From 514f18190504290aa3385d7de3cf5ee6e12a8819 Mon Sep 17 00:00:00 2001 From: lcawl Date: Fri, 13 Dec 2024 18:39:29 -0800 Subject: [PATCH 1/5] [DOCS] Add warning to create inference API --- specification/inference/put/PutRequest.ts | 11 ++++++++++- 1 file changed, 10 insertions(+), 1 deletion(-) diff --git a/specification/inference/put/PutRequest.ts b/specification/inference/put/PutRequest.ts index be83f3ace8..a14c366522 100644 --- a/specification/inference/put/PutRequest.ts +++ b/specification/inference/put/PutRequest.ts @@ -23,10 +23,19 @@ import { RequestBase } from '@_types/Base' import { Id } from '@_types/common' /** - * Create an inference endpoint + * Create an inference endpoint. + * When you create an inference endpoint, the associated machine learning model is automatically deployed if it is not already running. + * After creating the endpoint, wait for the model deployment to complete before using it. + * To verify the deployment status, use the get trained model statistics API. + * Look for `"state": "fully_allocated"` in the response and ensure that the `"allocation_count"` matches the `"target_allocation_count"`. + * Avoid creating multiple endpoints for the same model unless required, as each endpoint consumes significant resources. + * + * IMPORTANT: The inference APIs enable you to use certain services, such as built-in machine learning models (ELSER, E5), models uploaded through Eland, Cohere, OpenAI, Mistral, Azure OpenAI, Google AI Studio, Google Vertex AI, Anthropic, Watsonx.ai, or Hugging Face. + * For built-in models and models uploaded through Eland, the inference APIs offer an alternative way to use and manage trained models. However, if you do not plan to use the inference APIs to use these models or if you want to use non-NLP models, use the machine learning trained model APIs. * @rest_spec_name inference.put * @availability stack since=8.11.0 stability=stable visibility=public * @availability serverless stability=stable visibility=public + * @cluster_privileges manage_inference */ export interface Request extends RequestBase { path_parts: { From b637ad6a9d0c03f467414570e28cc38fb67578de Mon Sep 17 00:00:00 2001 From: lcawl Date: Fri, 13 Dec 2024 18:43:20 -0800 Subject: [PATCH 2/5] Generate output --- compiler/src/model/utils.ts | 2 +- output/openapi/elasticsearch-openapi.json | 2 ++ output/openapi/elasticsearch-serverless-openapi.json | 2 ++ output/schema/schema.json | 11 ++++++++--- 4 files changed, 13 insertions(+), 4 deletions(-) diff --git a/compiler/src/model/utils.ts b/compiler/src/model/utils.ts index 769de32bfc..9f4d3e669e 100644 --- a/compiler/src/model/utils.ts +++ b/compiler/src/model/utils.ts @@ -667,7 +667,7 @@ export function hoistRequestAnnotations ( } else if (tag === 'cluster_privileges') { const privileges = [ 'all', 'cancel_task', 'create_snapshot', 'grant_api_key', 'manage', 'manage_api_key', 'manage_ccr', - 'manage_enrich', 'manage_ilm', 'manage_index_templates', 'manage_ingest_pipelines', 'manage_logstash_pipelines', + 'manage_enrich', 'manage_ilm', 'manage_index_templates', 'manage_inference', 'manage_ingest_pipelines', 'manage_logstash_pipelines', 'manage_ml', 'manage_oidc', 'manage_own_api_key', 'manage_pipeline', 'manage_rollup', 'manage_saml', 'manage_security', 'manage_service_account', 'manage_slm', 'manage_token', 'manage_transform', 'manage_user_profile', 'manage_watcher', 'monitor', 'monitor_ml', 'monitor_rollup', 'monitor_snapshot', 'monitor_text_structure', diff --git a/output/openapi/elasticsearch-openapi.json b/output/openapi/elasticsearch-openapi.json index 67c12225ee..72bc65ae1e 100644 --- a/output/openapi/elasticsearch-openapi.json +++ b/output/openapi/elasticsearch-openapi.json @@ -14997,6 +14997,7 @@ "inference" ], "summary": "Create an inference endpoint", + "description": "When you create an inference endpoint, the associated machine learning model is automatically deployed if it is not already running.\nAfter creating the endpoint, wait for the model deployment to complete before using it.\nTo verify the deployment status, use the get trained model statistics API. \nLook for `\"state\": \"fully_allocated\"` in the response and ensure that the `\"allocation_count\"` matches the `\"target_allocation_count\"`.\nAvoid creating multiple endpoints for the same model unless required, as each endpoint consumes significant resources.\n\nIMPORTANT: The inference APIs enable you to use certain services, such as built-in machine learning models (ELSER, E5), models uploaded through Eland, Cohere, OpenAI, Mistral, Azure OpenAI, Google AI Studio, Google Vertex AI, Anthropic, Watsonx.ai, or Hugging Face.\nFor built-in models and models uploaded through Eland, the inference APIs offer an alternative way to use and manage trained models. However, if you do not plan to use the inference APIs to use these models or if you want to use non-NLP models, use the machine learning trained model APIs.", "operationId": "inference-put", "parameters": [ { @@ -15089,6 +15090,7 @@ "inference" ], "summary": "Create an inference endpoint", + "description": "When you create an inference endpoint, the associated machine learning model is automatically deployed if it is not already running.\nAfter creating the endpoint, wait for the model deployment to complete before using it.\nTo verify the deployment status, use the get trained model statistics API. \nLook for `\"state\": \"fully_allocated\"` in the response and ensure that the `\"allocation_count\"` matches the `\"target_allocation_count\"`.\nAvoid creating multiple endpoints for the same model unless required, as each endpoint consumes significant resources.\n\nIMPORTANT: The inference APIs enable you to use certain services, such as built-in machine learning models (ELSER, E5), models uploaded through Eland, Cohere, OpenAI, Mistral, Azure OpenAI, Google AI Studio, Google Vertex AI, Anthropic, Watsonx.ai, or Hugging Face.