3 ms·

Except the model would have been trained on the available corpus of known runners and will achieve the same average level of quality...