herurg/PPO-LunarLander-v2-Unit8

PPO Agent Playing LunarLander-v2

This is a PPO agent trained for Hugging Face Deep RL Course Unit 8 Part I.

The PPO implementation follows the CleanRL-style architecture and was adapted to Gymnasium.

Evaluation

• Mean reward: -106.451102
• Standard deviation: 72.517478
• Lower-bound score: -178.968580
• Evaluation episodes: 10

Training

• Environment: LunarLander-v2
• Algorithm: PPO
• Training timesteps: 51,200
• Framework: PyTorch
• Environment API: Gymnasium