Nemotron-RL Agentic Terminal Pivot

Nemotron-RL Agentic Terminal Pivot v1 is an open-source reinforcement learning dataset designed to train large language models for agentic command-line interface (CLI) tasks. It provides high-quality, expert-derived trajectories that enable models to perform complex software engineering operations, tool use, and reasoning within Linux environments.

Nemotron-RL Agentic Terminal Pivot v1 is an open-source reinforcement learning dataset designed to train large language models for agentic command-line interface (CLI) tasks. It provides high-quality, expert-derived trajectories that enable models to perform complex software engineering operations, tool use, and reasoning within Linux environments.

• 01Expert-derived trajectories for command-line agent training
• 02Supports reinforcement learning from verifiable reward (RLVR)
• 03Optimized for software engineering and tool-use reasoning
• 04Compatible with NVIDIA NeMo Gym and NeMo framework
• 05Permissive CC-BY-4.0 licensing for commercial and research use