Skip to main content

Pronounced as "musician", musicnn is a set of pre-trained deep convolutional neural networks for music audio tagging

Project description


Pronounced as "musician", musicnn is a set of pre-trained musically motivated convolutional neural networks for music audio tagging. This repository also includes some pre-trained vgg-like baselines.

Check the documentation and our basic / advanced examples to understand how to use musicnn.

Do you have questions? Check the FAQs.


pip install musicnn

or, to get bigger models and all the documentation (including jupyter notebooks), install from source:

git clone

python install

Predict tags

From within python, you can estimate the topN tags:

from musicnn.tagger import top_tags
top_tags('./audio/joram-moments_of_clarity-08-solipsism-59-88.mp3', model='MTT_musicnn', topN=10)

['techno', 'electronic', 'synth', 'fast', 'beat', 'drums', 'no vocals', 'no vocal', 'dance', 'ambient']

Let's try another song!


['guitar', 'piano', 'fast']

From the command-line, you can also print the topN tags on the screen:

python -m musicnn.tagger file_name.ogg --print
python -m musicnn.tagger --model 'MSD_musicnn' --topN 3 --length 3 --overlap 1.5 --print

or save to a file:

python -m musicnn.tagger file_name.wav --save out.tags
python -m musicnn.tagger file_name.mp3 --model 'MTT_musicnn' --topN 10 --length 3 --overlap 1 --print --save out.tags

Extract the Taggram

You can also compute the taggram using python (see our basic example for more details on how to depict it):

from musicnn.extractor import extractor
taggram, tags = extractor('./audio/joram-moments_of_clarity-08-solipsism-59-88.mp3', model='MTT_musicnn')


The above analyzed music clips are included in the ./audio/ folder of this repository.

You can listen to those and evaluate musicnn yourself!

Project details

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Files for musicnn, version 0.1.0
Filename, size File type Python version Upload date Hashes
Filename, size musicnn-0.1.0-py3-none-any.whl (29.3 MB) File type Wheel Python version py3 Upload date Hashes View
Filename, size musicnn-0.1.0.tar.gz (29.3 MB) File type Source Python version None Upload date Hashes View

Supported by

Pingdom Pingdom Monitoring Google Google Object Storage and Download Analytics Sentry Sentry Error logging AWS AWS Cloud computing DataDog DataDog Monitoring Fastly Fastly CDN DigiCert DigiCert EV certificate StatusPage StatusPage Status page