What's the difference between `torch.nn.functional` and `torch.nn`?

[Xixi]

Correct me if I’m wrong. Seems like when using nn.functional net.parameters() won’t find the parameters/weight. you need to specify them explicity