未验证 提交 a870f942 编写于 作者: L littletomatodonkey 提交者: GitHub

Update vision_transformer.py

上级 fb7c750c
......@@ -12,20 +12,18 @@
# See the License for the specific language governing permissions and
# limitations under the License.
import numpy as np
import paddle
import paddle.nn as nn
from paddle.nn.initializer import TruncatedNormal, Constant
__all__ = [
"VisionTransformer",
"ViT_small_patch16_224",
"ViT_base_patch16_224", "ViT_base_patch16_384", "ViT_base_patch32_384",
"ViT_large_patch16_224", "ViT_large_patch16_384", "ViT_large_patch32_384",
"ViT_huge_patch16_224", "ViT_huge_patch32_384"
"VisionTransformer", "ViT_small_patch16_224", "ViT_base_patch16_224",
"ViT_base_patch16_384", "ViT_base_patch32_384", "ViT_large_patch16_224",
"ViT_large_patch16_384", "ViT_large_patch32_384", "ViT_huge_patch16_224",
"ViT_huge_patch32_384"
]
trunc_normal_ = TruncatedNormal(std=.02)
zeros_ = Constant(value=0.)
ones_ = Constant(value=1.)
......@@ -43,12 +41,13 @@ def drop_path(x, drop_prob=0., training=False):
if drop_prob == 0. or not training:
return x
keep_prob = paddle.to_tensor(1 - drop_prob)
shape = (x.shape[0],) + (1,) * (x.ndim - 1)
shape = (paddle.shape(x)[0], ) + (1, ) * (x.ndim - 1)
random_tensor = keep_prob + paddle.rand(shape, dtype=x.dtype)
random_tensor = paddle.floor(random_tensor) # binarize
random_tensor = paddle.floor(random_tensor) # binarize
output = x.divide(keep_prob) * random_tensor
return output
class DropPath(nn.Layer):
"""Drop paths (Stochastic Depth) per sample (when applied in main path of residual blocks).
"""
......@@ -70,7 +69,12 @@ class Identity(nn.Layer):
class Mlp(nn.Layer):
def __init__(self, in_features, hidden_features=None, out_features=None, act_layer=nn.GELU, drop=0.):
def __init__(self,
in_features,
hidden_features=None,
out_features=None,
act_layer=nn.GELU,
drop=0.):
super().__init__()
out_features = out_features or in_features
hidden_features = hidden_features or in_features
......@@ -89,11 +93,17 @@ class Mlp(nn.Layer):
class Attention(nn.Layer):
def __init__(self, dim, num_heads=8, qkv_bias=False, qk_scale=None, attn_drop=0., proj_drop=0.):
def __init__(self,
dim,
num_heads=8,
qkv_bias=False,
qk_scale=None,
attn_drop=0.,
proj_drop=0.):
super().__init__()
self.num_heads = num_heads
head_dim = dim // num_heads
self.scale = qk_scale or head_dim ** -0.5
self.scale = qk_scale or head_dim**-0.5
self.qkv = nn.Linear(dim, dim * 3, bias_attr=qkv_bias)
self.attn_drop = nn.Dropout(attn_drop)
......@@ -101,8 +111,9 @@ class Attention(nn.Layer):
self.proj_drop = nn.Dropout(proj_drop)
def forward(self, x):
B, N, C = x.shape
qkv = self.qkv(x).reshape((B, N, 3, self.num_heads, C //
# B= paddle.shape(x)[0]
N, C = x.shape[1:]
qkv = self.qkv(x).reshape((-1, N, 3, self.num_heads, C //
self.num_heads)).transpose((2, 0, 3, 1, 4))
q, k, v = qkv[0], qkv[1], qkv[2]
......@@ -110,26 +121,42 @@ class Attention(nn.Layer):
attn = nn.functional.softmax(attn, axis=-1)
attn = self.attn_drop(attn)
x = (attn.matmul(v)).transpose((0, 2, 1, 3)).reshape((B, N, C))
x = (attn.matmul(v)).transpose((0, 2, 1, 3)).reshape((-1, N, C))
x = self.proj(x)
x = self.proj_drop(x)
return x
class Block(nn.Layer):
def __init__(self, dim, num_heads, mlp_ratio=4., qkv_bias=False, qk_scale=None, drop=0., attn_drop=0.,
drop_path=0., act_layer=nn.GELU, norm_layer='nn.LayerNorm', epsilon=1e-5):
def __init__(self,
dim,
num_heads,
mlp_ratio=4.,
qkv_bias=False,
qk_scale=None,
drop=0.,
