Closed vieting closed 2 years ago
I have a case where the output of a module is not a tensor. A simple example could be:
class DummyModule(torch.nn.Module): def forward(self, y): return {"a": 1, "b": y + 1}
I'm not exactly sure about all the logic in CallEntry, but just focusing on the Tensor outputs seems to work here.
CallEntry
Tensor
I have a case where the output of a module is not a tensor. A simple example could be:
I'm not exactly sure about all the logic in
CallEntry
, but just focusing on theTensor
outputs seems to work here.