Merge pull request #121 from Azure/main-dec31

Merge main-dec31 into main
2023-01-05 09:24:12 -06:00 · 2023-01-05 09:24:12 -06:00 · 09470468e1
--- a/classical/aml-cli-v2/data-science/environment/train-conda.yml
+++ b/classical/aml-cli-v2/data-science/environment/train-conda.yml
@ -7,7 +7,8 @@ dependencies:
  - pip
  - pip:
      - azureml-mlflow==1.38.0
-      - azureml-sdk==1.38.0
+      - azure-ai-ml==1.0.0
+      - pyarrow==10.0.0
      - scikit-learn==0.24.1
      - pandas==1.2.1
      - joblib==1.0.0
--- a/classical/aml-cli-v2/data-science/src/prep.py
+++ b/classical/aml-cli-v2/data-science/src/prep.py
@ -4,6 +4,12 @@
 Prepares raw data and provides training, validation and test datasets
 """

+# Copyright (c) Microsoft Corporation. All rights reserved.
+# Licensed under the MIT License.
+"""
+Prepares raw data and provides training, validation and test datasets
+"""
+
 import argparse

 from pathlib import Path
@ -41,21 +47,52 @@ CAT_NOM_COLS = [
    "vendor",
 ]

+CAT_ORD_COLS = [
+]
+TARGET_COL = "cost"
+
+NUMERIC_COLS = [
+    "distance",
+    "dropoff_latitude",
+    "dropoff_longitude",
+    "passengers",
+    "pickup_latitude",
+    "pickup_longitude",
+    "pickup_weekday",
+    "pickup_month",
+    "pickup_monthday",
+    "pickup_hour",
+    "pickup_minute",
+    "pickup_second",
+    "dropoff_weekday",
+    "dropoff_month",
+    "dropoff_monthday",
+    "dropoff_hour",
+    "dropoff_minute",
+    "dropoff_second",
+]
+
+CAT_NOM_COLS = [
+    "store_forward",
+    "vendor",
+]
+
 CAT_ORD_COLS = [
 ]

 def parse_args():
    '''Parse input arguments'''
+    '''Parse input arguments'''

    parser = argparse.ArgumentParser("prep")
    parser.add_argument("--raw_data", type=str, help="Path to raw data")
    parser.add_argument("--train_data", type=str, help="Path to train dataset")
    parser.add_argument("--val_data", type=str, help="Path to test dataset")
    parser.add_argument("--test_data", type=str, help="Path to test dataset")
-    
+
    parser.add_argument("--enable_monitoring", type=str, help="enable logging to ADX")
    parser.add_argument("--table_name", type=str, default="mlmonitoring", help="Table name in ADX for logging")
-    
+
    args = parser.parse_args()

    return args
@ -65,17 +102,15 @@ def log_training_data(df, table_name):
    collector = Online_Collector(table_name)
    collector.batch_collect(df)

+def main(args):
+    '''Read, split, and save datasets'''
 def main(args):
    '''Read, split, and save datasets'''

    # ------------ Reading Data ------------ #
    # -------------------------------------- #

-    print("mounted_path files: ")
-    arr = os.listdir(args.raw_data)
-    print(arr)
-
-    data = pd.read_csv((Path(args.raw_data) / 'taxi-data.csv'))
+    data = pd.read_csv((Path(args.raw_data)))
    data = data[NUMERIC_COLS + CAT_NOM_COLS + CAT_ORD_COLS + [TARGET_COL]]

    # ------------- Split Data ------------- #
@ -92,19 +127,24 @@ def main(args):
    train = data[msk_train]
    val = data[msk_val]
    test = data[msk_test]
+    test = data[msk_test]

    mlflow.log_metric('train size', train.shape[0])
    mlflow.log_metric('val size', val.shape[0])
    mlflow.log_metric('test size', test.shape[0])

+    train.to_parquet((Path(args.train_data) / "train.parquet"))
+    val.to_parquet((Path(args.val_data) / "val.parquet"))
+    test.to_parquet((Path(args.test_data) / "test.parquet"))
    train.to_parquet((Path(args.train_data) / "train.parquet"))
    val.to_parquet((Path(args.val_data) / "val.parquet"))
    test.to_parquet((Path(args.test_data) / "test.parquet"))

-    if (args.enable_monitoring.lower == 'true' or args.enable_monitoring == '1' or args.enable_monitoring.lower == 'yes'):
+    if (args.enable_monitoring.lower() == 'true' or args.enable_monitoring == '1' or args.enable_monitoring.lower() == 'yes'):
        log_training_data(data, args.table_name)


+
 if __name__ == "__main__":

    mlflow.start_run()
@ -124,9 +164,9 @@ if __name__ == "__main__":

    for line in lines:
        print(line)
-    
+
    main(args)

    mlflow.end_run()

-    
+
--- a/classical/aml-cli-v2/mlops/azureml/train/data.yml
+++ b/classical/aml-cli-v2/mlops/azureml/train/data.yml
@ -0,0 +1,11 @@
+$schema: https://azuremlschemas.azureedge.net/latest/data.schema.json
+
+# Supported paths include:
+# local: ./<path>
+# blob:  https://<account_name>.blob.core.windows.net/<container_name>/<path>
+# ADLS gen2: abfss://<file_system>@<account_name>.dfs.core.windows.net/<path>/
+# Datastore: azureml://datastores/<data_store_name>/paths/<path>
+type: uri_file
+name: taxi-data
+description: taxi dataset
+path: ../../../data/taxi-data.csv
--- a/classical/aml-cli-v2/mlops/azureml/train/pipeline.yml
+++ b/classical/aml-cli-v2/mlops/azureml/train/pipeline.yml
@ -1,17 +1,18 @@
 $schema: https://azuremlschemas.azureedge.net/latest/pipelineJob.schema.json
 type: pipeline
 experiment_name: taxi-fare-training
+experiment_name: taxi-fare-training
 description: Training Pipeline to train a model that predicts taxi fare price

 # <inputs_and_outputs>
 inputs:
-  input: #using local data, will create an anonymous data asset
-    type: uri_folder
-    path: ../../../data/
+  input:
+    type: uri_file
+    path: azureml:taxi-data@latest
  enable_monitoring: 'false'
  table_name: 'taximonitoring'

-outputs: 
+outputs:
  train_data:
  val_data:
  test_data:
@ -32,8 +33,8 @@ jobs:
    display_name: prep-data
    code: ../../../data-science/src
    command: >-
-      python prep.py 
-      --raw_data ${{inputs.raw_data}} 
+      python prep.py
+      --raw_data ${{inputs.raw_data}}
      --train_data ${{outputs.train_data}}
      --val_data ${{outputs.val_data}}
      --test_data ${{outputs.test_data}}
@ -54,24 +55,24 @@ jobs:
    display_name: train-model
    code: ../../../data-science/src
    command: >-
-      python train.py 
-      --train_data ${{inputs.train_data}} 
+      python train.py
+      --prepared_data ${{inputs.prepared_data}}
      --model_output ${{outputs.model_output}}
    environment: azureml:taxi-train-env@latest
    inputs:
      train_data: ${{parent.jobs.prep_data.outputs.train_data}}
    outputs:
      model_output: ${{parent.outputs.trained_model}}
-  
+
  evaluate_model:
    name: evaluate_model
    display_name: evaluate-model
    code: ../../../data-science/src
    command: >-
-      python evaluate.py 
-      --model_name ${{inputs.model_name}} 
-      --model_input ${{inputs.model_input}} 
-      --test_data ${{inputs.test_data}} 
+      python evaluate.py
+      --model_name ${{inputs.model_name}}
+      --model_input ${{inputs.model_input}}
+      --test_data ${{inputs.test_data}}
      --evaluation_output ${{outputs.evaluation_output}}
    environment: azureml:taxi-train-env@latest
    inputs:
@ -86,9 +87,9 @@ jobs:
    display_name: register-model
    code: ../../../data-science/src
    command: >-
-      python register.py 
-      --model_name ${{inputs.model_name}} 
-      --model_path ${{inputs.model_path}} 
+      python register.py
+      --model_name ${{inputs.model_name}}
+      --model_path ${{inputs.model_path}}
      --evaluation_output ${{inputs.evaluation_output}}
      --model_info_output_path ${{outputs.model_info_output_path}}
    environment: azureml:taxi-train-env@latest
--- a/classical/aml-cli-v2/mlops/devops-pipelines/deploy-model-training-pipeline.yml
+++ b/classical/aml-cli-v2/mlops/devops-pipelines/deploy-model-training-pipeline.yml
@ -2,61 +2,62 @@
 # Licensed under the MIT License.

 variables:
- ${{ if eq(variables['Build.SourceBranchName'], 'main') }}:
-    # 'main' branch: PRD environment
-    - template: ../../config-infra-prod.yml
- ${{ if ne(variables['Build.SourceBranchName'], 'main') }}:  
-    # 'develop' or feature branches: DEV environment
-    - template: ../../config-infra-dev.yml
- name: version
-  value: aml-cli-v2
-
+  - ${{ if eq(variables['Build.SourceBranchName'], 'main') }}:
+      # 'main' branch: PRD environment
+      - template: ../../config-infra-prod.yml
+  - ${{ if ne(variables['Build.SourceBranchName'], 'main') }}:
+      # 'develop' or feature branches: DEV environment
+      - template: ../../config-infra-dev.yml
+  - name: version
+    value: aml-cli-v2

 trigger:
- none
+  - none

 pool:
  vmImage: ubuntu-20.04

-
 resources:
  repositories:
-    - repository: mlops-templates  # Template Repo
+    - repository: mlops-templates # Template Repo
      name: Azure/mlops-templates # need to change org name from "Azure" to your own org
      endpoint: github-connection # need to set up and hardcode
      type: github
      ref: main

 stages:
- stage: DeployTrainingPipeline
-  displayName: Deploy Training Pipeline
-  jobs:
-    - job: DeployTrainingPipeline
-      timeoutInMinutes: 120 # how long to run the job before automatically cancelling
-      steps:
-      - checkout: self
-        path: s/
-      - checkout: mlops-templates
-        path: s/templates/
-      - template: templates/${{ variables.version }}/install-az-cli.yml@mlops-templates
-      - template: templates/${{ variables.version }}/install-aml-cli.yml@mlops-templates
-      - template: templates/${{ variables.version }}/connect-to-workspace.yml@mlops-templates
-      - template: templates/${{ variables.version }}/register-environment.yml@mlops-templates
-        parameters:
-          build_type: conda
-          environment_name: taxi-train-env
-          environment_file: mlops/azureml/train/train-env.yml
-          enable_monitoring: $(enable_monitoring)
-      - template: templates/${{ variables.version }}/create-compute.yml@mlops-templates
-        parameters:
-          cluster_name: cpu-cluster
-          size: Standard_DS3_v2
-          min_instances: 0
-          max_instances: 4
-          cluster_tier: LowPriority
-      - template: templates/${{ variables.version }}/run-pipeline.yml@mlops-templates
-        parameters: 
-          pipeline_file: mlops/azureml/train/pipeline.yml
-          experiment_name: $(environment)_taxi_fare_train_$(Build.SourceBranchName)
-          display_name: $(environment)_taxi_fare_run_$(Build.BuildID)
-          enable_monitoring: $(enable_monitoring)
+  - stage: DeployTrainingPipeline
+    displayName: Deploy Training Pipeline
+    jobs:
+      - job: DeployTrainingPipeline
+        timeoutInMinutes: 120 # how long to run the job before automatically cancelling
+        steps:
+          - checkout: self
+            path: s/
+          - checkout: mlops-templates
+            path: s/templates/
+          - template: templates/${{ variables.version }}/install-az-cli.yml@mlops-templates
+          - template: templates/${{ variables.version }}/install-aml-cli.yml@mlops-templates
+          - template: templates/${{ variables.version }}/connect-to-workspace.yml@mlops-templates
+          - template: templates/${{ variables.version }}/register-environment.yml@mlops-templates
+            parameters:
+              environment_name: taxi-train-env
+              environment_file: mlops/azureml/train/train-env.yml
+          - template: templates/${{ variables.version }}/create-compute.yml@mlops-templates
+            parameters:
+              cluster_name: cpu-cluster
+              size: Standard_DS3_v2
+              min_instances: 0
+              max_instances: 4
+              cluster_tier: low_priority
+          - template: templates/${{ variables.version }}/register-data.yml@mlops-templates
+            parameters:
+              data_type: uri_file
+              data_name: taxi-data
+              data_file: mlops/azureml/train/data.yml
+          - template: templates/${{ variables.version }}/run-pipeline.yml@mlops-templates
+            parameters:
+              pipeline_file: mlops/azureml/train/pipeline.yml
+              experiment_name: $(environment)_taxi_fare_train_$(Build.SourceBranchName)
+              display_name: $(environment)_taxi_fare_run_$(Build.BuildID)
+              enable_monitoring: $(enable_monitoring)
--- a/classical/aml-cli-v2/mlops/github-actions/deploy-batch-endpoint-pipeline-classical.yml
+++ b/classical/aml-cli-v2/mlops/github-actions/deploy-batch-endpoint-pipeline-classical.yml
@ -1,44 +1,63 @@
 name: deploy-batch-endpoint-pipeline

-on: 
+on:
  workflow_dispatch:
 jobs:
+  set-env-branch:
+    runs-on: ubuntu-latest
+    outputs:
+      config-file: ${{ steps.set-output-defaults.outputs.config-file }}
+    steps:
+      - id: set-prod-branch
+        name: set-prod-branch
+        if: ${{ github.ref == 'refs/heads/main'}}
+        run: echo "config_env=config-infra-prod.yml" >> $GITHUB_ENV;
+      - id: set-dev-branch
+        name: setdevbranch
+        if: ${{ github.ref != 'refs/heads/main'}}
+        run: echo "config_env=config-infra-dev.yml" >> $GITHUB_ENV;
+      - id: set-output-defaults
+        name: set-output-defaults
+        run: |
+          echo "config-file=$config_env" >> $GITHUB_OUTPUT;
  get-config:
-      uses: Azure/mlops-templates/.github/workflows/read-yaml.yml@main
-      with:
-        file_name: config-infra-prod.yml
+    needs: set-env-branch
+    uses: Azure/mlops-templates/.github/workflows/read-yaml.yml@main
+    with:
+      file_name: ${{ needs.set-env-branch.outputs.config-file}}
  create-compute:
-      needs: get-config
-      uses: Azure/mlops-templates/.github/workflows/create-compute.yml@main
-      with:
-        cluster_name: batch-cluster
-        size: STANDARD_DS3_V2
-        min_instances: 0
-        max_instances: 5
-        resource_group: ${{ needs.get-config.outputs.resource_group }}
-        workspace_name: ${{ needs.get-config.outputs.aml_workspace }}     
-      secrets:
-          creds: ${{secrets.AZURE_CREDENTIALS}}
+    needs: get-config
+    uses: Azure/mlops-templates/.github/workflows/create-compute.yml@main
+    with:
+      cluster_name: batch-cluster
+      size: STANDARD_DS3_V2
+      min_instances: 0
+      max_instances: 5
+      cluster_tier: low_priority
+      resource_group: ${{ needs.get-config.outputs.resource_group }}
+      workspace_name: ${{ needs.get-config.outputs.aml_workspace }}
+    secrets:
+      creds: ${{secrets.AZURE_CREDENTIALS}}
  create-endpoint:
-      needs: [get-config,create-compute]
-      uses: Azure/mlops-templates/.github/workflows/create-endpoint.yml@main
-      with:
-        resource_group: ${{ needs.get-config.outputs.resource_group }}
-        workspace_name: ${{ needs.get-config.outputs.aml_workspace }} 
-        endpoint_file: mlops/azureml/deploy/batch/batch-endpoint.yml
-        endpoint_name:  ${{ format('taxi-gha-{0}', needs.get-config.outputs.bep) }} 
-        endpoint_type: batch
-      secrets:
-        creds: ${{secrets.AZURE_CREDENTIALS}}
+    needs: [get-config, create-compute]
+    uses: Azure/mlops-templates/.github/workflows/create-endpoint.yml@main
+    with:
+      resource_group: ${{ needs.get-config.outputs.resource_group }}
+      workspace_name: ${{ needs.get-config.outputs.aml_workspace }}
+      endpoint_file: mlops/azureml/deploy/batch/batch-endpoint.yml
+      endpoint_name: ${{ format('taxi-gha-{0}', needs.get-config.outputs.bep) }}
+      endpoint_type: batch
+    secrets:
+      creds: ${{secrets.AZURE_CREDENTIALS}}
  create-deployment:
-      uses: Azure/mlops-templates/.github/workflows/create-deployment.yml@main
-      needs: [get-config,create-endpoint]
-      with:
-        resource_group: ${{ needs.get-config.outputs.resource_group }}
-        workspace_name: ${{ needs.get-config.outputs.aml_workspace }} 
-        endpoint_file: mlops/azureml/deploy/batch/batch-deployment.yml
-        endpoint_name: ${{ format('taxi-gha-{0}', needs.get-config.outputs.bep) }} 
-        endpoint_type: batch
-        deployment_name: eptestdeploy
-      secrets:
-        creds: ${{secrets.AZURE_CREDENTIALS}}
+    uses: Azure/mlops-templates/.github/workflows/create-deployment.yml@main
+    needs: [get-config, create-endpoint]
+    with:
+      resource_group: ${{ needs.get-config.outputs.resource_group }}
+      workspace_name: ${{ needs.get-config.outputs.aml_workspace }}
+      endpoint_file: mlops/azureml/deploy/batch/batch-deployment.yml
+      endpoint_name: ${{ format('taxi-gha-{0}', needs.get-config.outputs.bep) }}
+      endpoint_type: batch
+      deployment_name: eptestdeploy
+    secrets:
+      creds: ${{secrets.AZURE_CREDENTIALS}}
--- a/classical/aml-cli-v2/mlops/github-actions/deploy-model-training-pipeline-classical.yml
+++ b/classical/aml-cli-v2/mlops/github-actions/deploy-model-training-pipeline-classical.yml
@ -1,42 +1,70 @@
 name: deploy-model-training-pipeline

