Cao_enjun

NAN source?

1

https://github.com/abisee/pointer-generator/blob/b29e986f24fdd01a6b6d6008187c5c887f0be282/attention_decoder.py#L101

A data set with 1300K pairs of articles and abstracts!

1

Hi, see! I am working on a new data set with 6 times bigger than your CNN data set . Whether need I to add more parameters of the model...

Difference in paper and codes.

1

https://github.com/abisee/pointer-generator/blob/a7317f573d01b944c31a76bde7218bcfc890ef6a/attention_decoder.py#L173 Hi, abisee! I am confused about the codes in this line. In your paper, the function is V[s_t, h*_t] + b But in the codes, you use rnn_output instead....

Hi, abisee! Thanks for sharing this good work! I have some question about your paper in Experiments part. You said that with the coverage loss ratio set to 1.0 the...

Hi , zhiguowang! I found that you use dropout in training and when testing with out dropout. When the model is trained, add dropout. If the test is not added,...

why not use prob after softmax?

1

https://github.com/zhiguowang/BiMPM/blob/33cc8fe5d450f432a6843bc05cad29c6ce9f5714/src/SentenceMatchModelGraph.py#L158 hi zhiguowang ! why not use prob after softmax?

cat(): argument 'tensors' (position 1) must be tuple of Tensors, not generator

3

https://github.com/lanwuwei/SPM_toolkit/blob/3e2cb35681e9f31bfcc66afde2159bbf394056df/DecAtt/main_quora.py#L56 cat(): argument 'tensors' (position 1) must be tuple of Tensors, not generator can I fix this bug use this line? torch.cat(tuple([dict[word].view(1, -1) for word in lsent]))

hello, I have some questions about dropout in VAT. If I use dropout in VAT, the output distribution will change even without perturbation. thanks!

https://github.com/thinline72/toxic/blob/c4f50a0560d312da8ab60e2492a2e72c54cc3f55/sergeif/bigru_focal_loss.py#L52 Thanks for sharing this good work ! I have some problems with this line, how can I generate this file ? @ifserge I will appreciate for your help !

您好老师，想请教一下，在做序列标注任务时，标注序列Y是我们可以获得的，那么为什么不能用极大似然估计或者贝叶斯估计那样直接求，而是把Y看作隐变量用em算法呢？直接把Y看作分类标签作为观测数据是否合理？