[ML] Adds E5 model configs

[darnautov]

Summary

• Adds E5 model configurations available for download, portable and x86 linux optimized.
• Adds getCuratedModelConfig shared service to retrieve the model ID and configuration appropriate for the current cluster architecture.
• Updates description for the ELSER model
• Renames tabs in the "Add trained model" flyout
• Renames the name property in the ModelDefinitionResponse interface with model_id

Checklist

• Any text added follows EUI's writing guidelines, uses sentence case text and includes i18n support
• Documentation was added for features that require explanation or tutorials
• Unit or functional tests were updated or added to match the most common scenarios

[elasticmachine]

Pinging @elastic/ml-ui (:ml)

[saikatsarkar056]

LGTM

[serenachou]

The Search team is going to use the follow text for copy, proposed by Steve D, it'd be nice if the flyout had the same copy :)
ELSER (Elastic Learned Sparse EncodeR) is Elastic's NLP model for English semantic search, utilizing sparse vectors. It prioritizes intent and contextual meaning over literal term matching, optimized specifically for English documents and queries on the Elastic platform.

E5 (EmbEddings from bidirEctional Encoder rEpresentations) is an NLP model that enables you to perform multi-lingual semantic search by using dense vector representations. This model performs best for non-English language documents and queries.

[jgowdyelastic]

LGTM

[demjened]

LGTM

[spong]

Security Solution/Assistant changes LGTM! Thanks @darnautov!

[darnautov]

@elasticmachine merge upstream

[kibana-ci]

💚 Build Succeeded

• Buildkite Build
• Commit: 9216fa2

Metrics [docs]

Public APIs missing comments

Total count of every public API that lacks a comment. Target amount is 0. Run node scripts/build_api_docs --plugin [yourplugin] --stats comments for more detailed information.

id	before	after	diff
@kbn/ml-trained-models-utils	27	30	+3

Async chunks

Total size of all lazy-loaded chunks that will be downloaded as the user navigates the app

id	before	after	diff
enterpriseSearch	2.6MB	2.6MB	+823.0B
lists	148.4KB	148.4KB	+27.0B
ml	3.6MB	3.6MB	+2.5KB
total			+3.3KB

Page load bundle

Size of the bundles that are downloaded on every page load. Target size is below 100kb

id	before	after	diff
core	374.5KB	374.6KB	+27.0B

Unknown metric groups

API count

id	before	after	diff
@kbn/ml-trained-models-utils	29	32	+3

History

• 💔 Build #180659 failed de7154e
• 💚 Build #180615 succeeded 56fc7cb
• 💔 Build #180561 failed a0b45f2
• 💔 Build #180480 failed 6407bc0
• 💔 Build #180434 failed 232d691

To update your PR or re-run it, just comment with:
@elasticmachine merge upstream

cc @darnautov

[mdefazio]

Apologies for just now looking at this. I know this is now merged (and we are at FF). So I'm dropping these comments here for visibility to those participating on this PR and happy to move them to a follow-up issue if it's easier.

A few thoughts as we look at this flyout with both models added:

1. The tab labels are a bit ambiguous to me and wonder if we can start referring to E5 and ELSER as 'prepared' models?
2. While potentially not a blocker, but with the options segmented into two groups, I'm wondering if its clear that they can only select one of these? Granted once they click on the other model, it should be obvious. However, the keyboard focus sequence jumps to the documentation and then the license link before getting to the second set of radio buttons. This could be problematic.

Here are two updated mocks (again, late to the party...sorry).

Option 1

• Maintain the segmentation of radio buttons with headings + description
• Update tab labels

Option 2

• Move platform choice to be shown after model is selected
• Update tab labels
• Add a one-liner describing 'Prepared models'

[serenachou]

@mdefazio - let's do a bit of user testing on it, and keep the tabs names as is. We can iterate quickly on this so we don't do unnecessary churn on FF right now. as long as we get the license link and documentation as we needed I feel comfortable taking what we have into the release. https://github.com/elastic/ml-team/issues/1075 (WDYT @arisonl ? )

[mdefazio]

@serenachou Sounds good.

One last question though: Should E5 have a download link in the table, like ELSER?

[serenachou]

Nah, no need with this new button to add a model, we strike a good balance of having the primary callout for ELSER but the secondary be for e5 @mdefazio