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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
17,507 | /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/_features.py | 342,343 | 313,346 | forward | FeatureListNet | 343,346 | 343,346 | class FeatureListNet(FeatureDictNet):
""" Feature extractor with list return
A specialization of FeatureDictNet that always returns features as a list (values() of dict).
"""
def __init__(
self,
model: nn.Module,
out_indices: OutIndicesT = (0, 1, 2, 3, 4),
... | class FeatureListNet(FeatureDictNet):
""" Feature extractor with list return
A specialization of FeatureDictNet that always returns features as a list (values() of dict).
"""
def __init__(
self,
model: nn.Module,
out_indices: OutIndicesT = (0, 1, 2, 3, 4),
... | return list(self._collect(x).values())
| return list(self._collect(x).values())
| 2 | 330 | 18 | 348 | """ PyTorch Feature Extraction Helpers
A collection of classes, functions, modules to help extract features from models
and provide a common interface for describing them.
The return_layers, module re-writing idea inspired by torchvision IntermediateLayerGetter
https://github.com/pytorch/vision/blob/d88d8961ae51507d0... | """ PyTorch Feature Extraction Helpers
A collection of classes, functions, modules to help extract features from models
and provide a common interface for describing them.
The return_layers, module re-writing idea inspired by torchvision IntermediateLayerGetter
https://github.com/pytorch/vision/blob/d88d8961ae51507d0... |
17,508 | /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/_features_fx.py | 95,96 | 1,154 | create_feature_extractor | 96,103 | 96,103 | """ PyTorch FX Based Feature Extraction Helpers
Using https://pytorch.org/vision/stable/feature_extraction.html
"""
from typing import Callable, Dict, List, Optional, Union, Tuple, Type
import torch
from torch import nn
from ._features import _get_feature_info, _get_return_layers
try:
from torchvision.models.fea... | """ PyTorch FX Based Feature Extraction Helpers
Using https://pytorch.org/vision/stable/feature_extraction.html
"""
from typing import Callable, Dict, List, Optional, Union, Tuple, Type
import torch
from torch import nn
from ._features import _get_feature_info, _get_return_layers
try:
from torchvision.models.fea... | assert has_fx_feature_extraction, 'Please update to PyTorch 1.10+, torchvision 0.11+ for FX feature extraction'
return _create_feature_extractor(
model, return_nodes,
tracer_kwargs={'leaf_modules': list(_leaf_modules), 'autowrap_functions': list(_autowrap_functions)}
)
| assert has_fx_feature_extraction, 'Please update to PyTorch 1.10+, torchvision 0.11+ for FX feature extraction'
return _create_feature_extractor(
model, return_nodes,
tracer_kwargs={'leaf_modules': list(_leaf_modules), 'autowrap_functions': list(_autowrap_functions)}
)
| 22 | 1,575 | 105 | 1,680 | """ PyTorch FX Based Feature Extraction Helpers
Using https://pytorch.org/vision/stable/feature_extraction.html
"""
from typing import Callable, Dict, List, Optional, Union, Tuple, Type
import torch
from torch import nn
from ._features import _get_feature_info, _get_return_layers
try:
from torchvision.models.fea... | """ PyTorch FX Based Feature Extraction Helpers
Using https://pytorch.org/vision/stable/feature_extraction.html
"""
from typing import Callable, Dict, List, Optional, Union, Tuple, Type
import torch
from torch import nn
from ._features import _get_feature_info, _get_return_layers
try:
from torchvision.models.fea... | |
17,509 | /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/_features_fx.py | 122,123 | 103,126 | forward | FeatureGraphNet | 123,126 | 123,126 | class FeatureGraphNet(nn.Module):
""" A FX Graph based feature extractor that works with the model feature_info metadata
"""
def __init__(
self,
model: nn.Module,
out_indices: Tuple[int, ...],
out_map: Optional[Dict] = None,
output_fmt: str = 'NCHW... | class FeatureGraphNet(nn.Module):
""" A FX Graph based feature extractor that works with the model feature_info metadata
"""
def __init__(
self,
model: nn.Module,
out_indices: Tuple[int, ...],
out_map: Optional[Dict] = None,
output_fmt: str = 'NCHW... | return list(self.graph_module(x).values())
| return list(self.graph_module(x).values())
| 2 | 276 | 20 | 296 | """ PyTorch FX Based Feature Extraction Helpers
Using https://pytorch.org/vision/stable/feature_extraction.html
"""
from typing import Callable, Dict, List, Optional, Union, Tuple, Type
import torch
from torch import nn
from ._features import _get_feature_info, _get_return_layers
try:
from torchvision.models.fea... | """ PyTorch FX Based Feature Extraction Helpers
Using https://pytorch.org/vision/stable/feature_extraction.html
"""
from typing import Callable, Dict, List, Optional, Union, Tuple, Type
import torch
from torch import nn
from ._features import _get_feature_info, _get_return_layers
try:
from torchvision.models.fea... |
17,510 | /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/_features_fx.py | 149,150 | 126,154 | forward | GraphExtractNet | 150,154 | 150,154 | class GraphExtractNet(nn.Module):
""" A standalone feature extraction wrapper that maps dict -> list or single tensor
NOTE:
* one can use feature_extractor directly if dictionary output is desired
* unlike FeatureGraphNet, this is intended to be used standalone and not with model feature_info
... | class GraphExtractNet(nn.Module):
""" A standalone feature extraction wrapper that maps dict -> list or single tensor
NOTE:
* one can use feature_extractor directly if dictionary output is desired
* unlike FeatureGraphNet, this is intended to be used standalone and not with model feature_info
... | out = list(self.graph_module(x).values())
if self.squeeze_out and len(out) == 1:
return out[0]
return out
| out = list(self.graph_module(x).values())
if self.squeeze_out and len(out) == 1:
return out[0]
return out
| 13 | 293 | 49 | 342 | """ PyTorch FX Based Feature Extraction Helpers
Using https://pytorch.org/vision/stable/feature_extraction.html
"""
from typing import Callable, Dict, List, Optional, Union, Tuple, Type
import torch
from torch import nn
from ._features import _get_feature_info, _get_return_layers
try:
from torchvision.models.fea... | """ PyTorch FX Based Feature Extraction Helpers
Using https://pytorch.org/vision/stable/feature_extraction.html
"""
from typing import Callable, Dict, List, Optional, Union, Tuple, Type
import torch
from torch import nn
from ._features import _get_feature_info, _get_return_layers
try:
from torchvision.models.fea... |
17,511 | /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/_helpers.py | 88,93 | 1,151 | remap_state_dict | 96,108 | 93,108 | """ Model creation / weight loading / state_dict helpers
Hacked together by / Copyright 2020 Ross Wightman
"""
import logging
import os
from collections import OrderedDict
from typing import Any, Callable, Dict, Optional, Union
import torch
try:
import safetensors.torch
_has_safetensors = True
except ImportEr... | """ Model creation / weight loading / state_dict helpers
Hacked together by / Copyright 2020 Ross Wightman
"""
import logging
import os
from collections import OrderedDict
from typing import Any, Callable, Dict, Optional, Union
import torch
try:
import safetensors.torch
_has_safetensors = True
except ImportEr... | out_dict = {}
for (ka, va), (kb, vb) in zip(model.state_dict().items(), state_dict.items()):
assert va.numel() == vb.numel(), f'Tensor size mismatch {ka}: {va.shape} vs {kb}: {vb.shape}. Remap failed.'
if va.shape != vb.shape:
if allow_reshape:
vb = vb.reshape(va.shap... | """ remap checkpoint by iterating over state dicts in order (ignoring original keys).
This assumes models (and originating state dict) were created with params registered in same order.
"""
out_dict = {}
for (ka, va), (kb, vb) in zip(model.state_dict().items(), state_dict.items()):
assert va... | 52 | 1,636 | 189 | 1,825 | """ Model creation / weight loading / state_dict helpers
Hacked together by / Copyright 2020 Ross Wightman
"""
import logging
import os
from collections import OrderedDict
from typing import Any, Callable, Dict, Optional, Union
import torch
try:
import safetensors.torch
_has_safetensors = True
except ImportEr... | """ Model creation / weight loading / state_dict helpers
Hacked together by / Copyright 2020 Ross Wightman
"""
import logging
import os
from collections import OrderedDict
from typing import Any, Callable, Dict, Optional, Union
import torch
try:
import safetensors.torch
_has_safetensors = True
except ImportEr... | |
17,513 | /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/_pretrained.py | 69,70 | 1,95 | filter_pretrained_cfg | 70,81 | 70,81 | import copy
from collections import deque, defaultdict
from dataclasses import dataclass, field, replace, asdict
from typing import Any, Deque, Dict, Tuple, Optional, Union
__all__ = ['PretrainedCfg', 'filter_pretrained_cfg', 'DefaultCfg']
@dataclass
class PretrainedCfg:
"""
"""
# weight source location... | import copy
from collections import deque, defaultdict
from dataclasses import dataclass, field, replace, asdict
from typing import Any, Deque, Dict, Tuple, Optional, Union
__all__ = ['PretrainedCfg', 'filter_pretrained_cfg', 'DefaultCfg']
@dataclass
class PretrainedCfg:
"""
"""
# weight source location... | filtered_cfg = {}
keep_null = {'pool_size', 'first_conv', 'classifier'} # always keep these keys, even if none
for k, v in cfg.items():
if remove_source and k in {'url', 'file', 'hf_hub_id', 'hf_hub_id', 'hf_hub_filename', 'source'}:
continue
if remove_null and v is None and k n... | filtered_cfg = {}
keep_null = {'pool_size', 'first_conv', 'classifier'} # always keep these keys, even if none
for k, v in cfg.items():
if remove_source and k in {'url', 'file', 'hf_hub_id', 'hf_hub_id', 'hf_hub_filename', 'source'}:
continue
if remove_null and v is None and k n... | 50 | 1,055 | 148 | 1,203 | import copy
from collections import deque, defaultdict
from dataclasses import dataclass, field, replace, asdict
from typing import Any, Deque, Dict, Tuple, Optional, Union
__all__ = ['PretrainedCfg', 'filter_pretrained_cfg', 'DefaultCfg']
@dataclass
class PretrainedCfg:
"""
"""
# weight source location... | import copy
from collections import deque, defaultdict
from dataclasses import dataclass, field, replace, asdict
from typing import Any, Deque, Dict, Tuple, Optional, Union
__all__ = ['PretrainedCfg', 'filter_pretrained_cfg', 'DefaultCfg']
@dataclass
class PretrainedCfg:
"""
"""
# weight source location... | |
17,514 | /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/_pretrained.py | 91,93 | 81,95 | default_with_tag | DefaultCfg | 93,95 | 93,95 | @dataclass
class DefaultCfg:
tags: Deque[str] = field(default_factory=deque) # priority queue of tags (first is default)
cfgs: Dict[str, PretrainedCfg] = field(default_factory=dict) # pretrained cfgs by tag
is_pretrained: bool = False # at least one of the configs has a pretrained source set
@proper... | @dataclass
class DefaultCfg:
tags: Deque[str] = field(default_factory=deque) # priority queue of tags (first is default)
cfgs: Dict[str, PretrainedCfg] = field(default_factory=dict) # pretrained cfgs by tag
is_pretrained: bool = False # at least one of the configs has a pretrained source set
@proper... | tag = self.tags[0]
return tag, self.cfgs[tag]
| tag = self.tags[0]
return tag, self.cfgs[tag]
| 6 | 152 | 24 | 176 | import copy
from collections import deque, defaultdict
from dataclasses import dataclass, field, replace, asdict
from typing import Any, Deque, Dict, Tuple, Optional, Union
__all__ = ['PretrainedCfg', 'filter_pretrained_cfg', 'DefaultCfg']
@dataclass
class PretrainedCfg:
"""
"""
# weight source location... | import copy
from collections import deque, defaultdict
from dataclasses import dataclass, field, replace, asdict
from typing import Any, Deque, Dict, Tuple, Optional, Union
__all__ = ['PretrainedCfg', 'filter_pretrained_cfg', 'DefaultCfg']
@dataclass
class PretrainedCfg:
"""
"""
# weight source location... |
17,515 | /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/_pretrained.py | 57,59 | 10,69 | has_weights | PretrainedCfg | 59,61 | 59,61 | @dataclass
class PretrainedCfg:
"""
"""
# weight source locations
url: Optional[Union[str, Tuple[str, str]]] = None # remote URL
file: Optional[str] = None # local / shared filesystem path
state_dict: Optional[Dict[str, Any]] = None # in-memory state dict
hf_hub_id: Optional[str] = None ... | @dataclass
class PretrainedCfg:
"""
"""
# weight source locations
url: Optional[Union[str, Tuple[str, str]]] = None # remote URL
file: Optional[str] = None # local / shared filesystem path
state_dict: Optional[Dict[str, Any]] = None # in-memory state dict
hf_hub_id: Optional[str] = None ... | return self.url or self.file or self.hf_hub_id
| return self.url or self.file or self.hf_hub_id
| 6 | 770 | 23 | 793 | import copy
from collections import deque, defaultdict
from dataclasses import dataclass, field, replace, asdict
from typing import Any, Deque, Dict, Tuple, Optional, Union
__all__ = ['PretrainedCfg', 'filter_pretrained_cfg', 'DefaultCfg']
@dataclass
class PretrainedCfg:
"""
"""
# weight source location... | import copy
from collections import deque, defaultdict
from dataclasses import dataclass, field, replace, asdict
from typing import Any, Deque, Dict, Tuple, Optional, Union
__all__ = ['PretrainedCfg', 'filter_pretrained_cfg', 'DefaultCfg']
@dataclass
class PretrainedCfg:
"""
"""
# weight source location... |
17,516 | /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/_prune.py | 12,13 | 1,114 | extract_layer | 13,30 | 13,30 | import os
import pkgutil
from copy import deepcopy
from torch import nn as nn
from timm.layers import Conv2dSame, BatchNormAct2d, Linear
__all__ = ['extract_layer', 'set_layer', 'adapt_model_from_string', 'adapt_model_from_file']
def extract_layer(model, layer):
FILL_FUNC_BODYdef set_layer(model, layer, val):
... | import os
import pkgutil
from copy import deepcopy
from torch import nn as nn
from timm.layers import Conv2dSame, BatchNormAct2d, Linear
__all__ = ['extract_layer', 'set_layer', 'adapt_model_from_string', 'adapt_model_from_file']
def extract_layer(model, layer):
FILL_FUNC_BODYdef set_layer(model, layer, val):
... | layer = layer.split('.')
module = model
if hasattr(model, 'module') and layer[0] != 'module':
module = model.module
if not hasattr(model, 'module') and layer[0] == 'module':
layer = layer[1:]
for l in layer:
if hasattr(module, l):
if not l.isdigit():
... | layer = layer.split('.')
module = model
if hasattr(model, 'module') and layer[0] != 'module':
module = model.module
if not hasattr(model, 'module') and layer[0] == 'module':
layer = layer[1:]
for l in layer:
if hasattr(module, l):
if not l.isdigit():
... | 50 | 1,214 | 154 | 1,368 | import os
import pkgutil
from copy import deepcopy
from torch import nn as nn
from timm.layers import Conv2dSame, BatchNormAct2d, Linear
__all__ = ['extract_layer', 'set_layer', 'adapt_model_from_string', 'adapt_model_from_file']
def extract_layer(model, layer):
FILL_FUNC_BODYdef set_layer(model, layer, val):
... | import os
import pkgutil
from copy import deepcopy
from torch import nn as nn
from timm.layers import Conv2dSame, BatchNormAct2d, Linear
__all__ = ['extract_layer', 'set_layer', 'adapt_model_from_string', 'adapt_model_from_file']
def extract_layer(model, layer):
FILL_FUNC_BODYdef set_layer(model, layer, val):
... | |
17,520 | /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/byobnet.py | 253,254 | 213,271 | init_weights | BasicBlock | 254,260 | 254,260 | class BasicBlock(nn.Module):
""" ResNet Basic Block - kxk + kxk
"""
def __init__(
self,
in_chs: int,
out_chs: int,
kernel_size: int = 3,
stride: int = 1,
dilation: Tuple[int, int] = (1, 1),
group_size: Optional[int] = None,... | class BasicBlock(nn.Module):
""" ResNet Basic Block - kxk + kxk
"""
def __init__(
self,
in_chs: int,
out_chs: int,
kernel_size: int = 3,
stride: int = 1,
dilation: Tuple[int, int] = (1, 1),
group_size: Optional[int] = None,... | if zero_init_last and self.shortcut is not None and getattr(self.conv2_kxk.bn, 'weight', None) is not None:
nn.init.zeros_(self.conv2_kxk.bn.weight)
for attn in (self.attn, self.attn_last):
if hasattr(attn, 'reset_parameters'):
attn.reset_parameters()
| if zero_init_last and self.shortcut is not None and getattr(self.conv2_kxk.bn, 'weight', None) is not None:
nn.init.zeros_(self.conv2_kxk.bn.weight)
for attn in (self.attn, self.attn_last):
if hasattr(attn, 'reset_parameters'):
attn.reset_parameters()
| 24 | 682 | 113 | 795 | """ Bring-Your-Own-Blocks Network
A flexible network w/ dataclass based config for stacking those NN blocks.
This model is currently used to implement the following networks:
GPU Efficient (ResNets) - gernet_l/m/s (original versions called genet, but this was already used (by SENet author)).
Paper: `Neural Architect... | """ Bring-Your-Own-Blocks Network
A flexible network w/ dataclass based config for stacking those NN blocks.
This model is currently used to implement the following networks:
GPU Efficient (ResNets) - gernet_l/m/s (original versions called genet, but this was already used (by SENet author)).
Paper: `Neural Architect... |
17,521 | /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/byobnet.py | 319,320 | 271,340 | init_weights | BottleneckBlock | 320,326 | 320,326 | class BottleneckBlock(nn.Module):
""" ResNet-like Bottleneck Block - 1x1 - kxk - 1x1
"""
def __init__(
self,
in_chs: int,
out_chs: int,
kernel_size: int = 3,
stride: int = 1,
dilation: Tuple[int, int] = (1, 1),
bottle_ratio... | class BottleneckBlock(nn.Module):
""" ResNet-like Bottleneck Block - 1x1 - kxk - 1x1
"""
def __init__(
self,
in_chs: int,
out_chs: int,
kernel_size: int = 3,
stride: int = 1,
dilation: Tuple[int, int] = (1, 1),
bottle_ratio... | if zero_init_last and self.shortcut is not None and getattr(self.conv3_1x1.bn, 'weight', None) is not None:
nn.init.zeros_(self.conv3_1x1.bn.weight)
for attn in (self.attn, self.attn_last):
if hasattr(attn, 'reset_parameters'):
attn.reset_parameters()
| if zero_init_last and self.shortcut is not None and getattr(self.conv3_1x1.bn, 'weight', None) is not None:
nn.init.zeros_(self.conv3_1x1.bn.weight)
for attn in (self.attn, self.attn_last):
if hasattr(attn, 'reset_parameters'):
attn.reset_parameters()
| 24 | 873 | 113 | 986 | """ Bring-Your-Own-Blocks Network
A flexible network w/ dataclass based config for stacking those NN blocks.
This model is currently used to implement the following networks:
GPU Efficient (ResNets) - gernet_l/m/s (original versions called genet, but this was already used (by SENet author)).
Paper: `Neural Architect... | """ Bring-Your-Own-Blocks Network
A flexible network w/ dataclass based config for stacking those NN blocks.
This model is currently used to implement the following networks:
GPU Efficient (ResNets) - gernet_l/m/s (original versions called genet, but this was already used (by SENet author)).
Paper: `Neural Architect... |
17,522 | /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/byobnet.py | 1222,1224 | 1141,1262 | group_matcher | ByobNet | 1224,1233 | 1224,1233 | class ByobNet(nn.Module):
""" 'Bring-your-own-blocks' Net
A flexible network backbone that allows building model stem + blocks via
dataclass cfg definition w/ factory functions for module instantiation.
Current assumption is that both stem and blocks are in conv-bn-act order (w/ block ending in act).
... | class ByobNet(nn.Module):
""" 'Bring-your-own-blocks' Net
A flexible network backbone that allows building model stem + blocks via
dataclass cfg definition w/ factory functions for module instantiation.
Current assumption is that both stem and blocks are in conv-bn-act order (w/ block ending in act).
... | matcher = dict(
stem=r'^stem',
blocks=[
(r'^stages\.(\d+)' if coarse else r'^stages\.(\d+)\.(\d+)', None),
(r'^final_conv', (99999,))
]
)
return matcher
| matcher = dict(
stem=r'^stem',
blocks=[
(r'^stages\.(\d+)' if coarse else r'^stages\.(\d+)\.(\d+)', None),
(r'^final_conv', (99999,))
]
)
return matcher
| 17 | 1,333 | 97 | 1,430 | """ Bring-Your-Own-Blocks Network
A flexible network w/ dataclass based config for stacking those NN blocks.
This model is currently used to implement the following networks:
GPU Efficient (ResNets) - gernet_l/m/s (original versions called genet, but this was already used (by SENet author)).
Paper: `Neural Architect... | """ Bring-Your-Own-Blocks Network
A flexible network w/ dataclass based config for stacking those NN blocks.
This model is currently used to implement the following networks:
GPU Efficient (ResNets) - gernet_l/m/s (original versions called genet, but this was already used (by SENet author)).
Paper: `Neural Architect... |
17,523 | /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/byobnet.py | 394,395 | 340,406 | forward | DarkBlock | 395,406 | 395,406 | class DarkBlock(nn.Module):
""" DarkNet-like (1x1 + 3x3 w/ stride) block
The GE-Net impl included a 1x1 + 3x3 block in their search space. It was not used in the feature models.
