openpecha/tibetan_voice

TibetanVoice: 6.5 hours of validated transcribed speech data from 9 audio book in lhasa dialect. The dataset is in tsv format with two columns, path and sentence. The path column contains the path to the audio file and the sentence column contains the corresponding sentence spoken in the audio file.