James Betker d1007ccfe7 Adjustments to pixpro to allow training against networks with arbitrarily large structural latents ... - The pixpro latent now rescales the latent space instead of using a "coordinate vector", which **might** have performance implications. - The latent against which the pixel loss is computed can now be a small, randomly sampled patch out of the entire latent, allowing further memory/computational discounts. Since the loss computation does not have a receptive field, this should not alter the loss. - The instance projection size can now be separate from the pixel projection size. - PixContrast removed entirely. - ResUnet with full resolution added.		2021-01-12 09:17:45 -07:00
..
byol	Support training imagenet classifier	2021-01-11 20:09:16 -07:00
stylegan2	GLEAN!	2020-12-18 16:04:19 -07:00
ui/image_labeler	More refactoring	2020-12-18 09:18:34 -07:00
compute_fdpl_perceptual_weights.py	Directly use dataset keys	2020-12-04 20:14:53 -07:00
create_lmdb.py	Directly use dataset keys	2020-12-04 20:14:53 -07:00
extract_square_images.py	Adjustments to pixpro to allow training against networks with arbitrarily large structural latents	2021-01-12 09:17:45 -07:00
extract_subimages_with_ref.py	Add lucidrains pixpro trainer	2021-01-05 20:14:22 -07:00
extract_subimages.py	Misc	2020-10-20 12:56:52 -06:00
folderize_imagenet_val.py	Support training imagenet classifier	2021-01-11 20:09:16 -07:00
recover_tensorboard_log.py	Codebase cleanup	2020-10-13 20:56:39 -06:00
rename.py	Codebase cleanup	2020-10-13 20:56:39 -06:00
srflow_latent_space_playground.py	Script updates	2020-12-29 20:24:41 -07:00
test_psnr_approximator.py	More refactoring	2020-12-18 09:18:34 -07:00
use_discriminator_as_filter.py	Directly use dataset keys	2020-12-04 20:14:53 -07:00
use_generator_as_filter.py	More refactoring	2020-12-18 09:18:34 -07:00
validate_data.py	Fixes and additional support for progressive zoom	2020-10-30 09:59:54 -06:00