feat(rsl-rl): support DM Control Suite environments

[sylvesterkaczmarek]

Fixes #111

Summary

• add a baseline RSL-RL PPO config for DeepMind Control Suite environments
• reuse the existing DM Control Brax PPO learning rate, discount, entropy, and rollout-length settings where they map directly to RSL-RL
• route DM Control environments through train_rsl_rl.py
• add regression coverage for the config mapping and required RSL-RL runner fields

Notes

This provides a runnable baseline rather than claiming per-task RSL-RL tuning. The task-specific settings already present in the DM Control Brax PPO config are preserved where they have direct RSL-RL equivalents, while the remaining runner defaults follow the existing RSL-RL configs in this repository.

Testing

Added focused tests in mujoco_playground/_src/dm_control_suite/dm_control_suite_params_test.py.