Is your feature request related to a problem? Please describe.
We need to add a PyTorch worker that accepts the name of an PyTorch model file as an argument and compiles and evaluates it.

Describe the solution you'd like

• install pytorch and make the build succeed
• Get a small model () to run instead of the modified resnet50 of PtZendnn
• add a test for PyTorch worker
• add a new stage in the Dockerfile to build the PyTorch

Describe alternatives you've considered
NA

Additional context
NA