def _log(self, artifact):
"""
Log an artifact.
This method expects `artifact` to have
- ARTIFACT_TYPE_NAME string field the form of
<namespace>/<name>.
- serialization() method to return a openapi_client.MlMetadataArtifact.
This method will set artifact.id.
"""
serialization = artifact.serialization()
if serialization.custom_properties is None:
serialization.custom_properties = {}
if WORKSPACE_PROPERTY_NAME in serialization.custom_properties:
raise ValueError("custom_properties contains reserved key %s"
% WORKSPACE_PROPERTY_NAME)
if RUN_PROPERTY_NAME in serialization.custom_properties:
raise ValueError("custom_properties contains reserved key %s"
% RUN_PROPERTY_NAME)
if self.workspace is not None:
serialization.custom_properties[
WORKSPACE_PROPERTY_NAME] = openapi_client.MlMetadataValue(
string_value=self.workspace.name)
if self.run is not None:
serialization.custom_properties[
RUN_PROPERTY_NAME] = openapi_client.MlMetadataValue(
string_value=self.run.name)
response = self.workspace.client.create_artifact(
parent=artifact.ARTIFACT_TYPE_NAME,
body=serialization,
)
artifact.id = response.artifact.id
return artifact
def log(self, artifact):
"""
Log an artifact as an input or output of this run to
metadata backend serivce.
This method expects `artifact` to have
- ARTIFACT_TYPE_NAME string field the form of
<namespace>/<name>.
- serialization() method to return a openapi_client.MlMetadataArtifact.
This method will set artifact.id.
"""
# TODO(zhenghui): log this artifact as the input or output of an execution.
serialization = artifact.serialization()
if serialization.custom_properties == None:
serialization.custom_properties = {}
if WORKSPACE_PROPERTY_NAME in serialization.custom_properties:
raise ValueError("custom_properties contains reserved key %s" %
WORKSPACE_PROPERTY_NAME)
if RUN_PROPERTY_NAME in serialization.custom_properties:
raise ValueError("custom_properties contains reserved key %s" %
RUN_PROPERTY_NAME)
serialization.custom_properties[
WORKSPACE_PROPERTY_NAME] = openapi_client.MlMetadataValue(
string_value=self.workspace.name)
serialization.custom_properties[
RUN_PROPERTY_NAME] = openapi_client.MlMetadataValue(
string_value=self.name)
response = self.workspace._client.create_artifact(
parent=artifact.ARTIFACT_TYPE_NAME,
body=serialization,
)
artifact.id = response.artifact.id
return artifact
def test_log_invalid_artifacts_should_fail(self):
r = metadata.Run(name="test run")
artifact1 = ArtifactFixture(
openapi_client.MlMetadataArtifact(
uri="gs://uri",
custom_properties={
metadata.WORKSPACE_PROPERTY_NAME:
openapi_client.MlMetadataValue(string_value="ws1"),
}))
self.assertRaises(ValueError, r.log, artifact1)
artifact2 = ArtifactFixture(
openapi_client.MlMetadataArtifact(
uri="gs://uri",
custom_properties={
metadata.RUN_PROPERTY_NAME:
openapi_client.MlMetadataValue(string_value="run1"),
}))
self.assertRaises(ValueError, r.log, artifact2)
def test_log_invalid_artifacts_should_fail(self):
ws = metadata.Workspace(backend_url_prefix="127.0.0.1:8080",
name="ws_1",
description="a workspace for testing",
labels={"n1": "v1"})
e = metadata.Execution(name="test execution", workspace=ws)
artifact1 = ArtifactFixture(
openapi_client.MlMetadataArtifact(
uri="gs://uri",
custom_properties={
metadata.WORKSPACE_PROPERTY_NAME:
openapi_client.MlMetadataValue(string_value="ws1"),
}))
self.assertRaises(ValueError, e.log_input, artifact1)
artifact2 = ArtifactFixture(
openapi_client.MlMetadataArtifact(
uri="gs://uri",
custom_properties={
metadata.RUN_PROPERTY_NAME:
openapi_client.MlMetadataValue(string_value="run1"),
}))
self.assertRaises(ValueError, e.log_output, artifact2)
def serialized(self):
execution = openapi_client.MlMetadataExecution(
properties={
"name":
openapi_client.MlMetadataValue(string_value=self.name),
"create_time":
openapi_client.MlMetadataValue(string_value=self.create_time),
"description":
openapi_client.MlMetadataValue(string_value=self.description),
})
if self.description is None:
del execution.properties["description"]
execution.custom_properties = {}
if self.workspace is not None:
execution.custom_properties[
WORKSPACE_PROPERTY_NAME] = openapi_client.MlMetadataValue(
string_value=self.workspace.name)
if self.run is not None:
execution.custom_properties[
RUN_PROPERTY_NAME] = openapi_client.MlMetadataValue(
string_value=self.run.name)
return execution
def serialization(self):
model_artifact = openapi_client.MlMetadataArtifact(
uri=self.uri,
properties={
"name":
openapi_client.MlMetadataValue(string_value=self.name),
"create_time":
openapi_client.MlMetadataValue(string_value=self.create_time),
"description":
openapi_client.MlMetadataValue(string_value=self.description),
"metrics_type":
openapi_client.MlMetadataValue(string_value=self.metrics_type),
"data_set_id":
openapi_client.MlMetadataValue(string_value=self.data_set_id),
"model_id":
openapi_client.MlMetadataValue(string_value=self.model_id),
"owner":
openapi_client.MlMetadataValue(string_value=self.owner),
ALL_META_PROPERTY_NAME:
openapi_client.MlMetadataValue(string_value=json.dumps(self.__dict__)),
})
return model_artifact
def serialization(self):
data_set_artifact = openapi_client.MlMetadataArtifact(
uri=self.uri,
properties={
"name":
openapi_client.MlMetadataValue(string_value=self.name),
"create_time":
openapi_client.MlMetadataValue(string_value=self.create_time),
"description":
openapi_client.MlMetadataValue(string_value=self.description),
"query":
openapi_client.MlMetadataValue(string_value=self.query),
"version":
openapi_client.MlMetadataValue(string_value=self.version),
"owner":
openapi_client.MlMetadataValue(string_value=self.owner),
ALL_META_PROPERTY_NAME:
openapi_client.MlMetadataValue(string_value=json.dumps(self.__dict__)),
})
return data_set_artifact
