Sagemaker
Notebook Instances
Notebook instances in AWS SageMaker are used when you need an interactive development environment based on Jupyter Notebooks in the cloud.
AWS Sagemaker notebook instances work similarly to a Jupyter Lab or Notebooks that we use daily, but running on AWS. They will be particularly useful because they do not require extensive local resources when analyzing data.
This way you won't have to worry about whether your computer has enough RAM or processing power!
Furthermore, integration with AWS S3 and other resources will be easier, due to being in the AWS environment.
Create a Notebooks Instance
Tip! 1
Ensure that you have configured AWS credentials. Check this if you get any permission error.
Question! 1
You will need to wait a few minutes for the instance to be available.
Question! 2
Access the Instance
In the output returned by the aws sagemaker describe-notebook-instance
command, look for the Url
.
Question! 3
Tip! 3
After accessing Jupyter Notebook, replace the end of the URL /tree
with /lab
if you want to access the Jupyter Lab version!
You are now accessing a Jupyter Notebook that is running on AWS! Let's bring some resources into this environment:
Initial Exploration!
Question! 4
Question! 5
Question! 6
Question! 7
AWS Sagemaker Examples
Question! 8
Stop Instance
Important!
Delete resources at the end of class!
To prevent unnecessary resource expenditure, instances can be stopped and restarted as needed.
To start:
Important!
Replace YOUR_INSPER_USERNAME
with your Insper user.
To stop:
Important!
Replace YOUR_INSPER_USERNAME
with your Insper user.
To delete:
Important!
Replace YOUR_INSPER_USERNAME
with your Insper user.
References
- Beginning MLOps with MLFlow. Chapter 5.
- https://docs.aws.amazon.com/sagemaker/latest/dg/whatis.html