index int64 0 50.7k | project_path stringclasses 279
values | completion_path stringlengths 57 199 | signature_postion stringlengths 3 11 | class_position stringlengths 3 11 | function_name stringlengths 1 62 | class_name stringlengths 0 55 | body_position stringlengths 3 11 | body_position_with_comments stringlengths 3 11 | masked_class stringlengths 14 8.32k | masked_class_with_comment stringlengths 14 8.32k | func_body stringlengths 9 6.72k | func_body_with_comment stringlengths 9 6.72k | len_func_body int64 1 643 | len_input int64 10 2.02k | len_output int64 4 1.9k | total int64 20 2.05k | file_code stringlengths 14 442k | file_code_with_comment stringlengths 14 442k |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
17,633 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/efficientvit_mit.py | 217,218 | 170,224 | forward | MBConv | 218,224 | 218,224 | class MBConv(nn.Module):
def __init__(
self,
in_channels: int,
out_channels: int,
kernel_size=3,
stride=1,
mid_channels=None,
expand_ratio=6,
use_bias=False,
norm_layer=(nn.BatchNorm2d, nn.BatchNorm2d, nn.BatchNorm2d),
act_layer=(nn.ReL... | class MBConv(nn.Module):
def __init__(
self,
in_channels: int,
out_channels: int,
kernel_size=3,
stride=1,
mid_channels=None,
expand_ratio=6,
use_bias=False,
norm_layer=(nn.BatchNorm2d, nn.BatchNorm2d, nn.BatchNorm2d),
act_layer=(nn.ReL... | x = self.inverted_conv(x)
x = self.depth_conv(x)
x = self.point_conv(x)
return x
| x = self.inverted_conv(x)
x = self.depth_conv(x)
x = self.point_conv(x)
return x
| 11 | 485 | 49 | 534 | """ EfficientViT (by MIT Song Han's Lab)
Paper: `Efficientvit: Enhanced linear attention for high-resolution low-computation visual recognition`
- https://arxiv.org/abs/2205.14756
Adapted from official impl at https://github.com/mit-han-lab/efficientvit
"""
__all__ = ['EfficientVit']
from typing import Optional
... | """ EfficientViT (by MIT Song Han's Lab)
Paper: `Efficientvit: Enhanced linear attention for high-resolution low-computation visual recognition`
- https://arxiv.org/abs/2205.14756
Adapted from official impl at https://github.com/mit-han-lab/efficientvit
"""
__all__ = ['EfficientVit']
from typing import Optional
... |
17,634 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/efficientvit_mit.py | 419,420 | 407,426 | forward | ResidualBlock | 420,426 | 420,426 | class ResidualBlock(nn.Module):
def __init__(
self,
main: Optional[nn.Module],
shortcut: Optional[nn.Module] = None,
pre_norm: Optional[nn.Module] = None,
):
super(ResidualBlock, self).__init__()
self.pre_norm = pre_norm if pre_norm is not None else nn.Identity()
... | class ResidualBlock(nn.Module):
def __init__(
self,
main: Optional[nn.Module],
shortcut: Optional[nn.Module] = None,
pre_norm: Optional[nn.Module] = None,
):
super(ResidualBlock, self).__init__()
self.pre_norm = pre_norm if pre_norm is not None else nn.Identity()
... | res = self.main(self.pre_norm(x))
if self.shortcut is not None:
res = res + self.shortcut(x)
return res
| res = self.main(self.pre_norm(x))
if self.shortcut is not None:
res = res + self.shortcut(x)
return res
| 15 | 135 | 52 | 187 | """ EfficientViT (by MIT Song Han's Lab)
Paper: `Efficientvit: Enhanced linear attention for high-resolution low-computation visual recognition`
- https://arxiv.org/abs/2205.14756
Adapted from official impl at https://github.com/mit-han-lab/efficientvit
"""
__all__ = ['EfficientVit']
from typing import Optional
... | """ EfficientViT (by MIT Song Han's Lab)
Paper: `Efficientvit: Enhanced linear attention for high-resolution low-computation visual recognition`
- https://arxiv.org/abs/2205.14756
Adapted from official impl at https://github.com/mit-han-lab/efficientvit
"""
__all__ = ['EfficientVit']
from typing import Optional
... |
17,636 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/efficientvit_msra.py | 186,187 | 116,210 | forward | CascadedGroupAttention | 187,210 | 187,210 | class CascadedGroupAttention(torch.nn.Module):
attention_bias_cache: Dict[str, torch.Tensor]
r""" Cascaded Group Attention.
Args:
dim (int): Number of input channels.
key_dim (int): The dimension for query and key.
num_heads (int): Number of attention heads.
attn_ratio (int... | class CascadedGroupAttention(torch.nn.Module):
attention_bias_cache: Dict[str, torch.Tensor]
r""" Cascaded Group Attention.
Args:
dim (int): Number of input channels.
key_dim (int): The dimension for query and key.
num_heads (int): Number of attention heads.
attn_ratio (int... | B, C, H, W = x.shape
feats_in = x.chunk(len(self.qkvs), dim=1)
feats_out = []
feat = feats_in[0]
attn_bias = self.get_attention_biases(x.device)
for head_idx, (qkv, dws) in enumerate(zip(self.qkvs, self.dws)):
if head_idx > 0:
feat = feat + fea... | B, C, H, W = x.shape
feats_in = x.chunk(len(self.qkvs), dim=1)
feats_out = []
feat = feats_in[0]
attn_bias = self.get_attention_biases(x.device)
for head_idx, (qkv, dws) in enumerate(zip(self.qkvs, self.dws)):
if head_idx > 0:
feat = feat + fea... | 97 | 940 | 387 | 1,327 | """ EfficientViT (by MSRA)
Paper: `EfficientViT: Memory Efficient Vision Transformer with Cascaded Group Attention`
- https://arxiv.org/abs/2305.07027
Adapted from official impl at https://github.com/microsoft/Cream/tree/main/EfficientViT
"""
__all__ = ['EfficientVitMsra']
import itertools
from collections impor... | """ EfficientViT (by MSRA)
Paper: `EfficientViT: Memory Efficient Vision Transformer with Cascaded Group Attention`
- https://arxiv.org/abs/2305.07027
Adapted from official impl at https://github.com/microsoft/Cream/tree/main/EfficientViT
"""
__all__ = ['EfficientVitMsra']
import itertools
from collections impor... |
17,637 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/efficientvit_msra.py | 111,112 | 104,116 | forward | ConvMlp | 112,116 | 112,116 | class ConvMlp(torch.nn.Module):
def __init__(self, ed, h):
super().__init__()
self.pw1 = ConvNorm(ed, h)
self.act = torch.nn.ReLU()
self.pw2 = ConvNorm(h, ed, bn_weight_init=0)
def forward(self, x):
FILL_FUNC_BODY | class ConvMlp(torch.nn.Module):
def __init__(self, ed, h):
super().__init__()
self.pw1 = ConvNorm(ed, h)
self.act = torch.nn.ReLU()
self.pw2 = ConvNorm(h, ed, bn_weight_init=0)
def forward(self, x):
FILL_FUNC_BODY | x = self.pw2(self.act(self.pw1(x)))
return x
| x = self.pw2(self.act(self.pw1(x)))
return x
| 5 | 110 | 32 | 142 | """ EfficientViT (by MSRA)
Paper: `EfficientViT: Memory Efficient Vision Transformer with Cascaded Group Attention`
- https://arxiv.org/abs/2305.07027
Adapted from official impl at https://github.com/microsoft/Cream/tree/main/EfficientViT
"""
__all__ = ['EfficientVitMsra']
import itertools
from collections impor... | """ EfficientViT (by MSRA)
Paper: `EfficientViT: Memory Efficient Vision Transformer with Cascaded Group Attention`
- https://arxiv.org/abs/2305.07027
Adapted from official impl at https://github.com/microsoft/Cream/tree/main/EfficientViT
"""
__all__ = ['EfficientVitMsra']
import itertools
from collections impor... |
17,638 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/efficientvit_msra.py | 32,34 | 24,47 | fuse | ConvNorm | 34,47 | 34,47 | class ConvNorm(torch.nn.Sequential):
def __init__(self, in_chs, out_chs, ks=1, stride=1, pad=0, dilation=1, groups=1, bn_weight_init=1):
super().__init__()
self.conv = nn.Conv2d(in_chs, out_chs, ks, stride, pad, dilation, groups, bias=False)
self.bn = nn.BatchNorm2d(out_chs)
torch.nn... | class ConvNorm(torch.nn.Sequential):
def __init__(self, in_chs, out_chs, ks=1, stride=1, pad=0, dilation=1, groups=1, bn_weight_init=1):
super().__init__()
self.conv = nn.Conv2d(in_chs, out_chs, ks, stride, pad, dilation, groups, bias=False)
self.bn = nn.BatchNorm2d(out_chs)
torch.nn... | c, bn = self.conv, self.bn
w = bn.weight / (bn.running_var + bn.eps)**0.5
w = c.weight * w[:, None, None, None]
b = bn.bias - bn.running_mean * bn.weight / \
(bn.running_var + bn.eps)**0.5
m = torch.nn.Conv2d(
w.size(1) * self.conv.groups, w.size(0), w.sha... | c, bn = self.conv, self.bn
w = bn.weight / (bn.running_var + bn.eps)**0.5
w = c.weight * w[:, None, None, None]
b = bn.bias - bn.running_mean * bn.weight / \
(bn.running_var + bn.eps)**0.5
m = torch.nn.Conv2d(
w.size(1) * self.conv.groups, w.size(0), w.sha... | 48 | 206 | 221 | 427 | """ EfficientViT (by MSRA)
Paper: `EfficientViT: Memory Efficient Vision Transformer with Cascaded Group Attention`
- https://arxiv.org/abs/2305.07027
Adapted from official impl at https://github.com/microsoft/Cream/tree/main/EfficientViT
"""
__all__ = ['EfficientVitMsra']
import itertools
from collections impor... | """ EfficientViT (by MSRA)
Paper: `EfficientViT: Memory Efficient Vision Transformer with Cascaded Group Attention`
- https://arxiv.org/abs/2305.07027
Adapted from official impl at https://github.com/microsoft/Cream/tree/main/EfficientViT
"""
__all__ = ['EfficientVitMsra']
import itertools
from collections impor... |
17,639 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/efficientvit_msra.py | 315,316 | 275,319 | forward | EfficientVitBlock | 316,319 | 316,319 | class EfficientVitBlock(torch.nn.Module):
""" A basic EfficientVit building block.
Args:
dim (int): Number of input channels.
key_dim (int): Dimension for query and key in the token mixer.
num_heads (int): Number of attention heads.
attn_ratio (int): Multiplier for the query dim... | class EfficientVitBlock(torch.nn.Module):
""" A basic EfficientVit building block.
Args:
dim (int): Number of input channels.
key_dim (int): Dimension for query and key in the token mixer.
num_heads (int): Number of attention heads.
attn_ratio (int): Multiplier for the query dim... | return self.ffn1(self.dw1(self.mixer(self.ffn0(self.dw0(x)))))
| return self.ffn1(self.dw1(self.mixer(self.ffn0(self.dw0(x)))))
| 2 | 455 | 40 | 495 | """ EfficientViT (by MSRA)
Paper: `EfficientViT: Memory Efficient Vision Transformer with Cascaded Group Attention`
- https://arxiv.org/abs/2305.07027
Adapted from official impl at https://github.com/microsoft/Cream/tree/main/EfficientViT
"""
__all__ = ['EfficientVitMsra']
import itertools
from collections impor... | """ EfficientViT (by MSRA)
Paper: `EfficientViT: Memory Efficient Vision Transformer with Cascaded Group Attention`
- https://arxiv.org/abs/2305.07027
Adapted from official impl at https://github.com/microsoft/Cream/tree/main/EfficientViT
"""
__all__ = ['EfficientVitMsra']
import itertools
from collections impor... |
17,640 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/efficientvit_msra.py | 490,491 | 382,530 | forward | EfficientVitMsra | 491,530 | 491,530 | class EfficientVitMsra(nn.Module):
def __init__(
self,
img_size=224,
in_chans=3,
num_classes=1000,
embed_dim=(64, 128, 192),
key_dim=(16, 16, 16),
depth=(1, 2, 3),
num_heads=(4, 4, 4),
window_size=(7, 7, 7),
... | class EfficientVitMsra(nn.Module):
def __init__(
self,
img_size=224,
in_chans=3,
num_classes=1000,
embed_dim=(64, 128, 192),
key_dim=(16, 16, 16),
depth=(1, 2, 3),
num_heads=(4, 4, 4),
window_size=(7, 7, 7),
... | x = self.forward_features(x)
x = self.forward_head(x)
return x
# def checkpoint_filter_fn(state_dict, model):
# if 'model' in state_dict.keys():
# state_dict = state_dict['model']
# tmp_dict = {}
# out_dict = {}
# target_keys = model.state_dict().keys()
# target_key... | x = self.forward_features(x)
x = self.forward_head(x)
return x
# def checkpoint_filter_fn(state_dict, model):
# if 'model' in state_dict.keys():
# state_dict = state_dict['model']
# tmp_dict = {}
# out_dict = {}
# target_keys = model.state_dict().keys()
# target_key... | 158 | 1,310 | 482 | 1,792 | """ EfficientViT (by MSRA)
Paper: `EfficientViT: Memory Efficient Vision Transformer with Cascaded Group Attention`
- https://arxiv.org/abs/2305.07027
Adapted from official impl at https://github.com/microsoft/Cream/tree/main/EfficientViT
"""
__all__ = ['EfficientVitMsra']
import itertools
from collections impor... | """ EfficientViT (by MSRA)
Paper: `EfficientViT: Memory Efficient Vision Transformer with Cascaded Group Attention`
- https://arxiv.org/abs/2305.07027
Adapted from official impl at https://github.com/microsoft/Cream/tree/main/EfficientViT
"""
__all__ = ['EfficientVitMsra']
import itertools
from collections impor... |
17,641 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/efficientvit_msra.py | 363,364 | 319,369 | forward | EfficientVitStage | 364,369 | 364,369 | class EfficientVitStage(torch.nn.Module):
def __init__(
self,
in_dim,
out_dim,
key_dim,
downsample=('', 1),
num_heads=8,
attn_ratio=4,
resolution=14,
window_resolution=7,
kernels=[5, 5, 5, 5],
... | class EfficientVitStage(torch.nn.Module):
def __init__(
self,
in_dim,
out_dim,
key_dim,
downsample=('', 1),
num_heads=8,
attn_ratio=4,
resolution=14,
window_resolution=7,
kernels=[5, 5, 5, 5],
... | x = self.downsample(x)
x = self.blocks(x)
return x
| x = self.downsample(x)
x = self.blocks(x)
return x
| 8 | 513 | 32 | 545 | """ EfficientViT (by MSRA)
Paper: `EfficientViT: Memory Efficient Vision Transformer with Cascaded Group Attention`
- https://arxiv.org/abs/2305.07027
Adapted from official impl at https://github.com/microsoft/Cream/tree/main/EfficientViT
"""
__all__ = ['EfficientVitMsra']
import itertools
from collections impor... | """ EfficientViT (by MSRA)
Paper: `EfficientViT: Memory Efficient Vision Transformer with Cascaded Group Attention`
- https://arxiv.org/abs/2305.07027
Adapted from official impl at https://github.com/microsoft/Cream/tree/main/EfficientViT
"""
__all__ = ['EfficientVitMsra']
import itertools
from collections impor... |
17,642 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/efficientvit_msra.py | 246,247 | 210,275 | forward | LocalWindowAttention | 247,275 | 247,275 | class LocalWindowAttention(torch.nn.Module):
r""" Local Window Attention.
Args:
dim (int): Number of input channels.
key_dim (int): The dimension for query and key.
num_heads (int): Number of attention heads.
attn_ratio (int): Multiplier for the query dim for value dimension.
... | class LocalWindowAttention(torch.nn.Module):
r""" Local Window Attention.
Args:
dim (int): Number of input channels.
key_dim (int): The dimension for query and key.
num_heads (int): Number of attention heads.
attn_ratio (int): Multiplier for the query dim for value dimension.
... | H = W = self.resolution
B, C, H_, W_ = x.shape
# Only check this for classifcation models
_assert(H == H_, f'input feature has wrong size, expect {(H, W)}, got {(H_, W_)}')
_assert(W == W_, f'input feature has wrong size, expect {(H, W)}, got {(H_, W_)}')
if H <= self.win... | H = W = self.resolution
B, C, H_, W_ = x.shape
# Only check this for classifcation models
_assert(H == H_, f'input feature has wrong size, expect {(H, W)}, got {(H_, W_)}')
_assert(W == W_, f'input feature has wrong size, expect {(H, W)}, got {(H_, W_)}')
if H <= self.win... | 190 | 362 | 639 | 1,001 | """ EfficientViT (by MSRA)
Paper: `EfficientViT: Memory Efficient Vision Transformer with Cascaded Group Attention`
- https://arxiv.org/abs/2305.07027
Adapted from official impl at https://github.com/microsoft/Cream/tree/main/EfficientViT
"""
__all__ = ['EfficientVitMsra']
import itertools
from collections impor... | """ EfficientViT (by MSRA)
Paper: `EfficientViT: Memory Efficient Vision Transformer with Cascaded Group Attention`
- https://arxiv.org/abs/2305.07027
Adapted from official impl at https://github.com/microsoft/Cream/tree/main/EfficientViT
"""
__all__ = ['EfficientVitMsra']
import itertools
from collections impor... |
17,643 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/efficientvit_msra.py | 58,60 | 47,75 | fuse | NormLinear | 60,75 | 60,75 | class NormLinear(torch.nn.Sequential):
def __init__(self, in_features, out_features, bias=True, std=0.02, drop=0.):
super().__init__()
self.bn = nn.BatchNorm1d(in_features)
self.drop = nn.Dropout(drop)
self.linear = nn.Linear(in_features, out_features, bias=bias)
trunc_norma... | class NormLinear(torch.nn.Sequential):
def __init__(self, in_features, out_features, bias=True, std=0.02, drop=0.):
super().__init__()
self.bn = nn.BatchNorm1d(in_features)
self.drop = nn.Dropout(drop)
self.linear = nn.Linear(in_features, out_features, bias=bias)
trunc_norma... | bn, linear = self.bn, self.linear
w = bn.weight / (bn.running_var + bn.eps)**0.5
b = bn.bias - self.bn.running_mean * \
self.bn.weight / (bn.running_var + bn.eps)**0.5
w = linear.weight * w[None, :]
if linear.bias is None:
b = b @ self.linear.weight.T
... | bn, linear = self.bn, self.linear
w = bn.weight / (bn.running_var + bn.eps)**0.5
b = bn.bias - self.bn.running_mean * \
self.bn.weight / (bn.running_var + bn.eps)**0.5
w = linear.weight * w[None, :]
if linear.bias is None:
b = b @ self.linear.weight.T
... | 56 | 187 | 218 | 405 | """ EfficientViT (by MSRA)
Paper: `EfficientViT: Memory Efficient Vision Transformer with Cascaded Group Attention`
- https://arxiv.org/abs/2305.07027
Adapted from official impl at https://github.com/microsoft/Cream/tree/main/EfficientViT
"""
__all__ = ['EfficientVitMsra']
import itertools
from collections impor... | """ EfficientViT (by MSRA)
Paper: `EfficientViT: Memory Efficient Vision Transformer with Cascaded Group Attention`
- https://arxiv.org/abs/2305.07027
Adapted from official impl at https://github.com/microsoft/Cream/tree/main/EfficientViT
"""
__all__ = ['EfficientVitMsra']
import itertools
from collections impor... |
17,644 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/efficientvit_msra.py | 85,86 | 75,90 | forward | PatchMerging | 86,90 | 86,90 | class PatchMerging(torch.nn.Module):
def __init__(self, dim, out_dim):
super().__init__()
hid_dim = int(dim * 4)
self.conv1 = ConvNorm(dim, hid_dim, 1, 1, 0)
self.act = torch.nn.ReLU()
self.conv2 = ConvNorm(hid_dim, hid_dim, 3, 2, 1, groups=hid_dim)
self.se = SqueezeE... | class PatchMerging(torch.nn.Module):
def __init__(self, dim, out_dim):
super().__init__()
hid_dim = int(dim * 4)
self.conv1 = ConvNorm(dim, hid_dim, 1, 1, 0)
self.act = torch.nn.ReLU()
self.conv2 = ConvNorm(hid_dim, hid_dim, 3, 2, 1, groups=hid_dim)
self.se = SqueezeE... | x = self.conv3(self.se(self.act(self.conv2(self.act(self.conv1(x))))))
return x
| x = self.conv3(self.se(self.act(self.conv2(self.act(self.conv1(x))))))
return x
| 5 | 194 | 45 | 239 | """ EfficientViT (by MSRA)
Paper: `EfficientViT: Memory Efficient Vision Transformer with Cascaded Group Attention`
- https://arxiv.org/abs/2305.07027
Adapted from official impl at https://github.com/microsoft/Cream/tree/main/EfficientViT
"""
__all__ = ['EfficientVitMsra']
import itertools
from collections impor... | """ EfficientViT (by MSRA)
Paper: `EfficientViT: Memory Efficient Vision Transformer with Cascaded Group Attention`
- https://arxiv.org/abs/2305.07027
Adapted from official impl at https://github.com/microsoft/Cream/tree/main/EfficientViT
"""
__all__ = ['EfficientVitMsra']
import itertools
from collections impor... |
17,645 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/efficientvit_msra.py | 96,97 | 90,104 | forward | ResidualDrop | 97,104 | 97,104 | class ResidualDrop(torch.nn.Module):
def __init__(self, m, drop=0.):
super().__init__()
self.m = m
self.drop = drop
def forward(self, x):
FILL_FUNC_BODY | class ResidualDrop(torch.nn.Module):
def __init__(self, m, drop=0.):
super().__init__()
self.m = m
self.drop = drop
def forward(self, x):
FILL_FUNC_BODY | if self.training and self.drop > 0:
return x + self.m(x) * torch.rand(
x.size(0), 1, 1, 1, device=x.device).ge_(self.drop).div(1 - self.drop).detach()
else:
return x + self.m(x)
| if self.training and self.drop > 0:
return x + self.m(x) * torch.rand(
x.size(0), 1, 1, 1, device=x.device).ge_(self.drop).div(1 - self.drop).detach()
else:
return x + self.m(x)
| 24 | 70 | 94 | 164 | """ EfficientViT (by MSRA)
Paper: `EfficientViT: Memory Efficient Vision Transformer with Cascaded Group Attention`
- https://arxiv.org/abs/2305.07027
Adapted from official impl at https://github.com/microsoft/Cream/tree/main/EfficientViT
"""
__all__ = ['EfficientVitMsra']
import itertools
from collections impor... | """ EfficientViT (by MSRA)
Paper: `EfficientViT: Memory Efficient Vision Transformer with Cascaded Group Attention`
- https://arxiv.org/abs/2305.07027
Adapted from official impl at https://github.com/microsoft/Cream/tree/main/EfficientViT
"""
__all__ = ['EfficientVitMsra']
import itertools
from collections impor... |
17,647 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/fastvit.py | 503,504 | 466,534 | forward | Attention | 504,534 | 504,534 | class Attention(nn.Module):
"""Multi-headed Self Attention module.
Source modified from:
https://github.com/rwightman/pytorch-image-models/blob/master/timm/models/vision_transformer.py
"""
fused_attn: torch.jit.Final[bool]
def __init__(
self,
dim: int,
head_dim: int = 3... | class Attention(nn.Module):
"""Multi-headed Self Attention module.
Source modified from:
https://github.com/rwightman/pytorch-image-models/blob/master/timm/models/vision_transformer.py
"""
fused_attn: torch.jit.Final[bool]
def __init__(
self,
dim: int,
head_dim: int = 3... | B, C, H, W = x.shape
N = H * W
x = x.flatten(2).transpose(-2, -1) # (B, N, C)
qkv = (
self.qkv(x)
.reshape(B, N, 3, self.num_heads, self.head_dim)
.permute(2, 0, 3, 1, 4)
)
q, k, v = qkv.unbind(0) # make torchscript happy (cannot use ... | B, C, H, W = x.shape
N = H * W
x = x.flatten(2).transpose(-2, -1) # (B, N, C)
qkv = (
self.qkv(x)
.reshape(B, N, 3, self.num_heads, self.head_dim)
.permute(2, 0, 3, 1, 4)
)
q, k, v = qkv.unbind(0) # make torchscript happy (cannot use ... | 106 | 467 | 390 | 857 | # FastViT for PyTorch
#
# Original implementation and weights from https://github.com/apple/ml-fastvit
#
# For licensing see accompanying LICENSE file at https://github.com/apple/ml-fastvit/tree/main
# Original work is copyright (C) 2023 Apple Inc. All Rights Reserved.
#
import os
from functools import partial
from typ... | # FastViT for PyTorch
#
# Original implementation and weights from https://github.com/apple/ml-fastvit
#
# For licensing see accompanying LICENSE file at https://github.com/apple/ml-fastvit/tree/main
# Original work is copyright (C) 2023 Apple Inc. All Rights Reserved.
#
import os
from functools import partial
from typ... |
17,648 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/fastvit.py | 982,983 | 931,988 | forward | AttentionBlock | 983,988 | 983,988 | class AttentionBlock(nn.Module):
"""Implementation of metaformer block with MHSA as token mixer.
For more details on Metaformer structure, please refer to:
`MetaFormer Is Actually What You Need for Vision <https://arxiv.org/pdf/2111.11418.pdf>`_
"""
def __init__(
self,
dim: int,
... | class AttentionBlock(nn.Module):
"""Implementation of metaformer block with MHSA as token mixer.
For more details on Metaformer structure, please refer to:
`MetaFormer Is Actually What You Need for Vision <https://arxiv.org/pdf/2111.11418.pdf>`_
"""
def __init__(
self,
dim: int,
... | x = x + self.drop_path1(self.layer_scale_1(self.token_mixer(self.norm(x))))
x = x + self.drop_path2(self.layer_scale_2(self.mlp(x)))
return x
| x = x + self.drop_path1(self.layer_scale_1(self.token_mixer(self.norm(x))))
x = x + self.drop_path2(self.layer_scale_2(self.mlp(x)))
return x
| 12 | 634 | 74 | 708 | # FastViT for PyTorch
#
# Original implementation and weights from https://github.com/apple/ml-fastvit
#
# For licensing see accompanying LICENSE file at https://github.com/apple/ml-fastvit/tree/main
# Original work is copyright (C) 2023 Apple Inc. All Rights Reserved.
#
import os
from functools import partial
from typ... | # FastViT for PyTorch
#
# Original implementation and weights from https://github.com/apple/ml-fastvit
#
# For licensing see accompanying LICENSE file at https://github.com/apple/ml-fastvit/tree/main
# Original work is copyright (C) 2023 Apple Inc. All Rights Reserved.
#
import os
from functools import partial
from typ... |
17,649 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/fastvit.py | 747,748 | 705,757 | forward | ConvMlp | 748,757 | 748,757 | class ConvMlp(nn.Module):
"""Convolutional FFN Module."""
def __init__(
self,
in_chs: int,
hidden_channels: Optional[int] = None,
out_chs: Optional[int] = None,
act_layer: nn.Module = nn.GELU,
drop: float = 0.0,
) -> None:
"""Build convolutional FFN m... | class ConvMlp(nn.Module):
"""Convolutional FFN Module."""
def __init__(
self,
in_chs: int,
hidden_channels: Optional[int] = None,
out_chs: Optional[int] = None,
act_layer: nn.Module = nn.GELU,
drop: float = 0.0,
) -> None:
"""Build convolutional FFN m... | x = self.conv(x)
x = self.fc1(x)
x = self.act(x)
x = self.drop(x)
x = self.fc2(x)
x = self.drop(x)
return x
| x = self.conv(x)
x = self.fc1(x)
x = self.act(x)
x = self.drop(x)
x = self.fc2(x)
x = self.drop(x)
return x
| 20 | 509 | 77 | 586 | # FastViT for PyTorch
#
# Original implementation and weights from https://github.com/apple/ml-fastvit
#
# For licensing see accompanying LICENSE file at https://github.com/apple/ml-fastvit/tree/main
# Original work is copyright (C) 2023 Apple Inc. All Rights Reserved.
#
import os
from functools import partial
from typ... | # FastViT for PyTorch
#
# Original implementation and weights from https://github.com/apple/ml-fastvit
#
# For licensing see accompanying LICENSE file at https://github.com/apple/ml-fastvit/tree/main
# Original work is copyright (C) 2023 Apple Inc. All Rights Reserved.
#
import os
from functools import partial
from typ... |
17,651 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/fastvit.py | 1075,1076 | 988,1085 | forward | FastVitStage | 1076,1085 | 1076,1085 | class FastVitStage(nn.Module):
def __init__(
self,
dim: int,
dim_out: int,
depth: int,
token_mixer_type: str,
downsample: bool = True,
down_patch_size: int = 7,
down_stride: int = 2,
pos_emb_layer: Optional[n... | class FastVitStage(nn.Module):
def __init__(
self,
dim: int,
dim_out: int,
depth: int,
token_mixer_type: str,
downsample: bool = True,
down_patch_size: int = 7,
down_stride: int = 2,
pos_emb_layer: Optional[n... | x = self.downsample(x)
x = self.pos_emb(x)
if self.grad_checkpointing and not torch.jit.is_scripting():
x = checkpoint_seq(self.blocks, x)
else:
x = self.blocks(x)
return x
| x = self.downsample(x)
x = self.pos_emb(x)
if self.grad_checkpointing and not torch.jit.is_scripting():
x = checkpoint_seq(self.blocks, x)
else:
x = self.blocks(x)
return x
| 21 | 916 | 90 | 1,006 | # FastViT for PyTorch
#
# Original implementation and weights from https://github.com/apple/ml-fastvit
#
# For licensing see accompanying LICENSE file at https://github.com/apple/ml-fastvit/tree/main
# Original work is copyright (C) 2023 Apple Inc. All Rights Reserved.
#
import os
from functools import partial
from typ... | # FastViT for PyTorch
#
# Original implementation and weights from https://github.com/apple/ml-fastvit
#
# For licensing see accompanying LICENSE file at https://github.com/apple/ml-fastvit/tree/main
# Original work is copyright (C) 2023 Apple Inc. All Rights Reserved.
#
import os
from functools import partial
from typ... |
17,652 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/fastvit.py | 589,590 | 583,593 | forward | LayerScale2d | 590,593 | 590,593 | class LayerScale2d(nn.Module):
def __init__(self, dim, init_values=1e-5, inplace=False):
super().__init__()
self.inplace = inplace
self.gamma = nn.Parameter(init_values * torch.ones(dim, 1, 1))
def forward(self, x):
FILL_FUNC_BODY | class LayerScale2d(nn.Module):
def __init__(self, dim, init_values=1e-5, inplace=False):
super().__init__()
self.inplace = inplace
self.gamma = nn.Parameter(init_values * torch.ones(dim, 1, 1))
def forward(self, x):
FILL_FUNC_BODY | return x.mul_(self.gamma) if self.inplace else x * self.gamma
| return x.mul_(self.gamma) if self.inplace else x * self.gamma
| 8 | 101 | 28 | 129 | # FastViT for PyTorch
#
# Original implementation and weights from https://github.com/apple/ml-fastvit
#
# For licensing see accompanying LICENSE file at https://github.com/apple/ml-fastvit/tree/main
# Original work is copyright (C) 2023 Apple Inc. All Rights Reserved.
#
import os
from functools import partial
from typ... | # FastViT for PyTorch
#
# Original implementation and weights from https://github.com/apple/ml-fastvit
#
# For licensing see accompanying LICENSE file at https://github.com/apple/ml-fastvit/tree/main
# Original work is copyright (C) 2023 Apple Inc. All Rights Reserved.
#
import os
from functools import partial
from typ... |
17,654 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/fastvit.py | 578,579 | 534,583 | forward | PatchEmbed | 579,583 | 579,583 | class PatchEmbed(nn.Module):
"""Convolutional patch embedding layer."""
def __init__(
self,
patch_size: int,
stride: int,
in_chs: int,
embed_dim: int,
act_layer: nn.Module = nn.GELU,
lkc_use_act: bool = False,
inference_mode: bool = False,
) -... | class PatchEmbed(nn.Module):
"""Convolutional patch embedding layer."""
def __init__(
self,
patch_size: int,
stride: int,
in_chs: int,
embed_dim: int,
act_layer: nn.Module = nn.GELU,
lkc_use_act: bool = False,
inference_mode: bool = False,
) -... | x = self.proj(x)
return x
| x = self.proj(x)
return x
| 5 | 441 | 20 | 461 | # FastViT for PyTorch
#
# Original implementation and weights from https://github.com/apple/ml-fastvit
#
# For licensing see accompanying LICENSE file at https://github.com/apple/ml-fastvit/tree/main
# Original work is copyright (C) 2023 Apple Inc. All Rights Reserved.
#
import os
from functools import partial
from typ... | # FastViT for PyTorch
#
# Original implementation and weights from https://github.com/apple/ml-fastvit
#
# For licensing see accompanying LICENSE file at https://github.com/apple/ml-fastvit/tree/main
# Original work is copyright (C) 2023 Apple Inc. All Rights Reserved.
#
import os
from functools import partial
from typ... |
17,655 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/fastvit.py | 827,828 | 757,873 | reparameterize | RepConditionalPosEnc | 828,873 | 828,873 | class RepConditionalPosEnc(nn.Module):
"""Implementation of conditional positional encoding.
For more details refer to paper:
`Conditional Positional Encodings for Vision Transformers <https://arxiv.org/pdf/2102.10882.pdf>`_
In our implementation, we can reparameterize this module to eliminate a skip ... | class RepConditionalPosEnc(nn.Module):
"""Implementation of conditional positional encoding.
For more details refer to paper:
`Conditional Positional Encodings for Vision Transformers <https://arxiv.org/pdf/2102.10882.pdf>`_
In our implementation, we can reparameterize this module to eliminate a skip ... | # Build equivalent Id tensor
input_dim = self.dim // self.groups
kernel_value = torch.zeros(
(
self.dim,
input_dim,
self.spatial_shape[0],
self.spatial_shape[1],
),
dtype=self.pos_enc.weight.dtype... | # Build equivalent Id tensor
input_dim = self.dim // self.groups
kernel_value = torch.zeros(
(
self.dim,
input_dim,
self.spatial_shape[0],
self.spatial_shape[1],
),
dtype=self.pos_enc.weight.dtype... | 92 | 716 | 447 | 1,163 | # FastViT for PyTorch
#
# Original implementation and weights from https://github.com/apple/ml-fastvit
#
# For licensing see accompanying LICENSE file at https://github.com/apple/ml-fastvit/tree/main
# Original work is copyright (C) 2023 Apple Inc. All Rights Reserved.
#
import os
from functools import partial
from typ... | # FastViT for PyTorch
#
# Original implementation and weights from https://github.com/apple/ml-fastvit
#
# For licensing see accompanying LICENSE file at https://github.com/apple/ml-fastvit/tree/main
# Original work is copyright (C) 2023 Apple Inc. All Rights Reserved.
#
import os
from functools import partial
from typ... |
17,656 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/fastvit.py | 660,661 | 593,705 | reparameterize | RepMixer | 664,705 | 661,705 | class RepMixer(nn.Module):
"""Reparameterizable token mixer.
For more details, please refer to our paper:
`FastViT: A Fast Hybrid Vision Transformer using Structural Reparameterization <https://arxiv.org/pdf/2303.14189.pdf>`_
"""
def __init__(
self,
dim,
kernel_size=3,
... | class RepMixer(nn.Module):
"""Reparameterizable token mixer.
For more details, please refer to our paper:
`FastViT: A Fast Hybrid Vision Transformer using Structural Reparameterization <https://arxiv.org/pdf/2303.14189.pdf>`_
"""
def __init__(
self,
dim,
kernel_size=3,
... | if self.inference_mode:
return
self.mixer.reparameterize()
self.norm.reparameterize()
if isinstance(self.layer_scale, LayerScale2d):
w = self.mixer.id_tensor + self.layer_scale.gamma.unsqueeze(-1) * (
self.mixer.reparam_conv.weight - self.norm.re... | """Reparameterize mixer and norm into a single
convolutional layer for efficient inference.
"""
if self.inference_mode:
return
self.mixer.reparameterize()
self.norm.reparameterize()
if isinstance(self.layer_scale, LayerScale2d):
w = self.... | 73 | 692 | 444 | 1,136 | # FastViT for PyTorch
#
# Original implementation and weights from https://github.com/apple/ml-fastvit
#
# For licensing see accompanying LICENSE file at https://github.com/apple/ml-fastvit/tree/main
# Original work is copyright (C) 2023 Apple Inc. All Rights Reserved.
#
import os
from functools import partial
from typ... | # FastViT for PyTorch
#
# Original implementation and weights from https://github.com/apple/ml-fastvit
#
# For licensing see accompanying LICENSE file at https://github.com/apple/ml-fastvit/tree/main
# Original work is copyright (C) 2023 Apple Inc. All Rights Reserved.
#
import os
from functools import partial
from typ... |
17,657 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/fastvit.py | 925,926 | 873,931 | forward | RepMixerBlock | 926,931 | 926,931 | class RepMixerBlock(nn.Module):
"""Implementation of Metaformer block with RepMixer as token mixer.
For more details on Metaformer structure, please refer to:
`MetaFormer Is Actually What You Need for Vision <https://arxiv.org/pdf/2111.11418.pdf>`_
"""
def __init__(
self,
dim: int,... | class RepMixerBlock(nn.Module):
"""Implementation of Metaformer block with RepMixer as token mixer.
For more details on Metaformer structure, please refer to:
`MetaFormer Is Actually What You Need for Vision <https://arxiv.org/pdf/2111.11418.pdf>`_
"""
def __init__(
self,
dim: int,... | x = self.token_mixer(x)
x = x + self.drop_path(self.layer_scale(self.mlp(x)))
return x
| x = self.token_mixer(x)
x = x + self.drop_path(self.layer_scale(self.mlp(x)))
return x
| 10 | 593 | 49 | 642 | # FastViT for PyTorch
#
# Original implementation and weights from https://github.com/apple/ml-fastvit
#
# For licensing see accompanying LICENSE file at https://github.com/apple/ml-fastvit/tree/main
# Original work is copyright (C) 2023 Apple Inc. All Rights Reserved.
#
import os
from functools import partial
from typ... | # FastViT for PyTorch
#
# Original implementation and weights from https://github.com/apple/ml-fastvit
#
# For licensing see accompanying LICENSE file at https://github.com/apple/ml-fastvit/tree/main
# Original work is copyright (C) 2023 Apple Inc. All Rights Reserved.
#
import os
from functools import partial
from typ... |
17,658 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/fastvit.py | 374,375 | 273,421 | reparameterize | ReparamLargeKernelConv | 381,397 | 375,397 | class ReparamLargeKernelConv(nn.Module):
"""Building Block of RepLKNet
This class defines overparameterized large kernel conv block
introduced in `RepLKNet <https://arxiv.org/abs/2203.06717>`_
Reference: https://github.com/DingXiaoH/RepLKNet-pytorch
"""
def __init__(
self,
in_... | class ReparamLargeKernelConv(nn.Module):
"""Building Block of RepLKNet
This class defines overparameterized large kernel conv block
introduced in `RepLKNet <https://arxiv.org/abs/2203.06717>`_
Reference: https://github.com/DingXiaoH/RepLKNet-pytorch
"""
def __init__(
self,
in_... | eq_k, eq_b = self.get_kernel_bias()
self.reparam_conv = create_conv2d(
self.in_chs,
self.out_chs,
kernel_size=self.kernel_size,
stride=self.stride,
groups=self.groups,
bias=True,
)
self.reparam_conv.weight.data = eq... | """
Following works like `RepVGG: Making VGG-style ConvNets Great Again` -
https://arxiv.org/pdf/2101.03697.pdf. We re-parameterize multi-branched
architecture used at training time to obtain a plain CNN-like structure
for inference.
"""
eq_k, eq_b = self.get_kern... | 25 | 1,424 | 180 | 1,604 | # FastViT for PyTorch
#
# Original implementation and weights from https://github.com/apple/ml-fastvit
#
# For licensing see accompanying LICENSE file at https://github.com/apple/ml-fastvit/tree/main
# Original work is copyright (C) 2023 Apple Inc. All Rights Reserved.
#
import os
from functools import partial
from typ... | # FastViT for PyTorch
#
# Original implementation and weights from https://github.com/apple/ml-fastvit
#
# For licensing see accompanying LICENSE file at https://github.com/apple/ml-fastvit/tree/main
# Original work is copyright (C) 2023 Apple Inc. All Rights Reserved.
#
import os
from functools import partial
from typ... |
17,660 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/focalnet.py | 315,316 | 283,321 | forward | Downsample | 316,321 | 316,321 | class Downsample(nn.Module):
def __init__(
self,
in_chs: int,
out_chs: int,
stride: int = 4,
overlap: bool = False,
norm_layer: Optional[Callable] = None,
):
"""
Args:
in_chs: Number of input image channels.
... | class Downsample(nn.Module):
def __init__(
self,
in_chs: int,
out_chs: int,
stride: int = 4,
overlap: bool = False,
norm_layer: Optional[Callable] = None,
):
"""
Args:
in_chs: Number of input image channels.
... | x = self.proj(x)
x = self.norm(x)
return x
| x = self.proj(x)
x = self.norm(x)
return x
| 8 | 322 | 31 | 353 | """ FocalNet
As described in `Focal Modulation Networks` - https://arxiv.org/abs/2203.11926
Significant modifications and refactoring from the original impl at https://github.com/microsoft/FocalNet
This impl is/has:
* fully convolutional, NCHW tensor layout throughout, seemed to have minimal performance impact but m... | """ FocalNet
As described in `Focal Modulation Networks` - https://arxiv.org/abs/2203.11926
Significant modifications and refactoring from the original impl at https://github.com/microsoft/FocalNet
This impl is/has:
* fully convolutional, NCHW tensor layout throughout, seemed to have minimal performance impact but m... |
17,661 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/focalnet.py | 77,78 | 36,104 | forward | FocalModulation | 78,104 | 78,104 | class FocalModulation(nn.Module):
def __init__(
self,
dim: int,
focal_window,
focal_level: int,
focal_factor: int = 2,
bias: bool = True,
use_post_norm: bool = False,
normalize_modulator: bool = False,
proj_d... | class FocalModulation(nn.Module):
def __init__(
self,
dim: int,
focal_window,
focal_level: int,
focal_factor: int = 2,
bias: bool = True,
use_post_norm: bool = False,
normalize_modulator: bool = False,
proj_d... | # pre linear projection
x = self.f(x)
q, ctx, gates = torch.split(x, self.input_split, 1)
# context aggreation
ctx_all = 0
for l, focal_layer in enumerate(self.focal_layers):
ctx = focal_layer(ctx)
ctx_all = ctx_all + ctx * gates[:, l:l + 1]
... | # pre linear projection
x = self.f(x)
q, ctx, gates = torch.split(x, self.input_split, 1)
# context aggreation
ctx_all = 0
for l, focal_layer in enumerate(self.focal_layers):
ctx = focal_layer(ctx)
ctx_all = ctx_all + ctx * gates[:, l:l + 1]
... | 86 | 536 | 299 | 835 | """ FocalNet
As described in `Focal Modulation Networks` - https://arxiv.org/abs/2203.11926
Significant modifications and refactoring from the original impl at https://github.com/microsoft/FocalNet
This impl is/has:
* fully convolutional, NCHW tensor layout throughout, seemed to have minimal performance impact but m... | """ FocalNet
As described in `Focal Modulation Networks` - https://arxiv.org/abs/2203.11926
Significant modifications and refactoring from the original impl at https://github.com/microsoft/FocalNet
This impl is/has:
* fully convolutional, NCHW tensor layout throughout, seemed to have minimal performance impact but m... |
17,662 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/focalnet.py | 460,461 | 321,475 | forward_features | FocalNet | 461,466 | 461,466 | class FocalNet(nn.Module):
"""" Focal Modulation Networks (FocalNets)
"""
def __init__(
self,
in_chans: int = 3,
num_classes: int = 1000,
global_pool: str = 'avg',
embed_dim: int = 96,
depths: Tuple[int, ...] = (2, 2, 6, 2),
... | class FocalNet(nn.Module):
"""" Focal Modulation Networks (FocalNets)
"""
def __init__(
self,
in_chans: int = 3,
num_classes: int = 1000,
global_pool: str = 'avg',
embed_dim: int = 96,
depths: Tuple[int, ...] = (2, 2, 6, 2),
... | x = self.stem(x)
x = self.layers(x)
x = self.norm(x)
return x
| x = self.stem(x)
x = self.layers(x)
x = self.norm(x)
return x
| 11 | 1,709 | 41 | 1,750 | """ FocalNet
As described in `Focal Modulation Networks` - https://arxiv.org/abs/2203.11926
Significant modifications and refactoring from the original impl at https://github.com/microsoft/FocalNet
This impl is/has:
* fully convolutional, NCHW tensor layout throughout, seemed to have minimal performance impact but m... | """ FocalNet
As described in `Focal Modulation Networks` - https://arxiv.org/abs/2203.11926
Significant modifications and refactoring from the original impl at https://github.com/microsoft/FocalNet
This impl is/has:
* fully convolutional, NCHW tensor layout throughout, seemed to have minimal performance impact but m... |
17,663 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/focalnet.py | 182,183 | 115,197 | forward | FocalNetBlock | 183,197 | 183,197 | class FocalNetBlock(nn.Module):
""" Focal Modulation Network Block.
