Synthetics
This section covers the model training and generation APIs shared across all Gretel models.
Last updated
Was this helpful?
This section covers the model training and generation APIs shared across all Gretel models.
Last updated
Was this helpful?
Gretel offers the following synthetics models:
- Gretel’s flagship LLM-based model for generating privacy-preserving, real-world quality synthetic data across numeric, categorical, text, JSON, and event-based tabular data with up to ~50 columns.
Data types: Numeric, categorical, text, JSON, event-based
Differential privacy: Optional
- Gretel’s model for generating privacy-preserving synthetic text using your choice of top performing open-source models.
Data types: Text
Differential privacy: Optional
- Gretel’s model for quickly generating synthetic numeric and categorical data for high-dimensional datasets (>50 columns) while preserving relationships between numeric and categorical columns.
Data types: Numeric, categorical
Differential privacy: NOT supported
- Gretel’s model for generating differentially-private data with very low epsilon values (maximum privacy). It is best for basic analytics use cases (e.g. pairwise modeling), and runs on CPU. If your use case is training an ML model to learn deep insights in the data, Tabular Fine-Tuning is your best option.
Data types: Numeric, categorical
Differential privacy: Required; you cannot run without differential privacy
This section compares features of different generative data models supported by Gretel APIs.
✅ = Supported
✖️ = Not yet supported
Tag
tabular_ft
text_ft
tabular_gan
Type
Language Model
Language Model
Generative Adversarial Network
Model
Pre-trained Transformer
Pre-trained Transformer
GAN
Privacy filters
✖️
✖️
✅
Privacy metrics
✅
✖️
✅
Differential privacy
✖️
✅
✖️
✅
✅
✅
Tabular
✅
✖️
✅
Time-series
✅
✖️
✖️
Natural language
✅
✅
✖️
Conditional generation
✖️
✅
✅
Pre-trained
✅
✅
✖️
Gretel cloud
✅
✅
✅
Hybrid cloud
✅
✅
✅
Requires GPU
✅
✅
✅