Deep Learning Notebook

Start training your deep learning model in minutes.

Packaged as an Amazon Machine Image and available on the AWS marketplace, this software will give you access to a Jupyter Notebook enabled for interactive Python development, with the most popular data science and deep learning packages pre-installed.

Run Jupyter in the Cloud

Get access to a Jupyter Notebook running in the Cloud within minutes.

The Deep Learning Notebook is an AMI (Amazon Machine Image) available to run on your own AWS account.

AMIs can be launched from the EC2 console (see Launching an AMI from the EC2 console) or from the AWS Marketplace

Features

  • Very short boot time (under 2 minutes)
  • Easy access to your own Jupyter Notebook
  • Keras, Tensorflow, Pytorch pre-installed (see below)
  • Automatic CUDA activation when running on AWS instances with NVIDIA GPUs

Quick Start

  • Subscribe to the AMI: Deep Learning Notebook AMI
  • Configure the EC2 instance type (ex p3 for a GPU enabled instance)
  • Select Launch. The AMI will start momentarily
  • In the EC2 console, find your instance IP (ex: 31.2.3.4) and your instance ID (ex: i-0fe0dc95980ed33f3)
  • Open a web browser to https://<your-instance-ip>:8888. Your password is the instance ID

You can find detailed instructions with screenshots here

GPU Acceleration

Some Title

When launching the AMI on an EC2 instance with a NVIDIA GPU (see AWS EC2 instance types), the NVIDIA drivers and CUDA toolkit will automatically be activated. This lets Tensorflow and PyTorch use GPU acceleration to speed up model training.

PyTorch, Tensorflow, Keras, Scikit-Learn are pre-installed

The latest versions of these popular machine learning frameworks are pre-installed

List of installed software and their versions:

  • Python 3.7
  • Jupyter 1.0
  • Tensorflow 2.4
  • Keras 2.4.3
  • Pytorch 1.7
  • Nvidia CUDA 10.1.243
  • Nvidia CUDNN 7.6.5
  • Scikit-Learn 0.24.1

FAQ

  • I’m unable to connect to the notebook. I get a connection error in my browser

    Please make sure the url you are using starts with https.

    Your url should look like https://1.2.3.4:8888

    This is to ensure that the traffic between your computer and the notebook is encrypted

  • What is an AMI?

    Amazon Machine Images are OS snapshots that can be launched on new AWS instances and configured to run on your own AWS account. AMIs can be customized with specific software and services.

  • How much does it cost?

    The Deep Learning Notebook runs on top of AWS. The costs to run the software depend on the type of the instance being used and the uptime. For more information please see Amazon’s pricing calculator for the product, or contact Amazon.