关于#pytorch tensor#的问题，如何解决？

我在跑pytorch代码的时候遇到了这样的问题
代码：
model_params_file = open('model_param_best_%sway%sshot.pkl' % (3, 5), 'rb')
model_params = pickle.load(model_params_file)
model_params_file.close()
meta = MetaLearner(args).to(device)
meta.net.vars = [torch.tensor(item, requires_grad=False) for item in model_params]

报错：
File "E:/meta-learning/maml_classification/test.py", line 15, in main
meta.net.vars = [torch.tensor(item, requires_grad=False) for item in model_params]
File "E:\anaconda\envs\pytorch\lib\site-packages\torch\nn\modules\module.py", line 627, in setattr
.format(torch.typename(value), name))
TypeError: cannot assign 'list' as child module 'vars' (torch.nn.Module or None expected)

请问有人会吗？最近再做这个实验，调了一天都没调出来。