[CPU] [FP8 SDPA] Enable FP8 SDPA pattern match #3076
+297
−41
Conversation
This file contains hidden or 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
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/ao/3076
Note: Links to docs will display an error until the docs builds have been completed. ❗ 1 Active SEVsThere are 1 currently active SEVs. If your PR is affected, please view them below: ✅ No FailuresAs of commit 1ad747c with merge base 233cfc1 ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
meta-cla
bot
added
the
CLA Signed
Valentine233
added
topic: not user facing
meta-cla
bot
added
the
CLA Signed
Valentine233
force-pushed
the
fp8_sdpa_pattern
branch
from
7fc9ea4 to
1ad747c
Compare
|
@mingfeima Please help take a look at the PR, thanks! |
mingfeima
approved these changes
Oct 10, 2025
Comment on lines
+141
to
+146
| self.q_out_scale = 1.5 | ||
| self.k_out_scale = 1.5 | ||
| self.attn_weights_scale = 1.5 | ||
| self.v_out_scale = 1.5 | ||
| self.attn_out_scale = 1.5 | ||
| self.qk_out_scale = 1.5 |
There was a problem hiding this comment.
any particular reason we use the magic number 1.5? and FP8QDQLinear is using 2.0
There was a problem hiding this comment.
These values are just randomly selected.
jerryzh168
reviewed
Oct 14, 2025
| if qtype == torch.uint8: | ||
| assert zp is not None, "Zero point must be provided for uint8 dequantization" | ||
| return CallFunction( | ||
| torch.ops.quantized_decomposed.dequantize_per_tensor.default, |
There was a problem hiding this comment.
this is still using the old op, what is the e2e for quantized sdpa?
jerryzh168
approved these changes
Oct 14, 2025
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
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.
Support the FP8 SDPA pattern match.
Depend on #2565.