Update QMoE kernel with optimizations #26091
apsonawane
force pushed
from
426e1ac7
to
509e17c4
138 days ago
Fix merge conflicts
83ffc5d9
Re-enable quantized Mlas
d399c66d
apsonawane
force pushed
from
509e17c4
to
41133b7d
135 days ago
Add overflow safety changes
8289fcb1
apsonawane
force pushed
from
41133b7d
to
8289fcb1
135 days ago
Disable quantized Mlas, still not giving good tps
d57a7c3e
Assignees
No one assigned
Login to write a write a comment.
Login via GitHub