Pre-trained speech ASR/MT model and use-cases - V1

Summary
First version of pretrained speech ASRMT model and usecases