/
train_feedforward_simple.py
126 lines (110 loc) · 3.51 KB
/
train_feedforward_simple.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
import chainer
import numpy as np
import chainer.functions as F
from chainer import optimizers
def create_vocab():
vocab = dict()
for f in [train_file, test_file]:
for line in open(f):
for char in ''.join(line.strip().split()):
if char not in vocab:
vocab[char] = len(vocab)
vocab['<s>'] = len(vocab)
vocab['</s>'] = len(vocab)
return vocab
def init_model(vocab_size):
model = chainer.FunctionSet(
embed=F.EmbedID(vocab_size, embed_units),
hidden1=F.Linear(window * embed_units, hidden_units),
output=F.Linear(hidden_units, label_num),
)
opt = optimizers.AdaGrad(lr=learning_rate)
opt.setup(model)
return model, opt
def make_label(sent):
labels = list()
for char in sent:
if not char == ' ':
labels.append(1)
return labels
def train(char2id, model, optimizer):
for epoch in range(n_epoch):
batch_count = 0
accum_loss = 0
for line in open(train_file):
x = ''.join(line.strip().split())
t = make_label(line.strip())
for target in range(len(x)):
label = t[target]
pred, loss = forward_one(x, target, label)
accum_loss += loss
batch_count += 1
if batch_count == batch_size:
optimizer.zero_grads()
accum_loss.backward()
optimizer.update()
accum_loss = 0
batch_count = 0
if not batch_count == 0:
optimizer.zero_grads()
accum_loss.backward()
optimizer.update()
accum_loss = 0
batch_count = 0
def test(char2id, model, optimizer):
sum_accuracy = 0
sum_loss = 0
batch_count = 0
accum_loss = 0
labels = list()
for line in open(train_file):
x = ''.join(line.strip().split())
t = make_label(line.strip())
for target in range(len(x)):
label = t[target]
labels.append(label)
loss, acc = forward_one(x, target, label)
print (x)
print (labels)
print (t)
#sum_loss += float(loss.data) * len(y)
#sum_accuracy += float(acc.data) * len(y)
#print('test mean loss={}, accuracy={}'.format(
# sum_loss / N_test, sum_accuracy / N_test))
def forward_one(x, target, label):
# make input window vector
distance = window // 2
char_vecs = list()
x = list(x)
for i in range(distance):
x.append('</s>')
x.insert(0,'<s>')
for i in range(-distance, distance + 1):
char = x[target + i]
char_id = char2id[char]
char_vec = model.embed(get_onehot(char_id))
char_vecs.append(char_vec)
concat = F.concat(tuple(char_vecs))
hidden = model.hidden1(F.sigmoid(concat))
pred = model.output(hidden)
correct = get_onehot(label)
return np.argmax(pred), F.softmax_cross_entropy(pred, correct)
def get_onehot(num):
#return chainer.Variable(np.array([[num]], dtype=np.int32))
return chainer.Variable(np.array([num], dtype=np.int32))
def decode():
pass
if __name__ == '__main__':
train_file = '../data/train.txt'
test_file = '../data/test.txt'
window = 3
embed_units = 100
hidden_units = 50
label_num = 4
batch_size = 30
learning_rate = 0.1
n_epoch = 10
char2id = create_vocab()
model, opt = init_model(len(char2id))
train(char2id, model, opt)
test(char2id, model, opt)