Stonesjtu Pytorch-NCE issues

Stonesjtu / Pytorch-NCE

The Noise Contrastive Estimation for softmax output written in Pytorch

MIT License

318 stars 46 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

--nce

#20 drtonyr opened 7 months ago
2
How to select negative samples for NCE loss

#19 SRL94 closed 3 years ago
1
Error in NCE expression?

#18 bczhu closed 3 years ago
1
Is the implementation batched NCE?

#17 zhuang-li closed 3 years ago
1
Can I use this loss on my customization model ?

#16 aceprojectx closed 3 years ago
0
why the labels in sampled_softmax_loss func are all zero?

#15 universewill closed 4 years ago
1
Why the nec_linear output loss while output prob for testing?

#14 universewill closed 3 years ago
1
Why the target index is not removed from noise samples?

#13 universewill closed 4 years ago
1
Why need to sub math.log(self.noise_ratio)

#12 gbuion closed 3 years ago
1
Add setup.py for pypi packaging

#11 mgraczyk closed 4 years ago
3
Add LICENSE

#10 mgraczyk closed 4 years ago
2
Add importance sampling as another loss function and batched NCE/IS for speed

#9 Stonesjtu closed 6 years ago
0
Target Sample can be included in Noise sample

#8 adonisues closed 3 years ago
1
why squeeze here?

#7 chaoqing opened 6 years ago
3
num_layers does't work

#6 EricAugust opened 6 years ago
9
Use logger instead of plain `print` for training log

#5 Stonesjtu closed 6 years ago
0
Generic nce

#4 Stonesjtu closed 6 years ago
0
Update README.md

#3 xiaoyunwu closed 6 years ago
1
truncated bptt without padding?

#2 eric-haibin-lin closed 6 years ago
2
main.py does not run 'as is' on penn data

#1 francoishernandez closed 7 years ago
1