Used to ingest training metrics into SageMaker. These metrics can be visualized in SageMaker Studio.
Parameters:
-
trial_component_name
(Optional[ExperimentEntityName], default:
None
)
–
The name of the Trial Component to associate with the metrics. The Trial Component name must be entirely lowercase.
-
metric_data
(Optional[RawMetricDataList], default:
None
)
–
A list of raw metric values to put.
Raises:
Source code in docs_src/capo_sagemaker_metrics/_services/sage_maker_metrics.py
| def batch_put_metrics(
self,
*,
config_overrides: Optional[SageMakerMetricsClientConfig] = None,
trial_component_name: Optional[
"capo_sagemaker_metrics.types.experiment_entity_name.ExperimentEntityName"
] = None,
metric_data: Optional[
"capo_sagemaker_metrics.types.raw_metric_data_list.RawMetricDataList"
] = None,
) -> "capo_sagemaker_metrics.types.batch_put_metrics_response.BatchPutMetricsResponse":
"""<p>Used to ingest training metrics into SageMaker. These metrics can be visualized in SageMaker Studio. </p>
Args:
trial_component_name: <p>The name of the Trial Component to associate with the metrics. The Trial Component name must be entirely lowercase.</p>
metric_data: <p>A list of raw metric values to put.</p>
Raises:
capo_sagemaker_metrics.errors.UnknownServiceError: The service returned an error code this client does not model.
"""
def _handler(
req: "OperationRequest[capo_sagemaker_metrics.types.batch_put_metrics_request.BatchPutMetricsRequest]",
) -> OperationResponse[
"capo_sagemaker_metrics.types.batch_put_metrics_response.BatchPutMetricsResponse"
]:
import capo_sagemaker_metrics._operations.sage_maker_metrics_service.batch_put_metrics
output, http_response = (
capo_sagemaker_metrics._operations.sage_maker_metrics_service.batch_put_metrics.batch_put_metrics(
req.options, req.input
)
)
return OperationResponse(output=output, response=http_response)
interceptors_, options_ = self.operation_options(
config_overrides, OperationLogging("BatchPutMetrics")
)
input_: capo_sagemaker_metrics.types.batch_put_metrics_request.BatchPutMetricsRequest = {}
if trial_component_name is not None:
input_["trial_component_name"] = trial_component_name
if metric_data is not None:
input_["metric_data"] = metric_data
response = execute_pipeline(
OperationRequest(input=input_, options=options_),
handler=_handler,
interceptors=list(interceptors_),
)
response.response.close()
return response.output
|