This block is pretty much a DarkNet block (also DenseNet) hence the name. Neither DarkNet or DenseNet
uses strides within the... | class DarkBlock(nn.Module):
""" DarkNet-like (1x1 + 3x3 w/ stride) block
The GE-Net impl included a 1x1 + 3x3 block in their search space. It was not used in the feature models.
This block is pretty much a DarkNet block (also DenseNet) hence the name. Neither DarkNet or DenseNet
uses strides within the... | shortcut = x
x = self.conv1_1x1(x)
x = self.attn(x)
x = self.conv2_kxk(x)
x = self.attn_last(x)
x = self.drop_path(x)
if self.shortcut is not None:
x = x + self.shortcut(shortcut)
return self.act(x)
| shortcut = x
x = self.conv1_1x1(x)
x = self.attn(x)
x = self.conv2_kxk(x)
x = self.attn_last(x)
x = self.drop_path(x)
if self.shortcut is not None:
x = x + self.shortcut(shortcut)
return self.act(x)
| 30 | 831 | 118 | 949 | """ Bring-Your-Own-Blocks Network
A flexible network w/ dataclass based config for stacking those NN blocks.
This model is currently used to implement the following networks:
GPU Efficient (ResNets) - gernet_l/m/s (original versions called genet, but this was already used (by SENet author)).
Paper: `Neural Architect... | """ Bring-Your-Own-Blocks Network
A flexible network w/ dataclass based config for stacking those NN blocks.
This model is currently used to implement the following networks:
GPU Efficient (ResNets) - gernet_l/m/s (original versions called genet, but this was already used (by SENet author)).
Paper: `Neural Architect... |
17,524 | /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/byobnet.py | 188,189 | 167,192 | forward | DownsampleAvg | 189,192 | 189,192 | class DownsampleAvg(nn.Module):
def __init__(
self,
in_chs: int,
out_chs: int,
stride: int = 1,
dilation: int = 1,
apply_act: bool = False,
layers: LayerFn = None,
):
""" AvgPool Downsampling as in 'D' ResNet variants.""... | class DownsampleAvg(nn.Module):
def __init__(
self,
in_chs: int,
out_chs: int,
stride: int = 1,
dilation: int = 1,
apply_act: bool = False,
layers: LayerFn = None,
):
""" AvgPool Downsampling as in 'D' ResNet variants.""... | return self.conv(self.pool(x))
| return self.conv(self.pool(x))
| 2 | 298 | 18 | 316 | """ Bring-Your-Own-Blocks Network
A flexible network w/ dataclass based config for stacking those NN blocks.
This model is currently used to implement the following networks:
GPU Efficient (ResNets) - gernet_l/m/s (original versions called genet, but this was already used (by SENet author)).
Paper: `Neural Architect... | """ Bring-Your-Own-Blocks Network
A flexible network w/ dataclass based config for stacking those NN blocks.
This model is currently used to implement the following networks:
GPU Efficient (ResNets) - gernet_l/m/s (original versions called genet, but this was already used (by SENet author)).
Paper: `Neural Architect... |
17,525 | /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/byobnet.py | 459,460 | 406,471 | forward | EdgeBlock | 460,471 | 460,471 | class EdgeBlock(nn.Module):
""" EdgeResidual-like (3x3 + 1x1) block
A two layer block like DarkBlock, but with the order of the 3x3 and 1x1 convs reversed.
Very similar to the EfficientNet Edge-Residual block but this block it ends with activations, is
intended to be used with either expansion or bottl... | class EdgeBlock(nn.Module):
""" EdgeResidual-like (3x3 + 1x1) block
A two layer block like DarkBlock, but with the order of the 3x3 and 1x1 convs reversed.
Very similar to the EfficientNet Edge-Residual block but this block it ends with activations, is
intended to be used with either expansion or bottl... | shortcut = x
x = self.conv1_kxk(x)
x = self.attn(x)
x = self.conv2_1x1(x)
x = self.attn_last(x)
x = self.drop_path(x)
if self.shortcut is not None:
x = x + self.shortcut(shortcut)
return self.act(x)
| shortcut = x
x = self.conv1_kxk(x)
x = self.attn(x)
x = self.conv2_1x1(x)
x = self.attn_last(x)
x = self.drop_path(x)
if self.shortcut is not None:
x = x + self.shortcut(shortcut)
return self.act(x)
| 30 | 802 | 118 | 920 | """ Bring-Your-Own-Blocks Network
A flexible network w/ dataclass based config for stacking those NN blocks.
This model is currently used to implement the following networks:
GPU Efficient (ResNets) - gernet_l/m/s (original versions called genet, but this was already used (by SENet author)).
Paper: `Neural Architect... | """ Bring-Your-Own-Blocks Network
A flexible network w/ dataclass based config for stacking those NN blocks.
This model is currently used to implement the following networks:
GPU Efficient (ResNets) - gernet_l/m/s (original versions called genet, but this was already used (by SENet author)).
Paper: `Neural Architect... |
17,528 | /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/byobnet.py | 879,880 | 823,892 | forward | SelfAttnBlock | 880,892 | 880,892 | class SelfAttnBlock(nn.Module):
""" ResNet-like Bottleneck Block - 1x1 - optional kxk - self attn - 1x1
"""
def __init__(
self,
in_chs: int,
out_chs: int,
kernel_size: int = 3,
stride: int = 1,
dilation: Tuple[int, int] = (1, 1),
... | class SelfAttnBlock(nn.Module):
""" ResNet-like Bottleneck Block - 1x1 - optional kxk - self attn - 1x1
"""
def __init__(
self,
in_chs: int,
out_chs: int,
kernel_size: int = 3,
stride: int = 1,
dilation: Tuple[int, int] = (1, 1),
... | shortcut = x
x = self.conv1_1x1(x)
x = self.conv2_kxk(x)
x = self.self_attn(x)
x = self.post_attn(x)
x = self.conv3_1x1(x)
x = self.drop_path(x)
if self.shortcut is not None:
x = x + self.shortcut(shortcut)
return self.act(x)
| shortcut = x
x = self.conv1_1x1(x)
x = self.conv2_kxk(x)
x = self.self_attn(x)
x = self.post_attn(x)
x = self.conv3_1x1(x)
x = self.drop_path(x)
if self.shortcut is not None:
x = x + self.shortcut(shortcut)
return self.act(x)
| 33 | 857 | 136 | 993 | """ Bring-Your-Own-Blocks Network
A flexible network w/ dataclass based config for stacking those NN blocks.
This model is currently used to implement the following networks:
GPU Efficient (ResNets) - gernet_l/m/s (original versions called genet, but this was already used (by SENet author)).
Paper: `Neural Architect... | """ Bring-Your-Own-Blocks Network
A flexible network w/ dataclass based config for stacking those NN blocks.
This model is currently used to implement the following networks:
GPU Efficient (ResNets) - gernet_l/m/s (original versions called genet, but this was already used (by SENet author)).
Paper: `Neural Architect... |
17,531 | /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/cait.py | 46,47 | 27,71 | forward | ClassAttn | 47,71 | 47,71 | class ClassAttn(nn.Module):
# taken from https://github.com/rwightman/pytorch-image-models/blob/master/timm/models/vision_transformer.py
# with slight modifications to do CA
fused_attn: torch.jit.Final[bool]
def __init__(self, dim, num_heads=8, qkv_bias=False, attn_drop=0., proj_drop=0.):
super... | class ClassAttn(nn.Module):
# taken from https://github.com/rwightman/pytorch-image-models/blob/master/timm/models/vision_transformer.py
# with slight modifications to do CA
fused_attn: torch.jit.Final[bool]
def __init__(self, dim, num_heads=8, qkv_bias=False, attn_drop=0., proj_drop=0.):
super... | B, N, C = x.shape
q = self.q(x[:, 0]).unsqueeze(1).reshape(B, 1, self.num_heads, C // self.num_heads).permute(0, 2, 1, 3)
k = self.k(x).reshape(B, N, self.num_heads, C // self.num_heads).permute(0, 2, 1, 3)
v = self.v(x).reshape(B, N, self.num_heads, C // self.num_heads).permute(0, 2, 1,... | B, N, C = x.shape
q = self.q(x[:, 0]).unsqueeze(1).reshape(B, 1, self.num_heads, C // self.num_heads).permute(0, 2, 1, 3)
k = self.k(x).reshape(B, N, self.num_heads, C // self.num_heads).permute(0, 2, 1, 3)
v = self.v(x).reshape(B, N, self.num_heads, C // self.num_heads).permute(0, 2, 1,... | 90 | 328 | 400 | 728 | """ Class-Attention in Image Transformers (CaiT)
Paper: 'Going deeper with Image Transformers' - https://arxiv.org/abs/2103.17239
Original code and weights from https://github.com/facebookresearch/deit, copyright below
Modifications and additions for timm hacked together by / Copyright 2021, Ross Wightman
"""
# Copy... | """ Class-Attention in Image Transformers (CaiT)
Paper: 'Going deeper with Image Transformers' - https://arxiv.org/abs/2103.17239
Original code and weights from https://github.com/facebookresearch/deit, copyright below
Modifications and additions for timm hacked together by / Copyright 2021, Ross Wightman
"""
# Copy... |
17,532 | /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/cait.py | 198,199 | 159,204 | forward | LayerScaleBlock | 199,204 | 199,204 | class LayerScaleBlock(nn.Module):
# taken from https://github.com/rwightman/pytorch-image-models/blob/master/timm/models/vision_transformer.py
# with slight modifications to add layerScale
def __init__(
self,
dim,
num_heads,
mlp_ratio=4.,
qkv_bias=... | class LayerScaleBlock(nn.Module):
# taken from https://github.com/rwightman/pytorch-image-models/blob/master/timm/models/vision_transformer.py
# with slight modifications to add layerScale
def __init__(
self,
dim,
num_heads,
mlp_ratio=4.,
qkv_bias=... | x = x + self.drop_path(self.gamma_1 * self.attn(self.norm1(x)))
x = x + self.drop_path(self.gamma_2 * self.mlp(self.norm2(x)))
return x
| x = x + self.drop_path(self.gamma_1 * self.attn(self.norm1(x)))
x = x + self.drop_path(self.gamma_2 * self.mlp(self.norm2(x)))
return x
| 16 | 460 | 71 | 531 | """ Class-Attention in Image Transformers (CaiT)
Paper: 'Going deeper with Image Transformers' - https://arxiv.org/abs/2103.17239
Original code and weights from https://github.com/facebookresearch/deit, copyright below
Modifications and additions for timm hacked together by / Copyright 2021, Ross Wightman
"""
# Copy... | """ Class-Attention in Image Transformers (CaiT)
Paper: 'Going deeper with Image Transformers' - https://arxiv.org/abs/2103.17239
Original code and weights from https://github.com/facebookresearch/deit, copyright below
Modifications and additions for timm hacked together by / Copyright 2021, Ross Wightman
"""
# Copy... |
17,533 | /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/cait.py | 110,111 | 71,117 | forward | LayerScaleBlockClassAttn | 111,117 | 111,117 | class LayerScaleBlockClassAttn(nn.Module):
# taken from https://github.com/rwightman/pytorch-image-models/blob/master/timm/models/vision_transformer.py
# with slight modifications to add CA and LayerScale
def __init__(
self,
dim,
num_heads,
mlp_ratio=4.,
... | class LayerScaleBlockClassAttn(nn.Module):
# taken from https://github.com/rwightman/pytorch-image-models/blob/master/timm/models/vision_transformer.py
# with slight modifications to add CA and LayerScale
def __init__(
self,
dim,
num_heads,
mlp_ratio=4.,
... | u = torch.cat((x_cls, x), dim=1)
x_cls = x_cls + self.drop_path(self.gamma_1 * self.attn(self.norm1(u)))
x_cls = x_cls + self.drop_path(self.gamma_2 * self.mlp(self.norm2(x_cls)))
return x_cls
| u = torch.cat((x_cls, x), dim=1)
x_cls = x_cls + self.drop_path(self.gamma_1 * self.attn(self.norm1(u)))
x_cls = x_cls + self.drop_path(self.gamma_2 * self.mlp(self.norm2(x_cls)))
return x_cls
| 21 | 467 | 103 | 570 | """ Class-Attention in Image Transformers (CaiT)
Paper: 'Going deeper with Image Transformers' - https://arxiv.org/abs/2103.17239
Original code and weights from https://github.com/facebookresearch/deit, copyright below
Modifications and additions for timm hacked together by / Copyright 2021, Ross Wightman
"""
# Copy... | """ Class-Attention in Image Transformers (CaiT)
Paper: 'Going deeper with Image Transformers' - https://arxiv.org/abs/2103.17239
Original code and weights from https://github.com/facebookresearch/deit, copyright below
Modifications and additions for timm hacked together by / Copyright 2021, Ross Wightman
"""
# Copy... |
17,534 | /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/cait.py | 139,140 | 117,159 | forward | TalkingHeadAttn | 140,159 | 140,159 | class TalkingHeadAttn(nn.Module):
# taken from https://github.com/rwightman/pytorch-image-models/blob/master/timm/models/vision_transformer.py
# with slight modifications to add Talking Heads Attention (https://arxiv.org/pdf/2003.02436v1.pdf)
def __init__(self, dim, num_heads=8, qkv_bias=False, attn_drop=0.... | class TalkingHeadAttn(nn.Module):
# taken from https://github.com/rwightman/pytorch-image-models/blob/master/timm/models/vision_transformer.py
# with slight modifications to add Talking Heads Attention (https://arxiv.org/pdf/2003.02436v1.pdf)
def __init__(self, dim, num_heads=8, qkv_bias=False, attn_drop=0.... | B, N, C = x.shape
qkv = self.qkv(x).reshape(B, N, 3, self.num_heads, C // self.num_heads).permute(2, 0, 3, 1, 4)
q, k, v = qkv[0] * self.scale, qkv[1], qkv[2]
attn = q @ k.transpose(-2, -1)
attn = self.proj_l(attn.permute(0, 2, 3, 1)).permute(0, 3, 1, 2)
attn = attn.so... | B, N, C = x.shape
qkv = self.qkv(x).reshape(B, N, 3, self.num_heads, C // self.num_heads).permute(2, 0, 3, 1, 4)
q, k, v = qkv[0] * self.scale, qkv[1], qkv[2]
attn = q @ k.transpose(-2, -1)
attn = self.proj_l(attn.permute(0, 2, 3, 1)).permute(0, 3, 1, 2)
attn = attn.so... | 73 | 333 | 315 | 648 | """ Class-Attention in Image Transformers (CaiT)
Paper: 'Going deeper with Image Transformers' - https://arxiv.org/abs/2103.17239
Original code and weights from https://github.com/facebookresearch/deit, copyright below
Modifications and additions for timm hacked together by / Copyright 2021, Ross Wightman
"""
# Copy... | """ Class-Attention in Image Transformers (CaiT)
Paper: 'Going deeper with Image Transformers' - https://arxiv.org/abs/2103.17239
Original code and weights from https://github.com/facebookresearch/deit, copyright below
Modifications and additions for timm hacked together by / Copyright 2021, Ross Wightman
"""
# Copy... |
17,537 | /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/coat.py | 149,150 | 141,168 | forward | ConvPosEnc | 150,168 | 150,168 | class ConvPosEnc(nn.Module):
""" Convolutional Position Encoding.
Note: This module is similar to the conditional position encoding in CPVT.
"""
def __init__(self, dim, k=3):
super(ConvPosEnc, self).__init__()
self.proj = nn.Conv2d(dim, dim, k, 1, k//2, groups=dim)
def for... | class ConvPosEnc(nn.Module):
""" Convolutional Position Encoding.
Note: This module is similar to the conditional position encoding in CPVT.
"""
def __init__(self, dim, k=3):
super(ConvPosEnc, self).__init__()
self.proj = nn.Conv2d(dim, dim, k, 1, k//2, groups=dim)
def for... | B, N, C = x.shape
H, W = size
_assert(N == 1 + H * W, '')
# Extract CLS token and image tokens.
cls_token, img_tokens = x[:, :1], x[:, 1:] # [B, 1, C], [B, H*W, C]
# Depthwise convolution.
feat = img_tokens.transpose(1, 2).view(B, C, H, W)
x = s... | B, N, C = x.shape
H, W = size
_assert(N == 1 + H * W, '')
# Extract CLS token and image tokens.
cls_token, img_tokens = x[:, :1], x[:, 1:] # [B, 1, C], [B, H*W, C]
# Depthwise convolution.
feat = img_tokens.transpose(1, 2).view(B, C, H, W)
x = s... | 69 | 134 | 213 | 347 | """
CoaT architecture.
Paper: Co-Scale Conv-Attentional Image Transformers - https://arxiv.org/abs/2104.06399
Official CoaT code at: https://github.com/mlpc-ucsd/CoaT
Modified from timm/models/vision_transformer.py
"""
from functools import partial
from typing import Tuple, List, Union
import torch
import torch.nn... | """
CoaT architecture.
Paper: Co-Scale Conv-Attentional Image Transformers - https://arxiv.org/abs/2104.06399
Official CoaT code at: https://github.com/mlpc-ucsd/CoaT
Modified from timm/models/vision_transformer.py
"""
from functools import partial
from typing import Tuple, List, Union
import torch
import torch.nn... |
17,538 | /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/coat.py | 69,70 | 25,91 | forward | ConvRelPosEnc | 70,91 | 70,91 | class ConvRelPosEnc(nn.Module):
""" Convolutional relative position encoding. """
def __init__(self, head_chs, num_heads, window):
"""
Initialization.
Ch: Channels per head.
h: Number of heads.
window: Window size(s) in convolutional relative positional encodi... | class ConvRelPosEnc(nn.Module):
""" Convolutional relative position encoding. """
def __init__(self, head_chs, num_heads, window):
"""
Initialization.
Ch: Channels per head.
h: Number of heads.
window: Window size(s) in convolutional relative positional encodi... | B, num_heads, N, C = q.shape
H, W = size
_assert(N == 1 + H * W, '')
# Convolutional relative position encoding.
q_img = q[:, :, 1:, :] # [B, h, H*W, Ch]
v_img = v[:, :, 1:, :] # [B, h, H*W, Ch]
v_img = v_img.transpose(-1, -2).reshape(B, num_heads * C, H, W)
... | B, num_heads, N, C = q.shape
H, W = size
_assert(N == 1 + H * W, '')
# Convolutional relative position encoding.
q_img = q[:, :, 1:, :] # [B, h, H*W, Ch]
v_img = v[:, :, 1:, :] # [B, h, H*W, Ch]
v_img = v_img.transpose(-1, -2).reshape(B, num_heads * C, H, W)
... | 107 | 592 | 380 | 972 | """
CoaT architecture.
Paper: Co-Scale Conv-Attentional Image Transformers - https://arxiv.org/abs/2104.06399
Official CoaT code at: https://github.com/mlpc-ucsd/CoaT
Modified from timm/models/vision_transformer.py
"""
from functools import partial
from typing import Tuple, List, Union
import torch
import torch.nn... | """
CoaT architecture.
Paper: Co-Scale Conv-Attentional Image Transformers - https://arxiv.org/abs/2104.06399
Official CoaT code at: https://github.com/mlpc-ucsd/CoaT
Modified from timm/models/vision_transformer.py
"""
from functools import partial
from typing import Tuple, List, Union
import torch
import torch.nn... |
17,539 | /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/coat.py | 115,116 | 91,141 | forward | FactorAttnConvRelPosEnc | 116,141 | 116,141 | class FactorAttnConvRelPosEnc(nn.Module):
""" Factorized attention with convolutional relative position encoding class. """
def __init__(
self,
dim,
num_heads=8,
qkv_bias=False,
attn_drop=0.,
proj_drop=0.,
shared_crpe=None,
... | class FactorAttnConvRelPosEnc(nn.Module):
""" Factorized attention with convolutional relative position encoding class. """
def __init__(
self,
dim,
num_heads=8,
qkv_bias=False,
attn_drop=0.,
proj_drop=0.,
shared_crpe=None,
... | B, N, C = x.shape
# Generate Q, K, V.
qkv = self.qkv(x).reshape(B, N, 3, self.num_heads, C // self.num_heads).permute(2, 0, 3, 1, 4)
q, k, v = qkv.unbind(0) # [B, h, N, Ch]
# Factorized attention.
k_softmax = k.softmax(dim=2)
factor_att = k_softmax.transpose(-1... | B, N, C = x.shape
# Generate Q, K, V.
qkv = self.qkv(x).reshape(B, N, 3, self.num_heads, C // self.num_heads).permute(2, 0, 3, 1, 4)
q, k, v = qkv.unbind(0) # [B, h, N, Ch]
# Factorized attention.
k_softmax = k.softmax(dim=2)
factor_att = k_softmax.transpose(-1... | 107 | 293 | 346 | 639 | """
CoaT architecture.
Paper: Co-Scale Conv-Attentional Image Transformers - https://arxiv.org/abs/2104.06399
Official CoaT code at: https://github.com/mlpc-ucsd/CoaT
Modified from timm/models/vision_transformer.py
"""
from functools import partial
from typing import Tuple, List, Union
import torch
import torch.nn... | """
CoaT architecture.
