File size: 11,418 Bytes
878264b |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 |
# Copyright (c) Facebook, Inc. and its affiliates.
#
# This source code is licensed under the MIT license found in the
# LICENSE file in the root directory of this source tree.
from typing import Optional, Tuple
import torch
from torch import Tensor
from torch.nn.functional import (
linear, softmax, dropout, pad,
has_torch_function,
handle_torch_function,
_in_projection_packed,
)
import math
import warnings
def _scaled_dot_product_attention(
q: Tensor,
k: Tensor,
v: Tensor,
attn_mask: Optional[Tensor] = None,
dropout_p: float = 0.0,
bsz: int = 1,
subset_heads: Optional[Tensor] = None,
subset_weights: Optional[Tensor] = None,
) -> Tuple[Tensor, Tensor]:
B, Nt, E = q.shape
q = q / math.sqrt(E)
# B: bsz * total_num_heads
# (B, Nt, E) x (B, E, Ns) -> (B, Nt, Ns)
attn = torch.bmm(q, k.transpose(-2, -1))
if attn_mask is not None:
attn += attn_mask
attn = softmax(attn, dim=-1)
if dropout_p > 0.0:
attn = dropout(attn, p=dropout_p)
if subset_heads is None:
# (B, Nt, Ns) x (B, Ns, E) -> (B, Nt, E)
output = torch.bmm(attn, v)
else:
mixed_output = torch.bmm(attn, v).contiguous().view(bsz, -1, Nt, E)
output = torch.stack(
[mixed_output[torch.arange(bsz), subset_heads[:, col], :, :] for col in range(subset_heads.size(1))],
dim=1
)
output = output * subset_weights.unsqueeze(2).unsqueeze(3)
output = output.contiguous().view(-1, Nt, E)
if subset_heads is not None:
_, Nt, Ns = attn.size()
mixed_attn = attn.view(bsz, -1, Nt, Ns)
attn = torch.stack(
[mixed_attn[torch.arange(bsz), subset_heads[:, col], :, :] for col in range(subset_heads.size(1))], dim=1
)
return output, attn
def _in_projection(
q: Tensor,
k: Tensor,
v: Tensor,
w_q: Tensor,
w_k: Tensor,
w_v: Tensor,
b_q: Optional[Tensor] = None,
b_k: Optional[Tensor] = None,
b_v: Optional[Tensor] = None,
) -> Tuple[Tensor, Tensor, Tensor]:
return linear(q, w_q, b_q), linear(k, w_k, b_k), linear(v, w_v, b_v)
def multi_head_attention_forward(
query: Tensor,
key: Tensor,
value: Tensor,
embed_dim_to_check: int,
total_num_heads: int,
num_heads: int,
in_proj_weight: Tensor,
in_proj_bias: Optional[Tensor],
bias_k: Optional[Tensor],
bias_v: Optional[Tensor],
add_zero_attn: bool,
dropout_p: float,
out_proj_weight: Tensor,
out_proj_bias: Optional[Tensor],
training: bool = True,
key_padding_mask: Optional[Tensor] = None,
need_weights: bool = True,
attn_mask: Optional[Tensor] = None,
use_separate_proj_weight: bool = False,
q_proj_weight: Optional[Tensor] = None,
k_proj_weight: Optional[Tensor] = None,
v_proj_weight: Optional[Tensor] = None,
static_k: Optional[Tensor] = None,
static_v: Optional[Tensor] = None,
subset_heads: Optional[Tensor] = None,
subset_weights: Optional[Tensor] = None,
):
tens_ops = (query, key, value, in_proj_weight, in_proj_bias, bias_k, bias_v, out_proj_weight, out_proj_bias)
if has_torch_function(tens_ops):
return handle_torch_function(
multi_head_attention_forward,
tens_ops,
query,
key,
value,
embed_dim_to_check,
total_num_heads,
num_heads,
in_proj_weight,
in_proj_bias,
bias_k,
bias_v,
add_zero_attn,
dropout_p,
out_proj_weight,
out_proj_bias,
training=training,
key_padding_mask=key_padding_mask,
need_weights=need_weights,
attn_mask=attn_mask,
use_separate_proj_weight=use_separate_proj_weight,
q_proj_weight=q_proj_weight,
k_proj_weight=k_proj_weight,
v_proj_weight=v_proj_weight,
static_k=static_k,
static_v=static_v,
subset_heads=subset_heads,
subset_weights=subset_weights
)
# set up shape vars
tgt_len, bsz, embed_dim = query.shape
src_len, _, _ = key.shape
assert embed_dim == embed_dim_to_check, \
f"was expecting embedding dimension of {embed_dim_to_check}, but got {embed_dim}"
if isinstance(embed_dim, torch.Tensor):
# embed_dim can be a tensor when JIT tracing
head_dim = embed_dim.div(num_heads, rounding_mode='trunc')
else:
head_dim = embed_dim // num_heads
assert head_dim * num_heads == embed_dim, f"embed_dim {embed_dim} not divisible by num_heads {num_heads}"
if use_separate_proj_weight:
# allow MHA to have different embedding dimensions when separate projection weights are used
assert key.shape[:2] == value.shape[:2], \
f"key's sequence and batch dims {key.shape[:2]} do not match value's {value.shape[:2]}"
else:
assert key.shape == value.shape, f"key shape {key.shape} does not match value shape {value.shape}"
#
# compute in-projection
#
if not use_separate_proj_weight:
q, k, v = _in_projection_packed(query, key, value, in_proj_weight, in_proj_bias)
else:
assert q_proj_weight is not None, "use_separate_proj_weight is True but q_proj_weight is None"
assert k_proj_weight is not None, "use_separate_proj_weight is True but k_proj_weight is None"
assert v_proj_weight is not None, "use_separate_proj_weight is True but v_proj_weight is None"
if in_proj_bias is None:
b_q = b_k = b_v = None
else:
b_q, b_k, b_v = in_proj_bias.chunk(3)
q, k, v = _in_projection(query, key, value, q_proj_weight, k_proj_weight, v_proj_weight, b_q, b_k, b_v)
# prep attention mask
if attn_mask is not None:
if attn_mask.dtype == torch.uint8:
warnings.warn("Byte tensor for attn_mask in nn.MultiheadAttention is deprecated. Use bool tensor instead.")
attn_mask = attn_mask.to(torch.bool)
else:
assert attn_mask.is_floating_point() or attn_mask.dtype == torch.bool, \
f"Only float, byte, and bool types are supported for attn_mask, not {attn_mask.dtype}"
# ensure attn_mask's dim is 3
if attn_mask.dim() == 2:
correct_2d_size = (tgt_len, src_len)
if attn_mask.shape != correct_2d_size:
raise RuntimeError(f"The shape of the 2D attn_mask is {attn_mask.shape}, but should be {correct_2d_size}.")
attn_mask = attn_mask.unsqueeze(0)
elif attn_mask.dim() == 3:
correct_3d_size = (bsz * total_num_heads, tgt_len, src_len)
if attn_mask.shape != correct_3d_size:
raise RuntimeError(f"The shape of the 3D attn_mask is {attn_mask.shape}, but should be {correct_3d_size}.")
else:
raise RuntimeError(f"attn_mask's dimension {attn_mask.dim()} is not supported")
# prep key padding mask
if key_padding_mask is not None and key_padding_mask.dtype == torch.uint8:
warnings.warn("Byte tensor for key_padding_mask in nn.MultiheadAttention is deprecated. Use bool tensor instead.")
key_padding_mask = key_padding_mask.to(torch.bool)
# add bias along batch dimension (currently second)
if bias_k is not None and bias_v is not None:
assert static_k is None, "bias cannot be added to static key."
assert static_v is None, "bias cannot be added to static value."
