[ML] add new trained_models/{model_id}/_infer endpoint for all supervised models and deprecate deployment infer api

docs/changelog/86361.yaml

@@ -0,0 +1,5 @@
+pr: 86361
+summary: "Add new _infer endpoint for all supervised models and deprecate deployment infer api"
+area: Machine Learning
+type: enhancement
+issues: []

docs/reference/ml/trained-models/apis/index.asciidoc

@@ -11,6 +11,7 @@ include::delete-trained-models.asciidoc[leveloffset=+2]
 include::get-trained-models.asciidoc[leveloffset=+2]
 include::get-trained-models-stats.asciidoc[leveloffset=+2]
 //INFER
+include::infer-trained-model.asciidoc[leveloffset=+2]
 include::infer-trained-model-deployment.asciidoc[leveloffset=+2]
 //START/STOP
 include::start-trained-model-deployment.asciidoc[leveloffset=+2]

docs/reference/ml/trained-models/apis/infer-trained-model-deployment.asciidoc

@@ -8,6 +8,8 @@
 
 Evaluates a trained model.
 
+deprecated::[8.3.0,Replaced by <<infer-trained-model>>.]
+
 [[infer-trained-model-deployment-request]]
 == {api-request-title}
 
@@ -38,7 +40,7 @@ include::{es-repo-dir}/ml/ml-shared.asciidoc[tag=model-id]
 (Optional, time)
 Controls the amount of time to wait for {infer} results. Defaults to 10 seconds.
 
-[[infer-trained-model-request-body]]
+[[infer-trained-model-deployment-request-body]]
 == {api-request-body-title}
 
 `docs`::

