Some questions about Serverless workers and custom workflows
Hi all, i'm very newbie, please help me with this questions.
1)How long will it take for a serverless worker to start with models around ~60 GB, and is it better to store the models on a network volume or bake them into the Docker container?
2) What is the simplest and fastest way to create my own serverless worker if I already have a ComfyUI workflow with custom nodes?
P.S. I've already spend some time for reading docs and even tried to make my own container from https://github.com/runpod-workers/worker-comfyui
Or may be should i use this somehow? https://hub.docker.com/r/timpietruskyblibla/runpod-worker-comfy
GitHub
GitHub - runpod-workers/worker-comfyui: ComfyUI as a serverless API...
ComfyUI as a serverless API on RunPod. Contribute to runpod-workers/worker-comfyui development by creating an account on GitHub.
Solution:Jump to solution
https://github.com/runpod-workers/worker-comfyui has docs on customization
GitHub
GitHub - runpod-workers/worker-comfyui: ComfyUI as a serverless API...
ComfyUI as a serverless API on RunPod. Contribute to runpod-workers/worker-comfyui development by creating an account on GitHub.
3 Replies
1) you gotta put them on a network volume, 60GB is too much for docker
2) I would start with the official runpod template and customize the Dockerfile. you need to install the dependencies for the custom nodes, i.e. run pip install -r requirements in each folder
Solution
https://github.com/runpod-workers/worker-comfyui has docs on customization
GitHub
GitHub - runpod-workers/worker-comfyui: ComfyUI as a serverless API...
ComfyUI as a serverless API on RunPod. Contribute to runpod-workers/worker-comfyui development by creating an account on GitHub.
thanks mate