Reinforcement Learning for Reasoning and Vision
19 repos
Libraries, frameworks, and research implementations for training language models and vision-language agents using reinforcement learning techniques, with emphasis on reasoning tasks and R1-style approaches. This cluster covers RL training pipelines, reasoning-focused model development, and vision-language action systems that leverage policy optimization and reward modeling. Repositories include foundational RL training frameworks, reasoning-specific implementations, and practical applications of RL to multimodal agent tasks.
Python · 10
rl ·7,065
reasoning ·7,065
llm ·5,162
vla ·1,857
r1-zero ·1,277
trl ·503
verl ·503
agentic-ai ·503
grpo ·503
chameleon ·158
