{ "effective_tokens_per_sec": 6271.3017324294215, "epoch": 4.0, "total_flos": 3.438047841308639e+18, "train_loss": 0.8972434957087567, "train_runtime": 10632.4216, "train_samples_per_second": 4.93, "train_steps_per_second": 0.308 }