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/audiolm-pytorch
. Pull mirroring updated
Sep 19, 2024
.
2.0.7
42da76b6
·
patch
·
Jan 27, 2024
2.0.6
1a888d2f
·
fix dep
·
Jan 12, 2024
2.0.5
0d49ccb6
·
patch
·
Jan 03, 2024
2.0.4
562ed54f
·
patch
·
Dec 22, 2023
2.0.3
c4fb6f6b
·
use latest gateloop
·
Dec 21, 2023
2.0.2
6c031a44
·
fix for residual fsq
·
Dec 21, 2023
2.0.1
00f96814
·
2.0.1
·
Dec 19, 2023
2.0.0
8119f4fa
·
2.0
·
Dec 17, 2023
1.9.1
b5ef1b6f
·
only gradient sync on last gradient accum step for all 3 transformer training
·
Dec 11, 2023
1.9.0
01f00088
·
add warmup for soundstream as well as all discriminators
·
Dec 11, 2023
1.8.7
31b0b7e5
·
fix for splitting out valid dataset when dataset is directly passed into soundstream trainer
·
Dec 10, 2023
1.8.6
05c23d9b
·
1.8.6
·
Dec 10, 2023
1.8.5
c9e54cb9
·
1.8.5
·
Dec 01, 2023
1.8.4
37076fdb
·
address
https://github.com/lucidrains/audiolm-pytorch/issues/256
·
Nov 30, 2023
1.8.3
898f6479
·
use an assert to guide researchers
·
Nov 23, 2023
1.8.2
44a57ffb
·
allow for using jax associative scan for gate loop
·
Nov 20, 2023
1.8.1
e7b845a9
·
stft window fn requires device
·
Nov 19, 2023
1.8.0
2587134e
·
default window fn for complex stft discriminator to hann
·
Nov 18, 2023
1.7.10
32c3dd74
·
apply proposed gateloop to soundstream
·
Nov 18, 2023
1.7.9
48051631
·
minimize gradient syncs for soundstream distributed training
·
Nov 15, 2023
Prev
1
2
3
4
5
…
17
Next