{ "effective_tokens_per_sec": 2246.4632334123307, "epoch": 4.0, "total_flos": 3.954132003103703e+18, "train_loss": 0.909145103641359, "train_runtime": 34871.4276, "train_samples_per_second": 1.727, "train_steps_per_second": 0.029 }