index_error about audio-classification HOT 5 CLOSED

Can you print the len(labels) after line 79 of train.py?

Also maybe show a couple examples of labels although I don't think it matters.

from audio-classification.

Do we have give specific path to this line
self.labels = labels
like we gave to this
self.wav_paths =('C:/Users/i00504285/Desktop/Aditya/Sound_Analysis/CSF48/wave_files/clean')

from audio-classification.

Can you print the len(labels) after line 79 of train.py?

Also maybe show a couple examples of labels although I don't think it matters.

Ya I did it shows 4 to me

https://drive.google.com/drive/folders/1tdUN92kf6oDC3-h2eCh_IZjVMMLdC-OA?usp=sharing

above link where i uploaded all my files.

Thank you.

from audio-classification.

The issue is, len(labels) needs to be >= batch_size. So after you run clean.py, you will need at least 16 files if using a batch_size of 16. Your data size is too small for this type of network. You can reduce batch_size to 4 but the model will never be able to generalize to new data.

from audio-classification.

Ok one quick question seth, If i will give 30 files in one class and as you know i have 2 classes so batch_size = 16 will work for me or any other number will work for me??

from audio-classification.