Send nn.parameter to optimizer

KaiyangZhou / Dassl.pytorch

A PyTorch toolbox for domain generalization, domain adaptation and semi-supervised learning.

MIT License

1.21k stars 169 forks source link

Send nn.parameter to optimizer #67

Open Presleyliao opened 4 months ago

Presleyliao commented 4 months ago

Hello, thank you for your great work!

I have an issue when I want to send nn.parameter to the optimizer, the error is as follow: TypeError: params argument given to the optimizer should be an iterable of Tensors or dicts, but got torch.cuda.FloatTensor

Thank you!