Integrate huggingface_hub inference support #651
Merged
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
|
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. |
clefourrier
reviewed
Mar 28, 2025
|
|
||
| case _: | ||
| return lambda x: x | ||
| raise ValueError(f"Unsupported backend: {self.backend}") |
There was a problem hiding this comment.
@NathanHB were you using this case for some specific use cases?
There was a problem hiding this comment.
Oh it was because I wanted the details to be saved even with a failure in the judge so that we don't have to rerun everything.
I forgot to add a warning that's my bad :/
|
Cool PR! A couple nits, and I'd like @NathanHB to take a look since he knows the llm judge section better than I do, but very useful! |
clefourrier
approved these changes
Mar 31, 2025
alozowski
added a commit
that referenced
this pull request
Mar 31, 2025
This reverts commit e611a91.
hynky1999
pushed a commit
that referenced
this pull request
May 22, 2025
- Integrated Hugging Face Inference (hf-inference) as a new backend option for judge model evaluation. - Added async inference logic using AsyncInferenceClient only for hf-inference, with support for retries, timeouts, and provider specification. - Extended JudgeLM and JudgeLLM classes to support new configuration parameters like hf_provider, base_url, and max_tokens.
NathanHB
pushed a commit
that referenced
this pull request
Sep 19, 2025
- Integrated Hugging Face Inference (hf-inference) as a new backend option for judge model evaluation. - Added async inference logic using AsyncInferenceClient only for hf-inference, with support for retries, timeouts, and provider specification. - Extended JudgeLM and JudgeLLM classes to support new configuration parameters like hf_provider, base_url, and max_tokens.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 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.
hf-inference) as a new backend option for judge model evaluation.AsyncInferenceClientonly forhf-inference, with support for retries, timeouts, and provider specification.JudgeLMandJudgeLLMclasses to support new configuration parameters likehf_provider,base_url, andmax_tokens.