\nFor built-in models and models uploaded through Eland, the inference APIs offer an alternative way to use and manage trained models. However, if you do not plan to use the inference APIs to use these models or if you want to use non-NLP models, use the machine learning trained model APIs.", "operationId": "inference-put-1", "parameters": [ { diff --git a/output/openapi/elasticsearch-serverless-openapi.json b/output/openapi/elasticsearch-serverless-openapi.json index 7e639ab6c6..66bb399180 100644 --- a/output/openapi/elasticsearch-serverless-openapi.json +++ b/output/openapi/elasticsearch-serverless-openapi.json @@ -8579,6 +8579,7 @@ "inference" ], "summary": "Create an inference endpoint", + "description": "When you create an inference endpoint, the associated machine learning model is automatically deployed if it is not already running.\nAfter creating the endpoint, wait for the model deployment to complete before using it.\nTo verify the deployment status, use the get trained model statistics API. \nLook for `\"state\": \"fully_allocated\"` in the response and ensure that the `\"allocation_count\"` matches the `\"target_allocation_count\"`.\nAvoid creating multiple endpoints for the same model unless required, as each endpoint consumes significant resources.\n\nIMPORTANT: The inference APIs enable you to use certain services, such as built-in machine learning models (ELSER, E5), models uploaded through Eland, Cohere, OpenAI, Mistral, Azure OpenAI, Google AI Studio, Google Vertex AI, Anthropic, Watsonx.ai, or Hugging Face.\nFor built-in models and models uploaded through Eland, the inference APIs offer an alternative way to use and manage trained models. However, if you do not plan to use the inference APIs to use these models or if you want to use non-NLP models, use the machine learning trained model APIs.", "operationId": "inference-put", "parameters": [ { @@ -8671,6 +8672,7 @@ "inference" ], "summary": "Create an inference endpoint", + "description": "When you create an inference endpoint, the associated machine learning model is automatically deployed if it is not already running.\nAfter creating the endpoint, wait for the model deployment to complete before using it.\nTo verify the deployment status, use the get trained model statistics API. \nLook for `\"state\": \"fully_allocated\"` in the response and ensure that the `\"allocation_count\"` matches the `\"target_allocation_count\"`.\nAvoid creating multiple endpoints for the same model unless required, as each endpoint consumes significant resources.\n\nIMPORTANT: The inference APIs enable you to use certain services, such as built-in machine learning models (ELSER, E5), models uploaded through Eland, Cohere, OpenAI, Mistral, Azure OpenAI, Google AI Studio, Google Vertex AI, Anthropic, Watsonx.ai, or Hugging Face.\nFor built-in models and models uploaded through Eland, the inference APIs offer an alternative way to use and manage trained models. However, if you do not plan to use the inference APIs to use these models or if you want to use non-NLP models, use the machine learning trained model APIs.", "operationId": "inference-put-1", "parameters": [ { diff --git a/output/schema/schema.json b/output/schema/schema.json index 478520ccf2..7a2d4e5b23 100644 --- a/output/schema/schema.json +++ b/output/schema/schema.json @@ -8525,9 +8525,14 @@ "visibility": "public" } }, - "description": "Create an inference endpoint", + "description": "Create an inference endpoint.\nWhen you create an inference endpoint, the associated machine learning model is automatically deployed if it is not already running.\nAfter creating the endpoint, wait for the model deployment to complete before using it.\nTo verify the deployment status, use the get trained model statistics API. \nLook for `\"state\": \"fully_allocated\"` in the response and ensure that the `\"allocation_count\"` matches the `\"target_allocation_count\"`.\nAvoid creating multiple endpoints for the same model unless required, as each endpoint consumes significant resources.\n\nIMPORTANT: The inference APIs enable you to use certain services, such as built-in machine learning models (ELSER, E5), models uploaded through Eland, Cohere, OpenAI, Mistral, Azure OpenAI, Google AI Studio, Google Vertex AI, Anthropic, Watsonx.ai, or Hugging Face.\nFor built-in models and models uploaded through Eland, the inference APIs offer an alternative way to use and manage trained models. However, if you do not plan to use the inference APIs to use these models or if you want to use non-NLP models, use the machine learning trained model APIs.", "docUrl": "https://www.elastic.co/guide/en/elasticsearch/reference/master/put-inference-api.html", "name": "inference.put", + "privileges": { + "cluster": [ + "manage_inference" + ] + }, "request": { "name": "Request", "namespace": "inference.put" @@ -141784,7 +141789,7 @@ } } }, - "description": "Create an inference endpoint", + "description": "Create an inference endpoint.\nWhen you create an inference endpoint, the associated machine learning model is automatically deployed if it is not already running.\nAfter creating the endpoint, wait for the model deployment to complete before using it.\nTo verify the deployment status, use the get trained model statistics API. \nLook for `\"state\": \"fully_allocated\"` in the response and ensure that the `\"allocation_count\"` matches the `\"target_allocation_count\"`.\nAvoid creating multiple endpoints for the same model unless required, as each endpoint consumes significant resources.\n\nIMPORTANT: The inference APIs enable you to use certain services, such as built-in machine learning models (ELSER, E5), models uploaded through Eland, Cohere, OpenAI, Mistral, Azure OpenAI, Google AI Studio, Google Vertex AI, Anthropic, Watsonx.ai, or Hugging Face.\nFor built-in models and models uploaded through Eland, the inference APIs offer an alternative way to use and manage trained models. However, if you do not plan to use the inference APIs to use these models or if you want to use non-NLP models, use the machine learning trained model APIs.", "inherits": { "type": { "name": "RequestBase", @@ -141822,7 +141827,7 @@ } ], "query": [], - "specLocation": "inference/put/PutRequest.ts#L25-L44" + "specLocation": "inference/put/PutRequest.ts#L25-L53" }, { "kind": "response", From 6656cb4bab01a915879d027ebb0b7182185cf239 Mon Sep 17 00:00:00 2001 From: lcawl Date: Fri, 13 Dec 2024 18:52:31 -0800 Subject: [PATCH 3/5] Style fix --- output/openapi/elasticsearch-openapi.json | 4 ++-- output/openapi/elasticsearch-serverless-openapi.json | 4 ++-- output/schema/schema.json | 4 ++-- specification/inference/put/PutRequest.ts | 2 +- 4 files changed, 7 insertions(+), 7 deletions(-) diff --git a/output/openapi/elasticsearch-openapi.json b/output/openapi/elasticsearch-openapi.json index 72bc65ae1e..42f8b0870e 100644 --- a/output/openapi/elasticsearch-openapi.json +++ b/output/openapi/elasticsearch-openapi.json @@ -14997,7 +14997,7 @@ "inference" ], "summary": "Create an inference endpoint", - "description": "When you create an inference endpoint, the associated machine learning model is automatically deployed if it is not already running.\nAfter creating the endpoint, wait for the model deployment to complete before using it.\nTo verify the deployment status, use the get trained model statistics API. \nLook for `\"state\": \"fully_allocated\"` in the response and ensure that the `\"allocation_count\"` matches the `\"target_allocation_count\"`.\nAvoid creating multiple endpoints for the same model unless required, as each endpoint consumes significant resources.\n\nIMPORTANT: The inference APIs enable you to use certain services, such as built-in machine learning models (ELSER, E5), models uploaded through Eland, Cohere, OpenAI, Mistral, Azure OpenAI, Google AI Studio, Google Vertex AI, Anthropic, Watsonx.ai, or Hugging Face.\nFor built-in models and models uploaded through Eland, the inference APIs offer an alternative way to use and manage trained models. However, if you do not plan to use the inference APIs to use these models or if you want to use non-NLP models, use the machine learning trained model APIs.", + "description": "When you create an inference endpoint, the associated machine learning model is automatically deployed if it is not already running.\nAfter creating the endpoint, wait for the model deployment to complete before using it.\nTo verify the deployment status, use the get trained model statistics API.\nLook for `\"state\": \"fully_allocated\"` in the response and ensure that the `\"allocation_count\"` matches the `\"target_allocation_count\"`.\nAvoid creating multiple endpoints for the same model unless required, as each endpoint consumes significant resources.\n\nIMPORTANT: The inference APIs enable you to use certain services, such as built-in machine learning models (ELSER, E5), models uploaded through Eland, Cohere, OpenAI, Mistral, Azure OpenAI, Google AI Studio, Google Vertex AI, Anthropic, Watsonx.ai, or Hugging Face.\nFor built-in models and models uploaded through Eland, the inference APIs offer an alternative way to use and manage trained models. However, if you do not plan to use the inference APIs to use these models or if you want to use non-NLP models, use the machine learning trained model APIs.", "operationId": "inference-put", "parameters": [ { @@ -15090,7 +15090,7 @@ "inference" ], "summary": "Create an inference endpoint", - "description": "When you create an inference endpoint, the associated machine learning model is automatically deployed if it is not already running.\nAfter creating the endpoint, wait for the model deployment to complete before using it.\nTo verify the deployment status, use the get trained model statistics API. \nLook for `\"state\": \"fully_allocated\"` in the response and ensure that the `\"allocation_count\"` matches the `\"target_allocation_count\"`.\nAvoid creating multiple endpoints for the same model unless required, as each endpoint consumes significant resources.\n\nIMPORTANT: The inference APIs enable you to use certain services, such as built-in machine learning models (ELSER, E5), models uploaded through Eland, Cohere, OpenAI, Mistral, Azure OpenAI, Google AI Studio, Google Vertex AI, Anthropic, Watsonx.ai, or Hugging Face.\nFor built-in models and models uploaded through Eland, the inference APIs offer an alternative way to use and manage trained models. However, if you do not plan to use the inference APIs to use these models or if you want to use non-NLP models, use the machine learning trained model APIs.", + "description": "When you create an inference endpoint, the associated machine learning model is automatically deployed if it is not already running.\nAfter creating the endpoint, wait for the model deployment to complete before using it.\nTo verify the deployment status, use the get trained model statistics API.\nLook for `\"state\": \"fully_allocated\"` in the response and ensure that the `\"allocation_count\"` matches the `\"target_allocation_count\"`.\nAvoid creating multiple endpoints for the same model unless required, as each endpoint consumes significant resources.\n\nIMPORTANT: The inference APIs enable you to use certain services, such as built-in machine learning models (ELSER, E5), models uploaded through Eland, Cohere, OpenAI, Mistral, Azure OpenAI, Google AI Studio, Google Vertex AI, Anthropic, Watsonx.ai, or Hugging Face.