SearchModelFactory.HnswParameters Method
Definition
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.
Contains the parameters specific to the HNSW algorithm.
public static Azure.Search.Documents.Indexes.Models.HnswParameters HnswParameters(int? m = default, int? efConstruction = default, int? efSearch = default, Azure.Search.Documents.Indexes.Models.VectorSearchAlgorithmMetric? metric = default);static member HnswParameters : Nullable<int> * Nullable<int> * Nullable<int> * Nullable<Azure.Search.Documents.Indexes.Models.VectorSearchAlgorithmMetric> -> Azure.Search.Documents.Indexes.Models.HnswParametersPublic Shared Function HnswParameters (Optional m As Nullable(Of Integer) = Nothing, Optional efConstruction As Nullable(Of Integer) = Nothing, Optional efSearch As Nullable(Of Integer) = Nothing, Optional metric As Nullable(Of VectorSearchAlgorithmMetric) = Nothing) As HnswParametersParameters
The number of bi-directional links created for every new element during construction. Increasing this parameter value may improve recall and reduce retrieval times for datasets with high intrinsic dimensionality at the expense of increased memory consumption and longer indexing time.
The size of the dynamic list containing the nearest neighbors, which is used during index time. Increasing this parameter may improve index quality, at the expense of increased indexing time. At a certain point, increasing this parameter leads to diminishing returns.
The size of the dynamic list containing the nearest neighbors, which is used during search time. Increasing this parameter may improve search results, at the expense of slower search. At a certain point, increasing this parameter leads to diminishing returns.
- metric
- Nullable<VectorSearchAlgorithmMetric>
The similarity metric to use for vector comparisons.
Returns
A new HnswParameters instance for mocking.
