update config to train 5M steps

Project/Assets/ML-Agents/Examples/PushBlock/Scripts/PushBlockEnvController.cs

         m_ResetTimer += 1;
         if (m_ResetTimer > MaxEnvironmentSteps)
         {
-            m_ResetTimer = 0;
             ResetScene();
         }
     }
 
     void ResetScene()
     {
+        m_ResetTimer = 0;
+
         //Random platform rot
         var rotation = Random.Range(0, 4);
         var rotationAngle = rotation * 90f;

config/ppo/PushBlock.yaml

         gamma: 0.99
         strength: 1.0
     keep_checkpoints: 5
-    max_steps: 2000000
+    max_steps: 5000000 #2000000
     time_horizon: 64
     summary_freq: 60000
     threaded: true