The winning model I used in MDX 2021 Challenge.
python setup.py install
usage: danna_sep [-h] [--outdir OUTDIR] [--fast] infile
positional arguments:
infile input audio file
optional arguments:
-h, --help show this help message and exit
--outdir OUTDIR output directory. Default to current working directory
--fast faster inference using only two of the models
Given an audio file, the program will split it into 4 stems, which are drums, bass, vocals and other, and store them in the given directory as .wav files.
When execute the first time it will download our pre-trained models (around 1 to 2 Gb) to the directory specified by the environment variable DANNA_CHECKPOINTS, which by default is ~/danna-sep-checkpoints.
This process is very time comsuming and require at least 16 Gb of RAM.
Please refer to our training repo.
Try Danna-Sep on Huggingface Spaces.
- convert to ONNX format
- Pack it as standalone app