sheffieldnlp/stance-conditional

Issues

• Replicating the results.

  #21 opened 5 years ago by shubham068
  0

• generating stance results with new dataset?

  #20 opened 6 years ago by Howiezhu
  1

• Version issue

  #19 opened 6 years ago by km1994
  1

• cannot read the son file when do word2vec training

  #17 opened 6 years ago by eckolemon
  1

• can not find automatic labeling method

  #18 opened 6 years ago by sxhmilyoyo
  1

• cannot download corpus

  #16 opened 7 years ago by eckolemon
  2

• Attention

  #2 opened 7 years ago by rockt
  4

• fullp variable

  #15 opened 8 years ago by andreasvlachos
  0

• File not found

  #14 opened 8 years ago by andreasvlachos
  5

• problem with imports

  #13 opened 8 years ago by andreasvlachos
  4

• downloaded_Donald_Trump.txt

  #12 opened 8 years ago by andreasvlachos
  1

• Missing twokenize_wrapper module?

  #11 opened 8 years ago by andreasvlachos
  2

• Separate embeddings matrices for tweet and target

  #10 opened 8 years ago by isabelleaugenstein
  0

• Bi-directional conditional encoding

  #1 opened 8 years ago by rockt
  2

• Feeding target representation during tweet processing

  #3 opened 8 years ago by rockt
  5

• Stopping criterion for training

  #6 opened 8 years ago by isabelleaugenstein
  0

• Alternative accuracy hook that ignores NONE

  #7 opened 8 years ago by isabelleaugenstein
  0

• Make sure that initialized parameters are small (e.g. between -0.1 and 0.1) to avoid extreme predictions in the beginning

  #9 opened 8 years ago by rockt
  0

• Set a random seed in TF for reproducible runs

  #8 opened 8 years ago by rockt
  1

• Add parameter for changing loss for NONE

  #4 opened 8 years ago by isabelleaugenstein
  0

• Use DropoutWrapper from tensorflow.models.rnn.rnn_cell

  #5 opened 8 years ago by rockt
  1