-
Notifications
You must be signed in to change notification settings - Fork 0
/
transformer_model.py
152 lines (131 loc) · 5.58 KB
/
transformer_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
143
144
145
146
147
148
149
150
151
152
import torch
import torch.nn as nn
import torch.nn.functional as F
import math
class Transformer(nn.Module):
def __init__(self, input_dim, output_dim, d_model, n_layers, n_heads, dropout=0.1):
super(Transformer, self).__init__()
self.encoder = Encoder(input_dim, d_model, n_layers, n_heads, dropout)
self.decoder = Decoder(output_dim, d_model, n_layers, n_heads, dropout)
self.linear = nn.Linear(d_model, output_dim)
def forward(self, src, trg):
enc_output = self.encoder(src)
dec_output = self.decoder(trg, enc_output)
output = self.linear(dec_output)
return output
class Encoder(nn.Module):
def __init__(self, input_dim, d_model, n_layers, n_heads, dropout):
super(Encoder, self).__init__()
self.embedding = nn.Linear(input_dim, d_model)
self.pos_encoding = PositionalEncoding(d_model, dropout)
self.layers = nn.ModuleList([EncoderLayer(d_model, n_heads, dropout) for _ in range(n_layers)])
def forward(self, src):
# x = self.embedding(src.float())
x = self.embedding(src)
x = self.pos_encoding(x)
for layer in self.layers:
x = layer(x)
return x
class Decoder(nn.Module):
def __init__(self, output_dim, d_model, n_layers, n_heads, dropout):
super(Decoder, self).__init__()
self.embedding = nn.Linear(output_dim, d_model)
self.pos_encoding = PositionalEncoding(d_model, dropout)
self.layers = nn.ModuleList([DecoderLayer(d_model, n_heads, dropout) for _ in range(n_layers)])
def forward(self, trg, enc_output):
x = self.embedding(trg)
x = self.pos_encoding(x)
for layer in self.layers:
x = layer(x, enc_output)
return x
class EncoderLayer(nn.Module):
def __init__(self, d_model, n_heads, dropout):
super(EncoderLayer, self).__init__()
self.self_attn = MultiHeadAttention(d_model, n_heads)
self.feed_forward = FeedForward(d_model)
self.layer_norm1 = nn.LayerNorm(d_model)
self.layer_norm2 = nn.LayerNorm(d_model)
self.dropout = nn.Dropout(dropout)
def forward(self, src):
src2 = self.self_attn(src, src, src)
src = src + self.dropout(src2)
src = self.layer_norm1(src)
src2 = self.feed_forward(src)
src = src + self.dropout(src2)
src = self.layer_norm2(src)
return src
class DecoderLayer(nn.Module):
def __init__(self, d_model, n_heads, dropout):
super(DecoderLayer, self).__init__()
self.self_attn = MultiHeadAttention(d_model, n_heads)
self.enc_dec_attn = MultiHeadAttention(d_model, n_heads)
self.feed_forward = FeedForward(d_model)
self.layer_norm1 = nn.LayerNorm(d_model)
self.layer_norm2 = nn.LayerNorm(d_model)
self.layer_norm3 = nn.LayerNorm(d_model)
self.dropout = nn.Dropout(dropout)
def forward(self, trg, enc_output):
trg2 = self.self_attn(trg, trg, trg)
trg = trg + self.dropout(trg2)
trg = self.layer_norm1(trg)
trg2 = self.enc_dec_attn(trg, enc_output, enc_output)
trg = trg + self.dropout(trg2)
trg = self.layer_norm2(trg)
trg2 = self.feed_forward(trg)
trg = trg + self.dropout(trg2)
trg = self.layer_norm3(trg)
return trg
class MultiHeadAttention(nn.Module):
def __init__(self, d_model, n_heads):
super(MultiHeadAttention, self).__init__()
self.d_model = d_model
self.n_heads = n_heads
self.head_dim = d_model // n_heads
assert d_model % n_heads == 0
self.fc_q = nn.Linear(d_model, d_model)
self.fc_k = nn.Linear(d_model, d_model)
self.fc_v = nn.Linear(d_model, d_model)
self.fc_o = nn.Linear(d_model, d_model)
self.dropout = nn.Dropout(0.1)
def forward(self, query, key, value):
batch_size = query.shape[0]
Q = self.fc_q(query)
K = self.fc_k(key)
V = self.fc_v(value)
def split_heads(x, batch_size):
x = x.view(batch_size, -1, self.n_heads, self.head_dim)
return x.permute(0, 2, 1, 3)
Q = split_heads(Q, batch_size)
K = split_heads(K, batch_size)
V = split_heads(V, batch_size)
energy = torch.matmul(Q, K.permute(0, 1, 3, 2)) / torch.sqrt(torch.tensor(self.head_dim).float())
attention = F.softmax(energy, dim=-1)
x = torch.matmul(self.dropout(attention), V)
x = x.permute(0, 2, 1, 3).contiguous().view(batch_size, -1, self.d_model)
x = self.fc_o(x)
return x
class FeedForward(nn.Module):
def __init__(self, d_model, d_ff=2048, dropout=0.1):
super(FeedForward, self).__init__()
self.fc1 = nn.Linear(d_model, d_ff)
self.fc2 = nn.Linear(d_ff, d_model)
self.dropout = nn.Dropout(dropout)
def forward(self, x):
x = F.relu(self.fc1(x))
x = self.dropout(x)
x = self.fc2(x)
return x
class PositionalEncoding(nn.Module):
def __init__(self, d_model, dropout, max_len=5000):
super(PositionalEncoding, self).__init__()
self.dropout = nn.Dropout(p=dropout)
pe = torch.zeros(max_len, d_model)
position = torch.arange(0, max_len, dtype=torch.float).unsqueeze(1)
div_term = torch.exp(torch.arange(0, d_model, 2).float() * (-math.log(10000.0) / d_model))
pe[:, 0::2] = torch.sin(position * div_term)
pe[:, 1::2] = torch.cos(position * div_term)
pe = pe.unsqueeze(0).transpose(0, 1)
self.register_buffer('pe', pe)
def forward(self, x):
x = x + self.pe[:x.size(0), :]
return self.dropout(x)