| { | |
| "best_metric": 0.5128712871287129, | |
| "best_model_checkpoint": "tiny-bert-sst2-distilled/run-2/checkpoint-594", | |
| "epoch": 9.0, | |
| "eval_steps": 500, | |
| "global_step": 2673, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.6258772015571594, | |
| "learning_rate": 1.171336966673164e-05, | |
| "loss": 0.303, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.4910891089108911, | |
| "eval_f1": 0.0, | |
| "eval_loss": 0.28316354751586914, | |
| "eval_mcc": -0.09980217586956908, | |
| "eval_precision": 0.0, | |
| "eval_recall": 0.0, | |
| "eval_runtime": 0.9281, | |
| "eval_samples_per_second": 544.113, | |
| "eval_steps_per_second": 17.239, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.44756296277046204, | |
| "learning_rate": 1.0249198458390185e-05, | |
| "loss": 0.2807, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.5128712871287129, | |
| "eval_f1": 0.13986013986013984, | |
| "eval_loss": 0.2766987085342407, | |
| "eval_mcc": 0.047945551909639166, | |
| "eval_precision": 0.5882352941176471, | |
| "eval_recall": 0.07936507936507936, | |
| "eval_runtime": 0.9353, | |
| "eval_samples_per_second": 539.951, | |
| "eval_steps_per_second": 17.107, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.431320458650589, | |
| "learning_rate": 8.785027250048729e-06, | |
| "loss": 0.2776, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.5128712871287129, | |
| "eval_f1": 0.14583333333333331, | |
| "eval_loss": 0.27507784962654114, | |
| "eval_mcc": 0.046724423554504804, | |
| "eval_precision": 0.5833333333333334, | |
| "eval_recall": 0.08333333333333333, | |
| "eval_runtime": 0.9247, | |
| "eval_samples_per_second": 546.126, | |
| "eval_steps_per_second": 17.303, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.3838063180446625, | |
| "learning_rate": 7.320856041707276e-06, | |
| "loss": 0.2774, | |
| "step": 1188 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.5128712871287129, | |
| "eval_f1": 0.14583333333333331, | |
| "eval_loss": 0.273956835269928, | |
| "eval_mcc": 0.046724423554504804, | |
| "eval_precision": 0.5833333333333334, | |
| "eval_recall": 0.08333333333333333, | |
| "eval_runtime": 0.9281, | |
| "eval_samples_per_second": 544.106, | |
| "eval_steps_per_second": 17.239, | |
| "step": 1188 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.5975275039672852, | |
| "learning_rate": 5.85668483336582e-06, | |
| "loss": 0.2766, | |
| "step": 1485 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.5108910891089109, | |
| "eval_f1": 0.13937282229965156, | |
| "eval_loss": 0.27307695150375366, | |
| "eval_mcc": 0.039524509362779174, | |
| "eval_precision": 0.5714285714285714, | |
| "eval_recall": 0.07936507936507936, | |
| "eval_runtime": 0.9415, | |
| "eval_samples_per_second": 536.374, | |
| "eval_steps_per_second": 16.994, | |
| "step": 1485 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 0.4973074495792389, | |
| "learning_rate": 4.3925136250243645e-06, | |
| "loss": 0.2759, | |
| "step": 1782 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.5069306930693069, | |
| "eval_f1": 0.1384083044982699, | |
| "eval_loss": 0.27315467596054077, | |
| "eval_mcc": 0.023354867176075732, | |
| "eval_precision": 0.5405405405405406, | |
| "eval_recall": 0.07936507936507936, | |
| "eval_runtime": 0.9268, | |
| "eval_samples_per_second": 544.884, | |
| "eval_steps_per_second": 17.264, | |
| "step": 1782 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 0.440789133310318, | |
| "learning_rate": 2.92834241668291e-06, | |
| "loss": 0.2755, | |
| "step": 2079 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.5108910891089109, | |
| "eval_f1": 0.13937282229965156, | |
| "eval_loss": 0.27257436513900757, | |
| "eval_mcc": 0.039524509362779174, | |
| "eval_precision": 0.5714285714285714, | |
| "eval_recall": 0.07936507936507936, | |
| "eval_runtime": 0.9289, | |
| "eval_samples_per_second": 543.634, | |
| "eval_steps_per_second": 17.224, | |
| "step": 2079 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.6993103623390198, | |
| "learning_rate": 1.464171208341455e-06, | |
| "loss": 0.2751, | |
| "step": 2376 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.5089108910891089, | |
| "eval_f1": 0.1267605633802817, | |
| "eval_loss": 0.27230069041252136, | |
| "eval_mcc": 0.03302792352019969, | |
| "eval_precision": 0.5625, | |
| "eval_recall": 0.07142857142857142, | |
| "eval_runtime": 0.9319, | |
| "eval_samples_per_second": 541.932, | |
| "eval_steps_per_second": 17.17, | |
| "step": 2376 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 0.5155681371688843, | |
| "learning_rate": 0.0, | |
| "loss": 0.2752, | |
| "step": 2673 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.5069306930693069, | |
| "eval_f1": 0.1263157894736842, | |
| "eval_loss": 0.2723012864589691, | |
| "eval_mcc": 0.024561362814430283, | |
| "eval_precision": 0.5454545454545454, | |
| "eval_recall": 0.07142857142857142, | |
| "eval_runtime": 0.9301, | |
| "eval_samples_per_second": 542.98, | |
| "eval_steps_per_second": 17.203, | |
| "step": 2673 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 2673, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 9, | |
| "save_steps": 500, | |
| "total_flos": 6576311393820.0, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.33782488262757904, | |
| "learning_rate": 1.3177540875073095e-05, | |
| "num_train_epochs": 9, | |
| "temperature": 26 | |
| } | |
| } | |