feat: refactor build process, drop embedded backends (#5875)

* feat: split remaining backends and drop embedded backends - Drop silero-vad, huggingface, and stores backend from embedded binaries - Refactor Makefile and Dockerfile to avoid building grpc backends - Drop golang code that was used to embed backends - Simplify building by using goreleaser Signed-off-by: Ettore Di Giacinto <mudler@localai.io> * chore(gallery): be specific with llama-cpp backend templates Signed-off-by: Ettore Di Giacinto <mudler@localai.io> * chore(docs): update Signed-off-by: Ettore Di Giacinto <mudler@localai.io> * chore(ci): minor fixes Signed-off-by: Ettore Di Giacinto <mudler@localai.io> * chore: drop all ffmpeg references Signed-off-by: Ettore Di Giacinto <mudler@localai.io> * fix: run protogen-go Signed-off-by: Ettore Di Giacinto <mudler@localai.io> * Always enable p2p mode Signed-off-by: Ettore Di Giacinto <mudler@localai.io> * Update gorelease file Signed-off-by: Ettore Di Giacinto <mudler@localai.io> * fix(stores): do not always load Signed-off-by: Ettore Di Giacinto <mudler@localai.io> * Fix linting issues Signed-off-by: Ettore Di Giacinto <mudler@localai.io> * Simplify Signed-off-by: Ettore Di Giacinto <mudler@localai.io> * Mac OS fixup Signed-off-by: Ettore Di Giacinto <mudler@localai.io> --------- Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
2026-05-24 16:51:44 -04:00 · 2025-07-22 16:31:04 +02:00
parent e29b2c3aff
commit 98e5291afc
118 changed files with 631 additions and 1339 deletions
--- a/core/application/startup.go
+++ b/core/application/startup.go
@@ -9,9 +9,7 @@ import (
 	"github.com/mudler/LocalAI/core/gallery"
 	"github.com/mudler/LocalAI/core/services"
 	"github.com/mudler/LocalAI/internal"
-	"github.com/mudler/LocalAI/pkg/assets"

-	"github.com/mudler/LocalAI/pkg/library"
 	"github.com/mudler/LocalAI/pkg/model"
 	pkgStartup "github.com/mudler/LocalAI/pkg/startup"
 	"github.com/mudler/LocalAI/pkg/xsysinfo"
@@ -103,23 +101,6 @@ func New(opts ...config.AppOption) (*Application, error) {
 		}
 	}

-	if options.AssetsDestination != "" {
-		// Extract files from the embedded FS
-		err := assets.ExtractFiles(options.BackendAssets, options.AssetsDestination)
-		log.Debug().Msgf("Extracting backend assets files to %s", options.AssetsDestination)
-		if err != nil {
-			log.Warn().Msgf("Failed extracting backend assets files: %s (might be required for some backends to work properly)", err)
-		}
-	}
-
-	if options.LibPath != "" {
-		// If there is a lib directory, set LD_LIBRARY_PATH to include it
-		err := library.LoadExternal(options.LibPath)
-		if err != nil {
-			log.Error().Err(err).Str("LibPath", options.LibPath).Msg("Error while loading external libraries")
-		}
-	}
-
 	// turn off any process that was started by GRPC if the context is canceled
 	go func() {
 		<-options.Context.Done()
--- a/core/backend/options.go
+++ b/core/backend/options.go
@@ -20,7 +20,6 @@ func ModelOptions(c config.BackendConfig, so *config.ApplicationConfig, opts ...
 	defOpts := []model.Option{
 		model.WithBackendString(c.Backend),
 		model.WithModel(c.Model),
-		model.WithAssetDir(so.AssetsDestination),
 		model.WithContext(so.Context),
 		model.WithModelID(name),
 	}
--- a/core/backend/stores.go
+++ b/core/backend/stores.go
@@ -7,14 +7,12 @@ import (
 	"github.com/mudler/LocalAI/pkg/model"
 )

-func StoreBackend(sl *model.ModelLoader, appConfig *config.ApplicationConfig, storeName string) (grpc.Backend, error) {
-	if storeName == "" {
-		storeName = "default"
+func StoreBackend(sl *model.ModelLoader, appConfig *config.ApplicationConfig, storeName string, backend string) (grpc.Backend, error) {
+	if backend == "" {
+		backend = model.LocalStoreBackend
 	}
-
 	sc := []model.Option{
-		model.WithBackendString(model.LocalStoreBackend),
-		model.WithAssetDir(appConfig.AssetsDestination),
+		model.WithBackendString(backend),
 		model.WithModel(storeName),
 	}

--- a/core/cli/context/context.go
+++ b/core/cli/context/context.go
@@ -1,13 +1,6 @@
 package cliContext

