matatonic / openedai-speech

An OpenAI API compatible text to speech server using Coqui AI's xtts_v2 and/or piper tts as the backend.
GNU Affero General Public License v3.0
193 stars 32 forks source link

Error response from daemon: unknown or invalid runtime name: nvidia #1

Closed alexonpeace closed 2 months ago

alexonpeace commented 2 months ago
docker compose up
[+] Running 0/0
 - Container openedai-speech-server-1  Creating                                                                                                                                                             0.0s
Error response from daemon: unknown or invalid runtime name: nvidia
alexonpeace commented 2 months ago

I'm trying to run this on docker desktop

alexonpeace commented 2 months ago

It works without the

runtime: nvidia
deploy:
  resources:
    reservations:
      devices:
        - driver: nvidia
          #device_ids: ['0', '1'] # Select a gpu, or
          count: all
          capabilities: [gpu]

but how do I make it run with GPU

matatonic commented 2 months ago

I think all you need to do is setup wsl2 and enable wsl2 in docker desktop and run it from there. As I understand it, in windows the 'runtime' and 'deploy' sections are not required after you do this. Here is a video that should detail the steps: https://www.youtube.com/watch?v=PB7zM3JrgkI

jmtatsch commented 2 months ago

the runtime: nvidia throws the same error for me as well and I have a bunch of nvidia accelerated containers running just fine.

as per https://stackoverflow.com/questions/52865988/nvidia-docker-unknown-runtime-specified-nvidia

runtime: nvidia is just for nvidia-docker2 which is outdated.

matatonic commented 2 months ago

@alexonpeace - can you confirm this? Looks like the deploy section is still required to enable gpus even with an nvidia compatible runtime

alexonpeace commented 1 month ago

it works 👍