Search Criteria
Package Details: python-flash-attention 2.8.3-1
Package Actions
| Git Clone URL: | https://aur.archlinux.org/python-flash-attention.git (read-only, click to copy) |
|---|---|
| Package Base: | python-flash-attention |
| Description: | Fast and memory-efficient exact attention. CUDA version. |
| Upstream URL: | https://github.com/Dao-AILab/flash-attention |
| Keywords: | ai |
| Licenses: | BSD-3-Clause |
| Submitter: | daskol |
| Maintainer: | lubosz |
| Last Packager: | lubosz |
| Votes: | 4 |
| Popularity: | 0.001994 |
| First Submitted: | 2023-06-21 07:27 (UTC) |
| Last Updated: | 2026-04-09 22:59 (UTC) |
Dependencies (9)
- python-einopsAUR
- python-pytorch-cuda (python-pytorch-cuda12.9AUR, python-pytorch-opt-cuda12.9AUR, python-pytorch-opt-cuda)
- ninja (ninja-gitAUR, ninja-memAUR, ninja-noemacs-gitAUR, ninja-kitwareAUR, ninja-fuchsia-gitAUR, n2-gitAUR) (make)
- python-build (make)
- python-installer (make)
- python-packaging (make)
- python-psutil (make)
- python-setuptools (make)
- python-wheel (make)
Latest Comments
lubosz commented on 2025-10-13 20:46 (UTC)
Feel free to take my patches that drop GCC 13, fix OOM and bump to 2.8.3
https://github.com/lubosz/python-flash-attention/commits/v2.8.3/
lubosz commented on 2025-10-13 13:11 (UTC)
v2.8.3 has been released 2 months ago.
https://github.com/Dao-AILab/flash-attention/releases/tag/v2.8.3
hottea commented on 2025-04-21 13:28 (UTC)
arch should be x86_64, and according to spdx, license should be Apache-2.0