mirrors
/
pytorch-image-models
mirror of https://github.com/huggingface/pytorch-image-models.git
1
0
Fork 0
Code Issues Projects Releases Wiki Activity

Fix arg merging of sknet, old seresnet. Fix #2470

pull/2474/head^2
Ross Wightman 2025-04-14 09:24:31 -07:00 committed by Ross Wightman
parent e44f14d7d2
commit 681be882e8
2 changed files with 28 additions and 28 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

36
timm/models/senet.py
View File

@ -404,62 +404,62 @@ default_cfgs = generate_default_cfgs({
@register_model
def legacy_seresnet18(pretrained=False, **kwargs) -> SENet:
model_args = dict(
block=SEResNetBlock, layers=[2, 2, 2, 2], groups=1, reduction=16, **kwargs)
return _create_senet('legacy_seresnet18', pretrained, **model_args)
block=SEResNetBlock, layers=[2, 2, 2, 2], groups=1, reduction=16)
return _create_senet('legacy_seresnet18', pretrained, **dict(model_args, **kwargs))
@register_model
def legacy_seresnet34(pretrained=False, **kwargs) -> SENet:
model_args = dict(
block=SEResNetBlock, layers=[3, 4, 6, 3], groups=1, reduction=16, **kwargs)
return _create_senet('legacy_seresnet34', pretrained, **model_args)
block=SEResNetBlock, layers=[3, 4, 6, 3], groups=1, reduction=16)
return _create_senet('legacy_seresnet34', pretrained, **dict(model_args, **kwargs))
@register_model
def legacy_seresnet50(pretrained=False, **kwargs) -> SENet:
model_args = dict(
block=SEResNetBottleneck, layers=[3, 4, 6, 3], groups=1, reduction=16, **kwargs)
return _create_senet('legacy_seresnet50', pretrained, **model_args)
block=SEResNetBottleneck, layers=[3, 4, 6, 3], groups=1, reduction=16)
return _create_senet('legacy_seresnet50', pretrained, **dict(model_args, **kwargs))
@register_model
def legacy_seresnet101(pretrained=False, **kwargs) -> SENet:
model_args = dict(
block=SEResNetBottleneck, layers=[3, 4, 23, 3], groups=1, reduction=16, **kwargs)
return _create_senet('legacy_seresnet101', pretrained, **model_args)
block=SEResNetBottleneck, layers=[3, 4, 23, 3], groups=1, reduction=16)
return _create_senet('legacy_seresnet101', pretrained, **dict(model_args, **kwargs))
@register_model
def legacy_seresnet152(pretrained=False, **kwargs) -> SENet:
model_args = dict(
block=SEResNetBottleneck, layers=[3, 8, 36, 3], groups=1, reduction=16, **kwargs)
return _create_senet('legacy_seresnet152', pretrained, **model_args)
block=SEResNetBottleneck, layers=[3, 8, 36, 3], groups=1, reduction=16)
return _create_senet('legacy_seresnet152', pretrained, **dict(model_args, **kwargs))
@register_model
def legacy_senet154(pretrained=False, **kwargs) -> SENet:
model_args = dict(
block=SEBottleneck, layers=[3, 8, 36, 3], groups=64, reduction=16,
downsample_kernel_size=3, downsample_padding=1, inplanes=128, input_3x3=True, **kwargs)
return _create_senet('legacy_senet154', pretrained, **model_args)
downsample_kernel_size=3, downsample_padding=1, inplanes=128, input_3x3=True)
return _create_senet('legacy_senet154', pretrained, **dict(model_args, **kwargs))
@register_model
def legacy_seresnext26_32x4d(pretrained=False, **kwargs) -> SENet:
model_args = dict(
block=SEResNeXtBottleneck, layers=[2, 2, 2, 2], groups=32, reduction=16, **kwargs)
return _create_senet('legacy_seresnext26_32x4d', pretrained, **model_args)
block=SEResNeXtBottleneck, layers=[2, 2, 2, 2], groups=32, reduction=16)
return _create_senet('legacy_seresnext26_32x4d', pretrained, **dict(model_args, **kwargs))
@register_model
def legacy_seresnext50_32x4d(pretrained=False, **kwargs) -> SENet:
model_args = dict(
block=SEResNeXtBottleneck, layers=[3, 4, 6, 3], groups=32, reduction=16, **kwargs)
return _create_senet('legacy_seresnext50_32x4d', pretrained, **model_args)
block=SEResNeXtBottleneck, layers=[3, 4, 6, 3], groups=32, reduction=16)
return _create_senet('legacy_seresnext50_32x4d', pretrained, **dict(model_args, **kwargs))
@register_model
def legacy_seresnext101_32x4d(pretrained=False, **kwargs) -> SENet:
model_args = dict(
block=SEResNeXtBottleneck, layers=[3, 4, 23, 3], groups=32, reduction=16, **kwargs)
return _create_senet('legacy_seresnext101_32x4d', pretrained, **model_args)
block=SEResNeXtBottleneck, layers=[3, 4, 23, 3], groups=32, reduction=16)
return _create_senet('legacy_seresnext101_32x4d', pretrained, **dict(model_args, **kwargs))

