I downloaded the Common Voice Russian corpus to use this data to train my neural network. However, after unzipping, I received a file of an unknown type. I tried to convert it with bin/import_cv2.py, but doing everything according to the instructions I get errors with the sox and deepspeech_training packages.
Can you tell me if I can download already converted files from somewhere, or maybe there is a ready-made code on Colab?
Okay, I get the problem. For some unknown reason, instead of ru.tar.gz ru.tar is downloaded and everything breaks. If you add it .gz archive unpacks normally.
@ks-sav Could you please move that to the Common Voice section? https://discourse.mozilla.org/c/voice/239