{ "best_metric": null, "best_model_checkpoint": null, "epoch": 13.333333333333334, "eval_steps": 500, "global_step": 500, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 1.33, "grad_norm": 0.14105422794818878, "learning_rate": 0.00019851186053243666, "loss": 1.8081, "step": 50 }, { "epoch": 2.67, "grad_norm": 0.2578440308570862, "learning_rate": 0.00019036283606085053, "loss": 1.6886, "step": 100 }, { "epoch": 4.0, "grad_norm": 0.5536919832229614, "learning_rate": 0.00017567128158176953, "loss": 1.6036, "step": 150 }, { "epoch": 5.33, "grad_norm": 0.2558369040489197, "learning_rate": 0.000155500908021347, "loss": 1.4033, "step": 200 }, { "epoch": 6.67, "grad_norm": 0.497671514749527, "learning_rate": 0.00013131210861240026, "loss": 1.3546, "step": 250 }, { "epoch": 8.0, "grad_norm": 0.7663662433624268, "learning_rate": 0.00010485622221144484, "loss": 1.2003, "step": 300 }, { "epoch": 9.33, "grad_norm": 0.3598242700099945, "learning_rate": 7.804873131325954e-05, "loss": 1.0896, "step": 350 }, { "epoch": 10.67, "grad_norm": 0.7714847326278687, "learning_rate": 5.283057559252341e-05, "loss": 1.0081, "step": 400 }, { "epoch": 12.0, "grad_norm": 0.9072046875953674, "learning_rate": 3.102762227218957e-05, "loss": 0.9447, "step": 450 }, { "epoch": 13.33, "grad_norm": 0.40007126331329346, "learning_rate": 1.4218468069322578e-05, "loss": 0.9571, "step": 500 } ], "logging_steps": 50, "max_steps": 600, "num_input_tokens_seen": 0, "num_train_epochs": 17, "save_steps": 500, "total_flos": 2.4480902516598374e+17, "train_batch_size": 4, "trial_name": null, "trial_params": null }