NVIDIA NeMo Framework¶
NeMo Framework is NVIDIA's GPU accelerated, end-to-end training framework for large language models (LLMs), multi-modal models and speech models. It enables seamless scaling of training (both pretraining and post-training) workloads from single GPU to thousand-node clusters for both 🤗Hugging Face, Megatron, and PyTorch models.
This site hosts documentation, tutorials, and insights about NeMo's core components and integrations.
Latest Blog Posts¶
Reinforcement Learning with NVIDIA NeMo-RL: Reproducing a DeepScaleR Recipe Using GRPO¶
July 8, 2025
Learn how to use NVIDIA NeMo-RL to reproduce a DeepScaleR recipe using Group Relative Policy Optimization (GRPO) for training high-performing reasoning models. This comprehensive guide covers the step-by-step process of setting up, training, and evaluating models using the DeepScaleR methodology.
NeMo Framework Components¶
-
🚀 NeMo-RL
Scalable toolkit for efficient model reinforcement learning and post-training. Includes algorithms like DPO, GRPO, and support for everything from single-GPU prototypes to thousand-GPU deployments.
License¶
Apache 2.0 licensed with third-party attributions documented in each repository.