-import (
-	rice "github.com/GeertJohan/go.rice"
-)
-
 type Context struct {
 	Debug    bool    `env:"LOCALAI_DEBUG,DEBUG" default:"false" hidden:"" help:"DEPRECATED, use --log-level=debug instead. Enable debug logging"`
 	LogLevel *string `env:"LOCALAI_LOG_LEVEL" enum:"error,warn,info,debug,trace" help:"Set the level of logs to output [${enum}]"`
-
-	// This field is not a command line argument/flag, the struct tag excludes it from the parsed CLI
-	BackendAssets *rice.Box `kong:"-"`
 }
--- a/core/cli/run.go
+++ b/core/cli/run.go
@@ -23,7 +23,6 @@ type RunCMD struct {
 	ExternalBackends             []string      `env:"LOCALAI_EXTERNAL_BACKENDS,EXTERNAL_BACKENDS" help:"A list of external backends to load from gallery on boot" group:"backends"`
 	BackendsPath                 string        `env:"LOCALAI_BACKENDS_PATH,BACKENDS_PATH" type:"path" default:"${basepath}/backends" help:"Path containing backends used for inferencing" group:"backends"`
 	ModelsPath                   string        `env:"LOCALAI_MODELS_PATH,MODELS_PATH" type:"path" default:"${basepath}/models" help:"Path containing models used for inferencing" group:"storage"`
-	BackendAssetsPath            string        `env:"LOCALAI_BACKEND_ASSETS_PATH,BACKEND_ASSETS_PATH" type:"path" default:"/tmp/localai/backend_data" help:"Path used to extract libraries that are required by some of the backends in runtime" group:"storage"`
 	GeneratedContentPath         string        `env:"LOCALAI_GENERATED_CONTENT_PATH,GENERATED_CONTENT_PATH" type:"path" default:"/tmp/generated/content" help:"Location for generated content (e.g. images, audio, videos)" group:"storage"`
 	UploadPath                   string        `env:"LOCALAI_UPLOAD_PATH,UPLOAD_PATH" type:"path" default:"/tmp/localai/upload" help:"Path to store uploads from files api" group:"storage"`
 	ConfigPath                   string        `env:"LOCALAI_CONFIG_PATH,CONFIG_PATH" default:"/tmp/localai/config" group:"storage"`
@@ -46,7 +45,6 @@ type RunCMD struct {
 	Address                            string   `env:"LOCALAI_ADDRESS,ADDRESS" default:":8080" help:"Bind address for the API server" group:"api"`
 	CORS                               bool     `env:"LOCALAI_CORS,CORS" help:"" group:"api"`
 	CORSAllowOrigins                   string   `env:"LOCALAI_CORS_ALLOW_ORIGINS,CORS_ALLOW_ORIGINS" group:"api"`
-	LibraryPath                        string   `env:"LOCALAI_LIBRARY_PATH,LIBRARY_PATH" help:"Path to the library directory (for e.g. external libraries used by backends)" default:"/usr/share/local-ai/libs" group:"backends"`
 	CSRF                               bool     `env:"LOCALAI_CSRF" help:"Enables fiber CSRF middleware" group:"api"`
 	UploadLimit                        int      `env:"LOCALAI_UPLOAD_LIMIT,UPLOAD_LIMIT" default:"15" help:"Default upload-limit in MB" group:"api"`
 	APIKeys                            []string `env:"LOCALAI_API_KEY,API_KEY" help:"List of API Keys to enable API authentication. When this is set, all the requests must be authenticated with one of these API keys" group:"api"`
@@ -99,10 +97,7 @@ func (r *RunCMD) Run(ctx *cliContext.Context) error {
 		config.WithCors(r.CORS),
 		config.WithCorsAllowOrigins(r.CORSAllowOrigins),
 		config.WithCsrf(r.CSRF),
-		config.WithLibPath(r.LibraryPath),
 		config.WithThreads(r.Threads),
-		config.WithBackendAssets(ctx.BackendAssets),
-		config.WithBackendAssetsOutput(r.BackendAssetsPath),
 		config.WithUploadLimitMB(r.UploadLimit),
 		config.WithApiKeys(r.APIKeys),
 		config.WithModelsURL(append(r.Models, r.ModelArgs...)...),
--- a/core/cli/soundgeneration.go
+++ b/core/cli/soundgeneration.go
@@ -27,7 +27,6 @@ type SoundGenerationCMD struct {
 	DoSample               bool     `short:"s" default:"true" help:"Enables sampling from the model. Better quality at the cost of speed. Defaults to enabled."`
 	OutputFile             string   `short:"o" type:"path" help:"The path to write the output wav file"`
 	ModelsPath             string   `env:"LOCALAI_MODELS_PATH,MODELS_PATH" type:"path" default:"${basepath}/models" help:"Path containing models used for inferencing" group:"storage"`
-	BackendAssetsPath      string   `env:"LOCALAI_BACKEND_ASSETS_PATH,BACKEND_ASSETS_PATH" type:"path" default:"/tmp/localai/backend_data" help:"Path used to extract libraries that are required by some of the backends in runtime" group:"storage"`
 	ExternalGRPCBackends   []string `env:"LOCALAI_EXTERNAL_GRPC_BACKENDS,EXTERNAL_GRPC_BACKENDS" help:"A list of external grpc backends" group:"backends"`
 }

@@ -51,11 +50,10 @@ func parseToInt32Ptr(input string) *int32 {

 func (t *SoundGenerationCMD) Run(ctx *cliContext.Context) error {
 	outputFile := t.OutputFile
-	outputDir := t.BackendAssetsPath
+	outputDir := os.TempDir()
 	if outputFile != "" {
 		outputDir = filepath.Dir(outputFile)
 	}
-
 	text := strings.Join(t.Text, " ")

 	externalBackends := make(map[string]string)
@@ -71,7 +69,6 @@ func (t *SoundGenerationCMD) Run(ctx *cliContext.Context) error {
 		ModelPath:            t.ModelsPath,
 		Context:              context.Background(),
 		GeneratedContentDir:  outputDir,
-		AssetsDestination:    t.BackendAssetsPath,
 		ExternalGRPCBackends: externalBackends,
 	}
 	ml := model.NewModelLoader(opts.ModelPath, opts.SingleBackend)
--- a/core/cli/transcript.go
+++ b/core/cli/transcript.go
@@ -15,20 +15,18 @@ import (
 type TranscriptCMD struct {
 	Filename string `arg:""`

-	Backend           string `short:"b" default:"whisper" help:"Backend to run the transcription model"`
-	Model             string `short:"m" required:"" help:"Model name to run the TTS"`
-	Language          string `short:"l" help:"Language of the audio file"`
-	Translate         bool   `short:"c" help:"Translate the transcription to english"`
-	Threads           int    `short:"t" default:"1" help:"Number of threads used for parallel computation"`
-	ModelsPath        string `env:"LOCALAI_MODELS_PATH,MODELS_PATH" type:"path" default:"${basepath}/models" help:"Path containing models used for inferencing" group:"storage"`
-	BackendAssetsPath string `env:"LOCALAI_BACKEND_ASSETS_PATH,BACKEND_ASSETS_PATH" type:"path" default:"/tmp/localai/backend_data" help:"Path used to extract libraries that are required by some of the backends in runtime" group:"storage"`
+	Backend    string `short:"b" default:"whisper" help:"Backend to run the transcription model"`
+	Model      string `short:"m" required:"" help:"Model name to run the TTS"`
+	Language   string `short:"l" help:"Language of the audio file"`
+	Translate  bool   `short:"c" help:"Translate the transcription to english"`
+	Threads    int    `short:"t" default:"1" help:"Number of threads used for parallel computation"`
+	ModelsPath string `env:"LOCALAI_MODELS_PATH,MODELS_PATH" type:"path" default:"${basepath}/models" help:"Path containing models used for inferencing" group:"storage"`
 }

 func (t *TranscriptCMD) Run(ctx *cliContext.Context) error {
 	opts := &config.ApplicationConfig{
-		ModelPath:         t.ModelsPath,
-		Context:           context.Background(),
-		AssetsDestination: t.BackendAssetsPath,
+		ModelPath: t.ModelsPath,
+		Context:   context.Background(),
 	}

 	cl := config.NewBackendConfigLoader(t.ModelsPath)
--- a/core/cli/tts.go
+++ b/core/cli/tts.go
@@ -17,18 +17,17 @@ import (
 type TTSCMD struct {
 	Text []string `arg:""`

-	Backend           string `short:"b" default:"piper" help:"Backend to run the TTS model"`
-	Model             string `short:"m" required:"" help:"Model name to run the TTS"`
-	Voice             string `short:"v" help:"Voice name to run the TTS"`
-	Language          string `short:"l" help:"Language to use with the TTS"`
-	OutputFile        string `short:"o" type:"path" help:"The path to write the output wav file"`
-	ModelsPath        string `env:"LOCALAI_MODELS_PATH,MODELS_PATH" type:"path" default:"${basepath}/models" help:"Path containing models used for inferencing" group:"storage"`
-	BackendAssetsPath string `env:"LOCALAI_BACKEND_ASSETS_PATH,BACKEND_ASSETS_PATH" type:"path" default:"/tmp/localai/backend_data" help:"Path used to extract libraries that are required by some of the backends in runtime" group:"storage"`
+	Backend    string `short:"b" default:"piper" help:"Backend to run the TTS model"`
+	Model      string `short:"m" required:"" help:"Model name to run the TTS"`
+	Voice      string `short:"v" help:"Voice name to run the TTS"`
+	Language   string `short:"l" help:"Language to use with the TTS"`
+	OutputFile string `short:"o" type:"path" help:"The path to write the output wav file"`
+	ModelsPath string `env:"LOCALAI_MODELS_PATH,MODELS_PATH" type:"path" default:"${basepath}/models" help:"Path containing models used for inferencing" group:"storage"`
 }

 func (t *TTSCMD) Run(ctx *cliContext.Context) error {
 	outputFile := t.OutputFile
-	outputDir := t.BackendAssetsPath
+	outputDir := os.TempDir()
 	if outputFile != "" {
 		outputDir = filepath.Dir(outputFile)
 	}
@@ -39,7 +38,6 @@ func (t *TTSCMD) Run(ctx *cliContext.Context) error {
 		ModelPath:           t.ModelsPath,
 		Context:             context.Background(),
 		GeneratedContentDir: outputDir,
-		AssetsDestination:   t.BackendAssetsPath,
 	}
 	ml := model.NewModelLoader(opts.ModelPath, opts.SingleBackend)

