Skip to content

BaseResourceVersionControl

Overview

A resource_version_control is a component that is able to version control resources in the machine learning workflow, such as datasets and machine learning models. A resource_version_control system is able to save resources to persistent storage as well as retrieve those items in the future.

Attributes

BaseResourceVersionControl contains no default attributes.

Configuration

BaseResourceVersionControl contains no default or required configuration.

Interface

The following methods are part of BaseResourceVersionControl and should be implemented in any class that inherits from this base class:

version_data

Versions a dataset. 

def version_data(self, dataset_version, data, *args, **kwargs) -> dict[str, Any]

Arguments:

  • dataset_version (object): The dataset version being versioned
  • data (object): The data to version

Returns:

  • dict: Attributes returned from the resource version control system, such as a commit hash. The returned information should be able to be used to retrieve the object in the future and may very likely be logged in the metadata_tracker

get_data

Versions a dataset. 

def get_data(self, dataset_version, *args, **kwargs) -> Any

Arguments:

  • dataset_version (object): The dataset version to retrieve data from

Returns:

  • data: The data of the object

version_model

Versions a model. 

def version_model(self, experiment, model, *args, **kwargs) -> dict[str, Any]

Arguments:

  • experiment (object): The experiment being verisoned
  • model (object): The model to version

Returns:

  • dict: Attributes returned from the resource version control system, such as a commit hash. The returned information should be able to be used to retrieve the object in the future and may very likely be logged in the metadata_tracker

get_model

Returns a model object from an experiment. 

def get_model(self, experiment, *args, **kwargs) -> Any

Arguments:

  • id (str): The experiment to retrieve the model from

Returns:

  • model: The model object from the experiment.

version_feature_transformer

Versions a feature transformer. 

def version_feature_transformer(self, experiment, transformer, *args, **kwargs) -> dict[str, Any]

Arguments:

  • experiment (object): The experiment being verisoned
  • transformer (object): The feature transformer to version

Returns:

  • dict: Attributes returned from the resource version control system, such as a commit hash. The returned information should be able to be used to retrieve the object in the future and may very likely be logged in the metadata_tracker

get_feature_transformer

Returns a feature transformer object from an experiment. 

def get_feature_transformer(self, experiment, *args, **kwargs) -> Any

Arguments:

  • experiment (object): The experiment to retrieve the feature_transformer from

Returns:

  • feature_transformer: The feature transformer object from the experiment.