Adding sampling parameters for vllm generation #3210
+152
−6
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
|
@qgallouedec This PR is ready for review :) |
qgallouedec
reviewed
Apr 5, 2025
Comment on lines
+147
to
+149
| stop: list[str] = [], | ||
| stop_token_ids: list[int] = [], | ||
| bad_words: list[str] = [], |
There was a problem hiding this comment.
Using mutable default arguments can lead to unexpected behavior because the same list is shared across all function calls.
Maybe replace it with None, and in the function:
stop = stop or []
Comment on lines
+169
to
+170
| "help": "Minimum length of the prompt. If the prompt is shorter than this value, it will be truncated left." | ||
| }, |
There was a problem hiding this comment.
How truncation would solve it?
| repetition_penalty: int = field( | ||
| default=1, | ||
| metadata={ | ||
| "help": "List of text prompts for which the model will generate completions." |
There was a problem hiding this comment.
I think there is a mismatch here
There was a problem hiding this comment.
For parameters specific to vllm, I would name it with a vllm prefix. Ideally some are shared by transformers and vllm (like temperature), in such case, they don't need a prefix
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Outdated suggestions cannot be applied.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?
Fixes issue #3201 . Added support in GRPO config files to set additional parameters of vllm sampling.