{ "best_metric": 0.671207419944122, "best_model_checkpoint": "/nfs/production/literature/amina-mardiyyah/new_data/Ner_Pipeline/src/ner_pipeline/model_outputs/NER/CeLLaTe_3class_no_vague_3.0/BiomedNLP_BiomedBERT_base_uncased_abstract_fulltext/BaseStrategy/StandardNerHead/BaseTrainer/no_data_aug/label_all_tokens=False/checkpoint-1841", "epoch": 12.0, "eval_steps": 100, "global_step": 3156, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 1.0, "grad_norm": 1.5280296802520752, "learning_rate": 1.9204916458613406e-05, "loss": 0.3388, "step": 263 }, { "epoch": 1.0, "eval_accuracy": 0.9798439531859557, "eval_loss": 0.06261003017425537, "eval_macro_f1": 0.5871795975352566, "eval_micro_f1": 0.6030368763557484, "eval_precision": 0.569672131147541, "eval_recall": 0.6405529953917051, "eval_weighted_f1": 0.5999065069916109, "step": 263 }, { "epoch": 2.0, "grad_norm": 0.535184919834137, "learning_rate": 1.819473785289034e-05, "loss": 0.0448, "step": 526 }, { "epoch": 2.0, "eval_accuracy": 0.9825629507033928, "eval_loss": 0.05994148924946785, "eval_macro_f1": 0.665341975469488, "eval_micro_f1": 0.6762656758012077, "eval_precision": 0.6816479400749064, "eval_recall": 0.6709677419354839, "eval_weighted_f1": 0.673744334364318, "step": 526 }, { "epoch": 3.0, "grad_norm": 1.0914032459259033, "learning_rate": 1.7184559247167276e-05, "loss": 0.0307, "step": 789 }, { "epoch": 3.0, "eval_accuracy": 0.9808093943334516, "eval_loss": 0.0685759037733078, "eval_macro_f1": 0.6430273316967005, "eval_micro_f1": 0.6576086956521738, "eval_precision": 0.6464826357969724, "eval_recall": 0.6691244239631337, "eval_weighted_f1": 0.6531152541245993, "step": 789 }, { "epoch": 4.0, "grad_norm": 0.7419782876968384, "learning_rate": 1.617438064144421e-05, "loss": 0.0215, "step": 1052 }, { "epoch": 4.0, "eval_accuracy": 0.9815186980336525, "eval_loss": 0.0823308676481247, "eval_macro_f1": 0.6424893700582263, "eval_micro_f1": 0.662894860914663, "eval_precision": 0.6785714285714286, "eval_recall": 0.647926267281106, "eval_weighted_f1": 0.6557439600659847, "step": 1052 }, { "epoch": 5.0, "grad_norm": 1.4448457956314087, "learning_rate": 1.5164202035721145e-05, "loss": 0.017, "step": 1315 }, { "epoch": 5.0, "eval_accuracy": 0.9819324585254364, "eval_loss": 0.08244974166154861, "eval_macro_f1": 0.6430321181667845, "eval_micro_f1": 0.663763909046928, "eval_precision": 0.6985743380855397, "eval_recall": 0.632258064516129, "eval_weighted_f1": 0.6565625815951753, "step": 1315 }, { "epoch": 6.0, "grad_norm": 0.29027673602104187, "learning_rate": 1.415402342999808e-05, "loss": 0.0134, "step": 1578 }, { "epoch": 6.0, "eval_accuracy": 0.9826417622256374, "eval_loss": 0.07894663512706757, "eval_macro_f1": 0.6670742796238143, "eval_micro_f1": 0.679296346414073, "eval_precision": 0.6651943462897526, "eval_recall": 0.6940092165898617, "eval_weighted_f1": 0.6771928419768851, "step": 1578 }, { "epoch": 7.0, "grad_norm": 0.07943511009216309, "learning_rate": 1.3143844824275017e-05, "loss": 0.0097, "step": 1841 }, { "epoch": 7.0, "eval_accuracy": 0.9826811679867596, "eval_loss": 0.08737103641033173, "eval_macro_f1": 0.671207419944122, "eval_micro_f1": 0.6872979214780601, "eval_precision": 0.6888888888888889, "eval_recall": 0.6857142857142857, "eval_weighted_f1": 0.683550717613432, "step": 1841 }, { "epoch": 8.0, "grad_norm": 0.01676645316183567, "learning_rate": 1.213366621855195e-05, "loss": 0.0075, "step": 2104 }, { "epoch": 8.0, "eval_accuracy": 0.9822477046144146, "eval_loss": 0.09638739377260208, "eval_macro_f1": 0.6623525909690223, "eval_micro_f1": 0.6805170821791321, "eval_precision": 0.6817761332099908, "eval_recall": 0.6792626728110599, "eval_weighted_f1": 0.6754201718608044, "step": 2104 }, { "epoch": 9.0, "grad_norm": 0.2930453419685364, "learning_rate": 1.1123487612828885e-05, "loss": 0.0065, "step": 2367 }, { "epoch": 9.0, "eval_accuracy": 0.9823462190172203, "eval_loss": 0.09757288545370102, "eval_macro_f1": 0.6621652839617043, "eval_micro_f1": 0.6763466042154567, "eval_precision": 0.6876190476190476, "eval_recall": 0.6654377880184332, "eval_weighted_f1": 0.6741236231853154, "step": 2367 }, { "epoch": 10.0, "grad_norm": 0.007754873484373093, "learning_rate": 1.011330900710582e-05, "loss": 0.0055, "step": 2630 }, { "epoch": 10.0, "eval_accuracy": 0.9823462190172203, "eval_loss": 0.10528703033924103, "eval_macro_f1": 0.6644604090505265, "eval_micro_f1": 0.6832473017362742, "eval_precision": 0.6959847036328872, "eval_recall": 0.6709677419354839, "eval_weighted_f1": 0.6775409426460733, "step": 2630 }, { "epoch": 11.0, "grad_norm": 0.024706775322556496, "learning_rate": 9.103130401382755e-06, "loss": 0.0044, "step": 2893 }, { "epoch": 11.0, "eval_accuracy": 0.9827993852701264, "eval_loss": 0.10210014879703522, "eval_macro_f1": 0.6630466914471599, "eval_micro_f1": 0.6817343173431735, "eval_precision": 0.6823638042474608, "eval_recall": 0.6811059907834102, "eval_weighted_f1": 0.6779281758066938, "step": 2893 }, { "epoch": 12.0, "grad_norm": 0.028232423588633537, "learning_rate": 8.092951795659689e-06, "loss": 0.0041, "step": 3156 }, { "epoch": 12.0, "eval_accuracy": 0.982306813256098, "eval_loss": 0.10995829105377197, "eval_macro_f1": 0.6550023777979893, "eval_micro_f1": 0.6725417439703155, "eval_precision": 0.676937441643324, "eval_recall": 0.6682027649769585, "eval_weighted_f1": 0.6697272214604535, "step": 3156 }, { "epoch": 12.0, "step": 3156, "total_flos": 2554302996301254.0, "train_loss": 0.04199507938289521, "train_runtime": 361.4294, "train_samples_per_second": 465.319, "train_steps_per_second": 14.553 } ], "logging_steps": 10, "max_steps": 5260, "num_input_tokens_seen": 0, "num_train_epochs": 20, "save_steps": 100, "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": 2554302996301254.0, "train_batch_size": 16, "trial_name": null, "trial_params": null }