{ "effective_tokens_per_sec": 4202.352901714637, "epoch": 3.0, "total_flos": 4.871467377855824e+18, "train_loss": 1.1715838705861805, "train_runtime": 22979.7614, "train_samples_per_second": 2.523, "train_steps_per_second": 0.079 }