Dataset versioning (microsoft#218)

Author: eedorenko

.env.example

@@ -32,6 +32,9 @@ EVALUATE_SCRIPT_PATH = 'evaluate/evaluate_model.py'
 REGISTER_SCRIPT_PATH = 'register/register_model.py'
 SOURCES_DIR_TRAIN = 'diabetes_regression'
 DATASET_NAME = 'diabetes_ds'
+DATASET_VERSION = 'latest'
+# Optional. Set it if you have configured non default datastore to point to your data
+DATASTORE_NAME = ''
 SCORE_SCRIPT = 'scoring/score.py'
 
 # Optional. Used by a training pipeline with R on Databricks

.pipelines/diabetes_regression-variables-template.yml

@@ -23,6 +23,11 @@ variables:
     value: mlopspython
   - name: DATASET_NAME
     value: diabetes_ds
+  # Uncomment DATASTORE_NAME if you have configured non default datastore to point to your data   
+  # - name: DATASTORE_NAME
+  #   value: datablobstore
+  - name: DATASET_VERSION
+    value: latest
   - name: TRAINING_PIPELINE_NAME
     value: "diabetes-Training-Pipeline"
   - name: MODEL_NAME

diabetes_regression/register/register_model.py

@@ -28,7 +28,7 @@
 import argparse
 import traceback
 import joblib
-from azureml.core import Run, Experiment, Workspace
+from azureml.core import Run, Experiment, Workspace, Dataset
 from azureml.core.model import Model as AMLModel
 
 
@@ -105,15 +105,23 @@ def main():
         print("Tags present: {parent_tags}")
 
     if (model is not None):
+        dataset_id = parent_tags["dataset_id"]
         if (build_id is None):
-            register_aml_model(model_file, model_name, model_mse, exp, run_id)
+            register_aml_model(
+                model_file,
+                model_name,
+                model_mse,
+                exp,
+                run_id,
+                dataset_id)
         elif (build_uri is None):
             register_aml_model(
                 model_file,
                 model_name,
                 model_mse,
                 exp,
                 run_id,
+                dataset_id,
                 build_id)
         else:
             register_aml_model(
@@ -122,6 +130,7 @@ def main():
                 model_mse,
                 exp,
                 run_id,
+                dataset_id,
                 build_id,
                 build_uri)
     else:
@@ -146,6 +155,7 @@ def register_aml_model(
     model_mse,
     exp,
     run_id,
+    dataset_id,
     build_id: str = 'none',
     build_uri=None
 ):
@@ -164,7 +174,9 @@ def register_aml_model(
             workspace=exp.workspace,
             model_name=model_name,
             model_path=model_path,
-            tags=tagsValue)
+            tags=tagsValue,
+            datasets=[('training data',
+                       Dataset.get_by_id(exp.workspace, dataset_id))])
         os.chdir("..")
         print(
             "Model registered: {} \nModel Description: {} "

diabetes_regression/training/train.py

@@ -31,6 +31,22 @@
 from sklearn.model_selection import train_test_split
 import joblib
 import json
+from azureml.core import Dataset, Datastore, Workspace
+
+
+def register_dataset(
+    aml_workspace: Workspace,
+    dataset_name: str,
+    datastore_name: str,
+    file_path: str
+) -> Dataset:
+    datastore = Datastore.get(aml_workspace, datastore_name)
+    dataset = Dataset.Tabular.from_delimited_files(path=(datastore, file_path))
+    dataset = dataset.register(workspace=aml_workspace,
+                               name=dataset_name,
+                               create_new_version=True)
+
+    return dataset
 
 
 def train_model(run, data, alpha):
@@ -64,13 +80,47 @@ def main():
         help=("output for passing data to next step")
     )
 
+    parser.add_argument(
+        "--dataset_version",
+        type=str,
+        help=("dataset version")
+    )
+
+    parser.add_argument(
+        "--data_file_path",
+        type=str,
+        help=("data file path, if specified,\
+               a new version of the dataset will be registered")
+    )
+
+    parser.add_argument(
+        "--caller_run_id",
+        type=str,
+        help=("caller run id, for example ADF pipeline run id")
+    )
+
+    parser.add_argument(
+        "--dataset_name",
+        type=str,
+        help=("Dataset name. Dataset must be passed by name\
+              to always get the desired dataset version\
+              rather than the one used while the pipeline creation")
+    )
+
     args = parser.parse_args()
 
     print("Argument [model_name]: %s" % args.model_name)
     print("Argument [step_output]: %s" % args.step_output)
+    print("Argument [dataset_version]: %s" % args.dataset_version)
+    print("Argument [data_file_path]: %s" % args.data_file_path)
+    print("Argument [caller_run_id]: %s" % args.caller_run_id)
+    print("Argument [dataset_name]: %s" % args.dataset_name)
 
     model_name = args.model_name
     step_output_path = args.step_output
+    dataset_version = args.dataset_version
+    data_file_path = args.data_file_path
+    dataset_name = args.dataset_name
 
     print("Getting training parameters")
 
@@ -86,16 +136,27 @@ def main():
     run = Run.get_context()
 
