caikit.core

Caikit Core AI Framework library. This is the base framework for core AI/ML libraries.

Submodules

• caikit.core.augmentors
• caikit.core.data_model
• caikit.core.exceptions
• caikit.core.model_management
• caikit.core.model_manager
• caikit.core.module_backends
• caikit.core.modules
• caikit.core.registries
• caikit.core.signature_parsing
• caikit.core.task
• caikit.core.toolkit

Attributes

MODEL_MANAGER
extract
load
resolve_and_load
train
start_prediction_job
get_model_future
get_prediction_future

Exceptions

DataValidationError

This error is used for data validation problems during training

Classes

DataObjectBase	A DataObject is a data model class that is backed by a @dataclass.
ModelManager	Manage the models or resources for library.
BackendBase	Interface for creating configuration setup for backends
LocalBackend	Interface for creating configuration setup for backends
ModuleBase	Abstract base class from which all modules should inherit.
ModuleConfig	Config object used by all modules for config loading, saving, etc.
ModuleLoader
ModuleSaver	A module saver that provides common functionality used for saving modules and also a context
TaskBase	The TaskBase defines the interface for an abstract AI task
EvalTypes	Enum that contains set of all possible evaluation types.
F1Metrics
F1MetricsContainer
QualityEvaluator	Class that holds all evaluation logic for now. May eventually be broken up into
ObjectSerializer	Abstract class for serializing an object to disk.
JSONSerializer	An ObjectSerializer for serializing to a JSON file.
TextSerializer	An ObjectSerializer for serializing a python list to a text file.
YAMLSerializer	An ObjectSerializer for serializing to a YAML file.
CSVSerializer	An ObjectSerializer for serializing to a CSV file.
PickleSerializer	An ObjectSerializer for pickling arbitrary Python objects.

Functions

get_valid_module_ids()	Get a dictionary mapping all module IDs to the string names of the
module([id, name, version, task, tasks, backend_type, ...])	Apply this decorator to any class that should be treated as a caikit module
task(→ Callable[[Type[TaskBase]], Type[TaskBase]])	The decorator for AI Task classes.
load_txt(filename)	Load a string from a file with utf8 encoding.
load_txt_lines(filename)	Load a list of files from a text file with utf8 encoding
save_txt(text, filename[, mode])	Write a string to a text file with utf8 encoding.
load_binary(filename)	Load a binary string from a file.
save_binary(data, filename)	Write a binary buffer to a file.
load_csv(filename)	Load a csv into a list-of-lists.
save_csv(text_list, filename[, mode])	Write a list-of-lists to a csv file.
load_dict_csv(filename)	Load a csv into a list-of-dicts.
save_dict_csv(dict_list, filename[, mode])	Write a list of dicts to a csv file.
load_json(filename)	Load a json file into a dictionary.
save_json(save_dict, filename[, mode])	Save a dictionary into a json file.
load_yaml(filename)	Load a yaml file into a dictionary.
save_yaml(save_dict, filename[, mode])	Save a dictionary into a yaml file.
load_pickle(filename)	Load an object from a pickle file.
save_pickle(obj, filename[, mode])	Save an object to a pickle file.
save_raw(save_content, filename[, mode])	Write the given raw string content to output file.
compress(dir_path[, output_path, extension])	Compress a given folder recursively to an archive with a given extension format

Package Contents

class caikit.core.DataObjectBase

Bases: caikit.core.data_model.base.DataBase

A DataObject is a data model class that is backed by a @dataclass.

Data model classes that use the @dataobject decorator must derive from this base class.

exception caikit.core.DataValidationError(reason: str, item_number: int | None = None)

Bases: Exception

This error is used for data validation problems during training

_reason

_item_number = None

property reason: str

The reason given for this data validation error

property item_number: int | None

The index of the training data item that failed validation. Probably zero indexed

caikit.core.get_valid_module_ids()

Get a dictionary mapping all module IDs to the string names of the implementing classes.

class caikit.core.ModelManager

Manage the models or resources for library.

_singleton_module_cache

_trainers

_finders

_job_predictors

_initializers

__singleton_lock

initialize_components()

Proactively initialize all configured trainer/finder/initializer component instances. This is a separate call to enable explicit config.

train(module: Type[caikit.core.modules.base.ModuleBase] | str, *args, trainer: str | caikit.core.model_management.ModelTrainerBase = 'default', save_path: str | caikit.interfaces.common.data_model.stream_sources.S3Path | None = None, save_with_id: bool = False, model_name: str | None = None, wait: bool = False, **kwargs) → caikit.core.model_management.ModelTrainerFutureBase

Train an instance of the given module with the given args and kwargs using the given trainer.

Each module’s train function encapsulates the code needed to perform the training locally. This top-level train function provides the wrapper functionality to delegate the execution of the module’s train function to an alternate framework using a ModelTrainerBase. It also allows training to be launched asynchronously.

Args:

module (Union[Type[ModuleBase], str]): The module class or guid for

the module to train

*args: Additional positional args to pass through to the module’s

train function

Kwargs:

trainer (Union[str, ModelTrainerBase]): The trainer to use. If given

as a string, this is a key in the global config at model_management.trainers.

save_path (Optional[Union[str, S3Path]]): Base path where the model should be

saved (may be relative to a remote trainer’s filesystem, or link to S3 storage)

save_with_id (bool): Inject the training ID into the save path for

the output model

model_name (Optional[str]): Name of model that will be appended

to the end of the save_path

wait (bool): Wait for training to complete before returning **kwargs: Additional keyword arguments to pass through to the

modules’s train function

Returns:

model_future (ModelFutureBase): The future handle

to the model which holds the status of the in-flight training.

start_prediction_job(model: caikit.core.modules.base.ModuleBase, prediction_func_name: str, *args, predictor: str | caikit.core.model_management.JobPredictorBase = 'default', wait: bool = False, **kwargs) → caikit.core.model_management.JobPredictorFutureBase

Start a prediction job using a job_predictor.

Args:

model (ModuleBase): Loaded model to run prediction on prediction_func_name (str): String reference to name of function to run predictor (Union[str, JobPredictorBase], optional): Which job_predictor to use.

Defaults to “default”.

wait (bool, optional): Weather to wait for job to finish. Defaults to False.

