Text_Matching/model.py at master · Ricardokevins/Text_Matching · GitHub

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
import pandas as pd
import numpy as np
from sklearn.preprocessing import MinMaxScaler
import time
import copy
import torch.nn as nn
import torch.nn.functional as F
import torch.optim as optim
import torch.autograd as autograd
import torch.nn.functional
from torch.utils.data import Dataset, DataLoader
from torchvision import transforms
import warnings
import torch
import time
import get_data

USE_CUDA = torch.cuda.is_available()
class BiLSTM_Match(nn.Module):

    def __init__(self, embedding_dim, hidden_dim, vocab_size, tagset_size,batch_size,str_len):
        super(BiLSTM_Match, self).__init__()

        self.hidden_dim = hidden_dim
        self.str_len=str_len
        self.batch_size=batch_size
        self.word_embeddings = nn.Embedding(vocab_size, embedding_dim)
        self.lstma = nn.LSTM(embedding_dim, hidden_dim,bidirectional=True)
        self.lstmb = nn.LSTM(embedding_dim, hidden_dim,bidirectional=True)

        # self.dropout1 = nn.Dropout(0.5)
        self.densea= nn.Linear(2*str_len*hidden_dim, hidden_dim)
        self.denseb = nn.Linear(2*str_len * hidden_dim, hidden_dim)

        self.op_prelu=nn.PReLU()
        self.dropout2 = nn.Dropout(0.5)
        self.hidden2taga = nn.Linear(hidden_dim, tagset_size)
        self.hidden2tagb = nn.Linear(hidden_dim, tagset_size)

        self.op_tanh=nn.Tanh()
        self.hiddena = self.init_hidden()
        self.hiddenb = self.init_hidden()

    def init_hidden(self):
        if USE_CUDA:
            return (torch.zeros(2, self.batch_size, self.hidden_dim).cuda(),torch.zeros(2, self.batch_size, self.hidden_dim).cuda())
        else:
            return (torch.zeros(2, self.batch_size, self.hidden_dim),torch.zeros(2, self.batch_size, self.hidden_dim))

    def forward(self, sentencea , sentenceb , statea,stateb , train_flag):
        embedsa = self.word_embeddings(sentencea)
        embedsb = self.word_embeddings(sentenceb)
        #print(embeds.shape)
        self.hiddena=statea
        self.hiddenb = stateb

        lstm_outa, self.hiddena = self.lstma(embedsa.view(self.str_len, len(sentencea), -1), self.hiddena)
        lstm_outb, self.hiddenb = self.lstmb(embedsb.view(self.str_len, len(sentenceb), -1), self.hiddenb)

        tag_spacea = self.densea(lstm_outa.view(self.batch_size,-1))
        tag_spaceb = self.denseb(lstm_outb.view(self.batch_size, -1))

        tag_spacea = self.op_prelu(tag_spacea)
        tag_spaceb = self.op_prelu(tag_spaceb)

        if train_flag:
            tag_spacea = self.dropout2(tag_spacea)
            tag_spaceb = self.dropout2(tag_spaceb)

        tag_spacea=self.hidden2taga(tag_spacea)
        tag_spaceb = self.hidden2tagb(tag_spaceb)

        tag_spacea=self.op_tanh(tag_spacea)
        tag_spaceb = self.op_tanh(tag_spaceb)

        #similarity = torch.cosine_similarity(tag_spacea, tag_spaceb, dim=1)
        similarity=(F.pairwise_distance(tag_spacea, tag_spaceb, p=2)) # pytorch求欧氏距离
        similarity=similarity.view(-1,1)
        similarity=similarity.float()
        return similarity ,self.hiddena,self.hiddenb


class LSTM_Match(nn.Module):

    def __init__(self, embedding_dim, hidden_dim, vocab_size, tagset_size,batch_size,str_len):
        super(LSTM_Match, self).__init__()

        self.hidden_dim = hidden_dim
        self.str_len=str_len
        self.batch_size=batch_size

        self.word_embeddings = nn.Embedding(vocab_size, embedding_dim)
        self.lstma = nn.LSTM(embedding_dim, hidden_dim)
        self.lstmb = nn.LSTM(embedding_dim, hidden_dim)

