index 0000000..ce566c4 --- /dev/null +++ b/checkpoint-6125/chat_template.jinja @@ -0,0 +1,9 @@ +{% set system_message = 'Below is an instruction that describes a task. Write a response that appropriately completes the request. + +' %}{% if messages[0]['role'] == 'system' %}{% set loop_messages = messages[1:] %}{% set system_message = messages[0]['content'] %}{% else %}{% set loop_messages = messages %}{% endif %}{% if system_message is defined %}{{ system_message }}{% endif %}{% for message in loop_messages %}{% set content = message['content'] %}{% if message['role'] == 'user' %}{{ '### Instruction: +' + content + ' + +### Response: +' }}{% elif message['role'] == 'assistant' %}{{ content + '<|endoftext|>' + ' + +' }}{% endif %}{% endfor %} \ No newline at end of file diff --git a/checkpoint-6125/config.json b/checkpoint-6125/config.json new file mode 100644 index 0000000..642be54 --- /dev/null +++ b/checkpoint-6125/config.json @@ -0,0 +1,58 @@ +{ + "architectures": [ + "Qwen2ForCausalLM" + ], + "attention_dropout": 0.0, + "bos_token_id": null, + "dtype": "float32", + "eos_token_id": 151643, + "hidden_act": "silu", + "hidden_size": 896, + "initializer_range": 0.02, + "intermediate_size": 4864, + "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" + ], + "max_position_embeddings": 32768, + "max_window_layers": 24, + "model_type": "qwen2", + "num_attention_heads": 14, + "num_hidden_layers": 24, + "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_mrope": false, + "use_sliding_window": false, + "vocab_size": 151936 +} diff --git a/checkpoint-6125/generation_config.json b/checkpoint-6125/generation_config.json new file mode 100644 index 0000000..795a75e --- /dev/null +++ b/checkpoint-6125/generation_config.json @@ -0,0 +1,9 @@ +{ + "do_sample": false, + "eos_token_id": [ + 151643 + ], + "max_new_tokens": 2048, + "pad_token_id": 151643, + "transformers_version": "5.2.0" +} diff --git a/checkpoint-6125/model.safetensors b/checkpoint-6125/model.safetensors new file mode 100644 index 0000000..665761c --- /dev/null +++ b/checkpoint-6125/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1edd8c632dc7b25b22dc2c60673de1f976549056f1258360ed58b7276a7cd06c +size 1976163472 diff --git a/checkpoint-6125/tokenizer.json b/checkpoint-6125/tokenizer.json new file mode 100644 index 0000000..34510ff --- /dev/null +++ b/checkpoint-6125/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3fd169731d2cbde95e10bf356d66d5997fd885dd8dbb6fb4684da3f23b2585d8 +size 11421892 diff --git a/checkpoint-6125/tokenizer_config.json b/checkpoint-6125/tokenizer_config.json new file mode 100644 index 0000000..392ded1 --- /dev/null +++ b/checkpoint-6125/tokenizer_config.json @@ -0,0 +1,30 @@ +{ + "add_prefix_space": false, + "backend": "tokenizers", + "bos_token": null, + "clean_up_tokenization_spaces": false, + "eos_token": "<|endoftext|>", + "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/checkpoint-6125/trainer_state.json b/checkpoint-6125/trainer_state.json new file mode 100644 index 0000000..92c9656 --- /dev/null +++ b/checkpoint-6125/trainer_state.json @@ -0,0 +1,4588 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.0, + "eval_steps": 200, + "global_step": 6125, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0016326530612244899, + "grad_norm": 7.955283164978027, + "learning_rate": 4.5000000000000003e-07, + "loss": 4.391213607788086, + "step": 10 + }, + { + "epoch": 0.0032653061224489797, + "grad_norm": 2.4868686199188232, + "learning_rate": 9.500000000000001e-07, + "loss": 1.9876020431518555, + "step": 20 + }, + { + "epoch": 0.004897959183673469, + "grad_norm": 0.25438952445983887, + "learning_rate": 1.45e-06, + "loss": 0.34127225875854494, + "step": 30 + }, + { + "epoch": 0.006530612244897959, + "grad_norm": 0.13044093549251556, + "learning_rate": 1.9500000000000004e-06, + "loss": 0.23643033504486083, + "step": 40 + }, + { + "epoch": 0.00816326530612245, + "grad_norm": 0.17984068393707275, + "learning_rate": 2.4500000000000003e-06, + "loss": 0.23313720226287843, + "step": 50 + }, + { + "epoch": 0.009795918367346938, + "grad_norm": 0.15249308943748474, + "learning_rate": 2.95e-06, + "loss": 0.22351114749908446, + "step": 60 + }, + { + "epoch": 0.011428571428571429, + "grad_norm": 0.07123949378728867, + "learning_rate": 3.45e-06, + "loss": 0.21615636348724365, + "step": 70 + }, + { + "epoch": 0.013061224489795919, + "grad_norm": 0.21456502377986908, + "learning_rate": 3.95e-06, + "loss": 0.21677422523498535, + "step": 80 + }, + { + "epoch": 0.014693877551020407, + "grad_norm": 0.22391553223133087, + "learning_rate": 4.450000000000001e-06, + "loss": 0.21422204971313477, + "step": 90 + }, + { + "epoch": 0.0163265306122449, + "grad_norm": 0.1745220273733139, + "learning_rate": 4.95e-06, + "loss": 0.19350455999374389, + "step": 100 + }, + { + "epoch": 0.017959183673469388, + "grad_norm": 0.18820920586585999, + "learning_rate": 4.9999724716692435e-06, + "loss": 0.1981007933616638, + "step": 110 + }, + { + "epoch": 0.019591836734693877, + "grad_norm": 0.21733109652996063, + "learning_rate": 4.9998773127856985e-06, + "loss": 0.16994532346725463, + "step": 120 + }, + { + "epoch": 0.02122448979591837, + "grad_norm": 0.40475547313690186, + "learning_rate": 4.999714186080037e-06, + "loss": 0.1539763927459717, + "step": 130 + }, + { + "epoch": 0.022857142857142857, + "grad_norm": 0.3676598370075226, + "learning_rate": 4.999483095987423e-06, + "loss": 0.16321685314178466, + "step": 140 + }, + { + "epoch": 0.024489795918367346, + "grad_norm": 0.19484871625900269, + "learning_rate": 4.9991840487908426e-06, + "loss": 0.16523454189300538, + "step": 150 + }, + { + "epoch": 0.026122448979591838, + "grad_norm": 0.21928830444812775, + "learning_rate": 4.998817052620928e-06, + "loss": 0.14424897432327272, + "step": 160 + }, + { + "epoch": 0.027755102040816326, + "grad_norm": 0.2790975868701935, + "learning_rate": 4.9983821174557415e-06, + "loss": 0.13810645341873168, + "step": 170 + }, + { + "epoch": 0.029387755102040815, + "grad_norm": 0.21702228486537933, + "learning_rate": 4.997879255120501e-06, + "loss": 0.1318126440048218, + "step": 180 + }, + { + "epoch": 0.031020408163265307, + "grad_norm": 0.16967810690402985, + "learning_rate": 4.997308479287263e-06, + "loss": 0.14249314069747926, + "step": 190 + }, + { + "epoch": 0.0326530612244898, + "grad_norm": 0.14007994532585144, + "learning_rate": 4.996669805474542e-06, + "loss": 0.12708702087402343, + "step": 200 + }, + { + "epoch": 0.0326530612244898, + "eval_accuracy": 0.9408333333333335, + "eval_loss": 0.12691152095794678, + "eval_runtime": 156.68, + "eval_samples_per_second": 25.53, + "eval_steps_per_second": 25.53, + "step": 200 + }, + { + "epoch": 0.03428571428571429, + "grad_norm": 0.2674005627632141, + "learning_rate": 4.9959632510469e-06, + "loss": 0.11019121408462525, + "step": 210 + }, + { + "epoch": 0.035918367346938776, + "grad_norm": 0.12073638290166855, + "learning_rate": 4.995188835214465e-06, + "loss": 0.12032746076583863, + "step": 220 + }, + { + "epoch": 0.037551020408163265, + "grad_norm": 0.11186162382364273, + "learning_rate": 4.994346579032413e-06, + "loss": 0.12122929096221924, + "step": 230 + }, + { + "epoch": 0.03918367346938775, + "grad_norm": 0.09840800613164902, + "learning_rate": 4.993436505400396e-06, + "loss": 0.11416752338409424, + "step": 240 + }, + { + "epoch": 0.04081632653061224, + "grad_norm": 0.19076374173164368, + "learning_rate": 4.992458639061916e-06, + "loss": 0.11376930475234985, + "step": 250 + }, + { + "epoch": 0.04244897959183674, + "grad_norm": 0.14405065774917603, + "learning_rate": 4.991413006603657e-06, + "loss": 0.10997685194015502, + "step": 260 + }, + { + "epoch": 0.044081632653061226, + "grad_norm": 0.23844055831432343, + "learning_rate": 4.990299636454759e-06, + "loss": 0.1000083327293396, + "step": 270 + }, + { + "epoch": 0.045714285714285714, + "grad_norm": 0.23128052055835724, + "learning_rate": 4.989118558886043e-06, + "loss": 0.1243711233139038, + "step": 280 + }, + { + "epoch": 0.0473469387755102, + "grad_norm": 0.19849570095539093, + "learning_rate": 4.9878698060091955e-06, + "loss": 0.11728224754333497, + "step": 290 + }, + { + "epoch": 0.04897959183673469, + "grad_norm": 0.13039851188659668, + "learning_rate": 4.986553411775886e-06, + "loss": 0.09353847503662109, + "step": 300 + }, + { + "epoch": 0.05061224489795919, + "grad_norm": 0.19887588918209076, + "learning_rate": 4.985169411976852e-06, + "loss": 0.11291944980621338, + "step": 310 + }, + { + "epoch": 0.052244897959183675, + "grad_norm": 0.25038304924964905, + "learning_rate": 4.98371784424092e-06, + "loss": 0.11149277687072753, + "step": 320 + }, + { + "epoch": 0.053877551020408164, + "grad_norm": 0.238360196352005, + "learning_rate": 4.982198748033986e-06, + "loss": 0.11344953775405883, + "step": 330 + }, + { + "epoch": 0.05551020408163265, + "grad_norm": 0.13686658442020416, + "learning_rate": 4.980612164657943e-06, + "loss": 0.09844974875450134, + "step": 340 + }, + { + "epoch": 0.05714285714285714, + "grad_norm": 0.0829666405916214, + "learning_rate": 4.978958137249551e-06, + "loss": 0.10898438692092896, + "step": 350 + }, + { + "epoch": 0.05877551020408163, + "grad_norm": 0.10389993339776993, + "learning_rate": 4.977236710779277e-06, + "loss": 0.09371156096458436, + "step": 360 + }, + { + "epoch": 0.060408163265306125, + "grad_norm": 0.18182680010795593, + "learning_rate": 4.9754479320500584e-06, + "loss": 0.1049533724784851, + "step": 370 + }, + { + "epoch": 0.062040816326530614, + "grad_norm": 0.14488115906715393, + "learning_rate": 4.973591849696041e-06, + "loss": 0.10510741472244263, + "step": 380 + }, + { + "epoch": 0.0636734693877551, + "grad_norm": 0.14816738665103912, + "learning_rate": 4.971668514181252e-06, + "loss": 0.10956356525421143, + "step": 390 + }, + { + "epoch": 0.0653061224489796, + "grad_norm": 0.20584072172641754, + "learning_rate": 4.969677977798225e-06, + "loss": 0.11344989538192748, + "step": 400 + }, + { + "epoch": 0.0653061224489796, + "eval_accuracy": 0.9496250000000002, + "eval_loss": 0.11409253627061844, + "eval_runtime": 165.7682, + "eval_samples_per_second": 24.13, + "eval_steps_per_second": 24.13, + "step": 400 + }, + { + "epoch": 0.06693877551020408, + "grad_norm": 0.1342787891626358, + "learning_rate": 4.967620294666589e-06, + "loss": 0.09949527978897095, + "step": 410 + }, + { + "epoch": 0.06857142857142857, + "grad_norm": 0.17336596548557281, + "learning_rate": 4.965495520731583e-06, + "loss": 0.10606559514999389, + "step": 420 + }, + { + "epoch": 0.07020408163265306, + "grad_norm": 0.19437003135681152, + "learning_rate": 4.963303713762547e-06, + "loss": 0.11677042245864869, + "step": 430 + }, + { + "epoch": 0.07183673469387755, + "grad_norm": 0.12076073884963989, + "learning_rate": 4.961044933351342e-06, + "loss": 0.1019243597984314, + "step": 440 + }, + { + "epoch": 0.07346938775510205, + "grad_norm": 0.10272067785263062, + "learning_rate": 4.958719240910737e-06, + "loss": 0.10125980377197266, + "step": 450 + }, + { + "epoch": 0.07510204081632653, + "grad_norm": 0.1080622524023056, + "learning_rate": 4.956326699672736e-06, + "loss": 0.09888594150543213, + "step": 460 + }, + { + "epoch": 0.07673469387755102, + "grad_norm": 0.1307847797870636, + "learning_rate": 4.953867374686856e-06, + "loss": 0.10753540992736817, + "step": 470 + }, + { + "epoch": 0.0783673469387755, + "grad_norm": 0.2128073275089264, + "learning_rate": 4.951341332818365e-06, + "loss": 0.11452326774597169, + "step": 480 + }, + { + "epoch": 0.08, + "grad_norm": 0.21006478369235992, + "learning_rate": 4.9487486427464594e-06, + "loss": 0.09555010795593262, + "step": 490 + }, + { + "epoch": 0.08163265306122448, + "grad_norm": 0.24889807403087616, + "learning_rate": 4.946089374962396e-06, + "loss": 0.09771155118942261, + "step": 500 + }, + { + "epoch": 0.08326530612244898, + "grad_norm": 0.13042858242988586, + "learning_rate": 4.943363601767582e-06, + "loss": 0.10659778118133545, + "step": 510 + }, + { + "epoch": 0.08489795918367347, + "grad_norm": 0.2210501879453659, + "learning_rate": 4.940571397271598e-06, + "loss": 0.10161921977996827, + "step": 520 + }, + { + "epoch": 0.08653061224489796, + "grad_norm": 0.11963305622339249, + "learning_rate": 4.937712837390194e-06, + "loss": 0.09155921936035157, + "step": 530 + }, + { + "epoch": 0.08816326530612245, + "grad_norm": 0.07435746490955353, + "learning_rate": 4.93478799984322e-06, + "loss": 0.11164016723632812, + "step": 540 + }, + { + "epoch": 0.08979591836734693, + "grad_norm": 0.16348940134048462, + "learning_rate": 4.9317969641525154e-06, + "loss": 0.10738091468811035, + "step": 550 + }, + { + "epoch": 0.09142857142857143, + "grad_norm": 0.12221129983663559, + "learning_rate": 4.928739811639741e-06, + "loss": 0.09491395950317383, + "step": 560 + }, + { + "epoch": 0.09306122448979592, + "grad_norm": 0.1107218861579895, + "learning_rate": 4.9256166254241775e-06, + "loss": 0.08408294320106506, + "step": 570 + }, + { + "epoch": 0.0946938775510204, + "grad_norm": 0.207329660654068, + "learning_rate": 4.922427490420457e-06, + "loss": 0.10272791385650634, + "step": 580 + }, + { + "epoch": 0.0963265306122449, + "grad_norm": 0.14269684255123138, + "learning_rate": 4.919172493336259e-06, + "loss": 0.10189608335494996, + "step": 590 + }, + { + "epoch": 0.09795918367346938, + "grad_norm": 0.10970286279916763, + "learning_rate": 4.915851722669952e-06, + "loss": 0.09506216645240784, + "step": 600 + }, + { + "epoch": 0.09795918367346938, + "eval_accuracy": 0.9561666666666667, + "eval_loss": 0.09913250058889389, + "eval_runtime": 158.3659, + "eval_samples_per_second": 25.258, + "eval_steps_per_second": 25.258, + "step": 600 + }, + { + "epoch": 0.09959183673469388, + "grad_norm": 0.11382903158664703, + "learning_rate": 4.912465268708186e-06, + "loss": 0.09665942788124085, + "step": 610 + }, + { + "epoch": 0.10122448979591837, + "grad_norm": 0.24785904586315155, + "learning_rate": 4.90901322352344e-06, + "loss": 0.10539349317550659, + "step": 620 + }, + { + "epoch": 0.10285714285714286, + "grad_norm": 0.18585072457790375, + "learning_rate": 4.905495680971518e-06, + "loss": 0.10907092094421386, + "step": 630 + }, + { + "epoch": 0.10448979591836735, + "grad_norm": 0.16385921835899353, + "learning_rate": 4.901912736688995e-06, + "loss": 0.09973281621932983, + "step": 640 + }, + { + "epoch": 0.10612244897959183, + "grad_norm": 0.10820458829402924, + "learning_rate": 4.898264488090619e-06, + "loss": 0.09869267344474793, + "step": 650 + }, + { + "epoch": 0.10775510204081633, + "grad_norm": 0.18020115792751312, + "learning_rate": 4.894551034366665e-06, + "loss": 0.10499387979507446, + "step": 660 + }, + { + "epoch": 0.10938775510204081, + "grad_norm": 0.10356947779655457, + "learning_rate": 4.890772476480229e-06, + "loss": 0.10153095722198487, + "step": 670 + }, + { + "epoch": 0.1110204081632653, + "grad_norm": 0.08888879418373108, + "learning_rate": 4.886928917164496e-06, + "loss": 0.08613221645355225, + "step": 680 + }, + { + "epoch": 0.1126530612244898, + "grad_norm": 0.11978956311941147, + "learning_rate": 4.8830204609199335e-06, + "loss": 0.09545036554336547, + "step": 690 + }, + { + "epoch": 0.11428571428571428, + "grad_norm": 0.11924679577350616, + "learning_rate": 4.8790472140114635e-06, + "loss": 0.0950542688369751, + "step": 700 + }, + { + "epoch": 0.11591836734693878, + "grad_norm": 0.0637463852763176, + "learning_rate": 4.875009284465561e-06, + "loss": 0.10127921104431152, + "step": 710 + }, + { + "epoch": 0.11755102040816326, + "grad_norm": 0.0894392654299736, + "learning_rate": 4.8709067820673235e-06, + "loss": 0.09835844039916992, + "step": 720 + }, + { + "epoch": 0.11918367346938775, + "grad_norm": 0.06312194466590881, + "learning_rate": 4.866739818357489e-06, + "loss": 0.08583117127418519, + "step": 730 + }, + { + "epoch": 0.12081632653061225, + "grad_norm": 0.1375141739845276, + "learning_rate": 4.862508506629394e-06, + "loss": 0.08735222220420838, + "step": 740 + }, + { + "epoch": 0.12244897959183673, + "grad_norm": 0.12715794146060944, + "learning_rate": 4.858212961925902e-06, + "loss": 0.09324318170547485, + "step": 750 + }, + { + "epoch": 0.12408163265306123, + "grad_norm": 0.14359286427497864, + "learning_rate": 4.8538533010362735e-06, + "loss": 0.09482367634773255, + "step": 760 + }, + { + "epoch": 0.12571428571428572, + "grad_norm": 0.1228424683213234, + "learning_rate": 4.849429642492985e-06, + "loss": 0.09595230221748352, + "step": 770 + }, + { + "epoch": 0.1273469387755102, + "grad_norm": 0.08908221125602722, + "learning_rate": 4.844942106568517e-06, + "loss": 0.08924583196640015, + "step": 780 + }, + { + "epoch": 0.12897959183673469, + "grad_norm": 0.15417414903640747, + "learning_rate": 4.84039081527207e-06, + "loss": 0.09454164505004883, + "step": 790 + }, + { + "epoch": 0.1306122448979592, + "grad_norm": 0.1259608417749405, + "learning_rate": 4.835775892346264e-06, + "loss": 0.08659166097640991, + "step": 800 + }, + { + "epoch": 0.1306122448979592, + "eval_accuracy": 0.9564166666666667, + "eval_loss": 0.09297937154769897, + "eval_runtime": 157.5491, + "eval_samples_per_second": 25.389, + "eval_steps_per_second": 25.389, + "step": 800 + }, + { + "epoch": 0.13224489795918368, + "grad_norm": 0.15664513409137726, + "learning_rate": 4.8310974632637584e-06, + "loss": 0.09094155430793763, + "step": 810 + }, + { + "epoch": 0.13387755102040816, + "grad_norm": 0.12366823107004166, + "learning_rate": 4.82635565522385e-06, + "loss": 0.09823386669158936, + "step": 820 + }, + { + "epoch": 0.13551020408163264, + "grad_norm": 0.09367755055427551, + "learning_rate": 4.821550597149012e-06, + "loss": 0.09080367088317871, + "step": 830 + }, + { + "epoch": 0.13714285714285715, + "grad_norm": 0.15230220556259155, + "learning_rate": 4.816682419681389e-06, + "loss": 0.09046205282211303, + "step": 840 + }, + { + "epoch": 0.13877551020408163, + "grad_norm": 0.09214162826538086, + "learning_rate": 4.811751255179244e-06, + "loss": 0.09502801299095154, + "step": 850 + }, + { + "epoch": 0.1404081632653061, + "grad_norm": 0.09514431655406952, + "learning_rate": 4.806757237713361e-06, + "loss": 0.09261927604675294, + "step": 860 + }, + { + "epoch": 0.14204081632653062, + "grad_norm": 0.13052736222743988, + "learning_rate": 4.8017005030634e-06, + "loss": 0.08243635892868043, + "step": 870 + }, + { + "epoch": 0.1436734693877551, + "grad_norm": 0.05511025711894035, + "learning_rate": 4.796581188714205e-06, + "loss": 0.10203326940536499, + "step": 880 + }, + { + "epoch": 0.14530612244897959, + "grad_norm": 0.09896261245012283, + "learning_rate": 4.7913994338520655e-06, + "loss": 0.08544076681137085, + "step": 890 + }, + { + "epoch": 0.1469387755102041, + "grad_norm": 0.18726249039173126, + "learning_rate": 4.786155379360932e-06, + "loss": 0.0944068193435669, + "step": 900 + }, + { + "epoch": 0.14857142857142858, + "grad_norm": 0.09349548071622849, + "learning_rate": 4.7808491678185855e-06, + "loss": 0.0877715528011322, + "step": 910 + }, + { + "epoch": 0.15020408163265306, + "grad_norm": 0.13616853952407837, + "learning_rate": 4.775480943492764e-06, + "loss": 0.09196346998214722, + "step": 920 + }, + { + "epoch": 0.15183673469387754, + "grad_norm": 0.061041463166475296, + "learning_rate": 4.770050852337236e-06, + "loss": 0.09890784621238709, + "step": 930 + }, + { + "epoch": 0.15346938775510205, + "grad_norm": 0.09928455948829651, + "learning_rate": 4.764559041987836e-06, + "loss": 0.09260086417198181, + "step": 940 + }, + { + "epoch": 0.15510204081632653, + "grad_norm": 0.10996410250663757, + "learning_rate": 4.759005661758446e-06, + "loss": 0.09412651658058166, + "step": 950 + }, + { + "epoch": 0.156734693877551, + "grad_norm": 0.09731809049844742, + "learning_rate": 4.753390862636939e-06, + "loss": 0.09459404945373535, + "step": 960 + }, + { + "epoch": 0.15836734693877552, + "grad_norm": 0.12398040294647217, + "learning_rate": 4.747714797281076e-06, + "loss": 0.09960640072822571, + "step": 970 + }, + { + "epoch": 0.16, + "grad_norm": 0.14135490357875824, + "learning_rate": 4.741977620014347e-06, + "loss": 0.09338371157646179, + "step": 980 + }, + { + "epoch": 0.16163265306122448, + "grad_norm": 0.09384343028068542, + "learning_rate": 4.736179486821788e-06, + "loss": 0.08515415191650391, + "step": 990 + }, + { + "epoch": 0.16326530612244897, + "grad_norm": 0.10298261791467667, + "learning_rate": 4.730320555345728e-06, + "loss": 0.09090787768363953, + "step": 1000 + }, + { + "epoch": 0.16326530612244897, + "eval_accuracy": 0.9607916666666667, + "eval_loss": 0.08832322806119919, + "eval_runtime": 166.8275, + "eval_samples_per_second": 23.977, + "eval_steps_per_second": 23.977, + "step": 1000 + }, + { + "epoch": 0.16489795918367348, + "grad_norm": 0.05896759778261185, + "learning_rate": 4.724400984881508e-06, + "loss": 0.08005568385124207, + "step": 1010 + }, + { + "epoch": 0.16653061224489796, + "grad_norm": 0.07907038927078247, + "learning_rate": 4.7184209363731546e-06, + "loss": 0.08133251667022705, + "step": 1020 + }, + { + "epoch": 0.16816326530612244, + "grad_norm": 0.07444760203361511, + "learning_rate": 4.7123805724089935e-06, + "loss": 0.08690541982650757, + "step": 1030 + }, + { + "epoch": 0.16979591836734695, + "grad_norm": 0.16517433524131775, + "learning_rate": 4.706280057217239e-06, + "loss": 0.09946454763412475, + "step": 1040 + }, + { + "epoch": 0.17142857142857143, + "grad_norm": 0.18107500672340393, + "learning_rate": 4.700119556661521e-06, + "loss": 0.08886862993240356, + "step": 1050 + }, + { + "epoch": 0.1730612244897959, + "grad_norm": 0.15386350452899933, + "learning_rate": 4.693899238236383e-06, + "loss": 0.09117826819419861, + "step": 1060 + }, + { + "epoch": 0.17469387755102042, + "grad_norm": 0.1098460704088211, + "learning_rate": 4.687619271062725e-06, + "loss": 0.09286551475524903, + "step": 1070 + }, + { + "epoch": 0.1763265306122449, + "grad_norm": 0.09872590750455856, + "learning_rate": 4.681279825883201e-06, + "loss": 0.09102073907852173, + "step": 1080 + }, + { + "epoch": 0.17795918367346938, + "grad_norm": 0.12001907080411911, + "learning_rate": 4.674881075057581e-06, + "loss": 0.09481253027915955, + "step": 1090 + }, + { + "epoch": 0.17959183673469387, + "grad_norm": 0.13000014424324036, + "learning_rate": 4.668423192558069e-06, + "loss": 0.08999797105789184, + "step": 1100 + }, + { + "epoch": 0.18122448979591838, + "grad_norm": 0.07537361979484558, + "learning_rate": 4.661906353964563e-06, + "loss": 0.07345216870307922, + "step": 1110 + }, + { + "epoch": 0.18285714285714286, + "grad_norm": 0.10346294194459915, + "learning_rate": 4.65533073645989e-06, + "loss": 0.08096543550491334, + "step": 1120 + }, + { + "epoch": 0.18448979591836734, + "grad_norm": 0.12576258182525635, + "learning_rate": 4.6486965188249825e-06, + "loss": 0.08680164813995361, + "step": 1130 + }, + { + "epoch": 0.18612244897959185, + "grad_norm": 0.17314419150352478, + "learning_rate": 4.642003881434024e-06, + "loss": 0.08470291495323182, + "step": 1140 + }, + { + "epoch": 0.18775510204081633, + "grad_norm": 0.1576225906610489, + "learning_rate": 4.635253006249536e-06, + "loss": 0.10852066278457642, + "step": 1150 + }, + { + "epoch": 0.1893877551020408, + "grad_norm": 0.15372557938098907, + "learning_rate": 4.628444076817444e-06, + "loss": 0.1022921085357666, + "step": 1160 + }, + { + "epoch": 0.1910204081632653, + "grad_norm": 0.15021173655986786, + "learning_rate": 4.621577278262072e-06, + "loss": 0.09645735621452331, + "step": 1170 + }, + { + "epoch": 0.1926530612244898, + "grad_norm": 0.07714739441871643, + "learning_rate": 4.614652797281118e-06, + "loss": 0.08636590242385864, + "step": 1180 + }, + { + "epoch": 0.19428571428571428, + "grad_norm": 0.17956240475177765, + "learning_rate": 4.6076708221405806e-06, + "loss": 0.09293715357780456, + "step": 1190 + }, + { + "epoch": 0.19591836734693877, + "grad_norm": 0.13412988185882568, + "learning_rate": 4.600631542669633e-06, + "loss": 0.09799563884735107, + "step": 1200 + }, + { + "epoch": 0.19591836734693877, + "eval_accuracy": 0.963, + "eval_loss": 0.08563050627708435, + "eval_runtime": 171.1854, + "eval_samples_per_second": 23.366, + "eval_steps_per_second": 23.366, + "step": 1200 + }, + { + "epoch": 0.19755102040816327, + "grad_norm": 0.1682359278202057, + "learning_rate": 4.593535150255463e-06, + "loss": 0.08345703482627868, + "step": 1210 + }, + { + "epoch": 0.19918367346938776, + "grad_norm": 0.09281504899263382, + "learning_rate": 4.586381837838076e-06, + "loss": 0.08114275336265564, + "step": 1220 + }, + { + "epoch": 0.20081632653061224, + "grad_norm": 0.12355420738458633, + "learning_rate": 4.579171799905044e-06, + "loss": 0.0780484676361084, + "step": 1230 + }, + { + "epoch": 0.20244897959183675, + "grad_norm": 0.12121636420488358, + "learning_rate": 4.571905232486214e-06, + "loss": 0.07816853523254394, + "step": 1240 + }, + { + "epoch": 0.20408163265306123, + "grad_norm": 0.09326840937137604, + "learning_rate": 4.56458233314839e-06, + "loss": 0.08001847267150879, + "step": 1250 + }, + { + "epoch": 0.2057142857142857, + "grad_norm": 0.11423090845346451, + "learning_rate": 4.557203300989948e-06, + "loss": 0.08720387816429138, + "step": 1260 + }, + { + "epoch": 0.2073469387755102, + "grad_norm": 0.124808169901371, + "learning_rate": 4.549768336635432e-06, + "loss": 0.08812607526779175, + "step": 1270 + }, + { + "epoch": 0.2089795918367347, + "grad_norm": 0.08311110734939575, + "learning_rate": 4.542277642230096e-06, + "loss": 0.0880801498889923, + "step": 1280 + }, + { + "epoch": 0.21061224489795918, + "grad_norm": 0.1179559975862503, + "learning_rate": 4.53473142143441e-06, + "loss": 0.09044411182403564, + "step": 1290 + }, + { + "epoch": 0.21224489795918366, + "grad_norm": 0.13122901320457458, + "learning_rate": 4.527129879418519e-06, + "loss": 0.06792758107185363, + "step": 1300 + }, + { + "epoch": 0.21387755102040817, + "grad_norm": 0.10990814864635468, + "learning_rate": 4.519473222856667e-06, + "loss": 0.07861824035644531, + "step": 1310 + }, + { + "epoch": 0.21551020408163266, + "grad_norm": 0.14613105356693268, + "learning_rate": 4.51176165992158e-06, + "loss": 0.08266589641571045, + "step": 1320 + }, + { + "epoch": 0.21714285714285714, + "grad_norm": 0.07217947393655777, + "learning_rate": 4.503995400278801e-06, + "loss": 0.08272708654403686, + "step": 1330 + }, + { + "epoch": 0.21877551020408162, + "grad_norm": 0.12690335512161255, + "learning_rate": 4.4961746550809945e-06, + "loss": 0.08644292950630188, + "step": 1340 + }, + { + "epoch": 0.22040816326530613, + "grad_norm": 0.11895330995321274, + "learning_rate": 4.488299636962203e-06, + "loss": 0.08812489509582519, + "step": 1350 + }, + { + "epoch": 0.2220408163265306, + "grad_norm": 0.08727742731571198, + "learning_rate": 4.4803705600320655e-06, + "loss": 0.08752686977386474, + "step": 1360 + }, + { + "epoch": 0.2236734693877551, + "grad_norm": 0.1101253554224968, + "learning_rate": 4.472387639869998e-06, + "loss": 0.08365487456321716, + "step": 1370 + }, + { + "epoch": 0.2253061224489796, + "grad_norm": 0.0881471335887909, + "learning_rate": 4.464351093519332e-06, + "loss": 0.08101181387901306, + "step": 1380 + }, + { + "epoch": 0.22693877551020408, + "grad_norm": 0.1676328480243683, + "learning_rate": 4.45626113948141e-06, + "loss": 0.08873040080070496, + "step": 1390 + }, + { + "epoch": 0.22857142857142856, + "grad_norm": 0.13493730127811432, + "learning_rate": 4.44811799770965e-06, + "loss": 0.08360225558280945, + "step": 1400 + }, + { + "epoch": 0.22857142857142856, + "eval_accuracy": 0.961125, + "eval_loss": 0.0838993564248085, + "eval_runtime": 159.3888, + "eval_samples_per_second": 25.096, + "eval_steps_per_second": 25.096, + "step": 1400 + }, + { + "epoch": 0.23020408163265307, + "grad_norm": 0.18087805807590485, + "learning_rate": 4.4399218896035615e-06, + "loss": 0.07942168116569519, + "step": 1410 + }, + { + "epoch": 0.23183673469387756, + "grad_norm": 0.15980401635169983, + "learning_rate": 4.431673038002726e-06, + "loss": 0.07959137558937072, + "step": 1420 + }, + { + "epoch": 0.23346938775510204, + "grad_norm": 0.09756626188755035, + "learning_rate": 4.423371667180743e-06, + "loss": 0.08047830462455749, + "step": 1430 + }, + { + "epoch": 0.23510204081632652, + "grad_norm": 0.11922340840101242, + "learning_rate": 4.415018002839127e-06, + "loss": 0.07832802534103393, + "step": 1440 + }, + { + "epoch": 0.23673469387755103, + "grad_norm": 0.09443783015012741, + "learning_rate": 4.40661227210117e-06, + "loss": 0.07132846713066102, + "step": 1450 + }, + { + "epoch": 0.2383673469387755, + "grad_norm": 0.15397194027900696, + "learning_rate": 4.398154703505776e-06, + "loss": 0.09222683906555176, + "step": 1460 + }, + { + "epoch": 0.24, + "grad_norm": 0.06159692630171776, + "learning_rate": 4.389645527001234e-06, + "loss": 0.07937920093536377, + "step": 1470 + }, + { + "epoch": 0.2416326530612245, + "grad_norm": 0.11855911463499069, + "learning_rate": 4.381084973938976e-06, + "loss": 0.07650941014289855, + "step": 1480 + }, + { + "epoch": 0.24326530612244898, + "grad_norm": 0.1305486559867859, + "learning_rate": 4.372473277067285e-06, + "loss": 0.07730157375335693, + "step": 1490 + }, + { + "epoch": 0.24489795918367346, + "grad_norm": 0.09036920964717865, + "learning_rate": 4.363810670524962e-06, + "loss": 0.07192737460136414, + "step": 1500 + }, + { + "epoch": 0.24653061224489795, + "grad_norm": 0.0936962217092514, + "learning_rate": 4.355097389834966e-06, + "loss": 0.09040454030036926, + "step": 1510 + }, + { + "epoch": 0.24816326530612245, + "grad_norm": 0.0778910219669342, + "learning_rate": 4.346333671898007e-06, + "loss": 0.06948431730270385, + "step": 1520 + }, + { + "epoch": 0.24979591836734694, + "grad_norm": 0.0979679748415947, + "learning_rate": 4.337519754986108e-06, + "loss": 0.07317689657211304, + "step": 1530 + }, + { + "epoch": 0.25142857142857145, + "grad_norm": 0.15620048344135284, + "learning_rate": 4.328655878736123e-06, + "loss": 0.09126524925231934, + "step": 1540 + }, + { + "epoch": 0.2530612244897959, + "grad_norm": 0.09080824255943298, + "learning_rate": 4.319742284143226e-06, + "loss": 0.09065211415290833, + "step": 1550 + }, + { + "epoch": 0.2546938775510204, + "grad_norm": 0.046968940645456314, + "learning_rate": 4.310779213554353e-06, + "loss": 0.0718291699886322, + "step": 1560 + }, + { + "epoch": 0.2563265306122449, + "grad_norm": 0.11479081958532333, + "learning_rate": 4.301766910661619e-06, + "loss": 0.07913097143173217, + "step": 1570 + }, + { + "epoch": 0.25795918367346937, + "grad_norm": 0.1042279377579689, + "learning_rate": 4.292705620495689e-06, + "loss": 0.06828063130378723, + "step": 1580 + }, + { + "epoch": 0.2595918367346939, + "grad_norm": 0.17096509039402008, + "learning_rate": 4.283595589419115e-06, + "loss": 0.07042163014411926, + "step": 1590 + }, + { + "epoch": 0.2612244897959184, + "grad_norm": 0.10892639309167862, + "learning_rate": 4.274437065119643e-06, + "loss": 0.07090482115745544, + "step": 1600 + }, + { + "epoch": 0.2612244897959184, + "eval_accuracy": 0.964375, + "eval_loss": 0.08112655580043793, + "eval_runtime": 158.0164, + "eval_samples_per_second": 25.314, + "eval_steps_per_second": 25.314, + "step": 1600 + }, + { + "epoch": 0.26285714285714284, + "grad_norm": 0.09731441736221313, + "learning_rate": 4.265230296603471e-06, + "loss": 0.08803416490554809, + "step": 1610 + }, + { + "epoch": 0.26448979591836735, + "grad_norm": 0.0956069752573967, + "learning_rate": 4.255975534188488e-06, + "loss": 0.0734250545501709, + "step": 1620 + }, + { + "epoch": 0.26612244897959186, + "grad_norm": 0.14361435174942017, + "learning_rate": 4.246673029497458e-06, + "loss": 0.07703657150268554, + "step": 1630 + }, + { + "epoch": 0.2677551020408163, + "grad_norm": 0.08011971414089203, + "learning_rate": 4.237323035451188e-06, + "loss": 0.08410758376121522, + "step": 1640 + }, + { + "epoch": 0.2693877551020408, + "grad_norm": 0.1244175136089325, + "learning_rate": 4.2279258062616455e-06, + "loss": 0.0808764636516571, + "step": 1650 + }, + { + "epoch": 0.2710204081632653, + "grad_norm": 0.11162611097097397, + "learning_rate": 4.21848159742505e-06, + "loss": 0.07636536359786987, + "step": 1660 + }, + { + "epoch": 0.2726530612244898, + "grad_norm": 0.16819386184215546, + "learning_rate": 4.208990665714927e-06, + "loss": 0.07789689898490906, + "step": 1670 + }, + { + "epoch": 0.2742857142857143, + "grad_norm": 0.07522564381361008, + "learning_rate": 4.19945326917512e-06, + "loss": 0.07442367672920228, + "step": 1680 + }, + { + "epoch": 0.27591836734693875, + "grad_norm": 0.08352494984865189, + "learning_rate": 4.1898696671127855e-06, + "loss": 0.07646024227142334, + "step": 1690 + }, + { + "epoch": 0.27755102040816326, + "grad_norm": 0.12580762803554535, + "learning_rate": 4.180240120091333e-06, + "loss": 0.07484800815582275, + "step": 1700 + }, + { + "epoch": 0.27918367346938777, + "grad_norm": 0.14690619707107544, + "learning_rate": 4.170564889923348e-06, + "loss": 0.07452926635742188, + "step": 1710 + }, + { + "epoch": 0.2808163265306122, + "grad_norm": 0.08760543167591095, + "learning_rate": 4.1608442396634675e-06, + "loss": 0.07579211592674255, + "step": 1720 + }, + { + "epoch": 0.28244897959183674, + "grad_norm": 0.05263754725456238, + "learning_rate": 4.151078433601234e-06, + "loss": 0.08277950882911682, + "step": 1730 + }, + { + "epoch": 0.28408163265306124, + "grad_norm": 0.12929993867874146, + "learning_rate": 4.141267737253905e-06, + "loss": 0.07354526519775391, + "step": 1740 + }, + { + "epoch": 0.2857142857142857, + "grad_norm": 0.1444639414548874, + "learning_rate": 4.1314124173592356e-06, + "loss": 0.07377716898918152, + "step": 1750 + }, + { + "epoch": 0.2873469387755102, + "grad_norm": 0.15250442922115326, + "learning_rate": 4.121512741868229e-06, + "loss": 0.0802634596824646, + "step": 1760 + }, + { + "epoch": 0.2889795918367347, + "grad_norm": 0.14867356419563293, + "learning_rate": 4.111568979937845e-06, + "loss": 0.08141970038414001, + "step": 1770 + }, + { + "epoch": 0.29061224489795917, + "grad_norm": 0.1100233644247055, + "learning_rate": 4.10158140192369e-06, + "loss": 0.07370007634162903, + "step": 1780 + }, + { + "epoch": 0.2922448979591837, + "grad_norm": 0.10663364082574844, + "learning_rate": 4.091550279372659e-06, + "loss": 0.0754891276359558, + "step": 1790 + }, + { + "epoch": 0.2938775510204082, + "grad_norm": 0.07148909568786621, + "learning_rate": 4.081475885015558e-06, + "loss": 0.07567381858825684, + "step": 1800 + }, + { + "epoch": 0.2938775510204082, + "eval_accuracy": 0.9665416666666667, + "eval_loss": 0.07648575305938721, + "eval_runtime": 176.7261, + "eval_samples_per_second": 22.634, + "eval_steps_per_second": 22.634, + "step": 1800 + }, + { + "epoch": 0.29551020408163264, + "grad_norm": 0.07446923106908798, + "learning_rate": 4.071358492759686e-06, + "loss": 0.06706757545471191, + "step": 1810 + }, + { + "epoch": 0.29714285714285715, + "grad_norm": 0.144475519657135, + "learning_rate": 4.061198377681391e-06, + "loss": 0.07407833337783813, + "step": 1820 + }, + { + "epoch": 0.2987755102040816, + "grad_norm": 0.16540028154850006, + "learning_rate": 4.050995816018585e-06, + "loss": 0.07593105435371399, + "step": 1830 + }, + { + "epoch": 0.3004081632653061, + "grad_norm": 0.09237829595804214, + "learning_rate": 4.040751085163239e-06, + "loss": 0.08375502228736878, + "step": 1840 + }, + { + "epoch": 0.3020408163265306, + "grad_norm": 0.11202957481145859, + "learning_rate": 4.0304644636538435e-06, + "loss": 0.0692280650138855, + "step": 1850 + }, + { + "epoch": 0.3036734693877551, + "grad_norm": 0.16075952351093292, + "learning_rate": 4.020136231167824e-06, + "loss": 0.07824563384056091, + "step": 1860 + }, + { + "epoch": 0.3053061224489796, + "grad_norm": 0.12567447125911713, + "learning_rate": 4.0097666685139516e-06, + "loss": 0.07255304455757142, + "step": 1870 + }, + { + "epoch": 0.3069387755102041, + "grad_norm": 0.08912669867277145, + "learning_rate": 3.999356057624696e-06, + "loss": 0.06304961442947388, + "step": 1880 + }, + { + "epoch": 0.30857142857142855, + "grad_norm": 0.15736177563667297, + "learning_rate": 3.988904681548568e-06, + "loss": 0.07315694689750671, + "step": 1890 + }, + { + "epoch": 0.31020408163265306, + "grad_norm": 0.13224704563617706, + "learning_rate": 3.978412824442421e-06, + "loss": 0.07607222199440003, + "step": 1900 + }, + { + "epoch": 0.31183673469387757, + "grad_norm": 0.0942244678735733, + "learning_rate": 3.9678807715637245e-06, + "loss": 0.06320942640304565, + "step": 1910 + }, + { + "epoch": 0.313469387755102, + "grad_norm": 0.1153193786740303, + "learning_rate": 3.95730880926281e-06, + "loss": 0.07607839703559875, + "step": 1920 + }, + { + "epoch": 0.31510204081632653, + "grad_norm": 0.24464164674282074, + "learning_rate": 3.946697224975083e-06, + "loss": 0.07458690404891968, + "step": 1930 + }, + { + "epoch": 0.31673469387755104, + "grad_norm": 0.14584548771381378, + "learning_rate": 3.936046307213215e-06, + "loss": 0.07876120805740357, + "step": 1940 + }, + { + "epoch": 0.3183673469387755, + "grad_norm": 0.0880698636174202, + "learning_rate": 3.925356345559288e-06, + "loss": 0.07082821726799012, + "step": 1950 + }, + { + "epoch": 0.32, + "grad_norm": 0.08619305491447449, + "learning_rate": 3.914627630656931e-06, + "loss": 0.06648402214050293, + "step": 1960 + }, + { + "epoch": 0.3216326530612245, + "grad_norm": 0.1341329663991928, + "learning_rate": 3.903860454203413e-06, + "loss": 0.07000230550765991, + "step": 1970 + }, + { + "epoch": 0.32326530612244897, + "grad_norm": 0.1504107564687729, + "learning_rate": 3.8930551089417145e-06, + "loss": 0.06887786984443664, + "step": 1980 + }, + { + "epoch": 0.3248979591836735, + "grad_norm": 0.1262267827987671, + "learning_rate": 3.882211888652566e-06, + "loss": 0.08991954326629639, + "step": 1990 + }, + { + "epoch": 0.32653061224489793, + "grad_norm": 0.2052707076072693, + "learning_rate": 3.871331088146463e-06, + "loss": 0.07011507153511047, + "step": 2000 + }, + { + "epoch": 0.32653061224489793, + "eval_accuracy": 0.9654583333333334, + "eval_loss": 0.07913381606340408, + "eval_runtime": 161.4221, + "eval_samples_per_second": 24.78, + "eval_steps_per_second": 24.78, + "step": 2000 + }, + { + "epoch": 0.32816326530612244, + "grad_norm": 0.11643362790346146, + "learning_rate": 3.860413003255648e-06, + "loss": 0.08266008496284485, + "step": 2010 + }, + { + "epoch": 0.32979591836734695, + "grad_norm": 0.1472826600074768, + "learning_rate": 3.8494579308260695e-06, + "loss": 0.07043538689613342, + "step": 2020 + }, + { + "epoch": 0.3314285714285714, + "grad_norm": 0.10538945347070694, + "learning_rate": 3.838466168709312e-06, + "loss": 0.06780247688293457, + "step": 2030 + }, + { + "epoch": 0.3330612244897959, + "grad_norm": 0.14366035163402557, + "learning_rate": 3.827438015754491e-06, + "loss": 0.06872215867042542, + "step": 2040 + }, + { + "epoch": 0.3346938775510204, + "grad_norm": 0.11022552847862244, + "learning_rate": 3.81637377180014e-06, + "loss": 0.06861716508865356, + "step": 2050 + }, + { + "epoch": 0.3363265306122449, + "grad_norm": 0.07990925014019012, + "learning_rate": 3.8052737376660446e-06, + "loss": 0.07096996307373046, + "step": 2060 + }, + { + "epoch": 0.3379591836734694, + "grad_norm": 0.11924420297145844, + "learning_rate": 3.7941382151450756e-06, + "loss": 0.07092551589012146, + "step": 2070 + }, + { + "epoch": 0.3395918367346939, + "grad_norm": 0.08096475899219513, + "learning_rate": 3.782967506994975e-06, + "loss": 0.06780472993850709, + "step": 2080 + }, + { + "epoch": 0.34122448979591835, + "grad_norm": 0.14714907109737396, + "learning_rate": 3.7717619169301302e-06, + "loss": 0.07661795020103454, + "step": 2090 + }, + { + "epoch": 0.34285714285714286, + "grad_norm": 0.14839571714401245, + "learning_rate": 3.7605217496133127e-06, + "loss": 0.07939797639846802, + "step": 2100 + }, + { + "epoch": 0.34448979591836737, + "grad_norm": 0.15081271529197693, + "learning_rate": 3.749247310647398e-06, + "loss": 0.07869908213615417, + "step": 2110 + }, + { + "epoch": 0.3461224489795918, + "grad_norm": 0.13346275687217712, + "learning_rate": 3.737938906567053e-06, + "loss": 0.07526676058769226, + "step": 2120 + }, + { + "epoch": 0.34775510204081633, + "grad_norm": 0.112598717212677, + "learning_rate": 3.7265968448304057e-06, + "loss": 0.07366414070129394, + "step": 2130 + }, + { + "epoch": 0.34938775510204084, + "grad_norm": 0.12423531711101532, + "learning_rate": 3.7152214338106827e-06, + "loss": 0.06439908146858216, + "step": 2140 + }, + { + "epoch": 0.3510204081632653, + "grad_norm": 0.21758921444416046, + "learning_rate": 3.703812982787828e-06, + "loss": 0.08098108768463134, + "step": 2150 + }, + { + "epoch": 0.3526530612244898, + "grad_norm": 0.1298121213912964, + "learning_rate": 3.6923718019400906e-06, + "loss": 0.08134381771087647, + "step": 2160 + }, + { + "epoch": 0.35428571428571426, + "grad_norm": 0.09018377214670181, + "learning_rate": 3.6808982023355934e-06, + "loss": 0.07070590257644653, + "step": 2170 + }, + { + "epoch": 0.35591836734693877, + "grad_norm": 0.1329157054424286, + "learning_rate": 3.6693924959238774e-06, + "loss": 0.06200396418571472, + "step": 2180 + }, + { + "epoch": 0.3575510204081633, + "grad_norm": 0.08813945204019547, + "learning_rate": 3.657854995527416e-06, + "loss": 0.06886230111122131, + "step": 2190 + }, + { + "epoch": 0.35918367346938773, + "grad_norm": 0.13459829986095428, + "learning_rate": 3.646286014833114e-06, + "loss": 0.07170915603637695, + "step": 2200 + }, + { + "epoch": 0.35918367346938773, + "eval_accuracy": 0.970625, + "eval_loss": 0.06810569018125534, + "eval_runtime": 157.4227, + "eval_samples_per_second": 25.409, + "eval_steps_per_second": 25.409, + "step": 2200 + }, + { + "epoch": 0.36081632653061224, + "grad_norm": 0.13405098021030426, + "learning_rate": 3.634685868383776e-06, + "loss": 0.07438129186630249, + "step": 2210 + }, + { + "epoch": 0.36244897959183675, + "grad_norm": 0.1342102289199829, + "learning_rate": 3.6230548715695555e-06, + "loss": 0.07625046968460084, + "step": 2220 + }, + { + "epoch": 0.3640816326530612, + "grad_norm": 0.1532502919435501, + "learning_rate": 3.611393340619381e-06, + "loss": 0.06719872951507569, + "step": 2230 + }, + { + "epoch": 0.3657142857142857, + "grad_norm": 0.11218152195215225, + "learning_rate": 3.5997015925923563e-06, + "loss": 0.07131683230400085, + "step": 2240 + }, + { + "epoch": 0.3673469387755102, + "grad_norm": 0.09200239181518555, + "learning_rate": 3.587979945369141e-06, + "loss": 0.06948224902153015, + "step": 2250 + }, + { + "epoch": 0.3689795918367347, + "grad_norm": 0.08099328726530075, + "learning_rate": 3.5762287176433097e-06, + "loss": 0.07263018488883972, + "step": 2260 + }, + { + "epoch": 0.3706122448979592, + "grad_norm": 0.09097494184970856, + "learning_rate": 3.564448228912682e-06, + "loss": 0.06066258549690247, + "step": 2270 + }, + { + "epoch": 0.3722448979591837, + "grad_norm": 0.07033897191286087, + "learning_rate": 3.5526387994706428e-06, + "loss": 0.07041062712669373, + "step": 2280 + }, + { + "epoch": 0.37387755102040815, + "grad_norm": 0.07659374922513962, + "learning_rate": 3.5408007503974283e-06, + "loss": 0.07272242307662964, + "step": 2290 + }, + { + "epoch": 0.37551020408163266, + "grad_norm": 0.0836353451013565, + "learning_rate": 3.5289344035513996e-06, + "loss": 0.07268807291984558, + "step": 2300 + }, + { + "epoch": 0.37714285714285717, + "grad_norm": 0.1426168978214264, + "learning_rate": 3.5170400815602894e-06, + "loss": 0.05777261257171631, + "step": 2310 + }, + { + "epoch": 0.3787755102040816, + "grad_norm": 0.07479120045900345, + "learning_rate": 3.5051181078124336e-06, + "loss": 0.06174240708351135, + "step": 2320 + }, + { + "epoch": 0.38040816326530613, + "grad_norm": 0.14598114788532257, + "learning_rate": 3.4931688064479745e-06, + "loss": 0.06616324782371522, + "step": 2330 + }, + { + "epoch": 0.3820408163265306, + "grad_norm": 0.22399863600730896, + "learning_rate": 3.4811925023500524e-06, + "loss": 0.06692227125167846, + "step": 2340 + }, + { + "epoch": 0.3836734693877551, + "grad_norm": 0.11600285023450851, + "learning_rate": 3.4691895211359693e-06, + "loss": 0.0634126603603363, + "step": 2350 + }, + { + "epoch": 0.3853061224489796, + "grad_norm": 0.0577060841023922, + "learning_rate": 3.457160189148337e-06, + "loss": 0.06884999275207519, + "step": 2360 + }, + { + "epoch": 0.38693877551020406, + "grad_norm": 0.09160040318965912, + "learning_rate": 3.445104833446205e-06, + "loss": 0.05986430644989014, + "step": 2370 + }, + { + "epoch": 0.38857142857142857, + "grad_norm": 0.1284518539905548, + "learning_rate": 3.433023781796167e-06, + "loss": 0.06759394407272339, + "step": 2380 + }, + { + "epoch": 0.3902040816326531, + "grad_norm": 0.08572858572006226, + "learning_rate": 3.4209173626634494e-06, + "loss": 0.06813116669654846, + "step": 2390 + }, + { + "epoch": 0.39183673469387753, + "grad_norm": 0.11550725251436234, + "learning_rate": 3.4087859052029837e-06, + "loss": 0.06371633410453796, + "step": 2400 + }, + { + "epoch": 0.39183673469387753, + "eval_accuracy": 0.9685833333333334, + "eval_loss": 0.07035119831562042, + "eval_runtime": 156.9729, + "eval_samples_per_second": 25.482, + "eval_steps_per_second": 25.482, + "step": 2400 + }, + { + "epoch": 0.39346938775510204, + "grad_norm": 0.22328943014144897, + "learning_rate": 3.3966297392504535e-06, + "loss": 0.06804299950599671, + "step": 2410 + }, + { + "epoch": 0.39510204081632655, + "grad_norm": 0.18518126010894775, + "learning_rate": 3.3844491953133272e-06, + "loss": 0.07678125500679016, + "step": 2420 + }, + { + "epoch": 0.396734693877551, + "grad_norm": 0.09818598628044128, + "learning_rate": 3.372244604561875e-06, + "loss": 0.06537411212921143, + "step": 2430 + }, + { + "epoch": 0.3983673469387755, + "grad_norm": 0.08255010843276978, + "learning_rate": 3.360016298820163e-06, + "loss": 0.05702770948410034, + "step": 2440 + }, + { + "epoch": 0.4, + "grad_norm": 0.10319415479898453, + "learning_rate": 3.3477646105570293e-06, + "loss": 0.06334798336029053, + "step": 2450 + }, + { + "epoch": 0.4016326530612245, + "grad_norm": 0.06523872911930084, + "learning_rate": 3.3354898728770503e-06, + "loss": 0.051438117027282716, + "step": 2460 + }, + { + "epoch": 0.403265306122449, + "grad_norm": 0.15883173048496246, + "learning_rate": 3.3231924195114774e-06, + "loss": 0.06427289247512817, + "step": 2470 + }, + { + "epoch": 0.4048979591836735, + "grad_norm": 0.1734374463558197, + "learning_rate": 3.3108725848091685e-06, + "loss": 0.06652372479438781, + "step": 2480 + }, + { + "epoch": 0.40653061224489795, + "grad_norm": 0.10557752102613449, + "learning_rate": 3.2985307037274937e-06, + "loss": 0.07374808788299561, + "step": 2490 + }, + { + "epoch": 0.40816326530612246, + "grad_norm": 0.12942200899124146, + "learning_rate": 3.286167111823234e-06, + "loss": 0.06349012851715088, + "step": 2500 + }, + { + "epoch": 0.4097959183673469, + "grad_norm": 0.1858324259519577, + "learning_rate": 3.2737821452434502e-06, + "loss": 0.06953533291816712, + "step": 2510 + }, + { + "epoch": 0.4114285714285714, + "grad_norm": 0.10682573169469833, + "learning_rate": 3.26137614071635e-06, + "loss": 0.05826004147529602, + "step": 2520 + }, + { + "epoch": 0.41306122448979593, + "grad_norm": 0.12533792853355408, + "learning_rate": 3.2489494355421293e-06, + "loss": 0.05837242007255554, + "step": 2530 + }, + { + "epoch": 0.4146938775510204, + "grad_norm": 0.10952792316675186, + "learning_rate": 3.2365023675838027e-06, + "loss": 0.06359312534332276, + "step": 2540 + }, + { + "epoch": 0.4163265306122449, + "grad_norm": 0.08605755865573883, + "learning_rate": 3.22403527525802e-06, + "loss": 0.07361164093017578, + "step": 2550 + }, + { + "epoch": 0.4179591836734694, + "grad_norm": 0.15942610800266266, + "learning_rate": 3.211548497525861e-06, + "loss": 0.05989238619804382, + "step": 2560 + }, + { + "epoch": 0.41959183673469386, + "grad_norm": 0.16340239346027374, + "learning_rate": 3.1990423738836206e-06, + "loss": 0.07593320608139038, + "step": 2570 + }, + { + "epoch": 0.42122448979591837, + "grad_norm": 0.11011330783367157, + "learning_rate": 3.1865172443535797e-06, + "loss": 0.06850495338439941, + "step": 2580 + }, + { + "epoch": 0.4228571428571429, + "grad_norm": 0.10529608279466629, + "learning_rate": 3.173973449474762e-06, + "loss": 0.056969082355499266, + "step": 2590 + }, + { + "epoch": 0.42448979591836733, + "grad_norm": 0.12615667283535004, + "learning_rate": 3.1614113302936716e-06, + "loss": 0.06189748048782349, + "step": 2600 + }, + { + "epoch": 0.42448979591836733, + "eval_accuracy": 0.9746666666666668, + "eval_loss": 0.0618240088224411, + "eval_runtime": 157.0858, + "eval_samples_per_second": 25.464, + "eval_steps_per_second": 25.464, + "step": 2600 + }, + { + "epoch": 0.42612244897959184, + "grad_norm": 0.1343226134777069, + "learning_rate": 3.1488312283550216e-06, + "loss": 0.061647194623947146, + "step": 2610 + }, + { + "epoch": 0.42775510204081635, + "grad_norm": 0.11777225136756897, + "learning_rate": 3.136233485692451e-06, + "loss": 0.05916886925697327, + "step": 2620 + }, + { + "epoch": 0.4293877551020408, + "grad_norm": 0.14721977710723877, + "learning_rate": 3.12361844481922e-06, + "loss": 0.059158307313919065, + "step": 2630 + }, + { + "epoch": 0.4310204081632653, + "grad_norm": 0.12912684679031372, + "learning_rate": 3.110986448718902e-06, + "loss": 0.05914759635925293, + "step": 2640 + }, + { + "epoch": 0.4326530612244898, + "grad_norm": 0.13003695011138916, + "learning_rate": 3.0983378408360597e-06, + "loss": 0.059078723192214966, + "step": 2650 + }, + { + "epoch": 0.4342857142857143, + "grad_norm": 0.12714017927646637, + "learning_rate": 3.085672965066899e-06, + "loss": 0.05707222819328308, + "step": 2660 + }, + { + "epoch": 0.4359183673469388, + "grad_norm": 0.13708427548408508, + "learning_rate": 3.0729921657499294e-06, + "loss": 0.05138048529624939, + "step": 2670 + }, + { + "epoch": 0.43755102040816324, + "grad_norm": 0.0949573740363121, + "learning_rate": 3.0602957876565943e-06, + "loss": 0.050283265113830564, + "step": 2680 + }, + { + "epoch": 0.43918367346938775, + "grad_norm": 0.08723263442516327, + "learning_rate": 3.0475841759819023e-06, + "loss": 0.06342648863792419, + "step": 2690 + }, + { + "epoch": 0.44081632653061226, + "grad_norm": 0.09944969415664673, + "learning_rate": 3.034857676335038e-06, + "loss": 0.06127093434333801, + "step": 2700 + }, + { + "epoch": 0.4424489795918367, + "grad_norm": 0.10576151311397552, + "learning_rate": 3.0221166347299673e-06, + "loss": 0.05412284731864929, + "step": 2710 + }, + { + "epoch": 0.4440816326530612, + "grad_norm": 0.1598891168832779, + "learning_rate": 3.009361397576031e-06, + "loss": 0.05579071640968323, + "step": 2720 + }, + { + "epoch": 0.44571428571428573, + "grad_norm": 0.16953758895397186, + "learning_rate": 2.9965923116685224e-06, + "loss": 0.06472888588905334, + "step": 2730 + }, + { + "epoch": 0.4473469387755102, + "grad_norm": 0.10883133113384247, + "learning_rate": 2.9838097241792646e-06, + "loss": 0.06263615489006043, + "step": 2740 + }, + { + "epoch": 0.4489795918367347, + "grad_norm": 0.08914396911859512, + "learning_rate": 2.9710139826471646e-06, + "loss": 0.0624904453754425, + "step": 2750 + }, + { + "epoch": 0.4506122448979592, + "grad_norm": 0.1059069111943245, + "learning_rate": 2.9582054349687706e-06, + "loss": 0.05488716959953308, + "step": 2760 + }, + { + "epoch": 0.45224489795918366, + "grad_norm": 0.09883642196655273, + "learning_rate": 2.9453844293888067e-06, + "loss": 0.06263139843940735, + "step": 2770 + }, + { + "epoch": 0.45387755102040817, + "grad_norm": 0.11156519502401352, + "learning_rate": 2.932551314490712e-06, + "loss": 0.06324111819267272, + "step": 2780 + }, + { + "epoch": 0.4555102040816327, + "grad_norm": 0.08826332539319992, + "learning_rate": 2.9197064391871565e-06, + "loss": 0.06384775042533875, + "step": 2790 + }, + { + "epoch": 0.45714285714285713, + "grad_norm": 0.0698089748620987, + "learning_rate": 2.9068501527105595e-06, + "loss": 0.05722553730010986, + "step": 2800 + }, + { + "epoch": 0.45714285714285713, + "eval_accuracy": 0.977375, + "eval_loss": 0.05622806400060654, + "eval_runtime": 170.7758, + "eval_samples_per_second": 23.423, + "eval_steps_per_second": 23.423, + "step": 2800 + }, + { + "epoch": 0.45877551020408164, + "grad_norm": 0.07031046599149704, + "learning_rate": 2.893982804603592e-06, + "loss": 0.060957270860672, + "step": 2810 + }, + { + "epoch": 0.46040816326530615, + "grad_norm": 0.08772005140781403, + "learning_rate": 2.8811047447096714e-06, + "loss": 0.05667365789413452, + "step": 2820 + }, + { + "epoch": 0.4620408163265306, + "grad_norm": 0.10329847782850266, + "learning_rate": 2.8682163231634557e-06, + "loss": 0.06144403219223023, + "step": 2830 + }, + { + "epoch": 0.4636734693877551, + "grad_norm": 0.06306952983140945, + "learning_rate": 2.8553178903813193e-06, + "loss": 0.04665087461471558, + "step": 2840 + }, + { + "epoch": 0.46530612244897956, + "grad_norm": 0.11146903783082962, + "learning_rate": 2.842409797051825e-06, + "loss": 0.06363347172737122, + "step": 2850 + }, + { + "epoch": 0.4669387755102041, + "grad_norm": 0.04504350945353508, + "learning_rate": 2.829492394126192e-06, + "loss": 0.05332474112510681, + "step": 2860 + }, + { + "epoch": 0.4685714285714286, + "grad_norm": 0.06039620563387871, + "learning_rate": 2.8165660328087546e-06, + "loss": 0.07094461917877197, + "step": 2870 + }, + { + "epoch": 0.47020408163265304, + "grad_norm": 0.08084474503993988, + "learning_rate": 2.8036310645474096e-06, + "loss": 0.058167701959609984, + "step": 2880 + }, + { + "epoch": 0.47183673469387755, + "grad_norm": 0.05352522060275078, + "learning_rate": 2.7906878410240655e-06, + "loss": 0.0486674964427948, + "step": 2890 + }, + { + "epoch": 0.47346938775510206, + "grad_norm": 0.09431054443120956, + "learning_rate": 2.777736714145077e-06, + "loss": 0.062025409936904904, + "step": 2900 + }, + { + "epoch": 0.4751020408163265, + "grad_norm": 0.09863542020320892, + "learning_rate": 2.7647780360316787e-06, + "loss": 0.061118441820144656, + "step": 2910 + }, + { + "epoch": 0.476734693877551, + "grad_norm": 0.082881860435009, + "learning_rate": 2.7518121590104145e-06, + "loss": 0.061808472871780394, + "step": 2920 + }, + { + "epoch": 0.47836734693877553, + "grad_norm": 0.06983684748411179, + "learning_rate": 2.7388394356035524e-06, + "loss": 0.05849882960319519, + "step": 2930 + }, + { + "epoch": 0.48, + "grad_norm": 0.10832090675830841, + "learning_rate": 2.7258602185195043e-06, + "loss": 0.06612529158592224, + "step": 2940 + }, + { + "epoch": 0.4816326530612245, + "grad_norm": 0.1516750454902649, + "learning_rate": 2.7128748606432353e-06, + "loss": 0.058327901363372806, + "step": 2950 + }, + { + "epoch": 0.483265306122449, + "grad_norm": 0.14055661857128143, + "learning_rate": 2.699883715026669e-06, + "loss": 0.055737996101379396, + "step": 2960 + }, + { + "epoch": 0.48489795918367345, + "grad_norm": 0.0793638750910759, + "learning_rate": 2.686887134879089e-06, + "loss": 0.051851028203964235, + "step": 2970 + }, + { + "epoch": 0.48653061224489796, + "grad_norm": 0.07475428283214569, + "learning_rate": 2.6738854735575358e-06, + "loss": 0.06357839703559875, + "step": 2980 + }, + { + "epoch": 0.4881632653061225, + "grad_norm": 0.13335047662258148, + "learning_rate": 2.660879084557198e-06, + "loss": 0.06160399913787842, + "step": 2990 + }, + { + "epoch": 0.4897959183673469, + "grad_norm": 0.12583893537521362, + "learning_rate": 2.6478683215018025e-06, + "loss": 0.060502588748931885, + "step": 3000 + }, + { + "epoch": 0.4897959183673469, + "eval_accuracy": 0.9767083333333334, + "eval_loss": 0.0561000294983387, + "eval_runtime": 156.4202, + "eval_samples_per_second": 25.572, + "eval_steps_per_second": 25.572, + "step": 3000 + }, + { + "epoch": 0.49142857142857144, + "grad_norm": 0.1411438137292862, + "learning_rate": 2.6348535381340025e-06, + "loss": 0.060459482669830325, + "step": 3010 + }, + { + "epoch": 0.4930612244897959, + "grad_norm": 0.12584605813026428, + "learning_rate": 2.6218350883057546e-06, + "loss": 0.0632048487663269, + "step": 3020 + }, + { + "epoch": 0.4946938775510204, + "grad_norm": 0.10522248595952988, + "learning_rate": 2.608813325968702e-06, + "loss": 0.04655081033706665, + "step": 3030 + }, + { + "epoch": 0.4963265306122449, + "grad_norm": 0.11343304812908173, + "learning_rate": 2.59578860516455e-06, + "loss": 0.05672039389610291, + "step": 3040 + }, + { + "epoch": 0.49795918367346936, + "grad_norm": 0.12666337192058563, + "learning_rate": 2.5827612800154382e-06, + "loss": 0.047120341658592226, + "step": 3050 + }, + { + "epoch": 0.4995918367346939, + "grad_norm": 0.08930598944425583, + "learning_rate": 2.5697317047143177e-06, + "loss": 0.06063341498374939, + "step": 3060 + }, + { + "epoch": 0.5012244897959184, + "grad_norm": 0.07818852365016937, + "learning_rate": 2.556700233515316e-06, + "loss": 0.05240688323974609, + "step": 3070 + }, + { + "epoch": 0.5028571428571429, + "grad_norm": 0.1425711065530777, + "learning_rate": 2.5436672207241064e-06, + "loss": 0.053180736303329465, + "step": 3080 + }, + { + "epoch": 0.5044897959183674, + "grad_norm": 0.1457373946905136, + "learning_rate": 2.5306330206882764e-06, + "loss": 0.06092001795768738, + "step": 3090 + }, + { + "epoch": 0.5061224489795918, + "grad_norm": 0.12738314270973206, + "learning_rate": 2.5175979877876935e-06, + "loss": 0.06613186597824097, + "step": 3100 + }, + { + "epoch": 0.5077551020408163, + "grad_norm": 0.09169036149978638, + "learning_rate": 2.5045624764248687e-06, + "loss": 0.06115126609802246, + "step": 3110 + }, + { + "epoch": 0.5093877551020408, + "grad_norm": 0.09637737274169922, + "learning_rate": 2.491526841015321e-06, + "loss": 0.044120225310325625, + "step": 3120 + }, + { + "epoch": 0.5110204081632653, + "grad_norm": 0.15029804408550262, + "learning_rate": 2.478491435977945e-06, + "loss": 0.07098687291145325, + "step": 3130 + }, + { + "epoch": 0.5126530612244898, + "grad_norm": 0.1228741705417633, + "learning_rate": 2.4654566157253683e-06, + "loss": 0.05738852024078369, + "step": 3140 + }, + { + "epoch": 0.5142857142857142, + "grad_norm": 0.09150952100753784, + "learning_rate": 2.452422734654321e-06, + "loss": 0.053479206562042234, + "step": 3150 + }, + { + "epoch": 0.5159183673469387, + "grad_norm": 0.14547918736934662, + "learning_rate": 2.4393901471359977e-06, + "loss": 0.056794995069503786, + "step": 3160 + }, + { + "epoch": 0.5175510204081633, + "grad_norm": 0.10099287331104279, + "learning_rate": 2.4263592075064238e-06, + "loss": 0.05802353620529175, + "step": 3170 + }, + { + "epoch": 0.5191836734693878, + "grad_norm": 0.10859636962413788, + "learning_rate": 2.41333027005682e-06, + "loss": 0.05282890200614929, + "step": 3180 + }, + { + "epoch": 0.5208163265306123, + "grad_norm": 0.12420635670423508, + "learning_rate": 2.400303689023972e-06, + "loss": 0.05190218687057495, + "step": 3190 + }, + { + "epoch": 0.5224489795918368, + "grad_norm": 0.1902378350496292, + "learning_rate": 2.3872798185805986e-06, + "loss": 0.06482383608818054, + "step": 3200 + }, + { + "epoch": 0.5224489795918368, + "eval_accuracy": 0.9744583333333335, + "eval_loss": 0.058392543345689774, + "eval_runtime": 156.8655, + "eval_samples_per_second": 25.5, + "eval_steps_per_second": 25.5, + "step": 3200 + }, + { + "epoch": 0.5240816326530612, + "grad_norm": 0.15229448676109314, + "learning_rate": 2.37425901282572e-06, + "loss": 0.06062687039375305, + "step": 3210 + }, + { + "epoch": 0.5257142857142857, + "grad_norm": 0.08411400765180588, + "learning_rate": 2.3612416257750335e-06, + "loss": 0.06494970321655273, + "step": 3220 + }, + { + "epoch": 0.5273469387755102, + "grad_norm": 0.114021435379982, + "learning_rate": 2.348228011351287e-06, + "loss": 0.05140092372894287, + "step": 3230 + }, + { + "epoch": 0.5289795918367347, + "grad_norm": 0.08826510608196259, + "learning_rate": 2.335218523374656e-06, + "loss": 0.041680914163589475, + "step": 3240 + }, + { + "epoch": 0.5306122448979592, + "grad_norm": 0.10536732524633408, + "learning_rate": 2.3222135155531248e-06, + "loss": 0.04369679093360901, + "step": 3250 + }, + { + "epoch": 0.5322448979591837, + "grad_norm": 0.16883279383182526, + "learning_rate": 2.309213341472868e-06, + "loss": 0.050543594360351565, + "step": 3260 + }, + { + "epoch": 0.5338775510204081, + "grad_norm": 0.15277455747127533, + "learning_rate": 2.29621835458864e-06, + "loss": 0.05258342027664185, + "step": 3270 + }, + { + "epoch": 0.5355102040816326, + "grad_norm": 0.07162481546401978, + "learning_rate": 2.283228908214163e-06, + "loss": 0.055678969621658324, + "step": 3280 + }, + { + "epoch": 0.5371428571428571, + "grad_norm": 0.11578263342380524, + "learning_rate": 2.2702453555125182e-06, + "loss": 0.048620349168777464, + "step": 3290 + }, + { + "epoch": 0.5387755102040817, + "grad_norm": 0.21185705065727234, + "learning_rate": 2.2572680494865527e-06, + "loss": 0.046547812223434445, + "step": 3300 + }, + { + "epoch": 0.5404081632653062, + "grad_norm": 0.19949115812778473, + "learning_rate": 2.2442973429692705e-06, + "loss": 0.04967870116233826, + "step": 3310 + }, + { + "epoch": 0.5420408163265306, + "grad_norm": 0.10855370759963989, + "learning_rate": 2.231333588614248e-06, + "loss": 0.0508593499660492, + "step": 3320 + }, + { + "epoch": 0.5436734693877551, + "grad_norm": 0.11827342212200165, + "learning_rate": 2.2183771388860427e-06, + "loss": 0.04854269027709961, + "step": 3330 + }, + { + "epoch": 0.5453061224489796, + "grad_norm": 0.1081443578004837, + "learning_rate": 2.2054283460506096e-06, + "loss": 0.05934192538261414, + "step": 3340 + }, + { + "epoch": 0.5469387755102041, + "grad_norm": 0.14441704750061035, + "learning_rate": 2.1924875621657256e-06, + "loss": 0.05332346558570862, + "step": 3350 + }, + { + "epoch": 0.5485714285714286, + "grad_norm": 0.11531820893287659, + "learning_rate": 2.179555139071415e-06, + "loss": 0.04938814043998718, + "step": 3360 + }, + { + "epoch": 0.5502040816326531, + "grad_norm": 0.08365523815155029, + "learning_rate": 2.166631428380387e-06, + "loss": 0.05734975934028626, + "step": 3370 + }, + { + "epoch": 0.5518367346938775, + "grad_norm": 0.09165070950984955, + "learning_rate": 2.1537167814684728e-06, + "loss": 0.047475004196166994, + "step": 3380 + }, + { + "epoch": 0.553469387755102, + "grad_norm": 0.13237863779067993, + "learning_rate": 2.140811549465074e-06, + "loss": 0.062284767627716064, + "step": 3390 + }, + { + "epoch": 0.5551020408163265, + "grad_norm": 0.1698858141899109, + "learning_rate": 2.127916083243615e-06, + "loss": 0.06585791110992431, + "step": 3400 + }, + { + "epoch": 0.5551020408163265, + "eval_accuracy": 0.9770833333333334, + "eval_loss": 0.05396538972854614, + "eval_runtime": 156.044, + "eval_samples_per_second": 25.634, + "eval_steps_per_second": 25.634, + "step": 3400 + }, + { + "epoch": 0.556734693877551, + "grad_norm": 0.1104230135679245, + "learning_rate": 2.1150307334120033e-06, + "loss": 0.04635375440120697, + "step": 3410 + }, + { + "epoch": 0.5583673469387755, + "grad_norm": 0.21372850239276886, + "learning_rate": 2.1021558503030973e-06, + "loss": 0.05940789580345154, + "step": 3420 + }, + { + "epoch": 0.56, + "grad_norm": 0.2013176679611206, + "learning_rate": 2.0892917839651818e-06, + "loss": 0.047165653109550475, + "step": 3430 + }, + { + "epoch": 0.5616326530612245, + "grad_norm": 0.09189072996377945, + "learning_rate": 2.076438884152449e-06, + "loss": 0.04550817906856537, + "step": 3440 + }, + { + "epoch": 0.563265306122449, + "grad_norm": 0.10521821677684784, + "learning_rate": 2.0635975003154934e-06, + "loss": 0.045033201575279236, + "step": 3450 + }, + { + "epoch": 0.5648979591836735, + "grad_norm": 0.10311131924390793, + "learning_rate": 2.0507679815918043e-06, + "loss": 0.04502847194671631, + "step": 3460 + }, + { + "epoch": 0.566530612244898, + "grad_norm": 0.07487176358699799, + "learning_rate": 2.0379506767962788e-06, + "loss": 0.04697479605674744, + "step": 3470 + }, + { + "epoch": 0.5681632653061225, + "grad_norm": 0.12441752105951309, + "learning_rate": 2.0251459344117345e-06, + "loss": 0.052119261026382445, + "step": 3480 + }, + { + "epoch": 0.5697959183673469, + "grad_norm": 0.12823422253131866, + "learning_rate": 2.0123541025794397e-06, + "loss": 0.05213663578033447, + "step": 3490 + }, + { + "epoch": 0.5714285714285714, + "grad_norm": 0.11340193450450897, + "learning_rate": 1.9995755290896414e-06, + "loss": 0.05319364666938782, + "step": 3500 + }, + { + "epoch": 0.5730612244897959, + "grad_norm": 0.10469839721918106, + "learning_rate": 1.986810561372113e-06, + "loss": 0.046560952067375184, + "step": 3510 + }, + { + "epoch": 0.5746938775510204, + "grad_norm": 0.07797456532716751, + "learning_rate": 1.9740595464867092e-06, + "loss": 0.05037369728088379, + "step": 3520 + }, + { + "epoch": 0.5763265306122449, + "grad_norm": 0.13010293245315552, + "learning_rate": 1.961322831113927e-06, + "loss": 0.0543134331703186, + "step": 3530 + }, + { + "epoch": 0.5779591836734694, + "grad_norm": 0.13981281220912933, + "learning_rate": 1.9486007615454837e-06, + "loss": 0.052860957384109494, + "step": 3540 + }, + { + "epoch": 0.5795918367346938, + "grad_norm": 0.14817261695861816, + "learning_rate": 1.9358936836748972e-06, + "loss": 0.05510561466217041, + "step": 3550 + }, + { + "epoch": 0.5812244897959183, + "grad_norm": 0.15079492330551147, + "learning_rate": 1.9232019429880874e-06, + "loss": 0.06125714182853699, + "step": 3560 + }, + { + "epoch": 0.5828571428571429, + "grad_norm": 0.10553149878978729, + "learning_rate": 1.910525884553976e-06, + "loss": 0.051503753662109374, + "step": 3570 + }, + { + "epoch": 0.5844897959183674, + "grad_norm": 0.09094442427158356, + "learning_rate": 1.8978658530151113e-06, + "loss": 0.05377998948097229, + "step": 3580 + }, + { + "epoch": 0.5861224489795919, + "grad_norm": 0.09810740500688553, + "learning_rate": 1.8852221925782935e-06, + "loss": 0.05506981611251831, + "step": 3590 + }, + { + "epoch": 0.5877551020408164, + "grad_norm": 0.13397352397441864, + "learning_rate": 1.872595247005218e-06, + "loss": 0.056353437900543216, + "step": 3600 + }, + { + "epoch": 0.5877551020408164, + "eval_accuracy": 0.979875, + "eval_loss": 0.04787343740463257, + "eval_runtime": 157.8457, + "eval_samples_per_second": 25.341, + "eval_steps_per_second": 25.341, + "step": 3600 + }, + { + "epoch": 0.5893877551020408, + "grad_norm": 0.07899880409240723, + "learning_rate": 1.8599853596031289e-06, + "loss": 0.05329249501228332, + "step": 3610 + }, + { + "epoch": 0.5910204081632653, + "grad_norm": 0.10414254665374756, + "learning_rate": 1.8473928732154847e-06, + "loss": 0.04905630648136139, + "step": 3620 + }, + { + "epoch": 0.5926530612244898, + "grad_norm": 0.1037079319357872, + "learning_rate": 1.8348181302126367e-06, + "loss": 0.04863021373748779, + "step": 3630 + }, + { + "epoch": 0.5942857142857143, + "grad_norm": 0.06960397213697433, + "learning_rate": 1.8222614724825222e-06, + "loss": 0.04666702449321747, + "step": 3640 + }, + { + "epoch": 0.5959183673469388, + "grad_norm": 0.11159648001194, + "learning_rate": 1.8097232414213662e-06, + "loss": 0.05730679631233215, + "step": 3650 + }, + { + "epoch": 0.5975510204081632, + "grad_norm": 0.10231669992208481, + "learning_rate": 1.7972037779244e-06, + "loss": 0.04433938264846802, + "step": 3660 + }, + { + "epoch": 0.5991836734693877, + "grad_norm": 0.1256829798221588, + "learning_rate": 1.784703422376598e-06, + "loss": 0.04502029418945312, + "step": 3670 + }, + { + "epoch": 0.6008163265306122, + "grad_norm": 0.1006830632686615, + "learning_rate": 1.7722225146434146e-06, + "loss": 0.04091633260250092, + "step": 3680 + }, + { + "epoch": 0.6024489795918367, + "grad_norm": 0.12239495664834976, + "learning_rate": 1.7597613940615492e-06, + "loss": 0.0417140930891037, + "step": 3690 + }, + { + "epoch": 0.6040816326530613, + "grad_norm": 0.09386949241161346, + "learning_rate": 1.7473203994297197e-06, + "loss": 0.045112240314483645, + "step": 3700 + }, + { + "epoch": 0.6057142857142858, + "grad_norm": 0.1845707893371582, + "learning_rate": 1.7348998689994492e-06, + "loss": 0.04915813505649567, + "step": 3710 + }, + { + "epoch": 0.6073469387755102, + "grad_norm": 0.1345887929201126, + "learning_rate": 1.7225001404658714e-06, + "loss": 0.05195118188858032, + "step": 3720 + }, + { + "epoch": 0.6089795918367347, + "grad_norm": 0.16226522624492645, + "learning_rate": 1.7101215509585482e-06, + "loss": 0.047248026728630065, + "step": 3730 + }, + { + "epoch": 0.6106122448979592, + "grad_norm": 0.12012234330177307, + "learning_rate": 1.6977644370323032e-06, + "loss": 0.05229860544204712, + "step": 3740 + }, + { + "epoch": 0.6122448979591837, + "grad_norm": 0.16154533624649048, + "learning_rate": 1.6854291346580737e-06, + "loss": 0.048448923230171206, + "step": 3750 + }, + { + "epoch": 0.6138775510204082, + "grad_norm": 0.08616314828395844, + "learning_rate": 1.673115979213773e-06, + "loss": 0.04779277443885803, + "step": 3760 + }, + { + "epoch": 0.6155102040816327, + "grad_norm": 0.20649370551109314, + "learning_rate": 1.6608253054751739e-06, + "loss": 0.04001966416835785, + "step": 3770 + }, + { + "epoch": 0.6171428571428571, + "grad_norm": 0.22419355809688568, + "learning_rate": 1.6485574476068056e-06, + "loss": 0.04915740489959717, + "step": 3780 + }, + { + "epoch": 0.6187755102040816, + "grad_norm": 0.11812295019626617, + "learning_rate": 1.6363127391528694e-06, + "loss": 0.055313241481781, + "step": 3790 + }, + { + "epoch": 0.6204081632653061, + "grad_norm": 0.12400855869054794, + "learning_rate": 1.6240915130281696e-06, + "loss": 0.05452631115913391, + "step": 3800 + }, + { + "epoch": 0.6204081632653061, + "eval_accuracy": 0.981875, + "eval_loss": 0.04318515956401825, + "eval_runtime": 158.5838, + "eval_samples_per_second": 25.223, + "eval_steps_per_second": 25.223, + "step": 3800 + }, + { + "epoch": 0.6220408163265306, + "grad_norm": 0.094053253531456, + "learning_rate": 1.6118941015090614e-06, + "loss": 0.047327810525894166, + "step": 3810 + }, + { + "epoch": 0.6236734693877551, + "grad_norm": 0.11286808550357819, + "learning_rate": 1.599720836224419e-06, + "loss": 0.04089818894863129, + "step": 3820 + }, + { + "epoch": 0.6253061224489795, + "grad_norm": 0.0909293070435524, + "learning_rate": 1.5875720481466162e-06, + "loss": 0.04378651976585388, + "step": 3830 + }, + { + "epoch": 0.626938775510204, + "grad_norm": 0.1207573264837265, + "learning_rate": 1.5754480675825295e-06, + "loss": 0.055971544981002805, + "step": 3840 + }, + { + "epoch": 0.6285714285714286, + "grad_norm": 0.09171696752309799, + "learning_rate": 1.563349224164556e-06, + "loss": 0.04364039599895477, + "step": 3850 + }, + { + "epoch": 0.6302040816326531, + "grad_norm": 0.1100732609629631, + "learning_rate": 1.5512758468416561e-06, + "loss": 0.04227602481842041, + "step": 3860 + }, + { + "epoch": 0.6318367346938776, + "grad_norm": 0.08587154746055603, + "learning_rate": 1.5392282638704032e-06, + "loss": 0.03987776041030884, + "step": 3870 + }, + { + "epoch": 0.6334693877551021, + "grad_norm": 0.14273083209991455, + "learning_rate": 1.5272068028060616e-06, + "loss": 0.05040489435195923, + "step": 3880 + }, + { + "epoch": 0.6351020408163265, + "grad_norm": 0.12788064777851105, + "learning_rate": 1.5152117904936827e-06, + "loss": 0.04129888415336609, + "step": 3890 + }, + { + "epoch": 0.636734693877551, + "grad_norm": 0.14224086701869965, + "learning_rate": 1.5032435530592165e-06, + "loss": 0.04145504236221313, + "step": 3900 + }, + { + "epoch": 0.6383673469387755, + "grad_norm": 0.07414636015892029, + "learning_rate": 1.4913024159006444e-06, + "loss": 0.03938417434692383, + "step": 3910 + }, + { + "epoch": 0.64, + "grad_norm": 0.1443425565958023, + "learning_rate": 1.4793887036791338e-06, + "loss": 0.047359606623649596, + "step": 3920 + }, + { + "epoch": 0.6416326530612245, + "grad_norm": 0.1462288498878479, + "learning_rate": 1.4675027403102107e-06, + "loss": 0.047661393880844116, + "step": 3930 + }, + { + "epoch": 0.643265306122449, + "grad_norm": 0.15937893092632294, + "learning_rate": 1.4556448489549519e-06, + "loss": 0.04829471111297608, + "step": 3940 + }, + { + "epoch": 0.6448979591836734, + "grad_norm": 0.10469378530979156, + "learning_rate": 1.4438153520111978e-06, + "loss": 0.047904521226882935, + "step": 3950 + }, + { + "epoch": 0.6465306122448979, + "grad_norm": 0.10817313939332962, + "learning_rate": 1.4320145711047894e-06, + "loss": 0.04648992121219635, + "step": 3960 + }, + { + "epoch": 0.6481632653061224, + "grad_norm": 0.12307990342378616, + "learning_rate": 1.4202428270808256e-06, + "loss": 0.044715133309364316, + "step": 3970 + }, + { + "epoch": 0.649795918367347, + "grad_norm": 0.14425775408744812, + "learning_rate": 1.408500439994933e-06, + "loss": 0.042470496892929074, + "step": 3980 + }, + { + "epoch": 0.6514285714285715, + "grad_norm": 0.11694067716598511, + "learning_rate": 1.3967877291045706e-06, + "loss": 0.03910423815250397, + "step": 3990 + }, + { + "epoch": 0.6530612244897959, + "grad_norm": 0.13050039112567902, + "learning_rate": 1.3851050128603454e-06, + "loss": 0.046834564208984374, + "step": 4000 + }, + { + "epoch": 0.6530612244897959, + "eval_accuracy": 0.9817083333333334, + "eval_loss": 0.04485878720879555, + "eval_runtime": 155.2113, + "eval_samples_per_second": 25.771, + "eval_steps_per_second": 25.771, + "step": 4000 + }, + { + "epoch": 0.6546938775510204, + "grad_norm": 0.16441842913627625, + "learning_rate": 1.3734526088973576e-06, + "loss": 0.048332995176315306, + "step": 4010 + }, + { + "epoch": 0.6563265306122449, + "grad_norm": 0.16489046812057495, + "learning_rate": 1.3618308340265628e-06, + "loss": 0.05632186532020569, + "step": 4020 + }, + { + "epoch": 0.6579591836734694, + "grad_norm": 0.0776519775390625, + "learning_rate": 1.3502400042261583e-06, + "loss": 0.0413711816072464, + "step": 4030 + }, + { + "epoch": 0.6595918367346939, + "grad_norm": 0.07065974175930023, + "learning_rate": 1.3386804346329913e-06, + "loss": 0.04443484246730804, + "step": 4040 + }, + { + "epoch": 0.6612244897959184, + "grad_norm": 0.058040838688611984, + "learning_rate": 1.3271524395339958e-06, + "loss": 0.03936767876148224, + "step": 4050 + }, + { + "epoch": 0.6628571428571428, + "grad_norm": 0.12576180696487427, + "learning_rate": 1.31565633235764e-06, + "loss": 0.03893322050571442, + "step": 4060 + }, + { + "epoch": 0.6644897959183673, + "grad_norm": 0.1022111177444458, + "learning_rate": 1.30419242566541e-06, + "loss": 0.05221283435821533, + "step": 4070 + }, + { + "epoch": 0.6661224489795918, + "grad_norm": 0.07897438853979111, + "learning_rate": 1.2927610311433098e-06, + "loss": 0.04261190891265869, + "step": 4080 + }, + { + "epoch": 0.6677551020408163, + "grad_norm": 0.16535215079784393, + "learning_rate": 1.2813624595933885e-06, + "loss": 0.04804243743419647, + "step": 4090 + }, + { + "epoch": 0.6693877551020408, + "grad_norm": 0.10754700005054474, + "learning_rate": 1.2699970209252877e-06, + "loss": 0.03677414059638977, + "step": 4100 + }, + { + "epoch": 0.6710204081632654, + "grad_norm": 0.13121150434017181, + "learning_rate": 1.2586650241478173e-06, + "loss": 0.04266786575317383, + "step": 4110 + }, + { + "epoch": 0.6726530612244898, + "grad_norm": 0.11523906886577606, + "learning_rate": 1.2473667773605542e-06, + "loss": 0.045382776856422426, + "step": 4120 + }, + { + "epoch": 0.6742857142857143, + "grad_norm": 0.16150203347206116, + "learning_rate": 1.236102587745464e-06, + "loss": 0.0483401358127594, + "step": 4130 + }, + { + "epoch": 0.6759183673469388, + "grad_norm": 0.2037627100944519, + "learning_rate": 1.2248727615585501e-06, + "loss": 0.0531754195690155, + "step": 4140 + }, + { + "epoch": 0.6775510204081633, + "grad_norm": 0.08756938576698303, + "learning_rate": 1.2136776041215293e-06, + "loss": 0.046069496870040895, + "step": 4150 + }, + { + "epoch": 0.6791836734693878, + "grad_norm": 0.10196162015199661, + "learning_rate": 1.2025174198135255e-06, + "loss": 0.03938464820384979, + "step": 4160 + }, + { + "epoch": 0.6808163265306122, + "grad_norm": 0.10243439674377441, + "learning_rate": 1.1913925120627987e-06, + "loss": 0.04410702586174011, + "step": 4170 + }, + { + "epoch": 0.6824489795918367, + "grad_norm": 0.16230984032154083, + "learning_rate": 1.1803031833384923e-06, + "loss": 0.03496443629264832, + "step": 4180 + }, + { + "epoch": 0.6840816326530612, + "grad_norm": 0.1036972925066948, + "learning_rate": 1.1692497351424117e-06, + "loss": 0.04019207060337067, + "step": 4190 + }, + { + "epoch": 0.6857142857142857, + "grad_norm": 0.19966299831867218, + "learning_rate": 1.1582324680008252e-06, + "loss": 0.03875584900379181, + "step": 4200 + }, + { + "epoch": 0.6857142857142857, + "eval_accuracy": 0.9839583333333334, + "eval_loss": 0.0395464189350605, + "eval_runtime": 153.8599, + "eval_samples_per_second": 25.998, + "eval_steps_per_second": 25.998, + "step": 4200 + }, + { + "epoch": 0.6873469387755102, + "grad_norm": 0.15349307656288147, + "learning_rate": 1.1472516814562945e-06, + "loss": 0.04495754837989807, + "step": 4210 + }, + { + "epoch": 0.6889795918367347, + "grad_norm": 0.16430068016052246, + "learning_rate": 1.1363076740595294e-06, + "loss": 0.0422965407371521, + "step": 4220 + }, + { + "epoch": 0.6906122448979591, + "grad_norm": 0.1903417408466339, + "learning_rate": 1.1254007433612705e-06, + "loss": 0.04501302838325501, + "step": 4230 + }, + { + "epoch": 0.6922448979591836, + "grad_norm": 0.05951374024152756, + "learning_rate": 1.1145311859042026e-06, + "loss": 0.03593409359455109, + "step": 4240 + }, + { + "epoch": 0.6938775510204082, + "grad_norm": 0.05894222855567932, + "learning_rate": 1.1036992972148865e-06, + "loss": 0.034275355935096743, + "step": 4250 + }, + { + "epoch": 0.6955102040816327, + "grad_norm": 0.09788545221090317, + "learning_rate": 1.092905371795729e-06, + "loss": 0.039912134408950806, + "step": 4260 + }, + { + "epoch": 0.6971428571428572, + "grad_norm": 0.2608252763748169, + "learning_rate": 1.0821497031169729e-06, + "loss": 0.03563873171806335, + "step": 4270 + }, + { + "epoch": 0.6987755102040817, + "grad_norm": 0.10350881516933441, + "learning_rate": 1.0714325836087191e-06, + "loss": 0.036586591601371767, + "step": 4280 + }, + { + "epoch": 0.7004081632653061, + "grad_norm": 0.10214894264936447, + "learning_rate": 1.0607543046529763e-06, + "loss": 0.03775157630443573, + "step": 4290 + }, + { + "epoch": 0.7020408163265306, + "grad_norm": 0.1326843649148941, + "learning_rate": 1.050115156575738e-06, + "loss": 0.03171868622303009, + "step": 4300 + }, + { + "epoch": 0.7036734693877551, + "grad_norm": 0.12274391204118729, + "learning_rate": 1.0395154286390896e-06, + "loss": 0.028169712424278258, + "step": 4310 + }, + { + "epoch": 0.7053061224489796, + "grad_norm": 0.12580160796642303, + "learning_rate": 1.028955409033342e-06, + "loss": 0.03762204349040985, + "step": 4320 + }, + { + "epoch": 0.7069387755102041, + "grad_norm": 0.26383766531944275, + "learning_rate": 1.0184353848692005e-06, + "loss": 0.04035975933074951, + "step": 4330 + }, + { + "epoch": 0.7085714285714285, + "grad_norm": 0.101371631026268, + "learning_rate": 1.0079556421699533e-06, + "loss": 0.04394418001174927, + "step": 4340 + }, + { + "epoch": 0.710204081632653, + "grad_norm": 0.12708403170108795, + "learning_rate": 9.97516465863698e-07, + "loss": 0.04101270735263825, + "step": 4350 + }, + { + "epoch": 0.7118367346938775, + "grad_norm": 0.08838815987110138, + "learning_rate": 9.87118139775595e-07, + "loss": 0.04232141673564911, + "step": 4360 + }, + { + "epoch": 0.713469387755102, + "grad_norm": 0.13831272721290588, + "learning_rate": 9.767609466201494e-07, + "loss": 0.03110981285572052, + "step": 4370 + }, + { + "epoch": 0.7151020408163266, + "grad_norm": 0.14169402420520782, + "learning_rate": 9.664451679935252e-07, + "loss": 0.036100244522094725, + "step": 4380 + }, + { + "epoch": 0.7167346938775511, + "grad_norm": 0.09364918619394302, + "learning_rate": 9.561710843658892e-07, + "loss": 0.028025591373443605, + "step": 4390 + }, + { + "epoch": 0.7183673469387755, + "grad_norm": 0.08178938925266266, + "learning_rate": 9.459389750737854e-07, + "loss": 0.03276517987251282, + "step": 4400 + }, + { + "epoch": 0.7183673469387755, + "eval_accuracy": 0.9830833333333334, + "eval_loss": 0.03968822583556175, + "eval_runtime": 170.9863, + "eval_samples_per_second": 23.394, + "eval_steps_per_second": 23.394, + "step": 4400 + }, + { + "epoch": 0.72, + "grad_norm": 0.21406538784503937, + "learning_rate": 9.357491183125398e-07, + "loss": 0.04240003228187561, + "step": 4410 + }, + { + "epoch": 0.7216326530612245, + "grad_norm": 0.2528787851333618, + "learning_rate": 9.256017911286958e-07, + "loss": 0.03953196704387665, + "step": 4420 + }, + { + "epoch": 0.723265306122449, + "grad_norm": 0.1817523092031479, + "learning_rate": 9.154972694124866e-07, + "loss": 0.03630510270595551, + "step": 4430 + }, + { + "epoch": 0.7248979591836735, + "grad_norm": 0.19527313113212585, + "learning_rate": 9.054358278903272e-07, + "loss": 0.04581010937690735, + "step": 4440 + }, + { + "epoch": 0.726530612244898, + "grad_norm": 0.3551980257034302, + "learning_rate": 8.954177401173491e-07, + "loss": 0.04633161723613739, + "step": 4450 + }, + { + "epoch": 0.7281632653061224, + "grad_norm": 0.1003197506070137, + "learning_rate": 8.854432784699624e-07, + "loss": 0.03686977326869965, + "step": 4460 + }, + { + "epoch": 0.7297959183673469, + "grad_norm": 0.1250677853822708, + "learning_rate": 8.755127141384495e-07, + "loss": 0.03494169116020203, + "step": 4470 + }, + { + "epoch": 0.7314285714285714, + "grad_norm": 0.23732195794582367, + "learning_rate": 8.656263171195931e-07, + "loss": 0.04174305498600006, + "step": 4480 + }, + { + "epoch": 0.7330612244897959, + "grad_norm": 0.2553883194923401, + "learning_rate": 8.557843562093329e-07, + "loss": 0.04263312518596649, + "step": 4490 + }, + { + "epoch": 0.7346938775510204, + "grad_norm": 0.1312720775604248, + "learning_rate": 8.459870989954611e-07, + "loss": 0.03155354261398315, + "step": 4500 + }, + { + "epoch": 0.7363265306122448, + "grad_norm": 0.2632088363170624, + "learning_rate": 8.362348118503424e-07, + "loss": 0.05083600878715515, + "step": 4510 + }, + { + "epoch": 0.7379591836734694, + "grad_norm": 0.1293715387582779, + "learning_rate": 8.265277599236785e-07, + "loss": 0.031102049350738525, + "step": 4520 + }, + { + "epoch": 0.7395918367346939, + "grad_norm": 0.15807363390922546, + "learning_rate": 8.168662071352918e-07, + "loss": 0.033837145566940306, + "step": 4530 + }, + { + "epoch": 0.7412244897959184, + "grad_norm": 0.08791180700063705, + "learning_rate": 8.072504161679534e-07, + "loss": 0.0337544322013855, + "step": 4540 + }, + { + "epoch": 0.7428571428571429, + "grad_norm": 0.10295934230089188, + "learning_rate": 7.976806484602417e-07, + "loss": 0.03291972577571869, + "step": 4550 + }, + { + "epoch": 0.7444897959183674, + "grad_norm": 0.11660359799861908, + "learning_rate": 7.881571641994331e-07, + "loss": 0.032550466060638425, + "step": 4560 + }, + { + "epoch": 0.7461224489795918, + "grad_norm": 0.1896156519651413, + "learning_rate": 7.786802223144286e-07, + "loss": 0.030972787737846376, + "step": 4570 + }, + { + "epoch": 0.7477551020408163, + "grad_norm": 0.2084398865699768, + "learning_rate": 7.69250080468712e-07, + "loss": 0.0415867954492569, + "step": 4580 + }, + { + "epoch": 0.7493877551020408, + "grad_norm": 0.12618854641914368, + "learning_rate": 7.59866995053348e-07, + "loss": 0.0360225111246109, + "step": 4590 + }, + { + "epoch": 0.7510204081632653, + "grad_norm": 0.12291006743907928, + "learning_rate": 7.505312211800081e-07, + "loss": 0.03627174496650696, + "step": 4600 + }, + { + "epoch": 0.7510204081632653, + "eval_accuracy": 0.9855833333333334, + "eval_loss": 0.03568050265312195, + "eval_runtime": 156.9074, + "eval_samples_per_second": 25.493, + "eval_steps_per_second": 25.493, + "step": 4600 + }, + { + "epoch": 0.7526530612244898, + "grad_norm": 0.1491072177886963, + "learning_rate": 7.41243012674035e-07, + "loss": 0.033873635530471805, + "step": 4610 + }, + { + "epoch": 0.7542857142857143, + "grad_norm": 0.14268337190151215, + "learning_rate": 7.320026220675447e-07, + "loss": 0.03949807286262512, + "step": 4620 + }, + { + "epoch": 0.7559183673469387, + "grad_norm": 0.15548810362815857, + "learning_rate": 7.228103005925557e-07, + "loss": 0.03478931188583374, + "step": 4630 + }, + { + "epoch": 0.7575510204081632, + "grad_norm": 0.14298392832279205, + "learning_rate": 7.136662981741613e-07, + "loss": 0.030326524376869203, + "step": 4640 + }, + { + "epoch": 0.7591836734693878, + "grad_norm": 0.11854545772075653, + "learning_rate": 7.045708634237341e-07, + "loss": 0.034136393666267396, + "step": 4650 + }, + { + "epoch": 0.7608163265306123, + "grad_norm": 0.1346571296453476, + "learning_rate": 6.955242436321663e-07, + "loss": 0.040317881107330325, + "step": 4660 + }, + { + "epoch": 0.7624489795918368, + "grad_norm": 0.13217571377754211, + "learning_rate": 6.865266847631463e-07, + "loss": 0.02895120084285736, + "step": 4670 + }, + { + "epoch": 0.7640816326530612, + "grad_norm": 0.0761382132768631, + "learning_rate": 6.775784314464717e-07, + "loss": 0.029726022481918336, + "step": 4680 + }, + { + "epoch": 0.7657142857142857, + "grad_norm": 0.17583715915679932, + "learning_rate": 6.686797269713971e-07, + "loss": 0.02383600324392319, + "step": 4690 + }, + { + "epoch": 0.7673469387755102, + "grad_norm": 0.13434812426567078, + "learning_rate": 6.598308132800221e-07, + "loss": 0.03801000118255615, + "step": 4700 + }, + { + "epoch": 0.7689795918367347, + "grad_norm": 0.0841183140873909, + "learning_rate": 6.510319309607099e-07, + "loss": 0.030053699016571046, + "step": 4710 + }, + { + "epoch": 0.7706122448979592, + "grad_norm": 0.10911499708890915, + "learning_rate": 6.422833192415476e-07, + "loss": 0.03296325206756592, + "step": 4720 + }, + { + "epoch": 0.7722448979591837, + "grad_norm": 0.08562543988227844, + "learning_rate": 6.335852159838423e-07, + "loss": 0.029071077704429626, + "step": 4730 + }, + { + "epoch": 0.7738775510204081, + "grad_norm": 0.24464724957942963, + "learning_rate": 6.249378576756538e-07, + "loss": 0.03307548463344574, + "step": 4740 + }, + { + "epoch": 0.7755102040816326, + "grad_norm": 0.17294660210609436, + "learning_rate": 6.163414794253641e-07, + "loss": 0.03507235944271088, + "step": 4750 + }, + { + "epoch": 0.7771428571428571, + "grad_norm": 0.18331816792488098, + "learning_rate": 6.077963149552868e-07, + "loss": 0.038254004716873166, + "step": 4760 + }, + { + "epoch": 0.7787755102040816, + "grad_norm": 0.10756120830774307, + "learning_rate": 5.993025965953103e-07, + "loss": 0.03428310751914978, + "step": 4770 + }, + { + "epoch": 0.7804081632653062, + "grad_norm": 0.16071061789989471, + "learning_rate": 5.908605552765833e-07, + "loss": 0.032243117690086365, + "step": 4780 + }, + { + "epoch": 0.7820408163265307, + "grad_norm": 0.14897944033145905, + "learning_rate": 5.824704205252343e-07, + "loss": 0.03812848925590515, + "step": 4790 + }, + { + "epoch": 0.7836734693877551, + "grad_norm": 0.1536232829093933, + "learning_rate": 5.741324204561335e-07, + "loss": 0.031389710307121274, + "step": 4800 + }, + { + "epoch": 0.7836734693877551, + "eval_accuracy": 0.985, + "eval_loss": 0.03615976870059967, + "eval_runtime": 153.2713, + "eval_samples_per_second": 26.098, + "eval_steps_per_second": 26.098, + "step": 4800 + }, + { + "epoch": 0.7853061224489796, + "grad_norm": 0.15687023103237152, + "learning_rate": 5.65846781766688e-07, + "loss": 0.027873021364212037, + "step": 4810 + }, + { + "epoch": 0.7869387755102041, + "grad_norm": 0.14541009068489075, + "learning_rate": 5.576137297306794e-07, + "loss": 0.0407156765460968, + "step": 4820 + }, + { + "epoch": 0.7885714285714286, + "grad_norm": 0.16398967802524567, + "learning_rate": 5.494334881921393e-07, + "loss": 0.03936888575553894, + "step": 4830 + }, + { + "epoch": 0.7902040816326531, + "grad_norm": 0.2177809625864029, + "learning_rate": 5.413062795592633e-07, + "loss": 0.03059300184249878, + "step": 4840 + }, + { + "epoch": 0.7918367346938775, + "grad_norm": 0.1487113982439041, + "learning_rate": 5.332323247983632e-07, + "loss": 0.038730043172836306, + "step": 4850 + }, + { + "epoch": 0.793469387755102, + "grad_norm": 0.1111542284488678, + "learning_rate": 5.252118434278605e-07, + "loss": 0.033898717164993285, + "step": 4860 + }, + { + "epoch": 0.7951020408163265, + "grad_norm": 0.10296906530857086, + "learning_rate": 5.172450535123166e-07, + "loss": 0.033659690618515016, + "step": 4870 + }, + { + "epoch": 0.796734693877551, + "grad_norm": 0.11844902485609055, + "learning_rate": 5.093321716565064e-07, + "loss": 0.0293354332447052, + "step": 4880 + }, + { + "epoch": 0.7983673469387755, + "grad_norm": 0.08401476591825485, + "learning_rate": 5.014734129995258e-07, + "loss": 0.027682244777679443, + "step": 4890 + }, + { + "epoch": 0.8, + "grad_norm": 0.15870651602745056, + "learning_rate": 4.936689912089443e-07, + "loss": 0.029118484258651732, + "step": 4900 + }, + { + "epoch": 0.8016326530612244, + "grad_norm": 0.10739132016897202, + "learning_rate": 4.859191184749968e-07, + "loss": 0.031624987721443176, + "step": 4910 + }, + { + "epoch": 