     # Get the dataset
-    dataset = run.input_datasets['training_data']
-    if (dataset):
-        df = dataset.to_pandas_dataframe()
-        X = df.drop('Y', axis=1).values
-        y = df['Y'].values
+    if (dataset_name):
+        if (data_file_path == 'none'):
+            dataset = Dataset.get_by_name(run.experiment.workspace, dataset_name, dataset_version)  # NOQA: E402, E501
+        else:
+            dataset = register_dataset(run.experiment.workspace,
+                                       dataset_name,
+                                       os.environ.get("DATASTORE_NAME"),
+                                       data_file_path)
     else:
         e = ("No dataset provided")
         print(e)
         raise Exception(e)
 
+    # Link dataset to the step run so it is trackable in the UI
+    run.input_datasets['training_data'] = dataset
+    run.parent.tag("dataset_id", value=dataset.id)
+
+    df = dataset.to_pandas_dataframe()
+    X = df.drop('Y', axis=1).values
+    y = df['Y'].values
+
     X_train, X_test, y_train, y_test = train_test_split(
         X, y, test_size=0.2, random_state=0)
     data = {"train": {"X": X_train, "y": y_train},

ml_service/pipelines/diabetes_regression_build_train_pipeline.py

@@ -1,9 +1,8 @@
 from azureml.pipeline.core.graph import PipelineParameter
 from azureml.pipeline.steps import PythonScriptStep
 from azureml.pipeline.core import Pipeline, PipelineData
-from azureml.core import Workspace
+from azureml.core import Workspace, Dataset, Datastore
 from azureml.core.runconfig import RunConfiguration
-from azureml.core import Dataset
 from ml_service.util.attach_compute import get_compute
 from ml_service.util.env_variables import Env
 from ml_service.util.manage_environment import get_environment
@@ -39,8 +38,20 @@ def main():
     run_config = RunConfiguration()
     run_config.environment = environment
 
+    if (e.datastore_name):
+        datastore_name = e.datastore_name
+    else:
+        datastore_name = aml_workspace.get_default_datastore().name
+    run_config.environment.environment_variables["DATASTORE_NAME"] = datastore_name  # NOQA: E501
+
     model_name_param = PipelineParameter(
         name="model_name", default_value=e.model_name)
+    dataset_version_param = PipelineParameter(
+        name="dataset_version", default_value=e.dataset_version)
+    data_file_path_param = PipelineParameter(
+        name="data_file_path", default_value="none")
+    caller_run_id_param = PipelineParameter(
+        name="caller_run_id", default_value="none")
 
     # Get dataset name
     dataset_name = e.dataset_name
@@ -57,9 +68,9 @@ def main():
         df.to_csv(file_name, index=False)
 
         # Upload file to default datastore in workspace
-        default_ds = aml_workspace.get_default_datastore()
+        datatstore = Datastore.get(aml_workspace, datastore_name)
         target_path = 'training-data/'
-        default_ds.upload_files(
+        datatstore.upload_files(
             files=[file_name],
             target_path=target_path,
             overwrite=True,
@@ -68,17 +79,14 @@ def main():
         # Register dataset
         path_on_datastore = os.path.join(target_path, file_name)
         dataset = Dataset.Tabular.from_delimited_files(
-            path=(default_ds, path_on_datastore))
+            path=(datatstore, path_on_datastore))
         dataset = dataset.register(
             workspace=aml_workspace,
             name=dataset_name,
             description='diabetes training data',
             tags={'format': 'CSV'},
             create_new_version=True)
 
-    # Get the dataset
-    dataset = Dataset.get_by_name(aml_workspace, dataset_name)
-
     # Create a PipelineData to pass data between steps
     pipeline_data = PipelineData(
         'pipeline_data',
@@ -89,11 +97,14 @@ def main():
         script_name=e.train_script_path,
         compute_target=aml_compute,
         source_directory=e.sources_directory_train,
-        inputs=[dataset.as_named_input('training_data')],
         outputs=[pipeline_data],
         arguments=[
             "--model_name", model_name_param,
-            "--step_output", pipeline_data
+            "--step_output", pipeline_data,
+            "--dataset_version", dataset_version_param,
+            "--data_file_path", data_file_path_param,
+            "--caller_run_id", caller_run_id_param,
+            "--dataset_name", dataset_name,
         ],
         runconfig=run_config,
         allow_reuse=False,

ml_service/util/env_variables.py

@@ -40,6 +40,8 @@ def __init__(self):
         self._score_script = os.environ.get("SCORE_SCRIPT")
         self._build_uri = os.environ.get("BUILD_URI")
         self._dataset_name = os.environ.get("DATASET_NAME")
+        self._datastore_name = os.environ.get("DATASTORE_NAME")
+        self._dataset_version = os.environ.get("DATASET_VERSION")
         self._run_evaluation = os.environ.get("RUN_EVALUATION", "true")
         self._allow_run_cancel = os.environ.get(
             "ALLOW_RUN_CANCEL", "true")
@@ -145,6 +147,14 @@ def build_uri(self):
     def dataset_name(self):
         return self._dataset_name
 
+    @property
+    def datastore_name(self):
+        return self._datastore_name
+
+    @property
+    def dataset_version(self):
+        return self._dataset_version
+
     @property
     def run_evaluation(self):
         return self._run_evaluation