{ "effective_tokens_per_sec": 2270.7816778928122, "epoch": 4.0, "total_flos": 3.954132003103703e+18, "train_loss": 0.8280324278013461, "train_runtime": 34497.9796, "train_samples_per_second": 1.746, "train_steps_per_second": 0.029 }