Skip to content

Pull requests: Dao-AILab/flash-attention

New pull request New
78 Open 286 Closed
78 Open 286 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Assigned to nobody Loading
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

Minor changes to get cute-dsl bwd running (divisibility assumptions, returning correct number of args, etc.)
#1919 opened Sep 30, 2025 by imbr92 Loading…
feat: add to support float8 kvcache in fa4
#1914 opened Sep 28, 2025 by yicwang Loading…
3
fix forward and backward kernel
#1907 opened Sep 24, 2025 by rz2778 Loading…
Add flash_attn_varlen_qkvpacked_func to hopper (flash_attn_3)
#1902 opened Sep 22, 2025 by foreverYoungGitHub Loading…
Feature/varlen rotray
#1899 opened Sep 19, 2025 by mhoangvslev Loading…
Fix the torch.compile failure of flash_attn_varlen_func
#1894 opened Sep 17, 2025 by zhenwendai Loading…
3
Improve setup.py
#1859 opened Sep 3, 2025 by cyyever Loading…
Refactors to enable FlexAttention
#1840 opened Aug 26, 2025 by drisspg Loading…
35
feat: Implement Sink Attention
#1819 opened Aug 18, 2025 by aoxy Loading…
1
15
fix race condition bug in cute _flash_attn_fwd in multiple gpu env
#1793 opened Aug 1, 2025 by beiw-nv Loading…
3
feat: blocksparse support
#1784 opened Jul 30, 2025 by guangyunh-nv • Draft
[CI] build upon manylinux, improve compatibility
#1780 opened Jul 29, 2025 by zipzou Loading…
9
Change the update method of the sub-module
#1774 opened Jul 25, 2025 by RealTapeL Loading…
add var_len case for benchmark_mla_decode
#1770 opened Jul 22, 2025 by XiaobingSuper Loading…
1
Add torch.compile support to flash attention 3
#1769 opened Jul 22, 2025 by guilhermeleobas Loading…
51
Enable the deterministic mode option in the backward kernel
#1766 opened Jul 21, 2025 by GD06 Loading…
4
[AMD] Torch Compile Issues
#1756 opened Jul 15, 2025 by micmelesse Loading…
2
Suppress warnings in windows compilation
#1748 opened Jul 10, 2025 by XXXXRT666 Loading…
6
Fix illegal memory access through off-by-one error in num_splits_dynamic_ptr init
#1747 opened Jul 10, 2025 by klondenberg-bioptimus Loading…
Theoretically make compiling from pip quicker
#1703 opened Jun 8, 2025 by whrit Loading…
fix: fa3 backward check qkv with qkv_scale and dqkv
#1686 opened May 29, 2025 by yuyu5333 Loading…
1
[skip ci] libtorch agnostic FA3 north star proposal
#1685 opened May 28, 2025 by janeyx99 • Draft
Fix/deterministic dk dv
#1678 opened May 26, 2025 by yuWeiCute Loading…
10
Fix a bug in flash_attn_triton.py
#1668 opened May 15, 2025 by AminDarabi Loading…
Useuful command to install flash faster on behamoth clusters
#1660 opened May 10, 2025 by sleepingcat4 Loading…
ProTip! Find all pull requests that aren't related to any open issues with -linked:issue.