graph: backend: dnnl: not use decomp sdpa kernel for dynamic quantized cases #2458
+2
−5
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
wzt1997
force-pushed
the
zhitao/fix-decomp-sdp-with-dynamic-dequant
branch
from
ff023d7 to
8855900
Compare
gyhintel
approved these changes
Jan 21, 2025
xiang1guo
approved these changes
Jan 21, 2025
|
make test |
ElaineBao
approved these changes
Jan 21, 2025
rongzha1
approved these changes
Jan 21, 2025
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Description
Recently we supported a SDPA with compressed K using per_channel quantization case, which will run into the decomposed sdpa primitive kernel unexpectedly with
THREADPOOLcpu runtime( as it passes the check).As we do not aim to support dynamic quantized cases in decomposed sdpa kernel, the PR enhanced the restriction to avoid such case.