Ahmadzei's picture
added 3 more tables for large emb model
5fa1a76
ASR transcribes audio to text, so you will need a feature extractor and a tokenizer.