How to mask a multi-head attention layer?
I’m trying to make a Transformer model that could recibe sequences of a variable lenght. But I can’t
I’m trying to make a Transformer model that could recibe sequences of a variable lenght. But I can’t