Will you release the model trained on CelebA-HQ?

[gsygsy96]

Thanks.

[ShenYujun]

Sorry, but for face inversion, we only use the model trained on FF-HQ.

[gsygsy96]

Thanks for your reply! And can you tell me the test split of FFHQ and LSUN you used in your paper?

[zhujiapeng]

For FFHQ, we used the first 65k images as the training dataset and the last 5k images as the test datasets. For LSUN (tower and bedroom), we randomly sampled 100k images from the first 600k images in each dataset and sampled the test data from the rest.