| { | |
| "best_metric": 0.773286467486819, | |
| "best_model_checkpoint": "tiny-bert-sst2-distilled/run-31/checkpoint-2140", | |
| "epoch": 10.0, | |
| "eval_steps": 500, | |
| "global_step": 2140, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 7.312666416168213, | |
| "learning_rate": 0.00024913095817095677, | |
| "loss": 0.5995, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.715875805506737, | |
| "eval_f1": 0.4300822561692127, | |
| "eval_loss": 0.5564725995063782, | |
| "eval_mcc": 0.2978271442275633, | |
| "eval_precision": 0.648936170212766, | |
| "eval_recall": 0.3216168717047452, | |
| "eval_runtime": 3.1678, | |
| "eval_samples_per_second": 538.866, | |
| "eval_steps_per_second": 17.047, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.9955796003341675, | |
| "learning_rate": 0.00022421786235386112, | |
| "loss": 0.533, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.7445811364967779, | |
| "eval_f1": 0.5514403292181069, | |
| "eval_loss": 0.5220016837120056, | |
| "eval_mcc": 0.39114524585919325, | |
| "eval_precision": 0.6650124069478908, | |
| "eval_recall": 0.4710017574692443, | |
| "eval_runtime": 3.1849, | |
| "eval_samples_per_second": 535.964, | |
| "eval_steps_per_second": 16.955, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 2.413587808609009, | |
| "learning_rate": 0.00019930476653676543, | |
| "loss": 0.5076, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.7504393673110721, | |
| "eval_f1": 0.4766584766584766, | |
| "eval_loss": 0.5150462985038757, | |
| "eval_mcc": 0.3981605013144147, | |
| "eval_precision": 0.7918367346938775, | |
| "eval_recall": 0.3409490333919156, | |
| "eval_runtime": 3.1543, | |
| "eval_samples_per_second": 541.159, | |
| "eval_steps_per_second": 17.119, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 4.792643070220947, | |
| "learning_rate": 0.00017439167071966975, | |
| "loss": 0.4936, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.7568834212067955, | |
| "eval_f1": 0.5077105575326216, | |
| "eval_loss": 0.5015135407447815, | |
| "eval_mcc": 0.4152737161319117, | |
| "eval_precision": 0.781021897810219, | |
| "eval_recall": 0.37609841827768015, | |
| "eval_runtime": 3.1947, | |
| "eval_samples_per_second": 534.327, | |
| "eval_steps_per_second": 16.903, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 7.65413236618042, | |
| "learning_rate": 0.00014947857490257407, | |
| "loss": 0.4718, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.7662565905096661, | |
| "eval_f1": 0.5741728922091782, | |
| "eval_loss": 0.4888531267642975, | |
| "eval_mcc": 0.4422168885701626, | |
| "eval_precision": 0.7309782608695652, | |
| "eval_recall": 0.4727592267135325, | |
| "eval_runtime": 3.1595, | |
| "eval_samples_per_second": 540.283, | |
| "eval_steps_per_second": 17.092, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 7.2619218826293945, | |
| "learning_rate": 0.00012456547908547839, | |
| "loss": 0.4651, | |
| "step": 1284 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.7627416520210897, | |
| "eval_f1": 0.5837615621788285, | |
| "eval_loss": 0.48531395196914673, | |
| "eval_mcc": 0.43661648172458356, | |
| "eval_precision": 0.7029702970297029, | |
| "eval_recall": 0.4991212653778559, | |
| "eval_runtime": 3.3544, | |
| "eval_samples_per_second": 508.885, | |
| "eval_steps_per_second": 16.098, | |
| "step": 1284 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 9.458930969238281, | |
| "learning_rate": 9.965238326838272e-05, | |
| "loss": 0.4553, | |
| "step": 1498 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.7633274751025191, | |
| "eval_f1": 0.536697247706422, | |
| "eval_loss": 0.48649853467941284, | |
| "eval_mcc": 0.4325666465390762, | |
| "eval_precision": 0.7722772277227723, | |
| "eval_recall": 0.4112478031634446, | |
| "eval_runtime": 3.1637, | |
| "eval_samples_per_second": 539.56, | |
| "eval_steps_per_second": 17.069, | |
| "step": 1498 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 5.576714038848877, | |
| "learning_rate": 7.473928745128703e-05, | |
| "loss": 0.4423, | |
| "step": 1712 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.7668424135910955, | |
| "eval_f1": 0.5738758029978586, | |
| "eval_loss": 0.48117396235466003, | |
| "eval_mcc": 0.4435339132892942, | |
| "eval_precision": 0.7342465753424657, | |
| "eval_recall": 0.4710017574692443, | |
| "eval_runtime": 3.2168, | |
| "eval_samples_per_second": 530.659, | |
| "eval_steps_per_second": 16.787, | |
| "step": 1712 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 5.577918529510498, | |
| "learning_rate": 4.982619163419136e-05, | |
| "loss": 0.4406, | |
| "step": 1926 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.7721148213239601, | |
| "eval_f1": 0.6026557711950972, | |
| "eval_loss": 0.46937549114227295, | |
| "eval_mcc": 0.4605923234016756, | |
| "eval_precision": 0.7195121951219512, | |
| "eval_recall": 0.5184534270650264, | |
| "eval_runtime": 3.1822, | |
| "eval_samples_per_second": 536.419, | |
| "eval_steps_per_second": 16.969, | |
| "step": 1926 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 4.4660234451293945, | |
| "learning_rate": 2.491309581709568e-05, | |
| "loss": 0.4322, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.773286467486819, | |
| "eval_f1": 0.6063072227873855, | |
| "eval_loss": 0.47590258717536926, | |
| "eval_mcc": 0.4639025999570181, | |
| "eval_precision": 0.7198067632850241, | |
| "eval_recall": 0.523725834797891, | |
| "eval_runtime": 3.3263, | |
| "eval_samples_per_second": 513.185, | |
| "eval_steps_per_second": 16.234, | |
| "step": 2140 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 2354, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 11, | |
| "save_steps": 500, | |
| "total_flos": 5247756644400.0, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.995655146834002, | |
| "learning_rate": 0.00027404405398805245, | |
| "num_train_epochs": 11, | |
| "temperature": 14 | |
| } | |
| } | |