mirror of
https://github.com/huggingface/pytorch-image-models.git
synced 2025-06-03 15:01:08 +08:00
Merge pull request #1756 from huggingface/mw-resnet
ResNet models on HF hub, multi-weight support
This commit is contained in:
Ross Wightman
GitHub
commit
a09e240cd6
12
README.md
12
README.md
@ -24,6 +24,18 @@ And a big thanks to all GitHub sponsors who helped with some of my costs before
|
||||
* ❗Updates after Oct 10, 2022 are available in 0.8.x pre-releases (`pip install --pre timm`) or cloning main❗
|
||||
* Stable releases are 0.6.x and available by normal pip install or clone from [0.6.x](https://github.com/rwightman/pytorch-image-models/tree/0.6.x) branch.
|
||||
|
||||
### April 5, 2023
|
||||
* ALL ResNet models pushed to Hugging Face Hub with multi-weight support
|
||||
* All past `timm` trained weights added with recipe based tags to differentiate
|
||||
* All ResNet strikes back A1/A2/A3 (seed 0) and R50 example B/C1/C2/D weights available
|
||||
* Add torchvision v2 recipe weights to existing torchvision originals
|
||||
* See comparison table in https://huggingface.co/timm/seresnextaa101d_32x8d.sw_in12k_ft_in1k_288#model-comparison
|
||||
* New ImageNet-12k + ImageNet-1k fine-tunes available for a few anti-aliased ResNet models
|
||||
* `resnetaa50d.sw_in12k_ft_in1k` - 81.7 @ 224, 82.6 @ 288
|
||||
* `resnetaa101d.sw_in12k_ft_in1k` - 83.5 @ 224, 84.1 @ 288
|
||||
* `seresnextaa101d_32x8d.sw_in12k_ft_in1k` - 86.0 @ 224, 86.5 @ 288
|
||||
* `seresnextaa101d_32x8d.sw_in12k_ft_in1k_288` - 86.5 @ 288, 86.7 @ 320
|
||||
|
||||
### March 31, 2023
|
||||
* Add first ConvNext-XXLarge CLIP -> IN-1k fine-tune and IN-12k intermediate fine-tunes for convnext-base/large CLIP models.
|
||||
|
||||
|
||||
@ -21,7 +21,6 @@ from .eva import *
|
||||
from .focalnet import *
|
||||
from .gcvit import *
|
||||
from .ghostnet import *
|
||||
from .gluon_resnet import *
|
||||
from .gluon_xception import *
|
||||
from .hardcorenas import *
|
||||
from .hrnet import *
|
||||
|
||||
@ -134,7 +134,7 @@ def _deprecated_model_shim(deprecated_name: str, current_fn: Callable = None, cu
|
||||
def _fn(pretrained=False, **kwargs):
|
||||
assert current_fn is not None, f'Model {deprecated_name} has been removed with no replacement.'
|
||||
current_name = '.'.join([current_fn.__name__, current_tag]) if current_tag else current_fn.__name__
|
||||
warnings.warn(f'Mapping deprecated model {deprecated_name} to current {current_name}.', stacklevel=2)
|
||||
warnings.warn(f'Mapping deprecated model name {deprecated_name} to current {current_name}.', stacklevel=2)
|
||||
pretrained_cfg = kwargs.pop('pretrained_cfg', None)
|
||||
return current_fn(pretrained=pretrained, pretrained_cfg=pretrained_cfg or current_tag, **kwargs)
|
||||
return _fn
|
||||
|
||||
@ -1,247 +0,0 @@
|
||||
"""Pytorch impl of MxNet Gluon ResNet/(SE)ResNeXt variants
|
||||
This file evolved from https://github.com/pytorch/vision 'resnet.py' with (SE)-ResNeXt additions
|
||||
and ports of Gluon variations (https://github.com/dmlc/gluon-cv/blob/master/gluoncv/model_zoo/resnet.py)
|
||||
