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
..
.idea
data	Support training imagenet classifier	2021-01-11 20:09:16 -07:00
models	Adjustments to pixpro to allow training against networks with arbitrarily large structural latents	2021-01-12 09:17:45 -07:00
scripts	Adjustments to pixpro to allow training against networks with arbitrarily large structural latents	2021-01-12 09:17:45 -07:00
trainer	Support training imagenet classifier	2021-01-11 20:09:16 -07:00
utils	Enable vqvae to use a switched_conv variant	2021-01-09 20:53:14 -07:00
multi_modal_train.py
process_video.py	Fix process_video	2021-01-09 20:53:46 -07:00
requirements.txt	Support training imagenet classifier	2021-01-11 20:09:16 -07:00
test_image_patch_classifier.py
test.py
train.py	Adjustments to pixpro to allow training against networks with arbitrarily large structural latents	2021-01-12 09:17:45 -07:00