Skip to content

Releases: lucidrains/stylegan2-pytorch

1.0.0

24 Sep 17:15

Choose a tag to compare

ready for 1.0

0.22.3

24 Sep 01:32

Choose a tag to compare

add diffaugment script from zhaos paper, allow user to specify which …

0.22.2

23 Sep 20:18

Choose a tag to compare

no sync all DDPs that were usedeven if detached, to play it safe

0.22.1

23 Sep 18:48

Choose a tag to compare

properly manage gradient accumulations for distributed case

0.22.0

23 Sep 08:12

Choose a tag to compare

add distributed data parallel and offer multi-GPU training on one mac…

0.21.2

22 Sep 20:15

Choose a tag to compare

use entry points instead of scripts in setup.py, potentially fixing w…

0.21.1

22 Sep 06:12

Choose a tag to compare

allow for assigning a specific gpu id for training, readying for mult…

0.21.0

04 Sep 05:49

Choose a tag to compare

add relativistic discriminator loss

0.20.2

30 Aug 04:28

Choose a tag to compare

bump

0.20.1

23 Aug 00:15

Choose a tag to compare

make random crop in augmentation a bit less severe