mirror of
https://github.com/elastic/elasticsearch.git
synced 2025-06-29 01:44:36 -04:00
94adaa55c0
The feature branch contains changes to configure PyTorch models with a TrainedModelConfig and defines a format to store the binary models. The _start and _stop deployment actions control the model lifecycle and the model can be directly evaluated with the _infer endpoint. 2 Types of NLP tasks are supported: Named Entity Recognition and Fill Mask. The feature branch consists of these PRs: #73523, #72218, #71679 #71323, #71035, #71177, #70713
59 lines
2.1 KiB
Text
59 lines
2.1 KiB
Text
--
|
|
:api: put-trained-model
|
|
:request: PutTrainedModelRequest
|
|
:response: PutTrainedModelResponse
|
|
--
|
|
[role="xpack"]
|
|
[id="{upid}-{api}"]
|
|
=== Create trained models API
|
|
|
|
|
|
Creates a new trained model for inference.
|
|
The API accepts a +{request}+ object as a request and returns a +{response}+.
|
|
|
|
[id="{upid}-{api}-request"]
|
|
==== Request
|
|
|
|
A +{request}+ requires the following argument:
|
|
|
|
["source","java",subs="attributes,callouts,macros"]
|
|
--------------------------------------------------
|
|
include-tagged::{doc-tests-file}[{api}-request]
|
|
--------------------------------------------------
|
|
<1> The configuration of the {infer} trained model to create
|
|
|
|
[id="{upid}-{api}-config"]
|
|
==== Trained model configuration
|
|
|
|
The `TrainedModelConfig` object contains all the details about the trained model
|
|
configuration and contains the following arguments:
|
|
|
|
["source","java",subs="attributes,callouts,macros"]
|
|
--------------------------------------------------
|
|
include-tagged::{doc-tests-file}[{api}-config]
|
|
--------------------------------------------------
|
|
<1> The {infer} definition for the model
|
|
<2> Optionally, if the {infer} definition is large, you may choose to compress it for transport.
|
|
Do not supply both the compressed and uncompressed definitions.
|
|
<3> The unique model id
|
|
<4> The type of model being configured. If not set the type is inferred from the model definition
|
|
<5> The input field names for the model definition
|
|
<6> Optionally, a human-readable description
|
|
<7> Optionally, an object map contain metadata about the model
|
|
<8> Optionally, an array of tags to organize the model
|
|
<9> The default inference config to use with the model. Must match the underlying
|
|
definition target_type.
|
|
|
|
include::../execution.asciidoc[]
|
|
|
|
[id="{upid}-{api}-response"]
|
|
==== Response
|
|
|
|
The returned +{response}+ contains the newly created trained model.
|
|
The +{response}+ will omit the model definition as a precaution against
|
|
streaming large model definitions back to the client.
|
|
|
|
["source","java",subs="attributes,callouts,macros"]
|
|
--------------------------------------------------
|
|
include-tagged::{doc-tests-file}[{api}-response]
|
|
--------------------------------------------------
|