- download
LibriSpeech corpus: https://www.openslr.org/12/
WHAM! noise: https://wham.whisper.ai/
- environment
pip install -r requirements.txt
pip install -r prepare/requirements.txt
- prepare LibriMix & LibriMix-informed
cd prepare/librimix && bash gen_tiny.sh
cd prepare/librimix-informed && bash prepare_data.sh /home/zhao/data/librimix/Libri2Mix /home/zhao/code/spex+ /home/zhao/data/librimix-informed
- run
bash train.sh