predict_real_time

MultiModalCloudPredictor.predict_real_time(test_data: str | DataFrame, test_data_image_column: str | None = None, accept: str = 'application/x-parquet', **kwargs) → Series

Predict with the deployed endpoint. A deployed endpoint is required. This is intended to provide a low latency inference. If you want to inference on a large dataset, use predict() instead.

Deprecated since version Use: predict() of the endpoint returned by deploy() instead.

Parameters:

  • test_data (str | pd.DataFrame) – The test data to be inferenced. Can be a pd.DataFrame, or a local path to csv file.

  • test_data_image_column (default = None) – If provided a csv file or pd.DataFrame as the test_data and test_data involves image modality, you must specify the column name corresponding to image paths. The path MUST be an abspath

  • accept (str, default = application/x-parquet) – Type of accept output content. Valid options are application/x-parquet, text/csv, application/json

  • **kwargs (Any) – Additional args that you would pass to predict calls of an AutoGluon logic

Returns:

pd.Series – Predict results in pd.Series

SageMaker API

  • InvokeEndpoint: sends the data to the endpoint and returns the predictions. The payload is limited to 6 MB (4 MB for serverless endpoints).