From 0539b0da1e3ba56c994e30f72437a01831aa739a Mon Sep 17 00:00:00 2001
From: AlexisBogroff <31806900+AlexisBogroff@users.noreply.github.com>
Date: Tue, 6 Nov 2018 05:43:25 +0100
Subject: [PATCH] Clean way to stop training

- Avoid using sys.exit()
- Use a boolean instead
---
 2-cartpole/1-dqn/cartpole_dqn.py | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/2-cartpole/1-dqn/cartpole_dqn.py b/2-cartpole/1-dqn/cartpole_dqn.py
index 8b2baaf0..1440c749 100644
--- a/2-cartpole/1-dqn/cartpole_dqn.py
+++ b/2-cartpole/1-dqn/cartpole_dqn.py
@@ -1,4 +1,3 @@
-import sys
 import gym
 import pylab
 import random
@@ -121,6 +120,7 @@ def train_model(self):
     agent = DQNAgent(state_size, action_size)
 
     scores, episodes = [], []
+    complete = False
 
     for e in range(EPISODES):
         done = False
@@ -128,6 +128,8 @@ def train_model(self):
         state = env.reset()
         state = np.reshape(state, [1, state_size])
 
+        if complete: break
+
         while not done:
             if agent.render:
                 env.render()
@@ -162,7 +164,7 @@ def train_model(self):
                 # if the mean of scores of last 10 episode is bigger than 490
                 # stop training
                 if np.mean(scores[-min(10, len(scores)):]) > 490:
-                    sys.exit()
+                    complete = True
 
         # save the model
         if e % 50 == 0: