CNN training multilabel clasiffication error

1

Hi,

I am getting an error while training this CNN, the inputs are images and the labes are one-hot encoded vectors, and I am getting the following error when training the CNN

Thanks for the help in advance

image
image2777×1505 418 KB

2

nn.CrossEntropyLoss expects the target as a LongTensor and to have the shape [batch_size] containing the class indices in the range [0, nb_classes].
Since your target is one-hot encoded, you could use:

target = torch.argmax(target, dim=1)`

to create the expected shape and type.

PS: It’s better to post code snippets by wrapping them into three backticks ``` instead of screenshots. :wink: