wav2vec-u Common Voice Swedish - prepare tsv

Original here

import soundfile
input = {
    'train': '/kaggle/input/download-common-voice-swedish/cv-corpus-6.1-2020-12-11/sv-SE/train.tsv',
    'test': '/kaggle/input/download-common-voice-swedish/cv-corpus-6.1-2020-12-11/sv-SE/test.tsv',
    'valid': '/kaggle/input/download-common-voice-swedish/cv-corpus-6.1-2020-12-11/sv-SE/dev.tsv'
}
for split in input.keys():
    with open(input[split], 'r') as tsv:
        with open(f'/kaggle/working/{split}.tsv', 'w') as out:
            print('/kaggle/input/common-voice-swedish-16bit-wav/', file=out)
            for line in tsv.readlines():
                data = line.split('\t')
                if data[1] == 'path':
                    continue
                file = data[1]
                file = file.replace('.mp3', '.wav')
                path = f'/kaggle/input/common-voice-swedish-16bit-wav/{file}'
                frames = soundfile.info(path).frames
                print("{}\t{}".format(file, frames), file=out)