Merge pull request #276059 from ssalgadodev/patch-107

prmerger-automator[bot] · web-flow · commit fe05c055b26f · 2024-05-22T19:33:22.000Z
Update concept-model-catalog.md
diff --git a/articles/machine-learning/concept-data-privacy.md b/articles/machine-learning/concept-data-privacy.md
@@ -7,7 +7,7 @@ ms.service: azure-ai-studio
 ms.custom:
   - ignite-2023
 ms.topic: how-to
-ms.date: 5/02/2024
+ms.date: 5/22/2024
 ms.reviewer: jcioffi
 ms.author: ssalgado
 author: ssalgadodev
@@ -24,9 +24,9 @@ When you deploy models in Azure Machine Learning, the following types of data ar
 
 * **Uploaded data**. For models that support fine-tuning, customers can upload their data to the [Azure Machine Learning Datastore](./concept-data.md) for use for fine-tuning.
 
-## Generate inferencing outputs with real-time endpoints
+## Generate inferencing outputs with managed compute
 
-Deploying models to managed compute deploys model weights to dedicated Virtual Machines and exposes a REST API for real-time inference. Learn more about deploying models from the [Model Catalog to real-time endpoints](concept-model-catalog.md). You manage the infrastructure for these real-time endpoints, and Azure's data, privacy, and security commitments apply. Learn more about [Azure compliance offerings](https://servicetrust.microsoft.com/DocumentPage/7adf2d9e-d7b5-4e71-bad8-713e6a183cf3) applicable to Azure Machine Learning.
+Deploying models to managed compute deploys model weights to dedicated Virtual Machines and exposes a REST API for real-time inference. Learn more about deploying models from the [Model Catalog to managed compute](concept-model-catalog.md). You manage the infrastructure for these managed computes, and Azure's data, privacy, and security commitments apply. Learn more about [Azure compliance offerings](https://servicetrust.microsoft.com/DocumentPage/7adf2d9e-d7b5-4e71-bad8-713e6a183cf3) applicable to Azure Machine Learning.
 
 Although containers for models "Curated by Azure AI" are scanned for vulnerabilities that could exfiltrate data, not all models available through the model catalog have been scanned. To reduce the risk of data exfiltration, you can protect your deployment using virtual networks. Follow this link to [learn more](./how-to-network-isolation-model-catalog.md). You can also use [Azure Policy](./how-to-regulate-registry-deployments.md) to regulate the models that can be deployed by your users.
 
@@ -36,7 +36,9 @@ Although containers for models "Curated by Azure AI" are scanned for vulnerabili
 
 When you deploy a model from the model catalog (base or fine-tuned) as a serverless API for inferencing, an API is provisioned giving you access to the model hosted and managed by the Azure Machine Learning Service. Learn more about [Models-as-a-Service](concept-model-catalog.md). The model processes your input prompts and generates outputs based on the functionality of the model, as described in the model details provided for the model. While the model is provided by the model provider, and your use of the model (and the model provider's accountability for the model and its outputs) is subject to the license terms provided with the model, Microsoft provides and manages the hosting infrastructure and API endpoint. The models hosted in Models-as-a-Service are subject to Azure's data, privacy, and security commitments. Learn more about Azure compliance offerings applicable to Azure Machine Learning [here](https://servicetrust.microsoft.com/DocumentPage/7adf2d9e-d7b5-4e71-bad8-713e6a183cf3). 
 
-Microsoft acts as the data processor for prompts and outputs sent to and generated by a model deployed for pay-as-you-go inferencing (MaaS). Microsoft doesn't share these prompts and outputs with the model provider, and Microsoft doesn't use these prompts and outputs to train or improve Microsoft's, the model provider's, or any third party's models. Models are stateless and no prompts or outputs are stored in the model. If content filtering is enabled, prompts and outputs are screened for certain categories of harmful content by the Azure AI Content Safety service in real time; learn more about how Azure AI Content Safety processes data [here](/legal/cognitive-services/content-safety/data-privacy). Prompts and outputs are processed within the geography specified during deployment but may be processed between regions within the geography for operational purposes (including performance and capacity management).
+[!INCLUDE [machine-learning-preview-generic-disclaimer](includes/machine-learning-preview-generic-disclaimer.md)]
+
+Microsoft acts as the data processor for prompts and outputs sent to and generated by a model deployed for pay-as-you-go inferencing (MaaS). Microsoft doesn't share these prompts and outputs with the model provider, and Microsoft doesn't use these prompts and outputs to train or improve Microsoft's, the model provider's, or any third party's models. Models are stateless and no prompts or outputs are stored in the model. If content filtering (preview) is enabled, prompts and outputs are screened for certain categories of harmful content by the Azure AI Content Safety service in real time; learn more about how Azure AI Content Safety processes data [here](/legal/cognitive-services/content-safety/data-privacy). Prompts and outputs are processed within the geography specified during deployment but may be processed between regions within the geography for operational purposes (including performance and capacity management).
 
