What's the difference between one2one(Or CopyRNN*) and CopyRNN(meng deep keyphrase generation)?Are their models the same?If the model used by two are same, why result different? #10
They are the same model/architecture. The original is implemented with Theano, and there might be other factors affecting the results we reproduce (learning rate etc.).
They are the same model/architecture. The original is implemented with Theano, and there might be other factors affecting the results we reproduce (learning rate etc.).