Integrate VectorStore from Elasticsearch client

[maxjakob]

Description

We recently added a VectorStore abstraction to the Elasticsearch client library (elastic/elasticsearch-py#2528, see module) in order to centralize the development and ensure all GenAI library integrations work the same.

This PR integrates the new module into LlamaIndex. The LlamaIndex class keeps its existing interface, it is just extended with the option to specify more retrieval strategies.

Type of Change

• New feature
  • non-breaking change in default usage of the vector store
  • breaking change when using query modes that do not match the retrieval mode specified at init time
  • --> I would suggest making a breaking release with version 0.2.0
• This change requires a documentation update

How Has This Been Tested?

Please describe the tests that you ran to verify your changes. Provide instructions so we can reproduce. Please also list any relevant details for your test configuration

• Adapted existing unit and integration tests. Note that the abstract library already tests its own functionality with respect to the different retrieval strategies.
• I stared at the code and made sure it makes sense

Suggested Checklist:

• I have performed a self-review of my own code
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• I have added Google Colab support for the newly added notebooks.
• My changes generate no new warnings
• I have added tests that prove my fix is effective or that my feature works
• New and existing unit tests pass locally with my changes
• I ran make format; make lint to appease the lint gods

[review-notebook-app]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[logan-markewich]

This looks good to me. I tried running ElasticsearchIndexDemo though and got this error:

NotFoundError: NotFoundError(404, 'resource_not_found_exception', 'Could not find trained model [.elser_model_2]')

Did I miss a step?

[maxjakob]

@logan-markewich The ELSER model needs to be deployed first. I added a comment to this section.

[logan-markewich]

fyi this is why less filters are supported, elastic is only using the legacy exact match filters. If updated, it could use many filters (less than, greater than, contains, etc.) and operators (AND, OR)

[logan-markewich]

Seems like aiohttp needs to be an explicit dependency -- adding that now

[maxjakob]

How odd, CI ran successfully before, then you changed some text in a notebook and from then on it's failing. Let me trigger CI again (without explicit aiohttp installation).

[maxjakob]

Rebased on main.

[maxjakob]

Thank you so much for pushing this over the finish line @logan-markewich!