| { | |
| "best_metric": 0.7759295499021527, | |
| "best_model_checkpoint": "tiny-bert-sst2-distilled/run-0/checkpoint-840", | |
| "epoch": 9.0, | |
| "eval_steps": 500, | |
| "global_step": 1260, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.9402021765708923, | |
| "learning_rate": 2.0126613681226377e-05, | |
| "loss": 0.4767, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.5, | |
| "eval_f1": 0.0, | |
| "eval_loss": 0.46569836139678955, | |
| "eval_precision": 0.0, | |
| "eval_recall": 0.0, | |
| "eval_runtime": 28.6101, | |
| "eval_samples_per_second": 35.722, | |
| "eval_steps_per_second": 1.118, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 2.0842678546905518, | |
| "learning_rate": 1.761078697107308e-05, | |
| "loss": 0.4503, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.5851272015655578, | |
| "eval_f1": 0.3476923076923077, | |
| "eval_loss": 0.41868579387664795, | |
| "eval_precision": 0.8129496402877698, | |
| "eval_recall": 0.22113502935420742, | |
| "eval_runtime": 29.7024, | |
| "eval_samples_per_second": 34.408, | |
| "eval_steps_per_second": 1.077, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 3.5170960426330566, | |
| "learning_rate": 1.5094960260919783e-05, | |
| "loss": 0.4106, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.6986301369863014, | |
| "eval_f1": 0.6426914153132252, | |
| "eval_loss": 0.3922746479511261, | |
| "eval_precision": 0.7891737891737892, | |
| "eval_recall": 0.5420743639921722, | |
| "eval_runtime": 28.6537, | |
| "eval_samples_per_second": 35.667, | |
| "eval_steps_per_second": 1.117, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 5.289376735687256, | |
| "learning_rate": 1.2579133550766487e-05, | |
| "loss": 0.3982, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.723091976516634, | |
| "eval_f1": 0.6920565832426551, | |
| "eval_loss": 0.38318684697151184, | |
| "eval_precision": 0.7794117647058824, | |
| "eval_recall": 0.6223091976516634, | |
| "eval_runtime": 28.7822, | |
| "eval_samples_per_second": 35.508, | |
| "eval_steps_per_second": 1.112, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 4.148696422576904, | |
| "learning_rate": 1.0063306840613188e-05, | |
| "loss": 0.3917, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.7524461839530333, | |
| "eval_f1": 0.7420998980632009, | |
| "eval_loss": 0.3796561658382416, | |
| "eval_precision": 0.774468085106383, | |
| "eval_recall": 0.7123287671232876, | |
| "eval_runtime": 29.1866, | |
| "eval_samples_per_second": 35.016, | |
| "eval_steps_per_second": 1.096, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 4.664227485656738, | |
| "learning_rate": 7.547480130459891e-06, | |
| "loss": 0.3905, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.7759295499021527, | |
| "eval_f1": 0.7800192122958692, | |
| "eval_loss": 0.3800524175167084, | |
| "eval_precision": 0.7660377358490567, | |
| "eval_recall": 0.7945205479452054, | |
| "eval_runtime": 29.0066, | |
| "eval_samples_per_second": 35.233, | |
| "eval_steps_per_second": 1.103, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 3.6408803462982178, | |
| "learning_rate": 5.031653420306594e-06, | |
| "loss": 0.3881, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.7583170254403131, | |
| "eval_f1": 0.7522567703109329, | |
| "eval_loss": 0.37579989433288574, | |
| "eval_precision": 0.7716049382716049, | |
| "eval_recall": 0.7338551859099804, | |
| "eval_runtime": 28.8723, | |
| "eval_samples_per_second": 35.397, | |
| "eval_steps_per_second": 1.108, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 5.963382720947266, | |
| "learning_rate": 2.515826710153297e-06, | |
| "loss": 0.3872, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.7651663405088063, | |
| "eval_f1": 0.763779527559055, | |
| "eval_loss": 0.37569984793663025, | |
| "eval_precision": 0.7683168316831683, | |
| "eval_recall": 0.7592954990215264, | |
| "eval_runtime": 28.874, | |
| "eval_samples_per_second": 35.395, | |
| "eval_steps_per_second": 1.108, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 2.9294092655181885, | |
| "learning_rate": 0.0, | |
| "loss": 0.3822, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.7700587084148728, | |
| "eval_f1": 0.7711781888997078, | |
| "eval_loss": 0.3755285441875458, | |
| "eval_precision": 0.7674418604651163, | |
| "eval_recall": 0.7749510763209393, | |
| "eval_runtime": 28.6166, | |
| "eval_samples_per_second": 35.714, | |
| "eval_steps_per_second": 1.118, | |
| "step": 1260 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 1260, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 9, | |
| "save_steps": 500, | |
| "total_flos": 2121256775520.0, | |
| "train_batch_size": 22, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.6127720075177469, | |
| "learning_rate": 2.2642440391379676e-05, | |
| "num_train_epochs": 9, | |
| "per_device_train_batch_size": 22, | |
| "temperature": 21 | |
| } | |
| } | |