Their loss curve with the RL didn't level off much though, could be taken a lot further and scaled up to more parameters on the big nvidia mega clusters out the

Their loss curve with the RL didn't level off much though, could be taken a lot further and scaled up to more parameters on the big nvidia mega clusters out there. And the architecture is heavily tuned to nvidia optimizations.