REINFORCE CartPole-v1
REINFORCE agent trained from scratch with PyTorch on CartPole-v1.
Environment
CartPole-v1
Algorithm
REINFORCE / Monte Carlo Policy Gradient
Evaluation
Mean reward: 500.00
Standard deviation: 0.00
Result: 500.00
Result = mean reward - standard deviation.
Training
- Hidden layer size: 64
- Training episodes: 1500
- Maximum steps per episode: 500
- Learning rate: 0.01
- Gamma: 1.0
Evaluation results
- mean_reward on CartPole-v1self-reported500.00 +/- 0.00