Azure.Data.AI Namespace

Classes

Name Description
AzureDataAIContext

Context class which will be filled in by the System.ClientModel.SourceGeneration. For more information https://github.com/Azure/azure-sdk-for-net/blob/main/sdk/core/System.ClientModel/src/docs/ModelReaderWriterContext.md

EmbeddingInput

One item to embed, defaulting to text: an optional id and exactly one source, content or url with sasToken.

EmbeddingItem

Common metadata for one embedding result item, in request order. Please note this is the abstract base class. The derived classes available for instantiation are: MultimodalDenseAndMultiVectorEmbedding, MultimodalDenseEmbedding, and MultimodalMultiVectorEmbedding.

EmbeddingOutput

A delivery: link result: a JSON file with the complete response (the same body as an inline response, including data), readable with sasToken until expiresAt.

EmbeddingResult

The result of an embedding operation.

EmbeddingUsage

Common usage for an embedding operation. Token counts are omitted when the model doesn't report them. Please note this is the abstract base class. The derived classes available for instantiation are: MultimodalEmbeddingUsage.

EmbeddingWarning

Something about a successful multimodal embedding response the caller should know.

InferenceClient

Azure Data AI provides semantic reranking operations.

InferenceClientHostExtensions

Extension methods to add InferenceClient to an IHostApplicationBuilder.

InferenceClientOptions

Client options for InferenceClient.

InferenceClientSettings

Represents the settings used to configure a InferenceClient that can be loaded from an IConfigurationSection.

InferenceModelFactory

A factory class for creating instances of the models for mocking.

LatencyResult

The result of the latency in the semantic reranking operation.

MultimodalDenseAndMultiVectorEmbedding

The result for one item when vectorKind is dense_multi: a dense vector and one vector per token, from one model run.

MultimodalDenseEmbedding

The result for one item when vectorKind is dense: one vector.

MultimodalEmbeddingUsage

Usage for a multimodal embedding operation.

MultimodalMultiVectorEmbedding

The result for one item when vectorKind is multi: one vector per token, for late-interaction (MaxSim) scoring.

SemanticRerankingContent

The input for a semantic reranking operation.

SemanticRerankingMetadata

Additional metadata about the semantic reranking operation.

SemanticRerankingResult

The result of a semantic reranking operation.

SemanticRerankingScore

A relevance score assigned to a submitted document.

SentenceScore

A sentence-level relevance score for a submitted document.

TokenUsageResult

The result of the token usage in the semantic reranking operation.

Structs

Name Description
EmbeddingContentKind

The kind of an item to embed.

EmbeddingDelivery

How the results of a multimodal embedding operation are returned.

EmbeddingEncodingFormat

How vectors are encoded in the response.

EmbeddingImageFormat

The format of an embedded image.

EmbeddingOutputDataKind

The numeric type of the returned vector values.

EmbeddingVectorKind

The kind of vectors to return.

EmbeddingWarningCode

A machine-readable embedding warning code.

SemanticRerankingDocumentType

The format of a document submitted for semantic reranking.

Enums

Name Description
InferenceClientOptions.ServiceVersion

The version of the service to use.