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
|
"""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
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.embeddings.rotary import (
RotaryEmbedding,
rotate_half,
)
@attr.s(eq=False)
class LocalAttention(nn.Module):
"""Local windowed attention."""
dim: int = attr.ib()
num_heads: 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)
rotary_embedding: Optional[RotaryEmbedding] = attr.ib(default=None)
def __attrs_pre_init__(self) -> None:
"""Pre init constructor."""
super().__init__()
def __attrs_post_init__(self) -> None:
"""Post init constructor."""
self.scale = self.dim ** -0.5
inner_dim = self.num_heads * self.dim_head
self.to_qkv = nn.Linear(self.dim, 3 * inner_dim, bias=False)
self.dropout = nn.Dropout(p=self.dropout_rate)
# Feedforward
self.fc = nn.Linear(inner_dim, self.dim)
def _to_embeddings(self, x: Tensor) -> Tuple[Tensor, Tensor, Tensor]:
"""Convert input into query, key, and value."""
def _split_heads(t: Tensor) -> Tensor:
return _reshape_dim(t, -1, (-1, self.dim_head)).transpose(1, 2).contiguous()
def _merge_into_batch(t: Tensor) -> Tensor:
return t.reshape(-1, *t.shape[-2:])
qkv = self.to_qkv(x).chunk(3, dim=-1)
q, k, v = map(_split_heads, qkv)
q, k, v = map(_merge_into_batch, (q, k, v))
if self.rotary_embedding is not None:
embedding = self.rotary_embedding(q)
q, k = _apply_rotary_emb(q, k, embedding)
return q, k, v
def _create_buckets(
self, q: Tensor, k: Tensor, v: Tensor, n: int, b: int, num_windows: int
) -> Tuple[Tensor, Tensor, Tensor, Tensor]:
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.look_back)
bv = look_around(bv, backward=self.look_back)
bq_k = look_around(b_n, backward=self.look_back)
return b_n, bq, bk, bv, bq_k
def _apply_masks(
self,
b: int,
energy: Tensor,
b_n: Tensor,
bq_k: Tensor,
mask: Tensor,
num_windows: int,
) -> Tensor:
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.look_back,
window_size=self.window_size,
num_windows=num_windows,
mask_value=mask_value,
)
return energy
def forward(
self, x: Tensor, mask: Optional[Tensor] = None,
) -> Tuple[Tensor, Tensor]:
"""Computes windowed attention."""
b, n, _ = x.shape
if not n % self.window_size:
RuntimeError(
f"Sequence length {n} must be divisable with window size {self.window_size}"
)
num_windows = n // self.window_size
q, k, v = self._to_embeddings(x)
d = q.shape[-1]
# Compute buckets
b_n, bq, bk, bv, bq_k = self._create_buckets(q, k, v, n, b, num_windows)
# Compute the attention.
energy = einsum("b h i d, b h j d -> b h i j", bq, bk) * self.scale
energy = self._apply_masks(b, energy, b_n, bq_k, mask, num_windows)
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 = out.reshape(-1, n, d)
out = out.reshape(b, self.num_heads, n, -1).transpose(1, 2).reshape(b, n, -1)
out = self.fc(out)
return out, attn
def _apply_rotary_emb(q: Tensor, k: Tensor, freqs: Tensor) -> Tuple[Tensor, Tensor]:
q, k = map(lambda t: (t * freqs.cos()) + (rotate_half(t) * freqs.sin()), (q, k))
return q, k
def _reshape_dim(t: Tensor, dim: int, split_dims: Tuple[int, int]) -> Tensor:
shape = list(t.shape)
dims = len(t.shape)
dim = (dim + dims) % dims
shape[dim : dim + 1] = split_dims
return t.reshape(shape)
def merge_dims(ind_from: int, ind_to: int, tensor: Tensor) -> Tensor:
"""Merge dimensions."""
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: Tensor, dim: int, k: int, unsqueeze: bool = True) -> Tensor:
"""Expand tensors dimensions."""
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:
"""Apply windowing."""
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:
"""Applies input mask to energy tensor."""
h = b // mask.shape[0]
mask = mask.reshape(-1, num_windows, window_size)
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
|
