Bonsai_Tutorials/transcriber/README.md

transcriber

A python script which uses whisper to transcribe videos and outputs SRT subtitle text files.

• Github repository: https://github.com/dscoular/transcriber/
• Documentation https://dscoular.github.io/transcriber/

UV or Docker - that is the question.

Whether you choose to use uv or docker our Makefile will launch it to a series of interactive questions with sensible defaults. It will use the second smallest language model base.en by default, however I used the largest model medium.en for the SRT files I've commited to this repo. To be honest there wasn't much difference that I could discern.

Using python's uv

If you have uv in your PATH and ffmpeg installed in your operating system, you should be able to do the following to install and use this package.

1. make - runs the default target of make help to show help on all the make targets.
2. make install - installs the virtual environment and pre-commit hooks.
3. make check - optional, runs the code quality tools.
4. make test - optional, runs unit tests.
5. make docs-test - optional, generate HTML documents in the site directory.
6. make transcribe - by default, this converts the videos in the directory above our Makefile to .srt subtitle text files.

Using Docker

If you have docker installed on your system, you should be able to do the following to install and use this package.

1. make docker-build - Download and build the transcribe-app container image.
2. make docker-run - Run the transcriber interactively, by default, this converts the videos in the directory above our Makefile to .srt subtitle text files.