0.803265306122449, + "grad_norm": 0.1523323655128479, + "learning_rate": 4.782240055048113e-07, + "loss": 0.03415923118591309, + "step": 4920 + }, + { + "epoch": 0.8048979591836735, + "grad_norm": 0.31860172748565674, + "learning_rate": 4.705838615166836e-07, + "loss": 0.035749971866607666, + "step": 4930 + }, + { + "epoch": 0.806530612244898, + "grad_norm": 0.15043528378009796, + "learning_rate": 4.629988942343863e-07, + "loss": 0.029294773936271667, + "step": 4940 + }, + { + "epoch": 0.8081632653061225, + "grad_norm": 0.1007913127541542, + "learning_rate": 4.55469309881523e-07, + "loss": 0.026727235317230223, + "step": 4950 + }, + { + "epoch": 0.809795918367347, + "grad_norm": 0.10851600021123886, + "learning_rate": 4.479953131759199e-07, + "loss": 0.029011189937591553, + "step": 4960 + }, + { + "epoch": 0.8114285714285714, + "grad_norm": 0.19257904589176178, + "learning_rate": 4.4057710732406113e-07, + "loss": 0.040424668788909913, + "step": 4970 + }, + { + "epoch": 0.8130612244897959, + "grad_norm": 0.08985571563243866, + "learning_rate": 4.332148940155617e-07, + "loss": 0.031291928887367246, + "step": 4980 + }, + { + "epoch": 0.8146938775510204, + "grad_norm": 0.09874805808067322, + "learning_rate": 4.259088734176883e-07, + "loss": 0.028314629197120668, + "step": 4990 + }, + { + "epoch": 0.8163265306122449, + "grad_norm": 0.2056759148836136, + "learning_rate": 4.186592441699119e-07, + "loss": 0.028049173951148986, + "step": 5000 + }, + { + "epoch": 0.8163265306122449, + "eval_accuracy": 0.986125, + "eval_loss": 0.033585112541913986, + "eval_runtime": 153.2119, + "eval_samples_per_second": 26.108, + "eval_steps_per_second": 26.108, + "step": 5000 + }, + { + "epoch": 0.8179591836734694, + "grad_norm": 0.10125680267810822, + "learning_rate": 4.1146620337850965e-07, + "loss": 0.03612466752529144, + "step": 5010 + }, + { + "epoch": 0.8195918367346938, + "grad_norm": 0.10107982903718948, + "learning_rate": 4.043299466112058e-07, + "loss": 0.026968371868133546, + "step": 5020 + }, + { + "epoch": 0.8212244897959183, + "grad_norm": 0.13236716389656067, + "learning_rate": 3.9725066789185446e-07, + "loss": 0.030766445398330688, + "step": 5030 + }, + { + "epoch": 0.8228571428571428, + "grad_norm": 0.1182790994644165, + "learning_rate": 3.9022855969516384e-07, + "loss": 0.028759539127349854, + "step": 5040 + }, + { + "epoch": 0.8244897959183674, + "grad_norm": 0.1430918574333191, + "learning_rate": 3.832638129414637e-07, + "loss": 0.03126598298549652, + "step": 5050 + }, + { + "epoch": 0.8261224489795919, + "grad_norm": 0.24944688379764557, + "learning_rate": 3.7635661699151366e-07, + "loss": 0.030412834882736207, + "step": 5060 + }, + { + "epoch": 0.8277551020408164, + "grad_norm": 0.24371539056301117, + "learning_rate": 3.695071596413574e-07, + "loss": 0.03904336988925934, + "step": 5070 + }, + { + "epoch": 0.8293877551020408, + "grad_norm": 0.25066161155700684, + "learning_rate": 3.627156271172125e-07, + "loss": 0.032776132225990295, + "step": 5080 + }, + { + "epoch": 0.8310204081632653, + "grad_norm": 0.1635182797908783, + "learning_rate": 3.559822040704108e-07, + "loss": 0.027044805884361266, + "step": 5090 + }, + { + "epoch": 0.8326530612244898, + "grad_norm": 0.16698378324508667, + "learning_rate": 3.4930707357237666e-07, + "loss": 0.03271254003047943, + "step": 5100 + }, + { + "epoch": 0.8342857142857143, + "grad_norm": 0.18812412023544312, + "learning_rate": 3.4269041710964903e-07, + "loss": 0.030745366215705873, + "step": 5110 + }, + { + "epoch": 0.8359183673469388, + "grad_norm": 0.18739669024944305, + "learning_rate": 3.36132414578948e-07, + "loss": 0.032180759310722354, + "step": 5120 + }, + { + "epoch": 0.8375510204081633, + "grad_norm": 0.13260763883590698, + "learning_rate": 3.296332442822839e-07, + "loss": 0.026755687594413758, + "step": 5130 + }, + { + "epoch": 0.8391836734693877, + "grad_norm": 0.06771053373813629, + "learning_rate": 3.231930829221083e-07, + "loss": 0.02895594835281372, + "step": 5140 + }, + { + "epoch": 0.8408163265306122, + "grad_norm": 0.09441749006509781, + "learning_rate": 3.168121055965112e-07, + "loss": 0.033178862929344174, + "step": 5150 + }, + { + "epoch": 0.8424489795918367, + "grad_norm": 0.0863656997680664, + "learning_rate": 3.104904857944591e-07, + "loss": 0.02276728004217148, + "step": 5160 + }, + { + "epoch": 0.8440816326530612, + "grad_norm": 0.2047131359577179, + "learning_rate": 3.0422839539107794e-07, + "loss": 0.028258222341537475, + "step": 5170 + }, + { + "epoch": 0.8457142857142858, + "grad_norm": 0.2389051467180252, + "learning_rate": 2.980260046429834e-07, + "loss": 0.02725991904735565, + "step": 5180 + }, + { + "epoch": 0.8473469387755102, + "grad_norm": 0.1733924299478531, + "learning_rate": 2.9188348218364635e-07, + "loss": 0.0327379584312439, + "step": 5190 + }, + { + "epoch": 0.8489795918367347, + "grad_norm": 0.19450697302818298, + "learning_rate": 2.8580099501881285e-07, + "loss": 0.030778995156288146, + "step": 5200 + }, + { + "epoch": 0.8489795918367347, + "eval_accuracy": 0.9881666666666667, + "eval_loss": 0.02920014224946499, + "eval_runtime": 155.6487, + "eval_samples_per_second": 25.699, + "eval_steps_per_second": 25.699, + "step": 5200 + }, + { + "epoch": 0.8506122448979592, + "grad_norm": 0.14990773797035217, + "learning_rate": 2.7977870852195997e-07, + "loss": 0.03692179024219513, + "step": 5210 + }, + { + "epoch": 0.8522448979591837, + "grad_norm": 0.16051384806632996, + "learning_rate": 2.7381678642980303e-07, + "loss": 0.030528882145881654, + "step": 5220 + }, + { + "epoch": 0.8538775510204082, + "grad_norm": 0.1660623997449875, + "learning_rate": 2.6791539083784014e-07, + "loss": 0.025125756859779358, + "step": 5230 + }, + { + "epoch": 0.8555102040816327, + "grad_norm": 0.07568248361349106, + "learning_rate": 2.620746821959475e-07, + "loss": 0.02720560133457184, + "step": 5240 + }, + { + "epoch": 0.8571428571428571, + "grad_norm": 0.2634601593017578, + "learning_rate": 2.562948193040174e-07, + "loss": 0.033940815925598146, + "step": 5250 + }, + { + "epoch": 0.8587755102040816, + "grad_norm": 0.15926218032836914, + "learning_rate": 2.5057595930763873e-07, + "loss": 0.028036150336265563, + "step": 5260 + }, + { + "epoch": 0.8604081632653061, + "grad_norm": 0.19775289297103882, + "learning_rate": 2.4491825769382526e-07, + "loss": 0.02731768786907196, + "step": 5270 + }, + { + "epoch": 0.8620408163265306, + "grad_norm": 0.07538789510726929, + "learning_rate": 2.39321868286789e-07, + "loss": 0.0313712328672409, + "step": 5280 + }, + { + "epoch": 0.8636734693877551, + "grad_norm": 0.10647427290678024, + "learning_rate": 2.337869432437567e-07, + "loss": 0.02781860828399658, + "step": 5290 + }, + { + "epoch": 0.8653061224489796, + "grad_norm": 0.20678064227104187, + "learning_rate": 2.283136330508337e-07, + "loss": 0.028782907128334045, + "step": 5300 + }, + { + "epoch": 0.866938775510204, + "grad_norm": 0.18386024236679077, + "learning_rate": 2.2290208651891244e-07, + "loss": 0.02515394687652588, + "step": 5310 + }, + { + "epoch": 0.8685714285714285, + "grad_norm": 0.13139928877353668, + "learning_rate": 2.1755245077962618e-07, + "loss": 0.02002216875553131, + "step": 5320 + }, + { + "epoch": 0.8702040816326531, + "grad_norm": 0.10402868688106537, + "learning_rate": 2.1226487128134854e-07, + "loss": 0.03221684396266937, + "step": 5330 + }, + { + "epoch": 0.8718367346938776, + "grad_norm": 0.056439850479364395, + "learning_rate": 2.070394917852403e-07, + "loss": 0.023763810098171235, + "step": 5340 + }, + { + "epoch": 0.8734693877551021, + "grad_norm": 0.1421876698732376, + "learning_rate": 2.0187645436133913e-07, + "loss": 0.03192594051361084, + "step": 5350 + }, + { + "epoch": 0.8751020408163265, + "grad_norm": 0.20487430691719055, + "learning_rate": 1.967758993846966e-07, + "loss": 0.025955894589424135, + "step": 5360 + }, + { + "epoch": 0.876734693877551, + "grad_norm": 0.2183534801006317, + "learning_rate": 1.9173796553156487e-07, + "loss": 0.027435585856437683, + "step": 5370 + }, + { + "epoch": 0.8783673469387755, + "grad_norm": 0.19724565744400024, + "learning_rate": 1.8676278977562219e-07, + "loss": 0.028912261128425598, + "step": 5380 + }, + { + "epoch": 0.88, + "grad_norm": 0.21412917971611023, + "learning_rate": 1.818505073842508e-07, + "loss": 0.029015693068504333, + "step": 5390 + }, + { + "epoch": 0.8816326530612245, + "grad_norm": 0.08429288119077682, + "learning_rate": 1.770012519148598e-07, + "loss": 0.025249993801116942, + "step": 5400 + }, + { + "epoch": 0.8816326530612245, + "eval_accuracy": 0.989375, + "eval_loss": 0.02678457647562027, + "eval_runtime": 156.2763, + "eval_samples_per_second": 25.596, + "eval_steps_per_second": 25.596, + "step": 5400 + }, + { + "epoch": 0.883265306122449, + "grad_norm": 0.15489192306995392, + "learning_rate": 1.722151552112525e-07, + "loss": 0.03486272096633911, + "step": 5410 + }, + { + "epoch": 0.8848979591836734, + "grad_norm": 0.1594313681125641, + "learning_rate": 1.6749234740004227e-07, + "loss": 0.024751077592372894, + "step": 5420 + }, + { + "epoch": 0.8865306122448979, + "grad_norm": 0.12505307793617249, + "learning_rate": 1.6283295688711564e-07, + "loss": 0.025553828477859496, + "step": 5430 + }, + { + "epoch": 0.8881632653061224, + "grad_norm": 0.08852038532495499, + "learning_rate": 1.582371103541394e-07, + "loss": 0.025743791460990907, + "step": 5440 + }, + { + "epoch": 0.889795918367347, + "grad_norm": 0.08450407534837723, + "learning_rate": 1.5370493275511767e-07, + "loss": 0.02687264084815979, + "step": 5450 + }, + { + "epoch": 0.8914285714285715, + "grad_norm": 0.054211657494306564, + "learning_rate": 1.4923654731299363e-07, + "loss": 0.023073682188987733, + "step": 5460 + }, + { + "epoch": 0.893061224489796, + "grad_norm": 0.23101477324962616, + "learning_rate": 1.4483207551630045e-07, + "loss": 0.0280561625957489, + "step": 5470 + }, + { + "epoch": 0.8946938775510204, + "grad_norm": 0.19923920929431915, + "learning_rate": 1.4049163711585684e-07, + "loss": 0.02133965492248535, + "step": 5480 + }, + { + "epoch": 0.8963265306122449, + "grad_norm": 0.13227398693561554, + "learning_rate": 1.3621535012151244e-07, + "loss": 0.017637649178504945, + "step": 5490 + }, + { + "epoch": 0.8979591836734694, + "grad_norm": 0.11182940751314163, + "learning_rate": 1.3200333079893823e-07, + "loss": 0.02847398817539215, + "step": 5500 + }, + { + "epoch": 0.8995918367346939, + "grad_norm": 0.06762958317995071, + "learning_rate": 1.2785569366646628e-07, + "loss": 0.020397891104221345, + "step": 5510 + }, + { + "epoch": 0.9012244897959184, + "grad_norm": 0.06652984023094177, + "learning_rate": 1.2377255149197576e-07, + "loss": 0.03015478849411011, + "step": 5520 + }, + { + "epoch": 0.9028571428571428, + "grad_norm": 0.16087517142295837, + "learning_rate": 1.1975401528982718e-07, + "loss": 0.02614833414554596, + "step": 5530 + }, + { + "epoch": 0.9044897959183673, + "grad_norm": 0.0873415544629097, + "learning_rate": 1.1580019431784384e-07, + "loss": 0.023097904026508333, + "step": 5540 + }, + { + "epoch": 0.9061224489795918, + "grad_norm": 0.26114869117736816, + "learning_rate": 1.1191119607434076e-07, + "loss": 0.028036117553710938, + "step": 5550 + }, + { + "epoch": 0.9077551020408163, + "grad_norm": 0.08690405637025833, + "learning_rate": 1.0808712629520435e-07, + "loss": 0.027321887016296387, + "step": 5560 + }, + { + "epoch": 0.9093877551020408, + "grad_norm": 0.19979159533977509, + "learning_rate": 1.0432808895101426e-07, + "loss": 0.03543014228343964, + "step": 5570 + }, + { + "epoch": 0.9110204081632653, + "grad_norm": 0.09599421173334122, + "learning_rate": 1.0063418624421895e-07, + "loss": 0.02044544517993927, + "step": 5580 + }, + { + "epoch": 0.9126530612244897, + "grad_norm": 0.22291070222854614, + "learning_rate": 9.700551860635538e-08, + "loss": 0.027166444063186645, + "step": 5590 + }, + { + "epoch": 0.9142857142857143, + "grad_norm": 0.10956079512834549, + "learning_rate": 9.344218469532012e-08, + "loss": 0.02024113982915878, + "step": 5600 + }, + { + "epoch": 0.9142857142857143, + "eval_accuracy": 0.9900833333333334, + "eval_loss": 0.025807492434978485, + "eval_runtime": 154.9378, + "eval_samples_per_second": 25.817, + "eval_steps_per_second": 25.817, + "step": 5600 + }, + { + "epoch": 0.9159183673469388, + "grad_norm": 0.14949144423007965, + "learning_rate": 8.99442813926854e-08, + "loss": 0.024004274606704713, + "step": 5610 + }, + { + "epoch": 0.9175510204081633, + "grad_norm": 0.2315555214881897, + "learning_rate": 8.651190380106673e-08, + "loss": 0.024219857156276704, + "step": 5620 + }, + { + "epoch": 0.9191836734693878, + "grad_norm": 0.06734409183263779, + "learning_rate": 8.314514524153522e-08, + "loss": 0.02802177369594574, + "step": 5630 + }, + { + "epoch": 0.9208163265306123, + "grad_norm": 0.21427671611309052, + "learning_rate": 7.984409725108222e-08, + "loss": 0.021109704673290253, + "step": 5640 + }, + { + "epoch": 0.9224489795918367, + "grad_norm": 0.18013973534107208, + "learning_rate": 7.660884958012899e-08, + "loss": 0.020703735947608947, + "step": 5650 + }, + { + "epoch": 0.9240816326530612, + "grad_norm": 0.2695794701576233, + "learning_rate": 7.343949019008756e-08, + "loss": 0.03247661292552948, + "step": 5660 + }, + { + "epoch": 0.9257142857142857, + "grad_norm": 0.15175238251686096, + "learning_rate": 7.03361052509688e-08, + "loss": 0.021647430956363678, + "step": 5670 + }, + { + "epoch": 0.9273469387755102, + "grad_norm": 0.18325908482074738, + "learning_rate": 6.729877913903976e-08, + "loss": 0.028400209546089173, + "step": 5680 + }, + { + "epoch": 0.9289795918367347, + "grad_norm": 0.15859369933605194, + "learning_rate": 6.432759443452924e-08, + "loss": 0.014232246577739716, + "step": 5690 + }, + { + "epoch": 0.9306122448979591, + "grad_norm": 0.13438236713409424, + "learning_rate": 6.14226319193828e-08, + "loss": 0.02244553864002228, + "step": 5700 + }, + { + "epoch": 0.9322448979591836, + "grad_norm": 0.27041247487068176, + "learning_rate": 5.858397057506709e-08, + "loss": 0.029011306166648865, + "step": 5710 + }, + { + "epoch": 0.9338775510204081, + "grad_norm": 0.12215204536914825, + "learning_rate": 5.581168758042127e-08, + "loss": 0.028310227394104003, + "step": 5720 + }, + { + "epoch": 0.9355102040816327, + "grad_norm": 0.2103251814842224, + "learning_rate": 5.310585830955922e-08, + "loss": 0.021468037366867067, + "step": 5730 + }, + { + "epoch": 0.9371428571428572, + "grad_norm": 0.17152205109596252, + "learning_rate": 5.0466556329819835e-08, + "loss": 0.024149742722511292, + "step": 5740 + }, + { + "epoch": 0.9387755102040817, + "grad_norm": 0.08712399005889893, + "learning_rate": 4.789385339976832e-08, + "loss": 0.01809695065021515, + "step": 5750 + }, + { + "epoch": 0.9404081632653061, + "grad_norm": 0.15006859600543976, + "learning_rate": 4.53878194672433e-08, + "loss": 0.022831857204437256, + "step": 5760 + }, + { + "epoch": 0.9420408163265306, + "grad_norm": 0.06355110555887222, + "learning_rate": 4.2948522667455864e-08, + "loss": 0.015157052874565124, + "step": 5770 + }, + { + "epoch": 0.9436734693877551, + "grad_norm": 0.14330919086933136, + "learning_rate": 4.057602932113658e-08, + "loss": 0.02486245185136795, + "step": 5780 + }, + { + "epoch": 0.9453061224489796, + "grad_norm": 0.12216096371412277, + "learning_rate": 3.82704039327339e-08, + "loss": 0.024203602969646455, + "step": 5790 + }, + { + "epoch": 0.9469387755102041, + "grad_norm": 0.19883109629154205, + "learning_rate": 3.6031709188658336e-08, + "loss": 0.025162199139595033, + "step": 5800 + }, + { + "epoch": 0.9469387755102041, + "eval_accuracy": 0.99025, + "eval_loss": 0.025506427511572838, + "eval_runtime": 154.1326, + "eval_samples_per_second": 25.952, + "eval_steps_per_second": 25.952, + "step": 5800 + }, + { + "epoch": 0.9485714285714286, + "grad_norm": 0.25931984186172485, + "learning_rate": 3.386000595557964e-08, + "loss": 0.027550938725471496, + "step": 5810 + }, + { + "epoch": 0.950204081632653, + "grad_norm": 0.07933743298053741, + "learning_rate": 3.1755353278770936e-08, + "loss": 0.024577364325523376, + "step": 5820 + }, + { + "epoch": 0.9518367346938775, + "grad_norm": 0.13839955627918243, + "learning_rate": 2.9717808380503866e-08, + "loss": 0.021038763225078583, + "step": 5830 + }, + { + "epoch": 0.953469387755102, + "grad_norm": 0.14696243405342102, + "learning_rate": 2.7747426658492893e-08, + "loss": 0.02484375685453415, + "step": 5840 + }, + { + "epoch": 0.9551020408163265, + "grad_norm": 0.1538810133934021, + "learning_rate": 2.5844261684388473e-08, + "loss": 0.023458963632583617, + "step": 5850 + }, + { + "epoch": 0.9567346938775511, + "grad_norm": 0.08468759804964066, + "learning_rate": 2.4008365202321525e-08, + "loss": 0.03132525682449341, + "step": 5860 + }, + { + "epoch": 0.9583673469387755, + "grad_norm": 0.1627626270055771, + "learning_rate": 2.2239787127495695e-08, + "loss": 0.03181464076042175, + "step": 5870 + }, + { + "epoch": 0.96, + "grad_norm": 0.17413999140262604, + "learning_rate": 2.0538575544831186e-08, + "loss": 0.019847501814365388, + "step": 5880 + }, + { + "epoch": 0.9616326530612245, + "grad_norm": 0.23201079666614532, + "learning_rate": 1.8904776707655837e-08, + "loss": 0.02205444723367691, + "step": 5890 + }, + { + "epoch": 0.963265306122449, + "grad_norm": 0.17272613942623138, + "learning_rate": 1.7338435036449164e-08, + "loss": 0.022021512687206268, + "step": 5900 + }, + { + "epoch": 0.9648979591836735, + "grad_norm": 0.1446181833744049, + "learning_rate": 1.5839593117634166e-08, + "loss": 0.02300398051738739, + "step": 5910 + }, + { + "epoch": 0.966530612244898, + "grad_norm": 0.21913999319076538, + "learning_rate": 1.4408291702418808e-08, + "loss": 0.024825331568717957, + "step": 5920 + }, + { + "epoch": 0.9681632653061224, + "grad_norm": 0.12140090018510818, + "learning_rate": 1.3044569705689125e-08, + "loss": 0.02171301543712616, + "step": 5930 + }, + { + "epoch": 0.9697959183673469, + "grad_norm": 0.08353517204523087, + "learning_rate": 1.1748464204950072e-08, + "loss": 0.02564157545566559, + "step": 5940 + }, + { + "epoch": 0.9714285714285714, + "grad_norm": 0.2671544551849365, + "learning_rate": 1.052001043931855e-08, + "loss": 0.02467280924320221, + "step": 5950 + }, + { + "epoch": 0.9730612244897959, + "grad_norm": 0.16360966861248016, + "learning_rate": 9.3592418085639e-09, + "loss": 0.02429400235414505, + "step": 5960 + }, + { + "epoch": 0.9746938775510204, + "grad_norm": 0.08739493787288666, + "learning_rate": 8.266189872201957e-09, + "loss": 0.02331821173429489, + "step": 5970 + }, + { + "epoch": 0.976326530612245, + "grad_norm": 0.23632948100566864, + "learning_rate": 7.24088434863407e-09, + "loss": 0.02990926504135132, + "step": 5980 + }, + { + "epoch": 0.9779591836734693, + "grad_norm": 0.2005041390657425, + "learning_rate": 6.283353114342749e-09, + "loss": 0.023541297018527984, + "step": 5990 + }, + { + "epoch": 0.9795918367346939, + "grad_norm": 0.16744142770767212, + "learning_rate": 5.393622203129778e-09, + "loss": 0.020145678520202638, + "step": 6000 + }, + { + "epoch": 0.9795918367346939, + "eval_accuracy": 0.9904166666666667, + "eval_loss": 0.025301460176706314, + "eval_runtime": 157.9993, + "eval_samples_per_second": 25.317, + "eval_steps_per_second": 25.317, + "step": 6000 + }, + { + "epoch": 0.9812244897959184, + "grad_norm": 0.1079549789428711, + "learning_rate": 4.5717158054114916e-09, + "loss": 0.02092682272195816, + "step": 6010 + }, + { + "epoch": 0.9828571428571429, + "grad_norm": 0.28545212745666504, + "learning_rate": 3.817656267559866e-09, + "loss": 0.029862532019615175, + "step": 6020 + }, + { + "epoch": 0.9844897959183674, + "grad_norm": 0.09164315462112427, + "learning_rate": 3.1314640912943893e-09, + "loss": 0.021963220834732056, + "step": 6030 + }, + { + "epoch": 0.9861224489795918, + "grad_norm": 0.10882832109928131, + "learning_rate": 2.5131579331261204e-09, + "loss": 0.01937238872051239, + "step": 6040 + }, + { + "epoch": 0.9877551020408163, + "grad_norm": 0.2565634846687317, + "learning_rate": 1.9627546038492042e-09, + "loss": 0.03262064456939697, + "step": 6050 + }, + { + "epoch": 0.9893877551020408, + "grad_norm": 0.20641514658927917, + "learning_rate": 1.4802690680842947e-09, + "loss": 0.02464511841535568, + "step": 6060 + }, + { + "epoch": 0.9910204081632653, + "grad_norm": 0.15149660408496857, + "learning_rate": 1.065714443871102e-09, + "loss": 0.020028088986873627, + "step": 6070 + }, + { + "epoch": 0.9926530612244898, + "grad_norm": 0.24861690402030945, + "learning_rate": 7.191020023133988e-10, + "loss": 0.03540560901165009, + "step": 6080 + }, + { + "epoch": 0.9942857142857143, + "grad_norm": 0.1762499213218689, + "learning_rate": 4.404411672703779e-10, + "loss": 0.026075825095176697, + "step": 6090 + }, + { + "epoch": 0.9959183673469387, + "grad_norm": 0.23214563727378845, + "learning_rate": 2.2973951510185665e-10, + "loss": 0.024257904291152953, + "step": 6100 + }, + { + "epoch": 0.9975510204081632, + "grad_norm": 0.13295525312423706, + "learning_rate": 8.700277446233029e-11, + "loss": 0.021962395310401915, + "step": 6110 + }, + { + "epoch": 0.9991836734693877, + "grad_norm": 0.1011875718832016, + "learning_rate": 1.2234826143875388e-11, + "loss": 0.025374814867973328, + "step": 6120 + } + ], + "logging_steps": 10, + "max_steps": 6125, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 200, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 5.72407910760448e+16, + "train_batch_size": 1, + "trial_name": null, + "trial_params": null +} diff --git a/checkpoint-6125/training_args.bin b/checkpoint-6125/training_args.bin new file mode 100644 index 0000000..a9304c0 --- /dev/null +++ b/checkpoint-6125/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:10a2085a9c347d21774d35a2558a0376b0ffbb33ca9517fafbd4cd5d12552b02 +size 5585 diff --git a/config.json b/config.json new file mode 100644 index 0000000..642be54 --- /dev/null +++ b/config.json @@ -0,0 +1,58 @@ +{ + "architectures": [ + "Qwen2ForCausalLM" + ], + "attention_dropout": 0.0, + "bos_token_id": null, + "dtype": "float32", + "eos_token_id": 151643, + "hidden_act": "silu", + "hidden_size": 896, + "initializer_range": 0.02, + "intermediate_size": 4864, + "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" + ], + "max_position_embeddings": 32768, + "max_window_layers": 24, + "model_type": "qwen2", + "num_attention_heads": 14, + "num_hidden_layers": 24, + "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_mrope": false, + "use_sliding_window": false, + "vocab_size": 151936 +} diff --git a/eval_results.json b/eval_results.json new file mode 100644 index 0000000..a39142e --- /dev/null +++ b/eval_results.json @@ -0,0 +1,8 @@ +{ + "epoch": 1.0, + "eval_accuracy": 0.9905, + "eval_loss": 0.025262661278247833, + "eval_runtime": 156.0067, + "eval_samples_per_second": 25.64, + "eval_steps_per_second": 25.64 +} \ No newline at end of file diff --git a/generation_config.json b/generation_config.json new file mode 100644 index 0000000..795a75e --- /dev/null +++ b/generation_config.json @@ -0,0 +1,9 @@ +{ + "do_sample": false, + "eos_token_id": [ + 151643 + ], + "max_new_tokens": 2048, + "pad_token_id": 151643, + "transformers_version": "5.2.0" +} diff --git a/model.safetensors b/model.safetensors new file mode 100644 index 0000000..665761c --- /dev/null +++ b/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1edd8c632dc7b25b22dc2c60673de1f976549056f1258360ed58b7276a7cd06c +size 1976163472 diff --git a/tokenizer.json b/tokenizer.json new file mode 100644 index 0000000..34510ff --- /dev/null +++ b/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3fd169731d2cbde95e10bf356d66d5997fd885dd8dbb6fb4684da3f23b2585d8 +size 11421892 diff --git a/tokenizer_config.json b/tokenizer_config.json new file mode 100644 index 0000000..392ded1 --- /dev/null +++ b/tokenizer_config.json @@ -0,0 +1,30 @@ +{ + "add_prefix_space": false, + "backend": "tokenizers", + "bos_token": null, + "clean_up_tokenization_spaces": false, + "eos_token": "<|endoftext|>", + "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/train_results.json b/train_results.json new file mode 100644 index 0000000..d2aab5d --- /dev/null +++ b/train_results.json @@ -0,0 +1,8 @@ +{ + "epoch": 1.0, + "total_flos": 5.72407910760448e+16, + "train_loss": 0.07194560643361539, + "train_runtime": 34224.4204, + "train_samples_per_second": 5.727, + "train_steps_per_second": 0.179 +} \ No newline at end of file diff --git a/trainer_log.jsonl b/trainer_log.jsonl new file mode 100644 index 0000000..27a47d5 --- /dev/null +++ b/trainer_log.jsonl @@ -0,0 +1,643 @@ +{"current_steps": 10, "total_steps": 6125, "loss": 4.391213607788086, "lr": 4.5000000000000003e-07, "epoch": 0.0016326530612244899, "percentage": 0.16, "elapsed_time": "0:00:50", "remaining_time": "8:39:40"} +{"current_steps": 20, "total_steps": 6125, "loss": 1.9876020431518555, "lr": 9.500000000000001e-07, "epoch": 0.0032653061224489797, "percentage": 0.33, "elapsed_time": "0:01:40", "remaining_time": "8:31:52"} +{"current_steps": 30, "total_steps": 6125, "loss": 0.34127225875854494, "lr": 1.45e-06, "epoch": 0.004897959183673469, "percentage": 0.49, "elapsed_time": "0:02:31", "remaining_time": "8:31:44"} +{"current_steps": 40, "total_steps": 6125, "loss": 0.23643033504486083, "lr": 1.9500000000000004e-06, "epoch": 0.006530612244897959, "percentage": 0.65, "elapsed_time": "0:03:20", "remaining_time": "8:27:57"} +{"current_steps": 50, "total_steps": 6125, "loss": 0.23313720226287843, "lr": 2.4500000000000003e-06, "epoch": 0.00816326530612245, "percentage": 0.82, "elapsed_time": "0:04:10", "remaining_time": "8:26:23"} +{"current_steps": 60, "total_steps": 6125, "loss": 0.22351114749908446, "lr": 2.95e-06, "epoch": 0.009795918367346938, "percentage": 0.98, "elapsed_time": "0:05:01", "remaining_time": "8:27:10"} +{"current_steps": 70, "total_steps": 6125, "loss": 0.21615636348724365, "lr": 3.45e-06, "epoch": 0.011428571428571429, "percentage": 1.14, "elapsed_time": "0:05:51", "remaining_time": "8:26:18"} +{"current_steps": 80, "total_steps": 6125, "loss": 0.21677422523498535, "lr": 3.95e-06, "epoch": 0.013061224489795919, "percentage": 1.31, "elapsed_time": "0:06:38", "remaining_time": "8:21:27"} +{"current_steps": 90, "total_steps": 6125, "loss": 0.21422204971313477, "lr": 4.450000000000001e-06, "epoch": 0.014693877551020407, "percentage": 1.47, "elapsed_time": "0:07:25", "remaining_time": "8:17:28"} +{"current_steps": 100, "total_steps": 6125, "loss": 0.19350455999374389, "lr": 4.95e-06, "epoch": 0.0163265306122449, "percentage": 1.63, "elapsed_time": "0:08:11", "remaining_time": "8:13:42"} +{"current_steps": 110, "total_steps": 6125, "loss": 0.1981007933616638, "lr": 4.9999724716692435e-06, "epoch": 0.017959183673469388, "percentage": 1.8, "elapsed_time": "0:08:58", "remaining_time": "8:10:35"} +{"current_steps": 120, "total_steps": 6125, "loss": 0.16994532346725463, "lr": 4.9998773127856985e-06, "epoch": 0.019591836734693877, "percentage": 1.96, "elapsed_time": "0:09:44", "remaining_time": "8:07:43"} +{"current_steps": 130, "total_steps": 6125, "loss": 0.1539763927459717, "lr": 4.999714186080037e-06, "epoch": 0.02122448979591837, "percentage": 2.12, "elapsed_time": "0:10:31", "remaining_time": "8:05:14"} +{"current_steps": 140, "total_steps": 6125, "loss": 0.16321685314178466, "lr": 4.999483095987423e-06, "epoch": 0.022857142857142857, "percentage": 2.29, "elapsed_time": "0:11:17", "remaining_time": "8:02:50"} +{"current_steps": 150, "total_steps": 6125, "loss": 0.16523454189300538, "lr": 4.9991840487908426e-06, "epoch": 0.024489795918367346, "percentage": 2.45, "elapsed_time": "0:12:04", "remaining_time": "8:00:48"} +{"current_steps": 160, "total_steps": 6125, "loss": 0.14424897432327272, "lr": 4.998817052620928e-06, "epoch": 0.026122448979591838, "percentage": 2.61, "elapsed_time": "0:12:50", "remaining_time": "7:58:52"} +{"current_steps": 170, "total_steps": 6125, "loss": 0.13810645341873168, "lr": 4.9983821174557415e-06, "epoch": 0.027755102040816326, "percentage": 2.78, "elapsed_time": "0:13:37", "remaining_time": "7:57:24"} +{"current_steps": 180, "total_steps": 6125, "loss": 0.1318126440048218, "lr": 4.997879255120501e-06, "epoch": 0.029387755102040815, "percentage": 2.94, "elapsed_time": "0:14:24", "remaining_time": "7:55:45"} +{"current_steps": 190, "total_steps": 6125, "loss": 0.14249314069747926, "lr": 4.997308479287263e-06, "epoch": 0.031020408163265307, "percentage": 3.1, "elapsed_time": "0:15:10", "remaining_time": "7:54:15"} +{"current_steps": 200, "total_steps": 6125, "loss": 0.12708702087402343, "lr": 4.996669805474542e-06, "epoch": 0.0326530612244898, "percentage": 3.27, "elapsed_time": "0:15:57", "remaining_time": "7:52:47"} +{"current_steps": 200, "total_steps": 6125, "eval_loss": 0.12691152095794678, "epoch": 0.0326530612244898, "percentage": 3.27, "elapsed_time": "0:18:34", "remaining_time": "9:10:09"} +{"current_steps": 210, "total_steps": 6125, "loss": 0.11019121408462525, "lr": 4.9959632510469e-06, "epoch": 0.03428571428571429, "percentage": 3.43, "elapsed_time": "0:19:25", "remaining_time": "9:06:54"} +{"current_steps": 220, "total_steps": 6125, "loss": 0.12032746076583863, "lr": 4.995188835214465e-06, "epoch": 0.035918367346938776, "percentage": 3.59, "elapsed_time": "0:20:11", "remaining_time": "9:02:07"} +{"current_steps": 230, "total_steps": 6125, "loss": 0.12122929096221924, "lr": 4.994346579032413e-06, "epoch": 0.037551020408163265, "percentage": 3.76, "elapsed_time": "0:20:59", "remaining_time": "8:57:55"} +{"current_steps": 240, "total_steps": 6125, "loss": 0.11416752338409424, "lr": 4.993436505400396e-06, "epoch": 0.03918367346938775, "percentage": 3.92, "elapsed_time": "0:21:46", "remaining_time": "8:53:59"} +{"current_steps": 250, "total_steps": 6125, "loss": 0.11376930475234985, "lr": 4.992458639061916e-06, "epoch": 0.04081632653061224, "percentage": 4.08, "elapsed_time": "0:22:35", "remaining_time": "8:50:47"} +{"current_steps": 260, "total_steps": 6125, "loss": 0.10997685194015502, "lr": 4.991413006603657e-06, "epoch": 0.04244897959183674, "percentage": 4.24, "elapsed_time": "0:23:22", "remaining_time": "8:47:22"} +{"current_steps": 270, "total_steps": 6125, "loss": 0.1000083327293396, "lr": 4.990299636454759e-06, "epoch": 0.044081632653061226, "percentage": 4.41, "elapsed_time": "0:24:10", "remaining_time": "8:44:06"} +{"current_steps": 280, "total_steps": 6125, "loss": 0.1243711233139038, "lr": 4.989118558886043e-06, "epoch": 0.045714285714285714, "percentage": 4.57, "elapsed_time": "0:24:57", "remaining_time": "8:40:56"} +{"current_steps": 290, "total_steps": 6125, "loss": 0.11728224754333497, "lr": 4.9878698060091955e-06, "epoch": 0.0473469387755102, "percentage": 4.73, "elapsed_time": "0:25:44", "remaining_time": "8:37:59"} +{"current_steps": 300, "total_steps": 6125, "loss": 0.09353847503662109, "lr": 4.986553411775886e-06, "epoch": 0.04897959183673469, "percentage": 4.9, "elapsed_time": "0:26:31", "remaining_time": "8:34:58"} +{"current_steps": 310, "total_steps": 6125, "loss": 0.11291944980621338, "lr": 4.985169411976852e-06, "epoch": 0.05061224489795919, "percentage": 5.06, "elapsed_time": "0:27:18", "remaining_time": "8:32:17"} +{"current_steps": 320, "total_steps": 6125, "loss": 0.11149277687072753, "lr": 4.98371784424092e-06, "epoch": 0.052244897959183675, "percentage": 5.22, "elapsed_time": "0:28:06", "remaining_time": "8:29:47"} +{"current_steps": 330, "total_steps": 6125, "loss": 0.11344953775405883, "lr": 4.982198748033986e-06, "epoch": 0.053877551020408164, "percentage": 5.39, "elapsed_time": "0:28:53", "remaining_time": "8:27:25"} +{"current_steps": 340, "total_steps": 6125, "loss": 0.09844974875450134, "lr": 4.980612164657943e-06, "epoch": 0.05551020408163265, "percentage": 5.55, "elapsed_time": "0:29:41", "remaining_time": "8:25:08"} +{"current_steps": 350, "total_steps": 6125, "loss": 0.10898438692092896, "lr": 4.978958137249551e-06, "epoch": 0.05714285714285714, "percentage": 5.71, "elapsed_time": "0:30:29", "remaining_time": "8:23:10"} +{"current_steps": 360, "total_steps": 6125, "loss": 0.09371156096458436, "lr": 4.977236710779277e-06, "epoch": 0.05877551020408163, "percentage": 5.88, "elapsed_time": "0:31:17", "remaining_time": "8:21:05"} +{"current_steps": 370, "total_steps": 6125, "loss": 0.1049533724784851, "lr": 4.9754479320500584e-06, "epoch": 0.060408163265306125, "percentage": 6.04, "elapsed_time": "0:32:05", "remaining_time": "8:19:05"} +{"current_steps": 380, "total_steps": 6125, "loss": 0.10510741472244263, "lr": 4.973591849696041e-06, "epoch": 0.062040816326530614, "percentage": 6.2, "elapsed_time": "0:32:52", "remaining_time": "8:17:05"} +{"current_steps": 390, "total_steps": 6125, "loss": 0.10956356525421143, "lr": 4.971668514181252e-06, "epoch": 0.0636734693877551, "percentage": 6.37, "elapsed_time": "0:33:40", "remaining_time": "8:15:13"} +{"current_steps": 400, "total_steps": 6125, "loss": 0.11344989538192748, "lr": 4.969677977798225e-06, "epoch": 0.0653061224489796, "percentage": 6.53, "elapsed_time": "0:34:28", "remaining_time": "8:13:20"} +{"current_steps": 400, "total_steps": 6125, "eval_loss": 0.11409253627061844, "epoch": 0.0653061224489796, "percentage": 6.53, "elapsed_time": "0:37:13", "remaining_time": "8:52:53"} +{"current_steps": 410, "total_steps": 6125, "loss": 0.09949527978897095, "lr": 4.967620294666589e-06, "epoch": 0.06693877551020408, "percentage": 6.69, "elapsed_time": "0:38:05", "remaining_time": "8:50:55"} +{"current_steps": 420, "total_steps": 6125, "loss": 0.10606559514999389, "lr": 4.965495520731583e-06, "epoch": 0.06857142857142857, "percentage": 6.86, "elapsed_time": "0:38:54", "remaining_time": "8:48:26"} +{"current_steps": 430, "total_steps": 6125, "loss": 0.11677042245864869, "lr": 4.963303713762547e-06, "epoch": 0.07020408163265306, "percentage": 7.02, "elapsed_time": "0:39:41", "remaining_time": "8:45:44"} +{"current_steps": 440, "total_steps": 6125, "loss": 0.1019243597984314, "lr": 4.961044933351342e-06, "epoch": 0.07183673469387755, "percentage": 7.18, "elapsed_time": "0:40:29", "remaining_time": "8:43:10"} +{"current_steps": 450, "total_steps": 6125, "loss": 0.10125980377197266, "lr": 4.958719240910737e-06, "epoch": 0.07346938775510205, "percentage": 7.35, "elapsed_time": "0:41:16", "remaining_time": "8:40:33"} +{"current_steps": 460, "total_steps": 6125, "loss": 0.09888594150543213, "lr": 4.956326699672736e-06, "epoch": 0.07510204081632653, "percentage": 7.51, "elapsed_time": "0:42:03", "remaining_time": "8:38:00"} +{"current_steps": 470, "total_steps": 6125, "loss": 0.10753540992736817, "lr": 4.953867374686856e-06, "epoch": 0.07673469387755102, "percentage": 7.67, "elapsed_time": "0:42:50", "remaining_time": "8:35:27"} +{"current_steps": 480, "total_steps": 6125, "loss": 0.11452326774597169, "lr": 4.951341332818365e-06, "epoch": 0.0783673469387755, "percentage": 7.84, "elapsed_time": "0:43:38", "remaining_time": "8:33:09"} +{"current_steps": 490, "total_steps": 6125, "loss": 0.09555010795593262, "lr": 4.9487486427464594e-06, "epoch": 0.08, "percentage": 8.0, "elapsed_time": "0:44:25", "remaining_time": "8:30:48"} +{"current_steps": 500, "total_steps": 6125, "loss": 0.09771155118942261, "lr": 4.946089374962396e-06, "epoch": 0.08163265306122448, "percentage": 8.16, "elapsed_time": "0:45:12", "remaining_time": "8:28:38"} +{"current_steps": 510, "total_steps": 6125, "loss": 0.10659778118133545, "lr": 4.943363601767582e-06, "epoch": 0.08326530612244898, "percentage": 8.33, "elapsed_time": "0:46:00", "remaining_time": "8:26:27"} +{"current_steps": 520, "total_steps": 6125, "loss": 0.10161921977996827, "lr": 4.940571397271598e-06, "epoch": 0.08489795918367347, "percentage": 8.49, "elapsed_time": "0:46:47", "remaining_time": "8:24:20"} +{"current_steps": 530, "total_steps": 6125, "loss": 0.09155921936035157, "lr": 4.937712837390194e-06, "epoch": 0.08653061224489796, "percentage": 8.65, "elapsed_time": "0:47:35", "remaining_time": "8:22:21"} +{"current_steps": 540, "total_steps": 6125, "loss": 0.11164016723632812, "lr": 4.93478799984322e-06, "epoch": 0.08816326530612245, "percentage": 8.82, "elapsed_time": "0:48:23", "remaining_time": "8:20:25"} +{"current_steps": 550, "total_steps": 6125, "loss": 0.10738091468811035, "lr": 4.9317969641525154e-06, "epoch": 0.08979591836734693, "percentage": 8.98, "elapsed_time": "0:49:10", "remaining_time": "8:18:28"} +{"current_steps": 560, "total_steps": 6125, "loss": 0.09491395950317383, "lr": 4.928739811639741e-06, "epoch": 0.09142857142857143, "percentage": 9.14, "elapsed_time": "0:49:58", "remaining_time": "8:16:35"} +{"current_steps": 570, "total_steps": 6125, "loss": 0.08408294320106506, "lr": 4.9256166254241775e-06, "epoch": 0.09306122448979592, "percentage": 9.31, "elapsed_time": "0:50:45", "remaining_time": "8:14:44"} +{"current_steps": 580, "total_steps": 6125, "loss": 0.10272791385650634, "lr": 4.922427490420457e-06, "epoch": 0.0946938775510204, "percentage": 9.47, "elapsed_time": "0:51:33", "remaining_time": "8:12:57"} +{"current_steps": 590, "total_steps": 6125, "loss": 0.10189608335494996, "lr": 4.919172493336259e-06, "epoch": 0.0963265306122449, "percentage": 9.63, "elapsed_time": "0:52:21", "remaining_time": "8:11:08"} +{"current_steps": 600, "total_steps": 6125, "loss": 0.09506216645240784, "lr": 4.915851722669952e-06, "epoch": 0.09795918367346938, "percentage": 9.8, "elapsed_time": "0:53:08", "remaining_time": "8:09:21"} +{"current_steps": 600, "total_steps": 6125, "eval_loss": 0.09913250058889389, "epoch": 0.09795918367346938, "percentage": 9.8, "elapsed_time": "0:55:46", "remaining_time": "8:33:40"} +{"current_steps": 610, "total_steps": 6125, "loss": 0.09665942788124085, "lr": 4.912465268708186e-06, "epoch": 0.09959183673469388, "percentage": 9.96, "elapsed_time": "0:56:37", "remaining_time": "8:31:59"} +{"current_steps": 620, "total_steps": 6125, "loss": 0.10539349317550659, "lr": 4.90901322352344e-06, "epoch": 0.10122448979591837, "percentage": 10.12, "elapsed_time": "0:57:25", "remaining_time": "8:29:50"} +{"current_steps": 630, "total_steps": 6125, "loss": 0.10907092094421386, "lr": 4.905495680971518e-06, "epoch": 0.10285714285714286, "percentage": 10.29, "elapsed_time": "0:58:12", "remaining_time": "8:27:40"} +{"current_steps": 640, "total_steps": 6125, "loss": 0.09973281621932983, "lr": 4.901912736688995e-06, "epoch": 0.10448979591836735, "percentage": 10.45, "elapsed_time": "0:58:59", "remaining_time": "8:25:37"} +{"current_steps": 650, "total_steps": 6125, "loss": 0.09869267344474793, "lr": 4.898264488090619e-06, "epoch": 0.10612244897959183, "percentage": 10.61, "elapsed_time": "0:59:47", "remaining_time": "8:23:36"} +{"current_steps": 660, "total_steps": 6125, "loss": 0.10499387979507446, "lr": 4.894551034366665e-06, "epoch": 0.10775510204081633, "percentage": 10.78, "elapsed_time": "1:00:35", "remaining_time": "8:21:43"} +{"current_steps": 670, "total_steps": 6125, "loss": 0.10153095722198487, "lr": 4.890772476480229e-06, "epoch": 0.10938775510204081, "percentage": 10.94, "elapsed_time": "1:01:23", "remaining_time": "8:19:54"} +{"current_steps": 680, "total_steps": 6125, "loss": 0.08613221645355225, "lr": 4.886928917164496e-06, "epoch": 0.1110204081632653, "percentage": 11.1, "elapsed_time": "1:02:12", "remaining_time": "8:18:09"} +{"current_steps": 690, "total_steps": 6125, "loss": 0.09545036554336547, "lr": 4.8830204609199335e-06, "epoch": 0.1126530612244898, "percentage": 11.27, "elapsed_time": "1:03:01", "remaining_time": "8:16:22"} +{"current_steps": 700, "total_steps": 6125, "loss": 0.0950542688369751, "lr": 4.8790472140114635e-06, "epoch": 0.11428571428571428, "percentage": 11.43, "elapsed_time": "1:03:49", "remaining_time": "8:14:40"} +{"current_steps": 710, "total_steps": 6125, "loss": 0.10127921104431152, "lr": 4.875009284465561e-06, "epoch": 0.11591836734693878, "percentage": 11.59, "elapsed_time": "1:04:36", "remaining_time": "8:12:48"} +{"current_steps": 720, "total_steps": 6125, "loss": 0.09835844039916992, "lr": 4.8709067820673235e-06, "epoch": 0.11755102040816326, "percentage": 11.76, "elapsed_time": "1:05:24", "remaining_time": "8:10:58"} +{"current_steps": 730, "total_steps": 6125, "loss": 0.08583117127418519, "lr": 4.866739818357489e-06, "epoch": 0.11918367346938775, "percentage": 11.92, "elapsed_time": "1:06:11", "remaining_time": "8:09:12"} +{"current_steps": 740, "total_steps": 6125, "loss": 0.08735222220420838, "lr": 4.862508506629394e-06, "epoch": 0.12081632653061225, "percentage": 12.08, "elapsed_time": "1:06:59", "remaining_time": "8:07:28"} +{"current_steps": 750, "total_steps": 6125, "loss": 0.09324318170547485, "lr": 4.858212961925902e-06, "epoch": 0.12244897959183673, "percentage": 12.24, "elapsed_time": "1:07:46", "remaining_time": "8:05:41"} +{"current_steps": 760, "total_steps": 6125, "loss": 0.09482367634773255, "lr": 4.8538533010362735e-06, "epoch": 0.12408163265306123, "percentage": 12.41, "elapsed_time": "1:08:32", "remaining_time": "8:03:53"} +{"current_steps": 770, "total_steps": 6125, "loss": 0.09595230221748352, "lr": 4.849429642492985e-06, "epoch": 0.12571428571428572, "percentage": 12.57, "elapsed_time": "1:09:20", "remaining_time": "8:02:12"} +{"current_steps": 780, "total_steps": 6125, "loss": 0.08924583196640015, "lr": 4.844942106568517e-06, "epoch": 0.1273469387755102, "percentage": 12.73, "elapsed_time": "1:10:07", "remaining_time": "8:00:31"} +{"current_steps": 790, "total_steps": 6125, "loss": 0.09454164505004883, "lr": 4.84039081527207e-06, "epoch": 0.12897959183673469, "percentage": 12.9, "elapsed_time": "1:10:54", "remaining_time": "7:58:51"} +{"current_steps": 800, "total_steps": 6125, "loss": 0.08659166097640991, "lr": 4.835775892346264e-06, "epoch": 0.1306122448979592, "percentage": 13.06, "elapsed_time": "1:11:43", "remaining_time": "7:57:22"} +{"current_steps": 800, "total_steps": 6125, "eval_loss": 0.09297937154769897, "epoch": 0.1306122448979592, "percentage": 13.06, "elapsed_time": "1:14:20", "remaining_time": "8:14:50"} +{"current_steps": 810, "total_steps": 6125, "loss": 0.09094155430793763, "lr": 4.8310974632637584e-06, "epoch": 0.13224489795918368, "percentage": 13.22, "elapsed_time": "1:15:12", "remaining_time": "8:13:29"} +{"current_steps": 820, "total_steps": 6125, "loss": 0.09823386669158936, "lr": 4.82635565522385e-06, "epoch": 0.13387755102040816, "percentage": 13.39, "elapsed_time": "1:15:59", "remaining_time": "8:11:40"} +{"current_steps": 830, "total_steps": 6125, "loss": 0.09080367088317871, "lr": 4.821550597149012e-06, "epoch": 0.13551020408163264, "percentage": 13.55, "elapsed_time": "1:16:46", "remaining_time": "8:09:49"} +{"current_steps": 840, "total_steps": 6125, "loss": 0.09046205282211303, "lr": 4.816682419681389e-06, "epoch": 0.13714285714285715, "percentage": 13.71, "elapsed_time": "1:17:34", "remaining_time": "8:08:03"} +{"current_steps": 850, "total_steps": 6125, "loss": 0.09502801299095154, "lr": 4.811751255179244e-06, "epoch": 0.13877551020408163, "percentage": 13.88, "elapsed_time": "1:18:21", "remaining_time": "8:06:18"} +{"current_steps": 860, "total_steps": 6125, "loss": 0.09261927604675294, "lr": 4.806757237713361e-06, "epoch": 0.1404081632653061, "percentage": 14.04, "elapsed_time": "1:19:08", "remaining_time": "8:04:30"} +{"current_steps": 870, "total_steps": 6125, "loss": 0.08243635892868043, "lr": 4.8017005030634e-06, "epoch": 0.14204081632653062, "percentage": 14.2, "elapsed_time": "1:19:55", "remaining_time": "8:02:44"} +{"current_steps": 880, "total_steps": 6125, "loss": 0.10203326940536499, "lr": 4.796581188714205e-06, "epoch": 0.1436734693877551, "percentage": 14.37, "elapsed_time": "1:20:43", "remaining_time": "8:01:06"} +{"current_steps": 890, "total_steps": 6125, "loss": 0.08544076681137085, "lr": 4.7913994338520655e-06, "epoch": 0.14530612244897959, "percentage": 14.53, "elapsed_time": "1:21:30", "remaining_time": "7:59:28"} +{"current_steps": 900, "total_steps": 6125, "loss": 0.0944068193435669, "lr": 4.786155379360932e-06, "epoch": 0.1469387755102041, "percentage": 14.69, "elapsed_time": "1:22:18", "remaining_time": "7:57:49"} +{"current_steps": 910, "total_steps": 6125, "loss": 0.0877715528011322, "lr": 4.7808491678185855e-06, "epoch": 0.14857142857142858, "percentage": 14.86, "elapsed_time": "1:23:06", "remaining_time": "7:56:13"} +{"current_steps": 920, "total_steps": 6125, "loss": 0.09196346998214722, "lr": 4.775480943492764e-06, "epoch": 0.15020408163265306, "percentage": 15.02, "elapsed_time": "1:23:53", "remaining_time": "7:54:36"} +{"current_steps": 930, "total_steps": 6125, "loss": 0.09890784621238709, "lr": 4.770050852337236e-06, "epoch": 