[DOC]Add Documentation for creating Neural Sparse Remote Model

[saimedhi]

What do you want to do?

• Request a change to existing documentation
• Add new documentation
• Report a technical problem with the documentation
• Other

Tell us about your request. Provide a summary of the request.

• Currently there is no documentation for creating Neural Sparse Remote Model in OpenSearch documentation website. (Example by using SageMaker)
• Below documentation in github is useful for customers to create Neural Sparse Remote Model . @zhichao-aws could you please move the github doc to OpenSearch documentation website.

https://github.com/zhichao-aws/opensearch-neural-sparse-sample/blob/main/examples/deploy_on_sagemaker/run.ipynb

Version: Main branch

What other resources are available? Provide links to related issues, POCs, steps for testing, etc.

[zhichao-aws]

Agree we should provide guide for creating remote neural sparse connectors. One question is, do we support jupyter notebooks at documentation website? If not, is it a valid option to put the guide at ml-commons repo, and link it in documentation website?

[epugh]

There are some tools that let you take a jupter notebook and render them that enables a richer rendering than the say what Github does by default with a .ipynb file. I'm going to look up the one we've used in the past.... nbconvert is another path. We've shipped some scripts that ideally would have been jupter notebooks in the https://github.com/opensearch-project/user-behavior-insights/ repo if we had better support for them in the project!

[kolchfa-aws]

As a first step, we can move the file into ML commons' documentation section. Then we can explore integrating Jekyll with Jupyter notebooks.

[saimedhi]

Hello @zhichao-aws, Do you have any update/ETA on adding doc for "Deploy neural sparse model using sage maker". Could you please prioritize this?

[zhichao-aws]

@saimedhi I've created a PR in ml-commons docs opensearch-project/ml-commons#3857

Since we're having a collaboration with SentenceTransformer to support sparse models, I'd want to wait for their 5.0 release to simplify the deployment scripts before merge.

[zhichao-aws]

The PR is merged. Thanks.