{ "best_metric": null, "best_model_checkpoint": null, "epoch": 5.994570135746606, "eval_steps": 500, "global_step": 5796, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.52, "learning_rate": 4.568668046928917e-05, "loss": 0.7867, "step": 500 }, { "epoch": 1.03, "learning_rate": 4.1373360938578333e-05, "loss": 0.543, "step": 1000 }, { "epoch": 1.55, "learning_rate": 3.70600414078675e-05, "loss": 0.5058, "step": 1500 }, { "epoch": 2.07, "learning_rate": 3.274672187715666e-05, "loss": 0.4939, "step": 2000 }, { "epoch": 2.59, "learning_rate": 2.8433402346445827e-05, "loss": 0.4717, "step": 2500 }, { "epoch": 3.1, "learning_rate": 2.412008281573499e-05, "loss": 0.4665, "step": 3000 }, { "epoch": 3.62, "learning_rate": 1.9806763285024154e-05, "loss": 0.4507, "step": 3500 }, { "epoch": 4.14, "learning_rate": 1.549344375431332e-05, "loss": 0.4455, "step": 4000 }, { "epoch": 4.65, "learning_rate": 1.1180124223602485e-05, "loss": 0.436, "step": 4500 }, { "epoch": 5.17, "learning_rate": 6.86680469289165e-06, "loss": 0.4314, "step": 5000 }, { "epoch": 5.69, "learning_rate": 2.5534851621808144e-06, "loss": 0.4271, "step": 5500 }, { "epoch": 5.99, "step": 5796, "total_flos": 5.39900007388545e+17, "train_loss": 0.4925878707748845, "train_runtime": 10794.6781, "train_samples_per_second": 68.786, "train_steps_per_second": 0.537 } ], "logging_steps": 500, "max_steps": 5796, "num_input_tokens_seen": 0, "num_train_epochs": 6, "save_steps": 500, "total_flos": 5.39900007388545e+17, "train_batch_size": 8, "trial_name": null, "trial_params": null }