faster-whisper

v3.8.1-ls10

faster-whisper is a reimplementation of OpenAI's Whisper model using CTranslate2, which is a fast inference engine for Transformer models. This container provides a Wyoming protocol server for faster-whisper.

docker pull lscr.io/linuxserver/faster-whisper:latest

Quick start

Generated from the image's own configuration. Replace the /path/to folders with real paths on your host.

---
services:
  faster-whisper:
    image: lscr.io/linuxserver/faster-whisper:latest
    container_name: faster-whisper
    environment:
      - PUID=1000
      - PGID=1000
      - TZ=Etc/UTC
      - DEBUG= # optional
      - LOCAL_ONLY= # optional
      - WHISPER_BEAM=1 # optional
      - WHISPER_LANG=auto # optional
      - WHISPER_MODEL=auto # optional
    volumes:
      - /path/to/faster-whisper/data:/config
    ports:
      - 10300:10300
    restart: unless-stopped

Parameters

Ports
PortFunctionRequired
10300:10300Wyoming connection port.required
Environment variables
VariableFunctionRequired
PUID=1000User IDrequired
PGID=1000Group IDrequired
TZ=Etc/UTCTimezonerequired
DEBUG=If set to `true`, or any other value, the container will output debug logs.optional
LOCAL_ONLY=If set to `true`, or any other value, the container will not attempt to download models from HuggingFace and will only use locally-provided models.optional
WHISPER_BEAM=1Number of candidates to consider simultaneously during transcription.optional
WHISPER_LANG=autoTwo character code for the language that you will speak to the add-on.optional
WHISPER_MODEL=autoWhisper model that will be used for transcription. From [here](https://github.com/SYSTRAN/faster-whisper/blob/master/faster_whisper/utils.py#L11-L31).optional
Volumes
PathFunctionRequired
/configLocal path for Whisper config files.required

Architectures

Pulling :latest picks the right one for your host automatically.

  • x86_64amd64-latest
  • arm64arm64v8-latest

Tags

  • :latestStable releases
  • :gpuReleases with Nvidia GPU support (amd64 only)
  • :gpu-legacyLegacy releases with Nvidia GPU support for pre-Turing cards (amd64 only)

Changelog

  1. Pin `av` to 18.x.

  2. Rebase to Ubuntu Resolute.

  3. Default to `auto` for model and language if not set.

Related

Similar images

Browse all

piper

2.5.2-ls134

piper is a fast, local neural text to speech system that sounds great and is optimized for the Raspberry Pi 4. This container provides a Wyoming protocol server for Piper.

212K pulls per month70 GitHub starsarm64