Pre-trained speech ASR/MT model and use-cases - V2

Summary
Second version of pre-trained speech ASR/MT model and use-cases