Alibaba-NLP/HiAGM

Issues

• rcv1 sample replaced with the original data but not getting result.

  #7 opened 3 years ago by Arousha
  6

• 关于使用WOS数据集复现结果差的问题

  #17 opened 5 months ago by supersomnr
  0

• 作者您好，看了您的论文收获颇大！感谢您提供代码学习，请问Note that first change the Root.child List怎么操作，有点点懵

  #16 opened 6 months ago by gunquxuexia
  0

• the code displays the not good results unlike the paper

  #13 opened 2 years ago by Harrison-cc
  1

• How to get the vocab files of wos dataset?

  #6 opened 3 years ago by liug0324
  1

• glove.6B.300d.txt位置问题

  #11 opened 2 years ago by yangyuanjiao
  4

• 复现效果较差，跟文中效果不一致

  #9 opened 3 years ago by jiahenghuang
  2

• HIAGM-LA中的label dim是不是应该设置为100，设置为100后面的label attention计算的代码才合理

  #15 opened a year ago by ydwasd
  0

• HIAGM的LA中注意力计算代码部分和论文描述的不太相符

  #14 opened a year ago by ydwasd
  0

• Common samples across train/test splits in NYT dataset?

  #10 opened 2 years ago by DushyantaDhyani
  0

• AttributeError: 'Tree' object has no attribute 'bottom_up_state'

  #8 opened 3 years ago by coco11563
  0

• Missing "../glove.6B/glove.6B.300d.txt" file.

  #5 opened 3 years ago by HongtaoYang
  1

• 如果用transformer encoder

  #4 opened 3 years ago by ZTurboX
  0

• HiAGM-TP中的M矩阵

  #3 opened 4 years ago by moluchase
  0

• when train HiAGM-LA mode, raise error

  #2 opened 4 years ago by hdulbj
  1

• cannot achieve good performance on dev and test dataset when running the train.py script

  #1 opened 4 years ago by DanqingZ
  2