{ "effective_tokens_per_sec": 2340.6354596146743, "epoch": 3.0, "total_flos": 2.73366675523371e+17, "train_loss": 0.053853037605258475, "train_runtime": 459.8341, "train_samples_per_second": 48.755, "train_steps_per_second": 0.763 }