Adds new parameters to the elasticsearch inference rerank API

Author: kosabogi

specification/inference/_types/CommonTypes.ts

@@ -1306,6 +1306,23 @@ export class ElasticsearchServiceSettings {
    * The maximum value is 32.
    */
   num_threads: integer
+  /**
+   * Only for the `rerank` task type.
+   * Controls the strategy used for processing long documents during inference.
+   *
+   * Possible values:
+   * - `truncate` (default): Processes only the beginning of each document.
+   * - `chunk`: Splits long documents into smaller parts (chunks) before inference.
+   *
+   * To enable chunking, set this value to `chunk`.
+   */
+  long_document_strategy?: string
+  /**
+   * Only for the `rerank` task type.
+   * Limits the number of chunks per document that are sent for inference when chunking is enabled.
+   * If not set, all chunks generated for the document are processed.
+   */
+  max_chunks_per_doc?: integer
 }
 
 export class ElasticsearchTaskSettings {

specification/inference/put_elasticsearch/PutElasticsearchRequest.ts

@@ -26,7 +26,7 @@ import {
   ElasticsearchTaskSettings,
   ElasticsearchTaskType
 } from '@inference/_types/CommonTypes'
-import { InferenceChunkingSettings } from '@inference/_types/Services'
+import { ElasticsearchInferenceChunkingSettings } from '@inference/_types/Services'
 
 /**
  * Create an Elasticsearch inference endpoint.
@@ -78,10 +78,10 @@ export interface Request extends RequestBase {
   }
   body: {
     /**
-     * The chunking configuration object.
+     * The chunking configuration object. For the `rerank` task type, you can enable chunking by setting the `long_document_strategy` parameter to `chunk` in the `service_settings` object.
      * @ext_doc_id inference-chunking
      */
-    chunking_settings?: InferenceChunkingSettings
+    chunking_settings?: ElasticsearchInferenceChunkingSettings
     /**
      * The type of service supported for the specified task type. In this case, `elasticsearch`.
      */