Paper: Co-Scale Conv-Attentional Image Transformers - https://arxiv.org/abs/2104.06399
Official CoaT code at: https://github.com/mlpc-ucsd/CoaT
Modified from timm/models/vision_transformer.py
"""
from functools import partial
from typing import Tuple, List, Union
import torch
import torch.nn... |
17,540 | /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/coat.py | 319,320 | 226,354 | forward | ParallelBlock | 320,354 | 320,354 | class ParallelBlock(nn.Module):
""" Parallel block class. """
def __init__(
self,
dims,
num_heads,
mlp_ratios=[],
qkv_bias=False,
proj_drop=0.,
attn_drop=0.,
drop_path=0.,
act_layer=nn.GELU,
n... | class ParallelBlock(nn.Module):
""" Parallel block class. """
def __init__(
self,
dims,
num_heads,
mlp_ratios=[],
qkv_bias=False,
proj_drop=0.,
attn_drop=0.,
drop_path=0.,
act_layer=nn.GELU,
n... | _, S2, S3, S4 = sizes
cur2 = self.norm12(x2)
cur3 = self.norm13(x3)
cur4 = self.norm14(x4)
cur2 = self.factoratt_crpe2(cur2, size=S2)
cur3 = self.factoratt_crpe3(cur3, size=S3)
cur4 = self.factoratt_crpe4(cur4, size=S4)
upsample3_2 = self.upsample(cur3, fa... | _, S2, S3, S4 = sizes
cur2 = self.norm12(x2)
cur3 = self.norm13(x3)
cur4 = self.norm14(x4)
cur2 = self.factoratt_crpe2(cur2, size=S2)
cur3 = self.factoratt_crpe3(cur3, size=S3)
cur4 = self.factoratt_crpe4(cur4, size=S4)
upsample3_2 = self.upsample(cur3, fa... | 133 | 1,144 | 577 | 1,721 | """
CoaT architecture.
Paper: Co-Scale Conv-Attentional Image Transformers - https://arxiv.org/abs/2104.06399
Official CoaT code at: https://github.com/mlpc-ucsd/CoaT
Modified from timm/models/vision_transformer.py
"""
from functools import partial
from typing import Tuple, List, Union
import torch
import torch.nn... | """
CoaT architecture.
Paper: Co-Scale Conv-Attentional Image Transformers - https://arxiv.org/abs/2104.06399
Official CoaT code at: https://github.com/mlpc-ucsd/CoaT
Modified from timm/models/vision_transformer.py
"""
from functools import partial
from typing import Tuple, List, Union
import torch
import torch.nn... |
17,541 | /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/coat.py | 211,212 | 168,226 | forward | SerialBlock | 212,226 | 212,226 | class SerialBlock(nn.Module):
""" Serial block class.
Note: In this implementation, each serial block only contains a conv-attention and a FFN (MLP) module. """
def __init__(
self,
dim,
num_heads,
mlp_ratio=4.,
qkv_bias=False,
proj_... | class SerialBlock(nn.Module):
""" Serial block class.
Note: In this implementation, each serial block only contains a conv-attention and a FFN (MLP) module. """
def __init__(
self,
dim,
num_heads,
mlp_ratio=4.,
qkv_bias=False,
proj_... | # Conv-Attention.
x = self.cpe(x, size)
cur = self.norm1(x)
cur = self.factoratt_crpe(cur, size)
x = x + self.drop_path(cur)
# MLP.
cur = self.norm2(x)
cur = self.mlp(cur)
x = x + self.drop_path(cur)
return x
| # Conv-Attention.
x = self.cpe(x, size)
cur = self.norm1(x)
cur = self.factoratt_crpe(cur, size)
x = x + self.drop_path(cur)
# MLP.
cur = self.norm2(x)
cur = self.mlp(cur)
x = x + self.drop_path(cur)
return x
| 33 | 437 | 129 | 566 | """
CoaT architecture.
Paper: Co-Scale Conv-Attentional Image Transformers - https://arxiv.org/abs/2104.06399
Official CoaT code at: https://github.com/mlpc-ucsd/CoaT
Modified from timm/models/vision_transformer.py
"""
from functools import partial
from typing import Tuple, List, Union
import torch
import torch.nn... | """
CoaT architecture.
Paper: Co-Scale Conv-Attentional Image Transformers - https://arxiv.org/abs/2104.06399
Official CoaT code at: https://github.com/mlpc-ucsd/CoaT
Modified from timm/models/vision_transformer.py
"""
from functools import partial
from typing import Tuple, List, Union
import torch
import torch.nn... |
17,543 | /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/convit.py | 234,235 | 188,240 | forward | Block | 235,240 | 235,240 | class Block(nn.Module):
def __init__(
self,
dim,
num_heads,
mlp_ratio=4.,
qkv_bias=False,
proj_drop=0.,
attn_drop=0.,
drop_path=0.,
act_layer=nn.GELU,
norm_layer=LayerNorm,
use_gpsa=T... | class Block(nn.Module):
def __init__(
self,
dim,
num_heads,
mlp_ratio=4.,
qkv_bias=False,
proj_drop=0.,
attn_drop=0.,
drop_path=0.,
act_layer=nn.GELU,
norm_layer=LayerNorm,
use_gpsa=T... | x = x + self.drop_path(self.attn(self.norm1(x)))
x = x + self.drop_path(self.mlp(self.norm2(x)))
return x
| x = x + self.drop_path(self.attn(self.norm1(x)))
x = x + self.drop_path(self.mlp(self.norm2(x)))
return x
| 12 | 454 | 59 | 513 | """ ConViT Model
@article{d2021convit,
title={ConViT: Improving Vision Transformers with Soft Convolutional Inductive Biases},
author={d'Ascoli, St{\'e}phane and Touvron, Hugo and Leavitt, Matthew and Morcos, Ari and Biroli, Giulio and Sagun, Levent},
journal={arXiv preprint arXiv:2103.10697},
year={2021}
}
P... | """ ConViT Model
@article{d2021convit,
title={ConViT: Improving Vision Transformers with Soft Convolutional Inductive Biases},
author={d'Ascoli, St{\'e}phane and Touvron, Hugo and Leavitt, Matthew and Morcos, Ari and Biroli, Giulio and Sagun, Levent},
journal={arXiv preprint arXiv:2103.10697},
year={2021}
}
P... |
17,544 | /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/convit.py | 359,360 | 240,384 | forward_features | ConVit | 360,372 | 360,372 | class ConVit(nn.Module):
""" Vision Transformer with support for patch or hybrid CNN input stage
"""
def __init__(
self,
img_size=224,
patch_size=16,
in_chans=3,
num_classes=1000,
global_pool='token',
embed_dim=768,
... | class ConVit(nn.Module):
""" Vision Transformer with support for patch or hybrid CNN input stage
"""
def __init__(
self,
img_size=224,
patch_size=16,
in_chans=3,
num_classes=1000,
global_pool='token',
embed_dim=768,
... | x = self.patch_embed(x)
if self.use_pos_embed:
x = x + self.pos_embed
x = self.pos_drop(x)
cls_tokens = self.cls_token.expand(x.shape[0], -1, -1)
for u, blk in enumerate(self.blocks):
if u == self.local_up_to_layer:
x = torch.cat((cls_token... | x = self.patch_embed(x)
if self.use_pos_embed:
x = x + self.pos_embed
x = self.pos_drop(x)
cls_tokens = self.cls_token.expand(x.shape[0], -1, -1)
for u, blk in enumerate(self.blocks):
if u == self.local_up_to_layer:
x = torch.cat((cls_token... | 40 | 1,575 | 160 | 1,735 | """ ConViT Model
@article{d2021convit,
title={ConViT: Improving Vision Transformers with Soft Convolutional Inductive Biases},
author={d'Ascoli, St{\'e}phane and Touvron, Hugo and Leavitt, Matthew and Morcos, Ari and Biroli, Giulio and Sagun, Levent},
journal={arXiv preprint arXiv:2103.10697},
year={2021}
}
P... | """ ConViT Model
@article{d2021convit,
title={ConViT: Improving Vision Transformers with Soft Convolutional Inductive Biases},
author={d'Ascoli, St{\'e}phane and Touvron, Hugo and Leavitt, Matthew and Morcos, Ari and Biroli, Giulio and Sagun, Levent},
journal={arXiv preprint arXiv:2103.10697},
year={2021}
}
P... |
17,545 | /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/convit.py | 69,70 | 41,133 | forward | GPSA | 70,80 | 70,80 | @register_notrace_module # reason: FX can't symbolically trace control flow in forward method
class GPSA(nn.Module):
def __init__(
self,
dim,
num_heads=8,
qkv_bias=False,
attn_drop=0.,
proj_drop=0.,
locality_strength=1.,
):
... | @register_notrace_module # reason: FX can't symbolically trace control flow in forward method
class GPSA(nn.Module):
def __init__(
self,
dim,
num_heads=8,
qkv_bias=False,
attn_drop=0.,
proj_drop=0.,
locality_strength=1.,
):
... | B, N, C = x.shape
if self.rel_indices is None or self.rel_indices.shape[1] != N:
self.rel_indices = self.get_rel_indices(N)
attn = self.get_attention(x)
v = self.v(x).reshape(B, N, self.num_heads, C // self.num_heads).permute(0, 2, 1, 3)
x = (attn @ v).transpose(1, 2)... | B, N, C = x.shape
if self.rel_indices is None or self.rel_indices.shape[1] != N:
self.rel_indices = self.get_rel_indices(N)
attn = self.get_attention(x)
v = self.v(x).reshape(B, N, self.num_heads, C // self.num_heads).permute(0, 2, 1, 3)
x = (attn @ v).transpose(1, 2)... | 46 | 1,361 | 183 | 1,544 | """ ConViT Model
@article{d2021convit,
title={ConViT: Improving Vision Transformers with Soft Convolutional Inductive Biases},
author={d'Ascoli, St{\'e}phane and Touvron, Hugo and Leavitt, Matthew and Morcos, Ari and Biroli, Giulio and Sagun, Levent},
journal={arXiv preprint arXiv:2103.10697},
year={2021}
}
P... | """ ConViT Model
@article{d2021convit,
title={ConViT: Improving Vision Transformers with Soft Convolutional Inductive Biases},
author={d'Ascoli, St{\'e}phane and Touvron, Hugo and Leavitt, Matthew and Morcos, Ari and Biroli, Giulio and Sagun, Levent},
journal={arXiv preprint arXiv:2103.10697},
year={2021}
}
P... |
17,546 | /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/convit.py | 173,174 | 133,188 | forward | MHSA | 174,188 | 174,188 | class MHSA(nn.Module):
def __init__(
self,
dim,
num_heads=8,
qkv_bias=False,
attn_drop=0.,
proj_drop=0.,
):
super().__init__()
self.num_heads = num_heads
head_dim = dim // num_heads
self.scale = head_dim ** -... | class MHSA(nn.Module):
def __init__(
self,
dim,
num_heads=8,
qkv_bias=False,
attn_drop=0.,
proj_drop=0.,
):
super().__init__()
self.num_heads = num_heads
head_dim = dim // num_heads
self.scale = head_dim ** -... | B, N, C = x.shape
qkv = self.qkv(x).reshape(B, N, 3, self.num_heads, C // self.num_heads).permute(2, 0, 3, 1, 4)
q, k, v = qkv.unbind(0)
attn = (q @ k.transpose(-2, -1)) * self.scale
attn = attn.softmax(dim=-1)
attn = self.attn_drop(attn)
x = (attn @ v).transpos... | B, N, C = x.shape
qkv = self.qkv(x).reshape(B, N, 3, self.num_heads, C // self.num_heads).permute(2, 0, 3, 1, 4)
q, k, v = qkv.unbind(0)
attn = (q @ k.transpose(-2, -1)) * self.scale
attn = attn.softmax(dim=-1)
attn = self.attn_drop(attn)
x = (attn @ v).transpos... | 53 | 574 | 210 | 784 | """ ConViT Model
@article{d2021convit,
title={ConViT: Improving Vision Transformers with Soft Convolutional Inductive Biases},
author={d'Ascoli, St{\'e}phane and Touvron, Hugo and Leavitt, Matthew and Morcos, Ari and Biroli, Giulio and Sagun, Levent},
journal={arXiv preprint arXiv:2103.10697},
year={2021}
}
P... | """ ConViT Model
@article{d2021convit,
title={ConViT: Improving Vision Transformers with Soft Convolutional Inductive Biases},
author={d'Ascoli, St{\'e}phane and Touvron, Hugo and Leavitt, Matthew and Morcos, Ari and Biroli, Giulio and Sagun, Levent},
journal={arXiv preprint arXiv:2103.10697},
year={2021}
}
P... |
17,547 | /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/convmixer.py | 135,137 | 1,145 | convmixer_768_32 | 137,141 | 137,141 | """ ConvMixer
"""
import torch
import torch.nn as nn
from timm.data import IMAGENET_DEFAULT_MEAN, IMAGENET_DEFAULT_STD
from timm.layers import SelectAdaptivePool2d
from ._registry import register_model, generate_default_cfgs
from ._builder import build_model_with_cfg
from ._manipulate import checkpoint_seq
__all__ =... | """ ConvMixer
"""
import torch
import torch.nn as nn
from timm.data import IMAGENET_DEFAULT_MEAN, IMAGENET_DEFAULT_STD
from timm.layers import SelectAdaptivePool2d
from ._registry import register_model, generate_default_cfgs
from ._builder import build_model_with_cfg
from ._manipulate import checkpoint_seq
__all__ =... | model_args = dict(dim=768, depth=32, kernel_size=7, patch_size=7, act_layer=nn.ReLU, **kwargs)
return _create_convmixer('convmixer_768_32', pretrained, **model_args)
| model_args = dict(dim=768, depth=32, kernel_size=7, patch_size=7, act_layer=nn.ReLU, **kwargs)
return _create_convmixer('convmixer_768_32', pretrained, **model_args)
| 12 | 1,611 | 79 | 1,690 | """ ConvMixer
"""
import torch
import torch.nn as nn
from timm.data import IMAGENET_DEFAULT_MEAN, IMAGENET_DEFAULT_STD
from timm.layers import SelectAdaptivePool2d
from ._registry import register_model, generate_default_cfgs
from ._builder import build_model_with_cfg
from ._manipulate import checkpoint_seq
__all__ =... | """ ConvMixer
"""
import torch
import torch.nn as nn
from timm.data import IMAGENET_DEFAULT_MEAN, IMAGENET_DEFAULT_STD
from timm.layers import SelectAdaptivePool2d
from ._registry import register_model, generate_default_cfgs
from ._builder import build_model_with_cfg
from ._manipulate import checkpoint_seq
__all__ =... | |
17,548 | /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/convmixer.py | 92,93 | 25,103 | forward_head | ConvMixer | 93,97 | 93,97 | class ConvMixer(nn.Module):
def __init__(
self,
dim,
depth,
kernel_size=9,
patch_size=7,
in_chans=3,
num_classes=1000,
global_pool='avg',
drop_rate=0.,
act_layer=nn.GELU,
**kwargs,
... | class ConvMixer(nn.Module):
def __init__(
self,
dim,
depth,
kernel_size=9,
patch_size=7,
in_chans=3,
num_classes=1000,
global_pool='avg',
drop_rate=0.,
act_layer=nn.GELU,
**kwargs,
... | x = self.pooling(x)
x = self.head_drop(x)
return x if pre_logits else self.head(x)
| x = self.pooling(x)
x = self.head_drop(x)
return x if pre_logits else self.head(x)
| 12 | 798 | 44 | 842 | """ ConvMixer
"""
import torch
import torch.nn as nn
from timm.data import IMAGENET_DEFAULT_MEAN, IMAGENET_DEFAULT_STD
from timm.layers import SelectAdaptivePool2d
from ._registry import register_model, generate_default_cfgs
from ._builder import build_model_with_cfg
from ._manipulate import checkpoint_seq
__all__ =... | """ ConvMixer
"""
import torch
import torch.nn as nn
from timm.data import IMAGENET_DEFAULT_MEAN, IMAGENET_DEFAULT_STD
from timm.layers import SelectAdaptivePool2d
from ._registry import register_model, generate_default_cfgs
from ._builder import build_model_with_cfg
from ._manipulate import checkpoint_seq
__all__ =... |
17,549 | /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/convmixer.py | 21,22 | 16,25 | forward | Residual | 22,25 | 22,25 | class Residual(nn.Module):
def __init__(self, fn):
super().__init__()
self.fn = fn
def forward(self, x):
FILL_FUNC_BODY | class Residual(nn.Module):
def __init__(self, fn):
super().__init__()
self.fn = fn
def forward(self, x):
FILL_FUNC_BODY | return self.fn(x) + x
| return self.fn(x) + x
| 4 | 54 | 16 | 70 | """ ConvMixer
"""
import torch
import torch.nn as nn
from timm.data import IMAGENET_DEFAULT_MEAN, IMAGENET_DEFAULT_STD
from timm.layers import SelectAdaptivePool2d
from ._registry import register_model, generate_default_cfgs
from ._builder import build_model_with_cfg
from ._manipulate import checkpoint_seq
__all__ =... | """ ConvMixer
"""
import torch
import torch.nn as nn
from timm.data import IMAGENET_DEFAULT_MEAN, IMAGENET_DEFAULT_STD
from timm.layers import SelectAdaptivePool2d
from ._registry import register_model, generate_default_cfgs
from ._builder import build_model_with_cfg
from ._manipulate import checkpoint_seq
__all__ =... |
17,552 | /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/convnext.py | 151,152 | 81,169 | forward | ConvNeXtBlock | 152,169 | 152,169 | class ConvNeXtBlock(nn.Module):
""" ConvNeXt Block
There are two equivalent implementations:
(1) DwConv -> LayerNorm (channels_first) -> 1x1 Conv -> GELU -> 1x1 Conv; all in (N, C, H, W)
(2) DwConv -> Permute to (N, H, W, C); LayerNorm (channels_last) -> Linear -> GELU -> Linear; Permute back
U... | class ConvNeXtBlock(nn.Module):
""" ConvNeXt Block
There are two equivalent implementations:
(1) DwConv -> LayerNorm (channels_first) -> 1x1 Conv -> GELU -> 1x1 Conv; all in (N, C, H, W)
(2) DwConv -> Permute to (N, H, W, C); LayerNorm (channels_last) -> Linear -> GELU -> Linear; Permute back
U... | shortcut = x
x = self.conv_dw(x)
if self.use_conv_mlp:
x = self.norm(x)
x = self.mlp(x)
else:
x = x.permute(0, 2, 3, 1)
x = self.norm(x)
x = self.mlp(x)
x = x.permute(0, 3, 1, 2)
if self.gamma is not None:
... | shortcut = x
x = self.conv_dw(x)
if self.use_conv_mlp:
x = self.norm(x)
x = self.mlp(x)
else:
x = x.permute(0, 2, 3, 1)
x = self.norm(x)
x = self.mlp(x)
x = x.permute(0, 3, 1, 2)
if self.gamma is not None:
... | 51 | 1,069 | 200 | 1,269 | """ ConvNeXt
Papers:
* `A ConvNet for the 2020s` - https://arxiv.org/pdf/2201.03545.pdf
@Article{liu2022convnet,
author = {Zhuang Liu and Hanzi Mao and Chao-Yuan Wu and Christoph Feichtenhofer and Trevor Darrell and Saining Xie},
title = {A ConvNet for the 2020s},
journal = {Proceedings of the IEEE/CVF Confer... | """ ConvNeXt
Papers:
* `A ConvNet for the 2020s` - https://arxiv.org/pdf/2201.03545.pdf
@Article{liu2022convnet,
author = {Zhuang Liu and Hanzi Mao and Chao-Yuan Wu and Christoph Feichtenhofer and Trevor Darrell and Saining Xie},
title = {A ConvNet for the 2020s},
journal = {Proceedings of the IEEE/CVF Confer... |
17,553 | /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/convnext.py | 229,230 | 169,238 | forward | ConvNeXtStage | 230,238 | 230,238 | class ConvNeXtStage(nn.Module):
def __init__(
self,
in_chs,
out_chs,
kernel_size=7,
stride=2,
depth=2,
dilation=(1, 1),
drop_path_rates=None,
ls_init_value=1.0,
conv_mlp=False,
conv_b... | class ConvNeXtStage(nn.Module):
def __init__(
self,
in_chs,
out_chs,
kernel_size=7,
stride=2,
depth=2,
dilation=(1, 1),
drop_path_rates=None,
ls_init_value=1.0,
conv_mlp=False,
conv_b... | 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 | 623 | 77 | 700 | """ ConvNeXt
Papers:
* `A ConvNet for the 2020s` - https://arxiv.org/pdf/2201.03545.pdf
@Article{liu2022convnet,
author = {Zhuang Liu and Hanzi Mao and Chao-Yuan Wu and Christoph Feichtenhofer and Trevor Darrell and Saining Xie},
title = {A ConvNet for the 2020s},
journal = {Proceedings of the IEEE/CVF Confer... | """ ConvNeXt
Papers:
* `A ConvNet for the 2020s` - https://arxiv.org/pdf/2201.03545.pdf
@Article{liu2022convnet,
author = {Zhuang Liu and Hanzi Mao and Chao-Yuan Wu and Christoph Feichtenhofer and Trevor Darrell and Saining Xie},
title = {A ConvNet for the 2020s},
journal = {Proceedings of the IEEE/CVF Confer... |
17,554 | /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/convnext.py | 75,76 | 59,81 | forward | Downsample | 76,81 | 76,81 | class Downsample(nn.Module):
def __init__(self, in_chs, out_chs, stride=1, dilation=1):
super().__init__()
avg_stride = stride if dilation == 1 else 1
if stride > 1 or dilation > 1:
avg_pool_fn = AvgPool2dSame if avg_stride == 1 and dilation > 1 else nn.AvgPool2d
sel... | class Downsample(nn.Module):
def __init__(self, in_chs, out_chs, stride=1, dilation=1):
super().__init__()
avg_stride = stride if dilation == 1 else 1
if stride > 1 or dilation > 1:
avg_pool_fn = AvgPool2dSame if avg_stride == 1 and dilation > 1 else nn.AvgPool2d
sel... | x = self.pool(x)
x = self.conv(x)
return x
| x = self.pool(x)
x = self.conv(x)
return x
| 8 | 245 | 31 | 276 | """ ConvNeXt
Papers:
* `A ConvNet for the 2020s` - https://arxiv.org/pdf/2201.03545.pdf
@Article{liu2022convnet,
author = {Zhuang Liu and Hanzi Mao and Chao-Yuan Wu and Christoph Feichtenhofer and Trevor Darrell and Saining Xie},
title = {A ConvNet for the 2020s},
journal = {Proceedings of the IEEE/CVF Confer... | """ ConvNeXt
Papers:
* `A ConvNet for the 2020s` - https://arxiv.org/pdf/2201.03545.pdf
@Article{liu2022convnet,
