{ "best_global_step": 3136, "best_metric": 0.8876387938129007, "best_model_checkpoint": "/app/checkpoints/checkpoint-3136", "epoch": 2.0, "eval_steps": 500, "global_step": 3136, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0, "eval_accuracy": 0.43732057416267944, "eval_class_metrics": { "0": { "accuracy": 0.9419457735247209, "confusion_matrix": { "fn": 182, "fp": 0, "tn": 2953, "tp": 0 }, "f_beta_score": 0.0, "precision": 0, "recall": 0.0 }, "1": { "accuracy": 0.7435406698564593, "confusion_matrix": { "fn": 803, "fp": 1, "tn": 2331, "tp": 0 }, "f_beta_score": 0.0, "precision": 0.0, "recall": 0.0 }, "2": { "accuracy": 0.9151515151515152, "confusion_matrix": { "fn": 266, "fp": 0, "tn": 2869, "tp": 0 }, "f_beta_score": 0.0, "precision": 0, "recall": 0.0 }, "3": { "accuracy": 0.4376395534290271, "confusion_matrix": { "fn": 1, "fp": 1762, "tn": 1, "tp": 1371 }, "f_beta_score": 0.6086570477247503, "precision": 0.43759974465368656, "recall": 0.999271137026239 }, "4": { "accuracy": 0.8363636363636363, "confusion_matrix": { "fn": 512, "fp": 1, "tn": 2622, "tp": 0 }, "f_beta_score": 0.0, "precision": 0.0, "recall": 0.0 } }, "eval_confusion_matrix": { "fn": 1764, "fp": 1764, "tn": 10776, "tp": 1371 }, "eval_f_beta_score": 0.26637239855768974, "eval_label_confusion_matrix": { "labels": [ "FYI", "NEEDS_RESPONSE", "PERSONAL", "PROMOTIONAL", "URGENT" ], "matrix": [ [ 0, 1, 0, 181, 0 ], [ 0, 0, 0, 803, 0 ], [ 0, 0, 0, 266, 0 ], [ 0, 0, 0, 1371, 1 ], [ 0, 0, 0, 512, 0 ] ] }, "eval_loss": 1.6771581172943115, "eval_precision": 0.1915109568308957, "eval_recall": 0.43732057416267944, "eval_runtime": 18.7918, "eval_sample_size": 3135, "eval_samples_per_second": 166.828, "eval_steps_per_second": 20.86, "step": 0 }, { "epoch": 1.0, "grad_norm": 100.88322448730469, "hard_loss": 0.6793025164542439, "learning_rate": 4.3964845118848066e-05, "loss": 0.6793, "step": 1568 }, { "epoch": 1.0, "eval_accuracy": 0.8650717703349282, "eval_class_metrics": { "0": { "accuracy": 0.9639553429027113, "confusion_matrix": { "fn": 65, "fp": 48, "tn": 2905, "tp": 117 }, "f_beta_score": 0.6743515850144092, "precision": 0.7090909090909091, "recall": 0.6428571428571429 }, "1": { "accuracy": 0.9036682615629984, "confusion_matrix": { "fn": 167, "fp": 135, "tn": 2197, "tp": 636 }, "f_beta_score": 0.8081321473951716, "precision": 0.8249027237354085, "recall": 0.7920298879202988 }, "2": { "accuracy": 0.9808612440191388, "confusion_matrix": { "fn": 21, "fp": 39, "tn": 2830, "tp": 245 }, "f_beta_score": 0.8909090909090909, "precision": 0.8626760563380281, "recall": 0.9210526315789473 }, "3": { "accuracy": 0.9502392344497608, "confusion_matrix": { "fn": 101, "fp": 55, "tn": 1708, "tp": 1271 }, "f_beta_score": 0.9421793921423276, "precision": 0.9585218702865762, "recall": 0.9263848396501457 }, "4": { "accuracy": 0.9314194577352473, "confusion_matrix": { "fn": 69, "fp": 146, "tn": 2477, "tp": 443 }, "f_beta_score": 0.8047229791099001, "precision": 