fix: enable caching default behavior (#177)

deployer/cli.py

@@ -203,11 +203,17 @@ def deploy(  # noqa: C901
         ),
     ] = None,
     enable_caching: Annotated[
-        bool,
+        Optional[bool],
         typer.Option(
-            "--enable-caching", "-ec", help="Whether to enable caching when running the pipeline."
+            "--enable-caching / --no-cache",
+            "-ec / -nec",
+            help="Whether to turn on caching for the run."
+            "If this is not set, defaults to the compile time settings, which are True for all"
+            "tasks by default, while users may specify different caching options for individual"
+            "tasks. If this is set, the setting applies to all tasks in the pipeline."
+            "Overrides the compile time settings. Defaults to None.",
         ),
-    ] = False,
+    ] = None,
     experiment_name: Annotated[
         Optional[str],
         typer.Option(

deployer/pipeline_deployer.py

@@ -110,10 +110,34 @@ def _check_experiment_name(self, experiment_name: Optional[str] = None) -> str:
     def _create_pipeline_job(
         self,
         template_path: str,
-        enable_caching: bool = False,
+        enable_caching: Optional[bool] = None,
         parameter_values: Optional[dict] = None,
         input_artifacts: Optional[dict] = None,
     ) -> aiplatform.PipelineJob:
+        """Create a pipeline job object
+
+        Args:
+            template_path (str): The path of PipelineJob or PipelineSpec JSON or YAML file. If the
+                Artifact Registry host is provided, this is the path to the pipeline template in
+                the Artifact Registry. Otherwise, this is the path to the pipeline template in
+                the local package.
+            enable_caching (Optional[bool], optional): Whether to turn on caching for the run.
+                If this is not set, defaults to the compile time settings, which are True for all
+                tasks by default, while users may specify different caching options for individual
+                tasks.
+                If this is set, the setting applies to all tasks in the pipeline.
+                Overrides the compile time settings. Defaults to None.
+            parameter_values (Optional[dict], optional): The mapping from runtime parameter names
+                to its values that control the pipeline run. Defaults to None.
+            input_artifacts (Optional[dict], optional): The mapping from the runtime parameter
+                name for this artifact to its resource id.
+                For example: "vertex_model":"456".
+                Note: full resource name ("projects/123/locations/us-central1/metadataStores/default/artifacts/456")
+                    cannot be used. Defaults to None.
+
+        Returns:
+            aiplatform.PipelineJob: The pipeline job object
+        """  # noqa: E501
         job = aiplatform.PipelineJob(
             display_name=self.pipeline_name,
             template_path=template_path,
@@ -156,7 +180,7 @@ def upload_to_registry(
 
     def run(
         self,
-        enable_caching: bool = False,
+        enable_caching: Optional[bool] = None,
         parameter_values: Optional[dict] = None,
         input_artifacts: Optional[dict] = None,
         experiment_name: Optional[str] = None,
@@ -170,16 +194,33 @@ def run(
         provided. Otherwise, use the pipeline file in the local package.
 
         Args:
-            enable_caching (bool, optional): Whether to enable caching. Defaults to False.
-            parameter_values (dict, optional): Pipeline parameter values. Defaults to None.
-            input_artifacts (dict, optional): Pipeline input artifacts. Defaults to None.
+            enable_caching (Optional[bool], optional): Whether to turn on caching for the run.
+                If this is not set, defaults to the compile time settings, which are True for all
+                tasks by default, while users may specify different caching options for individual
+                tasks.
+                If this is set, the setting applies to all tasks in the pipeline.
+                Overrides the compile time settings. Defaults to None.
+            parameter_values (Optional[dict], optional): The mapping from runtime parameter names
+                to its values that control the pipeline run. Defaults to None.
+            input_artifacts (Optional[dict], optional): The mapping from the runtime parameter
+                name for this artifact to its resource id.
+                For example: "vertex_model":"456".
+                Note: full resource name ("projects/123/locations/us-central1/metadataStores/default/artifacts/456")
+                    cannot be used. Defaults to None.
             experiment_name (str, optional): Experiment name. Defaults to None.
             tag (str, optional): Tag of the pipeline template. Defaults to None.
-        """
+        """  # noqa: E501
         experiment_name = self._check_experiment_name(experiment_name)
 
         template_path = self._get_template_path(tag)
 
+        logger.debug(
+            f"Running pipeline '{self.pipeline_name}' with settings:"
+            f"\n {'template_path':<20} {template_path:<30}"
+            f"\n {'enable_caching':<20} {enable_caching!s:<30}"
+            f"\n {'experiment_name':<20} {experiment_name:<30}"
+        )
+
         job = self._create_pipeline_job(
             template_path=template_path,
             enable_caching=enable_caching,
@@ -207,7 +248,7 @@ def run(
 
     def compile_upload_run(
         self,
-        enable_caching: bool = False,
+        enable_caching: Optional[bool] = None,
         parameter_values: Optional[dict] = None,
         experiment_name: Optional[str] = None,
         tags: Optional[List[str]] = None,
@@ -229,7 +270,7 @@ def compile_upload_run(
     def schedule(
         self,
         cron: str,
-        enable_caching: bool = False,
+        enable_caching: Optional[bool] = None,
         parameter_values: Optional[dict] = None,
         tag: Optional[str] = None,
         delete_last_schedule: bool = False,

deployer/settings.py

@@ -28,7 +28,7 @@ class _DeployerDeploySettings(CustomBaseModel):
     tags: Optional[List[str]] = constants.DEFAULT_TAGS
     config_filepath: Optional[Path] = None
     config_name: Optional[str] = None
-    enable_caching: bool = False
+    enable_caching: Optional[bool] = None
     experiment_name: Optional[str] = None
     local_package_path: Path = constants.DEFAULT_LOCAL_PACKAGE_PATH
     skip_validation: bool = True

docs/CLI_REFERENCE.md

@@ -121,7 +121,7 @@ $ vertex-deployer deploy [OPTIONS] PIPELINE_NAMES...
 * `--tags TEXT`: The tags to use when uploading the pipeline.
 * `-cfp, --config-filepath FILE`: Path to the json/py file with parameter values and input artifacts to use when running the pipeline.
 * `-cn, --config-name TEXT`: Name of the json/py file with parameter values and input artifacts to use when running the pipeline. It must be in the pipeline config dir. e.g. `config_dev.json` for `./vertex/configs/{pipeline-name}/config_dev.json`.
-* `-ec, --enable-caching`: Whether to enable caching when running the pipeline.
+* `-ec, --enable-caching / -nec, --no-cache`: Whether to turn on caching for the run.If this is not set, defaults to the compile time settings, which are True for alltasks by default, while users may specify different caching options for individualtasks. If this is set, the setting applies to all tasks in the pipeline.Overrides the compile time settings. Defaults to None.
 * `-en, --experiment-name TEXT`: The name of the experiment to run the pipeline in.Defaults to '{pipeline_name}-experiment'.
 * `-lpp, --local-package-path DIRECTORY`: Local dir path where pipelines will be compiled.  [default: vertex/pipelines/compiled_pipelines]
 * `-y, --skip-validation / -n, --no-skip`: Whether to continue without user validation of the settings.  [default: skip-validation]