Returns:

JobPredictorFutureBase: Future to track job result

get_model_future(training_id: str) → caikit.core.model_management.ModelTrainerFutureBase

Get the future handle to an in-progress training

Args:

training_id (str): The ID string from the original training

submission’s ModelFuture

Returns:

model_future (ModelTrainerFutureBase): The future handle

to the model which holds the status of the in-flight training.

get_prediction_future(prediction_id: str) → caikit.core.model_management.JobPredictorFutureBase

Get the future handle to an in-progress prediction job

Args:

prediction_id (str): The ID string from the original prediction

submission’s ModelFuture

Returns:

prediction_future (JobPredictorFutureBase): The future handle

to the job which holds the status of the in-flight prediction.

load(module_path: str | io.BytesIO | bytes, *, load_singleton: bool = False, finder: str | caikit.core.model_management.ModelFinderBase = 'default', initializer: str | caikit.core.model_management.ModelInitializerBase = 'default', **kwargs)

Load a model and return an instantiated object on which we can run inference.

Args:

module_path (str | BytesIO | bytes): A module path (identifier) to

one of the following: 1. A directory containing a yaml config file in the top level. 2. A zip archive containing either a yaml config file in the

top level when extracted, or a directory containing a yaml config file in the top level.

3. A BytesIO object corresponding to a zip archive containing

   either a yaml config file in the top level when extracted, or a directory containing a yaml config file in the top level.

4. A bytes object corresponding to a zip archive containing

   either a yaml config file in the top level when extracted, or a directory containing a yaml config file in the top level.

5. A string that is understood by the configured

   finder/initializer

Kwargs:

load_singleton (bool): Load this model as a singleton finder (Union[str, ModelFinderBase]): Finder to use when loading

this model. If passed as a string, this names the finder in the global config model_management.finders section.

initializer (Union[str, ModelInitializerBase]): Loader to use when

initializint this model. If passed as a string, this is the name of the initializer in the global config model_management.initializers section.

Returns:

model (ModuleBase) Model object that is loaded, configured, and

ready for prediction.

extract(zip_path: str, model_path: str, force_overwrite: bool = False) → str

Method to extract a downloaded archive to a specified directory.

Args:

zip_path (str): Location of .zip file to extract. model_path (str): Model directory where the archive should be

unzipped unzipped.

force_overwrite: bool (Defaults to false)

Force an overwrite to model_path, even if the folder exists

Returns:

str: Output path where the model archive is unzipped.

resolve_and_load(path_or_name_or_model_reference: str | caikit.core.modules.base.ModuleBase, **kwargs)

Try our best to load a model, given a path or a name. Simply returns any loaded model passed in. This exists to ease the burden on workflow developers who need to accept individual modules in their API, where users may have references to custom models or may only have the ability to give the name of a stock model.

Args:

path_or_name_or_model_reference (str, ModuleBase): Either a

• Path to a model on disk

• Name of a model that the catalog knows about

• Loaded module

**kwargs: Any keyword arguments to pass along to ModelManager.load()

or ModelManager.download()

e.g. parent_dir

Returns:

A loaded module

Examples:

>>> stock_syntax_model = manager.resolve_and_load('syntax_izumo_en_stock')
>>> local_categories_model = manager.resolve_and_load('path/to/categories/model')
>>> some_custom_model = manager.resolve_and_load(some_custom_model)

get_singleton_model_cache_info()

Returns information about the singleton cache in {hash: module type} format

Returns:

Dict[str, type]: A dictionary of model hashes to model types

clear_singleton_cache()

Clears the cache of singleton models. Useful to release references of models, as long as you know that they are no longer held elsewhere and you won’t be loading them again.

Returns:

None

get_trainer(trainer: str | caikit.core.model_management.ModelTrainerBase) → caikit.core.model_management.ModelTrainerBase

Get the configured model trainer or the one passed by value

get_finder(finder: str | caikit.core.model_management.ModelFinderBase) → caikit.core.model_management.ModelFinderBase

Get the configured model finder or the one passed by value

get_initializer(initializer: str | caikit.core.model_management.ModelInitializerBase) → caikit.core.model_management.ModelInitializerBase

Get the configured model initializer or the one passed by value

get_predictor(inferencer: str | caikit.core.model_management.JobPredictorBase) → caikit.core.model_management.JobPredictorBase

Get the configured job predictor or the one passed by value

get_module_backends(initialize: bool = True) → List[caikit.core.module_backends.base.BackendBase]

Convenience method to get access to the configured module backends if any have been configured

Args:

initialize (bool): Initialize the components from config

Returns:

backends (List[BackendBase]): The list of backend instances that

have been configured

_do_load(module_path, load_singleton, finder, initializer, **kwargs)

Load a model from a directory.

Args:

module_path (str): Path to directory. At the top level of directory

is config.yml which holds info about the model.

load_singleton (bool): Load this model as a singleton finder (Union[str, ModelFinderBase]): Finder to use when loading

this model. If passed as a string, this names the finder in the global config model_management.finders section.

initializer (Union[str, ModelInitializerBase]): Loader to use when

loading this model. If passed as a string, this is the name of the initializer in the global config model_management.initializers section.

Returns:

subclass of caikit.core.modules.ModuleBase: Model object that is

loaded, configured, and ready for prediction.

_load_from_zipfile(module_path, load_singleton, finder, initializer, **kwargs)

Load a model from a zip archive.

Args:

module_path (str): Path to directory. At the top level of directory

is config.yml which holds info about the model.

load_singleton (bool): Load this model as a singleton finder (Union[str, ModelFinderBase]): Finder to use when loading

this model. If passed as a string, this names the finder in the global config model_management.finders section.

initializer (Union[str, ModelInitializerBase]): Loader to use when

loading this model. If passed as a string, this is the name of the initializer in the global config model_management.initializers section.

Returns:

subclass of caikit.core.modules.ModuleBase: Model object that is

loaded, configured, and ready for prediction.

_singleton_lock(load_singleton: bool)

Helper contextmanager that will only lock the singleton cache if this load is a singleton load

static _get_component(component: str | caikit.core.toolkit.factory.FactoryConstructible, component_dict: Dict[str, caikit.core.toolkit.factory.FactoryConstructible], component_factory: caikit.core.toolkit.factory.Factory, component_name: str, component_cfg: dict, component_type: type) → caikit.core.toolkit.factory.FactoryConstructible

Common logic for resolving components from config

NOTE: This is done lazily to avoid relying on import order and to allow

for dynamic config changes

class caikit.core.BackendBase(config: aconfig.Config | None = None)

Bases: abc.ABC

Interface for creating configuration setup for backends

config

_started = False

_start_lock

class property backend_type

Abstractmethod:

Property storing type of the backend

property is_started

abstract register_config(config)

Function to allow dynamic merging of configs. This can be useful, if there are explicit configurations particular implementations (modules) need to register before the starting the backend.

abstract start()

Function to start a distributed backend. This function should set self._started variable to True

abstract stop()

Function to stop a distributed backend. This function should set self._started variable to False

start_lock()

handle_runtime_context(model_id: str, runtime_context: caikit.core.data_model.runtime_context.RuntimeServerContextType)

Update backend state for the given model based on a runtime request.

