SageMaker Housekeeping

S3 Bucket Setup (Lesson 17)

Purpose

• Store data files for model training.
• Save trained models and artifacts post-training.

Steps for Bucket Creation

1. Sign-In:

   • Log in to AWS console using my_admin account.
   • Ensure using the N. Virginia region.

2. Access S3 Service:

   • In AWS Services, search and open the S3 Management Console.

3. Create Bucket:

   • Click on Create bucket.
   • Choose a globally unique name following the convention: prefix-ml-sagemaker.
   • If name conflict occurs, adjust the prefix for uniqueness.
   • Select N. Virginia as the region.

4. Bucket Creation Complete:

   • Click on Create to finalize the bucket setup.
   • The bucket will automatically replicate data across multiple Availability Zones in N. Virginia.

Setup SageMaker Notebook Instance (Lesson 18)

Steps for Setup

1. Sign-In:

   • Log in with the my_admin account to AWS Management Console.

2. Access SageMaker Service:

   • Find SageMaker service and select Notebook instances.

3. Select Region:

   • Choose N. Virginia or a region close to you. Use the same region throughout the course.

4. Create Notebook Instance:

   • Click on Create notebook instance.
   • Name the instance (e.g., SageMakerCourse).
   • Select T3 medium for the server configuration.

5. IAM Role Configuration:

   • Create a new IAM role during instance setup.
   • Grant access to any S3 bucket or specific ones as needed.

6. Instance Creation and Access:

   • Once the instance status is 'In service', access it by clicking Open Jupyter.
   • The homepage of the Jupyter notebook environment will appear.

Key Benefits

• AWS manages patching and maintenance of the notebook instance.
• Stop the instance when not in use to avoid charges and restart as needed.