author = {Zhuang Liu and Hanzi Mao and Chao-Yuan Wu and Christoph Feichtenhofer and Trevor Darrell and Saining Xie},
title = {A ConvNet for the 2020s},
journal = {Proceedings of the IEEE/CVF Confer... |
17,556 | /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/crossvit.py | 110,111 | 88,129 | forward | CrossAttention | 111,129 | 111,129 | class CrossAttention(nn.Module):
def __init__(
self,
dim,
num_heads=8,
qkv_bias=False,
attn_drop=0.,
proj_drop=0.,
):
super().__init__()
self.num_heads = num_heads
head_dim = dim // num_heads
# NOTE scale fac... | class CrossAttention(nn.Module):
def __init__(
self,
dim,
num_heads=8,
qkv_bias=False,
attn_drop=0.,
proj_drop=0.,
):
super().__init__()
self.num_heads = num_heads
head_dim = dim // num_heads
# NOTE scale fac... | B, N, C = x.shape
# B1C -> B1H(C/H) -> BH1(C/H)
q = self.wq(x[:, 0:1, ...]).reshape(B, 1, self.num_heads, C // self.num_heads).permute(0, 2, 1, 3)
# BNC -> BNH(C/H) -> BHN(C/H)
k = self.wk(x).reshape(B, N, self.num_heads, C // self.num_heads).permute(0, 2, 1, 3)
# BNC -> ... | B, N, C = x.shape
# B1C -> B1H(C/H) -> BH1(C/H)
q = self.wq(x[:, 0:1, ...]).reshape(B, 1, self.num_heads, C // self.num_heads).permute(0, 2, 1, 3)
# BNC -> BNH(C/H) -> BHN(C/H)
k = self.wk(x).reshape(B, N, self.num_heads, C // self.num_heads).permute(0, 2, 1, 3)
# BNC -> ... | 104 | 285 | 435 | 720 | """ CrossViT Model
@inproceedings{
chen2021crossvit,
title={{CrossViT: Cross-Attention Multi-Scale Vision Transformer for Image Classification}},
author={Chun-Fu (Richard) Chen and Quanfu Fan and Rameswar Panda},
booktitle={International Conference on Computer Vision (ICCV)},
year={2021}
}
Paper l... | """ CrossViT Model
@inproceedings{
chen2021crossvit,
title={{CrossViT: Cross-Attention Multi-Scale Vision Transformer for Image Classification}},
author={Chun-Fu (Richard) Chen and Quanfu Fan and Rameswar Panda},
booktitle={International Conference on Computer Vision (ICCV)},
year={2021}
}
Paper l... |
17,557 | /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/crossvit.py | 155,156 | 129,160 | forward | CrossAttentionBlock | 156,160 | 156,160 | class CrossAttentionBlock(nn.Module):
def __init__(
self,
dim,
num_heads,
mlp_ratio=4.,
qkv_bias=False,
proj_drop=0.,
attn_drop=0.,
drop_path=0.,
act_layer=nn.GELU,
norm_layer=nn.LayerNorm,
)... | class CrossAttentionBlock(nn.Module):
def __init__(
self,
dim,
num_heads,
mlp_ratio=4.,
qkv_bias=False,
proj_drop=0.,
attn_drop=0.,
drop_path=0.,
act_layer=nn.GELU,
norm_layer=nn.LayerNorm,
)... | x = x[:, 0:1, ...] + self.drop_path(self.attn(self.norm1(x)))
return x
| x = x[:, 0:1, ...] + self.drop_path(self.attn(self.norm1(x)))
return x
| 9 | 267 | 43 | 310 | """ CrossViT Model
@inproceedings{
chen2021crossvit,
title={{CrossViT: Cross-Attention Multi-Scale Vision Transformer for Image Classification}},
author={Chun-Fu (Richard) Chen and Quanfu Fan and Rameswar Panda},
booktitle={International Conference on Computer Vision (ICCV)},
year={2021}
}
Paper l... | """ CrossViT Model
@inproceedings{
chen2021crossvit,
title={{CrossViT: Cross-Attention Multi-Scale Vision Transformer for Image Classification}},
author={Chun-Fu (Richard) Chen and Quanfu Fan and Rameswar Panda},
booktitle={International Conference on Computer Vision (ICCV)},
year={2021}
}
Paper l... |
17,559 | /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/crossvit.py | 249,250 | 160,271 | forward | MultiScaleBlock | 250,271 | 250,271 | class MultiScaleBlock(nn.Module):
def __init__(
self,
dim,
patches,
depth,
num_heads,
mlp_ratio,
qkv_bias=False,
proj_drop=0.,
attn_drop=0.,
drop_path=0.,
act_layer=nn.GELU,
... | class MultiScaleBlock(nn.Module):
def __init__(
self,
dim,
patches,
depth,
num_heads,
mlp_ratio,
qkv_bias=False,
proj_drop=0.,
attn_drop=0.,
drop_path=0.,
act_layer=nn.GELU,
... |
outs_b = []
for i, block in enumerate(self.blocks):
outs_b.append(block(x[i]))
# only take the cls token out
proj_cls_token = torch.jit.annotate(List[torch.Tensor], [])
for i, proj in enumerate(self.projs):
proj_cls_token.append(proj(outs_b[i][:, 0:1, ..... |
outs_b = []
for i, block in enumerate(self.blocks):
outs_b.append(block(x[i]))
# only take the cls token out
proj_cls_token = torch.jit.annotate(List[torch.Tensor], [])
for i, proj in enumerate(self.projs):
proj_cls_token.append(proj(outs_b[i][:, 0:1, ..... | 70 | 995 | 318 | 1,313 | """ CrossViT Model
@inproceedings{
chen2021crossvit,
title={{CrossViT: Cross-Attention Multi-Scale Vision Transformer for Image Classification}},
author={Chun-Fu (Richard) Chen and Quanfu Fan and Rameswar Panda},
booktitle={International Conference on Computer Vision (ICCV)},
year={2021}
}
Paper l... | """ CrossViT Model
@inproceedings{
chen2021crossvit,
title={{CrossViT: Cross-Attention Multi-Scale Vision Transformer for Image Classification}},
author={Chun-Fu (Richard) Chen and Quanfu Fan and Rameswar Panda},
booktitle={International Conference on Computer Vision (ICCV)},
year={2021}
}
Paper l... |
17,560 | /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/crossvit.py | 77,78 | 45,88 | forward | PatchEmbed | 78,88 | 78,88 | class PatchEmbed(nn.Module):
""" Image to Patch Embedding
"""
def __init__(self, img_size=224, patch_size=16, in_chans=3, embed_dim=768, multi_conv=False):
super().__init__()
img_size = to_2tuple(img_size)
patch_size = to_2tuple(patch_size)
num_patches = (img_size[1] // patc... | class PatchEmbed(nn.Module):
""" Image to Patch Embedding
"""
def __init__(self, img_size=224, patch_size=16, in_chans=3, embed_dim=768, multi_conv=False):
super().__init__()
img_size = to_2tuple(img_size)
patch_size = to_2tuple(patch_size)
num_patches = (img_size[1] // patc... | B, C, H, W = x.shape
# FIXME look at relaxing size constraints
_assert(H == self.img_size[0],
f"Input image size ({H}*{W}) doesn't match model ({self.img_size[0]}*{self.img_size[1]}).")
_assert(W == self.img_size[1],
f"Input image size ({H}*{W}) doesn't ma... | B, C, H, W = x.shape
# FIXME look at relaxing size constraints
_assert(H == self.img_size[0],
f"Input image size ({H}*{W}) doesn't match model ({self.img_size[0]}*{self.img_size[1]}).")
_assert(W == self.img_size[1],
f"Input image size ({H}*{W}) doesn't ma... | 41 | 587 | 173 | 760 | """ CrossViT Model
@inproceedings{
chen2021crossvit,
title={{CrossViT: Cross-Attention Multi-Scale Vision Transformer for Image Classification}},
author={Chun-Fu (Richard) Chen and Quanfu Fan and Rameswar Panda},
booktitle={International Conference on Computer Vision (ICCV)},
year={2021}
}
Paper l... | """ CrossViT Model
@inproceedings{
chen2021crossvit,
title={{CrossViT: Cross-Attention Multi-Scale Vision Transformer for Image Classification}},
author={Chun-Fu (Richard) Chen and Quanfu Fan and Rameswar Panda},
booktitle={International Conference on Computer Vision (ICCV)},
year={2021}
}
Paper l... |
17,562 | /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/cspnet.py | 172,173 | 135,186 | forward | BottleneckBlock | 173,186 | 173,186 | class BottleneckBlock(nn.Module):
""" ResNe(X)t Bottleneck Block
"""
def __init__(
self,
in_chs,
out_chs,
dilation=1,
bottle_ratio=0.25,
groups=1,
act_layer=nn.ReLU,
norm_layer=nn.BatchNorm2d,
attn_l... | class BottleneckBlock(nn.Module):
""" ResNe(X)t Bottleneck Block
"""
def __init__(
self,
in_chs,
out_chs,
dilation=1,
bottle_ratio=0.25,
groups=1,
act_layer=nn.ReLU,
norm_layer=nn.BatchNorm2d,
attn_l... | shortcut = x
x = self.conv1(x)
x = self.conv2(x)
x = self.attn2(x)
x = self.conv3(x)
x = self.attn3(x)
x = self.drop_path(x) + shortcut
# FIXME partial shortcut needed if first block handled as per original, not used for my current impl
#x[:, :shor... | shortcut = x
x = self.conv1(x)
x = self.conv2(x)
x = self.attn2(x)
x = self.conv3(x)
x = self.attn3(x)
x = self.drop_path(x) + shortcut
# FIXME partial shortcut needed if first block handled as per original, not used for my current impl
#x[:, :shor... | 50 | 537 | 145 | 682 | """PyTorch CspNet
A PyTorch implementation of Cross Stage Partial Networks including:
* CSPResNet50
* CSPResNeXt50
* CSPDarkNet53
* and DarkNet53 for good measure
Based on paper `CSPNet: A New Backbone that can Enhance Learning Capability of CNN` - https://arxiv.org/abs/1911.11929
Reference impl via darknet cfg file... | """PyTorch CspNet
A PyTorch implementation of Cross Stage Partial Networks including:
* CSPResNet50
* CSPResNeXt50
* CSPDarkNet53
* and DarkNet53 for good measure
Based on paper `CSPNet: A New Backbone that can Enhance Learning Capability of CNN` - https://arxiv.org/abs/1911.11929
Reference impl via darknet cfg file... |
17,563 | /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/cspnet.py | 333,334 | 266,343 | forward | CrossStage | 334,343 | 334,343 | class CrossStage(nn.Module):
"""Cross Stage."""
def __init__(
self,
in_chs,
out_chs,
stride,
dilation,
depth,
block_ratio=1.,
bottle_ratio=1.,
expand_ratio=1.,
groups=1,
first_dila... | class CrossStage(nn.Module):
"""Cross Stage."""
def __init__(
self,
in_chs,
out_chs,
stride,
dilation,
depth,
block_ratio=1.,
bottle_ratio=1.,
expand_ratio=1.,
groups=1,
first_dila... | x = self.conv_down(x)
x = self.conv_exp(x)
xs, xb = x.split(self.expand_chs // 2, dim=1)
xb = self.blocks(xb)
xb = self.conv_transition_b(xb).contiguous()
out = self.conv_transition(torch.cat([xs, xb], dim=1))
return out
| x = self.conv_down(x)
x = self.conv_exp(x)
xs, xb = x.split(self.expand_chs // 2, dim=1)
xb = self.blocks(xb)
xb = self.conv_transition_b(xb).contiguous()
out = self.conv_transition(torch.cat([xs, xb], dim=1))
return out
| 26 | 923 | 121 | 1,044 | """PyTorch CspNet
A PyTorch implementation of Cross Stage Partial Networks including:
* CSPResNet50
* CSPResNeXt50
* CSPDarkNet53
* and DarkNet53 for good measure
Based on paper `CSPNet: A New Backbone that can Enhance Learning Capability of CNN` - https://arxiv.org/abs/1911.11929
Reference impl via darknet cfg file... | """PyTorch CspNet
A PyTorch implementation of Cross Stage Partial Networks including:
* CSPResNet50
* CSPResNeXt50
* CSPDarkNet53
* and DarkNet53 for good measure
Based on paper `CSPNet: A New Backbone that can Enhance Learning Capability of CNN` - https://arxiv.org/abs/1911.11929
Reference impl via darknet cfg file... |
17,564 | /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/cspnet.py | 409,410 | 343,418 | forward | CrossStage3 | 410,418 | 410,418 | class CrossStage3(nn.Module):
"""Cross Stage 3.
Similar to CrossStage, but with only one transition conv for the output.
"""
def __init__(
self,
in_chs,
out_chs,
stride,
dilation,
depth,
block_ratio=1.,
bottl... | class CrossStage3(nn.Module):
"""Cross Stage 3.
Similar to CrossStage, but with only one transition conv for the output.
"""
def __init__(
self,
in_chs,
out_chs,
stride,
dilation,
depth,
block_ratio=1.,
bottl... | x = self.conv_down(x)
x = self.conv_exp(x)
x1, x2 = x.split(self.expand_chs // 2, dim=1)
x1 = self.blocks(x1)
out = self.conv_transition(torch.cat([x1, x2], dim=1))
return out
| x = self.conv_down(x)
x = self.conv_exp(x)
x1, x2 = x.split(self.expand_chs // 2, dim=1)
x1 = self.blocks(x1)
out = self.conv_transition(torch.cat([x1, x2], dim=1))
return out
| 23 | 841 | 101 | 942 | """PyTorch CspNet
A PyTorch implementation of Cross Stage Partial Networks including:
* CSPResNet50
* CSPResNeXt50
* CSPDarkNet53
* and DarkNet53 for good measure
Based on paper `CSPNet: A New Backbone that can Enhance Learning Capability of CNN` - https://arxiv.org/abs/1911.11929
Reference impl via darknet cfg file... | """PyTorch CspNet
A PyTorch implementation of Cross Stage Partial Networks including:
* CSPResNet50
* CSPResNeXt50
* CSPDarkNet53
* and DarkNet53 for good measure
Based on paper `CSPNet: A New Backbone that can Enhance Learning Capability of CNN` - https://arxiv.org/abs/1911.11929
Reference impl via darknet cfg file... |
17,565 | /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/cspnet.py | 715,716 | 616,721 | forward | CspNet | 716,721 | 716,721 | class CspNet(nn.Module):
"""Cross Stage Partial base model.
Paper: `CSPNet: A New Backbone that can Enhance Learning Capability of CNN` - https://arxiv.org/abs/1911.11929
Ref Impl: https://github.com/WongKinYiu/CrossStagePartialNetworks
NOTE: There are differences in the way I handle the 1x1 'expansio... | class CspNet(nn.Module):
"""Cross Stage Partial base model.
Paper: `CSPNet: A New Backbone that can Enhance Learning Capability of CNN` - https://arxiv.org/abs/1911.11929
Ref Impl: https://github.com/WongKinYiu/CrossStagePartialNetworks
NOTE: There are differences in the way I handle the 1x1 'expansio... | x = self.forward_features(x)
x = self.forward_head(x)
return x
| x = self.forward_features(x)
x = self.forward_head(x)
return x
| 8 | 1,240 | 35 | 1,275 | """PyTorch CspNet
A PyTorch implementation of Cross Stage Partial Networks including:
* CSPResNet50
* CSPResNeXt50
* CSPDarkNet53
* and DarkNet53 for good measure
Based on paper `CSPNet: A New Backbone that can Enhance Learning Capability of CNN` - https://arxiv.org/abs/1911.11929
Reference impl via darknet cfg file... | """PyTorch CspNet
A PyTorch implementation of Cross Stage Partial Networks including:
* CSPResNet50
* CSPResNeXt50
* CSPDarkNet53
* and DarkNet53 for good measure
Based on paper `CSPNet: A New Backbone that can Enhance Learning Capability of CNN` - https://arxiv.org/abs/1911.11929
Reference impl via darknet cfg file... |
17,566 | /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/cspnet.py | 217,218 | 186,226 | forward | DarkBlock | 218,226 | 218,226 | class DarkBlock(nn.Module):
""" DarkNet Block
"""
def __init__(
self,
in_chs,
out_chs,
dilation=1,
bottle_ratio=0.5,
groups=1,
act_layer=nn.ReLU,
norm_layer=nn.BatchNorm2d,
attn_layer=None,
... | class DarkBlock(nn.Module):
""" DarkNet Block
"""
def __init__(
self,
in_chs,
out_chs,
dilation=1,
bottle_ratio=0.5,
groups=1,
act_layer=nn.ReLU,
norm_layer=nn.BatchNorm2d,
attn_layer=None,
... | shortcut = x
x = self.conv1(x)
x = self.attn(x)
x = self.conv2(x)
x = self.drop_path(x) + shortcut
return x
| shortcut = x
x = self.conv1(x)
x = self.attn(x)
x = self.conv2(x)
x = self.drop_path(x) + shortcut
return x
| 19 | 381 | 66 | 447 | """PyTorch CspNet
A PyTorch implementation of Cross Stage Partial Networks including:
* CSPResNet50
* CSPResNeXt50
* CSPDarkNet53
* and DarkNet53 for good measure
Based on paper `CSPNet: A New Backbone that can Enhance Learning Capability of CNN` - https://arxiv.org/abs/1911.11929
Reference impl via darknet cfg file... | """PyTorch CspNet
A PyTorch implementation of Cross Stage Partial Networks including:
* CSPResNet50
* CSPResNeXt50
* CSPDarkNet53
* and DarkNet53 for good measure
Based on paper `CSPNet: A New Backbone that can Enhance Learning Capability of CNN` - https://arxiv.org/abs/1911.11929
Reference impl via darknet cfg file... |
17,567 | /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/cspnet.py | 467,468 | 418,473 | forward | DarkStage | 468,473 | 468,473 | class DarkStage(nn.Module):
"""DarkNet stage."""
def __init__(
self,
in_chs,
out_chs,
stride,
dilation,
depth,
block_ratio=1.,
bottle_ratio=1.,
groups=1,
first_dilation=None,
avg_... | class DarkStage(nn.Module):
"""DarkNet stage."""
def __init__(
self,
in_chs,
out_chs,
stride,
dilation,
depth,
block_ratio=1.,
bottle_ratio=1.,
groups=1,
first_dilation=None,
avg_... | x = self.conv_down(x)
x = self.blocks(x)
return x
| x = self.conv_down(x)
x = self.blocks(x)
return x
| 8 | 577 | 33 | 610 | """PyTorch CspNet
A PyTorch implementation of Cross Stage Partial Networks including:
* CSPResNet50
* CSPResNeXt50
* CSPDarkNet53
* and DarkNet53 for good measure
Based on paper `CSPNet: A New Backbone that can Enhance Learning Capability of CNN` - https://arxiv.org/abs/1911.11929
Reference impl via darknet cfg file... | """PyTorch CspNet
A PyTorch implementation of Cross Stage Partial Networks including:
* CSPResNet50
* CSPResNeXt50
* CSPDarkNet53
* and DarkNet53 for good measure
Based on paper `CSPNet: A New Backbone that can Enhance Learning Capability of CNN` - https://arxiv.org/abs/1911.11929
Reference impl via darknet cfg file... |
17,568 | /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/cspnet.py | 257,258 | 226,266 | forward | EdgeBlock | 258,266 | 258,266 | class EdgeBlock(nn.Module):
""" EdgeResidual / Fused-MBConv / MobileNetV1-like 3x3 + 1x1 block (w/ activated output)
"""
def __init__(
self,
in_chs,
out_chs,
dilation=1,
bottle_ratio=0.5,
groups=1,
act_layer=nn.ReLU,
... | class EdgeBlock(nn.Module):
""" EdgeResidual / Fused-MBConv / MobileNetV1-like 3x3 + 1x1 block (w/ activated output)
"""
def __init__(
self,
in_chs,
out_chs,
dilation=1,
bottle_ratio=0.5,
groups=1,
act_layer=nn.ReLU,
... | shortcut = x
x = self.conv1(x)
x = self.attn(x)
x = self.conv2(x)
x = self.drop_path(x) + shortcut
return x
| shortcut = x
x = self.conv1(x)
x = self.attn(x)
x = self.conv2(x)
x = self.drop_path(x) + shortcut
return x
| 19 | 409 | 66 | 475 | """PyTorch CspNet
A PyTorch implementation of Cross Stage Partial Networks including:
* CSPResNet50
* CSPResNeXt50
* CSPDarkNet53
* and DarkNet53 for good measure
Based on paper `CSPNet: A New Backbone that can Enhance Learning Capability of CNN` - https://arxiv.org/abs/1911.11929
Reference impl via darknet cfg file... | """PyTorch CspNet
A PyTorch implementation of Cross Stage Partial Networks including:
* CSPResNet50
* CSPResNeXt50
* CSPDarkNet53
* and DarkNet53 for good measure
Based on paper `CSPNet: A New Backbone that can Enhance Learning Capability of CNN` - https://arxiv.org/abs/1911.11929
Reference impl via darknet cfg file... |
17,570 | /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/davit.py | 122,123 | 111,137 | forward | ChannelAttention | 123,137 | 123,137 | class ChannelAttention(nn.Module):
def __init__(self, dim, num_heads=8, qkv_bias=False):
super().__init__()
self.num_heads = num_heads
head_dim = dim // num_heads
self.scale = head_dim ** -0.5
self.qkv = nn.Linear(dim, dim * 3, bias=qkv_bias)
self.proj = nn.Linear(d... | class ChannelAttention(nn.Module):
def __init__(self, dim, num_heads=8, qkv_bias=False):
super().__init__()
self.num_heads = num_heads
head_dim = dim // num_heads
self.scale = head_dim ** -0.5
self.qkv = nn.Linear(dim, dim * 3, bias=qkv_bias)
self.proj = nn.Linear(d... | B, N, C = x.shape
qkv = self.qkv(x).reshape(B, N, 3, self.num_heads, C // self.num_heads).permute(2, 0, 3, 1, 4)
q, k, v = qkv.unbind(0)
k = k * self.scale
attention = k.transpose(-1, -2) @ v
attention = attention.softmax(dim=-1)
x = (attention @ q.transpose(-1,... | B, N, C = x.shape
qkv = self.qkv(x).reshape(B, N, 3, self.num_heads, C // self.num_heads).permute(2, 0, 3, 1, 4)
q, k, v = qkv.unbind(0)
k = k * self.scale
attention = k.transpose(-1, -2) @ v
attention = attention.softmax(dim=-1)
x = (attention @ q.transpose(-1,... | 55 | 152 | 205 | 357 | """ DaViT: Dual Attention Vision Transformers
As described in https://arxiv.org/abs/2204.03645
Input size invariant transformer architecture that combines channel and spacial
attention in each block. The attention mechanisms used are linear in complexity.
DaViT model defs and weights adapted from https://github.com/... | """ DaViT: Dual Attention Vision Transformers
As described in https://arxiv.org/abs/2204.03645
Input size invariant transformer architecture that combines channel and spacial
attention in each block. The attention mechanisms used are linear in complexity.
DaViT model defs and weights adapted from https://github.com/... |
17,571 | /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/davit.py | 173,174 | 137,192 | forward | ChannelBlock | 174,192 | 174,192 | class ChannelBlock(nn.Module):
def __init__(
self,
dim,
num_heads,
mlp_ratio=4.,
qkv_bias=False,
drop_path=0.,
act_layer=nn.GELU,
norm_layer=nn.LayerNorm,
ffn=True,
cpe_act=False,
):
... | class ChannelBlock(nn.Module):
def __init__(
self,
dim,
num_heads,
mlp_ratio=4.,
qkv_bias=False,
drop_path=0.,
act_layer=nn.GELU,
norm_layer=nn.LayerNorm,
ffn=True,
cpe_act=False,
):
... | B, C, H, W = x.shape
x = self.cpe1(x).flatten(2).transpose(1, 2)
cur = self.norm1(x)
cur = self.attn(cur)
x = x + self.drop_path1(cur)
x = self.cpe2(x.transpose(1, 2).view(B, C, H, W))
if self.mlp is not None:
x = x.flatten(2).transpose(1, 2)
... | B, C, H, W = x.shape
x = self.cpe1(x).flatten(2).transpose(1, 2)
cur = self.norm1(x)
cur = self.attn(cur)
x = x + self.drop_path1(cur)
x = self.cpe2(x.transpose(1, 2).view(B, C, H, W))
if self.mlp is not None:
x = x.flatten(2).transpose(1, 2)
... | 51 | 405 | 213 | 618 | """ DaViT: Dual Attention Vision Transformers
As described in https://arxiv.org/abs/2204.03645
Input size invariant transformer architecture that combines channel and spacial
attention in each block. The attention mechanisms used are linear in complexity.
DaViT model defs and weights adapted from https://github.com/... | """ DaViT: Dual Attention Vision Transformers
As described in https://arxiv.org/abs/2204.03645
Input size invariant transformer architecture that combines channel and spacial
attention in each block. The attention mechanisms used are linear in complexity.
DaViT model defs and weights adapted from https://github.com/... |
17,572 | /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/davit.py | 40,41 | 33,46 | forward | ConvPosEnc | 41,46 | 41,46 | class ConvPosEnc(nn.Module):
def __init__(self, dim: int, k: int = 3, act: bool = False):
super(ConvPosEnc, self).__init__()
self.proj = nn.Conv2d(dim, dim, k, 1, k // 2, groups=dim)
self.act = nn.GELU() if act else nn.Identity()
def forward(self, x: Tensor):
FILL_FUNC_BODY | class ConvPosEnc(nn.Module):
def __init__(self, dim: int, k: int = 3, act: bool = False):
super(ConvPosEnc, self).__init__()
self.proj = nn.Conv2d(dim, dim, k, 1, k // 2, groups=dim)
self.act = nn.GELU() if act else nn.Identity()
def forward(self, x: Tensor):
FILL_FUNC_BODY | feat = self.proj(x)
x = x + self.act(feat)
return x
| feat = self.proj(x)
x = x + self.act(feat)
return x
| 10 | 127 | 34 | 161 | """ DaViT: Dual Attention Vision Transformers
As described in https://arxiv.org/abs/2204.03645
Input size invariant transformer architecture that combines channel and spacial
attention in each block. The attention mechanisms used are linear in complexity.
DaViT model defs and weights adapted from https://github.com/... | """ DaViT: Dual Attention Vision Transformers
As described in https://arxiv.org/abs/2204.03645
Input size invariant transformer architecture that combines channel and spacial
attention in each block. The attention mechanisms used are linear in complexity.
DaViT model defs and weights adapted from https://github.com/... |
17,573 | /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/davit.py | 574,575 | 443,592 | forward_features | DaVit | 575,583 | 575,583 | class DaVit(nn.Module):
r""" DaViT
A PyTorch implementation of `DaViT: Dual Attention Vision Transformers` - https://arxiv.org/abs/2204.03645
Supports arbitrary input sizes and pyramid feature extraction
Args:
in_chans (int): Number of input image channels. Default: 3
n... | class DaVit(nn.Module):
r""" DaViT
A PyTorch implementation of `DaViT: Dual Attention Vision Transformers` - https://arxiv.org/abs/2204.03645
Supports arbitrary input sizes and pyramid feature extraction
Args:
in_chans (int): Number of input image channels. Default: 3
n... | x = self.stem(x)
if self.grad_checkpointing and not torch.jit.is_scripting():
x = checkpoint_seq(self.stages, x)
else:
x = self.stages(x)
x = self.norm_pre(x)
return x
| x = self.stem(x)
if self.grad_checkpointing and not torch.jit.is_scripting():
x = checkpoint_seq(self.stages, x)
else:
x = self.stages(x)
x = self.norm_pre(x)
return x
| 21 | 1,797 | 90 | 1,887 | """ DaViT: Dual Attention Vision Transformers
As described in https://arxiv.org/abs/2204.03645
Input size invariant transformer architecture that combines channel and spacial
attention in each block. The attention mechanisms used are linear in complexity.
DaViT model defs and weights adapted from https://github.com/... | """ DaViT: Dual Attention Vision Transformers
As described in https://arxiv.org/abs/2204.03645
Input size invariant transformer architecture that combines channel and spacial
attention in each block. The attention mechanisms used are linear in complexity.
DaViT model defs and weights adapted from https://github.com/... |
17,574 | /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/davit.py | 434,435 | 365,443 | forward | DaVitStage | 435,443 | 435,443 | class DaVitStage(nn.Module):
def __init__(
self,
in_chs,
out_chs,
depth=1,
downsample=True,
attn_types=('spatial', 'channel'),
num_heads=3,
window_size=7,
mlp_ratio=4,
qkv_bias=True,
d... | class DaVitStage(nn.Module):
def __init__(
self,
in_chs,
out_chs,
depth=1,
downsample=True,
attn_types=('spatial', 'channel'),
num_heads=3,
window_size=7,
mlp_ratio=4,
qkv_bias=True,
d... | 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 | 708 | 77 | 785 | """ DaViT: Dual Attention Vision Transformers
As described in https://arxiv.org/abs/2204.03645
Input size invariant transformer architecture that combines channel and spacial
attention in each block. The attention mechanisms used are linear in complexity.
DaViT model defs and weights adapted from https://github.com/... | """ DaViT: Dual Attention Vision Transformers
As described in https://arxiv.org/abs/2204.03645
Input size invariant transformer architecture that combines channel and spacial
attention in each block. The attention mechanisms used are linear in complexity.
DaViT model defs and weights adapted from https://github.com/... |
17,575 | /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/davit.py | 102,103 | 82,111 | forward | Downsample | 103,111 | 103,111 | class Downsample(nn.Module):
def __init__(
self,
in_chs,
out_chs,
norm_layer=LayerNorm2d,
):
super().__init__()
self.in_chs = in_chs
self.out_chs = out_chs
self.norm = norm_layer(in_chs)
self.conv = nn.Conv2d(
i... | class Downsample(nn.Module):
def __init__(
self,
in_chs,
out_chs,
norm_layer=LayerNorm2d,
):
super().__init__()
self.in_chs = in_chs
self.out_chs = out_chs
self.norm = norm_layer(in_chs)
self.conv = nn.Conv2d(
i... | B, C, H, W = x.shape
x = self.norm(x)
x = F.pad(x, (0, (2 - W % 2) % 2))
x = F.pad(x, (0, 0, 0, (2 - H % 2) % 2))
x = self.conv(x)
return x
| B, C, H, W = x.shape
x = self.norm(x)
x = F.pad(x, (0, (2 - W % 2) % 2))
x = F.pad(x, (0, 0, 0, (2 - H % 2) % 2))
x = self.conv(x)
return x
| 38 | 173 | 103 | 276 | """ DaViT: Dual Attention Vision Transformers
As described in https://arxiv.org/abs/2204.03645
Input size invariant transformer architecture that combines channel and spacial
attention in each block. The attention mechanisms used are linear in complexity.
DaViT model defs and weights adapted from https://github.com/... | """ DaViT: Dual Attention Vision Transformers
As described in https://arxiv.org/abs/2204.03645
Input size invariant transformer architecture that combines channel and spacial
attention in each block. The attention mechanisms used are linear in complexity.
DaViT model defs and weights adapted from https://github.com/... |
17,576 | /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/davit.py | 325,326 | 267,365 | forward | SpatialBlock | 326,365 | 326,365 | class SpatialBlock(nn.Module):
r""" Windows Block.
Args:
dim (int): Number of input channels.
num_heads (int): Number of attention heads.
window_size (int): Window size.
mlp_ratio (float): Ratio of mlp hidden dim to embedding dim.
qkv_bias (bool, optional): If True, add a... | class SpatialBlock(nn.Module):
r""" Windows Block.
Args:
dim (int): Number of input channels.
num_heads (int): Number of attention heads.
window_size (int): Window size.
mlp_ratio (float): Ratio of mlp hidden dim to embedding dim.
qkv_bias (bool, optional): If True, add a... | B, C, H, W = x.shape
shortcut = self.cpe1(x).flatten(2).transpose(1, 2)
x = self.norm1(shortcut)
x = x.view(B, H, W, C)
pad_l = pad_t = 0
pad_r = (self.window_size[1] - W % self.window_size[1]) % self.window_size[1]
pad_b = (self.window_size[0] - H % self.windo... | B, C, H, W = x.shape
shortcut = self.cpe1(x).flatten(2).transpose(1, 2)
x = self.norm1(shortcut)
x = x.view(B, H, W, C)
pad_l = pad_t = 0
pad_r = (self.window_size[1] - W % self.window_size[1]) % self.window_size[1]
pad_b = (self.window_size[0] - H % self.windo... | 145 | 670 | 568 | 1,238 | """ DaViT: Dual Attention Vision Transformers
As described in https://arxiv.org/abs/2204.03645
Input size invariant transformer architecture that combines channel and spacial
attention in each block. The attention mechanisms used are linear in complexity.
DaViT model defs and weights adapted from https://github.com/... | """ DaViT: Dual Attention Vision Transformers
As described in https://arxiv.org/abs/2204.03645
Input size invariant transformer architecture that combines channel and spacial
attention in each block. The attention mechanisms used are linear in complexity.
DaViT model defs and weights adapted from https://github.com/... |
17,577 | /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/davit.py | 73,74 | 46,82 | forward | Stem | 74,82 | 74,82 | class Stem(nn.Module):
""" Size-agnostic implementation of 2D image to patch embedding,
allowing input size to be adjusted during model forward operation
"""
def __init__(
self,
in_chs=3,
out_chs=96,
stride=4,
norm_layer=LayerNorm2d,
)... | class Stem(nn.Module):
""" Size-agnostic implementation of 2D image to patch embedding,
allowing input size to be adjusted during model forward operation
"""
def __init__(
self,
in_chs=3,
out_chs=96,
stride=4,
norm_layer=LayerNorm2d,
)... | B, C, H, W = x.shape
x = F.pad(x, (0, (self.stride[1] - W % self.stride[1]) % self.stride[1]))
x = F.pad(x, (0, 0, 0, (self.stride[0] - H % self.stride[0]) % self.stride[0]))
x = self.conv(x)
x = self.norm(x)
return x
| B, C, H, W = x.shape
x = F.pad(x, (0, (self.stride[1] - W % self.stride[1]) % self.stride[1]))
x = F.pad(x, (0, 0, 0, (self.stride[0] - H % self.stride[0]) % self.stride[0]))
x = self.conv(x)
x = self.norm(x)
return x
| 38 | 260 | 131 | 391 | """ DaViT: Dual Attention Vision Transformers
As described in https://arxiv.org/abs/2204.03645
Input size invariant transformer architecture that combines channel and spacial
attention in each block. The attention mechanisms used are linear in complexity.
DaViT model defs and weights adapted from https://github.com/... | """ DaViT: Dual Attention Vision Transformers
As described in https://arxiv.org/abs/2204.03645
Input size invariant transformer architecture that combines channel and spacial
attention in each block. The attention mechanisms used are linear in complexity.
DaViT model defs and weights adapted from https://github.com/... |
17,578 | /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/davit.py | 248,249 | 223,267 | forward | WindowAttention | 249,267 | 249,267 | class WindowAttention(nn.Module):
r""" Window based multi-head self attention (W-MSA) module with relative position bias.
It supports both of shifted and non-shifted window.
Args:
dim (int): Number of input channels.
window_size (tuple[int]): The height and width of the window.
num_h... | class WindowAttention(nn.Module):
r""" Window based multi-head self attention (W-MSA) module with relative position bias.
It supports both of shifted and non-shifted window.
Args:
dim (int): Number of input channels.
window_size (tuple[int]): The height and width of the window.
num_h... | B_, N, C = x.shape
qkv = self.qkv(x).reshape(B_, N, 3, self.num_heads, C // self.num_heads).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)
else:
q = q * self.scale
attn = (q @ k.... | B_, N, C = x.shape
qkv = self.qkv(x).reshape(B_, N, 3, self.num_heads, C // self.num_heads).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)
else:
q = q * self.scale
attn = (q @ k.... | 61 | 349 | 230 | 579 | """ DaViT: Dual Attention Vision Transformers
As described in https://arxiv.org/abs/2204.03645
Input size invariant transformer architecture that combines channel and spacial
attention in each block. The attention mechanisms used are linear in complexity.
DaViT model defs and weights adapted from https://github.com/... | """ DaViT: Dual Attention Vision Transformers
As described in https://arxiv.org/abs/2204.03645
Input size invariant transformer architecture that combines channel and spacial
attention in each block. The attention mechanisms used are linear in complexity.
