If your VRAM is insufficient for Flux, you need to run a quantized version. This is a really simple workflow with LoRA load and upscale. Keep in mind that the quantized versions need slightly higher strength values than the normal ones.
This workflow is based on the GGUF model loader in ComfyUI:
https://github.com/city96/ComfyUI-GGUF.
Update:
Added upgraded "Simple" version. It will requre 2 Custom Nodes to be installed. What is differrent in it:
Added Multi-LoRA suppor with the rgthree LoRA stacker. This is the best pick for low end video cards I've been able to find.
Added Civit-AI friendly file saver with the requred supporting nodes.
Orginised everything in groups a little bit.
It is still reqlly easy to use and now it is a good starting point for more complex Workflows as the generation info will be saved for Civit even if you do more complex operations.
Description
FAQ
Comments (12)
what upscale model I should use?
I still use Ultra Sharp 4x .
thank you sir!
Why don't you put the Dual Clip Loader GGUF version too?
2 Reasons for not using GGUF Q versions of T5
1. It gets offloaded from the VRAM , so the overall delay in generation is not that big
2. T5 GGUFs experience dramatic decrease in quality below Q8 .
Every LoRA I tried gave a massive amount of error messages in the cli and they did not work.
great workflow, but my civitai image saver is alwais red frame?why is that? what do i do to find it?
I had to right click and node and select Reload Node then it started working
Good workflow. However, I get faster generation results on laptop RTX 4060 8gb using flux_dev-fp8.safetensors rather than GGUF 4.0 format. 156 sec vs 186 sec.
The smaller format is not faster possibly because of additional overhead for unpacking/conversion.
great workflow, but image saver doesnt work for me. can anybody help?
It is a little bit outdated at this point. Must upload what I am currently using , but you can also find it in all of the Flux Images I upload.
This isn't simple at all, especially not the one called "simple". Better change the name to whatever. Where can I find a basic workflow for flux GGUF?