-
Notifications
You must be signed in to change notification settings - Fork 49
/
cnnmodel.py
348 lines (245 loc) · 13.5 KB
/
cnnmodel.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
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
import tensorflow as tf
import numpy as np
import random
import tensorflow.contrib.layers as layers
from tqdm import tqdm
import time
class Settings(object):
def __init__(self):
self.vocab_size = 114042
self.len_sentence = 70
self.num_epochs = 3
self.num_classes = 53
self.cnn_size = 230
self.num_layers = 1
self.pos_size = 5
self.pos_num = 123
self.word_embedding = 50
self.keep_prob = 0.5
self.batch_size = 300
self.num_steps = 10000
self.lr= 0.001
class CNN():
def __init__(self, word_embeddings, setting):
self.vocab_size = setting.vocab_size
self.len_sentence= len_sentence = setting.len_sentence
self.num_epochs = setting.num_epochs
self.num_classes = num_classes =setting.num_classes
self.cnn_size = setting.cnn_size
self.num_layers = setting.num_layers
self.pos_size = setting.pos_size
self.pos_num = setting.pos_num
self.word_embedding = setting.word_embedding
self.lr = setting.lr
word_embedding = tf.get_variable(initializer=word_embeddings, name='word_embedding')
pos1_embedding = tf.get_variable('pos1_embedding', [self.pos_num, self.pos_size])
pos2_embedding = tf.get_variable('pos2_embedding', [self.pos_num, self.pos_size])
#relation_embedding = tf.get_variable('relation_embedding', [self.num_classes, self.cnn_size])
self.input_word = tf.placeholder(dtype=tf.int32, shape=[None, len_sentence], name='input_word')
self.input_pos1 = tf.placeholder(dtype=tf.int32, shape=[None, len_sentence], name='input_pos1')
self.input_pos2 = tf.placeholder(dtype=tf.int32, shape=[None, len_sentence], name='input_pos2')
self.input_y = tf.placeholder(dtype=tf.float32, shape=[None, num_classes], name='input_y')
self.keep_prob = tf.placeholder(tf.float32)
self.input_word_ebd = tf.nn.embedding_lookup(word_embedding, self.input_word)
self.input_pos1_ebd = tf.nn.embedding_lookup(pos1_embedding, self.input_pos1)
self.input_pos2_ebd = tf.nn.embedding_lookup(pos2_embedding, self.input_pos2)
self.inputs = tf.concat(axis=2,values=[self.input_word_ebd,self.input_pos1_ebd,self.input_pos2_ebd])
self.inputs = tf.reshape(self.inputs, [-1,self.len_sentence,self.word_embedding+self.pos_size*2,1] )
conv = layers.conv2d(inputs =self.inputs ,num_outputs = self.cnn_size ,kernel_size = [3,60],stride=[1,60],padding='SAME')
max_pool = layers.max_pool2d(conv,kernel_size = [70,1],stride=[1,1])
self.sentence = tf.reshape(max_pool, [-1, self.cnn_size])
tanh = tf.nn.tanh(self.sentence)
drop = layers.dropout(tanh,keep_prob=self.keep_prob)
self.outputs = layers.fully_connected(inputs = drop,num_outputs = self.num_classes,activation_fn = tf.nn.softmax)
'''
self.y_index = tf.argmax(self.input_y,1,output_type=tf.int32)
self.indexes = tf.range(0, tf.shape(self.outputs)[0]) * tf.shape(self.outputs)[1] + self.y_index
self.responsible_outputs = - tf.reduce_mean(tf.log(tf.gather(tf.reshape(self.outputs, [-1]),self.indexes)))
'''
#loss
#self.cross_loss = -tf.reduce_mean( tf.log(tf.reduce_sum( self.input_y * self.outputs ,axis=1)))
self.cross_loss = -tf.reduce_mean(tf.reduce_sum( self.input_y * tf.log( self.outputs),axis=1))
self.reward = tf.log(tf.reduce_sum( self.input_y * self.outputs ,axis=1))
self.l2_loss = tf.contrib.layers.apply_regularization(regularizer=tf.contrib.layers.l2_regularizer(0.0001),
weights_list=tf.trainable_variables())
self.final_loss = self.cross_loss + self.l2_loss
#accuracy
self.pred = tf.argmax(self.outputs,axis=1)
self.pred_prob = tf.reduce_max(self.outputs,axis=1)
self.y_label = tf.argmax(self.input_y,axis=1)
self.accuracy = tf.reduce_mean(tf.cast( tf.equal(self.pred,self.y_label), 'float'))
#minimize loss
optimizer = tf.train.AdamOptimizer(self.lr)
self.train_op = optimizer.minimize(self.final_loss)
