flash-linear-attention has 499K monthly downloads and 5K GitHub stars. Version 0.5.1 supports Python 3.10+, first published in 2025.
It is actively maintained under a permissive license, last released in Jun 2026.
It is built on top of fla-core and transformers. It is used by 6 tracked packages, including lmdeploy, axolotl and megatron-bridge.
GitHub topics include large-language-models, machine-learning-systems and natural-language-processing.
Package Insights
((week_daily_avg - month_daily_avg) / month_daily_avg) * 100Weekly Downloads
GitHub Stars
Downloads by OS
Python Versions
Top Countries
Dependencies
- fla-core ==0.5.1
- transformers >=4.45.0
10 optional dependencies
- causal-conv1d[conv1d]
- datasets[benchmark]
- fla-core[npu]
- fla-core[rocm]
- fla-core[xpu]
- fla-core[cuda]
- fla-core[cpu]
- matplotlib[benchmark]
- pytest[test]
- tilelang[tilelang]