Sequence to sequence入门详解:从RNN, LSTM到Encoder-Decoder, Attention, transformer
文章目录1.前馈神经网络的缺点2.循环神经网络RNN2.1.RNN的基本结构与数学定义2.2.输入输出长度的讨论2.2.1.nx=ny=nn_x=n_y=nnx=ny=n2.2.2.nx=n,ny=1n_x=n,n_y=1nx=n,ny=12.2.3.nx=1,ny=nn_x=1,n_y=nnx=1,ny=n2.2.4.nx=n,ny=mn_x=n,n_y=mnx=n,ny=m,Encoder-D