Challenge Overview
Challenge Overview
In this challenge, we are going to create maching learning workflow with AWS Sagemaker
- Upload data to S3 bucket
- AWS Sagemaker trains a model https://docs.aws.amazon.com/sagemaker/latest/dg/how-it-works-training.html
- Deploy the model on AWS Sagemaker Hosting Service https://docs.aws.amazon.com/sagemaker/latest/dg/how-it-works-hosting.html
Project Background
Topcoder has been running over 52k challenges over the past few years. Thanks to the amazing work from the community, most of these challenges have completed successfully.
Topcoder is reaching out to our creative Topcoder community to
- Research on the historical challenge data
- Build ML workflow using AWS Sagemaker
- Use ML technology to help with
- Easy Topcoder challenge management
- Increase challenge success rates. Challenge success here means one challenge gets one successful submission (winner) in the end.
Technology Stack
- AWS Sagemaker
- Python 3.6
- Sagemaker SDK https://sagemaker.readthedocs.io/en/stable/
Individual requirements
Past challenge data in csv format and sample code are shared in the forum.
The following workflow should be created
- Create S3 bucket and upload csv data to S3 bucket.
- Use Sagemaker SDK to build a model. Sample code is provided only to show the algorithm right now is very trivial, right now we focus on the workflow creatation. Feel free to update the method for better result, but it's not one requirement in this challenge.
- Deploy the model
- Validate deployed model by sending requests to the AWS SageMaker HTTPS endpoint (https://docs.aws.amazon.com/sagemaker/latest/dg/ex1-test-model-endpoint.html)
We don't expect an auto workflow, meaning after uploading csv to S3, it triggers model training then deploying. As long as we follow README.md and we will able to validate result through Sagemaker HTTPS endpoint, it's acceptable.
Please make sure all steps are well documented, and all code should be in separated files.
You can create training job in AWS Sagemaker web console or API, either is acceptable.
If you are not able to launch training instance because of AWS account limit, please send support ticket, normally it will be solved within one hour.
To increase your Sagemaker notebook quota, create a new case on AWS Support (<https://console.aws.amazon.com/support/home>).
Choose - Service Limit Increase
Limit Type - Sagemaker
Resource Type - Sagemaker training
Final Submission Guidelines
- All source code
- README.md including deployment and verification guide