librosa 0.8 does not support output function so need to switch to soundfile package for writing audio file

meokz / looking-to-listen

Deep neural network (DNN) for noise reduction, removal of background music, and speech separation

MIT License

169 stars 19 forks source link

librosa 0.8 does not support output function so need to switch to soundfile package for writing audio file #15

Closed saumyaborwankar closed 3 years ago

saumyaborwankar commented 3 years ago

File "/home/saumya/looking-to-listen/network/src/common/util.py", line 81, in istft_and_save librosa.output.write_wav(path=filepath, y=signal, sr=settings.SR, norm=True) AttributeError: module 'librosa' has no attribute 'output'

saumyaborwankar commented 3 years ago

I have written the code to remove this error but im not sure how to make a pull request. Im quite new to github Maybe you can include it in the source code so the package will be compatible with librosa 0.8

meokz commented 3 years ago

We fixed the version of librosa at 0.6.3 You can see requirements.txt in https://github.com/meokz/looking-to-listen/blob/master/preprocess/requirements.txt.

Versions different from requirements.txt are not guaranteed.