Zhanxuan Hu

> @ClayZhang0403 I just tested the codes with wideresnet and resnet50, but resnet50 behaved much worse than wideresnet. different network may need different parameters, such as batch size and lr

> 我设置的batch_size=64，正常1s就训练完，有些时候得1分钟，感觉不是很正常 @CaiQiuYu 你好，你训练完作者的代码有调整参数吗？

@CaiQiuYu 我设置成了【80，5】之后，训练的ACC为0，Precision和Recall很高

Thanks for your reply. I will check the source codes of IR.

Yes! I find that 'Z' will finally be cancelled out in Eq.(3).

The pre-trained source models are consistent with your paper 'Exploiting the intrinsic neighborhood structure for source-free domain adaptation'?

ok, Thank you!