attn_drop=0.,
drop_path=0.,
act_layer=nn.GELU,
norm_layer='nn.LayerNorm',
epsilon=1e-5):
super().__init__()
self.norm1 = eval(norm_layer)(dim, epsilon=epsilon)
self.attn = Attention(
dim, num_heads=num_heads, qkv_bias=qkv_bias, qk_scale=qk_scale, attn_drop=attn_drop, proj_drop=drop)
dim,
num_heads=num_heads,
qkv_bias=qkv_bias,
qk_scale=qk_scale,
attn_drop=attn_drop,
proj_drop=drop)
# NOTE: drop path for stochastic depth, we shall see if this is better than dropout here
self.drop_path = DropPath(drop_path) if drop_path > 0. else Identity()
self.norm2 = eval(norm_layer)(dim, epsilon=epsilon)
mlp_hidden_dim = int(dim * mlp_ratio)
self.mlp = Mlp(in_features=dim, hidden_features=mlp_hidden_dim,
act_layer=act_layer, drop=drop)
self.mlp = Mlp(in_features=dim,
hidden_features=mlp_hidden_dim,
act_layer=act_layer,
drop=drop)
def forward(self, x):
x = x + self.drop_path(self.attn(self.norm1(x)))
......@@ -151,13 +178,13 @@ class PatchEmbed(nn.Layer):
self.patch_size = patch_size
self.num_patches = num_patches
self.proj = nn.Conv2D(in_chans, embed_dim,
kernel_size=patch_size, stride=patch_size)
self.proj = nn.Conv2D(
in_chans, embed_dim, kernel_size=patch_size, stride=patch_size)
def forward(self, x):
B, C, H, W = x.shape
assert H == self.img_size[0] and W == self.img_size[1], \
f"Input image size ({H}*{W}) doesn't match model ({self.img_size[0]}*{self.img_size[1]})."
"Input image size ({H}*{W}) doesn't match model ({self.img_size[0]}*{self.img_size[1]})."
x = self.proj(x).flatten(2).transpose((0, 2, 1))
return x
......@@ -167,16 +194,33 @@ class VisionTransformer(nn.Layer):
""" Vision Transformer with support for patch input
"""
def __init__(self, img_size=224, patch_size=16, in_chans=3, class_dim=1000, embed_dim=768, depth=12,
num_heads=12, mlp_ratio=4, qkv_bias=False, qk_scale=None, drop_rate=0., attn_drop_rate=0.,
drop_path_rate=0., norm_layer='nn.LayerNorm', epsilon=1e-5, **args):
def __init__(self,
img_size=224,
patch_size=16,
in_chans=3,
class_dim=1000,
embed_dim=768,
depth=12,
num_heads=12,
mlp_ratio=4,
qkv_bias=False,
qk_scale=None,
drop_rate=0.,
attn_drop_rate=0.,
drop_path_rate=0.,
norm_layer='nn.LayerNorm',
epsilon=1e-5,
**args):
super().__init__()
self.class_dim = class_dim
self.num_features = self.embed_dim = embed_dim
self.patch_embed = PatchEmbed(
img_size=img_size, patch_size=patch_size, in_chans=in_chans, embed_dim=embed_dim)
img_size=img_size,
patch_size=patch_size,
in_chans=in_chans,
embed_dim=embed_dim)
num_patches = self.patch_embed.num_patches
self.pos_embed = self.create_parameter(
......@@ -187,23 +231,33 @@ class VisionTransformer(nn.Layer):
self.add_parameter("cls_token", self.cls_token)
self.pos_drop = nn.Dropout(p=drop_rate)
dpr = [x for x in paddle.linspace(0, drop_path_rate, depth)]
dpr = np.linspace(0, drop_path_rate, depth)
self.blocks = nn.LayerList([
Block(
dim=embed_dim, num_heads=num_heads, mlp_ratio=mlp_ratio, qkv_bias=qkv_bias, qk_scale=qk_scale,
drop=drop_rate, attn_drop=attn_drop_rate, drop_path=dpr[i], norm_layer=norm_layer, epsilon=epsilon)
for i in range(depth)])
dim=embed_dim,
num_heads=num_heads,
mlp_ratio=mlp_ratio,
qkv_bias=qkv_bias,
qk_scale=qk_scale,
drop=drop_rate,
attn_drop=attn_drop_rate,
drop_path=dpr[i],
norm_layer=norm_layer,
epsilon=epsilon) for i in range(depth)
])
self.norm = eval(norm_layer)(embed_dim, epsilon=epsilon)
# Classifier head
self.head = nn.Linear(
embed_dim, class_dim) if class_dim > 0 else Identity()
self.head = nn.Linear(embed_dim,
class_dim) if class_dim > 0 else Identity()
trunc_normal_(self.pos_embed)
trunc_normal_(self.cls_token)
self.apply(self._init_weights)
# TODO(littletomatodonkey): same init in static mode
if paddle.in_dynamic_mode():
trunc_normal_(self.pos_embed)
trunc_normal_(self.cls_token)
self.apply(self._init_weights)
def _init_weights(self, m):
if isinstance(m, nn.Linear):
......@@ -215,7 +269,8 @@ class VisionTransformer(nn.Layer):
ones_(m.weight)
def forward_features(self, x):
B = x.shape[0]
# B = x.shape[0]
B = paddle.shape(x)[0]
x = self.patch_embed(x)
cls_tokens = self.cls_token.expand((B, -1, -1))
x = paddle.concat((cls_tokens, x), axis=1)
......@@ -234,59 +289,116 @@ class VisionTransformer(nn.Layer):
def ViT_small_patch16_224(**kwargs):
model = VisionTransformer(
patch_size=16, embed_dim=768, depth=8, num_heads=8, mlp_ratio=3, qk_scale=768**-0.5, **kwargs)
patch_size=16,
embed_dim=768,
depth=8,
num_heads=8,
mlp_ratio=3,
qk_scale=768**-0.5,
**kwargs)
return model
def ViT_base_patch16_224(**kwargs):
model = VisionTransformer(
patch_size=16, embed_dim=768, depth=12, num_heads=12, mlp_ratio=4, qkv_bias=True,
epsilon=1e-6, **kwargs)
patch_size=16,
embed_dim=768,
depth=12,
num_heads=12,
mlp_ratio=4,
qkv_bias=True,
epsilon=1e-6,
**kwargs)
return model
def ViT_base_patch16_384(**kwargs):
model = VisionTransformer(
img_size=384, patch_size=16, embed_dim=768, depth=12, num_heads=12, mlp_ratio=4,
qkv_bias=True, epsilon=1e-6, **kwargs)
img_size=384,
patch_size=16,
embed_dim=768,
depth=12,
num_heads=12,
mlp_ratio=4,
qkv_bias=True,
epsilon=1e-6,
**kwargs)
return model
def ViT_base_patch32_384(**kwargs):
model = VisionTransformer(
img_size=384, patch_size=32, embed_dim=768, depth=12, num_heads=12, mlp_ratio=4,
qkv_bias=True, epsilon=1e-6, **kwargs)
img_size=384,
patch_size=32,
embed_dim=768,
depth=12,
num_heads=12,
mlp_ratio=4,
qkv_bias=True,
epsilon=1e-6,
**kwargs)
return model
def ViT_large_patch16_224(**kwargs):
model = VisionTransformer(
patch_size=16, embed_dim=1024, depth=24, num_heads=16, mlp_ratio=4, qkv_bias=True,
epsilon=1e-6, **kwargs)
patch_size=16,
embed_dim=1024,
depth=24,
num_heads=16,
mlp_ratio=4,
qkv_bias=True,
epsilon=1e-6,
**kwargs)
return model
def ViT_large_patch16_384(**kwargs):
model = VisionTransformer(
img_size=384, patch_size=16, embed_dim=1024, depth=24, num_heads=16, mlp_ratio=4,
qkv_bias=True, epsilon=1e-6, **kwargs)
img_size=384,
patch_size=16,
embed_dim=1024,
depth=24,
num_heads=16,
mlp_ratio=4,
qkv_bias=True,
epsilon=1e-6,
**kwargs)
return model
def ViT_large_patch32_384(**kwargs):
model = VisionTransformer(
img_size=384, patch_size=32, embed_dim=1024, depth=24, num_heads=16, mlp_ratio=4,
qkv_bias=True, epsilon=1e-6, **kwargs)
img_size=384,
patch_size=32,
embed_dim=1024,
depth=24,
num_heads=16,
mlp_ratio=4,
qkv_bias=True,
epsilon=1e-6,
**kwargs)
return model
def ViT_huge_patch16_224(**kwargs):
model = VisionTransformer(
patch_size=16, embed_dim=1280, depth=32, num_heads=16, mlp_ratio=4, **kwargs)
patch_size=16,
embed_dim=1280,
depth=32,
num_heads=16,
mlp_ratio=4,
**kwargs)
return model
def ViT_huge_patch32_384(**kwargs):
model = VisionTransformer(
img_size=384, patch_size=32, embed_dim=1280, depth=32, num_heads=16, mlp_ratio=4, **kwargs)
img_size=384,
patch_size=32,
embed_dim=1280,
depth=32,
num_heads=16,
mlp_ratio=4,
**kwargs)
return model
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册