Module for orchestration service handling requests and responses.
 
Provides synchronous and asynchronous methods to run orchestration pipelines.

Modules
		asyncio dacite httpx logging random time

Classes
		builtins.object OrchestrationService gen_ai_hub.orchestration.models.base.JSONSerializable(abc.ABC) OrchestrationRequest   class OrchestrationRequest(gen_ai_hub.orchestration.models.base.JSONSerializable)     OrchestrationRequest(config: gen_ai_hub.orchestration.models.config.OrchestrationConfig, template_values: List[gen_ai_hub.orchestration.models.template.TemplateValue], history: List[gen_ai_hub.orchestration.models.message.Message]) -&gt; None   Represents a request for the orchestration process, including configuration, template values, and message history.     Method resolution order: OrchestrationRequest gen_ai_hub.orchestration.models.base.JSONSerializable abc.ABC builtins.object Methods defined here: __eq__(self, other) Return self==value. __init__(self, config: gen_ai_hub.orchestration.models.config.OrchestrationConfig, template_values: List[gen_ai_hub.orchestration.models.template.TemplateValue], history: List[gen_ai_hub.orchestration.models.message.Message]) -> None Initialize self.  See help(type(self)) for accurate signature. __repr__(self) Return repr(self). to_dict(self) Converts the OrchestrationRequest instance to a dictionary.   :return: Dictionary representation of the OrchestrationRequest :rtype: dict Data and other attributes defined here: __abstractmethods__ = frozenset() __annotations__ = {'config': <class 'gen_ai_hub.orchestration.models.config.OrchestrationConfig'>, 'history': typing.List[gen_ai_hub.orchestration.models.message.Message], 'template_values': typing.List[gen_ai_hub.orchestration.models.template.TemplateValue]} __dataclass_fields__ = {'config': Field(name='config',type=<class 'gen_ai_hub.orch...appingproxy({}),kw_only=False,_field_type=_FIELD), 'history': Field(name='history',type=typing.List[gen_ai_hub...appingproxy({}),kw_only=False,_field_type=_FIELD), 'template_values': Field(name='template_values',type=typing.List[ge...appingproxy({}),kw_only=False,_field_type=_FIELD)} __dataclass_params__ = _DataclassParams(init=True,repr=True,eq=True,order=False,unsafe_hash=False,frozen=False) __hash__ = None __match_args__ = ('config', 'template_values', 'history') Data descriptors inherited from gen_ai_hub.orchestration.models.base.JSONSerializable: __dict__ dictionary for instance variables (if defined) __weakref__ list of weak references to the object (if defined)   class OrchestrationService(builtins.object)     OrchestrationService(api_url: Optional[str] = None, config: Optional[gen_ai_hub.orchestration.models.config.OrchestrationConfig] = None, proxy_client: Optional[gen_ai_hub.proxy.gen_ai_hub_proxy.client.GenAIHubProxyClient] = None, deployment_id: Optional[str] = None, config_name: Optional[str] = None, config_id: Optional[str] = None, timeout: Union[int, float, openai.Timeout, NoneType] = None)   A service for executing orchestration requests, allowing for the generation of  LLM-generated content through a pipeline of configured modules. This service supports both synchronous and asynchronous request execution. For streaming responses, special care is taken to not close the underlying  HTTP stream prematurely.   https://api.sap.com/api/ORCHESTRATION_API/overview     Methods defined here: __init__(self, api_url: Optional[str] = None, config: Optional[gen_ai_hub.orchestration.models.config.OrchestrationConfig] = None, proxy_client: Optional[gen_ai_hub.proxy.gen_ai_hub_proxy.client.GenAIHubProxyClient] = None, deployment_id: Optional[str] = None, config_name: Optional[str] = None, config_id: Optional[str] = None, timeout: Union[int, float, openai.Timeout, NoneType] = None) Initializes the OrchestrationService with the provided parameters.   :param api_url: The base URL for the orchestration API, defaults to None :type api_url: Optional[str], optional :param config: The default orchestration configuration, defaults to None :type config: Optional[OrchestrationConfig], optional :param proxy_client: The GenAIHubProxyClient instance, defaults to None :type proxy_client: Optional[GenAIHubProxyClient], optional :param deployment_id: the deployment ID, defaults to None :type deployment_id: Optional[str], optional :param config_name: the configuration name, defaults to None :type config_name: Optional[str], optional :param config_id: the configuration ID, defaults to None :type config_id: Optional[str], optional :param timeout: the timeout for HTTP requests, defaults to None :type timeout: Union[int, float, httpx.Timeout, None], optional async aclose_http_connection(self) Closes the httpx asynchronous client. async arun(self, config: Optional[gen_ai_hub.orchestration.models.config.OrchestrationConfig] = None, template_values: Optional[List[gen_ai_hub.orchestration.models.template.TemplateValue]] = None, history: Optional[List[gen_ai_hub.orchestration.models.message.Message]] = None, timeout: Union[int, float, openai.Timeout, NoneType] = None) -> gen_ai_hub.orchestration.models.response.OrchestrationResponse Executes an orchestration request asynchronously (non-streaming).   :param config: the orchestration configuration, defaults to None :type config: Optional[OrchestrationConfig], optional :param template_values: the template values for the request, defaults to None :type template_values: Optional[List[TemplateValue]], optional :param history: the message history, defaults to None :type history: Optional[List[Message]], optional :param timeout: the timeout for the request, defaults to None :type timeout: Union[int, float, httpx.Timeout, None], optional :return: An OrchestrationResponse object. :rtype: OrchestrationResponse async arun_with_retries(self, config: Optional[gen_ai_hub.orchestration.models.config.OrchestrationConfig] = None, template_values: Optional[List[gen_ai_hub.orchestration.models.template.TemplateValue]] = None, history: Optional[List[gen_ai_hub.orchestration.models.message.Message]] = None, timeout: Union[int, float, openai.Timeout, NoneType] = None, max_retries: int = 10, base_delay: float = 1.0) -> gen_ai_hub.orchestration.models.response.OrchestrationResponseWithRetries | None Executes an orchestration request asynchronously with automatic retry on rate limits (429) and  server errors. Uses exponential backoff with jitter to handle rate limiting gracefully.   :param config: the orchestration configuration, defaults to None :type config: Optional[OrchestrationConfig], optional :param template_values: the template values for the request, defaults to None :type template_values: Optional[List[TemplateValue]], optional :param history: the message history, defaults to None :type history: Optional[List[Message]], optional :param timeout: the timeout for the request, defaults to None :type timeout: Union[int, float, httpx.Timeout, None], optional :param max_retries: the maximum number of retry attempts, defaults to 10 :type max_retries: int, optional :param base_delay: the initial delay between retries in seconds, defaults to 1.0 :type base_delay: float, optional :return: An OrchestrationResponseWithRetries with retry count information. :rtype: OrchestrationResponseWithRetries | None :raises OrchestrationError: If the request fails after all retries (includes retry count). :raises ValueError: If no configuration is provided. async astream(self, config: Optional[gen_ai_hub.orchestration.models.config.OrchestrationConfig] = None, template_values: Optional[List[gen_ai_hub.orchestration.models.template.TemplateValue]] = None, history: Optional[List[gen_ai_hub.orchestration.models.message.Message]] = None, stream_options: Optional[dict] = None, timeout: Union[int, float, openai.Timeout, NoneType] = None) -> gen_ai_hub.orchestration.sse_client.AsyncSSEClient Executes an orchestration request asynchronously in streaming mode.   :param config: the orchestration configuration, defaults to None :type config: Optional[OrchestrationConfig], optional :param template_values: the template values for the request, defaults to None :type template_values: Optional[List[TemplateValue]], optional :param history: the message history, defaults to None :type history: Optional[List[Message]], optional :param stream_options: the additional streaming options, defaults to None :type stream_options: Optional[dict], optional :param timeout: the timeout for the request, defaults to None :type timeout: Union[int, float, httpx.Timeout, None], optional :return: An AsyncSSEClient instance for iterating over the streaming response. :rtype: AsyncSSEClient close_http_connection(self) Closes the httpx synchronous client. handle_retry(self, retry_count: int, base_delay: float, error: gen_ai_hub.orchestration.exceptions.OrchestrationError, max_retries: int) -> float Handles retry logic with exponential backoff and jitter. If Retry-After header exists, use it as min_delay to add jitter on top   :param retry_count: the current retry attempt number :type retry_count: int :param base_delay: the initial delay between retries in seconds :type base_delay: float :param error: the exception that occurred :type error: OrchestrationError :param max_retries: the maximum number of retry attempts :type max_retries: int :raises error: Raises the original error if no more retries should be attempted :return: number of seconds to wait before next retry :rtype: float run(self, config: Optional[gen_ai_hub.orchestration.models.config.OrchestrationConfig] = None, template_values: Optional[List[gen_ai_hub.orchestration.models.template.TemplateValue]] = None, history: Optional[List[gen_ai_hub.orchestration.models.message.Message]] = None, timeout: Union[int, float, openai.Timeout, NoneType] = None) -> gen_ai_hub.orchestration.models.response.OrchestrationResponse Executes an orchestration request synchronously (non-streaming).   :param config: the orchestration configuration, defaults to None :type config: Optional[OrchestrationConfig], optional :param template_values: the template values for the request, defaults to None :type template_values: Optional[List[TemplateValue]], optional :param history: the message history, defaults to None :type history: Optional[List[Message]], optional :param timeout: the timeout for the request, defaults to None :type timeout: Union[int, float, httpx.Timeout, None], optional :return: An OrchestrationResponse object. :rtype: OrchestrationResponse run_with_retries(self, config: Optional[gen_ai_hub.orchestration.models.config.OrchestrationConfig] = None, template_values: Optional[List[gen_ai_hub.orchestration.models.template.TemplateValue]] = None, history: Optional[List[gen_ai_hub.orchestration.models.message.Message]] = None, timeout: Union[int, float, openai.Timeout, NoneType] = None, max_retries: int = 10, base_delay: float = 1.0) -> gen_ai_hub.orchestration.models.response.OrchestrationResponseWithRetries | None Executes an orchestration request with automatic retry on rate limits (429) and server errors.   :param config: the orchestration configuration, defaults to None :type config: Optional[OrchestrationConfig], optional :param template_values: the template values for the request, defaults to None :type template_values: Optional[List[TemplateValue]], optional :param history: the message history, defaults to None :type history: Optional[List[Message]], optional :param timeout: the timeout for the request, defaults to None :type timeout: Union[int, float, httpx.Timeout, None], optional :param max_retries: the maximum number of retry attempts, defaults to 10 :type max_retries: int, optional :param base_delay: the initial delay between retries in seconds, defaults to 1.0 :type base_delay: float, optional :return: An OrchestrationResponseWithRetries with retry count information. :rtype: OrchestrationResponseWithRetries | None :raises OrchestrationError: If the request fails after all retries (includes retry count). :raises ValueError: If no configuration is provided. stream(self, config: Optional[gen_ai_hub.orchestration.models.config.OrchestrationConfig] = None, template_values: Optional[List[gen_ai_hub.orchestration.models.template.TemplateValue]] = None, history: Optional[List[gen_ai_hub.orchestration.models.message.Message]] = None, stream_options: Optional[dict] = None, timeout: Union[int, float, openai.Timeout, NoneType] = None) -> gen_ai_hub.orchestration.sse_client.SSEClient Executes an orchestration request in streaming mode (synchronously).   :param config: the orchestration configuration, defaults to None :type config: Optional[OrchestrationConfig], optional :param template_values: the template values for the request, defaults to None :type template_values: Optional[List[TemplateValue]], optional :param history: the message history, defaults to None :type history: Optional[List[Message]], optional :param stream_options: the additional streaming options, defaults to None :type stream_options: Optional[dict], optional :param timeout: the timeout for the request, defaults to None :type timeout: Union[int, float, httpx.Timeout, None], optional :return: An SSEClient instance for iterating over the streaming response. :rtype: SSEClient Data descriptors defined here: __dict__ dictionary for instance variables (if defined) __weakref__ list of weak references to the object (if defined)