DaViT model defs and weights adapted from https://github.com/... |
17,580 | /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/deit.py | 54,56 | 29,121 | group_matcher | VisionTransformerDistilled | 56,63 | 56,63 | class VisionTransformerDistilled(VisionTransformer):
""" Vision Transformer w/ Distillation Token and Head
Distillation token & head support for `DeiT: Data-efficient Image Transformers`
- https://arxiv.org/abs/2012.12877
"""
def __init__(self, *args, **kwargs):
weight_init = kwargs.po... | class VisionTransformerDistilled(VisionTransformer):
""" Vision Transformer w/ Distillation Token and Head
Distillation token & head support for `DeiT: Data-efficient Image Transformers`
- https://arxiv.org/abs/2012.12877
"""
def __init__(self, *args, **kwargs):
weight_init = kwargs.po... | return dict(
stem=r'^cls_token|pos_embed|patch_embed|dist_token',
blocks=[
(r'^blocks\.(\d+)', None),
(r'^norm', (99999,))] # final norm w/ last block
)
| return dict(
stem=r'^cls_token|pos_embed|patch_embed|dist_token',
blocks=[
(r'^blocks\.(\d+)', None),
(r'^norm', (99999,))] # final norm w/ last block
)
| 15 | 1,107 | 85 | 1,192 | """ DeiT - Data-efficient Image Transformers
DeiT model defs and weights from https://github.com/facebookresearch/deit, original copyright below
paper: `DeiT: Data-efficient Image Transformers` - https://arxiv.org/abs/2012.12877
paper: `DeiT III: Revenge of the ViT` - https://arxiv.org/abs/2204.07118
Modifications ... | """ DeiT - Data-efficient Image Transformers
DeiT model defs and weights from https://github.com/facebookresearch/deit, original copyright below
paper: `DeiT: Data-efficient Image Transformers` - https://arxiv.org/abs/2012.12877
paper: `DeiT III: Revenge of the ViT` - https://arxiv.org/abs/2204.07118
Modifications ... |
17,582 | /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/dla.py | 335,337 | 261,386 | group_matcher | DLA | 337,348 | 337,348 | class DLA(nn.Module):
def __init__(
self,
levels,
channels,
output_stride=32,
num_classes=1000,
in_chans=3,
global_pool='avg',
cardinality=1,
base_width=64,
block=DlaBottle2neck,
short... | class DLA(nn.Module):
def __init__(
self,
levels,
channels,
output_stride=32,
num_classes=1000,
in_chans=3,
global_pool='avg',
cardinality=1,
base_width=64,
block=DlaBottle2neck,
short... | matcher = dict(
stem=r'^base_layer',
blocks=r'^level(\d+)' if coarse else [
# an unusual arch, this achieves somewhat more granularity without getting super messy
(r'^level(\d+)\.tree(\d+)', None),
(r'^level(\d+)\.root', (2,)),
... | matcher = dict(
stem=r'^base_layer',
blocks=r'^level(\d+)' if coarse else [
# an unusual arch, this achieves somewhat more granularity without getting super messy
(r'^level(\d+)\.tree(\d+)', None),
(r'^level(\d+)\.root', (2,)),
... | 32 | 1,502 | 131 | 1,633 | """ Deep Layer Aggregation and DLA w/ Res2Net
DLA original adapted from Official Pytorch impl at: https://github.com/ucbdrive/dla
DLA Paper: `Deep Layer Aggregation` - https://arxiv.org/abs/1707.06484
Res2Net additions from: https://github.com/gasvn/Res2Net/
Res2Net Paper: `Res2Net: A New Multi-scale Backbone Architec... | """ Deep Layer Aggregation and DLA w/ Res2Net
DLA original adapted from Official Pytorch impl at: https://github.com/ucbdrive/dla
DLA Paper: `Deep Layer Aggregation` - https://arxiv.org/abs/1707.06484
Res2Net additions from: https://github.com/gasvn/Res2Net/
Res2Net Paper: `Res2Net: A New Multi-scale Backbone Architec... |
17,583 | /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/dla.py | 39,40 | 23,56 | forward | DlaBasic | 40,56 | 40,56 | class DlaBasic(nn.Module):
"""DLA Basic"""
def __init__(self, inplanes, planes, stride=1, dilation=1, **_):
super(DlaBasic, self).__init__()
self.conv1 = nn.Conv2d(
inplanes, planes, kernel_size=3,
stride=stride, padding=dilation, bias=False, dilation=dilation)
s... | class DlaBasic(nn.Module):
"""DLA Basic"""
def __init__(self, inplanes, planes, stride=1, dilation=1, **_):
super(DlaBasic, self).__init__()
self.conv1 = nn.Conv2d(
inplanes, planes, kernel_size=3,
stride=stride, padding=dilation, bias=False, dilation=dilation)
s... | if shortcut is None:
shortcut = x
out = self.conv1(x)
out = self.bn1(out)
out = self.relu(out)
out = self.conv2(out)
out = self.bn2(out)
out += shortcut
out = self.relu(out)
return out
| if shortcut is None:
shortcut = x
out = self.conv1(x)
out = self.bn1(out)
out = self.relu(out)
out = self.conv2(out)
out = self.bn2(out)
out += shortcut
out = self.relu(out)
return out
| 30 | 272 | 109 | 381 | """ Deep Layer Aggregation and DLA w/ Res2Net
DLA original adapted from Official Pytorch impl at: https://github.com/ucbdrive/dla
DLA Paper: `Deep Layer Aggregation` - https://arxiv.org/abs/1707.06484
Res2Net additions from: https://github.com/gasvn/Res2Net/
Res2Net Paper: `Res2Net: A New Multi-scale Backbone Architec... | """ Deep Layer Aggregation and DLA w/ Res2Net
DLA original adapted from Official Pytorch impl at: https://github.com/ucbdrive/dla
DLA Paper: `Deep Layer Aggregation` - https://arxiv.org/abs/1707.06484
Res2Net additions from: https://github.com/gasvn/Res2Net/
Res2Net Paper: `Res2Net: A New Multi-scale Backbone Architec... |
17,584 | /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/dla.py | 130,131 | 97,166 | forward | DlaBottle2neck | 131,166 | 131,166 | class DlaBottle2neck(nn.Module):
""" Res2Net/Res2NeXT DLA Bottleneck
Adapted from https://github.com/gasvn/Res2Net/blob/master/dla.py
"""
expansion = 2
def __init__(self, inplanes, outplanes, stride=1, dilation=1, scale=4, cardinality=8, base_width=4):
super(DlaBottle2neck, self).__init__()... | class DlaBottle2neck(nn.Module):
""" Res2Net/Res2NeXT DLA Bottleneck
Adapted from https://github.com/gasvn/Res2Net/blob/master/dla.py
"""
expansion = 2
def __init__(self, inplanes, outplanes, stride=1, dilation=1, scale=4, cardinality=8, base_width=4):
super(DlaBottle2neck, self).__init__()... | if shortcut is None:
shortcut = x
out = self.conv1(x)
out = self.bn1(out)
out = self.relu(out)
spx = torch.split(out, self.width, 1)
spo = []
sp = spx[0] # redundant, for torchscript
for i, (conv, bn) in enumerate(zip(self.convs, self.bns)):... | if shortcut is None:
shortcut = x
out = self.conv1(x)
out = self.bn1(out)
out = self.relu(out)
spx = torch.split(out, self.width, 1)
spo = []
sp = spx[0] # redundant, for torchscript
for i, (conv, bn) in enumerate(zip(self.convs, self.bns)):... | 101 | 583 | 352 | 935 | """ Deep Layer Aggregation and DLA w/ Res2Net
DLA original adapted from Official Pytorch impl at: https://github.com/ucbdrive/dla
DLA Paper: `Deep Layer Aggregation` - https://arxiv.org/abs/1707.06484
Res2Net additions from: https://github.com/gasvn/Res2Net/
Res2Net Paper: `Res2Net: A New Multi-scale Backbone Architec... | """ Deep Layer Aggregation and DLA w/ Res2Net
DLA original adapted from Official Pytorch impl at: https://github.com/ucbdrive/dla
DLA Paper: `Deep Layer Aggregation` - https://arxiv.org/abs/1707.06484
Res2Net additions from: https://github.com/gasvn/Res2Net/
Res2Net Paper: `Res2Net: A New Multi-scale Backbone Architec... |
17,585 | /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/dla.py | 76,77 | 56,97 | forward | DlaBottleneck | 77,97 | 77,97 | class DlaBottleneck(nn.Module):
"""DLA/DLA-X Bottleneck"""
expansion = 2
def __init__(self, inplanes, outplanes, stride=1, dilation=1, cardinality=1, base_width=64):
super(DlaBottleneck, self).__init__()
self.stride = stride
mid_planes = int(math.floor(outplanes * (base_width / 64))... | class DlaBottleneck(nn.Module):
"""DLA/DLA-X Bottleneck"""
expansion = 2
def __init__(self, inplanes, outplanes, stride=1, dilation=1, cardinality=1, base_width=64):
super(DlaBottleneck, self).__init__()
self.stride = stride
mid_planes = int(math.floor(outplanes * (base_width / 64))... | if shortcut is None:
shortcut = x
out = self.conv1(x)
out = self.bn1(out)
out = self.relu(out)
out = self.conv2(out)
out = self.bn2(out)
out = self.relu(out)
out = self.conv3(out)
out = self.bn3(out)
out += shortcut
... | if shortcut is None:
shortcut = x
out = self.conv1(x)
out = self.bn1(out)
out = self.relu(out)
out = self.conv2(out)
out = self.bn2(out)
out = self.relu(out)
out = self.conv3(out)
out = self.bn3(out)
out += shortcut
... | 39 | 404 | 147 | 551 | """ Deep Layer Aggregation and DLA w/ Res2Net
DLA original adapted from Official Pytorch impl at: https://github.com/ucbdrive/dla
DLA Paper: `Deep Layer Aggregation` - https://arxiv.org/abs/1707.06484
Res2Net additions from: https://github.com/gasvn/Res2Net/
Res2Net Paper: `Res2Net: A New Multi-scale Backbone Architec... | """ Deep Layer Aggregation and DLA w/ Res2Net
DLA original adapted from Official Pytorch impl at: https://github.com/ucbdrive/dla
DLA Paper: `Deep Layer Aggregation` - https://arxiv.org/abs/1707.06484
Res2Net additions from: https://github.com/gasvn/Res2Net/
Res2Net Paper: `Res2Net: A New Multi-scale Backbone Architec... |
17,586 | /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/dla.py | 175,176 | 166,185 | forward | DlaRoot | 176,185 | 176,185 | class DlaRoot(nn.Module):
def __init__(self, in_channels, out_channels, kernel_size, shortcut):
super(DlaRoot, self).__init__()
self.conv = nn.Conv2d(
in_channels, out_channels, 1, stride=1, bias=False, padding=(kernel_size - 1) // 2)
self.bn = nn.BatchNorm2d(out_channels)
... | class DlaRoot(nn.Module):
def __init__(self, in_channels, out_channels, kernel_size, shortcut):
super(DlaRoot, self).__init__()
self.conv = nn.Conv2d(
in_channels, out_channels, 1, stride=1, bias=False, padding=(kernel_size - 1) // 2)
self.bn = nn.BatchNorm2d(out_channels)
... | x = self.conv(torch.cat(x_children, 1))
x = self.bn(x)
if self.shortcut:
x += x_children[0]
x = self.relu(x)
return x
| x = self.conv(torch.cat(x_children, 1))
x = self.bn(x)
if self.shortcut:
x += x_children[0]
x = self.relu(x)
return x
| 17 | 177 | 75 | 252 | """ Deep Layer Aggregation and DLA w/ Res2Net
DLA original adapted from Official Pytorch impl at: https://github.com/ucbdrive/dla
DLA Paper: `Deep Layer Aggregation` - https://arxiv.org/abs/1707.06484
Res2Net additions from: https://github.com/gasvn/Res2Net/
Res2Net Paper: `Res2Net: A New Multi-scale Backbone Architec... | """ Deep Layer Aggregation and DLA w/ Res2Net
DLA original adapted from Official Pytorch impl at: https://github.com/ucbdrive/dla
DLA Paper: `Deep Layer Aggregation` - https://arxiv.org/abs/1707.06484
Res2Net additions from: https://github.com/gasvn/Res2Net/
Res2Net Paper: `Res2Net: A New Multi-scale Backbone Architec... |
17,587 | /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/dla.py | 244,245 | 185,261 | forward | DlaTree | 245,261 | 245,261 | class DlaTree(nn.Module):
def __init__(
self,
levels,
block,
in_channels,
out_channels,
stride=1,
dilation=1,
cardinality=1,
base_width=64,
level_root=False,
root_dim=0,
ro... | class DlaTree(nn.Module):
def __init__(
self,
levels,
block,
in_channels,
out_channels,
stride=1,
dilation=1,
cardinality=1,
base_width=64,
level_root=False,
root_dim=0,
ro... | if children is None:
children = []
bottom = self.downsample(x)
shortcut = self.project(bottom)
if self.level_root:
children.append(bottom)
x1 = self.tree1(x, shortcut)
if self.root is not None: # levels == 1
x2 = self.tree2(x1)
... | if children is None:
children = []
bottom = self.downsample(x)
shortcut = self.project(bottom)
if self.level_root:
children.append(bottom)
x1 = self.tree1(x, shortcut)
if self.root is not None: # levels == 1
x2 = self.tree2(x1)
... | 47 | 696 | 160 | 856 | """ Deep Layer Aggregation and DLA w/ Res2Net
DLA original adapted from Official Pytorch impl at: https://github.com/ucbdrive/dla
DLA Paper: `Deep Layer Aggregation` - https://arxiv.org/abs/1707.06484
Res2Net additions from: https://github.com/gasvn/Res2Net/
Res2Net Paper: `Res2Net: A New Multi-scale Backbone Architec... | """ Deep Layer Aggregation and DLA w/ Res2Net
DLA original adapted from Official Pytorch impl at: https://github.com/ucbdrive/dla
DLA Paper: `Deep Layer Aggregation` - https://arxiv.org/abs/1707.06484
Res2Net additions from: https://github.com/gasvn/Res2Net/
Res2Net Paper: `Res2Net: A New Multi-scale Backbone Architec... |
17,589 | /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/edgenext.py | 91,92 | 67,108 | forward | ConvBlock | 92,108 | 92,108 | class ConvBlock(nn.Module):
def __init__(
self,
dim,
dim_out=None,
kernel_size=7,
stride=1,
conv_bias=True,
expand_ratio=4,
ls_init_value=1e-6,
norm_layer=partial(nn.LayerNorm, eps=1e-6),
act_laye... | class ConvBlock(nn.Module):
def __init__(
self,
dim,
dim_out=None,
kernel_size=7,
stride=1,
conv_bias=True,
expand_ratio=4,
ls_init_value=1e-6,
norm_layer=partial(nn.LayerNorm, eps=1e-6),
act_laye... | shortcut = x
x = self.conv_dw(x)
if self.shortcut_after_dw:
shortcut = x
x = x.permute(0, 2, 3, 1) # (N, C, H, W) -> (N, H, W, C)
x = self.norm(x)
x = self.mlp(x)
if self.gamma is not None:
x = self.gamma * x
x = x.permute(0, 3, 1... | shortcut = x
x = self.conv_dw(x)
if self.shortcut_after_dw:
shortcut = x
x = x.permute(0, 2, 3, 1) # (N, C, H, W) -> (N, H, W, C)
x = self.norm(x)
x = self.mlp(x)
if self.gamma is not None:
x = self.gamma * x
x = x.permute(0, 3, 1... | 66 | 342 | 200 | 542 | """ EdgeNeXt
Paper: `EdgeNeXt: Efficiently Amalgamated CNN-Transformer Architecture for Mobile Vision Applications`
- https://arxiv.org/abs/2206.10589
Original code and weights from https://github.com/mmaaz60/EdgeNeXt
Modifications and additions for timm by / Copyright 2022, Ross Wightman
"""
import math
from colle... | """ EdgeNeXt
Paper: `EdgeNeXt: Efficiently Amalgamated CNN-Transformer Architecture for Mobile Vision Applications`
- https://arxiv.org/abs/2206.10589
Original code and weights from https://github.com/mmaaz60/EdgeNeXt
Modifications and additions for timm by / Copyright 2022, Ross Wightman
"""
import math
from colle... |
17,590 | /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/edgenext.py | 126,127 | 108,147 | forward | CrossCovarianceAttn | 127,142 | 127,142 | class CrossCovarianceAttn(nn.Module):
def __init__(
self,
dim,
num_heads=8,
qkv_bias=False,
attn_drop=0.,
proj_drop=0.
):
super().__init__()
self.num_heads = num_heads
self.temperature = nn.Parameter(torch.ones(num_h... | class CrossCovarianceAttn(nn.Module):
def __init__(
self,
dim,
num_heads=8,
qkv_bias=False,
attn_drop=0.,
proj_drop=0.
):
super().__init__()
self.num_heads = num_heads
self.temperature = nn.Parameter(torch.ones(num_h... | B, N, C = x.shape
qkv = self.qkv(x).reshape(B, N, 3, self.num_heads, -1).permute(2, 0, 3, 4, 1)
q, k, v = qkv.unbind(0)
# NOTE, this is NOT spatial attn, q, k, v are B, num_heads, C, L --> C x C attn map
attn = (F.normalize(q, dim=-1) @ F.normalize(k, dim=-1).transpose(-2, -1))... | B, N, C = x.shape
qkv = self.qkv(x).reshape(B, N, 3, self.num_heads, -1).permute(2, 0, 3, 4, 1)
q, k, v = qkv.unbind(0)
# NOTE, this is NOT spatial attn, q, k, v are B, num_heads, C, L --> C x C attn map
attn = (F.normalize(q, dim=-1) @ F.normalize(k, dim=-1).transpose(-2, -1))... | 79 | 251 | 275 | 526 | """ EdgeNeXt
Paper: `EdgeNeXt: Efficiently Amalgamated CNN-Transformer Architecture for Mobile Vision Applications`
- https://arxiv.org/abs/2206.10589
Original code and weights from https://github.com/mmaaz60/EdgeNeXt
Modifications and additions for timm by / Copyright 2022, Ross Wightman
"""
import math
from colle... | """ EdgeNeXt
Paper: `EdgeNeXt: Efficiently Amalgamated CNN-Transformer Architecture for Mobile Vision Applications`
- https://arxiv.org/abs/2206.10589
Original code and weights from https://github.com/mmaaz60/EdgeNeXt
Modifications and additions for timm by / Copyright 2022, Ross Wightman
"""
import math
from colle... |
17,591 | /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/edgenext.py | 430,431 | 300,436 | forward | EdgeNeXt | 431,436 | 431,436 | class EdgeNeXt(nn.Module):
def __init__(
self,
in_chans=3,
num_classes=1000,
global_pool='avg',
dims=(24, 48, 88, 168),
depths=(3, 3, 9, 3),
global_block_counts=(0, 1, 1, 1),
kernel_sizes=(3, 5, 7, 9),
heads=... | class EdgeNeXt(nn.Module):
def __init__(
self,
in_chans=3,
num_classes=1000,
global_pool='avg',
dims=(24, 48, 88, 168),
depths=(3, 3, 9, 3),
global_block_counts=(0, 1, 1, 1),
kernel_sizes=(3, 5, 7, 9),
heads=... | x = self.forward_features(x)
x = self.forward_head(x)
return x
| x = self.forward_features(x)
x = self.forward_head(x)
return x
| 8 | 1,558 | 35 | 1,593 | """ EdgeNeXt
Paper: `EdgeNeXt: Efficiently Amalgamated CNN-Transformer Architecture for Mobile Vision Applications`
- https://arxiv.org/abs/2206.10589
Original code and weights from https://github.com/mmaaz60/EdgeNeXt
Modifications and additions for timm by / Copyright 2022, Ross Wightman
"""
import math
from colle... | """ EdgeNeXt
Paper: `EdgeNeXt: Efficiently Amalgamated CNN-Transformer Architecture for Mobile Vision Applications`
- https://arxiv.org/abs/2206.10589
Original code and weights from https://github.com/mmaaz60/EdgeNeXt
Modifications and additions for timm by / Copyright 2022, Ross Wightman
"""
import math
from colle... |
17,592 | /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/edgenext.py | 291,292 | 223,300 | forward | EdgeNeXtStage | 292,300 | 292,300 | class EdgeNeXtStage(nn.Module):
def __init__(
self,
in_chs,
out_chs,
stride=2,
depth=2,
num_global_blocks=1,
num_heads=4,
scales=2,
kernel_size=7,
expand_ratio=4,
use_pos_emb=False,
... | class EdgeNeXtStage(nn.Module):
def __init__(
self,
in_chs,
out_chs,
stride=2,
depth=2,
num_global_blocks=1,
num_heads=4,
scales=2,
kernel_size=7,
expand_ratio=4,
use_pos_emb=False,
... | 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 | 679 | 77 | 756 | """ EdgeNeXt
Paper: `EdgeNeXt: Efficiently Amalgamated CNN-Transformer Architecture for Mobile Vision Applications`
- https://arxiv.org/abs/2206.10589
Original code and weights from https://github.com/mmaaz60/EdgeNeXt
Modifications and additions for timm by / Copyright 2022, Ross Wightman
"""
import math
from colle... | """ EdgeNeXt
Paper: `EdgeNeXt: Efficiently Amalgamated CNN-Transformer Architecture for Mobile Vision Applications`
- https://arxiv.org/abs/2206.10589
Original code and weights from https://github.com/mmaaz60/EdgeNeXt
Modifications and additions for timm by / Copyright 2022, Ross Wightman
"""
import math
from colle... |
17,593 | /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/edgenext.py | 40,41 | 30,67 | forward | PositionalEncodingFourier | 41,67 | 41,67 | @register_notrace_module # reason: FX can't symbolically trace torch.arange in forward method
class PositionalEncodingFourier(nn.Module):
def __init__(self, hidden_dim=32, dim=768, temperature=10000):
super().__init__()
self.token_projection = nn.Conv2d(hidden_dim * 2, dim, kernel_size=1)
s... | @register_notrace_module # reason: FX can't symbolically trace torch.arange in forward method
class PositionalEncodingFourier(nn.Module):
def __init__(self, hidden_dim=32, dim=768, temperature=10000):
super().__init__()
self.token_projection = nn.Conv2d(hidden_dim * 2, dim, kernel_size=1)
s... | device = self.token_projection.weight.device
dtype = self.token_projection.weight.dtype
inv_mask = ~torch.zeros(shape).to(device=device, dtype=torch.bool)
y_embed = inv_mask.cumsum(1, dtype=torch.float32)
x_embed = inv_mask.cumsum(2, dtype=torch.float32)
eps = 1e-6
... | device = self.token_projection.weight.device
dtype = self.token_projection.weight.dtype
inv_mask = ~torch.zeros(shape).to(device=device, dtype=torch.bool)
y_embed = inv_mask.cumsum(1, dtype=torch.float32)
x_embed = inv_mask.cumsum(2, dtype=torch.float32)
eps = 1e-6
... | 112 | 176 | 505 | 681 | """ EdgeNeXt
Paper: `EdgeNeXt: Efficiently Amalgamated CNN-Transformer Architecture for Mobile Vision Applications`
- https://arxiv.org/abs/2206.10589
Original code and weights from https://github.com/mmaaz60/EdgeNeXt
Modifications and additions for timm by / Copyright 2022, Ross Wightman
"""
import math
from colle... | """ EdgeNeXt
Paper: `EdgeNeXt: Efficiently Amalgamated CNN-Transformer Architecture for Mobile Vision Applications`
- https://arxiv.org/abs/2206.10589
Original code and weights from https://github.com/mmaaz60/EdgeNeXt
Modifications and additions for timm by / Copyright 2022, Ross Wightman
"""
import math
from colle... |
17,594 | /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/edgenext.py | 187,188 | 147,223 | forward | SplitTransposeBlock | 188,223 | 188,223 | class SplitTransposeBlock(nn.Module):
def __init__(
self,
dim,
num_scales=1,
num_heads=8,
expand_ratio=4,
use_pos_emb=True,
conv_bias=True,
qkv_bias=True,
ls_init_value=1e-6,
norm_layer=partial(nn... | class SplitTransposeBlock(nn.Module):
def __init__(
self,
dim,
num_scales=1,
num_heads=8,
expand_ratio=4,
use_pos_emb=True,
conv_bias=True,
qkv_bias=True,
ls_init_value=1e-6,
norm_layer=partial(nn... | shortcut = x
# scales code re-written for torchscript as per my res2net fixes -rw
# NOTE torch.split(x, self.width, 1) causing issues with ONNX export
spx = x.chunk(len(self.convs) + 1, dim=1)
spo = []
sp = spx[0]
for i, conv in enumerate(self.convs):
... | shortcut = x
# scales code re-written for torchscript as per my res2net fixes -rw
# NOTE torch.split(x, self.width, 1) causing issues with ONNX export
spx = x.chunk(len(self.convs) + 1, dim=1)
spo = []
sp = spx[0]
for i, conv in enumerate(self.convs):
... | 151 | 582 | 482 | 1,064 | """ EdgeNeXt
Paper: `EdgeNeXt: Efficiently Amalgamated CNN-Transformer Architecture for Mobile Vision Applications`
- https://arxiv.org/abs/2206.10589
Original code and weights from https://github.com/mmaaz60/EdgeNeXt
Modifications and additions for timm by / Copyright 2022, Ross Wightman
"""
import math
from colle... | """ EdgeNeXt
Paper: `EdgeNeXt: Efficiently Amalgamated CNN-Transformer Architecture for Mobile Vision Applications`
- https://arxiv.org/abs/2206.10589
Original code and weights from https://github.com/mmaaz60/EdgeNeXt
Modifications and additions for timm by / Copyright 2022, Ross Wightman
"""
import math
from colle... |
17,596 | /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/efficientformer.py | 74,76 | 43,104 | train | Attention | 76,80 | 76,80 | class Attention(torch.nn.Module):
attention_bias_cache: Dict[str, torch.Tensor]
def __init__(
self,
dim=384,
key_dim=32,
num_heads=8,
attn_ratio=4,
resolution=7
):
super().__init__()
self.num_heads = num_heads
s... | class Attention(torch.nn.Module):
attention_bias_cache: Dict[str, torch.Tensor]
def __init__(
self,
dim=384,
key_dim=32,
num_heads=8,
attn_ratio=4,
resolution=7
):
super().__init__()
self.num_heads = num_heads
s... | super().train(mode)
if mode and self.attention_bias_cache:
self.attention_bias_cache = {} # clear ab cache
| super().train(mode)
if mode and self.attention_bias_cache:
self.attention_bias_cache = {} # clear ab cache
| 12 | 866 | 44 | 910 | """ EfficientFormer
@article{li2022efficientformer,
title={EfficientFormer: Vision Transformers at MobileNet Speed},
author={Li, Yanyu and Yuan, Geng and Wen, Yang and Hu, Eric and Evangelidis, Georgios and Tulyakov,
Sergey and Wang, Yanzhi and Ren, Jian},
journal={arXiv preprint arXiv:2206.01191},
year={20... | """ EfficientFormer
@article{li2022efficientformer,
title={EfficientFormer: Vision Transformers at MobileNet Speed},
author={Li, Yanyu and Yuan, Geng and Wen, Yang and Hu, Eric and Evangelidis, Georgios and Tulyakov,
Sergey and Wang, Yanzhi and Ren, Jian},
journal={arXiv preprint arXiv:2206.01191},
year={20... |
17,597 | /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/efficientformer.py | 186,187 | 161,197 | forward | ConvMlpWithNorm | 187,197 | 187,197 | class ConvMlpWithNorm(nn.Module):
"""
Implementation of MLP with 1*1 convolutions.
Input: tensor with shape [B, C, H, W]
"""
def __init__(
self,
in_features,
hidden_features=None,
out_features=None,
act_layer=nn.GELU,
norm_laye... | class ConvMlpWithNorm(nn.Module):
"""
Implementation of MLP with 1*1 convolutions.
