TAO CLI Pre-Requisite Installation Guide

(Note!!!) There are 2 Methods to get TAO running on a system (CLI launch and Container launch). This guide only covers the CLI Launch Methods.

  • Installing Docker:

    Docker is an open source platform for building, deploying and managing container as applications.

    Within the terminal install the most recent docker:

    curl https://get.docker.com | sh \
> && sudo systemctl --now enable docker

  • Installing NVIDIA container:

    distribution=$(. /etc/os-release;echo $ID$VERSION_ID) \
> && curl -s -L https://nvidia.github.io/nvidia-docker/gpgkey | sudo apt-key add - \
> && curl -s -L https://nvidia.github.io/nvidia-docker/$distribution/nvidia-docker.list | sudo tee /etc/apt sources.list.d/nvidia-docker.list

    sudo apt-get update

    sudo apt-get install -y nvidia-docker2

    sudo systemctl restart docker

    IMPORTANT!!! Make sure the system cuda version is the same as the one being installed on the docker cuda

    IMPORTANT!!! Install the latest version of cuda ([https://hub.docker.com/r/nvidia/cuda](https://hub.docker.com/r/nvidia/cuda))

    sudo docker run --rm --gpus all nvidia/cuda:12.0.0-base-ubuntu20.04 nvidia-smi

    • Test all the installation with

    docker login nvcr.io
    • NGC Account

      • For this step, one must log into the NGC NVIDIA account and gain their individual api token. After the token is inputted into the (http://nvcr.io) (next step)

    • Python Virtual Environment (conda installation recommended)
      • Create a virtual environment with python ≥ 3.6.9

        • for conda initialization bash path might need to be established

        source ~/miniconda3/etc/profile.d/conda.sh

    (Note!!!) The NVIDIA Provides getting started pack with necessary libraries

  • Installing pre-requisite files

    wget --content-disposition https://api.ngc.nvidia.com/v2/resources/nvidia/tao/tao-getting-started/versions/4.0.0/zip -O getting_started_v4.0.0.zip
unzip -u getting_started_v4.0.0.zip  -d ./getting_started_v4.0.0 && rm -rf getting_started_v4.0.0.zip && cd ./getting_started_v4.0.0

  • Install TAO launcher with the getting started pack

    bash setup/quickstart_launcher.sh --install

    • Check for TAO version. If There exists errors or dependency problems when tao info line is run, check the cuda version of the host file and cuda version of docker.

    • When running the TAO launcher, some dependency issue might appear.

  • Update the launcher

    bash setup/quickstart_launcher.sh --upgrade

    • Make sure that there are no warnings, (especially GPU dependency warning!!!)

      Example output (tao —help):

      usage: tao [-h]
        {list,stop,info,action_recognition,augment,bpnet,classification_tf1,classification_tf2,converter,deformable_detr,detectnet_v2,dssd,efficientdet_tf1,efficientdet_tf2,emotionnet,faster_rcnn,fpenet,gazenet,gesturenet,heartratenet,intent_slot_classification,lprnet,mask_rcnn,multitask_classification,n_gram,pointpillars,pose_classification,punctuation_and_capitalization,question_answering,re_identification,retinanet,segformer,spectro_gen,speech_to_text,speech_to_text_citrinet,speech_to_text_conformer,ssd,text_classification,token_classification,unet,vocoder,yolo_v3,yolo_v4,yolo_v4_tiny}
        ...

Launcher for TAO Toolkit.

optional arguments:
-h, --help            show this help message and exit

tasks:
{list,stop,info,action_recognition,augment,bpnet,classification_tf1,classification_tf2,converter,deformable_detr,detectnet_v2,dssd,efficientdet_tf1,efficientdet_tf2,emotionnet,faster_rcnn,fpenet,gazenet,gesturenet,heartratenet,intent_slot_classification,lprnet,mask_rcnn,multitask_classification,n_gram,pointpillars,pose_classification,punctuation_and_capitalization,question_answering,re_identification,retinanet,segformer,spectro_gen,speech_to_text,speech_to_text_citrinet,speech_to_text_conformer,ssd,text_classification,token_classification,unet,vocoder,yolo_v3,yolo_v4,yolo_v4_tiny}