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/DALLE2-pytorch
. Pull mirroring updated
Sep 19, 2024
.
1.6.1
94404119
·
make self conditioning technique work with diffusion prior
·
Aug 12, 2022
1.6.0
7c5477b2
·
bet on the new self-conditioning technique out of geoffrey hintons group
·
Aug 12, 2022
1.5.0
be3bb868
·
add gradient checkpointing for all resnet blocks
·
Aug 02, 2022
1.4.6
451de348
·
enforce clip anytorch version
·
Jul 30, 2022
1.4.5
f22e8c87
·
make open clip available for use with dalle2 pytorch
·
Jul 30, 2022
1.4.4
87432e93
·
quick fix for linear attention
·
Jul 29, 2022
1.4.3
d1673784
·
add cosine sim for self attention as well, as a setting
·
Jul 29, 2022
1.4.2
2d67d582
·
change up epsilon in layernorm the case of using fp16, thanks to @Veldrovive...
·
Jul 29, 2022
1.4.1
dbb52cea
·
change up epsilon in layernorm the case of using fp16, thanks to @Veldrovive...
·
Jul 29, 2022
1.4.0
748c7fe7
·
allow for cosine sim cross attention, modify linear attention in attempt to resolve issue on fp16
·
Jul 29, 2022
1.2.2
80046334
·
make sure entire readme runs without errors
·
Jul 28, 2022
1.2.1
36fb46a9
·
fix readme and a small bug in DALLE2 class
·
Jul 28, 2022
1.2.0
07abfcf4
·
rescale values in linear attention to mitigate overflows in fp16 setting
·
Jul 27, 2022
1.1.0
406e7504
·
add upsample combiner feature for the unets
·
Jul 26, 2022
1.0.6
9646dfc0
·
fix path_or_state bug
·
Jul 26, 2022
1.0.5
62043acb
·
fix repaint
·
Jul 24, 2022
1.0.4
9008531d
·
fix repaint
·
Jul 24, 2022
1.0.3
417ff808
·
1.0.3
·
Jul 22, 2022
1.0.1
76d08498
·
diffusion prior training updates from @nousr
·
Jul 20, 2022
1.0.0
06c65b60
·
1.0.0
·
Jul 19, 2022
Prev
1
2
3
4
5
6
7
…
18
Next