README.md

metadata

tags:
  - Pixelcopter-PLE-v0
  - reinforce
  - policy-gradient
  - reinforcement-learning
  - pixelcopter
  - custom-implementation
model-index:
  - name: Reinforce-Pixelcopter-PLE-v0
    results:
      - task:
          type: reinforcement-learning
          name: reinforcement-learning
        dataset:
          name: Pixelcopter-PLE-v0
          type: Pixelcopter-PLE-v0
        metrics:
          - type: mean_reward
            value: 48.95 +/- 42.79
            name: mean_reward
            verified: false

REINFORCE Agent on Pixelcopter-PLE-v0

This repository contains a REINFORCE (policy gradient) agent trained on Pixelcopter-PLE-v0.

Evaluation

• Mean reward: 48.95 ± 42.79
• Episodes: 20

Algorithm

• Monte Carlo Policy Gradient
• Stochastic policy
• PyTorch implementation