Method
createAlgorithm POST
Copy POST

Creates a custom algorithm that is re-usable for model training.

Arguments:

REQUIRED KEY TYPE DESCRIPTION
Yes name str The name to identify the algorithm; only uppercase letters, numbers, and underscores are allowed.
Yes problemType str The type of problem this algorithm will work on.
No sourceCode str Contents of a valid Python source code file. The source code should contain the train/predict/predict_many/initialize functions. A list of allowed import and system libraries for each language is specified in the user functions documentation section.
No trainingDataParameterNamesMapping dict The mapping from feature group types to training data parameter names in the train function.
No trainingConfigParameterName str The train config parameter name in the train function.
No trainFunctionName str Name of the function found in the source code that will be executed to train the model. It is not executed when this function is run.
No predictFunctionName str Name of the function found in the source code that will be executed to run predictions through the model. It is not executed when this function is run.
No predictManyFunctionName str Name of the function found in the source code that will be executed for batch prediction of the model. It is not executed when this function is run.
No initializeFunctionName str Name of the function found in the source code to initialize the trained model before using it to make predictions using the model.
No configOptions dict Map dataset types and configs to train function parameter names.
No isDefaultEnabled bool Whether to train with the algorithm by default.
No projectId str The unique version ID of the project.
No useGpu bool Whether this algorithm needs to run on GPU.
No packageRequirements list List of package requirement strings. For example: ['numpy==1.2.3', 'pandas>=1.4.0'].
Note: The arguments for the API methods follow camelCase but for Python SDK underscore_case is followed.

Response:

KEY TYPE DESCRIPTION
success Boolean true if the call succeeded, false if there was an error
result Algorithm
KEY TYPE Description
name str The name of the algorithm
problemType str The type of the problem this algorithm will work on
createdAt str When the algorithm was created
updatedAt str When the algorithm was last updated
isDefaultEnabled bool Whether train with the algorithm by default
trainingInputMappings dict The mappings for train function parameters' names, e.g. names for training data, name for training config
trainFunctionName str Name of the function found in the source code that will be executed to train the model. It is not executed when this function is run.
predictFunctionName str Name of the function found in the source code that will be executed run predictions through model. It is not executed when this function is run.
predictManyFunctionName str Name of the function found in the source code that will be executed for batch prediction of the model. It is not executed when this function is run.
initializeFunctionName str Name of the function found in the source code to initialize the trained model before using it to make predictions using the model
configOptions dict Map dataset types and configs to train function parameter names
algorithmId str The unique identifier of the algorithm
useGpu bool Whether to use gpu for model training
algorithmTrainingConfig dict The algorithm specific training config
onlyOfflineDeployable bool Whether or not the algorithm is only allowed to be deployed offline
codeSource CodeSource Info about the source code of the algorithm
KEY TYPE Description
sourceType str The type of the source, one of TEXT, PYTHON, FILE_UPLOAD, or APPLICATION_CONNECTOR
sourceCode str If the type of the source is TEXT, the raw text of the function
applicationConnectorId str The Application Connector to fetch the code from
applicationConnectorInfo str Args passed to the application connector to fetch the code
packageRequirements list The pip package dependencies required to run the code
status str The status of the code and validations
error str If the status is failed, an error message describing what went wrong
publishingMsg dict Warnings in the source code
moduleDependencies list The list of internal modules dependencies required to run the code

Exceptions:

TYPE WHEN
DataNotFoundError

projectId is not found.
InvalidEnumParameterError

An invalid value is passed for problemType.
ConflictError

The source code is invalid.
InvalidParameterError

The algorithm name is invalid.
Language:

© Copyright