dragen1860 / deepembeddingmodel_zsl-pytorch Goto Github PK

View Code? Open in Web Editor NEW

58.0 58.0 19.0 42.02 MB

Pytorch code for CVPR 2017 paper: Learning a Deep Embedding Model for Zero-Shot Learning

Python 100.00%

deepembeddingmodel_zsl-pytorch's People

Stargazers

Watchers

Forkers

shunsunsun pengboxiangshang wang1104014663 zcrwind suzyzd jkhlot feifanrensheng wayne980 indussky8 xinshengwang rohitkeshari 459548764 thoamsdong szsbyx ruizhaoz buptlmk harpavatkeerti xugaoxu1205 946958884

deepembeddingmodel_zsl-pytorch's Issues

AwA2 dataset

Nice work, I wanted to check if I can find this implementation for AwA2 dataset?
Thanks in advance.

No L2 regularization

Hi,

I was exploring your code and I came across that in the original paper and tensorflow version, l2 regularization is used. But I did not see that in pytorch version. Do you have a reason for that?

Kind regards

Hi, it's pleasure for me to read your code. However, could you explanation that in the line 83, "w1.data.normal_(0, 0.02)", why the w's std is 0.02? How you choose this parameter? And I am trying to use the "xavier_uniform", it seems that the ans is better thaner "normal_(0, 0.02)".

Recommend Projects