 :::image type="content" source="media/concept-data-privacy/model-publisher-cycle.png" alt-text="A diagram showing model publisher service cycle." lightbox="media/concept-data-privacy/model-publisher-cycle.png":::
 
diff --git a/articles/machine-learning/concept-model-catalog.md b/articles/machine-learning/concept-model-catalog.md
@@ -58,6 +58,9 @@ Model | Managed compute | Serverless API (pay-as-you-go)
 Llama family models  | Llama-2-7b <br> Llama-2-7b-chat <br> Llama-2-13b <br> Llama-2-13b-chat <br> Llama-2-70b <br> Llama-2-70b-chat <br> Llama-3-8B-Instruct <br> Llama-3-70B-Instruct <br> Llama-3-8B <br> Llama-3-70B | Llama-3-70B-Instruct <br> Llama-3-8B-Instruct <br> Llama-2-7b <br> Llama-2-7b-chat <br> Llama-2-13b <br> Llama-2-13b-chat <br> Llama-2-70b <br> Llama-2-70b-chat 
 Mistral family models | mistralai-Mixtral-8x22B-v0-1 <br> mistralai-Mixtral-8x22B-Instruct-v0-1 <br> mistral-community-Mixtral-8x22B-v0-1 <br> mistralai-Mixtral-8x7B-v01 <br> mistralai-Mistral-7B-Instruct-v0-2 <br> mistralai-Mistral-7B-v01 <br> mistralai-Mixtral-8x7B-Instruct-v01 <br> mistralai-Mistral-7B-Instruct-v01 | Mistral-large <br> Mistral-small
 Cohere family models | Not available | Cohere-command-r-plus <br> Cohere-command-r <br> Cohere-embed-v3-english <br> Cohere-embed-v3-multilingual
+JAIS | Not available | jais-30b-chat
+Phi3 family models | Phi-3-small-128k-Instruct <br> Phi-3-small-8k-Instruct <br> Phi-3-mini-4k-Instruct <br> Phi-3-mini-128k-Instruct <br> Phi3-medium-128k-instruct <br> Phi3-medium-4k-instruct | Phi-3-mini-4k-Instruct <br> Phi-3-mini-128k-Instruct <br> Phi3-medium-128k-instruct <br> Phi3-medium-4k-instruct <br> Phi-3-vision-128k-instruct 
+Nixtla | Not available | TimeGEN-1
 Other models | Available | Not available
 
 :::image type="content" source="./media/concept-model-catalog/platform-service-cycle.png" alt-text="A diagram showing models as a service and Real time end points service cycle." lightbox="media/concept-model-catalog/platform-service-cycle.png":::
@@ -90,7 +93,7 @@ Models available for deployment with managed compute can be deployed to Azure Ma
 
 ### Build Generative AI Apps with managed compute
 
-Prompt flow offers capabilities for prototyping, experimenting, iterating, and deploying your AI applications. You can use models deployed with managed compute in Prompt Flow with the [Open Model LLM tool](./prompt-flow/tools-reference/open-model-llm-tool.md).  You can also use the REST API exposed by the Real-time endpoints in popular LLM tools like LangChain with the [Azure Machine Learning extension](https://python.langchain.com/docs/integrations/chat/azureml_chat_endpoint/).  
+Prompt flow offers capabilities for prototyping, experimenting, iterating, and deploying your AI applications. You can use models deployed with managed compute in Prompt Flow with the [Open Model LLM tool](./prompt-flow/tools-reference/open-model-llm-tool.md).  You can also use the REST API exposed by the managed computes in popular LLM tools like LangChain with the [Azure Machine Learning extension](https://python.langchain.com/docs/integrations/chat/azureml_chat_endpoint/).  
 
 
 ### Content safety for models deployed with managed compute 
@@ -142,7 +145,9 @@ Cohere-command-r-plus <br> Cohere-command-r <br> Cohere-embed-v3-english <br> Co
 
 ### Content safety for models deployed via MaaS  
 
-Azure Machine Learning implements a default configuration of [Azure AI Content Safety](../ai-services/content-safety/overview.md) text moderation filters for harmful content (hate, self-harm, sexual, and violence) for language models deployed with MaaS. To learn more about content filtering, see [harm categories in Azure AI Content Safety](../ai-services/content-safety/concepts/harm-categories.md). Content filtering occurs synchronously as the service processes prompts to generate content, and you may be billed separately as per [AACS pricing](https://azure.microsoft.com/pricing/details/cognitive-services/content-safety/) for such use. You can disable content filtering for individual serverless endpoints when you first deploy a language model or in the deployment details page by clicking the content filtering toggle. You may be at higher risk of exposing users to harmful content if you turn off content filters. 
+[!INCLUDE [machine-learning-preview-generic-disclaimer](includes/machine-learning-preview-generic-disclaimer.md)]
+
+Azure Machine Learning implements a default configuration of [Azure AI Content Safety](../ai-services/content-safety/overview.md) text moderation filters for harmful content (hate, self-harm, sexual, and violence) for language models deployed with MaaS. To learn more about content filtering (preview), see [harm categories in Azure AI Content Safety](../ai-services/content-safety/concepts/harm-categories.md). Content filtering (preview) occurs synchronously as the service processes prompts to generate content, and you may be billed separately as per [AACS pricing](https://azure.microsoft.com/pricing/details/cognitive-services/content-safety/) for such use. You can disable content filtering (preview) for individual serverless endpoints when you first deploy a language model or in the deployment details page by selecting the content filtering toggle. You may be at higher risk of exposing users to harmful content if you turn off content filters. 
 
 ## Learn more
 