by Ross Wightman
|
||||
"""
|
||||
|
||||
from timm.data import IMAGENET_DEFAULT_MEAN, IMAGENET_DEFAULT_STD
|
||||
from timm.layers import SEModule
|
||||
from ._builder import build_model_with_cfg
|
||||
from ._registry import register_model
|
||||
from .resnet import ResNet, Bottleneck, BasicBlock
|
||||
|
||||
__all__ = []
|
||||
|
||||
|
||||
def _cfg(url='', **kwargs):
|
||||
return {
|
||||
'url': url,
|
||||
'num_classes': 1000, 'input_size': (3, 224, 224), 'pool_size': (7, 7),
|
||||
'crop_pct': 0.875, 'interpolation': 'bicubic',
|
||||
'mean': IMAGENET_DEFAULT_MEAN, 'std': IMAGENET_DEFAULT_STD,
|
||||
'first_conv': 'conv1', 'classifier': 'fc',
|
||||
**kwargs
|
||||
}
|
||||
|
||||
|
||||
default_cfgs = {
|
||||
'gluon_resnet18_v1b': _cfg(url='https://github.com/rwightman/pytorch-pretrained-gluonresnet/releases/download/v0.1/gluon_resnet18_v1b-0757602b.pth'),
|
||||
'gluon_resnet34_v1b': _cfg(url='https://github.com/rwightman/pytorch-pretrained-gluonresnet/releases/download/v0.1/gluon_resnet34_v1b-c6d82d59.pth'),
|
||||
'gluon_resnet50_v1b': _cfg(url='https://github.com/rwightman/pytorch-pretrained-gluonresnet/releases/download/v0.1/gluon_resnet50_v1b-0ebe02e2.pth'),
|
||||
'gluon_resnet101_v1b': _cfg(url='https://github.com/rwightman/pytorch-pretrained-gluonresnet/releases/download/v0.1/gluon_resnet101_v1b-3b017079.pth'),
|
||||
'gluon_resnet152_v1b': _cfg(url='https://github.com/rwightman/pytorch-pretrained-gluonresnet/releases/download/v0.1/gluon_resnet152_v1b-c1edb0dd.pth'),
|
||||
'gluon_resnet50_v1c': _cfg(url='https://github.com/rwightman/pytorch-pretrained-gluonresnet/releases/download/v0.1/gluon_resnet50_v1c-48092f55.pth',
|
||||
first_conv='conv1.0'),
|
||||
'gluon_resnet101_v1c': _cfg(url='https://github.com/rwightman/pytorch-pretrained-gluonresnet/releases/download/v0.1/gluon_resnet101_v1c-1f26822a.pth',
|
||||
first_conv='conv1.0'),
|
||||
'gluon_resnet152_v1c': _cfg(url='https://github.com/rwightman/pytorch-pretrained-gluonresnet/releases/download/v0.1/gluon_resnet152_v1c-a3bb0b98.pth',
|
||||
first_conv='conv1.0'),
|
||||
'gluon_resnet50_v1d': _cfg(url='https://github.com/rwightman/pytorch-pretrained-gluonresnet/releases/download/v0.1/gluon_resnet50_v1d-818a1b1b.pth',
|
||||
first_conv='conv1.0'),
|
||||
'gluon_resnet101_v1d': _cfg(url='https://github.com/rwightman/pytorch-pretrained-gluonresnet/releases/download/v0.1/gluon_resnet101_v1d-0f9c8644.pth',
|
||||
first_conv='conv1.0'),
|
||||
'gluon_resnet152_v1d': _cfg(url='https://github.com/rwightman/pytorch-pretrained-gluonresnet/releases/download/v0.1/gluon_resnet152_v1d-bd354e12.pth',
|
||||
first_conv='conv1.0'),
|
||||
'gluon_resnet50_v1s': _cfg(url='https://github.com/rwightman/pytorch-pretrained-gluonresnet/releases/download/v0.1/gluon_resnet50_v1s-1762acc0.pth',
|
||||
first_conv='conv1.0'),
|
||||
'gluon_resnet101_v1s': _cfg(url='https://github.com/rwightman/pytorch-pretrained-gluonresnet/releases/download/v0.1/gluon_resnet101_v1s-60fe0cc1.pth',
|
||||
first_conv='conv1.0'),
|
||||