"""
def __init__(
self,
dim: int,
mlp_ratio: float = 4.,
focal_level: int = 1,
focal_window: int = 3,
use_post_norm: bool = False,
use_post_norm_in_modulat... | class FocalNetBlock(nn.Module):
""" Focal Modulation Network Block.
"""
def __init__(
self,
dim: int,
mlp_ratio: float = 4.,
focal_level: int = 1,
focal_window: int = 3,
use_post_norm: bool = False,
use_post_norm_in_modulat... | shortcut = x
# Focal Modulation
x = self.norm1(x)
x = self.modulation(x)
x = self.norm1_post(x)
x = shortcut + self.drop_path1(self.ls1(x))
# FFN
x = x + self.drop_path2(self.ls2(self.norm2_post(self.mlp(self.norm2(x)))))
return x
| shortcut = x
# Focal Modulation
x = self.norm1(x)
x = self.modulation(x)
x = self.norm1_post(x)
x = shortcut + self.drop_path1(self.ls1(x))
# FFN
x = x + self.drop_path2(self.ls2(self.norm2_post(self.mlp(self.norm2(x)))))
return x
| 29 | 822 | 133 | 955 | """ FocalNet
As described in `Focal Modulation Networks` - https://arxiv.org/abs/2203.11926
Significant modifications and refactoring from the original impl at https://github.com/microsoft/FocalNet
This impl is/has:
* fully convolutional, NCHW tensor layout throughout, seemed to have minimal performance impact but m... | """ FocalNet
As described in `Focal Modulation Networks` - https://arxiv.org/abs/2203.11926
Significant modifications and refactoring from the original impl at https://github.com/microsoft/FocalNet
This impl is/has:
* fully convolutional, NCHW tensor layout throughout, seemed to have minimal performance impact but m... |
17,664 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/focalnet.py | 273,274 | 197,283 | forward | FocalNetStage | 274,283 | 274,283 | class FocalNetStage(nn.Module):
""" A basic Focal Transformer layer for one stage.
"""
def __init__(
self,
dim: int,
out_dim: int,
depth: int,
mlp_ratio: float = 4.,
downsample: bool = True,
focal_level: int = 1,
... | class FocalNetStage(nn.Module):
""" A basic Focal Transformer layer for one stage.
"""
def __init__(
self,
dim: int,
out_dim: int,
depth: int,
mlp_ratio: float = 4.,
downsample: bool = True,
focal_level: int = 1,
... | x = self.downsample(x)
for blk in self.blocks:
if self.grad_checkpointing and not torch.jit.is_scripting():
x = checkpoint.checkpoint(blk, x)
else:
x = blk(x)
return x
| x = self.downsample(x)
for blk in self.blocks:
if self.grad_checkpointing and not torch.jit.is_scripting():
x = checkpoint.checkpoint(blk, x)
else:
x = blk(x)
return x
| 22 | 775 | 87 | 862 | """ FocalNet
As described in `Focal Modulation Networks` - https://arxiv.org/abs/2203.11926
Significant modifications and refactoring from the original impl at https://github.com/microsoft/FocalNet
This impl is/has:
* fully convolutional, NCHW tensor layout throughout, seemed to have minimal performance impact but m... | """ FocalNet
As described in `Focal Modulation Networks` - https://arxiv.org/abs/2203.11926
Significant modifications and refactoring from the original impl at https://github.com/microsoft/FocalNet
This impl is/has:
* fully convolutional, NCHW tensor layout throughout, seemed to have minimal performance impact but m... |
17,665 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/focalnet.py | 110,111 | 104,115 | forward | LayerScale2d | 111,115 | 111,115 | class LayerScale2d(nn.Module):
def __init__(self, dim, init_values=1e-5, inplace=False):
super().__init__()
self.inplace = inplace
self.gamma = nn.Parameter(init_values * torch.ones(dim))
def forward(self, x):
FILL_FUNC_BODY | class LayerScale2d(nn.Module):
def __init__(self, dim, init_values=1e-5, inplace=False):
super().__init__()
self.inplace = inplace
self.gamma = nn.Parameter(init_values * torch.ones(dim))
def forward(self, x):
FILL_FUNC_BODY | gamma = self.gamma.view(1, -1, 1, 1)
return x.mul_(gamma) if self.inplace else x * gamma
| gamma = self.gamma.view(1, -1, 1, 1)
return x.mul_(gamma) if self.inplace else x * gamma
| 14 | 95 | 46 | 141 | """ FocalNet
As described in `Focal Modulation Networks` - https://arxiv.org/abs/2203.11926
Significant modifications and refactoring from the original impl at https://github.com/microsoft/FocalNet
This impl is/has:
* fully convolutional, NCHW tensor layout throughout, seemed to have minimal performance impact but m... | """ FocalNet
As described in `Focal Modulation Networks` - https://arxiv.org/abs/2203.11926
Significant modifications and refactoring from the original impl at https://github.com/microsoft/FocalNet
This impl is/has:
* fully convolutional, NCHW tensor layout throughout, seemed to have minimal performance impact but m... |
17,667 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/gcvit.py | 102,103 | 77,110 | forward | Downsample2d | 103,110 | 103,110 | class Downsample2d(nn.Module):
def __init__(
self,
dim,
dim_out=None,
reduction='conv',
act_layer=nn.GELU,
norm_layer=LayerNorm2d, # NOTE in NCHW
):
super().__init__()
dim_out = dim_out or dim
self.norm1 = norm_lay... | class Downsample2d(nn.Module):
def __init__(
self,
dim,
dim_out=None,
reduction='conv',
act_layer=nn.GELU,
norm_layer=LayerNorm2d, # NOTE in NCHW
):
super().__init__()
dim_out = dim_out or dim
self.norm1 = norm_lay... | x = self.norm1(x)
x = self.conv_block(x)
x = self.reduction(x)
x = self.norm2(x)
return x
| x = self.norm1(x)
x = self.conv_block(x)
x = self.reduction(x)
x = self.norm2(x)
return x
| 14 | 330 | 58 | 388 | """ Global Context ViT
From scratch implementation of GCViT in the style of timm swin_transformer_v2_cr.py
Global Context Vision Transformers -https://arxiv.org/abs/2206.09959
@article{hatamizadeh2022global,
title={Global Context Vision Transformers},
author={Hatamizadeh, Ali and Yin, Hongxu and Kautz, Jan and M... | """ Global Context ViT
From scratch implementation of GCViT in the style of timm swin_transformer_v2_cr.py
Global Context Vision Transformers -https://arxiv.org/abs/2206.09959
@article{hatamizadeh2022global,
title={Global Context Vision Transformers},
author={Hatamizadeh, Ali and Yin, Hongxu and Kautz, Jan and M... |
17,668 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/gcvit.py | 132,133 | 110,136 | forward | FeatureBlock | 133,136 | 133,136 | class FeatureBlock(nn.Module):
def __init__(
self,
dim,
levels=0,
reduction='max',
act_layer=nn.GELU,
):
super().__init__()
reductions = levels
levels = max(1, levels)
if reduction == 'avg':
pool_fn = partial... | class FeatureBlock(nn.Module):
def __init__(
self,
dim,
levels=0,
reduction='max',
act_layer=nn.GELU,
):
super().__init__()
reductions = levels
levels = max(1, levels)
if reduction == 'avg':
pool_fn = partial... | return self.blocks(x)
| return self.blocks(x)
| 2 | 255 | 14 | 269 | """ Global Context ViT
From scratch implementation of GCViT in the style of timm swin_transformer_v2_cr.py
Global Context Vision Transformers -https://arxiv.org/abs/2206.09959
@article{hatamizadeh2022global,
title={Global Context Vision Transformers},
author={Hatamizadeh, Ali and Yin, Hongxu and Kautz, Jan and M... | """ Global Context ViT
From scratch implementation of GCViT in the style of timm swin_transformer_v2_cr.py
Global Context Vision Transformers -https://arxiv.org/abs/2206.09959
@article{hatamizadeh2022global,
title={Global Context Vision Transformers},
author={Hatamizadeh, Ali and Yin, Hongxu and Kautz, Jan and M... |
17,669 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/gcvit.py | 475,477 | 372,512 | group_matcher | GlobalContextVit | 477,483 | 477,483 | class GlobalContextVit(nn.Module):
def __init__(
self,
in_chans: int = 3,
num_classes: int = 1000,
global_pool: str = 'avg',
img_size: Tuple[int, int] = 224,
window_ratio: Tuple[int, ...] = (32, 32, 16, 32),
window_size: Tuple[int, ... | class GlobalContextVit(nn.Module):
def __init__(
self,
in_chans: int = 3,
num_classes: int = 1000,
global_pool: str = 'avg',
img_size: Tuple[int, int] = 224,
window_ratio: Tuple[int, ...] = (32, 32, 16, 32),
window_size: Tuple[int, ... | matcher = dict(
stem=r'^stem', # stem and embed
blocks=r'^stages\.(\d+)'
)
return matcher
| matcher = dict(
stem=r'^stem', # stem and embed
blocks=r'^stages\.(\d+)'
)
return matcher
| 12 | 1,713 | 52 | 1,765 | """ Global Context ViT
From scratch implementation of GCViT in the style of timm swin_transformer_v2_cr.py
Global Context Vision Transformers -https://arxiv.org/abs/2206.09959
@article{hatamizadeh2022global,
title={Global Context Vision Transformers},
author={Hatamizadeh, Ali and Yin, Hongxu and Kautz, Jan and M... | """ Global Context ViT
From scratch implementation of GCViT in the style of timm swin_transformer_v2_cr.py
Global Context Vision Transformers -https://arxiv.org/abs/2206.09959
@article{hatamizadeh2022global,
title={Global Context Vision Transformers},
author={Hatamizadeh, Ali and Yin, Hongxu and Kautz, Jan and M... |
17,670 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/gcvit.py | 278,279 | 236,292 | _window_attn | GlobalContextVitBlock | 279,286 | 279,286 | class GlobalContextVitBlock(nn.Module):
def __init__(
self,
dim: int,
feat_size: Tuple[int, int],
num_heads: int,
window_size: int = 7,
mlp_ratio: float = 4.,
use_global: bool = True,
qkv_bias: bool = True,
l... | class GlobalContextVitBlock(nn.Module):
def __init__(
self,
dim: int,
feat_size: Tuple[int, int],
num_heads: int,
window_size: int = 7,
mlp_ratio: float = 4.,
use_global: bool = True,
qkv_bias: bool = True,
l... | B, H, W, C = x.shape
x_win = window_partition(x, self.window_size)
x_win = x_win.view(-1, self.window_size[0] * self.window_size[1], C)
attn_win = self.attn(x_win, q_global)
x = window_reverse(attn_win, self.window_size, (H, W))
return x
| B, H, W, C = x.shape
x_win = window_partition(x, self.window_size)
x_win = x_win.view(-1, self.window_size[0] * self.window_size[1], C)
attn_win = self.attn(x_win, q_global)
x = window_reverse(attn_win, self.window_size, (H, W))
return x
| 29 | 696 | 121 | 817 | """ Global Context ViT
From scratch implementation of GCViT in the style of timm swin_transformer_v2_cr.py
Global Context Vision Transformers -https://arxiv.org/abs/2206.09959
@article{hatamizadeh2022global,
title={Global Context Vision Transformers},
author={Hatamizadeh, Ali and Yin, Hongxu and Kautz, Jan and M... | """ Global Context ViT
From scratch implementation of GCViT in the style of timm swin_transformer_v2_cr.py
Global Context Vision Transformers -https://arxiv.org/abs/2206.09959
@article{hatamizadeh2022global,
title={Global Context Vision Transformers},
author={Hatamizadeh, Ali and Yin, Hongxu and Kautz, Jan and M... |
17,671 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/gcvit.py | 354,355 | 292,372 | forward | GlobalContextVitStage | 355,372 | 355,372 | class GlobalContextVitStage(nn.Module):
def __init__(
self,
dim,
depth: int,
num_heads: int,
feat_size: Tuple[int, int],
window_size: Tuple[int, int],
downsample: bool = True,
global_norm: bool = False,
stage... | class GlobalContextVitStage(nn.Module):
def __init__(
self,
dim,
depth: int,
num_heads: int,
feat_size: Tuple[int, int],
window_size: Tuple[int, int],
downsample: bool = True,
global_norm: bool = False,
stage... | # input NCHW, downsample & global block are 2d conv + pooling
x = self.downsample(x)
global_query = self.global_block(x)
# reshape NCHW --> NHWC for transformer blocks
x = x.permute(0, 2, 3, 1)
global_query = self.global_norm(global_query.permute(0, 2, 3, 1))
for... | # input NCHW, downsample & global block are 2d conv + pooling
x = self.downsample(x)
global_query = self.global_block(x)
# reshape NCHW --> NHWC for transformer blocks
x = x.permute(0, 2, 3, 1)
global_query = self.global_norm(global_query.permute(0, 2, 3, 1))
for... | 71 | 711 | 247 | 958 | """ Global Context ViT
From scratch implementation of GCViT in the style of timm swin_transformer_v2_cr.py
Global Context Vision Transformers -https://arxiv.org/abs/2206.09959
@article{hatamizadeh2022global,
title={Global Context Vision Transformers},
author={Hatamizadeh, Ali and Yin, Hongxu and Kautz, Jan and M... | """ Global Context ViT
From scratch implementation of GCViT in the style of timm swin_transformer_v2_cr.py
Global Context Vision Transformers -https://arxiv.org/abs/2206.09959
@article{hatamizadeh2022global,
title={Global Context Vision Transformers},
author={Hatamizadeh, Ali and Yin, Hongxu and Kautz, Jan and M... |
17,672 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/gcvit.py | 232,233 | 226,236 | forward | LayerScale | 233,236 | 233,236 | class LayerScale(nn.Module):
def __init__(self, dim, init_values=1e-5, inplace=False):
super().__init__()
self.inplace = inplace
self.gamma = nn.Parameter(init_values * torch.ones(dim))
def forward(self, x):
FILL_FUNC_BODY | class LayerScale(nn.Module):
def __init__(self, dim, init_values=1e-5, inplace=False):
super().__init__()
self.inplace = inplace
self.gamma = nn.Parameter(init_values * torch.ones(dim))
def forward(self, x):
FILL_FUNC_BODY | return x.mul_(self.gamma) if self.inplace else x * self.gamma
| return x.mul_(self.gamma) if self.inplace else x * self.gamma
| 8 | 93 | 28 | 121 | """ Global Context ViT
From scratch implementation of GCViT in the style of timm swin_transformer_v2_cr.py
Global Context Vision Transformers -https://arxiv.org/abs/2206.09959
@article{hatamizadeh2022global,
title={Global Context Vision Transformers},
author={Hatamizadeh, Ali and Yin, Hongxu and Kautz, Jan and M... | """ Global Context ViT
From scratch implementation of GCViT in the style of timm swin_transformer_v2_cr.py
Global Context Vision Transformers -https://arxiv.org/abs/2206.09959
@article{hatamizadeh2022global,
title={Global Context Vision Transformers},
author={Hatamizadeh, Ali and Yin, Hongxu and Kautz, Jan and M... |
17,673 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/gcvit.py | 67,68 | 41,77 | forward | MbConvBlock | 68,77 | 68,77 | class MbConvBlock(nn.Module):
""" A depthwise separable / fused mbconv style residual block with SE, `no norm.
"""
def __init__(
self,
in_chs,
out_chs=None,
expand_ratio=1.0,
attn_layer='se',
bias=False,
act_layer=nn.GELU,
... | class MbConvBlock(nn.Module):
""" A depthwise separable / fused mbconv style residual block with SE, `no norm.