docs/reference/ml/trained-models/apis/infer-trained-model.asciidoc

@@ -0,0 +1,271 @@
+[role="xpack"]
+[[infer-trained-model]]
+= Infer trained model API
+[subs="attributes"]
+++++
+<titleabbrev>Infer trained model</titleabbrev>
+++++
+
+Evaluates a trained model. The model may be any supervised model either trained by {dfanalytics} or imported.
+
+[[infer-trained-model-request]]
+== {api-request-title}
+
+`POST _ml/trained_models/<model_id>/_infer`
+
+////
+[[infer-trained-model-prereq]]
+== {api-prereq-title}
+
+////
+////
+[[infer-trained-model-desc]]
+== {api-description-title}
+
+////
+
+[[infer-trained-model-path-params]]
+== {api-path-parms-title}
+
+`<model_id>`::
+(Required, string)
+include::{es-repo-dir}/ml/ml-shared.asciidoc[tag=model-id]
+
+[[infer-trained-model-query-params]]
+== {api-query-parms-title}
+
+`timeout`::
+(Optional, time)
+Controls the amount of time to wait for {infer} results. Defaults to 10 seconds.
+
+[[infer-trained-model-request-body]]
+== {api-request-body-title}
+
+`docs`::
+(Required, array)
+An array of objects to pass to the model for inference. The objects should
+contain the fields matching your configured trained model input. Typically for NLP models, the field
+name is `text_field`. Currently for NLP models, only a single value is allowed. For {dfanalytics} or
+imported classification or regression models, more than one value is allowed.
+
+////
+[[infer-trained-model-results]]
+== {api-response-body-title}
+////
+////
+[[ml-get-trained-models-response-codes]]
+== {api-response-codes-title}
+
+////
+
+[[infer-trained-model-example]]
+== {api-examples-title}
+
+The response depends on the kind of model.
+
+For example, for language identification the response is the predicted language and the score:
+
+[source,console]
+--------------------------------------------------
+POST _ml/trained_models/lang_ident_model_1/_infer
+{
+  "docs":[{"text": "The fool doth think he is wise, but the wise man knows himself to be a fool."}]
+}
+--------------------------------------------------
+// TEST[skip:TBD]
+
+Here are the results predicting english with a high probability.
+
+[source,console-result]
+----
+{
+  "inference_results": [
+    {
+      "predicted_value": "en",
+      "prediction_probability": 0.9999658805366392,
+      "prediction_score": 0.9999658805366392
+    }
+  ]
+}
+----
+// NOTCONSOLE
+
+
+When it is a text classification model, the response is the score and predicted classification.
+
+For example:
+
+[source,console]
+--------------------------------------------------
+POST _ml/trained_models/model2/_infer
+{
+	"docs": [{"text_field": "The movie was awesome!!"}]
+}
+--------------------------------------------------
+// TEST[skip:TBD]
+
+The API returns the predicted label and the confidence.
+
+[source,console-result]
+----
+{
+  "inference_results": [{
+    "predicted_value" : "POSITIVE",
+    "prediction_probability" : 0.9998667964092964
+  }]
+}
+----
+// NOTCONSOLE
+
+For named entity recognition (NER) models, the response contains the annotated
+text output and the recognized entities.
+
+[source,console]
+--------------------------------------------------
+POST _ml/trained_models/model2/_infer
+{
+	"docs": [{"text_field": "Hi my name is Josh and I live in Berlin"}]
+}
+--------------------------------------------------
+// TEST[skip:TBD]
+
+The API returns in this case:
+
+[source,console-result]
+----
+{
+  "inference_results": [{
+    "predicted_value" : "Hi my name is [Josh](PER&Josh) and I live in [Berlin](LOC&Berlin)",
+    "entities" : [
+      {
+        "entity" : "Josh",
+        "class_name" : "PER",
+        "class_probability" : 0.9977303419824,
+        "start_pos" : 14,
+        "end_pos" : 18
+      },
+      {
+        "entity" : "Berlin",
+        "class_name" : "LOC",
+        "class_probability" : 0.9992474323902818,
+        "start_pos" : 33,
+        "end_pos" : 39
+      }
+    ]
+  }]
+}
+----
+// NOTCONSOLE
+
+Zero-shot classification models require extra configuration defining the class labels.
+These labels are passed in the zero-shot inference config.
+
+[source,console]
+--------------------------------------------------
+POST _ml/trained_models/model2/_infer
+{
+  "docs": [
+    {
+      "text_field": "This is a very happy person"
+    }
+  ],
+  "inference_config": {
+    "zero_shot_classification": {
+      "labels": [
+        "glad",
+        "sad",
+        "bad",
+        "rad"
+      ],
+      "multi_label": false
+    }
+  }
+}
+--------------------------------------------------
+// TEST[skip:TBD]
+
+The API returns the predicted label and the confidence, as well as the top classes:
+
+[source,console-result]
+----
+{
+  "inference_results": [{
+    "predicted_value" : "glad",
+    "top_classes" : [
+      {
+        "class_name" : "glad",
+        "class_probability" : 0.8061155063386439,
+        "class_score" : 0.8061155063386439
+      },
+      {
+        "class_name" : "rad",
+        "class_probability" : 0.18218006158387956,
+        "class_score" : 0.18218006158387956
+      },
+      {
+        "class_name" : "bad",
+        "class_probability" : 0.006325615787634201,
+        "class_score" : 0.006325615787634201
+      },
+      {
+        "class_name" : "sad",
+        "class_probability" : 0.0053788162898424545,
+        "class_score" : 0.0053788162898424545
+      }
+    ],
+    "prediction_probability" : 0.8061155063386439
+  }]
+}
+----
+// NOTCONSOLE
+
+
+The tokenization truncate option can be overridden when calling the API:
+
+[source,console]
+--------------------------------------------------
+POST _ml/trained_models/model2/_infer
+{
+  "docs": [{"text_field": "The Amazon rainforest covers most of the Amazon basin in South America"}],
+  "inference_config": {
+    "ner": {
+      "tokenization": {
+        "bert": {
+          "truncate": "first"
+        }
+      }
+    }
+  }
+}
+--------------------------------------------------
+// TEST[skip:TBD]
+
+When the input has been truncated due to the limit imposed by the model's `max_sequence_length`
+the `is_truncated` field appears in the response.
+
+[source,console-result]
+----
+{
+  "inference_results": [{
+    "predicted_value" : "The [Amazon](LOC&Amazon) rainforest covers most of the [Amazon](LOC&Amazon) basin in [South America](LOC&South+America)",
+    "entities" : [
+      {
+        "entity" : "Amazon",
+        "class_name" : "LOC",
+        "class_probability" : 0.9505460915724254,
+        "start_pos" : 4,
+        "end_pos" : 10
+      },
+      {
+        "entity" : "Amazon",
+        "class_name" : "LOC",
+        "class_probability" : 0.9969992804311777,
+        "start_pos" : 41,
+        "end_pos" : 47
+      }
+    ],
+    "is_truncated" : true
+  }]
+}
+----
+// NOTCONSOLE

