hirocaster / whisper-docker

whisper with docker

whisper-docker

whisper with docker

Setup

Install Docker.

$ git clone https://github.com/karaage0703/whisper-docker
$ cd whisper-docker
$ docker build -t whisper .

Usage

Voice recognition with microphone

Execute following command in whisper-docker directory.

$ docker run -it -d -v $(pwd):/workspace/ --net host --name whisper whisper
$ docker exec -it whisper bash
root@hostname:/workspace# python whisper-server.py

Open new terminal and execute following command:

$ python mic.py

Transcribe

Prepare audio file (ex: input.mp3) and execute following command in whisper-docker directory.

$ docker run -it -d -v $(pwd):/workspace/ --net host --name whisper whisper
$ docker exec -it whisper bash
root@hostname:/workspace# python transcribe.py --model='base' --input_file='input.mp3' --output_file='output.txt' --language='ja'

Reference

• https://zenn.dev/kento1109/articles/d7d8f512802935