This dataset was originally created for the purposes of a bachelors thesis in linguistics. The dataset consists of over 100 transcriptions made by Whisper V3 (model turbo) and Microsoft's Dictation service. The transcriptions were made based on sound files found at https://swedia.ling.gu.se/, and 14 different Swedish dialects, where each dialect has four speakers each
