llama.cpp icon indicating copy to clipboard operation
llama.cpp copied to clipboard

Published 20 hours ago • verified publisher icon ggml-org

qwen3-vl not have mmproj when using the convert_hf_to_gguf.py script

Open yuqikong opened this issue 2 weeks ago • 2 comments

yuqikong avatar Nov 14 '25 09:11 yuqikong

After using convert_hf_to_gguf.py on the Qwen3-VL model that I fine-tuned myself, I didn't get the mmproj file, making it impossible to use the visual features.

yuqikong avatar Nov 14 '25 09:11 yuqikong

You have to make the mmproj GGUF separately, using the --mmproj option.

CISC avatar Nov 14 '25 10:11 CISC