pytorch加载模型报错Unexpected key(s) in state_dict: module.conv1.weight, module.bn1

代码

import models

arch = 'resnet50'
model = models.__dict__[arch]()

checkpoint = torch.load(ckptFile)
model.load_state_dict(checkpoint['state_dict'])

model = torch.nn.DataParallel(model).cuda()

报错

Traceback (most recent call last):
  File "/home/user1/project1/utils/eval.py", line 193, in <module>
    test_models(r'/home/user1/models')
  File "/home/user1/project1/utils/eval.py", line 157, in test_models
    model.load_state_dict(checkpoint['state_dict'])
  File "/home/user1/miniconda3/lib/python3.7/site-packages/torch/nn/modules/module.py", line 777, in load_state_dict
    self.__class__.__name__, "\n\t".join(error_msgs)))
RuntimeError: Error(s) in loading state_dict for ResNet:
	Missing key(s) in state_dict: "conv1.weight", "bn1.weight", "bn1.bias", ...". 
	Unexpected key(s) in state_dict: "module.conv1.weight", "module.bn1.weight", "module.bn1.bias"...

原因

哥们，你顺序搞错了，应该先把模型放到GPU上，然后再开始load参数

解决

正确的顺序

import models

arch = 'resnet50'
model = models.__dict__[arch]()

model = torch.nn.DataParallel(model).cuda()

checkpoint = torch.load(ckptFile)
model.load_state_dict(checkpoint['state_dict'])