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
.
0.20.0
d4680bed
·
another local attention bump
·
Mar 02, 2023
0.19.2
06602256
·
allow for configuring xpos scale base, as @ilya16 reports that at about 3...
·
Mar 02, 2023
0.19.1
7e957b90
·
add quantize embeddings to facilitate learning and some product management
·
Mar 01, 2023
0.19.0
65495ad5
·
remove use of eos in fine transformer
·
Mar 01, 2023
0.18.2
b6eeb93f
·
add some asserts
·
Mar 01, 2023
0.18.1
1b15751a
·
fix length extrapolation for rotary embeddings in local attention package
·
Feb 28, 2023
0.18.0
a94e32f4
·
remove multihead EMA based on feedback from @ilya16
·
Feb 28, 2023
0.17.1
4daad9e7
·
expose the soundstream configs as a property
·
Feb 27, 2023
0.17.0
8d8b255c
·
release saving of soundstream configurations within checkpoint, and...
·
Feb 24, 2023
0.16.2
5889e518
·
make sure soundstream load function loads the ema version if available, also...
·
Feb 24, 2023
0.16.1
21a07ab9
·
get the default parameters for soundstream, as well as AudioLMSoundStream and...
·
Feb 23, 2023
0.16.0
1a44fd7a
·
get the default parameters for soundstream, as well as AudioLMSoundStream and...
·
Feb 23, 2023
0.15.9
9b8e702f
·
allow researchers to easily use original audiolm soundstream config
·
Feb 23, 2023
0.15.8
043dae08
·
missing import for soundstream file
·
Feb 23, 2023
0.15.7
1b5fdefc
·
always save version wherever possible
·
Feb 23, 2023
0.15.6
b612ff9b
·
always save version wherever possible
·
Feb 23, 2023
0.15.5
7596ae0c
·
default hubert to 16khz
·
Feb 23, 2023
0.15.4a
cd94fae5
·
start saving version numbers along with model, so researchers can trace back...
·
Feb 23, 2023
0.15.4
f3e80ad2
·
start saving version numbers along with model, so researchers can trace back...
·
Feb 23, 2023
0.15.3
bee80b5e
·
patch
·
Feb 21, 2023
Prev
1
…
4
5
6
7
8
9
10
11
12
…
17
Next