mirrors/pytorch-image-models
1
0
Fork 0
mirror of https://github.com/huggingface/pytorch-image-models.git synced 2025-06-03 15:01:08 +08:00
Code Issues Projects Releases Wiki Activity

Adding some configs to sknet, incl ResNet50 variants from 'Compounding ... Assembled Techniques' paper and original SKNet50

Browse Source
This commit is contained in:
Ross Wightman 2020-02-01 23:28:48 -08:00
parent a9d2424fd1
commit 7d07ebb660
1 changed files with 76 additions and 20 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

96
timm/models/sknet.py
View File

@ -22,7 +22,10 @@ def _cfg(url='', **kwargs):
default_cfgs = {
'skresnet18': _cfg(url=''),
'skresnet26d': _cfg()
'skresnet26d': _cfg(),
'skresnet50': _cfg(),
'skresnet50d': _cfg(),
'skresnext50_32x4d': _cfg(),
}
@ -131,24 +134,6 @@ class SelectiveKernelBottleneck(nn.Module):
return x
@register_model
def skresnet26d(pretrained=False, num_classes=1000, in_chans=3, **kwargs):
"""Constructs a ResNet-26 model.
"""
default_cfg = default_cfgs['skresnet26d']
sk_kwargs = dict(
keep_3x3=False,
)
model = ResNet(
SelectiveKernelBottleneck, [2, 2, 2, 2], stem_width=32, stem_type='deep', avg_down=True,
num_classes=num_classes, in_chans=in_chans, block_args=dict(sk_kwargs=sk_kwargs),
**kwargs)
model.default_cfg = default_cfg
if pretrained:
load_pretrained(model, default_cfg, num_classes, in_chans)
return model
@register_model
def skresnet18(pretrained=False, num_classes=1000, in_chans=3, **kwargs):
"""Constructs a ResNet-18 model.
@ -181,4 +166,75 @@ def sksresnet18(pretrained=False, num_classes=1000, in_chans=3, **kwargs):
model.default_cfg = default_cfg
if pretrained:
load_pretrained(model, default_cfg, num_classes, in_chans)
return model
return model
@register_model
def skresnet26d(pretrained=False, num_classes=1000, in_chans=3, **kwargs):
"""Constructs a ResNet-26 model.
"""
default_cfg = default_cfgs['skresnet26d']
sk_kwargs = dict(
keep_3x3=False,
)
model = ResNet(
SelectiveKernelBottleneck, [2, 2, 2, 2], stem_width=32, stem_type='deep', avg_down=True,
num_classes=num_classes, in_chans=in_chans, block_args=dict(sk_kwargs=sk_kwargs),
**kwargs)
model.default_cfg = default_cfg
if pretrained:
load_pretrained(model, default_cfg, num_classes, in_chans)
return model
@register_model
def skresnet50(pretrained=False, num_classes=1000, in_chans=3, **kwargs):
"""Constructs a Select Kernel ResNet-50 model.
Based on config in "Compounding the Performance Improvements of Assembled Techniques in a
Convolutional Neural Network"
"""
sk_kwargs = dict(
attn_reduction=2,
)
default_cfg = default_cfgs['skresnet50']
model = ResNet(
SelectiveKernelBottleneck, [3, 4, 6, 3], num_classes=num_classes, in_chans=in_chans,
block_args=dict(sk_kwargs=sk_kwargs), **kwargs)
model.default_cfg = default_cfg
if pretrained:
load_pretrained(model, default_cfg, num_classes, in_chans)
return model
@register_model
def skresnet50d(pretrained=False, num_classes=1000, in_chans=3, **kwargs):
"""Constructs a Select Kernel ResNet-50-D model.
Based on config in "Compounding the Performance Improvements of Assembled Techniques in a
Convolutional Neural Network"
"""
sk_kwargs = dict(
attn_reduction=2,
)
default_cfg = default_cfgs['skresnet50d']
model = ResNet(
SelectiveKernelBottleneck, [3, 4, 6, 3], stem_width=32, stem_type='deep', avg_down=True,
num_classes=num_classes, in_chans=in_chans, block_args=dict(sk_kwargs=sk_kwargs), **kwargs)
model.default_cfg = default_cfg
if pretrained:
load_pretrained(model, default_cfg, num_classes, in_chans)
return model
@register_model
def skresnext50_32x4d(pretrained=False, num_classes=1000, in_chans=3, **kwargs):
"""Constructs a Select Kernel ResNeXt50-32x4d model. This should be equivalent to
the SKNet50 model in the Select Kernel Paper
"""
default_cfg = default_cfgs['skresnext50_32x4d']
model = ResNet(
SelectiveKernelBottleneck, [3, 4, 6, 3], cardinality=32, base_width=4,
num_classes=num_classes, in_chans=in_chans, **kwargs)
model.default_cfg = default_cfg
if pretrained:
load_pretrained(model, default_cfg, num_classes, in_chans)
return model