Some backends may need to handle runtime context information for the target model in order to correctly configure the backend before finding and loading the model. By default, this is a No-Op.

Args:

model_id (str): The unique ID of the model that is referenced by the

runtime context

runtime_context (RuntimeServerContextType): The context for the

given runtime request

class caikit.core.LocalBackend(config: aconfig.Config | None = None)

Bases: caikit.core.module_backends.base.BackendBase

Interface for creating configuration setup for backends

backend_type = 'LOCAL'

Property storing type of the backend

register_config(config) → None

Function to merge configs with existing configurations

start()

Start local backend. This is a no-op function

stop()

Stop local backend. This is a no-op

class caikit.core.ModuleBase

Abstract base class from which all modules should inherit.

_metadata

_load_backend = None

property metadata: Dict[str, Any]

This module’s metadata.

Returns:

Dict[str, Any]: A dictionary of this module’s metadata

TODO: Can this be a ModuleConfig object instead? (or aconfig.Config)?

property module_metadata: Dict[str, Any]

Helper property to return metadata about a Module. This function is separate from metadata as this is specific for the class module. This function also requires a flat metadata structure without nested dictionaries.

NOTE: This should be a @classmethod but using @property/@classmethod together has been deprecated

Returns:

Dict[str, str]: A dictionary of this ModuleBases’s metadata

property public_model_info: Dict[str, Any]

Helper property to return public metadata about a specific Model. This function is separate from metdata as that contains the entire ModelConfig which might not want to be shared/exposed.

Returns:

Dict[str, str]: A dictionary of this models’s public metadata

set_load_backend(load_backend)

Method used by the model manager to indicate the load backend that was used to load this module

classmethod get_inference_signature(input_streaming: bool, output_streaming: bool, task: Type[caikit.core.TaskBase] = None) → caikit.core.signature_parsing.CaikitMethodSignature | None

Returns the inference method signature that is capable of running the module’s task for the given flavors of input and output streaming

classmethod get_inference_signatures(task: Type[caikit.core.TaskBase]) → List[Tuple[bool, bool, caikit.core.signature_parsing.CaikitMethodSignature]]

Returns inference method signatures for all supported flavors of input and output streaming for a given task

property load_backend

Get the backend instance used to load this module. This can be used in module implementations that require use of a specific backend at inference time.

classmethod bootstrap(*args, **kwargs)

Bootstrap a module. This method can be used to initialize the module from artifacts created outside of a particular caikit library

classmethod load(model_path: str | caikit.core.modules.config.ModuleConfig, *args, **kwargs) → ModuleBase

Load a new instance of workflow from a given model_path

Args:

model_path (Union[str, ModuleConfig]): Path to saved model or

in-memory ModuleConfig

Returns:

model (ModuleBase): A new instance of this module class

classmethod _load(module_loader, *args, **kwargs)

Load a model.

classmethod timed_load(*args, **kwargs)

Time a model load call.

Args:

*args (list): Will be passed to self.load. **kwargs (dict): Will be passed to self.load – the only way to

pass arbitrary arguments to self.load from this function.

Returns:

int, caikit.core._ModuleBase: The first return value is the total

time spent in the self.load call. The second return value is the loaded model.

Notes:

You can pass everything that should go to the run function normally using args/kwargs. Example: model.timed_load(“/model/path/dir”)

validate_loaded_model(*args)

Validate a loaded model.

save(model_path: str, *args, **kwargs)

Save a model.

Args:

model_path (str): Path on disk to export the model to.

as_file_like_object(*args, **kwargs) → io.BytesIO

Produces a file-like object corresponding to a zip archive affiliated with a given model. This method wraps is functionally similar to .save() - it saves a model into a temporary directory and produces a zip archive, then loads the result as a io.BytesIO object. The result of this function is also compatible with .load(), and cleanup is handled automatically.

Args:

*args, **kwargs (dict): Optional keyword arguments for saving.

Returns:

io.BytesIO: File like object holding an exported model in memory as

a io.BytesIO object.

as_bytes(*args, **kwargs) → bytes

Produces a bytes object corresponding to a zip archive affiliated with a given model. This method wraps is functionally similar to .save() - it saves a model into a temporary directory and produces a zip archive, then loads the result as a bytes object. The result of this function is also compatible with .load(), and cleanup is handled automatically.

Args:

*args, **kwargs (dict): Optional keyword arguments for saving.

Returns:

bytes: bytes object holding an exported model in memory.

run(*args, **kwargs)

Run a model - this typically makes a single prediction and returns an object from the data model.

run_batch(*args, **kwargs)

Run a model in batch mode - this typically ingests an iterable of inputs that can be applied to run & returns a list of data model objects that run ordinarily returns. A module may override this method to provide faster evaluation capabilities, e.g., by leveraging vectorization during prediction.

All provided args and kwargs that should be expanded with the batch should be provided as prebatched iterables. If a provided arg/kwarg is not provided as an iterable, it will be passed as is to all self contained run calls, which may be the case in some rare cases, such as runtime explanability enablement.

This function is intentionally kept as simple as possible. In order to maintain its simplicity, all argument iterables must be the same length, where the length of every provided iterable is presumed to be the batch size. If an iterable must be passed as arg to each run call, batch run must be called by wrapping it in another iterable and duplicating the iterable arg to match the size, or (ideally) overridden in the subclass as necessary.

Args:

*args: Variable length argument list to be passed directly to run(). **kwargs: Arbitrary keyword arguments to be passed directly to run().

Returns:

tuple: Iterable of prediction outputs, run as a batch.

timed_run(*args, num_seconds=None, num_iterations=None, **kwargs)

Time a number of runs over set seconds or iterations.

Args:

*args (list): Will be passed to self.run. num_seconds (int): Minimum number of seconds to run timed_run over.

Will most likely be more than this value due to its waiting for the each call to self.run to finish.

num_iterations (int): Minimum number of iterations to run timed_run

over. Will run exactly this many times.

