restore SDPA in gpt neo after 4.45 #1092
Open
+119
−1
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
|
The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update. |
eaidova
force-pushed
the
ea/gpt_neo_sdpa
branch
from
e2ceb43 to
03803c5
Compare
AlexKoff88
approved these changes
Dec 24, 2024
|
I am ok with the changes but I am thinking that next time we can just drop the support of this model as it is the legacy one. |
|
failed tests are not related to changes, it is issue with nonbackward compatible changes in openvino (should be passed in next nightly) |
eaidova
commented
Dec 25, 2024
|
diffusers tests fixes are covered by #1094 |
eaidova
force-pushed
the
ea/gpt_neo_sdpa
branch
from
7428615 to
1e00f2b
Compare
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Outdated suggestions cannot be applied.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.
What does this PR do?
bettertransformers is not compatible with gpt-neo starting 4.45, but transformers only provide eager and flash_attn attention implementations, this PR adds sdpa for gpt_neo