ComfyUI
ComfyUI copied to clipboard
The most powerful and modular diffusion model GUI, api and backend with a graph/nodes interface.
Add support for loading Kandinsky models, which allow for multilanguage prompts and image fusion https://github.com/ai-forever/Kandinsky-2 https://huggingface.co/ai-forever/Kandinsky_2.1
Same feature as webui and would save some manual install actions
Inpaint mask content settings. While inpainting to fix small issues with color or location of an object, only being able to inpaint with latent noise makes it very hard to...
My PC configureation CPU: AMD Ryzen 5 3600 6-Core GPU: NVIDA GeForce GTX 1050 Ti SSD: 512G I lauched the webui using run_nvidia_gpu.bat but the nvidia card isn't working and...
Currently, in ComfyUI-Manager, automatic installation of dependencies is performed when installing custom nodes. However, this can lead to module conflicts or failure to install due to file locking issues, especially...
I haven't used Comfyui since the last day of the last month and I just use it again yesterday but I notice a huge difference in generation speed. When I...
the rendering speed is extremely slow. I'm checking to see if there are any errors causing this issue. It took approximately 2 hours to process 22 images with the 'ComfyUI'...
and got this error when trying to run main.py ./main.py ./main.py: line 1: import: command not found ./main.py: line 2: import: command not found ./main.py: line 3: import: command not...
Thank you so much for providing this excellent tool! During my last generation, the PC crashed and the queue did not finish. Every time now I start the run_nvidia_gpu.bat, the...
This PR allows ComfyUI to fulfill one of my last remaining uses for automatic1111's web UI -- easy integration with external tools. The existing websocket-based API is ideal for interactive...