jefflai108 Contrastive-Predictive-Coding-PyTorch issues

jefflai108 / Contrastive-Predictive-Coding-PyTorch

Contrastive Predictive Coding for Automatic Speaker Verification

MIT License

472 stars 96 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

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

#21 Ieremie opened 2 years ago
0
Use of Batch Normalisation

#20 Ieremie opened 2 years ago
0
Softmax uses by default dimension 1

#19 Ieremie opened 2 years ago
0
Second last tilmestep as the c_t in the baseline model?

#18 KinWaiCheuk opened 3 years ago
0
Feed entire input to encoder??

#17 NeteeraAF opened 3 years ago
1
About negative sample?

#16 aceprojectx closed 3 years ago
0
Hey should 'correct' only be if the final prediction was correct? Same for accuracy?

#15 colinator opened 3 years ago
1
NCE accuracy calculation

#14 JaejinCho closed 8 months ago
1
the self.softmax() don't update the paras in the CDCK2?

#13 super-wcg opened 4 years ago
1
Train the spk_class.py

#12 super-wcg opened 4 years ago
0
why torch.diag in nce loss?

#11 artificertxj1 closed 4 years ago
0
Train and test data not available

#10 martinmamql opened 4 years ago
1
What is the format of "list" file?

#9 couragelfyang opened 4 years ago
2
how to format h5 files for input?

#8 kachiem closed 4 years ago
3
The implementation of loss might be wrong

#7 aayushP closed 4 years ago
1
Threre might some wrong in validation.py

#6 cyl250 closed 4 years ago
1
Some Trouble in Understanding

#5 cyl250 closed 4 years ago
5
May I ask which version of torch do you use?

#4 ChenJiaDong9219 closed 4 years ago
2
How combine MFCC and CPCfeatures

#3 cyl250 closed 4 years ago
4
Code question in src/model/model.py#L113

#2 joeyy5588 closed 5 years ago
1
Can you provide the train & test dataset?

#1 guanyuelee closed 4 years ago
3