"""
def __init__(
self,
in_chs,
out_chs=None,
expand_ratio=1.0,
attn_layer='se',
bias=False,
act_layer=nn.GELU,
... | shortcut = x
x = self.conv_dw(x)
x = self.act(x)
x = self.se(x)
x = self.conv_pw(x)
x = x + shortcut
return x
| shortcut = x
x = self.conv_dw(x)
x = self.act(x)
x = self.se(x)
x = self.conv_pw(x)
x = x + shortcut
return x
| 22 | 384 | 73 | 457 | """ Global Context ViT
From scratch implementation of GCViT in the style of timm swin_transformer_v2_cr.py
Global Context Vision Transformers -https://arxiv.org/abs/2206.09959
@article{hatamizadeh2022global,
title={Global Context Vision Transformers},
author={Hatamizadeh, Ali and Yin, Hongxu and Kautz, Jan and M... | """ Global Context ViT
From scratch implementation of GCViT in the style of timm swin_transformer_v2_cr.py
Global Context Vision Transformers -https://arxiv.org/abs/2206.09959
@article{hatamizadeh2022global,
title={Global Context Vision Transformers},
author={Hatamizadeh, Ali and Yin, Hongxu and Kautz, Jan and M... |
17,674 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/gcvit.py | 148,149 | 136,154 | forward | Stem | 149,154 | 149,154 | class Stem(nn.Module):
def __init__(
self,
in_chs: int = 3,
out_chs: int = 96,
act_layer: Callable = nn.GELU,
norm_layer: Callable = LayerNorm2d, # NOTE stem in NCHW
):
super().__init__()
self.conv1 = nn.Conv2d(in_chs, out_chs, kernel_... | class Stem(nn.Module):
def __init__(
self,
in_chs: int = 3,
out_chs: int = 96,
act_layer: Callable = nn.GELU,
norm_layer: Callable = LayerNorm2d, # NOTE stem in NCHW
):
super().__init__()
self.conv1 = nn.Conv2d(in_chs, out_chs, kernel_... | x = self.conv1(x)
x = self.down(x)
return x
| x = self.conv1(x)
x = self.down(x)
return x
| 8 | 186 | 32 | 218 | """ Global Context ViT
From scratch implementation of GCViT in the style of timm swin_transformer_v2_cr.py
Global Context Vision Transformers -https://arxiv.org/abs/2206.09959
@article{hatamizadeh2022global,
title={Global Context Vision Transformers},
author={Hatamizadeh, Ali and Yin, Hongxu and Kautz, Jan and M... | """ Global Context ViT
From scratch implementation of GCViT in the style of timm swin_transformer_v2_cr.py
Global Context Vision Transformers -https://arxiv.org/abs/2206.09959
@article{hatamizadeh2022global,
title={Global Context Vision Transformers},
author={Hatamizadeh, Ali and Yin, Hongxu and Kautz, Jan and M... |
17,675 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/gcvit.py | 183,184 | 154,210 | forward | WindowAttentionGlobal | 184,210 | 184,210 | class WindowAttentionGlobal(nn.Module):
def __init__(
self,
dim: int,
num_heads: int,
window_size: Tuple[int, int],
use_global: bool = True,
qkv_bias: bool = True,
attn_drop: float = 0.,
proj_drop: float = 0.,
):
... | class WindowAttentionGlobal(nn.Module):
def __init__(
self,
dim: int,
num_heads: int,
window_size: Tuple[int, int],
use_global: bool = True,
qkv_bias: bool = True,
attn_drop: float = 0.,
proj_drop: float = 0.,
):
... | B, N, C = x.shape
if self.use_global and q_global is not None:
_assert(x.shape[-1] == q_global.shape[-1], 'x and q_global seq lengths should be equal')
kv = self.qkv(x)
kv = kv.reshape(B, N, 2, self.num_heads, self.head_dim).permute(2, 0, 3, 1, 4)
k, v = ... | B, N, C = x.shape
if self.use_global and q_global is not None:
_assert(x.shape[-1] == q_global.shape[-1], 'x and q_global seq lengths should be equal')
kv = self.qkv(x)
kv = kv.reshape(B, N, 2, self.num_heads, self.head_dim).permute(2, 0, 3, 1, 4)
k, v = ... | 122 | 379 | 464 | 843 | """ Global Context ViT
From scratch implementation of GCViT in the style of timm swin_transformer_v2_cr.py
Global Context Vision Transformers -https://arxiv.org/abs/2206.09959
@article{hatamizadeh2022global,
title={Global Context Vision Transformers},
author={Hatamizadeh, Ali and Yin, Hongxu and Kautz, Jan and M... | """ Global Context ViT
From scratch implementation of GCViT in the style of timm swin_transformer_v2_cr.py
Global Context Vision Transformers -https://arxiv.org/abs/2206.09959
@article{hatamizadeh2022global,
title={Global Context Vision Transformers},
author={Hatamizadeh, Ali and Yin, Hongxu and Kautz, Jan and M... |
17,677 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/ghostnet.py | 164,165 | 111,186 | forward | GhostBottleneck | 165,186 | 165,186 | class GhostBottleneck(nn.Module):
""" Ghost bottleneck w/ optional SE"""
def __init__(
self,
in_chs,
mid_chs,
out_chs,
dw_kernel_size=3,
stride=1,
act_layer=nn.ReLU,
se_ratio=0.,
mode='original',
):
... | class GhostBottleneck(nn.Module):
""" Ghost bottleneck w/ optional SE"""
def __init__(
self,
in_chs,
mid_chs,
out_chs,
dw_kernel_size=3,
stride=1,
act_layer=nn.ReLU,
se_ratio=0.,
mode='original',
):
... | shortcut = x
# 1st ghost bottleneck
x = self.ghost1(x)
# Depth-wise convolution
if self.conv_dw is not None:
x = self.conv_dw(x)
x = self.bn_dw(x)
# Squeeze-and-excitation
if self.se is not None:
x = self.se(x)
# 2nd... | shortcut = x
# 1st ghost bottleneck
x = self.ghost1(x)
# Depth-wise convolution
if self.conv_dw is not None:
x = self.conv_dw(x)
x = self.bn_dw(x)
# Squeeze-and-excitation
if self.se is not None:
x = self.se(x)
# 2nd... | 46 | 668 | 175 | 843 | """
An implementation of GhostNet & GhostNetV2 Models as defined in:
GhostNet: More Features from Cheap Operations. https://arxiv.org/abs/1911.11907
GhostNetV2: Enhance Cheap Operation with Long-Range Attention. https://proceedings.neurips.cc/paper_files/paper/2022/file/40b60852a4abdaa696b5a1a78da34635-Paper-Conference... | """
An implementation of GhostNet & GhostNetV2 Models as defined in:
GhostNet: More Features from Cheap Operations. https://arxiv.org/abs/1911.11907
GhostNetV2: Enhance Cheap Operation with Long-Range Attention. https://proceedings.neurips.cc/paper_files/paper/2022/file/40b60852a4abdaa696b5a1a78da34635-Paper-Conference... |
17,678 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/ghostnet.py | 59,60 | 30,66 | forward | GhostModule | 60,66 | 60,66 | class GhostModule(nn.Module):
def __init__(
self,
in_chs,
out_chs,
kernel_size=1,
ratio=2,
dw_size=3,
stride=1,
use_act=True,
act_layer=nn.ReLU,
):
super(GhostModule, self).__init__()
self... | class GhostModule(nn.Module):
def __init__(
self,
in_chs,
out_chs,
kernel_size=1,
ratio=2,
dw_size=3,
stride=1,
use_act=True,
act_layer=nn.ReLU,
):
super(GhostModule, self).__init__()
self... | x1 = self.primary_conv(x)
x2 = self.cheap_operation(x1)
out = torch.cat([x1, x2], dim=1)
return out[:, :self.out_chs, :, :]
| x1 = self.primary_conv(x)
x2 = self.cheap_operation(x1)
out = torch.cat([x1, x2], dim=1)
return out[:, :self.out_chs, :, :]
| 16 | 362 | 73 | 435 | """
An implementation of GhostNet & GhostNetV2 Models as defined in:
GhostNet: More Features from Cheap Operations. https://arxiv.org/abs/1911.11907
GhostNetV2: Enhance Cheap Operation with Long-Range Attention. https://proceedings.neurips.cc/paper_files/paper/2022/file/40b60852a4abdaa696b5a1a78da34635-Paper-Conference... | """
An implementation of GhostNet & GhostNetV2 Models as defined in:
GhostNet: More Features from Cheap Operations. https://arxiv.org/abs/1911.11907
GhostNetV2: Enhance Cheap Operation with Long-Range Attention. https://proceedings.neurips.cc/paper_files/paper/2022/file/40b60852a4abdaa696b5a1a78da34635-Paper-Conference... |
17,679 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/ghostnet.py | 102,103 | 66,111 | forward | GhostModuleV2 | 103,111 | 103,111 | class GhostModuleV2(nn.Module):
def __init__(
self,
in_chs,
out_chs,
kernel_size=1,
ratio=2,
dw_size=3,
stride=1,
use_act=True,
act_layer=nn.ReLU,
):
super().__init__()
self.gate_fn = nn.S... | class GhostModuleV2(nn.Module):
def __init__(
self,
in_chs,
out_chs,
kernel_size=1,
ratio=2,
dw_size=3,
stride=1,
use_act=True,
act_layer=nn.ReLU,
):
super().__init__()
self.gate_fn = nn.S... | res = self.short_conv(F.avg_pool2d(x, kernel_size=2, stride=2))
x1 = self.primary_conv(x)
x2 = self.cheap_operation(x1)
out = torch.cat([x1, x2], dim=1)
return out[:, :self.out_chs, :, :] * F.interpolate(
self.gate_fn(res), size=(out.shape[-2], out.shape[-1]), mode='n... | res = self.short_conv(F.avg_pool2d(x, kernel_size=2, stride=2))
x1 = self.primary_conv(x)
x2 = self.cheap_operation(x1)
out = torch.cat([x1, x2], dim=1)
return out[:, :self.out_chs, :, :] * F.interpolate(
self.gate_fn(res), size=(out.shape[-2], out.shape[-1]), mode='n... | 27 | 581 | 146 | 727 | """
An implementation of GhostNet & GhostNetV2 Models as defined in:
GhostNet: More Features from Cheap Operations. https://arxiv.org/abs/1911.11907
GhostNetV2: Enhance Cheap Operation with Long-Range Attention. https://proceedings.neurips.cc/paper_files/paper/2022/file/40b60852a4abdaa696b5a1a78da34635-Paper-Conference... | """
An implementation of GhostNet & GhostNetV2 Models as defined in:
GhostNet: More Features from Cheap Operations. https://arxiv.org/abs/1911.11907
GhostNetV2: Enhance Cheap Operation with Long-Range Attention. https://proceedings.neurips.cc/paper_files/paper/2022/file/40b60852a4abdaa696b5a1a78da34635-Paper-Conference... |
17,680 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/ghostnet.py | 274,275 | 186,307 | reset_classifier | GhostNet | 275,281 | 275,281 | class GhostNet(nn.Module):
def __init__(
self,
cfgs,
num_classes=1000,
width=1.0,
in_chans=3,
output_stride=32,
global_pool='avg',
drop_rate=0.2,
version='v1',
):
super(GhostNet, self).__init__()
... | class GhostNet(nn.Module):
def __init__(
self,
cfgs,
num_classes=1000,
width=1.0,
in_chans=3,
output_stride=32,
global_pool='avg',
drop_rate=0.2,
version='v1',
):
super(GhostNet, self).__init__()
... | self.num_classes = num_classes
# cannot meaningfully change pooling of efficient head after creation
self.global_pool = SelectAdaptivePool2d(pool_type=global_pool)
self.flatten = nn.Flatten(1) if global_pool else nn.Identity() # don't flatten if pooling disabled
self.classifier ... | self.num_classes = num_classes
# cannot meaningfully change pooling of efficient head after creation
self.global_pool = SelectAdaptivePool2d(pool_type=global_pool)
self.flatten = nn.Flatten(1) if global_pool else nn.Identity() # don't flatten if pooling disabled
self.classifier ... | 39 | 1,397 | 125 | 1,522 | """
An implementation of GhostNet & GhostNetV2 Models as defined in:
GhostNet: More Features from Cheap Operations. https://arxiv.org/abs/1911.11907
GhostNetV2: Enhance Cheap Operation with Long-Range Attention. https://proceedings.neurips.cc/paper_files/paper/2022/file/40b60852a4abdaa696b5a1a78da34635-Paper-Conference... | """
An implementation of GhostNet & GhostNetV2 Models as defined in:
GhostNet: More Features from Cheap Operations. https://arxiv.org/abs/1911.11907
GhostNetV2: Enhance Cheap Operation with Long-Range Attention. https://proceedings.neurips.cc/paper_files/paper/2022/file/40b60852a4abdaa696b5a1a78da34635-Paper-Conference... |
17,683 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/inception_next.py | 83,84 | 57,92 | forward | ConvMlp | 84,92 | 84,92 | class ConvMlp(nn.Module):
""" MLP using 1x1 convs that keeps spatial dims
copied from timm: https://github.com/huggingface/pytorch-image-models/blob/v0.6.11/timm/models/layers/mlp.py
"""
def __init__(
self,
in_features,
hidden_features=None,
out_features=... | class ConvMlp(nn.Module):
""" MLP using 1x1 convs that keeps spatial dims
copied from timm: https://github.com/huggingface/pytorch-image-models/blob/v0.6.11/timm/models/layers/mlp.py
"""
def __init__(
self,
in_features,
hidden_features=None,
out_features=... | x = self.fc1(x)
x = self.norm(x)
x = self.act(x)
x = self.drop(x)
x = self.fc2(x)
return x
| x = self.fc1(x)
x = self.norm(x)
x = self.act(x)
x = self.drop(x)
x = self.fc2(x)
return x
| 17 | 335 | 66 | 401 | """
InceptionNeXt paper: https://arxiv.org/abs/2303.16900
Original implementation & weights from: https://github.com/sail-sg/inceptionnext
"""
from functools import partial
import torch
import torch.nn as nn
from timm.data import IMAGENET_DEFAULT_MEAN, IMAGENET_DEFAULT_STD
from timm.layers import trunc_normal_, Drop... | """
InceptionNeXt paper: https://arxiv.org/abs/2303.16900
Original implementation & weights from: https://github.com/sail-sg/inceptionnext
"""
from functools import partial
import torch
import torch.nn as nn
from timm.data import IMAGENET_DEFAULT_MEAN, IMAGENET_DEFAULT_STD
from timm.layers import trunc_normal_, Drop... |
17,684 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/inception_next.py | 46,47 | 18,57 | forward | InceptionDWConv2d | 47,57 | 47,57 | class InceptionDWConv2d(nn.Module):
""" Inception depthwise convolution
"""
def __init__(
self,
in_chs,
square_kernel_size=3,
band_kernel_size=11,
branch_ratio=0.125,
dilation=1,
):
super().__init__()
gc = int(in_c... | class InceptionDWConv2d(nn.Module):
""" Inception depthwise convolution
"""
def __init__(
self,
in_chs,
square_kernel_size=3,
band_kernel_size=11,
branch_ratio=0.125,
dilation=1,
):
super().__init__()
gc = int(in_c... | x_id, x_hw, x_w, x_h = torch.split(x, self.split_indexes, dim=1)
return torch.cat((
x_id,
self.dwconv_hw(x_hw),
self.dwconv_w(x_w),
self.dwconv_h(x_h)
), dim=1,
)
| x_id, x_hw, x_w, x_h = torch.split(x, self.split_indexes, dim=1)
return torch.cat((
x_id,
self.dwconv_hw(x_hw),
self.dwconv_w(x_w),
self.dwconv_h(x_h)
), dim=1,
)
| 17 | 388 | 114 | 502 | """
InceptionNeXt paper: https://arxiv.org/abs/2303.16900
Original implementation & weights from: https://github.com/sail-sg/inceptionnext
"""
from functools import partial
import torch
import torch.nn as nn
from timm.data import IMAGENET_DEFAULT_MEAN, IMAGENET_DEFAULT_STD
from timm.layers import trunc_normal_, Drop... | """
InceptionNeXt paper: https://arxiv.org/abs/2303.16900
Original implementation & weights from: https://github.com/sail-sg/inceptionnext
"""
from functools import partial
import torch
import torch.nn as nn
from timm.data import IMAGENET_DEFAULT_MEAN, IMAGENET_DEFAULT_STD
from timm.layers import trunc_normal_, Drop... |
17,685 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/inception_next.py | 335,336 | 221,375 | reset_classifier | MetaNeXt | 336,348 | 336,348 | class MetaNeXt(nn.Module):
r""" MetaNeXt
A PyTorch impl of : `InceptionNeXt: When Inception Meets ConvNeXt` - https://arxiv.org/abs/2303.16900
Args:
in_chans (int): Number of input image channels. Default: 3
num_classes (int): Number of classes for classification head. Default: 1000
... | class MetaNeXt(nn.Module):
r""" MetaNeXt
A PyTorch impl of : `InceptionNeXt: When Inception Meets ConvNeXt` - https://arxiv.org/abs/2303.16900
Args:
in_chans (int): Number of input image channels. Default: 3
num_classes (int): Number of classes for classification head. Default: 1000
... | if global_pool is not None:
self.global_pool = global_pool
if num_classes > 0:
if issubclass(head_fn, MlpClassifierHead):
assert self.global_pool, 'Cannot disable global pooling with MLP head present.'
self.head = head_fn(self.num_features, num_classes... | if global_pool is not None:
self.global_pool = global_pool
if num_classes > 0:
if issubclass(head_fn, MlpClassifierHead):
assert self.global_pool, 'Cannot disable global pooling with MLP head present.'
self.head = head_fn(self.num_features, num_classes... | 42 | 1,808 | 176 | 1,984 | """
InceptionNeXt paper: https://arxiv.org/abs/2303.16900
Original implementation & weights from: https://github.com/sail-sg/inceptionnext
"""
from functools import partial
import torch
import torch.nn as nn
from timm.data import IMAGENET_DEFAULT_MEAN, IMAGENET_DEFAULT_STD
from timm.layers import trunc_normal_, Drop... | """
InceptionNeXt paper: https://arxiv.org/abs/2303.16900
Original implementation & weights from: https://github.com/sail-sg/inceptionnext
"""
from functools import partial
import torch
import torch.nn as nn
from timm.data import IMAGENET_DEFAULT_MEAN, IMAGENET_DEFAULT_STD
from timm.layers import trunc_normal_, Drop... |
17,686 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/inception_next.py | 155,156 | 127,166 | forward | MetaNeXtBlock | 156,166 | 156,166 | class MetaNeXtBlock(nn.Module):
""" MetaNeXtBlock Block
Args:
dim (int): Number of input channels.
drop_path (float): Stochastic depth rate. Default: 0.0
ls_init_value (float): Init value for Layer Scale. Default: 1e-6.
"""
def __init__(
self,
dim,
... | class MetaNeXtBlock(nn.Module):
""" MetaNeXtBlock Block
Args:
dim (int): Number of input channels.
drop_path (float): Stochastic depth rate. Default: 0.0
ls_init_value (float): Init value for Layer Scale. Default: 1e-6.
"""
def __init__(
self,
dim,
... | shortcut = x
x = self.token_mixer(x)
x = self.norm(x)
x = self.mlp(x)
if self.gamma is not None:
x = x.mul(self.gamma.reshape(1, -1, 1, 1))
x = self.drop_path(x) + shortcut
return x
| shortcut = x
x = self.token_mixer(x)
x = self.norm(x)
x = self.mlp(x)
if self.gamma is not None:
x = x.mul(self.gamma.reshape(1, -1, 1, 1))
x = self.drop_path(x) + shortcut
return x
| 30 | 350 | 106 | 456 | """
InceptionNeXt paper: https://arxiv.org/abs/2303.16900
Original implementation & weights from: https://github.com/sail-sg/inceptionnext
"""
from functools import partial
import torch
import torch.nn as nn
from timm.data import IMAGENET_DEFAULT_MEAN, IMAGENET_DEFAULT_STD
from timm.layers import trunc_normal_, Drop... | """
InceptionNeXt paper: https://arxiv.org/abs/2303.16900
Original implementation & weights from: https://github.com/sail-sg/inceptionnext
"""
from functools import partial
import torch
import torch.nn as nn
from timm.data import IMAGENET_DEFAULT_MEAN, IMAGENET_DEFAULT_STD
from timm.layers import trunc_normal_, Drop... |
17,687 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/inception_next.py | 212,213 | 166,221 | forward | MetaNeXtStage | 213,221 | 213,221 | class MetaNeXtStage(nn.Module):
def __init__(
self,
in_chs,
out_chs,
stride=2,
depth=2,
dilation=(1, 1),
drop_path_rates=None,
ls_init_value=1.0,
token_mixer=InceptionDWConv2d,
act_layer=nn.GELU,
... | class MetaNeXtStage(nn.Module):
def __init__(
self,
in_chs,
out_chs,
stride=2,
depth=2,
dilation=(1, 1),
drop_path_rates=None,
ls_init_value=1.0,
token_mixer=InceptionDWConv2d,
act_layer=nn.GELU,
... | x = self.downsample(x)
if self.grad_checkpointing and not torch.jit.is_scripting():
x = checkpoint_seq(self.blocks, x)
else:
x = self.blocks(x)
return x
| x = self.downsample(x)
if self.grad_checkpointing and not torch.jit.is_scripting():
x = checkpoint_seq(self.blocks, x)
else:
x = self.blocks(x)
return x
| 18 | 450 | 77 | 527 | """
InceptionNeXt paper: https://arxiv.org/abs/2303.16900
Original implementation & weights from: https://github.com/sail-sg/inceptionnext
"""
from functools import partial
import torch
import torch.nn as nn
from timm.data import IMAGENET_DEFAULT_MEAN, IMAGENET_DEFAULT_STD
from timm.layers import trunc_normal_, Drop... | """
InceptionNeXt paper: https://arxiv.org/abs/2303.16900
Original implementation & weights from: https://github.com/sail-sg/inceptionnext
"""
from functools import partial
import torch
import torch.nn as nn
from timm.data import IMAGENET_DEFAULT_MEAN, IMAGENET_DEFAULT_STD
from timm.layers import trunc_normal_, Drop... |
17,688 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/inception_next.py | 117,118 | 92,127 | forward | MlpClassifierHead | 118,127 | 118,127 | class MlpClassifierHead(nn.Module):
""" MLP classification head
"""
def __init__(
self,
dim,
num_classes=1000,
pool_type='avg',
mlp_ratio=3,
act_layer=nn.GELU,
norm_layer=partial(nn.LayerNorm, eps=1e-6),
drop=0.... | class MlpClassifierHead(nn.Module):
""" MLP classification head
"""
def __init__(
self,
dim,
num_classes=1000,
pool_type='avg',
mlp_ratio=3,
act_layer=nn.GELU,
norm_layer=partial(nn.LayerNorm, eps=1e-6),
drop=0.... | x = self.global_pool(x)
x = self.fc1(x)
x = self.act(x)
x = self.norm(x)
x = self.drop(x)
x = self.fc2(x)
return x
| x = self.global_pool(x)
x = self.fc1(x)
x = self.act(x)
x = self.norm(x)
x = self.drop(x)
x = self.fc2(x)
return x
| 20 | 296 | 79 | 375 | """
InceptionNeXt paper: https://arxiv.org/abs/2303.16900
Original implementation & weights from: https://github.com/sail-sg/inceptionnext
"""
from functools import partial
import torch
import torch.nn as nn
from timm.data import IMAGENET_DEFAULT_MEAN, IMAGENET_DEFAULT_STD
from timm.layers import trunc_normal_, Drop... | """
InceptionNeXt paper: https://arxiv.org/abs/2303.16900
Original implementation & weights from: https://github.com/sail-sg/inceptionnext
"""
from functools import partial
import torch
import torch.nn as nn
from timm.data import IMAGENET_DEFAULT_MEAN, IMAGENET_DEFAULT_STD
from timm.layers import trunc_normal_, Drop... |
17,690 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/inception_resnet_v2.py | 124,125 | 107,134 | forward | Block17 | 125,134 | 125,134 | class Block17(nn.Module):
def __init__(self, scale=1.0, conv_block=None):
super(Block17, self).__init__()
self.scale = scale
conv_block = conv_block or ConvNormAct
self.branch0 = conv_block(1088, 192, kernel_size=1, stride=1)
self.branch1 = nn.Sequential(
conv_b... | class Block17(nn.Module):
def __init__(self, scale=1.0, conv_block=None):
super(Block17, self).__init__()
self.scale = scale
conv_block = conv_block or ConvNormAct
self.branch0 = conv_block(1088, 192, kernel_size=1, stride=1)
self.branch1 = nn.Sequential(
conv_b... | x0 = self.branch0(x)
x1 = self.branch1(x)
out = torch.cat((x0, x1), 1)
out = self.conv2d(out)
out = out * self.scale + x
out = self.act(out)
return out
| x0 = self.branch0(x)
x1 = self.branch1(x)
out = torch.cat((x0, x1), 1)
out = self.conv2d(out)
out = out * self.scale + x
out = self.act(out)
return out
| 26 | 290 | 90 | 380 | """ Pytorch Inception-Resnet-V2 implementation
Sourced from https://github.com/Cadene/tensorflow-model-zoo.torch (MIT License) which is
based upon Google's Tensorflow implementation and pretrained weights (Apache 2.0 License)
"""
from functools import partial
import torch
import torch.nn as nn
import torch.nn.functiona... | """ Pytorch Inception-Resnet-V2 implementation
Sourced from https://github.com/Cadene/tensorflow-model-zoo.torch (MIT License) which is
based upon Google's Tensorflow implementation and pretrained weights (Apache 2.0 License)
"""
from functools import partial
import torch
import torch.nn as nn
import torch.nn.functiona... |
17,691 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/inception_resnet_v2.py | 73,74 | 51,84 | forward | Block35 | 74,84 | 74,84 | class Block35(nn.Module):
def __init__(self, scale=1.0, conv_block=None):
super(Block35, self).__init__()
self.scale = scale
conv_block = conv_block or ConvNormAct
self.branch0 = conv_block(320, 32, kernel_size=1, stride=1)
self.branch1 = nn.Sequential(
conv_blo... | class Block35(nn.Module):
def __init__(self, scale=1.0, conv_block=None):
super(Block35, self).__init__()
self.scale = scale
conv_block = conv_block or ConvNormAct
self.branch0 = conv_block(320, 32, kernel_size=1, stride=1)
self.branch1 = nn.Sequential(
conv_blo... | x0 = self.branch0(x)
x1 = self.branch1(x)
x2 = self.branch2(x)
out = torch.cat((x0, x1, x2), 1)
out = self.conv2d(out)
out = out * self.scale + x
out = self.act(out)
