Policy Gradient (REINFORCE) Agent Playing CartPole-v1
This model is a custom PyTorch implementation of the REINFORCE (Monte Carlo Policy Gradient) algorithm, developed for the Hugging Face Deep Reinforcement Learning Course (Unit 4).
🚀 Model Details
- Environment: Gymnasium
CartPole-v1 - Algorithm: REINFORCE with Normalized Returns
- Mean Reward: 470.0 (Passing score: >= 350.0)
- Framework: Pure PyTorch
- Status: Officially Verified & Certified
Evaluation results
- mean_reward on CartPole-v1self-reported470.0 +/- 25.0