yasalma/tat_hackathon_asr

Hackathon Tatar ASR Dataset Summary Hackathon Tatar ASR is a speech dataset distributed during the "Татар.Бу Хакатон" (Tatar.Bu Hackathon) held in Tatarstan in May 2024. This dataset likely consists of newly collected crowdsourced recordings created after the last release of TatSC (Tatar Speech Corpus), although some intersections with TatSC might be present. While TatSC contains 269.1 hours of transcribed speech with 271,914 utterances, this hackathon dataset… See the full description on the dataset page: https://huggingface.co/datasets/yasalma/tat_hackathon_asr.