return out
| x0 = self.branch0(x)
x1 = self.branch1(x)
x2 = self.branch2(x)
out = torch.cat((x0, x1, x2), 1)
out = self.conv2d(out)
out = out * self.scale + x
out = self.act(out)
return out
| 30 | 336 | 106 | 442 | """ Pytorch Inception-Resnet-V2 implementation
Sourced from https://github.com/Cadene/tensorflow-model-zoo.torch (MIT License) which is
based upon Google's Tensorflow implementation and pretrained weights (Apache 2.0 License)
"""
from functools import partial
import torch
import torch.nn as nn
import torch.nn.functiona... | """ Pytorch Inception-Resnet-V2 implementation
Sourced from https://github.com/Cadene/tensorflow-model-zoo.torch (MIT License) which is
based upon Google's Tensorflow implementation and pretrained weights (Apache 2.0 License)
"""
from functools import partial
import torch
import torch.nn as nn
import torch.nn.functiona... |
17,692 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/inception_resnet_v2.py | 184,185 | 166,195 | forward | Block8 | 185,195 | 185,195 | class Block8(nn.Module):
def __init__(self, scale=1.0, no_relu=False, conv_block=None):
super(Block8, self).__init__()
self.scale = scale
conv_block = conv_block or ConvNormAct
self.branch0 = conv_block(2080, 192, kernel_size=1, stride=1)
self.branch1 = nn.Sequential(
... | class Block8(nn.Module):
def __init__(self, scale=1.0, no_relu=False, conv_block=None):
super(Block8, self).__init__()
self.scale = scale
conv_block = conv_block or ConvNormAct
self.branch0 = conv_block(2080, 192, kernel_size=1, stride=1)
self.branch1 = nn.Sequential(
... | x0 = self.branch0(x)
x1 = self.branch1(x)
out = torch.cat((x0, x1), 1)
out = self.conv2d(out)
out = out * self.scale + x
if self.relu is not None:
out = self.relu(out)
return out
| x0 = self.branch0(x)
x1 = self.branch1(x)
out = torch.cat((x0, x1), 1)
out = self.conv2d(out)
out = out * self.scale + x
if self.relu is not None:
out = self.relu(out)
return out
| 31 | 304 | 103 | 407 | """ Pytorch Inception-Resnet-V2 implementation
Sourced from https://github.com/Cadene/tensorflow-model-zoo.torch (MIT License) which is
based upon Google's Tensorflow implementation and pretrained weights (Apache 2.0 License)
"""
from functools import partial
import torch
import torch.nn as nn
import torch.nn.functiona... | """ Pytorch Inception-Resnet-V2 implementation
Sourced from https://github.com/Cadene/tensorflow-model-zoo.torch (MIT License) which is
based upon Google's Tensorflow implementation and pretrained weights (Apache 2.0 License)
"""
from functools import partial
import torch
import torch.nn as nn
import torch.nn.functiona... |
17,693 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/inception_resnet_v2.py | 280,281 | 195,309 | forward_features | InceptionResnetV2 | 281,298 | 281,298 | class InceptionResnetV2(nn.Module):
def __init__(
self,
num_classes=1000,
in_chans=3,
drop_rate=0.,
output_stride=32,
global_pool='avg',
norm_layer='batchnorm2d',
norm_eps=1e-3,
act_layer='relu',
):
... | class InceptionResnetV2(nn.Module):
def __init__(
self,
num_classes=1000,
in_chans=3,
drop_rate=0.,
output_stride=32,
global_pool='avg',
norm_layer='batchnorm2d',
norm_eps=1e-3,
act_layer='relu',
):
... | x = self.conv2d_1a(x)
x = self.conv2d_2a(x)
x = self.conv2d_2b(x)
x = self.maxpool_3a(x)
x = self.conv2d_3b(x)
x = self.conv2d_4a(x)
x = self.maxpool_5a(x)
x = self.mixed_5b(x)
x = self.repeat(x)
x = self.mixed_6a(x)
x = self.repeat... | x = self.conv2d_1a(x)
x = self.conv2d_2a(x)
x = self.conv2d_2b(x)
x = self.maxpool_3a(x)
x = self.conv2d_3b(x)
x = self.conv2d_4a(x)
x = self.maxpool_5a(x)
x = self.mixed_5b(x)
x = self.repeat(x)
x = self.mixed_6a(x)
x = self.repeat... | 47 | 1,464 | 227 | 1,691 | """ Pytorch Inception-Resnet-V2 implementation
Sourced from https://github.com/Cadene/tensorflow-model-zoo.torch (MIT License) which is
based upon Google's Tensorflow implementation and pretrained weights (Apache 2.0 License)
"""
from functools import partial
import torch
import torch.nn as nn
import torch.nn.functiona... | """ Pytorch Inception-Resnet-V2 implementation
Sourced from https://github.com/Cadene/tensorflow-model-zoo.torch (MIT License) which is
based upon Google's Tensorflow implementation and pretrained weights (Apache 2.0 License)
"""
from functools import partial
import torch
import torch.nn as nn
import torch.nn.functiona... |
17,694 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/inception_resnet_v2.py | 42,43 | 19,51 | forward | Mixed_5b | 43,51 | 43,51 | class Mixed_5b(nn.Module):
def __init__(self, conv_block=None):
super(Mixed_5b, self).__init__()
conv_block = conv_block or ConvNormAct
self.branch0 = conv_block(192, 96, kernel_size=1, stride=1)
self.branch1 = nn.Sequential(
conv_block(192, 48, kernel_size=1, stride=1)... | class Mixed_5b(nn.Module):
def __init__(self, conv_block=None):
super(Mixed_5b, self).__init__()
conv_block = conv_block or ConvNormAct
self.branch0 = conv_block(192, 96, kernel_size=1, stride=1)
self.branch1 = nn.Sequential(
conv_block(192, 48, kernel_size=1, stride=1)... | x0 = self.branch0(x)
x1 = self.branch1(x)
x2 = self.branch2(x)
x3 = self.branch3(x)
out = torch.cat((x0, x1, x2, x3), 1)
return out
| x0 = self.branch0(x)
x1 = self.branch1(x)
x2 = self.branch2(x)
x3 = self.branch3(x)
out = torch.cat((x0, x1, x2, x3), 1)
return out
| 21 | 350 | 86 | 436 | """ Pytorch Inception-Resnet-V2 implementation
Sourced from https://github.com/Cadene/tensorflow-model-zoo.torch (MIT License) which is
based upon Google's Tensorflow implementation and pretrained weights (Apache 2.0 License)
"""
from functools import partial
import torch
import torch.nn as nn
import torch.nn.functiona... | """ Pytorch Inception-Resnet-V2 implementation
Sourced from https://github.com/Cadene/tensorflow-model-zoo.torch (MIT License) which is
based upon Google's Tensorflow implementation and pretrained weights (Apache 2.0 License)
"""
from functools import partial
import torch
import torch.nn as nn
import torch.nn.functiona... |
17,695 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/inception_resnet_v2.py | 99,100 | 84,107 | forward | Mixed_6a | 100,107 | 100,107 | class Mixed_6a(nn.Module):
def __init__(self, conv_block=None):
super(Mixed_6a, self).__init__()
conv_block = conv_block or ConvNormAct
self.branch0 = conv_block(320, 384, kernel_size=3, stride=2)
self.branch1 = nn.Sequential(
conv_block(320, 256, kernel_size=1, stride=... | class Mixed_6a(nn.Module):
def __init__(self, conv_block=None):
super(Mixed_6a, self).__init__()
conv_block = conv_block or ConvNormAct
self.branch0 = conv_block(320, 384, kernel_size=3, stride=2)
self.branch1 = nn.Sequential(
conv_block(320, 256, kernel_size=1, stride=... | x0 = self.branch0(x)
x1 = self.branch1(x)
x2 = self.branch2(x)
out = torch.cat((x0, x1, x2), 1)
return out
| x0 = self.branch0(x)
x1 = self.branch1(x)
x2 = self.branch2(x)
out = torch.cat((x0, x1, x2), 1)
return out
| 17 | 231 | 70 | 301 | """ Pytorch Inception-Resnet-V2 implementation
Sourced from https://github.com/Cadene/tensorflow-model-zoo.torch (MIT License) which is
based upon Google's Tensorflow implementation and pretrained weights (Apache 2.0 License)
"""
from functools import partial
import torch
import torch.nn as nn
import torch.nn.functiona... | """ Pytorch Inception-Resnet-V2 implementation
Sourced from https://github.com/Cadene/tensorflow-model-zoo.torch (MIT License) which is
based upon Google's Tensorflow implementation and pretrained weights (Apache 2.0 License)
"""
from functools import partial
import torch
import torch.nn as nn
import torch.nn.functiona... |
17,696 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/inception_resnet_v2.py | 157,158 | 134,166 | forward | Mixed_7a | 158,166 | 158,166 | class Mixed_7a(nn.Module):
def __init__(self, conv_block=None):
super(Mixed_7a, self).__init__()
conv_block = conv_block or ConvNormAct
self.branch0 = nn.Sequential(
conv_block(1088, 256, kernel_size=1, stride=1),
conv_block(256, 384, kernel_size=3, stride=2)
... | class Mixed_7a(nn.Module):
def __init__(self, conv_block=None):
super(Mixed_7a, self).__init__()
conv_block = conv_block or ConvNormAct
self.branch0 = nn.Sequential(
conv_block(1088, 256, kernel_size=1, stride=1),
conv_block(256, 384, kernel_size=3, stride=2)
... | x0 = self.branch0(x)
x1 = self.branch1(x)
x2 = self.branch2(x)
x3 = self.branch3(x)
out = torch.cat((x0, x1, x2, x3), 1)
return out
| x0 = self.branch0(x)
x1 = self.branch1(x)
x2 = self.branch2(x)
x3 = self.branch3(x)
out = torch.cat((x0, x1, x2, x3), 1)
return out
| 21 | 342 | 86 | 428 | """ Pytorch Inception-Resnet-V2 implementation
Sourced from https://github.com/Cadene/tensorflow-model-zoo.torch (MIT License) which is
based upon Google's Tensorflow implementation and pretrained weights (Apache 2.0 License)
"""
from functools import partial
import torch
import torch.nn as nn
import torch.nn.functiona... | """ Pytorch Inception-Resnet-V2 implementation
Sourced from https://github.com/Cadene/tensorflow-model-zoo.torch (MIT License) which is
based upon Google's Tensorflow implementation and pretrained weights (Apache 2.0 License)
"""
from functools import partial
import torch
import torch.nn as nn
import torch.nn.functiona... |
17,698 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/inception_v3.py | 54,55 | 22,59 | forward | InceptionA | 55,59 | 55,59 | class InceptionA(nn.Module):
def __init__(self, in_channels, pool_features, conv_block=None):
super(InceptionA, self).__init__()
conv_block = conv_block or ConvNormAct
self.branch1x1 = conv_block(in_channels, 64, kernel_size=1)
self.branch5x5_1 = conv_block(in_channels, 48, kernel_... | class InceptionA(nn.Module):
def __init__(self, in_channels, pool_features, conv_block=None):
super(InceptionA, self).__init__()
conv_block = conv_block or ConvNormAct
self.branch1x1 = conv_block(in_channels, 64, kernel_size=1)
self.branch5x5_1 = conv_block(in_channels, 48, kernel_... | outputs = self._forward(x)
return torch.cat(outputs, 1)
| outputs = self._forward(x)
return torch.cat(outputs, 1)
| 6 | 526 | 30 | 556 | """ Inception-V3
Originally from torchvision Inception3 model
Licensed BSD-Clause 3 https://github.com/pytorch/vision/blob/master/LICENSE
"""
from functools import partial
import torch
import torch.nn as nn
import torch.nn.functional as F
from timm.data import IMAGENET_DEFAULT_STD, IMAGENET_DEFAULT_MEAN, IMAGENET_IN... | """ Inception-V3
Originally from torchvision Inception3 model
Licensed BSD-Clause 3 https://github.com/pytorch/vision/blob/master/LICENSE
"""
from functools import partial
import torch
import torch.nn as nn
import torch.nn.functional as F
from timm.data import IMAGENET_DEFAULT_STD, IMAGENET_DEFAULT_MEAN, IMAGENET_IN... |
17,699 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/inception_v3.py | 220,221 | 209,238 | forward | InceptionAux | 221,238 | 221,238 | class InceptionAux(nn.Module):
def __init__(self, in_channels, num_classes, conv_block=None):
super(InceptionAux, self).__init__()
conv_block = conv_block or ConvNormAct
self.conv0 = conv_block(in_channels, 128, kernel_size=1)
self.conv1 = conv_block(128, 768, kernel_size=5)
... | class InceptionAux(nn.Module):
def __init__(self, in_channels, num_classes, conv_block=None):
super(InceptionAux, self).__init__()
conv_block = conv_block or ConvNormAct
self.conv0 = conv_block(in_channels, 128, kernel_size=1)
self.conv1 = conv_block(128, 768, kernel_size=5)
... | # N x 768 x 17 x 17
x = F.avg_pool2d(x, kernel_size=5, stride=3)
# N x 768 x 5 x 5
x = self.conv0(x)
# N x 128 x 5 x 5
x = self.conv1(x)
# N x 768 x 1 x 1
# Adaptive average pooling
x = F.adaptive_avg_pool2d(x, (1, 1))
# N x 768 x 1 x 1
... | # N x 768 x 17 x 17
x = F.avg_pool2d(x, kernel_size=5, stride=3)
# N x 768 x 5 x 5
x = self.conv0(x)
# N x 128 x 5 x 5
x = self.conv1(x)
# N x 768 x 1 x 1
# Adaptive average pooling
x = F.adaptive_avg_pool2d(x, (1, 1))
# N x 768 x 1 x 1
... | 77 | 182 | 225 | 407 | """ Inception-V3
Originally from torchvision Inception3 model
Licensed BSD-Clause 3 https://github.com/pytorch/vision/blob/master/LICENSE
"""
from functools import partial
import torch
import torch.nn as nn
import torch.nn.functional as F
from timm.data import IMAGENET_DEFAULT_STD, IMAGENET_DEFAULT_MEAN, IMAGENET_IN... | """ Inception-V3
Originally from torchvision Inception3 model
Licensed BSD-Clause 3 https://github.com/pytorch/vision/blob/master/LICENSE
"""
from functools import partial
import torch
import torch.nn as nn
import torch.nn.functional as F
from timm.data import IMAGENET_DEFAULT_STD, IMAGENET_DEFAULT_MEAN, IMAGENET_IN... |
17,700 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/inception_v3.py | 70,71 | 59,87 | _forward | InceptionB | 71,82 | 71,82 | class InceptionB(nn.Module):
def __init__(self, in_channels, conv_block=None):
super(InceptionB, self).__init__()
conv_block = conv_block or ConvNormAct
self.branch3x3 = conv_block(in_channels, 384, kernel_size=3, stride=2)
self.branch3x3dbl_1 = conv_block(in_channels, 64, kernel_s... | class InceptionB(nn.Module):
def __init__(self, in_channels, conv_block=None):
super(InceptionB, self).__init__()
conv_block = conv_block or ConvNormAct
self.branch3x3 = conv_block(in_channels, 384, kernel_size=3, stride=2)
self.branch3x3dbl_1 = conv_block(in_channels, 64, kernel_s... | branch3x3 = self.branch3x3(x)
branch3x3dbl = self.branch3x3dbl_1(x)
branch3x3dbl = self.branch3x3dbl_2(branch3x3dbl)
branch3x3dbl = self.branch3x3dbl_3(branch3x3dbl)
branch_pool = F.max_pool2d(x, kernel_size=3, stride=2)
outputs = [branch3x3, branch3x3dbl, branch_pool]... | branch3x3 = self.branch3x3(x)
branch3x3dbl = self.branch3x3dbl_1(x)
branch3x3dbl = self.branch3x3dbl_2(branch3x3dbl)
branch3x3dbl = self.branch3x3dbl_3(branch3x3dbl)
branch_pool = F.max_pool2d(x, kernel_size=3, stride=2)
outputs = [branch3x3, branch3x3dbl, branch_pool]... | 24 | 250 | 159 | 409 | """ Inception-V3
Originally from torchvision Inception3 model
Licensed BSD-Clause 3 https://github.com/pytorch/vision/blob/master/LICENSE
"""
from functools import partial
import torch
import torch.nn as nn
import torch.nn.functional as F
from timm.data import IMAGENET_DEFAULT_STD, IMAGENET_DEFAULT_MEAN, IMAGENET_IN... | """ Inception-V3
Originally from torchvision Inception3 model
Licensed BSD-Clause 3 https://github.com/pytorch/vision/blob/master/LICENSE
"""
from functools import partial
import torch
import torch.nn as nn
import torch.nn.functional as F
from timm.data import IMAGENET_DEFAULT_STD, IMAGENET_DEFAULT_MEAN, IMAGENET_IN... |
17,701 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/inception_v3.py | 107,108 | 87,131 | _forward | InceptionC | 108,126 | 108,126 | class InceptionC(nn.Module):
def __init__(self, in_channels, channels_7x7, conv_block=None):
super(InceptionC, self).__init__()
conv_block = conv_block or ConvNormAct
self.branch1x1 = conv_block(in_channels, 192, kernel_size=1)
c7 = channels_7x7
self.branch7x7_1 = conv_bloc... | class InceptionC(nn.Module):
def __init__(self, in_channels, channels_7x7, conv_block=None):
super(InceptionC, self).__init__()
conv_block = conv_block or ConvNormAct
self.branch1x1 = conv_block(in_channels, 192, kernel_size=1)
c7 = channels_7x7
self.branch7x7_1 = conv_bloc... | branch1x1 = self.branch1x1(x)
branch7x7 = self.branch7x7_1(x)
branch7x7 = self.branch7x7_2(branch7x7)
branch7x7 = self.branch7x7_3(branch7x7)
branch7x7dbl = self.branch7x7dbl_1(x)
branch7x7dbl = self.branch7x7dbl_2(branch7x7dbl)
branch7x7dbl = self.branch7x7dbl_... | branch1x1 = self.branch1x1(x)
branch7x7 = self.branch7x7_1(x)
branch7x7 = self.branch7x7_2(branch7x7)
branch7x7 = self.branch7x7_3(branch7x7)
branch7x7dbl = self.branch7x7dbl_1(x)
branch7x7dbl = self.branch7x7dbl_2(branch7x7dbl)
branch7x7dbl = self.branch7x7dbl_... | 44 | 496 | 307 | 803 | """ Inception-V3
Originally from torchvision Inception3 model
Licensed BSD-Clause 3 https://github.com/pytorch/vision/blob/master/LICENSE
"""
from functools import partial
import torch
import torch.nn as nn
import torch.nn.functional as F
from timm.data import IMAGENET_DEFAULT_STD, IMAGENET_DEFAULT_MEAN, IMAGENET_IN... | """ Inception-V3
Originally from torchvision Inception3 model
Licensed BSD-Clause 3 https://github.com/pytorch/vision/blob/master/LICENSE
"""
from functools import partial
import torch
import torch.nn as nn
import torch.nn.functional as F
from timm.data import IMAGENET_DEFAULT_STD, IMAGENET_DEFAULT_MEAN, IMAGENET_IN... |
17,702 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/inception_v3.py | 144,145 | 131,162 | _forward | InceptionD | 145,157 | 145,157 | class InceptionD(nn.Module):
def __init__(self, in_channels, conv_block=None):
super(InceptionD, self).__init__()
conv_block = conv_block or ConvNormAct
self.branch3x3_1 = conv_block(in_channels, 192, kernel_size=1)
self.branch3x3_2 = conv_block(192, 320, kernel_size=3, stride=2)
... | class InceptionD(nn.Module):
def __init__(self, in_channels, conv_block=None):
super(InceptionD, self).__init__()
conv_block = conv_block or ConvNormAct
self.branch3x3_1 = conv_block(in_channels, 192, kernel_size=1)
self.branch3x3_2 = conv_block(192, 320, kernel_size=3, stride=2)
... | branch3x3 = self.branch3x3_1(x)
branch3x3 = self.branch3x3_2(branch3x3)
branch7x7x3 = self.branch7x7x3_1(x)
branch7x7x3 = self.branch7x7x3_2(branch7x7x3)
branch7x7x3 = self.branch7x7x3_3(branch7x7x3)
branch7x7x3 = self.branch7x7x3_4(branch7x7x3)
branch_pool = F.... | branch3x3 = self.branch3x3_1(x)
branch3x3 = self.branch3x3_2(branch3x3)
branch7x7x3 = self.branch7x7x3_1(x)
branch7x7x3 = self.branch7x7x3_2(branch7x7x3)
branch7x7x3 = self.branch7x7x3_3(branch7x7x3)
branch7x7x3 = self.branch7x7x3_4(branch7x7x3)
branch_pool = F.... | 30 | 339 | 209 | 548 | """ Inception-V3
Originally from torchvision Inception3 model
Licensed BSD-Clause 3 https://github.com/pytorch/vision/blob/master/LICENSE
"""
from functools import partial
import torch
import torch.nn as nn
import torch.nn.functional as F
from timm.data import IMAGENET_DEFAULT_STD, IMAGENET_DEFAULT_MEAN, IMAGENET_IN... | """ Inception-V3
Originally from torchvision Inception3 model
Licensed BSD-Clause 3 https://github.com/pytorch/vision/blob/master/LICENSE
"""
from functools import partial
import torch
import torch.nn as nn
import torch.nn.functional as F
from timm.data import IMAGENET_DEFAULT_STD, IMAGENET_DEFAULT_MEAN, IMAGENET_IN... |
17,703 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/inception_v3.py | 180,181 | 162,209 | _forward | InceptionE | 181,204 | 181,204 | class InceptionE(nn.Module):
def __init__(self, in_channels, conv_block=None):
super(InceptionE, self).__init__()
conv_block = conv_block or ConvNormAct
self.branch1x1 = conv_block(in_channels, 320, kernel_size=1)
self.branch3x3_1 = conv_block(in_channels, 384, kernel_size=1)
... | class InceptionE(nn.Module):
def __init__(self, in_channels, conv_block=None):
super(InceptionE, self).__init__()
conv_block = conv_block or ConvNormAct
self.branch1x1 = conv_block(in_channels, 320, kernel_size=1)
self.branch3x3_1 = conv_block(in_channels, 384, kernel_size=1)
... | branch1x1 = self.branch1x1(x)
branch3x3 = self.branch3x3_1(x)
branch3x3 = [
self.branch3x3_2a(branch3x3),
self.branch3x3_2b(branch3x3),
]
branch3x3 = torch.cat(branch3x3, 1)
branch3x3dbl = self.branch3x3dbl_1(x)
branch3x3dbl = self.branch... | branch1x1 = self.branch1x1(x)
branch3x3 = self.branch3x3_1(x)
branch3x3 = [
self.branch3x3_2a(branch3x3),
self.branch3x3_2b(branch3x3),
]
branch3x3 = torch.cat(branch3x3, 1)
branch3x3dbl = self.branch3x3dbl_1(x)
branch3x3dbl = self.branch... | 49 | 450 | 333 | 783 | """ Inception-V3
Originally from torchvision Inception3 model
Licensed BSD-Clause 3 https://github.com/pytorch/vision/blob/master/LICENSE
"""
from functools import partial
import torch
import torch.nn as nn
import torch.nn.functional as F
from timm.data import IMAGENET_DEFAULT_STD, IMAGENET_DEFAULT_MEAN, IMAGENET_IN... | """ Inception-V3
Originally from torchvision Inception3 model
Licensed BSD-Clause 3 https://github.com/pytorch/vision/blob/master/LICENSE
"""
from functools import partial
import torch
import torch.nn as nn
import torch.nn.functional as F
from timm.data import IMAGENET_DEFAULT_STD, IMAGENET_DEFAULT_MEAN, IMAGENET_IN... |
17,706 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/inception_v4.py | 88,89 | 67,97 | forward | InceptionA | 89,97 | 89,97 | class InceptionA(nn.Module):
def __init__(self, conv_block=ConvNormAct):
super(InceptionA, self).__init__()
self.branch0 = conv_block(384, 96, kernel_size=1, stride=1)
self.branch1 = nn.Sequential(
conv_block(384, 64, kernel_size=1, stride=1),
conv_block(64, 96, kern... | class InceptionA(nn.Module):
def __init__(self, conv_block=ConvNormAct):
super(InceptionA, self).__init__()
self.branch0 = conv_block(384, 96, kernel_size=1, stride=1)
self.branch1 = nn.Sequential(
conv_block(384, 64, kernel_size=1, stride=1),
conv_block(64, 96, kern... | x0 = self.branch0(x)
x1 = self.branch1(x)
x2 = self.branch2(x)
x3 = self.branch3(x)
out = torch.cat((x0, x1, x2, x3), 1)
return out
| x0 = self.branch0(x)
x1 = self.branch1(x)
x2 = self.branch2(x)
x3 = self.branch3(x)
out = torch.cat((x0, x1, x2, x3), 1)
return out
| 21 | 334 | 86 | 420 | """ Pytorch Inception-V4 implementation
Sourced from https://github.com/Cadene/tensorflow-model-zoo.torch (MIT License) which is
based upon Google's Tensorflow implementation and pretrained weights (Apache 2.0 License)
"""
from functools import partial
import torch
import torch.nn as nn
from timm.data import IMAGENET... | """ Pytorch Inception-V4 implementation
Sourced from https://github.com/Cadene/tensorflow-model-zoo.torch (MIT License) which is
based upon Google's Tensorflow implementation and pretrained weights (Apache 2.0 License)
"""
from functools import partial
import torch
import torch.nn as nn
from timm.data import IMAGENET... |
17,707 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/inception_v4.py | 142,143 | 118,151 | forward | InceptionB | 143,151 | 143,151 | class InceptionB(nn.Module):
def __init__(self, conv_block=ConvNormAct):
super(InceptionB, self).__init__()
self.branch0 = conv_block(1024, 384, kernel_size=1, stride=1)
self.branch1 = nn.Sequential(
conv_block(1024, 192, kernel_size=1, stride=1),
conv_block(192, 224... | class InceptionB(nn.Module):
def __init__(self, conv_block=ConvNormAct):
super(InceptionB, self).__init__()
self.branch0 = conv_block(1024, 384, kernel_size=1, stride=1)
self.branch1 = nn.Sequential(
conv_block(1024, 192, kernel_size=1, stride=1),
conv_block(192, 224... | x0 = self.branch0(x)
x1 = self.branch1(x)
x2 = self.branch2(x)
x3 = self.branch3(x)
out = torch.cat((x0, x1, x2, x3), 1)
return out
| x0 = self.branch0(x)
x1 = self.branch1(x)
x2 = self.branch2(x)
x3 = self.branch3(x)
out = torch.cat((x0, x1, x2, x3), 1)
return out
| 21 | 486 | 86 | 572 | """ Pytorch Inception-V4 implementation
Sourced from https://github.com/Cadene/tensorflow-model-zoo.torch (MIT License) which is
based upon Google's Tensorflow implementation and pretrained weights (Apache 2.0 License)
"""
from functools import partial
import torch
import torch.nn as nn
