-
Notifications
You must be signed in to change notification settings - Fork 0
/
openai_esdeep2.py
225 lines (191 loc) · 7.1 KB
/
openai_esdeep2.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
# Copyright 2015 The TensorFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ==============================================================================
"""A deep MNIST classifier using convolutional layers.
See extensive documentation at
https://www.tensorflow.org/get_started/mnist/pros
"""
# Disable linter warnings to maintain consistency with tutorial.
# pylint: disable=invalid-name
# pylint: disable=g-bad-import-order
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function
import argparse
import sys
import tempfile
from tensorflow.examples.tutorials.mnist import input_data
import tensorflow as tf
import numpy as np
FLAGS = None
# hyperparameters
npop = 50 # population size
sigma = 0.1 # noise standard deviation
alpha = 0.001 # learning rate
scaling = 2**125
np.random.seed(0)
def get_logits(x,params):
scaling = 2**125
h1 = tf.nn.bias_add(tf.matmul(x , params[0]), params[1]) / scaling
h2 = tf.nn.bias_add(tf.matmul(h1, params[2]) , params[3] / scaling)
o = tf.nn.bias_add(tf.matmul(h2, params[4]), params[5]/ scaling)*scaling
return o
def p2v(par,sess):
v = []
Nout= []
for tp in par:
p = sess.run(tp)
if len(p.shape)==2:
N = np.random.randn(npop, p.shape[0],p.shape[1])
if len(p.shape)==1:
N = np.random.randn(npop, p.shape[0])
Nout.append(N)
v.append(p)
return v,Nout
def v2p(v,par,sess):
start = 0
ov = []
# print(v)
for tp in par:
p = sess.run(tp)
if len(p.shape)==2:
s0,s1 = p.shape
end = start+(s0*s1)
if len(p.shape)==1:
s0 = p.shape
end = start+s0[0]
vcut = v[start:end]
vcut = np.reshape(vcut,p.shape)
ov.append(vcut)
start = end
# updatep = tf.assign(tp,vcut)
# sess.run(updatep)
return ov
def lossfun(y_,y_conv):
cross_entropy = tf.nn.softmax_cross_entropy_with_logits(labels=y_,
logits=y_conv)
cross_entropy = tf.reduce_mean(cross_entropy)
return cross_entropy
def main(_):
# Import data
mnist = input_data.read_data_sets(FLAGS.data_dir, one_hot=True)
x = tf.placeholder(tf.float32, [None, 784])
y_ = tf.placeholder(tf.float32, [None, 10])
# Create the model
w1 = tf.Variable(np.random.normal(scale=np.sqrt(2./784),size=[784,512]).astype(np.float32))
b1 = tf.Variable(np.zeros(512,dtype=np.float32))
w2 = tf.Variable(np.random.normal(scale=np.sqrt(2./512),size=[512,512]).astype(np.float32))
b2 = tf.Variable(np.zeros(512,dtype=np.float32))
w3 = tf.Variable(np.random.normal(scale=np.sqrt(2./512),size=[512,10]).astype(np.float32))
b3 = tf.Variable(np.zeros(10,dtype=np.float32))
params = [w1,b1,w2,b2,w3,b3]
h1 = tf.nn.bias_add(tf.matmul(x , params[0]), params[1]) / scaling
h2 = tf.nn.bias_add(tf.matmul(h1, params[2]) , params[3] / scaling)
y_conv = tf.nn.bias_add(tf.matmul(h2, params[4]), params[5]/ scaling)*scaling
# optimize model
w1_try = tf.placeholder(tf.float32,[784,512])
b1_try = tf.placeholder(tf.float32,512)
w2_try = tf.placeholder(tf.float32,[512,512])
b2_try = tf.placeholder(tf.float32,512)
w3_try = tf.placeholder(tf.float32,[512,10])
b3_try = tf.placeholder(tf.float32,10)
h1_try = tf.nn.bias_add(tf.matmul(x , w1_try ), b1_try) / scaling
h2_try = tf.nn.bias_add(tf.matmul(h1_try, w2_try) , b2_try / scaling)
reword = tf.nn.bias_add(tf.matmul(h2_try , w3_try), b3_try/ scaling)*scaling
# optimize model
w1new = tf.placeholder(tf.float32,[784,512])
b1new = tf.placeholder(tf.float32,512)
w2new = tf.placeholder(tf.float32,[512,512])
b2new = tf.placeholder(tf.float32,512)
w3new = tf.placeholder(tf.float32,[512,10])
b3new = tf.placeholder(tf.float32,10)
uppar1 = tf.assign(w1,w1new)
uppar2 = tf.assign(b1,b1new)
uppar3 = tf.assign(w2,w2new)
uppar4 = tf.assign(b2,b2new)
uppar5 = tf.assign(w3,w3new)
uppar6 = tf.assign(b3,b3new)
# Define loss and optimizer
with tf.name_scope('loss'):
# cross_entropy = lossfun(y_,y_conv)
loss_try= lossfun(y_,reword)
with tf.name_scope('accuracy'):
correct_prediction = tf.equal(tf.argmax(y_conv, 1), tf.argmax(y_, 1))
correct_prediction = tf.cast(correct_prediction, tf.float32)
accuracy = tf.reduce_mean(correct_prediction)
# with tf.Graph().as_default():
with tf.Session() as sess:
# global sess
sess.run(tf.global_variables_initializer())
for i in range(10000):
batch = mnist.train.next_batch(50)
# print("i=========:",i)
if i % 200== 0:
train_accuracy = sess.run(accuracy,feed_dict={
x: batch[0], y_: batch[1]})
print('step %d, training accuracy %g' % (i, train_accuracy))
# es step 1 jitter par using gaussian of sigma 0.1
vp,N = p2v(params,sess)
R = np.zeros(npop)
for j in range(npop):
ov = []
for m in range(len(params)):
vp_try = np.add(vp[m],sigma*N[m][j])
ov.append(vp_try)
R[j] = -1*sess.run(loss_try,feed_dict={x:batch[0], y_: batch[1],
w1_try:ov[0],b1_try:ov[1],
w2_try:ov[2],b2_try:ov[3],
w3_try:ov[4],b3_try:ov[5]}
)
if i==0:
print("R==================")
print(R)
print("vp==================")
print(vp[1][0:100])
# es step 2 update par by R
A = (R - np.mean(R)) / np.std(R)
if i==0:
# print("dw==================")
# print(alpha/(npop*sigma) * np.dot(N.T, A))
print("A==================")
print(A)
ovnew = []
for m in range(len(params)):
D = np.zeros(vp[m].shape)
for k in range(npop):
D =D +N[m][k]*A[k]
# K = N[i]
# K = alpha/(npop*sigma) * np.dot(N[m].T, A)
# print(A.shape)
vp[m] = vp[m] + alpha/(npop*sigma) * D
ovnew.append(vp[m])
if i==0:
print("dw==============")
print(vp[1][0:100])
# ovnew = v2p(vp,params,sess)
sess.run(uppar1,feed_dict={w1new:ovnew[0]})
sess.run(uppar2,feed_dict={b1new:ovnew[1]})
sess.run(uppar3,feed_dict={w2new:ovnew[2]})
sess.run(uppar4,feed_dict={b2new:ovnew[3]})
sess.run(uppar5,feed_dict={w3new:ovnew[4]})
sess.run(uppar6,feed_dict={b3new:ovnew[5]})
print('test accuracy %g' % accuracy.eval(feed_dict={
x: mnist.test.images, y_: mnist.test.labels}))
if __name__ == '__main__':
parser = argparse.ArgumentParser()
parser.add_argument('--data_dir', type=str,
default='/tmp/tensorflow/mnist/input_data',
help='Directory for storing input data')
FLAGS, unparsed = parser.parse_known_args()
tf.app.run(main=main, argv=[sys.argv[0]] + unparsed)