V1.1 BC state and visuomotor policies with various configs for kitchen task

Attached the pytorch models for BC policies under various configs of state based and visuomotor based (normal, fast and ultrafast with tradeoffs on accuracy and speed)

V1.0 BC state and visuomotor policy for kitchen task

Added the BC state and visuomotor policy models for kitchen task - picking up a tomota can from the fridge and placing inside the microwave cavity.

-- This release contains the code and imitation learning pytorch models which was showcased in the demo presentation on Sept 5th.