from timm.data import IMAGENET... | """ Pytorch Inception-V4 implementation
Sourced from https://github.com/Cadene/tensorflow-model-zoo.torch (MIT License) which is
based upon Google's Tensorflow implementation and pretrained weights (Apache 2.0 License)
"""
from functools import partial
import torch
import torch.nn as nn
from timm.data import IMAGENET... |
17,708 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/inception_v4.py | 198,199 | 177,219 | forward | InceptionC | 199,219 | 199,219 | class InceptionC(nn.Module):
def __init__(self, conv_block=ConvNormAct):
super(InceptionC, self).__init__()
self.branch0 = conv_block(1536, 256, kernel_size=1, stride=1)
self.branch1_0 = conv_block(1536, 384, kernel_size=1, stride=1)
self.branch1_1a = conv_block(384, 256, kernel_si... | class InceptionC(nn.Module):
def __init__(self, conv_block=ConvNormAct):
super(InceptionC, self).__init__()
self.branch0 = conv_block(1536, 256, kernel_size=1, stride=1)
self.branch1_0 = conv_block(1536, 384, kernel_size=1, stride=1)
self.branch1_1a = conv_block(384, 256, kernel_si... | x0 = self.branch0(x)
x1_0 = self.branch1_0(x)
x1_1a = self.branch1_1a(x1_0)
x1_1b = self.branch1_1b(x1_0)
x1 = torch.cat((x1_1a, x1_1b), 1)
x2_0 = self.branch2_0(x)
x2_1 = self.branch2_1(x2_0)
x2_2 = self.branch2_2(x2_1)
x2_3a = self.branch2_3a(x... | x0 = self.branch0(x)
x1_0 = self.branch1_0(x)
x1_1a = self.branch1_1a(x1_0)
x1_1b = self.branch1_1b(x1_0)
x1 = torch.cat((x1_1a, x1_1b), 1)
x2_0 = self.branch2_0(x)
x2_1 = self.branch2_1(x2_0)
x2_2 = self.branch2_2(x2_1)
x2_3a = self.branch2_3a(x... | 49 | 513 | 282 | 795 | """ Pytorch Inception-V4 implementation
Sourced from https://github.com/Cadene/tensorflow-model-zoo.torch (MIT License) which is
based upon Google's Tensorflow implementation and pretrained weights (Apache 2.0 License)
"""
from functools import partial
import torch
import torch.nn as nn
from timm.data import IMAGENET... | """ Pytorch Inception-V4 implementation
Sourced from https://github.com/Cadene/tensorflow-model-zoo.torch (MIT License) which is
based upon Google's Tensorflow implementation and pretrained weights (Apache 2.0 License)
"""
from functools import partial
import torch
import torch.nn as nn
from timm.data import IMAGENET... |
17,709 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/inception_v4.py | 275,277 | 219,302 | set_grad_checkpointing | InceptionV4 | 277,279 | 277,279 | class InceptionV4(nn.Module):
def __init__(
self,
num_classes=1000,
in_chans=3,
output_stride=32,
drop_rate=0.,
global_pool='avg',
norm_layer='batchnorm2d',
norm_eps=1e-3,
act_layer='relu',
):
sup... | class InceptionV4(nn.Module):
def __init__(
self,
num_classes=1000,
in_chans=3,
output_stride=32,
drop_rate=0.,
global_pool='avg',
norm_layer='batchnorm2d',
norm_eps=1e-3,
act_layer='relu',
):
sup... | assert not enable, 'gradient checkpointing not supported'
| assert not enable, 'gradient checkpointing not supported'
| 7 | 979 | 17 | 996 | """ Pytorch Inception-V4 implementation
Sourced from https://github.com/Cadene/tensorflow-model-zoo.torch (MIT License) which is
based upon Google's Tensorflow implementation and pretrained weights (Apache 2.0 License)
"""
from functools import partial
import torch
import torch.nn as nn
from timm.data import IMAGENET... | """ Pytorch Inception-V4 implementation
Sourced from https://github.com/Cadene/tensorflow-model-zoo.torch (MIT License) which is
based upon Google's Tensorflow implementation and pretrained weights (Apache 2.0 License)
"""
from functools import partial
import torch
import torch.nn as nn
from timm.data import IMAGENET... |
17,710 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/inception_v4.py | 24,25 | 18,31 | forward | Mixed3a | 25,31 | 25,31 | class Mixed3a(nn.Module):
def __init__(self, conv_block=ConvNormAct):
super(Mixed3a, self).__init__()
self.maxpool = nn.MaxPool2d(3, stride=2)
self.conv = conv_block(64, 96, kernel_size=3, stride=2)
def forward(self, x):
FILL_FUNC_BODY | class Mixed3a(nn.Module):
def __init__(self, conv_block=ConvNormAct):
super(Mixed3a, self).__init__()
self.maxpool = nn.MaxPool2d(3, stride=2)
self.conv = conv_block(64, 96, kernel_size=3, stride=2)
def forward(self, x):
FILL_FUNC_BODY | x0 = self.maxpool(x)
x1 = self.conv(x)
out = torch.cat((x0, x1), 1)
return out
| x0 = self.maxpool(x)
x1 = self.conv(x)
out = torch.cat((x0, x1), 1)
return out
| 13 | 113 | 53 | 166 | """ Pytorch Inception-V4 implementation
Sourced from https://github.com/Cadene/tensorflow-model-zoo.torch (MIT License) which is
based upon Google's Tensorflow implementation and pretrained weights (Apache 2.0 License)
"""
from functools import partial
import torch
import torch.nn as nn
from timm.data import IMAGENET... | """ Pytorch Inception-V4 implementation
Sourced from https://github.com/Cadene/tensorflow-model-zoo.torch (MIT License) which is
based upon Google's Tensorflow implementation and pretrained weights (Apache 2.0 License)
"""
from functools import partial
import torch
import torch.nn as nn
from timm.data import IMAGENET... |
17,711 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/inception_v4.py | 47,48 | 31,54 | forward | Mixed4a | 48,54 | 48,54 | class Mixed4a(nn.Module):
def __init__(self, conv_block=ConvNormAct):
super(Mixed4a, self).__init__()
self.branch0 = nn.Sequential(
conv_block(160, 64, kernel_size=1, stride=1),
conv_block(64, 96, kernel_size=3, stride=1)
)
self.branch1 = nn.Sequential(
... | class Mixed4a(nn.Module):
def __init__(self, conv_block=ConvNormAct):
super(Mixed4a, self).__init__()
self.branch0 = nn.Sequential(
conv_block(160, 64, kernel_size=1, stride=1),
conv_block(64, 96, kernel_size=3, stride=1)
)
self.branch1 = nn.Sequential(
... | x0 = self.branch0(x)
x1 = self.branch1(x)
out = torch.cat((x0, x1), 1)
return out
| x0 = self.branch0(x)
x1 = self.branch1(x)
out = torch.cat((x0, x1), 1)
return out
| 13 | 273 | 54 | 327 | """ Pytorch Inception-V4 implementation
Sourced from https://github.com/Cadene/tensorflow-model-zoo.torch (MIT License) which is
based upon Google's Tensorflow implementation and pretrained weights (Apache 2.0 License)
"""
from functools import partial
import torch
import torch.nn as nn
from timm.data import IMAGENET... | """ Pytorch Inception-V4 implementation
Sourced from https://github.com/Cadene/tensorflow-model-zoo.torch (MIT License) which is
based upon Google's Tensorflow implementation and pretrained weights (Apache 2.0 License)
"""
from functools import partial
import torch
import torch.nn as nn
from timm.data import IMAGENET... |
17,712 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/inception_v4.py | 60,61 | 54,67 | forward | Mixed5a | 61,67 | 61,67 | class Mixed5a(nn.Module):
def __init__(self, conv_block=ConvNormAct):
super(Mixed5a, self).__init__()
self.conv = conv_block(192, 192, kernel_size=3, stride=2)
self.maxpool = nn.MaxPool2d(3, stride=2)
def forward(self, x):
FILL_FUNC_BODY | class Mixed5a(nn.Module):
def __init__(self, conv_block=ConvNormAct):
super(Mixed5a, self).__init__()
self.conv = conv_block(192, 192, kernel_size=3, stride=2)
self.maxpool = nn.MaxPool2d(3, stride=2)
def forward(self, x):
FILL_FUNC_BODY | x0 = self.conv(x)
x1 = self.maxpool(x)
out = torch.cat((x0, x1), 1)
return out
| x0 = self.conv(x)
x1 = self.maxpool(x)
out = torch.cat((x0, x1), 1)
return out
| 13 | 115 | 53 | 168 | """ Pytorch Inception-V4 implementation
Sourced from https://github.com/Cadene/tensorflow-model-zoo.torch (MIT License) which is
based upon Google's Tensorflow implementation and pretrained weights (Apache 2.0 License)
"""
from functools import partial
import torch
import torch.nn as nn
from timm.data import IMAGENET... | """ Pytorch Inception-V4 implementation
Sourced from https://github.com/Cadene/tensorflow-model-zoo.torch (MIT License) which is
based upon Google's Tensorflow implementation and pretrained weights (Apache 2.0 License)
"""
from functools import partial
import torch
import torch.nn as nn
from timm.data import IMAGENET... |
17,713 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/inception_v4.py | 110,111 | 97,118 | forward | ReductionA | 111,118 | 111,118 | class ReductionA(nn.Module):
def __init__(self, conv_block=ConvNormAct):
super(ReductionA, self).__init__()
self.branch0 = conv_block(384, 384, kernel_size=3, stride=2)
self.branch1 = nn.Sequential(
conv_block(384, 192, kernel_size=1, stride=1),
conv_block(192, 224, ... | class ReductionA(nn.Module):
def __init__(self, conv_block=ConvNormAct):
super(ReductionA, self).__init__()
self.branch0 = conv_block(384, 384, kernel_size=3, stride=2)
self.branch1 = nn.Sequential(
conv_block(384, 192, kernel_size=1, stride=1),
conv_block(192, 224, ... | x0 = self.branch0(x)
x1 = self.branch1(x)
x2 = self.branch2(x)
out = torch.cat((x0, x1, x2), 1)
return out
| x0 = self.branch0(x)
x1 = self.branch1(x)
x2 = self.branch2(x)
out = torch.cat((x0, x1, x2), 1)
return out
| 17 | 215 | 70 | 285 | """ Pytorch Inception-V4 implementation
Sourced from https://github.com/Cadene/tensorflow-model-zoo.torch (MIT License) which is
based upon Google's Tensorflow implementation and pretrained weights (Apache 2.0 License)
"""
from functools import partial
import torch
import torch.nn as nn
from timm.data import IMAGENET... | """ Pytorch Inception-V4 implementation
Sourced from https://github.com/Cadene/tensorflow-model-zoo.torch (MIT License) which is
based upon Google's Tensorflow implementation and pretrained weights (Apache 2.0 License)
"""
from functools import partial
import torch
import torch.nn as nn
from timm.data import IMAGENET... |
17,714 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/inception_v4.py | 169,170 | 151,177 | forward | ReductionB | 170,177 | 170,177 | class ReductionB(nn.Module):
def __init__(self, conv_block=ConvNormAct):
super(ReductionB, self).__init__()
self.branch0 = nn.Sequential(
conv_block(1024, 192, kernel_size=1, stride=1),
conv_block(192, 192, kernel_size=3, stride=2)
)
self.branch1 = nn.Sequen... | class ReductionB(nn.Module):
def __init__(self, conv_block=ConvNormAct):
super(ReductionB, self).__init__()
self.branch0 = nn.Sequential(
conv_block(1024, 192, kernel_size=1, stride=1),
conv_block(192, 192, kernel_size=3, stride=2)
)
self.branch1 = nn.Sequen... | x0 = self.branch0(x)
x1 = self.branch1(x)
x2 = self.branch2(x)
out = torch.cat((x0, x1, x2), 1)
return out
| x0 = self.branch0(x)
x1 = self.branch1(x)
x2 = self.branch2(x)
out = torch.cat((x0, x1, x2), 1)
return out
| 17 | 302 | 70 | 372 | """ Pytorch Inception-V4 implementation
Sourced from https://github.com/Cadene/tensorflow-model-zoo.torch (MIT License) which is
based upon Google's Tensorflow implementation and pretrained weights (Apache 2.0 License)
"""
from functools import partial
import torch
import torch.nn as nn
from timm.data import IMAGENET... | """ Pytorch Inception-V4 implementation
Sourced from https://github.com/Cadene/tensorflow-model-zoo.torch (MIT License) which is
based upon Google's Tensorflow implementation and pretrained weights (Apache 2.0 License)
"""
from functools import partial
import torch
import torch.nn as nn
from timm.data import IMAGENET... |
17,716 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/levit.py | 210,211 | 166,245 | get_attention_biases | Attention | 211,219 | 211,219 | class Attention(nn.Module):
attention_bias_cache: Dict[str, torch.Tensor]
def __init__(
self,
dim,
key_dim,
num_heads=8,
attn_ratio=4.,
resolution=14,
use_conv=False,
act_layer=nn.SiLU,
):
super().__init... | class Attention(nn.Module):
attention_bias_cache: Dict[str, torch.Tensor]
def __init__(
self,
dim,
key_dim,
num_heads=8,
attn_ratio=4.,
resolution=14,
use_conv=False,
act_layer=nn.SiLU,
):
super().__init... | if torch.jit.is_tracing() or self.training:
return self.attention_biases[:, self.attention_bias_idxs]
else:
device_key = str(device)
if device_key not in self.attention_bias_cache:
self.attention_bias_cache[device_key] = self.attention_biases[:, self.a... | if torch.jit.is_tracing() or self.training:
return self.attention_biases[:, self.attention_bias_idxs]
else:
device_key = str(device)
if device_key not in self.attention_bias_cache:
self.attention_bias_cache[device_key] = self.attention_biases[:, self.a... | 22 | 1,028 | 130 | 1,158 | """ LeViT
Paper: `LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference`
- https://arxiv.org/abs/2104.01136
@article{graham2021levit,
title={LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference},
author={Benjamin Graham and Alaaeldin El-Nouby and Hugo Touvron and Pierre Stoc... | """ LeViT
Paper: `LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference`
- https://arxiv.org/abs/2104.01136
@article{graham2021levit,
title={LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference},
author={Benjamin Graham and Alaaeldin El-Nouby and Hugo Touvron and Pierre Stoc... |
17,717 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/levit.py | 311,312 | 245,346 | get_attention_biases | AttentionDownsample | 312,320 | 312,320 | class AttentionDownsample(nn.Module):
attention_bias_cache: Dict[str, torch.Tensor]
def __init__(
self,
in_dim,
out_dim,
key_dim,
num_heads=8,
attn_ratio=2.0,
stride=2,
resolution=14,
use_conv=False,
... | class AttentionDownsample(nn.Module):
attention_bias_cache: Dict[str, torch.Tensor]
def __init__(
self,
in_dim,
out_dim,
key_dim,
num_heads=8,
attn_ratio=2.0,
stride=2,
resolution=14,
use_conv=False,
... | if torch.jit.is_tracing() or self.training:
return self.attention_biases[:, self.attention_bias_idxs]
else:
device_key = str(device)
if device_key not in self.attention_bias_cache:
self.attention_bias_cache[device_key] = self.attention_biases[:, self.a... | if torch.jit.is_tracing() or self.training:
return self.attention_biases[:, self.attention_bias_idxs]
else:
device_key = str(device)
if device_key not in self.attention_bias_cache:
self.attention_bias_cache[device_key] = self.attention_biases[:, self.a... | 22 | 1,363 | 130 | 1,493 | """ LeViT
Paper: `LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference`
- https://arxiv.org/abs/2104.01136
@article{graham2021levit,
title={LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference},
author={Benjamin Graham and Alaaeldin El-Nouby and Hugo Touvron and Pierre Stoc... | """ LeViT
Paper: `LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference`
- https://arxiv.org/abs/2104.01136
@article{graham2021levit,
title={LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference},
author={Benjamin Graham and Alaaeldin El-Nouby and Hugo Touvron and Pierre Stoc... |
17,718 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/levit.py | 52,54 | 43,69 | fuse | ConvNorm | 54,65 | 54,65 | class ConvNorm(nn.Module):
def __init__(
self, in_chs, out_chs, kernel_size=1, stride=1, padding=0, dilation=1, groups=1, bn_weight_init=1):
super().__init__()
self.linear = nn.Conv2d(in_chs, out_chs, kernel_size, stride, padding, dilation, groups, bias=False)
self.bn = nn.BatchN... | class ConvNorm(nn.Module):
def __init__(
self, in_chs, out_chs, kernel_size=1, stride=1, padding=0, dilation=1, groups=1, bn_weight_init=1):
super().__init__()
self.linear = nn.Conv2d(in_chs, out_chs, kernel_size, stride, padding, dilation, groups, bias=False)
self.bn = nn.BatchN... | c, bn = self.linear, self.bn
w = bn.weight / (bn.running_var + bn.eps) ** 0.5
w = c.weight * w[:, None, None, None]
b = bn.bias - bn.running_mean * bn.weight / (bn.running_var + bn.eps) ** 0.5
m = nn.Conv2d(
w.size(1), w.size(0), w.shape[2:], stride=self.linear.stride... | c, bn = self.linear, self.bn
w = bn.weight / (bn.running_var + bn.eps) ** 0.5
w = c.weight * w[:, None, None, None]
b = bn.bias - bn.running_mean * bn.weight / (bn.running_var + bn.eps) ** 0.5
m = nn.Conv2d(
w.size(1), w.size(0), w.shape[2:], stride=self.linear.stride... | 49 | 207 | 208 | 415 | """ LeViT
Paper: `LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference`
- https://arxiv.org/abs/2104.01136
@article{graham2021levit,
title={LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference},
author={Benjamin Graham and Alaaeldin El-Nouby and Hugo Touvron and Pierre Stoc... | """ LeViT
Paper: `LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference`
- https://arxiv.org/abs/2104.01136
@article{graham2021levit,
title={LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference},
author={Benjamin Graham and Alaaeldin El-Nouby and Hugo Touvron and Pierre Stoc... |
17,719 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/levit.py | 156,157 | 149,166 | forward | Downsample | 157,166 | 157,166 | class Downsample(nn.Module):
def __init__(self, stride, resolution, use_pool=False):
super().__init__()
self.stride = stride
self.resolution = to_2tuple(resolution)
self.pool = nn.AvgPool2d(3, stride=stride, padding=1, count_include_pad=False) if use_pool else None
def forward(s... | class Downsample(nn.Module):
def __init__(self, stride, resolution, use_pool=False):
super().__init__()
self.stride = stride
self.resolution = to_2tuple(resolution)
self.pool = nn.AvgPool2d(3, stride=stride, padding=1, count_include_pad=False) if use_pool else None
def forward(s... | B, N, C = x.shape
x = x.view(B, self.resolution[0], self.resolution[1], C)
if self.pool is not None:
x = self.pool(x.permute(0, 3, 1, 2)).permute(0, 2, 3, 1)
else:
x = x[:, ::self.stride, ::self.stride]
return x.reshape(B, -1, C)
| B, N, C = x.shape
x = x.view(B, self.resolution[0], self.resolution[1], C)
if self.pool is not None:
x = self.pool(x.permute(0, 3, 1, 2)).permute(0, 2, 3, 1)
else:
x = x[:, ::self.stride, ::self.stride]
return x.reshape(B, -1, C)
| 35 | 120 | 134 | 254 | """ LeViT
Paper: `LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference`
- https://arxiv.org/abs/2104.01136
@article{graham2021levit,
title={LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference},
author={Benjamin Graham and Alaaeldin El-Nouby and Hugo Touvron and Pierre Stoc... | """ LeViT
Paper: `LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference`
- https://arxiv.org/abs/2104.01136
@article{graham2021levit,
title={LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference},
author={Benjamin Graham and Alaaeldin El-Nouby and Hugo Touvron and Pierre Stoc... |
17,721 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/levit.py | 457,458 | 421,463 | forward | LevitBlock | 458,463 | 458,463 | class LevitBlock(nn.Module):
def __init__(
self,
dim,
key_dim,
num_heads=8,
attn_ratio=4.,
mlp_ratio=2.,
resolution=14,
use_conv=False,
act_layer=nn.SiLU,
attn_act_layer=None,
drop_pat... | class LevitBlock(nn.Module):
def __init__(
self,
dim,
key_dim,
num_heads=8,
attn_ratio=4.,
mlp_ratio=2.,
resolution=14,
use_conv=False,
act_layer=nn.SiLU,
attn_act_layer=None,
drop_pat... | x = x + self.drop_path1(self.attn(x))
x = x + self.drop_path2(self.mlp(x))
return x
| x = x + self.drop_path1(self.attn(x))
x = x + self.drop_path2(self.mlp(x))
return x
| 12 | 345 | 51 | 396 | """ LeViT
Paper: `LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference`
- https://arxiv.org/abs/2104.01136
@article{graham2021levit,
title={LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference},
author={Benjamin Graham and Alaaeldin El-Nouby and Hugo Touvron and Pierre Stoc... | """ LeViT
Paper: `LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference`
- https://arxiv.org/abs/2104.01136
@article{graham2021levit,
title={LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference},
author={Benjamin Graham and Alaaeldin El-Nouby and Hugo Touvron and Pierre Stoc... |
17,722 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/levit.py | 733,734 | 723,759 | reset_classifier | LevitDistilled | 734,741 | 734,741 | class LevitDistilled(Levit):
def __init__(self, *args, **kwargs):
super().__init__(*args, **kwargs)
self.head_dist = NormLinear(self.num_features, self.num_classes) if self.num_classes > 0 else nn.Identity()
self.distilled_training = False # must set this True to train w/ distillation token... | class LevitDistilled(Levit):
def __init__(self, *args, **kwargs):
super().__init__(*args, **kwargs)
self.head_dist = NormLinear(self.num_features, self.num_classes) if self.num_classes > 0 else nn.Identity()
self.distilled_training = False # must set this True to train w/ distillation token... | self.num_classes = num_classes
if global_pool is not None:
self.global_pool = global_pool
self.head = NormLinear(
self.num_features, num_classes, drop=self.drop_rate) if num_classes > 0 else nn.Identity()
self.head_dist = NormLinear(self.num_features, num_classes)... | self.num_classes = num_classes
if global_pool is not None:
self.global_pool = global_pool
self.head = NormLinear(
self.num_features, num_classes, drop=self.drop_rate) if num_classes > 0 else nn.Identity()
self.head_dist = NormLinear(self.num_features, num_classes)... | 33 | 388 | 114 | 502 | """ LeViT
Paper: `LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference`
- https://arxiv.org/abs/2104.01136
@article{graham2021levit,
title={LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference},
author={Benjamin Graham and Alaaeldin El-Nouby and Hugo Touvron and Pierre Stoc... | """ LeViT
Paper: `LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference`
- https://arxiv.org/abs/2104.01136
@article{graham2021levit,
title={LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference},
author={Benjamin Graham and Alaaeldin El-Nouby and Hugo Touvron and Pierre Stoc... |
17,723 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/levit.py | 415,416 | 376,421 | forward | LevitDownsample | 416,421 | 416,421 | class LevitDownsample(nn.Module):
def __init__(
self,
in_dim,
out_dim,
key_dim,
num_heads=8,
attn_ratio=4.,
mlp_ratio=2.,
act_layer=nn.SiLU,
attn_act_layer=None,
resolution=14,
use_con... | class LevitDownsample(nn.Module):
def __init__(
self,
in_dim,
out_dim,
key_dim,
num_heads=8,
attn_ratio=4.,
mlp_ratio=2.,
act_layer=nn.SiLU,
attn_act_layer=None,
resolution=14,
use_con... | x = self.attn_downsample(x)
x = x + self.drop_path(self.mlp(x))
return x
| x = self.attn_downsample(x)
x = x + self.drop_path(self.mlp(x))
return x
| 10 | 365 | 44 | 409 | """ LeViT
Paper: `LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference`
- https://arxiv.org/abs/2104.01136
@article{graham2021levit,
title={LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference},
author={Benjamin Graham and Alaaeldin El-Nouby and Hugo Touvron and Pierre Stoc... | """ LeViT
Paper: `LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference`
- https://arxiv.org/abs/2104.01136
@article{graham2021levit,
title={LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference},
author={Benjamin Graham and Alaaeldin El-Nouby and Hugo Touvron and Pierre Stoc... |
17,724 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/levit.py | 368,369 | 346,376 | forward | LevitMlp | 369,376 | 369,376 | class LevitMlp(nn.Module):
""" MLP for Levit w/ normalization + ability to switch btw conv and linear
"""
def __init__(
self,
in_features,
hidden_features=None,
out_features=None,
use_conv=False,
act_layer=nn.SiLU,
drop=0.