0.15183673469387754, "percentage": 15.18, "elapsed_time": "1:24:40", "remaining_time": "7:52:59"} +{"current_steps": 940, "total_steps": 6125, "loss": 0.09260086417198181, "lr": 4.764559041987836e-06, "epoch": 0.15346938775510205, "percentage": 15.35, "elapsed_time": "1:25:27", "remaining_time": "7:51:23"} +{"current_steps": 950, "total_steps": 6125, "loss": 0.09412651658058166, "lr": 4.759005661758446e-06, "epoch": 0.15510204081632653, "percentage": 15.51, "elapsed_time": "1:26:14", "remaining_time": "7:49:50"} +{"current_steps": 960, "total_steps": 6125, "loss": 0.09459404945373535, "lr": 4.753390862636939e-06, "epoch": 0.156734693877551, "percentage": 15.67, "elapsed_time": "1:27:02", "remaining_time": "7:48:18"} +{"current_steps": 970, "total_steps": 6125, "loss": 0.09960640072822571, "lr": 4.747714797281076e-06, "epoch": 0.15836734693877552, "percentage": 15.84, "elapsed_time": "1:27:50", "remaining_time": "7:46:47"} +{"current_steps": 980, "total_steps": 6125, "loss": 0.09338371157646179, "lr": 4.741977620014347e-06, "epoch": 0.16, "percentage": 16.0, "elapsed_time": "1:28:37", "remaining_time": "7:45:17"} +{"current_steps": 990, "total_steps": 6125, "loss": 0.08515415191650391, "lr": 4.736179486821788e-06, "epoch": 0.16163265306122448, "percentage": 16.16, "elapsed_time": "1:29:26", "remaining_time": "7:43:57"} +{"current_steps": 1000, "total_steps": 6125, "loss": 0.09090787768363953, "lr": 4.730320555345728e-06, "epoch": 0.16326530612244897, "percentage": 16.33, "elapsed_time": "1:30:16", "remaining_time": "7:42:39"} +{"current_steps": 1000, "total_steps": 6125, "eval_loss": 0.08832322806119919, "epoch": 0.16326530612244897, "percentage": 16.33, "elapsed_time": "1:33:03", "remaining_time": "7:56:54"} +{"current_steps": 1010, "total_steps": 6125, "loss": 0.08005568385124207, "lr": 4.724400984881508e-06, "epoch": 0.16489795918367348, "percentage": 16.49, "elapsed_time": "1:33:57", "remaining_time": "7:55:49"} +{"current_steps": 1020, "total_steps": 6125, "loss": 0.08133251667022705, "lr": 4.7184209363731546e-06, "epoch": 0.16653061224489796, "percentage": 16.65, "elapsed_time": "1:34:46", "remaining_time": "7:54:19"} +{"current_steps": 1030, "total_steps": 6125, "loss": 0.08690541982650757, "lr": 4.7123805724089935e-06, "epoch": 0.16816326530612244, "percentage": 16.82, "elapsed_time": "1:35:33", "remaining_time": "7:52:40"} +{"current_steps": 1040, "total_steps": 6125, "loss": 0.09946454763412475, "lr": 4.706280057217239e-06, "epoch": 0.16979591836734695, "percentage": 16.98, "elapsed_time": "1:36:21", "remaining_time": "7:51:10"} +{"current_steps": 1050, "total_steps": 6125, "loss": 0.08886862993240356, "lr": 4.700119556661521e-06, "epoch": 0.17142857142857143, "percentage": 17.14, "elapsed_time": "1:37:09", "remaining_time": "7:49:35"} +{"current_steps": 1060, "total_steps": 6125, "loss": 0.09117826819419861, "lr": 4.693899238236383e-06, "epoch": 0.1730612244897959, "percentage": 17.31, "elapsed_time": "1:37:56", "remaining_time": "7:48:00"} +{"current_steps": 1070, "total_steps": 6125, "loss": 0.09286551475524903, "lr": 4.687619271062725e-06, "epoch": 0.17469387755102042, "percentage": 17.47, "elapsed_time": "1:38:44", "remaining_time": "7:46:30"} +{"current_steps": 1080, "total_steps": 6125, "loss": 0.09102073907852173, "lr": 4.681279825883201e-06, "epoch": 0.1763265306122449, "percentage": 17.63, "elapsed_time": "1:39:35", "remaining_time": "7:45:14"} +{"current_steps": 1090, "total_steps": 6125, "loss": 0.09481253027915955, "lr": 4.674881075057581e-06, "epoch": 0.17795918367346938, "percentage": 17.8, "elapsed_time": "1:40:25", "remaining_time": "7:43:51"} +{"current_steps": 1100, "total_steps": 6125, "loss": 0.08999797105789184, "lr": 4.668423192558069e-06, "epoch": 0.17959183673469387, "percentage": 17.96, "elapsed_time": "1:41:14", "remaining_time": "7:42:27"} +{"current_steps": 1110, "total_steps": 6125, "loss": 0.07345216870307922, "lr": 4.661906353964563e-06, "epoch": 0.18122448979591838, "percentage": 18.12, "elapsed_time": "1:42:10", "remaining_time": "7:41:37"} +{"current_steps": 1120, "total_steps": 6125, "loss": 0.08096543550491334, "lr": 4.65533073645989e-06, "epoch": 0.18285714285714286, "percentage": 18.29, "elapsed_time": "1:43:07", "remaining_time": "7:40:51"} +{"current_steps": 1130, "total_steps": 6125, "loss": 0.08680164813995361, "lr": 4.6486965188249825e-06, "epoch": 0.18448979591836734, "percentage": 18.45, "elapsed_time": "1:44:04", "remaining_time": "7:40:04"} +{"current_steps": 1140, "total_steps": 6125, "loss": 0.08470291495323182, "lr": 4.642003881434024e-06, "epoch": 0.18612244897959185, "percentage": 18.61, "elapsed_time": "1:45:02", "remaining_time": "7:39:18"} +{"current_steps": 1150, "total_steps": 6125, "loss": 0.10852066278457642, "lr": 4.635253006249536e-06, "epoch": 0.18775510204081633, "percentage": 18.78, "elapsed_time": "1:45:59", "remaining_time": "7:38:30"} +{"current_steps": 1160, "total_steps": 6125, "loss": 0.1022921085357666, "lr": 4.628444076817444e-06, "epoch": 0.1893877551020408, "percentage": 18.94, "elapsed_time": "1:46:56", "remaining_time": "7:37:45"} +{"current_steps": 1170, "total_steps": 6125, "loss": 0.09645735621452331, "lr": 4.621577278262072e-06, "epoch": 0.1910204081632653, "percentage": 19.1, "elapsed_time": "1:47:51", "remaining_time": "7:36:46"} +{"current_steps": 1180, "total_steps": 6125, "loss": 0.08636590242385864, "lr": 4.614652797281118e-06, "epoch": 0.1926530612244898, "percentage": 19.27, "elapsed_time": "1:48:40", "remaining_time": "7:35:26"} +{"current_steps": 1190, "total_steps": 6125, "loss": 0.09293715357780456, "lr": 4.6076708221405806e-06, "epoch": 0.19428571428571428, "percentage": 19.43, "elapsed_time": "1:49:28", "remaining_time": "7:34:01"} +{"current_steps": 1200, "total_steps": 6125, "loss": 0.09799563884735107, "lr": 4.600631542669633e-06, "epoch": 0.19591836734693877, "percentage": 19.59, "elapsed_time": "1:50:16", "remaining_time": "7:32:35"} +{"current_steps": 1200, "total_steps": 6125, "eval_loss": 0.08563050627708435, "epoch": 0.19591836734693877, "percentage": 19.59, "elapsed_time": "1:53:07", "remaining_time": "7:44:17"} +{"current_steps": 1210, "total_steps": 6125, "loss": 0.08345703482627868, "lr": 4.593535150255463e-06, "epoch": 0.19755102040816327, "percentage": 19.76, "elapsed_time": "1:53:58", "remaining_time": "7:42:56"} +{"current_steps": 1220, "total_steps": 6125, "loss": 0.08114275336265564, "lr": 4.586381837838076e-06, "epoch": 0.19918367346938776, "percentage": 19.92, "elapsed_time": "1:54:45", "remaining_time": "7:41:21"} +{"current_steps": 1230, "total_steps": 6125, "loss": 0.0780484676361084, "lr": 4.579171799905044e-06, "epoch": 0.20081632653061224, "percentage": 20.08, "elapsed_time": "1:55:31", "remaining_time": "7:39:45"} +{"current_steps": 1240, "total_steps": 6125, "loss": 0.07816853523254394, "lr": 4.571905232486214e-06, "epoch": 0.20244897959183675, "percentage": 20.24, "elapsed_time": "1:56:19", "remaining_time": "7:38:17"} +{"current_steps": 1250, "total_steps": 6125, "loss": 0.08001847267150879, "lr": 4.56458233314839e-06, "epoch": 0.20408163265306123, "percentage": 20.41, "elapsed_time": "1:57:10", "remaining_time": "7:36:57"} +{"current_steps": 1260, "total_steps": 6125, "loss": 0.08720387816429138, "lr": 4.557203300989948e-06, "epoch": 0.2057142857142857, "percentage": 20.57, "elapsed_time": "1:57:56", "remaining_time": "7:35:24"} +{"current_steps": 1270, "total_steps": 6125, "loss": 0.08812607526779175, "lr": 4.549768336635432e-06, "epoch": 0.2073469387755102, "percentage": 20.73, "elapsed_time": "1:58:43", "remaining_time": "7:33:53"} +{"current_steps": 1280, "total_steps": 6125, "loss": 0.0880801498889923, "lr": 4.542277642230096e-06, "epoch": 0.2089795918367347, "percentage": 20.9, "elapsed_time": "1:59:31", "remaining_time": "7:32:24"} +{"current_steps": 1290, "total_steps": 6125, "loss": 0.09044411182403564, "lr": 4.53473142143441e-06, "epoch": 0.21061224489795918, "percentage": 21.06, "elapsed_time": "2:00:19", "remaining_time": "7:30:57"} +{"current_steps": 1300, "total_steps": 6125, "loss": 0.06792758107185363, "lr": 4.527129879418519e-06, "epoch": 0.21224489795918366, "percentage": 21.22, "elapsed_time": "2:01:07", "remaining_time": "7:29:35"} +{"current_steps": 1310, "total_steps": 6125, "loss": 0.07861824035644531, "lr": 4.519473222856667e-06, "epoch": 0.21387755102040817, "percentage": 21.39, "elapsed_time": "2:01:55", "remaining_time": "7:28:09"} +{"current_steps": 1320, "total_steps": 6125, "loss": 0.08266589641571045, "lr": 4.51176165992158e-06, "epoch": 0.21551020408163266, "percentage": 21.55, "elapsed_time": "2:02:44", "remaining_time": "7:26:49"} +{"current_steps": 1330, "total_steps": 6125, "loss": 0.08272708654403686, "lr": 4.503995400278801e-06, "epoch": 0.21714285714285714, "percentage": 21.71, "elapsed_time": "2:03:34", "remaining_time": "7:25:31"} +{"current_steps": 1340, "total_steps": 6125, "loss": 0.08644292950630188, "lr": 4.4961746550809945e-06, "epoch": 0.21877551020408162, "percentage": 21.88, "elapsed_time": "2:04:23", "remaining_time": "7:24:10"} +{"current_steps": 1350, "total_steps": 6125, "loss": 0.08812489509582519, "lr": 4.488299636962203e-06, "epoch": 0.22040816326530613, "percentage": 22.04, "elapsed_time": "2:05:10", "remaining_time": "7:22:45"} +{"current_steps": 1360, "total_steps": 6125, "loss": 0.08752686977386474, "lr": 4.4803705600320655e-06, "epoch": 0.2220408163265306, "percentage": 22.2, "elapsed_time": "2:06:00", "remaining_time": "7:21:27"} +{"current_steps": 1370, "total_steps": 6125, "loss": 0.08365487456321716, "lr": 4.472387639869998e-06, "epoch": 0.2236734693877551, "percentage": 22.37, "elapsed_time": "2:06:50", "remaining_time": "7:20:14"} +{"current_steps": 1380, "total_steps": 6125, "loss": 0.08101181387901306, "lr": 4.464351093519332e-06, "epoch": 0.2253061224489796, "percentage": 22.53, "elapsed_time": "2:07:39", "remaining_time": "7:18:55"} +{"current_steps": 1390, "total_steps": 6125, "loss": 0.08873040080070496, "lr": 4.45626113948141e-06, "epoch": 0.22693877551020408, "percentage": 22.69, "elapsed_time": "2:08:27", "remaining_time": "7:17:35"} +{"current_steps": 1400, "total_steps": 6125, "loss": 0.08360225558280945, "lr": 4.44811799770965e-06, "epoch": 0.22857142857142856, "percentage": 22.86, "elapsed_time": "2:09:17", "remaining_time": "7:16:20"} +{"current_steps": 1400, "total_steps": 6125, "eval_loss": 0.0838993564248085, "epoch": 0.22857142857142856, "percentage": 22.86, "elapsed_time": "2:11:56", "remaining_time": "7:25:18"} +{"current_steps": 1410, "total_steps": 6125, "loss": 0.07942168116569519, "lr": 4.4399218896035615e-06, "epoch": 0.23020408163265307, "percentage": 23.02, "elapsed_time": "2:12:48", "remaining_time": "7:24:05"} +{"current_steps": 1420, "total_steps": 6125, "loss": 0.07959137558937072, "lr": 4.431673038002726e-06, "epoch": 0.23183673469387756, "percentage": 23.18, "elapsed_time": "2:13:35", "remaining_time": "7:22:38"} +{"current_steps": 1430, "total_steps": 6125, "loss": 0.08047830462455749, "lr": 4.423371667180743e-06, "epoch": 0.23346938775510204, "percentage": 23.35, "elapsed_time": "2:14:23", "remaining_time": "7:21:15"} +{"current_steps": 1440, "total_steps": 6125, "loss": 0.07832802534103393, "lr": 4.415018002839127e-06, "epoch": 0.23510204081632652, "percentage": 23.51, "elapsed_time": "2:15:11", "remaining_time": "7:19:49"} +{"current_steps": 1450, "total_steps": 6125, "loss": 0.07132846713066102, "lr": 4.40661227210117e-06, "epoch": 0.23673469387755103, "percentage": 23.67, "elapsed_time": "2:15:57", "remaining_time": "7:18:22"} +{"current_steps": 1460, "total_steps": 6125, "loss": 0.09222683906555176, "lr": 4.398154703505776e-06, "epoch": 0.2383673469387755, "percentage": 23.84, "elapsed_time": "2:16:44", "remaining_time": "7:16:56"} +{"current_steps": 1470, "total_steps": 6125, "loss": 0.07937920093536377, "lr": 4.389645527001234e-06, "epoch": 0.24, "percentage": 24.0, "elapsed_time": "2:17:31", "remaining_time": "7:15:29"} +{"current_steps": 1480, "total_steps": 6125, "loss": 0.07650941014289855, "lr": 4.381084973938976e-06, "epoch": 0.2416326530612245, "percentage": 24.16, "elapsed_time": "2:18:18", "remaining_time": "7:14:06"} +{"current_steps": 1490, "total_steps": 6125, "loss": 0.07730157375335693, "lr": 4.372473277067285e-06, "epoch": 0.24326530612244898, "percentage": 24.33, "elapsed_time": "2:19:07", "remaining_time": "7:12:45"} +{"current_steps": 1500, "total_steps": 6125, "loss": 0.07192737460136414, "lr": 4.363810670524962e-06, "epoch": 0.24489795918367346, "percentage": 24.49, "elapsed_time": "2:19:55", "remaining_time": "7:11:26"} +{"current_steps": 1510, "total_steps": 6125, "loss": 0.09040454030036926, "lr": 4.355097389834966e-06, "epoch": 0.24653061224489795, "percentage": 24.65, "elapsed_time": "2:20:44", "remaining_time": "7:10:07"} +{"current_steps": 1520, "total_steps": 6125, "loss": 0.06948431730270385, "lr": 4.346333671898007e-06, "epoch": 0.24816326530612245, "percentage": 24.82, "elapsed_time": "2:21:31", "remaining_time": "7:08:47"} +{"current_steps": 1530, "total_steps": 6125, "loss": 0.07317689657211304, "lr": 4.337519754986108e-06, "epoch": 0.24979591836734694, "percentage": 24.98, "elapsed_time": "2:22:19", "remaining_time": "7:07:26"} +{"current_steps": 1540, "total_steps": 6125, "loss": 0.09126524925231934, "lr": 4.328655878736123e-06, "epoch": 0.25142857142857145, "percentage": 25.14, "elapsed_time": "2:23:06", "remaining_time": "7:06:05"} +{"current_steps": 1550, "total_steps": 6125, "loss": 0.09065211415290833, "lr": 4.319742284143226e-06, "epoch": 0.2530612244897959, "percentage": 25.31, "elapsed_time": "2:23:54", "remaining_time": "7:04:45"} +{"current_steps": 1560, "total_steps": 6125, "loss": 0.0718291699886322, "lr": 4.310779213554353e-06, "epoch": 0.2546938775510204, "percentage": 25.47, "elapsed_time": "2:24:41", "remaining_time": "7:03:24"} +{"current_steps": 1570, "total_steps": 6125, "loss": 0.07913097143173217, "lr": 4.301766910661619e-06, "epoch": 0.2563265306122449, "percentage": 25.63, "elapsed_time": "2:25:29", "remaining_time": "7:02:05"} +{"current_steps": 1580, "total_steps": 6125, "loss": 0.06828063130378723, "lr": 4.292705620495689e-06, "epoch": 0.25795918367346937, "percentage": 25.8, "elapsed_time": "2:26:16", "remaining_time": "7:00:46"} +{"current_steps": 1590, "total_steps": 6125, "loss": 0.07042163014411926, "lr": 4.283595589419115e-06, "epoch": 0.2595918367346939, "percentage": 25.96, "elapsed_time": "2:27:03", "remaining_time": "6:59:27"} +{"current_steps": 1600, "total_steps": 6125, "loss": 0.07090482115745544, "lr": 4.274437065119643e-06, "epoch": 0.2612244897959184, "percentage": 26.12, "elapsed_time": "2:27:51", "remaining_time": "6:58:09"} +{"current_steps": 1600, "total_steps": 6125, "eval_loss": 0.08112655580043793, "epoch": 0.2612244897959184, "percentage": 26.12, "elapsed_time": "2:30:29", "remaining_time": "7:05:36"} +{"current_steps": 1610, "total_steps": 6125, "loss": 0.08803416490554809, "lr": 4.265230296603471e-06, "epoch": 0.26285714285714284, "percentage": 26.29, "elapsed_time": "2:31:21", "remaining_time": "7:04:26"} +{"current_steps": 1620, "total_steps": 6125, "loss": 0.0734250545501709, "lr": 4.255975534188488e-06, "epoch": 0.26448979591836735, "percentage": 26.45, "elapsed_time": "2:32:11", "remaining_time": "7:03:12"} +{"current_steps": 1630, "total_steps": 6125, "loss": 0.07703657150268554, "lr": 4.246673029497458e-06, "epoch": 0.26612244897959186, "percentage": 26.61, "elapsed_time": "2:33:00", "remaining_time": "7:01:57"} +{"current_steps": 1640, "total_steps": 6125, "loss": 0.08410758376121522, "lr": 4.237323035451188e-06, "epoch": 0.2677551020408163, "percentage": 26.78, "elapsed_time": "2:33:47", "remaining_time": "7:00:34"} +{"current_steps": 1650, "total_steps": 6125, "loss": 0.0808764636516571, "lr": 4.2279258062616455e-06, "epoch": 0.2693877551020408, "percentage": 26.94, "elapsed_time": "2:34:34", "remaining_time": "6:59:13"} +{"current_steps": 1660, "total_steps": 6125, "loss": 0.07636536359786987, "lr": 4.21848159742505e-06, "epoch": 0.2710204081632653, "percentage": 27.1, "elapsed_time": "2:35:28", "remaining_time": "6:58:12"} +{"current_steps": 1670, "total_steps": 6125, "loss": 0.07789689898490906, "lr": 4.208990665714927e-06, "epoch": 0.2726530612244898, "percentage": 27.27, "elapsed_time": "2:36:25", "remaining_time": "6:57:16"} +{"current_steps": 1680, "total_steps": 6125, "loss": 0.07442367672920228, "lr": 4.19945326917512e-06, "epoch": 0.2742857142857143, "percentage": 27.43, "elapsed_time": "2:37:21", "remaining_time": "6:56:19"} +{"current_steps": 1690, "total_steps": 6125, "loss": 0.07646024227142334, "lr": 4.1898696671127855e-06, "epoch": 0.27591836734693875, "percentage": 27.59, "elapsed_time": "2:38:17", "remaining_time": "6:55:23"} +{"current_steps": 1700, "total_steps": 6125, "loss": 0.07484800815582275, "lr": 4.180240120091333e-06, "epoch": 0.27755102040816326, "percentage": 27.76, "elapsed_time": "2:39:13", "remaining_time": "6:54:26"} +{"current_steps": 1710, "total_steps": 6125, "loss": 0.07452926635742188, "lr": 4.170564889923348e-06, "epoch": 0.27918367346938777, "percentage": 27.92, "elapsed_time": "2:40:09", "remaining_time": "6:53:29"} +{"current_steps": 1720, "total_steps": 6125, "loss": 0.07579211592674255, "lr": 4.1608442396634675e-06, "epoch": 0.2808163265306122, "percentage": 28.08, "elapsed_time": "2:41:05", "remaining_time": "6:52:34"} +{"current_steps": 1730, "total_steps": 6125, "loss": 0.08277950882911682, "lr": 4.151078433601234e-06, "epoch": 0.28244897959183674, "percentage": 28.24, "elapsed_time": "2:41:55", "remaining_time": "6:51:20"} +{"current_steps": 1740, "total_steps": 6125, "loss": 0.07354526519775391, "lr": 4.141267737253905e-06, "epoch": 0.28408163265306124, "percentage": 28.41, "elapsed_time": "2:42:43", "remaining_time": "6:50:03"} +{"current_steps": 1750, "total_steps": 6125, "loss": 0.07377716898918152, "lr": 4.1314124173592356e-06, "epoch": 0.2857142857142857, "percentage": 28.57, "elapsed_time": "2:43:30", "remaining_time": "6:48:46"} +{"current_steps": 1760, "total_steps": 6125, "loss": 0.0802634596824646, "lr": 4.121512741868229e-06, "epoch": 0.2873469387755102, "percentage": 28.73, "elapsed_time": "2:44:18", "remaining_time": "6:47:28"} +{"current_steps": 1770, "total_steps": 6125, "loss": 0.08141970038414001, "lr": 4.111568979937845e-06, "epoch": 0.2889795918367347, "percentage": 28.9, "elapsed_time": "2:45:04", "remaining_time": "6:46:10"} +{"current_steps": 1780, "total_steps": 6125, "loss": 0.07370007634162903, "lr": 4.10158140192369e-06, "epoch": 0.29061224489795917, "percentage": 29.06, "elapsed_time": "2:45:52", "remaining_time": "6:44:53"} +{"current_steps": 1790, "total_steps": 6125, "loss": 0.0754891276359558, "lr": 4.091550279372659e-06, "epoch": 0.2922448979591837, "percentage": 29.22, "elapsed_time": "2:46:39", "remaining_time": "6:43:36"} +{"current_steps": 1800, "total_steps": 6125, "loss": 0.07567381858825684, "lr": 4.081475885015558e-06, "epoch": 0.2938775510204082, "percentage": 29.39, "elapsed_time": "2:47:27", "remaining_time": "6:42:22"} +{"current_steps": 1800, "total_steps": 6125, "eval_loss": 0.07648575305938721, "epoch": 0.2938775510204082, "percentage": 29.39, "elapsed_time": "2:50:24", "remaining_time": "6:49:27"} +{"current_steps": 1810, "total_steps": 6125, "loss": 0.06706757545471191, "lr": 4.071358492759686e-06, "epoch": 0.29551020408163264, "percentage": 29.55, "elapsed_time": "2:51:17", "remaining_time": "6:48:21"} +{"current_steps": 1820, "total_steps": 6125, "loss": 0.07407833337783813, "lr": 4.061198377681391e-06, "epoch": 0.29714285714285715, "percentage": 29.71, "elapsed_time": "2:52:04", "remaining_time": "6:47:01"} +{"current_steps": 1830, "total_steps": 6125, "loss": 0.07593105435371399, "lr": 4.050995816018585e-06, "epoch": 0.2987755102040816, "percentage": 29.88, "elapsed_time": "2:52:51", "remaining_time": "6:45:41"} +{"current_steps": 1840, "total_steps": 6125, "loss": 0.08375502228736878, "lr": 4.040751085163239e-06, "epoch": 0.3004081632653061, "percentage": 30.04, "elapsed_time": "2:53:38", "remaining_time": "6:44:22"} +{"current_steps": 1850, "total_steps": 6125, "loss": 0.0692280650138855, "lr": 4.0304644636538435e-06, "epoch": 0.3020408163265306, "percentage": 30.2, "elapsed_time": "2:54:32", "remaining_time": "6:43:20"} +{"current_steps": 1860, "total_steps": 6125, "loss": 0.07824563384056091, "lr": 4.020136231167824e-06, "epoch": 0.3036734693877551, "percentage": 30.37, "elapsed_time": "2:55:24", "remaining_time": "6:42:13"} +{"current_steps": 1870, "total_steps": 6125, "loss": 0.07255304455757142, "lr": 4.0097666685139516e-06, "epoch": 0.3053061224489796, "percentage": 30.53, "elapsed_time": "2:56:11", "remaining_time": "6:40:53"} +{"current_steps": 1880, "total_steps": 6125, "loss": 0.06304961442947388, "lr": 3.999356057624696e-06, "epoch": 0.3069387755102041, "percentage": 30.69, "elapsed_time": "2:56:57", "remaining_time": "6:39:34"} +{"current_steps": 1890, "total_steps": 6125, "loss": 0.07315694689750671, "lr": 3.988904681548568e-06, "epoch": 0.30857142857142855, "percentage": 30.86, "elapsed_time": "2:57:45", "remaining_time": "6:38:17"} +{"current_steps": 1900, "total_steps": 6125, "loss": 0.07607222199440003, "lr": 3.978412824442421e-06, "epoch": 0.31020408163265306, "percentage": 31.02, "elapsed_time": "2:58:32", "remaining_time": "6:37:01"} +{"current_steps": 1910, "total_steps": 6125, "loss": 0.06320942640304565, "lr": 3.9678807715637245e-06, "epoch": 0.31183673469387757, "percentage": 31.18, "elapsed_time": "2:59:22", "remaining_time": "6:35:49"} +{"current_steps": 1920, "total_steps": 6125, "loss": 0.07607839703559875, "lr": 3.95730880926281e-06, "epoch": 0.313469387755102, "percentage": 31.35, "elapsed_time": "3:00:11", "remaining_time": "6:34:37"} +{"current_steps": 1930, "total_steps": 6125, "loss": 0.07458690404891968, "lr": 3.946697224975083e-06, "epoch": 0.31510204081632653, "percentage": 31.51, "elapsed_time": "3:00:59", "remaining_time": "6:33:24"} +{"current_steps": 1940, "total_steps": 6125, "loss": 0.07876120805740357, "lr": 3.936046307213215e-06, "epoch": 0.31673469387755104, "percentage": 31.67, "elapsed_time": "3:01:49", "remaining_time": "6:32:14"} +{"current_steps": 1950, "total_steps": 6125, "loss": 0.07082821726799012, "lr": 3.925356345559288e-06, "epoch": 0.3183673469387755, "percentage": 31.84, "elapsed_time": "3:02:40", "remaining_time": "6:31:06"} +{"current_steps": 1960, "total_steps": 6125, "loss": 0.06648402214050293, "lr": 3.914627630656931e-06, "epoch": 0.32, "percentage": 32.0, "elapsed_time": "3:03:31", "remaining_time": "6:29:58"} +{"current_steps": 1970, "total_steps": 6125, "loss": 0.07000230550765991, "lr": 3.903860454203413e-06, "epoch": 0.3216326530612245, "percentage": 32.16, "elapsed_time": "3:04:21", "remaining_time": "6:28:50"} +{"current_steps": 1980, "total_steps": 6125, "loss": 0.06887786984443664, "lr": 3.8930551089417145e-06, "epoch": 0.32326530612244897, "percentage": 32.33, "elapsed_time": "3:05:12", "remaining_time": "6:27:42"} +{"current_steps": 1990, "total_steps": 6125, "loss": 0.08991954326629639, "lr": 3.882211888652566e-06, "epoch": 0.3248979591836735, "percentage": 32.49, "elapsed_time": "3:06:00", "remaining_time": "6:26:29"} +{"current_steps": 2000, "total_steps": 6125, "loss": 0.07011507153511047, "lr": 3.871331088146463e-06, "epoch": 0.32653061224489793, "percentage": 32.65, "elapsed_time": "3:06:47", "remaining_time": "6:25:16"} +{"current_steps": 2000, "total_steps": 6125, "eval_loss": 0.07913381606340408, "epoch": 0.32653061224489793, "percentage": 32.65, "elapsed_time": "3:09:29", "remaining_time": "6:30:49"} +{"current_steps": 2010, "total_steps": 6125, "loss": 0.08266008496284485, "lr": 3.860413003255648e-06, "epoch": 0.32816326530612244, "percentage": 32.82, "elapsed_time": "3:10:20", "remaining_time": "6:29:39"} +{"current_steps": 2020, "total_steps": 6125, "loss": 0.07043538689613342, "lr": 3.8494579308260695e-06, "epoch": 0.32979591836734695, "percentage": 32.98, "elapsed_time": "3:11:07", "remaining_time": "6:28:23"} +{"current_steps": 2030, "total_steps": 6125, "loss": 0.06780247688293457, "lr": 3.838466168709312e-06, "epoch": 0.3314285714285714, "percentage": 33.14, "elapsed_time": "3:11:54", "remaining_time": "6:27:06"} +{"current_steps": 2040, "total_steps": 6125, "loss": 0.06872215867042542, "lr": 3.827438015754491e-06, "epoch": 0.3330612244897959, "percentage": 33.31, "elapsed_time": "3:12:40", "remaining_time": "6:25:49"} +{"current_steps": 2050, "total_steps": 6125, "loss": 0.06861716508865356, "lr": 3.81637377180014e-06, "epoch": 0.3346938775510204, "percentage": 33.47, "elapsed_time": "3:13:26", "remaining_time": "6:24:32"} +{"current_steps": 2060, "total_steps": 6125, "loss": 0.07096996307373046, "lr": 3.8052737376660446e-06, "epoch": 0.3363265306122449, "percentage": 33.63, "elapsed_time": "3:14:14", "remaining_time": "6:23:17"} +{"current_steps": 2070, "total_steps": 6125, "loss": 0.07092551589012146, "lr": 3.7941382151450756e-06, "epoch": 0.3379591836734694, "percentage": 33.8, "elapsed_time": "3:15:01", "remaining_time": "6:22:02"} +{"current_steps": 2080, "total_steps": 6125, "loss": 0.06780472993850709, "lr": 3.782967506994975e-06, "epoch": 0.3395918367346939, "percentage": 33.96, "elapsed_time": "3:15:48", "remaining_time": "6:20:46"} +{"current_steps": 2090, "total_steps": 6125, "loss": 0.07661795020103454, "lr": 3.7717619169301302e-06, "epoch": 0.34122448979591835, "percentage": 34.12, "elapsed_time": "3:16:35", "remaining_time": "6:19:32"} +{"current_steps": 2100, "total_steps": 6125, "loss": 0.07939797639846802, "lr": 3.7605217496133127e-06, "epoch": 0.34285714285714286, "percentage": 34.29, "elapsed_time": "3:17:22", "remaining_time": "6:18:18"} +{"current_steps": 2110, "total_steps": 6125, "loss": 0.07869908213615417, "lr": 3.749247310647398e-06, "epoch": 0.34448979591836737, "percentage": 34.45, "elapsed_time": "3:18:09", "remaining_time": "6:17:04"} +{"current_steps": 2120, "total_steps": 6125, "loss": 0.07526676058769226, "lr": 3.737938906567053e-06, "epoch": 0.3461224489795918, "percentage": 34.61, "elapsed_time": "3:18:57", "remaining_time": "6:15:51"} +{"current_steps": 2130, "total_steps": 6125, "loss": 0.07366414070129394, "lr": 3.7265968448304057e-06, "epoch": 0.34775510204081633, "percentage": 34.78, "elapsed_time": "3:19:44", "remaining_time": "6:14:38"} +{"current_steps": 2140, "total_steps": 6125, "loss": 0.06439908146858216, "lr": 3.7152214338106827e-06, "epoch": 0.34938775510204084, "percentage": 34.94, "elapsed_time": "3:20:31", "remaining_time": "6:13:24"} +{"current_steps": 2150, "total_steps": 6125, "loss": 0.08098108768463134, "lr": 3.703812982787828e-06, "epoch": 0.3510204081632653, "percentage": 35.1, "elapsed_time": "3:21:17", "remaining_time": "6:12:09"} +{"current_steps": 2160, "total_steps": 6125, "loss": 0.08134381771087647, "lr": 3.6923718019400906e-06, "epoch": 0.3526530612244898, "percentage": 35.27, "elapsed_time": "3:22:05", "remaining_time": "6:10:57"} +{"current_steps": 2170, "total_steps": 6125, "loss": 0.07070590257644653, "lr": 3.6808982023355934e-06, "epoch": 0.35428571428571426, "percentage": 35.43, "elapsed_time": "3:22:52", "remaining_time": "6:09:45"} +{"current_steps": 2180, "total_steps": 6125, "loss": 0.06200396418571472, "lr": 3.6693924959238774e-06, "epoch": 0.35591836734693877, "percentage": 35.59, "elapsed_time": "3:23:40", "remaining_time": "6:08:33"} +{"current_steps": 2190, "total_steps": 6125, "loss": 0.06886230111122131, "lr": 3.657854995527416e-06, "epoch": 0.3575510204081633, "percentage": 35.76, "elapsed_time": "3:24:27", "remaining_time": "6:07:22"} +{"current_steps": 2200, "total_steps": 6125, "loss": 0.07170915603637695, "lr": 3.646286014833114e-06, "epoch": 0.35918367346938773, "percentage": 35.92, "elapsed_time": "3:25:14", "remaining_time": "6:06:10"} +{"current_steps": 2200, "total_steps": 6125, "eval_loss": 0.06810569018125534, "epoch": 0.35918367346938773, "percentage": 35.92, "elapsed_time": "3:27:52", "remaining_time": "6:10:51"} +{"current_steps": 2210, "total_steps": 6125, "loss": 0.07438129186630249, "lr": 3.634685868383776e-06, "epoch": 0.36081632653061224, "percentage": 36.08, "elapsed_time": "3:28:43", "remaining_time": "6:09:45"} +{"current_steps": 2220, "total_steps": 6125, "loss": 0.07625046968460084, "lr": 3.6230548715695555e-06, "epoch": 0.36244897959183675, "percentage": 36.24, "elapsed_time": "3:29:33", "remaining_time": "6:08:36"} +{"current_steps": 2230, "total_steps": 6125, "loss": 0.06719872951507569, "lr": 3.611393340619381e-06, "epoch": 0.3640816326530612, "percentage": 36.41, "elapsed_time": "3:30:30", "remaining_time": "6:07:40"} +{"current_steps": 2240, "total_steps": 6125, "loss": 0.07131683230400085, "lr": 3.5997015925923563e-06, "epoch": 0.3657142857142857, "percentage": 36.57, "elapsed_time": "3:31:27", "remaining_time": "6:06:44"} +{"current_steps": 2250, "total_steps": 6125, "loss": 0.06948224902153015, "lr": 3.587979945369141e-06, "epoch": 0.3673469387755102, "percentage": 36.73, "elapsed_time": "3:32:23", "remaining_time": "6:05:47"} +{"current_steps": 2260, "total_steps": 6125, "loss": 0.07263018488883972, "lr": 3.5762287176433097e-06, "epoch": 0.3689795918367347, "percentage": 36.9, "elapsed_time": "3:33:20", "remaining_time": "6:04:51"} +{"current_steps": 2270, "total_steps": 6125, "loss": 0.06066258549690247, "lr": 3.564448228912682e-06, "epoch": 0.3706122448979592, "percentage": 37.06, "elapsed_time": "3:34:17", "remaining_time": "6:03:55"} +{"current_steps": 2280, "total_steps": 6125, "loss": 0.07041062712669373, "lr": 3.5526387994706428e-06, "epoch": 0.3722448979591837, "percentage": 37.22, "elapsed_time": "3:35:14", "remaining_time": "6:02:59"} +{"current_steps": 2290, "total_steps": 6125, "loss": 0.07272242307662964, "lr": 3.5408007503974283e-06, "epoch": 0.37387755102040815, "percentage": 37.39, "elapsed_time": "3:36:04", "remaining_time": "6:01:51"} +{"current_steps": 2300, "total_steps": 6125, "loss": 0.07268807291984558, "lr": 3.5289344035513996e-06, "epoch": 0.37551020408163266, "percentage": 37.55, "elapsed_time": "3:36:52", "remaining_time": "6:00:39"} +{"current_steps": 2310, "total_steps": 6125, "loss": 0.05777261257171631, "lr": 3.5170400815602894e-06, "epoch": 0.37714285714285717, "percentage": 37.71, "elapsed_time": "3:37:39", "remaining_time": "5:59:27"} +{"current_steps": 2320, "total_steps": 6125, "loss": 0.06174240708351135, "lr": 3.5051181078124336e-06, "epoch": 0.3787755102040816, "percentage": 37.88, "elapsed_time": "3:38:26", "remaining_time": "5:58:16"} +{"current_steps": 2330, "total_steps": 6125, "loss": 0.06616324782371522, "lr": 3.4931688064479745e-06, "epoch": 0.38040816326530613, "percentage": 38.04, "elapsed_time": "3:39:14", "remaining_time": "5:57:05"} +{"current_steps": 2340, "total_steps": 6125, "loss": 0.06692227125167846, "lr": 3.4811925023500524e-06, "epoch": 0.3820408163265306, "percentage": 38.2, "elapsed_time": "3:40:01", "remaining_time": "5:55:53"} +{"current_steps": 2350, "total_steps": 6125, "loss": 0.0634126603603363, "lr": 3.4691895211359693e-06, "epoch": 0.3836734693877551, "percentage": 38.37, "elapsed_time": "3:40:49", "remaining_time": "5:54:43"} +{"current_steps": 2360, "total_steps": 6125, "loss": 0.06884999275207519, "lr": 3.457160189148337e-06, "epoch": 0.3853061224489796, "percentage": 38.53, "elapsed_time": "3:41:35", "remaining_time": "5:53:31"} +{"current_steps": 2370, "total_steps": 6125, "loss": 0.05986430644989014, "lr": 3.445104833446205e-06, "epoch": 0.38693877551020406, "percentage": 38.69, "elapsed_time": "3:42:22", "remaining_time": "5:52:19"} +{"current_steps": 2380, "total_steps": 6125, "loss": 0.06759394407272339, "lr": 3.433023781796167e-06, "epoch": 0.38857142857142857, "percentage": 38.86, "elapsed_time": "3:43:09", "remaining_time": "5:51:08"} +{"current_steps": 2390, "total_steps": 6125, "loss": 0.06813116669654846, "lr": 3.4209173626634494e-06, "epoch": 0.3902040816326531, "percentage": 39.02, "elapsed_time": "3:43:57", "remaining_time": "5:49:59"} +{"current_steps": 2400, "total_steps": 6125, "loss": 0.06371633410453796, "lr": 3.4087859052029837e-06, "epoch": 0.39183673469387753, "percentage": 39.18, "elapsed_time": "3:44:44", "remaining_time": "5:48:48"} +{"current_steps": 2400, "total_steps": 6125, "eval_loss": 0.07035119831562042, "epoch": 0.39183673469387753, "percentage": 39.18, "elapsed_time": "3:47:21", "remaining_time": "5:52:52"} +{"current_steps": 2410, "total_steps": 6125, "loss": 0.06804299950599671, "lr": 3.3966297392504535e-06, "epoch": 0.39346938775510204, "percentage": 39.35, "elapsed_time": "3:48:12", "remaining_time": "5:51:46"} +{"current_steps": 2420, "total_steps": 6125, "loss": 0.07678125500679016, "lr": 3.3844491953133272e-06, "epoch": 0.39510204081632655, "percentage": 39.51, "elapsed_time": "3:48:58", "remaining_time": "5:50:33"} +{"current_steps": 2430, "total_steps": 6125, "loss": 0.06537411212921143, "lr": 3.372244604561875e-06, "epoch": 0.396734693877551, "percentage": 39.67, "elapsed_time": "3:49:45", "remaining_time": "5:49:21"} +{"current_steps": 2440, "total_steps": 6125, "loss": 0.05702770948410034, "lr": 3.360016298820163e-06, "epoch": 0.3983673469387755, "percentage": 39.84, "elapsed_time": "3:50:31", "remaining_time": "5:48:09"} +{"current_steps": 2450, "total_steps": 6125, "loss": 0.06334798336029053, "lr": 3.3477646105570293e-06, "epoch": 0.4, "percentage": 40.0, "elapsed_time": "3:51:19", "remaining_time": "5:46:59"} +{"current_steps": 2460, "total_steps": 6125, "loss": 0.051438117027282716, "lr": 3.3354898728770503e-06, "epoch": 0.4016326530612245, "percentage": 40.16, "elapsed_time": "3:52:07", "remaining_time": "5:45:49"} +{"current_steps": 2470, "total_steps": 6125, "loss": 0.06427289247512817, "lr": 3.3231924195114774e-06, "epoch": 0.403265306122449, "percentage": 40.33, "elapsed_time": "3:52:54", "remaining_time": "5:44:38"} +{"current_steps": 2480, "total_steps": 6125, "loss": 0.06652372479438781, "lr": 3.3108725848091685e-06, "epoch": 0.4048979591836735, "percentage": 40.49, "elapsed_time": "3:53:41", "remaining_time": "5:43:28"} +{"current_steps": 2490, "total_steps": 6125, "loss": 0.07374808788299561, "lr": 3.2985307037274937e-06, "epoch": 0.40653061224489795, "percentage": 40.65, "elapsed_time": "3:54:28", "remaining_time": "5:42:18"} +{"current_steps": 2500, "total_steps": 6125, "loss": 0.06349012851715088, "lr": 3.286167111823234e-06, "epoch": 0.40816326530612246, "percentage": 40.82, "elapsed_time": "3:55:16", "remaining_time": "5:41:08"} +{"current_steps": 2510, "total_steps": 6125, "loss": 0.06953533291816712, "lr": 3.2737821452434502e-06, "epoch": 0.4097959183673469, "percentage": 40.98, "elapsed_time": "3:56:03", "remaining_time": "5:39:58"} +{"current_steps": 2520, "total_steps": 6125, "loss": 0.05826004147529602, "lr": 3.26137614071635e-06, "epoch": 0.4114285714285714, "percentage": 41.14, "elapsed_time": "3:56:50", "remaining_time": "5:38:48"} +{"current_steps": 2530, "total_steps": 6125, "loss": 0.05837242007255554, "lr": 3.2489494355421293e-06, "epoch": 0.41306122448979593, "percentage": 41.31, "elapsed_time": "3:57:37", "remaining_time": "5:37:39"} +{"current_steps": 2540, "total_steps": 6125, "loss": 0.06359312534332276, "lr": 3.2365023675838027e-06, "epoch": 0.4146938775510204, "percentage": 41.47, "elapsed_time": "3:58:24", "remaining_time": "5:36:30"} +{"current_steps": 2550, "total_steps": 6125, "loss": 0.07361164093017578, "lr": 3.22403527525802e-06, "epoch": 0.4163265306122449, "percentage": 41.63, "elapsed_time": "3:59:12", "remaining_time": "5:35:21"} +{"current_steps": 2560, "total_steps": 6125, "loss": 0.05989238619804382, "lr": 3.211548497525861e-06, "epoch": 0.4179591836734694, "percentage": 41.8, "elapsed_time": "3:59:58", "remaining_time": "5:34:11"} +{"current_steps": 2570, "total_steps": 6125, "loss": 0.07593320608139038, "lr": 3.1990423738836206e-06, "epoch": 0.41959183673469386, "percentage": 41.96, "elapsed_time": "4:00:46", "remaining_time": "5:33:03"} +{"current_steps": 2580, "total_steps": 6125, "loss": 0.06850495338439941, "lr": 3.1865172443535797e-06, "epoch": 0.42122448979591837, "percentage": 42.12, "elapsed_time": "4:01:33", "remaining_time": "5:31:53"} +{"current_steps": 2590, "total_steps": 6125, "loss": 0.056969082355499266, "lr": 3.173973449474762e-06, "epoch": 0.4228571428571429, "percentage": 42.29, "elapsed_time": "4:02:20", "remaining_time": "5:30:45"} +{"current_steps": 2600, "total_steps": 6125, "loss": 0.06189748048782349, "lr": 3.1614113302936716e-06, "epoch": 0.42448979591836733, "percentage": 42.45, "elapsed_time": "4:03:07", "remaining_time": "5:29:37"} +{"current_steps": 2600, "total_steps": 6125, "eval_loss": 0.0618240088224411, "epoch": 0.42448979591836733, "percentage": 42.45, "elapsed_time": "4:05:44", "remaining_time": "5:33:10"} +{"current_steps": 2610, "total_steps": 6125, "loss": 0.061647194623947146, "lr": 3.1488312283550216e-06, "epoch": 0.42612244897959184, "percentage": 42.61, "elapsed_time": "4:06:35", "remaining_time": "5:32:05"} +{"current_steps": 2620, "total_steps": 6125, "loss": 0.05916886925697327, "lr": 3.136233485692451e-06, "epoch": 0.42775510204081635, "percentage": 42.78, "elapsed_time": "4:07:21", "remaining_time": "5:30:55"} +{"current_steps": 2630, "total_steps": 6125, "loss": 0.059158307313919065, "lr": 3.12361844481922e-06, "epoch": 0.4293877551020408, "percentage": 42.94, "elapsed_time": "4:08:08", "remaining_time": "5:29:44"} +{"current_steps": 2640, "total_steps": 6125, "loss": 0.05914759635925293, "lr": 3.110986448718902e-06, "epoch": 0.4310204081632653, "percentage": 43.1, "elapsed_time": "4:08:54", "remaining_time": "5:28:34"} +{"current_steps": 2650, "total_steps": 6125, "loss": 0.059078723192214966, "lr": 3.0983378408360597e-06, "epoch": 0.4326530612244898, "percentage": 43.27, "elapsed_time": "4:09:41", "remaining_time": "5:27:25"} +{"current_steps": 2660, "total_steps": 6125, "loss": 0.05707222819328308, "lr": 3.085672965066899e-06, "epoch": 0.4342857142857143, "percentage": 43.43, "elapsed_time": "4:10:28", "remaining_time": "5:26:17"} +{"current_steps": 2670, "total_steps": 6125, "loss": 0.05138048529624939, "lr": 3.0729921657499294e-06, "epoch": 0.4359183673469388, "percentage": 43.59, "elapsed_time": "4:11:17", "remaining_time": "5:25:10"} +{"current_steps": 2680, "total_steps": 6125, "loss": 0.050283265113830564, "lr": 3.0602957876565943e-06, "epoch": 0.43755102040816324, "percentage": 43.76, "elapsed_time": "4:12:06", "remaining_time": "5:24:04"} +{"current_steps": 2690, "total_steps": 6125, "loss": 0.06342648863792419, "lr": 3.0475841759819023e-06, "epoch": 0.43918367346938775, "percentage": 43.92, "elapsed_time": "4:12:54", "remaining_time": "5:22:57"} +{"current_steps": 2700, "total_steps": 6125, "loss": 0.06127093434333801, "lr": 3.034857676335038e-06, "epoch": 0.44081632653061226, "percentage": 44.08, "elapsed_time": "4:13:44", "remaining_time": "5:21:53"} +{"current_steps": 2710, "total_steps": 6125, "loss": 0.05412284731864929, "lr": 3.0221166347299673e-06, "epoch": 0.4424489795918367, "percentage": 44.24, "elapsed_time": "4:14:34", "remaining_time": "5:20:48"} +{"current_steps": 2720, "total_steps": 6125, "loss": 0.05579071640968323, "lr": 3.009361397576031e-06, "epoch": 0.4440816326530612, "percentage": 44.41, "elapsed_time": "4:15:24", "remaining_time": "5:19:44"} +{"current_steps": 2730, "total_steps": 6125, "loss": 0.06472888588905334, "lr": 2.9965923116685224e-06, "epoch": 0.44571428571428573, "percentage": 44.57, "elapsed_time": "4:16:14", "remaining_time": "5:18:39"} +{"current_steps": 2740, "total_steps": 6125, "loss": 0.06263615489006043, "lr": 2.9838097241792646e-06, "epoch": 0.4473469387755102, "percentage": 44.73, "elapsed_time": "4:17:04", "remaining_time": "5:17:35"} +{"current_steps": 2750, "total_steps": 6125, "loss": 0.0624904453754425, "lr": 2.9710139826471646e-06, "epoch": 0.4489795918367347, "percentage": 44.9, "elapsed_time": "4:17:53", "remaining_time": "5:16:30"} +{"current_steps": 2760, "total_steps": 6125, "loss": 0.05488716959953308, "lr": 2.9582054349687706e-06, "epoch": 0.4506122448979592, "percentage": 45.06, "elapsed_time": "4:18:44", "remaining_time": "5:15:26"} +{"current_steps": 2770, "total_steps": 6125, "loss": 0.06263139843940735, "lr": 2.9453844293888067e-06, "epoch": 0.45224489795918366, "percentage": 45.22, "elapsed_time": "4:19:33", "remaining_time": "5:14:23"} +{"current_steps": 2780, "total_steps": 6125, "loss": 0.06324111819267272, "lr": 2.932551314490712e-06, "epoch": 0.45387755102040817, "percentage": 45.39, "elapsed_time": "4:20:24", "remaining_time": "5:13:19"} +{"current_steps": 2790, "total_steps": 6125, "loss": 0.06384775042533875, "lr": 2.9197064391871565e-06, "epoch": 0.4555102040816327, "percentage": 45.55, "elapsed_time": "4:21:11", "remaining_time": "5:12:12"} +{"current_steps": 2800, "total_steps": 6125, "loss": 0.05722553730010986, "lr": 2.9068501527105595e-06, "epoch": 0.45714285714285713, "percentage": 45.71, "elapsed_time": "4:22:06", "remaining_time": "5:11:15"} +{"current_steps": 2800, "total_steps": 6125, "eval_loss": 0.05622806400060654, "epoch": 0.45714285714285713, "percentage": 45.71, "elapsed_time": "4:24:57", "remaining_time": "5:14:37"} +{"current_steps": 2810, "total_steps": 6125, "loss": 0.060957270860672, "lr": 2.893982804603592e-06, "epoch": 0.45877551020408164, "percentage": 45.88, "elapsed_time": "4:26:01", "remaining_time": "5:13:49"} +{"current_steps": 2820, "total_steps": 6125, "loss": 0.05667365789413452, "lr": 2.8811047447096714e-06, "epoch": 0.46040816326530615, "percentage": 46.04, "elapsed_time": "4:27:00", "remaining_time": "5:12:56"} +{"current_steps": 2830, "total_steps": 6125, "loss": 0.06144403219223023, "lr": 2.8682163231634557e-06, "epoch": 0.4620408163265306, "percentage": 46.2, "elapsed_time": "4:27:54", "remaining_time": "5:11:55"} +{"current_steps": 2840, "total_steps": 6125, "loss": 0.04665087461471558, "lr": 2.8553178903813193e-06, "epoch": 0.4636734693877551, "percentage": 46.37, "elapsed_time": "4:28:45", "remaining_time": "5:10:51"} +{"current_steps": 2850, "total_steps": 6125, "loss": 0.06363347172737122, "lr": 2.842409797051825e-06, "epoch": 0.46530612244897956, "percentage": 46.53, "elapsed_time": "4:29:35", "remaining_time": "5:09:47"} +{"current_steps": 2860, "total_steps": 6125, "loss": 0.05332474112510681, "lr": 2.829492394126192e-06, "epoch": 0.4669387755102041, "percentage": 46.69, "elapsed_time": "4:30:26", "remaining_time": "5:08:43"} +{"current_steps": 2870, "total_steps": 6125, "loss": 0.07094461917877197, "lr": 2.8165660328087546e-06, "epoch": 0.4685714285714286, "percentage": 46.86, "elapsed_time": "4:31:17", "remaining_time": "5:07:41"} +{"current_steps": 2880, "total_steps": 6125, "loss": 0.058167701959609984, "lr": 2.8036310645474096e-06, "epoch": 0.47020408163265304, "percentage": 47.02, "elapsed_time": "4:32:07", "remaining_time": "5:06:37"} +{"current_steps": 2890, "total_steps": 6125, "loss": 0.0486674964427948, "lr": 2.7906878410240655e-06, "epoch": 0.47183673469387755, "percentage": 47.18, "elapsed_time": "4:32:58", "remaining_time": "5:05:33"} +{"current_steps": 2900, "total_steps": 6125, "loss": 0.062025409936904904, "lr": 2.777736714145077e-06, "epoch": 0.47346938775510206, "percentage": 47.35, "elapsed_time": "4:33:48", "remaining_time": "5:04:30"} +{"current_steps": 2910, "total_steps": 6125, "loss": 0.061118441820144656, "lr": 2.7647780360316787e-06, "epoch": 0.4751020408163265, "percentage": 47.51, "elapsed_time": "4:34:36", "remaining_time": "5:03:23"} +{"current_steps": 2920, "total_steps": 6125, "loss": 0.061808472871780394, "lr": 2.7518121590104145e-06, "epoch": 0.476734693877551, "percentage": 47.67, "elapsed_time": "4:35:23", "remaining_time": "5:02:16"} +{"current_steps": 2930, "total_steps": 6125, "loss": 0.05849882960319519, "lr": 2.7388394356035524e-06, "epoch": 0.47836734693877553, "percentage": 47.84, "elapsed_time": "4:36:10", "remaining_time": "5:01:09"} +{"current_steps": 2940, "total_steps": 6125, "loss": 0.06612529158592224, "lr": 2.7258602185195043e-06, "epoch": 0.48, "percentage": 48.0, "elapsed_time": "4:36:56", "remaining_time": "5:00:01"} +{"current_steps": 2950, "total_steps": 6125, "loss": 0.058327901363372806, "lr": 2.7128748606432353e-06, "epoch": 0.4816326530612245, "percentage": 48.16, "elapsed_time": "4:37:44", "remaining_time": "4:58:55"} +{"current_steps": 2960, "total_steps": 6125, "loss": 0.055737996101379396, "lr": 2.699883715026669e-06, "epoch": 0.483265306122449, "percentage": 