Skip to content

About the validation dataset #3

@wenhancao

Description

@wenhancao

Hi, thank you for releasing the code!
In the paper, you mention evaluating on FFHQ-1k-validation and ImageNet-1k-validation datasets, each with 1000 images of size 256×256.

However, I could not find any script or instruction in this repo that extracts or defines these 1k subsets from the full FFHQ and ImageNet datasets.
Could you please clarify:

(a) Are these 1k validation sets publicly available or generated randomly?

(b) If they are generated locally, could you share the random seed, the image indices, or a sampling script to reproduce them?

Thanks in advance!

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions