This is a simple wrapper library for beat and down-beat tracking algorithm. Currently, this package only has madmom downbeat tracker The input audio format should be .wav. But you can use ffmpeg to convert .mp3 to .wav
docker run -ti --rm -v $(pwd):/beat_tracker python cli.py input/test.mp3
import tracker
args = tracker.TrackerArgument()
args.audio_path = 'input/slide.mp3'
args.output_dir_path = 'output/slide'
tracker.track(args)
ffmpeg -i input.wav -vn -ar 44100 -ac 2 -b:a 192k output.mp3
docker build -t beat_tracker:dev -f dev.dockerfile .