self.tvars = tf.trainable_variables()
# manual update parameters
self.tvars_holders = []
for idx, var in enumerate(self.tvars):
placeholder = tf.placeholder(tf.float32, name=str(idx) + '_holder')
self.tvars_holders.append(placeholder)
self.update_tvar_holder = []
for idx, var in enumerate(self.tvars):
update_tvar = tf.assign(var, self.tvars_holders[idx])
self.update_tvar_holder.append(update_tvar)
def train(path_train_word,path_train_pos1,path_train_pos2,path_train_y,save_path):
print('reading wordembedding')
wordembedding = np.load('./data/vec.npy')
print('reading training data')
cnn_train_word = np.load(path_train_word)
cnn_train_pos1 = np.load(path_train_pos1)
cnn_train_pos2 = np.load(path_train_pos2)
cnn_train_y = np.load(path_train_y)
settings = Settings()
settings.vocab_size = len(wordembedding)
settings.num_classes = len(cnn_train_y[0])
settings.num_steps = len(cnn_train_word) // settings.batch_size
with tf.Graph().as_default():
sess = tf.Session()
with sess.as_default():
initializer = tf.contrib.layers.xavier_initializer()
with tf.variable_scope("model", reuse=None, initializer=initializer):
model = CNN(word_embeddings=wordembedding, setting=settings)
sess.run(tf.global_variables_initializer())
saver = tf.train.Saver()
#saver.restore(sess,save_path=save_path)
for epoch in range(1,settings.num_epochs+1):
bar = tqdm(range(settings.num_steps), desc='epoch {}, loss=0.000000, accuracy=0.000000'.format(epoch))
for _ in bar:
sample_list = random.sample(range(len(cnn_train_y)),settings.batch_size)
batch_train_word = [cnn_train_word[x] for x in sample_list]
batch_train_pos1 = [cnn_train_pos1[x] for x in sample_list]
batch_train_pos2 = [cnn_train_pos2[x] for x in sample_list]
batch_train_y = [cnn_train_y[x] for x in sample_list]
feed_dict = {}
feed_dict[model.input_word] = batch_train_word
feed_dict[model.input_pos1] = batch_train_pos1
feed_dict[model.input_pos2] = batch_train_pos2
feed_dict[model.input_y] = batch_train_y
feed_dict[model.keep_prob] = settings.keep_prob
_,loss,accuracy=sess.run([model.train_op, model.final_loss, model.accuracy],feed_dict=feed_dict)
bar.set_description('epoch {} loss={:.6f} accuracy={:.6f}'.format(epoch, loss, accuracy))
#break
saver.save(sess, save_path=save_path)
#break
class interaction():
def __init__(self,sess,save_path ='model/model.ckpt3'):
self.settings = Settings()
wordembedding = np.load('./data/vec.npy')
self.sess = sess
with tf.variable_scope("model"):
self.model = CNN(word_embeddings=wordembedding, setting=self.settings)
self.saver = tf.train.Saver()
self.saver.restore(self.sess,save_path)
self.train_word = np.load('./data/train_word.npy')
self.train_pos1 = np.load('./data/train_pos1.npy')
self.train_pos2 = np.load('./data/train_pos2.npy')
self.y_train = np.load('data/train_y.npy')
def reward(self,batch_test_word,batch_test_pos1,batch_test_pos2,batch_test_y):
feed_dict = {}
feed_dict[self.model.input_word] = batch_test_word
feed_dict[self.model.input_pos1] = batch_test_pos1
feed_dict[self.model.input_pos2] = batch_test_pos2
feed_dict[self.model.input_y] = batch_test_y
feed_dict[self.model.keep_prob] = 1
outputs = (self.sess.run(self.model.reward,feed_dict = feed_dict))
return (outputs)
def sentence_ebd(self,batch_test_word,batch_test_pos1,batch_test_pos2,batch_test_y):
feed_dict = {}
feed_dict[self.model.input_word] = batch_test_word
feed_dict[self.model.input_pos1] = batch_test_pos1
feed_dict[self.model.input_pos2] = batch_test_pos2
feed_dict[self.model.input_y] = batch_test_y
feed_dict[self.model.keep_prob] = 1
outputs = self.sess.run(self.model.sentence,feed_dict = feed_dict)
return (outputs)
def test(self,batch_test_word,batch_test_pos1,batch_test_pos2):
feed_dict = {}
feed_dict[self.model.input_word] = batch_test_word
feed_dict[self.model.input_pos1] = batch_test_pos1
feed_dict[self.model.input_pos2] = batch_test_pos2
