Merge pull request #717 from MicrosoftDocs/main

10/9 11:00 AM IST Publish

Author: PhilKang0704

.github/policies/disallow-edits.yml

@@ -0,0 +1,21 @@
+name: GitOps.PullRequestIssueManagement
+description: GitOps.PullRequestIssueManagement primitive
+resource: repository
+
+configuration:
+  resourceManagementConfiguration:
+    eventResponderTasks:
+      - description: Close PRs that introduce an index file into the repository.
+        if:
+          - payloadType: Pull_Request
+          - isAction:
+              action: Opened
+          - or:              
+              - filesMatchPattern:
+                  pattern: articles/index.yml
+        then:
+          - addReply:
+              reply: >-
+                @${issueAuthor} - You tried to add an index file to this repository; this is not permitted so your pull request will be closed automatically. 
+          - closePullRequest
+          

articles/ai-services/openai/includes/ai-search-ingestion.md

@@ -3,17 +3,16 @@ manager: nitinme
 ms.service: azure-ai-studio
 ms.custom:
 ms.topic: include
-ms.date: 03/25/2024
+ms.date: 10/08/2024
 ms.author: aahi
 author: aahill
 ---
 
 ### How data is ingested into Azure AI search
 
-Data is ingested into Azure AI search using the following process:
+As of September 2024, the ingestion APIs switched to [integrated vectorization](/azure/search/vector-search-integrated-vectorization). This update does **not** alter the existing API contracts. Integrated vectorization, a new offering of Azure AI Search, utilizes prebuilt skills for chunking and embedding the input data. The Azure OpenAI On Your Data ingestion service no longer employs custom skills. Following the migration to integrated vectorization, the ingestion process has undergone some modifications and as a result only the following assets are created:
+   * `{job-id}-index`
+   * `{job-id}-indexer`, if an hourly or daily schedule is specified, otherwise, the indexer is cleaned-up at the end of the ingestion process.
+   * `{job-id}-datasource`
 
-1. Ingestion assets are created in Azure AI Search resource and Azure storage account. Currently these assets are: indexers, indexes, data sources, a [custom skill](/azure/search/cognitive-search-custom-skill-interface) in the search resource, and a container (later called the chunks container) in the Azure storage account. You can specify the input Azure storage container using the [Azure OpenAI studio](https://oai.azure.com/), or the [ingestion API (preview)](/rest/api/azureopenai/ingestion-jobs). By default, text is assumed to use the UTF-8 encoding. To specify a different encoding, use the encoding configuration property. See the [.NET documentation](/dotnet/fundamentals/runtime-libraries/system-text-encoding#list-of-encodings) for a list of supported encodings.
-
-2. Data is read from the input container, contents are opened and chunked into small chunks with a maximum of 1,024 tokens each. If vector search is enabled, the service calculates the vector representing the embeddings on each chunk. The output of this step (called the "preprocessed" or "chunked" data) is stored in the chunks container created in the previous step. 
-
-3. The preprocessed data is loaded from the chunks container, and indexed in the Azure AI Search index.
+The chunks container is no longer available, as this functionality is now inherently managed by Azure AI Search.

articles/ai-studio/how-to/develop/llama-index.md

@@ -67,7 +67,7 @@ export AZURE_INFERENCE_ENDPOINT="<your-model-endpoint-goes-here>"
 export AZURE_INFERENCE_CREDENTIAL="<your-key-goes-here>"
 ```
 
-Once configured, create a client to connect to the endpoint:
+Once configured, create a client to connect to the endpoint. The parameter `model_name` in the constructor is not required for endpoints serving a single model, like serverless endpoints.
 
 ```python
 import os
@@ -80,7 +80,7 @@ llm = AzureAICompletionsModel(
 ```
 
 > [!TIP]
-> The parameter `model_name` in the constructor is not required for endpoints serving a single model, like serverless endpoints).
+> If your model is an OpenAI model deployed to Azure OpenAI service or AI services resource, configure the client as indicated at [Azure OpenAI models](#azure-openai-models).
 
 Alternatively, if your endpoint support Microsoft Entra ID, you can use the following code to create the client:
 
@@ -112,6 +112,23 @@ llm = AzureAICompletionsModel(
 )
 ```
 
+### Azure OpenAI models
+
+If you are using Azure OpenAI models with key-based authentication, you need to pass the authentication key in the header `api-key`, which is the one expected in the Azure OpenAI service and in Azure AI Services. This configuration is not required if you are using Microsoft Entra ID (formerly known as Azure AD). The following example shows how to configure the client:
+
+```python
+import os
+from llama_index.llms.azure_inference import AzureAICompletionsModel
+
+llm = AzureAICompletionsModel(
+    endpoint=os.environ["AZURE_INFERENCE_ENDPOINT"],
+    credential="",
+    client_kwargs={"headers" : { "api-key": os.environ["AZURE_INFERENCE_CREDENTIAL"] } }
+)
+```
+
+Notice that `credentials` is still being passed with an empty value since it's a required parameter.
+
 ### Inference parameters
 
 You can configure how inference in performed for all the operations that are using this client by setting extra parameters. This helps avoid indicating them on each call you make to the model.