-
-
Notifications
You must be signed in to change notification settings - Fork 5.3k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[Frontend] Add logits_processors
as an extra completion argument
#11150
Conversation
…ment Signed-off-by: Brad Hilton <[email protected]>
👋 Hi! Thank you for contributing to the vLLM project. Once the PR is approved and ready to go, your PR reviewer(s) can run CI to test the changes comprehensively before merging. To run CI, PR reviewers can do one of these:
🚀 |
Signed-off-by: Brad Hilton <[email protected]>
Hmm, this seems a bit dangerous as it allows the client to execute any Python module server-side. A better solution would be to introduce a registry of logits processors which the client can reference by their key in the registry. |
Signed-off-by: Brad Hilton <[email protected]>
Signed-off-by: Brad Hilton <[email protected]>
Great feedback, thank you @DarkLight1337. I added an optional |
Signed-off-by: Brad Hilton <[email protected]>
Signed-off-by: Brad Hilton <[email protected]>
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Using a pattern to restrict access also works, thanks for adding this!
@DarkLight1337 |
Let me retry the test. |
Thanks @DarkLight1337 ! |
…llm-project#11150) Signed-off-by: Brad Hilton <[email protected]>
…llm-project#11150) Signed-off-by: Brad Hilton <[email protected]>
Adds
logits_processors
as an extra (optional) argument toChatCompletionRequest
andCompletionRequest
:This enables passing built-in or custom logits processors to direct the sampling process.