TruthfulQaDataLoaderOptions

Options & Configuration ยท AiDotNet.Data.Text.Benchmarks

Configuration for the TruthfulQA benchmark loader (Lin et al. 2022).

How It Works

TruthfulQA tests whether language models avoid common false beliefs in 38 categories โ€” 817 questions, each with a best correct answer and multiple correct/incorrect distractors. This loader exposes the generation-style version (best_answer as target).

Constructors

Constructor Summary
TruthfulQaDataLoaderOptions()

Properties

Property Summary
DatasetSplit Split { get; set; } Gets or sets which split of the dataset is loaded.
bool AutoDownload { get; set; } Gets or sets whether the dataset is downloaded into the data directory when it is not there.
int MaxAnswerLength { get; set; } Gets or sets the token length each answer is padded or truncated to.
int MaxQuestionLength { get; set; } Gets or sets the token length each question is padded or truncated to.
int VocabularySize { get; set; } Gets or sets the most distinct tokens in the vocabulary built from the loaded text.
int? MaxSamples { get; set; } Gets or sets the most samples loaded; null loads the whole split.
string DataPath { get; set; } Gets or sets the dataset root directory; null uses the default data directory for this dataset.

Methods

Method Summary
void Validate() Validates that all option values are within acceptable ranges.