-
Clone the repository in your local machine.
-
Install Azure ML CLI v2. If you don't have it, follow the installation instructions at Install and set up the CLI (v2).
-
Create a compute named
trainer-cpu
or rename the compute specified in .aml/jobs/carpricer.job.yml. -
Register the dataset:
az ml data create -f .aml/data/product-reviews-train.yml az ml data create -f .aml/data/product-reviews-eval.yml
-
Create the training job:
az ml job create -f .aml/jobs/carpricer.job.yml
(Optional)
-
Register the trained model in the registry:
JOB_NAME=$(az ml job list --query "[0].name" | tr -d '"') az ml model create --name "carpricer" \ --type "mlflow_model" \ --path "azureml://jobs/$JOB_NAME/outputs/artifacts/pipeline"
-
Deploy the model in an online endpoint:
az ml online-endpoint create -f .aml/endpoints/carpricer-online/endpoint.yml az ml online-deployment create -f .aml/endpoints/carpricer-online/deployments/default.yml --all-traffic
This project welcomes contributions and suggestions. Open an issue and start the discussion!