automl/Mighty

AI Summary: Add a configurable log path argument to the Mighty RL library's meta modules, allowing them to independently manage their logging output. This will improve modularity and flexibility.

AI Summary: The task involves enhancing the PPO and SAC models within the Mighty reinforcement learning library to accept keyword arguments for customizing the feature extractor and prediction head. This requires adding optional keyword argument dictionaries, updating configuration files with default values, and refactoring model initialization to utilize these dictionaries.