{ "effective_tokens_per_sec": 877.2247045451046, "epoch": 3.0, "total_flos": 8.4482141520606e+18, "train_loss": 1.0263564972723214, "train_runtime": 57008.8117, "train_samples_per_second": 0.69, "train_steps_per_second": 0.029 }