AlphaUncommon
LeapLabTHU/JustGRPO
[ICML 2026 Outstanding Paper] Minimalist RL for Diffusion LLMs. 89.1% on GSM8K.
- Owner
- LeapLabTHU (personal account)
- Stars
- Stars: 288
- Forks
- Forks: 16
- Language
- Python
- License
- MIT
- Created
- 2026
- Topics
- diffusion-language-modelsgrporeinforcement-learningminimalistic
- Tier
- Alpha (250 to 699 ★) What this tier means
More repos by LeapLabTHU
All 22 repos by LeapLabTHU- LeapLabTHU/Absolute-Zero-ReasonerStars: 1,925PythonBeta
- LeapLabTHU/cooragentStars: 1,671PythonBeta
- LeapLabTHU/DATStars: 944PythonBeta
- LeapLabTHU/Agent-AttentionStars: 669PythonAlpha
- LeapLabTHU/FLatten-TransformerStars: 449PythonAlpha
- LeapLabTHU/ACmixStars: 411PythonAlpha
Nearby in Python
All Python repos- yokel1121/muyang-flat-animationStars: 288PythonAlpha
- benjiyaya/CalliopeStars: 288PythonAlpha
- groundingpi/GroundingPIStars: 288PythonAlpha
- ScrumDoLLC/ScrumDoStars: 288PythonAlpha
- berkerpeksag/github-badgeStars: 288PythonAlpha
- jasperla/openbsd-wipStars: 288PythonAlpha