\nFor built-in models and models uploaded through Eland, the inference APIs offer an alternative way to use and manage trained models. However, if you do not plan to use the inference APIs to use these models or if you want to use non-NLP models, use the machine learning trained model APIs.", "operationId": "inference-put-1", "parameters": [ { diff --git a/output/openapi/elasticsearch-serverless-openapi.json b/output/openapi/elasticsearch-serverless-openapi.json index 66bb399180..a45034c8b5 100644 --- a/output/openapi/elasticsearch-serverless-openapi.json +++ b/output/openapi/elasticsearch-serverless-openapi.json @@ -8579,7 +8579,7 @@ "inference" ], "summary": "Create an inference endpoint", - "description": "When you create an inference endpoint, the associated machine learning model is automatically deployed if it is not already running.\nAfter creating the endpoint, wait for the model deployment to complete before using it.\nTo verify the deployment status, use the get trained model statistics API. \nLook for `\"state\": \"fully_allocated\"` in the response and ensure that the `\"allocation_count\"` matches the `\"target_allocation_count\"`.\nAvoid creating multiple endpoints for the same model unless required, as each endpoint consumes significant resources.\n\nIMPORTANT: The inference APIs enable you to use certain services, such as built-in machine learning models (ELSER, E5), models uploaded through Eland, Cohere, OpenAI, Mistral, Azure OpenAI, Google AI Studio, Google Vertex AI, Anthropic, Watsonx.ai, or Hugging Face.\nFor built-in models and models uploaded through Eland, the inference APIs offer an alternative way to use and manage trained models. However, if you do not plan to use the inference APIs to use these models or if you want to use non-NLP models, use the machine learning trained model APIs.", + "description": "When you create an inference endpoint, the associated machine learning model is automatically deployed if it is not already running.\nAfter creating the endpoint, wait for the model deployment to complete before using it.\nTo verify the deployment status, use the get trained model statistics API.\nLook for `\"state\": \"fully_allocated\"` in the response and ensure that the `\"allocation_count\"` matches the `\"target_allocation_count\"`.\nAvoid creating multiple endpoints for the same model unless required, as each endpoint consumes significant resources.\n\nIMPORTANT: The inference APIs enable you to use certain services, such as built-in machine learning models (ELSER, E5), models uploaded through Eland, Cohere, OpenAI, Mistral, Azure OpenAI, Google AI Studio, Google Vertex AI, Anthropic, Watsonx.ai, or Hugging Face.\nFor built-in models and models uploaded through Eland, the inference APIs offer an alternative way to use and manage trained models. However, if you do not plan to use the inference APIs to use these models or if you want to use non-NLP models, use the machine learning trained model APIs.", "operationId": "inference-put", "parameters": [ { @@ -8672,7 +8672,7 @@ "inference" ], "summary": "Create an inference endpoint", - "description": "When you create an inference endpoint, the associated machine learning model is automatically deployed if it is not already running.\nAfter creating the endpoint, wait for the model deployment to complete before using it.\nTo verify the deployment status, use the get trained model statistics API. \nLook for `\"state\": \"fully_allocated\"` in the response and ensure that the `\"allocation_count\"` matches the `\"target_allocation_count\"`.\nAvoid creating multiple endpoints for the same model unless required, as each endpoint consumes significant resources.\n\nIMPORTANT: The inference APIs enable you to use certain services, such as built-in machine learning models (ELSER, E5), models uploaded through Eland, Cohere, OpenAI, Mistral, Azure OpenAI, Google AI Studio, Google Vertex AI, Anthropic, Watsonx.ai, or Hugging Face.\nFor built-in models and models uploaded through Eland, the inference APIs offer an alternative way to use and manage trained models. However, if you do not plan to use the inference APIs to use these models or if you want to use non-NLP models, use the machine learning trained model APIs.", + "description": "When you create an inference endpoint, the associated machine learning model is automatically deployed if it is not already running.\nAfter creating the endpoint, wait for the model deployment to complete before using it.\nTo verify the deployment status, use the get trained model statistics API.\nLook for `\"state\": \"fully_allocated\"` in the response and ensure that the `\"allocation_count\"` matches the `\"target_allocation_count\"`.\nAvoid creating multiple endpoints for the same model unless required, as each endpoint consumes significant resources.\n\nIMPORTANT: The inference APIs enable you to use certain services, such as built-in machine learning models (ELSER, E5), models uploaded through Eland, Cohere, OpenAI, Mistral, Azure OpenAI, Google AI Studio, Google Vertex AI, Anthropic, Watsonx.ai, or Hugging Face.\nFor built-in models and models uploaded through Eland, the inference APIs offer an alternative way to use and manage trained models. However, if you do not plan to use the inference APIs to use these models or if you want to use non-NLP models, use the machine learning trained model APIs.", "operationId": "inference-put-1", "parameters": [ { diff --git a/output/schema/schema.json b/output/schema/schema.json index 7a2d4e5b23..c2f97f7f69 100644 --- a/output/schema/schema.json +++ b/output/schema/schema.json @@ -8525,7 +8525,7 @@ "visibility": "public" } }, - "description": "Create an inference endpoint.\nWhen you create an inference endpoint, the associated machine learning model is automatically deployed if it is not already running.