--- a/core/cli/worker/worker.go
+++ b/core/cli/worker/worker.go
@@ -1,7 +1,7 @@
 package worker

 type WorkerFlags struct {
-	BackendAssetsPath string `env:"LOCALAI_BACKEND_ASSETS_PATH,BACKEND_ASSETS_PATH" type:"path" default:"/tmp/localai/backend_data" help:"Path used to extract libraries that are required by some of the backends in runtime" group:"storage"`
+	BackendsPath      string `env:"LOCALAI_BACKENDS_PATH,BACKENDS_PATH" type:"path" default:"${basepath}/backends" help:"Path containing backends used for inferencing" group:"backends"`
 	ExtraLLamaCPPArgs string `name:"llama-cpp-args" env:"LOCALAI_EXTRA_LLAMA_CPP_ARGS,EXTRA_LLAMA_CPP_ARGS" help:"Extra arguments to pass to llama-cpp-rpc-server"`
 }

--- a/core/cli/worker/worker_llamacpp.go
+++ b/core/cli/worker/worker_llamacpp.go
@@ -9,8 +9,6 @@ import (

 	cliContext "github.com/mudler/LocalAI/core/cli/context"
 	"github.com/mudler/LocalAI/core/gallery"
-	"github.com/mudler/LocalAI/pkg/assets"
-	"github.com/mudler/LocalAI/pkg/library"
 	"github.com/rs/zerolog/log"
 )

@@ -47,24 +45,17 @@ func findLLamaCPPBackend(backendSystemPath string) (string, error) {
 }

 func (r *LLamaCPP) Run(ctx *cliContext.Context) error {
-	// Extract files from the embedded FS
-	err := assets.ExtractFiles(ctx.BackendAssets, r.BackendAssetsPath)
-	log.Debug().Msgf("Extracting backend assets files to %s", r.BackendAssetsPath)
-	if err != nil {
-		log.Warn().Msgf("Failed extracting backend assets files: %s (might be required for some backends to work properly)", err)
-	}

 	if len(os.Args) < 4 {
 		return fmt.Errorf("usage: local-ai worker llama-cpp-rpc -- <llama-rpc-server-args>")
 	}

-	grpcProcess, err := findLLamaCPPBackend(r.BackendAssetsPath)
+	grpcProcess, err := findLLamaCPPBackend(r.BackendsPath)
 	if err != nil {
 		return err
 	}

 	args := strings.Split(r.ExtraLLamaCPPArgs, " ")
-	args, grpcProcess = library.LoadLDSO(r.BackendAssetsPath, args, grpcProcess)

 	args = append([]string{grpcProcess}, args...)
 	return syscall.Exec(
--- a/core/cli/worker/worker_nop2p.go
+++ b/core/cli/worker/worker_nop2p.go
@@ -1,16 +0,0 @@
-//go:build !p2p
-// +build !p2p
-
-package worker
-
-import (
-	"fmt"
-
-	cliContext "github.com/mudler/LocalAI/core/cli/context"
-)
-
-type P2P struct{}
-
-func (r *P2P) Run(ctx *cliContext.Context) error {
-	return fmt.Errorf("p2p mode is not enabled in this build")
-}
--- a/core/cli/worker/worker_p2p.go
+++ b/core/cli/worker/worker_p2p.go
@@ -1,6 +1,3 @@
-//go:build p2p
-// +build p2p
-
 package worker

 import (
@@ -13,8 +10,6 @@ import (

 	cliContext "github.com/mudler/LocalAI/core/cli/context"
 	"github.com/mudler/LocalAI/core/p2p"
-	"github.com/mudler/LocalAI/pkg/assets"
-	"github.com/mudler/LocalAI/pkg/library"
 	"github.com/phayes/freeport"
 	"github.com/rs/zerolog/log"
 )
@@ -29,12 +24,6 @@ type P2P struct {
 }

 func (r *P2P) Run(ctx *cliContext.Context) error {
-	// Extract files from the embedded FS
-	err := assets.ExtractFiles(ctx.BackendAssets, r.BackendAssetsPath)
-	log.Debug().Msgf("Extracting backend assets files to %s", r.BackendAssetsPath)
-	if err != nil {
-		log.Warn().Msgf("Failed extracting backend assets files: %s (might be required for some backends to work properly)", err)
-	}

 	// Check if the token is set
 	// as we always need it.
@@ -71,7 +60,7 @@ func (r *P2P) Run(ctx *cliContext.Context) error {
 			for {
 				log.Info().Msgf("Starting llama-cpp-rpc-server on '%s:%d'", address, port)