**kwargs (dict): Will be passed to self.run.

Returns:

int, int, caikit.core.data_model.DataBase: The first return value is

the total time spent in the self.run loop. The second return value is the total number of calls to self.run were made. The return value is the output of the module’s run method

Notes:

You can pass everything that should go to the run function normally using args/kwargs. Example: model.timed_run(“some example text”, num_seconds=60)

By default it will run for greater than or equal to 120 seconds.

stream(data_stream, *args, **kwargs)

Lazily evaluate a run() on a given model by constructing a new data stream generator from the results. Note that we do not allow datastreams in args/kwargs. In rare cases, this may mean that stream() is not available, e.g., for keywords extraction. In these cases, stream() should be overridden in the subclass (module implementation) to allow and expand along multiple data streams.

Args:

data_stream (caikit.core.data_model.DataStream): Datastream to be

lazily sequentially processed by the module under consideration.

*args: Variable length argument list to be passed directly to run(). **kwargs: Arbitrary keyword arguments to be passed directly to run().

Returns:

protobufs: A DataBase object.

classmethod train(*args, **kwargs)

Train a model.

classmethod validate_training_data(training_data: str | caikit.core.data_model.DataStream, limit: int = -1) → List[caikit.core.exceptions.validation_error.DataValidationError]

Validate a set of training data, passed as a filename or as a data stream. Return up to limit number of DataValidationErrors

evaluation_type = None

evaluator = None

static find_label_func(*_args, **_kwargs)

Abstractmethod:

Function used to extract “label” from a prediction/result of a module’s .run method. Define if you wish to have more specific evaluation metrics. Implemented in subclass.

static find_label_data_func(*_args, **_kwargs)

Abstractmethod:

Function used to extract data belonging to class “label” from a prediction/result of a module’s .run method. Define if you wish to have more specific evaluation metrics. Implemented in subclass.

evaluate_quality(dataset_path, preprocess_func=None, detailed_metrics=False, labels=None, partial_match_metrics=False, max_hierarchy_levels=3, **kwargs)

Run quality evaluation for instance of module

Args:

dataset_path (str): Path to where the input “gold set” dataset

lives. Most often this is .json file.

preprocess_func (method): Function used as proxy for any preliminary

steps that need to be taken to run the model on the input text. This helper function ultimately leads to the input to this module and may involve executing other modules.

detailed_metrics: boolean (Optional, defaults to False)

Only for ‘keywords’. Include partial scores and scores over every text in document.

labels: list (Optional, defaults to None)

Optional list of class labels to evaluate quality on. By default evaluation is done over all class labels. Using this, you can explicitly mention only a subset of labels to include in the quality evaluation.

partial_match_metrics: boolean (Optional, defaults to False)

Include partial match micro avg F1.

max_hierarchy_levels (int): Used in hierarchical multilabel

multiclass evaluation only. The number of levels in the hierarchy to run model evaluation on, in addition to complete matches.

*args, **kwargs: Optional arguments which can be used by goldset/prediction

set extraction. Nonekeyword arguments: block_level: str

For any module that has pre processing steps in the middle of raw text and actual module input, use the input from gold standard labels instead of a pre-process function. Useful for measuring quality for the ‘block’ alone (instead of the module + pre_process pipeline)

Returns:

dict: Dictionary of results provided by the self.evaluator.run

function, depending on the associated evaluation_type. Reports things like precision, recall, and f1.

static _is_expandable_iterable(arg)

Check to see if something is a list / tuple of data model objects or strings. If it is, we consider it “expandable”, meaning that one element of the iterable to one run call. In contrast, if something is not expandable, it will be passed as is to each call.

Args:

arg (any): Argument to run_batch being considered.

Returns:

bool: True if the argument is a compatible iterable, False

otherwise.

_validate_and_extract_batch_size(*args, **kwargs)

Check to ensure that there’s at least one iterable whose length is well defined, i.e., no generators, and that if multiple iterable arg/kwarg values are provided, they are all the same length.

Args:

*args: Variable length argument list to be passed directly to run(). **kwargs: Arbitrary keyword arguments to be passed directly to run().

Returns:

int: Inferred batch size based on expandable iterables.

_validate_arg_and_verify_batch_size(val, current_batch_size)

Check an arg value from args/kwargs. If we find that it’s an expandable iterable, see if it conflicts with what we know about the inferred batch size so far.

args:

val (any): Argument / keyword argument value being inspected. current_batch_size (None | int): Current inferred batch size from

previous args/kwargs, or None if no inferences have been made on other expandable iterables yet.

Returns:

None | inferred batch size.

static _build_args_for_default_run_with_batch(fixed_args, expanded_args, idx)

Build the non keyword arguments for run_batch’s default implementation by expanding iterable args where possible, and grouping them with repeated noniterable arguments. The index correspondes to the current document under consideration.

Args:

fixed_args (dict): Noniterable args - common across all documents. expanded_args (dict): Iterable args - we’ll need to index into this

to get our doc arg.

idx (int): Index of the document being considered.

Returns:

list: Args to be run for document [idx].

static _build_kwargs_for_default_run_with_batch(fixed_kwargs, expanded_kwargs, idx)

Similar to the previous function, but for kwargs. Note that we can just clone our fixed kwargs instead of cycling through them, because order doesn’t matter here.

Args:

fixed_args (dict): Noniterable valued kwargs - common across all

documents.

expanded_args (dict): Iterable valued kwargs - we’ll need to index

into these to get our doc kwarg.

Returns:

dict: Kwargs to be run for document [idx].

_extract_gold_set(dataset)

Method for extracting gold set from dataset. Implemented in subclass.

Args:

dataset (object): In-memory version of whatever is loaded from on-

disk. May be json, txt, etc.

Returns:

list: List of labels in the format of the module_type that is being

called.

_extract_pred_set(dataset, preprocess_func=None, **kwargs)

Method for extracting pred set from dataset. Implemented in subclass.

Args:

dataset (object): In-memory version of whatever is loaded from on-

disk. May be json, txt, etc.

preprocess_func (method): Function used as proxy for any preliminary

steps that need to be taken to run the model on the input text. This helper function ultimately leads to the input to this module and may involve executing other modules.

**kwargs (dict): Optional keyword arguments for prediction set extraction.

Returns:

list: List of labels in the format of the module_type that is being

called.

static _load_evaluation_dataset(dataset_path)

Helper specifically for dataset loading.

