{ "effective_tokens_per_sec": 2581.7142880934816, "epoch": 3.0, "total_flos": 3.6341125395219743e+18, "train_loss": 1.5168476336773875, "train_runtime": 11672.6098, "train_samples_per_second": 2.817, "train_steps_per_second": 0.176 }