docs/reference/ml/trained-models/apis/ml-trained-models-apis.asciidoc

@@ -12,7 +12,8 @@ You can use the following APIs to perform model management operations:
 * <<delete-trained-models-aliases>>
 * <<get-trained-models>>
 * <<get-trained-models-stats>>
-* <<infer-trained-model-deployment>>
+* <<infer-trained-model>>
+* <<infer-trained-model-deployment>> deprecated:[8.3.0]
 * <<start-trained-model-deployment>>
 * <<stop-trained-model-deployment>>
 

docs/reference/ml/trained-models/apis/start-trained-model-deployment.asciidoc

@@ -26,7 +26,7 @@ Requires the `manage_ml` cluster privilege. This privilege is included in the
 Currently only `pytorch` models are supported for deployment. When deployed,
 the model attempts allocation to every machine learning node. Once deployed
 the model can be used by the <<inference-processor,{infer-cap} processor>>
-in an ingest pipeline or directly in the <<infer-trained-model-deployment>> API.
+in an ingest pipeline or directly in the <<infer-trained-model>> API.
 
 [[start-trained-model-deployment-path-params]]
 == {api-path-parms-title}

rest-api-spec/src/main/resources/rest-api-spec/api/ml.infer_trained_model_deployment.json → rest-api-spec/src/main/resources/rest-api-spec/api/ml.infer_trained_model.json

@@ -1,7 +1,7 @@
 {
-  "ml.infer_trained_model_deployment":{
+  "ml.infer_trained_model":{
     "documentation":{
-      "url":"https://www.elastic.co/guide/en/elasticsearch/reference/master/infer-trained-model-deployment.html",
+      "url":"https://www.elastic.co/guide/en/elasticsearch/reference/master/infer-trained-model.html",
       "description":"Evaluate a trained model."
     },
     "stability":"experimental",
@@ -12,6 +12,19 @@
     },
     "url":{
       "paths":[
+        {
+          "path":"/_ml/trained_models/{model_id}/_infer",
+          "methods":[
+            "POST"
+          ],
+          "parts":{
+            "model_id":{
+              "type":"string",
+              "description":"The unique identifier of the trained model.",
+              "required":true
+            }
+          }
+        },
         {
           "path":"/_ml/trained_models/{model_id}/deployment/_infer",
           "methods":[
@@ -23,6 +36,10 @@
               "description":"The unique identifier of the trained model.",
               "required":true
             }
+          },
+          "deprecated": {
+            "version":"8.3.0",
+            "description": "/_ml/trained_models/{model_id}/deployment/_infer is deprecated. Use /_ml/trained_models/{model_id}/_infer instead"
           }
         }
       ]
@@ -36,7 +53,7 @@
       }
     },
     "body":{
-      "description":"The docs to apply inference on",
+      "description":"The docs to apply inference on and inference configuration overrides",
       "required":true
     }
   }

x-pack/plugin/core/src/main/java/org/elasticsearch/xpack/core/XPackClientPlugin.java

@@ -109,7 +109,7 @@
 import org.elasticsearch.xpack.core.ml.action.GetRecordsAction;
 import org.elasticsearch.xpack.core.ml.action.GetTrainedModelsAction;
 import org.elasticsearch.xpack.core.ml.action.GetTrainedModelsStatsAction;
-import org.elasticsearch.xpack.core.ml.action.InternalInferModelAction;
+import org.elasticsearch.xpack.core.ml.action.InferModelAction;
 import org.elasticsearch.xpack.core.ml.action.IsolateDatafeedAction;
 import org.elasticsearch.xpack.core.ml.action.KillProcessAction;
 import org.elasticsearch.xpack.core.ml.action.MlInfoAction;
@@ -319,7 +319,8 @@ public List<ActionType<? extends ActionResponse>> getClientActions() {
                 StartDataFrameAnalyticsAction.INSTANCE,
                 EvaluateDataFrameAction.INSTANCE,
                 ExplainDataFrameAnalyticsAction.INSTANCE,
-                InternalInferModelAction.INSTANCE,
+                InferModelAction.INSTANCE,
+                InferModelAction.EXTERNAL_INSTANCE,
                 GetTrainedModelsAction.INSTANCE,
                 DeleteTrainedModelAction.INSTANCE,
                 GetTrainedModelsStatsAction.INSTANCE,