#42 run baselines in ci

torch_training/bla.py

@@ -148,15 +148,15 @@ def main(argv):
                     env_renderer.gl.saveImage("./Images/flatland_frame_{:04d}.bmp".format(frame_step))
                     frame_step += 1
             # print(step)
-        #     # Action
-        #     for a in range(env.get_num_agents()):
-        #         if demo:
-        #             eps = 0
-        #         # action = agent.act(np.array(obs[a]), eps=eps)
-        #         action = agent.act(agent_obs[a], eps=eps)
-        #         action_prob[action] += 1
-        #         action_dict.update({a: action})
-        #     # Environment step
+            # Action
+            for a in range(env.get_num_agents()):
+                if demo:
+                    eps = 0
+                # action = agent.act(np.array(obs[a]), eps=eps)
+                action = agent.act(agent_obs[a], eps=eps)
+                action_prob[action] += 1
+                action_dict.update({a: action})
+            # Environment step
         #
         #     next_obs, all_rewards, done, _ = env.step(action_dict)
         #     # print(all_rewards,action)