SageAttention-for-windows badgids · COMPLETED
Quantized Attention that achieves speedups of 2.1-3.1x and 2.7-5.1x compared to FlashAttention2 and xformers, respectively, without lossing end-to-end metrics across various models.
github.com/badgids/SageAttention-for-windows · ★ 1 · Forks 0 · Size 47.1 MB
SUMMARY
Technologies 5
Scored 5
Observed 5
Practices 6
Evidence 5
Skips 0
COVERAGE
Analyzed 93 files · 131 commits · 0 API calls
TECHNOLOGIES & DEPTH
C++ LANGUAGE Depth 70
9 files · PRODUCTION
Shell LANGUAGE Depth 70
1 files · PRODUCTION
Markdown LANGUAGE Depth 70
3 files · PRODUCTION
Python LANGUAGE Depth 70
24 files · PRODUCTION
TypeScript LANGUAGE Depth 70
3 files · PRODUCTION
PRACTICES
documentation · observedautomated_tests · absentcontinuous_integration · absentcontainerization · absentlinting · absentformatting · absent
ACTIVITY & OWNERSHIP
First commit 2024-10-03
Last commit 2025-03-01
Active months 6
Commits 131