Add training chat template for Qwen3-2507#5574
Merged
Merged
Conversation
Member
|
Can you add tests using |
There was a problem hiding this comment.
Cursor Bugbot has reviewed your changes and found 1 potential issue.
❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, enable autofix in the Cursor dashboard.
Reviewed by Cursor Bugbot for commit 7b6bdb5. Configure here.
|
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. |
qgallouedec
approved these changes
Apr 17, 2026
qgallouedec
left a comment
Member
There was a problem hiding this comment.
LGTM, just waiting for #5586 to be merged
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Models like
Qwen3-4B-Instruct-2507ship a chat template that is simpler than the standard Qwen3 template, it has noreasoning_contenthandling, no<think>block parsing, nomulti_step_tooltracking, and noenable_thinkingflag. Becauseget_training_chat_templateuses exact string matching, this template variant falls through and raises:This PR adds the
qwen3_instruct_2507base and training templates (with{% generation %}/{% endgeneration %}markers) and wires them intoget_training_chat_templateandadd_response_schema.Follows up on #5470.
cc: @qgallouedec
Note
Low Risk
Low risk: adds support for an additional Qwen3 template variant and updates exact-match routing/tests, with changes largely isolated to template selection and response parsing for that model family.
Overview
Adds first-class support for the
Qwen3-*-Instruct-2507chat-template variant by introducing new base and training Jinja templates, including{% generation %}/{% endgeneration %}markers forassistant_only_lossmasking.Wires this variant into
get_training_chat_templateandadd_response_schemavia exact template matching, expands the docs to list the new supported family/template, and extends the chat-template utility tests to cover the new tokenizer (including skippingreasoning_contentparsing where unsupported).Reviewed by Cursor Bugbot for commit 8cdbd28. Bugbot is set up for automated code reviews on this repo. Configure here.