From f0095aa997e1fb7ca1603be442532c4e918cf440 Mon Sep 17 00:00:00 2001 From: Yi Tang Date: Wed, 15 Jan 2025 13:49:02 -0800 Subject: [PATCH 1/2] add limitations on kubernetes compute target --- articles/machine-learning/how-to-attach-kubernetes-anywhere.md | 2 ++ 1 file changed, 2 insertions(+) diff --git a/articles/machine-learning/how-to-attach-kubernetes-anywhere.md b/articles/machine-learning/how-to-attach-kubernetes-anywhere.md index 939ab7d5437..dfcfb4b447e 100644 --- a/articles/machine-learning/how-to-attach-kubernetes-anywhere.md +++ b/articles/machine-learning/how-to-attach-kubernetes-anywhere.md @@ -74,6 +74,8 @@ A `KubernetesCompute` target in Azure Machine Learning workloads (training and m - The availability of **Preview features** in Azure Machine Learning isn't guaranteed. - Models (including the foundational model) from the **Model Catalog** and **Registry** aren't supported on Kubernetes online endpoints. - The process of creating a model inference deployment inside the cluster has a timeout limit of **20 minutes**. This includes downloading the image, downloading the model, and initializing the user scripts. +- Azure Machine Learning extension support Kubernetes Baseline Pod Security Standard. +- Training on Kubernetes compute doesn't support auto scale nodes. ## Recommended best practices From 9ed3010580ce795b60753e1e01c605ece594ae18 Mon Sep 17 00:00:00 2001 From: Yi Tang Date: Thu, 16 Jan 2025 13:32:45 -0800 Subject: [PATCH 2/2] fix typo --- articles/machine-learning/how-to-attach-kubernetes-anywhere.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/articles/machine-learning/how-to-attach-kubernetes-anywhere.md b/articles/machine-learning/how-to-attach-kubernetes-anywhere.md index dfcfb4b447e..2e4bd7b348d 100644 --- a/articles/machine-learning/how-to-attach-kubernetes-anywhere.md +++ b/articles/machine-learning/how-to-attach-kubernetes-anywhere.md @@ -74,7 +74,7 @@ A `KubernetesCompute` target in Azure Machine Learning workloads (training and m - The availability of **Preview features** in Azure Machine Learning isn't guaranteed. - Models (including the foundational model) from the **Model Catalog** and **Registry** aren't supported on Kubernetes online endpoints. - The process of creating a model inference deployment inside the cluster has a timeout limit of **20 minutes**. This includes downloading the image, downloading the model, and initializing the user scripts. -- Azure Machine Learning extension support Kubernetes Baseline Pod Security Standard. +- Azure Machine Learning extension supports Kubernetes Baseline Pod Security Standard. - Training on Kubernetes compute doesn't support auto scale nodes. ## Recommended best practices