search

gabrielmittag / NISQA

NISQA - Non-Intrusive Speech Quality and TTS Naturalness Assessment
MIT License
680 stars 117 forks source link

pip package #38

Open christophschuhmann opened 1 year ago

christophschuhmann commented 1 year ago

Could you make a pip package? :)

sheepymeh commented 8 months ago

I've created one unofficially at https://pypi.org/project/nisqa/2.0/ and the files required to build it are in #47

anuragkumar95 commented 6 months ago

Is there a python package for NISQA. In short, what would be the best way if we want to use the NISQA model during runtime? I would like something like, 

wav = librosa.load(wav_path)
mos = NISQA(wav)
sheepymeh commented 6 months ago

Sorry for the wrong link I posted previously (there was a packaging error so the version number is now 2.0.post2). You can just install the nisqa package using pip. You can use the same functions as shown in the run_train.py and run_evaluate.py files.

anuragkumar95 commented 6 months ago

Is there any way in which NISQA can read a loaded wav file during runtime instead of reading .wav file from disk?