-on: 
+on:
  workflow_dispatch:
 jobs:
+  set-env-branch:
+    runs-on: ubuntu-latest
+    outputs:
+      config-file: ${{ steps.set-output-defaults.outputs.config-file }}
+    steps:
+      - id: set-prod-branch
+        name: set-prod-branch
+        if: ${{ github.ref == 'refs/heads/main'}}
+        run: echo "config_env=config-infra-prod.yml" >> $GITHUB_ENV;
+      - id: set-dev-branch
+        name: setdevbranch
+        if: ${{ github.ref != 'refs/heads/main'}}
+        run: echo "config_env=config-infra-dev.yml" >> $GITHUB_ENV;
+      - id: set-output-defaults
+        name: set-output-defaults
+        run: |
+          echo "config-file=$config_env" >> $GITHUB_OUTPUT;
  get-config:
-      uses: Azure/mlops-templates/.github/workflows/read-yaml.yml@main
-      with:
-        file_name: config-infra-prod.yml
+    needs: set-env-branch
+    uses: Azure/mlops-templates/.github/workflows/read-yaml.yml@main
+    with:
+      file_name: ${{ needs.set-env-branch.outputs.config-file}}
  register-environment:
-      needs: get-config
-      uses: Azure/mlops-templates/.github/workflows/register-environment.yml@main
-      with:
-        resource_group: ${{ needs.get-config.outputs.resource_group }}
-        workspace_name: ${{ needs.get-config.outputs.aml_workspace }}     
-        environment_file:  mlops/azureml/train/train-env.yml
-        conda_file: data-science/environment/train-conda.yml
-      secrets:
-          creds: ${{secrets.AZURE_CREDENTIALS}}
+   needs: get-config
+    uses: Azure/mlops-templates/.github/workflows/register-environment.yml@main
+    with:
+      resource_group: ${{ needs.get-config.outputs.resource_group }}
+      workspace_name: ${{ needs.get-config.outputs.aml_workspace }}
+      environment_file: mlops/azureml/train/train-env.yml
+      conda_file: data-science/environment/train-conda.yml
+    secrets:
+      creds: ${{secrets.AZURE_CREDENTIALS}}
+  register-dataset:
+    needs: get-config
+    uses: Azure/mlops-templates/.github/workflows/register-dataset.yml@main
+    with:
+      resource_group: ${{ needs.get-config.outputs.resource_group }}
+      workspace_name: ${{ needs.get-config.outputs.aml_workspace }}
+      name: taxi-data
+      data_file: mlops/azureml/train/data.yml
+    secrets:
+      creds: ${{secrets.AZURE_CREDENTIALS}}
  create-compute:
-      needs: [get-config,register-environment]
-      uses: Azure/mlops-templates/.github/workflows/create-compute.yml@main
-      with:
-        cluster_name: cpu-cluster
-        size: Standard_DS3_v2
-        min_instances: 0
-        max_instances: 4
-        cluster_tier: LowPriority
-        resource_group: ${{ needs.get-config.outputs.resource_group }}
-        workspace_name: ${{ needs.get-config.outputs.aml_workspace }}     
-      secrets:
-          creds: ${{secrets.AZURE_CREDENTIALS}}
-  run-pipeline:
-      needs: [get-config,register-environment,create-compute]
-      uses: Azure/mlops-templates/.github/workflows/run-pipeline.yml@main
-      with:
-        resource_group: ${{ needs.get-config.outputs.resource_group }}
-        workspace_name: ${{ needs.get-config.outputs.aml_workspace }}     
-        parameters-file:  mlops/azureml/train/pipeline.yml
-        job-name: test
-      secrets:
-          creds: ${{secrets.AZURE_CREDENTIALS}}
+    needs: [get-config]
+    uses: Azure/mlops-templates/.github/workflows/create-compute.yml@main
+    with:
+      cluster_name: cpu-cluster
+      size: Standard_DS3_v2
+      min_instances: 0
+      max_instances: 4
+      cluster_tier: low_priority
+      resource_group: ${{ needs.get-config.outputs.resource_group }}
+      workspace_name: ${{ needs.get-config.outputs.aml_workspace }}
+    secrets:
+      creds: ${{secrets.AZURE_CREDENTIALS}}
+  run-model-training-pipeline:
+    needs: [get-config, register-environment, register-dataset, create-compute]
+    uses: Azure/mlops-templates/.github/workflows/run-pipeline.yml@main
+    with:
+      resource_group: ${{ needs.get-config.outputs.resource_group }}
+      workspace_name: ${{ needs.get-config.outputs.aml_workspace }}
+      parameters-file: mlops/azureml/train/pipeline.yml
+      job-name: test
+    secrets:
+      creds: ${{secrets.AZURE_CREDENTIALS}}
--- a/classical/aml-cli-v2/mlops/github-actions/deploy-online-endpoint-pipeline-classical.yml
+++ b/classical/aml-cli-v2/mlops/github-actions/deploy-online-endpoint-pipeline-classical.yml
@ -1,42 +1,60 @@
 name: deploy-online-endpoint-pipeline

-on: 
+on:
  workflow_dispatch:
 jobs:
+  set-env-branch:
+    runs-on: ubuntu-latest
+    outputs:
+      config-file: ${{ steps.set-output-defaults.outputs.config-file }}
+    steps:
+      - id: set-prod-branch
+        name: set-prod-branch
+        if: ${{ github.ref == 'refs/heads/main'}}
+        run: echo "config_env=config-infra-prod.yml" >> $GITHUB_ENV;
+      - id: set-dev-branch
+        name: setdevbranch
+        if: ${{ github.ref != 'refs/heads/main'}}
+        run: echo "config_env=config-infra-dev.yml" >> $GITHUB_ENV;
+      - id: set-output-defaults
+        name: set-output-defaults
+        run: |
+          echo "config-file=$config_env" >> $GITHUB_OUTPUT;
  get-config:
-      uses: Azure/mlops-templates/.github/workflows/read-yaml.yml@main
-      with:
-        file_name: config-infra-prod.yml
+    needs: set-env-branch
+    uses: Azure/mlops-templates/.github/workflows/read-yaml.yml@main
+    with:
+      file_name: ${{ needs.set-env-branch.outputs.config-file}}
  create-endpoint:
-      needs: get-config
-      uses: Azure/mlops-templates/.github/workflows/create-endpoint.yml@main
-      with:
-        resource_group: ${{ needs.get-config.outputs.resource_group }}
-        workspace_name: ${{ needs.get-config.outputs.aml_workspace }} 
-        endpoint_file: mlops/azureml/deploy/online/online-endpoint.yml
-        endpoint_name: ${{ format('taxi-gha-{0}', needs.get-config.outputs.oep) }}
-        endpoint_type: online
-      secrets:
-        creds: ${{secrets.AZURE_CREDENTIALS}}
+    needs: get-config
+    uses: Azure/mlops-templates/.github/workflows/create-endpoint.yml@main
+    with:
+      resource_group: ${{ needs.get-config.outputs.resource_group }}
+      workspace_name: ${{ needs.get-config.outputs.aml_workspace }}
+      endpoint_file: mlops/azureml/deploy/online/online-endpoint.yml
+      endpoint_name: ${{ format('taxi-gha-{0}', needs.get-config.outputs.oep) }}
+      endpoint_type: online
+    secrets:
+      creds: ${{secrets.AZURE_CREDENTIALS}}
  create-deployment:
-      uses: Azure/mlops-templates/.github/workflows/create-deployment.yml@main
-      needs: [get-config,create-endpoint]
-      with:
-        resource_group: ${{ needs.get-config.outputs.resource_group }}
-        workspace_name: ${{ needs.get-config.outputs.aml_workspace }} 
-        endpoint_file: mlops/azureml/deploy/online/online-deployment.yml
-        endpoint_name: ${{ format('taxi-gha-{0}', needs.get-config.outputs.oep) }}
-        endpoint_type: online
-        deployment_name: taxi-online-dp
-      secrets:
-        creds: ${{secrets.AZURE_CREDENTIALS}}
+    uses: Azure/mlops-templates/.github/workflows/create-deployment.yml@main
+    needs: [get-config, create-endpoint]
+    with:
+      resource_group: ${{ needs.get-config.outputs.resource_group }}
+      workspace_name: ${{ needs.get-config.outputs.aml_workspace }}
+      endpoint_file: mlops/azureml/deploy/online/online-deployment.yml
+      endpoint_name: ${{ format('taxi-gha-{0}', needs.get-config.outputs.oep) }}
+      endpoint_type: online
+      deployment_name: taxi-online-dp
+    secrets:
+      creds: ${{secrets.AZURE_CREDENTIALS}}
  allocate-traffic:
-      uses: Azure/mlops-templates/.github/workflows/allocate-traffic.yml@main
-      needs: [get-config,create-deployment]
-      with:
-        resource_group: ${{ needs.get-config.outputs.resource_group }}
-        workspace_name: ${{ needs.get-config.outputs.aml_workspace }} 
-        traffic_allocation: taxi-online-dp=100
-        endpoint_name: ${{ format('taxi-gha-{0}', needs.get-config.outputs.oep) }}
-      secrets:
-        creds: ${{secrets.AZURE_CREDENTIALS}}
+    uses: Azure/mlops-templates/.github/workflows/allocate-traffic.yml@main
+    needs: [get-config, create-deployment]
+    with:
+      resource_group: ${{ needs.get-config.outputs.resource_group }}
+      workspace_name: ${{ needs.get-config.outputs.aml_workspace }}
+      traffic_allocation: taxi-online-dp=100
+      endpoint_name: ${{ format('taxi-gha-{0}', needs.get-config.outputs.oep) }}
+    secrets:
+      creds: ${{secrets.AZURE_CREDENTIALS}}
--- a/classical/python-sdk-v1/config-aml.yml
+++ b/classical/python-sdk-v1/config-aml.yml
--- a/classical/python-sdk-v1/data-science/environment/batch.yml
+++ b/classical/python-sdk-v1/data-science/environment/batch.yml
--- a/classical/python-sdk-v1/data-science/environment/batch_monitor.yml
+++ b/classical/python-sdk-v1/data-science/environment/batch_monitor.yml
--- a/classical/python-sdk-v1/data-science/environment/train.yml
+++ b/classical/python-sdk-v1/data-science/environment/train.yml
--- a/classical/python-sdk-v1/data-science/environment/train_monitor.yml
+++ b/classical/python-sdk-v1/data-science/environment/train_monitor.yml
--- a/classical/python-sdk-v1/data-science/notebooks/experiment1.ipynb
+++ b/classical/python-sdk-v1/data-science/notebooks/experiment1.ipynb
--- a/classical/python-sdk-v1/data-science/src/evaluate.py
+++ b/classical/python-sdk-v1/data-science/src/evaluate.py
--- a/classical/python-sdk-v1/data-science/src/prep.py
+++ b/classical/python-sdk-v1/data-science/src/prep.py
--- a/classical/python-sdk-v1/data-science/src/score.py
+++ b/classical/python-sdk-v1/data-science/src/score.py
--- a/classical/python-sdk-v1/data-science/src/train.py
+++ b/classical/python-sdk-v1/data-science/src/train.py
--- a/classical/python-sdk-v1/data-science/tests/test.py
+++ b/classical/python-sdk-v1/data-science/tests/test.py
--- a/classical/python-sdk-v1/data/scoring/credit_batch.csv
+++ b/classical/python-sdk-v1/data/scoring/credit_batch.csv
--- a/classical/python-sdk-v1/data/training/credit.csv
+++ b/classical/python-sdk-v1/data/training/credit.csv
--- a/classical/python-sdk-v1/mlops/devops-pipelines/deploy-drift-detection.yml
+++ b/classical/python-sdk-v1/mlops/devops-pipelines/deploy-drift-detection.yml
@ -10,7 +10,7 @@ variables:
    # 'develop' or feature branches: DEV environment
    - template: ../../config-infra-dev.yml
 - name: version
-  value: python-sdk
+  value: python-sdk-v1

 trigger:
 - none
--- a/classical/python-sdk-v1/mlops/devops-pipelines/deploy-model-batch-scoring.yml
+++ b/classical/python-sdk-v1/mlops/devops-pipelines/deploy-model-batch-scoring.yml
@ -10,7 +10,7 @@ variables:
    # 'develop' or feature branches: DEV environment
    - template: ../../config-infra-dev.yml
 - name: version
-  value: python-sdk
+  value: python-sdk-v1

 trigger:
 - none
--- a/classical/python-sdk-v1/mlops/devops-pipelines/deploy-model-training-pipeline.yml
+++ b/classical/python-sdk-v1/mlops/devops-pipelines/deploy-model-training-pipeline.yml
@ -10,7 +10,7 @@ variables:
    # 'develop' or feature branches: DEV environment
    - template: ../../config-infra-dev.yml
 - name: version
-  value: python-sdk
+  value: python-sdk-v1

