This repository is used to create and push to `metricv/tl-whisper`.

# Usage
## Requirements
```
apt-get install ffmpeg # Do this however you want. GPU/nvenc not required, nor useful since we're only dealing with audio.
pip install -r requirements.pyr
```
## Append to dataset
```
python append_to_dataset.py AUDIO_FILE.m4a --entxt SUB.en.txt
# OR
python append_to_dataset.py AUDIO_FILE.m4a --ass SUB.ass
```
## Build and push
python ./push_self.py