{ "effective_tokens_per_sec": 8322.244592994637, "epoch": 3.0, "total_flos": 4.957143256761631e+18, "train_loss": 0.9666990600322765, "train_runtime": 11531.2166, "train_samples_per_second": 5.713, "train_steps_per_second": 0.095 }