Audio and transcript pairs for 50 languages. Each language is a Dataset Viewer configuration with train, validation, and test splits.
amharic: Amharic
arabic_msa: Arabic MSA
assamese: Assamese
bengali: Bengali
czech: Czech
dutch: Dutch
egyptian_arabic: Egyptian Arabic
english: English
farsi_persian: Farsi - Persian
filipino_tagalog: Filipino - Tagalog
french: French
german: German
greek: Greek
gujarati: Gujarati… See the full description on the dataset page:
https://huggingface.co/datasets/RidheshBhati/STT_MODEL.
