pinecone

Documentation

Index

• type Metric
• type Options
• type SparseValues
• type VectorStore
  • func NewVectorStore(ctx context.Context, embeddings flowllm.Embeddings, opts Options) (*VectorStore, error)
  • func (s *VectorStore) AddDocuments(ctx context.Context, documents ...flowllm.Document) error
  • func (s *VectorStore) SimilaritySearch(ctx context.Context, query string, k int) ([]flowllm.Document, error)
  • func (s *VectorStore) SimilaritySearchVectorWithScore(ctx context.Context, query []float32, k int) ([]flowllm.ScoredDocument, error)

Types

type Metric

type Metric string

const (
	Euclidean  Metric = "euclidean"
	Cosine     Metric = "cosine"
	DotProduct Metric = "dotproduct"
)

type Options

type Options struct {
	ApiKey      string
	Environment string
	Index       string
	NameSpace   string
	Pods        int
	Replicas    int
	PodType     string
	Metric      Metric
}

Options for the Pinecone vector store.

type SparseValues

type SparseValues struct {
	Indices []int     `json:"indices"`
	Values  []float32 `json:"values"`
}

type VectorStore

type VectorStore struct {
	// contains filtered or unexported fields
}

VectorStore is a vector store backed by Pinecone. It requires an already created Pinecone index, with the same dimensionality as the embeddings used to create the store.

func NewVectorStore

func NewVectorStore(ctx context.Context, embeddings flowllm.Embeddings, opts Options) (*VectorStore, error)

NewVectorStore creates a new Pinecone vector store.

func (*VectorStore) AddDocuments

func (s *VectorStore) AddDocuments(ctx context.Context, documents ...flowllm.Document) error

func (*VectorStore) SimilaritySearch

func (s *VectorStore) SimilaritySearch(ctx context.Context, query string, k int) ([]flowllm.Document, error)

func (*VectorStore) SimilaritySearchVectorWithScore

func (s *VectorStore) SimilaritySearchVectorWithScore(ctx context.Context, query []float32, k int) ([]flowllm.ScoredDocument, error)