Update llama specification

Author: Jan-Kazlouski-elastic

output/openapi/elasticsearch-openapi.json

@@ -1560,7 +1560,7 @@ export class LlamaServiceSettings {
   /**
    * The URL endpoint of the Llama stack endpoint.
    * URL must contain:
-   * * For `text_embedding` task - `/v1/openai/v1/embeddings`.
+   * * For `text_embedding` task - `/v1/inference/embeddings`.
    * * For `completion` and `chat_completion` tasks - `/v1/openai/v1/chat/completions`.
    */
   url: string
@@ -1573,26 +1573,10 @@ export class LlamaServiceSettings {
    * @ext_doc_id llama-api-models
    */
   model_id: string
-  /**
-   * A valid API key for accessing Llama stack endpoint that is going to be sent as part of Bearer authentication header.
-   * This field is optional because Llama stack doesn't provide authentication by default.
-   *
-   * IMPORTANT: You need to provide the API key only once, during the inference model creation.
-   * The get inference endpoint API does not retrieve your API key.
-   * After creating the inference model, you cannot change the associated API key.
-   * If you want to use a different API key, delete the inference model and recreate it with the same name and the updated API key.
-   */
-  api_key?: string
   /**
    * For a `text_embedding` task, the maximum number of tokens per input before chunking occurs.
    */
   max_input_tokens?: integer
-  /**
-   * For a `text_embedding` task, the number of dimensions the resulting output embeddings must have.
-   * It is supported only in `text-embedding-3` and later models. If it is not set by user, it defaults to the model returned dimensions.
-   * If model returns embeddings with a different number of dimensions, error is returned.
-   */
-  dimensions?: integer
   /**
    * For a `text_embedding` task, the similarity measure. One of cosine, dot_product, l2_norm.
    */
@@ -1604,14 +1588,6 @@ export class LlamaServiceSettings {
   rate_limit?: RateLimitSetting
 }
 
-export class LlamaTaskSettings {
-  /**
-   * For a `completion` or `text_embedding` task, specify the user issuing the request.
-   * This information can be used for abuse detection.
-   */
-  user?: string
-}
-
 export enum LlamaTaskType {
   text_embedding,
   completion,

output/openapi/elasticsearch-serverless-openapi.json

@@ -23,7 +23,6 @@ import { Duration } from '@_types/Time'
 import {
   LlamaServiceSettings,
   LlamaServiceType,
-  LlamaTaskSettings,
   LlamaTaskType
 } from '@inference/_types/CommonTypes'
 import { InferenceChunkingSettings } from '@inference/_types/Services'
@@ -76,10 +75,5 @@ export interface Request extends RequestBase {
      * Settings used to install the inference model. These settings are specific to the `llama` service.
      */
     service_settings: LlamaServiceSettings
-    /**
-     * Settings to configure the inference task.
-     * These settings are specific to the task type you specified.
-     */
-    task_settings?: LlamaTaskSettings
   }
 }

output/schema/schema.json

@@ -6,9 +6,8 @@ value: |-
   {
     "service": "llama",
     "service_settings": {
-      "url": "http://localhost:8321/v1/openai/v1/embeddings"
+      "url": "http://localhost:8321/v1/inference/embeddings"
       "dimensions": 384,
-      "api_key": "llama-api-key",
       "model_id": "all-MiniLM-L6-v2" 
     }
   }

output/typescript/types.ts

@@ -7,7 +7,6 @@ value: |-
     "service": "llama",
     "service_settings": {
       "url": "http://localhost:8321/v1/openai/v1/chat/completions"
-      "api_key": "llama-api-key",
       "model_id": "llama3.2:3b" 
     }
   }

specification/inference/_types/CommonTypes.ts

@@ -7,7 +7,6 @@ value: |-
     "service": "llama",
     "service_settings": {
       "url": "http://localhost:8321/v1/openai/v1/chat/completions"
-      "api_key": "llama-api-key",
       "model_id": "llama3.2:3b" 
     }
   }