IEmbeddingGenerator<TInput,TEmbedding> Interface

Reference

Definition

Namespace:: Microsoft.Extensions.AI

Assembly:: Microsoft.Extensions.AI.Abstractions.dll

Package:: Microsoft.Extensions.AI.Abstractions v9.0.0-preview.9.24556.5

Important

Some information relates to prerelease product that may be substantially modified before it’s released. Microsoft makes no warranties, express or implied, with respect to the information provided here.

Represents a generator of embeddings.

generic <typename TInput, typename TEmbedding>
 where TEmbedding : Embeddingpublic interface class IEmbeddingGenerator : IDisposable

public interface IEmbeddingGenerator<TInput,TEmbedding> : IDisposable where TEmbedding : Embedding

type IEmbeddingGenerator<'Input, 'Embedding (requires 'Embedding :> Embedding)> = interface
    interface IDisposable

Public Interface IEmbeddingGenerator(Of TInput, TEmbedding)
Implements IDisposable

Type Parameters

TInput

The type from which embeddings will be generated.

TEmbedding

The type of embeddings to generate.

Derived: Microsoft.Extensions.AI.AzureAIInferenceEmbeddingGenerator

Microsoft.Extensions.AI.DelegatingEmbeddingGenerator<TInput,TEmbedding>

Microsoft.Extensions.AI.OllamaEmbeddingGenerator

Microsoft.Extensions.AI.OpenAIEmbeddingGenerator

Implements: IDisposable

Remarks

Unless otherwise specified, all members of IEmbeddingGenerator<TInput,TEmbedding> are thread-safe for concurrent use. It is expected that all implementations of IEmbeddingGenerator<TInput,TEmbedding> support being used by multiple requests concurrently.

However, implementations of IEmbeddingGenerator<TInput,TEmbedding> may mutate the arguments supplied to GenerateAsync(IEnumerable<TInput>, EmbeddingGenerationOptions, CancellationToken), such as by adding additional values to the values list or configuring the options instance. Thus, consumers of the interface either should avoid using shared instances of these arguments for concurrent invocations or should otherwise ensure by construction that no IEmbeddingGenerator<TInput,TEmbedding> instances are used which might employ such mutation.

Properties

Metadata

Gets metadata that describes the IEmbeddingGenerator<TInput,TEmbedding>.

Methods

Dispose()	Performs application-defined tasks associated with freeing, releasing, or resetting unmanaged resources. (Inherited from IDisposable)
GenerateAsync(IEnumerable<TInput>, EmbeddingGenerationOptions, CancellationToken)	Generates embeddings for each of the supplied `values`.
GetService<TService>(Object)	Asks the IEmbeddingGenerator<TInput,TEmbedding> for an object of type `TService`.

Extension Methods

GenerateAndZipAsync<TInput,TEmbedding>(IEmbeddingGenerator<TInput,TEmbedding>, IEnumerable<TInput>, EmbeddingGenerationOptions, CancellationToken)	Generates embeddings for each of the supplied `values` and produces a list that pairs each input value with its resulting embedding.
GenerateEmbeddingAsync<TInput,TEmbedding>(IEmbeddingGenerator<TInput,TEmbedding>, TInput, EmbeddingGenerationOptions, CancellationToken)	Generates an embedding from the specified `value`.

Applies to

แชร์ผ่าน