Add all the latest models to hubconf (#2189)

* Add all the latest models to hubconf

* remove detection models from hubconf

* fix link error

hubconf.py

 # Optional list of dependencies required by the package
 dependencies = ['torch']
 
+# classification
 from torchvision.models.alexnet import alexnet
 from torchvision.models.densenet import densenet121, densenet169, densenet201, densenet161
 from torchvision.models.inception import inception_v3
@@ -8,7 +9,12 @@ from torchvision.models.resnet import resnet18, resnet34, resnet50, resnet101, r
     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
 from torchvision.models.googlenet import googlenet
 from torchvision.models.shufflenetv2 import shufflenet_v2_x0_5, shufflenet_v2_x1_0
 from torchvision.models.mobilenet import mobilenet_v2
+from torchvision.models.mnasnet import mnasnet0_5, mnasnet0_75, mnasnet1_0, \
+    mnasnet1_3
+
+# segmentation
+from torchvision.models.segmentation import fcn_resnet50, fcn_resnet101, \
+    deeplabv3_resnet50, deeplabv3_resnet101