feat: Added create_training_pipeline_custom_package_job_sample and create_training_pipeline_custom_container_job_sample and reworked create_training_pipeline_custom_job_sample

samples/model-builder/conftest.py

@@ -147,6 +147,18 @@ def mock_custom_training_job():
     yield mock
 
 
+@pytest.fixture
+def mock_custom_container_training_job():
+    mock = MagicMock(aiplatform.training_jobs.CustomContainerTrainingJob)
+    yield mock
+
+
+@pytest.fixture
+def mock_custom_package_training_job():
+    mock = MagicMock(aiplatform.training_jobs.CustomPythonPackageTrainingJob)
+    yield mock
+
+
 @pytest.fixture
 def mock_image_training_job():
     mock = MagicMock(aiplatform.training_jobs.AutoMLImageTrainingJob)
@@ -198,15 +210,41 @@ def mock_run_automl_image_training_job(mock_image_training_job):
 
 
 @pytest.fixture
-def mock_init_custom_training_job():
-    with patch.object(aiplatform.training_jobs.CustomTrainingJob, "__init__") as mock:
-        mock.return_value = None
+def mock_get_custom_training_job(mock_custom_training_job):
+    with patch.object(aiplatform, "CustomTrainingJob") as mock:
+        mock.return_value = mock_custom_training_job
+        yield mock
+
+
+@pytest.fixture
+def mock_get_custom_container_training_job(mock_custom_container_training_job):
+    with patch.object(aiplatform, "CustomContainerTrainingJob") as mock:
+        mock.return_value = mock_custom_container_training_job
+        yield mock
+
+
+@pytest.fixture
+def mock_get_custom_package_training_job(mock_custom_package_training_job):
+    with patch.object(aiplatform, "CustomPythonPackageTrainingJob") as mock:
+        mock.return_value = mock_custom_package_training_job
+        yield mock
+
+
+@pytest.fixture
+def mock_run_custom_training_job(mock_custom_training_job):
+    with patch.object(mock_custom_training_job, "run") as mock:
+        yield mock
+
+
+@pytest.fixture
+def mock_run_custom_container_training_job(mock_custom_container_training_job):
+    with patch.object(mock_custom_container_training_job, "run") as mock:
         yield mock
 
 
 @pytest.fixture
-def mock_run_custom_training_job():
-    with patch.object(aiplatform.training_jobs.CustomTrainingJob, "run") as mock:
+def mock_run_custom_package_training_job(mock_custom_package_training_job):
+    with patch.object(mock_custom_package_training_job, "run") as mock:
         yield mock
 
 

samples/model-builder/create_training_pipeline_custom_training_managed_dataset_sample.py → samples/model-builder/create_training_pipeline_custom_container_job_sample.py

@@ -17,18 +17,18 @@
 from google.cloud import aiplatform
 
 
-#  [START aiplatform_sdk_create_training_pipeline_custom_job_sample]
-def create_training_pipeline_custom_training_managed_dataset_sample(
+#  [START aiplatform_sdk_create_training_pipeline_custom_container_job_sample]
+def create_training_pipeline_custom_container_job_sample(
     project: str,
     location: str,
+    staging_bucket: str,
     display_name: str,
-    script_path: str,
     container_uri: str,
     model_serving_container_image_uri: str,
-    dataset_id: int,
+    dataset_id: Optional[str] = None,
     model_display_name: Optional[str] = None,
     args: Optional[List[Union[str, float, int]]] = None,
-    replica_count: int = 0,
+    replica_count: int = 1,
     machine_type: str = "n1-standard-4",
     accelerator_type: str = "ACCELERATOR_TYPE_UNSPECIFIED",
     accelerator_count: int = 0,
@@ -37,19 +37,19 @@ def create_training_pipeline_custom_training_managed_dataset_sample(
     test_fraction_split: float = 0.1,
     sync: bool = True,
 ):
-    aiplatform.init(project=project, location=location)
+    aiplatform.init(project=project, location=location, staging_bucket=staging_bucket)
 
-    job = aiplatform.CustomTrainingJob(
+    job = aiplatform.CustomContainerTrainingJob(
         display_name=display_name,
-        script_path=script_path,
         container_uri=container_uri,
         model_serving_container_image_uri=model_serving_container_image_uri,
     )
 
-    my_image_ds = aiplatform.ImageDataset(dataset_id)
+    # This example uses an ImageDataset, but you can use another type
+    dataset = aiplatform.ImageDataset(dataset_id) if dataset_id else None
 
     model = job.run(
-        dataset=my_image_ds,
+        dataset=dataset,
         model_display_name=model_display_name,
         args=args,
         replica_count=replica_count,
@@ -70,4 +70,4 @@ def create_training_pipeline_custom_training_managed_dataset_sample(
     return model
 
 
-#  [END aiplatform_sdk_create_training_pipeline_custom_job_sample]
+#  [END aiplatform_sdk_create_training_pipeline_custom_container_job_sample]

samples/model-builder/create_training_pipeline_custom_training_managed_dataset_test.py → samples/model-builder/create_training_pipeline_custom_container_job_test.py

@@ -13,24 +13,24 @@
 # limitations under the License.
 
 
-import create_training_pipeline_custom_training_managed_dataset_sample
+import create_training_pipeline_custom_container_job_sample
 import test_constants as constants
 
 
-def test_create_training_pipeline_custom_job_sample(
+def test_create_training_pipeline_custom_container_job_sample(
     mock_sdk_init,
     mock_image_dataset,
-    mock_init_custom_training_job,
-    mock_run_custom_training_job,
     mock_get_image_dataset,
+    mock_get_custom_container_training_job,
+    mock_run_custom_container_training_job,
 ):
 
-    create_training_pipeline_custom_training_managed_dataset_sample.create_training_pipeline_custom_training_managed_dataset_sample(
+    create_training_pipeline_custom_container_job_sample.create_training_pipeline_custom_container_job_sample(
         project=constants.PROJECT,
         location=constants.LOCATION,
+        staging_bucket=constants.STAGING_BUCKET,
         display_name=constants.DISPLAY_NAME,
         args=constants.ARGS,
-        script_path=constants.SCRIPT_PATH,
         container_uri=constants.CONTAINER_URI,
         model_serving_container_image_uri=constants.CONTAINER_URI,
         dataset_id=constants.RESOURCE_ID,
@@ -44,25 +44,25 @@ def test_create_training_pipeline_custom_job_sample(
         test_fraction_split=constants.TEST_FRACTION_SPLIT,
     )
 
-    mock_get_image_dataset.assert_called_once_with(constants.RESOURCE_ID)
-
     mock_sdk_init.assert_called_once_with(
-        project=constants.PROJECT, location=constants.LOCATION
+        project=constants.PROJECT,
+        location=constants.LOCATION,
+        staging_bucket=constants.STAGING_BUCKET,
     )
-    mock_init_custom_training_job.assert_called_once_with(
+
+    mock_get_custom_container_training_job.assert_called_once_with(
         display_name=constants.DISPLAY_NAME,
-        script_path=constants.SCRIPT_PATH,
         container_uri=constants.CONTAINER_URI,
         model_serving_container_image_uri=constants.CONTAINER_URI,
     )
-    mock_run_custom_training_job.assert_called_once_with(
+    mock_run_custom_container_training_job.assert_called_once_with(
         dataset=mock_image_dataset,
         model_display_name=constants.DISPLAY_NAME_2,
-        args=constants.ARGS,
         replica_count=constants.REPLICA_COUNT,
         machine_type=constants.MACHINE_TYPE,
         accelerator_type=constants.ACCELERATOR_TYPE,
         accelerator_count=constants.ACCELERATOR_COUNT,
+        args=constants.ARGS,
         training_fraction_split=constants.TRAINING_FRACTION_SPLIT,
         validation_fraction_split=constants.VALIDATION_FRACTION_SPLIT,
         test_fraction_split=constants.TEST_FRACTION_SPLIT,

samples/model-builder/create_training_pipeline_custom_job_sample.py

@@ -21,10 +21,12 @@
 def create_training_pipeline_custom_job_sample(
     project: str,
     location: str,
+    staging_bucket: str,
     display_name: str,
     script_path: str,
     container_uri: str,
     model_serving_container_image_uri: str,
+    dataset_id: Optional[str] = None,
     model_display_name: Optional[str] = None,
     args: Optional[List[Union[str, float, int]]] = None,
     replica_count: int = 0,
@@ -36,7 +38,7 @@ def create_training_pipeline_custom_job_sample(
     test_fraction_split: float = 0.1,
     sync: bool = True,
 ):
-    aiplatform.init(project=project, location=location)
+    aiplatform.init(project=project, location=location, staging_bucket=staging_bucket)
 
     job = aiplatform.CustomTrainingJob(
         display_name=display_name,
@@ -45,7 +47,11 @@ def create_training_pipeline_custom_job_sample(
         model_serving_container_image_uri=model_serving_container_image_uri,
     )
 
