tf_jetson_nano icon indicating copy to clipboard operation
tf_jetson_nano copied to clipboard

How to run Keras model on Jetson Nano

Run Keras/Tensorflow model on Jetson Nano

How to Run

Require Python 3.5+ and Jupyter notebook installed

Clone or download this repo

git clone https://github.com/Tony607/tf_jetson_nano

Install required libraries for your development machine

pip3 install -r requirements.txt

A. The Keras Image classification model

How to run Keras model on Jetson Nano | DLology Blog

Step1: Convert Keras model into TensorRT model

On development machine or Google Colab

To run locally, start a terminal, then run,

jupyter notebook

In the opened browser window open

Step1_Colab_TensorRT.ipynb

Step2: Make prediction (On Jetson Nano)

To run notebook on Jetson Nano and make it accessible from another machine, in its terminal run,

jupyter notebook --ip=0.0.0.0

In another machine's browser, open the notebook,

http://<Jetson Nano's LAN IP address>:8888/notebooks/Step2_keras-jetson-ImageNet-predict.ipynb

B.Tensorflow object detection model

How to run TensorFlow Object Detection model on Jetson Nano | DLology Blog

Step1: Convert Tensorflow object detection model into TensorRT model

On development machine or Google Colab Notebook

To run locally, start a terminal, then run,

jupyter notebook

In the opened browser window open

Step1_Object_detection_Colab_TensorRT.ipynb

Step2: Make prediction (On Jetson Nano)

To run notebook on Jetson Nano and make it accessible from another machine, in its terminal run,

jupyter notebook --ip=0.0.0.0

In another machine's browser, open the notebook,

http://<Jetson Nano's LAN IP address>:8888/notebooks/Step2_jetson-object-detection-predict.ipynb