{ "effective_tokens_per_sec": 7728.355210837788, "epoch": 3.0, "total_flos": 2.7756656592022405e+18, "train_loss": 1.0638441426315526, "train_runtime": 6969.7468, "train_samples_per_second": 5.992, "train_steps_per_second": 0.05 }