This guide will get you started with TensorPort. You will learn how to create projects and datasets, and how to run them on TensorPort. Finally, we will take you for a tour around TensorPort's graphical interface and show you how to access TensorBoard right from your project.
To keep things simple, we'll show you how to do all this using a ready-to-go demo model of a self-driving car and a pre-built dataset.
If you prefer video over text, you can also watch this tutorial as a video below. We recommend following the video along with the guide for the best learning experience.
Before we begin, make sure you have all your gear ready:
- A TensorPort account. Go create one if you haven't yet!
- Make sure you have Python installed. TensorPort supports both Python 2.7 and 3.5.
- Install the TensorPort Python package. It's as easy as
pip install tensorport
- Install Git Large File Storage (LFS). TensorPort uses it to handle large datasets.
For this guide you also need the self-driving car demo:
Get a quick overview of how TensorPort is structured.