k = torch.cat([k, bias_k.repeat(1, bsz, 1)])
v = torch.cat([v, bias_v.repeat(1, bsz, 1)])
if attn_mask is not None:
attn_mask = pad(attn_mask, (0, 1))
if key_padding_mask is not None:
key_padding_mask = pad(key_padding_mask, (0, 1))
else:
assert bias_k is None
assert bias_v is None
#
# reshape q, k, v for multihead attention and make em batch first
#
q = q.contiguous().view(tgt_len, bsz * total_num_heads, head_dim).transpose(0, 1)
if static_k is None:
k = k.contiguous().view(k.shape[0], bsz * total_num_heads, head_dim).transpose(0, 1)
else:
# TODO finish disentangling control flow so we don't do in-projections when statics are passed
assert static_k.size(0) == bsz * total_num_heads, \
f"expecting static_k.size(0) of {bsz * total_num_heads}, but got {static_k.size(0)}"
assert static_k.size(2) == head_dim, \
f"expecting static_k.size(2) of {head_dim}, but got {static_k.size(2)}"
k = static_k
if static_v is None:
v = v.contiguous().view(v.shape[0], bsz * total_num_heads, head_dim).transpose(0, 1)
else:
# TODO finish disentangling control flow so we don't do in-projections when statics are passed
assert static_v.size(0) == bsz * total_num_heads, \
f"expecting static_v.size(0) of {bsz * total_num_heads}, but got {static_v.size(0)}"
assert static_v.size(2) == head_dim, \
f"expecting static_v.size(2) of {head_dim}, but got {static_v.size(2)}"
v = static_v
# add zero attention along batch dimension (now first)
if add_zero_attn:
zero_attn_shape = (bsz * total_num_heads, 1, head_dim)
k = torch.cat([k, torch.zeros(zero_attn_shape, dtype=k.dtype, device=k.device)], dim=1)
v = torch.cat([v, torch.zeros(zero_attn_shape, dtype=v.dtype, device=v.device)], dim=1)
if attn_mask is not None:
attn_mask = pad(attn_mask, (0, 1))
if key_padding_mask is not None:
key_padding_mask = pad(key_padding_mask, (0, 1))
# update source sequence length after adjustments
src_len = k.size(1)
# merge key padding and attention masks
if key_padding_mask is not None:
assert key_padding_mask.shape == (bsz, src_len), \
f"expecting key_padding_mask shape of {(bsz, src_len)}, but got {key_padding_mask.shape}"
key_padding_mask = key_padding_mask.view(bsz, 1, 1, src_len). \
expand(-1, total_num_heads, -1, -1).reshape(bsz * total_num_heads, 1, src_len)
if attn_mask is None:
attn_mask = key_padding_mask
elif attn_mask.dtype == torch.bool:
attn_mask = attn_mask.logical_or(key_padding_mask)
else:
attn_mask = attn_mask.masked_fill(key_padding_mask, float("-inf"))
# convert mask to float
if attn_mask is not None and attn_mask.dtype == torch.bool:
new_attn_mask = torch.zeros_like(attn_mask, dtype=torch.float)
new_attn_mask.masked_fill_(attn_mask, float("-inf"))
attn_mask = new_attn_mask
# adjust dropout probability
if not training:
dropout_p = 0.0
#
# (deep breath) calculate attention and out projection
#
attn_output, attn_output_weights = _scaled_dot_product_attention(q, k, v, attn_mask, dropout_p, bsz, subset_heads, subset_weights)
attn_output = attn_output.transpose(0, 1).contiguous().view(tgt_len, bsz, embed_dim)
attn_output = linear(attn_output, out_proj_weight, out_proj_bias)
if need_weights:
# average attention weights over heads
attn_output_weights = attn_output_weights.view(bsz, num_heads, tgt_len, src_len)
return attn_output, attn_output_weights.sum(dim=1) / num_heads
else:
return attn_output, None
|