| { | |
| "best_metric": 0.5678136944770813, | |
| "best_model_checkpoint": "bert_tiny_lda_20_v1_book_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": 1.0802937746047974, | |
| "learning_rate": 4.9e-05, | |
| "loss": 0.6236, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.6936274509803921, | |
| "eval_combined_score": 0.7496131034062148, | |
| "eval_f1": 0.8055987558320373, | |
| "eval_loss": 0.5941258072853088, | |
| "eval_runtime": 0.1505, | |
| "eval_samples_per_second": 2710.301, | |
| "eval_steps_per_second": 13.286, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.9952943325042725, | |
| "learning_rate": 4.8e-05, | |
| "loss": 0.5876, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.7083333333333334, | |
| "eval_combined_score": 0.7638783510369247, | |
| "eval_f1": 0.819423368740516, | |
| "eval_loss": 0.5766649842262268, | |
| "eval_runtime": 0.135, | |
| "eval_samples_per_second": 3021.892, | |
| "eval_steps_per_second": 14.813, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 2.0742907524108887, | |
| "learning_rate": 4.7e-05, | |
| "loss": 0.5497, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.7132352941176471, | |
| "eval_combined_score": 0.7659199726402189, | |
| "eval_f1": 0.8186046511627908, | |
| "eval_loss": 0.5678136944770813, | |
| "eval_runtime": 0.137, | |
| "eval_samples_per_second": 2977.789, | |
| "eval_steps_per_second": 14.597, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 2.5984385013580322, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 0.503, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.7156862745098039, | |
| "eval_combined_score": 0.766934046345811, | |
| "eval_f1": 0.8181818181818181, | |
| "eval_loss": 0.5999492406845093, | |
| "eval_runtime": 0.137, | |
| "eval_samples_per_second": 2979.106, | |
| "eval_steps_per_second": 14.603, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 2.8568592071533203, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.4554, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.7156862745098039, | |
| "eval_combined_score": 0.767500458127176, | |
| "eval_f1": 0.8193146417445483, | |
| "eval_loss": 0.5997003316879272, | |
| "eval_runtime": 0.1442, | |
| "eval_samples_per_second": 2830.333, | |
| "eval_steps_per_second": 13.874, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 9.565511703491211, | |
| "learning_rate": 4.4000000000000006e-05, | |
| "loss": 0.3873, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.7034313725490197, | |
| "eval_combined_score": 0.7398857417273749, | |
| "eval_f1": 0.7763401109057301, | |
| "eval_loss": 0.6210173964500427, | |
| "eval_runtime": 0.1348, | |
| "eval_samples_per_second": 3025.776, | |
| "eval_steps_per_second": 14.832, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 9.428056716918945, | |
| "learning_rate": 4.3e-05, | |
| "loss": 0.3092, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.7303921568627451, | |
| "eval_combined_score": 0.7805806938159879, | |
| "eval_f1": 0.8307692307692308, | |
| "eval_loss": 0.7192116379737854, | |
| "eval_runtime": 0.1548, | |
| "eval_samples_per_second": 2635.25, | |
| "eval_steps_per_second": 12.918, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 6.240992546081543, | |
| "learning_rate": 4.2e-05, | |
| "loss": 0.2575, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.6593137254901961, | |
| "eval_combined_score": 0.6898178285398666, | |
| "eval_f1": 0.7203219315895372, | |
| "eval_loss": 0.7417659163475037, | |
| "eval_runtime": 0.1525, | |
| "eval_samples_per_second": 2675.349, | |
| "eval_steps_per_second": 13.114, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "step": 120, | |
| "total_flos": 769504676118528.0, | |
| "train_loss": 0.4591776688893636, | |
| "train_runtime": 25.4279, | |
| "train_samples_per_second": 7212.562, | |
| "train_steps_per_second": 29.495 | |
| } | |
| ], | |
| "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 | |
| } | |