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/magvit2-pytorch
. Pull mirroring updated
Sep 19, 2024
.
0.4.9
06edbd31
·
address
https://github.com/lucidrains/magvit2-pytorch/issues/44
·
Jul 23, 2024
0.4.8
e085ec3c
·
0.4.8
·
Jul 22, 2024
0.4.6
f2027d2a
·
add ability to train with bsq
·
Jun 26, 2024
0.4.7
f2027d2a
·
add ability to train with bsq
·
Jun 26, 2024
0.4.5
0a8e8349
·
make sure schedulers are prepared
·
May 16, 2024
0.4.4
31258005
·
add a soft clamping of value before LFQ step
·
May 08, 2024
0.4.3
d266deb0
·
address
https://github.com/lucidrains/magvit2-pytorch/issues/18
·
May 07, 2024
0.4.2
c4d073a3
·
bump vq lib
·
May 03, 2024
0.4.1
b17e1aed
·
address
https://github.com/lucidrains/vector-quantize-pytorch/issues/112
·
May 01, 2024
0.4.0
9f490741
·
start using a new linear attention from stanford
·
Jan 18, 2024
0.3.4
434f36e9
·
address issue with encoding first frame separately...
·
Jan 11, 2024
0.3.3
9b12d255
·
address issue with encoding first frame separately...
·
Jan 11, 2024
0.3.2
eef687ca
·
0.3.2
·
Jan 07, 2024
0.3.1
5cda1aa7
·
Patch
·
Jan 06, 2024
0.3.0
80cde122
·
attempt a metaprogramming solution for the unwrapped model issue in accelerate
·
Jan 06, 2024
0.2.2
290323c5
·
0.2.2
·
Jan 04, 2024
0.2.1
39e99293
·
default heinsen false
·
Jan 03, 2024
0.2.0
f9c64673
·
add warmup
·
Dec 10, 2023
0.1.45
c6b91c5f
·
take care of all caps image extensions, for
https://github.com/lucidrains/magvit2-pytorch/issues/19
·
Dec 06, 2023
0.1.44
58a7b688
·
fix lfq
·
Dec 01, 2023
Prev
1
2
3
4
5
…
7
Next