Changing worker-comfyui to get custom functionality
Hello, I am currently trying to set up a serverless api for my comfyui workflow. The thing is that it produces the videos, instead of images, though, I guess I cannot directly use the fast ComfyUI-to-API and cannot deploy the Docker via git, since i need custom "handler.py" logic. What would be the workaround for my case?
Should i clone the worker-comfyui change there Dockerfile by adding my custom_nodes, models and change a bit handler.py? And then just deploy this as github repo?
Would be nice to avoid too much of a setup.
There is a video, that does kind of what i expect my API to do, but i dont get what kind of folder do they submit during endpoint setup stage: https://www.youtube.com/watch?v=Q03YJEmBJvQ
Would be grateful for any help!
Should i clone the worker-comfyui change there Dockerfile by adding my custom_nodes, models and change a bit handler.py? And then just deploy this as github repo?
Would be nice to avoid too much of a setup.
There is a video, that does kind of what i expect my API to do, but i dont get what kind of folder do they submit during endpoint setup stage: https://www.youtube.com/watch?v=Q03YJEmBJvQ
Would be grateful for any help!
YouTubeAI Kitty
Learn how to generate videos from images using ComfyUI WAN2.2 and RunPod Serverless! All resources, code, and helper files are available for free — make sure to download them before following along: https://aikitty.cc/image-to-video-tutorial
For business inquiries or collaborations, contact me at contact@aikitty.cc
#Serverless #ComfyUI #Run...
For business inquiries or collaborations, contact me at contact@aikitty.cc
#Serverless #ComfyUI #Run...