'gluon_resnet152_v1s': _cfg(url='https://github.com/rwightman/pytorch-pretrained-gluonresnet/releases/download/v0.1/gluon_resnet152_v1s-dcc41b81.pth',
|
||||
first_conv='conv1.0'),
|
||||
'gluon_resnext50_32x4d': _cfg(url='https://github.com/rwightman/pytorch-pretrained-gluonresnet/releases/download/v0.1/gluon_resnext50_32x4d-e6a097c1.pth'),
|
||||
'gluon_resnext101_32x4d': _cfg(url='https://github.com/rwightman/pytorch-pretrained-gluonresnet/releases/download/v0.1/gluon_resnext101_32x4d-b253c8c4.pth'),
|
||||
'gluon_resnext101_64x4d': _cfg(url='https://github.com/rwightman/pytorch-pretrained-gluonresnet/releases/download/v0.1/gluon_resnext101_64x4d-f9a8e184.pth'),
|
||||
'gluon_seresnext50_32x4d': _cfg(url='https://github.com/rwightman/pytorch-pretrained-gluonresnet/releases/download/v0.1/gluon_seresnext50_32x4d-90cf2d6e.pth'),
|
||||
'gluon_seresnext101_32x4d': _cfg(url='https://github.com/rwightman/pytorch-pretrained-gluonresnet/releases/download/v0.1/gluon_seresnext101_32x4d-cf52900d.pth'),
|
||||
'gluon_seresnext101_64x4d': _cfg(url='https://github.com/rwightman/pytorch-pretrained-gluonresnet/releases/download/v0.1/gluon_seresnext101_64x4d-f9926f93.pth'),
|
||||
'gluon_senet154': _cfg(url='https://github.com/rwightman/pytorch-pretrained-gluonresnet/releases/download/v0.1/gluon_senet154-70a1a3c0.pth',
|
||||
first_conv='conv1.0'),
|
||||
}
|
||||
|
||||
|
||||
def _create_resnet(variant, pretrained=False, **kwargs):
|
||||
return build_model_with_cfg(ResNet, variant, pretrained, **kwargs)
|
||||
|
||||
|
||||
@register_model
|
||||
def gluon_resnet18_v1b(pretrained=False, **kwargs):
|
||||
"""Constructs a ResNet-18 model.
|
||||
"""
|
||||
model_args = dict(block=BasicBlock, layers=[2, 2, 2, 2], **kwargs)
|
||||
return _create_resnet('gluon_resnet18_v1b', pretrained, **model_args)
|
||||
|
||||
|
||||
@register_model
|
||||
def gluon_resnet34_v1b(pretrained=False, **kwargs):
|
||||
"""Constructs a ResNet-34 model.
|
||||
"""
|
||||
model_args = dict(block=BasicBlock, layers=[3, 4, 6, 3], **kwargs)
|
||||
return _create_resnet('gluon_resnet34_v1b', pretrained, **model_args)
|
||||
|
||||
|
||||
@register_model
|
||||
def gluon_resnet50_v1b(pretrained=False, **kwargs):
|
||||
"""Constructs a ResNet-50 model.
|
||||
"""
|
||||
model_args = dict(block=Bottleneck, layers=[3, 4, 6, 3], **kwargs)
|
||||
return _create_resnet('gluon_resnet50_v1b', pretrained, **model_args)
|
||||
|
||||
|
||||
@register_model
|
||||
def gluon_resnet101_v1b(pretrained=False, **kwargs):
|
||||
"""Constructs a ResNet-101 model.
|
||||
"""
|
||||
model_args = dict(block=Bottleneck, layers=[3, 4, 23, 3], **kwargs)
|
||||
return _create_resnet('gluon_resnet101_v1b', pretrained, **model_args)
|
||||
|
||||
|
||||
@register_model
|
||||
def gluon_resnet152_v1b(pretrained=False, **kwargs):
|
||||
"""Constructs a ResNet-152 model.
|
||||
"""
|
||||
model_args = dict(block=Bottleneck, layers=[3, 8, 36, 3], **kwargs)
|
||||
return _create_resnet('gluon_resnet152_v1b', pretrained, **model_args)
|
||||
|
||||
|
||||
@register_model
|
||||
def gluon_resnet50_v1c(pretrained=False, **kwargs):
|
||||
"""Constructs a ResNet-50 model.
