DocVqaDataLoader<T>

Models & Types · AiDotNet.Data.Vision.Benchmarks

Loads the DocVQA document visual question answering dataset.

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

How It Works

DocVQA expects:

Features are flattened image pixels Tensor[N, H * W * 3]. Labels are answer text encoded as character indices Tensor[N, MaxAnswerLength], where each element is the Unicode code point of the character.

Constructors

Constructor Summary
DocVqaDataLoader(DocVqaDataLoaderOptions options)

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)