llama.cpp

[Hermit6202]

When starting Comfyui i get this error: "Error loading module AILab_QwenVL_GGUF_PromptEnhancer: No module named 'llama_cpp'"
There is a folder llama.cpp in the Add-ons/Tools folder. How to use it?

About
ComfyUI 0.9.1
ComfyUI_frontend v1.38.4
Discord
ComfyOrg
ComfyUI-Ollama
LoRA Manager v0.9.11-stable
EasyUse v1.3.6
rgthree-comfy v1.0.2512112053
ComfyUI-Manager V3.39
System Info
OS
win32
Python Version
3.11.9 (tags/v3.11.9:de54cf5, Apr 2 2024, 10:12:12) [MSC v.1938 64 bit (AMD64)]
Embedded Python
true
Pytorch Version
2.9.1+cu130
Arguments
ComfyUI\main.py --windows-standalone-build --front-end-version Comfy-Org/ComfyUI_frontend@latest --fast fp16_accumulation --lowvram --reserve-vram 2 --use-sage-attention --disable-pinned-memory --disable-async-offload --output-directory E:\Comfyui_Output --input-directory E:\Comfyui_Input
RAM Total
31.93 GB
RAM Free
20.04 GB
Devices
Name
cuda:0 NVIDIA GeForce RTX 4060 : cudaMallocAsync
Type
cuda
VRAM Total
8 GB
VRAM Free
6.89 GB
Torch VRAM Total
32 MB
Torch VRAM Free
23.88 MB

[Tavris1]

Python 3.11 is no longer supported, so consider a fresh install. If only llama is the problem, this is the command you need to run from the python_embeded folder:

python.exe -m pip install https://github.com/JamePeng/llama-cpp-python/releases/download/v0.3.22-cu130-Basic-win-20260118/llama_cpp_python-0.3.22-cp311-cp311-win_amd64.whl

[Hermit6202]

Thanks. I installed the latest version, but I have to reinstall a lot of custom nodes. And they often don't work then fix and restart and fix and ...

[fu8765]

@Tavris1 could you include the installation of llama.cpp. Python ? Either as an optional, or bundled with QwenVL.....

[Tavris1]

ComfyUI-Easy-Install has llama-cpp built in. If for some reason you don't have it, try reinstalling ComfyUI-Easy-Install in a new folder, according to the instructions.