... | class LevitMlp(nn.Module):
""" MLP for Levit w/ normalization + ability to switch btw conv and linear
"""
def __init__(
self,
in_features,
hidden_features=None,
out_features=None,
use_conv=False,
act_layer=nn.SiLU,
drop=0.
... | x = self.ln1(x)
x = self.act(x)
x = self.drop(x)
x = self.ln2(x)
return x
| x = self.ln1(x)
x = self.act(x)
x = self.drop(x)
x = self.ln2(x)
return x
| 14 | 242 | 55 | 297 | """ LeViT
Paper: `LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference`
- https://arxiv.org/abs/2104.01136
@article{graham2021levit,
title={LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference},
author={Benjamin Graham and Alaaeldin El-Nouby and Hugo Touvron and Pierre Stoc... | """ LeViT
Paper: `LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference`
- https://arxiv.org/abs/2104.01136
@article{graham2021levit,
title={LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference},
author={Benjamin Graham and Alaaeldin El-Nouby and Hugo Touvron and Pierre Stoc... |
17,725 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/levit.py | 518,519 | 463,524 | forward | LevitStage | 519,524 | 519,524 | class LevitStage(nn.Module):
def __init__(
self,
in_dim,
out_dim,
key_dim,
depth=4,
num_heads=8,
attn_ratio=4.0,
mlp_ratio=4.0,
act_layer=nn.SiLU,
attn_act_layer=None,
resolution=14,
... | class LevitStage(nn.Module):
def __init__(
self,
in_dim,
out_dim,
key_dim,
depth=4,
num_heads=8,
attn_ratio=4.0,
mlp_ratio=4.0,
act_layer=nn.SiLU,
attn_act_layer=None,
resolution=14,
... | x = self.downsample(x)
x = self.blocks(x)
return x
| x = self.downsample(x)
x = self.blocks(x)
return x
| 8 | 492 | 32 | 524 | """ LeViT
Paper: `LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference`
- https://arxiv.org/abs/2104.01136
@article{graham2021levit,
title={LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference},
author={Benjamin Graham and Alaaeldin El-Nouby and Hugo Touvron and Pierre Stoc... | """ LeViT
Paper: `LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference`
- https://arxiv.org/abs/2104.01136
@article{graham2021levit,
title={LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference},
author={Benjamin Graham and Alaaeldin El-Nouby and Hugo Touvron and Pierre Stoc... |
17,726 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/levit.py | 88,89 | 69,93 | forward | LinearNorm | 89,93 | 89,93 | class LinearNorm(nn.Module):
def __init__(self, in_features, out_features, bn_weight_init=1):
super().__init__()
self.linear = nn.Linear(in_features, out_features, bias=False)
self.bn = nn.BatchNorm1d(out_features)
nn.init.constant_(self.bn.weight, bn_weight_init)
@torch.no_gra... | class LinearNorm(nn.Module):
def __init__(self, in_features, out_features, bn_weight_init=1):
super().__init__()
self.linear = nn.Linear(in_features, out_features, bias=False)
self.bn = nn.BatchNorm1d(out_features)
nn.init.constant_(self.bn.weight, bn_weight_init)
@torch.no_gra... | x = self.linear(x)
return self.bn(x.flatten(0, 1)).reshape_as(x)
| x = self.linear(x)
return self.bn(x.flatten(0, 1)).reshape_as(x)
| 6 | 287 | 40 | 327 | """ LeViT
Paper: `LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference`
- https://arxiv.org/abs/2104.01136
@article{graham2021levit,
title={LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference},
author={Benjamin Graham and Alaaeldin El-Nouby and Hugo Touvron and Pierre Stoc... | """ LeViT
Paper: `LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference`
- https://arxiv.org/abs/2104.01136
@article{graham2021levit,
title={LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference},
author={Benjamin Graham and Alaaeldin El-Nouby and Hugo Touvron and Pierre Stoc... |
17,727 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/levit.py | 104,106 | 93,123 | fuse | NormLinear | 106,119 | 106,119 | class NormLinear(nn.Module):
def __init__(self, in_features, out_features, bias=True, std=0.02, drop=0.):
super().__init__()
self.bn = nn.BatchNorm1d(in_features)
self.drop = nn.Dropout(drop)
self.linear = nn.Linear(in_features, out_features, bias=bias)
trunc_normal_(self.li... | class NormLinear(nn.Module):
def __init__(self, in_features, out_features, bias=True, std=0.02, drop=0.):
super().__init__()
self.bn = nn.BatchNorm1d(in_features)
self.drop = nn.Dropout(drop)
self.linear = nn.Linear(in_features, out_features, bias=bias)
trunc_normal_(self.li... | bn, l = self.bn, self.linear
w = bn.weight / (bn.running_var + bn.eps) ** 0.5
b = bn.bias - self.bn.running_mean * self.bn.weight / (bn.running_var + bn.eps) ** 0.5
w = l.weight * w[None, :]
if l.bias is None:
b = b @ self.linear.weight.T
else:
b =... | bn, l = self.bn, self.linear
w = bn.weight / (bn.running_var + bn.eps) ** 0.5
b = bn.bias - self.bn.running_mean * self.bn.weight / (bn.running_var + bn.eps) ** 0.5
w = l.weight * w[None, :]
if l.bias is None:
b = b @ self.linear.weight.T
else:
b =... | 59 | 210 | 212 | 422 | """ LeViT
Paper: `LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference`
- https://arxiv.org/abs/2104.01136
@article{graham2021levit,
title={LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference},
author={Benjamin Graham and Alaaeldin El-Nouby and Hugo Touvron and Pierre Stoc... | """ LeViT
Paper: `LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference`
- https://arxiv.org/abs/2104.01136
@article{graham2021levit,
title={LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference},
author={Benjamin Graham and Alaaeldin El-Nouby and Hugo Touvron and Pierre Stoc... |
17,729 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/metaformer.py | 193,194 | 158,218 | forward | Attention | 194,218 | 194,218 | class Attention(nn.Module):
"""
Vanilla self-attention from Transformer: https://arxiv.org/abs/1706.03762.
Modified from timm.
"""
fused_attn: Final[bool]
def __init__(
self,
dim,
head_dim=32,
num_heads=None,
qkv_bias=False,
... | class Attention(nn.Module):
"""
Vanilla self-attention from Transformer: https://arxiv.org/abs/1706.03762.
Modified from timm.
"""
fused_attn: Final[bool]
def __init__(
self,
dim,
head_dim=32,
num_heads=None,
qkv_bias=False,
... | B, N, C = x.shape
qkv = self.qkv(x).reshape(B, N, 3, self.num_heads, self.head_dim).permute(2, 0, 3, 1, 4)
q, k, v = qkv.unbind(0)
if self.fused_attn:
x = F.scaled_dot_product_attention(
q, k, v,
dropout_p=self.attn_drop.p if self.training els... | B, N, C = x.shape
qkv = self.qkv(x).reshape(B, N, 3, self.num_heads, self.head_dim).permute(2, 0, 3, 1, 4)
q, k, v = qkv.unbind(0)
if self.fused_attn:
x = F.scaled_dot_product_attention(
q, k, v,
dropout_p=self.attn_drop.p if self.training els... | 96 | 385 | 318 | 703 | """
Poolformer from MetaFormer is Actually What You Need for Vision https://arxiv.org/abs/2111.11418
IdentityFormer, RandFormer, PoolFormerV2, ConvFormer, and CAFormer
from MetaFormer Baselines for Vision https://arxiv.org/abs/2210.13452
All implemented models support feature extraction and variable input resolution.... | """
Poolformer from MetaFormer is Actually What You Need for Vision https://arxiv.org/abs/2111.11418
IdentityFormer, RandFormer, PoolFormerV2, ConvFormer, and CAFormer
from MetaFormer Baselines for Vision https://arxiv.org/abs/2210.13452
All implemented models support feature extraction and variable input resolution.... |
17,730 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/metaformer.py | 101,102 | 77,107 | forward | Downsampling | 102,107 | 102,107 | class Downsampling(nn.Module):
"""
Downsampling implemented by a layer of convolution.
"""
def __init__(
self,
in_channels,
out_channels,
kernel_size,
stride=1,
padding=0,
norm_layer=None,
):
super().__init_... | class Downsampling(nn.Module):
"""
Downsampling implemented by a layer of convolution.
"""
def __init__(
self,
in_channels,
out_channels,
kernel_size,
stride=1,
padding=0,
norm_layer=None,
):
super().__init_... | x = self.norm(x)
x = self.conv(x)
return x
| x = self.norm(x)
x = self.conv(x)
return x
| 8 | 191 | 31 | 222 | """
Poolformer from MetaFormer is Actually What You Need for Vision https://arxiv.org/abs/2111.11418
IdentityFormer, RandFormer, PoolFormerV2, ConvFormer, and CAFormer
from MetaFormer Baselines for Vision https://arxiv.org/abs/2210.13452
All implemented models support feature extraction and variable input resolution.... | """
Poolformer from MetaFormer is Actually What You Need for Vision https://arxiv.org/abs/2111.11418
IdentityFormer, RandFormer, PoolFormerV2, ConvFormer, and CAFormer
from MetaFormer Baselines for Vision https://arxiv.org/abs/2210.13452
All implemented models support feature extraction and variable input resolution.... |
17,732 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/metaformer.py | 362,363 | 320,378 | forward | MetaFormerBlock | 363,378 | 363,378 | class MetaFormerBlock(nn.Module):
"""
Implementation of one MetaFormer block.
"""
def __init__(
self,
dim,
token_mixer=Pooling,
mlp_act=StarReLU,
mlp_bias=False,
norm_layer=LayerNorm2d,
proj_drop=0.,
drop_pa... | class MetaFormerBlock(nn.Module):
"""
Implementation of one MetaFormer block.
"""
def __init__(
self,
dim,
token_mixer=Pooling,
mlp_act=StarReLU,
mlp_bias=False,
norm_layer=LayerNorm2d,
proj_drop=0.,
drop_pa... | x = self.res_scale1(x) + \
self.layer_scale1(
self.drop_path1(
self.token_mixer(self.norm1(x))
)
)
x = self.res_scale2(x) + \
self.layer_scale2(
self.drop_path2(
self.mlp(self.norm... | x = self.res_scale1(x) + \
self.layer_scale1(
self.drop_path1(
self.token_mixer(self.norm1(x))
)
)
x = self.res_scale2(x) + \
self.layer_scale2(
self.drop_path2(
self.mlp(self.norm... | 22 | 556 | 129 | 685 | """
Poolformer from MetaFormer is Actually What You Need for Vision https://arxiv.org/abs/2111.11418
IdentityFormer, RandFormer, PoolFormerV2, ConvFormer, and CAFormer
from MetaFormer Baselines for Vision https://arxiv.org/abs/2210.13452
All implemented models support feature extraction and variable input resolution.... | """
Poolformer from MetaFormer is Actually What You Need for Vision https://arxiv.org/abs/2111.11418
IdentityFormer, RandFormer, PoolFormerV2, ConvFormer, and CAFormer
from MetaFormer Baselines for Vision https://arxiv.org/abs/2210.13452
All implemented models support feature extraction and variable input resolution.... |
17,733 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/metaformer.py | 429,430 | 378,447 | forward | MetaFormerStage | 430,447 | 430,447 | class MetaFormerStage(nn.Module):
def __init__(
self,
in_chs,
out_chs,
depth=2,
token_mixer=nn.Identity,
mlp_act=StarReLU,
mlp_bias=False,
downsample_norm=LayerNorm2d,
norm_layer=LayerNorm2d,
pro... | class MetaFormerStage(nn.Module):
def __init__(
self,
in_chs,
out_chs,
depth=2,
token_mixer=nn.Identity,
mlp_act=StarReLU,
mlp_bias=False,
downsample_norm=LayerNorm2d,
norm_layer=LayerNorm2d,
pro... | x = self.downsample(x)
B, C, H, W = x.shape
if not self.use_nchw:
x = x.reshape(B, C, -1).transpose(1, 2)
if self.grad_checkpointing and not torch.jit.is_scripting():
x = checkpoint_seq(self.blocks, x)
else:
x = self.blocks(x)
if not... | x = self.downsample(x)
B, C, H, W = x.shape
if not self.use_nchw:
x = x.reshape(B, C, -1).transpose(1, 2)
if self.grad_checkpointing and not torch.jit.is_scripting():
x = checkpoint_seq(self.blocks, x)
else:
x = self.blocks(x)
if not... | 43 | 527 | 176 | 703 | """
Poolformer from MetaFormer is Actually What You Need for Vision https://arxiv.org/abs/2111.11418
IdentityFormer, RandFormer, PoolFormerV2, ConvFormer, and CAFormer
from MetaFormer Baselines for Vision https://arxiv.org/abs/2210.13452
All implemented models support feature extraction and variable input resolution.... | """
Poolformer from MetaFormer is Actually What You Need for Vision https://arxiv.org/abs/2111.11418
IdentityFormer, RandFormer, PoolFormerV2, ConvFormer, and CAFormer
from MetaFormer Baselines for Vision https://arxiv.org/abs/2210.13452
All implemented models support feature extraction and variable input resolution.... |
17,734 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/metaformer.py | 311,312 | 289,320 | forward | MlpHead | 312,320 | 312,320 | class MlpHead(nn.Module):
""" MLP classification head
"""
def __init__(
self,
dim,
num_classes=1000,
mlp_ratio=4,
act_layer=SquaredReLU,
norm_layer=LayerNorm,
drop_rate=0.,
bias=True
):
super().__ini... | class MlpHead(nn.Module):
""" MLP classification head
"""
def __init__(
self,
dim,
num_classes=1000,
mlp_ratio=4,
act_layer=SquaredReLU,
norm_layer=LayerNorm,
drop_rate=0.,
bias=True
):
super().__ini... | x = self.fc1(x)
x = self.act(x)
x = self.norm(x)
x = self.head_drop(x)
x = self.fc2(x)
return x
| x = self.fc1(x)
x = self.act(x)
x = self.norm(x)
x = self.head_drop(x)
x = self.fc2(x)
return x
| 17 | 227 | 68 | 295 | """
Poolformer from MetaFormer is Actually What You Need for Vision https://arxiv.org/abs/2111.11418
IdentityFormer, RandFormer, PoolFormerV2, ConvFormer, and CAFormer
from MetaFormer Baselines for Vision https://arxiv.org/abs/2210.13452
All implemented models support feature extraction and variable input resolution.... | """
Poolformer from MetaFormer is Actually What You Need for Vision https://arxiv.org/abs/2111.11418
IdentityFormer, RandFormer, PoolFormerV2, ConvFormer, and CAFormer
from MetaFormer Baselines for Vision https://arxiv.org/abs/2210.13452
All implemented models support feature extraction and variable input resolution.... |
17,735 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/metaformer.py | 284,285 | 274,289 | forward | Pooling | 285,289 | 285,289 | class Pooling(nn.Module):
"""
Implementation of pooling for PoolFormer: https://arxiv.org/abs/2111.11418
"""
def __init__(self, pool_size=3, **kwargs):
super().__init__()
self.pool = nn.AvgPool2d(
pool_size, stride=1, padding=pool_size // 2, count_include_pad=False)
def... | class Pooling(nn.Module):
"""
Implementation of pooling for PoolFormer: https://arxiv.org/abs/2111.11418
"""
def __init__(self, pool_size=3, **kwargs):
super().__init__()
self.pool = nn.AvgPool2d(
pool_size, stride=1, padding=pool_size // 2, count_include_pad=False)
def... | y = self.pool(x)
return y - x
| y = self.pool(x)
return y - x
| 7 | 136 | 22 | 158 | """
Poolformer from MetaFormer is Actually What You Need for Vision https://arxiv.org/abs/2111.11418
IdentityFormer, RandFormer, PoolFormerV2, ConvFormer, and CAFormer
from MetaFormer Baselines for Vision https://arxiv.org/abs/2210.13452
All implemented models support feature extraction and variable input resolution.... | """
Poolformer from MetaFormer is Actually What You Need for Vision https://arxiv.org/abs/2111.11418
IdentityFormer, RandFormer, PoolFormerV2, ConvFormer, and CAFormer
from MetaFormer Baselines for Vision https://arxiv.org/abs/2210.13452
All implemented models support feature extraction and variable input resolution.... |
17,736 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/metaformer.py | 117,118 | 107,121 | forward | Scale | 118,121 | 118,121 | class Scale(nn.Module):
"""
Scale vector by element multiplications.
"""
def __init__(self, dim, init_value=1.0, trainable=True, use_nchw=True):
super().__init__()
self.shape = (dim, 1, 1) if use_nchw else (dim,)
self.scale = nn.Parameter(init_value * torch.ones(dim), requires_g... | class Scale(nn.Module):
"""
Scale vector by element multiplications.
"""
def __init__(self, dim, init_value=1.0, trainable=True, use_nchw=True):
super().__init__()
self.shape = (dim, 1, 1) if use_nchw else (dim,)
self.scale = nn.Parameter(init_value * torch.ones(dim), requires_g... | return x * self.scale.view(self.shape)
| return x * self.scale.view(self.shape)
| 4 | 139 | 20 | 159 | """
Poolformer from MetaFormer is Actually What You Need for Vision https://arxiv.org/abs/2111.11418
IdentityFormer, RandFormer, PoolFormerV2, ConvFormer, and CAFormer
from MetaFormer Baselines for Vision https://arxiv.org/abs/2210.13452
All implemented models support feature extraction and variable input resolution.... | """
Poolformer from MetaFormer is Actually What You Need for Vision https://arxiv.org/abs/2111.11418
IdentityFormer, RandFormer, PoolFormerV2, ConvFormer, and CAFormer
from MetaFormer Baselines for Vision https://arxiv.org/abs/2210.13452
All implemented models support feature extraction and variable input resolution.... |
17,737 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/metaformer.py | 265,266 | 239,274 | forward | SepConv | 266,274 | 266,274 | class SepConv(nn.Module):
r"""
Inverted separable convolution from MobileNetV2: https://arxiv.org/abs/1801.04381.
"""
def __init__(
self,
dim,
expansion_ratio=2,
act1_layer=StarReLU,
act2_layer=nn.Identity,
bias=False,
... | class SepConv(nn.Module):
r"""
Inverted separable convolution from MobileNetV2: https://arxiv.org/abs/1801.04381.
