| { | |
| "best_metric": 0.5935130715370178, | |
| "best_model_checkpoint": "bert_tiny_lda_20_v1_mrpc/checkpoint-45", | |
| "epoch": 8.0, | |
| "eval_steps": 500, | |
| "global_step": 120, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.7700666189193726, | |
| "learning_rate": 4.9e-05, | |
| "loss": 0.6315, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.6862745098039216, | |
| "eval_combined_score": 0.7492956126732218, | |
| "eval_f1": 0.812316715542522, | |
| "eval_loss": 0.6003513932228088, | |
| "eval_runtime": 0.1367, | |
| "eval_samples_per_second": 2984.452, | |
| "eval_steps_per_second": 14.63, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.322261929512024, | |
| "learning_rate": 4.8e-05, | |
| "loss": 0.6013, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.6887254901960784, | |
| "eval_combined_score": 0.7462174591629542, | |
| "eval_f1": 0.8037094281298299, | |
| "eval_loss": 0.5958060622215271, | |
| "eval_runtime": 0.1396, | |
| "eval_samples_per_second": 2921.977, | |
| "eval_steps_per_second": 14.323, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 1.471887469291687, | |
| "learning_rate": 4.7e-05, | |
| "loss": 0.5707, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.6985294117647058, | |
| "eval_combined_score": 0.7542106100554595, | |
| "eval_f1": 0.8098918083462132, | |
| "eval_loss": 0.5935130715370178, | |
| "eval_runtime": 0.1333, | |
| "eval_samples_per_second": 3060.988, | |
| "eval_steps_per_second": 15.005, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 1.2280375957489014, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 0.5415, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.6985294117647058, | |
| "eval_combined_score": 0.750864705882353, | |
| "eval_f1": 0.8032000000000001, | |
| "eval_loss": 0.6068510413169861, | |
| "eval_runtime": 0.1378, | |
| "eval_samples_per_second": 2961.536, | |
| "eval_steps_per_second": 14.517, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 2.841067314147949, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.4866, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.678921568627451, | |
| "eval_combined_score": 0.7263347048664026, | |
| "eval_f1": 0.7737478411053541, | |
| "eval_loss": 0.6273573040962219, | |
| "eval_runtime": 0.1355, | |
| "eval_samples_per_second": 3011.787, | |
| "eval_steps_per_second": 14.764, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 7.474719524383545, | |
| "learning_rate": 4.4000000000000006e-05, | |
| "loss": 0.397, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.6985294117647058, | |
| "eval_combined_score": 0.7495888549909429, | |
| "eval_f1": 0.8006482982171799, | |
| "eval_loss": 0.7452647089958191, | |
| "eval_runtime": 0.137, | |
| "eval_samples_per_second": 2978.759, | |
| "eval_steps_per_second": 14.602, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 6.474092483520508, | |
| "learning_rate": 4.3e-05, | |
| "loss": 0.3039, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.6519607843137255, | |
| "eval_combined_score": 0.6968894830659537, | |
| "eval_f1": 0.7418181818181818, | |
| "eval_loss": 0.8151072859764099, | |
| "eval_runtime": 0.1328, | |
| "eval_samples_per_second": 3071.421, | |
| "eval_steps_per_second": 15.056, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 3.6470448970794678, | |
| "learning_rate": 4.2e-05, | |
| "loss": 0.2217, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.6225490196078431, | |
| "eval_combined_score": 0.6614690623331044, | |
| "eval_f1": 0.7003891050583657, | |
| "eval_loss": 0.986514687538147, | |
| "eval_runtime": 0.1369, | |
| "eval_samples_per_second": 2979.22, | |
| "eval_steps_per_second": 14.604, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "step": 120, | |
| "total_flos": 769504676118528.0, | |
| "train_loss": 0.4692632933457693, | |
| "train_runtime": 24.0661, | |
| "train_samples_per_second": 7620.676, | |
| "train_steps_per_second": 31.164 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 750, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 50, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 5 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 769504676118528.0, | |
| "train_batch_size": 256, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |