{ "best_metric": NaN, "best_model_checkpoint": "miner_id_24/checkpoint-10", "epoch": 0.0001649899562364141, "eval_steps": 5, "global_step": 10, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 1.649899562364141e-05, "eval_loss": NaN, "eval_runtime": 530.6325, "eval_samples_per_second": 48.094, "eval_steps_per_second": 24.047, "step": 1 }, { "epoch": 4.9496986870924236e-05, "grad_norm": NaN, "learning_rate": 0.00012, "loss": 0.0, "step": 3 }, { "epoch": 8.249497811820705e-05, "eval_loss": NaN, "eval_runtime": 532.5134, "eval_samples_per_second": 47.924, "eval_steps_per_second": 23.962, "step": 5 }, { "epoch": 9.899397374184847e-05, "grad_norm": NaN, "learning_rate": 0.0001992114701314478, "loss": 0.0, "step": 6 }, { "epoch": 0.00014849096061277269, "grad_norm": NaN, "learning_rate": 0.00018763066800438636, "loss": 0.0, "step": 9 }, { "epoch": 0.0001649899562364141, "eval_loss": NaN, "eval_runtime": 527.4181, "eval_samples_per_second": 48.387, "eval_steps_per_second": 24.193, "step": 10 } ], "logging_steps": 3, "max_steps": 30, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 10, "stateful_callbacks": { "EarlyStoppingCallback": { "args": { "early_stopping_patience": 1, "early_stopping_threshold": 0.0 }, "attributes": { "early_stopping_patience_counter": 0 } }, "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 194674081136640.0, "train_batch_size": 2, "trial_name": null, "trial_params": null }