Functions
		cache_if_not_none(func) Custom cache decorator that only caches non-None results   :param func: The function to be decorated. :type func: callable :return: The decorated function with caching behavior. :rtype: callable discover_orchestration_api_url(base_url: str, auth_url: str, client_id: str, client_secret: str, resource_group: str, config_id: Optional[str] = None, config_name: Optional[str] = None, orchestration_scenario: str = 'orchestration', executable_id: str = 'orchestration') -> Optional[str] Discovers the orchestration API URL based on provided configuration details.   :param base_url: the base URL for the AI Core API. :type base_url: str :param auth_url: the URL for the AI Core authentication service. :type auth_url: str :param client_id: the client ID for the AI Core API. :type client_id: str :param client_secret: the client secret for the AI Core API. :type client_secret: str :param resource_group: the resource group for the AI Core API. :type resource_group: str :param config_id: the configuration ID, defaults to None :type config_id: Optional[str], optional :param config_name: the configuration name, defaults to None :type config_name: Optional[str], optional :param orchestration_scenario: the orchestration scenario ID, defaults to "orchestration" :type orchestration_scenario: str, optional :param executable_id: the orchestration executable ID, defaults to "orchestration" :type executable_id: str, optional :return: The orchestration API URL or None if no deployment is found. :rtype: Optional[str] get_orchestration_api_url(proxy_client: gen_ai_hub.proxy.gen_ai_hub_proxy.client.GenAIHubProxyClient, deployment_id: Optional[str] = None, config_name: Optional[str] = None, config_id: Optional[str] = None) -> str Retrieves the orchestration API URL based on provided deployment or configuration details.   :param proxy_client: The GenAIHubProxyClient instance. :type proxy_client: GenAIHubProxyClient :param deployment_id: the deployment ID, defaults to None :type deployment_id: Optional[str], optional :param config_name: the configuration name, defaults to None :type config_name: Optional[str], optional :param config_id: the configuration ID, defaults to None :type config_id: Optional[str], optional :raises ValueError: If no orchestration deployment is found. :return: The orchestration API URL. :rtype: str

Data
		COMPLETION_SUFFIX = '/completion' Iterable = typing.Iterable List = typing.List Optional = typing.Optional Union = typing.Union