chenchen278

print('hs.append(h)', hs.append(h))这里出错啦，hs.append(h)的方法会再一次把h插入到hs中，造成后续的维度不匹配。同时该方法调用没有返回值，print了个寂寞hhh。

In the paper DDPM, it seems that var is equal to posterior_var or betas. And I don't know why they were concatenated here?

训练时并没有遍历所有的时间t，而是对时间进行采样，只对某部分t到t-1的过程进行降噪训练。在采样时才是按时间遍历进行降噪