improved-diffusion In my opinion, authors define L_{vlb} = L_0 + ... + L_T, not L

In my opinion, authors define L_{vlb} = L_0 + ... + L_T, not L_t.

Open unl1002 opened this issue 1 year ago • 2 comments

          In my opinion, authors define L_{vlb} = L_0 + ... + L_T, not L_t.

Thus, they may calculate the vlb loss with scale factor T (self.num_timestep).

Originally posted by @yhy258 in https://github.com/openai/improved-diffusion/issues/114#issuecomment-2261727529

Nov 23 '24 14:11 unl1002

so, which means we use L_t * T (self. num_timestep) to approximate L_ {vlb}?

Nov 23 '24 14:11 unl1002

Nov 27 '24 09:11 unl1002