RElbers/info-nce-pytorch

Issues

• How to use InfoNCE for prototypes with shape [b, num_cls, num_tokens, dim]?

  #19 opened 4 months ago by zzzyzh
  1

• Question about input data

  #18 opened 8 months ago by irowberryFS
  0

• [Issue] InfoCE rather than InfoNCE

  #16 opened a year ago by k920049
  0

• How to import info-nce-pytorch

  #14 opened a year ago by Skywalker-Harrison
  2

• Questions about codes for negative key is None

  #13 opened a year ago by Yuntian9708
  10

• multiple positive sampels per query

  #7 opened 2 years ago by YiJun-Yang
  4

• Contribute to PyTorch?

  #10 opened 2 years ago by adamjstewart
  2

• I wonder if the way of InfoNCE I used was wrong( ´•̥̥̥ω•̥̥̥` )

  #12 opened 2 years ago by evelynlee999
  7

• should I add parameters of InfoNCE to optimizer for optimizing?

  #11 opened 2 years ago by xinyaojiejie
  3

• Why do you use cross entropy ?

  #9 opened 2 years ago by shamanez
  4

• Minimizing this loss function will minimize or maximize mutual information?

  #8 opened 2 years ago by geolvr
  3

• Do you have infoNCE loss function cpde

  #5 opened 2 years ago by jiangxinke
  1

• calculate similarity for 3-dim input

  #4 opened 2 years ago by dangvansam
  1

• about loss parameter

  #6 opened 3 years ago by hrdxwandg
  1

• Correspondance between query and negative_key

  #3 opened 3 years ago by TedSIWEILIU
  2

• i have one question for the part of code

  #2 opened 3 years ago by mmmmmmrluo
  7

• reduction mode

  #1 opened 4 years ago by XA-kirino
  1