Update test2.py

test2.py

@@ -394,7 +394,7 @@ def train_pong():
     
     # Hyperparameters
     gamma = 0.99
-    learning_rate = 0.01  # Lower learning rate for stability
+    learning_rate = 0.001  # Lower learning rate for stability
     num_episodes = 1000  # Pong requires more episodes
     
     # Action mapping: policy outputs 0 or 1, map to RIGHT(2) or LEFT(3)