Ahmadzei's picture
added 3 more tables for large emb model
5fa1a76
Tips:
All ASR models accept a float array corresponding to the raw waveform of the speech signal.