{ "effective_tokens_per_sec": 7881.3747448722315, "epoch": 3.0, "total_flos": 4.931331991116186e+18, "train_loss": 0.9669766378315854, "train_runtime": 12176.252, "train_samples_per_second": 5.411, "train_steps_per_second": 0.09 }