| { | |
| "best_metric": 0.8131115459882583, | |
| "best_model_checkpoint": "tiny-bert-sst2-distilled/run-0/checkpoint-1344", | |
| "epoch": 10.0, | |
| "eval_steps": 500, | |
| "global_step": 1920, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.7846055030822754, | |
| "learning_rate": 4.610177667546352e-05, | |
| "loss": 0.4587, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.6976516634050881, | |
| "eval_f1": 0.6419466975666281, | |
| "eval_loss": 0.4008609354496002, | |
| "eval_mcc": 0.41595144404027, | |
| "eval_precision": 0.7869318181818182, | |
| "eval_recall": 0.5420743639921722, | |
| "eval_runtime": 66.5688, | |
| "eval_samples_per_second": 15.353, | |
| "eval_steps_per_second": 0.481, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 6.001947402954102, | |
| "learning_rate": 4.0979357044856464e-05, | |
| "loss": 0.4016, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.7181996086105675, | |
| "eval_f1": 0.6778523489932886, | |
| "eval_loss": 0.3820359408855438, | |
| "eval_mcc": 0.4507700437564526, | |
| "eval_precision": 0.7911227154046997, | |
| "eval_recall": 0.5929549902152642, | |
| "eval_runtime": 66.6005, | |
| "eval_samples_per_second": 15.345, | |
| "eval_steps_per_second": 0.48, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 2.8345351219177246, | |
| "learning_rate": 3.5856937414249405e-05, | |
| "loss": 0.3879, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8062622309197651, | |
| "eval_f1": 0.8159851301115242, | |
| "eval_loss": 0.37824752926826477, | |
| "eval_mcc": 0.6159734786715062, | |
| "eval_precision": 0.7769911504424779, | |
| "eval_recall": 0.8590998043052838, | |
| "eval_runtime": 66.7443, | |
| "eval_samples_per_second": 15.312, | |
| "eval_steps_per_second": 0.479, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 2.9280295372009277, | |
| "learning_rate": 3.0734517783642345e-05, | |
| "loss": 0.3803, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.7681017612524462, | |
| "eval_f1": 0.7569230769230769, | |
| "eval_loss": 0.37255340814590454, | |
| "eval_mcc": 0.5384860699767601, | |
| "eval_precision": 0.7952586206896551, | |
| "eval_recall": 0.7221135029354208, | |
| "eval_runtime": 67.0029, | |
| "eval_samples_per_second": 15.253, | |
| "eval_steps_per_second": 0.478, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 3.7051830291748047, | |
| "learning_rate": 2.561209815303529e-05, | |
| "loss": 0.3773, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.7964774951076321, | |
| "eval_f1": 0.7932405566600398, | |
| "eval_loss": 0.3688918650150299, | |
| "eval_mcc": 0.5932458672098203, | |
| "eval_precision": 0.806060606060606, | |
| "eval_recall": 0.7808219178082192, | |
| "eval_runtime": 68.0849, | |
| "eval_samples_per_second": 15.011, | |
| "eval_steps_per_second": 0.47, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 3.5727412700653076, | |
| "learning_rate": 2.0489678522428232e-05, | |
| "loss": 0.3741, | |
| "step": 1152 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.8101761252446184, | |
| "eval_f1": 0.8197026022304833, | |
| "eval_loss": 0.37157630920410156, | |
| "eval_mcc": 0.623845344213634, | |
| "eval_precision": 0.7805309734513274, | |
| "eval_recall": 0.863013698630137, | |
| "eval_runtime": 66.833, | |
| "eval_samples_per_second": 15.292, | |
| "eval_steps_per_second": 0.479, | |
| "step": 1152 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 1.87214994430542, | |
| "learning_rate": 1.5367258891821172e-05, | |
| "loss": 0.3712, | |
| "step": 1344 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.8131115459882583, | |
| "eval_f1": 0.8203198494825964, | |
| "eval_loss": 0.3685600161552429, | |
| "eval_mcc": 0.6282485725306769, | |
| "eval_precision": 0.7898550724637681, | |
| "eval_recall": 0.8532289628180039, | |
| "eval_runtime": 66.8069, | |
| "eval_samples_per_second": 15.298, | |
| "eval_steps_per_second": 0.479, | |
| "step": 1344 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 2.601350784301758, | |
| "learning_rate": 1.0244839261214116e-05, | |
| "loss": 0.3667, | |
| "step": 1536 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.8091976516634051, | |
| "eval_f1": 0.8179271708683473, | |
| "eval_loss": 0.36975762248039246, | |
| "eval_mcc": 0.6212581246070178, | |
| "eval_precision": 0.7821428571428571, | |
| "eval_recall": 0.8571428571428571, | |
| "eval_runtime": 67.182, | |
| "eval_samples_per_second": 15.212, | |
| "eval_steps_per_second": 0.476, | |
| "step": 1536 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 1.19636869430542, | |
| "learning_rate": 5.122419630607058e-06, | |
| "loss": 0.3678, | |
| "step": 1728 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.8062622309197651, | |
| "eval_f1": 0.8062622309197651, | |
| "eval_loss": 0.36616405844688416, | |
| "eval_mcc": 0.6125244618395304, | |
| "eval_precision": 0.8062622309197651, | |
| "eval_recall": 0.8062622309197651, | |
| "eval_runtime": 67.13, | |
| "eval_samples_per_second": 15.224, | |
| "eval_steps_per_second": 0.477, | |
| "step": 1728 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 3.5874884128570557, | |
| "learning_rate": 0.0, | |
| "loss": 0.3677, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.8052837573385518, | |
| "eval_f1": 0.8069835111542193, | |
| "eval_loss": 0.3663344383239746, | |
| "eval_mcc": 0.610662236046953, | |
| "eval_precision": 0.8, | |
| "eval_recall": 0.8140900195694716, | |
| "eval_runtime": 67.1725, | |
| "eval_samples_per_second": 15.215, | |
| "eval_steps_per_second": 0.476, | |
| "step": 1920 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 1920, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "total_flos": 2356951972800.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.6226480553446316, | |
| "learning_rate": 5.122419630607058e-05, | |
| "num_train_epochs": 10, | |
| "per_device_train_batch_size": 16, | |
| "temperature": 20 | |
| } | |
| } | |