{ "effective_tokens_per_sec": 7414.328680986972, "epoch": 4.0, "total_flos": 3.519329211045118e+18, "train_loss": 0.8784391032123708, "train_runtime": 9185.186, "train_samples_per_second": 6.558, "train_steps_per_second": 0.109 }