{ "effective_tokens_per_sec": 8317.189025716589, "epoch": 3.0, "total_flos": 4.957143256761631e+18, "train_loss": 0.8975460599354708, "train_runtime": 11538.2258, "train_samples_per_second": 5.71, "train_steps_per_second": 0.095 }