HellaswagDataLoader<T>

Models & Types · AiDotNet.Data.Text.Benchmarks

Loads the HellaSwag 4-way commonsense NLI benchmark (Zellers et al. 2019).

Inheritance: DataLoaderBase<T> → InputOutputDataLoaderBase<T, Tensor<T>, Tensor<T>> → HellaswagDataLoader<T>

How It Works

Expects:

Auto-download fetches the canonical Rowan Zellers GitHub release. Features Tensor[N, 4 * MaxSequenceLength] holds the 4 candidate (context + ending) sequences concatenated; labels are one-hot 4-class vectors Tensor[N, 4]. Test split labels are filled with the uniform prior since they aren’t released publicly.

Constructors

Constructor Summary
HellaswagDataLoader(HellaswagDataLoaderOptions options) Creates a new HellaSwag data loader.

Properties

Property Summary
int FeatureCount { get; }
int OutputDimension { get; }
int TotalCount { get; }
string Description { get; }
string Name { get; }

Methods

Method Summary
ValueTuple<IInputOutputDataLoader<T, Tensor<T>, Tensor<T>>, IInputOutputDataLoader<T, Tensor<T>, Tensor<T>>, IInputOutputDataLoader<T, Tensor<T>, Tensor<T>>> Split(double trainRatio, double validationRatio, int? seed)