\nAfter creating the endpoint, wait for the model deployment to complete before using it.\nTo verify the deployment status, use the get trained model statistics API. \nLook for `\"state\": \"fully_allocated\"` in the response and ensure that the `\"allocation_count\"` matches the `\"target_allocation_count\"`.\nAvoid creating multiple endpoints for the same model unless required, as each endpoint consumes significant resources.\n\nIMPORTANT: The inference APIs enable you to use certain services, such as built-in machine learning models (ELSER, E5), models uploaded through Eland, Cohere, OpenAI, Mistral, Azure OpenAI, Google AI Studio, Google Vertex AI, Anthropic, Watsonx.ai, or Hugging Face.\nFor built-in models and models uploaded through Eland, the inference APIs offer an alternative way to use and manage trained models. However, if you do not plan to use the inference APIs to use these models or if you want to use non-NLP models, use the machine learning trained model APIs.", + "description": "Create an inference endpoint.\nWhen you create an inference endpoint, the associated machine learning model is automatically deployed if it is not already running.\nAfter creating the endpoint, wait for the model deployment to complete before using it.\nTo verify the deployment status, use the get trained model statistics API.\nLook for `\"state\": \"fully_allocated\"` in the response and ensure that the `\"allocation_count\"` matches the `\"target_allocation_count\"`.\nAvoid creating multiple endpoints for the same model unless required, as each endpoint consumes significant resources.\n\nIMPORTANT: The inference APIs enable you to use certain services, such as built-in machine learning models (ELSER, E5), models uploaded through Eland, Cohere, OpenAI, Mistral, Azure OpenAI, Google AI Studio, Google Vertex AI, Anthropic, Watsonx.ai, or Hugging Face.\nFor built-in models and models uploaded through Eland, the inference APIs offer an alternative way to use and manage trained models. However, if you do not plan to use the inference APIs to use these models or if you want to use non-NLP models, use the machine learning trained model APIs.", "docUrl": "https://www.elastic.co/guide/en/elasticsearch/reference/master/put-inference-api.html", "name": "inference.put", "privileges": { @@ -141789,7 +141789,7 @@ } } }, - "description": "Create an inference endpoint.\nWhen you create an inference endpoint, the associated machine learning model is automatically deployed if it is not already running.\nAfter creating the endpoint, wait for the model deployment to complete before using it.\nTo verify the deployment status, use the get trained model statistics API. \nLook for `\"state\": \"fully_allocated\"` in the response and ensure that the `\"allocation_count\"` matches the `\"target_allocation_count\"`.\nAvoid creating multiple endpoints for the same model unless required, as each endpoint consumes significant resources.\n\nIMPORTANT: The inference APIs enable you to use certain services, such as built-in machine learning models (ELSER, E5), models uploaded through Eland, Cohere, OpenAI, Mistral, Azure OpenAI, Google AI Studio, Google Vertex AI, Anthropic, Watsonx.ai, or Hugging Face.\nFor built-in models and models uploaded through Eland, the inference APIs offer an alternative way to use and manage trained models. However, if you do not plan to use the inference APIs to use these models or if you want to use non-NLP models, use the machine learning trained model APIs.", + "description": "Create an inference endpoint.\nWhen you create an inference endpoint, the associated machine learning model is automatically deployed if it is not already running.\nAfter creating the endpoint, wait for the model deployment to complete before using it.\nTo verify the deployment status, use the get trained model statistics API.\nLook for `\"state\": \"fully_allocated\"` in the response and ensure that the `\"allocation_count\"` matches the `\"target_allocation_count\"`.\nAvoid creating multiple endpoints for the same model unless required, as each endpoint consumes significant resources.\n\nIMPORTANT: The inference APIs enable you to use certain services, such as built-in machine learning models (ELSER, E5), models uploaded through Eland, Cohere, OpenAI, Mistral, Azure OpenAI, Google AI Studio, Google Vertex AI, Anthropic, Watsonx.ai, or Hugging Face.\nFor built-in models and models uploaded through Eland, the inference APIs offer an alternative way to use and manage trained models. However, if you do not plan to use the inference APIs to use these models or if you want to use non-NLP models, use the machine learning trained model APIs.", "inherits": { "type": { "name": "RequestBase", diff --git a/specification/inference/put/PutRequest.ts b/specification/inference/put/PutRequest.ts index a14c366522..1fa26d2fc0 100644 --- a/specification/inference/put/PutRequest.ts +++ b/specification/inference/put/PutRequest.ts @@ -26,7 +26,7 @@ import { Id } from '@_types/common' * Create an inference endpoint. * When you create an inference endpoint, the associated machine learning model is automatically deployed if it is not already running. * After creating the endpoint, wait for the model deployment to complete before using it. - * To verify the deployment status, use the get trained model statistics API. + * To verify the deployment status, use the get trained model statistics API. * Look for `"state": "fully_allocated"` in the response and ensure that the `"allocation_count"` matches the `"target_allocation_count"`. * Avoid creating multiple endpoints for the same model unless required, as each endpoint consumes significant resources. * From 66b3d5773197be91adc299ba212fc187ba4d0397 Mon Sep 17 00:00:00 2001 From: lcawl Date: Mon, 16 Dec 2024 18:40:13 -0800 Subject: [PATCH 4/5] Edit description --- specification/inference/put/PutRequest.ts | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/specification/inference/put/PutRequest.ts b/specification/inference/put/PutRequest.ts index 1fa26d2fc0..3ddb505e53 100644 --- a/specification/inference/put/PutRequest.ts +++ b/specification/inference/put/PutRequest.ts @@ -31,7 +31,8 @@ import { Id } from '@_types/common' * Avoid creating multiple endpoints for the same model unless required, as each endpoint consumes significant resources. * * IMPORTANT: The inference APIs enable you to use certain services, such as built-in machine learning models (ELSER, E5), models uploaded through Eland, Cohere, OpenAI, Mistral, Azure OpenAI, Google AI Studio, Google Vertex AI, Anthropic, Watsonx.ai, or Hugging Face. - * For built-in models and models uploaded through Eland, the inference APIs offer an alternative way to use and manage trained models. However, if you do not plan to use the inference APIs to use these models or if you want to use non-NLP models, use the machine learning trained model APIs. + * For built-in models and models uploaded through Eland, the inference APIs offer an alternative way to use and manage trained models. + * However, if you do not plan to use the inference APIs to use these models or if you want to use non-NLP models, use the machine learning trained model APIs. * @rest_spec_name inference.put * @availability stack since=8.11.0 stability=stable visibility=public * @availability serverless stability=stable visibility=public From 8ee6452f07d9b78c35d817393a803d53dddd163d Mon Sep 17 00:00:00 2001 From: lcawl Date: Mon, 16 Dec 2024 18:41:20 -0800 Subject: [PATCH 5/5] Generate output --- output/openapi/elasticsearch-openapi.json | 4 ++-- output/openapi/elasticsearch-serverless-openapi.json | 4 ++-- output/schema/schema.json | 6 +++--- 3 files changed, 7 insertions(+), 7 deletions(-) diff --git a/output/openapi/elasticsearch-openapi.json b/output/openapi/elasticsearch-openapi.json index 1590a173bb..ff3f3af4b0 100644 --- a/output/openapi/elasticsearch-openapi.json +++ b/output/openapi/elasticsearch-openapi.json @@ -14997,7 +14997,7 @@ "inference" ], "summary": "Create an inference endpoint", - "description": "When you create an inference endpoint, the associated machine learning model is automatically deployed if it is not already running.\nAfter creating the endpoint, wait for the model deployment to complete before using it.\nTo verify the deployment status, use the get trained model statistics API.\nLook for `\"state\": \"fully_allocated\"` in the response and ensure that the `\"allocation_count\"` matches the `\"target_allocation_count\"`.\nAvoid creating multiple endpoints for the same model unless required, as each endpoint consumes significant resources.\n\nIMPORTANT: The inference APIs enable you to use certain services, such as built-in machine learning models (ELSER, E5), models uploaded through Eland, Cohere, OpenAI, Mistral, Azure OpenAI, Google AI Studio, Google Vertex AI, Anthropic, Watsonx.ai, or Hugging Face.\nFor built-in models and models uploaded through Eland, the inference APIs offer an alternative way to use and manage trained models. However, if you do not plan to use the inference APIs to use these models or if you want to use non-NLP models, use the machine learning trained model APIs.", + "description": "When you create an inference endpoint, the associated machine learning model is automatically deployed if it is not already running.\nAfter creating the endpoint, wait for the model deployment to complete before using it.\nTo verify the deployment status, use the get trained model statistics API.\nLook for `\"state\": \"fully_allocated\"` in the response and ensure that the `\"allocation_count\"` matches the `\"target_allocation_count\"`.\nAvoid creating multiple endpoints for the same model unless required, as each endpoint consumes significant resources.\n\nIMPORTANT: The inference APIs enable you to use certain services, such as built-in machine learning models (ELSER, E5), models uploaded through Eland, Cohere, OpenAI, Mistral, Azure OpenAI, Google AI Studio, Google Vertex AI, Anthropic, Watsonx.ai, or Hugging Face.\nFor built-in models and models uploaded through Eland, the inference APIs offer an alternative way to use and manage trained models.\nHowever, if you do not plan to use the inference APIs to use these models or if you want to use non-NLP models, use the machine learning trained model APIs.", "operationId": "inference-put", "parameters": [ { @@ -15090,7 +15090,7 @@ "inference" ], "summary": "Create an inference endpoint", - "description": "When you create an inference endpoint, the associated machine learning model is automatically deployed if it is not already running.\nAfter creating the endpoint, wait for the model deployment to complete before using it.\nTo verify the deployment status, use the get trained model statistics API.\nLook for `\"state\": \"fully_allocated\"` in the response and ensure that the `\"allocation_count\"` matches the `\"target_allocation_count\"`.\nAvoid creating multiple endpoints for the same model unless required, as each endpoint consumes significant resources.\n\nIMPORTANT: The inference APIs enable you to use certain services, such as built-in machine learning models (ELSER, E5), models uploaded through Eland, Cohere, OpenAI, Mistral, Azure OpenAI, Google AI Studio, Google Vertex AI, Anthropic, Watsonx.ai, or Hugging Face.\nFor built-in models and models uploaded through Eland, the inference APIs offer an alternative way to use and manage trained models. However, if you do not plan to use the inference APIs to use these models or if you want to use non-NLP models, use the machine learning trained model APIs.", + "description": "When you create an inference endpoint, the associated machine learning model is automatically deployed if it is not already running.