ARS Agent Playing BipedalWalker-v3
This is a trained model of a ARS agent playing BipedalWalker-v3.
Hyperparameters
{'v': 0.03
'N': 16
'b': 8
'lr': 0.02
'normalizer': <__main__.Normalizer object at 0x0000017A496510F0>
'writer': <torch.utils.tensorboard.writer.SummaryWriter object at 0x0000017A49651720>
'env_id': 'BipedalWalker-v3'
'algorithm': 'ARS'
'iterations': 3000
'episode_length': 5000
'test_iterations': 100}
Evaluation results
- mean_reward on BipedalWalker-v3self-reported-97.25 +/- 36.07