Datasetpublic

parambharat/kannada_asr_corpus

The corpus contains roughly 360 hours of audio and transcripts in Kannada language. The transcripts have beed de-duplicated using exact match deduplication.

sourceHugging Facecc-by-4.0updated 4y agoView on Hugging Face

0likes29downloads

root

dirdata/

file.gitattributes2 KBdownload

filekannada_asr_corpus.py6 KBdownload

fileREADME.md3 KBdownload

parambharat/kannada_asr_corpus · main · files are served by the source, never re-hosted here