lancopku/SGM

Issues

• 梯度没有下降

  #24 opened 5 years ago by ErxinYu
  10

• 如果要训练中文的模型有对应的预训练模型吗？

  #35 opened 10 months ago by xyxhchb
  0

• AAPD 数据集无法下载

  #25 opened 5 years ago by cjopengler
  2

• 关于运行效果到不到论文中进度的问题

  #31 opened 4 years ago by NLPER-luo
  13

• Why reverse the source sequence in the collate_function

  #34 opened 3 years ago by VRCMF
  0

• 使用AAPD数据集，验证时，微F1一直为1,看acc才60多。能够提供处理好的AAPD数据集呢？

  #33 opened 3 years ago by 20161105421
  0

• 请问在decode中前向传播的时候，为什么rnn的输入是标签向量和state，而没有经过attention后得到的c(t)呢，没看懂

  #16 opened 6 years ago by HaimianYu
  5

• About train/test split of RCV1

  #32 opened 4 years ago by yzhangcs
  0

• 关于论文图1

  #30 opened 4 years ago by Decalogue
  0

• The text in dataset seems no meaning

  #26 opened 5 years ago by MaYB2333
  3

• beam设置的问题

  #29 opened 4 years ago by terminator123
  0

• AAPD Dataset

  #28 opened 4 years ago by rajcscw
  1

• config.yaml是SGM的配置文件还是SGM+GE的配置文件？

  #21 opened 6 years ago by WrRan
  1

• 添加license

  #27 opened 4 years ago by shenlanyilang
  0

• 请问用preprocess处理后的训练数据集的标签有没有按照标签的频率排序呢

  #18 opened 6 years ago by HaimianYu
  2

• The size of tensor a (576) must match the size of tensor b (64) at non-singleton dimension 1

  #12 opened 5 years ago by mochp
  14

• 运行了你的代码，设置了GPU，但是GPU显示存在任务，但没有占用GPU，gpu空转

  #23 opened 5 years ago by cm-xcju
  1

• 运行preprocess.py时出错

  #10 opened 6 years ago by hanyegirl
  2

• 想问下关于评价的问题

  #22 opened 5 years ago by Canadalynx
  1

• why sorting matters?

  #7 opened 6 years ago by eveliao
  2

• 解码的时候LSTM输入的问题

  #17 opened 6 years ago by HaimianYu
  1

• 请问能不能提供一下基准模型的代码呢

  #19 opened 6 years ago by HaimianYu
  0

• Where is the tgt seq in loss computation?

  #9 opened 6 years ago by Decalogue
  1

• dataset words repeat so many times.

  #11 opened 6 years ago by Langjiaqi
  6

• i cant open the Google,can not get the dataset

  #8 opened 6 years ago by hanyegirl
  7

• Definition of loss function

  #1 opened 6 years ago by JianWenJun
  0

• RuntimeError: While copying the parameter named decoder.embedding.weight, whose dimensions in the model are torch.Size([58, 256]) and whose dimensions in the checkpoint are torch.Size([107, 256]).

  #3 opened 6 years ago by SeekPoint
  1

• What is the format of label_test?

  #5 opened 6 years ago by linxid
  1

• 预测为 label 序列，例如 MCAT ECAT M13 M131 E12

  #4 opened 6 years ago by Decalogue
  6

• 请问词向量是随机初始化的吗

  #15 opened 6 years ago by HaimianYu
  1

• 论文中的两个小疑问

  #14 opened 6 years ago by HaimianYu
  1

• Attention layer

  #13 opened 6 years ago by nooralahzadeh
  0

• does the order of predicted labels matter? and what if the length of prediction is not the same as ground trut?

  #6 opened 6 years ago by eveliao
  1

• Prune function in Dict.py may cause index out of range

  #2 opened 6 years ago by vicety
  1