Args:

dataset_path (str): Path to where the input ‘gold set’ dataset

lives. Most often this is .json file.

Returns:

object: list, dict, or other python object, depending on the input

dataset_path extension. Currently only supports .json and uses fileio from toolkit.

static _extract_gold_annotations(gold_set)

Extract the core list of annotations that is needed for quality evaluation

Args:

gold_set (list)

Returns:

gold_annotations: list

static _extract_pred_annotations(pred_set)

Extract the core list of predictions that is needed for quality evaluation

Args:

pred_set (list)

Returns:

pred_annotations: list

static _generate_report(report, gold_set)

Generate the quality report output Args:

report (dict) gold_set (list(dict))

class caikit.core.ModuleConfig(config_dict)

Bases: aconfig.Config

Config object used by all modules for config loading, saving, etc.

reserved_keys = ['model_path']

classmethod load(model_path: str | ModuleConfig) → ModuleConfig

Load a new module configuration from a directory on disk.

Args:

model_path (Union[str, ModuleConfig]): Path to model directory. At

the top level of directory is config.yml which holds info about the model. Note that the model_path here is assumed to be operating system correct as a consequence of the way this method is invoked by the model manager.

Returns:

model_config (ModuleConfig): Instantiated ModuleConfig for model

given model_path.

save(model_path)

Save this module configuration to a top-level config.yml file in the specified model path.

Args: str

Path to model directory. The config.yml file will be written to this location.

Notes:

model_path must already exist! This means you must create the directory outside of this routine.

class caikit.core.ModuleLoader(model_path: str | caikit.core.modules.config.ModuleConfig)

MODULE_PATHS_KEY = 'module_paths'

config

model_path

load_arg(arg)

Extract arg value from the loaded model’s config

load_args(*args)

Extract values from the loaded model’s config

load_module(module_paths_key, load_singleton=False)

Load a CaikitCore module from a module config.module_paths specification.

Args:

module_paths_key (str): key in config.module_paths looked at to

load a module

load_singleton (bool): singleton load flag to pass to individual

module loads

load_module_list(module_paths_key)

Load a list of CaikitCore module from a workflow config.module_paths specification.

Args:

module_paths_key (str): key in config.module_paths looked at to

load a list of modules

Returns:

list: list of loaded modules

class caikit.core.ModuleSaver(module: caikit.core.modules.base.ModuleBase, model_path, exist_ok=True)

A module saver that provides common functionality used for saving modules and also a context manager that cleans up in case an error is encountered during the save process for a model_path that did not already exist.

SAVED_KEY_NAME = 'saved'

CREATED_KEY_NAME = 'created'

TRACKING_KEY_NAME = 'tracking_id'

MODULE_VERSION_KEY_NAME = 'version'

MODULE_ID_KEY_NAME = 'module_id'

MODULE_CLASS_KEY_NAME = 'module_class'

model_path = b'.'

exist_ok = True

config

add_dir(relative_path, base_relative_path='')

Create a directory inside the model_path for this saver.

Args:

relative_path (str): A path relative to this saver’s model_path

denoting the directory to create.

base_relative_path (str): A path, relative to this saver’s

model_path, in which relative_path will be created.

Returns:

str, str: A tuple containing both the relative_path and

absolute_path to the directory created.

Examples:

>>> with ModelSaver('/path/to/model') as saver:
>>>     rel_path, abs_path = saver.add_dir('word_embeddings', 'model_data')
>>> print(rel_path)
model_data/word_embeddings
>>> print(abs_path)
/path/to/model/model_data/word_embeddings

copy_file(file_path, relative_path='')

Copy an external file into a subdirectory of the model_path for this saver.

Args:

file_path (str): Absolute path to the external file to copy. relative_path (str): The relative path inside of model_path where

the file will be copied to. If set to the empty string (default) then the file will be placed directly in the model_path directory.

Returns:

str, str: A tuple containing both the relative_path and

absolute_path to the copied file.

save_object(obj, filename, serializer, relative_path='')

Save a Python object using the provided ObjectSerializer.

Args:

obj (any): The Python object to save filename (str): The filename to use for the saved object serializer (ObjectSerializer): An ObjectSerializer instance (e.g.,

YAMLSerializer) that should be used to serialize the object

relative_path (str): The relative path inside of model_path where

the object will be saved

update_config(additional_config)

Add items to this saver’s config dictionary.

Args:

additional_config (dict): A dictionary of config options to add the

this saver’s configuration.

Notes:

The behavior of this method matches dict.update and is equivalent to calling saver.config.update. The saver.config dictionary may be accessed directly for more sophisticated manipulation of the configuration.

save_module(module, relative_path, **kwargs)

Save a CaikitCore module within a workflow artifact and add a reference to the config.

Args:

module (caikit.core.ModuleBase): The CaikitCore module to save as

part of this workflow

relative_path (str): The relative path inside of model_path where

the module will be saved

**kwargs: dict

key-value pair of parameters to be passed to module.save

save_module_list(modules, config_key, **kwargs)

Save a list of CaikitCore modules within a workflow artifact and add a reference to the config.

Args:

modules (dict{str -> caikit.core.ModuleBase}): A dict with module

relative path as key and a CaikitCore module as value to save as part of this workflow

config_key (str): The config key inside of model_path where the

modules’ relative path with be referenced

**kwargs: dict

key-value pair of parameters to be passed to module.save

Returns:

list_of_rel_path: list(str)

List of relative paths where the modules are saved

list_of_abs_path: list(str)

List of absolute paths where the modules are saved

__enter__()

Enter the module saver context. This creates the model_path directory. If this context successfully exits, then the model configuration and all files it contains will be written and saved to disk inside the model_path directory.

If exist_ok is False, an exception will be raised before touching existing model_path files.

If any uncaught exceptions are thrown inside this context, and exist_ok is False, then this new model_path will be removed. If exist_ok is True, the files will be kept and may include incomplete updates.

__exit__(exc_type, exc_val, exc_tb)

Exit the module saver context. If this context successfully exits, then the model configuration and all files it contains will be written and saved to disk inside the model_path directory.

If any uncaught exceptions are thrown inside this context, and exist_ok is False, then this new model_path will be removed. If exist_ok is True, the files will be kept and may include incomplete updates.

caikit.core.module(id=None, name=None, version=None, task: Type[caikit.core.task.TaskBase] = None, tasks: List[Type[caikit.core.task.TaskBase]] | None = None, backend_type='LOCAL', base_module: str | Type[caikit.core.modules.base.ModuleBase] = None, backend_config_override: Dict | None = None)

Apply this decorator to any class that should be treated as a caikit module

(i.e., extends`{caikit.core.ModuleBase}) and registered with caikit.core so that the library “knows” the class is a caikit module and is capable of loading instances of the module.

