regression fixes and tests for fix-5119-5121 branch#1
Open
rycerzes wants to merge 4 commits intoalbertvillanova:fix-5119-5121from
Open
regression fixes and tests for fix-5119-5121 branch#1rycerzes wants to merge 4 commits intoalbertvillanova:fix-5119-5121from
rycerzes wants to merge 4 commits intoalbertvillanova:fix-5119-5121from
Conversation
- add TestGRPORolloutDispatch tests: no extra fields, vLLM sync guard when step unchanged
- fix side effect where moving dispatch into _generate_single_turn caused _tool_call_loop to inadvertently route through rollout_func
0af8842 to
92848db
Compare
tests for conversational prompts pass through without apply_chat_template, plain string prompts pass through unchanged
92848db to
9db804a
Compare
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Pushed commits on top.
Main fix is server-mode chat routing,
generate()->chat()inopenenv/utils.pyso the vLLM server actually applies the template, withchat_template_kwargs,tools, andchat_templatethreaded through.Also added a
ValueErrorat init when bothrollout_funcandtoolsare set (tool-call loop was silently passing assembled histories to_generate_single_turn), clarified therollout_funcdocstring around the no-flattening contract, and new tests including a PIL Image through the full pipeline. This confirms fix for huggingface#5120