jordipons / musicnn

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

jordipons/musicnn Issues

Instruction how to install the package to solve dependency issues
Updated 2 months ago
Convert model to TensorFlow Lite
Closed 3 months ago2
Update the project to current numpy and librosa
Updated 7 months ago1
ValueError: Model not implemented! (Musicnn_big)
Updated 9 months ago
Incorrect features for MusicCNN models (pypi version)
Updated a year ago1
Can't I use `musicnn` on Windows10?
Closed a year ago4
UnboundLocalError: local variable 'batch' referenced before assignment
Updated 3 years ago2
Extractor value error from call as per README
Updated 3 years ago
COLA pretrain
Updated 4 years ago1
Inference with SavedModel format
Updated 4 years ago
how to use musicnn extractor in multiprocess?
Updated 4 years ago
musicnn-training model cannot used in this repo
Updated 4 years ago
Didn't find code for training vgg and musicnn from scratch
Updated 4 years ago
Running musicnn_test.ipynb throws exception
Updated 4 years ago
[QUESTION] About models
Closed 4 years ago2
Input from memory for tagger
Updated 4 years ago
Possible mistake in feature extraction?
Updated 4 years ago3
train/valid/test split for MagnaTagATune
Closed 4 years ago1
Beat detection
Closed 4 years ago1
Overlap is used as step size
Updated 4 years ago
dancing types, e.g. Walz, Swing,
Closed 5 years ago1