VisionLanguageInputOptions

Options & Configuration · AiDotNet.NeuralNetworks.Options

Shared image, text-context, and output dimensions for vision-language models.

Inheritance: ModelOptions → NeuralNetworkOptions → ModelHyperparameterOptions → VisionLanguageInputOptions

Known derived types: ClipOptions, VisionLanguageModelOptions

For Beginners

These settings describe the pictures and text supplied to a model and the width of its embeddings. They do not describe or rebuild its internal layers. For an ONNX model, the configured dimensions must match the loaded graphs.

How It Works

Native tower settings belong to VisionLanguageModelOptions. Inference-only models can use this input contract without advertising unused tower setters.

Properties

Property Summary
int EmbeddingDimension { get; set; } Gets or sets the width of the shared image/text embedding space.
int ImageSize { get; set; } Gets or sets the side length, in pixels, of the square input image.
int MaxSequenceLength { get; set; } Gets or sets the maximum text sequence length, in tokens.