Create spectrograms from audio for using Cyanite without exposing your source audio files.
127
The container expects mp3 audio files as input. All data to be processed must be available within the container at /tmp/files. Therefore you must bind an external volume into the containers /tmp/files location.
The audio files to process should be passed as arguments during execution. You can pass as many arguments as you like. You may also pass in folders as arguments in which case the container will process all files within that folder recursively.
The resulting spectrogram files will be created adjacent to the input audios with the appendix <file-name>.cyanite2.npz.
Generate .cyanite2.npz files for each file and folder in the current working directory, recursively.
docker run -v "$PWD:/tmp/files" cyaniteai/preprocessor2:<tag> .
Generate .cyanite2.npz files for specific audio files relative to the current working directory.
docker run -v "$PWD:/tmp/files" cyaniteai/preprocessor2:<tag> audios/audio1.mp3 audios/audio2.mp3
Content type
Image
Digest
sha256:879d1a957…
Size
131.9 MB
Last updated
over 1 year ago
docker pull cyaniteai/preprocessor2:1.0.0-amd64