[DOCS] Add Elastic Rerank usage docs #117625
Conversation
leemthompo
added
>docs
|
Documentation preview: |
| [discrete] | ||
| [[inference-example-elastic-reranker]] | ||
| ==== Elastic Rerank via the `elasticsearch` service | ||
|
|
||
| The following example shows how to create an {infer} endpoint called `my-elastic-rerank` to perform a `rerank` task type using the built-in Elastic Rerank cross-encoder model. | ||
|
|
||
| The API request below will automatically download the Elastic Rerank model if it isn't already downloaded and then deploy the model. | ||
| Once deployed, the model can be used for semantic re-ranking with a <<text-similarity-reranker-retriever-example-elastic-rerank,`text_similarity_reranker` retriever>>. | ||
|
|
||
| [source,console] | ||
| ------------------------------------------------------------ | ||
| PUT _inference/rerank/my-elastic-rerank | ||
| { | ||
| "service": "elasticsearch", | ||
| "service_settings": { | ||
| "model_id": ".rerank-v1", <1> | ||
| "num_threads": 1, | ||
| "adaptive_allocations": { <2> | ||
| "enabled": true, | ||
| "min_number_of_allocations": 1, | ||
| "max_number_of_allocations": 10 | ||
| } | ||
| } | ||
| } | ||
| ------------------------------------------------------------ | ||
| // TEST[skip:TBD] | ||
| <1> The `model_id` must be the ID of the built-in Elastic Rerank model: `.rerank-v1`. | ||
| <2> {ml-docs}/ml-nlp-auto-scale.html#nlp-model-adaptive-allocations[Adaptive allocations] will be enabled with the minimum of 1 and the maximum of 10 allocations. |
There was a problem hiding this comment.
this looks great! Maybe we should consider making the max allocations in this example a smaller number like 2 or 4.
|
|
||
| To use semantic re-ranking in {es}, you need to: | ||
|
|
||
| . *Choose a re-ranking model*. |
There was a problem hiding this comment.
What do you think about changing the wording from "Choose a re-ranking model" to "Decide which re-ranking model to use". I think the language in this section like "Use the built-in model" will make people think it is already available (without following step 2, which is the case for our "default" models, but isn't the case for elser rerank yet. Maybe theres a better way to phrase this.
There was a problem hiding this comment.
Absolutely, good point 👍
| ===== Prerequisites | ||
|
|
||
| To use `text_similarity_reranker` you must first set up a `rerank` task using the <<put-inference-api, Create {infer} API>>. | ||
| The `rerank` task should be set up with a machine learning model that can compute text similarity. |
There was a problem hiding this comment.
The
reranktask
we should be referring to the objects that are created with the create inference API as "endpoints" rather than "tasks". I don't think this change needs to be in this PR as many of such instances weren't created in this PR, but I think we need to open an issue to reword this across our documentation.
There was a problem hiding this comment.
I tried to reword AMAP in these files let me know if it looks OK
|
|
||
| Follow these steps: | ||
|
|
||
| . Create an inference endpoint for the `rerank` task using the <<put-inference-api, Create {infer} API>>. |
There was a problem hiding this comment.
in this case the "rerank task" refers to the task type rather than the endpoint-object, so we don't need to change this instance.
| "adaptive_allocations": { <1> | ||
| "enabled": true, | ||
| "min_number_of_allocations": 1, | ||
| "max_number_of_allocations": 10 |
There was a problem hiding this comment.
once again, we may want to decrease the max allocations in this example
Co-authored-by: Max Hniebergall <[email protected]>
|
Pinging @elastic/es-docs (Team:Docs) |
3 participants
https://github.com/elastic/search-team/issues/8757
Updates:
text-similarity-retrieverretriever APIs under Search APIMakes Elastic Rerank the go-to default choice for re-ranking use cases.
Follow up