A collection of easy to use, highly optimized Deep Learning Models for Language Modelling. Deep Learning Examples provides Data Scientist and Software Engineers with recipes to Train, fine-tune, and deploy State-of-the-Art Models

languagemodelling

DeepStream SDK delivers a complete streaming analytics toolkit for AI based video and image understanding and multi-sensor processing. The DeepStream SDK brings deep neural networks and other complex processing tasks into a stream processing pipeline.

deepstreamcomputervision

This collection contains NeMo models for Automatic Speech Recognition (ASR): Speech to Text, Speech Classification, Speaker Diarization, Speaker Verification, Speaker Recognition, Command Recognition, Voice Activity Detection

nemo_asr

A collection of TensorRT-LLM accelerated Windows RTX PC LLM models.

llms_optimized_for_rtx_pcs

Monitor GPUs in Kubernetes using NVIDIA DCGM. This is an exporter for a Prometheus monitoring solution in Kubernetes.

dcgm-exporter

The NVIDIA Kubernetes Device Plugin registers GPUs as compute resources in the Kubernetes cluster.

k8s-device-plugin

Deploy and Manage NVIDIA GPU resources in Kubernetes.

gpu-operator

gpu-operator-validator

Build and Run GPU Accelerated Docker Containers.

container-toolkit

Plugin for the Kubernetes Node Feature Discovery for adding GPU node labels.

gpu-feature-discovery

Docker containers distributed as part of the TAO Toolkit package

tao-toolkit

Provision NVIDIA GPU Driver as a Container

driver

Triton Inference Server is an open source software that lets teams deploy trained AI models from any framework, from local or cloud storage and on any GPU- or CPU-based infrastructure in the cloud, data center, or embedded devices.

tritonserver

Manages NVIDIA Driver upgrades in Kubernetes cluster.

k8s-driver-manager

CUDA is a parallel computing platform and programming model that enables dramatic increases in computing performance by harnessing the power of the NVIDIA GPUs.

cuda

NVIDIA Network Operator Helm Chart provides an easy way to install, configure and manage the lifecycle of NVIDIA Mellanox network operator.

network-operator

Manage and Monitor GPUs in Cluster Environments.

dcgm

PyTorch is a GPU accelerated tensor computational framework. Functionality can be extended with common Python libraries such as NumPy and SciPy. Automatic differentiation is done with a tape-based system at the functional and neural network layer levels.

pytorch

TensorFlow is an open source platform for machine learning. It provides comprehensive tools and libraries in a flexible architecture allowing easy deployment across a variety of platforms and devices.

tensorflow

Manage MIG partitions in Kubernetes with a simple label change to a node.

k8s-mig-manager

Llama 3.1 70B-Instruct NIM Production Branch October 2024 (PB 24h2) offers a 9-month lifecycle for API stability, with monthly patches for high and critical software vulnerabilities.

llama-3.1-70b-instruct-pb24h2

NVIDIA TensorRT is a C++ library that facilitates high-performance inference on NVIDIA graphics processing units (GPUs). TensorRT takes a trained network and produces a highly optimized runtime engine that performs inference for that network.

tensorrt

NVIDIA NIM for GPU accelerated Gemma-2-2B-IT inference through OpenAI compatible APIs

gemma-2-2b-instruct

NVIDIA NIM for GPU accelerated Llama-3-Swalow-70B-Instruct-v0.1 inference through OpenAI compatible APIs

llama-3-swallow-70b-instruct-v0.1

NVIDIA NIM for GPU accelerated Snowflake Arctic Embed Large Embedding inference

arctic-embed-l

The NVIDIA Retrieval QA Llama 1B Reranking NIM is a NIM optimized for providing a logit score that represents how relevant a document(s) is to a given query, fine-tuned for multilingual and cross-lingual text question-answering retrieval.

llama-3.2-nv-rerankqa-1b-v2

NVIDIA NIM for GPU accelerated Llama 3.1 Swallow 8B inference through OpenAI compatible APIs

llama-3.1-swallow-8b-instruct-v0.1

NVIDIA NIM for GPU accelerated Llama 2 70B inference through OpenAI compatible APIs

llama-2-70b-chat

NVIDIA NIM for GPU accelerated Mistral-7B-Instruct-v0.3 inference through OpenAI compatible APIs

mistral-7b-instruct-v0.3

NVIDIA NIM for GPU accelerated Llama 2 13B inference through OpenAI compatible APIs

llama-2-13b-chat

Nemo Retriever is a collection of NIMs reference architectures for RAG

nemo-retriever

Python Basic - AI Workbench Default Container

python-basic

NeuralVAD enabled Parakeet 1.1B ASR Model

rmir_asr_parakeet_1-1b_en_us_str_vad

The Default Avatar Scene is a pre-made 3D asset of an avatar scene compatible with the ACE Animation Pipeline.

default-avatar-scene

Access the Long-Term Support Branches (LTSB) of AI frameworks and SDKs. Supported for 36-months, with quarterly patches for high and critical software vulnerabilities.

long_term_support_branch_2

Access the production branches of AI frameworks and SDKs. Supported for 9 months with monthly security patches.

production_branch_october_2024

Access the latest infrastructure software and tools of the NVIDIA AI Enterprise 6 release, exclusively available with your NVIDIA AI Enterprise subscription. 

nvidia-ai-enterprise-infra-6

Access the production branches of Omniverse frameworks and SDKs. Supported for 9 months with monthly security patches.

production_branch_december_2024

Kit SDK is a toolkit for building native Omniverse applications and microservices.

kit-appstreaming-collection

AI-powered search for OpenUSD data, 3D models, images, and assets using text or image-based inputs.

usdsearch

Use a reference application to build a fully functional retrieval-augmented generation (RAG)-based AI chatbot built with NVIDIA NIMTM microservices

ai-chatbot-with-rag

A collection of easy to use, highly optimized Deep Learning Models for Recommender Systems. Deep Learning Examples provides Data Scientist and Software Engineers with recipes to Train, fine-tune, and deploy State-of-the-Art Models

automaticspeechrecognition

clara_holoscan

Clara Discovery is a collection of frameworks, applications, and AI models enabling GPU-accelerated computational drug discovery

claradiscovery

Clara NLP is a collection of SOTA biomedical pre-trained language models as well as highly optimized pipelines for training NLP models on biomedical and clinical text

claranlp

Clara Parabricks is a collection of software tools and notebooks for next generation sequencing, including short- and long-read applications.  These tools are designed to be scalable, generating highly accurate results in an accelerated compute environmen

claraparabricks

Cosmos World Foundation Models: A family of highly performant pre-trained world foundation models purpose-built for generating physics-aware videos and world states for physical AI development.

cosmos

PyTorch - AI Workbench Default Container (Beta)

ucs-tokkio-app-base-1-stream-llm-rag-3d-ov

This example showcases multi modal usecase in a RAG pipeline. It can understand any kind of images in PDF or .pptx (like graphs and plots) alongside text and tables.

rag-app-multimodal-chatbot

This example showcases a RAG workflow with multi-turn conversation capabilities.

rag-app-multiturn-chatbot

Sample RAG application which can handle question-answering from tabular data stored in CSV format.

rag-app-structured-data-chatbot

ChatGLM3-6B is the latest open-source model in the ChatGLM series. ChatGLM3-6B introduces the following features (1) More Powerful Base Model (2) More Comprehensive Function Support (3)  More Comprehensive Open-source Series.