James Betker 5f2c722a10 SRG2 revival ... Big update to SRG2 architecture to pull in a lot of things that have been learned: - Use group norm instead of batch norm - Initialize the weights on the transformations low like is done in RRDB rather than using the scalar. Models live or die by their early stages, and this ones early stage is pretty weak - Transform multiplexer to use u-net like architecture. - Just use one set of configuration variables instead of a list - flat networks performed fine in this regard.		2020-07-09 17:34:51 -06:00
..
__init__.py
arch_util.py	SRG2 revival	2020-07-09 17:34:51 -06:00
AttentionResnet.py	Add attention resnet	2020-05-29 20:02:10 -06:00
discriminator_vgg_arch.py	Fix up pixgan loss and pixdisc	2020-07-08 21:27:48 -06:00
DiscriminatorResnet_arch_passthrough.py	Allow passthrough discriminator to have passthrough disabled from config	2020-05-19 09:41:16 -06:00
DiscriminatorResnet_arch.py	Fixup upconv for the next attempt!	2020-05-01 19:56:14 -06:00
feature_arch.py	Fix process_video bugs	2020-05-29 12:47:22 -06:00
FlatProcessorNet_arch.py
FlatProcessorNetNew_arch.py
HighToLowResNet.py
NestedSwitchGenerator.py	SRG2 revival	2020-07-09 17:34:51 -06:00
ResGen_arch.py	More NSG improvements (v3)	2020-06-29 20:26:51 -06:00
RRDBNet_arch.py	Remove RRDB with switching	2020-07-01 12:08:32 -06:00
spinenet_arch.py	SRG3 work	2020-07-07 13:46:40 -06:00
SRG1_arch_new.py	Remove all biases from generator	2020-07-04 22:19:55 -06:00
SRG1_arch.py	Fix assertion error	2020-07-06 09:23:53 -06:00
SRResNet_arch.py
SwitchedResidualGenerator_arch.py	SRG2 revival	2020-07-09 17:34:51 -06:00