Add test and docs for multimodal tool responses#5448
Merged
Conversation
|
The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update. |
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
3 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Add test
test_training_with_tools_multimodal_responsecovering tools that return images introduced in Support multimodal tool responses inenvironment_factoryfor VLM training #5323Add "Multimodal Tool Responses" subsection to the Agent Training docs with a code example
Note
Low Risk
Low risk: changes are limited to documentation and an additional unit test covering image-returning tool outputs; no production training logic is modified.
Overview
Adds documentation for multimodal tool outputs in GRPO agent training, showing that tools can return a list of
{type: image/text}blocks and that images are injected into subsequent VLM turns.Adds a new vision-gated GRPO trainer test (
test_training_with_tools_multimodal_response) that mocks generation to trigger tool calls and verifies multimodal tool responses (PIL image + text) train successfully and log expected tool call/failure frequencies.Reviewed by Cursor Bugbot for commit c40f4f5. Bugbot is set up for automated code reviews on this repo. Configure here.