Updating model artifacts

This document explains how to update a model artifact in Fiddler using the client.update_model_artifact function, allowing you to replace a surrogate model or your own uploaded model.

If you need to update a model artifact already uploaded in Fiddler, you can use the update_artifact() function. This allows you to replace a surrogate model or your own uploaded model.

Once you have prepared the model artifacts directory, you can update your model

## "Create the Model"

PROJECT_ID = 'example_project'
MODEL_NAME = 'example_model'
DATASET_FILE_PATH = "Where dataset is available for the model"
MODEL_ARTIFACTS_DIR = 'path where model artefacts are put'

model = fdl.Model.from_data(
    source=DATASET_FILE_PATH, 
    name=MODEL_NAME, 
    project_id=project.id,
    spec=model_spec,
    task=fdl.ModelTask.BINARY_CLASSIFICATION,
    task_params=task_params,
    xai_params=xai_params,
)

Update the artifact


job = model.update_artifact(
    model_dir=MODEL_DIR,
)
job.wait()