|
||||
"""
|
||||
model_args = dict(block=Bottleneck, layers=[3, 4, 6, 3], stem_width=32, stem_type='deep', **kwargs)
|
||||
return _create_resnet('gluon_resnet50_v1c', pretrained, **model_args)
|
||||
|
||||
|
||||
@register_model
|
||||
def gluon_resnet101_v1c(pretrained=False, **kwargs):
|
||||
"""Constructs a ResNet-101 model.
|
||||
"""
|
||||
model_args = dict(block=Bottleneck, layers=[3, 4, 23, 3], stem_width=32, stem_type='deep', **kwargs)
|
||||
return _create_resnet('gluon_resnet101_v1c', pretrained, **model_args)
|
||||
|
||||
|
||||
@register_model
|
||||
def gluon_resnet152_v1c(pretrained=False, **kwargs):
|
||||
"""Constructs a ResNet-152 model.
|
||||
"""
|
||||
model_args = dict(block=Bottleneck, layers=[3, 8, 36, 3], stem_width=32, stem_type='deep', **kwargs)
|
||||
return _create_resnet('gluon_resnet152_v1c', pretrained, **model_args)
|
||||
|
||||
|
||||
@register_model
|
||||
def gluon_resnet50_v1d(pretrained=False, **kwargs):
|
||||
"""Constructs a ResNet-50 model.
|
||||
"""
|
||||
model_args = dict(
|
||||
block=Bottleneck, layers=[3, 4, 6, 3], stem_width=32, stem_type='deep', avg_down=True, **kwargs)
|
||||
return _create_resnet('gluon_resnet50_v1d', pretrained, **model_args)
|
||||
|
||||
|
||||
@register_model
|
||||
def gluon_resnet101_v1d(pretrained=False, **kwargs):
|
||||
"""Constructs a ResNet-101 model.
|
||||
"""
|
||||
model_args = dict(
|
||||
block=Bottleneck, layers=[3, 4, 23, 3], stem_width=32, stem_type='deep', avg_down=True, **kwargs)
|
||||
return _create_resnet('gluon_resnet101_v1d', pretrained, **model_args)
|
||||
|
||||
|
||||
@register_model
|
||||
def gluon_resnet152_v1d(pretrained=False, **kwargs):
|
||||
"""Constructs a ResNet-152 model.
|
||||
"""
|
||||
model_args = dict(
|
||||
block=Bottleneck, layers=[3, 8, 36, 3], stem_width=32, stem_type='deep', avg_down=True, **kwargs)
|
||||
return _create_resnet('gluon_resnet152_v1d', pretrained, **model_args)
|
||||
|
||||
|
||||
@register_model
|
||||
def gluon_resnet50_v1s(pretrained=False, **kwargs):
|
||||
"""Constructs a ResNet-50 model.
|
||||
"""
|
||||
model_args = dict(
|
||||
block=Bottleneck, layers=[3, 4, 6, 3], stem_width=64, stem_type='deep', **kwargs)
|
||||
return _create_resnet('gluon_resnet50_v1s', pretrained, **model_args)
|
||||
|
||||
|
||||
|
||||
@register_model
|
||||
def gluon_resnet101_v1s(pretrained=False, **kwargs):
|
||||
"""Constructs a ResNet-101 model.
|
||||
"""
|
||||
model_args = dict(
|
||||
block=Bottleneck, layers=[3, 4, 23, 3], stem_width=64, stem_type='deep', **kwargs)
|
||||
return _create_resnet('gluon_resnet101_v1s', pretrained, **model_args)
|
||||
|
||||
|
||||
@register_model
|
||||
def gluon_resnet152_v1s(pretrained=False, **kwargs):
|
||||
"""Constructs a ResNet-152 model.
|
||||
"""
|
||||
model_args = dict(
|
||||
block=Bottleneck, layers=[3, 8, 36, 3], stem_width=64, stem_type='deep', **kwargs)
|
||||
return _create_resnet('gluon_resnet152_v1s', pretrained, **model_args)
|
||||
|
||||
|
||||
|
||||
@register_model
|
||||
def gluon_resnext50_32x4d(pretrained=False, **kwargs):
|
||||
"""Constructs a ResNeXt50-32x4d model.
