celeba - 4GB. CelebA-HQ 256x256 dataset, as described in Progressive growing of GAN's.
For 1024x1024 version (120GB), use celeba-full-tfr.tar while downloading.
In celeba-4GB，there is {‘shape’， ‘data’， ’label‘} in one data, I can't use celeba's attribution to train the model.
In celeba-full-tfr, is 'attr' in one data?
In other words, to use celeba's attribution in 256x256x3, is there a prepared resource or need I generate it myself?
In celeba-4GB，there is {‘shape’， ‘data’， ’label‘} in one data, I can't use celeba's attribution to train the model.
In celeba-full-tfr, is 'attr' in one data?
In other words, to use celeba's attribution in 256x256x3, is there a prepared resource or need I generate it myself?
Thanks!