-
Notifications
You must be signed in to change notification settings - Fork 25
/
Copy pathSPSE_Training.py
166 lines (158 loc) · 7.77 KB
/
SPSE_Training.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
from __future__ import division
import numpy as np
import pickle
import random
import math
import sys
np.set_printoptions(threshold=np.nan)
#reload(sys)
#sys.setdefaultencoding("utf-8")
if (len(sys.argv)<5):
exit(0)
hownet_filename = sys.argv[1]
embedding_filename = sys.argv[2]
sememe_all_filename = sys.argv[3]
target_filename = sys.argv[4]
para_lambda = 0
max_iter = 20
with open(hownet_filename,'r',encoding='utf-8') as hownet:
with open(embedding_filename,'r',encoding='utf-8') as embedding_file:
with open(sememe_all_filename,'r',encoding='utf-8') as sememe_all:
sememes_buf = sememe_all.readlines()
sememes = sememes_buf[1].strip().strip('[]').split(' ')
sememes = [sememe.strip().strip('\'') for sememe in sememes]
sememe_size = len(sememes)
hownet_words = []
#read sememe complete
word2sememe = {}
while True:
word = hownet.readline().strip()
sememes_tmp = hownet.readline().strip().split()
if (word or sememes_tmp):
word2sememe[word] = []
hownet_words.append(word)
length = len(sememes_tmp)
for i in range(0,length):
word2sememe[word].append(sememes_tmp[i])
else: break
#read hownet complete
print("hownet reading complete")
line = embedding_file.readline()
arr = line.strip().split()
word_size = len(hownet_words)
dim_size = int(arr[1])
embedding_vec = {}
W = []
for line in embedding_file:
arr = line.strip().split()
float_arr = []
now_word = arr[0].strip()
if (now_word not in hownet_words):
continue
for i in range(1,dim_size+1):
float_arr.append(float(arr[i]))
regular = math.sqrt(sum([x*x for x in float_arr]))
word = arr[0].strip()
embedding_vec[word] = []
for i in range(1,dim_size+1):
embedding_vec[word].append(float(arr[i])/regular)
W.append(float(arr[i])/regular)
# sometimes, people use word_embeddings with fewer words than hownet
W = np.array(W).reshape(-1,dim_size)
word_size = len(embedding_vec)
#read embedding complete
print('Embedding reading complete')
with open('PMI.txt','r',encoding='utf-8') as PMI:
P = []
for line in PMI:
arr = line.strip().split()
arr = [float(e) for e in arr]
P.extend(arr)
P = np.array(P).reshape(sememe_size,sememe_size)
M = np.zeros((word_size,sememe_size))
se_index = 0
word_index = 0
for word in hownet_words:
if (word not in embedding_vec):
continue
try:
for sememe in word2sememe[word]:
se_index = sememes.index(sememe)
M[word_index][se_index] = 1
word_index += 1
except:
print(word)
sys.exit()
print("PMI calculating complete")
sememe_embedding = (np.random.randn(sememe_size*2,dim_size)-0.5) / dim_size
bias_sememe = (np.random.randn(sememe_size,1)-0.5) / dim_size
bias_word = (np.random.randn(word_size,1)-0.5) / dim_size
try:
print('Try to read from checkpoint')
target=open(target_filename,'rb')
sememe_embedding = pickle.load(target)
bias_word = pickle.load(target)
bias_sememe = pickle.load(target)
print('checkpoint reading complete')
target.close()
except:
print('checkpoint reading failed, initialize with random value')
with open(target_filename,'wb') as target:
sememe_embedding_dersum = np.ones((sememe_size*2,dim_size))
bias_sememe_dersum = np.ones((sememe_size,1))
bias_word_dersum = np.ones((word_size,1))
print('Initailization complete')
learning_rate = 0.01
for i in range(1,max_iter):
print("Process:%f" %(i/max_iter))
loss = 0
count = 0
for j in range(0,word_size):
for i in range(0,sememe_size):
sem0 = sememe_embedding[2 * i]
sem1 = sememe_embedding[2 * i + 1]
der = np.zeros((1,dim_size))
if (M[j][i] == 0):
rand = random.randint(1,1000)
if (rand>5):
continue
count += 1
w = W[j].reshape(1,dim_size)
delta = w.dot((sem0+sem1).transpose())+bias_sememe[i]+bias_word[j]-M[j][i]
loss += delta ** 2
der += delta * 2 * w
der = der.reshape(dim_size,)
sememe_embedding[2 * i] += -learning_rate * der / sememe_embedding_dersum[2 * i]
sememe_embedding[2 * i + 1] += -learning_rate * der / sememe_embedding_dersum[2 * i + 1]
sememe_embedding_dersum[2 * i] += der ** 2
sememe_embedding_dersum[2 * i + 1] += der ** 2
bias_word[j] += 2 * delta * learning_rate / bias_word_dersum[j]
bias_word_dersum[j] += 4 * delta ** 2
bias_sememe[i] += 2 * delta * learning_rate / bias_sememe_dersum[i]
bias_sememe_dersum[i] += 4 * delta ** 2
for j in range(0,sememe_size):
for i in range(0,sememe_size):
sem0 = sememe_embedding[2 * j]
sem1 = sememe_embedding[2 * i + 1]
der = np.zeros((1,dim_size))
der_out = np.zeros((1,dim_size))
if (P[j][i] == 0):
rand = random.randint(1,1000)
if (rand>5):
continue
count += 1
w = W[j].reshape(1,dim_size)
delta = sem0.dot((sem1).transpose())-P[j][i]
loss += para_lambda * delta ** 2
der += para_lambda * delta * 2 * sem0
der = der.reshape(dim_size,)
sememe_embedding[2 * i + 1] += -learning_rate * der / sememe_embedding_dersum[2 * i + 1]
sememe_embedding_dersum[2 * i + 1] += der ** 2
der_out += para_lambda * delta * 2 * sem1
der_out = der_out.reshape(dim_size,)
sememe_embedding[2 * j] += -learning_rate * der_out / sememe_embedding_dersum[2 * j]
sememe_embedding_dersum[2 * j] += der_out ** 2
print("loss:%f" %(loss / count,))
pickle.dump(sememe_embedding,target)
pickle.dump(bias_word,target)
pickle.dump(bias_sememe,target)