{ "effective_tokens_per_sec": 7437.991340194643, "epoch": 4.0, "total_flos": 4.038502240003031e+18, "train_loss": 0.7884150154599215, "train_runtime": 10532.0612, "train_samples_per_second": 5.72, "train_steps_per_second": 0.095 }