 trigger:
 - none
--- a/classical/aml-cli-v2/data-science/environment/train-monitor-conda.yml
+++ b/classical/aml-cli-v2/data-science/environment/train-monitor-conda.yml
@ -7,7 +7,8 @@ dependencies:
  - pip
  - pip:
      - azureml-mlflow==1.38.0
-      - azureml-sdk==1.38.0
+      - azure-ai-ml==1.0.0
+      - pyarrow==10.0.0
      - scikit-learn==0.24.1
      - pandas==1.2.1
      - joblib==1.0.0
--- a/classical/python-sdk-v2/data-science/src/evaluate/evaluate.py
+++ b/classical/python-sdk-v2/data-science/src/evaluate/evaluate.py
@ -0,0 +1,174 @@
+# Copyright (c) Microsoft Corporation. All rights reserved.
+# Licensed under the MIT License.
+"""
+Evaluates trained ML model using test dataset.
+Saves predictions, evaluation results and deploy flag.
+"""
+
+import argparse
+from pathlib import Path
+
+import numpy as np
+import pandas as pd
+from matplotlib import pyplot as plt
+
+from sklearn.metrics import r2_score, mean_absolute_error, mean_squared_error
+
+import mlflow
+import mlflow.sklearn
+import mlflow.pyfunc
+from mlflow.tracking import MlflowClient
+
+TARGET_COL = "cost"
+
+NUMERIC_COLS = [
+    "distance", "dropoff_latitude", "dropoff_longitude", "passengers", "pickup_latitude",
+    "pickup_longitude", "pickup_weekday", "pickup_month", "pickup_monthday", "pickup_hour",
+    "pickup_minute", "pickup_second", "dropoff_weekday", "dropoff_month", "dropoff_monthday",
+    "dropoff_hour", "dropoff_minute", "dropoff_second"
+]
+
+CAT_NOM_COLS = [
+    "store_forward", "vendor"
+]
+
+CAT_ORD_COLS = [
+]
+
+def parse_args():
+    '''Parse input arguments'''
+
+    parser = argparse.ArgumentParser("predict")
+    parser.add_argument("--model_name", type=str, help="Name of registered model")
+    parser.add_argument("--model_input", type=str, help="Path of input model")
+    parser.add_argument("--test_data", type=str, help="Path to test dataset")
+    parser.add_argument("--evaluation_output", type=str, help="Path of eval results")
+
+    args = parser.parse_args()
+
+    return args
+
+def main(args):
+    '''Read trained model and test dataset, evaluate model and save result'''
+
+    # Load the test data
+    test_data = pd.read_parquet(Path(args.test_data))
+
+    # Split the data into inputs and outputs
+    y_test = test_data[TARGET_COL]
+    X_test = test_data[NUMERIC_COLS + CAT_NOM_COLS + CAT_ORD_COLS]
+
+    # Load the model from input port
+    model =  mlflow.sklearn.load_model(args.model_input) 
+
+    # ---------------- Model Evaluation ---------------- #
+    yhat_test, score = model_evaluation(X_test, y_test, model, args.evaluation_output)
+
+    # ----------------- Model Promotion ---------------- #
+    predictions, deploy_flag = model_promotion(args.model_name, args.evaluation_output, X_test, y_test, yhat_test, score)
+
+
+
+def model_evaluation(X_test, y_test, model, evaluation_output):
+
+    # Get predictions to y_test (y_test)
+    yhat_test = model.predict(X_test)
+
+    # Save the output data with feature columns, predicted cost, and actual cost in csv file
+    output_data = X_test.copy()
+    output_data["real_label"] = y_test
+    output_data["predicted_label"] = yhat_test
+    output_data.to_csv((Path(evaluation_output) / "predictions.csv"))
+
+    # Evaluate Model performance with the test set
+    r2 = r2_score(y_test, yhat_test)
+    mse = mean_squared_error(y_test, yhat_test)
+    rmse = np.sqrt(mse)
+    mae = mean_absolute_error(y_test, yhat_test)
+
+    # Print score report to a text file
+    (Path(evaluation_output) / "score.txt").write_text(
+        f"Scored with the following model:\n{format(model)}"
+    )
+    with open((Path(evaluation_output) / "score.txt"), "a") as outfile:
+        outfile.write("Mean squared error: {mse.2f} \n")
+        outfile.write("Root mean squared error: {rmse.2f} \n")
+        outfile.write("Mean absolute error: {mae.2f} \n")
+        outfile.write("Coefficient of determination: {r2.2f} \n")
+
+    mlflow.log_metric("test r2", r2)
+    mlflow.log_metric("test mse", mse)
+    mlflow.log_metric("test rmse", rmse)
+    mlflow.log_metric("test mae", mae)
+
+    # Visualize results
+    plt.scatter(y_test, yhat_test,  color='black')
+    plt.plot(y_test, y_test, color='blue', linewidth=3)
+    plt.xlabel("Real value")
+    plt.ylabel("Predicted value")
+    plt.title("Comparing Model Predictions to Real values - Test Data")
+    plt.savefig("predictions.png")
+    mlflow.log_artifact("predictions.png")
+
+    return yhat_test, r2
+
+def model_promotion(model_name, evaluation_output, X_test, y_test, yhat_test, score):
+    
+    scores = {}
+    predictions = {}
+
+    client = MlflowClient()
+
+    for model_run in client.search_model_versions(f"name='{model_name}'"):
+        model_version = model_run.version
+        mdl = mlflow.pyfunc.load_model(
+            model_uri=f"models:/{model_name}/{model_version}")
+        predictions[f"{model_name}:{model_version}"] = mdl.predict(X_test)
+        scores[f"{model_name}:{model_version}"] = r2_score(
+            y_test, predictions[f"{model_name}:{model_version}"])
+
+    if scores:
+        if score >= max(list(scores.values())):
+            deploy_flag = 1
+        else:
+            deploy_flag = 0
+    else:
+        deploy_flag = 1
+    print(f"Deploy flag: {deploy_flag}")
+
+    with open((Path(evaluation_output) / "deploy_flag"), 'w') as outfile:
+        outfile.write(f"{int(deploy_flag)}")
+
+    # add current model score and predictions
+    scores["current model"] = score
+    predictions["currrent model"] = yhat_test
+
+    perf_comparison_plot = pd.DataFrame(
+        scores, index=["r2 score"]).plot(kind='bar', figsize=(15, 10))
+    perf_comparison_plot.figure.savefig("perf_comparison.png")
+    perf_comparison_plot.figure.savefig(Path(evaluation_output) / "perf_comparison.png")
+
+    mlflow.log_metric("deploy flag", bool(deploy_flag))
+    mlflow.log_artifact("perf_comparison.png")
+
+    return predictions, deploy_flag
+
+if __name__ == "__main__":
+
+    mlflow.start_run()
+
+    args = parse_args()
+
+    lines = [
+        f"Model name: {args.model_name}",
+        f"Model path: {args.model_input}",
+        f"Test data path: {args.test_data}",
+        f"Evaluation output path: {args.evaluation_output}",
+    ]
+
+    for line in lines:
+        print(line)
+    
+    main(args)
+
+    mlflow.end_run()
--- a/classical/python-sdk-v2/data-science/src/prep/prep.py
+++ b/classical/python-sdk-v2/data-science/src/prep/prep.py
@ -0,0 +1,113 @@
+# Copyright (c) Microsoft Corporation. All rights reserved.
+# Licensed under the MIT License.
+"""
+Prepares raw data and provides training, validation and test datasets
+"""
+
+import argparse
+
+from pathlib import Path
+import os
+import numpy as np
+import pandas as pd
+
+import mlflow
+
+TARGET_COL = "cost"
+
+NUMERIC_COLS = [
+    "distance", "dropoff_latitude", "dropoff_longitude", "passengers", "pickup_latitude",
+    "pickup_longitude", "pickup_weekday", "pickup_month", "pickup_monthday", "pickup_hour",
+    "pickup_minute", "pickup_second", "dropoff_weekday", "dropoff_month", "dropoff_monthday",
+    "dropoff_hour", "dropoff_minute", "dropoff_second"
+]
+
+CAT_NOM_COLS = [
+    "store_forward", "vendor"
+]
+
+CAT_ORD_COLS = [
+]
+
+def parse_args():
+    '''Parse input arguments'''
+
+    parser = argparse.ArgumentParser("prep")
+    parser.add_argument("--raw_data", type=str, help="Path to raw data")
+    parser.add_argument("--train_data", type=str, help="Path to train dataset")
+    parser.add_argument("--val_data", type=str, help="Path to test dataset")
+    parser.add_argument("--test_data", type=str, help="Path to test dataset")
+    
+    parser.add_argument("--enable_monitoring", type=str, help="enable logging to ADX")
+    parser.add_argument("--table_name", type=str, default="mlmonitoring", help="Table name in ADX for logging")
+    
+    args = parser.parse_args()
+
+    return args
+
+def log_training_data(df, table_name):
+    from obs.collector import Online_Collector
+    collector = Online_Collector(table_name)
+    collector.batch_collect(df)
+
+def main(args):
+    '''Read, split, and save datasets'''
+
+    # ------------ Reading Data ------------ #
+    # -------------------------------------- #
+
+    data = pd.read_csv((Path(args.raw_data)))
+    data = data[NUMERIC_COLS + CAT_NOM_COLS + CAT_ORD_COLS + [TARGET_COL]]
+
+    # ------------- Split Data ------------- #
+    # -------------------------------------- #
+
+    # Split data into train, val and test datasets
+
+    random_data = np.random.rand(len(data))
+
+    msk_train = random_data < 0.7
+    msk_val = (random_data >= 0.7) & (random_data < 0.85)
+    msk_test = random_data >= 0.85
+
+    train = data[msk_train]
+    val = data[msk_val]
+    test = data[msk_test]
+
+    mlflow.log_metric('train size', train.shape[0])
+    mlflow.log_metric('val size', val.shape[0])
+    mlflow.log_metric('test size', test.shape[0])
+
+    train.to_parquet((Path(args.train_data) / "train.parquet"))
+    val.to_parquet((Path(args.val_data) / "val.parquet"))
+    test.to_parquet((Path(args.test_data) / "test.parquet"))
+
+    if (args.enable_monitoring.lower == 'true' or args.enable_monitoring == '1' or args.enable_monitoring.lower == 'yes'):
+        log_training_data(data, args.table_name)
+
+
+if __name__ == "__main__":
+
+    mlflow.start_run()
+
+    # ---------- Parse Arguments ----------- #
+    # -------------------------------------- #
+
+    args = parse_args()
+
+    lines = [
+        f"Raw data path: {args.raw_data}",
+        f"Train dataset output path: {args.train_data}",
+        f"Val dataset output path: {args.val_data}",
+        f"Test dataset path: {args.test_data}",
+
+    ]
+
+    for line in lines:
+        print(line)
+    
+    main(args)
+
+    mlflow.end_run()
+
+    
--- a/classical/python-sdk-v2/data-science/src/register/register.py
+++ b/classical/python-sdk-v2/data-science/src/register/register.py
@ -0,0 +1,85 @@
+# Copyright (c) Microsoft Corporation. All rights reserved.
+# Licensed under the MIT License.
+"""
+Registers trained ML model if deploy flag is True.
+"""
+
+import argparse
+from pathlib import Path
+import pickle
+import mlflow
+
+import os 
+import json
+
+def parse_args():
+    '''Parse input arguments'''
+
+    parser = argparse.ArgumentParser()
+    parser.add_argument('--model_name', type=str, help='Name under which model will be registered')
+    parser.add_argument('--model_path', type=str, help='Model directory')
+    parser.add_argument('--evaluation_output', type=str, help='Path of eval results')
+    parser.add_argument(
+        "--model_info_output_path", type=str, help="Path to write model info JSON"
+    )
+    args, _ = parser.parse_known_args()
+    print(f'Arguments: {args}')
+
+    return args
+
+
+def main(args):
+    '''Loads model, registers it if deply flag is True'''
+
+    with open((Path(args.evaluation_output) / "deploy_flag"), 'rb') as infile:
+        deploy_flag = int(infile.read())
+        
+    mlflow.log_metric("deploy flag", int(deploy_flag))
+    deploy_flag=1
+    if deploy_flag==1:
+
+        print("Registering ", args.model_name)
+
+        # load model
+        model =  mlflow.sklearn.load_model(args.model_path) 
+
+        # log model using mlflow
+        mlflow.sklearn.log_model(model, args.model_name)
+
+        # register logged model using mlflow
+        run_id = mlflow.active_run().info.run_id
+        model_uri = f'runs:/{run_id}/{args.model_name}'
+        mlflow_model = mlflow.register_model(model_uri, args.model_name)
+        model_version = mlflow_model.version
+
+        # write model info
+        print("Writing JSON")
+        dict = {"id": "{0}:{1}".format(args.model_name, model_version)}
+        output_path = os.path.join(args.model_info_output_path, "model_info.json")
+        with open(output_path, "w") as of:
+            json.dump(dict, fp=of)
+
+    else:
+        print("Model will not be registered!")
+
+if __name__ == "__main__":
+
+    mlflow.start_run()
+    
+    # ---------- Parse Arguments ----------- #
+    # -------------------------------------- #
+
+    args = parse_args()
+    
+    lines = [
+        f"Model name: {args.model_name}",
+        f"Model path: {args.model_path}",
+        f"Evaluation output path: {args.evaluation_output}",
+    ]
+
+    for line in lines:
+        print(line)
+
+    main(args)
+
+    mlflow.end_run()
--- a/classical/python-sdk-v2/data-science/src/train/train.py
+++ b/classical/python-sdk-v2/data-science/src/train/train.py
@ -0,0 +1,147 @@
+# Copyright (c) Microsoft Corporation. All rights reserved.
+# Licensed under the MIT License.
+"""
+Trains ML model using training dataset. Saves trained model.
+"""
+
+import argparse
+
+from pathlib import Path
+
+import numpy as np
+import pandas as pd
+from matplotlib import pyplot as plt
+
+from sklearn.ensemble import RandomForestRegressor
+from sklearn.metrics import r2_score, mean_absolute_error, mean_squared_error
+
+import mlflow
+import mlflow.sklearn
+
+TARGET_COL = "cost"
+
+NUMERIC_COLS = [
+    "distance", "dropoff_latitude", "dropoff_longitude", "passengers", "pickup_latitude",
+    "pickup_longitude", "pickup_weekday", "pickup_month", "pickup_monthday", "pickup_hour",
+    "pickup_minute", "pickup_second", "dropoff_weekday", "dropoff_month", "dropoff_monthday",
+    "dropoff_hour", "dropoff_minute", "dropoff_second"
+]
+
+CAT_NOM_COLS = [
+    "store_forward", "vendor"
+]
+
+CAT_ORD_COLS = [
+]
+
+
+def parse_args():
+    '''Parse input arguments'''
+
+    parser = argparse.ArgumentParser("train")
+    parser.add_argument("--train_data", type=str, help="Path to train dataset")
+    parser.add_argument("--model_output", type=str, help="Path of output model")
+
+    # classifier specific arguments
+    parser.add_argument('--regressor__n_estimators', type=int, default=500,
+                        help='Number of trees')
+    parser.add_argument('--regressor__bootstrap', type=int, default=1,
+                        help='Method of selecting samples for training each tree')
+    parser.add_argument('--regressor__max_depth', type=int, default=10,
+                        help=' Maximum number of levels in tree')
+    parser.add_argument('--regressor__max_features', type=str, default='auto',
+                        help='Number of features to consider at every split')
+    parser.add_argument('--regressor__min_samples_leaf', type=int, default=4,
+                        help='Minimum number of samples required at each leaf node')
+    parser.add_argument('--regressor__min_samples_split', type=int, default=5,
+                        help='Minimum number of samples required to split a node')
+
+    args = parser.parse_args()
+
+    return args
+
+def main(args):
+    '''Read train dataset, train model, save trained model'''
+
+    # Read train data
+    train_data = pd.read_parquet(Path(args.train_data))
+
+    # Split the data into input(X) and output(y)
+    y_train = train_data[TARGET_COL]
+    X_train = train_data[NUMERIC_COLS + CAT_NOM_COLS + CAT_ORD_COLS]
+
+    # Train a Random Forest Regression Model with the training set
+    model = RandomForestRegressor(n_estimators = args.regressor__n_estimators,
+                                  bootstrap = args.regressor__bootstrap,
+                                  max_depth = args.regressor__max_depth,
+                                  max_features = args.regressor__max_features,
+                                  min_samples_leaf = args.regressor__min_samples_leaf,
+                                  min_samples_split = args.regressor__min_samples_split,
+                                  random_state=0)
+
+    # log model hyperparameters
+    mlflow.log_param("model", "RandomForestRegressor")
+    mlflow.log_param("n_estimators", args.regressor__n_estimators)
+    mlflow.log_param("bootstrap", args.regressor__bootstrap)
+    mlflow.log_param("max_depth", args.regressor__max_depth)
+    mlflow.log_param("max_features", args.regressor__max_features)
+    mlflow.log_param("min_samples_leaf", args.regressor__min_samples_leaf)
+    mlflow.log_param("min_samples_split", args.regressor__min_samples_split)
+
+    # Train model with the train set
+    model.fit(X_train, y_train)
+
+    # Predict using the Regression Model
+    yhat_train = model.predict(X_train)
+
+    # Evaluate Regression performance with the train set
+    r2 = r2_score(y_train, yhat_train)
+    mse = mean_squared_error(y_train, yhat_train)
+    rmse = np.sqrt(mse)
+    mae = mean_absolute_error(y_train, yhat_train)
+    
+    # log model performance metrics
+    mlflow.log_metric("train r2", r2)
+    mlflow.log_metric("train mse", mse)
+    mlflow.log_metric("train rmse", rmse)
+    mlflow.log_metric("train mae", mae)
+
+    # Visualize results
+    plt.scatter(y_train, yhat_train,  color='black')
+    plt.plot(y_train, y_train, color='blue', linewidth=3)
+    plt.xlabel("Real value")
+    plt.ylabel("Predicted value")
+    plt.savefig("regression_results.png")
+    mlflow.log_artifact("regression_results.png")
+
+    # Save the model
+    mlflow.sklearn.save_model(sk_model=model, path=args.model_output)
+
+
+if __name__ == "__main__":
+    
+    mlflow.start_run()
+
+    # ---------- Parse Arguments ----------- #
+    # -------------------------------------- #
+
+    args = parse_args()
+
+    lines = [
+        f"Train dataset input path: {args.train_data}",
+        f"Model output path: {args.model_output}",
+        f"n_estimators: {args.regressor__n_estimators}",
+        f"bootstrap: {args.regressor__bootstrap}",
+        f"max_depth: {args.regressor__max_depth}",
+        f"max_features: {args.regressor__max_features}",
+        f"min_samples_leaf: {args.regressor__min_samples_leaf}",
+        f"min_samples_split: {args.regressor__min_samples_split}"
+    ]
+
+    for line in lines:
+        print(line)
+
+    main(args)
+
+    mlflow.end_run()
+    
--- a/classical/python-sdk-v2/data/taxi-batch.csv
+++ b/classical/python-sdk-v2/data/taxi-batch.csv
--- a/classical/python-sdk-v2/data/taxi-data.csv
+++ b/classical/python-sdk-v2/data/taxi-data.csv
--- a/classical/python-sdk-v2/data/taxi-request.json
+++ b/classical/python-sdk-v2/data/taxi-request.json
@ -0,0 +1,3 @@
+{"input_data": [[2.86,40.66551971,-73.98258972,1,40.69801331,-73.97357178,0,2,1,1,19,21,3,56,1,1,19,21,21,57],
+                [3.98,40.68072128,-73.931633,1,40.6909523,-73.99185181,0,2,0,1,4,21,44,11,0,1,4,21,59,35]]
+}
--- a/classical/python-sdk-v2/mlops/azureml/train/run_pipeline.py
+++ b/classical/python-sdk-v2/mlops/azureml/train/run_pipeline.py
@ -0,0 +1,195 @@
+# Copyright (c) Microsoft Corporation. All rights reserved.
+# Licensed under the MIT License.
+
+import argparse
+
+from azure.ai.ml.entities import Data
+from azure.ai.ml.constants import AssetTypes
+
+from azure.identity import DefaultAzureCredential
+from azure.ai.ml import MLClient
+
+from azure.ai.ml.entities import AmlCompute
+from azure.ai.ml.entities import Data
+from azure.ai.ml.constants import AssetTypes
+from azure.ai.ml.entities import Environment
+from azure.ai.ml.dsl import pipeline
+from azure.ai.ml import Input, Output, command
+from azure.ai.ml.constants import AssetTypes, InputOutputModes
+
+import json
+import yaml
+import os
+
+
+def parse_args():
+    parser = argparse.ArgumentParser("Deploy Training Pipeline")
+    parser.add_argument("--experiment_name", type=str, help="Experiment Name")
+    parser.add_argument("--compute_name", type=str, help="Compute Cluster Name")
+    parser.add_argument("--data_name", type=str, help="Data Asset Name")
+    parser.add_argument("--environment_name", type=str, help="Registered Environment Name")
+    parser.add_argument("--enable_monitoring", type=str, help="Enable Monitoring", default="false")
+    parser.add_argument("--table_name", type=str, help="ADX Monitoring Table Name", default="taximonitoring")
+    
+    args = parser.parse_args()
+
+    return parser.parse_args()
+
+
+def main():
+    args = parse_args()
+    print(args)
+    
+    credential = DefaultAzureCredential()
+    try:
+        ml_client = MLClient.from_config(credential, path='config.json')
+
+    except Exception as ex:
+        print("HERE IN THE EXCEPTION BLOCK")
+        print(ex)
+
+    try:
+        print(ml_client.compute.get(args.compute_name))
+    except:
+        print("No compute found")
+
+
+
+    print(os.getcwd())
+    print('current', os.listdir())
+
+    # Create pipeline job
+
+    # 1. Define components
+    parent_dir = "data-science/src"
+    
+    prep_data = command( 
+        name="prep_data",
+        display_name="prep-data",
+        code=os.path.join(parent_dir, "prep"),
+        command="python prep.py \
+                --raw_data ${{inputs.raw_data}} \
+                --train_data ${{outputs.train_data}}  \
+                --val_data ${{outputs.val_data}} \
+                --test_data ${{outputs.test_data}} \
+                --enable_monitoring ${{inputs.enable_monitoring}} \
+                --table_name ${{inputs.table_name}}",
+        environment=args.environment_name+"@latest",
+        inputs={
+            "raw_data": Input(type="uri_file"),
+            "enable_monitoring": Input(type="string"),
+            "table_name": Input(type="string")
+            },
+        outputs={
+            "train_data": Output(type="uri_folder"),
+            "val_data": Output(type="uri_folder"),
+            "test_data": Output(type="uri_folder"),
+            }
+    )
+
+    train_model = command( 
+        name="train_model",
+        display_name="train-model",
+        code=os.path.join(parent_dir, "train"),
+        command="python train.py \
+                --train_data ${{inputs.train_data}} \
+                --model_output ${{outputs.model_output}}",
+        environment=args.environment_name+"@latest",
+        inputs={"train_data": Input(type="uri_folder")},
+        outputs={"model_output": Output(type="uri_folder")}
+    )
+
+    evaluate_model = command(
+        name="evaluate_model",
+        display_name="evaluate-model",
+        code=os.path.join(parent_dir, "evaluate"),
+        command="python evaluate.py \
+                --model_name ${{inputs.model_name}} \
+                --model_input ${{inputs.model_input}} \
+                --test_data ${{inputs.test_data}} \
+                --evaluation_output ${{outputs.evaluation_output}}",
+        environment=args.environment_name+"@latest",
+        inputs={
+            "model_name": Input(type="string"),
+            "model_input": Input(type="uri_folder"),
+            "test_data": Input(type="uri_folder")
+            },
+        outputs={
+            "evaluation_output": Output(type="uri_folder")
+            }
+    )
+
+    register_model = command(
+        name="register_model",
+        display_name="register-model",
+        code=os.path.join(parent_dir, "register"),
+        command="python register.py \
+                --model_name ${{inputs.model_name}} \
+                --model_path ${{inputs.model_path}} \
+                --evaluation_output ${{inputs.evaluation_output}} \
+                --model_info_output_path ${{outputs.model_info_output_path}}",
+        environment=args.environment_name+"@latest",
+        inputs={
+            "model_name": Input(type="string"),
+            "model_path": Input(type="uri_folder"),
+            "evaluation_output": Input(type="uri_folder")
+            },
+        outputs={
+            "model_info_output_path": Output(type="uri_folder")
+            }
+    )
+
+    # 2. Construct pipeline
+    @pipeline()
+    def taxi_training_pipeline(raw_data, enable_monitoring, table_name):
+        
+        prep = prep_data(
+            raw_data=raw_data,
+            enable_monitoring=enable_monitoring, 
+            table_name=table_name
+        )
+
+        train = train_model(
+            train_data=prep.outputs.train_data
+        )
+
+        evaluate = evaluate_model(
+            model_name="taxi-model",
+            model_input=train.outputs.model_output,
+            test_data=prep.outputs.test_data
+        )
+
+
+        register = register_model(
+            model_name="taxi-model",
+            model_path=train.outputs.model_output,
+            evaluation_output=evaluate.outputs.evaluation_output
+        )
+
+        return {
+            "pipeline_job_train_data": prep.outputs.train_data,
+            "pipeline_job_test_data": prep.outputs.test_data,
+            "pipeline_job_trained_model": train.outputs.model_output,
+            "pipeline_job_score_report": evaluate.outputs.evaluation_output,
+        }
+
+
+    pipeline_job = taxi_training_pipeline(
+        Input(path=args.data_name + "@latest", type="uri_file"), args.enable_monitoring, args.table_name
+    )
+
+    # set pipeline level compute
+    pipeline_job.settings.default_compute = args.compute_name
+    # set pipeline level datastore
+    pipeline_job.settings.default_datastore = "workspaceblobstore"
+
+    pipeline_job = ml_client.jobs.create_or_update(
+        pipeline_job, experiment_name=args.experiment_name
+    )
+
+    pipeline_job
+    ml_client.jobs.stream(pipeline_job.name)
+
+    
+if __name__ == "__main__":
+    main()
--- a/classical/python-sdk-v2/mlops/devops-pipelines/deploy-batch-endpoint-pipeline.yml
+++ b/classical/python-sdk-v2/mlops/devops-pipelines/deploy-batch-endpoint-pipeline.yml
@ -0,0 +1,66 @@
+# Copyright (c) Microsoft Corporation. All rights reserved.
+# Licensed under the MIT License.
+
+variables:
+  - ${{ if eq(variables['Build.SourceBranchName'], 'main') }}:
+      # 'main' branch: PRD environment
+      - template: ../../config-infra-prod.yml
+  - ${{ if ne(variables['Build.SourceBranchName'], 'main') }}:
+      # 'develop' or feature branches: DEV environment
+      - template: ../../config-infra-dev.yml
+  - name: version
+    value: python-sdk-v2
+  - name: endpoint_name
+    value: taxi-batch-$(namespace)$(postfix)$(environment)
+
+trigger:
+  - none
+
+pool:
+  vmImage: ubuntu-20.04
+
+resources:
+  repositories:
+    - repository: mlops-templates # Template Repo
+      name: Azure/mlops-templates # need to change org name from "Azure" to your own org
+      endpoint: github-connection # need to set up and hardcode
+      type: github
+      ref: main
+
+stages:
+  - stage: CreateBatchEndpoint
+    displayName: Create/Update Batch Endpoint
+    jobs:
+      - job: DeployBatchEndpoint
+        steps:
+          - checkout: self
+            path: s/
+          - checkout: mlops-templates
+            path: s/templates/
+          - template: templates/aml-cli-v2/install-az-cli.yml@mlops-templates
+          - template: templates/aml-cli-v2/install-aml-cli.yml@mlops-templates
+          - template: templates/aml-cli-v2/connect-to-workspace.yml@mlops-templates
+          - template: templates/aml-cli-v2/create-compute.yml@mlops-templates
+            parameters:
+              cluster_name: batch-cluster # name must match cluster name in deployment file below
+              size: STANDARD_DS3_V2
+              min_instances: 0
+              max_instances: 5
+              cluster_tier: dedicated
+          - template: templates/${{ variables.version }}/create-batch-endpoint.yml@mlops-templates
+            parameters:
+              endpoint_name: "${{ variables.endpoint_name }}"
+              endpoint_description: "Taxi batch endpoint"
+              auth_mode: "aad_token"
+          - template: templates/${{ variables.version }}/create-batch-deployment.yml@mlops-templates
+            parameters:
+              deployment_name: taxi-batch-dp
+              deployment_description: "Taxi batch deployment"
+              endpoint_name: "${{ variables.endpoint_name }}"
+              model_path: "taxi-model@latest"
+              compute: batch-cluster
+          - template: templates/${{ variables.version }}/test-batch-endpoint.yml@mlops-templates
+            parameters:
+              endpoint_name: "${{ variables.endpoint_name }}"
+              sample_request: data/taxi-batch.csv
+              request_type: uri_file #either uri_folder or uri_file
--- a/classical/python-sdk-v2/mlops/devops-pipelines/deploy-model-training-pipeline.yml
+++ b/classical/python-sdk-v2/mlops/devops-pipelines/deploy-model-training-pipeline.yml
@ -0,0 +1,70 @@
+# Copyright (c) Microsoft Corporation. All rights reserved.
+# Licensed under the MIT License.
+
+variables:
+  - ${{ if eq(variables['Build.SourceBranchName'], 'main') }}:
+      # 'main' branch: PRD environment
+      - template: ../../config-infra-prod.yml
+  - ${{ if ne(variables['Build.SourceBranchName'], 'main') }}:
+      # 'develop' or feature branches: DEV environment
+      - template: ../../config-infra-dev.yml
+  - name: version
+    value: python-sdk-v2
+
+trigger:
+  - none
+
+pool:
+  vmImage: ubuntu-20.04
+
+resources:
+  repositories:
+    - repository: mlops-templates # Template Repo
+      name: Azure/mlops-templates # need to change org name from "Azure" to your own org
+      endpoint: github-connection # need to set up and hardcode
+      type: github
+      ref: main
+
+stages:
+  - stage: DeployTrainingPipeline
+    displayName: Deploy Training Pipeline
+    jobs:
+      - job: DeployTrainingPipeline
+        timeoutInMinutes: 120 # how long to run the job before automatically cancelling
+        steps:
+          - checkout: self
+            path: s/
+          - checkout: mlops-templates
+            path: s/templates/
+          - template: templates/aml-cli-v2/install-az-cli.yml@mlops-templates
+          - template: templates/aml-cli-v2/install-aml-cli.yml@mlops-templates
+          - template: templates/python-sdk-v2/install-requirements.yml@mlops-templates
+          - template: templates/aml-cli-v2/connect-to-workspace.yml@mlops-templates
+          - template: templates/aml-cli-v2/create-compute.yml@mlops-templates
+            parameters:
+              cluster_name: cpu-cluster
+              size: Standard_DS3_v2
+              min_instances: 0
+              max_instances: 4
+              cluster_tier: low_priority
+          - template: templates/${{ variables.version }}/register-environment.yml@mlops-templates
+            parameters:
+              environment_name: taxi-train-env
+              environment_description: "Training Environment for Taxi Pipeline"
+              environment_path: data-science/environment/train-conda.yml
+              build_type: conda
+          - template: templates/${{ variables.version }}/register-data-asset.yml@mlops-templates
+            parameters:
+              data_name: taxi-data
+              data_description: taxi-training-dataset
+              data_path: data/taxi-data.csv
+              data_type: uri_file
+          - template: templates/${{ variables.version }}/run-pipeline.yml@mlops-templates
+            parameters:
+              pipeline_path: mlops/azureml/train/run_pipeline.py
+              experiment_name: taxi-train-pipeline
+              data_name: taxi-data
+              environment_name: taxi-train-env
+              compute_name: cpu-cluster
+              enable_monitoring: $(enable_monitoring)
+              table_name: "taximonitoring"
--- a/classical/python-sdk-v2/mlops/devops-pipelines/deploy-online-endpoint-pipeline.yml
+++ b/classical/python-sdk-v2/mlops/devops-pipelines/deploy-online-endpoint-pipeline.yml
@ -0,0 +1,57 @@
+# Copyright (c) Microsoft Corporation. All rights reserved.
+# Licensed under the MIT License.
+
+variables:
+  - ${{ if eq(variables['Build.SourceBranchName'], 'main') }}:
+      # 'main' branch: PRD environment
+      - template: ../../config-infra-prod.yml
+  - ${{ if ne(variables['Build.SourceBranchName'], 'main') }}:
+      # 'develop' or feature branches: DEV environment
+      - template: ../../config-infra-dev.yml
+  - name: version
+    value: python-sdk-v2
+  - name: endpoint_name
+    value: taxi-online-$(namespace)$(postfix)$(environment)
+
+trigger:
+  - none
+
+pool:
+  vmImage: ubuntu-20.04
+
+resources:
+  repositories:
+    - repository: mlops-templates # Template Repo
+      name: Azure/mlops-templates # need to change org name from "Azure" to your own org
+      endpoint: github-connection # need to set up and hardcode
+      type: github
+      ref: main
+
+stages:
+  - stage: CreateOnlineEndpoint
+    displayName: Create/Update Online Endpoint
+    jobs:
+      - job: DeployOnlineEndpoint
+        steps:
+          - checkout: self
+            path: s/
+          - checkout: mlops-templates
+            path: s/templates/
+          - template: templates/aml-cli-v2/install-az-cli.yml@mlops-templates
+          - template: templates/aml-cli-v2/install-aml-cli.yml@mlops-templates
+          - template: templates/aml-cli-v2/connect-to-workspace.yml@mlops-templates
+          - template: templates/${{ variables.version }}/create-online-endpoint.yml@mlops-templates
+            parameters:
+              endpoint_name: "${{ variables.endpoint_name }}"
+              endpoint_description: "Taxi Online Endpoint"
+              auth_mode: "aml_token"
+          - template: templates/${{ variables.version }}/create-online-deployment.yml@mlops-templates
+            parameters:
+              deployment_name: taxi-online-dp
+              endpoint_name: "${{ variables.endpoint_name }}"
+              model_path: "taxi-model@latest"
+              traffic_allocation: 100
+          - template: templates/${{ variables.version }}/test-online-endpoint.yml@mlops-templates
+            parameters:
+              endpoint_name: "${{ variables.endpoint_name }}"
+              sample_request: data/taxi-request.json
--- a/classical/rai-aml-cli-v2/data-science/environment/train-monitor-conda.yml
+++ b/classical/rai-aml-cli-v2/data-science/environment/train-monitor-conda.yml
@ -1,16 +0,0 @@
-channels:
-  - defaults
-  - anaconda
-  - conda-forge
-dependencies:
-  - python=3.7.5
-  - pip
-  - pip:
-      - azureml-mlflow==1.38.0
-      - azureml-sdk==1.38.0
-      - scikit-learn==0.24.1
-      - pandas==1.2.1
-      - joblib==1.0.0
-      - matplotlib==3.3.3
-      - git+https://github.com/microsoft/AzureML-Observability#subdirectory=aml-obs-client
-      - git+https://github.com/microsoft/AzureML-Observability#subdirectory=aml-obs-collector
--- a/classical/rai-aml-cli-v2/data-science/environment/train-requirements.txt
+++ b/classical/rai-aml-cli-v2/data-science/environment/train-requirements.txt
@ -1,6 +1,6 @@
 azureml-mlflow==1.38.0
-azureml-sdk==1.38.0
+pyarrow==10.0.0
 scikit-learn==0.24.1
 pandas==1.2.1
 joblib==1.2.0
-matplotlib==3.3.3
+matplotlib==3.3.3
--- a/classical/rai-aml-cli-v2/data-science/src/prep/prep.py
+++ b/classical/rai-aml-cli-v2/data-science/src/prep/prep.py
@ -71,11 +71,7 @@ def main(args):
    # ------------ Reading Data ------------ #
    # -------------------------------------- #

-    print("mounted_path files: ")
-    arr = os.listdir(args.raw_data)
-    print(arr)
-
-    data = pd.read_csv((Path(args.raw_data) / 'taxi-data.csv'))
+    data = pd.read_csv((Path(args.raw_data)))
    data = data[NUMERIC_COLS + CAT_NOM_COLS + CAT_ORD_COLS + [TARGET_COL]]

    # ------------- Split Data ------------- #
@ -101,7 +97,7 @@ def main(args):
    val.to_parquet((Path(args.val_data) / "val.parquet"))
    test.to_parquet((Path(args.test_data) / "test.parquet"))

-    if (args.enable_monitoring.lower == 'true' or args.enable_monitoring == '1' or args.enable_monitoring.lower == 'yes'):
+    if (args.enable_monitoring.lower() == 'true' or args.enable_monitoring == '1' or args.enable_monitoring.lower() == 'yes'):
        log_training_data(data, args.table_name)


@ -129,4 +125,4 @@ if __name__ == "__main__":

    mlflow.end_run()

-    
+    
--- a/classical/rai-aml-cli-v2/data-science/src/prep/test_prep.py
+++ b/classical/rai-aml-cli-v2/data-science/src/prep/test_prep.py
@ -82,7 +82,9 @@ def test_prep_data():
    df = pd.DataFrame(data)
    df.to_csv(os.path.join(raw_data, "taxi-data.csv"))
    
+    raw_data= os.path.join(raw_data, "taxi-data.csv")
    cmd = f"python data-science/src/prep/prep.py --raw_data={raw_data} --train_data={train_data} --val_data={val_data} --test_data={test_data}"
+
    p = subprocess.Popen(cmd, stdout=subprocess.PIPE, shell=True)
    out, err = p.communicate() 
    result = str(out).split('\\n')
--- a/classical/rai-aml-cli-v2/mlops/devops-pipelines/deploy-model-training-pipeline.yml
+++ b/classical/rai-aml-cli-v2/mlops/devops-pipelines/deploy-model-training-pipeline.yml
@ -2,63 +2,75 @@
 # Licensed under the MIT License.

 variables:
- ${{ if eq(variables['Build.SourceBranchName'], 'main') }}:
-    # 'main' branch: PRD environment
-    - template: ../../config-infra-prod.yml
- ${{ if ne(variables['Build.SourceBranchName'], 'main') }}:  
-    # 'develop' or feature branches: DEV environment
-    - template: ../../config-infra-dev.yml
- name: version
-  value: aml-cli-v2
-
+  - ${{ if eq(variables['Build.SourceBranchName'], 'main') }}:
+      # 'main' branch: PRD environment
+      - template: ../../config-infra-prod.yml
+  - ${{ if ne(variables['Build.SourceBranchName'], 'main') }}:
+      # 'develop' or feature branches: DEV environment
+      - template: ../../config-infra-dev.yml
+  - name: version
+    value: aml-cli-v2

 trigger:
- none
+  - none

 pool:
  vmImage: ubuntu-20.04

-
 resources:
  repositories:
-    - repository: mlops-templates  # Template Repo
+    - repository: mlops-templates # Template Repo
      name: Azure/mlops-templates # need to change org name from "Azure" to your own org
      endpoint: github-connection # need to set up and hardcode
      type: github
      ref: main
-    - repository: rai-vnext-preview  # Template Repo
+    - repository: rai-vnext-preview # Template Repo
      name: Azure/rai-vnext-preview # need to change org name from "Azure" to your own org
      endpoint: github-connection # need to set up and hardcode
      type: github
      ref: main

 stages:
- stage: DeployTrainingPipeline
-  displayName: Deploy Training Pipeline
-  jobs:
-    - job: DeployTrainingPipeline
-      timeoutInMinutes: 120 # how long to run the job before automatically cancelling
-      steps:
-      - checkout: self
-        path: s/
-      - checkout: mlops-templates
-        path: s/templates/
-      - checkout: rai-vnext-preview
-        path: s/
-      - template: templates/tests/unit-tests.yml@mlops-templates
-      - template: templates/${{ variables.version }}/install-az-cli.yml@mlops-templates
-      - template: templates/${{ variables.version }}/install-aml-cli.yml@mlops-templates
-      - template: templates/${{ variables.version }}/connect-to-workspace.yml@mlops-templates
-      - template: templates/${{ variables.version }}/register-environment.yml@mlops-templates
-        parameters:
-          build_type: conda
-          environment_name: taxi-train-env
-          environment_file: mlops/azureml/train/train-env.yml
-          enable_monitoring: $(enable_monitoring)
-      - template: register-rai-components.yml
-      - template: templates/${{ variables.version }}/run-pipeline.yml@mlops-templates
-        parameters: 
-          pipeline_file: mlops/azureml/train/pipeline.yml
-          experiment_name: $(environment)_taxi_fare_train_$(Build.SourceBranchName)
-          display_name: $(environment)_taxi_fare_run_$(Build.BuildID)
-          enable_monitoring: $(enable_monitoring)
+  - stage: DeployTrainingPipeline
+    displayName: Deploy Training Pipeline
+    jobs:
+      - job: DeployTrainingPipeline
+        timeoutInMinutes: 120 # how long to run the job before automatically cancelling
+        steps:
+          - checkout: self
+            path: s/
+          - checkout: mlops-templates
+            path: s/templates/
+          - template: templates/tests/unit-tests.yml@mlops-templates
+          - template: templates/${{ variables.version }}/install-az-cli.yml@mlops-templates
+          - template: templates/${{ variables.version }}/install-aml-cli.yml@mlops-templates
+          - template: templates/${{ variables.version }}/connect-to-workspace.yml@mlops-templates
+          - template: templates/${{ variables.version }}/register-environment.yml@mlops-templates
+            parameters:
+              environment_name: taxi-train-env
+              environment_file: mlops/azureml/train/train-env.yml
+          - template: templates/${{ variables.version }}/create-compute.yml@mlops-templates
+            parameters:
+              cluster_name: cpu-cluster
+              size: Standard_DS3_v2
+              min_instances: 0
+              max_instances: 4
+              cluster_tier: low_priority
+          - template: templates/${{ variables.version }}/register-data.yml@mlops-templates
+            parameters:
+              data_type: uri_file
+              data_name: taxi-data
+              data_file: mlops/azureml/train/data.yml
+          - checkout: rai-vnext-preview
+            path: s/
+          - template: register-rai-components.yml
+          - checkout: self
+            path: s/
+          - checkout: mlops-templates
+            path: s/templates/
+          - template: templates/${{ variables.version }}/run-pipeline.yml@mlops-templates
+            parameters:
+              pipeline_file: mlops/azureml/train/pipeline.yml
+              experiment_name: $(environment)_taxi_fare_train_$(Build.SourceBranchName)
+              display_name: $(environment)_taxi_fare_run_$(Build.BuildID)
+              enable_monitoring: $(enable_monitoring)
--- a/config-infra-dev.yml
+++ b/config-infra-dev.yml
@ -1,15 +1,14 @@
 # Copyright (c) Microsoft Corporation. All rights reserved.
 # Licensed under the MIT License.

