Embedding interface

List of embeddings with an optional timestamp.

One and only one of the two 'floatEmbedding' and 'quantizedEmbedding' will contain data, based on whether or not the embedder was configured to perform scalar quantization.

Signature:

export declare interface Embedding 

Properties

Property Type Description
floatEmbedding number[] Floating-point embedding. Empty if the embedder was configured to perform scalar-quantization.
headIndex number The index of the classifier head these categories refer to. This is useful for multi-head models.
headName string The name of the classifier head, which is the corresponding tensor metadata name.
quantizedEmbedding Uint8Array Scalar-quantized embedding. Empty if the embedder was not configured to perform scalar quantization.

Embedding.floatEmbedding

Floating-point embedding. Empty if the embedder was configured to perform scalar-quantization.

Signature:

floatEmbedding?: number[];

Embedding.headIndex

The index of the classifier head these categories refer to. This is useful for multi-head models.

Signature:

headIndex: number;

Embedding.headName

The name of the classifier head, which is the corresponding tensor metadata name.

Signature:

headName: string;

Embedding.quantizedEmbedding

Scalar-quantized embedding. Empty if the embedder was not configured to perform scalar quantization.

Signature:

quantizedEmbedding?: Uint8Array;