STOP behavior reward and learning masking.

[riccardosavorgnan]

This PR updates the code so that we don't accumulate returns nor update the policy from steps where the car is STOPPED (STOP behavior from either collission or offroad).

High level:

• we add a flag for the stopped condition passed to the python hierarchy
• we set rewards to 0 when the flag is true
• we set losses (grad) to 0 when the flag is true to prevent updating the policy in those instances.

NOTE: we do not mask the entropy bonus at the moment.