Merge pull request #252130 from dem108/patch-23

update on shared quota

Author: v-shils

articles/machine-learning/how-to-deploy-online-endpoints.md

@@ -8,7 +8,7 @@ ms.subservice: inferencing
 author: dem108
 ms.author: sehan
 ms.reviewer: mopeakande
-ms.date: 07/17/2023
+ms.date: 09/18/2023
 reviewer: msakande
 ms.topic: how-to
 ms.custom: how-to, devplatv2, ignite-fall-2021, cliv2, event-tier1-build-2022, sdkv2
@@ -71,10 +71,12 @@ Before following the steps in this article, make sure you have the following pre
 
 ### Virtual machine quota allocation for deployment
 
-For managed online endpoints, Azure Machine Learning reserves 20% of your compute resources for performing upgrades. Therefore, if you request a given number of instances in a deployment, you must have a quota for `ceil(1.2 * number of instances requested for deployment) * number of cores for the VM SKU` available to avoid getting an error. For example, if you request 10 instances of a [Standard_DS3_v2](/azure/virtual-machines/dv2-dsv2-series) VM (that comes with 4 cores) in a deployment, you should have a quota for 48 cores (`12 instances * 4 cores`) available. To view your usage and request quota increases, see [View your usage and quotas in the Azure portal](how-to-manage-quotas.md#view-your-usage-and-quotas-in-the-azure-portal).
+For managed online endpoints, Azure Machine Learning reserves 20% of your compute resources for performing upgrades on some VM SKUs. If you request a given number of instances in a deployment, you must have a quota for `ceil(1.2 * number of instances requested for deployment) * number of cores for the VM SKU` available to avoid getting an error. For example, if you request 10 instances of a [Standard_DS3_v2](/azure/virtual-machines/dv2-dsv2-series) VM (that comes with 4 cores) in a deployment, you should have a quota for 48 cores (`12 instances * 4 cores`) available. To view your usage and request quota increases, see [View your usage and quotas in the Azure portal](how-to-manage-quotas.md#view-your-usage-and-quotas-in-the-azure-portal).
 
-<!-- In this tutorial, you'll request one instance of a Standard_DS2_v2 VM SKU (that comes with 2 cores) in your deployment; therefore, you should have a minimum quota for 4 cores (`2 instances*2 cores`) available.  -->
----
+
+Azure Machine Learning provides a [shared quota](how-to-manage-quotas.md#azure-machine-learning-shared-quota) pool from which all users can access quota to perform testing for a limited time. When you use the studio to deploy Llama models (from the model catalog) to a managed online endpoint, Azure Machine Learning allows you to access this shared quota for a short time. 
+
+To deploy a _Llama-2-70b_ or _Llama-2-70b-chat_ model, however, you must have an [Enterprise Agreement subscription](/azure/cost-management-billing/manage/create-enterprise-subscription) before you can deploy using the shared quota. For more information on how to use the shared quota for online endpoint deployment, see [How to deploy foundation models using the studio](how-to-use-foundation-models.md#deploying-using-the-studio).
 
 ## Prepare your system
 

articles/machine-learning/how-to-use-foundation-models.md

@@ -138,6 +138,10 @@ Since the scoring script and environment are automatically included with the fou
 
 :::image type="content" source="./media/how-to-use-foundation-models/deploy-options.png" alt-text="Screenshot showing the deploy options on the foundation model card after user selects the deploy button.":::
 
+If you're deploying a Llama model from the model catalog but don't have enough quota available for the deployment, Azure Machine Learning allows you to use quota from a shared quota pool for a limited time. For _Llama-2-70b_ and _Llama-2-70b-chat_ model deployment, access to the shared quota is available only to customers with [Enterprise Agreement subscriptions](/azure/cost-management-billing/manage/create-enterprise-subscription). For more information on shared quota, see [Azure Machine Learning shared quota](how-to-manage-quotas.md#azure-machine-learning-shared-quota).
+
+:::image type="content" source="media/how-to-use-foundation-models/deploy-llama-model-with-shared-quota.png" alt-text="Screenshot showing the option to deploy a Llama model temporarily, using shared quota." lightbox="media/how-to-use-foundation-models/deploy-llama-model-with-shared-quota.png":::
+
 ### Deploying using code based samples
 
 To enable users to quickly get started with deployment and inferencing, we have published samples in the [Inference samples in the azureml-examples git repo](https://github.com/Azure/azureml-examples/tree/main/sdk/python/foundation-models/system/inference). The published samples include Python notebooks and CLI examples. Each model card also links to Inference samples for Real time and Batch inferencing.