Commits

Commits on Jan 21, 2023

Update README.md
arxyzan
authored
5bcb32f
Copy full SHA for 5bcb32f

Commits on Jan 2, 2023

Merge branch 'main' of https://github.com/AryanShekarlaban/data2vec-pytorch into main
arxyzan
committed
5cc7469
Copy full SHA for 5cc7469
fix ema cast to device bug
arxyzan
committed
97e7a78
Copy full SHA for 97e7a78

Commits on Jul 11, 2022

add disclaimer
arxyzan
authored
9ffa7b6
Copy full SHA for 9ffa7b6

Commits on Jul 7, 2022

fix #9: fix deepcopy
arxyzan
committed
58c9f56
Copy full SHA for 58c9f56
[text][dataset] remove unnecessary attrs from config
arxyzan
committed
f3785ce
Copy full SHA for f3785ce
fix #8: deep copy model in EMA
arxyzan
committed
20f1f2f
Copy full SHA for 20f1f2f

Commits on Jul 5, 2022

fix #8: deep copy model in EMA
arxyzan
authored
61afed5
Copy full SHA for 61afed5

Commits on May 12, 2022

Update README.md
arxyzan
authored
39245c4
Copy full SHA for 39245c4

Commits on May 9, 2022

Update README.md
arxyzan
authored
152c7e9
Copy full SHA for 152c7e9

Commits on May 6, 2022

Update README.md
arxyzan
authored
0743a70
Copy full SHA for 0743a70

Commits on May 5, 2022

Update README.md
arxyzan
authored
ed35d77
Copy full SHA for ed35d77
[readme] add pretrained weights for beit and roberta
arxyzan
committed
9987ec3
Copy full SHA for 9987ec3
[readme] add pretrained weights
arxyzan
committed
8ff2421
Copy full SHA for 8ff2421
[readme] add pretrained weights for beit and roberta
arxyzan
committed
b434a20
Copy full SHA for b434a20
[readme] add pretrained weights for beit and roberta
arxyzan
committed
11b8205
Copy full SHA for 11b8205
[readme] add pretrained weights for beit and roberta
arxyzan
committed
248eaad
Copy full SHA for 248eaad
[utils] save data2vec.encoder.encoder in checkpoint instead of data2vec.encoder
arxyzan
committed
867c537
Copy full SHA for 867c537

Commits on Apr 28, 2022

add weights section
arxyzan
authored
9cdb445
Copy full SHA for 9cdb445

Commits on Apr 25, 2022

[data2vec] more docstrings
arxyzan
committed
16de692
Copy full SHA for 16de692
[text] static regression head
arxyzan
committed
a3f84df
Copy full SHA for a3f84df

Commits on Apr 18, 2022

[audio] use cuda, [vision] minor modif in encoder docstring
arxyzan
committed
640fb85
Copy full SHA for 640fb85
modify save checkpoints
arxyzan
committed
8fbc43e
Copy full SHA for 8fbc43e
Merge pull request #1 from kabouzeid/main
Show description for 16c7fd8
arxyzan
authored
16c7fd8
Copy full SHA for 16c7fd8
complete requirements.txt
karimknaebel
committed
1439727
Copy full SHA for 1439727
fix ema cfg
karimknaebel
committed
292ed1d
Copy full SHA for 292ed1d

Commits on Apr 17, 2022

add audio pretraining instructions to readme
arxyzan
authored
4cd28d8
Copy full SHA for 4cd28d8
[audio] add trainer.py
arxyzan
committed
f6e1487
Copy full SHA for f6e1487
[data2vec] fix audio redundant permutations
arxyzan
committed
c1ad8f4
Copy full SHA for c1ad8f4
provide mask param for modalities in encoder.py files
arxyzan
committed
c8f6c47
Copy full SHA for c8f6c47
[text] minor name refactor
arxyzan
committed
630ed38
Copy full SHA for 630ed38
[text, vision] move log_dir to train in config
arxyzan
committed
bdb9057
Copy full SHA for bdb9057

Commits on Apr 16, 2022

Merge branch 'main' of https://github.com/AryanShekarlaban/data2vec-pytorch into main
arxyzan
committed
1cc4f7d
Copy full SHA for 1cc4f7d
[audio] remove feature extractor from encoder (moved to dataset instead)
arxyzan
committed
5543e3a
Copy full SHA for 5543e3a
[audio] add timit_asr dataset
arxyzan
committed
0f2552a
Copy full SHA for 0f2552a