embedding

package

v0.1.0-round1 Latest Latest Go to latest Published: Aug 25, 2026 License: Apache-2.0 Imports: 5 Imported by: 0

Details

Package embedding adapts the standalone vector engine to WAS domain contracts.

This section is empty.

This section is empty.

This section is empty.

type Config = vector.Config

Config identifies the pinned vector model, tokenizer, and runtime artifacts.

type E5 struct {
	// contains filtered or unexported fields
}

E5 translates WAS processing profiles and errors at the embedding seam.

func New(config Config) (*E5, error)

New opens the standalone vector engine behind the WAS embedding seam.

func (e *E5) Close() error

Close releases the local inference runtime.

func (e *E5) CountTokens(text string) (int, error)

CountTokens returns the encoded passage length used by the vector engine.

func (e *E5) EmbedPassages(
	ctx context.Context,
	passages []string,
	profile knowledge.Profile,
) ([][]float32, error)

EmbedPassages creates one normalized vector per passage for profile.

func (e *E5) EmbedQuery(ctx context.Context, query string, profile knowledge.Profile) ([]float32, error)

EmbedQuery creates one normalized query vector for profile.

func (e *E5) Ready(ctx context.Context) error

Ready reports whether the vector engine can accept work.