\nAfter creating the endpoint, wait for the model deployment to complete before using it.\nTo verify the deployment status, use the get trained model statistics API.\nLook for `\"state\": \"fully_allocated\"` in the response and ensure that the `\"allocation_count\"` matches the `\"target_allocation_count\"`.\nAvoid creating multiple endpoints for the same model unless required, as each endpoint consumes significant resources.\n\nIMPORTANT: The inference APIs enable you to use certain services, such as built-in machine learning models (ELSER, E5), models uploaded through Eland, Cohere, OpenAI, Mistral, Azure OpenAI, Google AI Studio, Google Vertex AI, Anthropic, Watsonx.ai, or Hugging Face.\nFor built-in models and models uploaded through Eland, the inference APIs offer an alternative way to use and manage trained models.\nHowever, if you do not plan to use the inference APIs to use these models or if you want to use non-NLP models, use the machine learning trained model APIs.", "operationId": "inference-put-1", "parameters": [ { diff --git a/output/openapi/elasticsearch-serverless-openapi.json b/output/openapi/elasticsearch-serverless-openapi.json index 46c233ed84..3bd73e4a41 100644 --- a/output/openapi/elasticsearch-serverless-openapi.json +++ b/output/openapi/elasticsearch-serverless-openapi.json @@ -8579,7 +8579,7 @@ "inference" ], "summary": "Create an inference endpoint", - "description": "When you create an inference endpoint, the associated machine learning model is automatically deployed if it is not already running.\nAfter creating the endpoint, wait for the model deployment to complete before using it.\nTo verify the deployment status, use the get trained model statistics API.\nLook for `\"state\": \"fully_allocated\"` in the response and ensure that the `\"allocation_count\"` matches the `\"target_allocation_count\"`.\nAvoid creating multiple endpoints for the same model unless required, as each endpoint consumes significant resources.\n\nIMPORTANT: The inference APIs enable you to use certain services, such as built-in machine learning models (ELSER, E5), models uploaded through Eland, Cohere, OpenAI, Mistral, Azure OpenAI, Google AI Studio, Google Vertex AI, Anthropic, Watsonx.ai, or Hugging Face.\nFor built-in models and models uploaded through Eland, the inference APIs offer an alternative way to use and manage trained models. However, if you do not plan to use the inference APIs to use these models or if you want to use non-NLP models, use the machine learning trained model APIs.", + "description": "When you create an inference endpoint, the associated machine learning model is automatically deployed if it is not already running.\nAfter creating the endpoint, wait for the model deployment to complete before using it.\nTo verify the deployment status, use the get trained model statistics API.\nLook for `\"state\": \"fully_allocated\"` in the response and ensure that the `\"allocation_count\"` matches the `\"target_allocation_count\"`.\nAvoid creating multiple endpoints for the same model unless required, as each endpoint consumes significant resources.\n\nIMPORTANT: The inference APIs enable you to use certain services, such as built-in machine learning models (ELSER, E5), models uploaded through Eland, Cohere, OpenAI, Mistral, Azure OpenAI, Google AI Studio, Google Vertex AI, Anthropic, Watsonx.ai, or Hugging Face.\nFor built-in models and models uploaded through Eland, the inference APIs offer an alternative way to use and manage trained models.\nHowever, if you do not plan to use the inference APIs to use these models or if you want to use non-NLP models, use the machine learning trained model APIs.", "operationId": "inference-put", "parameters": [ { @@ -8672,7 +8672,7 @@ "inference" ], "summary": "Create an inference endpoint", - "description": "When you create an inference endpoint, the associated machine learning model is automatically deployed if it is not already running.\nAfter creating the endpoint, wait for the model deployment to complete before using it.\nTo verify the deployment status, use the get trained model statistics API.\nLook for `\"state\": \"fully_allocated\"` in the response and ensure that the `\"allocation_count\"` matches the `\"target_allocation_count\"`.\nAvoid creating multiple endpoints for the same model unless required, as each endpoint consumes significant resources.\n\nIMPORTANT: The inference APIs enable you to use certain services, such as built-in machine learning models (ELSER, E5), models uploaded through Eland, Cohere, OpenAI, Mistral, Azure OpenAI, Google AI Studio, Google Vertex AI, Anthropic, Watsonx.ai, or Hugging Face.\nFor built-in models and models uploaded through Eland, the inference APIs offer an alternative way to use and manage trained models. However, if you do not plan to use the inference APIs to use these models or if you want to use non-NLP models, use the machine learning trained model APIs.", + "description": "When you create an inference endpoint, the associated machine learning model is automatically deployed if it is not already running.\nAfter creating the endpoint, wait for the model deployment to complete before using it.\nTo verify the deployment status, use the get trained model statistics API.\nLook for `\"state\": \"fully_allocated\"` in the response and ensure that the `\"allocation_count\"` matches the `\"target_allocation_count\"`.\nAvoid creating multiple endpoints for the same model unless required, as each endpoint consumes significant resources.