4 ms·
They mention that as state of the art, this one is supposed to be 14-18x better. By the way, has anyone ran these locally? Is the inference time also lower?
by daghamm 2y ago
They mention that as state of the art, this one is supposed to be 14-18x better.
By the way, has anyone ran these locally? Is the inference time also lower?