How to generate the npz files for the custom reference dataset?

openai / guided-diffusion

MIT License

6.11k stars 813 forks source link

How to generate the npz files for the custom reference dataset? #148

Open ChongjianGE opened 3 months ago

ChongjianGE commented 3 months ago

I notice that the the npz file for the reference images mainly include some information of _mu
sigma
mu_s
sigma_s
mu_clip
sigma_clip
arr0
Since I am currently considering on generating new reference npz files on my custom dataset, could anyone explain how to generate the above information on a new set of reference images?

Fushennn commented 3 months ago

I am thinking about it, too. Has it been solved yet?

TaylorJocelyn commented 1 month ago

I’m experiencing a similar issue. May I ask if you were able to resolve it?

ZachL1 commented 1 hour ago

If the FID statistics haven't been calculated for the provided reference batch, the evaluation script will automatically compute them.

see line 172: https://github.com/openai/guided-diffusion/blob/22e0df8183507e13a7813f8d38d51b072ca1e67c/evaluations/evaluator.py#L167-L172