feed_dict[self.model.keep_prob] = 1
relation,prob = self.sess.run([self.model.pred,self.model.pred_prob],feed_dict = feed_dict)
return (relation,prob)
def update_cnn(self,update_word,update_pos1,update_pos2,update_y,updaterate):
num_steps = len(update_word) // self.settings.batch_size
with self.sess.as_default():
tvars_old = self.sess.run(self.model.tvars)
for i in tqdm(range(num_steps)):
batch_word = update_word[i* self.settings.batch_size:(i+1)*self.settings.batch_size]
batch_pos1 = update_pos1[i* self.settings.batch_size:(i+1)*self.settings.batch_size]
batch_pos2 = update_pos2[i* self.settings.batch_size:(i+1)*self.settings.batch_size]
batch_y = update_y[i* self.settings.batch_size:(i+1)*self.settings.batch_size]
feed_dict = {}
feed_dict[self.model.input_word] = batch_word
feed_dict[self.model.input_pos1] = batch_pos1
feed_dict[self.model.input_pos2] = batch_pos2
feed_dict[self.model.input_y] = batch_y
feed_dict[self.model.keep_prob] = self.settings.keep_prob
#_, loss, accuracy = sess.run([self.model.train_op,self.model.final_loss, self.model.accuracy], feed_dict=feed_dict)
self.sess.run(self.model.train_op, feed_dict=feed_dict)
# get tvars_new
tvars_new = self.sess.run(self.model.tvars)
# update old variables of the target network
tvars_update = self.sess.run(self.model.tvars)
for index, var in enumerate(tvars_update):
tvars_update[index] = updaterate * tvars_new[index] + (1 - updaterate) * tvars_old[index]
feed_dict = dictionary = dict(zip(self.model.tvars_holders, tvars_update))
self.sess.run(self.model.update_tvar_holder, feed_dict)
def produce_new_embedding(self):
# produce reward sentence_ebd average_reward
train_word = self.train_word
train_pos1 = self.train_pos1
train_pos2 = self.train_pos2
y_train = self.y_train
all_sentence_ebd = []
all_reward = []
all_reward_list = []
len_batch = len(train_word)
with self.sess.as_default():
for batch in tqdm(range(len_batch)):
batch_word = train_word[batch]
batch_pos1 = train_pos1[batch]
batch_pos2 = train_pos2[batch]
# batch_y = train_y[batch]
batch_y = [y_train[batch] for x in range(len(batch_word))]
tmp_sentence_ebd = self.sentence_ebd(batch_word, batch_pos1, batch_pos2, batch_y)
tmp_reward = self.reward(batch_word, batch_pos1, batch_pos2, batch_y)
all_sentence_ebd.append(tmp_sentence_ebd)
all_reward.append(tmp_reward)
all_reward_list += list(tmp_reward)
all_reward_list = np.array(all_reward_list)
average_reward = np.mean(all_reward_list)
average_reward = np.array(average_reward)
all_sentence_ebd = np.array(all_sentence_ebd)
all_reward = np.array(all_reward)
return average_reward,all_sentence_ebd,all_reward
def save_cnnmodel(self,save_path):
with self.sess.as_default():
self.saver.save(self.sess, save_path=save_path)
def tvars(self):
with self.sess.as_default():
tvars = self.sess.run(self.model.tvars)
return tvars
def update_tvars(self,tvars_update):
with self.sess.as_default():
feed_dict = dictionary = dict(zip(self.model.tvars_holders, tvars_update))
self.sess.run(self.model.update_tvar_holder, feed_dict)
# produce reward sentence_ebd average_reward
def produce_rldata(save_path):
with tf.Graph().as_default():
sess = tf.Session()
with sess.as_default():
# start = time.time()
interact = interaction(sess, save_path)
average_reward, all_sentence_ebd, all_reward = interact.produce_new_embedding()
np.save('data/average_reward.npy', average_reward)
np.save('data/all_sentence_ebd.npy', all_sentence_ebd)
np.save('data/all_reward.npy', all_reward)
print (average_reward)
if __name__ == '__main__':
# train model
print ('train model')
train('cnndata/cnn_train_word.npy', 'cnndata/cnn_train_pos1.npy', 'cnndata/cnn_train_pos2.npy','cnndata/cnn_train_y.npy','model/origin_cnn_model.ckpt')
# produce reward sentence_ebd average_reward for rlmodel
print ('produce reward sentence_ebd average_reward for rlmodel')
produce_rldata(save_path='model/origin_cnn_model.ckpt')