Method optimize

src/google/adk/optimization/agent_optimizer.py:32–49

Runs the optimizer. Args: initial_agent: The initial agent to be optimized. sampler: The interface used to get training and validation example UIDs, request agent evaluations, and get useful data for optimizing the agent. Returns: The final result of the optimizat

(
      self,
      initial_agent: Agent,
      sampler: Sampler[SamplingResultT],
  )

  @abstractmethod
  async def optimize(
      self,
      initial_agent: Agent,
      sampler: Sampler[SamplingResultT],
  ) -> OptimizerResult[AgentWithScoresT]:
    """Runs the optimizer.

    Args:
      initial_agent: The initial agent to be optimized.
      sampler: The interface used to get training and validation example UIDs,
        request agent evaluations, and get useful data for optimizing the agent.

    Returns:
      The final result of the optimization process, containing the optimized
      agent instances along with their corresponding scores on the validation
      examples and any optimization metadata.
    """
    ...