fix: preserve UTF-8 in SageMaker streaming decode - #13181
Open
pacocartones wants to merge 1 commit into
Open
Conversation
The SageMaker _streamComplete and _streamChat generators decoded each
streamed binary chunk with a fresh TextDecoder and no streaming flag.
A multibyte UTF-8 character split across two network chunks was decoded
as U+FFFD replacement characters, corrupting non-ASCII model output.
Reuse a single TextDecoder per stream and decode with { stream: true }
so partial multibyte sequences at chunk boundaries are buffered until
completed. Adds regression tests covering a split emoji and a split
two-byte character.
Contributor
|
All contributors have signed the CLA ✍️ ✅ |
Author
|
I have read the CLA Document and I hereby sign the CLA |
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
_streamCompleteand_streamChatincore/llm/llms/SageMaker.tsdecoded each streamed binary chunk with a freshnew TextDecoder().decode(chunk)and no streaming flag. A multibyte UTF-8 character split across two network chunks was emitted as U+FFFD replacement characters, corrupting any non-ASCII model output (emoji, accents, CJK). The chunk content and boundaries come from the remote SageMaker endpoint, so this triggers on ordinary response fragmentation.Reuse a single
TextDecoderper stream and decode with{ stream: true }, so partial sequences at chunk boundaries are buffered until complete.Added regression tests for a split emoji and a split two-byte character (vitest): they fail before the fix (replacement chars) and pass after.