liuyisu

> 我遇到了同样的问题，报错的行是data_loader.py文件第223行，内容为： return {'imidx':torch.from_numpy(imidx), 'image': torch.from_numpy(tmpImg), 'label': torch.from_numpy(tmpLbl)} 我根据提示信息分别在tmplmg和tmpLbl后加上了.copy()，代码改为： return {'imidx':torch.from_numpy(imidx), 'image': torch.from_numpy(tmpImg.copy()), 'label': torch.from_numpy(tmpLbl.copy())} 然后可以正常训练，但是当跑到一个epoch结束后代码就会自动停止，而且也没有报任何错误，不知道是不是上面改动的原因 > > > without showing me the exact line of triggering this error, I...

> clone the repo and install 'improved_diffusion' using: pip install -e . Thank you very much for your reply. I encountered the above problem while testing on the server. Do...

Yes, I later discovered and corrected this issue, it seems to be still improved_ The issue of diffusion. I am a beginner, this is really too difficult

> 嗨，你解决了吗？我用我的自定义数据集进行了训练，遇到了同样的问题。我的模型过度拟合。 Hello, I am currently training on my own dataset, but at the beginning, I encountered the following problem. May I ask how to solve it? Thank you very...

> pytorch和CUDA版本不对应吧？ 感谢回复，已经放弃这个模型了🤦‍♀️