{ "best_metric": null, "best_model_checkpoint": null, "epoch": 9.433962264150944, "eval_steps": 500, "global_step": 3000, "is_hyper_param_search": true, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 1.0, "eval_accuracy": 0.6529032258064517, "eval_loss": 0.28555235266685486, "eval_runtime": 5.4121, "eval_samples_per_second": 572.786, "eval_steps_per_second": 12.01, "step": 318 }, { "epoch": 1.5723270440251573, "grad_norm": 0.6621501445770264, "learning_rate": 1.685534591194969e-05, "loss": 0.4579, "step": 500 }, { "epoch": 2.0, "eval_accuracy": 0.8451612903225807, "eval_loss": 0.1139635443687439, "eval_runtime": 5.9245, "eval_samples_per_second": 523.247, "eval_steps_per_second": 10.971, "step": 636 }, { "epoch": 3.0, "eval_accuracy": 0.8961290322580645, "eval_loss": 0.06579811871051788, "eval_runtime": 5.4507, "eval_samples_per_second": 568.736, "eval_steps_per_second": 11.925, "step": 954 }, { "epoch": 3.1446540880503147, "grad_norm": 0.5677153468132019, "learning_rate": 1.371069182389937e-05, "loss": 0.1346, "step": 1000 }, { "epoch": 4.0, "eval_accuracy": 0.9138709677419354, "eval_loss": 0.04725706949830055, "eval_runtime": 5.5473, "eval_samples_per_second": 558.83, "eval_steps_per_second": 11.717, "step": 1272 }, { "epoch": 4.716981132075472, "grad_norm": 0.3639396131038666, "learning_rate": 1.0566037735849058e-05, "loss": 0.0751, "step": 1500 }, { "epoch": 5.0, "eval_accuracy": 0.92, "eval_loss": 0.03804958611726761, "eval_runtime": 5.7319, "eval_samples_per_second": 540.831, "eval_steps_per_second": 11.34, "step": 1590 }, { "epoch": 6.0, "eval_accuracy": 0.927741935483871, "eval_loss": 0.03348824381828308, "eval_runtime": 5.4253, "eval_samples_per_second": 571.402, "eval_steps_per_second": 11.981, "step": 1908 }, { "epoch": 6.289308176100629, "grad_norm": 0.29170453548431396, "learning_rate": 7.421383647798742e-06, "loss": 0.0566, "step": 2000 }, { "epoch": 7.0, "eval_accuracy": 0.932258064516129, "eval_loss": 0.0307790357619524, "eval_runtime": 5.6345, "eval_samples_per_second": 550.182, "eval_steps_per_second": 11.536, "step": 2226 }, { "epoch": 7.861635220125786, "grad_norm": 0.32524076104164124, "learning_rate": 4.276729559748428e-06, "loss": 0.0488, "step": 2500 }, { "epoch": 8.0, "eval_accuracy": 0.9319354838709677, "eval_loss": 0.028849748894572258, "eval_runtime": 5.7512, "eval_samples_per_second": 539.022, "eval_steps_per_second": 11.302, "step": 2544 }, { "epoch": 9.0, "eval_accuracy": 0.9341935483870968, "eval_loss": 0.027793901041150093, "eval_runtime": 5.4474, "eval_samples_per_second": 569.075, "eval_steps_per_second": 11.932, "step": 2862 }, { "epoch": 9.433962264150944, "grad_norm": 0.28644028306007385, "learning_rate": 1.1320754716981133e-06, "loss": 0.0453, "step": 3000 } ], "logging_steps": 500, "max_steps": 3180, "num_input_tokens_seen": 0, "num_train_epochs": 10, "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": 780738843279612.0, "train_batch_size": 48, "trial_name": null, "trial_params": { "alpha": 0.31688267813599835, "num_train_epochs": 10, "temperature": 3 } }