User Guide

Jetson AGX™ Orin Developer Kit Reviewer's Guide
17
cd $HOME/Review/TAO-PTM
bash installTAOPTM.sh
Installation will take couple of minutes.
After installation is complete, start the jupyter-lab by following the commands below:
cd $HOME/Review/TAO-PTM/
export PATH="$HOME/.local/bin/:$PATH"
jupyter-lab
Click on the link in the end of the output of the above command and it will open Jupyter notebook on
the browser.
Deploy Pre-Trained Models on Jetson
In this section, you will deploy the Peoplenet deployable model using DeepStream and see it in action.
In the Jupyter notebook on the browser, navigate (navigation pane on the left) to
PreTrainedModel
directory and open the notebook named peoplenet_workflow.ipynb. You can read about the PeopleNet
model: the model architecture, training data and accuracy in the notebook.
Train using NVIDIA TAO and Deploy on Jetson
In this section, you will start from a trainable PeopleNet model and train in the cloud for an additional
detection class. After training, you will download the trained model to Jetson and deploy using
DeepStream. In the Jupyter notebook on the browser, navigate (navigation pane on the left) to
TrainAdaptOptimize directory and open the notebook named tao_workflow.ipynb
For your review, we have provided minimal set of data for training and training for 500 epochs. The
quality of detections benefits from the amount and quality of dataset. Since we have kept the data set
minimal in order to shorten the training time, the model generated will not be of a production quality.
NVIDIA RIVA
In this demo, you will experience NVIDIA RIVA ASR in action. You will need a headset with mic which
we have shipped to you in the reviewers package. Please connect the USB headset with mic to the
developer kit before starting on the demos below.
Create a directory and name it as “Review” in your home directory if already not done.
mkdir $HOME/Review