{ "effective_tokens_per_sec": 8041.289292236616, "epoch": 4.0, "total_flos": 3.5845266747435254e+18, "train_loss": 0.8250222063222683, "train_runtime": 8638.0168, "train_samples_per_second": 6.571, "train_steps_per_second": 0.219 }