Repository navigation
FIX: Replay content-filtered turns as text in the Responses and AzureML targets - #4160
Open
Utkarsh Bahuguna (u7k4rs6) wants to merge 1 commit into
Open
Utkarsh Bahuguna (u7k4rs6) wants to merge 1 commit into
Utkarsh Bahuguna (u7k4rs6) wants to merge 1 commit into
Conversation
This branch has not been deployed
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
A content-filtered response is stored as an error piece with response_error="blocked" and stays in history, but OpenAIResponseTarget and ChatMessageNormalizer (used by AzureMLChatTarget) rejected error pieces, so the next turn in that conversation always raised before reaching the provider. Now both replay a blocked piece as text, assistant output_text on Responses and a text part in the normalizer, which matches what OpenAIChatTarget already does. Processing and unknown errors are still rejected like before.
The new tests send a filtered turn and then a follow-up through PromptNormalizer on each target and check the second request carries the blocked turn as text, plus normalizer tests for single and multipart blocked pieces and that processing and unknown errors still raise. They fail on main and pass here.
Fixes #4155