VLMs: small clean-up for cache class #32417
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
|
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. |
4 tasks
amyeroberts
reviewed
Aug 14, 2024
There was a problem hiding this comment.
Thanks for fixing @zucchini-nlp!
Are there slow tests which should have caught this? If so, can we run them here to make sure they all pass? If not, can we add some?
ArthurZucker
approved these changes
Aug 15, 2024
|
Yep, after resolving conflicts the fix wasn't needed anymore. I'll rename PR and merge it :) |
zucchini-nlp
changed the title
zucchini-nlp
changed the title
zucchini-nlp
changed the title
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Outdated suggestions cannot be applied.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?
VideoLLaVa was generating garbage in beam search as reported by some users after moving to the new cache format. Thsi PR fixes it as we no longer will have
past_key_values is Nonein any generation step. Except for when caching isn't enabled at all.Also a little clean up on VLMs, after removing old cache format for all major LLMs we no longer need the
reorder cachemethod