James Betker 44a19cd37c ExtensibleTrainer mods to support advanced checkpointing for stylegan2 ... Basically: stylegan2 makes use of gradient-based normalizers. These make it so that I cannot use gradient checkpointing. But I love gradient checkpointing. It makes things really, really fast and memory conscious. So - only don't checkpoint when we run the regularizer loss. This is a bit messy, but speeds up training by at least 20%. Also: pytorch: please make checkpointing a first class citizen.		2020-11-12 15:45:07 -07:00
..
archs	stylegan2 in ml art school!	2020-11-12 15:42:05 -07:00
experiments
flownet2@db2b7899ea
steps	stylegan2 in ml art school!	2020-11-12 15:42:05 -07:00
__init__.py	Lots of new discriminator nets	2020-11-10 16:06:54 -07:00
base_model.py	Extra logging for teco_resgen	2020-10-28 15:21:22 -06:00
ExtensibleTrainer.py	ExtensibleTrainer mods to support advanced checkpointing for stylegan2	2020-11-12 15:45:07 -07:00
feature_model.py
loss.py
lr_scheduler.py
networks.py	stylegan2 in ml art school!	2020-11-12 15:42:05 -07:00