|
||||
"""
|
||||
model_args = dict(block=Bottleneck, layers=[3, 4, 6, 3], cardinality=32, base_width=4, **kwargs)
|
||||
return _create_resnet('gluon_resnext50_32x4d', pretrained, **model_args)
|
||||
|
||||
|
||||
@register_model
|
||||
def gluon_resnext101_32x4d(pretrained=False, **kwargs):
|
||||
"""Constructs a ResNeXt-101 model.
|
||||
"""
|
||||
model_args = dict(block=Bottleneck, layers=[3, 4, 23, 3], cardinality=32, base_width=4, **kwargs)
|
||||
return _create_resnet('gluon_resnext101_32x4d', pretrained, **model_args)
|
||||
|
||||
|
||||
@register_model
|
||||
def gluon_resnext101_64x4d(pretrained=False, **kwargs):
|
||||
"""Constructs a ResNeXt-101 model.
|
||||
"""
|
||||
model_args = dict(block=Bottleneck, layers=[3, 4, 23, 3], cardinality=64, base_width=4, **kwargs)
|
||||
return _create_resnet('gluon_resnext101_64x4d', pretrained, **model_args)
|
||||
|
||||
|
||||
@register_model
|
||||
def gluon_seresnext50_32x4d(pretrained=False, **kwargs):
|
||||
"""Constructs a SEResNeXt50-32x4d model.
|
||||
"""
|
||||
model_args = dict(
|
||||
block=Bottleneck, layers=[3, 4, 6, 3], cardinality=32, base_width=4,
|
||||
block_args=dict(attn_layer=SEModule), **kwargs)
|
||||
return _create_resnet('gluon_seresnext50_32x4d', pretrained, **model_args)
|
||||
|
||||
|
||||
@register_model
|
||||
def gluon_seresnext101_32x4d(pretrained=False, **kwargs):
|
||||
"""Constructs a SEResNeXt-101-32x4d model.
|
||||
"""
|
||||
model_args = dict(
|
||||
block=Bottleneck, layers=[3, 4, 23, 3], cardinality=32, base_width=4,
|
||||
block_args=dict(attn_layer=SEModule), **kwargs)
|
||||
return _create_resnet('gluon_seresnext101_32x4d', pretrained, **model_args)
|
||||
|
||||
|
||||
@register_model
|
||||
def gluon_seresnext101_64x4d(pretrained=False, **kwargs):
|
||||
"""Constructs a SEResNeXt-101-64x4d model.
|
||||
"""
|
||||
model_args = dict(
|
||||
block=Bottleneck, layers=[3, 4, 23, 3], cardinality=64, base_width=4,
|
||||
block_args=dict(attn_layer=SEModule), **kwargs)
|
||||
return _create_resnet('gluon_seresnext101_64x4d', pretrained, **model_args)
|
||||
|
||||
|
||||
@register_model
|
||||
def gluon_senet154(pretrained=False, **kwargs):
|
||||
"""Constructs an SENet-154 model.
|
||||
"""
|
||||
model_args = dict(
|
||||
block=Bottleneck, layers=[3, 8, 36, 3], cardinality=64, base_width=4, stem_type='deep',
|
||||
down_kernel_size=3, block_reduce_first=2, block_args=dict(attn_layer=SEModule), **kwargs)
|
||||
return _create_resnet('gluon_senet154', pretrained, **model_args)
|
||||
File diff suppressed because it is too large
Load Diff
@ -41,7 +41,7 @@ from timm.layers import PatchEmbed, Mlp, DropPath, trunc_normal_, lecun_normal_,
|
||||
resample_abs_pos_embed, RmsNorm
|
||||
from ._builder import build_model_with_cfg
|
||||
from ._manipulate import named_apply, checkpoint_seq, adapt_input_conv
|
||||
from ._registry import generate_default_cfgs, register_model
|
||||
from ._registry import generate_default_cfgs, register_model, register_model_deprecations
|
||||
|
||||
__all__ = ['VisionTransformer'] # model_registry will add each entrypoint fn to this
|
||||
|
||||
@ -1830,3 +1830,27 @@ def vit_huge_patch14_xp_224(pretrained=False, **kwargs):
