archae0pteryx / deepsearch-tag-audio

Deepsearch tagging

Development

audio needs to be in wav format and sample rate matched with model use .env to point to the things

with docker

docker-compose build
yarn shell

locally:

yarn install
yarn setup # OSX: This installs deps, sample models, and audio. takes a while

Notes:

• https://deepspeech.readthedocs.io/en/r0.9/TRAINING.html
• https://ffmpeg.org/ffmpeg.html