The Bengali.AI Split of the dataset can be found here: https://huggingface.co/datasets/bengaliAI/CommonVoiceBangla The Mozilla CV split and the Audio files can be obtained directly from the CommonVoice website: https://commonvoice.mozilla.org/en/datasets