Args:

id: str

A UUID to use when registering this module with caikit.core Not required if based on another caikit module using base_module

name: str

A human-readable name for the module Not required if based on another caikit module using base_module

version: str

A SemVer for the module Not required if based on another caikit module using base_module

task: Type[TaskBase]

An ML task class that this module is an implementation for Not required if based on another caikit module using base_module, or if multiple tasks are specified using tasks.

tasks: Optional[List[Type[TaskBase]]

List of ML task classes that this module implements.

backend_type: backend_type

Associated backend type for the module. Default: LOCAL

base_module: str | ModuleBase

If this module is based on a different caikit module, provide name of the base module. Default: None

backend_config_override: Dict

Dictionary containing configuration required for the specific backend. Default: None

Returns:

A decorated version of the class to which it was applied, after registering the class as a valid module with caikit.core

class caikit.core.TaskBase

The TaskBase defines the interface for an abstract AI task

An AI task is a logical function signature which, when implemented, performs a task in some AI domain. The key property of a task is that the set of required input argument types and the output value type are consistent across all implementations of the task.

class InferenceMethodPtr

Little container class that holds a method name and its flavor of streaming. i.e. the args to a @TaskClass.taskmethod decoration.

method_name: str

input_streaming: bool

output_streaming: bool

context_arg: str | None

deferred_method_decorators: Dict[Type[TaskBase], Dict[str, List[TaskBase]]]

classmethod taskmethod(input_streaming: bool = False, output_streaming: bool = False, context_arg: str | None = None) → Callable[[_InferenceMethodBaseT], _InferenceMethodBaseT]

Decorates a module instancemethod and indicates whether the inputs and outputs should be handled as streams. This will trigger validation that the signature of this method is compatible with the task’s definition of input and output types.

The actual handling of validating the method and registering it is deferred until after the module class is created, which happens outside the context of this decoration.

classmethod deferred_method_decoration(module: Type)

Runs the actual decoration logic that taskmethod would have run if the module class existed during its lifetime.

Validates that all decorated methods match the task’s API expectations, and stores the signatures on the module class for access later.

classmethod has_inference_method_decorators(module_class: Type) → bool

Utility that returns true iff a module has any @TaskClass.taskmethod decorations

classmethod validate_run_signature(signature: caikit.core.signature_parsing.CaikitMethodSignature, input_streaming: bool, output_streaming: bool) → None

Validates that the provided method signature meets the api constraints defined in this task, for the given streaming flavors.

Raises:

ValueError if no type annotations were provided on the method TypeError if the type annotations do not meet the task’s api constraints

classmethod get_required_parameters(input_streaming: bool) → Dict[str, ValidInputTypes | Type[Iterable[ValidInputTypes]]]

Get the set of input types required by this task

classmethod get_output_type(output_streaming: bool) → Type[caikit.core.data_model.base.DataBase]

Get the output type for this task

NOTE: This method is automatically configured by the @task decorator

and should not be overwritten by child classes.

classmethod get_visibility() → bool

Get the visibility for this task.

NOTE: defaults to True even if visibility wasn’t provided

classmethod get_metadata() → Dict[str, Any]

Get any metadata defined for this task

NOTE: defaults to an empty dict if one wasn’t provided

classmethod _raise_on_wrong_output_type(output_type, module, output_streaming: bool)

static _subclass_check(this_type, that_type)

Wrapper around issubclass that first checks if both args are classes. Returns True if the types are the same, or they are both classes and this_type is a subclass of that_type

static _is_iterable_type(typ: Type) → bool

Returns True if typ is an iterable type. Does not work for types like list, tuple, but we’re interested here in List[T] etc.

This is implemented this way to support older python versions where isinstance(typ, typing.Iterable) does not work

caikit.core.task(unary_parameters: Dict[str, ValidInputTypes] = None, streaming_parameters: Dict[str, Type[Iterable[ValidInputTypes]]] = None, unary_output_type: Type[caikit.core.data_model.base.DataBase] = None, streaming_output_type: Type[Iterable[Type[caikit.core.data_model.base.DataBase]]] = None, visible: bool = True, metadata: Dict[str, Any] | None = None, **kwargs) → Callable[[Type[TaskBase]], Type[TaskBase]]

The decorator for AI Task classes.

This defines an output data model type for the task, and a minimal set of required inputs that all public models implementing this task must accept.

As an example, the caikit.interfaces.nlp.SentimentTask might look like:

@task(
    unary_parameters={
        "raw_document": caikit.interfaces.nlp.RawDocument
    },
    streaming_parameters={
        "raw_documents": Iterable[caikit.interfaces.nlp.RawDocument]
    }
    unary_output_type=caikit.interfaces.nlp.SentimentPrediction
    streaming_output_type=Iterable[caikit.interfaces.nlp.SentimentPrediction]
)
class SentimentTask(caikit.TaskBase):
    pass

and a module that implements this task might have methods like:

@module(id="b9d98408-84c2-488c-8385-9d698effe60b", task=SentimentTask)
class MyModule(ModuleBase):

    @SentimentTask.taskmethod()
    def run(raw_document: caikit.interfaces.nlp.RawDocument,
            inference_mode: str = "fast") ->
                caikit.interfaces.nlp.SentimentPrediction:
        # impl

    @SentimentTask.taskmethod(input_streaming=True, output_streaming=True)
    def run_bidi_stream(raw_documents: DataStream[caikit.interfaces.nlp.RawDocument])
            -> DataStream[caikit.interfaces.nlp.SentimentPrediction]:
        # impl

Note the run function may include other arguments beyond the minimal required inputs for the task.

