{ "effective_tokens_per_sec": 6658.574016285121, "epoch": 4.0, "total_flos": 3.438047841308639e+18, "train_loss": 0.884952375524301, "train_runtime": 10014.0246, "train_samples_per_second": 5.234, "train_steps_per_second": 0.327 }