Repository navigation
fix: preserve streamed provider thinking blocks - #4177
Merged
Merged
Conversation
Co-authored-by: abhay-codes07 <abhaysingh0293@gmail.com>
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
This pull request fixes thinking-block preservation for LiteLLM-backed Chat Completions streams.
Streaming responses now retain the complete ordered sequence of signed and redacted thinking blocks for client-managed replay. Reasoning items are finalized before their done events, thinking-only streams produce replayable items, opaque provider metadata survives round trips, and assistant message content indexes remain consistent when reasoning occupies a separate output slot.
A runtime probe using LiteLLM 1.83.0 verified that multiple signed thinking blocks and an interleaved redacted block survive stream transformation, SDK event serialization, and outbound converter replay without losing their order, signatures, or provider metadata.
The implementation also preserves the released normalized reasoning text and final-signature behavior while rejecting malformed interpreted thinking fields before emitting an item.