{ "best_metric": 0.18560722470283508, "best_model_checkpoint": "biolinkbert_caption_reference_RTX6000/checkpoint-1230", "epoch": 2.0, "eval_steps": 500, "global_step": 2460, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.4065040650406504, "grad_norm": 4.879099369049072, "learning_rate": 2.75609756097561e-05, "loss": 0.2196, "step": 500 }, { "epoch": 0.8130081300813008, "grad_norm": 1.5588421821594238, "learning_rate": 2.5121951219512197e-05, "loss": 0.1697, "step": 1000 }, { "epoch": 1.0, "eval_accuracy": 0.92894387076986, "eval_f1": 0.9240870898542597, "eval_f1_medical": 0.9432884536923705, "eval_f1_non_medical": 0.9048857260161489, "eval_loss": 0.18560722470283508, "eval_precision": 0.9266766962431328, "eval_recall": 0.9217536238643151, "eval_runtime": 100.9349, "eval_samples_per_second": 96.904, "eval_steps_per_second": 3.032, "step": 1230 }, { "epoch": 1.2195121951219512, "grad_norm": 1.4866325855255127, "learning_rate": 2.2682926829268295e-05, "loss": 0.135, "step": 1500 }, { "epoch": 1.6260162601626016, "grad_norm": 2.8782901763916016, "learning_rate": 2.024390243902439e-05, "loss": 0.1139, "step": 2000 }, { "epoch": 2.0, "eval_accuracy": 0.9333401492689909, "eval_f1": 0.9292895437603008, "eval_f1_medical": 0.9462134961227521, "eval_f1_non_medical": 0.9123655913978495, "eval_loss": 0.2073008269071579, "eval_precision": 0.928674499317921, "eval_recall": 0.929922559124454, "eval_runtime": 93.6114, "eval_samples_per_second": 104.485, "eval_steps_per_second": 3.269, "step": 2460 } ], "logging_steps": 500, "max_steps": 6150, "num_input_tokens_seen": 0, "num_train_epochs": 5, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 2.071174050023808e+16, "train_batch_size": 32, "trial_name": null, "trial_params": null }