opendilab/LightRFT

Fine-tuning Tools

LightRFT: Light, Efficient, Omni-modal & Reward-model Driven Reinforcement Fine-Tuning Framework

GitHub →Website →

5.1

GitHub Metrics

Stars

340

Forks

Open Issues

Watchers

Contributors

Weekly Commits

Language

Python

License

Apache-2.0

Last Commit

Apr 29, 2026

Created

Dec 23, 2025

Latest Release

v0.1.1

Release Date

Jan 23, 2026

Synced: Jun 5, 2026

Quality Scores

Documentation Qualityw: 20%

5.6

Has docs site (https://opendilab.github.io/LightRFT). Description: 96 chars. Stars signal: 340. Contributors: 6. Score: 5.6/10

Community Healthw: 20%

3.0

Stars: 340. Contributors: 6. Watchers: 2. Forks: 11. Issue ratio: 3.8%. Score: 3/10

Maintenance Velocityw: 15%

6.5

Last commit: 37d ago. Weekly commits: 0. Latest release: v0.1.1. Score: 6.5/10

API Design & DXw: 20%

6.9

Stars/issues ratio: 26. Dynamic language: Python. Has documentation site. Permissive license: Apache-2.0. Popularity signal: 340 stars. Score: 6.9/10

Production Readinessw: 15%

3.3

Battle-tested: 340 stars. Peer review: 6 contributors. Versioned: v0.1.1. Licensed: Apache-2.0. Age: 0.5 years. Maintenance: last commit 37d ago. Score: 3.3/10

Ecosystem Integrationw: 10%

5.3

Fork interest: 11. Major ecosystem: Python. Integration-friendly: Apache-2.0. Adoption: 340 stars. Has web presence. Score: 5.3/10