| { | |
| "best_metric": 0.28948119282722473, | |
| "best_model_checkpoint": "bert_tiny_lda_5_v1_book_qqp/checkpoint-4266", | |
| "epoch": 8.0, | |
| "eval_steps": 500, | |
| "global_step": 11376, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 2.789701223373413, | |
| "learning_rate": 4.9e-05, | |
| "loss": 0.3935, | |
| "step": 1422 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8488003957457334, | |
| "eval_combined_score": 0.8166441605028685, | |
| "eval_f1": 0.7844879252600035, | |
| "eval_loss": 0.33461156487464905, | |
| "eval_runtime": 12.6619, | |
| "eval_samples_per_second": 3193.05, | |
| "eval_steps_per_second": 12.478, | |
| "step": 1422 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 3.348468542098999, | |
| "learning_rate": 4.8e-05, | |
| "loss": 0.294, | |
| "step": 2844 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8698243878308187, | |
| "eval_combined_score": 0.8491089398151779, | |
| "eval_f1": 0.828393491799537, | |
| "eval_loss": 0.2950727939605713, | |
| "eval_runtime": 12.5285, | |
| "eval_samples_per_second": 3227.053, | |
| "eval_steps_per_second": 12.611, | |
| "step": 2844 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 3.606611967086792, | |
| "learning_rate": 4.7e-05, | |
| "loss": 0.2397, | |
| "step": 4266 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8787286668315607, | |
| "eval_combined_score": 0.8569969645349551, | |
| "eval_f1": 0.8352652622383496, | |
| "eval_loss": 0.28948119282722473, | |
| "eval_runtime": 12.5737, | |
| "eval_samples_per_second": 3215.449, | |
| "eval_steps_per_second": 12.566, | |
| "step": 4266 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 4.67673921585083, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 0.1952, | |
| "step": 5688 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.881127875340094, | |
| "eval_combined_score": 0.8579184568611324, | |
| "eval_f1": 0.8347090383821709, | |
| "eval_loss": 0.30098268389701843, | |
| "eval_runtime": 12.9847, | |
| "eval_samples_per_second": 3113.663, | |
| "eval_steps_per_second": 12.168, | |
| "step": 5688 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 4.516249179840088, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.1584, | |
| "step": 7110 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.8838733613653228, | |
| "eval_combined_score": 0.8638766268137424, | |
| "eval_f1": 0.8438798922621621, | |
| "eval_loss": 0.313992977142334, | |
| "eval_runtime": 12.8907, | |
| "eval_samples_per_second": 3136.38, | |
| "eval_steps_per_second": 12.257, | |
| "step": 7110 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 4.016932964324951, | |
| "learning_rate": 4.4000000000000006e-05, | |
| "loss": 0.128, | |
| "step": 8532 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.8851842691070987, | |
| "eval_combined_score": 0.8668266452279634, | |
| "eval_f1": 0.8484690213488281, | |
| "eval_loss": 0.3402121961116791, | |
| "eval_runtime": 13.1699, | |
| "eval_samples_per_second": 3069.869, | |
| "eval_steps_per_second": 11.997, | |
| "step": 8532 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 2.571812868118286, | |
| "learning_rate": 4.3e-05, | |
| "loss": 0.1045, | |
| "step": 9954 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.8844917140737076, | |
| "eval_combined_score": 0.8643606001989619, | |
| "eval_f1": 0.8442294863242161, | |
| "eval_loss": 0.35473570227622986, | |
| "eval_runtime": 12.69, | |
| "eval_samples_per_second": 3185.972, | |
| "eval_steps_per_second": 12.451, | |
| "step": 9954 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 2.9188523292541504, | |
| "learning_rate": 4.2e-05, | |
| "loss": 0.0881, | |
| "step": 11376 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.8849121939154093, | |
| "eval_combined_score": 0.8663993445303615, | |
| "eval_f1": 0.8478864951453137, | |
| "eval_loss": 0.3847922384738922, | |
| "eval_runtime": 12.9115, | |
| "eval_samples_per_second": 3131.313, | |
| "eval_steps_per_second": 12.237, | |
| "step": 11376 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "step": 11376, | |
| "total_flos": 7.633075201391002e+16, | |
| "train_loss": 0.20017086369746512, | |
| "train_runtime": 1701.3381, | |
| "train_samples_per_second": 10692.936, | |
| "train_steps_per_second": 41.791 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 71100, | |
| "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": 7.633075201391002e+16, | |
| "train_batch_size": 256, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |