Language-generation-project/CNNModel.py at master · mRSun15/Language-generation-project · 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
import torch
import torch.nn as nn
import torch.legacy.nn as luann
import sys

class MaxPool(nn.Module):
    def __init__(self, dim=1):
        super(MaxPool, self).__init__()
        self.dim = dim

    def forward(self, input):
        return torch.max(input, self.dim)[0]

    def __repr__(self):
        return self.__class__.__name__ +'('+ 'dim=' + str(self.dim) + ')'

class View(nn.Module):
    def __init__(self, *sizes):
        super(View, self).__init__()
        self.sizes_list = sizes

    def forward(self, input):
        return input.view(*self.sizes_list)

    def __repr__(self):
        return self.__class__.__name__ + ' (' \
               + 'sizes=' + str(self.sizes_list) + ')'

class Transpose(nn.Module):
    def __init__(self, dim1=0, dim2=1):
        super(Transpose, self).__init__()
        self.dim1 = dim1
        self.dim2 = dim2

    def forward(self, input):
        return input.transpose(self.dim1, self.dim2).contiguous()

    def __repr__(self):
        return self.__class__.__name__ + ' (' \
               + 'between=' + str(self.dim1) + ',' + str(self.dim2) + ')'

class CNNModel(nn.Module):
    def __init__(self, vocab_size, num_labels, emb_size, w_hid_size, h_hid_size, win, batch_size,with_proj=False):
        super(CNNModel, self).__init__()

        self.model = nn.Sequential()
        self.model.add_module('transpose', Transpose())
        self.embed = nn.Embedding(num_embeddings=vocab_size, embedding_dim=emb_size)
        self.model.add_module('emb', self.embed)
        if with_proj:
            self.model.add_module('view1', View(-1, emb_size))
            self.model.add_module('linear1', nn.Linear(emb_size, w_hid_size))
            self.model.add_module('relu1', nn.ReLU())
        else:
            w_hid_size = emb_size

        self.model.add_module('trans2', Transpose(1, 2))

        conv_nn = nn.Conv1d(w_hid_size, h_hid_size, win, padding=1)
        self.model.add_module('conv', conv_nn)
        self.model.add_module('relu2', nn.ReLU())

        self.model.add_module('max', MaxPool(2))

        self.model.add_module('view4', View(-1, h_hid_size))
        self.model.add_module('linear2', nn.Linear(h_hid_size, num_labels))
        self.model.add_module('softmax', nn.LogSoftmax())


    def forward(self, x):

        output = self.model.forward(x)

        return output


# class CNNModel(nn.Module):
#     def __init__(self, vocab_size, num_labels, emb_size, w_hid_size, h_hid_size, win, batch_size, with_proj=False):
#         super(CNNModel, self).__init__()

#         self.model = nn.Sequential()
#         self.model.add_module('transpose', Transpose())
#         self.embed = nn.Embedding(num_embeddings=vocab_size, embedding_dim=emb_size)
#         self.model.add_module('emb', self.embed)
#         if with_proj:
#             self.model.add_module('view1', View(-1, emb_size))
#             self.model.add_module('linear1', nn.Linear(emb_size, w_hid_size))
#             # self.model.add_module('view2', View(batch_size, w_hid_size, -1))
#             self.model.add_module('relu1', nn.ReLU())
#         else:
#             w_hid_size = emb_size
#             # self.model.add_module('view2', View(batch_size, w_hid_size, -1))

#         self.model.add_module('trans2', Transpose(1, 2))

#         conv_nn = nn.Conv1d(w_hid_size, h_hid_size, win, padding=1)
#         self.model.add_module('conv', conv_nn)
#         self.model.add_module('relu2', nn.ReLU())

#         # # self.model.add_module('view3', View(batch_size, -1, h_hid_size))
#         # self.model.add_module('trans3', Transpose(1, 2))
#         # self.model.add_module('max', MaxPool(1))

#         # new implementation
#         self.model.add_module('max', MaxPool(2))

#         # old implementation
#         # self.model.add_module('transpose2', Transpose(1, 2))
#         # # self.model.add_module('view3', View(batch_size, -1, h_hid_size))
#         # self.model.add_module('max', MaxPool(1))

#         self.model.add_module('view4', View(batch_size, h_hid_size))
#         self.model.add_module('linear2', nn.Linear(h_hid_size, num_labels))
#         # m = nn.LogSoftmax()
#         self.model.add_module('softmax', nn.LogSoftmax())
#         # model:add(nn.Max(2))

#         # model:add(nn.Linear(opt.numFilters, opt.hiddenDim))
#         # model:add(nn.ReLU())
#         # if opt.dropout > 0:
#         #     model:add(nn.Dropout(opt.dropout))

#         # self.model2 = nn.Sequential()
#         # self.model2.add_module('linear2', nn.Linear(h_hid_size, num_labels))
#         # self.model2.add_module('softmax', nn.LogSoftMax())
#         # # Criterion
#         # self.criterion = nn.ClassNLLCriterion()

#     def forward(self, x):

#         # output = self.lookupTable.forward(x)
#         # output = x
#         # for i in range(9):
#         #     output = self.model[i].forward(output)
#         #     print output.size()
#         # sys.stdin.readline()
#         # output = self.model[1].forward(output)
#         # print output.size()
#         # output = self.model[2].forward(output)
#         # print output.size()
#         # output = self.model[3].forward(output)
#         # print output.size()
#         # output = self.model[4].forward(output)
#         # print output.size()
#         # output = self.model[5].forward(output)
#         # print output.size()

#         output = self.model.forward(x)
#         # output = torch.max(output, 1)[0]
#         # output = self.model2.forward(output)
#         return output

#     def num_flat_features(self, x):
#         size = x.size()[1:] # all dimensions except the batch dimension
#         num_features = 1
#         for s in size:
#             num_features *= s
#         return num_features