OutputRails

class nemo_microservices.types.shared.OutputRails(*args: Any, **kwargs: Any)

Bases: BaseModel

apply_to_reasoning_traces: bool | None = None

If True, output rails will apply guardrails to both reasoning traces and output response. If False, output rails will only apply guardrails to the output response excluding the reasoning traces, thus keeping reasoning traces unaltered.

flows: List[str] | None = None

The names of all the flows that implement output rails.

streaming: OutputRailsStreamingConfig | None = None

Configuration for managing streaming output of LLM tokens.