Opik Agent Optimizer API Reference

1
def optimize_prompt(
2
    self,
3
    prompt: ChatPrompt,
4
    dataset: Dataset,
5
    metric: Callable,
6
    experiment_config: dict | None = None,
7
    n_samples: int | None = None,
8
    auto_continue: bool = False,
9
    agent_class: type[OptimizableAgent] | None = None,
10
    **kwargs: Any,
11
) -> OptimizationResult

1
def optimize_mcp(
2
    self,
3
    prompt: ChatPrompt,
4
    dataset: Dataset,
5
    metric: Callable,
6
    *,
7
    tool_name: str,
8
    second_pass: Any,
9
    experiment_config: dict | None = None,
10
    n_samples: int | None = None,
11
    auto_continue: bool = False,
12
    agent_class: type[OptimizableAgent] | None = None,
13
    fallback_invoker: Callable[[dict[str, Any]], str] | None = None,
14
    fallback_arguments: Callable[[Any], dict[str, Any]] | None = None,
15
    allow_tool_use_on_second_pass: bool = False,
16
    **kwargs: Any,
17
) -> OptimizationResult

1
FewShotBayesianOptimizer(
2
    model: str,
3
    min_examples: int = 2,
4
    max_examples: int = 8,
5
    seed: int = 42,
6
    n_threads: int = 8,
7
    verbose: int = 1,
8
    model_kwargs: Any
9
)

1
cleanup()

1
configure_prompt_model(
2
    prompt: ChatPrompt
3
)

1
create_optimization_context(
2
    dataset: Dataset,
3
    metric: Callable,
4
    metadata: Optional = None
5
)

1
evaluate_prompt(
2
    prompt: ChatPrompt,
3
    dataset: Dataset,
4
    metric: Callable,
5
    n_threads: int,
6
    verbose: int = 1,
7
    dataset_item_ids: Optional = None,
8
    experiment_config: Optional = None,
9
    n_samples: Optional = None,
10
    seed: Optional = None,
11
    agent_class: Optional = None
12
)

1
get_history()

1
increment_llm_counter()

1
increment_tool_counter()

1
optimize_mcp(
2
    prompt: ChatPrompt,
3
    dataset: Dataset,
4
    metric: Callable,
5
    tool_name: str,
6
    second_pass: Any,
7
    experiment_config: Optional = None,
8
    n_samples: Optional = None,
9
    auto_continue: bool = False,
10
    agent_class: Optional = None,
11
    fallback_invoker: Optional = None,
12
    fallback_arguments: Optional = None,
13
    allow_tool_use_on_second_pass: bool = False,
14
    kwargs: Any
15
)

1
optimize_prompt(
2
    prompt: ChatPrompt,
3
    dataset: Dataset,
4
    metric: Callable,
5
    experiment_config: Optional = None,
6
    n_samples: Optional = None,
7
    auto_continue: bool = False,
8
    agent_class: Optional = None,
9
    kwargs: Any
10
)

1
reset_counters()

1
setup_agent_class(
2
    prompt: ChatPrompt,
3
    agent_class: Optional = None
4
)

1
update_optimization(
2
    optimization: Optimization,
3
    status: str
4
)

1
validate_optimization_inputs(
2
    prompt: ChatPrompt,
3
    dataset: Dataset,
4
    metric: Callable
5
)

1
MetaPromptOptimizer(
2
    model: str,
3
    reasoning_model: Optional = None,
4
    rounds: int = 3,
5
    num_prompts_per_round: int = 4,
6
    num_threads: Optional = None,
7
    verbose: int = 1,
8
    enable_context: bool = True,
9
    n_threads: int = 12,
10
    seed: int = 42,
11
    model_kwargs: Any
12
)

1
cleanup()

1
configure_prompt_model(
2
    prompt: ChatPrompt
3
)

1
create_optimization_context(
2
    dataset: Dataset,
3
    metric: Callable,
4
    metadata: Optional = None
5
)

1
evaluate_prompt(
2
    prompt: ChatPrompt,
3
    dataset: Dataset,
4
    metric: Callable,
5
    n_threads: int,
6
    verbose: int = 1,
7
    dataset_item_ids: Optional = None,
8
    experiment_config: Optional = None,
9
    n_samples: Optional = None,
10
    seed: Optional = None,
11
    agent_class: Optional = None
12
)

1
get_history()

1
increment_llm_counter()

1
increment_tool_counter()

1
optimize_mcp(
2
    prompt: ChatPrompt,
3
    dataset: Dataset,
4
    metric: Callable,
5
    tool_name: str,
6
    second_pass: MCPSecondPassCoordinator,
7
    experiment_config: Optional = None,
8
    n_samples: Optional = None,
9
    auto_continue: bool = False,
10
    agent_class: Optional = None,
11
    fallback_invoker: Optional = None,
12
    fallback_arguments: Optional = None,
13
    allow_tool_use_on_second_pass: bool = False,
14
    kwargs: Any
15
)

