pytorch 获取层权重,对特定层注入hook, 提取中间层输出的方法
如下所示：
#获取模型权重
fork,vinmodel_2.state_dict().iteritems():
print("Layer{}".format(k))
print(v)
#获取模型权重 forlayerinmodel_2.modules(): ifisinstance(layer,nn.Linear): print(layer.weight)
#将一个模型权重载入另一个模型
model=VGG(make_layers(cfg['E']),**kwargs)
ifpretrained:
load=torch.load('/home/huangqk/.torch/models/vgg19-dcbb9e9d.pth')
load_state={k:vfork,vinload.items()ifknotin['classifier.0.weight','classifier.0.bias','classifier.3.weight','classifier.3.bias','classifier.6.weight','classifier.6.bias']}
model_state=model.state_dict()
model_state.update(load_state)
model.load_state_dict(model_state)
returnmodel
#对特定层注入hook defhook_layers(model): defhook_function(module,inputs,outputs): recreate_image(inputs[0]) print(model.features._modules) first_layer=list(model.features._modules.items())[0][1] first_layer.register_forward_hook(hook_function)
#获取层 x=someinput forlinvgg.features.modules(): x=l(x) modulelist=list(vgg.features.modules()) forlinmodulelist[:5]: x=l(x) keep=x forlinmodulelist[5:]: x=l(x)
#提取vgg模型的中间层输出
#coding:utf8
importtorch
importtorch.nnasnn
fromtorchvision.modelsimportvgg16
fromcollectionsimportnamedtuple
classVgg16(torch.nn.Module):
def__init__(self):
super(Vgg16,self).__init__()
features=list(vgg16(pretrained=True).features)[:23]
#features的第3，8，15，22层分别是:relu1_2,relu2_2,relu3_3,relu4_3
self.features=nn.ModuleList(features).eval()
defforward(self,x):
results=[]
forii,modelinenumerate(self.features):
x=model(x)
ifiiin{3,8,15,22}:
results.append(x)
vgg_outputs=namedtuple("VggOutputs",['relu1_2','relu2_2','relu3_3','relu4_3'])
returnvgg_outputs(*results)
以上这篇pytorch获取层权重,对特定层注入hook,提取中间层输出的方法就是小编分享给大家的全部内容了，希望能给大家一个参考，也希望大家多多支持毛票票。