zhenglab/spiralnet

where to put dataset

Opened this issue · 1 comments

Hi, where should I put the dataset of celeba_hq?

Hi, you need to list the file names of training/testing images in a flist file, and modify the file path in config.yml here:

TRAIN_FLIST: ./datasets/Celeba-HQ/train.flist
VAL_FLIST: ./datasets/Celeba-HQ/val.flist
MASK_FLIST: ./datasets/Celeba-HQ/mask.flist

An example flist file like this:

{PATH}/celeba-256/5196.jpg
{PATH}/celeba-256/27106.jpg
{PATH}/celeba-256/14405.jpg
...

It is similar to other datasets. Thanks.