{ "effective_tokens_per_sec": 7977.5366389034325, "epoch": 2.0, "total_flos": 3.038711167334744e+18, "train_loss": 0.6924078768212223, "train_runtime": 3699.7491, "train_samples_per_second": 12.017, "train_steps_per_second": 0.401 }