M.I.M.I.R - Multi-agent Intelligent Memory & Insight Repository

//go:build cgo && !nolocalllm && (darwin || linux) // Package heimdall provides the Heimdall cognitive guardian for NornicDB. // This file provides the CGO-enabled generator using localllm. package heimdall import ( "context" "github.com/orneryd/nornicdb/pkg/localllm" ) func init() { // Register the CGO-enabled generator loader SetGeneratorLoader(cgoGeneratorLoader) } // cgoGeneratorLoader loads a generation model using localllm CGO bindings. func cgoGeneratorLoader(modelPath string, gpuLayers, contextSize, batchSize int) (Generator, error) { opts := localllm.DefaultGenerationOptions(modelPath) opts.GPULayers = gpuLayers opts.ContextSize = contextSize opts.BatchSize = batchSize model, err := localllm.LoadGenerationModel(opts) if err != nil { return nil, err } return &cgoGenerator{model: model}, nil } // cgoGenerator wraps localllm.GenerationModel to implement Generator interface. type cgoGenerator struct { model *localllm.GenerationModel } func (g *cgoGenerator) Generate(ctx context.Context, prompt string, params GenerateParams) (string, error) { llamaParams := localllm.GenerateParams{ MaxTokens: params.MaxTokens, Temperature: params.Temperature, TopP: params.TopP, TopK: params.TopK, StopTokens: params.StopTokens, } return g.model.Generate(ctx, prompt, llamaParams) } func (g *cgoGenerator) GenerateStream(ctx context.Context, prompt string, params GenerateParams, callback func(token string) error) error { llamaParams := localllm.GenerateParams{ MaxTokens: params.MaxTokens, Temperature: params.Temperature, TopP: params.TopP, TopK: params.TopK, StopTokens: params.StopTokens, } return g.model.GenerateStream(ctx, prompt, llamaParams, callback) } func (g *cgoGenerator) Close() error { return g.model.Close() } func (g *cgoGenerator) ModelPath() string { return g.model.ModelPath() }