feifeibear/long-context-attention

Fine-tuning Tools

USP: Unified (a.k.a. Hybrid, 2D) Sequence Parallel Attention for Long Context Transformers Model Training and Inference

GitHub →

5.8

GitHub Metrics

Stars

675

Forks

Open Issues

Watchers

Contributors

Weekly Commits

Language

Python

License

Apache-2.0

Last Commit

May 21, 2026

Created

Mar 27, 2024

Latest Release

0.6.4

Release Date

Jan 13, 2026

Synced: Jun 29, 2026

Quality Scores

Documentation Qualityw: 20%

4.7

No dedicated docs site. Description: 119 chars. Stars signal: 675. Contributors: 24. Score: 4.7/10

Community Healthw: 20%

5.0

Stars: 675. Contributors: 24. Watchers: 5. Forks: 80. Issue ratio: 1.8%. Score: 5/10

Maintenance Velocityw: 15%

6.3

Last commit: 39d ago. Weekly commits: 0. Latest release: 0.6.4. Maturity bonus: 2.3y old. Score: 6.3/10

API Design & DXw: 20%

6.7

Stars/issues ratio: 56. Dynamic language: Python. No dedicated API docs. Permissive license: Apache-2.0. Popularity signal: 675 stars. Score: 6.7/10

Production Readinessw: 15%

6.1

Battle-tested: 675 stars. Peer review: 24 contributors. Versioned: 0.6.4. Licensed: Apache-2.0. Age: 2.3 years. Maintenance: last commit 39d ago. Score: 6.1/10

Ecosystem Integrationw: 10%

6.5

Fork interest: 80. Major ecosystem: Python. Integration-friendly: Apache-2.0. Adoption: 675 stars. Score: 6.5/10