{ "effective_tokens_per_sec": 7374.398147293401, "epoch": 4.0, "total_flos": 3.519329211045118e+18, "train_loss": 0.8034457557749345, "train_runtime": 9234.9215, "train_samples_per_second": 6.523, "train_steps_per_second": 0.109 }