{ "effective_tokens_per_sec": 2260.7612660843506, "epoch": 4.0, "total_flos": 3.954132003103703e+18, "train_loss": 0.8403717435273041, "train_runtime": 34650.8856, "train_samples_per_second": 1.738, "train_steps_per_second": 0.029 }