Skip to content

Releases: Dao-AILab/flash-attention

v2.1.2.post3

04 Sep 06:46
Compare
Choose a tag to compare
Set single threaded compilation for CUDA 12.2 so CI doesn't OOM

v2.1.2.post2

04 Sep 06:00
Compare
Choose a tag to compare
Remove constexpr in launch template to fix CI compilation

v2.1.2.post1

04 Sep 05:46
Compare
Choose a tag to compare
Try switching back to Cutlass 3.2.0

v2.1.2

04 Sep 05:29
Compare
Choose a tag to compare
Bump to v2.1.2

v2.1.1

28 Aug 07:39
Compare
Choose a tag to compare
Update Cutlass to v3.2.0

v2.1.0

25 Aug 06:43
Compare
Choose a tag to compare
Change causal mask to be aligned to bottom-right instead of top-left

v2.0.9

22 Aug 07:21
Compare
Choose a tag to compare
Bump version to 2.0.9

v2.0.8

16 Aug 22:13
Compare
Choose a tag to compare
Bump to v2.0.8

v2.0.7

14 Aug 21:56
Compare
Choose a tag to compare
Bump to v2.0.7

v2.0.6.post2

14 Aug 17:28
Compare
Choose a tag to compare
[CI] Fix MATRIX_CUDA_VERSION check