Skip to content

Pull requests: ROCm/flash-attention

New pull request New
9 Open 51 Closed
9 Open 51 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
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

Enable MQA/GQA in backward
#100 opened Nov 14, 2024 by micmelesse Loading…
Added Support for Rotary Positional Embeddings (both non-fused and fused kernel)
#99 opened Nov 13, 2024 by alexkranias-amd Loading…
@alexkranias-amd
7
Added Dropout BWD
#95 opened Nov 5, 2024 by alexkranias-amd Loading…
@alexkranias-amd
1
Integrated Rotary Positional Embeddings (RoPEs) into flash_attn_kvcache
#83 opened Sep 27, 2024 by alexkranias-amd Loading…
@alexkranias-amd
Fix stride issues in flash_attn_interface
#58 opened May 31, 2024 by clintg6 Loading…
1
@clintg6
GPUAI-1250 - Flash Attention v2.04 two modules layer_norm cannot be used fixed
#52 opened Apr 3, 2024 by xiaoxiangAMD Loading…
add FA api benchmark csv
#48 opened Mar 7, 2024 by fsx950223 Loading…
2
GPUAI-1250 - Flash Attention v2.04 module rotary cannot be used code fixed
#47 opened Mar 1, 2024 by xiaoxiangAMD Loading…
@sabreshao
2
Flash attention for rocm
#1 opened Feb 17, 2023 by groenenboomj Loading…
3
ProTip! What’s not been updated in a month: updated:<2024-10-15.