support qwen3-vl handle requests with embeddings (#30037)

Signed-off-by: taoyun <1069423820@qq.com> Signed-off-by: Cyrus Leung <cyrus.tl.leung@gmail.com> Co-authored-by: Cyrus Leung <cyrus.tl.leung@gmail.com>
2025-12-05 01:34:06 +08:00
parent 46cbbca05c
commit 6dcb07f676
2 changed files with 7 additions and 2 deletions
--- a/docs/features/multimodal_inputs.md
+++ b/docs/features/multimodal_inputs.md
@@ -443,6 +443,8 @@ For Qwen2-VL and MiniCPM-V, we accept additional parameters alongside the embedd
        print(generated_text)
    ```

+For Qwen3-VL, the `image_embeds` should contain both the base image embedding and deepstack features.
+
 #### Audio Embeddings

 You can pass pre-computed audio embeddings similar to image embeddings: