adds enable_prefix_caching option to VLLMModelConfig - #945
Merged
Conversation
Collaborator
|
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. |
Contributor
Author
|
@NathanHB seems like a cuda compilation error, do you have any clue why ? |
Member
|
mhh not sure why. I launched again but if it does not work can you try and set the default value of enable prefix caching to None ? |
Contributor
Author
OK I've changed the default value |
NathanHB
changed the title
enable_prefix_caching option to VLLMModelConfig
NathanHB
approved these changes
Sep 23, 2025
rakkit
pushed a commit
to SDLAML/lighteval
that referenced
this pull request
Jun 21, 2026
* add enable_prefix_caching in VLLMModel * typo * None default value --------- Co-authored-by: Nathan Habib <30601243+NathanHB@users.noreply.github.com>
artemorloff
pushed a commit
to artemorloff/lighteval
that referenced
this pull request
Jul 16, 2026
* add enable_prefix_caching in VLLMModel * typo * None default value --------- Co-authored-by: Nathan Habib <30601243+NathanHB@users.noreply.github.com>
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Small patch to support LFM2 with vLLM.
Since LFM2 doesn’t support prefix caching with vLLM, I had to add enable_prefix_caching to both VLLMModelConfig and VLLMModel._create_auto_model to make it work.