Add pretrained Wide ResNet (#912)

* add wide resnet

* add docstring for wide resnet

* update WRN-50-2 model

* add docs

* extend WRN docstring

* use pytorch storage for WRN

* fix rebase

* fix typo in docs

docs/source/models.rst

@@ -24,6 +24,7 @@ architectures for image classification:
 -  `ShuffleNet`_ v2
 -  `MobileNet`_ v2
 -  `ResNeXt`_
+-  `Wide ResNet`_
 -  `MNASNet`_
 
 You can construct a model with random weights by calling its constructor:
@@ -41,6 +42,7 @@ You can construct a model with random weights by calling its constructor:
     shufflenet = models.shufflenet_v2_x1_0()
     mobilenet = models.mobilenet_v2()
     resnext50_32x4d = models.resnext50_32x4d()
+    wide_resnet50_2 = models.wide_resnet50_2()
     mnasnet = models.mnasnet1_0()
 
 We provide pre-trained models, using the PyTorch :mod:`torch.utils.model_zoo`.
@@ -59,6 +61,7 @@ These can be constructed by passing ``pretrained=True``:
     shufflenet = models.shufflenet_v2_x1_0(pretrained=True)
     mobilenet = models.mobilenet_v2(pretrained=True)
     resnext50_32x4d = models.resnext50_32x4d(pretrained=True)
+    wide_resnet50_2 = models.wide_resnet50_2(pretrained=True)
     mnasnet = models.mnasnet1_0(pretrained=True)
 
 Instancing a pre-trained model will download its weights to a cache directory.
@@ -114,6 +117,8 @@ ShuffleNet V2                     30.64           11.68
 MobileNet V2                      28.12           9.71
 ResNeXt-50-32x4d                  22.38           6.30
 ResNeXt-101-32x8d                 20.69           5.47
+Wide ResNet-50-2                  21.49           5.91
+Wide ResNet-101-2                 21.16           5.72
 MNASNet 1.0                       26.49           8.456
 ================================  =============   =============
 
@@ -202,6 +207,12 @@ ResNext
 .. autofunction:: resnext50_32x4d
 .. autofunction:: resnext101_32x8d
 
+Wide ResNet
+-----------
+
+.. autofunction:: wide_resnet50_2
+.. autofunction:: wide_resnet101_2
+
 MNASNet
 --------
 

hubconf.py

@@ -5,7 +5,7 @@ from torchvision.models.alexnet import alexnet
 from torchvision.models.densenet import densenet121, densenet169, densenet201, densenet161
 from torchvision.models.inception import inception_v3
 from torchvision.models.resnet import resnet18, resnet34, resnet50, resnet101, resnet152,\
-    resnext50_32x4d, resnext101_32x8d
+    resnext50_32x4d, resnext101_32x8d, wide_resnet50_2, wide_resnet101_2
 from torchvision.models.squeezenet import squeezenet1_0, squeezenet1_1
 from torchvision.models.vgg import vgg11, vgg13, vgg16, vgg19, vgg11_bn, vgg13_bn, vgg16_bn, vgg19_bn
 from torchvision.models.segmentation import fcn_resnet101, deeplabv3_resnet101

torchvision/models/resnet.py

@@ -3,7 +3,8 @@ from .utils import load_state_dict_from_url
 
 
 __all__ = ['ResNet', 'resnet18', 'resnet34', 'resnet50', 'resnet101',
-           'resnet152', 'resnext50_32x4d', 'resnext101_32x8d']
+           'resnet152', 'resnext50_32x4d', 'resnext101_32x8d',
+           'wide_resnet50_2', 'wide_resnet101_2']
 
 
 model_urls = {
@@ -14,6 +15,8 @@ model_urls = {
     'resnet152': 'https://download.pytorch.org/models/resnet152-b121ed2d.pth',
     'resnext50_32x4d': 'https://download.pytorch.org/models/resnext50_32x4d-7cdf4587.pth',
     'resnext101_32x8d': 'https://download.pytorch.org/models/resnext101_32x8d-8ba56ff5.pth',
+    'wide_resnet50_2': 'https://download.pytorch.org/models/wide_resnet50_2-95faca4d.pth',
+    'wide_resnet101_2': 'https://download.pytorch.org/models/wide_resnet101_2-32ee1156.pth',
 }
 
 
@@ -294,3 +297,37 @@ def resnext101_32x8d(pretrained=False, progress=True, **kwargs):
     kwargs['width_per_group'] = 8
     return _resnet('resnext101_32x8d', Bottleneck, [3, 4, 23, 3],
                    pretrained, progress, **kwargs)
+
+
+def wide_resnet50_2(pretrained=False, progress=True, **kwargs):
+    """Constructs a Wide ResNet-50-2 model.
+
+    The model is the same as ResNet except for the bottleneck number of channels
+    which is twice larger in every block. The number of channels in outer 1x1
+    convolutions is the same, e.g. last block in ResNet-50 has 2048-512-2048
+    channels, and in Wide ResNet-50-2 has 2048-1024-2048.
+
+    Args:
+        pretrained (bool): If True, returns a model pre-trained on ImageNet
+        progress (bool): If True, displays a progress bar of the download to stderr
+    """
+    kwargs['width_per_group'] = 64 * 2
+    return _resnet('wide_resnet50_2', Bottleneck, [3, 4, 6, 3],
+                   pretrained, progress, **kwargs)
+
+
+def wide_resnet101_2(pretrained=False, progress=True, **kwargs):
+    """Constructs a Wide ResNet-101-2 model.
+
+    The model is the same as ResNet except for the bottleneck number of channels
+    which is twice larger in every block. The number of channels in outer 1x1
+    convolutions is the same, e.g. last block in ResNet-50 has 2048-512-2048
+    channels, and in Wide ResNet-50-2 has 2048-1024-2048.
+
+    Args:
+        pretrained (bool): If True, returns a model pre-trained on ImageNet
+        progress (bool): If True, displays a progress bar of the download to stderr
+    """
+    kwargs['width_per_group'] = 64 * 2
+    return _resnet('wide_resnet101_2', Bottleneck, [3, 4, 23, 3],
+                   pretrained, progress, **kwargs)