Objectives:
* Get familiar with the Self-attention layers in Pytorch.
* Design a Transformer encoder to process aduio signal to recognize the speaker.