-
Notifications
You must be signed in to change notification settings - Fork 0
/
cartpole.py
205 lines (180 loc) · 7.24 KB
/
cartpole.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
"""
Created on Mon May 8 11:26:55 2017
cartpole control using RL
@author: minty
"""
import argparse
import sys
import gym
import tensorflow as tf
import numpy as np
import random
from collections import deque
# Hyper Parameters for DQN
GAMMA = 0.9 # discount factor for target Q
INITIAL_EPSILON = 0.5 # starting value of epsilon
FINAL_EPSILON = 0.01 # final value of epsilon
REPLAY_SIZE = 1000 # experience replay buffer size
BATCH_SIZE = 32 # size of minibatch
class DQN():
# DQN Agent
def __init__(self, env):
# init experience replay
self.replay_buffer = deque()
# init some parameters
self.time_step = 0
self.epsilon = INITIAL_EPSILON
self.state_dim = env.observation_space.shape[0]
self.action_dim = env.action_space.n
self.create_Q_network()
self.create_training_method()
# Init session
self.session = tf.InteractiveSession()
self.session.run(tf.global_variables_initializer())
self.merged=tf.summary.merge_all()
self.train_writer=tf.summary.FileWriter(FLAGS.log_dir+'/train',self.session.graph)
def create_Q_network(self):
# network weights
with tf.name_scope('weights1'):
W1 = self.weight_variable([self.state_dim,20])
tf.summary.histogram('histogram',W1)
with tf.name_scope('bias1'):
b1 = self.bias_variable([20])
tf.summary.histogram('histogram',b1)
with tf.name_scope('weights2'):
W2 = self.weight_variable([20,self.action_dim])
tf.summary.histogram('histogram',W2)
with tf.name_scope('bias2'):
b2 = self.bias_variable([self.action_dim])
tf.summary.histogram('histogram',b2)
# input layer
self.state_input = tf.placeholder("float",[None,self.state_dim])
# hidden layers
h_layer = tf.nn.relu(tf.matmul(self.state_input,W1) + b1)
self.Q_value = tf.matmul(h_layer,W2) + b2
tf.summary.histogram('Q_value',self.Q_value)
def create_training_method(self):
with tf.name_scope('input'):
self.action_input = tf.placeholder("float",[None,self.action_dim]) # one hot presentation
self.y_input = tf.placeholder("float",[None])#target Q value
# self.action_input=action_batch,self.Q_value=[float,float],
Q_action = tf.reduce_sum(tf.multiply(self.Q_value,self.action_input),reduction_indices = 1)
self.cost = tf.reduce_mean(tf.square(self.y_input - Q_action))
tf.summary.scalar('cost function',self.cost)
self.optimizer = tf.train.AdamOptimizer(0.0001).minimize(self.cost)
def perceive(self,state,action,reward,next_state,done):
one_hot_action = np.zeros(self.action_dim)
one_hot_action[action] = 1
self.replay_buffer.append((state,one_hot_action,reward,next_state,done))
if len(self.replay_buffer) > REPLAY_SIZE:
self.replay_buffer.popleft()
if len(self.replay_buffer) > BATCH_SIZE:
self.train_Q_network()
def train_Q_network(self):
self.time_step += 1
# Step 1: obtain random minibatch from replay memory
minibatch = random.sample(self.replay_buffer,BATCH_SIZE)
state_batch = [data[0] for data in minibatch]
action_batch = [data[1] for data in minibatch]
reward_batch = [data[2] for data in minibatch]
next_state_batch = [data[3] for data in minibatch]
# Step 2: calculate y
y_batch = []
Q_value_batch = self.Q_value.eval(feed_dict={self.state_input:next_state_batch})
for i in range(0,BATCH_SIZE):
done = minibatch[i][4]
if done:
y_batch.append(reward_batch[i])
else :
y_batch.append(reward_batch[i] + GAMMA * np.max(Q_value_batch[i]))
"""if self.time_step%100 == 99:
run_options = tf.RunOptions(trace_level=tf.RunOptions.FULL_TRACE)
run_metadata = tf.RunMetadata()
summary, _ = self.session.run([self.merged, self.optimizer],
feed_dict = {self.state_input:state_batch,
self.y_input:y_batch,
self.action_input:action_batch},
options=run_options,
run_metadata=run_metadata)
self.train_writer.add_run_metadata(run_metadata, 'step%03d' % self.time_step)
self.train_writer.add_summary(summary, self.time_step)
print('Adding run metadata for', self.time_step)"""
summary, _, cost_value = self.session.run([self.merged, self.optimizer, self.cost], feed_dict = {self.state_input:state_batch,
self.y_input:y_batch,
self.action_input:action_batch})
tf.summary.scalar('cost function value',cost_value)
if self.time_step%100 == 1:
self.train_writer.add_summary(summary, self.time_step)
def egreedy_action(self,state):
# output of self.Q_value.eval():[[-2.99488783 -0.5567829 ]]
# Q_value=[-2.99488783 -0.5567829 ]
Q_value = self.Q_value.eval(feed_dict = {
self.state_input:[state]
})[0]
# produce a float between [0,1] random
if random.random() <= self.epsilon:
return random.randint(0,self.action_dim - 1)
else:
return np.argmax(Q_value) #return ints
self.epsilon -= (INITIAL_EPSILON - FINAL_EPSILON)/10000
def action(self,state):
#return the action which will have the maximum Q
return np.argmax(self.Q_value.eval(feed_dict = {
self.state_input:[state]
})[0])
def weight_variable(self,shape):
initial = tf.truncated_normal(shape)
return tf.Variable(initial)
def bias_variable(self,shape):
initial = tf.constant(0.01, shape = shape)
return tf.Variable(initial)
# ---------------------------------------------------------
# Hyper Parameters
ENV_NAME = 'CartPole-v0'
EPISODE = 1000 # Episode limitation
STEP = 300 # Step limitation in an episode
TEST = 10 # The number of experiment test every 100 episode
def main(_):
# initialize OpenAI Gym env and dqn agent
if tf.gfile.Exists(FLAGS.log_dir):
tf.gfile.DeleteRecursively(FLAGS.log_dir)
tf.gfile.MakeDirs(FLAGS.log_dir)
env = gym.make(ENV_NAME)
agent = DQN(env)
for episode in xrange(EPISODE):
# initialize task
state = env.reset()
# Train
for step in xrange(STEP):
action = agent.egreedy_action(state) # e-greedy action for train
next_state,reward,done,_ = env.step(action)
# Define reward for agent
# reward_agent = -1 if done else 0.1
agent.perceive(state,action,reward,next_state,done)
state = next_state
if done:
break
# Test every 100 episodes
if episode % 100 == 0:
total_reward = 0
for i in xrange(TEST):
state = env.reset()
for j in xrange(STEP):
env.render()
action = agent.action(state) # direct action for test
state,reward,done,_ = env.step(action)
total_reward += reward
if done:
break
ave_reward = total_reward/TEST
print 'episode: ',episode,'Evaluation Average Reward:',ave_reward
if ave_reward >= 200:
break
agent.train_writer.close()
env.close()
if __name__ == '__main__':
parser = argparse.ArgumentParser()
parser.add_argument('--log_dir', type=str, default='/tmp/tensorflow/draft',
help='Summaries log directory')
FLAGS, unparsed = parser.parse_known_args()
tf.app.run(main=main, argv=[sys.argv[0]] + unparsed)