ExperimentalOptionalconcurrency?: numberThe number of dataset examples to run in parallel
The dataset to run the experiment on
OptionaldiagLogLevel?: DiagLogLevelLog level to set for the default DiagConsoleLogger when tracing. Omit to disable default diag logging, or to bring your own.
OptionaldryRun?: number | booleanWhether or not to run the experiment as a dry run. If a number is privided, n examples will be run.
Optionalevaluators?: Evaluator[]The evaluators to use
OptionalexperimentDescription?: stringThe description of the experiment
OptionalexperimentMetadata?: Record<string, unknown>Experiment metadata E.x. modelName
OptionalexperimentName?: stringAn optional name for the experiment. Defaults to the dataset name + a timestamp
Optionallogger?: LoggerThe logger to use
Optionalrecord?: booleanWhether to record the experiment results
Optionalrepetitions?: numberNumber of times to repeat each dataset example
OptionalsetGlobalTracerProvider?: booleanWhether to set the global tracer provider when running the task. If set to false, a locally scoped tracer will be created but will not get registered. This may cause certain spans to not be picked up by Phoenix. Notably libraries like the AI SDK that leverage the global tracer.
The task to run
OptionaluseBatchSpanProcessor?: boolean
Parameters for running an experiment.
This feature is not complete, and will change in the future.