Investigations - Open WebUI

Overview

…

Audio Configurations

(as of 2025/01/16 )

For STTモデル in the “Audio” in Admin Settings, which says

Open WebUI uses faster-whisper internally. Click here to learn more about faster-whisper and see the available models.

If you give wrong model, the logging shows an error e.g.

 raise ValueError(
ValueError: Invalid model size 'faster-distil-whisper-large-v3', expected one of: tiny.en, tiny, base.en, base, small.en, small, medium.en, medium, large-v1, large-v2, large-v3, large, distil-large-v2, distil-medium.en, distil-small.en, distil-large-v3

So they have

• tiny.en

• tiny

• base.en

• base

• small.en

• small

• medium.en

• medium

• large-v1

• large-v2

• large-v3

• large

• distil-large-v2

• distil-medium.en

• distil-small.en

• distil-large-v3

Logs

docker run -d -p 3000:8080 --add-host=host.docker.internal:host-gateway -v open-webui:/app/backend/data --name open-webui --restart always ghcr.io/open-webui/open-webui:main

from https://github.com/open-webui/open-webui

We may also specify the version of the image

$ docker rm --force open-webui
$ docker run --detach --publish 3000:8080 --volume open-webui:/app/backend/data --name open-webui ghcr.io/open-webui/open-webui:0.5.16

to avoid some issue on the specific version.

docker run --rm --volume /var/run/docker.sock:/var/run/docker.sock containrrr/watchtower --run-once open-webui

to update online.

References

• Open WebUI - https://github.com/open-webui/open-webui