48.33, "elapsed_time": "4:38:31", "remaining_time": "4:57:48"} +{"current_steps": 2970, "total_steps": 6125, "loss": 0.051851028203964235, "lr": 2.686887134879089e-06, "epoch": 0.48489795918367345, "percentage": 48.49, "elapsed_time": "4:39:18", "remaining_time": "4:56:42"} +{"current_steps": 2980, "total_steps": 6125, "loss": 0.06357839703559875, "lr": 2.6738854735575358e-06, "epoch": 0.48653061224489796, "percentage": 48.65, "elapsed_time": "4:40:05", "remaining_time": "4:55:35"} +{"current_steps": 2990, "total_steps": 6125, "loss": 0.06160399913787842, "lr": 2.660879084557198e-06, "epoch": 0.4881632653061225, "percentage": 48.82, "elapsed_time": "4:40:52", "remaining_time": "4:54:29"} +{"current_steps": 3000, "total_steps": 6125, "loss": 0.060502588748931885, "lr": 2.6478683215018025e-06, "epoch": 0.4897959183673469, "percentage": 48.98, "elapsed_time": "4:41:39", "remaining_time": "4:53:23"} +{"current_steps": 3000, "total_steps": 6125, "eval_loss": 0.0561000294983387, "epoch": 0.4897959183673469, "percentage": 48.98, "elapsed_time": "4:44:15", "remaining_time": "4:56:06"} +{"current_steps": 3010, "total_steps": 6125, "loss": 0.060459482669830325, "lr": 2.6348535381340025e-06, "epoch": 0.49142857142857144, "percentage": 49.14, "elapsed_time": "4:45:06", "remaining_time": "4:55:02"} +{"current_steps": 3020, "total_steps": 6125, "loss": 0.0632048487663269, "lr": 2.6218350883057546e-06, "epoch": 0.4930612244897959, "percentage": 49.31, "elapsed_time": "4:45:52", "remaining_time": "4:53:55"} +{"current_steps": 3030, "total_steps": 6125, "loss": 0.04655081033706665, "lr": 2.608813325968702e-06, "epoch": 0.4946938775510204, "percentage": 49.47, "elapsed_time": "4:46:39", "remaining_time": "4:52:48"} +{"current_steps": 3040, "total_steps": 6125, "loss": 0.05672039389610291, "lr": 2.59578860516455e-06, "epoch": 0.4963265306122449, "percentage": 49.63, "elapsed_time": "4:47:26", "remaining_time": "4:51:41"} +{"current_steps": 3050, "total_steps": 6125, "loss": 0.047120341658592226, "lr": 2.5827612800154382e-06, "epoch": 0.49795918367346936, "percentage": 49.8, "elapsed_time": "4:48:13", "remaining_time": "4:50:35"} +{"current_steps": 3060, "total_steps": 6125, "loss": 0.06063341498374939, "lr": 2.5697317047143177e-06, "epoch": 0.4995918367346939, "percentage": 49.96, "elapsed_time": "4:49:00", "remaining_time": "4:49:28"} +{"current_steps": 3070, "total_steps": 6125, "loss": 0.05240688323974609, "lr": 2.556700233515316e-06, "epoch": 0.5012244897959184, "percentage": 50.12, "elapsed_time": "4:49:47", "remaining_time": "4:48:22"} +{"current_steps": 3080, "total_steps": 6125, "loss": 0.053180736303329465, "lr": 2.5436672207241064e-06, "epoch": 0.5028571428571429, "percentage": 50.29, "elapsed_time": "4:50:34", "remaining_time": "4:47:16"} +{"current_steps": 3090, "total_steps": 6125, "loss": 0.06092001795768738, "lr": 2.5306330206882764e-06, "epoch": 0.5044897959183674, "percentage": 50.45, "elapsed_time": "4:51:22", "remaining_time": "4:46:11"} +{"current_steps": 3100, "total_steps": 6125, "loss": 0.06613186597824097, "lr": 2.5175979877876935e-06, "epoch": 0.5061224489795918, "percentage": 50.61, "elapsed_time": "4:52:09", "remaining_time": "4:45:05"} +{"current_steps": 3110, "total_steps": 6125, "loss": 0.06115126609802246, "lr": 2.5045624764248687e-06, "epoch": 0.5077551020408163, "percentage": 50.78, "elapsed_time": "4:52:56", "remaining_time": "4:43:59"} +{"current_steps": 3120, "total_steps": 6125, "loss": 0.044120225310325625, "lr": 2.491526841015321e-06, "epoch": 0.5093877551020408, "percentage": 50.94, "elapsed_time": "4:53:43", "remaining_time": "4:42:53"} +{"current_steps": 3130, "total_steps": 6125, "loss": 0.07098687291145325, "lr": 2.478491435977945e-06, "epoch": 0.5110204081632653, "percentage": 51.1, "elapsed_time": "4:54:30", "remaining_time": "4:41:47"} +{"current_steps": 3140, "total_steps": 6125, "loss": 0.05738852024078369, "lr": 2.4654566157253683e-06, "epoch": 0.5126530612244898, "percentage": 51.27, "elapsed_time": "4:55:17", "remaining_time": "4:40:42"} +{"current_steps": 3150, "total_steps": 6125, "loss": 0.053479206562042234, "lr": 2.452422734654321e-06, "epoch": 0.5142857142857142, "percentage": 51.43, "elapsed_time": "4:56:04", "remaining_time": "4:39:37"} +{"current_steps": 3160, "total_steps": 6125, "loss": 0.056794995069503786, "lr": 2.4393901471359977e-06, "epoch": 0.5159183673469387, "percentage": 51.59, "elapsed_time": "4:56:50", "remaining_time": "4:38:31"} +{"current_steps": 3170, "total_steps": 6125, "loss": 0.05802353620529175, "lr": 2.4263592075064238e-06, "epoch": 0.5175510204081633, "percentage": 51.76, "elapsed_time": "4:57:37", "remaining_time": "4:37:26"} +{"current_steps": 3180, "total_steps": 6125, "loss": 0.05282890200614929, "lr": 2.41333027005682e-06, "epoch": 0.5191836734693878, "percentage": 51.92, "elapsed_time": "4:58:24", "remaining_time": "4:36:21"} +{"current_steps": 3190, "total_steps": 6125, "loss": 0.05190218687057495, "lr": 2.400303689023972e-06, "epoch": 0.5208163265306123, "percentage": 52.08, "elapsed_time": "4:59:11", "remaining_time": "4:35:16"} +{"current_steps": 3200, "total_steps": 6125, "loss": 0.06482383608818054, "lr": 2.3872798185805986e-06, "epoch": 0.5224489795918368, "percentage": 52.24, "elapsed_time": "4:59:57", "remaining_time": "4:34:10"} +{"current_steps": 3200, "total_steps": 6125, "eval_loss": 0.058392543345689774, "epoch": 0.5224489795918368, "percentage": 52.24, "elapsed_time": "5:02:34", "remaining_time": "4:36:34"} +{"current_steps": 3210, "total_steps": 6125, "loss": 0.06062687039375305, "lr": 2.37425901282572e-06, "epoch": 0.5240816326530612, "percentage": 52.41, "elapsed_time": "5:03:25", "remaining_time": "4:35:32"} +{"current_steps": 3220, "total_steps": 6125, "loss": 0.06494970321655273, "lr": 2.3612416257750335e-06, "epoch": 0.5257142857142857, "percentage": 52.57, "elapsed_time": "5:04:12", "remaining_time": "4:34:26"} +{"current_steps": 3230, "total_steps": 6125, "loss": 0.05140092372894287, "lr": 2.348228011351287e-06, "epoch": 0.5273469387755102, "percentage": 52.73, "elapsed_time": "5:04:59", "remaining_time": "4:33:21"} +{"current_steps": 3240, "total_steps": 6125, "loss": 0.041680914163589475, "lr": 2.335218523374656e-06, "epoch": 0.5289795918367347, "percentage": 52.9, "elapsed_time": "5:05:46", "remaining_time": "4:32:16"} +{"current_steps": 3250, "total_steps": 6125, "loss": 0.04369679093360901, "lr": 2.3222135155531248e-06, "epoch": 0.5306122448979592, "percentage": 53.06, "elapsed_time": "5:06:33", "remaining_time": "4:31:11"} +{"current_steps": 3260, "total_steps": 6125, "loss": 0.050543594360351565, "lr": 2.309213341472868e-06, "epoch": 0.5322448979591837, "percentage": 53.22, "elapsed_time": "5:07:21", "remaining_time": "4:30:06"} +{"current_steps": 3270, "total_steps": 6125, "loss": 0.05258342027664185, "lr": 2.29621835458864e-06, "epoch": 0.5338775510204081, "percentage": 53.39, "elapsed_time": "5:08:07", "remaining_time": "4:29:01"} +{"current_steps": 3280, "total_steps": 6125, "loss": 0.055678969621658324, "lr": 2.283228908214163e-06, "epoch": 0.5355102040816326, "percentage": 53.55, "elapsed_time": "5:08:54", "remaining_time": "4:27:56"} +{"current_steps": 3290, "total_steps": 6125, "loss": 0.048620349168777464, "lr": 2.2702453555125182e-06, "epoch": 0.5371428571428571, "percentage": 53.71, "elapsed_time": "5:09:42", "remaining_time": "4:26:52"} +{"current_steps": 3300, "total_steps": 6125, "loss": 0.046547812223434445, "lr": 2.2572680494865527e-06, "epoch": 0.5387755102040817, "percentage": 53.88, "elapsed_time": "5:10:29", "remaining_time": "4:25:47"} +{"current_steps": 3310, "total_steps": 6125, "loss": 0.04967870116233826, "lr": 2.2442973429692705e-06, "epoch": 0.5404081632653062, "percentage": 54.04, "elapsed_time": "5:11:16", "remaining_time": "4:24:43"} +{"current_steps": 3320, "total_steps": 6125, "loss": 0.0508593499660492, "lr": 2.231333588614248e-06, "epoch": 0.5420408163265306, "percentage": 54.2, "elapsed_time": "5:12:02", "remaining_time": "4:23:38"} +{"current_steps": 3330, "total_steps": 6125, "loss": 0.04854269027709961, "lr": 2.2183771388860427e-06, "epoch": 0.5436734693877551, "percentage": 54.37, "elapsed_time": "5:12:49", "remaining_time": "4:22:34"} +{"current_steps": 3340, "total_steps": 6125, "loss": 0.05934192538261414, "lr": 2.2054283460506096e-06, "epoch": 0.5453061224489796, "percentage": 54.53, "elapsed_time": "5:13:36", "remaining_time": "4:21:30"} +{"current_steps": 3350, "total_steps": 6125, "loss": 0.05332346558570862, "lr": 2.1924875621657256e-06, "epoch": 0.5469387755102041, "percentage": 54.69, "elapsed_time": "5:14:23", "remaining_time": "4:20:26"} +{"current_steps": 3360, "total_steps": 6125, "loss": 0.04938814043998718, "lr": 2.179555139071415e-06, "epoch": 0.5485714285714286, "percentage": 54.86, "elapsed_time": "5:15:10", "remaining_time": "4:19:21"} +{"current_steps": 3370, "total_steps": 6125, "loss": 0.05734975934028626, "lr": 2.166631428380387e-06, "epoch": 0.5502040816326531, "percentage": 55.02, "elapsed_time": "5:15:57", "remaining_time": "4:18:18"} +{"current_steps": 3380, "total_steps": 6125, "loss": 0.047475004196166994, "lr": 2.1537167814684728e-06, "epoch": 0.5518367346938775, "percentage": 55.18, "elapsed_time": "5:16:44", "remaining_time": "4:17:14"} +{"current_steps": 3390, "total_steps": 6125, "loss": 0.062284767627716064, "lr": 2.140811549465074e-06, "epoch": 0.553469387755102, "percentage": 55.35, "elapsed_time": "5:17:31", "remaining_time": "4:16:10"} +{"current_steps": 3400, "total_steps": 6125, "loss": 0.06585791110992431, "lr": 2.127916083243615e-06, "epoch": 0.5551020408163265, "percentage": 55.51, "elapsed_time": "5:18:17", "remaining_time": "4:15:06"} +{"current_steps": 3400, "total_steps": 6125, "eval_loss": 0.05396538972854614, "epoch": 0.5551020408163265, "percentage": 55.51, "elapsed_time": "5:20:53", "remaining_time": "4:17:11"} +{"current_steps": 3410, "total_steps": 6125, "loss": 0.04635375440120697, "lr": 2.1150307334120033e-06, "epoch": 0.556734693877551, "percentage": 55.67, "elapsed_time": "5:21:45", "remaining_time": "4:16:10"} +{"current_steps": 3420, "total_steps": 6125, "loss": 0.05940789580345154, "lr": 2.1021558503030973e-06, "epoch": 0.5583673469387755, "percentage": 55.84, "elapsed_time": "5:22:32", "remaining_time": "4:15:06"} +{"current_steps": 3430, "total_steps": 6125, "loss": 0.047165653109550475, "lr": 2.0892917839651818e-06, "epoch": 0.56, "percentage": 56.0, "elapsed_time": "5:23:19", "remaining_time": "4:14:02"} +{"current_steps": 3440, "total_steps": 6125, "loss": 0.04550817906856537, "lr": 2.076438884152449e-06, "epoch": 0.5616326530612245, "percentage": 56.16, "elapsed_time": "5:24:05", "remaining_time": "4:12:57"} +{"current_steps": 3450, "total_steps": 6125, "loss": 0.045033201575279236, "lr": 2.0635975003154934e-06, "epoch": 0.563265306122449, "percentage": 56.33, "elapsed_time": "5:24:52", "remaining_time": "4:11:53"} +{"current_steps": 3460, "total_steps": 6125, "loss": 0.04502847194671631, "lr": 2.0507679815918043e-06, "epoch": 0.5648979591836735, "percentage": 56.49, "elapsed_time": "5:25:39", "remaining_time": "4:10:49"} +{"current_steps": 3470, "total_steps": 6125, "loss": 0.04697479605674744, "lr": 2.0379506767962788e-06, "epoch": 0.566530612244898, "percentage": 56.65, "elapsed_time": "5:26:26", "remaining_time": "4:09:46"} +{"current_steps": 3480, "total_steps": 6125, "loss": 0.052119261026382445, "lr": 2.0251459344117345e-06, "epoch": 0.5681632653061225, "percentage": 56.82, "elapsed_time": "5:27:13", "remaining_time": "4:08:42"} +{"current_steps": 3490, "total_steps": 6125, "loss": 0.05213663578033447, "lr": 2.0123541025794397e-06, "epoch": 0.5697959183673469, "percentage": 56.98, "elapsed_time": "5:28:00", "remaining_time": "4:07:38"} +{"current_steps": 3500, "total_steps": 6125, "loss": 0.05319364666938782, "lr": 1.9995755290896414e-06, "epoch": 0.5714285714285714, "percentage": 57.14, "elapsed_time": "5:28:46", "remaining_time": "4:06:34"} +{"current_steps": 3510, "total_steps": 6125, "loss": 0.046560952067375184, "lr": 1.986810561372113e-06, "epoch": 0.5730612244897959, "percentage": 57.31, "elapsed_time": "5:29:32", "remaining_time": "4:05:30"} +{"current_steps": 3520, "total_steps": 6125, "loss": 0.05037369728088379, "lr": 1.9740595464867092e-06, "epoch": 0.5746938775510204, "percentage": 57.47, "elapsed_time": "5:30:18", "remaining_time": "4:04:27"} +{"current_steps": 3530, "total_steps": 6125, "loss": 0.0543134331703186, "lr": 1.961322831113927e-06, "epoch": 0.5763265306122449, "percentage": 57.63, "elapsed_time": "5:31:05", "remaining_time": "4:03:23"} +{"current_steps": 3540, "total_steps": 6125, "loss": 0.052860957384109494, "lr": 1.9486007615454837e-06, "epoch": 0.5779591836734694, "percentage": 57.8, "elapsed_time": "5:31:52", "remaining_time": "4:02:20"} +{"current_steps": 3550, "total_steps": 6125, "loss": 0.05510561466217041, "lr": 1.9358936836748972e-06, "epoch": 0.5795918367346938, "percentage": 57.96, "elapsed_time": "5:32:39", "remaining_time": "4:01:17"} +{"current_steps": 3560, "total_steps": 6125, "loss": 0.06125714182853699, "lr": 1.9232019429880874e-06, "epoch": 0.5812244897959183, "percentage": 58.12, "elapsed_time": "5:33:25", "remaining_time": "4:00:14"} +{"current_steps": 3570, "total_steps": 6125, "loss": 0.051503753662109374, "lr": 1.910525884553976e-06, "epoch": 0.5828571428571429, "percentage": 58.29, "elapsed_time": "5:34:13", "remaining_time": "3:59:11"} +{"current_steps": 3580, "total_steps": 6125, "loss": 0.05377998948097229, "lr": 1.8978658530151113e-06, "epoch": 0.5844897959183674, "percentage": 58.45, "elapsed_time": "5:35:00", "remaining_time": "3:58:08"} +{"current_steps": 3590, "total_steps": 6125, "loss": 0.05506981611251831, "lr": 1.8852221925782935e-06, "epoch": 0.5861224489795919, "percentage": 58.61, "elapsed_time": "5:35:46", "remaining_time": "3:57:06"} +{"current_steps": 3600, "total_steps": 6125, "loss": 0.056353437900543216, "lr": 1.872595247005218e-06, "epoch": 0.5877551020408164, "percentage": 58.78, "elapsed_time": "5:36:33", "remaining_time": "3:56:03"} +{"current_steps": 3600, "total_steps": 6125, "eval_loss": 0.04787343740463257, "epoch": 0.5877551020408164, "percentage": 58.78, "elapsed_time": "5:39:11", "remaining_time": "3:57:54"} +{"current_steps": 3610, "total_steps": 6125, "loss": 0.05329249501228332, "lr": 1.8599853596031289e-06, "epoch": 0.5893877551020408, "percentage": 58.94, "elapsed_time": "5:40:02", "remaining_time": "3:56:53"} +{"current_steps": 3620, "total_steps": 6125, "loss": 0.04905630648136139, "lr": 1.8473928732154847e-06, "epoch": 0.5910204081632653, "percentage": 59.1, "elapsed_time": "5:40:49", "remaining_time": "3:55:51"} +{"current_steps": 3630, "total_steps": 6125, "loss": 0.04863021373748779, "lr": 1.8348181302126367e-06, "epoch": 0.5926530612244898, "percentage": 59.27, "elapsed_time": "5:41:37", "remaining_time": "3:54:48"} +{"current_steps": 3640, "total_steps": 6125, "loss": 0.04666702449321747, "lr": 1.8222614724825222e-06, "epoch": 0.5942857142857143, "percentage": 59.43, "elapsed_time": "5:42:25", "remaining_time": "3:53:45"} +{"current_steps": 3650, "total_steps": 6125, "loss": 0.05730679631233215, "lr": 1.8097232414213662e-06, "epoch": 0.5959183673469388, "percentage": 59.59, "elapsed_time": "5:43:12", "remaining_time": "3:52:43"} +{"current_steps": 3660, "total_steps": 6125, "loss": 0.04433938264846802, "lr": 1.7972037779244e-06, "epoch": 0.5975510204081632, "percentage": 59.76, "elapsed_time": "5:43:59", "remaining_time": "3:51:40"} +{"current_steps": 3670, "total_steps": 6125, "loss": 0.04502029418945312, "lr": 1.784703422376598e-06, "epoch": 0.5991836734693877, "percentage": 59.92, "elapsed_time": "5:44:47", "remaining_time": "3:50:38"} +{"current_steps": 3680, "total_steps": 6125, "loss": 0.04091633260250092, "lr": 1.7722225146434146e-06, "epoch": 0.6008163265306122, "percentage": 60.08, "elapsed_time": "5:45:34", "remaining_time": "3:49:36"} +{"current_steps": 3690, "total_steps": 6125, "loss": 0.0417140930891037, "lr": 1.7597613940615492e-06, "epoch": 0.6024489795918367, "percentage": 60.24, "elapsed_time": "5:46:21", "remaining_time": "3:48:33"} +{"current_steps": 3700, "total_steps": 6125, "loss": 0.045112240314483645, "lr": 1.7473203994297197e-06, "epoch": 0.6040816326530613, "percentage": 60.41, "elapsed_time": "5:47:09", "remaining_time": "3:47:31"} +{"current_steps": 3710, "total_steps": 6125, "loss": 0.04915813505649567, "lr": 1.7348998689994492e-06, "epoch": 0.6057142857142858, "percentage": 60.57, "elapsed_time": "5:47:56", "remaining_time": "3:46:29"} +{"current_steps": 3720, "total_steps": 6125, "loss": 0.05195118188858032, "lr": 1.7225001404658714e-06, "epoch": 0.6073469387755102, "percentage": 60.73, "elapsed_time": "5:48:44", "remaining_time": "3:45:27"} +{"current_steps": 3730, "total_steps": 6125, "loss": 0.047248026728630065, "lr": 1.7101215509585482e-06, "epoch": 0.6089795918367347, "percentage": 60.9, "elapsed_time": "5:49:32", "remaining_time": "3:44:26"} +{"current_steps": 3740, "total_steps": 6125, "loss": 0.05229860544204712, "lr": 1.6977644370323032e-06, "epoch": 0.6106122448979592, "percentage": 61.06, "elapsed_time": "5:50:19", "remaining_time": "3:43:24"} +{"current_steps": 3750, "total_steps": 6125, "loss": 0.048448923230171206, "lr": 1.6854291346580737e-06, "epoch": 0.6122448979591837, "percentage": 61.22, "elapsed_time": "5:51:07", "remaining_time": "3:42:22"} +{"current_steps": 3760, "total_steps": 6125, "loss": 0.04779277443885803, "lr": 1.673115979213773e-06, "epoch": 0.6138775510204082, "percentage": 61.39, "elapsed_time": "5:51:54", "remaining_time": "3:41:20"} +{"current_steps": 3770, "total_steps": 6125, "loss": 0.04001966416835785, "lr": 1.6608253054751739e-06, "epoch": 0.6155102040816327, "percentage": 61.55, "elapsed_time": "5:52:41", "remaining_time": "3:40:19"} +{"current_steps": 3780, "total_steps": 6125, "loss": 0.04915740489959717, "lr": 1.6485574476068056e-06, "epoch": 0.6171428571428571, "percentage": 61.71, "elapsed_time": "5:53:29", "remaining_time": "3:39:17"} +{"current_steps": 3790, "total_steps": 6125, "loss": 0.055313241481781, "lr": 1.6363127391528694e-06, "epoch": 0.6187755102040816, "percentage": 61.88, "elapsed_time": "5:54:16", "remaining_time": "3:38:16"} +{"current_steps": 3800, "total_steps": 6125, "loss": 0.05452631115913391, "lr": 1.6240915130281696e-06, "epoch": 0.6204081632653061, "percentage": 62.04, "elapsed_time": "5:55:04", "remaining_time": "3:37:15"} +{"current_steps": 3800, "total_steps": 6125, "eval_loss": 0.04318515956401825, "epoch": 0.6204081632653061, "percentage": 62.04, "elapsed_time": "5:57:43", "remaining_time": "3:38:52"} +{"current_steps": 3810, "total_steps": 6125, "loss": 0.047327810525894166, "lr": 1.6118941015090614e-06, "epoch": 0.6220408163265306, "percentage": 62.2, "elapsed_time": "5:58:35", "remaining_time": "3:37:53"} +{"current_steps": 3820, "total_steps": 6125, "loss": 0.04089818894863129, "lr": 1.599720836224419e-06, "epoch": 0.6236734693877551, "percentage": 62.37, "elapsed_time": "5:59:23", "remaining_time": "3:36:51"} +{"current_steps": 3830, "total_steps": 6125, "loss": 0.04378651976585388, "lr": 1.5875720481466162e-06, "epoch": 0.6253061224489795, "percentage": 62.53, "elapsed_time": "6:00:10", "remaining_time": "3:35:49"} +{"current_steps": 3840, "total_steps": 6125, "loss": 0.055971544981002805, "lr": 1.5754480675825295e-06, "epoch": 0.626938775510204, "percentage": 62.69, "elapsed_time": "6:00:58", "remaining_time": "3:34:47"} +{"current_steps": 3850, "total_steps": 6125, "loss": 0.04364039599895477, "lr": 1.563349224164556e-06, "epoch": 0.6285714285714286, "percentage": 62.86, "elapsed_time": "6:01:45", "remaining_time": "3:33:46"} +{"current_steps": 3860, "total_steps": 6125, "loss": 0.04227602481842041, "lr": 1.5512758468416561e-06, "epoch": 0.6302040816326531, "percentage": 63.02, "elapsed_time": "6:02:33", "remaining_time": "3:32:44"} +{"current_steps": 3870, "total_steps": 6125, "loss": 0.03987776041030884, "lr": 1.5392282638704032e-06, "epoch": 0.6318367346938776, "percentage": 63.18, "elapsed_time": "6:03:20", "remaining_time": "3:31:43"} +{"current_steps": 3880, "total_steps": 6125, "loss": 0.05040489435195923, "lr": 1.5272068028060616e-06, "epoch": 0.6334693877551021, "percentage": 63.35, "elapsed_time": "6:04:08", "remaining_time": "3:30:41"} +{"current_steps": 3890, "total_steps": 6125, "loss": 0.04129888415336609, "lr": 1.5152117904936827e-06, "epoch": 0.6351020408163265, "percentage": 63.51, "elapsed_time": "6:04:55", "remaining_time": "3:29:40"} +{"current_steps": 3900, "total_steps": 6125, "loss": 0.04145504236221313, "lr": 1.5032435530592165e-06, "epoch": 0.636734693877551, "percentage": 63.67, "elapsed_time": "6:05:43", "remaining_time": "3:28:38"} +{"current_steps": 3910, "total_steps": 6125, "loss": 0.03938417434692383, "lr": 1.4913024159006444e-06, "epoch": 0.6383673469387755, "percentage": 63.84, "elapsed_time": "6:06:30", "remaining_time": "3:27:37"} +{"current_steps": 3920, "total_steps": 6125, "loss": 0.047359606623649596, "lr": 1.4793887036791338e-06, "epoch": 0.64, "percentage": 64.0, "elapsed_time": "6:07:17", "remaining_time": "3:26:36"} +{"current_steps": 3930, "total_steps": 6125, "loss": 0.047661393880844116, "lr": 1.4675027403102107e-06, "epoch": 0.6416326530612245, "percentage": 64.16, "elapsed_time": "6:08:04", "remaining_time": "3:25:34"} +{"current_steps": 3940, "total_steps": 6125, "loss": 0.04829471111297608, "lr": 1.4556448489549519e-06, "epoch": 0.643265306122449, "percentage": 64.33, "elapsed_time": "6:08:51", "remaining_time": "3:24:33"} +{"current_steps": 3950, "total_steps": 6125, "loss": 0.047904521226882935, "lr": 1.4438153520111978e-06, "epoch": 0.6448979591836734, "percentage": 64.49, "elapsed_time": "6:09:39", "remaining_time": "3:23:32"} +{"current_steps": 3960, "total_steps": 6125, "loss": 0.04648992121219635, "lr": 1.4320145711047894e-06, "epoch": 0.6465306122448979, "percentage": 64.65, "elapsed_time": "6:10:26", "remaining_time": "3:22:31"} +{"current_steps": 3970, "total_steps": 6125, "loss": 0.044715133309364316, "lr": 1.4202428270808256e-06, "epoch": 0.6481632653061224, "percentage": 64.82, "elapsed_time": "6:11:14", "remaining_time": "3:21:31"} +{"current_steps": 3980, "total_steps": 6125, "loss": 0.042470496892929074, "lr": 1.408500439994933e-06, "epoch": 0.649795918367347, "percentage": 64.98, "elapsed_time": "6:12:01", "remaining_time": "3:20:30"} +{"current_steps": 3990, "total_steps": 6125, "loss": 0.03910423815250397, "lr": 1.3967877291045706e-06, "epoch": 0.6514285714285715, "percentage": 65.14, "elapsed_time": "6:12:48", "remaining_time": "3:19:28"} +{"current_steps": 4000, "total_steps": 6125, "loss": 0.046834564208984374, "lr": 1.3851050128603454e-06, "epoch": 0.6530612244897959, "percentage": 65.31, "elapsed_time": "6:13:34", "remaining_time": "3:18:27"} +{"current_steps": 4000, "total_steps": 6125, "eval_loss": 0.04485878720879555, "epoch": 0.6530612244897959, "percentage": 65.31, "elapsed_time": "6:16:09", "remaining_time": "3:19:50"} +{"current_steps": 4010, "total_steps": 6125, "loss": 0.048332995176315306, "lr": 1.3734526088973576e-06, "epoch": 0.6546938775510204, "percentage": 65.47, "elapsed_time": "6:17:00", "remaining_time": "3:18:50"} +{"current_steps": 4020, "total_steps": 6125, "loss": 0.05632186532020569, "lr": 1.3618308340265628e-06, "epoch": 0.6563265306122449, "percentage": 65.63, "elapsed_time": "6:17:46", "remaining_time": "3:17:48"} +{"current_steps": 4030, "total_steps": 6125, "loss": 0.0413711816072464, "lr": 1.3502400042261583e-06, "epoch": 0.6579591836734694, "percentage": 65.8, "elapsed_time": "6:18:32", "remaining_time": "3:16:47"} +{"current_steps": 4040, "total_steps": 6125, "loss": 0.04443484246730804, "lr": 1.3386804346329913e-06, "epoch": 0.6595918367346939, "percentage": 65.96, "elapsed_time": "6:19:19", "remaining_time": "3:15:45"} +{"current_steps": 4050, "total_steps": 6125, "loss": 0.03936767876148224, "lr": 1.3271524395339958e-06, "epoch": 0.6612244897959184, "percentage": 66.12, "elapsed_time": "6:20:06", "remaining_time": "3:14:44"} +{"current_steps": 4060, "total_steps": 6125, "loss": 0.03893322050571442, "lr": 1.31565633235764e-06, "epoch": 0.6628571428571428, "percentage": 66.29, "elapsed_time": "6:20:53", "remaining_time": "3:13:43"} +{"current_steps": 4070, "total_steps": 6125, "loss": 0.05221283435821533, "lr": 1.30419242566541e-06, "epoch": 0.6644897959183673, "percentage": 66.45, "elapsed_time": "6:21:40", "remaining_time": "3:12:42"} +{"current_steps": 4080, "total_steps": 6125, "loss": 0.04261190891265869, "lr": 1.2927610311433098e-06, "epoch": 0.6661224489795918, "percentage": 66.61, "elapsed_time": "6:22:27", "remaining_time": "3:11:41"} +{"current_steps": 4090, "total_steps": 6125, "loss": 0.04804243743419647, "lr": 1.2813624595933885e-06, "epoch": 0.6677551020408163, "percentage": 66.78, "elapsed_time": "6:23:14", "remaining_time": "3:10:41"} +{"current_steps": 4100, "total_steps": 6125, "loss": 0.03677414059638977, "lr": 1.2699970209252877e-06, "epoch": 0.6693877551020408, "percentage": 66.94, "elapsed_time": "6:24:00", "remaining_time": "3:09:39"} +{"current_steps": 4110, "total_steps": 6125, "loss": 0.04266786575317383, "lr": 1.2586650241478173e-06, "epoch": 0.6710204081632654, "percentage": 67.1, "elapsed_time": "6:24:46", "remaining_time": "3:08:38"} +{"current_steps": 4120, "total_steps": 6125, "loss": 0.045382776856422426, "lr": 1.2473667773605542e-06, "epoch": 0.6726530612244898, "percentage": 67.27, "elapsed_time": "6:25:32", "remaining_time": "3:07:37"} +{"current_steps": 4130, "total_steps": 6125, "loss": 0.0483401358127594, "lr": 1.236102587745464e-06, "epoch": 0.6742857142857143, "percentage": 67.43, "elapsed_time": "6:26:18", "remaining_time": "3:06:36"} +{"current_steps": 4140, "total_steps": 6125, "loss": 0.0531754195690155, "lr": 1.2248727615585501e-06, "epoch": 0.6759183673469388, "percentage": 67.59, "elapsed_time": "6:27:10", "remaining_time": "3:05:38"} +{"current_steps": 4150, "total_steps": 6125, "loss": 0.046069496870040895, "lr": 1.2136776041215293e-06, "epoch": 0.6775510204081633, "percentage": 67.76, "elapsed_time": "6:27:59", "remaining_time": "3:04:38"} +{"current_steps": 4160, "total_steps": 6125, "loss": 0.03938464820384979, "lr": 1.2025174198135255e-06, "epoch": 0.6791836734693878, "percentage": 67.92, "elapsed_time": "6:28:47", "remaining_time": "3:03:39"} +{"current_steps": 4170, "total_steps": 6125, "loss": 0.04410702586174011, "lr": 1.1913925120627987e-06, "epoch": 0.6808163265306122, "percentage": 68.08, "elapsed_time": "6:29:40", "remaining_time": "3:02:41"} +{"current_steps": 4180, "total_steps": 6125, "loss": 0.03496443629264832, "lr": 1.1803031833384923e-06, "epoch": 0.6824489795918367, "percentage": 68.24, "elapsed_time": "6:30:33", "remaining_time": "3:01:43"} +{"current_steps": 4190, "total_steps": 6125, "loss": 0.04019207060337067, "lr": 1.1692497351424117e-06, "epoch": 0.6840816326530612, "percentage": 68.41, "elapsed_time": "6:31:18", "remaining_time": "3:00:42"} +{"current_steps": 4200, "total_steps": 6125, "loss": 0.03875584900379181, "lr": 1.1582324680008252e-06, "epoch": 0.6857142857142857, "percentage": 68.57, "elapsed_time": "6:32:04", "remaining_time": "2:59:42"} +{"current_steps": 4200, "total_steps": 6125, "eval_loss": 0.0395464189350605, "epoch": 0.6857142857142857, "percentage": 68.57, "elapsed_time": "6:34:38", "remaining_time": "3:00:52"} +{"current_steps": 4210, "total_steps": 6125, "loss": 0.04495754837989807, "lr": 1.1472516814562945e-06, "epoch": 0.6873469387755102, "percentage": 68.73, "elapsed_time": "6:35:28", "remaining_time": "2:59:53"} +{"current_steps": 4220, "total_steps": 6125, "loss": 0.0422965407371521, "lr": 1.1363076740595294e-06, "epoch": 0.6889795918367347, "percentage": 68.9, "elapsed_time": "6:36:14", "remaining_time": "2:58:52"} +{"current_steps": 4230, "total_steps": 6125, "loss": 0.04501302838325501, "lr": 1.1254007433612705e-06, "epoch": 0.6906122448979591, "percentage": 69.06, "elapsed_time": "6:37:00", "remaining_time": "2:57:51"} +{"current_steps": 4240, "total_steps": 6125, "loss": 0.03593409359455109, "lr": 1.1145311859042026e-06, "epoch": 0.6922448979591836, "percentage": 69.22, "elapsed_time": "6:37:47", "remaining_time": "2:56:50"} +{"current_steps": 4250, "total_steps": 6125, "loss": 0.034275355935096743, "lr": 1.1036992972148865e-06, "epoch": 0.6938775510204082, "percentage": 69.39, "elapsed_time": "6:38:32", "remaining_time": "2:55:49"} +{"current_steps": 4260, "total_steps": 6125, "loss": 0.039912134408950806, "lr": 1.092905371795729e-06, "epoch": 0.6955102040816327, "percentage": 69.55, "elapsed_time": "6:39:17", "remaining_time": "2:54:48"} +{"current_steps": 4270, "total_steps": 6125, "loss": 0.03563873171806335, "lr": 1.0821497031169729e-06, "epoch": 0.6971428571428572, "percentage": 69.71, "elapsed_time": "6:40:03", "remaining_time": "2:53:47"} +{"current_steps": 4280, "total_steps": 6125, "loss": 0.036586591601371767, "lr": 1.0714325836087191e-06, "epoch": 0.6987755102040817, "percentage": 69.88, "elapsed_time": "6:40:49", "remaining_time": "2:52:47"} +{"current_steps": 4290, "total_steps": 6125, "loss": 0.03775157630443573, "lr": 1.0607543046529763e-06, "epoch": 0.7004081632653061, "percentage": 70.04, "elapsed_time": "6:41:35", "remaining_time": "2:51:46"} +{"current_steps": 4300, "total_steps": 6125, "loss": 0.03171868622303009, "lr": 1.050115156575738e-06, "epoch": 0.7020408163265306, "percentage": 70.2, "elapsed_time": "6:42:21", "remaining_time": "2:50:46"} +{"current_steps": 4310, "total_steps": 6125, "loss": 0.028169712424278258, "lr": 1.0395154286390896e-06, "epoch": 0.7036734693877551, "percentage": 70.37, "elapsed_time": "6:43:07", "remaining_time": "2:49:45"} +{"current_steps": 4320, "total_steps": 6125, "loss": 0.03762204349040985, "lr": 1.028955409033342e-06, "epoch": 0.7053061224489796, "percentage": 70.53, "elapsed_time": "6:43:53", "remaining_time": "2:48:45"} +{"current_steps": 4330, "total_steps": 6125, "loss": 0.04035975933074951, "lr": 1.0184353848692005e-06, "epoch": 0.7069387755102041, "percentage": 70.69, "elapsed_time": "6:44:38", "remaining_time": "2:47:44"} +{"current_steps": 4340, "total_steps": 6125, "loss": 0.04394418001174927, "lr": 1.0079556421699533e-06, "epoch": 0.7085714285714285, "percentage": 70.86, "elapsed_time": "6:45:25", "remaining_time": "2:46:44"} +{"current_steps": 4350, "total_steps": 6125, "loss": 0.04101270735263825, "lr": 9.97516465863698e-07, "epoch": 0.710204081632653, "percentage": 71.02, "elapsed_time": "6:46:11", "remaining_time": "2:45:44"} +{"current_steps": 4360, "total_steps": 6125, "loss": 0.04232141673564911, "lr": 9.87118139775595e-07, "epoch": 0.7118367346938775, "percentage": 71.18, "elapsed_time": "6:46:58", "remaining_time": "2:44:45"} +{"current_steps": 4370, "total_steps": 6125, "loss": 0.03110981285572052, "lr": 9.767609466201494e-07, "epoch": 0.713469387755102, "percentage": 71.35, "elapsed_time": "6:47:45", "remaining_time": "2:43:45"} +{"current_steps": 4380, "total_steps": 6125, "loss": 0.036100244522094725, "lr": 9.664451679935252e-07, "epoch": 0.7151020408163266, "percentage": 71.51, "elapsed_time": "6:48:31", "remaining_time": "2:42:45"} +{"current_steps": 4390, "total_steps": 6125, "loss": 0.028025591373443605, "lr": 9.561710843658892e-07, "epoch": 0.7167346938775511, "percentage": 71.67, "elapsed_time": "6:49:18", "remaining_time": "2:41:45"} +{"current_steps": 4400, "total_steps": 6125, "loss": 0.03276517987251282, "lr": 9.459389750737854e-07, "epoch": 0.7183673469387755, "percentage": 71.84, "elapsed_time": "6:50:04", "remaining_time": "2:40:46"} +{"current_steps": 4400, "total_steps": 6125, "eval_loss": 0.03968822583556175, "epoch": 0.7183673469387755, "percentage": 71.84, "elapsed_time": "6:52:55", "remaining_time": "2:41:53"} +{"current_steps": 4410, "total_steps": 6125, "loss": 0.04240003228187561, "lr": 9.357491183125398e-07, "epoch": 0.72, "percentage": 72.0, "elapsed_time": "6:53:45", "remaining_time": "2:40:54"} +{"current_steps": 4420, "total_steps": 6125, "loss": 0.03953196704387665, "lr": 9.256017911286958e-07, "epoch": 0.7216326530612245, "percentage": 72.16, "elapsed_time": "6:54:32", "remaining_time": "2:39:54"} +{"current_steps": 4430, "total_steps": 6125, "loss": 0.03630510270595551, "lr": 9.154972694124866e-07, "epoch": 0.723265306122449, "percentage": 72.33, "elapsed_time": "6:55:18", "remaining_time": "2:38:54"} +{"current_steps": 4440, "total_steps": 6125, "loss": 0.04581010937690735, "lr": 9.054358278903272e-07, "epoch": 0.7248979591836735, "percentage": 72.49, "elapsed_time": "6:56:04", "remaining_time": "2:37:54"} +{"current_steps": 4450, "total_steps": 6125, "loss": 0.04633161723613739, "lr": 8.954177401173491e-07, "epoch": 0.726530612244898, "percentage": 72.65, "elapsed_time": "6:56:50", "remaining_time": "2:36:54"} +{"current_steps": 4460, "total_steps": 6125, "loss": 0.03686977326869965, "lr": 8.854432784699624e-07, "epoch": 0.7281632653061224, "percentage": 72.82, "elapsed_time": "6:57:36", "remaining_time": "2:35:54"} +{"current_steps": 4470, "total_steps": 6125, "loss": 0.03494169116020203, "lr": 8.755127141384495e-07, "epoch": 0.7297959183673469, "percentage": 72.98, "elapsed_time": "6:58:22", "remaining_time": "2:34:54"} +{"current_steps": 4480, "total_steps": 6125, "loss": 0.04174305498600006, "lr": 8.656263171195931e-07, "epoch": 0.7314285714285714, "percentage": 73.14, "elapsed_time": "6:59:08", "remaining_time": "2:33:54"} +{"current_steps": 4490, "total_steps": 6125, "loss": 0.04263312518596649, "lr": 8.557843562093329e-07, "epoch": 0.7330612244897959, "percentage": 73.31, "elapsed_time": "6:59:54", "remaining_time": "2:32:54"} +{"current_steps": 4500, "total_steps": 6125, "loss": 0.03155354261398315, "lr": 8.459870989954611e-07, "epoch": 0.7346938775510204, "percentage": 73.47, "elapsed_time": "7:00:39", "remaining_time": "2:31:54"} +{"current_steps": 4510, "total_steps": 6125, "loss": 0.05083600878715515, "lr": 8.362348118503424e-07, "epoch": 0.7363265306122448, "percentage": 73.63, "elapsed_time": "7:01:25", "remaining_time": "2:30:54"} +{"current_steps": 4520, "total_steps": 6125, "loss": 0.031102049350738525, "lr": 8.265277599236785e-07, "epoch": 0.7379591836734694, "percentage": 73.8, "elapsed_time": "7:02:12", "remaining_time": "2:29:55"} +{"current_steps": 4530, "total_steps": 6125, "loss": 0.033837145566940306, "lr": 8.168662071352918e-07, "epoch": 0.7395918367346939, "percentage": 73.96, "elapsed_time": "7:02:58", "remaining_time": "2:28:55"} +{"current_steps": 4540, "total_steps": 6125, "loss": 0.0337544322013855, "lr": 8.072504161679534e-07, "epoch": 0.7412244897959184, "percentage": 74.12, "elapsed_time": "7:03:44", "remaining_time": "2:27:56"} +{"current_steps": 4550, "total_steps": 6125, "loss": 0.03291972577571869, "lr": 7.976806484602417e-07, "epoch": 0.7428571428571429, "percentage": 74.29, "elapsed_time": "7:04:38", "remaining_time": "2:26:59"} +{"current_steps": 4560, "total_steps": 6125, "loss": 0.032550466060638425, "lr": 7.881571641994331e-07, "epoch": 0.7444897959183674, "percentage": 74.45, "elapsed_time": "7:05:31", "remaining_time": "2:26:02"} +{"current_steps": 4570, "total_steps": 6125, "loss": 0.030972787737846376, "lr": 7.786802223144286e-07, "epoch": 0.7461224489795918, "percentage": 74.61, "elapsed_time": "7:06:26", "remaining_time": "2:25:06"} +{"current_steps": 4580, "total_steps": 6125, "loss": 0.0415867954492569, "lr": 7.69250080468712e-07, "epoch": 0.7477551020408163, "percentage": 74.78, "elapsed_time": "7:07:19", "remaining_time": "2:24:09"} +{"current_steps": 4590, "total_steps": 6125, "loss": 0.0360225111246109, "lr": 7.59866995053348e-07, "epoch": 0.7493877551020408, "percentage": 74.94, "elapsed_time": "7:08:09", "remaining_time": "2:23:11"} +{"current_steps": 4600, "total_steps": 6125, "loss": 0.03627174496650696, "lr": 7.505312211800081e-07, "epoch": 0.7510204081632653, "percentage": 75.1, "elapsed_time": "7:09:05", "remaining_time": "2:22:15"} +{"current_steps": 4600, "total_steps": 6125, "eval_loss": 0.03568050265312195, "epoch": 0.7510204081632653, "percentage": 75.1, "elapsed_time": "7:11:42", "remaining_time": "2:23:07"} +{"current_steps": 4610, "total_steps": 6125, "loss": 0.033873635530471805, "lr": 7.41243012674035e-07, "epoch": 0.7526530612244898, "percentage": 75.27, "elapsed_time": "7:12:36", "remaining_time": "2:22:10"} +{"current_steps": 4620, "total_steps": 6125, "loss": 0.03949807286262512, "lr": 7.320026220675447e-07, "epoch": 0.7542857142857143, "percentage": 75.43, "elapsed_time": "7:13:23", "remaining_time": "2:21:10"} +{"current_steps": 4630, "total_steps": 6125, "loss": 0.03478931188583374, "lr": 7.228103005925557e-07, "epoch": 0.7559183673469387, "percentage": 75.59, "elapsed_time": "7:14:09", "remaining_time": "2:20:11"} +{"current_steps": 4640, "total_steps": 6125, "loss": 0.030326524376869203, "lr": 7.136662981741613e-07, "epoch": 0.7575510204081632, "percentage": 75.76, "elapsed_time": "7:14:55", "remaining_time": "2:19:11"} +{"current_steps": 4650, "total_steps": 6125, "loss": 0.034136393666267396, "lr": 7.045708634237341e-07, "epoch": 0.7591836734693878, "percentage": 75.92, "elapsed_time": "7:15:41", "remaining_time": "2:18:12"} +{"current_steps": 4660, "total_steps": 6125, "loss": 0.040317881107330325, "lr": 6.955242436321663e-07, "epoch": 0.7608163265306123, "percentage": 76.08, "elapsed_time": "7:16:28", "remaining_time": "2:17:12"} +{"current_steps": 4670, "total_steps": 6125, "loss": 0.02895120084285736, "lr": 6.865266847631463e-07, "epoch": 0.7624489795918368, "percentage": 76.24, "elapsed_time": "7:17:14", "remaining_time": "2:16:13"} +{"current_steps": 4680, "total_steps": 6125, "loss": 0.029726022481918336, "lr": 6.775784314464717e-07, "epoch": 0.7640816326530612, "percentage": 76.41, "elapsed_time": "7:18:00", "remaining_time": "2:15:14"} +{"current_steps": 4690, "total_steps": 6125, "loss": 0.02383600324392319, "lr": 6.686797269713971e-07, "epoch": 0.7657142857142857, "percentage": 76.57, "elapsed_time": "7:18:46", "remaining_time": "2:14:15"} +{"current_steps": 4700, "total_steps": 6125, "loss": 0.03801000118255615, "lr": 6.598308132800221e-07, "epoch": 0.7673469387755102, "percentage": 76.73, "elapsed_time": "7:19:32", "remaining_time": "2:13:15"} +{"current_steps": 4710, "total_steps": 6125, "loss": 0.030053699016571046, "lr": 6.510319309607099e-07, "epoch": 0.7689795918367347, "percentage": 76.9, "elapsed_time": "7:20:18", "remaining_time": "2:12:16"} +{"current_steps": 4720, "total_steps": 6125, "loss": 0.03296325206756592, "lr": 6.422833192415476e-07, "epoch": 0.7706122448979592, "percentage": 77.06, "elapsed_time": "7:21:04", "remaining_time": "2:11:17"} +{"current_steps": 4730, "total_steps": 6125, "loss": 0.029071077704429626, "lr": 6.335852159838423e-07, "epoch": 0.7722448979591837, "percentage": 77.22, "elapsed_time": "7:21:50", "remaining_time": "2:10:18"} +{"current_steps": 4740, "total_steps": 6125, "loss": 0.03307548463344574, "lr": 6.249378576756538e-07, "epoch": 0.7738775510204081, "percentage": 77.39, "elapsed_time": "7:22:36", "remaining_time": "2:09:19"} +{"current_steps": 4750, "total_steps": 6125, "loss": 0.03507235944271088, "lr": 6.163414794253641e-07, "epoch": 0.7755102040816326, "percentage": 77.55, "elapsed_time": "7:23:22", "remaining_time": "2:08:20"} +{"current_steps": 4760, "total_steps": 6125, "loss": 0.038254004716873166, "lr": 6.077963149552868e-07, "epoch": 0.7771428571428571, "percentage": 77.71, "elapsed_time": "7:24:08", "remaining_time": "2:07:21"} +{"current_steps": 4770, "total_steps": 6125, "loss": 0.03428310751914978, "lr": 5.993025965953103e-07, "epoch": 0.7787755102040816, "percentage": 77.88, "elapsed_time": "7:24:53", "remaining_time": "2:06:22"} +{"current_steps": 4780, "total_steps": 6125, "loss": 0.032243117690086365, "lr": 5.908605552765833e-07, "epoch": 0.7804081632653062, "percentage": 78.04, "elapsed_time": "7:25:39", "remaining_time": "2:05:24"} +{"current_steps": 4790, "total_steps": 6125, "loss": 0.03812848925590515, "lr": 5.824704205252343e-07, "epoch": 0.7820408163265307, "percentage": 78.2, "elapsed_time": "7:26:25", "remaining_time": "2:04:25"} +{"current_steps": 4800, "total_steps": 6125, "loss": 0.031389710307121274, "lr": 5.741324204561335e-07, "epoch": 0.7836734693877551, "percentage": 78.37, "elapsed_time": "7:27:11", "remaining_time": "2:03:26"} +{"current_steps": 4800, "total_steps": 6125, "eval_loss": 0.03615976870059967, "epoch": 0.7836734693877551, "percentage": 78.37, "elapsed_time": "7:29:45", "remaining_time": "2:04:09"} +{"current_steps": 4810, "total_steps": 6125, "loss": 0.027873021364212037, "lr": 5.65846781766688e-07, "epoch": 0.7853061224489796, "percentage": 78.53, "elapsed_time": "7:30:35", "remaining_time": "2:03:11"} +{"current_steps": 4820, "total_steps": 6125, "loss": 0.0407156765460968, "lr": 5.576137297306794e-07, "epoch": 0.7869387755102041, "percentage": 78.69, "elapsed_time": "7:31:21", "remaining_time": "2:02:12"} +{"current_steps": 4830, "total_steps": 6125, "loss": 0.03936888575553894, "lr": 5.494334881921393e-07, "epoch": 0.7885714285714286, "percentage": 78.86, "elapsed_time": "7:32:07", "remaining_time": "2:01:13"} +{"current_steps": 4840, "total_steps": 6125, "loss": 0.03059300184249878, "lr": 5.413062795592633e-07, "epoch": 0.7902040816326531, "percentage": 79.02, "elapsed_time": "7:32:52", "remaining_time": "2:00:14"} +{"current_steps": 4850, "total_steps": 6125, "loss": 0.038730043172836306, "lr": 5.332323247983632e-07, "epoch": 0.7918367346938775, "percentage": 79.18, "elapsed_time": "7:33:38", "remaining_time": "1:59:15"} +{"current_steps": 4860, "total_steps": 6125, "loss": 0.033898717164993285, "lr": 5.252118434278605e-07, "epoch": 0.793469387755102, "percentage": 79.35, "elapsed_time": "7:34:24", "remaining_time": "1:58:16"} +{"current_steps": 4870, "total_steps": 6125, "loss": 0.033659690618515016, "lr": 5.172450535123166e-07, "epoch": 0.7951020408163265, "percentage": 79.51, "elapsed_time": "7:35:09", "remaining_time": "1:57:17"} +{"current_steps": 4880, "total_steps": 6125, "loss": 0.0293354332447052, "lr": 5.093321716565064e-07, "epoch": 0.796734693877551, "percentage": 79.67, "elapsed_time": "7:35:54", "remaining_time": "1:56:18"} +{"current_steps": 4890, "total_steps": 6125, "loss": 0.027682244777679443, "lr": 5.014734129995258e-07, "epoch": 0.7983673469387755, "percentage": 79.84, "elapsed_time": "7:36:41", "remaining_time": "1:55:20"} +{"current_steps": 4900, "total_steps": 6125, "loss": 0.029118484258651732, "lr": 4.936689912089443e-07, "epoch": 0.8, "percentage": 80.0, "elapsed_time": "7:37:27", "remaining_time": "1:54:21"} +{"current_steps": 4910, "total_steps": 6125, "loss": 0.031624987721443176, "lr": 4.859191184749968e-07, "epoch": 0.8016326530612244, "percentage": 80.16, "elapsed_time": "7:38:13", "remaining_time": "1:53:23"} +{"current_steps": 4920, "total_steps": 6125, "loss": 0.03415923118591309, "lr": 4.782240055048113e-07, "epoch": 0.803265306122449, "percentage": 80.33, "elapsed_time": "7:38:59", "remaining_time": "1:52:24"} +{"current_steps": 4930, "total_steps": 6125, "loss": 0.035749971866607666, "lr": 4.705838615166836e-07, "epoch": 0.8048979591836735, "percentage": 80.49, "elapsed_time": "7:39:45", "remaining_time": "1:51:26"} +{"current_steps": 4940, "total_steps": 6125, "loss": 0.029294773936271667, "lr": 4.629988942343863e-07, "epoch": 0.806530612244898, "percentage": 80.65, "elapsed_time": "7:40:31", "remaining_time": "1:50:28"} +{"current_steps": 4950, "total_steps": 6125, "loss": 0.026727235317230223, "lr": 4.55469309881523e-07, "epoch": 0.8081632653061225, "percentage": 80.82, "elapsed_time": "7:41:17", "remaining_time": "1:49:29"} +{"current_steps": 4960, "total_steps": 6125, "loss": 0.029011189937591553, "lr": 4.479953131759199e-07, "epoch": 0.809795918367347, "percentage": 80.98, "elapsed_time": "7:42:03", "remaining_time": "1:48:31"} +{"current_steps": 4970, "total_steps": 6125, "loss": 0.040424668788909913, "lr": 4.4057710732406113e-07, "epoch": 0.8114285714285714, "percentage": 81.14, "elapsed_time": "7:42:49", "remaining_time": "1:47:33"} +{"current_steps": 4980, "total_steps": 6125, "loss": 0.031291928887367246, "lr": 4.332148940155617e-07, "epoch": 0.8130612244897959, "percentage": 81.31, "elapsed_time": "7:43:35", "remaining_time": "1:46:35"} +{"current_steps": 4990, "total_steps": 6125, "loss": 0.028314629197120668, "lr": 4.259088734176883e-07, "epoch": 0.8146938775510204, "percentage": 