Hi, it this notebook,
https://github.com/Azure/Machine-Learning-Operationalization/blob/master/samples/python/tutorials/realtime/digit_classification.ipynb
Step 4. Use the CLI to deploy and manage your web service
mentions to create a cluster in Step (1) below...but I already have a resource group.
(1) Create the environment (you need to do this once per environment e.g. dev or prod)
az ml env setup -c -n --location <e.g. eastus2>
(2) Create a Model Management account (one time setup)
az ml account modelmanagement create --location <e.g. eastus2> -n -g
(3) Set the Model Management account
az ml account modelmanagement set -n -g
(4) Set the environment. The cluster name is the name used in step 1 above. The resource group name was the output of the same process and would be in the command window when the setup process is completed.
az ml env set -n -g
Questions
(1) How can I reuse the resource group and check if clusters already exist?
(2) Using the command you mention, what is a cluster being created...what resources are created?
(3) What is the relation between a cluster and the model management account?
(4) Do both the cluster and the model management account need to be in the same location?
Hi, it this notebook,
https://github.com/Azure/Machine-Learning-Operationalization/blob/master/samples/python/tutorials/realtime/digit_classification.ipynb
Step 4. Use the CLI to deploy and manage your web service
mentions to create a cluster in Step (1) below...but I already have a resource group.
(1) Create the environment (you need to do this once per environment e.g. dev or prod)
az ml env setup -c -n --location <e.g. eastus2>
(2) Create a Model Management account (one time setup)
az ml account modelmanagement create --location <e.g. eastus2> -n -g
(3) Set the Model Management account
az ml account modelmanagement set -n -g
(4) Set the environment. The cluster name is the name used in step 1 above. The resource group name was the output of the same process and would be in the command window when the setup process is completed.
az ml env set -n -g
Questions
(1) How can I reuse the resource group and check if clusters already exist?
(2) Using the command you mention, what is a cluster being created...what resources are created?
(3) What is the relation between a cluster and the model management account?
(4) Do both the cluster and the model management account need to be in the same location?