{ "effective_tokens_per_sec": 4889.172491836145, "epoch": 3.0, "total_flos": 2.7548743231466373e+18, "train_loss": 1.0648414060130886, "train_runtime": 11016.7324, "train_samples_per_second": 3.791, "train_steps_per_second": 0.063 }