81.47, "elapsed_time": "7:44:20", "remaining_time": "1:45:36"} +{"current_steps": 5000, "total_steps": 6125, "loss": 0.028049173951148986, "lr": 4.186592441699119e-07, "epoch": 0.8163265306122449, "percentage": 81.63, "elapsed_time": "7:45:06", "remaining_time": "1:44:38"} +{"current_steps": 5000, "total_steps": 6125, "eval_loss": 0.033585112541913986, "epoch": 0.8163265306122449, "percentage": 81.63, "elapsed_time": "7:47:39", "remaining_time": "1:45:13"} +{"current_steps": 5010, "total_steps": 6125, "loss": 0.03612466752529144, "lr": 4.1146620337850965e-07, "epoch": 0.8179591836734694, "percentage": 81.8, "elapsed_time": "7:48:29", "remaining_time": "1:44:15"} +{"current_steps": 5020, "total_steps": 6125, "loss": 0.026968371868133546, "lr": 4.043299466112058e-07, "epoch": 0.8195918367346938, "percentage": 81.96, "elapsed_time": "7:49:15", "remaining_time": "1:43:17"} +{"current_steps": 5030, "total_steps": 6125, "loss": 0.030766445398330688, "lr": 3.9725066789185446e-07, "epoch": 0.8212244897959183, "percentage": 82.12, "elapsed_time": "7:50:01", "remaining_time": "1:42:19"} +{"current_steps": 5040, "total_steps": 6125, "loss": 0.028759539127349854, "lr": 3.9022855969516384e-07, "epoch": 0.8228571428571428, "percentage": 82.29, "elapsed_time": "7:50:47", "remaining_time": "1:41:20"} +{"current_steps": 5050, "total_steps": 6125, "loss": 0.03126598298549652, "lr": 3.832638129414637e-07, "epoch": 0.8244897959183674, "percentage": 82.45, "elapsed_time": "7:51:33", "remaining_time": "1:40:22"} +{"current_steps": 5060, "total_steps": 6125, "loss": 0.030412834882736207, "lr": 3.7635661699151366e-07, "epoch": 0.8261224489795919, "percentage": 82.61, "elapsed_time": "7:52:19", "remaining_time": "1:39:24"} +{"current_steps": 5070, "total_steps": 6125, "loss": 0.03904336988925934, "lr": 3.695071596413574e-07, "epoch": 0.8277551020408164, "percentage": 82.78, "elapsed_time": "7:53:05", "remaining_time": "1:38:26"} +{"current_steps": 5080, "total_steps": 6125, "loss": 0.032776132225990295, "lr": 3.627156271172125e-07, "epoch": 0.8293877551020408, "percentage": 82.94, "elapsed_time": "7:53:51", "remaining_time": "1:37:28"} +{"current_steps": 5090, "total_steps": 6125, "loss": 0.027044805884361266, "lr": 3.559822040704108e-07, "epoch": 0.8310204081632653, "percentage": 83.1, "elapsed_time": "7:54:37", "remaining_time": "1:36:30"} +{"current_steps": 5100, "total_steps": 6125, "loss": 0.03271254003047943, "lr": 3.4930707357237666e-07, "epoch": 0.8326530612244898, "percentage": 83.27, "elapsed_time": "7:55:23", "remaining_time": "1:35:32"} +{"current_steps": 5110, "total_steps": 6125, "loss": 0.030745366215705873, "lr": 3.4269041710964903e-07, "epoch": 0.8342857142857143, "percentage": 83.43, "elapsed_time": "7:56:09", "remaining_time": "1:34:34"} +{"current_steps": 5120, "total_steps": 6125, "loss": 0.032180759310722354, "lr": 3.36132414578948e-07, "epoch": 0.8359183673469388, "percentage": 83.59, "elapsed_time": "7:56:55", "remaining_time": "1:33:36"} +{"current_steps": 5130, "total_steps": 6125, "loss": 0.026755687594413758, "lr": 3.296332442822839e-07, "epoch": 0.8375510204081633, "percentage": 83.76, "elapsed_time": "7:57:41", "remaining_time": "1:32:39"} +{"current_steps": 5140, "total_steps": 6125, "loss": 0.02895594835281372, "lr": 3.231930829221083e-07, "epoch": 0.8391836734693877, "percentage": 83.92, "elapsed_time": "7:58:27", "remaining_time": "1:31:41"} +{"current_steps": 5150, "total_steps": 6125, "loss": 0.033178862929344174, "lr": 3.168121055965112e-07, "epoch": 0.8408163265306122, "percentage": 84.08, "elapsed_time": "7:59:13", "remaining_time": "1:30:43"} +{"current_steps": 5160, "total_steps": 6125, "loss": 0.02276728004217148, "lr": 3.104904857944591e-07, "epoch": 0.8424489795918367, "percentage": 84.24, "elapsed_time": "7:59:59", "remaining_time": "1:29:45"} +{"current_steps": 5170, "total_steps": 6125, "loss": 0.028258222341537475, "lr": 3.0422839539107794e-07, "epoch": 0.8440816326530612, "percentage": 84.41, "elapsed_time": "8:00:45", "remaining_time": "1:28:48"} +{"current_steps": 5180, "total_steps": 6125, "loss": 0.02725991904735565, "lr": 2.980260046429834e-07, "epoch": 0.8457142857142858, "percentage": 84.57, "elapsed_time": "8:01:32", "remaining_time": "1:27:50"} +{"current_steps": 5190, "total_steps": 6125, "loss": 0.0327379584312439, "lr": 2.9188348218364635e-07, "epoch": 0.8473469387755102, "percentage": 84.73, "elapsed_time": "8:02:27", "remaining_time": "1:26:54"} +{"current_steps": 5200, "total_steps": 6125, "loss": 0.030778995156288146, "lr": 2.8580099501881285e-07, "epoch": 0.8489795918367347, "percentage": 84.9, "elapsed_time": "8:03:17", "remaining_time": "1:25:58"} +{"current_steps": 5200, "total_steps": 6125, "eval_loss": 0.02920014224946499, "epoch": 0.8489795918367347, "percentage": 84.9, "elapsed_time": "8:05:53", "remaining_time": "1:26:25"} +{"current_steps": 5210, "total_steps": 6125, "loss": 0.03692179024219513, "lr": 2.7977870852195997e-07, "epoch": 0.8506122448979592, "percentage": 85.06, "elapsed_time": "8:06:43", "remaining_time": "1:25:28"} +{"current_steps": 5220, "total_steps": 6125, "loss": 0.030528882145881654, "lr": 2.7381678642980303e-07, "epoch": 0.8522448979591837, "percentage": 85.22, "elapsed_time": "8:07:29", "remaining_time": "1:24:31"} +{"current_steps": 5230, "total_steps": 6125, "loss": 0.025125756859779358, "lr": 2.6791539083784014e-07, "epoch": 0.8538775510204082, "percentage": 85.39, "elapsed_time": "8:08:16", "remaining_time": "1:23:33"} +{"current_steps": 5240, "total_steps": 6125, "loss": 0.02720560133457184, "lr": 2.620746821959475e-07, "epoch": 0.8555102040816327, "percentage": 85.55, "elapsed_time": "8:09:02", "remaining_time": "1:22:35"} +{"current_steps": 5250, "total_steps": 6125, "loss": 0.033940815925598146, "lr": 2.562948193040174e-07, "epoch": 0.8571428571428571, "percentage": 85.71, "elapsed_time": "8:09:48", "remaining_time": "1:21:38"} +{"current_steps": 5260, "total_steps": 6125, "loss": 0.028036150336265563, "lr": 2.5057595930763873e-07, "epoch": 0.8587755102040816, "percentage": 85.88, "elapsed_time": "8:10:34", "remaining_time": "1:20:40"} +{"current_steps": 5270, "total_steps": 6125, "loss": 0.02731768786907196, "lr": 2.4491825769382526e-07, "epoch": 0.8604081632653061, "percentage": 86.04, "elapsed_time": "8:11:20", "remaining_time": "1:19:42"} +{"current_steps": 5280, "total_steps": 6125, "loss": 0.0313712328672409, "lr": 2.39321868286789e-07, "epoch": 0.8620408163265306, "percentage": 86.2, "elapsed_time": "8:12:06", "remaining_time": "1:18:45"} +{"current_steps": 5290, "total_steps": 6125, "loss": 0.02781860828399658, "lr": 2.337869432437567e-07, "epoch": 0.8636734693877551, "percentage": 86.37, "elapsed_time": "8:12:52", "remaining_time": "1:17:47"} +{"current_steps": 5300, "total_steps": 6125, "loss": 0.028782907128334045, "lr": 2.283136330508337e-07, "epoch": 0.8653061224489796, "percentage": 86.53, "elapsed_time": "8:13:39", "remaining_time": "1:16:50"} +{"current_steps": 5310, "total_steps": 6125, "loss": 0.02515394687652588, "lr": 2.2290208651891244e-07, "epoch": 0.866938775510204, "percentage": 86.69, "elapsed_time": "8:14:26", "remaining_time": "1:15:53"} +{"current_steps": 5320, "total_steps": 6125, "loss": 0.02002216875553131, "lr": 2.1755245077962618e-07, "epoch": 0.8685714285714285, "percentage": 86.86, "elapsed_time": "8:15:12", "remaining_time": "1:14:56"} +{"current_steps": 5330, "total_steps": 6125, "loss": 0.03221684396266937, "lr": 2.1226487128134854e-07, "epoch": 0.8702040816326531, "percentage": 87.02, "elapsed_time": "8:15:59", "remaining_time": "1:13:58"} +{"current_steps": 5340, "total_steps": 6125, "loss": 0.023763810098171235, "lr": 2.070394917852403e-07, "epoch": 0.8718367346938776, "percentage": 87.18, "elapsed_time": "8:16:45", "remaining_time": "1:13:01"} +{"current_steps": 5350, "total_steps": 6125, "loss": 0.03192594051361084, "lr": 2.0187645436133913e-07, "epoch": 0.8734693877551021, "percentage": 87.35, "elapsed_time": "8:17:31", "remaining_time": "1:12:04"} +{"current_steps": 5360, "total_steps": 6125, "loss": 0.025955894589424135, "lr": 1.967758993846966e-07, "epoch": 0.8751020408163265, "percentage": 87.51, "elapsed_time": "8:18:17", "remaining_time": "1:11:07"} +{"current_steps": 5370, "total_steps": 6125, "loss": 0.027435585856437683, "lr": 1.9173796553156487e-07, "epoch": 0.876734693877551, "percentage": 87.67, "elapsed_time": "8:19:03", "remaining_time": "1:10:09"} +{"current_steps": 5380, "total_steps": 6125, "loss": 0.028912261128425598, "lr": 1.8676278977562219e-07, "epoch": 0.8783673469387755, "percentage": 87.84, "elapsed_time": "8:19:50", "remaining_time": "1:09:12"} +{"current_steps": 5390, "total_steps": 6125, "loss": 0.029015693068504333, "lr": 1.818505073842508e-07, "epoch": 0.88, "percentage": 88.0, "elapsed_time": "8:20:36", "remaining_time": "1:08:15"} +{"current_steps": 5400, "total_steps": 6125, "loss": 0.025249993801116942, "lr": 1.770012519148598e-07, "epoch": 0.8816326530612245, "percentage": 88.16, "elapsed_time": "8:21:28", "remaining_time": "1:07:19"} +{"current_steps": 5400, "total_steps": 6125, "eval_loss": 0.02678457647562027, "epoch": 0.8816326530612245, "percentage": 88.16, "elapsed_time": "8:24:04", "remaining_time": "1:07:40"} +{"current_steps": 5410, "total_steps": 6125, "loss": 0.03486272096633911, "lr": 1.722151552112525e-07, "epoch": 0.883265306122449, "percentage": 88.33, "elapsed_time": "8:25:04", "remaining_time": "1:06:45"} +{"current_steps": 5420, "total_steps": 6125, "loss": 0.024751077592372894, "lr": 1.6749234740004227e-07, "epoch": 0.8848979591836734, "percentage": 88.49, "elapsed_time": "8:25:59", "remaining_time": "1:05:49"} +{"current_steps": 5430, "total_steps": 6125, "loss": 0.025553828477859496, "lr": 1.6283295688711564e-07, "epoch": 0.8865306122448979, "percentage": 88.65, "elapsed_time": "8:26:55", "remaining_time": "1:04:52"} +{"current_steps": 5440, "total_steps": 6125, "loss": 0.025743791460990907, "lr": 1.582371103541394e-07, "epoch": 0.8881632653061224, "percentage": 88.82, "elapsed_time": "8:27:51", "remaining_time": "1:03:56"} +{"current_steps": 5450, "total_steps": 6125, "loss": 0.02687264084815979, "lr": 1.5370493275511767e-07, "epoch": 0.889795918367347, "percentage": 88.98, "elapsed_time": "8:28:40", "remaining_time": "1:03:00"} +{"current_steps": 5460, "total_steps": 6125, "loss": 0.023073682188987733, "lr": 1.4923654731299363e-07, "epoch": 0.8914285714285715, "percentage": 89.14, "elapsed_time": "8:29:26", "remaining_time": "1:02:02"} +{"current_steps": 5470, "total_steps": 6125, "loss": 0.0280561625957489, "lr": 1.4483207551630045e-07, "epoch": 0.893061224489796, "percentage": 89.31, "elapsed_time": "8:30:13", "remaining_time": "1:01:05"} +{"current_steps": 5480, "total_steps": 6125, "loss": 0.02133965492248535, "lr": 1.4049163711585684e-07, "epoch": 0.8946938775510204, "percentage": 89.47, "elapsed_time": "8:30:59", "remaining_time": "1:00:08"} +{"current_steps": 5490, "total_steps": 6125, "loss": 0.017637649178504945, "lr": 1.3621535012151244e-07, "epoch": 0.8963265306122449, "percentage": 89.63, "elapsed_time": "8:31:45", "remaining_time": "0:59:11"} +{"current_steps": 5500, "total_steps": 6125, "loss": 0.02847398817539215, "lr": 1.3200333079893823e-07, "epoch": 0.8979591836734694, "percentage": 89.8, "elapsed_time": "8:32:31", "remaining_time": "0:58:14"} +{"current_steps": 5510, "total_steps": 6125, "loss": 0.020397891104221345, "lr": 1.2785569366646628e-07, "epoch": 0.8995918367346939, "percentage": 89.96, "elapsed_time": "8:33:17", "remaining_time": "0:57:17"} +{"current_steps": 5520, "total_steps": 6125, "loss": 0.03015478849411011, "lr": 1.2377255149197576e-07, "epoch": 0.9012244897959184, "percentage": 90.12, "elapsed_time": "8:34:03", "remaining_time": "0:56:20"} +{"current_steps": 5530, "total_steps": 6125, "loss": 0.02614833414554596, "lr": 1.1975401528982718e-07, "epoch": 0.9028571428571428, "percentage": 90.29, "elapsed_time": "8:34:49", "remaining_time": "0:55:23"} +{"current_steps": 5540, "total_steps": 6125, "loss": 0.023097904026508333, "lr": 1.1580019431784384e-07, "epoch": 0.9044897959183673, "percentage": 90.45, "elapsed_time": "8:35:35", "remaining_time": "0:54:26"} +{"current_steps": 5550, "total_steps": 6125, "loss": 0.028036117553710938, "lr": 1.1191119607434076e-07, "epoch": 0.9061224489795918, "percentage": 90.61, "elapsed_time": "8:36:21", "remaining_time": "0:53:29"} +{"current_steps": 5560, "total_steps": 6125, "loss": 0.027321887016296387, "lr": 1.0808712629520435e-07, "epoch": 0.9077551020408163, "percentage": 90.78, "elapsed_time": "8:37:07", "remaining_time": "0:52:32"} +{"current_steps": 5570, "total_steps": 6125, "loss": 0.03543014228343964, "lr": 1.0432808895101426e-07, "epoch": 0.9093877551020408, "percentage": 90.94, "elapsed_time": "8:37:53", "remaining_time": "0:51:36"} +{"current_steps": 5580, "total_steps": 6125, "loss": 0.02044544517993927, "lr": 1.0063418624421895e-07, "epoch": 0.9110204081632653, "percentage": 91.1, "elapsed_time": "8:38:39", "remaining_time": "0:50:39"} +{"current_steps": 5590, "total_steps": 6125, "loss": 0.027166444063186645, "lr": 9.700551860635538e-08, "epoch": 0.9126530612244897, "percentage": 91.27, "elapsed_time": "8:39:25", "remaining_time": "0:49:42"} +{"current_steps": 5600, "total_steps": 6125, "loss": 0.02024113982915878, "lr": 9.344218469532012e-08, "epoch": 0.9142857142857143, "percentage": 91.43, "elapsed_time": "8:40:11", "remaining_time": "0:48:46"} +{"current_steps": 5600, "total_steps": 6125, "eval_loss": 0.025807492434978485, "epoch": 0.9142857142857143, "percentage": 91.43, "elapsed_time": "8:42:46", "remaining_time": "0:49:00"} +{"current_steps": 5610, "total_steps": 6125, "loss": 0.024004274606704713, "lr": 8.99442813926854e-08, "epoch": 0.9159183673469388, "percentage": 91.59, "elapsed_time": "8:43:38", "remaining_time": "0:48:04"} +{"current_steps": 5620, "total_steps": 6125, "loss": 0.024219857156276704, "lr": 8.651190380106673e-08, "epoch": 0.9175510204081633, "percentage": 91.76, "elapsed_time": "8:44:28", "remaining_time": "0:47:07"} +{"current_steps": 5630, "total_steps": 6125, "loss": 0.02802177369594574, "lr": 8.314514524153522e-08, "epoch": 0.9191836734693878, "percentage": 91.92, "elapsed_time": "8:45:14", "remaining_time": "0:46:10"} +{"current_steps": 5640, "total_steps": 6125, "loss": 0.021109704673290253, "lr": 7.984409725108222e-08, "epoch": 0.9208163265306123, "percentage": 92.08, "elapsed_time": "8:46:01", "remaining_time": "0:45:14"} +{"current_steps": 5650, "total_steps": 6125, "loss": 0.020703735947608947, "lr": 7.660884958012899e-08, "epoch": 0.9224489795918367, "percentage": 92.24, "elapsed_time": "8:46:47", "remaining_time": "0:44:17"} +{"current_steps": 5660, "total_steps": 6125, "loss": 0.03247661292552948, "lr": 7.343949019008756e-08, "epoch": 0.9240816326530612, "percentage": 92.41, "elapsed_time": "8:47:33", "remaining_time": "0:43:20"} +{"current_steps": 5670, "total_steps": 6125, "loss": 0.021647430956363678, "lr": 7.03361052509688e-08, "epoch": 0.9257142857142857, "percentage": 92.57, "elapsed_time": "8:48:20", "remaining_time": "0:42:23"} +{"current_steps": 5680, "total_steps": 6125, "loss": 0.028400209546089173, "lr": 6.729877913903976e-08, "epoch": 0.9273469387755102, "percentage": 92.73, "elapsed_time": "8:49:06", "remaining_time": "0:41:27"} +{"current_steps": 5690, "total_steps": 6125, "loss": 0.014232246577739716, "lr": 6.432759443452924e-08, "epoch": 0.9289795918367347, "percentage": 92.9, "elapsed_time": "8:49:52", "remaining_time": "0:40:30"} +{"current_steps": 5700, "total_steps": 6125, "loss": 0.02244553864002228, "lr": 6.14226319193828e-08, "epoch": 0.9306122448979591, "percentage": 93.06, "elapsed_time": "8:50:39", "remaining_time": "0:39:34"} +{"current_steps": 5710, "total_steps": 6125, "loss": 0.029011306166648865, "lr": 5.858397057506709e-08, "epoch": 0.9322448979591836, "percentage": 93.22, "elapsed_time": "8:51:27", "remaining_time": "0:38:37"} +{"current_steps": 5720, "total_steps": 6125, "loss": 0.028310227394104003, "lr": 5.581168758042127e-08, "epoch": 0.9338775510204081, "percentage": 93.39, "elapsed_time": "8:52:14", "remaining_time": "0:37:41"} +{"current_steps": 5730, "total_steps": 6125, "loss": 0.021468037366867067, "lr": 5.310585830955922e-08, "epoch": 0.9355102040816327, "percentage": 93.55, "elapsed_time": "8:53:00", "remaining_time": "0:36:44"} +{"current_steps": 5740, "total_steps": 6125, "loss": 0.024149742722511292, "lr": 5.0466556329819835e-08, "epoch": 0.9371428571428572, "percentage": 93.71, "elapsed_time": "8:53:52", "remaining_time": "0:35:48"} +{"current_steps": 5750, "total_steps": 6125, "loss": 0.01809695065021515, "lr": 4.789385339976832e-08, "epoch": 0.9387755102040817, "percentage": 93.88, "elapsed_time": "8:54:38", "remaining_time": "0:34:52"} +{"current_steps": 5760, "total_steps": 6125, "loss": 0.022831857204437256, "lr": 4.53878194672433e-08, "epoch": 0.9404081632653061, "percentage": 94.04, "elapsed_time": "8:55:24", "remaining_time": "0:33:55"} +{"current_steps": 5770, "total_steps": 6125, "loss": 0.015157052874565124, "lr": 4.2948522667455864e-08, "epoch": 0.9420408163265306, "percentage": 94.2, "elapsed_time": "8:56:10", "remaining_time": "0:32:59"} +{"current_steps": 5780, "total_steps": 6125, "loss": 0.02486245185136795, "lr": 4.057602932113658e-08, "epoch": 0.9436734693877551, "percentage": 94.37, "elapsed_time": "8:56:57", "remaining_time": "0:32:03"} +{"current_steps": 5790, "total_steps": 6125, "loss": 0.024203602969646455, "lr": 3.82704039327339e-08, "epoch": 0.9453061224489796, "percentage": 94.53, "elapsed_time": "8:57:43", "remaining_time": "0:31:06"} +{"current_steps": 5800, "total_steps": 6125, "loss": 0.025162199139595033, "lr": 3.6031709188658336e-08, "epoch": 0.9469387755102041, "percentage": 94.69, "elapsed_time": "8:58:30", "remaining_time": "0:30:10"} +{"current_steps": 5800, "total_steps": 6125, "eval_loss": 0.025506427511572838, "epoch": 0.9469387755102041, "percentage": 94.69, "elapsed_time": "9:01:04", "remaining_time": "0:30:19"} +{"current_steps": 5810, "total_steps": 6125, "loss": 0.027550938725471496, "lr": 3.386000595557964e-08, "epoch": 0.9485714285714286, "percentage": 94.86, "elapsed_time": "9:01:54", "remaining_time": "0:29:22"} +{"current_steps": 5820, "total_steps": 6125, "loss": 0.024577364325523376, "lr": 3.1755353278770936e-08, "epoch": 0.950204081632653, "percentage": 95.02, "elapsed_time": "9:02:40", "remaining_time": "0:28:26"} +{"current_steps": 5830, "total_steps": 6125, "loss": 0.021038763225078583, "lr": 2.9717808380503866e-08, "epoch": 0.9518367346938775, "percentage": 95.18, "elapsed_time": "9:03:26", "remaining_time": "0:27:29"} +{"current_steps": 5840, "total_steps": 6125, "loss": 0.02484375685453415, "lr": 2.7747426658492893e-08, "epoch": 0.953469387755102, "percentage": 95.35, "elapsed_time": "9:04:12", "remaining_time": "0:26:33"} +{"current_steps": 5850, "total_steps": 6125, "loss": 0.023458963632583617, "lr": 2.5844261684388473e-08, "epoch": 0.9551020408163265, "percentage": 95.51, "elapsed_time": "9:04:58", "remaining_time": "0:25:37"} +{"current_steps": 5860, "total_steps": 6125, "loss": 0.03132525682449341, "lr": 2.4008365202321525e-08, "epoch": 0.9567346938775511, "percentage": 95.67, "elapsed_time": "9:05:44", "remaining_time": "0:24:40"} +{"current_steps": 5870, "total_steps": 6125, "loss": 0.03181464076042175, "lr": 2.2239787127495695e-08, "epoch": 0.9583673469387755, "percentage": 95.84, "elapsed_time": "9:06:30", "remaining_time": "0:23:44"} +{"current_steps": 5880, "total_steps": 6125, "loss": 0.019847501814365388, "lr": 2.0538575544831186e-08, "epoch": 0.96, "percentage": 96.0, "elapsed_time": "9:07:17", "remaining_time": "0:22:48"} +{"current_steps": 5890, "total_steps": 6125, "loss": 0.02205444723367691, "lr": 1.8904776707655837e-08, "epoch": 0.9616326530612245, "percentage": 96.16, "elapsed_time": "9:08:03", "remaining_time": "0:21:51"} +{"current_steps": 5900, "total_steps": 6125, "loss": 0.022021512687206268, "lr": 1.7338435036449164e-08, "epoch": 0.963265306122449, "percentage": 96.33, "elapsed_time": "9:08:49", "remaining_time": "0:20:55"} +{"current_steps": 5910, "total_steps": 6125, "loss": 0.02300398051738739, "lr": 1.5839593117634166e-08, "epoch": 0.9648979591836735, "percentage": 96.49, "elapsed_time": "9:09:35", "remaining_time": "0:19:59"} +{"current_steps": 5920, "total_steps": 6125, "loss": 0.024825331568717957, "lr": 1.4408291702418808e-08, "epoch": 0.966530612244898, "percentage": 96.65, "elapsed_time": "9:10:21", "remaining_time": "0:19:03"} +{"current_steps": 5930, "total_steps": 6125, "loss": 0.02171301543712616, "lr": 1.3044569705689125e-08, "epoch": 0.9681632653061224, "percentage": 96.82, "elapsed_time": "9:11:07", "remaining_time": "0:18:07"} +{"current_steps": 5940, "total_steps": 6125, "loss": 0.02564157545566559, "lr": 1.1748464204950072e-08, "epoch": 0.9697959183673469, "percentage": 96.98, "elapsed_time": "9:12:00", "remaining_time": "0:17:11"} +{"current_steps": 5950, "total_steps": 6125, "loss": 0.02467280924320221, "lr": 1.052001043931855e-08, "epoch": 0.9714285714285714, "percentage": 97.14, "elapsed_time": "9:12:57", "remaining_time": "0:16:15"} +{"current_steps": 5960, "total_steps": 6125, "loss": 0.02429400235414505, "lr": 9.3592418085639e-09, "epoch": 0.9730612244897959, "percentage": 97.31, "elapsed_time": "9:13:53", "remaining_time": "0:15:20"} +{"current_steps": 5970, "total_steps": 6125, "loss": 0.02331821173429489, "lr": 8.266189872201957e-09, "epoch": 0.9746938775510204, "percentage": 97.47, "elapsed_time": "9:14:49", "remaining_time": "0:14:24"} +{"current_steps": 5980, "total_steps": 6125, "loss": 0.02990926504135132, "lr": 7.24088434863407e-09, "epoch": 0.976326530612245, "percentage": 97.63, "elapsed_time": "9:15:45", "remaining_time": "0:13:28"} +{"current_steps": 5990, "total_steps": 6125, "loss": 0.023541297018527984, "lr": 6.283353114342749e-09, "epoch": 0.9779591836734693, "percentage": 97.8, "elapsed_time": "9:16:42", "remaining_time": "0:12:32"} +{"current_steps": 6000, "total_steps": 6125, "loss": 0.020145678520202638, "lr": 5.393622203129778e-09, "epoch": 0.9795918367346939, "percentage": 97.96, "elapsed_time": "9:17:39", "remaining_time": "0:11:37"} +{"current_steps": 6000, "total_steps": 6125, "eval_loss": 0.025301460176706314, "epoch": 0.9795918367346939, "percentage": 97.96, "elapsed_time": "9:20:17", "remaining_time": "0:11:40"} +{"current_steps": 6010, "total_steps": 6125, "loss": 0.02092682272195816, "lr": 4.5717158054114916e-09, "epoch": 0.9812244897959184, "percentage": 98.12, "elapsed_time": "9:21:15", "remaining_time": "0:10:44"} +{"current_steps": 6020, "total_steps": 6125, "loss": 0.029862532019615175, "lr": 3.817656267559866e-09, "epoch": 0.9828571428571429, "percentage": 98.29, "elapsed_time": "9:22:02", "remaining_time": "0:09:48"} +{"current_steps": 6030, "total_steps": 6125, "loss": 0.021963220834732056, "lr": 3.1314640912943893e-09, "epoch": 0.9844897959183674, "percentage": 98.45, "elapsed_time": "9:22:50", "remaining_time": "0:08:52"} +{"current_steps": 6040, "total_steps": 6125, "loss": 0.01937238872051239, "lr": 2.5131579331261204e-09, "epoch": 0.9861224489795918, "percentage": 98.61, "elapsed_time": "9:23:37", "remaining_time": "0:07:55"} +{"current_steps": 6050, "total_steps": 6125, "loss": 0.03262064456939697, "lr": 1.9627546038492042e-09, "epoch": 0.9877551020408163, "percentage": 98.78, "elapsed_time": "9:24:25", "remaining_time": "0:06:59"} +{"current_steps": 6060, "total_steps": 6125, "loss": 0.02464511841535568, "lr": 1.4802690680842947e-09, "epoch": 0.9893877551020408, "percentage": 98.94, "elapsed_time": "9:25:12", "remaining_time": "0:06:03"} +{"current_steps": 6070, "total_steps": 6125, "loss": 0.020028088986873627, "lr": 1.065714443871102e-09, "epoch": 0.9910204081632653, "percentage": 99.1, "elapsed_time": "9:25:59", "remaining_time": "0:05:07"} +{"current_steps": 6080, "total_steps": 6125, "loss": 0.03540560901165009, "lr": 7.191020023133988e-10, "epoch": 0.9926530612244898, "percentage": 99.27, "elapsed_time": "9:26:47", "remaining_time": "0:04:11"} +{"current_steps": 6090, "total_steps": 6125, "loss": 0.026075825095176697, "lr": 4.404411672703779e-10, "epoch": 0.9942857142857143, "percentage": 99.43, "elapsed_time": "9:27:35", "remaining_time": "0:03:15"} +{"current_steps": 6100, "total_steps": 6125, "loss": 0.024257904291152953, "lr": 2.2973951510185665e-10, "epoch": 0.9959183673469387, "percentage": 99.59, "elapsed_time": "9:28:22", "remaining_time": "0:02:19"} +{"current_steps": 6110, "total_steps": 6125, "loss": 0.021962395310401915, "lr": 8.700277446233029e-11, "epoch": 0.9975510204081632, "percentage": 99.76, "elapsed_time": "9:29:10", "remaining_time": "0:01:23"} +{"current_steps": 6120, "total_steps": 6125, "loss": 0.025374814867973328, "lr": 1.2234826143875388e-11, "epoch": 0.9991836734693877, "percentage": 99.92, "elapsed_time": "9:29:56", "remaining_time": "0:00:27"} +{"current_steps": 6125, "total_steps": 6125, "epoch": 1.0, "percentage": 100.0, "elapsed_time": "9:30:24", "remaining_time": "0:00:00"} diff --git a/trainer_state.json b/trainer_state.json new file mode 100644 index 0000000..411fe85 --- /dev/null +++ b/trainer_state.json @@ -0,0 +1,4597 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.0, + "eval_steps": 200, + "global_step": 6125, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0016326530612244899, + "grad_norm": 7.955283164978027, + "learning_rate": 4.5000000000000003e-07, + "loss": 4.391213607788086, + "step": 10 + }, + { + "epoch": 0.0032653061224489797, + "grad_norm": 2.4868686199188232, + "learning_rate": 9.500000000000001e-07, + "loss": 1.9876020431518555, + "step": 20 + }, + { + "epoch": 0.004897959183673469, + "grad_norm": 0.25438952445983887, + "learning_rate": 1.45e-06, + "loss": 0.34127225875854494, + "step": 30 + }, + { + "epoch": 0.006530612244897959, + "grad_norm": 0.13044093549251556, + "learning_rate": 1.9500000000000004e-06, + "loss": 0.23643033504486083, + "step": 40 + }, + { + "epoch": 0.00816326530612245, + "grad_norm": 0.17984068393707275, + "learning_rate": 2.4500000000000003e-06, + "loss": 0.23313720226287843, + "step": 50 + }, + { + "epoch": 0.009795918367346938, + "grad_norm": 0.15249308943748474, + "learning_rate": 2.95e-06, + "loss": 0.22351114749908446, + "step": 60 + }, + { + "epoch": 0.011428571428571429, + "grad_norm": 0.07123949378728867, + "learning_rate": 3.45e-06, + "loss": 0.21615636348724365, + "step": 70 + }, + { + "epoch": 0.013061224489795919, + "grad_norm": 0.21456502377986908, + "learning_rate": 3.95e-06, + "loss": 0.21677422523498535, + "step": 80 + }, + { + "epoch": 0.014693877551020407, + "grad_norm": 0.22391553223133087, + "learning_rate": 4.450000000000001e-06, + "loss": 0.21422204971313477, + "step": 90 + }, + { + "epoch": 0.0163265306122449, + "grad_norm": 0.1745220273733139, + "learning_rate": 4.95e-06, + "loss": 0.19350455999374389, + "step": 100 + }, + { + "epoch": 0.017959183673469388, + "grad_norm": 0.18820920586585999, + "learning_rate": 4.9999724716692435e-06, + "loss": 0.1981007933616638, + "step": 110 + }, + { + "epoch": 0.019591836734693877, + "grad_norm": 0.21733109652996063, + "learning_rate": 4.9998773127856985e-06, + "loss": 0.16994532346725463, + "step": 120 + }, + { + "epoch": 0.02122448979591837, + "grad_norm": 0.40475547313690186, + "learning_rate": 4.999714186080037e-06, + "loss": 0.1539763927459717, + "step": 130 + }, + { + "epoch": 0.022857142857142857, + "grad_norm": 0.3676598370075226, + "learning_rate": 4.999483095987423e-06, + "loss": 0.16321685314178466, + "step": 140 + }, + { + "epoch": 0.024489795918367346, + "grad_norm": 0.19484871625900269, + "learning_rate": 4.9991840487908426e-06, + "loss": 0.16523454189300538, + "step": 150 + }, + { + "epoch": 0.026122448979591838, + "grad_norm": 0.21928830444812775, + "learning_rate": 4.998817052620928e-06, + "loss": 0.14424897432327272, + "step": 160 + }, + { + "epoch": 0.027755102040816326, + "grad_norm": 0.2790975868701935, + "learning_rate": 4.9983821174557415e-06, + "loss": 0.13810645341873168, + "step": 170 + }, + { + "epoch": 0.029387755102040815, + "grad_norm": 0.21702228486537933, + "learning_rate": 4.997879255120501e-06, + "loss": 0.1318126440048218, + "step": 180 + }, + { + "epoch": 0.031020408163265307, + "grad_norm": 0.16967810690402985, + "learning_rate": 4.997308479287263e-06, + "loss": 0.14249314069747926, + "step": 190 + }, + { + "epoch": 0.0326530612244898, + "grad_norm": 0.14007994532585144, + "learning_rate": 4.996669805474542e-06, + "loss": 0.12708702087402343, + "step": 200 + }, + { + "epoch": 0.0326530612244898, + "eval_accuracy": 0.9408333333333335, + "eval_loss": 0.12691152095794678, + "eval_runtime": 156.68, + "eval_samples_per_second": 25.53, + "eval_steps_per_second": 25.53, + "step": 200 + }, + { + "epoch": 0.03428571428571429, + "grad_norm": 0.2674005627632141, + "learning_rate": 4.9959632510469e-06, + "loss": 0.11019121408462525, + "step": 210 + }, + { + "epoch": 0.035918367346938776, + "grad_norm": 0.12073638290166855, + "learning_rate": 4.995188835214465e-06, + "loss": 0.12032746076583863, + "step": 220 + }, + { + "epoch": 0.037551020408163265, + "grad_norm": 0.11186162382364273, + "learning_rate": 4.994346579032413e-06, + "loss": 0.12122929096221924, + "step": 230 + }, + { + "epoch": 0.03918367346938775, + "grad_norm": 0.09840800613164902, + "learning_rate": 4.993436505400396e-06, + "loss": 0.11416752338409424, + "step": 240 + }, + { + "epoch": 0.04081632653061224, + "grad_norm": 0.19076374173164368, + "learning_rate": 4.992458639061916e-06, + "loss": 0.11376930475234985, + "step": 250 + }, + { + "epoch": 0.04244897959183674, + "grad_norm": 0.14405065774917603, + "learning_rate": 4.991413006603657e-06, + "loss": 0.10997685194015502, + "step": 260 + }, + { + "epoch": 0.044081632653061226, + "grad_norm": 0.23844055831432343, + "learning_rate": 4.990299636454759e-06, + "loss": 0.1000083327293396, + "step": 270 + }, + { + "epoch": 0.045714285714285714, + "grad_norm": 0.23128052055835724, + "learning_rate": 4.989118558886043e-06, + "loss": 0.1243711233139038, + "step": 280 + }, + { + "epoch": 0.0473469387755102, + "grad_norm": 0.19849570095539093, + "learning_rate": 4.9878698060091955e-06, + "loss": 0.11728224754333497, + "step": 290 + }, + { + "epoch": 0.04897959183673469, + "grad_norm": 0.13039851188659668, + "learning_rate": 4.986553411775886e-06, + "loss": 0.09353847503662109, + "step": 300 + }, + { + "epoch": 0.05061224489795919, + "grad_norm": 0.19887588918209076, + "learning_rate": 4.985169411976852e-06, + "loss": 0.11291944980621338, + "step": 310 + }, + { + "epoch": 0.052244897959183675, + "grad_norm": 0.25038304924964905, + "learning_rate": 4.98371784424092e-06, + "loss": 0.11149277687072753, + "step": 320 + }, + { + "epoch": 0.053877551020408164, + "grad_norm": 0.238360196352005, + "learning_rate": 4.982198748033986e-06, + "loss": 0.11344953775405883, + "step": 330 + }, + { + "epoch": 0.05551020408163265, + "grad_norm": 0.13686658442020416, + "learning_rate": 4.980612164657943e-06, + "loss": 0.09844974875450134, + "step": 340 + }, + { + "epoch": 0.05714285714285714, + "grad_norm": 0.0829666405916214, + "learning_rate": 4.978958137249551e-06, + "loss": 0.10898438692092896, + "step": 350 + }, + { + "epoch": 0.05877551020408163, + "grad_norm": 0.10389993339776993, + "learning_rate": 4.977236710779277e-06, + "loss": 0.09371156096458436, + "step": 360 + }, + { + "epoch": 0.060408163265306125, + "grad_norm": 0.18182680010795593, + "learning_rate": 4.9754479320500584e-06, + "loss": 0.1049533724784851, + "step": 370 + }, + { + "epoch": 0.062040816326530614, + "grad_norm": 0.14488115906715393, + "learning_rate": 4.973591849696041e-06, + "loss": 0.10510741472244263, + "step": 380 + }, + { + "epoch": 0.0636734693877551, + "grad_norm": 0.14816738665103912, + "learning_rate": 4.971668514181252e-06, + "loss": 0.10956356525421143, + "step": 390 + }, + { + "epoch": 0.0653061224489796, + "grad_norm": 0.20584072172641754, + "learning_rate": 4.969677977798225e-06, + "loss": 0.11344989538192748, + "step": 400 + }, + { + "epoch": 0.0653061224489796, + "eval_accuracy": 0.9496250000000002, + "eval_loss": 0.11409253627061844, + "eval_runtime": 165.7682, + "eval_samples_per_second": 24.13, + "eval_steps_per_second": 24.13, + "step": 400 + }, + { + "epoch": 0.06693877551020408, + "grad_norm": 0.1342787891626358, + "learning_rate": 4.967620294666589e-06, + "loss": 0.09949527978897095, + "step": 410 + }, + { + "epoch": 0.06857142857142857, + "grad_norm": 0.17336596548557281, + "learning_rate": 4.965495520731583e-06, + "loss": 0.10606559514999389, + "step": 420 + }, + { + "epoch": 0.07020408163265306, + "grad_norm": 0.19437003135681152, + "learning_rate": 4.963303713762547e-06, + "loss": 0.11677042245864869, + "step": 430 + }, + { + "epoch": 0.07183673469387755, + "grad_norm": 0.12076073884963989, + "learning_rate": 4.961044933351342e-06, + "loss": 0.1019243597984314, + "step": 440 + }, + { + "epoch": 0.07346938775510205, + "grad_norm": 0.10272067785263062, + "learning_rate": 4.958719240910737e-06, + "loss": 0.10125980377197266, + "step": 450 + }, + { + "epoch": 0.07510204081632653, + "grad_norm": 0.1080622524023056, + "learning_rate": 4.956326699672736e-06, + "loss": 0.09888594150543213, + "step": 460 + }, + { + "epoch": 0.07673469387755102, + "grad_norm": 0.1307847797870636, + "learning_rate": 4.953867374686856e-06, + "loss": 0.10753540992736817, + "step": 470 + }, + { + "epoch": 0.0783673469387755, + "grad_norm": 0.2128073275089264, + "learning_rate": 4.951341332818365e-06, + "loss": 0.11452326774597169, + "step": 480 + }, + { + "epoch": 0.08, + "grad_norm": 0.21006478369235992, + "learning_rate": 4.9487486427464594e-06, + "loss": 0.09555010795593262, + "step": 490 + }, + { + "epoch": 0.08163265306122448, + "grad_norm": 0.24889807403087616, + "learning_rate": 4.946089374962396e-06, + "loss": 0.09771155118942261, + "step": 500 + }, + { + "epoch": 0.08326530612244898, + "grad_norm": 0.13042858242988586, + "learning_rate": 4.943363601767582e-06, + "loss": 0.10659778118133545, + "step": 510 + }, + { + "epoch": 0.08489795918367347, + "grad_norm": 0.2210501879453659, + "learning_rate": 4.940571397271598e-06, + "loss": 0.10161921977996827, + "step": 520 + }, + { + "epoch": 0.08653061224489796, + "grad_norm": 0.11963305622339249, + "learning_rate": 4.937712837390194e-06, + "loss": 0.09155921936035157, + "step": 530 + }, + { + "epoch": 0.08816326530612245, + "grad_norm": 0.07435746490955353, + "learning_rate": 4.93478799984322e-06, + "loss": 0.11164016723632812, + "step": 540 + }, + { + "epoch": 0.08979591836734693, + "grad_norm": 0.16348940134048462, + "learning_rate": 4.9317969641525154e-06, + "loss": 0.10738091468811035, + "step": 550 + }, + { + "epoch": 0.09142857142857143, + "grad_norm": 0.12221129983663559, + "learning_rate": 4.928739811639741e-06, + "loss": 0.09491395950317383, + "step": 560 + }, + { + "epoch": 0.09306122448979592, + "grad_norm": 0.1107218861579895, + "learning_rate": 4.9256166254241775e-06, + "loss": 0.08408294320106506, + "step": 570 + }, + { + "epoch": 0.0946938775510204, + "grad_norm": 0.207329660654068, + "learning_rate": 4.922427490420457e-06, + "loss": 0.10272791385650634, + "step": 580 + }, + { + "epoch": 0.0963265306122449, + "grad_norm": 0.14269684255123138, + "learning_rate": 4.919172493336259e-06, + "loss": 0.10189608335494996, + "step": 590 + }, + { + "epoch": 0.09795918367346938, + "grad_norm": 0.10970286279916763, + "learning_rate": 4.915851722669952e-06, + "loss": 0.09506216645240784, + "step": 600 + }, + { + "epoch": 0.09795918367346938, + "eval_accuracy": 0.9561666666666667, + "eval_loss": 0.09913250058889389, + "eval_runtime": 158.3659, + "eval_samples_per_second": 25.258, + "eval_steps_per_second": 25.258, + "step": 600 + }, + { + "epoch": 0.09959183673469388, + "grad_norm": 0.11382903158664703, + "learning_rate": 4.912465268708186e-06, + "loss": 0.09665942788124085, + "step": 610 + }, + { + "epoch": 0.10122448979591837, + "grad_norm": 0.24785904586315155, + "learning_rate": 4.90901322352344e-06, + "loss": 0.10539349317550659, + "step": 620 + }, + { + "epoch": 0.10285714285714286, + "grad_norm": 0.18585072457790375, + "learning_rate": 4.905495680971518e-06, + "loss": 0.10907092094421386, + "step": 630 + }, + { + "epoch": 0.10448979591836735, + "grad_norm": 0.16385921835899353, + "learning_rate": 4.901912736688995e-06, + "loss": 0.09973281621932983, + "step": 640 + }, + { + "epoch": 0.10612244897959183, + "grad_norm": 0.10820458829402924, + "learning_rate": 4.898264488090619e-06, + "loss": 0.09869267344474793, + "step": 650 + }, + { + "epoch": 0.10775510204081633, + "grad_norm": 0.18020115792751312, + "learning_rate": 4.894551034366665e-06, + "loss": 0.10499387979507446, + "step": 660 + }, + { + "epoch": 0.10938775510204081, + "grad_norm": 0.10356947779655457, + "learning_rate": 4.890772476480229e-06, + "loss": 0.10153095722198487, + "step": 670 + }, + { + "epoch": 0.1110204081632653, + "grad_norm": 0.08888879418373108, + "learning_rate": 4.886928917164496e-06, + "loss": 0.08613221645355225, + "step": 680 + }, + { + "epoch": 0.1126530612244898, + "grad_norm": 0.11978956311941147, + "learning_rate": 4.8830204609199335e-06, + "loss": 0.09545036554336547, + "step": 690 + }, + { + "epoch": 0.11428571428571428, + "grad_norm": 0.11924679577350616, + "learning_rate": 4.8790472140114635e-06, + "loss": 0.0950542688369751, + "step": 700 + }, + { + "epoch": 0.11591836734693878, + "grad_norm": 0.0637463852763176, + "learning_rate": 4.875009284465561e-06, + "loss": 0.10127921104431152, + "step": 710 + }, + { + "epoch": 0.11755102040816326, + "grad_norm": 0.0894392654299736, + "learning_rate": 4.8709067820673235e-06, + "loss": 0.09835844039916992, + "step": 720 + }, + { + "epoch": 0.11918367346938775, + "grad_norm": 0.06312194466590881, + "learning_rate": 4.866739818357489e-06, + "loss": 0.08583117127418519, + "step": 730 + }, + { + "epoch": 0.12081632653061225, + "grad_norm": 0.1375141739845276, + "learning_rate": 4.862508506629394e-06, + "loss": 0.08735222220420838, + "step": 740 + }, + { + "epoch": 0.12244897959183673, + "grad_norm": 0.12715794146060944, + "learning_rate": 4.858212961925902e-06, + "loss": 0.09324318170547485, + "step": 750 + }, + { + "epoch": 0.12408163265306123, + "grad_norm": 0.14359286427497864, + "learning_rate": 4.8538533010362735e-06, + "loss": 0.09482367634773255, + "step": 760 + }, + { + "epoch": 0.12571428571428572, + "grad_norm": 0.1228424683213234, + "learning_rate": 4.849429642492985e-06, + "loss": 0.09595230221748352, + "step": 770 + }, + { + "epoch": 0.1273469387755102, + "grad_norm": 0.08908221125602722, + "learning_rate": 4.844942106568517e-06, + "loss": 0.08924583196640015, + "step": 780 + }, + { + "epoch": 0.12897959183673469, + "grad_norm": 0.15417414903640747, + "learning_rate": 4.84039081527207e-06, + "loss": 0.09454164505004883, + "step": 790 + }, + { + "epoch": 0.1306122448979592, + "grad_norm": 0.1259608417749405, + "learning_rate": 4.835775892346264e-06, + "loss": 0.08659166097640991, + "step": 800 + }, + { + "epoch": 0.1306122448979592, + "eval_accuracy": 0.9564166666666667, + "eval_loss": 0.09297937154769897, + "eval_runtime": 157.5491, + "eval_samples_per_second": 25.389, + "eval_steps_per_second": 25.389, + "step": 800 + }, + { + "epoch": 0.13224489795918368, + "grad_norm": 0.15664513409137726, + "learning_rate": 4.8310974632637584e-06, + "loss": 0.09094155430793763, + "step": 810 + }, + { + "epoch": 0.13387755102040816, + "grad_norm": 0.12366823107004166, + "learning_rate": 4.82635565522385e-06, + "loss": 0.09823386669158936, + "step": 820 + }, + { + "epoch": 0.13551020408163264, + "grad_norm": 0.09367755055427551, + "learning_rate": 4.821550597149012e-06, + "loss": 0.09080367088317871, + "step": 830 + }, + { + "epoch": 0.13714285714285715, + "grad_norm": 0.15230220556259155, + "learning_rate": 4.816682419681389e-06, + "loss": 0.09046205282211303, + "step": 840 + }, + { + "epoch": 0.13877551020408163, + "grad_norm": 0.09214162826538086, + "learning_rate": 4.811751255179244e-06, + "loss": 0.09502801299095154, + "step": 850 + }, + { + "epoch": 0.1404081632653061, + "grad_norm": 0.09514431655406952, + "learning_rate": 4.806757237713361e-06, + "loss": 0.09261927604675294, + "step": 860 + }, + { + "epoch": 0.14204081632653062, + "grad_norm": 0.13052736222743988, + "learning_rate": 4.8017005030634e-06, + "loss": 0.08243635892868043, + "step": 870 + }, + { + "epoch": 0.1436734693877551, + "grad_norm": 0.05511025711894035, + "learning_rate": 4.796581188714205e-06, + "loss": 0.10203326940536499, + "step": 880 + }, + { + "epoch": 0.14530612244897959, + "grad_norm": 0.09896261245012283, + "learning_rate": 4.7913994338520655e-06, + "loss": 0.08544076681137085, + "step": 890 + }, + { + "epoch": 0.1469387755102041, + "grad_norm": 0.18726249039173126, + "learning_rate": 4.786155379360932e-06, + "loss": 0.0944068193435669, + "step": 900 + }, + { + "epoch": 0.14857142857142858, + "grad_norm": 0.09349548071622849, + "learning_rate": 4.7808491678185855e-06, + "loss": 0.0877715528011322, + "step": 910 + }, + { + "epoch": 0.15020408163265306, + "grad_norm": 0.13616853952407837, + "learning_rate": 4.775480943492764e-06, + "loss": 0.09196346998214722, + "step": 920 + }, + { + "epoch": 0.15183673469387754, + "grad_norm": 0.061041463166475296, + "learning_rate": 4.770050852337236e-06, + "loss": 0.09890784621238709, + "step": 930 + }, + { + "epoch": 0.15346938775510205, + "grad_norm": 0.09928455948829651, + "learning_rate": 4.764559041987836e-06, + "loss": 0.09260086417198181, + "step": 940 + }, + { + "epoch": 0.15510204081632653, + "grad_norm": 0.10996410250663757, + "learning_rate": 4.759005661758446e-06, + "loss": 0.09412651658058166, + "step": 950 + }, + { + "epoch": 0.156734693877551, + "grad_norm": 0.09731809049844742, + "learning_rate": 4.753390862636939e-06, + "loss": 0.09459404945373535, + "step": 960 + }, + { + "epoch": 0.15836734693877552, + "grad_norm": 0.12398040294647217, + "learning_rate": 4.747714797281076e-06, + "loss": 0.09960640072822571, + "step": 970 + }, + { + "epoch": 0.16, + "grad_norm": 0.14135490357875824, + "learning_rate": 4.741977620014347e-06, + "loss": 0.09338371157646179, + "step": 980 + }, + { + "epoch": 0.16163265306122448, + "grad_norm": 0.09384343028068542, + "learning_rate": 4.736179486821788e-06, + "loss": 0.08515415191650391, + "step": 990 + }, + { + "epoch": 0.16326530612244897, + "grad_norm": 0.10298261791467667, + "learning_rate": 4.730320555345728e-06, + "loss": 0.09090787768363953, + "step": 1000 + }, + { + "epoch": 0.16326530612244897, + "eval_accuracy": 0.9607916666666667, + "eval_loss": 0.08832322806119919, + "eval_runtime": 166.8275, + "eval_samples_per_second": 23.977, + "eval_steps_per_second": 23.977, + "step": 1000 + }, + { + "epoch": 0.16489795918367348, + "grad_norm": 0.05896759778261185, + "learning_rate": 4.724400984881508e-06, + "loss": 0.08005568385124207, + "step": 1010 + }, + { + "epoch": 0.16653061224489796, + "grad_norm": 0.07907038927078247, + "learning_rate": 4.7184209363731546e-06, + "loss": 0.08133251667022705, + "step": 1020 + }, + { + "epoch": 0.16816326530612244, + "grad_norm": 0.07444760203361511, + "learning_rate": 4.7123805724089935e-06, + "loss": 0.08690541982650757, + "step": 1030 + }, + { + "epoch": 0.16979591836734695, + "grad_norm": 0.16517433524131775, + "learning_rate": 4.706280057217239e-06, + "loss": 0.09946454763412475, + "step": 1040 + }, + { + "epoch": 0.17142857142857143, + "grad_norm": 0.18107500672340393, + "learning_rate": 4.700119556661521e-06, + "loss": 0.08886862993240356, + "step": 1050 + }, + { + "epoch": 0.1730612244897959, + "grad_norm": 0.15386350452899933, + "learning_rate": 4.693899238236383e-06, + "loss": 0.09117826819419861, + "step": 1060 + }, + { + "epoch": 0.17469387755102042, + "grad_norm": 0.1098460704088211, + "learning_rate": 4.687619271062725e-06, + "loss": 0.09286551475524903, + "step": 1070 + }, + { + "epoch": 0.1763265306122449, + "grad_norm": 0.09872590750455856, + "learning_rate": 4.681279825883201e-06, + "loss": 0.09102073907852173, + "step": 1080 + }, + { + "epoch": 0.17795918367346938, + "grad_norm": 0.12001907080411911, + "learning_rate": 4.674881075057581e-06, + "loss": 0.09481253027915955, + "step": 1090 + }, + { + "epoch": 0.17959183673469387, + "grad_norm": 0.13000014424324036, + "learning_rate": 4.668423192558069e-06, + "loss": 0.08999797105789184, + "step": 1100 + }, + { + "epoch": 0.18122448979591838, + "grad_norm": 0.07537361979484558, + "learning_rate": 4.661906353964563e-06, + "loss": 