{ "effective_tokens_per_sec": 4364.043462600282, "epoch": 3.0, "total_flos": 2.909158325117518e+18, "train_loss": 1.307906815331462, "train_runtime": 13215.9511, "train_samples_per_second": 2.488, "train_steps_per_second": 0.078 }