+    # This example uses an ImageDataset, but you can use another type
+    dataset = aiplatform.ImageDataset(dataset_id) if dataset_id else None
+
     model = job.run(
+        dataset=dataset,
         model_display_name=model_display_name,
         args=args,
         replica_count=replica_count,

samples/model-builder/create_training_pipeline_custom_job_test.py

@@ -18,17 +18,23 @@
 
 
 def test_create_training_pipeline_custom_job_sample(
-    mock_sdk_init, mock_init_custom_training_job, mock_run_custom_training_job,
+    mock_sdk_init,
+    mock_image_dataset,
+    mock_get_image_dataset,
+    mock_get_custom_training_job,
+    mock_run_custom_training_job,
 ):
 
     create_training_pipeline_custom_job_sample.create_training_pipeline_custom_job_sample(
         project=constants.PROJECT,
         location=constants.LOCATION,
+        staging_bucket=constants.STAGING_BUCKET,
         display_name=constants.DISPLAY_NAME,
         args=constants.ARGS,
         script_path=constants.SCRIPT_PATH,
         container_uri=constants.CONTAINER_URI,
         model_serving_container_image_uri=constants.CONTAINER_URI,
+        dataset_id=constants.RESOURCE_ID,
         model_display_name=constants.DISPLAY_NAME_2,
         replica_count=constants.REPLICA_COUNT,
         machine_type=constants.MACHINE_TYPE,
@@ -40,15 +46,18 @@ def test_create_training_pipeline_custom_job_sample(
     )
 
     mock_sdk_init.assert_called_once_with(
-        project=constants.PROJECT, location=constants.LOCATION
+        project=constants.PROJECT,
+        location=constants.LOCATION,
+        staging_bucket=constants.STAGING_BUCKET,
     )
-    mock_init_custom_training_job.assert_called_once_with(
+    mock_get_custom_training_job.assert_called_once_with(
         display_name=constants.DISPLAY_NAME,
         script_path=constants.SCRIPT_PATH,
         container_uri=constants.CONTAINER_URI,
         model_serving_container_image_uri=constants.CONTAINER_URI,
     )
     mock_run_custom_training_job.assert_called_once_with(
+        dataset=mock_image_dataset,
         model_display_name=constants.DISPLAY_NAME_2,
         replica_count=constants.REPLICA_COUNT,
         machine_type=constants.MACHINE_TYPE,

samples/model-builder/create_training_pipeline_custom_package_job_sample.py

@@ -0,0 +1,77 @@
+# Copyright 2021 Google LLC
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     https://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+from typing import List, Optional, Union
+
+from google.cloud import aiplatform
+
+
+#  [START aiplatform_sdk_create_training_pipeline_custom_package_job_sample]
+def create_training_pipeline_custom_package_job_sample(
+    project: str,
+    location: str,
+    staging_bucket: str,
+    display_name: str,
+    python_package_gcs_uri: str,
+    python_module_name: str,
+    container_uri: str,
+    model_serving_container_image_uri: str,
+    dataset_id: Optional[str] = None,
+    model_display_name: Optional[str] = None,
+    args: Optional[List[Union[str, float, int]]] = None,
+    replica_count: int = 1,
+    machine_type: str = "n1-standard-4",
+    accelerator_type: str = "ACCELERATOR_TYPE_UNSPECIFIED",
+    accelerator_count: int = 0,
+    training_fraction_split: float = 0.8,
+    validation_fraction_split: float = 0.1,
+    test_fraction_split: float = 0.1,
+    sync: bool = True,
+):
+    aiplatform.init(project=project, location=location, staging_bucket=staging_bucket)
+
+    job = aiplatform.CustomPythonPackageTrainingJob(
+        display_name=display_name,
+        python_package_gcs_uri=python_package_gcs_uri,
+        python_module_name=python_module_name,
+        container_uri=container_uri,
+        model_serving_container_image_uri=model_serving_container_image_uri,
+    )
+
+    # This example uses an ImageDataset, but you can use another type
+    dataset = aiplatform.ImageDataset(dataset_id) if dataset_id else None
+
+    model = job.run(
+        dataset=dataset,
+        model_display_name=model_display_name,
+        args=args,
+        replica_count=replica_count,
+        machine_type=machine_type,
+        accelerator_type=accelerator_type,
+        accelerator_count=accelerator_count,
+        training_fraction_split=training_fraction_split,
+        validation_fraction_split=validation_fraction_split,
+        test_fraction_split=test_fraction_split,
+        sync=sync,
+    )
+
+    model.wait()
+
+    print(model.display_name)
+    print(model.resource_name)
+    print(model.uri)
+    return model
+
+
+#  [END aiplatform_sdk_create_training_pipeline_custom_package_job_sample]