README.md

metadata

tags:
  - PandaReach-v3
  - ddpg
  - reinforcement-learning
  - custom-implementation
model-index:
  - name: ddpg-PandaReach-v3
    results:
      - task:
          type: reinforcement-learning
          name: reinforcement-learning
        dataset:
          name: PandaReach-v3
          type: PandaReach-v3
        metrics:
          - type: mean_reward
            value: '-1.68 +/- 0.81'
            name: mean_reward
            verified: false

DDPG Agent playing PandaReach-v3

This is a trained model of a DDPG agent playing PandaReach-v3.

Hyperparameters:

hyperparameters = {
  "env_id": "PandaReach-v3",
  "max_steps": 50000,
  "n_training_episodes": 9624,
  "n_eval_episodes": 3000,
  "learning_rate": 0.001,
}