MlRun
Type: Object
Telemetry-rich execution record for ML workflows
Fields
| Name | Type | Args | Description | Deprecated |
|---|---|---|---|---|
| canceledAt | DateTime | When the run was explicitly canceled | ||
| completedAt | DateTime | When the run finished successfully | ||
| context | JsonString! | Arbitrary context stored for Reactor steps | ||
| datasetReport | JsonString! | Dataset quality information captured during assembly | ||
| errorDetails | JsonString! | Structured error payload | ||
| externalId | String | Optional cross-system identifier for the run | ||
| id | ID! | |||
| initiatedBy | String | Human-friendly explanation of what triggered the run | ||
| initiatorMetadata | JsonString! | Structured metadata describing the run initiator | ||
| insertedAt | DateTime! | When the run record was created | ||
| kind | String! | Category of work performed by this run | ||
| metrics | JsonString! | Captured metrics for the run | ||
| modelVersion | MlModelVersion | |||
| modelVersionId | ID | Associated model version id, populated for training/evaluation runs | ||
| question | MlQuestion! | |||
| questionId | ID! | Owning ML question id | ||
| startedAt | DateTime | When the run started | ||
| status | String | Current execution status | ||
| updatedAt | DateTime! | When the run last changed | ||
| validationReport | JsonString! | Evaluation or QA summary |