Args:

unary_parameters (Dict[str, ValidInputTypes]): The required parameters that all module’s

unary-input inference methods must contain. A dictionary of parameter name to parameter type, where the types can be in the set of:

• Python primitives

• Caikit data models

• Iterable containers of the above

• Caikit model references (maybe?)

streaming_parameters: The same as unary_parameters, but for streaming-input inference

methods. All types must be in the form Iterable[T]

unary_output_type (Type[DataBase]): The unary output type of the task, which all modules’

unary-output inference methods must return. This must be a caikit data model type.

streaming_output_type (Type[Iterable[Type[DataBase]]]): The streaming output type of the

task, which all modules’ streaming-output inference methods must return. This must be in the form Iterable[T].

visible (bool): If this task should be exposed to the end user in documentation or if

it should only be used internally

metadata (Optional[Dict[str, Any]]): Any additional metadata that should

be included in the documentation for this task

Returns:

A decorator function for the task class, registering it with caikit’s core registry of

tasks.

class caikit.core.EvalTypes(*args, **kwds)

Bases: enum.Enum

Enum that contains set of all possible evaluation types.

SINGLELABEL_MULTICLASS = 1

MULTILABEL_MULTICLASS = 2

MULTILABEL_MULTICLASS_HIERARCHICAL = 3

class caikit.core.F1Metrics

true_positive: int | None = None

false_positive: int | None = None

false_negative: int | None = None

precision: float | None = None

recall: float | None = None

f1: float | None = None

class caikit.core.F1MetricsContainer

per_class_confusion_matrix: Dict[str, F1Metrics]

macro_metrics: F1Metrics

micro_metrics: F1Metrics

class caikit.core.QualityEvaluator(gold, pred)

Class that holds all evaluation logic for now. May eventually be broken up into subclasses.

gold

pred

run(evaluation_type, find_label_func=None, find_label_data_func=None, detailed_metrics=False, labels=None, partial_match_metrics=False, max_hierarchy_levels=3)

Main entry point for evaluation.

Args:

evaluation_type (str): Which type of evaluation to run. Only a few

are currently supported.

find_label_func: function to fetch labels from any one prediction, used in

multiclass multilabel evaluation. eg: if a prediction is of form (token, label), this function should be able to tell us how to extract the class labels from the prediction, in this case return the second element of the tuple.

find_label_data_func: function to fetch predictions that belongs to a certain label,

used only in multiclass multilabel eval type, e.g., if predictions for a data example looks like [(tok1, labX), (tok2, labY), (tok3, labX)], then the function should be able to return all predictions with a given label - labX return should look like [(tok1, labX), (tok3, labX)]

detailed_metrics: flag to indicate whether or not you want detailed metrics

(currently only for multiclass multilabel eval type) Detailed metrics give us metrics for every example, and metrics using a custom partial match function

labels: list (Optional, defaults to None)

Optional list of class labels to evaluate quality on. By default evaluation is done over all class labels. Using this, you can explicitly mention only a subset of labels to include in the quality evaluation.

partial_match_metrics: flag to indicate whether or not you want partial match

micro avg metrics. (currently only for multiclass multilabel eval type)

max_hierarchy_levels (int): Used in hierarchical multilabel

multiclass evaluation only. The number of levels in the hierarchy to run model evaluation on, in addition to complete matches.

Returns:

dict: Full results from evaluation on dataset and model.

singlelabel_multiclass_evaluation(labels=None) → dict

Obtain results of evaluation for a single-label, multi-class model.

Args:

Note: here class should be initialized with gold and pred in the following format self.gold (list): list of gold set labels for every example, where each example

can have only one label eg: [‘label1’,’label2’, ‘label3’,’label4’]

self.pred (list): Predicted-by-the-model set labels for every example. labels: list (Optional, defaults to None)

Optional list of class labels to evaluate quality on. By default evaluation is done over all class labels. Using this, you can explicitly mention only a subset of labels to include in the quality evaluation.

Returns:

dict: Dictionary looks like: { ‘per_class_confusion_matrix’:

{‘entity_type’: {‘true_positive’: int …}} ‘macro_precision’: 0 <= float <= 1, ‘macro_recall’: 0 <= float <= 1, ‘macro_f1’: 0 <= float <= 1, ‘micro_precision’: 0 <= float <= 1,, ‘micro_recall’: 0 <= float <= 1,, ‘micro_f1’: 0 <= float <= 1, ‘overall_tp’: int, ‘overall_fp’: int, ‘overall_fn’: int

}

multilabel_multiclass_evaluation(find_label_func, find_label_data_func, labels=None, detailed_metrics=False, partial_match_metrics=False, use_labels_for_matching=False) → dict

Obtain results of evaluation for a multi-label, multi-class model.

Args:

Note: here class should be initialized with gold and pred in the following format self.gold (list(list)): list of gold set labels for every example eg:

[[‘label1’,’label2’], [‘label1’, ‘label4’]]

self.pred (list(list)): Predicted-by-the-model set labels for every example. find_label_func: function to fetch labels from any one prediction find_label_data_func: function to fetch data that belongs to a certain class labels: list (Optional, defaults to None)

Optional list of class labels to evaluate quality on. By default evaluation is done over all class labels. Using this, you can explicitly mention only a subset of labels to include in the quality evaluation.

detailed_metrics: flag to indicate whether or not you want detailed metrics

Detailed metrics give us metrics for every example, and metrics using a custom partial match function

partial_match_metrics: flag to indicate whether or not you want partial match

micro avg metrics.

use_labels_for_matching (bool): Indicates whether or not we should

use the output of find_label_func for metric computations, or the raw data tuples.

Returns:

dict: Dictionary looks like: { ‘per_class_confusion_matrix’:

{‘entity_type’: {‘true_positive’: int …}} ‘macro_precision’: 0 <= float <= 1, ‘macro_recall’: 0 <= float <= 1, ‘macro_f1’: 0 <= float <= 1, ‘micro_precision’: micro_precision, ‘micro_recall’: micro_recall, ‘micro_f1’: micro_f1, ‘detailed_metrics’ : {‘exact_match_precision’..,’partial_match_precision’} ‘micro_precision_partial_match’: 0 <= float <= 1, ‘micro_recall_partial_match’: 0 <= float <= 1, ‘micro_f1_partial_match’: 0 <= float <= 1 }

multilabel_multiclass_hierarchical_evaluation(find_label_func_builder, find_label_data_func_builder, max_hierarchy_levels=3) → dict

Evaluate multilabel/multiclass over a hierarchy, e.g., for ESA categories. This method Evaluates over a set number of hierarchy levels.

Because each level in the hierarchy needs to be able to compare and extract differently, we use builder funcs that create the appropriate functions for a given level of the hierarchy.

Args:

find_label_func_builder (function): A function that takes in a level

number (or None if full hierarchy) and returns a find_label_func for this level that can be passed to the multilabel multiclass evaluator.

find_label_data_func_builder (function): A function that takes in a

level number (or None if full hierarchy) and returns a find_label_data_func for this level that can be passed to the multilabel multiclass evaluator.

max_hierarchy_levels (int): The number of levels to run in the

hierarchy, in addition to complete match.

