Add repetition_penalty
and kwargs
parameters to SamplingParams
#174
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Outdated suggestions cannot be applied.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.
Summary:
This update introduces new parameters to the
SamplingParams
class, allowing for more flexible and customizable behavior when using vLLM and Hugging Face (HF) transformers based text generation deployments.Key Changes:
repetition_penalty
Parameter: Added a new parameter,repetition_penalty
that is supported by both vLLM and HF transformers.kwargs
Parameter: Introducedkwargs
to pass extra parameters that are not directly implemented inSamplingParams
. These will be forwarded to theSamplingParams
of vLLM and the generation parameters of HF transformers, enhancing flexibility.Closes: #171