Documentation
¶
Overview ¶
Package localllm manages the lifecycle of the local llama.cpp HTTP server SpeechKit ships alongside the desktop bundle for offline LLM inference. Owns process spawn, health-probing, graceful shutdown, and the OpenAI-compatible client wrapper the Assist + Voice Agent paths route to when Local is the selected provider.
Audit 2026-05-24 maintainability sweep.
Index ¶
Constants ¶
This section is empty.
Variables ¶
This section is empty.
Functions ¶
func FindServerBinary ¶
func ValidateModelPath ¶
ValidateModelPath verifies that path points at a safe GGUF model filename.
Types ¶
type InstallStatus ¶
type Server ¶
type Server struct {
BaseURL string
Port int
ModelPath string
GPU string
ContextSize int
Threads int
Validation netsec.ValidationOptions
// contains filtered or unexported fields
}
Server manages SpeechKit's bundled llama.cpp OpenAI-compatible server.
func (*Server) RuntimeKey ¶
func (*Server) StopServer ¶
func (s *Server) StopServer()
func (*Server) VerifyInstallation ¶
func (s *Server) VerifyInstallation() InstallStatus
Click to show internal directories.
Click to hide internal directories.