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
|
"""Local attention module.
Also stolen from lucidrains from here:
https://github.com/lucidrains/local-attention/blob/master/local_attention/local_attention.py
"""
from functools import reduce
from operator import mul
from typing import Optional, Tuple
import attr
from einops import rearrange
import torch
from torch import einsum
from torch import nn
from torch import Tensor
import torch.nn.functional as F
from text_recognizer.networks.transformer.attention import apply_rotary_emb
@attr.s(eq=False)
class LocalAttention(nn.Module):
dim: int = attr.ib()
dim_head: int = attr.ib(default=64)
window_size: int = attr.ib(default=128)
look_back: int = attr.ib(default=1)
dropout_rate: float = attr.ib(default=0.0)
def __attrs_pre_init__(self) -> None:
super().__init__()
def __attrs_post_init__(self) -> None:
"""Post init configuration."""
self.scale = self.dim ** -0.5
inner_dim = self.dim * self.dim_head
self.query = nn.Linear(self.dim, inner_dim, bias=False)
self.key = nn.Linear(self.dim, inner_dim, bias=False)
self.value = nn.Linear(self.dim, inner_dim, bias=False)
self.dropout = nn.Dropout(p=self.dropout_rate)
# Feedforward
self.fc = nn.Linear(inner_dim, self.dim)
def forward(
self,
x: Tensor,
mask: Optional[Tensor] = None,
rotary_pos_emb: Optional[Tensor] = None,
) -> Tuple[Tensor, Tensor]:
b, n, _, device, dtype = *x.shape, x.device, x.dtype
assert (
n % self.window_size
), f"Sequence length {n} must be divisable with window size {self.window_size}"
q = self.query(x)
k = self.key(x)
v = self.value(x)
q, k, v = map(
lambda t: rearrange(t, "b n (h d) -> b h n d", h=self.num_heads), (q, k, v)
)
q, k, v = (
apply_rotary_emb(q, k, v, rotary_pos_emb)
if rotary_pos_emb is not None
else (q, k, v,)
)
num_windows = n // self.window_size
# Compute buckets
b_n = torch.arange(n).type_as(q).reshape(1, num_windows, self.window_size)
bq, bk, bv = map(
lambda t: t.reshape(b, num_windows, self.window_size, -1), (q, k, v)
)
bk = look_around(bk, backward=self.backward)
bv = look_around(bv, backward=self.backward)
bq_k = look_around(b_n, backward=self.backward)
# Compute the attention.
energy = einsum("b h i d, b h j d -> b h i j", bq, bk) * self.scale
mask_value = -torch.finfo(energy.dtype).max
# Causal mask.
causal_mask = b_n[:, :, :, None] < bq_k[:, :, None, :]
energy = energy.masked_fill_(causal_mask, mask_value)
del causal_mask
bucket_mask = bq_k[:, :, None, :] == -1
energy.masked_fill_(bucket_mask, mask_value)
del bucket_mask
energy = apply_input_mask(
b,
energy=energy,
mask=mask,
backward=self.backward,
window_size=self.window_size,
num_windows=num_windows,
mask_value=mask_value,
)
attn = F.softmax(energy, dim=-1)
attn = self.dropout(attn)
out = einsum("b h i j, b h j d -> b h i d", attn, bv)
out = rearrange(out, "b h n d -> b n (h d)")
out = self.fc(out)
return out, attn
def merge_dims(ind_from, ind_to, tensor):
shape = list(tensor.shape)
arr_slice = slice(ind_from, ind_to + 1)
shape[arr_slice] = [reduce(mul, shape[arr_slice])]
return tensor.reshape(*shape)
def expand_dim(t, dim, k, unsqueeze=True):
if unsqueeze:
t = t.unsqueeze(dim)
expand_shape = [-1] * len(t.shape)
expand_shape[dim] = k
return t.expand(*expand_shape)
def look_around(x: Tensor, backward: int, pad_value: int = -1, dim: int = 2) -> Tensor:
n = x.shape[1]
dims = (len(x.shape) - dim) * (0, 0)
x_pad = F.pad(x, (*dims, backward, 0), value=pad_value)
tensors = [x_pad[:, ind : (ind + n), ...] for ind in range(backward + 1)]
return torch.cat(tensors, dim=dim)
def apply_input_mask(
b: int,
energy: Tensor,
mask: Tensor,
backward: int,
window_size: int,
num_windows: int,
mask_value: Tensor,
) -> Tensor:
h = b // mask.shape[0]
mask = mask.reshape(-1, window_size, num_windows)
mq = mk = mask
mk = look_around(mk, pad_value=False, backward=backward)
mask = mq[:, :, :, None] * mk[:, :, None, :]
mask = merge_dims(0, 1, expand_dim(mask, 1, h))
energy.masked_fill_(~mask, mask_value)
del mask
return energy
|