+# Dev environment
 variables:
-
  # Global
  ap_vm_image: ubuntu-20.04

  namespace: mlopsv2 #Note: A namespace with many characters will cause storage account creation to fail due to storage account names having a limit of 24 characters.
-  postfix: 0659
-  location: westus
-
+  postfix: 0001
+  location: eastus
  environment: dev
  enable_aml_computecluster: true
  enable_monitoring: false
@ -29,8 +28,9 @@ variables:
  storage_account: st$(namespace)$(postfix)$(environment)

  # For terraform reference
-  terraform_version: 0.14.7
+  terraform_version: 1.3.6
  terraform_workingdir: infrastructure
+  terraform_st_location: $(location)
  terraform_st_resource_group: rg-$(namespace)-$(postfix)$(environment)-tf
  terraform_st_storage_account: st$(namespace)$(postfix)$(environment)tf
  terraform_st_container_name: default
--- a/config-infra-prod.yml
+++ b/config-infra-prod.yml
@ -3,18 +3,16 @@

 # Prod environment
 variables:
-
  # Global
  ap_vm_image: ubuntu-20.04

  namespace: mlopsv2 #Note: A namespace with many characters will cause storage account creation to fail due to storage account names having a limit of 24 characters.
-  postfix: 0518
-  location: westeurope
+  postfix: 0001
+  location: eastus
  environment: prod
  enable_aml_computecluster: true
  enable_monitoring: false

-
  # Azure DevOps
  ado_service_connection_rg: Azure-ARM-Prod
  ado_service_connection_aml_ws: Azure-ARM-Prod
@ -30,8 +28,9 @@ variables:
  storage_account: st$(namespace)$(postfix)$(environment)

  # For terraform reference
-  terraform_version: 0.14.7
+  terraform_version: 1.3.6
  terraform_workingdir: infrastructure
+  terraform_st_location: $(location)
  terraform_st_resource_group: rg-$(namespace)-$(postfix)$(environment)-tf
  terraform_st_storage_account: st$(namespace)$(postfix)$(environment)tf
  terraform_st_container_name: default
--- a/cv/aml-cli-v2/mlops/github-actions/deploy-model-training-pipeline.yml
+++ b/cv/aml-cli-v2/mlops/github-actions/deploy-model-training-pipeline.yml
@ -1,51 +1,90 @@
 name: deploy-cv-model-training-pipeline

-on: 
+on:
  workflow_dispatch:
 jobs:
+  set-env-branch:
+    runs-on: ubuntu-latest
+    outputs:
+      config-file: ${{ steps.set-output-defaults.outputs.config-file }}
+    steps:
+      - id: set-prod-branch
+        name: set-prod-branch
+        if: ${{ github.ref == 'refs/heads/main'}}
+        run: echo "config_env=config-infra-prod.yml" >> $GITHUB_ENV;
+      - id: set-dev-branch
+        name: setdevbranch
+        if: ${{ github.ref != 'refs/heads/main'}}
+        run: echo "config_env=config-infra-dev.yml" >> $GITHUB_ENV;
+      - id: set-output-defaults
+        name: set-output-defaults
+        run: |
+          echo "config-file=$config_env" >> $GITHUB_OUTPUT;
  get-config:
-      uses: Azure/mlops-templates/.github/workflows/read-yaml.yml@main
-      with:
-        file_name: config-infra-prod.yml
-  create-compute:
-      needs: get-config
-      uses: Azure/mlops-templates/.github/workflows/create-compute.yml@main
-      with:
-        cluster_name: gpu-cluster
-        size: Standard_NC6
-        min_instances: 0
-        max_instances: 1
-        resource_group: ${{ needs.get-config.outputs.resource_group }}
-        workspace_name: ${{ needs.get-config.outputs.aml_workspace }}     
-      secrets:
-          creds: ${{secrets.AZURE_CREDENTIALS}}
+    needs: set-env-branch
+    uses: Azure/mlops-templates/.github/workflows/read-yaml.yml@main
+    with:
+      file_name: ${{ needs.set-env-branch.outputs.config-file}}
+  create-dataprep-compute:
+    needs: [get-config]
+    uses: Azure/mlops-templates/.github/workflows/create-compute.yml@main
+    with:
+      cluster_name: cpu-cluster
+      size: Standard_DS3_v2
+      min_instances: 0
+      max_instances: 4
+      cluster_tier: low_priority
+      resource_group: ${{ needs.get-config.outputs.resource_group }}
+      workspace_name: ${{ needs.get-config.outputs.aml_workspace }}
+    secrets:
+      creds: ${{secrets.AZURE_CREDENTIALS}}
+  create-training-compute:
+    needs: get-config
+    uses: Azure/mlops-templates/.github/workflows/create-compute.yml@main
+    with:
+      cluster_name: gpu-cluster
+      size: Standard_NC6
+      min_instances: 0
+      max_instances: 1
+      cluster_tier: low_priority
+      resource_group: ${{ needs.get-config.outputs.resource_group }}
+      workspace_name: ${{ needs.get-config.outputs.aml_workspace }}
+    secrets:
+      creds: ${{secrets.AZURE_CREDENTIALS}}
  register-environment:
-      needs: [get-config,create-compute]
-      uses: ./.github/workflows/register-environment.yml
-      with:
-        resource_group: ${{ needs.get-config.outputs.resource_group }}
-        workspace_name: ${{ needs.get-config.outputs.aml_workspace }}     
-        environment_file: mlops/azureml/train/train-env.yaml 
-      secrets:
-        creds: ${{secrets.AZURE_CREDENTIALS}}
+    needs: [get-config, create-dataprep-compute, create-training-compute]
+    uses: Azure/mlops-templates/.github/workflows/register-environment.yml@main
+    with:
+      resource_group: ${{ needs.get-config.outputs.resource_group }}
+      workspace_name: ${{ needs.get-config.outputs.aml_workspace }}
+      environment_file: mlops/azureml/train/train-env.yaml
+    secrets:
+      creds: ${{secrets.AZURE_CREDENTIALS}}
  register-dataset:
-      needs: [get-config,register-environment]
-      uses: Azure/mlops-templates/.github/workflows/register-dataset.yml@main
-      with:
-        resource_group: ${{ needs.get-config.outputs.resource_group }}
-        workspace_name: ${{ needs.get-config.outputs.aml_workspace }} 
-        data_file: mlops/azureml/train/create_stanford_dogs_dataset.yaml
-        file_type: Training
-        name: stanford_dogs
-      secrets:
-        creds: ${{secrets.AZURE_CREDENTIALS}}
-  run-pipeline:
-      needs: [get-config,register-dataset]
-      uses: Azure/mlops-templates/.github/workflows/run-pipeline.yml@main
-      with:
-        resource_group: ${{ needs.get-config.outputs.resource_group }}
-        workspace_name: ${{ needs.get-config.outputs.aml_workspace }}     
-        parameters-file:  mlops/azureml/train/pipeline.yaml
-        job-name: cv-train
-      secrets:
-          creds: ${{secrets.AZURE_CREDENTIALS}}
+    needs: [get-config, register-environment]
+    uses: Azure/mlops-templates/.github/workflows/register-dataset.yml@main
+    with:
+      resource_group: ${{ needs.get-config.outputs.resource_group }}
+      workspace_name: ${{ needs.get-config.outputs.aml_workspace }}
+      data_file: mlops/azureml/train/create_stanford_dogs_dataset.yaml
+      file_type: Training
+      name: stanford_dogs
+    secrets:
+      creds: ${{secrets.AZURE_CREDENTIALS}}
+  run-model-training-pipeline:
+    needs:
+      [
+        get-config,
+        create-dataprep-compute,
+        create-training-compute,
+        register-environment,
+        register-dataset,
+      ]
+    uses: Azure/mlops-templates/.github/workflows/run-pipeline.yml@main
+    with:
+      resource_group: ${{ needs.get-config.outputs.resource_group }}
+      workspace_name: ${{ needs.get-config.outputs.aml_workspace }}
+      parameters-file: mlops/azureml/train/pipeline.yaml
+      job-name: cv-train
+    secrets:
+      creds: ${{secrets.AZURE_CREDENTIALS}}
--- a/cv/aml-cli-v2/mlops/github-actions/deploy-online-endpoint-pipeline.yml
+++ b/cv/aml-cli-v2/mlops/github-actions/deploy-online-endpoint-pipeline.yml
@ -1,42 +1,60 @@
 name: deploy-online-endpoint-pipeline

