-
Notifications
You must be signed in to change notification settings - Fork 4
/
trial_RL.py
196 lines (137 loc) · 6.27 KB
/
trial_RL.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
# -*- coding: utf-8 -*-
"""
Created on Mon Jan 18 16:19:39 2021
@author: mnguy
##
"""
import numpy as np
import tensorflow.compat.v1 as tf
tf.disable_v2_behavior()
import matplotlib.pylab as plt
import random
import pickle
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Dense, Dropout, Conv2D, MaxPooling2D, Activation, Flatten
from tensorflow.keras.optimizers import Adam
from collections import deque
import scipy.io as sio
from parameters import parameters, write_para
import pickle
REPLAY_MEMORY_SIZE = 10_000 # How many last steps to keep for model training
MINIBATCH_SIZE = 64 # How many steps (samples) to use for training
UPDATE_TARGET_EVERY = 10 # Terminal states (end of episodes)
MIN_REWARD = -200 # For model save
MEMORY_FRACTION = 0.20
LEARNING_RATE = 0.1
DISCOUNT = 0.99
# Environment settings
#EPISODES = 20_000
EPISODES = 10000
# Exploration settings
epsilon = 1 # not a constant, going to be decayed
EPSILON_DECAY = 0.99975
# Stats settings
AGGREGATE_STATS_EVERY = 50 # episodes
SHOW_PREVIEW = False
# # initialize
# write_para()
file_Name = "parameters_data" # depends on the file name that we want to save
fileObject = open(file_Name,'rb')
para = pickle.load(fileObject)
num_bits = np.log2(para.num_codework) + para.num_per_recent*para.num_recent + para.num_indicateDRLorDNN
num_bits_int = np.ceil(num_bits)
num_actions = para.num_actions
num_states = para.num_states
try:
# Disable all GPUS
tf.config.set_visible_devices([], 'GPU')
visible_devices = tf.config.get_visible_devices()
for device in visible_devices:
assert device.device_type != 'GPU'
except:
# Invalid device or cannot modify virtual devices once initialized.
pass
class DQNAgent:
def __init__(self):
self._num_states = num_states # input of DRL
self._num_actions = num_actions # num_actions = para.Ntotal_urllc*2^num_cluster, output of DRL
self._batch_size = MINIBATCH_SIZE
self.num_per_recent = para.num_per_recent
self.num_recent = para.num_recent
self.model = self.define_model()
self.target_model = self.define_model()
self.target_model.set_weights(self.model.get_weights()) # two networks with the same initial weights
# An array with last n steps for training
self.replay_memory = deque(maxlen=REPLAY_MEMORY_SIZE) #replay1#
# Used to count when to update target network with main network's weights
self.target_update_counter = 0
def define_model(self):
model = Sequential()
model.add(Dense(50, input_dim= self._num_states, activation='relu'))
model.add(Dense(50, activation='relu'))
#model.add(Dense(50, activation='relu'))
model.add(Dense(self._num_actions, activation='relu'))
model.compile(loss="mse", optimizer=Adam(lr=0.001), metrics=['accuracy'])
return model
def update_replay_memory(self, transition):
self.replay_memory.append(transition)
def convert_array_action(self, action):
# action is integer number
yy = bin(int(action))[2:]
xx = yy[::-1]
action_indicate = np.array(int(xx[0]))
action_recent = np.zeros((self.num_recent,1))
for kk in range(self.num_recent):
a2 = [int(x) for x in xx[kk*self.num_per_recent+1:(kk+1)*self.num_per_recent+1]]
a3 = np.asarray(a2)
ss = 0
for jj in range(a3.shape[0]):
ss = ss + 2**jj*a3[jj]
action_recent[kk] = int(2**self.num_per_recent*kk + ss)
a2 = [int(x) for x in xx[(self.num_recent)*self.num_per_recent+1:]]
a3 = np.asarray(a2)
ss = 0
for jj in range(len(a3)):
ss = ss + 2**jj*a3[jj]
action_refine = ss
#action_out = np.concatenate((a1,a4,a3))
return action_indicate, action_recent, action_refine
# Trains main network every step during episode
def train(self, terminal_state):
# Get a minibatch of random samples from memory replay table
minibatch = random.sample(self.replay_memory, self._batch_size)
# Get current states from minibatch, then query NN model for Q values
current_states = np.array([transition[0] for transition in minibatch])
#current_states = current_states.reshape((self._batch_size,self._num_states))
current_qs_list = self.model.predict(current_states)
# Get future states from minibatch, then query NN model for Q values
# When using target network, query it, otherwise main network should be queried
new_current_states = np.array([transition[3] for transition in minibatch])
#new_current_states = new_current_states.reshape((self._batch_size,self._num_states))
future_qs_list = self.target_model.predict(new_current_states)
X = []
y = []
# Now we need to enumerate our batches
for index, (current_state, action, reward, new_current_state, done) in enumerate(minibatch):
# If not a terminal state, get new q from future states, otherwise set it to 0
# almost like with Q Learning, but we use just part of equation here
if not done:
max_future_q = np.max(future_qs_list[index])
new_q = reward + DISCOUNT * max_future_q
else:
new_q = reward
# Update Q value for given state
current_qs = current_qs_list[index]
current_qs[action] = new_q
# And append to our training data
X.append(current_state)
y.append(current_qs)
# Fit on all samples as one batch, log only on terminal state
self.model.fit(np.array(X), np.array(y), epochs=50, batch_size=16, verbose=0)
# Update target network counter every episode
if terminal_state:
self.target_update_counter += 1
# If counter reaches set value, update target network with weights of main network
if self.target_update_counter > UPDATE_TARGET_EVERY:
self.target_model.set_weights(self.model.get_weights())
self.target_update_counter = 0