{ "best_metric": null, "best_model_checkpoint": null, "epoch": 3.0, "eval_steps": 500, "global_step": 21, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.14, "grad_norm": 10.338247707526135, "learning_rate": 3.333333333333333e-07, "loss": 1.7184, "step": 1 }, { "epoch": 0.71, "grad_norm": 8.76464917733351, "learning_rate": 9.698463103929541e-07, "loss": 1.7425, "step": 5 }, { "epoch": 1.0, "eval_loss": 1.8325284719467163, "eval_runtime": 1.5341, "eval_samples_per_second": 5.867, "eval_steps_per_second": 0.652, "step": 7 }, { "epoch": 1.43, "grad_norm": 5.930723640057177, "learning_rate": 6.710100716628344e-07, "loss": 1.6727, "step": 10 }, { "epoch": 2.0, "eval_loss": 1.8040801286697388, "eval_runtime": 1.3208, "eval_samples_per_second": 6.814, "eval_steps_per_second": 0.757, "step": 14 }, { "epoch": 2.14, "grad_norm": 4.985313025787138, "learning_rate": 2.500000000000001e-07, "loss": 1.6104, "step": 15 }, { "epoch": 2.86, "grad_norm": 4.707591365221186, "learning_rate": 7.59612349389599e-09, "loss": 1.5411, "step": 20 }, { "epoch": 3.0, "eval_loss": 1.7995264530181885, "eval_runtime": 1.3179, "eval_samples_per_second": 6.829, "eval_steps_per_second": 0.759, "step": 21 }, { "epoch": 3.0, "step": 21, "total_flos": 4344627855360.0, "train_loss": 1.6381536835715884, "train_runtime": 361.3139, "train_samples_per_second": 3.67, "train_steps_per_second": 0.058 } ], "logging_steps": 5, "max_steps": 21, "num_input_tokens_seen": 0, "num_train_epochs": 3, "save_steps": 500, "total_flos": 4344627855360.0, "train_batch_size": 4, "trial_name": null, "trial_params": null }