Skip to content

Latest commit

 

History

History

ar-cnn

README.md

Training an autoregressive convolutional neural network(AR-CNN) model

What is AWS DeepComposer?

AWS DeepComposer provides a creative, hands-on experience for learning generative AI and machine learning. With generative AI, one of the biggest recent advancements in artificial intelligence. With AWS DeepComposer, you can experiment with different generative AI architectures and models by creating and transforming musical inputs and accompaniments to create compositions.

Regardless of your experience with machine learning (ML) or music, you can use AWS DeepComposer to develop a working knowledge of generative AI. AWS DeepComposer includes learning capsules, sample code, and training data to help you understand and use generative AI models.

Note

To use the AWS DeepComposer console and other AWS services, you need an AWS account. If you don't have an account, go to aws.amazon.com and choose Create an AWS Account. For detailed instructions, see Create and Activate an AWS Account.

As a best practice, you should also create an AWS Identity and Access Management (AWS IAM) user with administrator permissions and use that user for all work that doesn’t require root credentials.

To use the AWS Command Line Tool (AWS CLI), create a password for console access, and access keys. For more information, see Creating Your First IAM Admin User and Group in the AWS IAM User Guide.

Prerequisites

The best way to understand how to use this notebook is to read the following learning capsules first:

  1. Introduction to autoregressive convolutional neural networks
  2. A deep dive into training an autoregressive convolutional neural network model AR-CNN model

If you’re unfamiliar with Amazon SageMaker, also read Getting started with Amazon SageMaker notebook instances and SDKsin the Amazon SageMaker Developer Guide.

Request a Service limit increase for use with a Amazon SageMaker Notebook Instance

To train the AR-CNN model using a compute optimized notebook instances will decrease the amount of time required to train the model. When you create your Notebook instance we recommend that you select the ml.c5.4xlarge instance. To use that instance you need to request a service Service limit increase.

  1. Open the AWS Support Center console

  2. On the AWS Support Center page, choose Create Case and then choose Service limit increase.

  3. In the Case classification panel under Limit type, search for SageMaker.

  4. In the Request panel, choose the Region that you are working in. For Resource Type, choose SageMaker Notebooks.

  5. For Limit choose ml.c5.4xlarge instances

  6. For New Limit Value, verify that the value is 1

  7. In Case description, provide a brief explanation of why you need the Service limit increase. For example, I need to use this compute optimized notebook instance to train a deep learning model using TensorFlow.

  8. In Contact options, provide some details about how you would like to be contacted by the AWS service support team on the status of your Service limit increase request.

  9. Choose submit.

After you submit your servive limit increase it can take 2-3 days for it to be approved.

Creating a notebook instance

To start creating a custom AR-CNN model, launch a SageMaker notebook instance.

Note

Jupyter notebooks are open-source web applications that you can use to create and share documents that contain live code, equations, visualizations, and instructions. The AWS DeepComposer Jupyter notebook in this repo contains code that shows how to train a custom autoregressive convolutional neural network (CNN) with SageMaker and AWS DeepComposer.

Creating a SageMaker training instance

  1. Open the Amazon SageMaker console.

  2. In the navigation pane, choose Notebook instances.

  3. On the Notebook instances page, choose Create notebook instance.

  4. On the Create notebook instance page, for Notebook instance name, enter your notebook name and then choose the ml.c5.4xlarge instance.

  5. Choose your IAM Role to set up the correct permissions and encryption. If you already have an Amazon SageMaker IAM role, choose it from the list. If you are new to SageMaker, create an IAM role by choosing Create a new role.

  6. On the Create an IAM role page, choose Any S3 bucket to give your new IAM role access to any S3 bucket you might create.

  7. Choose Create Role.

  8. On the Create notebook instance page, for IAM role, choose your newly created IAM role.

  9. Open the GitHub repositories panel. For Default repository, choose Clone a public Git repository to this notebook instance only.

  10. The GitHub repository called ar-cnn contains the Jupyter notebook required for this custom project. Copy this link and paste it into the field under GitHub repositories.

  11. Choose Create notebook instance.

  12. On the Notebook instances page, choose Open Jupyter to launch your new Jupyter notebook.

  13. On the Jupyter screen, choose *New and then choose the conda_tensorflow_p36 virtual environment. This launches your new Jupyter notebook.

Submitting to the Spin the Model Chartbusters challenge

To submit your composition(s), checkpoint file(s), and model to the Spin the model Chartbusters challenge you will first need to create a public repository on GitHub. Then download your notebook, checkpoint files, and compositions from SageMaker, and upload them to your public repository. Use the link from your public repository to make your submission to the Chartbusters challenge!

Troubleshooting

You can reach out to the AWS DeepComposer engineering team for additional support by visiting the AWS DeepComposer forum

Legal

# The MIT-Zero License

# Copyright 2020 Amazon.com, Inc. or its affiliates. All Rights Reserved.

# Permission is hereby granted, free of charge, to any person obtaining a copy
# of this software and associated documentation files (the "Software"), to deal
# in the Software without restriction, including without limitation the rights
# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
# copies of the Software, and to permit persons to whom the Software is
# furnished to do so.

# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
# THE SOFTWARE.