| { | |
| "best_metric": 0.7798434442270059, | |
| "best_model_checkpoint": "tiny-bert-sst2-distilled/run-3/checkpoint-1152", | |
| "epoch": 6.0, | |
| "eval_steps": 500, | |
| "global_step": 1152, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.7568773031234741, | |
| "learning_rate": 0.00039540069434609415, | |
| "loss": 0.3274, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.5430528375733855, | |
| "eval_f1": 0.20442930153321975, | |
| "eval_loss": 0.30545827746391296, | |
| "eval_mcc": 0.1640968755879385, | |
| "eval_precision": 0.7894736842105263, | |
| "eval_recall": 0.11741682974559686, | |
| "eval_runtime": 66.7048, | |
| "eval_samples_per_second": 15.321, | |
| "eval_steps_per_second": 0.48, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.822717547416687, | |
| "learning_rate": 0.0003295005786217451, | |
| "loss": 0.3076, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.5459882583170255, | |
| "eval_f1": 0.21355932203389827, | |
| "eval_loss": 0.2995118200778961, | |
| "eval_mcc": 0.17219814389174506, | |
| "eval_precision": 0.7974683544303798, | |
| "eval_recall": 0.1232876712328767, | |
| "eval_runtime": 66.8, | |
| "eval_samples_per_second": 15.299, | |
| "eval_steps_per_second": 0.479, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 1.0308061838150024, | |
| "learning_rate": 0.0002636004628973961, | |
| "loss": 0.2967, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.5694716242661448, | |
| "eval_f1": 0.29936305732484075, | |
| "eval_loss": 0.2949042022228241, | |
| "eval_mcc": 0.21819315809183495, | |
| "eval_precision": 0.8034188034188035, | |
| "eval_recall": 0.18395303326810175, | |
| "eval_runtime": 66.6296, | |
| "eval_samples_per_second": 15.339, | |
| "eval_steps_per_second": 0.48, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 1.6645914316177368, | |
| "learning_rate": 0.00019770034717304707, | |
| "loss": 0.296, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.6183953033268101, | |
| "eval_f1": 0.4298245614035087, | |
| "eval_loss": 0.29466578364372253, | |
| "eval_mcc": 0.3157246377506289, | |
| "eval_precision": 0.8497109826589595, | |
| "eval_recall": 0.2876712328767123, | |
| "eval_runtime": 67.3924, | |
| "eval_samples_per_second": 15.165, | |
| "eval_steps_per_second": 0.475, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 1.67695951461792, | |
| "learning_rate": 0.00013180023144869805, | |
| "loss": 0.2906, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.7240704500978473, | |
| "eval_f1": 0.6658767772511849, | |
| "eval_loss": 0.29085618257522583, | |
| "eval_mcc": 0.478083499336065, | |
| "eval_precision": 0.8438438438438438, | |
| "eval_recall": 0.5499021526418787, | |
| "eval_runtime": 66.3348, | |
| "eval_samples_per_second": 15.407, | |
| "eval_steps_per_second": 0.482, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 0.9722081422805786, | |
| "learning_rate": 6.590011572434902e-05, | |
| "loss": 0.2869, | |
| "step": 1152 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.7798434442270059, | |
| "eval_f1": 0.7583243823845327, | |
| "eval_loss": 0.29121583700180054, | |
| "eval_mcc": 0.5687784616694839, | |
| "eval_precision": 0.8404761904761905, | |
| "eval_recall": 0.6908023483365949, | |
| "eval_runtime": 66.588, | |
| "eval_samples_per_second": 15.348, | |
| "eval_steps_per_second": 0.481, | |
| "step": 1152 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 1344, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 7, | |
| "save_steps": 500, | |
| "total_flos": 1414171183680.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.4369516344275761, | |
| "learning_rate": 0.00046130081007044317, | |
| "num_train_epochs": 7, | |
| "per_device_train_batch_size": 16, | |
| "temperature": 15 | |
| } | |
| } | |