Explanation of variable comments

[AADeLucia]

I am having issues aligning the variables in the code to their respective meanings, mostly N in models.py.

Can you provide more information / references to the paper? Here is my interpretation:

• N: number of bags (i.e., batch size)
• K: number of instances in a bag (i.e., embeddings)

However, that doesn't make any sense with these lines from the Attention class:

        H = self.feature_extractor_part2(H)  # NxL

        A = self.attention(H)  # NxK
        A = torch.transpose(A, 1, 0)  # KxN
        A = F.softmax(A, dim=1)  # softmax over N

        M = torch.mm(A, H)  # KxL

because H = self.feature_extractor_part2(H) returns a KxL matrix ([13, 500]), no N involved.

Am I misinterpreting the meaning of N?