tempo.insights.manager module

class tempo.insights.manager.InsightsManager(worker_endpoint: str = '', batch_size: int = 1, parallelism: int = 1, retries: int = 3, window_time: int = None, in_asyncio: bool = False, mode_type: tempo.serve.metadata.InsightRequestModes = None)

Bases: object

log(data, insights_type: tempo.serve.metadata.InsightsTypes = <InsightsTypes.CUSTOM_INSIGHT: 'io.seldon.serving.inference.custominsight'>)

By default function doesn’t have any logic unless an endpoint is provided.

log_request()

By default function doesn’t have any logic unless an endpoint is provided.

log_response()

By default function doesn’t have any logic unless an endpoint is provided.