audio2face-2d runpod.io setup

audio2face.md

audio2face on runpod.io

• audio2face-2d
• runpod.io

1. Create an nvidia enterprise account.
   https://build.nvidia.com/nvidia/audio2face-2d/docker?signin=true

2. Create an nvidia API key.
   Go to https://build.nvidia.com/nvidia/audio2face-2d/docker and press get API key.

3. Go to runpod.io and deploy a pod

4. Pick a GPU. I picked the cheapest one just for experimentation.

5. Pick a random template and edit it.
   We want to use the audio2face-2d docker image

6. Set the NGC_API_KEY environvent variable and enter your nvidia API key.

7. Deploy

8. The container will not run correctly yet. Go to your account and under "Container Registry Auth" add your key again.

This is to authenticate pulling the docker image.

9. Select this authorization for your pod.

And use it in the pod's settings.

10. If you get the following error you're not using an nvidia enterprise account or you didn't set the authtoken correctly. Please revisit those steps.

11. After pulling you will find your server's ip in the connect tab.

12. The end of https://build.nvidia.com/nvidia/audio2face-2d/docker?signin=true&snippet_tab=Docker describes how to interact with this container. Follow the steps.