metadata
tags:
- FrozenLake-v1-4x4-no_slippery
- q-learning
- reinforcement-learning
- custom-implementation
model-index:
- name: q-FrozenLake-v1-4x4-noSlippery
results:
- metrics:
- type: mean_reward
value: 1.00 +/- 0.00
name: mean_reward
task:
type: reinforcement-learning
name: reinforcement-learning
dataset:
name: FrozenLake-v1-4x4-no_slippery
type: FrozenLake-v1-4x4-no_slippery
# 使用**Q-Learning**智能体来玩**FrozenLake-v1**
这是一个使用**Q-Learning**训练有素的模型玩**FrozenLake-v1**.
## 用法
```python
model = load_from_hub(repo_id='sun1638650145/q-FrozenLake-v1-4x4-noSlippery', filename='q-learning.pkl')
# 不要忘记检查是否需要添加额外的参数(例如is_slippery=False)
env = gym.make(model['env_id'])
evaluate_agent(env, model['max_steps'], model['n_eval_episodes'], model['qtable'], model['eval_seed'])
```