{ "best_metric": null, "best_model_checkpoint": null, "epoch": 4.0, "eval_steps": 500, "global_step": 252, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.79, "grad_norm": 2.052056312561035, "learning_rate": 9.059690028579284e-05, "loss": 1.8565, "step": 50 }, { "epoch": 1.0, "eval_loss": 1.6724138259887695, "eval_runtime": 10.7526, "eval_samples_per_second": 18.6, "eval_steps_per_second": 2.325, "step": 63 }, { "epoch": 1.59, "grad_norm": 2.2137675285339355, "learning_rate": 6.592433251258423e-05, "loss": 1.5298, "step": 100 }, { "epoch": 2.0, "eval_loss": 1.6357216835021973, "eval_runtime": 10.7663, "eval_samples_per_second": 18.576, "eval_steps_per_second": 2.322, "step": 126 }, { "epoch": 2.38, "grad_norm": 2.588240623474121, "learning_rate": 3.5262241279454785e-05, "loss": 1.3719, "step": 150 }, { "epoch": 3.0, "eval_loss": 1.664973497390747, "eval_runtime": 10.7606, "eval_samples_per_second": 18.586, "eval_steps_per_second": 2.323, "step": 189 }, { "epoch": 3.17, "grad_norm": 2.6214730739593506, "learning_rate": 1.0143374638853891e-05, "loss": 1.2514, "step": 200 }, { "epoch": 3.97, "grad_norm": 2.66404390335083, "learning_rate": 1.5540899959187727e-08, "loss": 1.1578, "step": 250 }, { "epoch": 4.0, "eval_loss": 1.7135345935821533, "eval_runtime": 10.7577, "eval_samples_per_second": 18.591, "eval_steps_per_second": 2.324, "step": 252 } ], "logging_steps": 50, "max_steps": 252, "num_input_tokens_seen": 0, "num_train_epochs": 4, "save_steps": 500, "total_flos": 2.1885823746048e+16, "train_batch_size": 16, "trial_name": null, "trial_params": null }