1
optimize_prompt(
2
    prompt: ChatPrompt,
3
    dataset: Dataset,
4
    metric: Callable,
5
    experiment_config: Optional = None,
6
    n_samples: Optional = None,
7
    auto_continue: bool = False,
8
    agent_class: Optional = None,
9
    kwargs: Any
10
)

1
reset_counters()

1
setup_agent_class(
2
    prompt: ChatPrompt,
3
    agent_class: Optional = None
4
)

1
update_optimization(
2
    optimization: Optimization,
3
    status: str
4
)

1
validate_optimization_inputs(
2
    prompt: ChatPrompt,
3
    dataset: Dataset,
4
    metric: Callable
5
)

1
MiproOptimizer(
2
    model: Any,
3
    project_name: Optional = None,
4
    verbose: int = 1,
5
    model_kwargs: Any
6
)

1
cleanup()

1
configure_prompt_model(
2
    prompt: ChatPrompt
3
)

1
continue_optimize_prompt()

1
create_optimization_context(
2
    dataset: Dataset,
3
    metric: Callable,
4
    metadata: Optional = None
5
)

1
evaluate_prompt(
2
    dataset: str | opik.api_objects.dataset.dataset.Dataset,
3
    metric: Callable,
4
    task_config: TaskConfig,
5
    prompt: Union = None,
6
    n_samples: int = 10,
7
    dataset_item_ids: Optional = None,
8
    experiment_config: Optional = None,
9
    verbose: int = 1,
10
    kwargs: Any
11
)

1
get_best(
2
    position: int = 0
3
)

1
get_history()

1
increment_llm_counter()

1
increment_tool_counter()

1
load_from_checkpoint(
2
    filename: Any
3
)

1
optimize_mcp(
2
    prompt: ChatPrompt,
3
    dataset: Dataset,
4
    metric: Callable,
5
    tool_name: str,
6
    second_pass: Any,
7
    experiment_config: Optional = None,
8
    n_samples: Optional = None,
9
    auto_continue: bool = False,
10
    agent_class: Optional = None,
11
    fallback_invoker: Optional = None,
12
    fallback_arguments: Optional = None,
13
    allow_tool_use_on_second_pass: bool = False,
14
    kwargs: Any
15
)

1
optimize_prompt(
2
    prompt: ChatPrompt,
3
    dataset: Dataset,
4
    metric: Callable,
5
    experiment_config: Optional = None,
6
    n_samples: int | None = 10,
7
    auto_continue: bool = False,
8
    agent_class: Optional = None,
9
    kwargs: Any
10
)

1
prepare_optimize_prompt(
2
    dataset: Any,
3
    metric: Any,
4
    task_config: Any,
5
    num_candidates: int = 10,
6
    experiment_config: Optional = None,
7
    optimization_id: Optional = None,
8
    num_trials: int | None = 3,
9
    n_samples: int | None = 10,
10
    auto: Optional = 'light',
11
    kwargs: Any
12
)

1
reset_counters()

1
setup_agent_class(
2
    prompt: ChatPrompt,
3
    agent_class: Optional = None
4
)

1
update_optimization(
2
    optimization: Optimization,
3
    status: str
4
)

1
validate_optimization_inputs(
2
    prompt: ChatPrompt,
3
    dataset: Dataset,
4
    metric: Callable
5
)

1
EvolutionaryOptimizer(
2
    model: str,
3
    population_size: int = 30,
4
    num_generations: int = 15,
5
    mutation_rate: float = 0.2,
6
    crossover_rate: float = 0.8,
7
    tournament_size: int = 4,
8
    num_threads: Optional = None,
9
    elitism_size: int = 3,
10
    adaptive_mutation: bool = True,
11
    enable_moo: bool = True,
12
    enable_llm_crossover: bool = True,
13
    seed: int | None = 42,
14
    output_style_guidance: Optional = None,
15
    infer_output_style: bool = False,
16
    verbose: int = 1,
17
    n_threads: int = 12,
18
    model_kwargs: Any
19
)

1
cleanup()

1
configure_prompt_model(
2
    prompt: ChatPrompt
3
)

1
create_optimization_context(
2
    dataset: Dataset,
3
    metric: Callable,
4
    metadata: Optional = None
5
)

1
evaluate_prompt(
2
    prompt: ChatPrompt,
3
    dataset: Dataset,
4
    metric: Callable,
5
    n_threads: int,
6
    verbose: int = 1,
7
    dataset_item_ids: Optional = None,
8
    experiment_config: Optional = None,
9
    n_samples: Optional = None,
10
    seed: Optional = None,
11
    agent_class: Optional = None
12
)

1
get_history()

1
get_llm_crossover_system_prompt()

1
increment_llm_counter()

1
increment_tool_counter()

