parambharat/bengali_asr_corpus
The corpus contains roughly 500 hours of audio and transcripts in Bangla language. The transcripts have beed de-duplicated using exact match deduplication and audio has be converted to 16000 samples
036
The corpus contains roughly 500 hours of audio and transcripts in Bangla language. The transcripts have beed de-duplicated using exact match deduplication and audio has be converted to 16000 samples