ningshixian

Results 9 comments of ningshixian

What do you specifically mean?

you can do it like this, ``` output = Bidirectional(LSTM()) output = AttentionDecoder()(output) output = TimeDistributed(Dense(num_class))(output) output = crf(output) ```

我认为AttentionDecoder的输入和输出的time_step数量一般情况是一样的! 不然无法对输入进行加权和操作吧? 另外,你说问题和答案词向量不同,这个没有关系的,可以通过不同大小的权重矩阵W来合并

CudnnRNN需要在GPU环境下才能跑,可以改为RNN/LSTM/GRU在CPU环境下运行

> 多谢!祝万事如意 conll 2003 命名实体中最常见的公开数据集,总共四类实体:persons, locations, organizations ,miscellaneous entities,加上O标签,共5个,哪来的9标签?详情参考官网: https://www.clips.uantwerpen.be/conll2003/ner/

> Hi, I solved this by pip uninstall transformer-engine nice! it works

> 我也遇到同样的问题 pip install scikit-learn==1.0.2 ,已解决