        #self.dropout1 = nn.Dropout(0.5)
        self.densea= nn.Linear(str_len*hidden_dim, hidden_dim)
        self.denseb = nn.Linear(str_len * hidden_dim, hidden_dim)

        self.op_prelu=nn.PReLU()
        self.dropout2 = nn.Dropout(0.5)
        self.hidden2taga = nn.Linear(hidden_dim, tagset_size)
        self.hidden2tagb = nn.Linear(hidden_dim, tagset_size)

        self.op_tanh=nn.Tanh()
        self.hiddena = self.init_hidden()
        self.hiddenb = self.init_hidden()

    def init_hidden(self):
        if USE_CUDA:
            return (torch.zeros(1, self.batch_size, self.hidden_dim).cuda(),torch.zeros(1, self.batch_size, self.hidden_dim).cuda())
        else:
            return (torch.zeros(1, self.batch_size, self.hidden_dim),torch.zeros(1, self.batch_size, self.hidden_dim))

    def forward(self, sentencea , sentenceb , statea,stateb , train_flag):
        embedsa = self.word_embeddings(sentencea)
        embedsb = self.word_embeddings(sentenceb)
        #print(embeds.shape)
        self.hiddena = statea
        self.hiddenb = stateb

        lstm_outa, self.hiddena = self.lstma(embedsa.view(self.str_len, len(sentencea), -1), self.hiddena)
        lstm_outb, self.hiddenb = self.lstmb(embedsb.view(self.str_len, len(sentenceb), -1), self.hiddenb)

        if train_flag:
            lstm_outa=self.dropout1(lstm_outa)
            lstm_outb = self.dropout1(lstm_outb)

        tag_spacea = self.densea(lstm_outa.view(self.batch_size,-1))
        tag_spaceb = self.denseb(lstm_outb.view(self.batch_size, -1))

        tag_spacea = self.op_prelu(tag_spacea)
        tag_spaceb = self.op_prelu(tag_spaceb)
        if train_flag:
            tag_spacea = self.dropout2(tag_spacea)
            tag_spaceb = self.dropout2(tag_spaceb)

        tag_spacea = self.hidden2taga(tag_spacea)
        tag_spaceb = self.hidden2tagb(tag_spaceb)

        tag_spacea = self.op_tanh(tag_spacea)
        tag_spaceb = self.op_tanh(tag_spaceb)


        #similarity = torch.cosine_similarity(tag_spacea, tag_spaceb, dim=1)
        similarity=(F.pairwise_distance(tag_spacea, tag_spaceb, p=2)) # pytorch求欧氏距离
        similarity=similarity.view(-1,1)
        similarity=similarity.float()
        #print(tag_spacea)
        #print(tag_spaceb)

        #print(similarity)
        #print("t", similarity.shape)
        #tag_scores = F.log_softmax(tag_space,dim=1)
        #exit()
        return similarity ,self.hiddena,self.hiddenb

class EmbeddingModel(nn.Module):
    def __init__(self, vocab_size, embed_size):
        ''' 初始化输出和输出embedding
        '''
        super(EmbeddingModel, self).__init__()
        self.vocab_size = vocab_size
        self.embed_size = embed_size

        initrange = 0.5 / self.embed_size
        self.out_embed = nn.Embedding(self.vocab_size, self.embed_size, sparse=False)
        self.out_embed.weight.data.uniform_(-initrange, initrange)

        self.in_embed = nn.Embedding(self.vocab_size, self.embed_size, sparse=False)
        self.in_embed.weight.data.uniform_(-initrange, initrange)

    def forward(self, input_labels, pos_labels, neg_labels):
        '''
        input_labels: 中心词, [batch_size]
        pos_labels: 中心词周围 context window 出现过的单词 [batch_size * (window_size * 2)]
        neg_labelss: 中心词周围没有出现过的单词，从 negative sampling 得到 [batch_size, (window_size * 2 * K)]

        return: loss, [batch_size]
        '''

        batch_size = input_labels.size(0)

        input_embedding = self.in_embed(input_labels)  # B * embed_size
        pos_embedding = self.out_embed(pos_labels)  # B * (2*C) * embed_size
        neg_embedding = self.out_embed(neg_labels)  # B * (2*C * K) * embed_size

        log_pos = torch.bmm(pos_embedding, input_embedding.unsqueeze(2)).squeeze()  # B * (2*C)
        log_neg = torch.bmm(neg_embedding, -input_embedding.unsqueeze(2)).squeeze()  # B * (2*C*K)

        log_pos = F.logsigmoid(log_pos).sum(1)
        log_neg = F.logsigmoid(log_neg).sum(1)  # batch_size

        loss = log_pos + log_neg

        return -loss

    def input_embeddings(self):
        return self.in_embed.weight.data.cpu().numpy()