gaozhiguang

Results 17 comments of gaozhiguang

您好,您在上面说的意思是这里作者并没有实现copy机制吗?就是指的pointer network论文里面的copy机制。 @HitAgain

> 你是使用了character而非word,并且删去了decoder扩充字典这一操作,是这样吗? 您说的character是指的英文字母吗?还是中文里的字。另外,这里有实现copy机制吗?就是与原论文pointer network一致的copy机制。因为最近赶论文,想找个transformer copy机制的实现快速套用,您知道有哪些好的资料吗?

> > > 你是使用了character而非word,并且删去了decoder扩充字典这一操作,是这样吗? > > > > > > 您说的character是指的英文字母吗?还是中文里的字。另外,这里有实现copy机制吗?就是与原论文pointer network一致的copy机制。因为最近赶论文,想找个transformer copy机制的实现快速套用,您知道有哪些好的资料吗? > > 我说的是中文汉字的字,然后这里可能缺少了扩充字典的操作,即将oov的词在其出现之后扩充到字典中,但是复制本身还是可以正常运作的。简单来说可以实现复制,但是不太能解决oov问题,HitAgain也说了,这里用的是字,oov现象可能本身就不太常见。 您说的扩充字典是什么意思呢?copy机制实现的应该是一个动态的词表吧,碰到oov的单词或字只需要暂时的扩充词表吧

> > > 你是使用了character而非word,并且删去了decoder扩充字典这一操作,是这样吗? > > > > > > 您说的character是指的英文字母吗?还是中文里的字。另外,这里有实现copy机制吗?就是与原论文pointer network一致的copy机制。因为最近赶论文,想找个transformer copy机制的实现快速套用,您知道有哪些好的资料吗? > > 我最近也需要找一个Transformer+Pointer的模型急用,不知道你有找到吗?或者可以交流一下? 已经在邮件中进行了回复,望知悉。

Hi, have you being sucessful in predict the single line, i am recently working on my graduation project which requiring a display system of automatic summarization.

As a fact, I also encountered such a problem. Just like the method mentioned above, I used the gold summary to evaluate and got the following result: ***** Eval results...

Hi, thanks. but i am still a little confused about what is a PTB tokenizer? the code in train_utils.py uses the stanford corenlp for tokenizer. In addition, for some case,...

Hi, this will cause an error, from pycorenlp import StanfordCoreNLP nlp = StanfordCoreNLP('http://localhost:9000') line="teams are scouring the depths of a remote part of the southern indian ocean . so far...

Hi, i have another try, when i set: line = "teams are scouring the depths of a remote part of the southern indian ocean . so far , they 've...

> The instructions are included in the 'Running pre-trained factuality models' section of the readme. Set $MODEL_TYPE to 'electra_sentence'. $INPUT_DIR should point to the location of the model. > >...