coderfeli 402a8443d9 port per token quant fix from dev

2025-03-06 11:02:50 +00:00

..

block

i4 support lds multiple shuffle

2025-03-04 11:06:15 +08:00

device

Revert "fix build"

2025-03-04 11:12:33 +08:00

element

commit a version for compiler debug.

2025-02-24 14:44:02 +08:00

grid

port per token quant fix from dev

2025-03-06 11:02:50 +00:00

thread

i4 support lds multiple shuffle

2025-03-04 11:06:15 +08:00

warp

Merge from internal (#1857)

2025-02-07 15:05:05 -07:00