pytorch 获取层权重，对特定层注入hook， 提取中间层输出

#获取模型权重
for k, v in model_2.state_dict().iteritems():
    print("Layer {}".format(k))
    print(v)

#获取模型权重
for layer in model_2.modules():
   if isinstance(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)
if pretrained:
    load = torch.load('/home/huangqk/.torch/models/vgg19-dcbb9e9d.pth')
    load_state = {k: v for k, v in load.items() if k not in ['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)
return model

# 对特定层注入hook
def hook_layers(model):
    def hook_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
for l in vgg.features.modules():
  x = l(x)
modulelist = list(vgg.features.modules())
for l in modulelist[:5]:
  x = l(x)
keep = x
for l in modulelist[5:]:
  x = l(x)

# 提取vgg模型的中间层输出
# coding:utf8
import torch
import torch.nn as nn
from torchvision.models import vgg16
from collections import namedtuple


class Vgg16(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()

    def forward(self, x):
        results = []
        for ii, model in enumerate(self.features):
            x = model(x)
            if ii in {3, 8, 15, 22}:
                results.append(x)

        vgg_outputs = namedtuple("VggOutputs", ['relu1_2', 'relu2_2', 'relu3_3', 'relu4_3'])
        return vgg_outputs(*results)