{ "effective_tokens_per_sec": 7357.671030733768, "epoch": 4.0, "total_flos": 6.453258552541708e+18, "train_loss": 0.5996934499091338, "train_runtime": 17171.6272, "train_samples_per_second": 4.923, "train_steps_per_second": 0.308 }