-
Notifications
You must be signed in to change notification settings - Fork 9
/
DQN_MLP.py
288 lines (208 loc) · 11.3 KB
/
DQN_MLP.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
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
"""
This part of code is the DQN brain, which is a brain of the agent.
All decisions are made in here.
Using Tensorflow to build the neural network.
View more on my tutorial page: https://morvanzhou.github.io/tutorials/
Using:
Tensorflow: 1.0
gym: 0.7.3
"""
import numpy as np
import tensorflow as tf
np.random.seed(1)
tf.set_random_seed(1)
# DQN + MLP with one hidden layer
class MLP1:
def __init__(
self,
n_actions,
n_features,
learning_rate=0.01,
reward_decay=0.9,
e_greedy=0.9,
replace_target_iter=300,
memory_size=500,
):
self.n_actions = n_actions
self.n_features = n_features
self.lr = learning_rate
self.gamma = reward_decay
self.epsilon_max = e_greedy
self.replace_target_iter = replace_target_iter
self.memory_size = memory_size
self.epsilon = e_greedy
# total learning step
self.learn_step_counter = 0
# initialize zero memory [s, a, r, s_]
self.memory = np.zeros((self.memory_size, n_features * 2 + 2))
# consist of [target_net, evaluate_net]
self.nInternalUnits = 64
self._build_net(self.nInternalUnits)
self.sess = tf.Session()
self.sess.run(tf.global_variables_initializer())
self.sess.run(self.replace_target_op)
self.cost_his = []
def _build_net(self, nInternalUnits):
# ------------------ build evaluate_net ------------------
self.s = tf.placeholder(tf.float32, [None, self.n_features]) # input
self.q_target = tf.placeholder(tf.float32, [None, self.n_actions]) # for calculating loss
self.w1_eval = tf.Variable(tf.random_normal([self.n_features, nInternalUnits], mean=0.0, stddev=0.3))
self.b1_eval = tf.Variable(tf.random_normal([1, nInternalUnits], mean=0.0, stddev=0.3))
self.l1_eval = tf.nn.tanh(tf.matmul(self.s, self.w1_eval) + self.b1_eval)
self.w2_eval = tf.Variable(tf.random_normal([nInternalUnits, self.n_actions], mean=0.0, stddev=0.3))
self.b2_eval = tf.Variable(tf.random_normal([1, self.n_actions], mean=0.0, stddev=0.3))
self.q_eval = tf.matmul(self.l1_eval, self.w2_eval) + self.b2_eval
self.loss = tf.reduce_mean(tf.squared_difference(self.q_target, self.q_eval))
self._train_op = tf.train.GradientDescentOptimizer(self.lr).minimize(self.loss)
# ------------------ build target_net ------------------
self.s_ = tf.placeholder(tf.float32, [None, self.n_features]) # input
self.w1_target = tf.Variable(tf.random_normal([self.n_features, nInternalUnits], mean=0.0, stddev=0.3))
self.b1_target = tf.Variable(tf.random_normal([1, nInternalUnits], mean=0.0, stddev=0.3))
self.l1_target = tf.nn.tanh(tf.matmul(self.s_, self.w1_target) + self.b1_target)
self.w2_target = tf.Variable(tf.random_normal([nInternalUnits, self.n_actions], mean=0.0, stddev=0.3))
self.b2_target = tf.Variable(tf.random_normal([1, self.n_actions], mean=0.0, stddev=0.3))
self.q_next = tf.matmul(self.l1_target, self.w2_target) + self.b2_target
self.replace_target_op = [tf.assign(self.w1_target, self.w1_eval), tf.assign(self.b1_target, self.b1_eval),
tf.assign(self.w2_target, self.w2_eval), tf.assign(self.b2_target, self.b2_eval)]
def store_transition(self, s, a, r, s_):
if not hasattr(self, 'memory_counter'):
self.memory_counter = 0
transition = np.hstack((s, [a, r], s_))
# replace the old memory with new memory
index = self.memory_counter % self.memory_size
self.memory[index, :] = transition
self.memory_counter += 1
def choose_action(self, observation):
# to have batch dimension when feed into tf placeholder
observation = observation[np.newaxis, :]
if np.random.uniform() < self.epsilon:
# forward feed the observation and get q value for every actions
actions_value = self.sess.run(self.q_eval, feed_dict={self.s: observation})
action = np.argmax(actions_value)
else:
action = np.random.randint(0, self.n_actions)
return action
def learn(self):
batch_memory = self.memory
q_next, q_eval = self.sess.run(
[self.q_next, self.q_eval],
feed_dict={
self.s_: batch_memory[:, -self.n_features:], # fixed params
self.s: batch_memory[:, :self.n_features], # newest params
})
# change q_target w.r.t q_eval's action
q_target = q_eval.copy()
eval_act_index = batch_memory[:, self.n_features].astype(int)
reward = batch_memory[:, self.n_features + 1]
next_q_value = self.gamma * np.max(q_next, axis=1)
for index in range(len(eval_act_index)):
q_target[index, eval_act_index[index]] = reward[index] + next_q_value[index]
# train eval network
_, self.cost = self.sess.run([self._train_op, self.loss],
feed_dict={self.s: batch_memory[:, :self.n_features], self.q_target: q_target})
self.cost_his.append(self.cost)
# replace target parameters
if self.learn_step_counter % self.replace_target_iter == 0:
self.sess.run(self.replace_target_op)
self.learn_step_counter += 1
