{ "effective_tokens_per_sec": 860.6135136203767, "epoch": 3.0, "total_flos": 8.4482141520606e+18, "train_loss": 0.9984596982616499, "train_runtime": 58109.1712, "train_samples_per_second": 0.677, "train_steps_per_second": 0.028 }