[Kernel][Triton][AMD] Use block size heuristic for avg 2.8x speedup for int8 models #1184
mgoinadd_label_automerge.yml
on: pull_request_target
add-label-on-auto-merge
3s
Annotations
1 warning
|
add-label-on-auto-merge
ubuntu-latest pipelines will use ubuntu-24.04 soon. For more details, see https://github.com/actions/runner-images/issues/10636
|