-
Notifications
You must be signed in to change notification settings - Fork 19
/
model.py
142 lines (117 loc) · 5.73 KB
/
model.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
import tensorflow as tf
import tensorflow.contrib.rnn as rnn
from data_utils import batch_review_normalize, batch_image_normalize
from layers import bidirectional_rnn, text_attention, visual_aspect_attention
from model_utils import get_shape, load_glove
from data_preprocess import VOCAB_SIZE
class VistaNet:
def __init__(self, hidden_dim, att_dim, emb_size, num_images, num_classes):
self.hidden_dim = hidden_dim
self.att_dim = att_dim
self.emb_size = emb_size
self.num_classes = num_classes
self.num_images = num_images
self.global_step = tf.Variable(0, name='global_step', trainable=False)
self.dropout_keep_prob = tf.placeholder(dtype=tf.float32, name='dropout_keep_prob')
self.documents = tf.placeholder(shape=(None, None, None), dtype=tf.int32, name='reviews')
self.document_lengths = tf.placeholder(shape=(None,), dtype=tf.int32, name='review_lengths')
self.sentence_lengths = tf.placeholder(shape=(None, None), dtype=tf.int32, name='sentence_lengths')
self.max_num_words = tf.placeholder(dtype=tf.int32, name='max_num_words')
self.max_num_sents = tf.placeholder(dtype=tf.int32, name='max_num_sents')
self.images = tf.placeholder(shape=(None, None, 4096), dtype=tf.float32, name='images')
self.labels = tf.placeholder(shape=(None), dtype=tf.int32, name='labels')
with tf.variable_scope('VistaNet'):
self._init_embedding()
self._init_word_encoder()
self._init_sent_encoder()
self._init_classifier()
def _init_embedding(self):
with tf.variable_scope('embedding'):
self.embedding_matrix = tf.get_variable(
name='embedding_matrix',
shape=[VOCAB_SIZE, self.emb_size],
initializer=tf.constant_initializer(load_glove(VOCAB_SIZE, self.emb_size)),
dtype=tf.float32
)
self.embedded_inputs = tf.nn.embedding_lookup(self.embedding_matrix, self.documents)
def _init_word_encoder(self):
with tf.variable_scope('word') as scope:
word_rnn_inputs = tf.reshape(
self.embedded_inputs,
[-1, self.max_num_words, self.emb_size]
)
sentence_lengths = tf.reshape(self.sentence_lengths, [-1])
# word encoder
cell_fw = rnn.GRUCell(self.hidden_dim)
cell_bw = rnn.GRUCell(self.hidden_dim)
init_state_fw = tf.tile(tf.get_variable('init_state_fw',
shape=[1, self.hidden_dim],
initializer=tf.constant_initializer(1.0)),
multiples=[get_shape(word_rnn_inputs)[0], 1])
init_state_bw = tf.tile(tf.get_variable('init_state_bw',
shape=[1, self.hidden_dim],
initializer=tf.constant_initializer(1.0)),
multiples=[get_shape(word_rnn_inputs)[0], 1])
word_rnn_outputs, _ = bidirectional_rnn(
cell_fw=cell_fw,
cell_bw=cell_bw,
inputs=word_rnn_inputs,
input_lengths=sentence_lengths,
initial_state_fw=init_state_fw,
initial_state_bw=init_state_bw,
scope=scope
)
self.word_outputs, self.word_att_weights = text_attention(inputs=word_rnn_outputs,
att_dim=self.att_dim,
sequence_lengths=sentence_lengths)
self.word_outputs = tf.nn.dropout(self.word_outputs, keep_prob=self.dropout_keep_prob)
def _init_sent_encoder(self):
with tf.variable_scope('sentence') as scope:
sentence_rnn_inputs = tf.reshape(self.word_outputs, [-1, self.max_num_sents, 2 * self.hidden_dim])
# sentence encoder
cell_fw = rnn.GRUCell(self.hidden_dim)
cell_bw = rnn.GRUCell(self.hidden_dim)
init_state_fw = tf.tile(tf.get_variable('init_state_fw',
shape=[1, self.hidden_dim],
initializer=tf.constant_initializer(1.0)),
multiples=[get_shape(sentence_rnn_inputs)[0], 1])
init_state_bw = tf.tile(tf.get_variable('init_state_bw',
shape=[1, self.hidden_dim],
initializer=tf.constant_initializer(1.0)),
multiples=[get_shape(sentence_rnn_inputs)[0], 1])
sentence_rnn_outputs, _ = bidirectional_rnn(
cell_fw=cell_fw,
cell_bw=cell_bw,
inputs=sentence_rnn_inputs,
input_lengths=self.document_lengths,
initial_state_fw=init_state_fw,
initial_state_bw=init_state_bw,
scope=scope
)
self.sentence_outputs, self.sent_att_weights, self.img_att_weights = visual_aspect_attention(
text_input=sentence_rnn_outputs,
visual_input=self.images,
att_dim=self.att_dim,
sequence_lengths=self.document_lengths
)
self.sentence_outputs = tf.nn.dropout(self.sentence_outputs, keep_prob=self.dropout_keep_prob)
def _init_classifier(self):
with tf.variable_scope('classifier'):
self.logits = tf.layers.dense(
inputs=self.sentence_outputs,
units=self.num_classes,
name='logits'
)
def get_feed_dict(self, reviews, images, labels, dropout_keep_prob=1.0):
norm_docs, doc_sizes, sent_sizes, max_num_sents, max_num_words = batch_review_normalize(reviews)
fd = {
self.documents: norm_docs,
self.document_lengths: doc_sizes,
self.sentence_lengths: sent_sizes,
self.max_num_sents: max_num_sents,
self.max_num_words: max_num_words,
self.images: batch_image_normalize(images, self.num_images),
self.labels: labels,
self.dropout_keep_prob: dropout_keep_prob
}
return fd