snorkelai.sdk.client.evaluation.create_evaluation_report

warning

This SDK function is not supported in Snorkel AI Data Development Platform v25.5 and later versions and will be removed in a future release. For assistance finding alternative approaches, please contact your Snorkel support team.

snorkelai.sdk.client.evaluation.create_evaluation_report(dataset, metric_schemas, split=None, slices=None, models=None)

Start a job to compute metrics for a dataset split.

Parameters

Parameters

Name	Type	Default	Info
dataset	Union[str, int]		Name or UID of the dataset to load unlabeled dataset from.
metric_schemas	List[MetricSchema]		Create built-in MetricSchema objects.
split	Optional[str]	None	[DEPRECATED] Name of the split to evaluate.
slices	Optional[List[int]]	None	UID of the slices for which to compute metrics If no slices are provide, all slices in the given dataset will be evaluated.
models	Optional[List[int]]	None	UID of the models for which to compute metrics If no models are provided, metrics for all models will be computed All models must map to sources associated with datasources in the given dataset.

Returns

Returns

A dictionary containing the evaluation results

Return type

Return type

Dict[str, Any]