然后用这个embedding和直接embedding的word vector拼到一起,放到第二个LSTM As far as I know, PyTorch does not inherently have masked tensor operations (such as ...
確定! 回上一頁