| { | |
| "best_metric": 0.576320939334638, | |
| "best_model_checkpoint": "tiny-bert-sst2-distilled/run-3/checkpoint-192", | |
| "epoch": 6.0, | |
| "eval_steps": 500, | |
| "global_step": 576, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.41195374727249146, | |
| "learning_rate": 7.986864922441338e-05, | |
| "loss": 0.2944, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.5547945205479452, | |
| "eval_f1": 0.2528735632183908, | |
| "eval_loss": 0.2588364779949188, | |
| "eval_precision": 0.7857142857142857, | |
| "eval_recall": 0.1506849315068493, | |
| "eval_runtime": 25.9352, | |
| "eval_samples_per_second": 39.406, | |
| "eval_steps_per_second": 1.234, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.7536277174949646, | |
| "learning_rate": 6.38949193795307e-05, | |
| "loss": 0.261, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.576320939334638, | |
| "eval_f1": 0.3244929797191888, | |
| "eval_loss": 0.25162196159362793, | |
| "eval_precision": 0.8, | |
| "eval_recall": 0.2035225048923679, | |
| "eval_runtime": 25.5955, | |
| "eval_samples_per_second": 39.929, | |
| "eval_steps_per_second": 1.25, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.5996448993682861, | |
| "learning_rate": 4.7921189534648024e-05, | |
| "loss": 0.2559, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.5684931506849316, | |
| "eval_f1": 0.29888712241653415, | |
| "eval_loss": 0.24895831942558289, | |
| "eval_precision": 0.7966101694915254, | |
| "eval_recall": 0.18395303326810175, | |
| "eval_runtime": 26.2664, | |
| "eval_samples_per_second": 38.909, | |
| "eval_steps_per_second": 1.218, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.5228096842765808, | |
| "learning_rate": 3.194745968976535e-05, | |
| "loss": 0.2542, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.550880626223092, | |
| "eval_f1": 0.24382207578253706, | |
| "eval_loss": 0.24970793724060059, | |
| "eval_precision": 0.7708333333333334, | |
| "eval_recall": 0.14481409001956946, | |
| "eval_runtime": 25.3347, | |
| "eval_samples_per_second": 40.34, | |
| "eval_steps_per_second": 1.263, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.459949791431427, | |
| "learning_rate": 1.5973729844882675e-05, | |
| "loss": 0.2522, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.5626223091976517, | |
| "eval_f1": 0.2848, | |
| "eval_loss": 0.24784404039382935, | |
| "eval_precision": 0.7807017543859649, | |
| "eval_recall": 0.17416829745596868, | |
| "eval_runtime": 25.7014, | |
| "eval_samples_per_second": 39.764, | |
| "eval_steps_per_second": 1.245, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 1.6193181276321411, | |
| "learning_rate": 0.0, | |
| "loss": 0.2516, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.5714285714285714, | |
| "eval_f1": 0.3134796238244514, | |
| "eval_loss": 0.24758201837539673, | |
| "eval_precision": 0.7874015748031497, | |
| "eval_recall": 0.19569471624266144, | |
| "eval_runtime": 26.0053, | |
| "eval_samples_per_second": 39.3, | |
| "eval_steps_per_second": 1.231, | |
| "step": 576 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 576, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 6, | |
| "save_steps": 500, | |
| "total_flos": 1414171183680.0, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.3513021160288001, | |
| "learning_rate": 9.584237906929605e-05, | |
| "num_train_epochs": 6, | |
| "temperature": 30 | |
| } | |
| } | |