Repository navigation
FIX: Infer the objective from the user request when tool calls precede the answer - #3517
Open
Lanre Shittu (Shizoqua) wants to merge 1 commit into
Open
Lanre Shittu (Shizoqua) wants to merge 1 commit into
Lanre Shittu (Shizoqua) 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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Description
Fixes #3515.
extract_objective_from_previous_turnand its async version only read the turn atsequence - 1. When a target runs a tool loop, that turn is thefunction_call_output, so a scorer inferring its objective from the request got an empty string.Both functions now go through
_objective_from_request_turn, which picks the latest earlier turn that is user-authored or marked as an outbound request, the same rule #3410 added toMemoryInterface.get_request_from_response_async. Tool calls and outputs between the request and the answer are skipped, turns after the scored response are still never read, and only user text pieces from that turn are joined, as before.Tests and Documentation
Added
test_skips_tool_call_turns_between_request_and_answer_asynctotests/unit/score/test_message_scorer.py, covering the sync and async functions with a user request, afunction_call, afunction_call_outputand the final answer. The existing objective extraction tests are unchanged.No docs or notebooks changed, so JupyText was not needed.