Training library for Megatron-based models with bi-directional Hugging Face conversion capability

Developer Asset Hub for NVIDIA Nemotron — A one-stop resource for training recipes, usage cookbooks, and full end-to-end reference examples to build with Nemotron models

Lyra: Generative 3D Scene Reconstruction via Video Diffusion Model Self-Distillation

🎨 NeMo Data Designer: A general library for generating high-quality synthetic data from scratch or based on seed data.

Build RL environments for LLM training

The official Python library for the OpenAI API

DC-Gen: Post-Training Diffusion Acceleration with Deeply Compressed Latent Space

A collection of various NVIDIA DGX Cloud code examples

Pytorch Distributed native training library for LLMs/VLMs with OOTB Hugging Face support

Scalable toolkit for efficient model reinforcement

Lets make video diffusion practical!

Lightweight coding agent that runs in your terminal

A Pythonic framework to simplify AI service building

A tool to configure, launch and manage your machine learning experiments.

s1: Simple test-time scaling

CloudAI Benchmark Framework

Open-Sora: Democratizing Efficient Video Production for All

Scalable data pre processing and curation toolkit for LLMs

NVIDIA ACE samples, workflows, and resources

Official inference library for Mistral models

A package to generate summaries of long-form text and evaluate the coherence of these summaries. Official package for our ICLR 2024 paper, "BooookScore: A systematic exploration of book-length summ…

Convert PDF to markdown + JSON quickly with high accuracy

This is the Personality Core for GLaDOS, the first steps towards a real-life implementation of the AI from the Portal series by Valve.

The open source codebase powering HuggingChat

A high-throughput and memory-efficient inference and serving engine for LLMs

David Attenborough narrates your life

Dolphin is a GameCube / Wii emulator, allowing you to play games for these two platforms on PC with improvements.

An API standard for single-agent reinforcement learning environments, with popular reference environments and related utilities (formerly Gym)

The TinyLlama project is an open endeavor to pretrain a 1.1B Llama model on 3 trillion tokens.