1
optimize_mcp(
2
    prompt: ChatPrompt,
3
    dataset: Dataset,
4
    metric: Callable,
5
    tool_name: str,
6
    second_pass: Any,
7
    experiment_config: Optional = None,
8
    n_samples: Optional = None,
9
    auto_continue: bool = False,
10
    agent_class: Optional = None,
11
    fallback_invoker: Optional = None,
12
    fallback_arguments: Optional = None,
13
    allow_tool_use_on_second_pass: bool = False,
14
    kwargs: Any
15
)

1
optimize_prompt(
2
    prompt: ChatPrompt,
3
    dataset: Dataset,
4
    metric: Callable,
5
    experiment_config: Optional = None,
6
    n_samples: Optional = None,
7
    auto_continue: bool = False,
8
    agent_class: Optional = None,
9
    kwargs: Any
10
)

1
reset_counters()

1
setup_agent_class(
2
    prompt: ChatPrompt,
3
    agent_class: Optional = None
4
)

1
update_optimization(
2
    optimization: Optimization,
3
    status: str
4
)

1
validate_optimization_inputs(
2
    prompt: ChatPrompt,
3
    dataset: Dataset,
4
    metric: Callable
5
)

1
GepaOptimizer(
2
    model: str,
3
    project_name: Optional = None,
4
    reflection_model: Optional = None,
5
    verbose: int = 1,
6
    seed: int = 42,
7
    model_kwargs: Any
8
)

1
cleanup()

1
configure_prompt_model(
2
    prompt: ChatPrompt
3
)

1
create_optimization_context(
2
    dataset: Dataset,
3
    metric: Callable,
4
    metadata: Optional = None
5
)

1
evaluate_prompt(
2
    prompt: ChatPrompt,
3
    dataset: Dataset,
4
    metric: Callable,
5
    n_threads: int,
6
    verbose: int = 1,
7
    dataset_item_ids: Optional = None,
8
    experiment_config: Optional = None,
9
    n_samples: Optional = None,
10
    seed: Optional = None,
11
    agent_class: Optional = None
12
)

1
get_history()

1
increment_llm_counter()

1
increment_tool_counter()

1
optimize_mcp(
2
    prompt: ChatPrompt,
3
    dataset: Dataset,
4
    metric: Callable,
5
    tool_name: str,
6
    second_pass: Any,
7
    experiment_config: Optional = None,
8
    n_samples: Optional = None,
9
    auto_continue: bool = False,
10
    agent_class: Optional = None,
11
    fallback_invoker: Optional = None,
12
    fallback_arguments: Optional = None,
13
    allow_tool_use_on_second_pass: bool = False,
14
    kwargs: Any
15
)

1
optimize_prompt(
2
    prompt: ChatPrompt,
3
    dataset: Dataset,
4
    metric: Callable,
5
    experiment_config: Optional = None,
6
    n_samples: Optional = None,
7
    auto_continue: bool = False,
8
    agent_class: Optional = None,
9
    kwargs: Any
10
)

1
reset_counters()

1
setup_agent_class(
2
    prompt: ChatPrompt,
3
    agent_class: Optional = None
4
)

1
update_optimization(
2
    optimization: Optimization,
3
    status: str
4
)

1
validate_optimization_inputs(
2
    prompt: ChatPrompt,
3
    dataset: Dataset,
4
    metric: Callable
5
)

1
ChatPrompt(
2
    name: str = 'chat-prompt',
3
    system: Optional = None,
4
    user: Optional = None,
5
    messages: Optional = None,
6
    tools: Optional = None,
7
    function_map: Optional = None,
8
    model: Optional = None,
9
    invoke: Optional = None,
10
    project_name: str | None = 'Default Project',
11
    model_kwargs: Any
12
)

1
copy()

1
get_messages(
2
    dataset_item: Optional = None
3
)

1
set_messages(
2
    messages: list
3
)

1
to_dict()

1
with_messages(
2
    messages: list
3
)

1
OptimizationResult(
2
    optimizer: <class 'str'> = 'Optimizer',
3
    prompt: list[dict[str, str]],
4
    score: <class 'float'>,
5
    metric_name: <class 'str'>,
6
    optimization_id: str | None = None,
7
    dataset_id: str | None = None,
8
    initial_prompt: list[dict[str, str]] | None = None,
9
    initial_score: float | None = None,
10
    details: dict[str, Any] = PydanticUndefined,
11
    history: list[dict[str, Any]] = [],
12
    llm_calls: int | None = None,
13
    demonstrations: list[dict[str, Any]] | None = None,
14
    mipro_prompt: str | None = None,
15
    tool_prompts: dict[str, str] | None = None
16
)

1
OptimizableAgent(
2
    prompt: Any
3
)

1
init_agent(
2
    prompt: Any
3
)

1
init_llm()

1
invoke(
2
    messages: list,
3
    seed: Optional = None
4
)

1
invoke_dataset_item(
2
    dataset_item: dict
3
)

1
llm_invoke(
2
    query: Optional = None,
3
    messages: Optional = None,
4
    seed: Optional = None,
5
    allow_tool_use: bool | None = False
6
)