update readme

README.md

@@ -21,15 +21,14 @@ pip install git+https://github.com/KohlerHECTOR/interpreter-py.git@v0.2.1
 ```
 
 ```python
-from interpreter import Interpreter
+ffrom interpreter import Interpreter
 from interpreter import ObliqueDTPolicy, SB3Policy, DTPolicy
 
 from stable_baselines3 import SAC
 from stable_baselines3.common.evaluation import evaluate_policy
 from stable_baselines3.common.monitor import Monitor
 
 import gymnasium as gym
-from gymnasium.wrappers.time_limit import TimeLimit
 from sklearn.tree import DecisionTreeRegressor
 from huggingface_sb3 import load_from_hub
 
@@ -60,21 +59,14 @@ interpret = Interpreter(oracle, tree_policy, env)
 interpret.fit(10)
 
 # Eval and save the best tree
-best_tree_policy, _ = interpret.get_best_tree_policy()
-final_tree_reward, _ = evaluate_policy(best_tree_policy, env=env, n_eval_episodes=10)
+final_tree_reward, _ = evaluate_policy(interpret._policy, env=env, n_eval_episodes=10)
 print(final_tree_reward)
 # Here you can replace pickle with joblib or cloudpickle
 with open("tree_halfcheetah.pkl", "wb") as f:
-    dump(best_tree_policy.clf, f, protocol=5)
+    dump(interpret._policy.clf, f)
 
 with open("tree_halfcheetah.pkl", "rb") as f:
     clf = load(f)
-# Render
-evaluate_policy(
-    ObliqueDTPolicy(clf, env),
-    env=Monitor(gym.make("HalfCheetah-v4", render_mode="human")),
-    render=True,
-)
 
 ```