| { | |
| "best_metric": 0.7006444053895724, | |
| "best_model_checkpoint": "tiny-bert-sst2-distilled/run-7/checkpoint-642", | |
| "epoch": 7.0, | |
| "eval_steps": 500, | |
| "global_step": 1498, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 5.4419636726379395, | |
| "learning_rate": 4.778212463646413e-05, | |
| "loss": 0.4867, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.6649091974223784, | |
| "eval_f1": 0.01718213058419244, | |
| "eval_loss": 0.4545574486255646, | |
| "eval_mcc": 0.009529862152017439, | |
| "eval_precision": 0.38461538461538464, | |
| "eval_recall": 0.008787346221441126, | |
| "eval_runtime": 3.147, | |
| "eval_samples_per_second": 542.424, | |
| "eval_steps_per_second": 17.159, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 2.8326942920684814, | |
| "learning_rate": 3.9818437197053446e-05, | |
| "loss": 0.4528, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.6959578207381371, | |
| "eval_f1": 0.40955631399317405, | |
| "eval_loss": 0.44587329030036926, | |
| "eval_mcc": 0.24713473389349372, | |
| "eval_precision": 0.5806451612903226, | |
| "eval_recall": 0.3163444639718805, | |
| "eval_runtime": 3.2126, | |
| "eval_samples_per_second": 531.351, | |
| "eval_steps_per_second": 16.809, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 2.2975902557373047, | |
| "learning_rate": 3.1854749757642754e-05, | |
| "loss": 0.4424, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.7006444053895724, | |
| "eval_f1": 0.396694214876033, | |
| "eval_loss": 0.4441916346549988, | |
| "eval_mcc": 0.25354492903122977, | |
| "eval_precision": 0.60431654676259, | |
| "eval_recall": 0.29525483304042177, | |
| "eval_runtime": 3.1517, | |
| "eval_samples_per_second": 541.618, | |
| "eval_steps_per_second": 17.134, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 3.1896305084228516, | |
| "learning_rate": 2.3891062318232066e-05, | |
| "loss": 0.4399, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.700058582308143, | |
| "eval_f1": 0.42471910112359557, | |
| "eval_loss": 0.4420276880264282, | |
| "eval_mcc": 0.2607870237342907, | |
| "eval_precision": 0.5887850467289719, | |
| "eval_recall": 0.3321616871704745, | |
| "eval_runtime": 3.2393, | |
| "eval_samples_per_second": 526.972, | |
| "eval_steps_per_second": 16.67, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 3.5341804027557373, | |
| "learning_rate": 1.5927374878821377e-05, | |
| "loss": 0.437, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.6988869361452841, | |
| "eval_f1": 0.38663484486873506, | |
| "eval_loss": 0.4399532675743103, | |
| "eval_mcc": 0.24671136320593817, | |
| "eval_precision": 0.6022304832713755, | |
| "eval_recall": 0.28471001757469244, | |
| "eval_runtime": 3.152, | |
| "eval_samples_per_second": 541.569, | |
| "eval_steps_per_second": 17.132, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 2.5217621326446533, | |
| "learning_rate": 7.963687439410689e-06, | |
| "loss": 0.4353, | |
| "step": 1284 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.700058582308143, | |
| "eval_f1": 0.3875598086124402, | |
| "eval_loss": 0.4395112693309784, | |
| "eval_mcc": 0.24974237662368784, | |
| "eval_precision": 0.6067415730337079, | |
| "eval_recall": 0.28471001757469244, | |
| "eval_runtime": 3.3343, | |
| "eval_samples_per_second": 511.957, | |
| "eval_steps_per_second": 16.195, | |
| "step": 1284 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 2.783250570297241, | |
| "learning_rate": 0.0, | |
| "loss": 0.4339, | |
| "step": 1498 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.6988869361452841, | |
| "eval_f1": 0.3938679245283018, | |
| "eval_loss": 0.43912431597709656, | |
| "eval_mcc": 0.2486977013771559, | |
| "eval_precision": 0.5985663082437276, | |
| "eval_recall": 0.29349736379613356, | |
| "eval_runtime": 3.1437, | |
| "eval_samples_per_second": 542.982, | |
| "eval_steps_per_second": 17.177, | |
| "step": 1498 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 1498, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 7, | |
| "save_steps": 500, | |
| "total_flos": 3673429651080.0, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.7570540466770024, | |
| "learning_rate": 5.574581207587482e-05, | |
| "num_train_epochs": 7, | |
| "temperature": 25 | |
| } | |
| } | |