{ "effective_tokens_per_sec": 7427.754507135659, "epoch": 3.0, "total_flos": 4.929269661905715e+18, "train_loss": 1.00607624289008, "train_runtime": 12927.7338, "train_samples_per_second": 5.096, "train_steps_per_second": 0.085 }