Merge pull request #61 from k223kim/k223kim/fix-training-sumo-rl-env

fix: training loop example for sumo rl

Author: pankit-eng

examples/sumo_rl_simple.py

@@ -56,7 +56,7 @@ def main():
             action_id = np.random.choice(result.observation.action_mask)
 
             # Take action
-            result = env.step(SumoAction(phase_id=action_id))
+            result = env.step(SumoAction(phase_id=int(action_id)))
 
             episode_reward += result.reward or 0
             steps += 1

src/envs/sumo_rl_env/README.md

@@ -220,7 +220,7 @@ for episode in range(10):
         action_id = np.random.choice(result.observation.action_mask)
 
         # Take action
-        result = env.step(SumoAction(phase_id=action_id))
+        result = env.step(SumoAction(phase_id=int(action_id)))
 
         episode_reward += result.reward or 0
         steps += 1