|
||||
model = _create_vision_transformer(
|
||||
'vit_huge_patch14_xp_224', pretrained=pretrained, **dict(model_kwargs, **kwargs))
|
||||
return model
|
||||
|
||||
|
||||
register_model_deprecations(__name__, {
|
||||
'vit_tiny_patch16_224_in21k': 'vit_tiny_patch16_224.augreg_in21k',
|
||||
'vit_small_patch32_224_in21k': 'vit_small_patch32_224.augreg_in21k',
|
||||
'vit_small_patch16_224_in21k': 'vit_small_patch16_224.augreg_in21k',
|
||||
'vit_base_patch32_224_in21k': 'vit_base_patch32_224.augreg_in21k',
|
||||
'vit_base_patch16_224_in21k': 'vit_base_patch16_224.augreg_in21k',
|
||||
'vit_base_patch8_224_in21k': 'vit_base_patch8_224.augreg_in21k',
|
||||
'vit_large_patch32_224_in21k': 'vit_large_patch32_224.orig_in21k',
|
||||
'vit_large_patch16_224_in21k': 'vit_large_patch16_224.augreg_in21k',
|
||||
'vit_huge_patch14_224_in21k': 'vit_huge_patch14_224.orig_in21k',
|
||||
'vit_base_patch32_224_sam': 'vit_base_patch32_224.sam',
|
||||
'vit_base_patch16_224_sam': 'vit_base_patch16_224.sam',
|
||||
'vit_small_patch16_224_dino': 'vit_small_patch16_224.dino',
|
||||
'vit_small_patch8_224_dino': 'vit_small_patch8_224.dino',
|
||||
'vit_base_patch16_224_dino': 'vit_base_patch16_224.dino',
|
||||
'vit_base_patch8_224_dino': 'vit_base_patch8_224.dino',
|
||||
'vit_base_patch16_224_miil_in21k': 'vit_base_patch16_224_miil.in21k',
|
||||
'vit_base_patch32_224_clip_laion2b': 'vit_base_patch32_clip_224.laion2b',
|
||||
'vit_large_patch14_224_clip_laion2b': 'vit_large_patch14_clip_224.laion2b',
|
||||
'vit_huge_patch14_224_clip_laion2b': 'vit_huge_patch14_clip_224.laion2b',
|
||||
'vit_giant_patch14_224_clip_laion2b': 'vit_giant_patch14_clip_224.laion2b',
|
||||
})
|
||||
|
||||
@ -20,7 +20,7 @@ import torch.nn as nn
|
||||
|
||||
from timm.data import IMAGENET_DEFAULT_MEAN, IMAGENET_DEFAULT_STD
|
||||
from timm.layers import StdConv2dSame, StdConv2d, to_2tuple
|
||||
from ._registry import generate_default_cfgs, register_model
|
||||
from ._registry import generate_default_cfgs, register_model, register_model_deprecations
|
||||
from .resnet import resnet26d, resnet50d
|
||||
from .resnetv2 import ResNetV2, create_resnetv2_stem
|
||||
from .vision_transformer import _create_vision_transformer
|
||||
@ -318,3 +318,13 @@ def vit_base_resnet50d_224(pretrained=False, **kwargs):
|
||||
model = _create_vision_transformer_hybrid(
|
||||
'vit_base_resnet50d_224', backbone=backbone, pretrained=pretrained, **model_kwargs)
|
||||
return model
|
||||
|
||||
|
||||
register_model_deprecations(__name__, {
|
||||
'vit_tiny_r_s16_p8_224_in21k': 'vit_tiny_r_s16_p8_224.augreg_in21k',
|
||||
'vit_small_r26_s32_224_in21k': 'vit_small_r26_s32_224.augreg_in21k',
|
||||
'vit_base_r50_s16_224_in21k': 'vit_base_r50_s16_224.orig_in21k',
|
||||
'vit_base_resnet50_224_in21k': 'vit_base_r50_s16_224.orig_in21k',
|
||||
'vit_large_r50_s32_224_in21k': 'vit_large_r50_s32_224.augreg_in21k',
|
||||
'vit_base_resnet50_384': 'vit_base_r50_s16_384.orig_in21k_ft_in1k'
|
||||
})
|
||||
|
||||
@ -1 +1 @@
|
||||
__version__ = '0.8.18dev0'
|
||||
__version__ = '0.8.19dev0'
|
||||
|
||||
Loading…
x
Reference in New Issue
Block a user