-				grpcProcess, err := findLLamaCPPBackend(r.BackendAssetsPath)
+				grpcProcess, err := findLLamaCPPBackend(r.BackendsPath)
 				if err != nil {
 					log.Error().Err(err).Msg("Failed to find llama-cpp-rpc-server")
 					return
@@ -85,8 +74,6 @@ func (r *P2P) Run(ctx *cliContext.Context) error {
 				args := append([]string{"--host", address, "--port", fmt.Sprint(port)}, extraArgs...)
 				log.Debug().Msgf("Starting llama-cpp-rpc-server on '%s:%d' with args: %+v (%d)", address, port, args, len(args))

-				args, grpcProcess = library.LoadLDSO(r.BackendAssetsPath, args, grpcProcess)
-
 				cmd := exec.Command(
 					grpcProcess, args...,
 				)
--- a/core/config/application_config.go
+++ b/core/config/application_config.go
@@ -6,7 +6,6 @@ import (
 	"regexp"
 	"time"

-	rice "github.com/GeertJohan/go.rice"
 	"github.com/mudler/LocalAI/pkg/xsysinfo"
 	"github.com/rs/zerolog/log"
 )
@@ -17,7 +16,6 @@ type ApplicationConfig struct {
 	ModelPath                           string
 	BackendsPath                        string
 	ExternalBackends                    []string
-	LibPath                             string
 	UploadLimitMB, Threads, ContextSize int
 	F16                                 bool
 	Debug                               bool
@@ -50,9 +48,6 @@ type ApplicationConfig struct {
 	Galleries        []Gallery
 	BackendGalleries []Gallery

-	BackendAssets     *rice.Box
-	AssetsDestination string
-
 	ExternalGRPCBackends map[string]string

 	AutoloadGalleries, AutoloadBackendGalleries bool
@@ -140,12 +135,6 @@ func WithP2PToken(s string) AppOption {
 	}
 }

-func WithLibPath(path string) AppOption {
-	return func(o *ApplicationConfig) {
-		o.LibPath = path
-	}
-}
-
 var EnableWatchDog = func(o *ApplicationConfig) {
 	o.WatchDog = true
 }
@@ -211,18 +200,6 @@ func WithCorsAllowOrigins(b string) AppOption {
 	}
 }

-func WithBackendAssetsOutput(out string) AppOption {
-	return func(o *ApplicationConfig) {
-		o.AssetsDestination = out
-	}
-}
-
-func WithBackendAssets(f *rice.Box) AppOption {
-	return func(o *ApplicationConfig) {
-		o.BackendAssets = f
-	}
-}
-
 func WithStringGalleries(galls string) AppOption {
 	return func(o *ApplicationConfig) {
 		if galls == "" {
--- a/core/gallery/models.go
+++ b/core/gallery/models.go
@@ -126,8 +126,9 @@ func InstallModelFromGallery(
 		if err != nil {
 			return err
 		}
-
+		log.Debug().Msgf("Installed model %q", installedModel.Name)
 		if automaticallyInstallBackend && installedModel.Backend != "" {
+			log.Debug().Msgf("Installing backend %q", installedModel.Backend)
 			systemState, err := system.GetSystemState()
 			if err != nil {
 				return err
--- a/core/http/app_test.go
+++ b/core/http/app_test.go
@@ -23,7 +23,6 @@ import (
 	. "github.com/onsi/gomega"
 	"gopkg.in/yaml.v3"

-	rice "github.com/GeertJohan/go.rice"
 	openaigo "github.com/otiai10/openaigo"
 	"github.com/sashabaranov/go-openai"
 	"github.com/sashabaranov/go-openai/jsonschema"
@@ -264,16 +263,6 @@ func getRequest(url string, header http.Header) (error, int, []byte) {

 const bertEmbeddingsURL = `https://gist.githubusercontent.com/mudler/0a080b166b87640e8644b09c2aee6e3b/raw/f0e8c26bb72edc16d9fbafbfd6638072126ff225/bert-embeddings-gallery.yaml`

-var backendAssets *rice.Box
-
-func init() {
-	var err error
-	backendAssets, err = rice.FindBox("backend-assets")
-	if err != nil {
-		panic(err)
-	}
-}
-
 var _ = Describe("API test", func() {

 	var app *fiber.App
@@ -300,9 +289,6 @@ var _ = Describe("API test", func() {
 			modelDir = filepath.Join(tmpdir, "models")
 			err = os.Mkdir(modelDir, 0750)
 			Expect(err).ToNot(HaveOccurred())
-			backendAssetsDir := filepath.Join(tmpdir, "backend-assets")
-			err = os.Mkdir(backendAssetsDir, 0750)
-			Expect(err).ToNot(HaveOccurred())

 			c, cancel = context.WithCancel(context.Background())

