-
Notifications
You must be signed in to change notification settings - Fork 0
/
dqn.py
176 lines (150 loc) · 6.57 KB
/
dqn.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
import random
import numpy as np
from collections import deque
from tensorflow.keras import Sequential
from tensorflow.keras.layers import Conv2D, MaxPooling2D, Flatten, Dense, ReLU, BatchNormalization
from tensorflow.keras.optimizers import Adam
# ACTION SPACE
# [steer, gas, brake]
# steer: -1.0 ~ 1.0 (left ~ right)
# gas: 0.0 ~ 1.0
# brake: 0.0 ~ 1.0
default_action_space = [
(0.0, 0.0, 0.0), (-1.0, 0.0, 0.0), (1.0, 0.0, 0.0), (0.0, 1.0, 0.0), (0.0, 0.0, 1.0), # full gas, left, right, brake and do nothing
(-0.5, 0.0, 0.0), (0.5, 0.0, 0.0), (0.0, 0.5, 0.0), (0.0, 0.0, 0.5), # half gas, left, right, brake
(-0.5, 0.5, 0.0), (0.5, 0.5, 0.0), (-0.5, 0.0, 0.5), (0.5, 0.0, 0.5), # half gas and left, right and brake
(-1.0, 0.0, 0.5), (1.0, 0.0, 0.5), # half brake and left, right
(-0.5, 0.5, 0.5), (0.5, 0.5, 0.5), # half gas, left, right and brake
]
normal_space = [
(-1, 1, 0.2), (0, 1, 0.2), (1, 1, 0.2), # Action Space Structure
(-1, 1, 0), (0, 1, 0), (1, 1, 0), # (Steering Wheel, Gas, Break)
(-1, 0, 0.2), (0, 0, 0.2), (1, 0, 0.2), # Range -1~1 0~1 0~1
(-1, 0, 0), (0, 0, 0), (1, 0, 0)
]
forward_action_space = [
(0.0, 1.0, 0.0),
(1.0, 1.0, 0.0), (-1.0, 1.0, 0.0),
(0.0, 0.5, 0.0),
(1.0, 0.5, 0.0), (-1.0, 0.5, 0.0),
(0.0, 1.0, 0.5),
(1.0, 1.0, 0.5), (-1.0, 1.0, 0.5),
(0.0, 1.0, 0.2),
(1.0, 1.0, 0.2), (-1.0, 1.0, 0.2),
(0.0, 1.0, 0.8),
(1.0, 1.0, 0.8), (-1.0, 1.0, 0.8),
(0.5, 1.0, 0.0), (-0.5, 1.0, 0.0),
(0.5, 0.5, 0.0), (-0.5, 0.5, 0.0),
(0.5, 1.0, 0.5), (-0.5, 1.0, 0.5),
(0.5, 1.0, 0.2), (-0.5, 1.0, 0.2),
(0.5, 1.0, 0.8), (-0.5, 1.0, 0.8),
(0.25, 1.0, 0.0), (-0.25, 1.0, 0.0),
(0.25, 0.5, 0.0), (-0.25, 0.5, 0.0),
(0.25, 1.0, 0.5), (-0.25, 1.0, 0.5),
(0.25, 1.0, 0.2), (-0.25, 1.0, 0.2),
(0.25, 1.0, 0.8), (-0.25, 1.0, 0.8),
]
# other action spaces can be defined here to emulate different driving styles
class CarRacingAgent:
def __init__(
self,
action_space=forward_action_space,
frame_stack_num=3,
memory_size=5000,
gamma=0.95,
epsilon=1.0,
epsilon_min=0.1,
epsilon_decay=0.9999,
learning_rate=0.001,
improved=False,
):
self.action_space = action_space
self.frame_stack_num = frame_stack_num
self.memory_size = memory_size
self.memory = self.build_memory()
self.gamma = gamma
self.epsilon = epsilon
self.epsilon_min = epsilon_min
self.epsilon_decay = epsilon_decay
self.learning_rate = learning_rate
self.model = self._build_model(improved=improved)
self.target_model = self._build_model(improved=improved)
self.update_target_model()
def get_params(self):
return {
"action_space": self.action_space,
"frame_stack_num": self.frame_stack_num,
"memory_size": self.memory_size,
"gamma": self.gamma,
"epsilon": self.epsilon,
"epsilon_min": self.epsilon_min,
"epsilon_decay": self.epsilon_decay,
"learning_rate": self.learning_rate,
}
def build_memory(self, memory_size=5000):
return deque(maxlen=memory_size)
def _build_model(self, improved=False):
model = Sequential()
model.add(Conv2D(filters=6, kernel_size=(4, 4), strides=(2,2), activation='relu', input_shape=(96, 96, self.frame_stack_num)))
model.add(MaxPooling2D(pool_size=(2, 2)))
if improved:
model.add(ReLU())
model.add(BatchNormalization())
model.add(Conv2D(filters=12, kernel_size=(4, 4), strides=(2,2), activation='relu'))
model.add(MaxPooling2D(pool_size=(2, 2)))
if improved:
model.add(ReLU())
model.add(BatchNormalization())
model.add(Flatten())
model.add(Dense(300, activation='relu'))
model.add(Dense(len(self.action_space), activation=None))
model.compile(loss='mse', optimizer=Adam(learning_rate=self.learning_rate, epsilon=1e-7))
return model
def update_target_model(self):
self.target_model.set_weights(self.model.get_weights())
def remember(self, state, action, reward, next_state, done):
self.memory.append((state, self.action_space.index(action), reward, next_state, done))
def act(self, state, explore=True):
if explore and np.random.rand() <= self.epsilon:
return random.choice(self.action_space)
act_values = self.model.predict(state, verbose=0)
return self.action_space[np.argmax(act_values[0])]
def replay(self, batch_size):
rng = np.random.default_rng()
"""
minibatch = random.sample(self.memory, batch_size)
train_state = []
train_target = []
for state, action_index, reward, next_state, done in minibatch:
target = self.model.predict(np.expand_dims(state, axis=0))[0]
if done:
target[action_index] = reward
else:
t = self.target_model.predict(np.expand_dims(next_state, axis=0))[0]
target[action_index] = reward + self.gamma * np.amax(t)
train_state.append(state)
train_target.append(target)
self.model.fit(np.array(train_state), np.array(train_target), epochs=1, verbose=0)
if self.epsilon > self.epsilon_min:
self.epsilon *= self.epsilon_decay
"""
minibatch = rng.choice(np.array(self.memory), batch_size, replace=False)
states, actions, rewards, next_states, dones = zip(*minibatch)
train_states = np.array([state for state in states])
train_targets = self.model.predict(train_states, verbose=0)
next_states = np.array([state for state in next_states])
next_state_values = np.amax(self.target_model.predict(next_states, verbose=0), axis=1)
dones = np.array(dones) # Convert dones to NumPy array
train_targets[np.arange(batch_size), actions] = rewards + self.gamma * next_state_values * (1 - dones)
self.model.fit(train_states, train_targets, epochs=1, verbose=0)
#print(f'epsilon: {self.epsilon}')
if self.epsilon > self.epsilon_min:
self.epsilon *= self.epsilon_decay
def load(self, name):
self.model.load_weights(name)
self.update_target_model()
def save(self, name):
self.model.save_weights(name)
if __name__ == '__main__':
agent = CarRacingAgent()
print(agent.model.summary())