| { | |
| "best_metric": 0.7346221441124781, | |
| "best_model_checkpoint": "tiny-bert-sst2-distilled/run-20/checkpoint-428", | |
| "epoch": 4.0, | |
| "eval_steps": 500, | |
| "global_step": 856, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 3.5011236667633057, | |
| "learning_rate": 0.00029874954647137283, | |
| "loss": 0.4553, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.6848271821909784, | |
| "eval_f1": 0.1567398119122257, | |
| "eval_loss": 0.4299508333206177, | |
| "eval_mcc": 0.17036819686661997, | |
| "eval_precision": 0.7246376811594203, | |
| "eval_recall": 0.08787346221441125, | |
| "eval_runtime": 3.1578, | |
| "eval_samples_per_second": 540.559, | |
| "eval_steps_per_second": 17.1, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.671453595161438, | |
| "learning_rate": 0.0002489579553928107, | |
| "loss": 0.4196, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.7346221441124781, | |
| "eval_f1": 0.4372670807453416, | |
| "eval_loss": 0.4135269522666931, | |
| "eval_mcc": 0.3504334189243189, | |
| "eval_precision": 0.7457627118644068, | |
| "eval_recall": 0.3093145869947276, | |
| "eval_runtime": 3.2195, | |
| "eval_samples_per_second": 530.204, | |
| "eval_steps_per_second": 16.773, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 1.788920521736145, | |
| "learning_rate": 0.00019916636431424855, | |
| "loss": 0.411, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.7106033977738723, | |
| "eval_f1": 0.27988338192419826, | |
| "eval_loss": 0.415696382522583, | |
| "eval_mcc": 0.2803429374077876, | |
| "eval_precision": 0.8205128205128205, | |
| "eval_recall": 0.1687170474516696, | |
| "eval_runtime": 3.1669, | |
| "eval_samples_per_second": 539.016, | |
| "eval_steps_per_second": 17.051, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 1.344914436340332, | |
| "learning_rate": 0.00014937477323568641, | |
| "loss": 0.4054, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.7193907439953134, | |
| "eval_f1": 0.37711313394018203, | |
| "eval_loss": 0.4086741507053375, | |
| "eval_mcc": 0.302678268200975, | |
| "eval_precision": 0.725, | |
| "eval_recall": 0.2548330404217926, | |
| "eval_runtime": 3.1334, | |
| "eval_samples_per_second": 544.772, | |
| "eval_steps_per_second": 17.234, | |
| "step": 856 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 1498, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 7, | |
| "save_steps": 500, | |
| "total_flos": 2099102657760.0, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.7240883100721838, | |
| "learning_rate": 0.000348541137549935, | |
| "num_train_epochs": 7, | |
| "temperature": 12 | |
| } | |
| } | |