SageMaker Migration Workshop

SageMaker Migration Workshop is designed to enable data scientists/ML engineers with no or little SageMaker experience to build, train, and deploy customer's own machine learning models using SageMaker. The primary outcome of the workshop will be:

• Hands-on experience with SageMaker and AWS ML ecosystem for customer's own ML use cases
• An AWS sandbox environment for on-going ML experiments and testing
• A set of working data science notebooks and training scripts for customer ML projects
• A process and mechanism for training and deploy customer's own machine learning models for online prediction and/or batch prediction

Pre-requisite

• Model training scripts (Python preferred)
• Training dataset
• Non-production AWS account with limits increased for notebook instance, training instance, and hosting instance
  • SageMaker Notebook: ml.t2.medium
  • SageMaker Training: ml.m4.xlarge
  • SageMkaer Hosting: ml.m4.xlarge
• A S3 Working bucket for storing data and output
• AWS User accounts with the following IAM permission
  • Managed policy: AmazonSageMakerFullAccess
  • Custom policy for access to a S3 working bucket
  • Custom policy for a subset of IAM permission

Sample policy for S3 bucket access

 {
  "Version": "2012-10-17",
  "Statement": [
    {
      "Effect": "Allow",
      "Action": [
        "s3:GetBucketLocation",
        "s3:ListAllMyBuckets"
      ],
      "Resource": "*"
    },
    {
      "Effect": "Allow",
      "Action": ["s3:ListBucket"],
      "Resource": ["<arn for the s3 bucket>"]
    },
    {
      "Effect": "Allow",
      "Action": [
        "s3:PutObject",
        "s3:GetObject",
        "s3:DeleteObject"
      ],
      "Resource": ["<arn for the S3 bucket>/*"]
    }
  ]
}

Sample policy for SageMaker role management

{
  "Version": "2012-10-17",
  "Statement": {
    "Effect": "Allow",
    "Action": ["iam:CreateRole", "iam:CreatePolicy", "iam:AttachRolePolicy"],
    "Resource": "*"
  }
}

Instruction

Follow the steps below to get started:

1. Follow the instruction below to create a SageMaker notebook

   • Open the Amazon SageMaker console at https://console.aws.amazon.com/sagemaker/.

   • Choose Notebook instances, then choose Create notebook instance.

   • On the Create notebook instance page, provide the following information:

     • For Notebook instance name, type a name for the noteboook instance.

     • For Instance type, choose ml.t2.medium.

     • For IAM role, create an IAM role. Choose Create a new role.

       • on the pop up screen, select Specific S3 Bucket and enter the bucket name for the lab

     • Change volume size in GB if need more than 5GB for data storage for the notebook instance

     • Leave other as defaults

     • Click on Create Notebook Instance

   • After the status changes to InService, click on Open JupyterLab to launch the Jupyter Notebook

2. Once you are inside the JupyterLab environment, follow the instruction below to download the content

   • Click the Terminal icon in the Launcher pane.
   • Type the command below inside the terminal to get into the CD SageMaker directory

   cd SageMaker

   • Type the command below to download the content

   git clone https://github.com/dping1/SageMaker_Migration_Workshop.git

3. You should see a new folder called SageMaker-Migration-Workshop is created inside the left pane. Double click on the folder to list its content. There are folders named step-x. And inside each folder, there is a notebook that starts with step-x-instruction.ipynb. Start with step-1 and follow the instruction in step-1-instruction.ipynb to continue with the workshop.