There was an error in the video test.？

[zhanghongyong123456]

I'm unable to accurately determine which project this problem originated from. Could you provide some suggestions for fixing it? 1038lab/ComfyUI-QwenVL#101

[JamePeng]

Could you provide your runtime environment (hardware), project direction, and code? Currently, llama.cpp does not support qwen3vl video processing. With standard parameters, it handles around 8-9 images when inputting multiple images. Increasing the context and batch parameters could expand processing capacity.

Please wait for the official llama.cpp adaptation; I will adapt it as soon as possible.

[zhanghongyong123456]

Could you provide your runtime environment (hardware), project direction, and code? Currently, llama.cpp does not support qwen3vl video processing. With standard parameters, it handles around 8-9 images when inputting multiple images. Increasing the context and batch parameters could expand processing capacity.

Please wait for the official llama.cpp adaptation; I will adapt it as soon as possible.

I replaced the original workflow modules with the gguf module myself. This error occurred. If the module itself does not support video parsing, it indicates that it does not support it.
https://github.com/1038lab/ComfyUI-QwenVL/blob/main/example_workflows/QWenVL.jpg