jefflai108/Contrastive-Predictive-Coding-PyTorch

Issues

• NCE accuracy calculation

  #14 opened 9 months ago by JaejinCho
  1

• About negative sample?

  #16 opened 3 years ago by aceprojectx
  0

• NCE should be calculated for each output of the GRU (and future positions) instead of only a random index

  #21 opened 3 years ago by Ieremie
  0

• Use of Batch Normalisation

  #20 opened 3 years ago by Ieremie
  0

• Softmax uses by default dimension 1

  #19 opened 3 years ago by Ieremie
  0

• Feed entire input to encoder??

  #17 opened 3 years ago by NeteeraAF
  1

• Train and test data not available

  #10 opened 4 years ago by martinmamql
  1

• What is the format of "list" file?

  #9 opened 4 years ago by couragelfyang
  2

• Some Trouble in Understanding

  #5 opened 5 years ago by cyl250
  5

• Second last tilmestep as the c_t in the baseline model?

  #18 opened 3 years ago by KinWaiCheuk
  0

• Hey should 'correct' only be if the final prediction was correct? Same for accuracy?

  #15 opened 4 years ago by colinator
  1

• the self.softmax() don't update the paras in the CDCK2?

  #13 opened 4 years ago by super-wcg
  1

• Train the spk_class.py

  #12 opened 4 years ago by super-wcg
  0

• why torch.diag in nce loss?

  #11 opened 4 years ago by artificertxj1
  0

• Can you provide the train & test dataset?

  #1 opened 5 years ago by guanyuelee
  3

• The implementation of loss might be wrong

  #7 opened 5 years ago by aayushP
  1

• how to format h5 files for input?

  #8 opened 5 years ago by kachiem
  3

• Threre might some wrong in validation.py

  #6 opened 5 years ago by cyl250
  1

• May I ask which version of torch do you use?

  #4 opened 5 years ago by ChenJiaDong9219
  2

• How combine MFCC and CPCfeatures

  #3 opened 5 years ago by cyl250
  4

• Code question in src/model/model.py#L113

  #2 opened 5 years ago by joeyy5588
  1