corenel/pytorch-adda
A PyTorch implementation for Adversarial Discriminative Domain Adaptation
PythonMIT
Issues
- 0
How to adapt when the dataset of the target domain is much smaller than source domain
#30 opened by universe27king - 5
- 2
Criterion problem
#11 opened by chihhuiho - 4
0% accuracy with pytorch >= 0.4.0
#10 opened by williford - 1
ERROR!!!
#20 opened by jiebiandeshitou - 2
Question about the USPS dataset
#25 opened by GaochangWu - 2
why in adapt.py line 91 label_tgt = make_variable(torch.ones(feat_tgt.size(0)).long())? I think label_tgt = make_variable(torch.zeros(feat_tgt.size(0)).long())
#28 opened by zzzpc - 1
RNN network invalid
#26 opened by zhangqingqing24630 - 2
- 8
d_loss and accuracy
#22 opened by Dr-Zhou - 1
The adaptation does not work
#23 opened by ZhangC2 - 2
USPS to MNIST
#3 opened by redhat12345 - 7
error about mnist data shape
#21 opened by SpiritBear000 - 7
Adaptation leads to lower precision.
#15 opened by TGISer - 0
something about eval
#24 opened by mrsempress - 0
Questions about zero_grad()
#19 opened by omg777 - 0
I can't download
#18 opened by OilGao - 0
I can't download
#17 opened by OilGao - 4
Losses
#2 opened by redhat12345 - 1
Question about the loss
#16 opened by taylover-pei - 8
accuracy just 13% after adaptation?
#8 opened by dupanfei1 - 3
- 0
Two feature maps output from discriminator?
#14 opened by aasharma90 - 0
Anybody visualized feature adaptation?
#13 opened by aasharma90 - 0
Pretrain+Finetune results?
#12 opened by jinfengr - 1
Can't load usps_28x28.pkl
#4 opened by IssamLaradji - 0