ViTOptions
Options & Configuration · AiDotNet.VisionLanguage.Encoders
Configuration options for the Vision Transformer (ViT) model.
Inheritance: ModelOptions → VisionEncoderOptions → ViTOptions
How It Works
ViT (Dosovitskiy et al., ICLR 2021) splits an image into fixed-size patches, linearly embeds them, adds position embeddings, and processes the sequence through a standard Transformer encoder. A [CLS] token aggregates information for classification.
Constructors
| Constructor | Summary |
|---|---|
ViTOptions() |
Initializes a new instance with default values. |
ViTOptions(ViTOptions other) |
Initializes a new instance by copying from another instance. |
Properties
| Property | Summary |
|---|---|
PositionalEmbeddingType PositionalEmbedding { get; set; } |
Gets or sets the positional embedding type. |
ViTVariant Variant { get; set; } |
Gets or sets the ViT model variant. |
bool UseClsToken { get; set; } |
Gets or sets whether to use a [CLS] token for classification. |