My TG Channel - https://t.me/StefanFalkokAI
My TG Chat - https://t.me/+y4R5JybDZcFjMjFi
Hi! I introduce my working workflows with Flux.2 D generating image for ComfyUI.
I have included workflow with Flux.2 GGUF Model
You need to have Flux 2.D GGUF Model (https://huggingface.co/city96/FLUX.2-dev-gguf/tree/main), clip (https://huggingface.co/Comfy-Org/flux2-dev/tree/main/split_files/text_encoders) and vae (https://huggingface.co/Comfy-Org/flux2-dev/tree/main/split_files/vae)
If you need FP8 model - https://huggingface.co/Comfy-Org/flux2-dev/tree/main/split_files/diffusion_models
If you need GGUF Encoders version - https://huggingface.co/unsloth/Mistral-Small-3.2-24B-Instruct-2506-GGUF/tree/main
download my ComfyUI Build https://huggingface.co/datasets/StefanFalkok/ComfyUI_portable_torch_2.10.0_cu130_cp313_sageattention_triton/tree/main, also u need download and install CUDA 13.0 (https://developer.nvidia.com/cuda-13-0-0-download-archive) and VS Code (https://visualstudio.microsoft.com/downloads/)
Leave comments if you have trouble or you found the problem with workflows
Description
FAQ
Comments (6)
What should I do if I get the error "PathchSageAttentionKJ" on the installed version of Windows (not the portable version)? This applies to the version "sageattention-2.2.0+cu128torch2.8.0-cp39-cp39-win_amd64". It works fine on the "sageattention-1.0.6.dist-info" version.
good. Then leave sageattention 1.0.6
@Stefan_Falkok I jinxed it)) Right now, I have the generation running only once, and the error appears the second time - KSamplerAdvanced 'RMSNorm' object has no attribute 'weight'
@AlG80 https://huggingface.co/datasets/StefanFalkok/ComfyUI_portable_torch_2.9.1_cu130_cp313_sageattention_triton - i put fixed distorch_2.py file for multigpu. Replace to ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-MultiGPU - it will work
@Stefan_Falkok ок, thanks
@Stefan_Falkok use it, thanks