{ "best_metric": null, "best_model_checkpoint": null, "epoch": 0.012503907471084715, "eval_steps": 2, "global_step": 5, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0025007814942169426, "grad_norm": NaN, "learning_rate": 1e-05, "loss": 129.6588, "step": 1 }, { "epoch": 0.0025007814942169426, "eval_loss": NaN, "eval_runtime": 439.112, "eval_samples_per_second": 1.535, "eval_steps_per_second": 0.767, "step": 1 }, { "epoch": 0.005001562988433885, "grad_norm": NaN, "learning_rate": 2e-05, "loss": 291.7946, "step": 2 }, { "epoch": 0.005001562988433885, "eval_loss": NaN, "eval_runtime": 300.1271, "eval_samples_per_second": 2.246, "eval_steps_per_second": 1.123, "step": 2 }, { "epoch": 0.007502344482650828, "grad_norm": NaN, "learning_rate": 3e-05, "loss": 0.2961, "step": 3 }, { "epoch": 0.01000312597686777, "grad_norm": NaN, "learning_rate": 4e-05, "loss": 77.4935, "step": 4 }, { "epoch": 0.01000312597686777, "eval_loss": NaN, "eval_runtime": 295.5873, "eval_samples_per_second": 2.28, "eval_steps_per_second": 1.14, "step": 4 }, { "epoch": 0.012503907471084715, "grad_norm": NaN, "learning_rate": 5e-05, "loss": 0.1346, "step": 5 } ], "logging_steps": 1, "max_steps": 5, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 2, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 6572656904110080.0, "train_batch_size": 2, "trial_name": null, "trial_params": null }