@@ -341,8 +327,7 @@ var _ = Describe("API test", func() {
 					config.WithModelPath(modelDir),
 					config.WithBackendsPath(backendPath),
 					config.WithApiKeys([]string{apiKey}),
-					config.WithBackendAssets(backendAssets),
-					config.WithBackendAssetsOutput(backendAssetsDir))...)
+				)...)
 			Expect(err).ToNot(HaveOccurred())

 			app, err = API(application)
@@ -545,8 +530,7 @@ var _ = Describe("API test", func() {
 					config.WithBackendsPath(backendPath),
 					config.WithGalleries(galleries),
 					config.WithModelPath(modelDir),
-					config.WithBackendAssets(backendAssets),
-					config.WithBackendAssetsOutput(tmpdir))...,
+				)...,
 			)
 			Expect(err).ToNot(HaveOccurred())
 			app, err = API(application)
@@ -803,6 +787,10 @@ var _ = Describe("API test", func() {
 		})

 		It("shows the external backend", func() {
+			// Only run on linux
+			if runtime.GOOS != "linux" {
+				Skip("test supported only on linux")
+			}
 			// do an http request to the /system endpoint
 			resp, err := http.Get("http://127.0.0.1:9090/system")
 			Expect(err).ToNot(HaveOccurred())
@@ -888,6 +876,13 @@ var _ = Describe("API test", func() {
 		// See tests/integration/stores_test
 		Context("Stores", Label("stores"), func() {

+			BeforeEach(func() {
+				// Only run on linux
+				if runtime.GOOS != "linux" {
+					Skip("test supported only on linux")
+				}
+			})
+
 			It("sets, gets, finds and deletes entries", func() {
 				ks := [][]float32{
 					{0.1, 0.2, 0.3},
--- a/core/http/endpoints/localai/stores.go
+++ b/core/http/endpoints/localai/stores.go
@@ -17,7 +17,7 @@ func StoresSetEndpoint(sl *model.ModelLoader, appConfig *config.ApplicationConfi
 			return err
 		}

-		sb, err := backend.StoreBackend(sl, appConfig, input.Store)
+		sb, err := backend.StoreBackend(sl, appConfig, input.Store, input.Backend)
 		if err != nil {
 			return err
 		}
@@ -45,7 +45,7 @@ func StoresDeleteEndpoint(sl *model.ModelLoader, appConfig *config.ApplicationCo
 			return err
 		}

-		sb, err := backend.StoreBackend(sl, appConfig, input.Store)
+		sb, err := backend.StoreBackend(sl, appConfig, input.Store, input.Backend)
 		if err != nil {
 			return err
 		}
@@ -67,7 +67,7 @@ func StoresGetEndpoint(sl *model.ModelLoader, appConfig *config.ApplicationConfi
 			return err
 		}

-		sb, err := backend.StoreBackend(sl, appConfig, input.Store)
+		sb, err := backend.StoreBackend(sl, appConfig, input.Store, input.Backend)
 		if err != nil {
 			return err
 		}
@@ -99,7 +99,7 @@ func StoresFindEndpoint(sl *model.ModelLoader, appConfig *config.ApplicationConf
 			return err
 		}

-		sb, err := backend.StoreBackend(sl, appConfig, input.Store)
+		sb, err := backend.StoreBackend(sl, appConfig, input.Store, input.Backend)
 		if err != nil {
 			return err
 		}
--- a/core/http/endpoints/localai/system.go
+++ b/core/http/endpoints/localai/system.go
@@ -13,10 +13,7 @@ import (
 // @Router /system [get]
 func SystemInformations(ml *model.ModelLoader, appConfig *config.ApplicationConfig) func(*fiber.Ctx) error {
 	return func(c *fiber.Ctx) error {
-		availableBackends, err := ml.ListAvailableBackends(appConfig.AssetsDestination)
-		if err != nil {
-			return err
-		}
+		availableBackends := []string{}
 		loadedModels := ml.ListModels()
 		for b := range appConfig.ExternalGRPCBackends {
 			availableBackends = append(availableBackends, b)
--- a/core/http/endpoints/localai/welcome.go
+++ b/core/http/endpoints/localai/welcome.go
@@ -5,7 +5,6 @@ import (
 	"github.com/mudler/LocalAI/core/config"
 	"github.com/mudler/LocalAI/core/gallery"
 	"github.com/mudler/LocalAI/core/http/utils"
-	"github.com/mudler/LocalAI/core/p2p"
 	"github.com/mudler/LocalAI/core/services"
 	"github.com/mudler/LocalAI/internal"
 	"github.com/mudler/LocalAI/pkg/model"
@@ -37,7 +36,6 @@ func WelcomeEndpoint(appConfig *config.ApplicationConfig,
 			"Models":            modelsWithoutConfig,
 			"ModelsConfig":      backendConfigs,
 			"GalleryConfig":     galleryConfigs,
-			"IsP2PEnabled":      p2p.IsP2PEnabled(),
 			"ApplicationConfig": appConfig,
 			"ProcessingModels":  processingModels,
 			"TaskTypes":         taskTypes,
--- a/core/http/routes/localai.go
+++ b/core/http/routes/localai.go
@@ -6,7 +6,6 @@ import (
 	"github.com/mudler/LocalAI/core/config"
 	"github.com/mudler/LocalAI/core/http/endpoints/localai"
 	"github.com/mudler/LocalAI/core/http/middleware"
-	"github.com/mudler/LocalAI/core/p2p"
 	"github.com/mudler/LocalAI/core/schema"
 	"github.com/mudler/LocalAI/core/services"
 	"github.com/mudler/LocalAI/internal"
@@ -80,10 +79,8 @@ func RegisterLocalAIRoutes(router *fiber.App,
 	router.Post("/v1/backend/shutdown", localai.BackendShutdownEndpoint(backendMonitorService))

