{ "effective_tokens_per_sec": 7734.120440513665, "epoch": 4.0, "total_flos": 4.038502240003031e+18, "train_loss": 0.8272027168051893, "train_runtime": 10128.8027, "train_samples_per_second": 5.947, "train_steps_per_second": 0.099 }