a deep cascade model for multi-document reading comprehension
1.先排序检索，后精过滤排序
2. 先 document paragraph rank, 再 model-rank
3.model-rank, self-attention,
embedding= embedding+cnn-embedding
a deep cascade model for multi-document reading comprehension 阅读理解
猜你喜欢
转载自blog.csdn.net/qq_34146899/article/details/95047249
今日推荐
周排行