Calculating MSE loss on two tensors of the same shape
I am trying to do a classification task but due to reasons I needed to delete the softmax and replace the loss module from cross entropy to MSE now to create a one hot tensor for the labels (target) I do the following: