aevaluate_run(
self,
run: Union[V2Run, ls_schemas.Run, ls_schemas.RunBase, str, uuid.| Name | Type | Description |
|---|---|---|
run* | Union[V2Run, Run, RunBase, str, UUID] | The run to evaluate. Passing |
evaluator* | RunEvaluator | The evaluator to use. |
project_id | Optional[Union[UUID, str]] | Default: None |
start_time | Optional[datetime] | Default: None |
source_info | Optional[Dict[str, Any]] | Default: None |
reference_example | Optional[Union[Example, str, dict, UUID]] | Default: None |
load_child_runs | bool, default=False | Default: False |
Evaluate a run asynchronously.
.. admonition:: Deprecated
There is no replacement. Run the evaluator yourself and log the
result with :meth:`langsmith.Client.create_feedback`.
Will be removed after Jan 31, 2027.
The ID of the project (session) that owns the run. Omitting it is deprecated and will raise in a future release.
The run's start time, passed to runs.retrieve when resolving
run from an ID. Required when resolving a run from an ID on
SmithDB-only backends (no ClickHouse query support).
Additional information about the source of the evaluation to log as feedback metadata.
The example to use as a reference for the evaluation. If not provided, the run's reference example will be used.
Whether to load child runs when resolving the run ID.
Deprecated: see read_run.