Commits

Commits on Sep 18, 2025

0.7.12
lucidrains
committed
1fd83f3
Copy full SHA for 1fd83f3
Modify RiNALMoWrapper to accept variant parameter (#326)
ZhiyuanChen
authored
a7704fc
Copy full SHA for a7704fc

Commits on Jul 1, 2025

fix another type issue
lucidrains
committed
a32ffd9
Copy full SHA for a32ffd9
0.7.10
lucidrains
committed
d733662
Copy full SHA for d733662

Commits on Jun 29, 2025

fix types in inputs
lucidrains
committed
976aec4
Copy full SHA for 976aec4

Commits on Jun 28, 2025

link to megafold
lucidrains
authored
8baeb9a
Copy full SHA for 8baeb9a

Commits on Dec 29, 2024

complete hyper connected alphafold3
lucidrains
committed
c42884e
Copy full SHA for c42884e

Commits on Dec 28, 2024

hyperconnect single repr in pairformerstack
lucidrains
committed
e18a330
Copy full SHA for e18a330
remove laser attention
lucidrains
committed
c5d1f7b
Copy full SHA for c5d1f7b
fix rmsnorm in hyper conn lib
lucidrains
committed
119fcb9
Copy full SHA for 119fcb9
add hyper connections to diffusion transformer
lucidrains
committed
ebaa6ee
Copy full SHA for ebaa6ee
upgrade value residual to learnt mixing per token / head
lucidrains
committed
85c0de3
Copy full SHA for 85c0de3

Commits on Dec 27, 2024

try to fix workflow
lucidrains
committed
af6f972
Copy full SHA for af6f972
complete value residual for pairformer stack
lucidrains
committed
1684d50
Copy full SHA for 1684d50

Commits on Dec 3, 2024

0.6.11
lucidrains
committed
49f7c96
Copy full SHA for 49f7c96
use softclamping to address numerical issues with laser instead
lucidrains
committed
0cee1cc
Copy full SHA for 0cee1cc

Commits on Dec 2, 2024

just tempt some student into trying laser attention
lucidrains
committed
34f6c59
Copy full SHA for 34f6c59

Commits on Nov 26, 2024

adopt Adopt
lucidrains
committed
0791dfe
Copy full SHA for 0791dfe

Commits on Nov 17, 2024

fix local windowed attention to be overlapping
lucidrains
committed
308c470
Copy full SHA for 308c470

Commits on Nov 2, 2024

value residual learning (#312)
Show description for 95d3ab6
lucidrains
authored
95d3ab6
Copy full SHA for 95d3ab6

Commits on Oct 17, 2024

0.6.5
lucidrains
committed
9e5bb92
Copy full SHA for 9e5bb92
Fix bugs related to pAE (#310)
qiaoqiaoLF
authored
2ad09e9
Copy full SHA for 2ad09e9
every machine keep track of ema if using switch ema
lucidrains
committed
3415855
Copy full SHA for 3415855

Commits on Oct 16, 2024

a new paper claims there is a free lunch by setting model weights to ema weights every epoch. allow researchers to experiment with this, conveniently already available in EMA-pytorch due to hare an…
Show description for 39d282f
lucidrains
committed
39d282f
Copy full SHA for 39d282f

Commits on Oct 11, 2024

fix MSA_TYPE
lucidrains
committed
4768a65
Copy full SHA for 4768a65

Commits on Oct 4, 2024

0.6.1
lucidrains
authored
2cdc53b
Copy full SHA for 2cdc53b
Update alphafold3.py (#306)
amorehead
authored
72183b4
Copy full SHA for 72183b4

Commits on Oct 3, 2024

0.6.0
lucidrains
authored
b6c8b17
Copy full SHA for b6c8b17
Remove zero init and constant biasing for `to_gate` (#304)
Show description for ffe6a89
amorehead
authored
ffe6a89
Copy full SHA for ffe6a89
0.5.58
lucidrains
authored
41aa91c
Copy full SHA for 41aa91c
Update alphafold3.py (#302)
amorehead
authored
665ffef
Copy full SHA for 665ffef
update nim version
lucidrains
committed
7d7e6d8
Copy full SHA for 7d7e6d8

Commits on Oct 1, 2024

0.5.57
lucidrains
authored
e4202e3
Copy full SHA for e4202e3
Update alphafold3.py (#300)
amorehead
authored
64bd90f
Copy full SHA for 64bd90f
Update inputs.py (#299)
amorehead
authored
7258c56
Copy full SHA for 7258c56