caikit.runtime.service_generation.create_service
================================================

.. py:module:: caikit.runtime.service_generation.create_service

.. autoapi-nested-parse::

   This script auto-generates the `caikit-runtime.proto` RPC definitions for a
   collection of caikit.core derived libraries



Attributes
----------

.. autoapisummary::

   caikit.runtime.service_generation.create_service.log
   caikit.runtime.service_generation.create_service.error
   caikit.runtime.service_generation.create_service.TRAIN_FUNCTION_NAME


Functions
---------

.. autoapisummary::

   caikit.runtime.service_generation.create_service.assert_compatible
   caikit.runtime.service_generation.create_service.create_inference_rpcs
   caikit.runtime.service_generation.create_service.create_job_inference_rpcs
   caikit.runtime.service_generation.create_service.create_training_rpcs
   caikit.runtime.service_generation.create_service._group_modules_by_task


Module Contents
---------------

.. py:data:: log

.. py:data:: error

.. py:data:: TRAIN_FUNCTION_NAME
   :value: 'train'


.. py:function:: assert_compatible(modules: List[str], previous_modules: List[str])

   Logic about whether it's okay to include this set of modules in service generation

   Args:
       modules: list of module IDs that we are considering in service generation
       previous_modules: list of module IDs that were supported in the previous service version

   Raises:
       If a new service should not be built with this set of modules


.. py:function:: create_inference_rpcs(modules: List[Type[caikit.core.ModuleBase]], caikit_config: aconfig.aconfig.Config = None) -> List[caikit.runtime.service_generation.rpcs.CaikitRPCBase]

   Handles the logic to create all the RPCs for inference


.. py:function:: create_job_inference_rpcs(modules: List[Type[caikit.core.ModuleBase]], caikit_config: aconfig.aconfig.Config = None) -> List[caikit.runtime.service_generation.rpcs.CaikitRPCBase]

   Handles the logic to create all the RPCs for inference jobs


.. py:function:: create_training_rpcs(modules: List[Type[caikit.core.ModuleBase]]) -> List[caikit.runtime.service_generation.rpcs.CaikitRPCBase]

   Handles the logic to create all the RPCs for training


.. py:function:: _group_modules_by_task(modules: List[Type[caikit.core.ModuleBase]], included_tasks: List[Type[caikit.core.TaskBase]], excluded_tasks: List[Type[caikit.core.TaskBase]]) -> Dict[Type[caikit.core.TaskBase], List[caikit.core.signature_parsing.module_signature.CaikitMethodSignature]]