"""
def __init__(
self,
dim,
expansion_ratio=2,
act1_layer=StarReLU,
act2_layer=nn.Identity,
bias=False,
... | x = self.pwconv1(x)
x = self.act1(x)
x = self.dwconv(x)
x = self.act2(x)
x = self.pwconv2(x)
return x
| x = self.pwconv1(x)
x = self.act1(x)
x = self.dwconv(x)
x = self.act2(x)
x = self.pwconv2(x)
return x
| 17 | 322 | 74 | 396 | """
Poolformer from MetaFormer is Actually What You Need for Vision https://arxiv.org/abs/2111.11418
IdentityFormer, RandFormer, PoolFormerV2, ConvFormer, and CAFormer
from MetaFormer Baselines for Vision https://arxiv.org/abs/2210.13452
All implemented models support feature extraction and variable input resolution.... | """
Poolformer from MetaFormer is Actually What You Need for Vision https://arxiv.org/abs/2111.11418
IdentityFormer, RandFormer, PoolFormerV2, ConvFormer, and CAFormer
from MetaFormer Baselines for Vision https://arxiv.org/abs/2210.13452
All implemented models support feature extraction and variable input resolution.... |
17,738 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/metaformer.py | 130,131 | 121,134 | forward | SquaredReLU | 131,134 | 131,134 | class SquaredReLU(nn.Module):
"""
Squared ReLU: https://arxiv.org/abs/2109.08668
"""
def __init__(self, inplace=False):
super().__init__()
self.relu = nn.ReLU(inplace=inplace)
def forward(self, x):
FILL_FUNC_BODY | class SquaredReLU(nn.Module):
"""
Squared ReLU: https://arxiv.org/abs/2109.08668
"""
def __init__(self, inplace=False):
super().__init__()
self.relu = nn.ReLU(inplace=inplace)
def forward(self, x):
FILL_FUNC_BODY | return torch.square(self.relu(x))
| return torch.square(self.relu(x))
| 2 | 110 | 20 | 130 | """
Poolformer from MetaFormer is Actually What You Need for Vision https://arxiv.org/abs/2111.11418
IdentityFormer, RandFormer, PoolFormerV2, ConvFormer, and CAFormer
from MetaFormer Baselines for Vision https://arxiv.org/abs/2210.13452
All implemented models support feature extraction and variable input resolution.... | """
Poolformer from MetaFormer is Actually What You Need for Vision https://arxiv.org/abs/2111.11418
IdentityFormer, RandFormer, PoolFormerV2, ConvFormer, and CAFormer
from MetaFormer Baselines for Vision https://arxiv.org/abs/2210.13452
All implemented models support feature extraction and variable input resolution.... |
17,739 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/metaformer.py | 154,155 | 134,158 | forward | StarReLU | 155,158 | 155,158 | class StarReLU(nn.Module):
"""
StarReLU: s * relu(x) ** 2 + b
"""
def __init__(
self,
scale_value=1.0,
bias_value=0.0,
scale_learnable=True,
bias_learnable=True,
mode=None,
inplace=False
):
super().__init__(... | class StarReLU(nn.Module):
"""
StarReLU: s * relu(x) ** 2 + b
"""
def __init__(
self,
scale_value=1.0,
bias_value=0.0,
scale_learnable=True,
bias_learnable=True,
mode=None,
inplace=False
):
super().__init__(... | return self.scale * self.relu(x) ** 2 + self.bias
| return self.scale * self.relu(x) ** 2 + self.bias
| 8 | 223 | 27 | 250 | """
Poolformer from MetaFormer is Actually What You Need for Vision https://arxiv.org/abs/2111.11418
IdentityFormer, RandFormer, PoolFormerV2, ConvFormer, and CAFormer
from MetaFormer Baselines for Vision https://arxiv.org/abs/2210.13452
All implemented models support feature extraction and variable input resolution.... | """
Poolformer from MetaFormer is Actually What You Need for Vision https://arxiv.org/abs/2111.11418
IdentityFormer, RandFormer, PoolFormerV2, ConvFormer, and CAFormer
from MetaFormer Baselines for Vision https://arxiv.org/abs/2210.13452
All implemented models support feature extraction and variable input resolution.... |
17,740 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/metaformer.py | 71,72 | 49,77 | forward | Stem | 72,77 | 72,77 | class Stem(nn.Module):
"""
Stem implemented by a layer of convolution.
Conv2d params constant across all models.
"""
def __init__(
self,
in_channels,
out_channels,
norm_layer=None,
):
super().__init__()
self.conv = nn.Conv2d(
... | class Stem(nn.Module):
"""
Stem implemented by a layer of convolution.
Conv2d params constant across all models.
"""
def __init__(
self,
in_channels,
out_channels,
norm_layer=None,
):
super().__init__()
self.conv = nn.Conv2d(
... | x = self.conv(x)
x = self.norm(x)
return x
| x = self.conv(x)
x = self.norm(x)
return x
| 8 | 179 | 31 | 210 | """
Poolformer from MetaFormer is Actually What You Need for Vision https://arxiv.org/abs/2111.11418
IdentityFormer, RandFormer, PoolFormerV2, ConvFormer, and CAFormer
from MetaFormer Baselines for Vision https://arxiv.org/abs/2210.13452
All implemented models support feature extraction and variable input resolution.... | """
Poolformer from MetaFormer is Actually What You Need for Vision https://arxiv.org/abs/2111.11418
IdentityFormer, RandFormer, PoolFormerV2, ConvFormer, and CAFormer
from MetaFormer Baselines for Vision https://arxiv.org/abs/2210.13452
All implemented models support feature extraction and variable input resolution.... |
17,742 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/mlp_mixer.py | 91,92 | 85,95 | forward | Affine | 92,95 | 92,95 | class Affine(nn.Module):
def __init__(self, dim):
super().__init__()
self.alpha = nn.Parameter(torch.ones((1, 1, dim)))
self.beta = nn.Parameter(torch.zeros((1, 1, dim)))
def forward(self, x):
FILL_FUNC_BODY | class Affine(nn.Module):
def __init__(self, dim):
super().__init__()
self.alpha = nn.Parameter(torch.ones((1, 1, dim)))
self.beta = nn.Parameter(torch.zeros((1, 1, dim)))
def forward(self, x):
FILL_FUNC_BODY | return torch.addcmul(self.beta, self.alpha, x)
| return torch.addcmul(self.beta, self.alpha, x)
| 4 | 94 | 25 | 119 | """ MLP-Mixer, ResMLP, and gMLP in PyTorch
This impl originally based on MLP-Mixer paper.
Official JAX impl: https://github.com/google-research/vision_transformer/blob/linen/vit_jax/models_mixer.py
Paper: 'MLP-Mixer: An all-MLP Architecture for Vision' - https://arxiv.org/abs/2105.01601
@article{tolstikhin2021,
t... | """ MLP-Mixer, ResMLP, and gMLP in PyTorch
This impl originally based on MLP-Mixer paper.
Official JAX impl: https://github.com/google-research/vision_transformer/blob/linen/vit_jax/models_mixer.py
Paper: 'MLP-Mixer: An all-MLP Architecture for Vision' - https://arxiv.org/abs/2105.01601
@article{tolstikhin2021,
t... |
17,743 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/mlp_mixer.py | 79,80 | 56,85 | forward | MixerBlock | 80,85 | 80,85 | class MixerBlock(nn.Module):
""" Residual Block w/ token mixing and channel MLPs
Based on: 'MLP-Mixer: An all-MLP Architecture for Vision' - https://arxiv.org/abs/2105.01601
"""
def __init__(
self,
dim,
seq_len,
mlp_ratio=(0.5, 4.0),
mlp_layer=... | class MixerBlock(nn.Module):
""" Residual Block w/ token mixing and channel MLPs
Based on: 'MLP-Mixer: An all-MLP Architecture for Vision' - https://arxiv.org/abs/2105.01601
"""
def __init__(
self,
dim,
seq_len,
mlp_ratio=(0.5, 4.0),
mlp_layer=... | x = x + self.drop_path(self.mlp_tokens(self.norm1(x).transpose(1, 2)).transpose(1, 2))
x = x + self.drop_path(self.mlp_channels(self.norm2(x)))
return x
| x = x + self.drop_path(self.mlp_tokens(self.norm1(x).transpose(1, 2)).transpose(1, 2))
x = x + self.drop_path(self.mlp_channels(self.norm2(x)))
return x
| 14 | 351 | 81 | 432 | """ MLP-Mixer, ResMLP, and gMLP in PyTorch
This impl originally based on MLP-Mixer paper.
Official JAX impl: https://github.com/google-research/vision_transformer/blob/linen/vit_jax/models_mixer.py
Paper: 'MLP-Mixer: An all-MLP Architecture for Vision' - https://arxiv.org/abs/2105.01601
@article{tolstikhin2021,
t... | """ MLP-Mixer, ResMLP, and gMLP in PyTorch
This impl originally based on MLP-Mixer paper.
Official JAX impl: https://github.com/google-research/vision_transformer/blob/linen/vit_jax/models_mixer.py
Paper: 'MLP-Mixer: An all-MLP Architecture for Vision' - https://arxiv.org/abs/2105.01601
@article{tolstikhin2021,
t... |
17,744 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/mlp_mixer.py | 253,254 | 179,281 | reset_classifier | MlpMixer | 254,260 | 254,260 | class MlpMixer(nn.Module):
def __init__(
self,
num_classes=1000,
img_size=224,
in_chans=3,
patch_size=16,
num_blocks=8,
embed_dim=512,
mlp_ratio=(0.5, 4.0),
block_layer=MixerBlock,
mlp_layer=Mlp,... | class MlpMixer(nn.Module):
def __init__(
self,
num_classes=1000,
img_size=224,
in_chans=3,
patch_size=16,
num_blocks=8,
embed_dim=512,
mlp_ratio=(0.5, 4.0),
block_layer=MixerBlock,
mlp_layer=Mlp,... | self.num_classes = num_classes
if global_pool is not None:
assert global_pool in ('', 'avg')
self.global_pool = global_pool
self.head = nn.Linear(self.embed_dim, num_classes) if num_classes > 0 else nn.Identity()
| self.num_classes = num_classes
if global_pool is not None:
assert global_pool in ('', 'avg')
self.global_pool = global_pool
self.head = nn.Linear(self.embed_dim, num_classes) if num_classes > 0 else nn.Identity()
| 26 | 1,036 | 85 | 1,121 | """ MLP-Mixer, ResMLP, and gMLP in PyTorch
This impl originally based on MLP-Mixer paper.
Official JAX impl: https://github.com/google-research/vision_transformer/blob/linen/vit_jax/models_mixer.py
Paper: 'MLP-Mixer: An all-MLP Architecture for Vision' - https://arxiv.org/abs/2105.01601
@article{tolstikhin2021,
t... | """ MLP-Mixer, ResMLP, and gMLP in PyTorch
This impl originally based on MLP-Mixer paper.
Official JAX impl: https://github.com/google-research/vision_transformer/blob/linen/vit_jax/models_mixer.py
Paper: 'MLP-Mixer: An all-MLP Architecture for Vision' - https://arxiv.org/abs/2105.01601
@article{tolstikhin2021,
t... |
17,745 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/mlp_mixer.py | 122,123 | 95,128 | forward | ResBlock | 123,128 | 123,128 | class ResBlock(nn.Module):
""" Residual MLP block w/ LayerScale and Affine 'norm'
Based on: `ResMLP: Feedforward networks for image classification...` - https://arxiv.org/abs/2105.03404
"""
def __init__(
self,
dim,
seq_len,
mlp_ratio=4,
mlp_la... | class ResBlock(nn.Module):
""" Residual MLP block w/ LayerScale and Affine 'norm'
Based on: `ResMLP: Feedforward networks for image classification...` - https://arxiv.org/abs/2105.03404
"""
def __init__(
self,
dim,
seq_len,
mlp_ratio=4,
mlp_la... | x = x + self.drop_path(self.ls1 * self.linear_tokens(self.norm1(x).transpose(1, 2)).transpose(1, 2))
x = x + self.drop_path(self.ls2 * self.mlp_channels(self.norm2(x)))
return x
| x = x + self.drop_path(self.ls1 * self.linear_tokens(self.norm1(x).transpose(1, 2)).transpose(1, 2))
x = x + self.drop_path(self.ls2 * self.mlp_channels(self.norm2(x)))
return x
| 18 | 361 | 90 | 451 | """ MLP-Mixer, ResMLP, and gMLP in PyTorch
This impl originally based on MLP-Mixer paper.
Official JAX impl: https://github.com/google-research/vision_transformer/blob/linen/vit_jax/models_mixer.py
Paper: 'MLP-Mixer: An all-MLP Architecture for Vision' - https://arxiv.org/abs/2105.01601
@article{tolstikhin2021,
t... | """ MLP-Mixer, ResMLP, and gMLP in PyTorch
This impl originally based on MLP-Mixer paper.
Official JAX impl: https://github.com/google-research/vision_transformer/blob/linen/vit_jax/models_mixer.py
Paper: 'MLP-Mixer: An all-MLP Architecture for Vision' - https://arxiv.org/abs/2105.01601
@article{tolstikhin2021,
t... |
17,746 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/mlp_mixer.py | 174,175 | 151,179 | forward | SpatialGatingBlock | 175,179 | 175,179 | class SpatialGatingBlock(nn.Module):
""" Residual Block w/ Spatial Gating
Based on: `Pay Attention to MLPs` - https://arxiv.org/abs/2105.08050
"""
def __init__(
self,
dim,
seq_len,
mlp_ratio=4,
mlp_layer=GatedMlp,
norm_layer=partia... | class SpatialGatingBlock(nn.Module):
""" Residual Block w/ Spatial Gating
Based on: `Pay Attention to MLPs` - https://arxiv.org/abs/2105.08050
"""
def __init__(
self,
dim,
seq_len,
mlp_ratio=4,
mlp_layer=GatedMlp,
norm_layer=partia... | x = x + self.drop_path(self.mlp_channels(self.norm(x)))
return x
| x = x + self.drop_path(self.mlp_channels(self.norm(x)))
return x
| 7 | 302 | 36 | 338 | """ MLP-Mixer, ResMLP, and gMLP in PyTorch
This impl originally based on MLP-Mixer paper.
Official JAX impl: https://github.com/google-research/vision_transformer/blob/linen/vit_jax/models_mixer.py
Paper: 'MLP-Mixer: An all-MLP Architecture for Vision' - https://arxiv.org/abs/2105.01601
@article{tolstikhin2021,
t... | """ MLP-Mixer, ResMLP, and gMLP in PyTorch
This impl originally based on MLP-Mixer paper.
Official JAX impl: https://github.com/google-research/vision_transformer/blob/linen/vit_jax/models_mixer.py
Paper: 'MLP-Mixer: An all-MLP Architecture for Vision' - https://arxiv.org/abs/2105.01601
@article{tolstikhin2021,
t... |
17,747 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/mlp_mixer.py | 144,145 | 128,151 | forward | SpatialGatingUnit | 145,151 | 145,151 | class SpatialGatingUnit(nn.Module):
""" Spatial Gating Unit
Based on: `Pay Attention to MLPs` - https://arxiv.org/abs/2105.08050
"""
def __init__(self, dim, seq_len, norm_layer=nn.LayerNorm):
super().__init__()
gate_dim = dim // 2
self.norm = norm_layer(gate_dim)
self.pr... | class SpatialGatingUnit(nn.Module):
""" Spatial Gating Unit
Based on: `Pay Attention to MLPs` - https://arxiv.org/abs/2105.08050
"""
def __init__(self, dim, seq_len, norm_layer=nn.LayerNorm):
super().__init__()
gate_dim = dim // 2
self.norm = norm_layer(gate_dim)
self.pr... | u, v = x.chunk(2, dim=-1)
v = self.norm(v)
v = self.proj(v.transpose(-1, -2))
return u * v.transpose(-1, -2)
| u, v = x.chunk(2, dim=-1)
v = self.norm(v)
v = self.proj(v.transpose(-1, -2))
return u * v.transpose(-1, -2)
| 17 | 221 | 69 | 290 | """ MLP-Mixer, ResMLP, and gMLP in PyTorch
This impl originally based on MLP-Mixer paper.
Official JAX impl: https://github.com/google-research/vision_transformer/blob/linen/vit_jax/models_mixer.py
Paper: 'MLP-Mixer: An all-MLP Architecture for Vision' - https://arxiv.org/abs/2105.01601
@article{tolstikhin2021,
t... | """ MLP-Mixer, ResMLP, and gMLP in PyTorch
This impl originally based on MLP-Mixer paper.
Official JAX impl: https://github.com/google-research/vision_transformer/blob/linen/vit_jax/models_mixer.py
Paper: 'MLP-Mixer: An all-MLP Architecture for Vision' - https://arxiv.org/abs/2105.01601
@article{tolstikhin2021,
t... |
17,749 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/mobilevit.py | 317,318 | 272,391 | _forward_self_attn | LinearSelfAttention | 318,341 | 318,341 | class LinearSelfAttention(nn.Module):
"""
This layer applies a self-attention with linear complexity, as described in `https://arxiv.org/abs/2206.02680`
This layer can be used for self- as well as cross-attention.
Args:
embed_dim (int): :math:`C` from an expected input of size :math:`(N, C, H, W... | class LinearSelfAttention(nn.Module):
"""
This layer applies a self-attention with linear complexity, as described in `https://arxiv.org/abs/2206.02680`
This layer can be used for self- as well as cross-attention.
Args:
embed_dim (int): :math:`C` from an expected input of size :math:`(N, C, H, W... | # [B, C, P, N] --> [B, h + 2d, P, N]
qkv = self.qkv_proj(x)
# Project x into query, key and value
# Query --> [B, 1, P, N]
# value, key --> [B, d, P, N]
query, key, value = qkv.split([1, self.embed_dim, self.embed_dim], dim=1)
# apply softmax along N dimension
... | # [B, C, P, N] --> [B, h + 2d, P, N]
qkv = self.qkv_proj(x)
# Project x into query, key and value
# Query --> [B, 1, P, N]
# value, key --> [B, d, P, N]
query, key, value = qkv.split([1, self.embed_dim, self.embed_dim], dim=1)
# apply softmax along N dimension
... | 123 | 1,400 | 364 | 1,764 | """ MobileViT
Paper:
V1: `MobileViT: Light-weight, General-purpose, and Mobile-friendly Vision Transformer` - https://arxiv.org/abs/2110.02178
V2: `Separable Self-attention for Mobile Vision Transformers` - https://arxiv.org/abs/2206.02680
MobileVitBlock and checkpoints adapted from https://github.com/apple/ml-cvnets... | """ MobileViT
Paper:
V1: `MobileViT: Light-weight, General-purpose, and Mobile-friendly Vision Transformer` - https://arxiv.org/abs/2110.02178
V2: `Separable Self-attention for Mobile Vision Transformers` - https://arxiv.org/abs/2206.02680
MobileVitBlock and checkpoints adapted from https://github.com/apple/ml-cvnets... |
17,750 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/mobilevit.py | 433,434 | 391,449 | forward | LinearTransformerBlock | 434,449 | 434,449 | class LinearTransformerBlock(nn.Module):
"""
This class defines the pre-norm transformer encoder with linear self-attention in `MobileViTv2 paper <>`_
Args:
embed_dim (int): :math:`C_{in}` from an expected input of size :math:`(B, C_{in}, P, N)`
mlp_ratio (float): Inner dimension ratio of th... | class LinearTransformerBlock(nn.Module):
"""
This class defines the pre-norm transformer encoder with linear self-attention in `MobileViTv2 paper <>`_
Args:
embed_dim (int): :math:`C_{in}` from an expected input of size :math:`(B, C_{in}, P, N)`
mlp_ratio (float): Inner dimension ratio of th... | if x_prev is None:
# self-attention
x = x + self.drop_path1(self.attn(self.norm1(x)))
else:
# cross-attention
res = x
x = self.norm1(x) # norm
x = self.attn(x, x_prev) # attn
x = self.drop_path1(x) + res # residual
... | if x_prev is None:
# self-attention
x = x + self.drop_path1(self.attn(self.norm1(x)))
else:
# cross-attention
res = x
x = self.norm1(x) # norm
x = self.attn(x, x_prev) # attn
x = self.drop_path1(x) + res # residual
... | 46 | 616 | 159 | 775 | """ MobileViT
Paper:
V1: `MobileViT: Light-weight, General-purpose, and Mobile-friendly Vision Transformer` - https://arxiv.org/abs/2110.02178
V2: `Separable Self-attention for Mobile Vision Transformers` - https://arxiv.org/abs/2206.02680
MobileVitBlock and checkpoints adapted from https://github.com/apple/ml-cvnets... | """ MobileViT
Paper:
V1: `MobileViT: Light-weight, General-purpose, and Mobile-friendly Vision Transformer` - https://arxiv.org/abs/2110.02178
V2: `Separable Self-attention for Mobile Vision Transformers` - https://arxiv.org/abs/2206.02680
MobileVitBlock and checkpoints adapted from https://github.com/apple/ml-cvnets... |
17,751 | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models | /home/hieuvd/dung_repo/raw/stars_01/huggingface_pytorch-image-models/pytorch-image-models/timm/models/mobilevit.py | 229,230 | 164,272 | forward | MobileVitBlock | 230,272 | 230,272 | @register_notrace_module
class MobileVitBlock(nn.Module):
""" MobileViT block
Paper: https://arxiv.org/abs/2110.02178?context=cs.LG
"""
def __init__(
self,
in_chs: int,
out_chs: Optional[int] = None,
kernel_size: int = 3,
stride: int = 1,
... | @register_notrace_module
class MobileVitBlock(nn.Module):
""" MobileViT block
Paper: https://arxiv.org/abs/2110.02178?context=cs.LG
"""
def __init__(
self,
in_chs: int,
out_chs: Optional[int] = None,
kernel_size: int = 3,
stride: int = 1,
... | shortcut = x
# Local representation
x = self.conv_kxk(x)
x = self.conv_1x1(x)
# Unfold (feature map -> patches)
patch_h, patch_w = self.patch_size
B, C, H, W = x.shape
new_h, new_w = math.ceil(H / patch_h) * patch_h, math.ceil(W / patch_w) * patch_w
... | shortcut = x
# Local representation
x = self.conv_kxk(x)
x = self.conv_1x1(x)
# Unfold (feature map -> patches)
patch_h, patch_w = self.patch_size
B, C, H, W = x.shape
new_h, new_w = math.ceil(H / patch_h) * patch_h, math.ceil(W / patch_w) * patch_w
... | 250 | 818 | 830 | 1,648 | """ MobileViT
Paper:
V1: `MobileViT: Light-weight, General-purpose, and Mobile-friendly Vision Transformer` - https://arxiv.org/abs/2110.02178
V2: `Separable Self-attention for Mobile Vision Transformers` - https://arxiv.org/abs/2206.02680
MobileVitBlock and checkpoints adapted from https://github.com/apple/ml-cvnets... | """ MobileViT
Paper:
V1: `MobileViT: Light-weight, General-purpose, and Mobile-friendly Vision Transformer` - https://arxiv.org/abs/2110.02178
V2: `Separable Self-attention for Mobile Vision Transformers` - https://arxiv.org/abs/2206.02680
MobileVitBlock and checkpoints adapted from https://github.com/apple/ml-cvnets... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.