-on: 
+on:
  workflow_dispatch:
 jobs:
+  set-env-branch:
+    runs-on: ubuntu-latest
+    outputs:
+      config-file: ${{ steps.set-output-defaults.outputs.config-file }}
+    steps:
+      - id: set-prod-branch
+        name: set-prod-branch
+        if: ${{ github.ref == 'refs/heads/main'}}
+        run: echo "config_env=config-infra-prod.yml" >> $GITHUB_ENV;
+      - id: set-dev-branch
+        name: setdevbranch
+        if: ${{ github.ref != 'refs/heads/main'}}
+        run: echo "config_env=config-infra-dev.yml" >> $GITHUB_ENV;
+      - id: set-output-defaults
+        name: set-output-defaults
+        run: |
+          echo "config-file=$config_env" >> $GITHUB_OUTPUT;
  get-config:
-      uses: Azure/mlops-templates/.github/workflows/read-yaml.yml@main
-      with:
-        file_name: config-infra-prod.yml
+    needs: set-env-branch
+    uses: Azure/mlops-templates/.github/workflows/read-yaml.yml@main
+    with:
+      file_name: ${{ needs.set-env-branch.outputs.config-file}}
  create-endpoint:
-      needs: get-config
-      uses: Azure/mlops-templates/.github/workflows/create-endpoint.yml@main
-      with:
-        resource_group: ${{ needs.get-config.outputs.resource_group }}
-        workspace_name: ${{ needs.get-config.outputs.aml_workspace }} 
-        endpoint_file: mlops/azureml/deploy/online/online-endpoint.yml
-        endpoint_name: dogs-classifier-online2
-        endpoint_type: online
-      secrets:
-        creds: ${{secrets.AZURE_CREDENTIALS}}
+    needs: get-config
+    uses: Azure/mlops-templates/.github/workflows/create-endpoint.yml@main
+    with:
+      resource_group: ${{ needs.get-config.outputs.resource_group }}
+      workspace_name: ${{ needs.get-config.outputs.aml_workspace }}
+      endpoint_file: mlops/azureml/deploy/online/online-endpoint.yml
+      endpoint_name: dogs-classifier-online2
+      endpoint_type: online
+    secrets:
+      creds: ${{secrets.AZURE_CREDENTIALS}}
  create-deployment:
-      uses: Azure/mlops-templates/.github/workflows/create-deployment.yml@main
-      needs: [get-config,create-endpoint]
-      with:
-        resource_group: ${{ needs.get-config.outputs.resource_group }}
-        workspace_name: ${{ needs.get-config.outputs.aml_workspace }} 
-        endpoint_file: mlops/azureml/deploy/online/online-deployment.yml
-        endpoint_name: dogs-classifier-online2
-        endpoint_type: online
-        deployment_name: dogs-online-dp
-      secrets:
-        creds: ${{secrets.AZURE_CREDENTIALS}}
+    uses: Azure/mlops-templates/.github/workflows/create-deployment.yml@main
+    needs: [get-config, create-endpoint]
+    with:
+      resource_group: ${{ needs.get-config.outputs.resource_group }}
+      workspace_name: ${{ needs.get-config.outputs.aml_workspace }}
+      endpoint_file: mlops/azureml/deploy/online/online-deployment.yml
+      endpoint_name: dogs-classifier-online2
+      endpoint_type: online
+      deployment_name: dogs-online-dp
+    secrets:
+      creds: ${{secrets.AZURE_CREDENTIALS}}
  allocate-traffic:
-      uses: Azure/mlops-templates/.github/workflows/allocate-traffic.yml@main
-      needs: [get-config,create-deployment]
-      with:
-        resource_group: ${{ needs.get-config.outputs.resource_group }}
-        workspace_name: ${{ needs.get-config.outputs.aml_workspace }} 
-        traffic_allocation: dogs-online-dp=100
-        endpoint_name: dogs-classifier-online2
-      secrets:
-        creds: ${{secrets.AZURE_CREDENTIALS}}
+    uses: Azure/mlops-templates/.github/workflows/allocate-traffic.yml@main
+    needs: [get-config, create-deployment]
+    with:
+      resource_group: ${{ needs.get-config.outputs.resource_group }}
+      workspace_name: ${{ needs.get-config.outputs.aml_workspace }}
+      traffic_allocation: dogs-online-dp=100
+      endpoint_name: dogs-classifier-online2
+    secrets:
+      creds: ${{secrets.AZURE_CREDENTIALS}}
--- a/cv/python-sdk-v1/config-aml.yml
+++ b/cv/python-sdk-v1/config-aml.yml
--- a/cv/python-sdk-v1/data-science/environment/training/azureml_environment.json
+++ b/cv/python-sdk-v1/data-science/environment/training/azureml_environment.json
--- a/cv/python-sdk-v1/data-science/environment/training/conda_dependencies.yml
+++ b/cv/python-sdk-v1/data-science/environment/training/conda_dependencies.yml
--- a/cv/python-sdk-v1/data-science/notebooks/.gitkeep
+++ b/cv/python-sdk-v1/data-science/notebooks/.gitkeep
--- a/cv/python-sdk-v1/data-science/src/evaluate.py
+++ b/cv/python-sdk-v1/data-science/src/evaluate.py
--- a/cv/python-sdk-v1/data-science/src/model/init.py
+++ b/cv/python-sdk-v1/data-science/src/model/init.py
--- a/cv/python-sdk-v1/data-science/src/model/dataset.py
+++ b/cv/python-sdk-v1/data-science/src/model/dataset.py
--- a/cv/python-sdk-v1/data-science/src/model/net.py
+++ b/cv/python-sdk-v1/data-science/src/model/net.py
--- a/cv/python-sdk-v1/data-science/src/prep.py
+++ b/cv/python-sdk-v1/data-science/src/prep.py
--- a/cv/python-sdk-v1/data-science/src/train.py
+++ b/cv/python-sdk-v1/data-science/src/train.py
--- a/cv/python-sdk-v1/data-science/tests/.gitkeep
+++ b/cv/python-sdk-v1/data-science/tests/.gitkeep
--- a/cv/python-sdk-v1/data/training/image_labels.csv
+++ b/cv/python-sdk-v1/data/training/image_labels.csv
--- a/cv/python-sdk-v1/mlops/devops-pipelines/deploy-model-training-pipeline.yml
+++ b/cv/python-sdk-v1/mlops/devops-pipelines/deploy-model-training-pipeline.yml
@ -0,0 +1,103 @@
+# Copyright (c) Microsoft Corporation. All rights reserved.
+# Licensed under the MIT License.
+
+variables:
+  - template: ../../config-aml.yml
+  - ${{ if eq(variables['Build.SourceBranchName'], 'main') }}:
+      # 'main' branch: PRD environment
+      - template: ../../config-infra-prod.yml
+  - ${{ if ne(variables['Build.SourceBranchName'], 'main') }}:
+      # 'develop' or feature branches: DEV environment
+      - template: ../../config-infra-dev.yml
+  - name: version
+    value: python-sdk-v1
+
+trigger:
+  - none
+
+pool:
+  vmImage: $(ap_vm_image)
+
+resources:
+  repositories:
+    - repository: mlops-templates # Template Repo
+      name: Azure/mlops-templates # need to change org name from "Azure" to your own org
+      endpoint: github-connection # need to set up and hardcode
+      type: github
+      ref: main
+
+stages:
+  - stage: DeployTrainingPipeline
+    displayName: Deploy Training Pipeline
+    jobs:
+      - job: DeployTrainingPipeline
+        steps:
+          # Setup
+          - checkout: self
+            path: s/
+          - checkout: mlops-templates
+            path: s/templates/
+          - template: templates/${{ variables.version }}/install-az-cli.yml@mlops-templates
+          - template: templates/${{ variables.version }}/install-aml-cli.yml@mlops-templates
+          - template: templates/${{ variables.version }}/connect-to-workspace.yml@mlops-templates
+
+          # Environment
+          - template: templates/${{ variables.version }}/create-environment.yml@mlops-templates
+            parameters:
+              environment_name: $(training_env_name)
+              build_type: folder
+              environment_file: $(training_env_path)
+
+          # Compute
+          - template: templates/${{ variables.version }}/get-compute.yml@mlops-templates
+            parameters:
+              compute_type: training
+
+          # Datasets (images + labels)
+          # Images dataset
+          - task: Bash@3
+            displayName: "Download data"
+            inputs:
+              targetType: inline
+              script: |
+                mkdir -p $(training_dataset_local_path)
+                curl $(training_dataset_storage_url) | tar xvf - --no-same-owner -C $(training_dataset_local_path)
+          - template: templates/${{ variables.version }}/register-dataset.yml@mlops-templates
+            parameters:
+              data_type: training
+          # Labels dataset
+          - template: templates/${{ variables.version }}/register-dataset.yml@mlops-templates
+            parameters:
+              data_type: training
+              datasetName: $(labels_dataset_name)
+              datasetDescription: $(labels_dataset_description)
+              datasetLocalPath: $(labels_dataset_local_path)
+              datasetPathOnDatastore: $(labels_dataset_path_on_datastore)
+              datasetType: $(labels_dataset_type)
+
+          # Deploy training pipeline
+          - template: templates/${{ variables.version }}/deploy-training-pipeline.yml@mlops-templates
+          - template: templates/${{ variables.version }}/add-pipeline-to-endpoint.yml@mlops-templates
+          - task: Bash@3
+            name: export_pipeline_id
+            displayName: "Export Pipeline ID"
+            inputs:
+              targetType: "inline"
+              script: |
+                echo "##vso[task.setvariable variable=pipeline_id;isOutput=true;]$(pipeline_id)"
+
+        # Run training
+      - job: invoke_pipeline
+        displayName: "Invoke pipeline"
+        pool: server
+        timeoutInMinutes: 0
+        dependsOn: DeployTrainingPipeline
+        variables:
+          pipeline_id: $[ dependencies.DeployTrainingPipeline.outputs['export_pipeline_id.pipeline_id'] ]
+        steps:
+          - task: ms-air-aiagility.vss-services-azureml.azureml-restApi-task.MLPublishedPipelineRestAPITask@0
+            displayName: "Invoke AML Pipeline"
+            inputs:
+              azureSubscription: "$(ado_service_connection_aml_ws)"
+              PipelineId: "$(PIPELINE_ID)"
+              ExperimentName: "$(training_experiment_name)"
--- a/cv/python-sdk/mlops/devops-pipelines/deploy-model-training-pipeline.yml
+++ b/cv/python-sdk/mlops/devops-pipelines/deploy-model-training-pipeline.yml
@ -1,105 +0,0 @@
-# Copyright (c) Microsoft Corporation. All rights reserved.
-# Licensed under the MIT License.
-
-variables:
- template: ../../config-aml.yml
- ${{ if eq(variables['Build.SourceBranchName'], 'main') }}:
-    # 'main' branch: PRD environment
-    - template: ../../config-infra-prod.yml
- ${{ if ne(variables['Build.SourceBranchName'], 'main') }}:  
-    # 'develop' or feature branches: DEV environment
-    - template: ../../config-infra-dev.yml
- name: version
-  value: python-sdk
-
-
-trigger:
- none
-
-pool:
-  vmImage: $(ap_vm_image)
-
-
-resources:
-  repositories:
-    - repository: mlops-templates  # Template Repo
-      name: Azure/mlops-templates # need to change org name from "Azure" to your own org
-      endpoint: github-connection # need to set up and hardcode
-      type: github
-
-stages:
- stage: DeployTrainingPipeline
-  displayName: Deploy Training Pipeline
-  jobs:
-    - job: DeployTrainingPipeline
-      steps:
-
-      # Setup
-      - checkout: self
-        path: s/
-      - checkout: mlops-templates
-        path: s/templates/
-      - template: templates/${{ variables.version }}/install-az-cli.yml@mlops-templates
-      - template: templates/${{ variables.version }}/install-aml-cli.yml@mlops-templates
-      - template: templates/${{ variables.version }}/connect-to-workspace.yml@mlops-templates
-
-      # Environment
-      - template: templates/${{ variables.version }}/create-environment.yml@mlops-templates
-        parameters:
-          environment_name: $(training_env_name)
-          build_type: folder
-          environment_file: $(training_env_path)
-
-      # Compute
-      - template: templates/${{ variables.version }}/get-compute.yml@mlops-templates
-        parameters:
-          compute_type: training
-
-      # Datasets (images + labels)
-      # Images dataset
-      - task: Bash@3
-        displayName: 'Download data'
-        inputs:
-          targetType: inline
-          script: |
-            mkdir -p $(training_dataset_local_path)
-            curl $(training_dataset_storage_url) | tar xvf - --no-same-owner -C $(training_dataset_local_path)
-      - template: templates/${{ variables.version }}/register-dataset.yml@mlops-templates
-        parameters:
-          data_type: training
-      # Labels dataset
-      - template: templates/${{ variables.version }}/register-dataset.yml@mlops-templates
-        parameters:
-          data_type: training
-          datasetName: $(labels_dataset_name)
-          datasetDescription: $(labels_dataset_description)
-          datasetLocalPath: $(labels_dataset_local_path)
-          datasetPathOnDatastore: $(labels_dataset_path_on_datastore)
-          datasetType: $(labels_dataset_type)
-
-      # Deploy training pipeline
-      - template: templates/${{ variables.version }}/deploy-training-pipeline.yml@mlops-templates
-      - template: templates/${{ variables.version }}/add-pipeline-to-endpoint.yml@mlops-templates
-      - task: Bash@3
-        name: export_pipeline_id
-        displayName: "Export Pipeline ID"
-        inputs:
-          targetType: "inline"
-          script: |
-            echo "##vso[task.setvariable variable=pipeline_id;isOutput=true;]$(pipeline_id)"
-
-      # Run training
-    - job: invoke_pipeline
-      displayName: 'Invoke pipeline'
-      pool: server
-      timeoutInMinutes: 0
-      dependsOn: DeployTrainingPipeline
-      variables: 
-        pipeline_id: $[ dependencies.DeployTrainingPipeline.outputs['export_pipeline_id.pipeline_id'] ]
-      steps:
-          - task: ms-air-aiagility.vss-services-azureml.azureml-restApi-task.MLPublishedPipelineRestAPITask@0
-            displayName: 'Invoke AML Pipeline'
-            inputs:
-              azureSubscription: '$(ado_service_connection_aml_ws)'
-              PipelineId: '$(PIPELINE_ID)'
-              ExperimentName: '$(training_experiment_name)'
--- a/infrastructure/terraform/devops-pipelines/tf-ado-deploy-infra.yml
+++ b/infrastructure/terraform/devops-pipelines/tf-ado-deploy-infra.yml
--- a/infrastructure/terraform/github-actions/tf-gha-deploy-infra.yml
+++ b/infrastructure/terraform/github-actions/tf-gha-deploy-infra.yml
@ -0,0 +1,63 @@
+name: tf-gha-deploy-infra.yml
+
+on:
+  #push:
+  workflow_dispatch:
+env:
+  config_env: "none"
+jobs:
+  set-env-branch:
+    runs-on: ubuntu-latest
+    outputs:
+      config-file: ${{ steps.set-output-defaults.outputs.config-file }}
+    steps:
+      - id: set-prod-branch
+        name: set-prod-branch
+        if: ${{ github.ref == 'refs/heads/main'}}
+        run: echo "config_env=config-infra-prod.yml" >> $GITHUB_ENV;
+      - id: set-dev-branch
+        name: setdevbranch
+        if: ${{ github.ref != 'refs/heads/main'}}
+        run: echo "config_env=config-infra-dev.yml" >> $GITHUB_ENV;
+      - id: set-output-defaults
+        name: set-output-defaults
+        run: |
+          echo "config-file=$config_env" >> $GITHUB_OUTPUT;
+  get-config:
+    needs: set-env-branch
+    uses: Azure/mlops-templates/.github/workflows/read-yaml.yml@main
+    with:
+      file_name: ${{ needs.set-env-branch.outputs.config-file}}
+  test-terraform-state-deployment:
+    needs: [get-config, set-env-branch]
+    uses: Azure/mlops-templates/.github/workflows/tf-gha-install-terraform.yml@main
+    with:
+      TFAction: "apply"
+      dply_environment: ${{ needs.set-env-branch.outputs.config-file }}
+      location: ${{ needs.get-config.outputs.location }}
+      namespace: ${{ needs.get-config.outputs.namespace }}
+      postfix: ${{ needs.get-config.outputs.postfix }}
+      environment: ${{ needs.get-config.outputs.environment }}
+      enable_aml_computecluster: ${{ needs.get-config.outputs.enable_aml_computecluster == true }} ## TODO review the evaluation of boolean
+      enable_monitoring: ${{ needs.get-config.outputs.enable_monitoring == true  }} ## TODO review the evaluation of boolean
+      terraform_version: ${{ needs.get-config.outputs.terraform_version }}
+      terraform_workingdir: ${{ needs.get-config.outputs.terraform_workingdir }}
+      terraform_st_location: ${{ needs.get-config.outputs.terraform_st_location }}
+      terraform_st_storage_account: ${{ needs.get-config.outputs.terraform_st_storage_account }}
+      terraform_st_resource_group: ${{ needs.get-config.outputs.terraform_st_resource_group }}
+      terraform_st_container_name: ${{ needs.get-config.outputs.terraform_st_container_name }}
+      terraform_st_key: ${{ needs.get-config.outputs.terraform_st_key }}
+      terraform_plan_location: ${{ needs.get-config.outputs.location }}
+      terraform_plan_vnet: "TBD" # TBD
+    secrets:
+      azure_creds: ${{ secrets.AZURE_CREDENTIALS }}
+      clientId: ${{ secrets.ARM_CLIENT_ID }}
+      clientSecret: ${{ secrets.ARM_CLIENT_SECRET }}
+      subscriptionId: ${{ secrets.ARM_SUBSCRIPTION_ID }}
+      tenantId: ${{ secrets.ARM_TENANT_ID }}
+  deploy-azureml-resources:
+    runs-on: ubuntu-latest
+    steps:
+      - id: deploy-aml-workspace
+        name: deploy-aml-workspace
+        run: echo "OK"
--- a/infrastructure/terraform/modules/data-explorer/main.tf
+++ b/infrastructure/terraform/modules/data-explorer/main.tf
@ -56,4 +56,4 @@ resource "azurerm_key_vault_secret" "ADX_DB" {
  value        = azurerm_kusto_database.database[0].name
  key_vault_id = var.key_vault_id
  count               = var.enable_monitoring ? 1 : 0
-}
+}
--- a/infrastructure/terraform/modules/storage-account/main.tf
+++ b/infrastructure/terraform/modules/storage-account/main.tf
@ -19,13 +19,13 @@ resource "azurerm_storage_account" "st" {
  is_hns_enabled           = var.hns_enabled

  tags = var.tags
+  
 }

 # Virtual Network & Firewall configuration

 resource "azurerm_storage_account_network_rules" "firewall_rules" {
-  resource_group_name  = var.rg_name
-  storage_account_name = azurerm_storage_account.st.name
+  storage_account_id = azurerm_storage_account.st.id

  default_action             = "Allow"
  ip_rules                   = [] # [data.http.ip.body]
--- a/nlp/aml-cli-v2/data-science/environments/training/requirements.txt
+++ b/nlp/aml-cli-v2/data-science/environments/training/requirements.txt
@ -16,3 +16,13 @@ psutil==5.9.0