\n\nIMPORTANT: The inference APIs enable you to use certain services, such as built-in machine learning models (ELSER, E5), models uploaded through Eland, Cohere, OpenAI, Mistral, Azure OpenAI, Google AI Studio, Google Vertex AI, Anthropic, Watsonx.ai, or Hugging Face.\nFor built-in models and models uploaded through Eland, the inference APIs offer an alternative way to use and manage trained models.\nHowever, if you do not plan to use the inference APIs to use these models or if you want to use non-NLP models, use the machine learning trained model APIs.", "operationId": "inference-put-1", "parameters": [ { diff --git a/output/schema/schema.json b/output/schema/schema.json index dd94ad6433..58e5eb1d7d 100644 --- a/output/schema/schema.json +++ b/output/schema/schema.json @@ -8525,7 +8525,7 @@ "visibility": "public" } }, - "description": "Create an inference endpoint.\nWhen you create an inference endpoint, the associated machine learning model is automatically deployed if it is not already running.\nAfter creating the endpoint, wait for the model deployment to complete before using it.\nTo verify the deployment status, use the get trained model statistics API.\nLook for `\"state\": \"fully_allocated\"` in the response and ensure that the `\"allocation_count\"` matches the `\"target_allocation_count\"`.\nAvoid creating multiple endpoints for the same model unless required, as each endpoint consumes significant resources.\n\nIMPORTANT: The inference APIs enable you to use certain services, such as built-in machine learning models (ELSER, E5), models uploaded through Eland, Cohere, OpenAI, Mistral, Azure OpenAI, Google AI Studio, Google Vertex AI, Anthropic, Watsonx.ai, or Hugging Face.\nFor built-in models and models uploaded through Eland, the inference APIs offer an alternative way to use and manage trained models. However, if you do not plan to use the inference APIs to use these models or if you want to use non-NLP models, use the machine learning trained model APIs.", + "description": "Create an inference endpoint.\nWhen you create an inference endpoint, the associated machine learning model is automatically deployed if it is not already running.\nAfter creating the endpoint, wait for the model deployment to complete before using it.\nTo verify the deployment status, use the get trained model statistics API.\nLook for `\"state\": \"fully_allocated\"` in the response and ensure that the `\"allocation_count\"` matches the `\"target_allocation_count\"`.\nAvoid creating multiple endpoints for the same model unless required, as each endpoint consumes significant resources.\n\nIMPORTANT: The inference APIs enable you to use certain services, such as built-in machine learning models (ELSER, E5), models uploaded through Eland, Cohere, OpenAI, Mistral, Azure OpenAI, Google AI Studio, Google Vertex AI, Anthropic, Watsonx.ai, or Hugging Face.\nFor built-in models and models uploaded through Eland, the inference APIs offer an alternative way to use and manage trained models.\nHowever, if you do not plan to use the inference APIs to use these models or if you want to use non-NLP models, use the machine learning trained model APIs.", "docUrl": "https://www.elastic.co/guide/en/elasticsearch/reference/master/put-inference-api.html", "name": "inference.put", "privileges": { @@ -141965,7 +141965,7 @@ } } }, - "description": "Create an inference endpoint.\nWhen you create an inference endpoint, the associated machine learning model is automatically deployed if it is not already running.\nAfter creating the endpoint, wait for the model deployment to complete before using it.\nTo verify the deployment status, use the get trained model statistics API.\nLook for `\"state\": \"fully_allocated\"` in the response and ensure that the `\"allocation_count\"` matches the `\"target_allocation_count\"`.\nAvoid creating multiple endpoints for the same model unless required, as each endpoint consumes significant resources.\n\nIMPORTANT: The inference APIs enable you to use certain services, such as built-in machine learning models (ELSER, E5), models uploaded through Eland, Cohere, OpenAI, Mistral, Azure OpenAI, Google AI Studio, Google Vertex AI, Anthropic, Watsonx.ai, or Hugging Face.\nFor built-in models and models uploaded through Eland, the inference APIs offer an alternative way to use and manage trained models. However, if you do not plan to use the inference APIs to use these models or if you want to use non-NLP models, use the machine learning trained model APIs.", + "description": "Create an inference endpoint.\nWhen you create an inference endpoint, the associated machine learning model is automatically deployed if it is not already running.\nAfter creating the endpoint, wait for the model deployment to complete before using it.\nTo verify the deployment status, use the get trained model statistics API.\nLook for `\"state\": \"fully_allocated\"` in the response and ensure that the `\"allocation_count\"` matches the `\"target_allocation_count\"`.\nAvoid creating multiple endpoints for the same model unless required, as each endpoint consumes significant resources.\n\nIMPORTANT: The inference APIs enable you to use certain services, such as built-in machine learning models (ELSER, E5), models uploaded through Eland, Cohere, OpenAI, Mistral, Azure OpenAI, Google AI Studio, Google Vertex AI, Anthropic, Watsonx.ai, or Hugging Face.\nFor built-in models and models uploaded through Eland, the inference APIs offer an alternative way to use and manage trained models.\nHowever, if you do not plan to use the inference APIs to use these models or if you want to use non-NLP models, use the machine learning trained model APIs.", "inherits": { "type": { "name": "RequestBase", @@ -142003,7 +142003,7 @@ } ], "query": [], - "specLocation": "inference/put/PutRequest.ts#L25-L53" + "specLocation": "inference/put/PutRequest.ts#L25-L54" }, { "kind": "response",