Evaluations

Run Eval

Evaluate a model against a eval spice specification

post/v1/evals/{name}

Path parameters

namestring required

Name of the evaluation to run

modelstring required

Evaluation run successfully

datasetstring required

The name of the dataset used for the evaluation

error_messagestring nullable

The error message if the evaluation failed, otherwise None

metricsobject required

A map of metric names to their corresponding values

modelstring required

The model used for the evaluation

primary_keystring required

Unique identifier for the evaluation run

scorersstring[] required

List of scorers used in the evaluation

statusstring required

The status of the evaluation (e.g., "completed", "failed", etc.)

time_columnstring required

Timestamp indicating when the evaluation was created or run