REaLTabFormerOptions<T>
Options & Configuration · AiDotNet.Models.Options
Configuration options for REaLTabFormer, a GPT-2 style autoregressive transformer for generating realistic tabular data by treating columns as a sequence of tokens.
Inheritance: ModelOptions → NeuralNetworkOptions → ModelHyperparameterOptions → FinancialNeuralNetworkOptions → RiskModelOptions<T> → SyntheticDataGeneratorOptions<T> → REaLTabFormerOptions<T>
For Beginners
REaLTabFormer treats each row like a sentence and each column value like a word. It generates data by predicting one column at a time:
- Start with a special [START] token
- Predict the first column’s value: “Age = 35”
- Given Age=35, predict next: “Income = 75000”
- Given Age=35, Income=75000, predict: “Education = MS”
- Continue until all columns are filled
This captures column dependencies naturally because each prediction is conditioned on all previously generated columns.
Example:
How It Works
REaLTabFormer tokenizes each column value and generates table rows left-to-right, one column at a time, using a causal (autoregressive) transformer architecture.
Reference: “REaLTabFormer: Generating Realistic Relational and Tabular Data using Transformers” (Solatorio and Dupriez, 2023)
Constructors
| Constructor | Summary |
|---|---|
REaLTabFormerOptions() |
Properties
| Property | Summary |
|---|---|
double DropoutRate { get; set; } |
Gets or sets the dropout rate for attention and feed-forward layers. |
double LearningRate { get; set; } |
Gets or sets the learning rate. |
double Temperature { get; set; } |
Gets or sets the sampling temperature for generation. |
int BatchSize { get; set; } |
Gets or sets the training batch size. |
int EmbeddingDimension { get; set; } |
Gets or sets the embedding dimension for tokens and positions. |
int Epochs { get; set; } |
Gets or sets the number of training epochs. |
int FeedForwardDimension { get; set; } |
Gets or sets the dimension of the feed-forward network in each transformer block. |
int NumBins { get; set; } |
Gets or sets the number of bins for discretizing continuous values. |
int NumHeads { get; set; } |
Gets or sets the number of attention heads per layer. |
int NumLayers { get; set; } |
Gets or sets the number of transformer layers. |