• Clip-l

• for GGUF t5 models

• OFFICIAL gguf models

• advance gguf workflow

• MY UNIQUE EDITION model

• MY MERGE

*** remember the gguf format not depend that much on GPU so the Q8 is the most closed for fp16 and its the best choice for PC users

STEPS:

• "download workflow > install missing nodes through manager > run it"

• manually

1. Clone the Repository

Clone the custom node repository to your ComfyUI custom nodes folder:

   git clone https://github.com/city96/ComfyUI-GGUF

2. Install Dependencies

Install the necessary dependency for inference:

   pip install --upgrade gguf

if you have this error during loading gguf loader

 newbyteorder was removed from the ndarray class in NumPy 2.0.

try:

pip install numpy==1.26.4

• article help you to choose which model to download

https://civarchive.com/articles/6730/flux-gguf