How to create an AD model object ?.¶
Following are the steps or information required to create an Amorphic machine learning model object:
This is the model name in the Amorphic portal.
This is the description of the model
There are three ways to integrate Sagemaker model with the Amorphic portal
Existing Model Resource¶
This is a way to import Sagemaker Marketplace subscribed model in the Amorphic Portal. In order to import a model from Sagemaker marketplace, raise a request to the Admin. The admin will create a support ticket for the AWS Marketplace model using firstname.lastname@example.org. Amorphic team will make sure the respective model is available to select.
This is a way to upload your sagemaker model file directly from an Amorphic S3 location. Amorphic users can use Amorphic Notebooks to create a model in a Amorphic S3 location. Please refer to the Notebook section for the respective bucket details.
This is a way to upload a sagemaker model tar file directly to the Amorphic portal. Upon selecting this option you can upload any model tar or tar.gz file directly into the Amorphic portal.
You have two options - Dataset Data or Metadata. Select Dataset data when the requirement is to run the model on a Dataset file. Select Metadata, when you would like to view “AI/ML Results” ie. metadata on a dataset files (explained later). Most of the time you would want to use Dataset data.
Input Schema and Output Schema¶
Dataset Data would require two additional inputs - Input and Output Schema.
This is the schema used to identify the schema for the Dataset on which the pre-processing ETL job or the model is to be run.
This is the schema used to identify the schema for the Dataset on which the post-processing job or the model output will be saved.
Both the schema should have the same following format matching the respective Datasets:
Note: You can import the schema of an Amorphic dataset using the “Import from Dataset” functionality
The platform currently supports all the major AWS Sagemaker models
Supported file formats¶
Select the respective file type for predictions. If you require a file format other than the available format, then select the others file type. It will default to no file type required for batch prediction purposes. Note: if a model is selected as “Others” file type, then it can only be run on a “Others” file type Dataset.
Preprocess Glue Job¶
Select the preprocessing ETL jobs created using Amorphic ETL functionality.
Postprocess Glue Job¶
Select the post process ETL jobs created using Amorphic ETL functionality.