craggy dome
I'm very new to RunPod, and this is likely a relatively simple issue, but I can't figure out what to do based on the docs/experimentation. Chatgpt and Claude have been no help.
I see in the docs that Network Volumes are mounted to /Workspace. I thought that meant that the Network Volume would have its own folder inside Workspace. I made a Network Volume titled InvokeAI (in hindsight that was bound to cause confusion, poor naming on my part) and after launching a pod, I saw on Jupyter that the folder /Workspace/InvokeAI existed. So I moved some models over to that folder, and when launching future pods, the models were still there, so everything was fine.
But now I'm trying to create a second Network Volume for another purpose, this one titled FramePackVolume. But when I created a pod using a template for the FramePack program, it quickly hit disk storage limits and wouldn't run properly. There was plenty of Container Disk storage, but none of it is being used--the pod is using 83mb of Container Disk storage and all 30gb of its Network Volume storage. I've used this pod template without Network Volumes without issues. The idea this time was to launch the pod normally, relying on the Container Disk storage when downloading FramePack and all the models, then move the models over to the Network Volume to speed up the setup process in the future.
But everything just downloaded to the Network Volume, which I didn't want. Also, Workspace/FramePackVolume doesn't exist. There's a FramePack folder in the same root as Workspace, and another FramePack folder inside the Workspace folder.
Note that all I did was launch the template with the Network Volumes mounted, I haven't done anything else. Here's the template: https://github.com/ashleykleynhans/framepack-docker
I'm guessing that I'm completely misunderstanding how Network Volumes work?
All I'm trying to do is drastically speed up the time it takes to launch a FramePack pod. How should I go about this?
Thank you! :)
scenic palm