{ "effective_tokens_per_sec": 7619.424096418565, "epoch": 4.0, "total_flos": 4.038502240003031e+18, "train_loss": 0.9239894755600221, "train_runtime": 10281.2731, "train_samples_per_second": 5.859, "train_steps_per_second": 0.098 }