Add continuous control environments

[dtch1997]

The goal of this PR is to extend the RL-CBF method to work on continuous control environments.

Environments:

• Gymnasium mujoco locomotion envs

Milestones

• Try RL-CBF directly by discretizing the environment.
• Work out theory for continuous envs
• Implement CBF training & verification
• Evaluate RL-CBF vs vanilla RL on safety performance