Fixes to alternating SWA layers in Gemma2 #31775
Merged
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
|
Thanks for your PR @turboderp, we're taking a look with @ArthurZucker |
4 tasks
|
Any updates on this? It's likely required to get the proper performance out of the Gemma 2 models |
ArthurZucker
approved these changes
Jul 10, 2024
This comment was marked as duplicate.
This comment was marked as duplicate.
Sorry, something went wrong.
|
The slow tests are gonna fail potentially cc @ydshieh if it's alright with you to update later on? I think a patch will include this! |
|
Thanks @turboderp |
ArthurZucker
pushed a commit
that referenced
this pull request
Jul 11, 2024
* HybridCache: Flip order of alternating global-attn/sliding-attn layers * HybridCache: Read sliding_window argument from cache_kwargs * Gemma2Model: Flip order of alternating global-attn/sliding-attn layers * Code formatting
amyeroberts
pushed a commit
to amyeroberts/transformers
that referenced
this pull request
Jul 19, 2024
* HybridCache: Flip order of alternating global-attn/sliding-attn layers * HybridCache: Read sliding_window argument from cache_kwargs * Gemma2Model: Flip order of alternating global-attn/sliding-attn layers * Code formatting
MHRDYN7
pushed a commit
to MHRDYN7/transformers
that referenced
this pull request
Jul 23, 2024
* HybridCache: Flip order of alternating global-attn/sliding-attn layers * HybridCache: Read sliding_window argument from cache_kwargs * Gemma2Model: Flip order of alternating global-attn/sliding-attn layers * Code formatting
zucchini-nlp
pushed a commit
to zucchini-nlp/transformers
that referenced
this pull request
Jul 24, 2024
* HybridCache: Flip order of alternating global-attn/sliding-attn layers * HybridCache: Read sliding_window argument from cache_kwargs * Gemma2Model: Flip order of alternating global-attn/sliding-attn layers * Code formatting
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?
Reverses the order of global and sliding attention layers in Gemma2. This brings it in line with Google's implementation in which sliding attention is used on layers 0, 2, 4.., whereas currently the Transformers implementation uses sliding attn on layers 1, 3, 5...
Changes
HybridCache.updateto read thesliding_windowargument fromcache_kwargssince it wasn't being parsed otherwise. The cache was created with alternating max seq lenghts of 4k and 8k, but all layers were being updated as if they were 8k, causing out-of-bounds errors and CUDA exceptions.Before submitting
Pull Request section?
to it if that's the case.
documentation guidelines, and
here are tips on formatting docstrings.
Who can review?
Anyone in the community is free to review the PR once the tests have passed. Feel free to tag
members/contributors who may be interested in your PR. @ArthurZucker