/
transformer_AAI.py
217 lines (185 loc) · 8.4 KB
/
transformer_AAI.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
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
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
import torch
import torch.nn as nn
import torch.nn.functional as F
import os
from common.utils import mask_from_lens
import numpy as np
class PositionalEmbedding(nn.Module):
def __init__(self, demb):
super(PositionalEmbedding, self).__init__()
self.demb = demb
inv_freq = 1 / (10000 ** (torch.arange(0.0, demb, 2.0) / demb))
self.register_buffer('inv_freq', inv_freq)
def forward(self, pos_seq, bsz=None):
sinusoid_inp = torch.ger(pos_seq, self.inv_freq)
pos_emb = torch.cat([sinusoid_inp.sin(), sinusoid_inp.cos()], dim=1)
if bsz is not None:
return pos_emb[None, :, :].expand(bsz, -1, -1)
else:
return pos_emb[None, :, :]
class PositionwiseConvFF(nn.Module):
def __init__(self, d_model, d_inner, kernel_size, dropout, pre_lnorm=False):
super(PositionwiseConvFF, self).__init__()
self.d_model = d_model
self.d_inner = d_inner
self.dropout = dropout
self.CoreNet = nn.Sequential(
nn.Conv1d(d_model, d_inner, kernel_size, 1, (kernel_size // 2)),
nn.ReLU(),
# nn.Dropout(dropout), # worse convergence
nn.Conv1d(d_inner, d_model, kernel_size, 1, (kernel_size // 2)),
nn.Dropout(dropout),
)
self.layer_norm = nn.LayerNorm(d_model)
self.pre_lnorm = pre_lnorm
def forward(self, inp):
return self._forward(inp)
def _forward(self, inp):
if self.pre_lnorm:
# layer normalization + positionwise feed-forward
core_out = inp.transpose(1, 2)
core_out = self.CoreNet(self.layer_norm(core_out))
core_out = core_out.transpose(1, 2)
# residual connection
output = core_out + inp
else:
# positionwise feed-forward
core_out = inp.transpose(1, 2)
core_out = self.CoreNet(core_out)
core_out = core_out.transpose(1, 2)
# residual connection + layer normalization
output = self.layer_norm(inp + core_out)
return output
class MultiHeadAttn(nn.Module):
def __init__(self, n_head, d_model, d_head, dropout, dropatt=0.1,
pre_lnorm=False, relative=None):
super(MultiHeadAttn, self).__init__()
self.n_head = n_head
self.d_model = d_model
self.d_head = d_head
self.scale = 1 / (d_head ** 0.5)
self.pre_lnorm = pre_lnorm
self.qkv_net = nn.Linear(d_model, 3 * n_head * d_head)
self.drop = nn.Dropout(dropout)
self.dropatt = nn.Dropout(dropatt)
self.o_net = nn.Linear(n_head * d_head, d_model, bias=False)
self.layer_norm = nn.LayerNorm(d_model)
self.k = relative
self.val_rel = relative
self.training = True
seq_len = 400
if relative is not None:
self.embed_idx = torch.tensor(np.array([max(-self.k, min(self.k, j-i)) for i in range(seq_len) for j in range(seq_len)]), device=device).view(-1, seq_len, seq_len).long().squeeze()+self.k
self.relative_embed = nn.Embedding(self.k*2+1, 8).float().to(device)
