CUDA: stream-k decomposition for MMQ (#8018)
* CUDA: stream-k decomposition for MMQ * fix undefined memory reads for small matrices
J
Johannes Gäßler committed
d50f8897a797a5a03f31228d1b5a7b8130ee1bc2
Parent: 2075a66
Committed by GitHub <noreply@github.com>
on 6/20/2024, 12:39:21 PM