# DQN + MLP with two hidden layer
class MLP2:
def __init__(
self,
n_actions,
n_features,
learning_rate=0.01,
reward_decay=0.9,
e_greedy=0.9,
replace_target_iter=300,
memory_size=500,
):
self.n_actions = n_actions
self.n_features = n_features
self.lr = learning_rate
self.gamma = reward_decay
self.epsilon_max = e_greedy
self.replace_target_iter = replace_target_iter
self.memory_size = memory_size
self.epsilon = e_greedy
# total learning step
self.learn_step_counter = 0
# initialize zero memory [s, a, r, s_]
self.memory = np.zeros((self.memory_size, n_features * 2 + 2))
# consist of [target_net, evaluate_net]
self.nInternalUnits = 64
self._build_net(self.nInternalUnits)
self.sess = tf.Session()
self.sess.run(tf.global_variables_initializer())
self.sess.run(self.replace_target_op)
self.cost_his = []
def _build_net(self, nInternalUnits):
# ------------------ build evaluate_net ------------------
self.s = tf.placeholder(tf.float32, [None, self.n_features]) # input
self.q_target = tf.placeholder(tf.float32, [None, self.n_actions]) # for calculating loss
self.w1_eval = tf.Variable(tf.random_normal([self.n_features, nInternalUnits], mean=0.0, stddev=0.3))
self.b1_eval = tf.Variable(tf.random_normal([1, nInternalUnits], mean=0.0, stddev=0.3))
self.l1_eval = tf.nn.tanh(tf.matmul(self.s, self.w1_eval)+self.b1_eval)
self.w2_eval = tf.Variable(tf.random_normal([nInternalUnits, nInternalUnits], mean=0.0, stddev=0.3))
self.b2_eval = tf.Variable(tf.random_normal([1, nInternalUnits], mean=0.0, stddev=0.3))
self.l2_eval = tf.nn.tanh(tf.matmul(self.l1_eval, self.w2_eval)+self.b2_eval)
self.w3_eval = tf.Variable(tf.random_normal([nInternalUnits, self.n_actions], mean=0.0, stddev=0.3))
self.b3_eval = tf.Variable(tf.random_normal([1, self.n_actions], mean=0.0, stddev=0.3))
self.q_eval = tf.matmul(self.l2_eval, self.w3_eval) + self.b3_eval
self.loss = tf.reduce_mean(tf.squared_difference(self.q_target, self.q_eval))
self._train_op = tf.train.GradientDescentOptimizer(self.lr).minimize(self.loss)
# ------------------ build target_net ------------------
self.s_ = tf.placeholder(tf.float32, [None, self.n_features]) # input
self.w1_target = tf.Variable(tf.random_normal([self.n_features, nInternalUnits], mean=0.0, stddev=0.3))
self.b1_target = tf.Variable(tf.random_normal([1, nInternalUnits], mean=0.0, stddev=0.3))
self.l1_target = tf.nn.tanh(tf.matmul(self.s, self.w1_target)+self.b1_target)
self.w2_target = tf.Variable(tf.random_normal([nInternalUnits, nInternalUnits], mean=0.0, stddev=0.3))
self.b2_target = tf.Variable(tf.random_normal([1, nInternalUnits], mean=0.0, stddev=0.3))
self.l2_target = tf.nn.tanh(tf.matmul(self.l1_target, self.w2_target)+self.b2_target)
self.w3_target = tf.Variable(tf.random_normal([nInternalUnits, self.n_actions], mean=0.0, stddev=0.3))
self.b3_target = tf.Variable(tf.random_normal([1, self.n_actions], mean=0.0, stddev=0.3))
self.q_next = tf.matmul(self.l2_target, self.w3_target)+self.b3_target
self.replace_target_op = [tf.assign(self.w1_target, self.w1_eval), tf.assign(self.b1_target, self.b1_eval),
tf.assign(self.w2_target, self.w2_eval), tf.assign(self.b2_target, self.b2_eval),
tf.assign(self.w3_target, self.w3_eval), tf.assign(self.b3_target, self.b3_eval)]
def store_transition(self, s, a, r, s_):
if not hasattr(self, 'memory_counter'):
self.memory_counter = 0
transition = np.hstack((s, [a, r], s_))
# replace the old memory with new memory
index = self.memory_counter % self.memory_size
self.memory[index, :] = transition
self.memory_counter += 1
def choose_action(self, observation):
# to have batch dimension when feed into tf placeholder
observation = observation[np.newaxis, :]
if np.random.uniform() < self.epsilon:
# forward feed the observation and get q value for every actions
actions_value = self.sess.run(self.q_eval, feed_dict={self.s: observation})
action = np.argmax(actions_value)
else:
action = np.random.randint(0, self.n_actions)
return action
def learn(self):
batch_memory = self.memory
q_next, q_eval = self.sess.run(
[self.q_next, self.q_eval],
feed_dict={
self.s_: batch_memory[:, -self.n_features:], # fixed params
self.s: batch_memory[:, :self.n_features], # newest params
})
# change q_target w.r.t q_eval's action
q_target = q_eval.copy()
eval_act_index = batch_memory[:, self.n_features].astype(int)
reward = batch_memory[:, self.n_features + 1]
next_q_value = self.gamma * np.max(q_next, axis=1)
for index in range(len(eval_act_index)):
q_target[index, eval_act_index[index]] = reward[index] + next_q_value[index]
# train eval network
_, self.cost = self.sess.run([self._train_op, self.loss],
feed_dict={self.s: batch_memory[:, :self.n_features], self.q_target: q_target})
self.cost_his.append(self.cost)
# replace target parameters
if self.learn_step_counter % self.replace_target_iter == 0:
self.sess.run(self.replace_target_op)
self.learn_step_counter += 1