Model Types

Gretel's models provide a vast number of possibilities.

Gretel provides a number of different model types which may be utilized directly or combined via workflows. This page will outline the different categories of models that Gretel offers.

Synthetics

Synthetic models are the core models you will use to generate privacy guaranteed synthetic data. Here are the Synthetic data models supported by Gretel APIs.

• Gretel ACTGAN - Adversarial model for tabular, structured numerical, high column count data.

• Gretel Tabular DP - Graph-based model for tabular data with differential privacy.

• Gretel GPT - Generative pre-trained transformer for natural language text.

• Gretel DGAN - Adversarial model for time-series data.

• Gretel Amplify - Statistical model for high volume tabular data.

• Gretel LSTM - Language model for tabular, time series, text data.

For more information, please refer to the Synthetics documentation.

Transform

Gretel’s Transform model combines data classification with data transformation to easily detect and anonymize or mutate sensitive data. Gretel’s data classification can detect a variety of Supported Entities such as PII, which can be used for defining transforms. Gretel Transform can be combined with Gretel Synthetics using workflows to transform data before training the synthetic model or after generating output synthetic data.

For more information, please refer to the Transform documentation.

Classify

The Classify model allows you to define a policy to discover and label sensitive data including personally identifiable information, credentials, and even custom regular expressions inside text, logs, and other structured data.

For more information, please refer to the Classify documentation.

Relational

With Gretel Relational, you can synthesize and transform multiple tables or entire SQL databases, while ensuring referential integrity, accuracy, and privacy.

For more information, please refer to the Relational documentation.

Tabular LLM

Tabular LLM is Gretel's first foundational model designed to generate, edit, and augment tabular data using natural language or SQL prompts. It's a tool for working with and enhancing datasets in a more intuitive and interactive way.

For more information, please refer to the Tabular LLM documentation.