Reposby GlyphNvidia open-source projects
12 projects · 111,559 stars · nvidia.com
Megatron-LMOngoing research training transformer models at scale.Nvidia · +52 a weekTensorRT-LLMTensorRT LLM provides users with an easy-to-use Python API to define Large Language Models (LLMs) and supports state-of-the-art optimizations to perform inference efficiently on NVIDIA GPUs.Nvidia · +53 a weekcosmosNVIDIA Cosmos is an open platform of world models, datasets, and tools that enables developers to build Physical AI for robots, autonomous vehicles, smart infrastructure, and more.Nvidia · +59 a weekcutlassCUDA Templates and Python DSLs for High-Performance Linear Algebra.Nvidia · +32 a weekcudfcuDF - GPU DataFrame Library.Nvidia · +5 a weekapexA PyTorch Extension: Tools for easy mixed precision and distributed training in Pytorch.Nvidia · +0 a weekIsaac-GR00TNVIDIA Isaac GR00T N1.7 - A Foundation Model for Generalist Robots.Nvidia · +34 a weekFasterTransformerTransformer related optimization, including BERT, GPT.Nvidia · +1 a week · release/v5.3_tag, Jan 23, 2023DALIA GPU-accelerated library containing highly optimized building blocks and an execution engine for data processing to accelerate deep learning training and inference applications.Nvidia · +4 a weekcumlNVIDIA cuML: GPU-Accelerated Machine Learning.Nvidia · +4 a weektacotron2Tacotron 2 - PyTorch implementation with faster-than-realtime inference.Nvidia · +2 a weekncclOptimized primitives for collective multi-GPU communication.Nvidia · +14 a week