{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 2813, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.17774617845716317, "grad_norm": 0.15928706526756287, "learning_rate": 1.8226093138997513e-05, "loss": 0.03903742980957031, "step": 500 }, { "epoch": 0.35549235691432635, "grad_norm": 0.5245766043663025, "learning_rate": 1.6448631354425882e-05, "loss": 0.03496078109741211, "step": 1000 }, { "epoch": 0.5332385353714895, "grad_norm": 0.19340601563453674, "learning_rate": 1.4671169569854249e-05, "loss": 0.03378898239135742, "step": 1500 }, { "epoch": 0.7109847138286527, "grad_norm": 0.14931844174861908, "learning_rate": 1.2893707785282616e-05, "loss": 0.032428707122802734, "step": 2000 }, { "epoch": 0.8887308922858158, "grad_norm": 0.1427156776189804, "learning_rate": 1.1116246000710985e-05, "loss": 0.03130688285827637, "step": 2500 }, { "epoch": 1.0, "eval_accuracy": 0.9922607455690136, "eval_f1": 0.9399925570272034, "eval_loss": 0.029042916372418404, "eval_precision": 0.9330253062608622, "eval_recall": 0.9470646448131097, "eval_runtime": 222.0032, "eval_samples_per_second": 90.089, "eval_steps_per_second": 1.41, "step": 2813 } ], "logging_steps": 500, "max_steps": 5626, "num_input_tokens_seen": 0, "num_train_epochs": 2, "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": 5911301099520000.0, "train_batch_size": 64, "trial_name": null, "trial_params": null }