# self.relative_embed2 = nn.Embedding(self.k*2+1, 8).float().to(device)
# padding_idx=self.padding_idx).
def forward(self, inp, attn_mask=None, return_attn=False):
return self._forward(inp, attn_mask, return_attn=return_attn)
def _forward(self, inp, attn_mask=None, return_attn=False):
residual = inp
if self.pre_lnorm:
# layer normalization
inp = self.layer_norm(inp)
n_head, d_head = self.n_head, self.d_head
head_q, head_k, head_v = torch.chunk(self.qkv_net(inp), 3, dim=-1)
head_q = head_q.view(inp.size(0), inp.size(1), n_head, d_head)
head_k = head_k.view(inp.size(0), inp.size(1), n_head, d_head)
head_v = head_v.view(inp.size(0), inp.size(1), n_head, d_head)
q = head_q.permute(0, 2, 1, 3).reshape(-1, inp.size(1), d_head)
k = head_k.permute(0, 2, 1, 3).reshape(-1, inp.size(1), d_head)
v = head_v.permute(0, 2, 1, 3).reshape(-1, inp.size(1), d_head)
attn_score = torch.bmm(q, k.transpose(1, 2))
attn_score.mul_(self.scale)
if self.k is not None:
seq_len = inp.shape[1]
if not self.training:
embed_idx = torch.from_numpy(np.array([max(-self.k, min(self.k, j-i)) for i in range(seq_len) for j in range(seq_len)])).view(-1, seq_len, seq_len).long().to(device).squeeze()+self.k
else:
embed_idx = self.embed_idx
rel_pos_vec = self.relative_embed(embed_idx)
rel_pos_vec = rel_pos_vec.permute(0, 2, 1)
q = q.permute(1, 0, 2)
rel_score = torch.matmul(q, rel_pos_vec).permute(1, 0, 2)
attn_score += rel_score
if attn_mask is not None:
attn_mask = attn_mask.unsqueeze(1)
attn_mask = attn_mask.repeat(n_head, attn_mask.size(2), 1)
attn_score.masked_fill_(attn_mask, -float('inf'))
attn_prob = F.softmax(attn_score, dim=2)
attn_prob = self.dropatt(attn_prob)
attn_vec = torch.bmm(attn_prob, v)
attn_vec = attn_vec.view(n_head, inp.size(0), inp.size(1), d_head)
attn_vec = attn_vec.permute(1, 2, 0, 3).contiguous().view(
inp.size(0), inp.size(1), n_head * d_head)
# linear projection
attn_out = self.o_net(attn_vec)
attn_out = self.drop(attn_out)
if self.pre_lnorm:
# residual connection
output = residual + attn_out
else:
# residual connection + layer normalization
output = self.layer_norm(residual + attn_out)
if return_attn:
return output, attn_prob
return output
class TransformerLayer(nn.Module):
def __init__(self, n_head, d_model, d_head, d_inner, kernel_size, dropout, relative
**kwargs):
super(TransformerLayer, self).__init__()
self.dec_attn = MultiHeadAttn(n_head, d_model, d_head, dropout, relative=relative, **kwargs)
self.pos_ff = PositionwiseConvFF(d_model, d_inner, kernel_size, dropout,
pre_lnorm=kwargs.get('pre_lnorm'))
def forward(self, dec_inp, mask=None, return_attn=False):
output = self.dec_attn(dec_inp, attn_mask=None, return_attn=return_attn)
if return_attn:
output, attn = output
output = self.pos_ff(output)
if return_attn:
return output, attn
else:
return output
class FFTransformer(nn.Module):
def __init__(self, n_layer, n_head, d_model, d_head, d_inner, kernel_size,
dropout, dropatt, dropemb=0.0, embed_input=True,
n_embed=None, d_embed=None, padding_idx=0, pre_lnorm=False,
pos_embed_type=None):
super(FFTransformer, self).__init__()
self.d_model = d_model
self.n_head = n_head
self.d_head = d_head
self.padding_idx = 40
self.word_emb = None
if pos_embed_type == 'concat':
self.pos_emb = PositionalEmbedding(32)
elif pos_embed_type == 'additive':
self.pos_emb = PositionalEmbedding(d_model)
self.drop = nn.Dropout(dropemb)
self.layers = nn.ModuleList()
self.ln = nn.LayerNorm(d_model)
self.attn = []
if pos_embed_type == 'relative':
relative = 10
else:
relative = None
for _ in range(n_layer):
self.layers.append(
TransformerLayer(
n_head, d_model, d_head, d_inner, kernel_size, dropout, relative
dropatt=dropatt, pre_lnorm=pre_lnorm)
)
def forward(self, dec_inp, seq_lens=None,pos_embed_type=None):
out = dec_inp
if pos_embed_type == 'additive':
pos_seq = torch.arange(out.size(1), device=out.device, dtype=out.dtype)
pos_emb = self.pos_emb(pos_seq)
pos_emb = torch.cat(out.shape[0]*[pos_emb])
out = self.ln(out) + pos_emb
elif pos_embed_type == 'concat':
pos_seq = torch.arange(out.size(1), device=out.device, dtype=out.dtype)
pos_emb = self.pos_emb(pos_seq)
pos_emb = torch.cat(out.shape[0]*[pos_emb])
out = torch.cat([out, pos_emb], dim=-1)
for idx, layer in enumerate(self.layers):
out = layer(out, mask=None)
return out