0.07345216870307922, + "step": 1110 + }, + { + "epoch": 0.18285714285714286, + "grad_norm": 0.10346294194459915, + "learning_rate": 4.65533073645989e-06, + "loss": 0.08096543550491334, + "step": 1120 + }, + { + "epoch": 0.18448979591836734, + "grad_norm": 0.12576258182525635, + "learning_rate": 4.6486965188249825e-06, + "loss": 0.08680164813995361, + "step": 1130 + }, + { + "epoch": 0.18612244897959185, + "grad_norm": 0.17314419150352478, + "learning_rate": 4.642003881434024e-06, + "loss": 0.08470291495323182, + "step": 1140 + }, + { + "epoch": 0.18775510204081633, + "grad_norm": 0.1576225906610489, + "learning_rate": 4.635253006249536e-06, + "loss": 0.10852066278457642, + "step": 1150 + }, + { + "epoch": 0.1893877551020408, + "grad_norm": 0.15372557938098907, + "learning_rate": 4.628444076817444e-06, + "loss": 0.1022921085357666, + "step": 1160 + }, + { + "epoch": 0.1910204081632653, + "grad_norm": 0.15021173655986786, + "learning_rate": 4.621577278262072e-06, + "loss": 0.09645735621452331, + "step": 1170 + }, + { + "epoch": 0.1926530612244898, + "grad_norm": 0.07714739441871643, + "learning_rate": 4.614652797281118e-06, + "loss": 0.08636590242385864, + "step": 1180 + }, + { + "epoch": 0.19428571428571428, + "grad_norm": 0.17956240475177765, + "learning_rate": 4.6076708221405806e-06, + "loss": 0.09293715357780456, + "step": 1190 + }, + { + "epoch": 0.19591836734693877, + "grad_norm": 0.13412988185882568, + "learning_rate": 4.600631542669633e-06, + "loss": 0.09799563884735107, + "step": 1200 + }, + { + "epoch": 0.19591836734693877, + "eval_accuracy": 0.963, + "eval_loss": 0.08563050627708435, + "eval_runtime": 171.1854, + "eval_samples_per_second": 23.366, + "eval_steps_per_second": 23.366, + "step": 1200 + }, + { + "epoch": 0.19755102040816327, + "grad_norm": 0.1682359278202057, + "learning_rate": 4.593535150255463e-06, + "loss": 0.08345703482627868, + "step": 1210 + }, + { + "epoch": 0.19918367346938776, + "grad_norm": 0.09281504899263382, + "learning_rate": 4.586381837838076e-06, + "loss": 0.08114275336265564, + "step": 1220 + }, + { + "epoch": 0.20081632653061224, + "grad_norm": 0.12355420738458633, + "learning_rate": 4.579171799905044e-06, + "loss": 0.0780484676361084, + "step": 1230 + }, + { + "epoch": 0.20244897959183675, + "grad_norm": 0.12121636420488358, + "learning_rate": 4.571905232486214e-06, + "loss": 0.07816853523254394, + "step": 1240 + }, + { + "epoch": 0.20408163265306123, + "grad_norm": 0.09326840937137604, + "learning_rate": 4.56458233314839e-06, + "loss": 0.08001847267150879, + "step": 1250 + }, + { + "epoch": 0.2057142857142857, + "grad_norm": 0.11423090845346451, + "learning_rate": 4.557203300989948e-06, + "loss": 0.08720387816429138, + "step": 1260 + }, + { + "epoch": 0.2073469387755102, + "grad_norm": 0.124808169901371, + "learning_rate": 4.549768336635432e-06, + "loss": 0.08812607526779175, + "step": 1270 + }, + { + "epoch": 0.2089795918367347, + "grad_norm": 0.08311110734939575, + "learning_rate": 4.542277642230096e-06, + "loss": 0.0880801498889923, + "step": 1280 + }, + { + "epoch": 0.21061224489795918, + "grad_norm": 0.1179559975862503, + "learning_rate": 4.53473142143441e-06, + "loss": 0.09044411182403564, + "step": 1290 + }, + { + "epoch": 0.21224489795918366, + "grad_norm": 0.13122901320457458, + "learning_rate": 4.527129879418519e-06, + "loss": 0.06792758107185363, + "step": 1300 + }, + { + "epoch": 0.21387755102040817, + "grad_norm": 0.10990814864635468, + "learning_rate": 4.519473222856667e-06, + "loss": 0.07861824035644531, + "step": 1310 + }, + { + "epoch": 0.21551020408163266, + "grad_norm": 0.14613105356693268, + "learning_rate": 4.51176165992158e-06, + "loss": 0.08266589641571045, + "step": 1320 + }, + { + "epoch": 0.21714285714285714, + "grad_norm": 0.07217947393655777, + "learning_rate": 4.503995400278801e-06, + "loss": 0.08272708654403686, + "step": 1330 + }, + { + "epoch": 0.21877551020408162, + "grad_norm": 0.12690335512161255, + "learning_rate": 4.4961746550809945e-06, + "loss": 0.08644292950630188, + "step": 1340 + }, + { + "epoch": 0.22040816326530613, + "grad_norm": 0.11895330995321274, + "learning_rate": 4.488299636962203e-06, + "loss": 0.08812489509582519, + "step": 1350 + }, + { + "epoch": 0.2220408163265306, + "grad_norm": 0.08727742731571198, + "learning_rate": 4.4803705600320655e-06, + "loss": 0.08752686977386474, + "step": 1360 + }, + { + "epoch": 0.2236734693877551, + "grad_norm": 0.1101253554224968, + "learning_rate": 4.472387639869998e-06, + "loss": 0.08365487456321716, + "step": 1370 + }, + { + "epoch": 0.2253061224489796, + "grad_norm": 0.0881471335887909, + "learning_rate": 4.464351093519332e-06, + "loss": 0.08101181387901306, + "step": 1380 + }, + { + "epoch": 0.22693877551020408, + "grad_norm": 0.1676328480243683, + "learning_rate": 4.45626113948141e-06, + "loss": 0.08873040080070496, + "step": 1390 + }, + { + "epoch": 0.22857142857142856, + "grad_norm": 0.13493730127811432, + "learning_rate": 4.44811799770965e-06, + "loss": 0.08360225558280945, + "step": 1400 + }, + { + "epoch": 0.22857142857142856, + "eval_accuracy": 0.961125, + "eval_loss": 0.0838993564248085, + "eval_runtime": 159.3888, + "eval_samples_per_second": 25.096, + "eval_steps_per_second": 25.096, + "step": 1400 + }, + { + "epoch": 0.23020408163265307, + "grad_norm": 0.18087805807590485, + "learning_rate": 4.4399218896035615e-06, + "loss": 0.07942168116569519, + "step": 1410 + }, + { + "epoch": 0.23183673469387756, + "grad_norm": 0.15980401635169983, + "learning_rate": 4.431673038002726e-06, + "loss": 0.07959137558937072, + "step": 1420 + }, + { + "epoch": 0.23346938775510204, + "grad_norm": 0.09756626188755035, + "learning_rate": 4.423371667180743e-06, + "loss": 0.08047830462455749, + "step": 1430 + }, + { + "epoch": 0.23510204081632652, + "grad_norm": 0.11922340840101242, + "learning_rate": 4.415018002839127e-06, + "loss": 0.07832802534103393, + "step": 1440 + }, + { + "epoch": 0.23673469387755103, + "grad_norm": 0.09443783015012741, + "learning_rate": 4.40661227210117e-06, + "loss": 0.07132846713066102, + "step": 1450 + }, + { + "epoch": 0.2383673469387755, + "grad_norm": 0.15397194027900696, + "learning_rate": 4.398154703505776e-06, + "loss": 0.09222683906555176, + "step": 1460 + }, + { + "epoch": 0.24, + "grad_norm": 0.06159692630171776, + "learning_rate": 4.389645527001234e-06, + "loss": 0.07937920093536377, + "step": 1470 + }, + { + "epoch": 0.2416326530612245, + "grad_norm": 0.11855911463499069, + "learning_rate": 4.381084973938976e-06, + "loss": 0.07650941014289855, + "step": 1480 + }, + { + "epoch": 0.24326530612244898, + "grad_norm": 0.1305486559867859, + "learning_rate": 4.372473277067285e-06, + "loss": 0.07730157375335693, + "step": 1490 + }, + { + "epoch": 0.24489795918367346, + "grad_norm": 0.09036920964717865, + "learning_rate": 4.363810670524962e-06, + "loss": 0.07192737460136414, + "step": 1500 + }, + { + "epoch": 0.24653061224489795, + "grad_norm": 0.0936962217092514, + "learning_rate": 4.355097389834966e-06, + "loss": 0.09040454030036926, + "step": 1510 + }, + { + "epoch": 0.24816326530612245, + "grad_norm": 0.0778910219669342, + "learning_rate": 4.346333671898007e-06, + "loss": 0.06948431730270385, + "step": 1520 + }, + { + "epoch": 0.24979591836734694, + "grad_norm": 0.0979679748415947, + "learning_rate": 4.337519754986108e-06, + "loss": 0.07317689657211304, + "step": 1530 + }, + { + "epoch": 0.25142857142857145, + "grad_norm": 0.15620048344135284, + "learning_rate": 4.328655878736123e-06, + "loss": 0.09126524925231934, + "step": 1540 + }, + { + "epoch": 0.2530612244897959, + "grad_norm": 0.09080824255943298, + "learning_rate": 4.319742284143226e-06, + "loss": 0.09065211415290833, + "step": 1550 + }, + { + "epoch": 0.2546938775510204, + "grad_norm": 0.046968940645456314, + "learning_rate": 4.310779213554353e-06, + "loss": 0.0718291699886322, + "step": 1560 + }, + { + "epoch": 0.2563265306122449, + "grad_norm": 0.11479081958532333, + "learning_rate": 4.301766910661619e-06, + "loss": 0.07913097143173217, + "step": 1570 + }, + { + "epoch": 0.25795918367346937, + "grad_norm": 0.1042279377579689, + "learning_rate": 4.292705620495689e-06, + "loss": 0.06828063130378723, + "step": 1580 + }, + { + "epoch": 0.2595918367346939, + "grad_norm": 0.17096509039402008, + "learning_rate": 4.283595589419115e-06, + "loss": 0.07042163014411926, + "step": 1590 + }, + { + "epoch": 0.2612244897959184, + "grad_norm": 0.10892639309167862, + "learning_rate": 4.274437065119643e-06, + "loss": 0.07090482115745544, + "step": 1600 + }, + { + "epoch": 0.2612244897959184, + "eval_accuracy": 0.964375, + "eval_loss": 0.08112655580043793, + "eval_runtime": 158.0164, + "eval_samples_per_second": 25.314, + "eval_steps_per_second": 25.314, + "step": 1600 + }, + { + "epoch": 0.26285714285714284, + "grad_norm": 0.09731441736221313, + "learning_rate": 4.265230296603471e-06, + "loss": 0.08803416490554809, + "step": 1610 + }, + { + "epoch": 0.26448979591836735, + "grad_norm": 0.0956069752573967, + "learning_rate": 4.255975534188488e-06, + "loss": 0.0734250545501709, + "step": 1620 + }, + { + "epoch": 0.26612244897959186, + "grad_norm": 0.14361435174942017, + "learning_rate": 4.246673029497458e-06, + "loss": 0.07703657150268554, + "step": 1630 + }, + { + "epoch": 0.2677551020408163, + "grad_norm": 0.08011971414089203, + "learning_rate": 4.237323035451188e-06, + "loss": 0.08410758376121522, + "step": 1640 + }, + { + "epoch": 0.2693877551020408, + "grad_norm": 0.1244175136089325, + "learning_rate": 4.2279258062616455e-06, + "loss": 0.0808764636516571, + "step": 1650 + }, + { + "epoch": 0.2710204081632653, + "grad_norm": 0.11162611097097397, + "learning_rate": 4.21848159742505e-06, + "loss": 0.07636536359786987, + "step": 1660 + }, + { + "epoch": 0.2726530612244898, + "grad_norm": 0.16819386184215546, + "learning_rate": 4.208990665714927e-06, + "loss": 0.07789689898490906, + "step": 1670 + }, + { + "epoch": 0.2742857142857143, + "grad_norm": 0.07522564381361008, + "learning_rate": 4.19945326917512e-06, + "loss": 0.07442367672920228, + "step": 1680 + }, + { + "epoch": 0.27591836734693875, + "grad_norm": 0.08352494984865189, + "learning_rate": 4.1898696671127855e-06, + "loss": 0.07646024227142334, + "step": 1690 + }, + { + "epoch": 0.27755102040816326, + "grad_norm": 0.12580762803554535, + "learning_rate": 4.180240120091333e-06, + "loss": 0.07484800815582275, + "step": 1700 + }, + { + "epoch": 0.27918367346938777, + "grad_norm": 0.14690619707107544, + "learning_rate": 4.170564889923348e-06, + "loss": 0.07452926635742188, + "step": 1710 + }, + { + "epoch": 0.2808163265306122, + "grad_norm": 0.08760543167591095, + "learning_rate": 4.1608442396634675e-06, + "loss": 0.07579211592674255, + "step": 1720 + }, + { + "epoch": 0.28244897959183674, + "grad_norm": 0.05263754725456238, + "learning_rate": 4.151078433601234e-06, + "loss": 0.08277950882911682, + "step": 1730 + }, + { + "epoch": 0.28408163265306124, + "grad_norm": 0.12929993867874146, + "learning_rate": 4.141267737253905e-06, + "loss": 0.07354526519775391, + "step": 1740 + }, + { + "epoch": 0.2857142857142857, + "grad_norm": 0.1444639414548874, + "learning_rate": 4.1314124173592356e-06, + "loss": 0.07377716898918152, + "step": 1750 + }, + { + "epoch": 0.2873469387755102, + "grad_norm": 0.15250442922115326, + "learning_rate": 4.121512741868229e-06, + "loss": 0.0802634596824646, + "step": 1760 + }, + { + "epoch": 0.2889795918367347, + "grad_norm": 0.14867356419563293, + "learning_rate": 4.111568979937845e-06, + "loss": 0.08141970038414001, + "step": 1770 + }, + { + "epoch": 0.29061224489795917, + "grad_norm": 0.1100233644247055, + "learning_rate": 4.10158140192369e-06, + "loss": 0.07370007634162903, + "step": 1780 + }, + { + "epoch": 0.2922448979591837, + "grad_norm": 0.10663364082574844, + "learning_rate": 4.091550279372659e-06, + "loss": 0.0754891276359558, + "step": 1790 + }, + { + "epoch": 0.2938775510204082, + "grad_norm": 0.07148909568786621, + "learning_rate": 4.081475885015558e-06, + "loss": 0.07567381858825684, + "step": 1800 + }, + { + "epoch": 0.2938775510204082, + "eval_accuracy": 0.9665416666666667, + "eval_loss": 0.07648575305938721, + "eval_runtime": 176.7261, + "eval_samples_per_second": 22.634, + "eval_steps_per_second": 22.634, + "step": 1800 + }, + { + "epoch": 0.29551020408163264, + "grad_norm": 0.07446923106908798, + "learning_rate": 4.071358492759686e-06, + "loss": 0.06706757545471191, + "step": 1810 + }, + { + "epoch": 0.29714285714285715, + "grad_norm": 0.144475519657135, + "learning_rate": 4.061198377681391e-06, + "loss": 0.07407833337783813, + "step": 1820 + }, + { + "epoch": 0.2987755102040816, + "grad_norm": 0.16540028154850006, + "learning_rate": 4.050995816018585e-06, + "loss": 0.07593105435371399, + "step": 1830 + }, + { + "epoch": 0.3004081632653061, + "grad_norm": 0.09237829595804214, + "learning_rate": 4.040751085163239e-06, + "loss": 0.08375502228736878, + "step": 1840 + }, + { + "epoch": 0.3020408163265306, + "grad_norm": 0.11202957481145859, + "learning_rate": 4.0304644636538435e-06, + "loss": 0.0692280650138855, + "step": 1850 + }, + { + "epoch": 0.3036734693877551, + "grad_norm": 0.16075952351093292, + "learning_rate": 4.020136231167824e-06, + "loss": 0.07824563384056091, + "step": 1860 + }, + { + "epoch": 0.3053061224489796, + "grad_norm": 0.12567447125911713, + "learning_rate": 4.0097666685139516e-06, + "loss": 0.07255304455757142, + "step": 1870 + }, + { + "epoch": 0.3069387755102041, + "grad_norm": 0.08912669867277145, + "learning_rate": 3.999356057624696e-06, + "loss": 0.06304961442947388, + "step": 1880 + }, + { + "epoch": 0.30857142857142855, + "grad_norm": 0.15736177563667297, + "learning_rate": 3.988904681548568e-06, + "loss": 0.07315694689750671, + "step": 1890 + }, + { + "epoch": 0.31020408163265306, + "grad_norm": 0.13224704563617706, + "learning_rate": 3.978412824442421e-06, + "loss": 0.07607222199440003, + "step": 1900 + }, + { + "epoch": 0.31183673469387757, + "grad_norm": 0.0942244678735733, + "learning_rate": 3.9678807715637245e-06, + "loss": 0.06320942640304565, + "step": 1910 + }, + { + "epoch": 0.313469387755102, + "grad_norm": 0.1153193786740303, + "learning_rate": 3.95730880926281e-06, + "loss": 0.07607839703559875, + "step": 1920 + }, + { + "epoch": 0.31510204081632653, + "grad_norm": 0.24464164674282074, + "learning_rate": 3.946697224975083e-06, + "loss": 0.07458690404891968, + "step": 1930 + }, + { + "epoch": 0.31673469387755104, + "grad_norm": 0.14584548771381378, + "learning_rate": 3.936046307213215e-06, + "loss": 0.07876120805740357, + "step": 1940 + }, + { + "epoch": 0.3183673469387755, + "grad_norm": 0.0880698636174202, + "learning_rate": 3.925356345559288e-06, + "loss": 0.07082821726799012, + "step": 1950 + }, + { + "epoch": 0.32, + "grad_norm": 0.08619305491447449, + "learning_rate": 3.914627630656931e-06, + "loss": 0.06648402214050293, + "step": 1960 + }, + { + "epoch": 0.3216326530612245, + "grad_norm": 0.1341329663991928, + "learning_rate": 3.903860454203413e-06, + "loss": 0.07000230550765991, + "step": 1970 + }, + { + "epoch": 0.32326530612244897, + "grad_norm": 0.1504107564687729, + "learning_rate": 3.8930551089417145e-06, + "loss": 0.06887786984443664, + "step": 1980 + }, + { + "epoch": 0.3248979591836735, + "grad_norm": 0.1262267827987671, + "learning_rate": 3.882211888652566e-06, + "loss": 0.08991954326629639, + "step": 1990 + }, + { + "epoch": 0.32653061224489793, + "grad_norm": 0.2052707076072693, + "learning_rate": 3.871331088146463e-06, + "loss": 0.07011507153511047, + "step": 2000 + }, + { + "epoch": 0.32653061224489793, + "eval_accuracy": 0.9654583333333334, + "eval_loss": 0.07913381606340408, + "eval_runtime": 161.4221, + "eval_samples_per_second": 24.78, + "eval_steps_per_second": 24.78, + "step": 2000 + }, + { + "epoch": 0.32816326530612244, + "grad_norm": 0.11643362790346146, + "learning_rate": 3.860413003255648e-06, + "loss": 0.08266008496284485, + "step": 2010 + }, + { + "epoch": 0.32979591836734695, + "grad_norm": 0.1472826600074768, + "learning_rate": 3.8494579308260695e-06, + "loss": 0.07043538689613342, + "step": 2020 + }, + { + "epoch": 0.3314285714285714, + "grad_norm": 0.10538945347070694, + "learning_rate": 3.838466168709312e-06, + "loss": 0.06780247688293457, + "step": 2030 + }, + { + "epoch": 0.3330612244897959, + "grad_norm": 0.14366035163402557, + "learning_rate": 3.827438015754491e-06, + "loss": 0.06872215867042542, + "step": 2040 + }, + { + "epoch": 0.3346938775510204, + "grad_norm": 0.11022552847862244, + "learning_rate": 3.81637377180014e-06, + "loss": 0.06861716508865356, + "step": 2050 + }, + { + "epoch": 0.3363265306122449, + "grad_norm": 0.07990925014019012, + "learning_rate": 3.8052737376660446e-06, + "loss": 0.07096996307373046, + "step": 2060 + }, + { + "epoch": 0.3379591836734694, + "grad_norm": 0.11924420297145844, + "learning_rate": 3.7941382151450756e-06, + "loss": 0.07092551589012146, + "step": 2070 + }, + { + "epoch": 0.3395918367346939, + "grad_norm": 0.08096475899219513, + "learning_rate": 3.782967506994975e-06, + "loss": 0.06780472993850709, + "step": 2080 + }, + { + "epoch": 0.34122448979591835, + "grad_norm": 0.14714907109737396, + "learning_rate": 3.7717619169301302e-06, + "loss": 0.07661795020103454, + "step": 2090 + }, + { + "epoch": 0.34285714285714286, + "grad_norm": 0.14839571714401245, + "learning_rate": 3.7605217496133127e-06, + "loss": 0.07939797639846802, + "step": 2100 + }, + { + "epoch": 0.34448979591836737, + "grad_norm": 0.15081271529197693, + "learning_rate": 3.749247310647398e-06, + "loss": 0.07869908213615417, + "step": 2110 + }, + { + "epoch": 0.3461224489795918, + "grad_norm": 0.13346275687217712, + "learning_rate": 3.737938906567053e-06, + "loss": 0.07526676058769226, + "step": 2120 + }, + { + "epoch": 0.34775510204081633, + "grad_norm": 0.112598717212677, + "learning_rate": 3.7265968448304057e-06, + "loss": 0.07366414070129394, + "step": 2130 + }, + { + "epoch": 0.34938775510204084, + "grad_norm": 0.12423531711101532, + "learning_rate": 3.7152214338106827e-06, + "loss": 0.06439908146858216, + "step": 2140 + }, + { + "epoch": 0.3510204081632653, + "grad_norm": 0.21758921444416046, + "learning_rate": 3.703812982787828e-06, + "loss": 0.08098108768463134, + "step": 2150 + }, + { + "epoch": 0.3526530612244898, + "grad_norm": 0.1298121213912964, + "learning_rate": 3.6923718019400906e-06, + "loss": 0.08134381771087647, + "step": 2160 + }, + { + "epoch": 0.35428571428571426, + "grad_norm": 0.09018377214670181, + "learning_rate": 3.6808982023355934e-06, + "loss": 0.07070590257644653, + "step": 2170 + }, + { + "epoch": 0.35591836734693877, + "grad_norm": 0.1329157054424286, + "learning_rate": 3.6693924959238774e-06, + "loss": 0.06200396418571472, + "step": 2180 + }, + { + "epoch": 0.3575510204081633, + "grad_norm": 0.08813945204019547, + "learning_rate": 3.657854995527416e-06, + "loss": 0.06886230111122131, + "step": 2190 + }, + { + "epoch": 0.35918367346938773, + "grad_norm": 0.13459829986095428, + "learning_rate": 3.646286014833114e-06, + "loss": 0.07170915603637695, + "step": 2200 + }, + { + "epoch": 0.35918367346938773, + "eval_accuracy": 0.970625, + "eval_loss": 0.06810569018125534, + "eval_runtime": 157.4227, + "eval_samples_per_second": 25.409, + "eval_steps_per_second": 25.409, + "step": 2200 + }, + { + "epoch": 0.36081632653061224, + "grad_norm": 0.13405098021030426, + "learning_rate": 3.634685868383776e-06, + "loss": 0.07438129186630249, + "step": 2210 + }, + { + "epoch": 0.36244897959183675, + "grad_norm": 0.1342102289199829, + "learning_rate": 3.6230548715695555e-06, + "loss": 0.07625046968460084, + "step": 2220 + }, + { + "epoch": 0.3640816326530612, + "grad_norm": 0.1532502919435501, + "learning_rate": 3.611393340619381e-06, + "loss": 0.06719872951507569, + "step": 2230 + }, + { + "epoch": 0.3657142857142857, + "grad_norm": 0.11218152195215225, + "learning_rate": 3.5997015925923563e-06, + "loss": 0.07131683230400085, + "step": 2240 + }, + { + "epoch": 0.3673469387755102, + "grad_norm": 0.09200239181518555, + "learning_rate": 3.587979945369141e-06, + "loss": 0.06948224902153015, + "step": 2250 + }, + { + "epoch": 0.3689795918367347, + "grad_norm": 0.08099328726530075, + "learning_rate": 3.5762287176433097e-06, + "loss": 0.07263018488883972, + "step": 2260 + }, + { + "epoch": 0.3706122448979592, + "grad_norm": 0.09097494184970856, + "learning_rate": 3.564448228912682e-06, + "loss": 0.06066258549690247, + "step": 2270 + }, + { + "epoch": 0.3722448979591837, + "grad_norm": 0.07033897191286087, + "learning_rate": 3.5526387994706428e-06, + "loss": 0.07041062712669373, + "step": 2280 + }, + { + "epoch": 0.37387755102040815, + "grad_norm": 0.07659374922513962, + "learning_rate": 3.5408007503974283e-06, + "loss": 0.07272242307662964, + "step": 2290 + }, + { + "epoch": 0.37551020408163266, + "grad_norm": 0.0836353451013565, + "learning_rate": 3.5289344035513996e-06, + "loss": 0.07268807291984558, + "step": 2300 + }, + { + "epoch": 0.37714285714285717, + "grad_norm": 0.1426168978214264, + "learning_rate": 3.5170400815602894e-06, + "loss": 0.05777261257171631, + "step": 2310 + }, + { + "epoch": 0.3787755102040816, + "grad_norm": 0.07479120045900345, + "learning_rate": 3.5051181078124336e-06, + "loss": 0.06174240708351135, + "step": 2320 + }, + { + "epoch": 0.38040816326530613, + "grad_norm": 0.14598114788532257, + "learning_rate": 3.4931688064479745e-06, + "loss": 0.06616324782371522, + "step": 2330 + }, + { + "epoch": 0.3820408163265306, + "grad_norm": 0.22399863600730896, + "learning_rate": 3.4811925023500524e-06, + "loss": 0.06692227125167846, + "step": 2340 + }, + { + "epoch": 0.3836734693877551, + "grad_norm": 0.11600285023450851, + "learning_rate": 3.4691895211359693e-06, + "loss": 0.0634126603603363, + "step": 2350 + }, + { + "epoch": 0.3853061224489796, + "grad_norm": 0.0577060841023922, + "learning_rate": 3.457160189148337e-06, + "loss": 0.06884999275207519, + "step": 2360 + }, + { + "epoch": 0.38693877551020406, + "grad_norm": 0.09160040318965912, + "learning_rate": 3.445104833446205e-06, + "loss": 0.05986430644989014, + "step": 2370 + }, + { + "epoch": 0.38857142857142857, + "grad_norm": 0.1284518539905548, + "learning_rate": 3.433023781796167e-06, + "loss": 0.06759394407272339, + "step": 2380 + }, + { + "epoch": 0.3902040816326531, + "grad_norm": 0.08572858572006226, + "learning_rate": 3.4209173626634494e-06, + "loss": 0.06813116669654846, + "step": 2390 + }, + { + "epoch": 0.39183673469387753, + "grad_norm": 0.11550725251436234, + "learning_rate": 3.4087859052029837e-06, + "loss": 0.06371633410453796, + "step": 2400 + }, + { + "epoch": 0.39183673469387753, + "eval_accuracy": 0.9685833333333334, + "eval_loss": 0.07035119831562042, + "eval_runtime": 156.9729, + "eval_samples_per_second": 25.482, + "eval_steps_per_second": 25.482, + "step": 2400 + }, + { + "epoch": 0.39346938775510204, + "grad_norm": 0.22328943014144897, + "learning_rate": 3.3966297392504535e-06, + "loss": 0.06804299950599671, + "step": 2410 + }, + { + "epoch": 0.39510204081632655, + "grad_norm": 0.18518126010894775, + "learning_rate": 3.3844491953133272e-06, + "loss": 0.07678125500679016, + "step": 2420 + }, + { + "epoch": 0.396734693877551, + "grad_norm": 0.09818598628044128, + "learning_rate": 3.372244604561875e-06, + "loss": 0.06537411212921143, + "step": 2430 + }, + { + "epoch": 0.3983673469387755, + "grad_norm": 0.08255010843276978, + "learning_rate": 3.360016298820163e-06, + "loss": 0.05702770948410034, + "step": 2440 + }, + { + "epoch": 0.4, + "grad_norm": 0.10319415479898453, + "learning_rate": 3.3477646105570293e-06, + "loss": 0.06334798336029053, + "step": 2450 + }, + { + "epoch": 0.4016326530612245, + "grad_norm": 0.06523872911930084, + "learning_rate": 3.3354898728770503e-06, + "loss": 0.051438117027282716, + "step": 2460 + }, + { + "epoch": 0.403265306122449, + "grad_norm": 0.15883173048496246, + "learning_rate": 3.3231924195114774e-06, + "loss": 0.06427289247512817, + "step": 2470 + }, + { + "epoch": 0.4048979591836735, + "grad_norm": 0.1734374463558197, + "learning_rate": 3.3108725848091685e-06, + "loss": 0.06652372479438781, + "step": 2480 + }, + { + "epoch": 0.40653061224489795, + "grad_norm": 0.10557752102613449, + "learning_rate": 3.2985307037274937e-06, + "loss": 0.07374808788299561, + "step": 2490 + }, + { + "epoch": 0.40816326530612246, + "grad_norm": 0.12942200899124146, + "learning_rate": 3.286167111823234e-06, + "loss": 0.06349012851715088, + "step": 2500 + }, + { + "epoch": 0.4097959183673469, + "grad_norm": 0.1858324259519577, + "learning_rate": 3.2737821452434502e-06, + "loss": 0.06953533291816712, + "step": 2510 + }, + { + "epoch": 0.4114285714285714, + "grad_norm": 0.10682573169469833, + "learning_rate": 3.26137614071635e-06, + "loss": 0.05826004147529602, + "step": 2520 + }, + { + "epoch": 0.41306122448979593, + "grad_norm": 0.12533792853355408, + "learning_rate": 3.2489494355421293e-06, + "loss": 0.05837242007255554, + "step": 2530 + }, + { + "epoch": 0.4146938775510204, + "grad_norm": 0.10952792316675186, + "learning_rate": 3.2365023675838027e-06, + "loss": 0.06359312534332276, + "step": 2540 + }, + { + "epoch": 0.4163265306122449, + "grad_norm": 0.08605755865573883, + "learning_rate": 3.22403527525802e-06, + "loss": 0.07361164093017578, + "step": 2550 + }, + { + "epoch": 0.4179591836734694, + "grad_norm": 0.15942610800266266, + "learning_rate": 3.211548497525861e-06, + "loss": 0.05989238619804382, + "step": 2560 + }, + { + "epoch": 0.41959183673469386, + "grad_norm": 0.16340239346027374, + "learning_rate": 3.1990423738836206e-06, + "loss": 0.07593320608139038, + "step": 2570 + }, + { + "epoch": 0.42122448979591837, + "grad_norm": 0.11011330783367157, + "learning_rate": 3.1865172443535797e-06, + "loss": 0.06850495338439941, + "step": 2580 + }, + { + "epoch": 0.4228571428571429, + "grad_norm": 0.10529608279466629, + "learning_rate": 3.173973449474762e-06, + "loss": 0.056969082355499266, + "step": 2590 + }, + { + "epoch": 0.42448979591836733, + "grad_norm": 0.12615667283535004, + "learning_rate": 3.1614113302936716e-06, + "loss": 0.06189748048782349, + "step": 2600 + }, + { + "epoch": 0.42448979591836733, + "eval_accuracy": 0.9746666666666668, + "eval_loss": 0.0618240088224411, + "eval_runtime": 157.0858, + "eval_samples_per_second": 25.464, + "eval_steps_per_second": 25.464, + "step": 2600 + }, + { + "epoch": 0.42612244897959184, + "grad_norm": 0.1343226134777069, + "learning_rate": 3.1488312283550216e-06, + "loss": 0.061647194623947146, + "step": 2610 + }, + { + "epoch": 0.42775510204081635, + "grad_norm": 0.11777225136756897, + "learning_rate": 3.136233485692451e-06, + "loss": 0.05916886925697327, + "step": 2620 + }, + { + "epoch": 0.4293877551020408, + "grad_norm": 0.14721977710723877, + "learning_rate": 3.12361844481922e-06, + "loss": 0.059158307313919065, + "step": 2630 + }, + { + "epoch": 0.4310204081632653, + "grad_norm": 0.12912684679031372, + "learning_rate": 3.110986448718902e-06, + "loss": 0.05914759635925293, + "step": 2640 + }, + { + "epoch": 0.4326530612244898, + "grad_norm": 0.13003695011138916, + "learning_rate": 3.0983378408360597e-06, + "loss": 0.059078723192214966, + "step": 2650 + }, + { + "epoch": 0.4342857142857143, + "grad_norm": 0.12714017927646637, + "learning_rate": 3.085672965066899e-06, + "loss": 0.05707222819328308, + "step": 2660 + }, + { + "epoch": 0.4359183673469388, + "grad_norm": 0.13708427548408508, + "learning_rate": 3.0729921657499294e-06, + "loss": 0.05138048529624939, + "step": 2670 + }, + { + "epoch": 0.43755102040816324, + "grad_norm": 0.0949573740363121, + "learning_rate": 3.0602957876565943e-06, + "loss": 0.050283265113830564, + "step": 2680 + }, + { + "epoch": 0.43918367346938775, + "grad_norm": 0.08723263442516327, + "learning_rate": 3.0475841759819023e-06, + "loss": 0.06342648863792419, + "step": 2690 + }, + { + "epoch": 0.44081632653061226, + "grad_norm": 0.09944969415664673, + "learning_rate": 3.034857676335038e-06, + "loss": 0.06127093434333801, + "step": 2700 + }, + { + "epoch": 0.4424489795918367, + "grad_norm": 0.10576151311397552, + "learning_rate": 3.0221166347299673e-06, + "loss": 0.05412284731864929, + "step": 2710 + }, + { + "epoch": 0.4440816326530612, + "grad_norm": 0.1598891168832779, + "learning_rate": 3.009361397576031e-06, + "loss": 0.05579071640968323, + "step": 2720 + }, + { + "epoch": 0.44571428571428573, + "grad_norm": 0.16953758895397186, + "learning_rate": 2.9965923116685224e-06, + "loss": 0.06472888588905334, + "step": 2730 + }, + { + "epoch": 0.4473469387755102, + "grad_norm": 0.10883133113384247, + "learning_rate": 2.9838097241792646e-06, + "loss": 0.06263615489006043, + "step": 2740 + }, + { + "epoch": 0.4489795918367347, + "grad_norm": 0.08914396911859512, + "learning_rate": 2.9710139826471646e-06, + "loss": 0.0624904453754425, + "step": 2750 + }, + { + "epoch": 0.4506122448979592, + "grad_norm": 0.1059069111943245, + "learning_rate": 2.9582054349687706e-06, + "loss": 0.05488716959953308, + "step": 2760 + }, + { + "epoch": 0.45224489795918366, + "grad_norm": 0.09883642196655273, + "learning_rate": 2.9453844293888067e-06, + "loss": 0.06263139843940735, + "step": 2770 + }, + { + "epoch": 0.45387755102040817, + "grad_norm": 0.11156519502401352, + "learning_rate": 2.932551314490712e-06, + "loss": 0.06324111819267272, + "step": 2780 + }, + { + "epoch": 0.4555102040816327, + "grad_norm": 0.08826332539319992, + "learning_rate": 2.9197064391871565e-06, + "loss": 0.06384775042533875, + "step": 2790 + }, + { + "epoch": 0.45714285714285713, + "grad_norm": 0.0698089748620987, + "learning_rate": 2.9068501527105595e-06, + "loss": 0.05722553730010986, + "step": 2800 + }, + { + "epoch": 0.45714285714285713, + "eval_accuracy": 0.977375, + "eval_loss": 0.05622806400060654, + "eval_runtime": 170.7758, + "eval_samples_per_second": 23.423, + "eval_steps_per_second": 23.423, + "step": 2800 + }, + { + "epoch": 0.45877551020408164, + "grad_norm": 0.07031046599149704, + "learning_rate": 2.893982804603592e-06, + "loss": 0.060957270860672, + "step": 2810 + }, + { + "epoch": 0.46040816326530615, + "grad_norm": 0.08772005140781403, + "learning_rate": 2.8811047447096714e-06, + "loss": 0.05667365789413452, + "step": 2820 + }, + { + "epoch": 0.4620408163265306, + "grad_norm": 0.10329847782850266, + "learning_rate": 2.8682163231634557e-06, + "loss": 0.06144403219223023, + "step": 2830 + }, + { + "epoch": 0.4636734693877551, + "grad_norm": 0.06306952983140945, + "learning_rate": 2.8553178903813193e-06, + "loss": 0.04665087461471558, + "step": 2840 + }, + { + "epoch": 0.46530612244897956, + "grad_norm": 0.11146903783082962, + "learning_rate": 2.842409797051825e-06, + "loss": 0.06363347172737122, + "step": 2850 + }, + { + "epoch": 0.4669387755102041, + "grad_norm": 0.04504350945353508, + "learning_rate": 2.829492394126192e-06, + "loss": 0.05332474112510681, + "step": 2860 + }, + { + "epoch": 0.4685714285714286, + "grad_norm": 0.06039620563387871, + "learning_rate": 2.8165660328087546e-06, + "loss": 0.07094461917877197, + "step": 2870 + }, + { + "epoch": 0.47020408163265304, + "grad_norm": 0.08084474503993988, + "learning_rate": 2.8036310645474096e-06, + "loss": 0.058167701959609984, + "step": 2880 + }, + { + "epoch": 0.47183673469387755, + "grad_norm": 0.05352522060275078, + "learning_rate": 2.7906878410240655e-06, + "loss": 0.0486674964427948, + "step": 2890 + }, + { + "epoch": 0.47346938775510206, + "grad_norm": 0.09431054443120956, + "learning_rate": 2.777736714145077e-06, + "loss": 0.062025409936904904, + "step": 2900 + }, + { + "epoch": 0.4751020408163265, + "grad_norm": 0.09863542020320892, + "learning_rate": 2.7647780360316787e-06, + "loss": 0.061118441820144656, + "step": 2910 + }, + { + "epoch": 0.476734693877551, + "grad_norm": 0.082881860435009, + "learning_rate": 2.7518121590104145e-06, + "loss": 0.061808472871780394, + "step": 2920 + }, + { + "epoch": 0.47836734693877553, + "grad_norm": 0.06983684748411179, + "learning_rate": 2.7388394356035524e-06, + "loss": 0.05849882960319519, + "step": 2930 + }, + { + "epoch": 0.48, + "grad_norm": 0.10832090675830841, + "learning_rate": 2.7258602185195043e-06, + "loss": 0.06612529158592224, + "step": 2940 + }, + { + "epoch": 0.4816326530612245, + "grad_norm": 0.1516750454902649, + "learning_rate": 2.7128748606432353e-06, + "loss": 0.058327901363372806, + "step": 2950 + }, + { + "epoch": 0.483265306122449, + "grad_norm": 0.14055661857128143, + "learning_rate": 2.699883715026669e-06, + "loss": 0.055737996101379396, + "step": 2960 + }, + { + "epoch": 0.48489795918367345, + "grad_norm": 0.0793638750910759, + "learning_rate": 2.686887134879089e-06, + "loss": 0.051851028203964235, + "step": 2970 + }, + { + "epoch": 0.48653061224489796, + "grad_norm": 0.07475428283214569, + "learning_rate": 2.6738854735575358e-06, + "loss": 0.06357839703559875, + "step": 2980 + }, + { + "epoch": 0.4881632653061225, + "grad_norm": 0.13335047662258148, + "learning_rate": 2.660879084557198e-06, + "loss": 0.06160399913787842, + "step": 2990 + }, + { + "epoch": 0.4897959183673469, + "grad_norm": 0.12583893537521362, + "learning_rate": 2.6478683215018025e-06, + "loss": 0.060502588748931885, + "step": 3000 + }, + { + "epoch": 0.4897959183673469, + "eval_accuracy": 0.9767083333333334, + "eval_loss": 0.0561000294983387, + "eval_runtime": 156.4202, + "eval_samples_per_second": 25.572, + "eval_steps_per_second": 25.572, + "step": 3000 + }, + { + "epoch": 0.49142857142857144, + "grad_norm": 0.1411438137292862, + "learning_rate": 2.6348535381340025e-06, + "loss": 0.060459482669830325, + "step": 3010 + }, + { + "epoch": 0.4930612244897959, + "grad_norm": 0.12584605813026428, + "learning_rate": 2.6218350883057546e-06, + "loss": 0.0632048487663269, + "step": 3020 + }, + { + "epoch": 0.4946938775510204, + "grad_norm": 0.10522248595952988, + "learning_rate": 2.608813325968702e-06, + "loss": 0.04655081033706665, + "step": 3030 + }, + { + "epoch": 0.4963265306122449, + "grad_norm": 0.11343304812908173, + "learning_rate": 2.59578860516455e-06, + "loss": 0.05672039389610291, + "step": 3040 + }, + { + "epoch": 0.49795918367346936, + "grad_norm": 0.12666337192058563, + "learning_rate": 2.5827612800154382e-06, + "loss": 0.047120341658592226, + "step": 3050 + }, + { + "epoch": 0.4995918367346939, + "grad_norm": 0.08930598944425583, + "learning_rate": 2.5697317047143177e-06, + "loss": 0.06063341498374939, + "step": 3060 + }, + { + "epoch": 0.5012244897959184, + "grad_norm": 0.07818852365016937, + "learning_rate": 2.556700233515316e-06, + "loss": 0.05240688323974609, + "step": 3070 + }, + { + "epoch": 0.5028571428571429, + "grad_norm": 0.1425711065530777, + "learning_rate": 2.5436672207241064e-06, + "loss": 0.053180736303329465, + "step": 3080 + }, + { + "epoch": 0.5044897959183674, + "grad_norm": 0.1457373946905136, + "learning_rate": 2.5306330206882764e-06, + "loss": 0.06092001795768738, + "step": 3090 + }, + { + "epoch": 0.5061224489795918, + "grad_norm": 0.12738314270973206, + "learning_rate": 2.5175979877876935e-06, + "loss": 0.06613186597824097, + "step": 3100 + }, + { + "epoch": 0.5077551020408163, + "grad_norm": 0.09169036149978638, + "learning_rate": 2.5045624764248687e-06, + "loss": 0.06115126609802246, + "step": 3110 + }, + { + "epoch": 0.5093877551020408, + "grad_norm": 0.09637737274169922, + "learning_rate": 2.491526841015321e-06, + "loss": 0.044120225310325625, + "step": 3120 + }, + { + "epoch": 0.5110204081632653, + "grad_norm": 0.15029804408550262, + "learning_rate": 2.478491435977945e-06, + "loss": 0.07098687291145325, + "step": 3130 + }, + { + "epoch": 0.5126530612244898, + "grad_norm": 0.1228741705417633, + "learning_rate": 2.4654566157253683e-06, + "loss": 0.05738852024078369, + "step": 3140 + }, + { + "epoch": 0.5142857142857142, + "grad_norm": 0.09150952100753784, + "learning_rate": 2.452422734654321e-06, + "loss": 0.053479206562042234, + "step": 3150 + }, + { + "epoch": 0.5159183673469387, + "grad_norm": 0.14547918736934662, + "learning_rate": 2.4393901471359977e-06, + "loss": 0.056794995069503786, + "step": 3160 + }, + { + "epoch": 0.5175510204081633, + "grad_norm": 0.10099287331104279, + "learning_rate": 2.4263592075064238e-06, + "loss": 0.05802353620529175, + "step": 3170 + }, + { + "epoch": 0.5191836734693878, + "grad_norm": 0.10859636962413788, + "learning_rate": 2.41333027005682e-06, + "loss": 0.05282890200614929, + "step": 3180 + }, + { + "epoch": 0.5208163265306123, + "grad_norm": 0.12420635670423508, + "learning_rate": 2.400303689023972e-06, + "loss": 0.05190218687057495, + "step": 3190 + }, + { + "epoch": 0.5224489795918368, + "grad_norm": 0.1902378350496292, + "learning_rate": 2.3872798185805986e-06, + "loss": 0.06482383608818054, + "step": 3200 + }, + { + "epoch": 0.5224489795918368, + "eval_accuracy": 0.9744583333333335, + "eval_loss": 0.058392543345689774, + "eval_runtime": 156.8655, + "eval_samples_per_second": 25.5, + "eval_steps_per_second": 25.5, + "step": 3200 + }, + { + "epoch": 0.5240816326530612, + "grad_norm": 0.15229448676109314, + "learning_rate": 2.37425901282572e-06, + "loss": 0.06062687039375305, + "step": 3210 + }, + { + "epoch": 0.5257142857142857, + "grad_norm": 0.08411400765180588, + "learning_rate": 2.3612416257750335e-06, + "loss": 0.06494970321655273, + "step": 3220 + }, + { + "epoch": 0.5273469387755102, + "grad_norm": 0.114021435379982, + "learning_rate": 2.348228011351287e-06, + "loss": 0.05140092372894287, + "step": 3230 + }, + { + "epoch": 0.5289795918367347, + "grad_norm": 0.08826510608196259, + "learning_rate": 2.335218523374656e-06, + "loss": 0.041680914163589475, + "step": 3240 + }, + { + "epoch": 0.5306122448979592, + "grad_norm": 0.10536732524633408, + "learning_rate": 2.3222135155531248e-06, + "loss": 0.04369679093360901, + "step": 3250 + }, + { + "epoch": 0.5322448979591837, + "grad_norm": 0.16883279383182526, + "learning_rate": 2.309213341472868e-06, + "loss": 0.050543594360351565, + "step": 3260 + }, + { + "epoch": 0.5338775510204081, + "grad_norm": 0.15277455747127533, + "learning_rate": 2.29621835458864e-06, + "loss": 0.05258342027664185, + "step": 3270 + }, + { + "epoch": 0.5355102040816326, + "grad_norm": 0.07162481546401978, + "learning_rate": 2.283228908214163e-06, + "loss": 0.055678969621658324, + "step": 3280 + }, + { + "epoch": 0.5371428571428571, + "grad_norm": 0.11578263342380524, + "learning_rate": 2.2702453555125182e-06, + "loss": 0.048620349168777464, + "step": 3290 + }, + { + "epoch": 0.5387755102040817, + "grad_norm": 0.21185705065727234, + "learning_rate": 2.2572680494865527e-06, + "loss": 0.046547812223434445, + "step": 3300 + }, + { + "epoch": 0.5404081632653062, + "grad_norm": 0.19949115812778473, + "learning_rate": 2.2442973429692705e-06, + "loss": 0.04967870116233826, + "step": 3310 + }, + { + "epoch": 0.5420408163265306, + "grad_norm": 0.10855370759963989, + "learning_rate": 2.231333588614248e-06, + "loss": 0.0508593499660492, + "step": 3320 + }, + { + "epoch": 0.5436734693877551, + "grad_norm": 0.11827342212200165, + "learning_rate": 2.2183771388860427e-06, + "loss": 0.04854269027709961, + "step": 3330 + }, + { + "epoch": 0.5453061224489796, + "grad_norm": 0.1081443578004837, + "learning_rate": 2.2054283460506096e-06, + "loss": 0.05934192538261414, + "step": 3340 + }, + { + "epoch": 0.5469387755102041, + "grad_norm": 0.14441704750061035, + "learning_rate": 2.1924875621657256e-06, + "loss": 0.05332346558570862, + "step": 3350 + }, + { + "epoch": 0.5485714285714286, + "grad_norm": 0.11531820893287659, + "learning_rate": 2.179555139071415e-06, + "loss": 0.04938814043998718, + "step": 3360 + }, + { + "epoch": 0.5502040816326531, + "grad_norm": 0.08365523815155029, + "learning_rate": 2.166631428380387e-06, + "loss": 0.05734975934028626, + "step": 3370 + }, + { + "epoch": 0.5518367346938775, + "grad_norm": 0.09165070950984955, + "learning_rate": 2.1537167814684728e-06, + "loss": 0.047475004196166994, + "step": 3380 + }, + { + "epoch": 0.553469387755102, + "grad_norm": 0.13237863779067993, + "learning_rate": 2.140811549465074e-06, + "loss": 0.062284767627716064, + "step": 3390 + }, + { + "epoch": 0.5551020408163265, + "grad_norm": 0.1698858141899109, + "learning_rate": 2.127916083243615e-06, + "loss": 0.06585791110992431, + "step": 3400 + }, + { + "epoch": 0.5551020408163265, + "eval_accuracy": 0.9770833333333334, + "eval_loss": 0.05396538972854614, + "eval_runtime": 156.044, + "eval_samples_per_second": 25.634, + "eval_steps_per_second": 25.634, + "step": 3400 + }, + { + "epoch": 0.556734693877551, + "grad_norm": 0.1104230135679245, + "learning_rate": 2.1150307334120033e-06, + "loss": 0.04635375440120697, + "step": 3410 + }, + { + "epoch": 0.5583673469387755, + "grad_norm": 0.21372850239276886, + "learning_rate": 2.1021558503030973e-06, + "loss": 0.05940789580345154, + "step": 3420 + }, + { + "epoch": 0.56, + "grad_norm": 0.2013176679611206, + "learning_rate": 2.0892917839651818e-06, + "loss": 0.047165653109550475, + "step": 3430 + }, + { + "epoch": 0.5616326530612245, + "grad_norm": 0.09189072996377945, + "learning_rate": 2.076438884152449e-06, + "loss": 0.04550817906856537, + "step": 3440 + }, + { + "epoch": 0.563265306122449, + "grad_norm": 0.10521821677684784, + "learning_rate": 2.0635975003154934e-06, + "loss": 0.045033201575279236, + "step": 3450 + }, + { + "epoch": 0.5648979591836735, + "grad_norm": 0.10311131924390793, + "learning_rate": 2.0507679815918043e-06, + "loss": 0.04502847194671631, + "step": 3460 + }, + { + "epoch": 0.566530612244898, + "grad_norm": 0.07487176358699799, + "learning_rate": 2.0379506767962788e-06, + "loss": 0.04697479605674744, + "step": 3470 + }, + { + "epoch": 0.5681632653061225, + "grad_norm": 0.12441752105951309, + "learning_rate": 2.0251459344117345e-06, + "loss": 0.052119261026382445, + "step": 3480 + }, + { + "epoch": 0.5697959183673469, + "grad_norm": 0.12823422253131866, + "learning_rate": 2.0123541025794397e-06, + "loss": 0.05213663578033447, + "step": 3490 + }, + { + "epoch": 0.5714285714285714, + "grad_norm": 0.11340193450450897, + "learning_rate": 1.9995755290896414e-06, + "loss": 0.05319364666938782, + "step": 3500 + }, + { + "epoch": 0.5730612244897959, + "grad_norm": 0.10469839721918106, + "learning_rate": 1.986810561372113e-06, + "loss": 0.046560952067375184, + "step": 3510 + }, + { + "epoch": 0.5746938775510204, + "grad_norm": 0.07797456532716751, + "learning_rate": 1.9740595464867092e-06, + "loss": 0.05037369728088379, + "step": 3520 + }, + { + "epoch": 0.5763265306122449, + "grad_norm": 0.13010293245315552, + "learning_rate": 1.961322831113927e-06, + "loss": 0.0543134331703186, + "step": 3530 + }, + { + "epoch": 0.5779591836734694, + "grad_norm": 0.13981281220912933, + "learning_rate": 1.9486007615454837e-06, + "loss": 0.052860957384109494, + "step": 3540 + }, + { + "epoch": 0.5795918367346938, + "grad_norm": 0.14817261695861816, + "learning_rate": 1.9358936836748972e-06, + "loss": 0.05510561466217041, + "step": 3550 + }, + { + "epoch": 0.5812244897959183, + "grad_norm": 0.15079492330551147, + "learning_rate": 1.9232019429880874e-06, + "loss": 0.06125714182853699, + "step": 3560 + }, + { + "epoch": 0.5828571428571429, + "grad_norm": 0.10553149878978729, + "learning_rate": 1.910525884553976e-06, + "loss": 0.051503753662109374, + "step": 3570 + }, + { + "epoch": 0.5844897959183674, + "grad_norm": 0.09094442427158356, + "learning_rate": 1.8978658530151113e-06, + "loss": 0.05377998948097229, + "step": 3580 + }, + { + "epoch": 0.5861224489795919, + "grad_norm": 0.09810740500688553, + "learning_rate": 1.8852221925782935e-06, + "loss": 0.05506981611251831, + "step": 3590 + }, + { + "epoch": 0.5877551020408164, + "grad_norm": 0.13397352397441864, + "learning_rate": 1.872595247005218e-06, + "loss": 0.056353437900543216, + "step": 3600 + }, + { + "epoch": 0.5877551020408164, + "eval_accuracy": 0.979875, + "eval_loss": 0.04787343740463257, + "eval_runtime": 157.8457, + "eval_samples_per_second": 25.341, + "eval_steps_per_second": 25.341, + "step": 3600 + }, + { + "epoch": 0.5893877551020408, + "grad_norm": 0.07899880409240723, + "learning_rate": 1.8599853596031289e-06, + "loss": 0.05329249501228332, + "step": 3610 + }, + { + "epoch": 0.5910204081632653, + "grad_norm": 0.10414254665374756, + "learning_rate": 1.8473928732154847e-06, + "loss": 0.04905630648136139, + "step": 3620 + }, + { + "epoch": 0.5926530612244898, + "grad_norm": 0.1037079319357872, + "learning_rate": 1.8348181302126367e-06, + "loss": 0.04863021373748779, + "step": 3630 + }, + { + "epoch": 0.5942857142857143, + "grad_norm": 0.06960397213697433, + "learning_rate": 1.8222614724825222e-06, + "loss": 0.04666702449321747, + "step": 3640 + }, + { + "epoch": 0.5959183673469388, + "grad_norm": 