2 ms·Maybe more data to train on while keeping model size the sameby jarbus 4y agoMaybe more data to train on while keeping model size the same