NVIDIA NeMo Gym
NVIDIA's RL-on-LLM framework inside NeMo - a library of agentic environments (code, math, search, games, robotics) that drop into the NeMo-RL trainer for large-scale, GPU-efficient post-training.
- Type
- Framework
- Publisher
- NVIDIA
- Capabilities
- Code GenerationMathPlanningTool Calling
- Runtime
custom- License
- Apache-2.0
- Size
- growing collection: SWE-Gym, Sokoban, math, search, code-IO, browser envs
- Published
- Mar 2025
- Canonical
- github.com/NVIDIA-NeMo/RL
Cite
Notes
Only stored in your browser.