Documentation
¶
Overview ¶
Package embed implements the optional embeddings path decided in https://github.com/cheetahbyte/lore/issues/7: vector search is opt-in, never generated in-process (no bundled/local model, per the research on issue #6) — always an external API call.
Index ¶
Constants ¶
This section is empty.
Variables ¶
This section is empty.
Functions ¶
This section is empty.
Types ¶
type OpenAICompatible ¶
type OpenAICompatible struct {
HTTPClient *http.Client
Endpoint string // e.g. https://api.openai.com/v1 or http://localhost:11434/v1
APIKey string
Model string
}
OpenAICompatible talks to any embeddings endpoint that speaks the OpenAI /v1/embeddings request/response shape — this covers both configured providers from issue #3 (config.toml's embeddings.provider: "openai" or "ollama", since Ollama serves an OpenAI-compatible endpoint).
func New ¶
func New(endpoint, apiKey, model string) *OpenAICompatible
Source Files
Click to show internal directories.
Click to hide internal directories.