first tests with new tree search observation. looks good :)

examples/training_navigation.py

@@ -55,9 +55,9 @@ scores = []
 dones_list = []
 action_prob = [0] * 4
 agent = Agent(state_size, action_size, "FC", 0)
-#agent.qnetwork_local.load_state_dict(torch.load('../flatland/baselines/Nets/avoid_checkpoint14900.pth'))
+agent.qnetwork_local.load_state_dict(torch.load('../flatland/baselines/Nets/avoid_checkpoint14900.pth'))
 
-demo = False
+demo = True
 
 
 def max_lt(seq, val):