def _resnet(arch, block, layers, pretrained, progress, **kwargs):
model = ResNet(block, layers, **kwargs)
if pretrained:
state_dict = load_state_dict_from_url(model_urls[arch],
progress=progress)
load_pretrained_params(model, state_dict)
return model
def l_resnext101(pretrained, baseWidth, cardinality):
"""
Construct ResNeXt-101.
"""
model = ResNeXt(baseWidth, cardinality, [3, 4, 23, 3], 1000)
if pretrained:
pretrained_dict = model_zoo.load_url(model_urls["resnext101"])
load_pretrained_params(model, pretrained_dict)
return model
def l_resnet101(pretrained=False, **kwargs):
"""Constructs a ResNet-101 model.
Args:
pretrained (bool): If True, returns a model pre-trained on ImageNet
"""
model = ResNet(Bottleneck, [3, 4, 23, 3], **kwargs)
if pretrained:
pretrained_dict = model_zoo.load_url(model_urls["resnet101"])
load_pretrained_params(model, pretrained_dict)
return model
def _load_model(backbone, pretrained, progress, in_channel, num_classes,
replace_stride_with_dilation):
model = _segm_resnet(backbone, in_channel, num_classes,
replace_stride_with_dilation)
if pretrained:
arch = "deeplabv3_" + backbone + "_coco"
model_url = model_urls[arch]
if model_url is None:
raise NotImplementedError(
"pretrained {} is not supported as of now".format(arch))
else:
print(f"Loading parameters from {model_url}")
state_dict = load_state_dict_from_url(model_url, progress=progress)
load_pretrained_params(model, state_dict)
return model
def Res_Deeplab(pretrained: bool = True) -> nn.Module:
model = ResNet(Bottleneck, [3, 4, 23, 3])
if pretrained:
pretrained_dict = torch.load(model_urls["res_deeplabv3"])
load_pretrained_params(model, pretrained_dict["model"])
return model