0.7521222410865874, "recall": 0.865234375 } }, "eval_confusion_matrix": { "fn": 423, "fp": 423, "tn": 12117, "tp": 2712 }, "eval_f_beta_score": 0.8654967184485823, "eval_label_confusion_matrix": { "labels": [ "FYI", "NEEDS_RESPONSE", "PERSONAL", "PROMOTIONAL", "URGENT" ], "matrix": [ [ 117, 36, 4, 20, 5 ], [ 40, 636, 10, 19, 98 ], [ 1, 5, 245, 7, 8 ], [ 6, 59, 1, 1271, 35 ], [ 1, 35, 24, 9, 443 ] ] }, "eval_loss": 0.5608769059181213, "eval_precision": 0.8679750740253618, "eval_recall": 0.8650717703349282, "eval_runtime": 19.0243, "eval_sample_size": 3135, "eval_samples_per_second": 164.789, "eval_steps_per_second": 20.605, "step": 1568 }, { "epoch": 2.0, "grad_norm": 21.096660614013672, "hard_loss": 0.4585136638052063, "learning_rate": 4.165098356927757e-05, "loss": 0.4585, "step": 3136 }, { "epoch": 2.0, "eval_accuracy": 0.8902711323763955, "eval_class_metrics": { "0": { "accuracy": 0.9728867623604466, "confusion_matrix": { "fn": 69, "fp": 16, "tn": 2937, "tp": 113 }, "f_beta_score": 0.7266881028938906, "precision": 0.875968992248062, "recall": 0.6208791208791209 }, "1": { "accuracy": 0.9247208931419457, "confusion_matrix": { "fn": 121, "fp": 115, "tn": 2217, "tp": 682 }, "f_beta_score": 0.8525, "precision": 0.8557089084065245, "recall": 0.8493150684931506 }, "2": { "accuracy": 0.980542264752791, "confusion_matrix": { "fn": 23, "fp": 38, "tn": 2831, "tp": 243 }, "f_beta_score": 0.8884826325411335, "precision": 0.8647686832740213, "recall": 0.9135338345864662 }, "3": { "accuracy": 0.9569377990430622, "confusion_matrix": { "fn": 26, "fp": 109, "tn": 1654, "tp": 1346 }, "f_beta_score": 0.9522461973823841, "precision": 0.925085910652921, "recall": 0.9810495626822158 }, "4": { "accuracy": 0.9454545454545454, "confusion_matrix": { "fn": 105, "fp": 66, "tn": 2557, "tp": 407 }, "f_beta_score": 0.8263959390862944, "precision": 0.8604651162790697, "recall": 0.794921875 } }, "eval_confusion_matrix": { "fn": 344, "fp": 344, "tn": 12196, "tp": 2791 }, "eval_f_beta_score": 0.8876387938129007, "eval_label_confusion_matrix": { "labels": [ "FYI", "NEEDS_RESPONSE", "PERSONAL", "PROMOTIONAL", "URGENT" ], "matrix": [ [ 113, 47, 2, 19, 1 ], [ 8, 682, 12, 58, 43 ], [ 0, 8, 243, 6, 9 ], [ 6, 6, 1, 1346, 13 ], [ 2, 54, 23, 26, 407 ] ] }, "eval_loss": 0.5647311210632324, "eval_precision": 0.8887926917834664, "eval_recall": 0.8902711323763955, "eval_runtime": 19.121, "eval_sample_size": 3135, "eval_samples_per_second": 163.956, "eval_steps_per_second": 20.501, "step": 3136 } ], "logging_steps": 500, "max_steps": 31360, "num_input_tokens_seen": 0, "num_train_epochs": 20, "save_steps": 500, "stateful_callbacks": { "EarlyStoppingCallback": { "args": { "early_stopping_patience": 3, "early_stopping_threshold": 0.01 }, "attributes": { "early_stopping_patience_counter": 0 } }, "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 5962158438930816.0, "train_batch_size": 8, "trial_name": null, "trial_params": null }