{ "effective_tokens_per_sec": 7561.8214712507715, "epoch": 4.0, "total_flos": 3.519329211045118e+18, "train_loss": 0.882850165816536, "train_runtime": 9006.0296, "train_samples_per_second": 6.689, "train_steps_per_second": 0.111 }