 # for unit testing
 pytest==7.1.2
+
+# for azure ml SDK v2
+azure-ai-ml==1.1.0
+azure-common==1.1.28
+azure-core==1.26.1
+azure-identity==1.10.0
+azure-mgmt-core==1.3.0 
+azure-storage-blob==12.14.1
+azure-storage-file-datalake==12.9.1
+azure-storage-file-share==12.7.0
--- a/nlp/aml-cli-v2/mlops/devops-pipelines/deploy-model-training-pipeline.yml
+++ b/nlp/aml-cli-v2/mlops/devops-pipelines/deploy-model-training-pipeline.yml
@ -63,7 +63,6 @@ stages:
          cluster_tier: dedicated
      - template: templates/${{ variables.version }}/register-environment.yml@mlops-templates
        parameters:
-          build_type: conda
          environment_name: nlp_summarization_train
          environment_file: mlops/azureml/train/train-env.yml
      - template: templates/${{ variables.version }}/run-pipeline.yml@mlops-templates
--- a/nlp/python-sdk-v2/data-science/environments/inference/conda_env.yml
+++ b/nlp/python-sdk-v2/data-science/environments/inference/conda_env.yml
@ -0,0 +1,24 @@
+name: nlp_inference_conda_env
+channels:
+  - pytorch
+  - anaconda
+  - defaults
+  - conda-forge
+dependencies:
+  - python=3.8
+  - pip=21.2.4
+  - pytorch=1.10.0
+  - torchvision=0.11.1
+  - torchaudio=0.10.0
+  - cudatoolkit=11.1.1
+  - nvidia-apex=0.1.0
+  - gxx_linux-64=8.5.0
+  - pip:
+      - azureml-defaults==1.39.0
+      - azureml-mlflow==1.39.0
+      - azureml-telemetry==1.39.0
+      - azureml-train-core==1.39.0
+      - mlflow==1.24.0
+      - transformers==4.17.0
+      - 'inference-schema[numpy-support]==1.3.0'
+      - applicationinsights==0.11.10
--- a/nlp/python-sdk-v2/data-science/environments/training/Dockerfile
+++ b/nlp/python-sdk-v2/data-science/environments/training/Dockerfile
@ -0,0 +1,7 @@
+# check release notes https://docs.nvidia.com/deeplearning/frameworks/pytorch-release-notes/index.html
+FROM nvcr.io/nvidia/pytorch:22.04-py3
+
+# Install dependencies missing in this container
+# NOTE: container already has matplotlib==3.5.1 tqdm==4.62.0
+COPY requirements.txt ./
+RUN pip install -r requirements.txt
--- a/nlp/python-sdk-v2/data-science/environments/training/requirements.txt
+++ b/nlp/python-sdk-v2/data-science/environments/training/requirements.txt
@ -0,0 +1,21 @@
+# data science requirements
+# torchvision==0.12.0
+# torch==1.11.0
+pytorch_lightning==1.6.4
+transformers==4.18.0
+datasets==2.3.2
+rouge_score==0.0.4
+sentencepiece==0.1.96
+
+# for metrics reporting/plotting
+mlflow==1.25.1
+azureml-mlflow==1.41.0
+# matplotlib==3.5.2
+# tqdm==4.64.0
+psutil==5.9.0
+
+# for unit testing
+pytest==7.1.2
+
+# for azure ml SDK v2
+azure-ai-ml==1.1.0
--- a/nlp/python-sdk-v2/data-science/src/summarization/compare.py
+++ b/nlp/python-sdk-v2/data-science/src/summarization/compare.py
@ -0,0 +1,88 @@
+import os
+import argparse
+import logging
+import mlflow
+import json
+from distutils.util import strtobool
+
+def main():
+    """Main function of the script."""
+    # initialize root logger
+    logger = logging.getLogger()
+    logger.setLevel(logging.INFO)
+    console_handler = logging.StreamHandler()
+    formatter = logging.Formatter(
+        "%(asctime)s : %(levelname)s : %(name)s : %(message)s"
+    )
+    console_handler.setFormatter(formatter)
+    logger.addHandler(console_handler)
+
+    # input and output arguments
+    parser = argparse.ArgumentParser()
+    parser.add_argument(
+        "--baseline_metrics",
+        type=str,
+        required=True,
+        help="path to baseline metrics folder containing all_results.json",
+    )
+    parser.add_argument(
+        "--candidate_metrics",
+        type=str,
+        required=True,
+        help="path to candidate metrics folder containing all_results.json",
+    )
+    parser.add_argument(
+        "--reference_metric",
+        type=str,
+        default="predict_rougeLsum",
+        help="name of reference metric for shipping flag (default: predict_rougeLsum)",
+    )
+    parser.add_argument(
+        "--force_comparison", type=strtobool, default=False, help="set to True to bypass comparison and set --deploy_flag to True"
+    )
+    parser.add_argument(
+        "--deploy_flag", type=str, help="a deploy flag whether to deploy or not"
+    )
+
+    args = parser.parse_args()
+
+    # Start Logging
+    mlflow.start_run()
+
+    logger.info(f"Running with arguments: {args}")
+
+    # open metrics on both sides
+    with open(os.path.join(args.baseline_metrics, "all_results.json")) as in_file:
+        baseline_metrics = json.loads(in_file.read())
+    with open(os.path.join(args.candidate_metrics, "all_results.json")) as in_file:
+        candidate_metrics = json.loads(in_file.read())
+
+    # should we ship or not?
+    if args.force_comparison:
+        deploy_flag = True
+    else:
+        deploy_flag = (
+            candidate_metrics[args.reference_metric]
+            > baseline_metrics[args.reference_metric]
+        )
+
+    logger.info("baseline_metrics[{}]={}, candidate_metrics[{}]={}, deploy_flag={} (force_comparison={})".format(
+        args.reference_metric,
+        baseline_metrics[args.reference_metric],
+        args.reference_metric,
+        candidate_metrics[args.reference_metric],
+        deploy_flag,
+        args.force_comparison
+    ))
+
+    # save deploy_flag as a file
+    os.makedirs(args.deploy_flag, exist_ok=True)
+    with open(os.path.join(args.deploy_flag, "deploy_flag"), "w") as out_file:
+        out_file.write("%d" % int(deploy_flag))
+
+    # Stop Logging
+    mlflow.end_run()
+
+
+if __name__ == "__main__":
+    main()
--- a/nlp/python-sdk-v2/data-science/src/summarization/prepare.py
+++ b/nlp/python-sdk-v2/data-science/src/summarization/prepare.py
@ -0,0 +1,151 @@
+"""
+This scripts prepares a HuggingFace dataset to be used
+for fine-tuning. It encodes the train/val/test tests and
+outputs as JSONL files.
+"""
+import os
+import argparse
+import logging
+from datasets import load_dataset, DatasetDict
+from transformers import AutoTokenizer
+import mlflow
+
+
+def main():
+    """Main function of the script."""
+    # initialize root logger
+    logger = logging.getLogger()
+    logger.setLevel(logging.INFO)
+    console_handler = logging.StreamHandler()
+    formatter = logging.Formatter(
+        "%(asctime)s : %(levelname)s : %(name)s : %(message)s"
+    )
+    console_handler.setFormatter(formatter)
+    logger.addHandler(console_handler)
+
+    # input and output arguments
+    parser = argparse.ArgumentParser()
+    parser.add_argument(
+        "--dataset_name", type=str, help="name of dataset or path to input dataset_name"
+    )
+    parser.add_argument(
+        "--dataset_config", type=str, help="config for huggingface dataset"
+    )
+    parser.add_argument("--text_column", type=str, help="name of text_column")
+    parser.add_argument("--summary_column", type=str, help="name of summary_column")
+    parser.add_argument(
+        "--max_input_length", type=int, default=512, help="max_input_length"
+    )
+    parser.add_argument(
+        "--max_target_length", type=int, default=40, help="max_target_length"
+    )
+    parser.add_argument(
+        "--padding", type=str, default="max_length", help="padding type"
+    )
+    parser.add_argument(
+        "--model_arch",
+        type=str,
+        help="name of the model to prepare for in HF model library",
+    )
+    parser.add_argument(
+        "--limit_samples", type=int, default=-1, help="sample size from input dataset"
+    )
+    parser.add_argument("--encodings", type=str, help="path to tokenized dataset")
+    parser.add_argument(
+        "--source_prefix",
+        type=str,
+        help="A prefix to add before every source text (useful for T5 models).",
+    )
+
+    args = parser.parse_args()
+
+    # Start Logging
+    mlflow.start_run()
+
+    logger.info(f"Running with arguments: {args}")
+
+    # get tokenizer ready
+    tokenizer = AutoTokenizer.from_pretrained(args.model_arch)
+    logger.info(f"tokenizer: {tokenizer}")
+
+    prefix = args.source_prefix if args.source_prefix is not None else ""
+    if args.source_prefix is None and "t5" in args.model_arch.lower():
+        logger.warning(
+            "You're running a t5 model but didn't provide a source prefix, which is the expected, e.g. with "
+            "`--source_prefix 'summarize: ' `"
+        )
+
+    # Load HuggingFace dataset
+    raw_dataset = load_dataset(args.dataset_name, args.dataset_config)
+
+    logger.info(f"raw dataset length: {raw_dataset.num_rows}")
+    mlflow.log_metric("train_samples", raw_dataset["train"].shape[0])
+    mlflow.log_metric("test_samples", raw_dataset["test"].shape[0])
+    mlflow.log_metric("validation_samples", raw_dataset["validation"].shape[0])
+
+    if args.limit_samples > 0:
+        sample_sizes = {
+            k: min(len(raw_dataset[k]), args.limit_samples) for k in raw_dataset.keys()
+        }
+        raw_dataset = DatasetDict(
+            {
+                k: raw_dataset[k].select([i for i in range(sample_sizes[k])])
+                for k in raw_dataset.keys()
+            }
+        )
+        logger.info("sampled raw dataset:")
+        logger.info(raw_dataset.num_rows)
+
+    def preprocess_function(examples):
+        # remove pairs where at least one record is None
+
+        inputs, targets = [], []
+        for i in range(len(examples[args.text_column])):
+            if (
+                examples[args.text_column][i] is not None
+                and examples[args.summary_column][i] is not None
+            ):
+                inputs.append(examples[args.text_column][i])
+                targets.append(examples[args.summary_column][i])
+
+        inputs = [prefix + inp for inp in inputs]
+        model_inputs = tokenizer(
+            inputs,
+            max_length=args.max_input_length,
+            padding=args.padding,
+            truncation=True,
+        )
+        # Set up the tokenizer for targets
+        with tokenizer.as_target_tokenizer():
+            labels = tokenizer(
+                examples[args.summary_column],
+                max_length=args.max_target_length,
+                padding=args.padding,
+                truncation=True,
+            )
+
+        # replace all tokenizer.pad_token_id in the labels by -100 to ignore padding in the loss.
+        if args.padding == "max_length":
+            labels["input_ids"] = [
+                [(l if l != tokenizer.pad_token_id else -100) for l in label]
+                for label in labels["input_ids"]
+            ]
+        model_inputs["labels"] = labels["input_ids"]
+        return model_inputs
+
+    preprocessed_datasets = raw_dataset.map(preprocess_function, batched=True)
+    logger.info(f"preprocessed_datasets: {preprocessed_datasets}")
+
+    output_path = os.path.join(args.encodings)
+    os.makedirs(output_path, exist_ok=True)
+
+    preprocessed_datasets.save_to_disk(output_path)
+
+    logger.info(f"tokenized data is saved to {output_path}")
+
+    # Stop Logging
+    mlflow.end_run()
+
+
+if __name__ == "__main__":
+    main()
--- a/nlp/python-sdk-v2/data-science/src/summarization/register.py
+++ b/nlp/python-sdk-v2/data-science/src/summarization/register.py
@ -0,0 +1,82 @@
+from azureml.core import Run
+from azureml.core.model import Model
+
+import os
+import argparse
+import logging
+import mlflow
+
+
+def main():
+    """Main function of the script."""
+    # initialize root logger
+    logger = logging.getLogger()
+    logger.setLevel(logging.INFO)
+    console_handler = logging.StreamHandler()
+    formatter = logging.Formatter(
+        "%(asctime)s : %(levelname)s : %(name)s : %(message)s"
+    )
+    console_handler.setFormatter(formatter)
+    logger.addHandler(console_handler)
+
+    # input and output arguments
+    parser = argparse.ArgumentParser()
+    parser.add_argument(
+        "--model_folder",
+        type=str,
+        required=True,
+        help="folder containing model",
+    )
+    parser.add_argument(
+        "--register_as",
+        type=str,
+        required=True,
+        help="name to use for model registration in AzureML",
+    )
+    parser.add_argument(
+        "--deploy_flag", type=str, required=True, help="a deploy flag whether to deploy or not"
+    )
+
+    args = parser.parse_args()
+    logger.info(f"Running with arguments: {args}")
+
+    # Start Logging
+    mlflow.start_run()
+
+    if os.path.isfile(args.deploy_flag):
+        deploy_flag_file_path = args.deploy_flag
+    else:
+        deploy_flag_file_path = os.path.join(args.deploy_flag, "deploy_flag")
+
+    logger.info(f"Opening deploy_flag file from {deploy_flag_file_path}")
+    with open(deploy_flag_file_path, 'rb') as in_file:
+        deploy_flag = bool(int(in_file.read()))
+
+    if deploy_flag:
+        logger.info(f"Deploy flag is True, registering model as {args.register_as}...")
+        run = Run.get_context()
+
+        # if we're running locally, except
+        if run.__class__.__name__ == "_OfflineRun":
+            raise Exception("You can't run this script locally, you will need to run it as an AzureML job.")
+
+        _ = Model.register(
+            run.experiment.workspace,
+            model_name=args.register_as,
+            model_path=args.model_folder,
+            tags={
+                "type": "huggingface",
+                "task": "summarization"
+            },
+            description="Huggingface model finetuned for summarization",
+        )
+    else:
+        logger.info(f"Deploy flag is False, pass.")
+
+    # Stop Logging
+    mlflow.end_run()
+
+
+if __name__ == "__main__":
+    main()
+
--- a/nlp/python-sdk-v2/data-science/src/summarization/run.py
+++ b/nlp/python-sdk-v2/data-science/src/summarization/run.py
@ -0,0 +1,269 @@
+import logging
+import os
+from datasets import load_metric, load_from_disk
+from transformers import (
+    AutoModelForSeq2SeqLM,
+    AutoTokenizer,
+    DataCollatorForSeq2Seq,
+    Seq2SeqTrainingArguments,
+    Seq2SeqTrainer,
+    HfArgumentParser,
+    IntervalStrategy,
+)
+from transformers.trainer_callback import TrainerCallback
+
+import torch
+import nltk
+from dataclasses import dataclass, field
+from typing import Optional
+import numpy as np
+import mlflow
+from pynvml import *
+import time
+
+
+# Input arguments are set with dataclass. Huggingface library stores the default training args in TrainingArguments dataclass
+# user args are also defined in dataclasses, we will then load arguments from a tuple of user defined and built-in dataclasses.
+@dataclass
+class DataArgs:
+    # Inputs
+    preprocessed_datasets: str = field(
+        default=None, metadata={"help": "path to preprocesed datasets"}
+    )
+
+    # Processing parameters
+    max_target_length: Optional[int] = field(
+        default=128,
+        metadata={"help": "maxi sequence length for target text after tokenization."},
+    )
+    limit_samples: Optional[int] = field(
+        default=-1,
+        metadata={"help": "limit the number of samples for faster run."},
+    )
+
+
+@dataclass
+class ModelArgs:
+    model_name: Optional[str] = field(default=None, metadata={"help": "model name"})
+    model_path: Optional[str] = field(
+        default=None, metadata={"help": "path to existing model file to load"}
+    )
+    model_output: Optional[str] = field(
+        default=None, metadata={"help": "path to save the model"}
+    )
+
+
+nltk.download("punkt")
+
+
+def print_gpu_utilization():
+    nvmlInit()
+    handle = nvmlDeviceGetHandleByIndex(0)
+    info = nvmlDeviceGetMemoryInfo(handle)
+    print(f"GPU memory occupied: {info.used//1024**2} MB.")
+
+
+def print_summary(result):
+    print(f"Time: {result.metrics['train_runtime']:.2f}")
+    print(f"Samples/second: {result.metrics['train_samples_per_second']:.2f}")
+    print_gpu_utilization()
+
+
+def postprocess_text(preds, labels):
+    """Postprocess output for computing metrics"""
+    preds = [pred.strip() for pred in preds]
+    labels = [label.strip() for label in labels]
+
+    # rougeLSum expects newline after each sentence
+    preds = ["\n".join(nltk.sent_tokenize(pred)) for pred in preds]
+    labels = ["\n".join(nltk.sent_tokenize(label)) for label in labels]
+
+    return preds, labels
+
+
+def compute_metrics(eval_preds, tokenizer, metric):
+    """Compute metric based on predictions from evaluation"""
+    preds, labels = eval_preds
+    if isinstance(preds, tuple):
+        preds = preds[0]
+    decoded_preds = tokenizer.batch_decode(preds, skip_special_tokens=True)
+
+    # Replace -100 in the labels as we can't decode them.
+    labels = np.where(labels != -100, labels, tokenizer.pad_token_id)
+    decoded_labels = tokenizer.batch_decode(labels, skip_special_tokens=True)
+
+    # Some simple post-processing
+    decoded_preds, decoded_labels = postprocess_text(decoded_preds, decoded_labels)
+
+    result = metric.compute(
+        predictions=decoded_preds, references=decoded_labels, use_stemmer=True
+    )
+    # Extract a few results from ROUGE
+    result = {key: value.mid.fmeasure * 100 for key, value in result.items()}
+
+    prediction_lens = [
+        np.count_nonzero(pred != tokenizer.pad_token_id) for pred in preds
+    ]
+    result["gen_len"] = np.mean(prediction_lens)
+    result = {k: round(v, 4) for k, v in result.items()}
+    return result
+
+
+class CustomCallback(TrainerCallback):
+    """A [`TrainerCallback`] that sends the logs to [AzureML](https://pypi.org/project/azureml-sdk/).
+
+    This is a hotfix for the issue raised here:
+    https://github.com/huggingface/transformers/issues/18870
+    """
+    def on_log(self, args, state, control, logs=None, **kwargs):
+        if state.is_world_process_zero:
+            metrics = {}
+            for k, v in logs.items():
+                if isinstance(v, (int, float)):
+                    metrics[k] = v
+            mlflow.log_metrics(metrics=metrics, step=state.global_step)
+
+
+def main():
+    # Setup logging
+    logger = logging.getLogger()
+    logger.setLevel(logging.INFO)
+    console_handler = logging.StreamHandler()
+    formatter = logging.Formatter(
+        "%(asctime)s : %(levelname)s : %(name)s : %(message)s"
+    )
+    console_handler.setFormatter(formatter)
+    logger.addHandler(console_handler)
+
+    # initialize the mlflow session
+    mlflow.start_run()
+
+    parser = HfArgumentParser((ModelArgs, DataArgs, Seq2SeqTrainingArguments))
+    model_args, data_args, training_args = parser.parse_args_into_dataclasses()
+    logger.info(f"Running with arguments: {model_args}, {data_args}, {training_args}")
+
+    # Check if this is the main node
+    is_this_main_node = int(os.environ.get("RANK", "0")) == 0
+    if is_this_main_node:
+        logger.info("This is the main Node")
+
+    input_datasets = load_from_disk(data_args.preprocessed_datasets)
+    logger.info(f"preprocessed dataset is loaded")
+
+    if model_args.model_path:
+        logger.info("using a saved model")
+        model = AutoModelForSeq2SeqLM.from_pretrained(model_args.model_path)
+        tokenizer = AutoTokenizer.from_pretrained(model_args.model_path)
+    else:
+        logger.info("using a model from model library")
+        model = AutoModelForSeq2SeqLM.from_pretrained(model_args.model_name)
+        tokenizer = AutoTokenizer.from_pretrained(model_args.model_name)
+
+    # Artificially limit the number of samples (for testing)
+    if training_args.do_train:  # if using --do-train from Seq2SeqTrainingArguments
+        if data_args.limit_samples > 0:
+            max_train_samples = min(len(input_datasets["train"]), data_args.limit_samples)
+            train_dataset = input_datasets["train"].select(range(max_train_samples))
+            logger.info(f"train: making a {max_train_samples} sample of the data")
+        else:
+            train_dataset = input_datasets["train"]
+
+    if training_args.do_eval:
+        if data_args.limit_samples > 0:
+            max_eval_samples = min(
+                len(input_datasets["validation"]), data_args.limit_samples
+            )
+            eval_dataset = input_datasets["validation"].select(range(max_eval_samples))
+            logger.info(f"eval: making a {max_eval_samples} sample of the data")
+        else:
+            eval_dataset = input_datasets["validation"]
+
+    if training_args.do_predict:
+        if data_args.limit_samples > 0:
+            max_predict_samples = min(
+                len(input_datasets["test"]), data_args.limit_samples
+            )
+            predict_dataset = input_datasets["test"].select(range(max_predict_samples))
+            logger.info(f"predict: making a {max_predict_samples} sample of the data")
+        else:
+            predict_dataset = input_datasets["test"]
+
+    # Data collator
+    label_pad_token_id = -100
+
+    data_collator = DataCollatorForSeq2Seq(
+        tokenizer,
+        model=model,
+        label_pad_token_id=label_pad_token_id,
+    )
+
+    # Metric
+    metric = load_metric("rouge")
+
+    if training_args.do_train:
+        logging_steps = len(train_dataset) // training_args.per_device_train_batch_size
+        training_args.logging_steps = logging_steps
+    #training_args.output_dir = "outputs"
+    training_args.save_strategy = "epoch"
+    training_args.evaluation_strategy = IntervalStrategy.EPOCH
+    training_args.predict_with_generate = True
+    training_args.report_to = [] # use our own callback
+    logger.info(f"training args: {training_args}")
+
+    # Initialize our Trainer
+    trainer = Seq2SeqTrainer(
+        model=model,
+        args=training_args,
+        train_dataset=train_dataset if training_args.do_train else None,
+        eval_dataset=eval_dataset if training_args.do_eval else None,
+        tokenizer=tokenizer,