 	// p2p
-	if p2p.IsP2PEnabled() {
-		router.Get("/api/p2p", localai.ShowP2PNodes(appConfig))
-		router.Get("/api/p2p/token", localai.ShowP2PToken(appConfig))
-	}
+	router.Get("/api/p2p", localai.ShowP2PNodes(appConfig))
+	router.Get("/api/p2p/token", localai.ShowP2PToken(appConfig))

 	router.Get("/version", func(c *fiber.Ctx) error {
 		return c.JSON(struct {
--- a/core/http/routes/ui.go
+++ b/core/http/routes/ui.go
@@ -25,38 +25,39 @@ func RegisterUIRoutes(app *fiber.App,

 	app.Get("/", localai.WelcomeEndpoint(appConfig, cl, ml, processingOps))

-	if p2p.IsP2PEnabled() {
-		app.Get("/p2p", func(c *fiber.Ctx) error {
-			summary := fiber.Map{
-				"Title":   "LocalAI - P2P dashboard",
-				"BaseURL": utils.BaseURL(c),
-				"Version": internal.PrintableVersion(),
-				//"Nodes":          p2p.GetAvailableNodes(""),
-				//"FederatedNodes": p2p.GetAvailableNodes(p2p.FederatedID),
-				"IsP2PEnabled": p2p.IsP2PEnabled(),
-				"P2PToken":     appConfig.P2PToken,
-				"NetworkID":    appConfig.P2PNetworkID,
-			}
+	// P2P
+	app.Get("/p2p", func(c *fiber.Ctx) error {
+		summary := fiber.Map{
+			"Title":   "LocalAI - P2P dashboard",
+			"BaseURL": utils.BaseURL(c),
+			"Version": internal.PrintableVersion(),
+			//"Nodes":          p2p.GetAvailableNodes(""),
+			//"FederatedNodes": p2p.GetAvailableNodes(p2p.FederatedID),

-			// Render index
-			return c.Render("views/p2p", summary)
-		})
+			"P2PToken":  appConfig.P2PToken,
+			"NetworkID": appConfig.P2PNetworkID,
+		}

-		/* show nodes live! */
-		app.Get("/p2p/ui/workers", func(c *fiber.Ctx) error {
-			return c.SendString(elements.P2PNodeBoxes(p2p.GetAvailableNodes(p2p.NetworkID(appConfig.P2PNetworkID, p2p.WorkerID))))
-		})
-		app.Get("/p2p/ui/workers-federation", func(c *fiber.Ctx) error {
-			return c.SendString(elements.P2PNodeBoxes(p2p.GetAvailableNodes(p2p.NetworkID(appConfig.P2PNetworkID, p2p.FederatedID))))
-		})
+		// Render index
+		return c.Render("views/p2p", summary)
+	})

-		app.Get("/p2p/ui/workers-stats", func(c *fiber.Ctx) error {
-			return c.SendString(elements.P2PNodeStats(p2p.GetAvailableNodes(p2p.NetworkID(appConfig.P2PNetworkID, p2p.WorkerID))))
-		})
-		app.Get("/p2p/ui/workers-federation-stats", func(c *fiber.Ctx) error {
-			return c.SendString(elements.P2PNodeStats(p2p.GetAvailableNodes(p2p.NetworkID(appConfig.P2PNetworkID, p2p.FederatedID))))
-		})
-	}
+	/* show nodes live! */
+	app.Get("/p2p/ui/workers", func(c *fiber.Ctx) error {
+		return c.SendString(elements.P2PNodeBoxes(p2p.GetAvailableNodes(p2p.NetworkID(appConfig.P2PNetworkID, p2p.WorkerID))))
+	})
+	app.Get("/p2p/ui/workers-federation", func(c *fiber.Ctx) error {
+		return c.SendString(elements.P2PNodeBoxes(p2p.GetAvailableNodes(p2p.NetworkID(appConfig.P2PNetworkID, p2p.FederatedID))))
+	})
+
+	app.Get("/p2p/ui/workers-stats", func(c *fiber.Ctx) error {
+		return c.SendString(elements.P2PNodeStats(p2p.GetAvailableNodes(p2p.NetworkID(appConfig.P2PNetworkID, p2p.WorkerID))))
+	})
+	app.Get("/p2p/ui/workers-federation-stats", func(c *fiber.Ctx) error {
+		return c.SendString(elements.P2PNodeStats(p2p.GetAvailableNodes(p2p.NetworkID(appConfig.P2PNetworkID, p2p.FederatedID))))
+	})
+
+	// End P2P

