RuntimeError: Input type (torch.cuda.FloatTensor) and weight type (torch.FloatTensor) should be the same.已经解决

显示这里有错误：

model = nn.Sequential(*list(trained_model.children())[:-1], #[b, 512, 1, 1]  取前17层
                      Flatten(), # [b, 512, 1, 1] => [b, 512]
                      nn.Linear(9216, 3)
                      )

仔细看了一下，原来没有把模型to到cuda上，那么在末尾加上.to(torch.device('cuda'))就可以，即：

model = nn.Sequential(*list(trained_model.children())[:-1], #[b, 512, 1, 1]  取前17层
                      Flatten(), # [b, 512, 1, 1] => [b, 512]
                      nn.Linear(9216, 3)
                      ).to(device)

问题解决。