{ "effective_tokens_per_sec": 6675.040163681728, "epoch": 3.0, "total_flos": 4.2988160857187287e+18, "train_loss": 0.7978645538875685, "train_runtime": 6311.8591, "train_samples_per_second": 9.034, "train_steps_per_second": 0.565 }