0.11159648001194, + "learning_rate": 1.8097232414213662e-06, + "loss": 0.05730679631233215, + "step": 3650 + }, + { + "epoch": 0.5975510204081632, + "grad_norm": 0.10231669992208481, + "learning_rate": 1.7972037779244e-06, + "loss": 0.04433938264846802, + "step": 3660 + }, + { + "epoch": 0.5991836734693877, + "grad_norm": 0.1256829798221588, + "learning_rate": 1.784703422376598e-06, + "loss": 0.04502029418945312, + "step": 3670 + }, + { + "epoch": 0.6008163265306122, + "grad_norm": 0.1006830632686615, + "learning_rate": 1.7722225146434146e-06, + "loss": 0.04091633260250092, + "step": 3680 + }, + { + "epoch": 0.6024489795918367, + "grad_norm": 0.12239495664834976, + "learning_rate": 1.7597613940615492e-06, + "loss": 0.0417140930891037, + "step": 3690 + }, + { + "epoch": 0.6040816326530613, + "grad_norm": 0.09386949241161346, + "learning_rate": 1.7473203994297197e-06, + "loss": 0.045112240314483645, + "step": 3700 + }, + { + "epoch": 0.6057142857142858, + "grad_norm": 0.1845707893371582, + "learning_rate": 1.7348998689994492e-06, + "loss": 0.04915813505649567, + "step": 3710 + }, + { + "epoch": 0.6073469387755102, + "grad_norm": 0.1345887929201126, + "learning_rate": 1.7225001404658714e-06, + "loss": 0.05195118188858032, + "step": 3720 + }, + { + "epoch": 0.6089795918367347, + "grad_norm": 0.16226522624492645, + "learning_rate": 1.7101215509585482e-06, + "loss": 0.047248026728630065, + "step": 3730 + }, + { + "epoch": 0.6106122448979592, + "grad_norm": 0.12012234330177307, + "learning_rate": 1.6977644370323032e-06, + "loss": 0.05229860544204712, + "step": 3740 + }, + { + "epoch": 0.6122448979591837, + "grad_norm": 0.16154533624649048, + "learning_rate": 1.6854291346580737e-06, + "loss": 0.048448923230171206, + "step": 3750 + }, + { + "epoch": 0.6138775510204082, + "grad_norm": 0.08616314828395844, + "learning_rate": 1.673115979213773e-06, + "loss": 0.04779277443885803, + "step": 3760 + }, + { + "epoch": 0.6155102040816327, + "grad_norm": 0.20649370551109314, + "learning_rate": 1.6608253054751739e-06, + "loss": 0.04001966416835785, + "step": 3770 + }, + { + "epoch": 0.6171428571428571, + "grad_norm": 0.22419355809688568, + "learning_rate": 1.6485574476068056e-06, + "loss": 0.04915740489959717, + "step": 3780 + }, + { + "epoch": 0.6187755102040816, + "grad_norm": 0.11812295019626617, + "learning_rate": 1.6363127391528694e-06, + "loss": 0.055313241481781, + "step": 3790 + }, + { + "epoch": 0.6204081632653061, + "grad_norm": 0.12400855869054794, + "learning_rate": 1.6240915130281696e-06, + "loss": 0.05452631115913391, + "step": 3800 + }, + { + "epoch": 0.6204081632653061, + "eval_accuracy": 0.981875, + "eval_loss": 0.04318515956401825, + "eval_runtime": 158.5838, + "eval_samples_per_second": 25.223, + "eval_steps_per_second": 25.223, + "step": 3800 + }, + { + "epoch": 0.6220408163265306, + "grad_norm": 0.094053253531456, + "learning_rate": 1.6118941015090614e-06, + "loss": 0.047327810525894166, + "step": 3810 + }, + { + "epoch": 0.6236734693877551, + "grad_norm": 0.11286808550357819, + "learning_rate": 1.599720836224419e-06, + "loss": 0.04089818894863129, + "step": 3820 + }, + { + "epoch": 0.6253061224489795, + "grad_norm": 0.0909293070435524, + "learning_rate": 1.5875720481466162e-06, + "loss": 0.04378651976585388, + "step": 3830 + }, + { + "epoch": 0.626938775510204, + "grad_norm": 0.1207573264837265, + "learning_rate": 1.5754480675825295e-06, + "loss": 0.055971544981002805, + "step": 3840 + }, + { + "epoch": 0.6285714285714286, + "grad_norm": 0.09171696752309799, + "learning_rate": 1.563349224164556e-06, + "loss": 0.04364039599895477, + "step": 3850 + }, + { + "epoch": 0.6302040816326531, + "grad_norm": 0.1100732609629631, + "learning_rate": 1.5512758468416561e-06, + "loss": 0.04227602481842041, + "step": 3860 + }, + { + "epoch": 0.6318367346938776, + "grad_norm": 0.08587154746055603, + "learning_rate": 1.5392282638704032e-06, + "loss": 0.03987776041030884, + "step": 3870 + }, + { + "epoch": 0.6334693877551021, + "grad_norm": 0.14273083209991455, + "learning_rate": 1.5272068028060616e-06, + "loss": 0.05040489435195923, + "step": 3880 + }, + { + "epoch": 0.6351020408163265, + "grad_norm": 0.12788064777851105, + "learning_rate": 1.5152117904936827e-06, + "loss": 0.04129888415336609, + "step": 3890 + }, + { + "epoch": 0.636734693877551, + "grad_norm": 0.14224086701869965, + "learning_rate": 1.5032435530592165e-06, + "loss": 0.04145504236221313, + "step": 3900 + }, + { + "epoch": 0.6383673469387755, + "grad_norm": 0.07414636015892029, + "learning_rate": 1.4913024159006444e-06, + "loss": 0.03938417434692383, + "step": 3910 + }, + { + "epoch": 0.64, + "grad_norm": 0.1443425565958023, + "learning_rate": 1.4793887036791338e-06, + "loss": 0.047359606623649596, + "step": 3920 + }, + { + "epoch": 0.6416326530612245, + "grad_norm": 0.1462288498878479, + "learning_rate": 1.4675027403102107e-06, + "loss": 0.047661393880844116, + "step": 3930 + }, + { + "epoch": 0.643265306122449, + "grad_norm": 0.15937893092632294, + "learning_rate": 1.4556448489549519e-06, + "loss": 0.04829471111297608, + "step": 3940 + }, + { + "epoch": 0.6448979591836734, + "grad_norm": 0.10469378530979156, + "learning_rate": 1.4438153520111978e-06, + "loss": 0.047904521226882935, + "step": 3950 + }, + { + "epoch": 0.6465306122448979, + "grad_norm": 0.10817313939332962, + "learning_rate": 1.4320145711047894e-06, + "loss": 0.04648992121219635, + "step": 3960 + }, + { + "epoch": 0.6481632653061224, + "grad_norm": 0.12307990342378616, + "learning_rate": 1.4202428270808256e-06, + "loss": 0.044715133309364316, + "step": 3970 + }, + { + "epoch": 0.649795918367347, + "grad_norm": 0.14425775408744812, + "learning_rate": 1.408500439994933e-06, + "loss": 0.042470496892929074, + "step": 3980 + }, + { + "epoch": 0.6514285714285715, + "grad_norm": 0.11694067716598511, + "learning_rate": 1.3967877291045706e-06, + "loss": 0.03910423815250397, + "step": 3990 + }, + { + "epoch": 0.6530612244897959, + "grad_norm": 0.13050039112567902, + "learning_rate": 1.3851050128603454e-06, + "loss": 0.046834564208984374, + "step": 4000 + }, + { + "epoch": 0.6530612244897959, + "eval_accuracy": 0.9817083333333334, + "eval_loss": 0.04485878720879555, + "eval_runtime": 155.2113, + "eval_samples_per_second": 25.771, + "eval_steps_per_second": 25.771, + "step": 4000 + }, + { + "epoch": 0.6546938775510204, + "grad_norm": 0.16441842913627625, + "learning_rate": 1.3734526088973576e-06, + "loss": 0.048332995176315306, + "step": 4010 + }, + { + "epoch": 0.6563265306122449, + "grad_norm": 0.16489046812057495, + "learning_rate": 1.3618308340265628e-06, + "loss": 0.05632186532020569, + "step": 4020 + }, + { + "epoch": 0.6579591836734694, + "grad_norm": 0.0776519775390625, + "learning_rate": 1.3502400042261583e-06, + "loss": 0.0413711816072464, + "step": 4030 + }, + { + "epoch": 0.6595918367346939, + "grad_norm": 0.07065974175930023, + "learning_rate": 1.3386804346329913e-06, + "loss": 0.04443484246730804, + "step": 4040 + }, + { + "epoch": 0.6612244897959184, + "grad_norm": 0.058040838688611984, + "learning_rate": 1.3271524395339958e-06, + "loss": 0.03936767876148224, + "step": 4050 + }, + { + "epoch": 0.6628571428571428, + "grad_norm": 0.12576180696487427, + "learning_rate": 1.31565633235764e-06, + "loss": 0.03893322050571442, + "step": 4060 + }, + { + "epoch": 0.6644897959183673, + "grad_norm": 0.1022111177444458, + "learning_rate": 1.30419242566541e-06, + "loss": 0.05221283435821533, + "step": 4070 + }, + { + "epoch": 0.6661224489795918, + "grad_norm": 0.07897438853979111, + "learning_rate": 1.2927610311433098e-06, + "loss": 0.04261190891265869, + "step": 4080 + }, + { + "epoch": 0.6677551020408163, + "grad_norm": 0.16535215079784393, + "learning_rate": 1.2813624595933885e-06, + "loss": 0.04804243743419647, + "step": 4090 + }, + { + "epoch": 0.6693877551020408, + "grad_norm": 0.10754700005054474, + "learning_rate": 1.2699970209252877e-06, + "loss": 0.03677414059638977, + "step": 4100 + }, + { + "epoch": 0.6710204081632654, + "grad_norm": 0.13121150434017181, + "learning_rate": 1.2586650241478173e-06, + "loss": 0.04266786575317383, + "step": 4110 + }, + { + "epoch": 0.6726530612244898, + "grad_norm": 0.11523906886577606, + "learning_rate": 1.2473667773605542e-06, + "loss": 0.045382776856422426, + "step": 4120 + }, + { + "epoch": 0.6742857142857143, + "grad_norm": 0.16150203347206116, + "learning_rate": 1.236102587745464e-06, + "loss": 0.0483401358127594, + "step": 4130 + }, + { + "epoch": 0.6759183673469388, + "grad_norm": 0.2037627100944519, + "learning_rate": 1.2248727615585501e-06, + "loss": 0.0531754195690155, + "step": 4140 + }, + { + "epoch": 0.6775510204081633, + "grad_norm": 0.08756938576698303, + "learning_rate": 1.2136776041215293e-06, + "loss": 0.046069496870040895, + "step": 4150 + }, + { + "epoch": 0.6791836734693878, + "grad_norm": 0.10196162015199661, + "learning_rate": 1.2025174198135255e-06, + "loss": 0.03938464820384979, + "step": 4160 + }, + { + "epoch": 0.6808163265306122, + "grad_norm": 0.10243439674377441, + "learning_rate": 1.1913925120627987e-06, + "loss": 0.04410702586174011, + "step": 4170 + }, + { + "epoch": 0.6824489795918367, + "grad_norm": 0.16230984032154083, + "learning_rate": 1.1803031833384923e-06, + "loss": 0.03496443629264832, + "step": 4180 + }, + { + "epoch": 0.6840816326530612, + "grad_norm": 0.1036972925066948, + "learning_rate": 1.1692497351424117e-06, + "loss": 0.04019207060337067, + "step": 4190 + }, + { + "epoch": 0.6857142857142857, + "grad_norm": 0.19966299831867218, + "learning_rate": 1.1582324680008252e-06, + "loss": 0.03875584900379181, + "step": 4200 + }, + { + "epoch": 0.6857142857142857, + "eval_accuracy": 0.9839583333333334, + "eval_loss": 0.0395464189350605, + "eval_runtime": 153.8599, + "eval_samples_per_second": 25.998, + "eval_steps_per_second": 25.998, + "step": 4200 + }, + { + "epoch": 0.6873469387755102, + "grad_norm": 0.15349307656288147, + "learning_rate": 1.1472516814562945e-06, + "loss": 0.04495754837989807, + "step": 4210 + }, + { + "epoch": 0.6889795918367347, + "grad_norm": 0.16430068016052246, + "learning_rate": 1.1363076740595294e-06, + "loss": 0.0422965407371521, + "step": 4220 + }, + { + "epoch": 0.6906122448979591, + "grad_norm": 0.1903417408466339, + "learning_rate": 1.1254007433612705e-06, + "loss": 0.04501302838325501, + "step": 4230 + }, + { + "epoch": 0.6922448979591836, + "grad_norm": 0.05951374024152756, + "learning_rate": 1.1145311859042026e-06, + "loss": 0.03593409359455109, + "step": 4240 + }, + { + "epoch": 0.6938775510204082, + "grad_norm": 0.05894222855567932, + "learning_rate": 1.1036992972148865e-06, + "loss": 0.034275355935096743, + "step": 4250 + }, + { + "epoch": 0.6955102040816327, + "grad_norm": 0.09788545221090317, + "learning_rate": 1.092905371795729e-06, + "loss": 0.039912134408950806, + "step": 4260 + }, + { + "epoch": 0.6971428571428572, + "grad_norm": 0.2608252763748169, + "learning_rate": 1.0821497031169729e-06, + "loss": 0.03563873171806335, + "step": 4270 + }, + { + "epoch": 0.6987755102040817, + "grad_norm": 0.10350881516933441, + "learning_rate": 1.0714325836087191e-06, + "loss": 0.036586591601371767, + "step": 4280 + }, + { + "epoch": 0.7004081632653061, + "grad_norm": 0.10214894264936447, + "learning_rate": 1.0607543046529763e-06, + "loss": 0.03775157630443573, + "step": 4290 + }, + { + "epoch": 0.7020408163265306, + "grad_norm": 0.1326843649148941, + "learning_rate": 1.050115156575738e-06, + "loss": 0.03171868622303009, + "step": 4300 + }, + { + "epoch": 0.7036734693877551, + "grad_norm": 0.12274391204118729, + "learning_rate": 1.0395154286390896e-06, + "loss": 0.028169712424278258, + "step": 4310 + }, + { + "epoch": 0.7053061224489796, + "grad_norm": 0.12580160796642303, + "learning_rate": 1.028955409033342e-06, + "loss": 0.03762204349040985, + "step": 4320 + }, + { + "epoch": 0.7069387755102041, + "grad_norm": 0.26383766531944275, + "learning_rate": 1.0184353848692005e-06, + "loss": 0.04035975933074951, + "step": 4330 + }, + { + "epoch": 0.7085714285714285, + "grad_norm": 0.101371631026268, + "learning_rate": 1.0079556421699533e-06, + "loss": 0.04394418001174927, + "step": 4340 + }, + { + "epoch": 0.710204081632653, + "grad_norm": 0.12708403170108795, + "learning_rate": 9.97516465863698e-07, + "loss": 0.04101270735263825, + "step": 4350 + }, + { + "epoch": 0.7118367346938775, + "grad_norm": 0.08838815987110138, + "learning_rate": 9.87118139775595e-07, + "loss": 0.04232141673564911, + "step": 4360 + }, + { + "epoch": 0.713469387755102, + "grad_norm": 0.13831272721290588, + "learning_rate": 9.767609466201494e-07, + "loss": 0.03110981285572052, + "step": 4370 + }, + { + "epoch": 0.7151020408163266, + "grad_norm": 0.14169402420520782, + "learning_rate": 9.664451679935252e-07, + "loss": 0.036100244522094725, + "step": 4380 + }, + { + "epoch": 0.7167346938775511, + "grad_norm": 0.09364918619394302, + "learning_rate": 9.561710843658892e-07, + "loss": 0.028025591373443605, + "step": 4390 + }, + { + "epoch": 0.7183673469387755, + "grad_norm": 0.08178938925266266, + "learning_rate": 9.459389750737854e-07, + "loss": 0.03276517987251282, + "step": 4400 + }, + { + "epoch": 0.7183673469387755, + "eval_accuracy": 0.9830833333333334, + "eval_loss": 0.03968822583556175, + "eval_runtime": 170.9863, + "eval_samples_per_second": 23.394, + "eval_steps_per_second": 23.394, + "step": 4400 + }, + { + "epoch": 0.72, + "grad_norm": 0.21406538784503937, + "learning_rate": 9.357491183125398e-07, + "loss": 0.04240003228187561, + "step": 4410 + }, + { + "epoch": 0.7216326530612245, + "grad_norm": 0.2528787851333618, + "learning_rate": 9.256017911286958e-07, + "loss": 0.03953196704387665, + "step": 4420 + }, + { + "epoch": 0.723265306122449, + "grad_norm": 0.1817523092031479, + "learning_rate": 9.154972694124866e-07, + "loss": 0.03630510270595551, + "step": 4430 + }, + { + "epoch": 0.7248979591836735, + "grad_norm": 0.19527313113212585, + "learning_rate": 9.054358278903272e-07, + "loss": 0.04581010937690735, + "step": 4440 + }, + { + "epoch": 0.726530612244898, + "grad_norm": 0.3551980257034302, + "learning_rate": 8.954177401173491e-07, + "loss": 0.04633161723613739, + "step": 4450 + }, + { + "epoch": 0.7281632653061224, + "grad_norm": 0.1003197506070137, + "learning_rate": 8.854432784699624e-07, + "loss": 0.03686977326869965, + "step": 4460 + }, + { + "epoch": 0.7297959183673469, + "grad_norm": 0.1250677853822708, + "learning_rate": 8.755127141384495e-07, + "loss": 0.03494169116020203, + "step": 4470 + }, + { + "epoch": 0.7314285714285714, + "grad_norm": 0.23732195794582367, + "learning_rate": 8.656263171195931e-07, + "loss": 0.04174305498600006, + "step": 4480 + }, + { + "epoch": 0.7330612244897959, + "grad_norm": 0.2553883194923401, + "learning_rate": 8.557843562093329e-07, + "loss": 0.04263312518596649, + "step": 4490 + }, + { + "epoch": 0.7346938775510204, + "grad_norm": 0.1312720775604248, + "learning_rate": 8.459870989954611e-07, + "loss": 0.03155354261398315, + "step": 4500 + }, + { + "epoch": 0.7363265306122448, + "grad_norm": 0.2632088363170624, + "learning_rate": 8.362348118503424e-07, + "loss": 0.05083600878715515, + "step": 4510 + }, + { + "epoch": 0.7379591836734694, + "grad_norm": 0.1293715387582779, + "learning_rate": 8.265277599236785e-07, + "loss": 0.031102049350738525, + "step": 4520 + }, + { + "epoch": 0.7395918367346939, + "grad_norm": 0.15807363390922546, + "learning_rate": 8.168662071352918e-07, + "loss": 0.033837145566940306, + "step": 4530 + }, + { + "epoch": 0.7412244897959184, + "grad_norm": 0.08791180700063705, + "learning_rate": 8.072504161679534e-07, + "loss": 0.0337544322013855, + "step": 4540 + }, + { + "epoch": 0.7428571428571429, + "grad_norm": 0.10295934230089188, + "learning_rate": 7.976806484602417e-07, + "loss": 0.03291972577571869, + "step": 4550 + }, + { + "epoch": 0.7444897959183674, + "grad_norm": 0.11660359799861908, + "learning_rate": 7.881571641994331e-07, + "loss": 0.032550466060638425, + "step": 4560 + }, + { + "epoch": 0.7461224489795918, + "grad_norm": 0.1896156519651413, + "learning_rate": 7.786802223144286e-07, + "loss": 0.030972787737846376, + "step": 4570 + }, + { + "epoch": 0.7477551020408163, + "grad_norm": 0.2084398865699768, + "learning_rate": 7.69250080468712e-07, + "loss": 0.0415867954492569, + "step": 4580 + }, + { + "epoch": 0.7493877551020408, + "grad_norm": 0.12618854641914368, + "learning_rate": 7.59866995053348e-07, + "loss": 0.0360225111246109, + "step": 4590 + }, + { + "epoch": 0.7510204081632653, + "grad_norm": 0.12291006743907928, + "learning_rate": 7.505312211800081e-07, + "loss": 0.03627174496650696, + "step": 4600 + }, + { + "epoch": 0.7510204081632653, + "eval_accuracy": 0.9855833333333334, + "eval_loss": 0.03568050265312195, + "eval_runtime": 156.9074, + "eval_samples_per_second": 25.493, + "eval_steps_per_second": 25.493, + "step": 4600 + }, + { + "epoch": 0.7526530612244898, + "grad_norm": 0.1491072177886963, + "learning_rate": 7.41243012674035e-07, + "loss": 0.033873635530471805, + "step": 4610 + }, + { + "epoch": 0.7542857142857143, + "grad_norm": 0.14268337190151215, + "learning_rate": 7.320026220675447e-07, + "loss": 0.03949807286262512, + "step": 4620 + }, + { + "epoch": 0.7559183673469387, + "grad_norm": 0.15548810362815857, + "learning_rate": 7.228103005925557e-07, + "loss": 0.03478931188583374, + "step": 4630 + }, + { + "epoch": 0.7575510204081632, + "grad_norm": 0.14298392832279205, + "learning_rate": 7.136662981741613e-07, + "loss": 0.030326524376869203, + "step": 4640 + }, + { + "epoch": 0.7591836734693878, + "grad_norm": 0.11854545772075653, + "learning_rate": 7.045708634237341e-07, + "loss": 0.034136393666267396, + "step": 4650 + }, + { + "epoch": 0.7608163265306123, + "grad_norm": 0.1346571296453476, + "learning_rate": 6.955242436321663e-07, + "loss": 0.040317881107330325, + "step": 4660 + }, + { + "epoch": 0.7624489795918368, + "grad_norm": 0.13217571377754211, + "learning_rate": 6.865266847631463e-07, + "loss": 0.02895120084285736, + "step": 4670 + }, + { + "epoch": 0.7640816326530612, + "grad_norm": 0.0761382132768631, + "learning_rate": 6.775784314464717e-07, + "loss": 0.029726022481918336, + "step": 4680 + }, + { + "epoch": 0.7657142857142857, + "grad_norm": 0.17583715915679932, + "learning_rate": 6.686797269713971e-07, + "loss": 0.02383600324392319, + "step": 4690 + }, + { + "epoch": 0.7673469387755102, + "grad_norm": 0.13434812426567078, + "learning_rate": 6.598308132800221e-07, + "loss": 0.03801000118255615, + "step": 4700 + }, + { + "epoch": 0.7689795918367347, + "grad_norm": 0.0841183140873909, + "learning_rate": 6.510319309607099e-07, + "loss": 0.030053699016571046, + "step": 4710 + }, + { + "epoch": 0.7706122448979592, + "grad_norm": 0.10911499708890915, + "learning_rate": 6.422833192415476e-07, + "loss": 0.03296325206756592, + "step": 4720 + }, + { + "epoch": 0.7722448979591837, + "grad_norm": 0.08562543988227844, + "learning_rate": 6.335852159838423e-07, + "loss": 0.029071077704429626, + "step": 4730 + }, + { + "epoch": 0.7738775510204081, + "grad_norm": 0.24464724957942963, + "learning_rate": 6.249378576756538e-07, + "loss": 0.03307548463344574, + "step": 4740 + }, + { + "epoch": 0.7755102040816326, + "grad_norm": 0.17294660210609436, + "learning_rate": 6.163414794253641e-07, + "loss": 0.03507235944271088, + "step": 4750 + }, + { + "epoch": 0.7771428571428571, + "grad_norm": 0.18331816792488098, + "learning_rate": 6.077963149552868e-07, + "loss": 0.038254004716873166, + "step": 4760 + }, + { + "epoch": 0.7787755102040816, + "grad_norm": 0.10756120830774307, + "learning_rate": 5.993025965953103e-07, + "loss": 0.03428310751914978, + "step": 4770 + }, + { + "epoch": 0.7804081632653062, + "grad_norm": 0.16071061789989471, + "learning_rate": 5.908605552765833e-07, + "loss": 0.032243117690086365, + "step": 4780 + }, + { + "epoch": 0.7820408163265307, + "grad_norm": 0.14897944033145905, + "learning_rate": 5.824704205252343e-07, + "loss": 0.03812848925590515, + "step": 4790 + }, + { + "epoch": 0.7836734693877551, + "grad_norm": 0.1536232829093933, + "learning_rate": 5.741324204561335e-07, + "loss": 0.031389710307121274, + "step": 4800 + }, + { + "epoch": 0.7836734693877551, + "eval_accuracy": 0.985, + "eval_loss": 0.03615976870059967, + "eval_runtime": 153.2713, + "eval_samples_per_second": 26.098, + "eval_steps_per_second": 26.098, + "step": 4800 + }, + { + "epoch": 0.7853061224489796, + "grad_norm": 0.15687023103237152, + "learning_rate": 5.65846781766688e-07, + "loss": 0.027873021364212037, + "step": 4810 + }, + { + "epoch": 0.7869387755102041, + "grad_norm": 0.14541009068489075, + "learning_rate": 5.576137297306794e-07, + "loss": 0.0407156765460968, + "step": 4820 + }, + { + "epoch": 0.7885714285714286, + "grad_norm": 0.16398967802524567, + "learning_rate": 5.494334881921393e-07, + "loss": 0.03936888575553894, + "step": 4830 + }, + { + "epoch": 0.7902040816326531, + "grad_norm": 0.2177809625864029, + "learning_rate": 5.413062795592633e-07, + "loss": 0.03059300184249878, + "step": 4840 + }, + { + "epoch": 0.7918367346938775, + "grad_norm": 0.1487113982439041, + "learning_rate": 5.332323247983632e-07, + "loss": 0.038730043172836306, + "step": 4850 + }, + { + "epoch": 0.793469387755102, + "grad_norm": 0.1111542284488678, + "learning_rate": 5.252118434278605e-07, + "loss": 0.033898717164993285, + "step": 4860 + }, + { + "epoch": 0.7951020408163265, + "grad_norm": 0.10296906530857086, + "learning_rate": 5.172450535123166e-07, + "loss": 0.033659690618515016, + "step": 4870 + }, + { + "epoch": 0.796734693877551, + "grad_norm": 0.11844902485609055, + "learning_rate": 5.093321716565064e-07, + "loss": 0.0293354332447052, + "step": 4880 + }, + { + "epoch": 0.7983673469387755, + "grad_norm": 0.08401476591825485, + "learning_rate": 5.014734129995258e-07, + "loss": 0.027682244777679443, + "step": 4890 + }, + { + "epoch": 0.8, + "grad_norm": 0.15870651602745056, + "learning_rate": 4.936689912089443e-07, + "loss": 0.029118484258651732, + "step": 4900 + }, + { + "epoch": 0.8016326530612244, + "grad_norm": 0.10739132016897202, + "learning_rate": 4.859191184749968e-07, + "loss": 0.031624987721443176, + "step": 4910 + }, + { + "epoch": 0.803265306122449, + "grad_norm": 0.1523323655128479, + "learning_rate": 4.782240055048113e-07, + "loss": 0.03415923118591309, + "step": 4920 + }, + { + "epoch": 0.8048979591836735, + "grad_norm": 0.31860172748565674, + "learning_rate": 4.705838615166836e-07, + "loss": 0.035749971866607666, + "step": 4930 + }, + { + "epoch": 0.806530612244898, + "grad_norm": 0.15043528378009796, + "learning_rate": 4.629988942343863e-07, + "loss": 0.029294773936271667, + "step": 4940 + }, + { + "epoch": 0.8081632653061225, + "grad_norm": 0.1007913127541542, + "learning_rate": 4.55469309881523e-07, + "loss": 0.026727235317230223, + "step": 4950 + }, + { + "epoch": 0.809795918367347, + "grad_norm": 0.10851600021123886, + "learning_rate": 4.479953131759199e-07, + "loss": 0.029011189937591553, + "step": 4960 + }, + { + "epoch": 0.8114285714285714, + "grad_norm": 0.19257904589176178, + "learning_rate": 4.4057710732406113e-07, + "loss": 0.040424668788909913, + "step": 4970 + }, + { + "epoch": 0.8130612244897959, + "grad_norm": 0.08985571563243866, + "learning_rate": 4.332148940155617e-07, + "loss": 0.031291928887367246, + "step": 4980 + }, + { + "epoch": 0.8146938775510204, + "grad_norm": 0.09874805808067322, + "learning_rate": 4.259088734176883e-07, + "loss": 0.028314629197120668, + "step": 4990 + }, + { + "epoch": 0.8163265306122449, + "grad_norm": 0.2056759148836136, + "learning_rate": 4.186592441699119e-07, + "loss": 0.028049173951148986, + "step": 5000 + }, + { + "epoch": 0.8163265306122449, + "eval_accuracy": 0.986125, + "eval_loss": 0.033585112541913986, + "eval_runtime": 153.2119, + "eval_samples_per_second": 26.108, + "eval_steps_per_second": 26.108, + "step": 5000 + }, + { + "epoch": 0.8179591836734694, + "grad_norm": 0.10125680267810822, + "learning_rate": 4.1146620337850965e-07, + "loss": 0.03612466752529144, + "step": 5010 + }, + { + "epoch": 0.8195918367346938, + "grad_norm": 0.10107982903718948, + "learning_rate": 4.043299466112058e-07, + "loss": 0.026968371868133546, + "step": 5020 + }, + { + "epoch": 0.8212244897959183, + "grad_norm": 0.13236716389656067, + "learning_rate": 3.9725066789185446e-07, + "loss": 0.030766445398330688, + "step": 5030 + }, + { + "epoch": 0.8228571428571428, + "grad_norm": 0.1182790994644165, + "learning_rate": 3.9022855969516384e-07, + "loss": 0.028759539127349854, + "step": 5040 + }, + { + "epoch": 0.8244897959183674, + "grad_norm": 0.1430918574333191, + "learning_rate": 3.832638129414637e-07, + "loss": 0.03126598298549652, + "step": 5050 + }, + { + "epoch": 0.8261224489795919, + "grad_norm": 0.24944688379764557, + "learning_rate": 3.7635661699151366e-07, + "loss": 0.030412834882736207, + "step": 5060 + }, + { + "epoch": 0.8277551020408164, + "grad_norm": 0.24371539056301117, + "learning_rate": 3.695071596413574e-07, + "loss": 0.03904336988925934, + "step": 5070 + }, + { + "epoch": 0.8293877551020408, + "grad_norm": 0.25066161155700684, + "learning_rate": 3.627156271172125e-07, + "loss": 0.032776132225990295, + "step": 5080 + }, + { + "epoch": 0.8310204081632653, + "grad_norm": 0.1635182797908783, + "learning_rate": 3.559822040704108e-07, + "loss": 0.027044805884361266, + "step": 5090 + }, + { + "epoch": 0.8326530612244898, + "grad_norm": 0.16698378324508667, + "learning_rate": 3.4930707357237666e-07, + "loss": 0.03271254003047943, + "step": 5100 + }, + { + "epoch": 0.8342857142857143, + "grad_norm": 0.18812412023544312, + "learning_rate": 3.4269041710964903e-07, + "loss": 0.030745366215705873, + "step": 5110 + }, + { + "epoch": 0.8359183673469388, + "grad_norm": 0.18739669024944305, + "learning_rate": 3.36132414578948e-07, + "loss": 0.032180759310722354, + "step": 5120 + }, + { + "epoch": 0.8375510204081633, + "grad_norm": 0.13260763883590698, + "learning_rate": 3.296332442822839e-07, + "loss": 0.026755687594413758, + "step": 5130 + }, + { + "epoch": 0.8391836734693877, + "grad_norm": 0.06771053373813629, + "learning_rate": 3.231930829221083e-07, + "loss": 0.02895594835281372, + "step": 5140 + }, + { + "epoch": 0.8408163265306122, + "grad_norm": 0.09441749006509781, + "learning_rate": 3.168121055965112e-07, + "loss": 0.033178862929344174, + "step": 5150 + }, + { + "epoch": 0.8424489795918367, + "grad_norm": 0.0863656997680664, + "learning_rate": 3.104904857944591e-07, + "loss": 0.02276728004217148, + "step": 5160 + }, + { + "epoch": 0.8440816326530612, + "grad_norm": 0.2047131359577179, + "learning_rate": 3.0422839539107794e-07, + "loss": 0.028258222341537475, + "step": 5170 + }, + { + "epoch": 0.8457142857142858, + "grad_norm": 0.2389051467180252, + "learning_rate": 2.980260046429834e-07, + "loss": 0.02725991904735565, + "step": 5180 + }, + { + "epoch": 0.8473469387755102, + "grad_norm": 0.1733924299478531, + "learning_rate": 2.9188348218364635e-07, + "loss": 0.0327379584312439, + "step": 5190 + }, + { + "epoch": 0.8489795918367347, + "grad_norm": 0.19450697302818298, + "learning_rate": 2.8580099501881285e-07, + "loss": 0.030778995156288146, + "step": 5200 + }, + { + "epoch": 0.8489795918367347, + "eval_accuracy": 0.9881666666666667, + "eval_loss": 0.02920014224946499, + "eval_runtime": 155.6487, + "eval_samples_per_second": 25.699, + "eval_steps_per_second": 25.699, + "step": 5200 + }, + { + "epoch": 0.8506122448979592, + "grad_norm": 0.14990773797035217, + "learning_rate": 2.7977870852195997e-07, + "loss": 0.03692179024219513, + "step": 5210 + }, + { + "epoch": 0.8522448979591837, + "grad_norm": 0.16051384806632996, + "learning_rate": 2.7381678642980303e-07, + "loss": 0.030528882145881654, + "step": 5220 + }, + { + "epoch": 0.8538775510204082, + "grad_norm": 0.1660623997449875, + "learning_rate": 2.6791539083784014e-07, + "loss": 0.025125756859779358, + "step": 5230 + }, + { + "epoch": 0.8555102040816327, + "grad_norm": 0.07568248361349106, + "learning_rate": 2.620746821959475e-07, + "loss": 0.02720560133457184, + "step": 5240 + }, + { + "epoch": 0.8571428571428571, + "grad_norm": 0.2634601593017578, + "learning_rate": 2.562948193040174e-07, + "loss": 0.033940815925598146, + "step": 5250 + }, + { + "epoch": 0.8587755102040816, + "grad_norm": 0.15926218032836914, + "learning_rate": 2.5057595930763873e-07, + "loss": 0.028036150336265563, + "step": 5260 + }, + { + "epoch": 0.8604081632653061, + "grad_norm": 0.19775289297103882, + "learning_rate": 2.4491825769382526e-07, + "loss": 0.02731768786907196, + "step": 5270 + }, + { + "epoch": 0.8620408163265306, + "grad_norm": 0.07538789510726929, + "learning_rate": 2.39321868286789e-07, + "loss": 0.0313712328672409, + "step": 5280 + }, + { + "epoch": 0.8636734693877551, + "grad_norm": 0.10647427290678024, + "learning_rate": 2.337869432437567e-07, + "loss": 0.02781860828399658, + "step": 5290 + }, + { + "epoch": 0.8653061224489796, + "grad_norm": 0.20678064227104187, + "learning_rate": 2.283136330508337e-07, + "loss": 0.028782907128334045, + "step": 5300 + }, + { + "epoch": 0.866938775510204, + "grad_norm": 0.18386024236679077, + "learning_rate": 2.2290208651891244e-07, + "loss": 0.02515394687652588, + "step": 5310 + }, + { + "epoch": 0.8685714285714285, + "grad_norm": 0.13139928877353668, + "learning_rate": 2.1755245077962618e-07, + "loss": 0.02002216875553131, + "step": 5320 + }, + { + "epoch": 0.8702040816326531, + "grad_norm": 0.10402868688106537, + "learning_rate": 2.1226487128134854e-07, + "loss": 0.03221684396266937, + "step": 5330 + }, + { + "epoch": 0.8718367346938776, + "grad_norm": 0.056439850479364395, + "learning_rate": 2.070394917852403e-07, + "loss": 0.023763810098171235, + "step": 5340 + }, + { + "epoch": 0.8734693877551021, + "grad_norm": 0.1421876698732376, + "learning_rate": 2.0187645436133913e-07, + "loss": 0.03192594051361084, + "step": 5350 + }, + { + "epoch": 0.8751020408163265, + "grad_norm": 0.20487430691719055, + "learning_rate": 1.967758993846966e-07, + "loss": 0.025955894589424135, + "step": 5360 + }, + { + "epoch": 0.876734693877551, + "grad_norm": 0.2183534801006317, + "learning_rate": 1.9173796553156487e-07, + "loss": 0.027435585856437683, + "step": 5370 + }, + { + "epoch": 0.8783673469387755, + "grad_norm": 0.19724565744400024, + "learning_rate": 1.8676278977562219e-07, + "loss": 0.028912261128425598, + "step": 5380 + }, + { + "epoch": 0.88, + "grad_norm": 0.21412917971611023, + "learning_rate": 1.818505073842508e-07, + "loss": 0.029015693068504333, + "step": 5390 + }, + { + "epoch": 0.8816326530612245, + "grad_norm": 0.08429288119077682, + "learning_rate": 1.770012519148598e-07, + "loss": 0.025249993801116942, + "step": 5400 + }, + { + "epoch": 0.8816326530612245, + "eval_accuracy": 0.989375, + "eval_loss": 0.02678457647562027, + "eval_runtime": 156.2763, + "eval_samples_per_second": 25.596, + "eval_steps_per_second": 25.596, + "step": 5400 + }, + { + "epoch": 0.883265306122449, + "grad_norm": 0.15489192306995392, + "learning_rate": 1.722151552112525e-07, + "loss": 0.03486272096633911, + "step": 5410 + }, + { + "epoch": 0.8848979591836734, + "grad_norm": 0.1594313681125641, + "learning_rate": 1.6749234740004227e-07, + "loss": 0.024751077592372894, + "step": 5420 + }, + { + "epoch": 0.8865306122448979, + "grad_norm": 0.12505307793617249, + "learning_rate": 1.6283295688711564e-07, + "loss": 0.025553828477859496, + "step": 5430 + }, + { + "epoch": 0.8881632653061224, + "grad_norm": 0.08852038532495499, + "learning_rate": 1.582371103541394e-07, + "loss": 0.025743791460990907, + "step": 5440 + }, + { + "epoch": 0.889795918367347, + "grad_norm": 0.08450407534837723, + "learning_rate": 1.5370493275511767e-07, + "loss": 0.02687264084815979, + "step": 5450 + }, + { + "epoch": 0.8914285714285715, + "grad_norm": 0.054211657494306564, + "learning_rate": 1.4923654731299363e-07, + "loss": 0.023073682188987733, + "step": 5460 + }, + { + "epoch": 0.893061224489796, + "grad_norm": 0.23101477324962616, + "learning_rate": 1.4483207551630045e-07, + "loss": 0.0280561625957489, + "step": 5470 + }, + { + "epoch": 0.8946938775510204, + "grad_norm": 0.19923920929431915, + "learning_rate": 1.4049163711585684e-07, + "loss": 0.02133965492248535, + "step": 5480 + }, + { + "epoch": 0.8963265306122449, + "grad_norm": 0.13227398693561554, + "learning_rate": 1.3621535012151244e-07, + "loss": 0.017637649178504945, + "step": 5490 + }, + { + "epoch": 0.8979591836734694, + "grad_norm": 0.11182940751314163, + "learning_rate": 1.3200333079893823e-07, + "loss": 0.02847398817539215, + "step": 5500 + }, + { + "epoch": 0.8995918367346939, + "grad_norm": 0.06762958317995071, + "learning_rate": 1.2785569366646628e-07, + "loss": 0.020397891104221345, + "step": 5510 + }, + { + "epoch": 0.9012244897959184, + "grad_norm": 0.06652984023094177, + "learning_rate": 1.2377255149197576e-07, + "loss": 0.03015478849411011, + "step": 5520 + }, + { + "epoch": 0.9028571428571428, + "grad_norm": 0.16087517142295837, + "learning_rate": 1.1975401528982718e-07, + "loss": 0.02614833414554596, + "step": 5530 + }, + { + "epoch": 0.9044897959183673, + "grad_norm": 0.0873415544629097, + "learning_rate": 1.1580019431784384e-07, + "loss": 0.023097904026508333, + "step": 5540 + }, + { + "epoch": 0.9061224489795918, + "grad_norm": 0.26114869117736816, + "learning_rate": 1.1191119607434076e-07, + "loss": 0.028036117553710938, + "step": 5550 + }, + { + "epoch": 0.9077551020408163, + "grad_norm": 0.08690405637025833, + "learning_rate": 1.0808712629520435e-07, + "loss": 0.027321887016296387, + "step": 5560 + }, + { + "epoch": 0.9093877551020408, + "grad_norm": 0.19979159533977509, + "learning_rate": 1.0432808895101426e-07, + "loss": 0.03543014228343964, + "step": 5570 + }, + { + "epoch": 0.9110204081632653, + "grad_norm": 0.09599421173334122, + "learning_rate": 1.0063418624421895e-07, + "loss": 0.02044544517993927, + "step": 5580 + }, + { + "epoch": 0.9126530612244897, + "grad_norm": 0.22291070222854614, + "learning_rate": 9.700551860635538e-08, + "loss": 0.027166444063186645, + "step": 5590 + }, + { + "epoch": 0.9142857142857143, + "grad_norm": 0.10956079512834549, + "learning_rate": 9.344218469532012e-08, + "loss": 0.02024113982915878, + "step": 5600 + }, + { + "epoch": 0.9142857142857143, + "eval_accuracy": 0.9900833333333334, + "eval_loss": 0.025807492434978485, + "eval_runtime": 154.9378, + "eval_samples_per_second": 25.817, + "eval_steps_per_second": 25.817, + "step": 5600 + }, + { + "epoch": 0.9159183673469388, + "grad_norm": 0.14949144423007965, + "learning_rate": 8.99442813926854e-08, + "loss": 0.024004274606704713, + "step": 5610 + }, + { + "epoch": 0.9175510204081633, + "grad_norm": 0.2315555214881897, + "learning_rate": 8.651190380106673e-08, + "loss": 0.024219857156276704, + "step": 5620 + }, + { + "epoch": 0.9191836734693878, + "grad_norm": 0.06734409183263779, + "learning_rate": 8.314514524153522e-08, + "loss": 0.02802177369594574, + "step": 5630 + }, + { + "epoch": 0.9208163265306123, + "grad_norm": 0.21427671611309052, + "learning_rate": 7.984409725108222e-08, + "loss": 0.021109704673290253, + "step": 5640 + }, + { + "epoch": 0.9224489795918367, + "grad_norm": 0.18013973534107208, + "learning_rate": 7.660884958012899e-08, + "loss": 0.020703735947608947, + "step": 5650 + }, + { + "epoch": 0.9240816326530612, + "grad_norm": 0.2695794701576233, + "learning_rate": 7.343949019008756e-08, + "loss": 0.03247661292552948, + "step": 5660 + }, + { + "epoch": 0.9257142857142857, + "grad_norm": 0.15175238251686096, + "learning_rate": 7.03361052509688e-08, + "loss": 0.021647430956363678, + "step": 5670 + }, + { + "epoch": 0.9273469387755102, + "grad_norm": 0.18325908482074738, + "learning_rate": 6.729877913903976e-08, + "loss": 0.028400209546089173, + "step": 5680 + }, + { + "epoch": 0.9289795918367347, + "grad_norm": 0.15859369933605194, + "learning_rate": 6.432759443452924e-08, + "loss": 0.014232246577739716, + "step": 5690 + }, + { + "epoch": 0.9306122448979591, + "grad_norm": 0.13438236713409424, + "learning_rate": 6.14226319193828e-08, + "loss": 0.02244553864002228, + "step": 5700 + }, + { + "epoch": 0.9322448979591836, + "grad_norm": 0.27041247487068176, + "learning_rate": 5.858397057506709e-08, + "loss": 0.029011306166648865, + "step": 5710 + }, + { + "epoch": 0.9338775510204081, + "grad_norm": 0.12215204536914825, + "learning_rate": 5.581168758042127e-08, + "loss": 0.028310227394104003, + "step": 5720 + }, + { + "epoch": 0.9355102040816327, + "grad_norm": 0.2103251814842224, + "learning_rate": 5.310585830955922e-08, + "loss": 0.021468037366867067, + "step": 5730 + }, + { + "epoch": 0.9371428571428572, + "grad_norm": 0.17152205109596252, + "learning_rate": 5.0466556329819835e-08, + "loss": 0.024149742722511292, + "step": 5740 + }, + { + "epoch": 0.9387755102040817, + "grad_norm": 0.08712399005889893, + "learning_rate": 4.789385339976832e-08, + "loss": 0.01809695065021515, + "step": 5750 + }, + { + "epoch": 0.9404081632653061, + "grad_norm": 0.15006859600543976, + "learning_rate": 4.53878194672433e-08, + "loss": 0.022831857204437256, + "step": 5760 + }, + { + "epoch": 0.9420408163265306, + "grad_norm": 0.06355110555887222, + "learning_rate": 4.2948522667455864e-08, + "loss": 0.015157052874565124, + "step": 5770 + }, + { + "epoch": 0.9436734693877551, + "grad_norm": 0.14330919086933136, + "learning_rate": 4.057602932113658e-08, + "loss": 0.02486245185136795, + "step": 5780 + }, + { + "epoch": 0.9453061224489796, + "grad_norm": 0.12216096371412277, + "learning_rate": 3.82704039327339e-08, + "loss": 0.024203602969646455, + "step": 5790 + }, + { + "epoch": 0.9469387755102041, + "grad_norm": 0.19883109629154205, + "learning_rate": 3.6031709188658336e-08, + "loss": 0.025162199139595033, + "step": 5800 + }, + { + "epoch": 0.9469387755102041, + "eval_accuracy": 0.99025, + "eval_loss": 0.025506427511572838, + "eval_runtime": 154.1326, + "eval_samples_per_second": 25.952, + "eval_steps_per_second": 25.952, + "step": 5800 + }, + { + "epoch": 0.9485714285714286, + "grad_norm": 0.25931984186172485, + "learning_rate": 3.386000595557964e-08, + "loss": 0.027550938725471496, + "step": 5810 + }, + { + "epoch": 0.950204081632653, + "grad_norm": 0.07933743298053741, + "learning_rate": 3.1755353278770936e-08, + "loss": 0.024577364325523376, + "step": 5820 + }, + { + "epoch": 0.9518367346938775, + "grad_norm": 0.13839955627918243, + "learning_rate": 2.9717808380503866e-08, + "loss": 0.021038763225078583, + "step": 5830 + }, + { + "epoch": 0.953469387755102, + "grad_norm": 0.14696243405342102, + "learning_rate": 2.7747426658492893e-08, + "loss": 0.02484375685453415, + "step": 5840 + }, + { + "epoch": 0.9551020408163265, + "grad_norm": 0.1538810133934021, + "learning_rate": 2.5844261684388473e-08, + "loss": 0.023458963632583617, + "step": 5850 + }, + { + "epoch": 0.9567346938775511, + "grad_norm": 0.08468759804964066, + "learning_rate": 2.4008365202321525e-08, + "loss": 0.03132525682449341, + "step": 5860 + }, + { + "epoch": 0.9583673469387755, + "grad_norm": 0.1627626270055771, + "learning_rate": 2.2239787127495695e-08, + "loss": 0.03181464076042175, + "step": 5870 + }, + { + "epoch": 0.96, + "grad_norm": 0.17413999140262604, + "learning_rate": 2.0538575544831186e-08, + "loss": 0.019847501814365388, + "step": 5880 + }, + { + "epoch": 0.9616326530612245, + "grad_norm": 0.23201079666614532, + "learning_rate": 1.8904776707655837e-08, + "loss": 0.02205444723367691, + "step": 5890 + }, + { + "epoch": 0.963265306122449, + "grad_norm": 0.17272613942623138, + "learning_rate": 1.7338435036449164e-08, + "loss": 0.022021512687206268, + "step": 5900 + }, + { + "epoch": 0.9648979591836735, + "grad_norm": 0.1446181833744049, + "learning_rate": 1.5839593117634166e-08, + "loss": 0.02300398051738739, + "step": 5910 + }, + { + "epoch": 0.966530612244898, + "grad_norm": 0.21913999319076538, + "learning_rate": 1.4408291702418808e-08, + "loss": 0.024825331568717957, + "step": 5920 + }, + { + "epoch": 0.9681632653061224, + "grad_norm": 0.12140090018510818, + "learning_rate": 1.3044569705689125e-08, + "loss": 0.02171301543712616, + "step": 5930 + }, + { + "epoch": 0.9697959183673469, + "grad_norm": 0.08353517204523087, + "learning_rate": 1.1748464204950072e-08, + "loss": 0.02564157545566559, + "step": 5940 + }, + { + "epoch": 0.9714285714285714, + "grad_norm": 0.2671544551849365, + "learning_rate": 1.052001043931855e-08, + "loss": 0.02467280924320221, + "step": 5950 + }, + { + "epoch": 0.9730612244897959, + "grad_norm": 0.16360966861248016, + "learning_rate": 9.3592418085639e-09, + "loss": 0.02429400235414505, + "step": 5960 + }, + { + "epoch": 0.9746938775510204, + "grad_norm": 0.08739493787288666, + "learning_rate": 8.266189872201957e-09, + "loss": 0.02331821173429489, + "step": 5970 + }, + { + "epoch": 0.976326530612245, + "grad_norm": 0.23632948100566864, + "learning_rate": 7.24088434863407e-09, + "loss": 0.02990926504135132, + "step": 5980 + }, + { + "epoch": 0.9779591836734693, + "grad_norm": 0.2005041390657425, + "learning_rate": 6.283353114342749e-09, + "loss": 0.023541297018527984, + "step": 5990 + }, + { + "epoch": 0.9795918367346939, + "grad_norm": 0.16744142770767212, + "learning_rate": 5.393622203129778e-09, + "loss": 0.020145678520202638, + "step": 6000 + }, + { + "epoch": 0.9795918367346939, + "eval_accuracy": 0.9904166666666667, + "eval_loss": 0.025301460176706314, + "eval_runtime": 157.9993, + "eval_samples_per_second": 25.317, + "eval_steps_per_second": 25.317, + "step": 6000 + }, + { + "epoch": 0.9812244897959184, + "grad_norm": 0.1079549789428711, + "learning_rate": 4.5717158054114916e-09, + "loss": 0.02092682272195816, + "step": 6010 + }, + { + "epoch": 0.9828571428571429, + "grad_norm": 0.28545212745666504, + "learning_rate": 3.817656267559866e-09, + "loss": 0.029862532019615175, + "step": 6020 + }, + { + "epoch": 0.9844897959183674, + "grad_norm": 0.09164315462112427, + "learning_rate": 3.1314640912943893e-09, + "loss": 0.021963220834732056, + "step": 6030 + }, + { + "epoch": 0.9861224489795918, + "grad_norm": 0.10882832109928131, + "learning_rate": 2.5131579331261204e-09, + "loss": 0.01937238872051239, + "step": 6040 + }, + { + "epoch": 0.9877551020408163, + "grad_norm": 0.2565634846687317, + "learning_rate": 1.9627546038492042e-09, + "loss": 0.03262064456939697, + "step": 6050 + }, + { + "epoch": 0.9893877551020408, + "grad_norm": 0.20641514658927917, + "learning_rate": 1.4802690680842947e-09, + "loss": 0.02464511841535568, + "step": 6060 + }, + { + "epoch": 0.9910204081632653, + "grad_norm": 0.15149660408496857, + "learning_rate": 1.065714443871102e-09, + "loss": 0.020028088986873627, + "step": 6070 + }, + { + "epoch": 0.9926530612244898, + "grad_norm": 0.24861690402030945, + "learning_rate": 7.191020023133988e-10, + "loss": 0.03540560901165009, + "step": 6080 + }, + { + "epoch": 0.9942857142857143, + "grad_norm": 0.1762499213218689, + "learning_rate": 4.404411672703779e-10, + "loss": 0.026075825095176697, + "step": 6090 + }, + { + "epoch": 0.9959183673469387, + "grad_norm": 0.23214563727378845, + "learning_rate": 2.2973951510185665e-10, + "loss": 0.024257904291152953, + "step": 6100 + }, + { + "epoch": 0.9975510204081632, + "grad_norm": 0.13295525312423706, + "learning_rate": 8.700277446233029e-11, + "loss": 0.021962395310401915, + "step": 6110 + }, + { + "epoch": 0.9991836734693877, + "grad_norm": 0.1011875718832016, + "learning_rate": 1.2234826143875388e-11, + "loss": 0.025374814867973328, + "step": 6120 + }, + { + "epoch": 1.0, + "step": 6125, + "total_flos": 5.72407910760448e+16, + "train_loss": 0.07194560643361539, + "train_runtime": 34224.4204, + "train_samples_per_second": 5.727, + "train_steps_per_second": 0.179 + } + ], + "logging_steps": 10, + "max_steps": 6125, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 200, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 5.72407910760448e+16, + "train_batch_size": 1, + "trial_name": null, + "trial_params": null +} diff --git a/training_args.bin b/training_args.bin new file mode 100644 index 0000000..a9304c0 --- /dev/null +++ b/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:10a2085a9c347d21774d35a2558a0376b0ffbb33ca9517fafbd4cd5d12552b02 +size 5585 diff --git a/training_eval_accuracy.png b/training_eval_accuracy.png new file mode 100644 index 0000000..d380359 Binary files /dev/null and b/training_eval_accuracy.png differ diff --git a/training_eval_loss.png b/training_eval_loss.png new file mode 100644 index 0000000..337dda6 Binary files /dev/null and b/training_eval_loss.png differ diff --git a/training_loss.png b/training_loss.png new file mode 100644 index 0000000..1e9516f Binary files /dev/null and b/training_loss.png differ