Getting Started#

This section provides a video- and text-based quick start guide for installing and running TAO.

Requirements#

Hardware Requirements#

TAO is supported on discrete GPUs, such as B200, H100, A100, A100x, A30x, T4, Titan-RTX, and Quadro-RTX.

Note

TAO is not supported on GPUs from or before the Volta generation.

Minimum System Configuration#

  • 8 GB system RAM

  • 4 GB of GPU RAM

  • 8 core CPU

  • At least one NVIDIA GPU

  • 100 GB of SSD space

TAO Getting Started Resource#

To help you get started with the TAO, NVIDIA distributes a package of setup scripts and tutorial notebooks on GitHub, in the tao_tutorials repository.

Enter the following command to download this resource by cloning the repository to your local machine:

git clone https://github.com/NVIDIA/tao_tutorials.git

Note

The tao_tutorials repo provides quickstart scripts for TAO using Launcher CLI and notebooks for both Finetuning Microservice (FTMS) and Launcher modes.

Before using the FTMS notebooks, follow the separate FTMS deployment documentation from Microservices Setup or the video from Getting Started.

The FTMS kubernetes interface setup is not included in the GitHub quickstart scripts by default.

This is the file hierarchy and contents of the package:

setup

   |--> quickstart_launcher.sh

notebooks

   |--> tao_api_starter_kit

      |--> api

            |--> automl
            |--> end2end
            |--> dataset_prepare

      |--> client

            |--> automl
            |--> end2end
            |--> dataset_prepare

   |--> tao_launcher_starter_kit

      |--> dino
      |--> deformable_detr
      |--> classification_pyt
      |--> ocdnet
      |-->  ...

   |--> tao_data_services

      |--> data
      |-->  ...

The tao_tutorials repository has two major components:

  • setup: A set of quick-start scripts to help you install and deploy the TAO launcher and the TAO APIs on various cloud service providers.

  • notebooks: Beginner-friendly end-to-end tutorial notebooks to help you hit the ground running with TAO. The notebooks install TAO, download the required data, and run TAO commands end-to-end for various use cases.

    There are three categories of notebooks:

    • tao_api_starter_kit: End-to-end notebooks that help you learn the features supported by the TAO API model of execution. The notebooks under the api directory use REST API requests, while those in the client directory use the TAO Client CLI to interact with the API server.

    • tao_launcher_starter_kit: Sample notebooks that walk you through the end-to-end workflow for all the computer vision models supported in TAO. You can interact with TAO using the TAO launcher CLI.

    • tao_data_services: Sample notebooks that walk you through the end-to-end workflow of the different dataset manipulation and annotation tools that are included as part of TAO.

Running TAO#

You can run TAO in five different ways:

More advanced users who want to modify functionality and customize the TAO experience can interact with the source code repositories for the different components of TAO. More details on this can be found in the section Running TAO from source.

On this page