[Litellm Enhancement] Enable extra sampling args for litellm backend#1195
Open
eldarkurtic wants to merge 1 commit intohuggingface:mainfrom
Open
[Litellm Enhancement] Enable extra sampling args for litellm backend#1195eldarkurtic wants to merge 1 commit intohuggingface:mainfrom
eldarkurtic wants to merge 1 commit intohuggingface:mainfrom
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
When models are served with vllm, vllm's OpenAI endpoint supports some additional sampling args like
top_k,min_p, andpresence_penalty. This PR enables forwarding these as standardgeneration_parameterswithout breaking litellm usage for other endpoints like openai, togetherai, etc which don't support these extra args.This PR is a bit less invasive approach than #1194
Motivation for these extra sampling args: newer models like Qwen3.5 suggest using
top_k/presence_penaltyas sampling args for the "best behavior" of their models. Without this PR, it was not possible to evaluate vllm-hosted Qwen3.5 models with recommended sampling args.