20
timm/models/sknet.py
View File

@ -181,8 +181,8 @@ def skresnet18(pretrained=False, **kwargs) -> ResNet:
sk_kwargs = dict(rd_ratio=1 / 8, rd_divisor=16, split_input=True)
model_args = dict(
block=SelectiveKernelBasic, layers=[2, 2, 2, 2], block_args=dict(sk_kwargs=sk_kwargs),
zero_init_last=False, **kwargs)
return _create_skresnet('skresnet18', pretrained, **model_args)
zero_init_last=False)
return _create_skresnet('skresnet18', pretrained, **dict(model_args, **kwargs))
@register_model
@ -195,8 +195,8 @@ def skresnet34(pretrained=False, **kwargs) -> ResNet:
sk_kwargs = dict(rd_ratio=1 / 8, rd_divisor=16, split_input=True)
model_args = dict(
block=SelectiveKernelBasic, layers=[3, 4, 6, 3], block_args=dict(sk_kwargs=sk_kwargs),
zero_init_last=False, **kwargs)
return _create_skresnet('skresnet34', pretrained, **model_args)
zero_init_last=False)
return _create_skresnet('skresnet34', pretrained, **dict(model_args, **kwargs))
@register_model
@ -209,8 +209,8 @@ def skresnet50(pretrained=False, **kwargs) -> ResNet:
sk_kwargs = dict(split_input=True)
model_args = dict(
block=SelectiveKernelBottleneck, layers=[3, 4, 6, 3], block_args=dict(sk_kwargs=sk_kwargs),
zero_init_last=False, **kwargs)
return _create_skresnet('skresnet50', pretrained, **model_args)
zero_init_last=False)
return _create_skresnet('skresnet50', pretrained, **dict(model_args, **kwargs))
@register_model
@ -223,8 +223,8 @@ def skresnet50d(pretrained=False, **kwargs) -> ResNet:
sk_kwargs = dict(split_input=True)
model_args = dict(
block=SelectiveKernelBottleneck, layers=[3, 4, 6, 3], stem_width=32, stem_type='deep', avg_down=True,
block_args=dict(sk_kwargs=sk_kwargs), zero_init_last=False, **kwargs)
return _create_skresnet('skresnet50d', pretrained, **model_args)
block_args=dict(sk_kwargs=sk_kwargs), zero_init_last=False)
return _create_skresnet('skresnet50d', pretrained, **dict(model_args, **kwargs))
@register_model
@ -235,6 +235,6 @@ def skresnext50_32x4d(pretrained=False, **kwargs) -> ResNet:
sk_kwargs = dict(rd_ratio=1/16, rd_divisor=32, split_input=False)
model_args = dict(
block=SelectiveKernelBottleneck, layers=[3, 4, 6, 3], cardinality=32, base_width=4,
block_args=dict(sk_kwargs=sk_kwargs), zero_init_last=False, **kwargs)
return _create_skresnet('skresnext50_32x4d', pretrained, **model_args)
block_args=dict(sk_kwargs=sk_kwargs), zero_init_last=False)
return _create_skresnet('skresnext50_32x4d', pretrained, **dict(model_args, **kwargs))