Can I create a CNN sentence classifier without standardizing (padding) a sentence length?

nlp
2

I’m unsure if you are asking about the pooling operation creating the outputs in a fixed size, but it so you could use nn.AdaptiveMaxPool2d and define the desired output size.