Fix: flush partial stop string when <EOG> is reached in /completion endpoint in streaming mode #15007
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
matteoserva
changed the title
matteoserva
force-pushed
the
fix_stop_bug
branch
from
1c4809a to
a39eac1
Compare
matteoserva
force-pushed
the
fix_stop_bug
branch
from
a39eac1 to
d3dd9e0
Compare
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Outdated suggestions cannot be applied.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.
In the
/completionendpoint the server correctly holds the partial stop string until a decision can be made.If EOG is reached and a stop string was not found, then the generated content should be flushed to the user.
To reproduce:
query:
Repeat exactly the following text: UNO DUE TREstop:
["TRE\nAAAAAAAA"]stream:
TruePrevious result:
UNO DUEAfter this PR:
UNO DUE TRENote:
The openAI compatible api has the opposite problem, the partial stop string is always flushed even if the full stop string is found.
This is not addressed in this PR