Skip to content
GitLab
Explore
Sign in
Tags
Tags give the ability to mark specific points in history as being important
This project is mirrored from
https://github.com/lucidrains/gigagan-pytorch
. Pull mirroring updated
Sep 19, 2024
.
0.1.1b
a5d942df
·
0.1.1
·
Jul 20, 2023
0.1.1
302b5d0d
·
add mixed precision, validate some memory saved, but still need to figure out...
·
Jul 20, 2023
0.1.1a
302b5d0d
·
add mixed precision, validate some memory saved, but still need to figure out...
·
Jul 20, 2023
0.1.0
9eb43610
·
some more progress, just release 0.1.0 as non-distributed seems to be working
·
Jul 20, 2023
0.0.92
5854e40f
·
potential fix for base generator
·
Jul 19, 2023
0.0.91
266016dc
·
fix multiscale input
·
Jul 19, 2023
0.0.90
8548f578
·
sanity checks
·
Jul 19, 2023
0.0.89
e7387795
·
remove the `real_images` keyword argument to discriminator and simplify. just...
·
Jul 19, 2023
0.0.88
3aa2b083
·
bump
·
Jul 19, 2023
0.0.87
ba3a6af0
·
just do auxiliary reconstruction loss on input image, not the multiscales;...
·
Jul 19, 2023
0.0.86
99e089e2
·
add some inner residuals to the multi-output predictor, deviating from paper
·
Jul 19, 2023
0.0.85
06bd4d48
·
print out number of parameters for generator and discriminator to make sure roughly balanced
·
Jul 19, 2023
0.0.84
92ff8cc7
·
do not calculate aux recon loss when training generator, add an old technique...
·
Jul 19, 2023
0.0.83
dd11b8aa
·
yet another fix for multiscale input
·
Jul 19, 2023
0.0.82
abee81a2
·
also make sure auxiliary reconstruction is not done for multiscale inputs of the same stage
·
Jul 19, 2023
0.0.81
d77d3e66
·
redo the conv layer that takes in multi-scale inputs and converts to feature,...
·
Jul 19, 2023
0.0.80
6c4572d4
·
nonlinearity in multiscale output predictor, add train() to make sure generator is in train state
·
Jul 19, 2023
0.0.79
3e4088ce
·
0.0.79
·
Jul 19, 2023
0.0.77
48e6eca9
·
get correct losses
·
Jul 19, 2023
0.0.76
3575974f
·
yet another fix
·
Jul 19, 2023
Prev
1
2
3
4
5
6
7
Next