Input: tensor with shape [B, C, H, W]
"""
def __init__(
self,
in_features,
hidden_features=None,
out_features=None,
act_layer=nn.GELU,
norm_laye... | x = self.fc1(x)
x = self.norm1(x)
x = self.act(x)
x = self.drop(x)
x = self.fc2(x)
x = self.norm2(x)
x = self.drop(x)
return x
| x = self.fc1(x)
x = self.norm1(x)
x = self.act(x)
x = self.drop(x)
x = self.fc2(x)
x = self.norm2(x)
x = self.drop(x)
return x
| 23 | 305 | 90 | 395 | """ EfficientFormer
@article{li2022efficientformer,
title={EfficientFormer: Vision Transformers at MobileNet Speed},
author={Li, Yanyu and Yuan, Geng and Wen, Yang and Hu, Eric and Evangelidis, Georgios and Tulyakov,
Sergey and Wang, Yanzhi and Ren, Jian},
journal={arXiv preprint arXiv:2206.01191},
year={20... | """ EfficientFormer
@article{li2022efficientformer,
title={EfficientFormer: Vision Transformers at MobileNet Speed},
author={Li, Yanyu and Yuan, Geng and Wen, Yang and Hu, Eric and Evangelidis, Georgios and Tulyakov,
Sergey and Wang, Yanzhi and Ren, Jian},
journal={arXiv preprint arXiv:2206.01191},
year={20... |
17,598 | /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/efficientformer.py | 131,132 | 117,137 | forward | Downsample | 132,137 | 132,137 | class Downsample(nn.Module):
"""
Downsampling via strided conv w/ norm
Input: tensor in shape [B, C, H, W]
Output: tensor in shape [B, C, H/stride, W/stride]
"""
def __init__(self, in_chs, out_chs, kernel_size=3, stride=2, padding=None, norm_layer=nn.BatchNorm2d):
super().__init__()
... | class Downsample(nn.Module):
"""
Downsampling via strided conv w/ norm
Input: tensor in shape [B, C, H, W]
Output: tensor in shape [B, C, H/stride, W/stride]
"""
def __init__(self, in_chs, out_chs, kernel_size=3, stride=2, padding=None, norm_layer=nn.BatchNorm2d):
super().__init__()
... | x = self.conv(x)
x = self.norm(x)
return x
| x = self.conv(x)
x = self.norm(x)
return x
| 8 | 212 | 31 | 243 | """ EfficientFormer
@article{li2022efficientformer,
title={EfficientFormer: Vision Transformers at MobileNet Speed},
author={Li, Yanyu and Yuan, Geng and Wen, Yang and Hu, Eric and Evangelidis, Georgios and Tulyakov,
Sergey and Wang, Yanzhi and Ren, Jian},
journal={arXiv preprint arXiv:2206.01191},
year={20... | """ EfficientFormer
@article{li2022efficientformer,
title={EfficientFormer: Vision Transformers at MobileNet Speed},
author={Li, Yanyu and Yuan, Geng and Wen, Yang and Hu, Eric and Evangelidis, Georgios and Tulyakov,
Sergey and Wang, Yanzhi and Ren, Jian},
journal={arXiv preprint arXiv:2206.01191},
year={20... |
17,600 | /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/efficientformer.py | 347,348 | 285,356 | forward | EfficientFormerStage | 348,356 | 348,356 | class EfficientFormerStage(nn.Module):
def __init__(
self,
dim,
dim_out,
depth,
downsample=True,
num_vit=1,
pool_size=3,
mlp_ratio=4.,
act_layer=nn.GELU,
norm_layer=nn.BatchNorm2d,
no... | class EfficientFormerStage(nn.Module):
def __init__(
self,
dim,
dim_out,
depth,
downsample=True,
num_vit=1,
pool_size=3,
mlp_ratio=4.,
act_layer=nn.GELU,
norm_layer=nn.BatchNorm2d,
no... | 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 | 580 | 77 | 657 | """ EfficientFormer
@article{li2022efficientformer,
title={EfficientFormer: Vision Transformers at MobileNet Speed},
author={Li, Yanyu and Yuan, Geng and Wen, Yang and Hu, Eric and Evangelidis, Georgios and Tulyakov,
Sergey and Wang, Yanzhi and Ren, Jian},
journal={arXiv preprint arXiv:2206.01191},
year={20... | """ EfficientFormer
@article{li2022efficientformer,
title={EfficientFormer: Vision Transformers at MobileNet Speed},
author={Li, Yanyu and Yuan, Geng and Wen, Yang and Hu, Eric and Evangelidis, Georgios and Tulyakov,
Sergey and Wang, Yanzhi and Ren, Jian},
journal={arXiv preprint arXiv:2206.01191},
year={20... |
17,601 | /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/efficientformer.py | 142,143 | 137,147 | forward | Flat | 143,147 | 143,147 | class Flat(nn.Module):
def __init__(self, ):
super().__init__()
def forward(self, x):
FILL_FUNC_BODY | class Flat(nn.Module):
def __init__(self, ):
super().__init__()
def forward(self, x):
FILL_FUNC_BODY | x = x.flatten(2).transpose(1, 2)
return x
| x = x.flatten(2).transpose(1, 2)
return x
| 6 | 47 | 29 | 76 | """ EfficientFormer
@article{li2022efficientformer,
title={EfficientFormer: Vision Transformers at MobileNet Speed},
author={Li, Yanyu and Yuan, Geng and Wen, Yang and Hu, Eric and Evangelidis, Georgios and Tulyakov,
Sergey and Wang, Yanzhi and Ren, Jian},
journal={arXiv preprint arXiv:2206.01191},
year={20... | """ EfficientFormer
@article{li2022efficientformer,
title={EfficientFormer: Vision Transformers at MobileNet Speed},
author={Li, Yanyu and Yuan, Geng and Wen, Yang and Hu, Eric and Evangelidis, Georgios and Tulyakov,
Sergey and Wang, Yanzhi and Ren, Jian},
journal={arXiv preprint arXiv:2206.01191},
year={20... |
17,602 | /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/efficientformer.py | 203,204 | 197,207 | forward | LayerScale | 204,207 | 204,207 | 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 | """ EfficientFormer
@article{li2022efficientformer,
title={EfficientFormer: Vision Transformers at MobileNet Speed},
author={Li, Yanyu and Yuan, Geng and Wen, Yang and Hu, Eric and Evangelidis, Georgios and Tulyakov,
Sergey and Wang, Yanzhi and Ren, Jian},
journal={arXiv preprint arXiv:2206.01191},
year={20... | """ EfficientFormer
@article{li2022efficientformer,
title={EfficientFormer: Vision Transformers at MobileNet Speed},
author={Li, Yanyu and Yuan, Geng and Wen, Yang and Hu, Eric and Evangelidis, Georgios and Tulyakov,
Sergey and Wang, Yanzhi and Ren, Jian},
journal={arXiv preprint arXiv:2206.01191},
year={20... |
17,603 | /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/efficientformer.py | 246,247 | 240,251 | forward | LayerScale2d | 247,251 | 247,251 | 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 | """ EfficientFormer
@article{li2022efficientformer,
title={EfficientFormer: Vision Transformers at MobileNet Speed},
author={Li, Yanyu and Yuan, Geng and Wen, Yang and Hu, Eric and Evangelidis, Georgios and Tulyakov,
Sergey and Wang, Yanzhi and Ren, Jian},
journal={arXiv preprint arXiv:2206.01191},
year={20... | """ EfficientFormer
@article{li2022efficientformer,
title={EfficientFormer: Vision Transformers at MobileNet Speed},
author={Li, Yanyu and Yuan, Geng and Wen, Yang and Hu, Eric and Evangelidis, Georgios and Tulyakov,
Sergey and Wang, Yanzhi and Ren, Jian},
journal={arXiv preprint arXiv:2206.01191},
year={20... |
17,604 | /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/efficientformer.py | 234,235 | 207,240 | forward | MetaBlock1d | 235,240 | 235,240 | class MetaBlock1d(nn.Module):
def __init__(
self,
dim,
mlp_ratio=4.,
act_layer=nn.GELU,
norm_layer=nn.LayerNorm,
proj_drop=0.,
drop_path=0.,
layer_scale_init_value=1e-5
):
super().__init__()
self.nor... | class MetaBlock1d(nn.Module):
def __init__(
self,
dim,
mlp_ratio=4.,
act_layer=nn.GELU,
norm_layer=nn.LayerNorm,
proj_drop=0.,
drop_path=0.,
layer_scale_init_value=1e-5
):
super().__init__()
self.nor... | x = x + self.drop_path(self.ls1(self.token_mixer(self.norm1(x))))
x = x + self.drop_path(self.ls2(self.mlp(self.norm2(x))))
return x
| x = x + self.drop_path(self.ls1(self.token_mixer(self.norm1(x))))
x = x + self.drop_path(self.ls2(self.mlp(self.norm2(x))))
return x
| 12 | 287 | 73 | 360 | """ EfficientFormer
@article{li2022efficientformer,
title={EfficientFormer: Vision Transformers at MobileNet Speed},
author={Li, Yanyu and Yuan, Geng and Wen, Yang and Hu, Eric and Evangelidis, Georgios and Tulyakov,
Sergey and Wang, Yanzhi and Ren, Jian},
journal={arXiv preprint arXiv:2206.01191},
year={20... | """ EfficientFormer
@article{li2022efficientformer,
title={EfficientFormer: Vision Transformers at MobileNet Speed},
author={Li, Yanyu and Yuan, Geng and Wen, Yang and Hu, Eric and Evangelidis, Georgios and Tulyakov,
Sergey and Wang, Yanzhi and Ren, Jian},
journal={arXiv preprint arXiv:2206.01191},
year={20... |
17,605 | /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/efficientformer.py | 279,280 | 251,285 | forward | MetaBlock2d | 280,285 | 280,285 | class MetaBlock2d(nn.Module):
def __init__(
self,
dim,
pool_size=3,
mlp_ratio=4.,
act_layer=nn.GELU,
norm_layer=nn.BatchNorm2d,
proj_drop=0.,
drop_path=0.,
layer_scale_init_value=1e-5
):
super().... | class MetaBlock2d(nn.Module):
def __init__(
self,
dim,
pool_size=3,
mlp_ratio=4.,
act_layer=nn.GELU,
norm_layer=nn.BatchNorm2d,
proj_drop=0.,
drop_path=0.,
layer_scale_init_value=1e-5
):
super().... | x = x + self.drop_path1(self.ls1(self.token_mixer(x)))
x = x + self.drop_path2(self.ls2(self.mlp(x)))
return x
| x = x + self.drop_path1(self.ls1(self.token_mixer(x)))
x = x + self.drop_path2(self.ls2(self.mlp(x)))
return x
| 12 | 321 | 63 | 384 | """ EfficientFormer
@article{li2022efficientformer,
title={EfficientFormer: Vision Transformers at MobileNet Speed},
author={Li, Yanyu and Yuan, Geng and Wen, Yang and Hu, Eric and Evangelidis, Georgios and Tulyakov,
Sergey and Wang, Yanzhi and Ren, Jian},
journal={arXiv preprint arXiv:2206.01191},
year={20... | """ EfficientFormer
@article{li2022efficientformer,
title={EfficientFormer: Vision Transformers at MobileNet Speed},
author={Li, Yanyu and Yuan, Geng and Wen, Yang and Hu, Eric and Evangelidis, Georgios and Tulyakov,
Sergey and Wang, Yanzhi and Ren, Jian},
journal={arXiv preprint arXiv:2206.01191},
year={20... |
17,606 | /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/efficientformer.py | 157,158 | 147,161 | forward | Pooling | 158,161 | 158,161 | class Pooling(nn.Module):
"""
Implementation of pooling for PoolFormer
--pool_size: pooling size
"""
def __init__(self, pool_size=3):
super().__init__()
self.pool = nn.AvgPool2d(pool_size, stride=1, padding=pool_size // 2, count_include_pad=False)
def forward(self, x):
FILL_FUN... | class Pooling(nn.Module):
"""
Implementation of pooling for PoolFormer
--pool_size: pooling size
"""
def __init__(self, pool_size=3):
super().__init__()
self.pool = nn.AvgPool2d(pool_size, stride=1, padding=pool_size // 2, count_include_pad=False)
def forward(self, x):
FILL_FUN... | return self.pool(x) - x
| return self.pool(x) - x
| 4 | 120 | 16 | 136 | """ EfficientFormer
@article{li2022efficientformer,
title={EfficientFormer: Vision Transformers at MobileNet Speed},
author={Li, Yanyu and Yuan, Geng and Wen, Yang and Hu, Eric and Evangelidis, Georgios and Tulyakov,
Sergey and Wang, Yanzhi and Ren, Jian},
journal={arXiv preprint arXiv:2206.01191},
year={20... | """ EfficientFormer
@article{li2022efficientformer,
title={EfficientFormer: Vision Transformers at MobileNet Speed},
author={Li, Yanyu and Yuan, Geng and Wen, Yang and Hu, Eric and Evangelidis, Georgios and Tulyakov,
Sergey and Wang, Yanzhi and Ren, Jian},
journal={arXiv preprint arXiv:2206.01191},
year={20... |
17,608 | /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/efficientformer_v2.py | 154,155 | 88,181 | forward | Attention2d | 155,181 | 155,181 | class Attention2d(torch.nn.Module):
attention_bias_cache: Dict[str, torch.Tensor]
def __init__(
self,
dim=384,
key_dim=32,
num_heads=8,
attn_ratio=4,
resolution=7,
act_layer=nn.GELU,
stride=None,
):
supe... | class Attention2d(torch.nn.Module):
attention_bias_cache: Dict[str, torch.Tensor]
def __init__(
self,
dim=384,
key_dim=32,
num_heads=8,
attn_ratio=4,
resolution=7,
act_layer=nn.GELU,
stride=None,
):
supe... | B, C, H, W = x.shape
if self.stride_conv is not None:
x = self.stride_conv(x)
q = self.q(x).reshape(B, self.num_heads, -1, self.N).permute(0, 1, 3, 2)
k = self.k(x).reshape(B, self.num_heads, -1, self.N).permute(0, 1, 2, 3)
v = self.v(x)
v_local = self.v_loca... | B, C, H, W = x.shape
if self.stride_conv is not None:
x = self.stride_conv(x)
q = self.q(x).reshape(B, self.num_heads, -1, self.N).permute(0, 1, 3, 2)
k = self.k(x).reshape(B, self.num_heads, -1, self.N).permute(0, 1, 2, 3)
v = self.v(x)
v_local = self.v_loca... | 98 | 980 | 401 | 1,381 | """ EfficientFormer-V2
@article{
li2022rethinking,
title={Rethinking Vision Transformers for MobileNet Size and Speed},
author={Li, Yanyu and Hu, Ju and Wen, Yang and Evangelidis, Georgios and Salahi, Kamyar and Wang, Yanzhi and Tulyakov, Sergey and Ren, Jian},
journal={arXiv preprint arXiv:2212.08059}... | """ EfficientFormer-V2
@article{
li2022rethinking,
title={Rethinking Vision Transformers for MobileNet Size and Speed},
author={Li, Yanyu and Hu, Ju and Wen, Yang and Evangelidis, Georgios and Salahi, Kamyar and Wang, Yanzhi and Tulyakov, Sergey and Ren, Jian},
journal={arXiv preprint arXiv:2212.08059}... |
17,609 | /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/efficientformer_v2.py | 259,260 | 196,279 | forward | Attention2dDownsample | 260,279 | 260,279 | class Attention2dDownsample(torch.nn.Module):
attention_bias_cache: Dict[str, torch.Tensor]
def __init__(
self,
dim=384,
key_dim=16,
num_heads=8,
attn_ratio=4,
resolution=7,
out_dim=None,
act_layer=nn.GELU,
):
... | class Attention2dDownsample(torch.nn.Module):
attention_bias_cache: Dict[str, torch.Tensor]
def __init__(
self,
dim=384,
key_dim=16,
num_heads=8,
attn_ratio=4,
resolution=7,
out_dim=None,
act_layer=nn.GELU,
):
... | B, C, H, W = x.shape
q = self.q(x).reshape(B, self.num_heads, -1, self.N2).permute(0, 1, 3, 2)
k = self.k(x).reshape(B, self.num_heads, -1, self.N).permute(0, 1, 2, 3)
v = self.v(x)
v_local = self.v_local(v)
v = v.reshape(B, self.num_heads, -1, self.N).permute(0, 1, 3, 2... | B, C, H, W = x.shape
q = self.q(x).reshape(B, self.num_heads, -1, self.N2).permute(0, 1, 3, 2)
k = self.k(x).reshape(B, self.num_heads, -1, self.N).permute(0, 1, 2, 3)
v = self.v(x)
v_local = self.v_local(v)
v = v.reshape(B, self.num_heads, -1, self.N).permute(0, 1, 3, 2... | 76 | 924 | 316 | 1,240 | """ EfficientFormer-V2
@article{
li2022rethinking,
title={Rethinking Vision Transformers for MobileNet Size and Speed},
author={Li, Yanyu and Hu, Ju and Wen, Yang and Evangelidis, Georgios and Salahi, Kamyar and Wang, Yanzhi and Tulyakov, Sergey and Ren, Jian},
journal={arXiv preprint arXiv:2212.08059}... | """ EfficientFormer-V2
@article{
li2022rethinking,
title={Rethinking Vision Transformers for MobileNet Size and Speed},
author={Li, Yanyu and Hu, Ju and Wen, Yang and Evangelidis, Georgios and Salahi, Kamyar and Wang, Yanzhi and Tulyakov, Sergey and Ren, Jian},
journal={arXiv preprint arXiv:2212.08059}... |
17,610 | /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/efficientformer_v2.py | 356,357 | 324,365 | forward | ConvMlpWithNorm | 357,365 | 357,365 | class ConvMlpWithNorm(nn.Module):
"""
Implementation of MLP with 1*1 convolutions.
Input: tensor with shape [B, C, H, W]
"""
def __init__(
self,
in_features,
hidden_features=None,
out_features=None,
act_layer=nn.GELU,
norm_laye... | class ConvMlpWithNorm(nn.Module):
"""
Implementation of MLP with 1*1 convolutions.
