Skip to content

Fixed a small bug in the Qwen3-VL chat template. - #37

Merged
JamePeng merged 1 commit into
JamePeng:mainfrom
TAO71-AI:main
Dec 26, 2025
Merged

Fixed a small bug in the Qwen3-VL chat template.#37
JamePeng merged 1 commit into
JamePeng:mainfrom
TAO71-AI:main

Conversation

@alcoftTAO

Copy link
Copy Markdown

Sometimes, Qwen3-VL models (specially the smallest models like Qwen3-VL-4B and Qwen3-VL-2B) generated the <response-for-user> example tag in the template when doing inference with tools.

Fixed replacing <response-for-user> with RESPONSE FOR THE USER HERE for better understanding for the model.

Models sometimes generate a <im_end> tag.

I've noticed that I had left a typo in the PR I made some time ago. This PR fixes that typo.

@JamePeng
JamePeng merged commit 79e4db1 into JamePeng:main Dec 26, 2025
16 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants