Releases: kingbri1/flash-attention
Releases · kingbri1/flash-attention
v2.8.3
v2.8.2
Actions: Add torch 2.8.0 Signed-off-by: kingbri <8082010+kingbri1@users.noreply.github.com>
v2.7.4.post1
Actions: Remove artifacts No need to upload an artifact when uploading to a release anyway Signed-off-by: kingbri <8082010+kingbri1@users.noreply.github.com>
v2.7.1.post1
Actions: Bump Cuda 12.4 Signed-off-by: kingbri <8082010+bdashore3@users.noreply.github.com>
v2.7.0.post2
Actions: Bump Cuda 12.4 Signed-off-by: kingbri <8082010+bdashore3@users.noreply.github.com>
v2.6.3
Synced to Upstream version
NOTE: Backward and Dropout are disabled meaning that this release is INFERENCE ONLY.
This is because including these features more than doubles the build time and makes the github action time itself out. Please raise an issue to the parent repo to help reduce the build times if you want these features.
v2.6.1
Actions: Switch to CUDA 12.3 Signed-off-by: kingbri <bdashore3@proton.me>
v2.5.9.post2
Quick release to add softcapping commits. Does not have backward, dropout, or alibi support.
v2.5.9.post1
Actions: Clarify dispatch formatting Signed-off-by: kingbri <bdashore3@proton.me>
v2.5.8
Same as Upstream tag
Now built for only torch 2.2.2 and 2.3.0