Loss divergency in S2_adv

[dg618k]

Dear friend! During the training for S2_adv, the gen_loss and disc_loss are divergency by Wasserstein distance. However, it is fine with cross_entropy loss referenced by the adda project. Can you help with it? Thank you!

[AIJIA000]

亲爱的朋友！ 在 S2_adv 的训练过程中，gen_loss 和 disc_loss 在 Wasserstein 距离上是发散的。但是，adda 项目引用的 cross_entropy 损失很好。你能帮忙吗？谢谢！

hello，“we preprocess all of these data into a json file ” May I ask if the processed data set in jison format can be uploaded, or tell me how to handle it.

[jiang18-318]

Dear friend! During the training for S2_adv, the gen_loss and disc_loss are divergency by Wasserstein distance. However, it is fine with cross_entropy loss referenced by the adda project. Can you help with it? Thank you!

do u have the processed datasets, i need it. thamk u.