Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

时间t的选择 #81

Open
X-fxx opened this issue May 10, 2024 · 1 comment
Open

时间t的选择 #81

X-fxx opened this issue May 10, 2024 · 1 comment

Comments

@X-fxx
Copy link

X-fxx commented May 10, 2024

你好,有两个问题想请教一下

  1. 论文中提到的“随机选择t而不使用最后一步的原因是,如果只保留最后一步去噪的梯度,则训练非常不稳定,结果很差。”得出这个结论是通过实验证明的吗,还是有什么理论依据?
  2. 直接从T时刻的噪声变成x_0会存在模糊的生成结果,为什么在前t+1个时刻中,代码实现中不直接使用噪声呢?
@xujz18
Copy link
Member

xujz18 commented Aug 1, 2024

  1. 是实验发现的。
  2. 因为ImageReward需要输入图片进行打分。
  1. It is found in experiments.
  2. Because ImageReward requires an input image for scoring.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

No branches or pull requests

2 participants