+        data_collator=data_collator,
+        compute_metrics=lambda preds : compute_metrics(preds, tokenizer, metric),
+        callbacks=[CustomCallback]
+    )
+
+    # Start the actual training (to include evaluation use --do-eval)
+    if training_args.do_train:
+        logger.info("Start training")
+        start = time.time()
+        train_result = trainer.train()
+
+        mlflow.log_metric(
+            "time/epoch", (time.time() - start) / 60 / training_args.num_train_epochs
+        )
+        logger.info(
+            "training is done"
+        )  # Only print gpu utilization if gpu is available
+        if torch.cuda.is_available():
+            print_summary(train_result)
+
+    # Save the model as an output
+    if model_args.model_output and is_this_main_node:
+        logger.info(f"Saving the model at {model_args.model_output}")
+        os.makedirs(model_args.model_output, exist_ok=True)
+        trainer.save_model(model_args.model_output)
+
+    # Just run the predictions
+    if training_args.do_predict:
+        logger.info("*** Predict ***")
+        max_length = (
+            training_args.generation_max_length
+            if training_args.generation_max_length is not None
+            else data_args.max_target_length
+        )
+
+        predict_results = trainer.predict(
+            predict_dataset, metric_key_prefix="predict", max_length=max_length
+        )
+        metrics = predict_results.metrics
+        metrics["predict_samples"] = len(predict_dataset)
+
+        trainer.log_metrics("predict", metrics)
+        trainer.save_metrics("predict", metrics)
+
+    # Stop Logging
+    mlflow.end_run()
+
+
+if __name__ == "__main__":
+    main()
--- a/nlp/python-sdk-v2/data-science/src/summarization/score.py
+++ b/nlp/python-sdk-v2/data-science/src/summarization/score.py
@ -0,0 +1,44 @@
+import os
+import logging
+import json
+from transformers import (
+    AutoModelForSeq2SeqLM,
+    AutoTokenizer,
+)
+
+def init():
+    """
+    This function is called when the container is initialized/started, typically after create/update of the deployment.
+    You can write the logic here to perform init operations like caching the model in memory
+    """
+    global model, tokenizer
+    # AZUREML_MODEL_DIR is an environment variable created during deployment.
+    # It is the path to the model folder (./azureml-models/$MODEL_NAME/$VERSION)
+    model_path = os.path.join(os.getenv("AZUREML_MODEL_DIR"), os.listdir(os.getenv("AZUREML_MODEL_DIR"))[0])
+    print("model_path")
+    print(os.listdir(model_path))
+    model = AutoModelForSeq2SeqLM.from_pretrained(model_path)
+    tokenizer = AutoTokenizer.from_pretrained(model_path)
+    print("Init complete")
+
+
+def run(raw_data):
+    global model, tokenizer
+    """
+    This function is called for every invocation of the endpoint to perform the actual scoring/prediction.
+    In the example we extract the data from the json input and call the scikit-learn model's predict()
+    method and return the result back
+    """
+    logging.info("Request received")
+    article = json.loads(raw_data)["data"]
+    if "t5" in model.config.architectures[0].lower():
+        article= "summarize:" + article
+    
+    inputs = tokenizer(article, return_tensors="pt", max_length=512, truncation=True)
+    outputs = model.generate(
+        inputs["input_ids"], max_length=150, min_length=40, length_penalty=2.0, num_beams=4, early_stopping=True
+    )
+    result = tokenizer.decode(outputs[0])
+    print(result)
+    logging.info("Request processed")
+    return result
--- a/nlp/python-sdk-v2/data/nlp-summarization-request.json
+++ b/nlp/python-sdk-v2/data/nlp-summarization-request.json
--- a/nlp/python-sdk-v2/mlops/azureml/train/pipeline-train.py
+++ b/nlp/python-sdk-v2/mlops/azureml/train/pipeline-train.py
@ -0,0 +1,405 @@
+"""MLOps v2 NLP Python SDK training submission script."""
+import os
+import argparse
+
+# Azure ML sdk v2 imports
+from azure.identity import DefaultAzureCredential, InteractiveBrowserCredential
+from azure.ai.ml import MLClient
+from azure.ai.ml import command
+from azure.ai.ml import Input, Output
+from azure.ai.ml import dsl, Input, Output
+
+def get_config_parger(parser: argparse.ArgumentParser = None):
+    """Builds the argument parser for the script."""
+    if parser is None:
+        parser = argparse.ArgumentParser(description=__doc__)
+
+    group = parser.add_argument_group("Azure ML references")
+    group.add_argument(
+        "--config_location",
+        type=str,
+        required=False,
+        help="Subscription ID",
+    )
+    group.add_argument(
+        "--subscription_id",
+        type=str,
+        required=False,
+        help="Subscription ID",
+    )
+    group.add_argument(
+        "--resource_group",
+        type=str,
+        required=False,
+        help="Resource group name",
+    )
+    group.add_argument(
+        "--workspace_name",
+        type=str,
+        required=False,
+        help="Workspace name",
+    )
+    # Experiment Name
+    group.add_argument(
+        "-n",
+        type=str,
+        required=True,
+        default="nlp_summarization_train",
+        help="Experiment name",
+    )
+    parser.add_argument(
+        "--wait",
+        default=False,
+        action="store_true",
+        help="wait for the job to finish",
+    )
+
+    group = parser.add_argument_group("Training parameters")
+    group.add_argument(
+        "--limit_samples",
+        type=int,
+        default=1000,
+    )
+    group.add_argument(
+        "--pretrained_model_name",
+        type=str,
+        default="t5-small",
+    )
+    group.add_argument(
+        "--num_train_epochs",
+        type=int,
+        default=5,
+    )
+    group.add_argument(
+        "--batch_size",
+        type=int,
+        default=8,
+    )
+    group.add_argument(
+        "--learning_rate",
+        type=float,
+        default=0.00005,
+    )
+    group.add_argument(
+        "--model_registration_name",
+        type=str,
+        default="pubmed-summarization",
+    )
+
+    group = parser.add_argument_group("Compute parameters")
+    group.add_argument(
+        "--cpu_compute",
+        type=str,
+        default="cpu-cluster",
+    )
+    group.add_argument(
+        "--cpu_compute_large",
+        type=str,
+        default="cpu-cluster-lg",
+    )
+    group.add_argument(
+        "--gpu_compute",
+        type=str,
+        default="gpu-cluster",
+    )
+    group.add_argument(
+        "--training_nodes",
+        type=int,
+        default=1,
+    )
+    group.add_argument(
+        "--gpus_per_node",
+        type=int,
+        default=1,
+    )
+
+    return parser
+
+
+def connect_to_aml(args):
+    """Connect to Azure ML workspace using provided cli arguments."""
+    try:
+        credential = DefaultAzureCredential()
+        # Check if given credential can get token successfully.
+        credential.get_token("https://management.azure.com/.default")
+    except Exception as ex:
+        # Fall back to InteractiveBrowserCredential in case DefaultAzureCredential not work
+        credential = InteractiveBrowserCredential()
+
+    # Get a handle to workspace
+    try:
+        # ml_client to connect using local config.json
+        ml_client = MLClient.from_config(credential, path='config.json')
+
+    except Exception as ex:
+        print(
+            "Could not find config.json, using config.yaml refs to Azure ML workspace instead."
+        )
+
+        # tries to connect using cli args if provided else using config.yaml
+        ml_client = MLClient(
+            subscription_id=args.subscription_id,
+            resource_group_name=args.resource_group,
+            workspace_name=args.workspace_name,
+            credential=credential,
+        )
+    return ml_client
+
+
+def build_components(args):
+    """Builds the components for the pipeline."""
+    DATA_SCIENCE_FOLDER = os.path.join(
+        os.path.dirname(os.path.abspath(__file__)), "..","..", "..", "data-science", "src"
+    )
+
+    prep_finetuning_dataset = command(
+        name="prep_finetuning_dataset",
+        display_name="Prepare dataset for training",
+        inputs={
+            "dataset_name": Input(type="string"),
+            "dataset_config": Input(type="string"),
+            "text_column": Input(type="string"),
+            "summary_column": Input(type="string"),
+            "limit_samples": Input(type="integer"),
+            "max_input_length": Input(type="integer"),
+            "max_target_length": Input(type="integer"),
+            "padding": Input(type="string"),
+            "pretrained_model_name": Input(type="string"),
+        },
+        outputs=dict(
+            encodings=Output(type="uri_folder", mode="rw_mount"),
+        ),
+        code=DATA_SCIENCE_FOLDER,
+        command="""python summarization/prepare.py \
+                        --dataset_name ${{inputs.dataset_name}} \
+                        --dataset_config ${{inputs.dataset_config}} \
+                        --text_column ${{inputs.text_column}} \
+                        --summary_column ${{inputs.summary_column}} \
+                        --limit_samples ${{inputs.limit_samples}} \
+                        --model_arch ${{inputs.pretrained_model_name}} \
+                        --max_input_length ${{inputs.max_input_length}} \
+                        --max_target_length ${{inputs.max_target_length}} \
+                        --padding ${{inputs.padding}} \
+                        --encodings ${{outputs.encodings}}\
+                """,
+        environment="nlp_summarization_train@latest",
+    )
+
+    finetune_model = command(
+        name="finetune_model",
+        display_name="Fine-tune summarization model",
+        inputs={
+            "preprocessed_datasets": Input(type="uri_folder"),
+            "pretrained_model_name": Input(type="string"),
+            "limit_samples": Input(type="integer"),
+            "learning_rate": Input(type="number"),
+            "num_train_epochs": Input(type="integer"),
+            "per_device_train_batch_size": Input(type="integer"),
+            "per_device_eval_batch_size": Input(type="integer"),
+        },
+        outputs=dict(
+            finetuned_model=Output(type="uri_folder", mode="rw_mount"),
+        ),
+        code=DATA_SCIENCE_FOLDER,
+        command="""python summarization/run.py \
+                    --preprocessed_datasets ${{inputs.preprocessed_datasets}} \
+                    --learning_rate ${{inputs.learning_rate}} \
+                    --per_device_train_batch_size ${{inputs.per_device_train_batch_size}} \
+                    --per_device_eval_batch_size ${{inputs.per_device_eval_batch_size}} \
+                    --limit_samples ${{inputs.limit_samples}} \
+                    --model_name ${{inputs.pretrained_model_name}} \
+                    --model_output ${{outputs.finetuned_model}}\
+                    --output_dir outputs \
+                    --num_train_epochs ${{inputs.num_train_epochs}} \
+                    --do_train --do_eval \
+                """,
+        environment="nlp_summarization_train@latest",
+        distribution={
+            "type": "PyTorch",
+            # set process count to the number of gpus on the node
+            "process_count_per_instance": args.gpus_per_node,
+        },
+        # set instance count to the number of nodes you want to use
+        instance_count=args.training_nodes,
+    )
+
+    evaluate_model = command(
+        name="evaluate_model",
+        display_name="Run eval on a model",
+        inputs={
+            "preprocessed_datasets": Input(type="uri_folder"),
+            "model_path": Input(type="uri_folder", optional=True),
+            "model_name": Input(type="string", optional=True),
+            "limit_samples": Input(type="integer"),
+            "max_target_length": Input(type="integer"),
+        },
+        outputs=dict(
+            metrics=Output(type="uri_folder", mode="rw_mount"),
+        ),
+        code=DATA_SCIENCE_FOLDER,
+        command="""python summarization/run.py \
+                    --preprocessed_datasets ${{inputs.preprocessed_datasets}} \
+                    --limit_samples ${{inputs.limit_samples}} \
+                    --output_dir ${{outputs.metrics}} \
+                    $[[--model_path ${{inputs.model_path}}]] \
+                    $[[--model_name ${{inputs.model_name}}]] \
+                    --max_target_length ${{inputs.max_target_length}} \
+                    --do_predict \
+                """,
+        environment="nlp_summarization_train@latest",
+    )
+
+    compare_models = command(
+        name="compare_models",
+        display_name="Compare finetuned to baseline",
+        inputs={
+            "baseline_metrics": Input(type="uri_folder"),
+            "candidate_metrics": Input(type="uri_folder"),
+            "reference_metric": Input(type="string"),
+        },
+        outputs=dict(
+            deploy_flag=Output(type="uri_folder", mode="rw_mount"),
+        ),
+        code=DATA_SCIENCE_FOLDER,
+        command="""python summarization/compare.py \
+                    --baseline_metrics ${{inputs.baseline_metrics}} \
+                    --candidate_metrics ${{inputs.candidate_metrics}} \
+                    --reference_metric ${{inputs.reference_metric}} \
+                    --deploy_flag ${{outputs.deploy_flag}} \
+                    --force_comparison True\
+                """,
+        environment="AzureML-sklearn-1.0-ubuntu20.04-py38-cpu@latest",
+    )
+
+    register_model = command(
+        name="register_model",
+        display_name="Register model",
+        inputs={
+            "model": Input(type="uri_folder"),
+            "deploy_flag": Input(type="uri_folder"),
+            "model_registration_name": Input(type="string"),
+        },
+        code=DATA_SCIENCE_FOLDER,
+        command="""python summarization/register.py \
+                    --model_folder ${{inputs.model}} \
+                    --deploy_flag ${{inputs.deploy_flag}} \
+                    --register_as ${{inputs.model_registration_name}} \
+                """,
+        environment="AzureML-sklearn-1.0-ubuntu20.04-py38-cpu@latest",
+    )
+
+    return {
+        "prep_finetuning_dataset": prep_finetuning_dataset,
+        "finetune_model": finetune_model,
+        "evaluate_model": evaluate_model,
+        "compare_models": compare_models,
+        "register_model": register_model,
+    }
+
+
+def main():
+    """Main entry point for the script."""
+    parser = get_config_parger()
+    args, _ = parser.parse_known_args()
+    ml_client = connect_to_aml(args)
+
+    # get components from build function
+    components_dict = build_components(args)
+    prep_finetuning_dataset = components_dict["prep_finetuning_dataset"]
+    finetune_model = components_dict["finetune_model"]
+    evaluate_model = components_dict["evaluate_model"]
+    compare_models = components_dict["compare_models"]
+    register_model = components_dict["register_model"]
+
+    # build the pipeline using Azure ML SDK v2
+    @dsl.pipeline(
+        name="NLP Training Pipeline",
+        description="NLP Training Pipeline",
+    )
+    def nlp_training_pipeline(
+        limit_samples: int,
+        pretrained_model_name: str,
+        num_train_epochs: int,
+        batch_size: int,
+        learning_rate: float,
+        model_registration_name: str,
+    ):
+        prep_finetuning_dataset_step = prep_finetuning_dataset(
+            dataset_name="ccdv/pubmed-summarization",
+            dataset_config="section",
+            text_column="article",
+            summary_column="abstract",
+            limit_samples=limit_samples,
+            max_input_length=512,
+            max_target_length=40,
+            padding="max_length",
+            pretrained_model_name=pretrained_model_name,
+        )
+        prep_finetuning_dataset_step.compute = args.cpu_compute_large
+
+        finetune_model_step = finetune_model(
+            preprocessed_datasets=prep_finetuning_dataset_step.outputs.encodings,
+            pretrained_model_name=pretrained_model_name,
+            limit_samples=limit_samples,
+            learning_rate=learning_rate,
+            num_train_epochs=num_train_epochs,
+            per_device_train_batch_size=batch_size,
+            per_device_eval_batch_size=batch_size,
+        )
+        finetune_model_step.compute = args.gpu_compute
+
+        evaluate_finetuned_model_step = evaluate_model(
+            preprocessed_datasets=prep_finetuning_dataset_step.outputs.encodings,
+            model_path=finetune_model_step.outputs.finetuned_model,
+            limit_samples=limit_samples,
+            max_target_length=40,
+        )
+        evaluate_finetuned_model_step.compute = args.gpu_compute
+
+        evaluate_baseline_model_step = evaluate_model(
+            preprocessed_datasets=prep_finetuning_dataset_step.outputs.encodings,
+            model_name=pretrained_model_name,
+            limit_samples=limit_samples,
+            max_target_length=40,
+        )
+        evaluate_baseline_model_step.compute = args.gpu_compute
+
+        compare_models_step = compare_models(
+            baseline_metrics=evaluate_finetuned_model_step.outputs.metrics,
+            candidate_metrics=evaluate_baseline_model_step.outputs.metrics,
+            reference_metric="predict_rougeLsum",
+        )
+        compare_models_step.compute = args.cpu_compute
+
+        register_model_step = register_model(
+            model=finetune_model_step.outputs.finetuned_model,
+            deploy_flag=compare_models_step.outputs.deploy_flag,
+            model_registration_name=model_registration_name,
+        )
+        register_model_step.compute = args.cpu_compute
+
+    # instanciates the job
+    pipeline_job = nlp_training_pipeline(
+        limit_samples=args.limit_samples,
+        pretrained_model_name=args.pretrained_model_name,
+        num_train_epochs=args.num_train_epochs,
+        batch_size=args.batch_size,
+        learning_rate=args.learning_rate,
+        model_registration_name=args.model_registration_name,
+    )
+
+    # submits the job
+    print("Submitting the pipeline job to your AzureML workspace...")
+    pipeline_job = ml_client.jobs.create_or_update(
+        pipeline_job, experiment_name=args.n
+    )
+
+    print("The url to see your live job running is returned by the sdk:")
+    print(pipeline_job.services["Studio"].endpoint)
+
+    if args.wait:
+        ml_client.jobs.stream(pipeline_job.name)
+
+
+if __name__ == "__main__":
+    main()
--- a/nlp/python-sdk-v2/mlops/devops-pipelines/deploy-model-training-pipeline.yml
+++ b/nlp/python-sdk-v2/mlops/devops-pipelines/deploy-model-training-pipeline.yml
@ -0,0 +1,71 @@
+# Copyright (c) Microsoft Corporation. All rights reserved.
+# Licensed under the MIT License.
+
+variables:
+  - ${{ if eq(variables['Build.SourceBranchName'], 'main') }}:
+      # 'main' branch: PRD environment
+      - template: ../../config-infra-prod.yml
+  - ${{ if ne(variables['Build.SourceBranchName'], 'main') }}:
+      # 'develop' or feature branches: DEV environment
+      - template: ../../config-infra-dev.yml
+  - name: version
+    value: python-sdk-v2
+
+trigger:
+  - none
+
+pool:
+  vmImage: ubuntu-20.04
+
+resources:
+  repositories:
+    - repository: mlops-templates # Template Repo
+      name: Azure/mlops-templates # need to change org name from "Azure" to your own org
+      endpoint: github-connection # need to set up and hardcode
+      type: github
+      ref: main
+
+stages:
+  - stage: DeployTrainingPipeline
+    displayName: Deploy Training Pipeline
+    jobs:
+      - job: DeployTrainingPipeline
+        steps:
+          - checkout: self
+            path: s/
+          - checkout: mlops-templates
+            path: s/templates/
+          - template: templates/aml-cli-v2/install-az-cli.yml@mlops-templates
+          - template: templates/aml-cli-v2/install-aml-cli.yml@mlops-templates
+          - template: templates/python-sdk-v2/install-requirements.yml@mlops-templates
+          - template: templates/aml-cli-v2/connect-to-workspace.yml@mlops-templates
+          - template: templates/${{ variables.version }}/create-compute.yml@mlops-templates
+            parameters:
+              cluster_name: cpu-cluster
+              size: STANDARD_DS3_V2
+              min_instances: 0
+              max_instances: 1
+              cluster_tier: dedicated
+          - template: templates/${{ variables.version }}/create-compute.yml@mlops-templates
+            parameters:
+              cluster_name: cpu-cluster-lg
+              size: Standard_D14_v2
+              min_instances: 0
+              max_instances: 1
+              cluster_tier: dedicated
+          - template: templates/${{ variables.version }}/create-compute.yml@mlops-templates
+            parameters:
+              cluster_name: gpu-cluster
+              size: Standard_NV6
+              min_instances: 0
+              max_instances: 1
+              cluster_tier: dedicated
+          - template: templates/${{ variables.version }}/register-environment.yml@mlops-templates
+            parameters:
+              build_type: docker
+              environment_name: nlp_summarization_train
+              environment_path: data-science/environments/training
+          - template: templates/${{ variables.version }}/run-pipeline.yml@mlops-templates
+            parameters:
+              pipeline_path: mlops/azureml/train/pipeline-train.py
+              experiment_name: $(environment)_nlp_summarization_$(Build.SourceBranchName)