AI4Finance-Foundation / FinRL

Open-source Python framework for training and backtesting financial trading agents with deep reinforcement learning.

NVIDIA-NeMo / labs-molt

A scalable, agentic-first, and HuggingFace-native RL framework for research (9k lines).

Frank0415 / mini-RL-Framework

Implementing a tiny RL framework from scratch with as little dependencies as possible, can run RL on real LLMS using PC/laptops with limited resources.