Documentation
¶
Overview ¶
Package embedder defines the interface for generating vector embeddings from text.
Index ¶
Constants ¶
This section is empty.
Variables ¶
var ErrNotImplemented = errors.New("weave: not implemented")
ErrNotImplemented is returned when a feature is not yet implemented.
Functions ¶
This section is empty.
Types ¶
type EmbedResult ¶
type EmbedResult struct {
// Vector is the embedding vector.
Vector []float32 `json:"vector"`
// TokenCount is the number of tokens consumed.
TokenCount int `json:"token_count"`
}
EmbedResult contains the embedding vector for a single text input.
type Embedder ¶
type Embedder interface {
// Embed generates embeddings for the given texts.
Embed(ctx context.Context, texts []string) ([]EmbedResult, error)
// Dimensions returns the dimensionality of the embeddings produced
// by this embedder.
Dimensions() int
}
Embedder generates vector embeddings from text.
type LocalEmbedder ¶
type LocalEmbedder struct {
// contains filtered or unexported fields
}
LocalEmbedder is a placeholder for local model embedding (e.g., ONNX).
func NewLocalEmbedder ¶
func NewLocalEmbedder(dimensions int) *LocalEmbedder
NewLocalEmbedder creates a placeholder local embedder.
func (*LocalEmbedder) Describe ¶
func (*LocalEmbedder) Describe() weave.ComponentInfo
Describe reports the local embedder, which is not implemented yet: every Embed call returns ErrNotImplemented.
func (*LocalEmbedder) Dimensions ¶
func (e *LocalEmbedder) Dimensions() int
Dimensions returns the configured dimensionality.
func (*LocalEmbedder) Embed ¶
func (e *LocalEmbedder) Embed(_ context.Context, _ []string) ([]EmbedResult, error)
Embed is not yet implemented.
type OpenAIEmbedder ¶
type OpenAIEmbedder struct {
// contains filtered or unexported fields
}
OpenAIEmbedder generates embeddings using the OpenAI Embeddings API.
func NewOpenAIEmbedder ¶
func NewOpenAIEmbedder(apiKey string, opts ...OpenAIOption) *OpenAIEmbedder
NewOpenAIEmbedder creates an OpenAI embedder.
func (*OpenAIEmbedder) Describe ¶
func (e *OpenAIEmbedder) Describe() weave.ComponentInfo
Describe reports the OpenAI embedder and its model.
func (*OpenAIEmbedder) Dimensions ¶
func (e *OpenAIEmbedder) Dimensions() int
Dimensions returns the embedding dimensionality.
func (*OpenAIEmbedder) Embed ¶
func (e *OpenAIEmbedder) Embed(ctx context.Context, texts []string) ([]EmbedResult, error)
Embed generates embeddings for the given texts.
type OpenAIOption ¶
type OpenAIOption func(*OpenAIEmbedder)
OpenAIOption configures the OpenAI embedder.
func WithOpenAIBaseURL ¶
func WithOpenAIBaseURL(baseURL string) OpenAIOption
WithOpenAIBaseURL overrides the base URL for API-compatible endpoints.
func WithOpenAIClient ¶
func WithOpenAIClient(c *http.Client) OpenAIOption
WithOpenAIClient sets a custom HTTP client. The embedder traces each call itself, so a client whose transport is already traced (forge.HTTPClient, otelhttp) records the call twice.
func WithOpenAIDimensions ¶
func WithOpenAIDimensions(dims int) OpenAIOption
WithOpenAIDimensions sets the output dimensions.
func WithOpenAIModel ¶
func WithOpenAIModel(model string) OpenAIOption
WithOpenAIModel sets the model name (default: text-embedding-3-small).
type StatusError ¶
type StatusError struct {
// StatusCode is the HTTP status the endpoint returned.
StatusCode int
}
StatusError is returned when the embeddings endpoint answers with a status other than 200 OK. Telemetry names a failed call by this type and the status code, never by the response body.
func (*StatusError) Error ¶
func (e *StatusError) Error() string
Source Files
Directories