forked from Karylab-cklius/vllm
[Bugfix] Use enable_sm120_family for per-tensor FP8 CUTLASS kernels on SM12.1 (#41215)
Signed-off-by: j9smith <j.smith9103@outlook.com> Signed-off-by: Joel Smith <j.smith9103@outlook.com> Co-authored-by: Shengqi Chen <harry-chen@outlook.com>
This commit is contained in:
co-authored by
Shengqi Chen
parent
ded871201a
commit
644b2a28e7