Returns:

dict: Dictionary, where each key is a level number, or ‘FULL’, and

maps to the dict returned by multilabel_multiclass_evaluation for that level of the hierarchy.

static calc_f1_score(gold, pred, match_fun=None)

Calculates F1 score Args:

gold (list): List of gold annotations pred (list): List of predictions match_fun: Function that finds the matches and returns tuple of matched gold, preds

Returns:

tuple: Precision, Recall, F1 score

static find_partial_matches(groundtruth, prediction)

Function to do find partial match between predicted phrases and the ground truth.

partial match means a complete predicted phrase is a part of any ground truth phrase or a complete ground truth phrase is a part of any predicted phrase. Overlaps are not considered.

Args:

groundtruth (list): Groundtruth data prediction (list): Predictions returned by the model

Returns:

tuple: gold_matched: set, pred_matched: set gold annotations that

were matched Predictions that partially or fully matched with groundtruth

static calc_metrics_from_confusion_matrix(per_class_confusion_matrix: Dict[str, F1Metrics]) → F1MetricsContainer

Function to calculate precision, recall, F1 metrics using a confusion matrix containing

statistics per class label.

Args:

per_class_confusion_matrix (Dict[str, F1Metrics]): Dictionary of

statistics per class label. Class labels are keys for the dictionary. For each class label, there should be a F1Metrics class object with values true positive, false_positive , false_negative representating the count of these per class. The dictionary looks like: per_class_confusion_matrix[label] = F1Metrics(true_positive = val 1, false_positive = val 2, false_negative = val 3)

Returns:

Returns: metrics_summary: F1MetricsContainer An instance of F1MetricsContainer dataclass containing summary of F1 metrics

convert_F1MetricsContainer_to_dict() → dict

Args:

metrics_summary (F1MetricsContainer): An object of dataclass

F1MetricsContainer

Returns:

Returns: dict

Dictionary looks like: {

‘per_class_confusion_matrix’: {‘entity_type’: {‘true_positive’: int …}} ‘macro_precision’: 0 <= float <= 1, ‘macro_recall’: 0 <= float <= 1, ‘macro_f1’: 0 <= float <= 1, ‘micro_precision’: 0 <= float <= 1,, ‘micro_recall’: 0 <= float <= 1,, ‘micro_f1’: 0 <= float <= 1, ‘overall_tp’: int, ‘overall_fp’: int, ‘overall_fn’: int

}

caikit.core.load_txt(filename)

Load a string from a file with utf8 encoding.

caikit.core.load_txt_lines(filename)

Load a list of files from a text file with utf8 encoding

caikit.core.save_txt(text, filename, mode='w')

Write a string to a text file with utf8 encoding.

caikit.core.load_binary(filename)

Load a binary string from a file.

caikit.core.save_binary(data, filename)

Write a binary buffer to a file.

caikit.core.load_csv(filename)

Load a csv into a list-of-lists.

caikit.core.save_csv(text_list, filename, mode='w')

Write a list-of-lists to a csv file.

caikit.core.load_dict_csv(filename)

Load a csv into a list-of-dicts.

caikit.core.save_dict_csv(dict_list, filename, mode='w')

Write a list of dicts to a csv file.

caikit.core.load_json(filename)

Load a json file into a dictionary.

caikit.core.save_json(save_dict, filename, mode='w')

Save a dictionary into a json file.

caikit.core.load_yaml(filename)

Load a yaml file into a dictionary.

caikit.core.save_yaml(save_dict, filename, mode='w')

Save a dictionary into a yaml file.

caikit.core.load_pickle(filename)

Load an object from a pickle file.

caikit.core.save_pickle(obj, filename, mode='wb')

Save an object to a pickle file.

caikit.core.save_raw(save_content, filename, mode='w')

Write the given raw string content to output file.

caikit.core.compress(dir_path, output_path=None, extension='zip')

Compress a given folder recursively to an archive with a given extension format

Args:

dir_path (str): Path of directory to compress output_path: (Optional) str

Output path where the archive is created. Defaults to current path + ‘archive’ + format extension >>> compress(‘.’, ‘my/path’, ‘tar’) >>> # saves to ‘my/path/archive.tar’

extension: (Optional) (one of: zip/tar/gztar/bztar/xztar depending on module availability)

Defaults to .zip

Returns:

str: Path to created archive

class caikit.core.ObjectSerializer

Bases: abc.ABC

Abstract class for serializing an object to disk.

abstract serialize(obj, file_path)

Serialize the provided object to the specified file path.

Args:

obj (object): The object to serialize file_path (str): Absolute path to which the object should be

serialized

class caikit.core.JSONSerializer

Bases: ObjectSerializer

An ObjectSerializer for serializing to a JSON file.

serialize(obj, file_path)

Serialize the provided object to a JSON file.

Args:

obj (object): The object to serialize file_path (str): Absolute path to which the object should be

serialized

class caikit.core.TextSerializer

Bases: ObjectSerializer

An ObjectSerializer for serializing a python list to a text file.

serialize(obj, file_path)

Serialize the provided python list to a text file.

Args:

obj (list(str)): The list to serialize file_path (str): Absolute path to which the object should be

serialized

class caikit.core.YAMLSerializer

Bases: ObjectSerializer

An ObjectSerializer for serializing to a YAML file.

serialize(obj, file_path)

Serialize the provided object to a YAML file.

Args:

obj (object): The object to serialize file_path (str): Absolute path to which the object should be

serialized

class caikit.core.CSVSerializer

Bases: ObjectSerializer

An ObjectSerializer for serializing to a CSV file.

serialize(obj, file_path)

Serialize the provided object to a CSV file.

Args:

obj (object): The object to serialize file_path (str): Absolute path to which the object should be

serialized

class caikit.core.PickleSerializer

Bases: ObjectSerializer

An ObjectSerializer for pickling arbitrary Python objects.

serialize(obj, file_path)

Serialize the provided object to a CSV file.

Args:

obj (any): The object to serialize file_path (str): Absolute path to which the object should be

serialized

caikit.core.MODEL_MANAGER

caikit.core.extract

caikit.core.load

caikit.core.resolve_and_load

caikit.core.train

caikit.core.start_prediction_job

caikit.core.get_model_future

caikit.core.get_prediction_future