Termination is used to stop an episode when the robot enters an invalid or unhelpful state, so the policy does not keep learning from messy failure motions. Here we use two termination conditions:
- fell over termination
if the robot base tilts more than 70 degrees, the robot is considered fallen or highly unstable, so the episode is stopped.
- time out termination
Time-out termination means if the robot survives until the maximum episode length (normally defined as 1000 sim steps, 20s) , the episode ends normally.