Input: tensor with shape [B, C, H, W]
"""
def __init__(
self,
in_features,
hidden_features=None,
out_features=None,
act_layer=nn.GELU,
norm_laye... | x = self.fc1(x)
x = self.mid(x)
x = self.drop1(x)
x = self.fc2(x)
x = self.drop2(x)
return x
| x = self.fc1(x)
x = self.mid(x)
x = self.drop1(x)
x = self.fc2(x)
x = self.drop2(x)
return x
| 17 | 362 | 68 | 430 | """ EfficientFormer-V2
@article{
li2022rethinking,
title={Rethinking Vision Transformers for MobileNet Size and Speed},
author={Li, Yanyu and Hu, Ju and Wen, Yang and Evangelidis, Georgios and Salahi, Kamyar and Wang, Yanzhi and Tulyakov, Sergey and Ren, Jian},
journal={arXiv preprint arXiv:2212.08059}... | """ EfficientFormer-V2
@article{
li2022rethinking,
title={Rethinking Vision Transformers for MobileNet Size and Speed},
author={Li, Yanyu and Hu, Ju and Wen, Yang and Evangelidis, Georgios and Salahi, Kamyar and Wang, Yanzhi and Tulyakov, Sergey and Ren, Jian},
journal={arXiv preprint arXiv:2212.08059}... |
17,611 | /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/efficientformer_v2.py | 82,83 | 54,88 | forward | ConvNorm | 83,88 | 83,88 | class ConvNorm(nn.Module):
def __init__(
self,
in_channels,
out_channels,
kernel_size=1,
stride=1,
padding='',
dilation=1,
groups=1,
bias=True,
norm_layer='batchnorm2d',
norm_kwargs=No... | class ConvNorm(nn.Module):
def __init__(
self,
in_channels,
out_channels,
kernel_size=1,
stride=1,
padding='',
dilation=1,
groups=1,
bias=True,
norm_layer='batchnorm2d',
norm_kwargs=No... | x = self.conv(x)
x = self.bn(x)
return x
| x = self.conv(x)
x = self.bn(x)
return x
| 8 | 238 | 31 | 269 | """ EfficientFormer-V2
@article{
li2022rethinking,
title={Rethinking Vision Transformers for MobileNet Size and Speed},
author={Li, Yanyu and Hu, Ju and Wen, Yang and Evangelidis, Georgios and Salahi, Kamyar and Wang, Yanzhi and Tulyakov, Sergey and Ren, Jian},
journal={arXiv preprint arXiv:2212.08059}... | """ EfficientFormer-V2
@article{
li2022rethinking,
title={Rethinking Vision Transformers for MobileNet Size and Speed},
author={Li, Yanyu and Hu, Ju and Wen, Yang and Evangelidis, Georgios and Salahi, Kamyar and Wang, Yanzhi and Tulyakov, Sergey and Ren, Jian},
journal={arXiv preprint arXiv:2212.08059}... |
17,612 | /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/efficientformer_v2.py | 317,318 | 279,324 | forward | Downsample | 318,324 | 318,324 | class Downsample(nn.Module):
def __init__(
self,
in_chs,
out_chs,
kernel_size=3,
stride=2,
padding=1,
resolution=7,
use_attn=False,
act_layer=nn.GELU,
norm_layer=nn.BatchNorm2d,
):
sup... | class Downsample(nn.Module):
def __init__(
self,
in_chs,
out_chs,
kernel_size=3,
stride=2,
padding=1,
resolution=7,
use_attn=False,
act_layer=nn.GELU,
norm_layer=nn.BatchNorm2d,
):
sup... | out = self.conv(x)
if self.attn is not None:
return self.attn(x) + out
return out
| out = self.conv(x)
if self.attn is not None:
return self.attn(x) + out
return out
| 14 | 319 | 45 | 364 | """ EfficientFormer-V2
@article{
li2022rethinking,
title={Rethinking Vision Transformers for MobileNet Size and Speed},
author={Li, Yanyu and Hu, Ju and Wen, Yang and Evangelidis, Georgios and Salahi, Kamyar and Wang, Yanzhi and Tulyakov, Sergey and Ren, Jian},
journal={arXiv preprint arXiv:2212.08059}... | """ EfficientFormer-V2
@article{
li2022rethinking,
title={Rethinking Vision Transformers for MobileNet Size and Speed},
author={Li, Yanyu and Hu, Ju and Wen, Yang and Evangelidis, Georgios and Salahi, Kamyar and Wang, Yanzhi and Tulyakov, Sergey and Ren, Jian},
journal={arXiv preprint arXiv:2212.08059}... |
17,613 | /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/efficientformer_v2.py | 615,616 | 508,652 | reset_classifier | EfficientFormerV2 | 616,622 | 616,622 | class EfficientFormerV2(nn.Module):
def __init__(
self,
depths,
in_chans=3,
img_size=224,
global_pool='avg',
embed_dims=None,
downsamples=None,
mlp_ratios=4,
norm_layer='batchnorm2d',
norm_eps=1e-... | class EfficientFormerV2(nn.Module):
def __init__(
self,
depths,
in_chans=3,
img_size=224,
global_pool='avg',
embed_dims=None,
downsamples=None,
mlp_ratios=4,
norm_layer='batchnorm2d',
norm_eps=1e-... | self.num_classes = num_classes
if global_pool is not None:
self.global_pool = global_pool
self.head = nn.Linear(self.num_features, num_classes) if num_classes > 0 else nn.Identity()
self.head_dist = nn.Linear(self.num_features, num_classes) if num_classes > 0 else nn.Identity... | self.num_classes = num_classes
if global_pool is not None:
self.global_pool = global_pool
self.head = nn.Linear(self.num_features, num_classes) if num_classes > 0 else nn.Identity()
self.head_dist = nn.Linear(self.num_features, num_classes) if num_classes > 0 else nn.Identity... | 31 | 1,598 | 107 | 1,705 | """ EfficientFormer-V2
@article{
li2022rethinking,
title={Rethinking Vision Transformers for MobileNet Size and Speed},
author={Li, Yanyu and Hu, Ju and Wen, Yang and Evangelidis, Georgios and Salahi, Kamyar and Wang, Yanzhi and Tulyakov, Sergey and Ren, Jian},
journal={arXiv preprint arXiv:2212.08059}... | """ EfficientFormer-V2
@article{
li2022rethinking,
title={Rethinking Vision Transformers for MobileNet Size and Speed},
author={Li, Yanyu and Hu, Ju and Wen, Yang and Evangelidis, Georgios and Salahi, Kamyar and Wang, Yanzhi and Tulyakov, Sergey and Ren, Jian},
journal={arXiv preprint arXiv:2212.08059}... |
17,614 | /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/efficientformer_v2.py | 419,420 | 376,426 | forward | EfficientFormerV2Block | 420,426 | 420,426 | class EfficientFormerV2Block(nn.Module):
def __init__(
self,
dim,
mlp_ratio=4.,
act_layer=nn.GELU,
norm_layer=nn.BatchNorm2d,
proj_drop=0.,
drop_path=0.,
layer_scale_init_value=1e-5,
resolution=7,
... | class EfficientFormerV2Block(nn.Module):
def __init__(
self,
dim,
mlp_ratio=4.,
act_layer=nn.GELU,
norm_layer=nn.BatchNorm2d,
proj_drop=0.,
drop_path=0.,
layer_scale_init_value=1e-5,
resolution=7,
... | if self.token_mixer is not None:
x = x + self.drop_path1(self.ls1(self.token_mixer(x)))
x = x + self.drop_path2(self.ls2(self.mlp(x)))
return x
| if self.token_mixer is not None:
x = x + self.drop_path1(self.ls1(self.token_mixer(x)))
x = x + self.drop_path2(self.ls2(self.mlp(x)))
return x
| 17 | 453 | 77 | 530 | """ EfficientFormer-V2
@article{
li2022rethinking,
title={Rethinking Vision Transformers for MobileNet Size and Speed},
author={Li, Yanyu and Hu, Ju and Wen, Yang and Evangelidis, Georgios and Salahi, Kamyar and Wang, Yanzhi and Tulyakov, Sergey and Ren, Jian},
journal={arXiv preprint arXiv:2212.08059}... | """ EfficientFormer-V2
@article{
li2022rethinking,
title={Rethinking Vision Transformers for MobileNet Size and Speed},
author={Li, Yanyu and Hu, Ju and Wen, Yang and Evangelidis, Georgios and Salahi, Kamyar and Wang, Yanzhi and Tulyakov, Sergey and Ren, Jian},
journal={arXiv preprint arXiv:2212.08059}... |
17,615 | /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/efficientformer_v2.py | 499,500 | 440,508 | forward | EfficientFormerV2Stage | 500,508 | 500,508 | class EfficientFormerV2Stage(nn.Module):
def __init__(
self,
dim,
dim_out,
depth,
resolution=7,
downsample=True,
block_stride=None,
downsample_use_attn=False,
block_use_attn=False,
num_vit=1,
... | class EfficientFormerV2Stage(nn.Module):
def __init__(
self,
dim,
dim_out,
depth,
resolution=7,
downsample=True,
block_stride=None,
downsample_use_attn=False,
block_use_attn=False,
num_vit=1,
... | 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 | 548 | 77 | 625 | """ EfficientFormer-V2
@article{
li2022rethinking,
title={Rethinking Vision Transformers for MobileNet Size and Speed},
author={Li, Yanyu and Hu, Ju and Wen, Yang and Evangelidis, Georgios and Salahi, Kamyar and Wang, Yanzhi and Tulyakov, Sergey and Ren, Jian},
journal={arXiv preprint arXiv:2212.08059}... | """ EfficientFormer-V2
@article{
li2022rethinking,
title={Rethinking Vision Transformers for MobileNet Size and Speed},
author={Li, Yanyu and Hu, Ju and Wen, Yang and Evangelidis, Georgios and Salahi, Kamyar and Wang, Yanzhi and Tulyakov, Sergey and Ren, Jian},
journal={arXiv preprint arXiv:2212.08059}... |
17,616 | /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/efficientformer_v2.py | 371,372 | 365,376 | forward | LayerScale2d | 372,376 | 372,376 | 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 | """ EfficientFormer-V2
@article{
li2022rethinking,
title={Rethinking Vision Transformers for MobileNet Size and Speed},
author={Li, Yanyu and Hu, Ju and Wen, Yang and Evangelidis, Georgios and Salahi, Kamyar and Wang, Yanzhi and Tulyakov, Sergey and Ren, Jian},
journal={arXiv preprint arXiv:2212.08059}... | """ EfficientFormer-V2
@article{
li2022rethinking,
title={Rethinking Vision Transformers for MobileNet Size and Speed},
author={Li, Yanyu and Hu, Ju and Wen, Yang and Evangelidis, Georgios and Salahi, Kamyar and Wang, Yanzhi and Tulyakov, Sergey and Ren, Jian},
journal={arXiv preprint arXiv:2212.08059}... |
17,617 | /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/efficientformer_v2.py | 188,189 | 181,196 | forward | LocalGlobalQuery | 189,196 | 189,196 | class LocalGlobalQuery(torch.nn.Module):
def __init__(self, in_dim, out_dim):
super().__init__()
self.pool = nn.AvgPool2d(1, 2, 0)
self.local = nn.Conv2d(in_dim, in_dim, kernel_size=3, stride=2, padding=1, groups=in_dim)
self.proj = ConvNorm(in_dim, out_dim, 1)
def forward(self,... | class LocalGlobalQuery(torch.nn.Module):
def __init__(self, in_dim, out_dim):
super().__init__()
self.pool = nn.AvgPool2d(1, 2, 0)
self.local = nn.Conv2d(in_dim, in_dim, kernel_size=3, stride=2, padding=1, groups=in_dim)
self.proj = ConvNorm(in_dim, out_dim, 1)
def forward(self,... | local_q = self.local(x)
pool_q = self.pool(x)
q = local_q + pool_q
q = self.proj(q)
return q
| local_q = self.local(x)
pool_q = self.pool(x)
q = local_q + pool_q
q = self.proj(q)
return q
| 16 | 144 | 58 | 202 | """ EfficientFormer-V2
@article{
li2022rethinking,
title={Rethinking Vision Transformers for MobileNet Size and Speed},
author={Li, Yanyu and Hu, Ju and Wen, Yang and Evangelidis, Georgios and Salahi, Kamyar and Wang, Yanzhi and Tulyakov, Sergey and Ren, Jian},
journal={arXiv preprint arXiv:2212.08059}... | """ EfficientFormer-V2
@article{
li2022rethinking,
title={Rethinking Vision Transformers for MobileNet Size and Speed},
author={Li, Yanyu and Hu, Ju and Wen, Yang and Evangelidis, Georgios and Salahi, Kamyar and Wang, Yanzhi and Tulyakov, Sergey and Ren, Jian},
journal={arXiv preprint arXiv:2212.08059}... |
17,620 | /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/efficientnet.py | 330,331 | 265,351 | forward | EfficientNetFeatures | 331,351 | 331,351 | class EfficientNetFeatures(nn.Module):
""" EfficientNet Feature Extractor
A work-in-progress feature extraction module for EfficientNet, to use as a backbone for segmentation
and object detection models.
"""
def __init__(
self,
block_args,
out_indices=(0, 1, 2, ... | class EfficientNetFeatures(nn.Module):
""" EfficientNet Feature Extractor
A work-in-progress feature extraction module for EfficientNet, to use as a backbone for segmentation
and object detection models.
"""
def __init__(
self,
block_args,
out_indices=(0, 1, 2, ... | x = self.conv_stem(x)
x = self.bn1(x)
if self.feature_hooks is None:
features = []
if 0 in self._stage_out_idx:
features.append(x) # add stem out
for i, b in enumerate(self.blocks):
if self.grad_checkpointing and not torch.jit.... | x = self.conv_stem(x)
x = self.bn1(x)
if self.feature_hooks is None:
features = []
if 0 in self._stage_out_idx:
features.append(x) # add stem out
for i, b in enumerate(self.blocks):
if self.grad_checkpointing and not torch.jit.... | 56 | 782 | 217 | 999 | """ The EfficientNet Family in PyTorch
An implementation of EfficienNet that covers variety of related models with efficient architectures:
* EfficientNet-V2
- `EfficientNetV2: Smaller Models and Faster Training` - https://arxiv.org/abs/2104.00298
* EfficientNet (B0-B8, L2 + Tensorflow pretrained AutoAug/RandAug/A... | """ The EfficientNet Family in PyTorch
An implementation of EfficienNet that covers variety of related models with efficient architectures:
* EfficientNet-V2
- `EfficientNetV2: Smaller Models and Faster Training` - https://arxiv.org/abs/2104.00298
* EfficientNet (B0-B8, L2 + Tensorflow pretrained AutoAug/RandAug/A... |
17,622 | /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 | 655,656 | 632,664 | forward | ClassifierHead | 656,664 | 656,664 | class ClassifierHead(nn.Module):
def __init__(
self,
in_channels,
widths,
n_classes=1000,
dropout=0.,
norm_layer=nn.BatchNorm2d,
act_layer=nn.Hardswish,
global_pool='avg',
norm_eps=1e-5,
):
super(ClassifierHead, self).__init__()
... | class ClassifierHead(nn.Module):
def __init__(
self,
in_channels,
widths,
n_classes=1000,
dropout=0.,
norm_layer=nn.BatchNorm2d,
act_layer=nn.Hardswish,
global_pool='avg',
norm_eps=1e-5,
):
super(ClassifierHead, self).__init__()
... | x = self.in_conv(x)
x = self.global_pool(x)
if pre_logits:
return x
x = self.classifier(x)
return x
| x = self.in_conv(x)
x = self.global_pool(x)
if pre_logits:
return x
x = self.classifier(x)
return x
| 15 | 342 | 61 | 403 | """ 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
... |
17,623 | /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 | 164,165 | 126,170 | forward | ConvBlock | 165,170 | 165,170 | class ConvBlock(nn.Module):
def __init__(
self,
in_channels: int,
out_channels: int,
kernel_size=3,
stride=1,
mid_channels=None,
expand_ratio=1,
use_bias=False,
norm_layer=(nn.BatchNorm2d, nn.BatchNorm2d),
act_layer=(nn.ReLU6, None),
... | class ConvBlock(nn.Module):
def __init__(
self,
in_channels: int,
out_channels: int,
kernel_size=3,
stride=1,
mid_channels=None,
expand_ratio=1,
use_bias=False,
norm_layer=(nn.BatchNorm2d, nn.BatchNorm2d),
act_layer=(nn.ReLU6, None),
... | x = self.conv1(x)
x = self.conv2(x)
return x
| x = self.conv1(x)
x = self.conv2(x)
return x
| 8 | 384 | 33 | 417 | """ 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
... |
17,624 | /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 | 77,78 | 49,85 | forward | ConvNormAct | 78,85 | 78,85 | class ConvNormAct(nn.Module):
def __init__(
self,
in_channels: int,
out_channels: int,
kernel_size=3,
stride=1,
dilation=1,
groups=1,
bias=False,
dropout=0.,
norm_layer=nn.BatchNorm2d,
act_layer=nn.ReLU,
):
super(Con... | class ConvNormAct(nn.Module):
def __init__(
self,
in_channels: int,
out_channels: int,
kernel_size=3,
stride=1,
dilation=1,
groups=1,
bias=False,
dropout=0.,
norm_layer=nn.BatchNorm2d,
act_layer=nn.ReLU,
):
super(Con... | x = self.dropout(x)
x = self.conv(x)
x = self.norm(x)
x = self.act(x)
return x
| x = self.dropout(x)
x = self.conv(x)
x = self.norm(x)
x = self.act(x)
return x
| 14 | 290 | 54 | 344 | """ 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
... |
17,625 | /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 | 120,121 | 85,126 | forward | DSConv | 121,126 | 121,126 | class DSConv(nn.Module):
def __init__(
self,
in_channels: int,
out_channels: int,
kernel_size=3,
stride=1,
use_bias=False,
norm_layer=(nn.BatchNorm2d, nn.BatchNorm2d),
act_layer=(nn.ReLU6, None),
):
super(DSConv, self).__init__()
us... | class DSConv(nn.Module):
def __init__(
self,
in_channels: int,
out_channels: int,
kernel_size=3,
stride=1,
use_bias=False,
norm_layer=(nn.BatchNorm2d, nn.BatchNorm2d),
act_layer=(nn.ReLU6, None),
):
super(DSConv, self).__init__()
us... | x = self.depth_conv(x)
x = self.point_conv(x)
return x
| x = self.depth_conv(x)
x = self.point_conv(x)
return x
| 8 | 349 | 35 | 384 | """ 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
... |
17,626 | /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 | 735,737 | 664,778 | set_grad_checkpointing | EfficientVit | 737,739 | 737,739 | class EfficientVit(nn.Module):
def __init__(
self,
in_chans=3,
widths=(),
depths=(),
head_dim=32,
expand_ratio=4,
norm_layer=nn.BatchNorm2d,
act_layer=nn.Hardswish,
global_pool='avg',
head_widths=(),
drop_rate=0.0,
num_c... | class EfficientVit(nn.Module):
def __init__(
self,
in_chans=3,
widths=(),
depths=(),
head_dim=32,
expand_ratio=4,
norm_layer=nn.BatchNorm2d,
act_layer=nn.Hardswish,
global_pool='avg',
head_widths=(),
drop_rate=0.0,
num_c... | self.grad_checkpointing = enable
| self.grad_checkpointing = enable
| 3 | 1,151 | 14 | 1,165 | """ 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
... |
17,627 | /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 | 401,402 | 368,407 | forward | EfficientVitBlock | 402,407 | 402,407 | class EfficientVitBlock(nn.Module):
def __init__(
self,
in_channels,
heads_ratio=1.0,
head_dim=32,
expand_ratio=4,
norm_layer=nn.BatchNorm2d,
act_layer=nn.Hardswish,
):
super(EfficientVitBlock, self).__init__()
self.context_module = Residua... | class EfficientVitBlock(nn.Module):
def __init__(
self,
in_channels,
heads_ratio=1.0,
head_dim=32,
expand_ratio=4,
norm_layer=nn.BatchNorm2d,
act_layer=nn.Hardswish,
):
super(EfficientVitBlock, self).__init__()
self.context_module = Residua... | x = self.context_module(x)
x = self.local_module(x)
return x
| x = self.context_module(x)
x = self.local_module(x)
return x
| 8 | 320 | 35 | 355 | """ 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
... |
17,628 | /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 | 861,862 | 778,897 | reset_classifier | EfficientVitLarge | 862,880 | 862,880 | class EfficientVitLarge(nn.Module):
def __init__(
self,
in_chans=3,
widths=(),
depths=(),
head_dim=32,
norm_layer=nn.BatchNorm2d,
act_layer=GELUTanh,
global_pool='avg',
head_widths=(),
drop_rate=0.0,
num_classes=1000,
no... | class EfficientVitLarge(nn.Module):
def __init__(
self,
in_chans=3,
widths=(),
depths=(),
head_dim=32,
norm_layer=nn.BatchNorm2d,
act_layer=GELUTanh,
global_pool='avg',
head_widths=(),
drop_rate=0.0,
num_classes=1000,
no... | self.num_classes = num_classes
if global_pool is not None:
self.global_pool = global_pool
if num_classes > 0:
self.head = ClassifierHead(
self.num_features,
self.head_widths,
n_classes=num_classes,
dropout=se... | self.num_classes = num_classes
if global_pool is not None:
self.global_pool = global_pool
if num_classes > 0:
self.head = ClassifierHead(
self.num_features,
self.head_widths,
n_classes=num_classes,
dropout=se... | 38 | 1,063 | 196 | 1,259 | """ 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
... |
17,629 | /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 | 628,629 | 569,632 | forward | EfficientVitLargeStage | 629,632 | 629,632 | class EfficientVitLargeStage(nn.Module):
def __init__(
self,
in_chs,
out_chs,
depth,
norm_layer,
act_layer,
head_dim,
vit_stage=False,
fewer_norm=False,
):
super(EfficientVitLargeStage, self).__in... | class EfficientVitLargeStage(nn.Module):
def __init__(
self,
in_chs,
out_chs,
depth,
norm_layer,
act_layer,
head_dim,
vit_stage=False,
fewer_norm=False,
):
super(EfficientVitLargeStage, self).__in... | return self.blocks(x)
| return self.blocks(x)
| 2 | 514 | 14 | 528 | """ 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
... |
17,630 | /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 | 565,566 | 509,569 | forward | EfficientVitStage | 566,569 | 566,569 | class EfficientVitStage(nn.Module):
def __init__(
self,
in_chs,
out_chs,
depth,
norm_layer,
act_layer,
expand_ratio,
head_dim,
vit_stage=False,
):
super(EfficientVitStage, self).__init__()
... | class EfficientVitStage(nn.Module):
def __init__(
self,
in_chs,
out_chs,
depth,
norm_layer,
act_layer,
expand_ratio,
head_dim,
vit_stage=False,
):
super(EfficientVitStage, self).__init__()
... | return self.blocks(x)
| return self.blocks(x)
| 2 | 457 | 14 | 471 | """ 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
... |
17,631 | /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 | 263,264 | 224,269 | forward | FusedMBConv | 264,269 | 264,269 | class FusedMBConv(nn.Module):
def __init__(
self,
in_channels: int,
out_channels: int,
kernel_size=3,
stride=1,
mid_channels=None,
expand_ratio=6,
groups=1,
use_bias=False,
norm_layer=(nn.BatchNorm2d, nn.BatchNorm2d),
act_layer=... | class FusedMBConv(nn.Module):
def __init__(
self,
in_channels: int,
out_channels: int,
kernel_size=3,
stride=1,
mid_channels=None,
expand_ratio=6,
groups=1,
use_bias=False,
norm_layer=(nn.BatchNorm2d, nn.BatchNorm2d),
act_layer=... | x = self.spatial_conv(x)
x = self.point_conv(x)
return x
| x = self.spatial_conv(x)
x = self.point_conv(x)
return x
| 8 | 400 | 36 | 436 | """ 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
... |
17,632 | /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 | 336,337 | 269,365 | forward | LiteMLA | 337,365 | 337,365 | class LiteMLA(nn.Module):
"""Lightweight multi-scale linear attention"""
def __init__(
self,
in_channels: int,
out_channels: int,
heads: int or None = None,
heads_ratio: float = 1.0,
dim=8,
use_bias=False,
norm_layer=(None, nn.BatchNorm2d),
... | class LiteMLA(nn.Module):
"""Lightweight multi-scale linear attention"""
def __init__(
self,
in_channels: int,
out_channels: int,
heads: int or None = None,
heads_ratio: float = 1.0,
dim=8,
use_bias=False,
norm_layer=(None, nn.BatchNorm2d),
... | B, _, H, W = x.shape
# generate multi-scale q, k, v
qkv = self.qkv(x)
multi_scale_qkv = [qkv]
for op in self.aggreg:
multi_scale_qkv.append(op(qkv))
multi_scale_qkv = torch.cat(multi_scale_qkv, dim=1)
multi_scale_qkv = multi_scale_qkv.reshape(B, -1, 3... | B, _, H, W = x.shape
# generate multi-scale q, k, v
qkv = self.qkv(x)
multi_scale_qkv = [qkv]
for op in self.aggreg:
multi_scale_qkv.append(op(qkv))
multi_scale_qkv = torch.cat(multi_scale_qkv, dim=1)
multi_scale_qkv = multi_scale_qkv.reshape(B, -1, 3... | 93 | 696 | 386 | 1,082 | """ 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
... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.