 	if !appConfig.DisableGalleryEndpoint {
 		registerGalleryRoutes(app, cl, appConfig, galleryService, processingOps)
@@ -76,8 +77,8 @@ func RegisterUIRoutes(app *fiber.App,
 			"BaseURL":      utils.BaseURL(c),
 			"ModelsConfig": backendConfigs,
 			"Model":        backendConfigs[0],
-			"IsP2PEnabled": p2p.IsP2PEnabled(),
-			"Version":      internal.PrintableVersion(),
+
+			"Version": internal.PrintableVersion(),
 		}

 		// Render index
@@ -121,7 +122,6 @@ func RegisterUIRoutes(app *fiber.App,
 			"ModelsConfig":        backendConfigs,
 			"Model":               modelThatCanBeUsed,
 			"Version":             internal.PrintableVersion(),
-			"IsP2PEnabled":        p2p.IsP2PEnabled(),
 		}

 		// Render index
@@ -151,7 +151,6 @@ func RegisterUIRoutes(app *fiber.App,
 			"ModelsWithoutConfig": modelsWithoutConfig,
 			"Model":               c.Params("model"),
 			"Version":             internal.PrintableVersion(),
-			"IsP2PEnabled":        p2p.IsP2PEnabled(),
 		}

 		// Render index
@@ -169,7 +168,6 @@ func RegisterUIRoutes(app *fiber.App,
 			"ModelsWithoutConfig": modelsWithoutConfig,
 			"Model":               c.Params("model"),
 			"Version":             internal.PrintableVersion(),
-			"IsP2PEnabled":        p2p.IsP2PEnabled(),
 		}

 		// Render index
@@ -203,7 +201,6 @@ func RegisterUIRoutes(app *fiber.App,
 			"ModelsWithoutConfig": modelsWithoutConfig,
 			"Model":               modelThatCanBeUsed,
 			"Version":             internal.PrintableVersion(),
-			"IsP2PEnabled":        p2p.IsP2PEnabled(),
 		}

 		// Render index
@@ -221,7 +218,6 @@ func RegisterUIRoutes(app *fiber.App,
 			"ModelsWithoutConfig": modelsWithoutConfig,
 			"Model":               c.Params("model"),
 			"Version":             internal.PrintableVersion(),
-			"IsP2PEnabled":        p2p.IsP2PEnabled(),
 		}

 		// Render index
@@ -253,7 +249,6 @@ func RegisterUIRoutes(app *fiber.App,
 			"ModelsConfig":        backendConfigs,
 			"ModelsWithoutConfig": modelsWithoutConfig,
 			"Model":               modelThatCanBeUsed,
-			"IsP2PEnabled":        p2p.IsP2PEnabled(),
 			"Version":             internal.PrintableVersion(),
 		}

--- a/core/http/routes/ui_backend_gallery.go
+++ b/core/http/routes/ui_backend_gallery.go
@@ -15,7 +15,6 @@ import (
 	"github.com/mudler/LocalAI/core/gallery"
 	"github.com/mudler/LocalAI/core/http/elements"
 	"github.com/mudler/LocalAI/core/http/utils"
-	"github.com/mudler/LocalAI/core/p2p"
 	"github.com/mudler/LocalAI/core/services"
 	"github.com/mudler/LocalAI/internal"
 	"github.com/rs/zerolog/log"
@@ -71,7 +70,6 @@ func registerBackendGalleryRoutes(app *fiber.App, appConfig *config.ApplicationC
 			"ProcessingBackends": processingBackendsData,
 			"AvailableBackends":  len(backends),
 			"TaskTypes":          taskTypes,
-			"IsP2PEnabled":       p2p.IsP2PEnabled(),
 		}

 		if page == "" {
--- a/core/http/routes/ui_gallery.go
+++ b/core/http/routes/ui_gallery.go
@@ -15,7 +15,6 @@ import (
 	"github.com/mudler/LocalAI/core/gallery"
 	"github.com/mudler/LocalAI/core/http/elements"
 	"github.com/mudler/LocalAI/core/http/utils"
-	"github.com/mudler/LocalAI/core/p2p"
 	"github.com/mudler/LocalAI/core/services"
 	"github.com/mudler/LocalAI/internal"
 	"github.com/rs/zerolog/log"
@@ -70,9 +69,7 @@ func registerGalleryRoutes(app *fiber.App, cl *config.BackendConfigLoader, appCo
 			"AllTags":          tags,
 			"ProcessingModels": processingModelsData,
 			"AvailableModels":  len(models),
-			"IsP2PEnabled":     p2p.IsP2PEnabled(),
-
-			"TaskTypes": taskTypes,
+			"TaskTypes":        taskTypes,
 			//	"ApplicationConfig": appConfig,
 		}

--- a/core/http/views/explorer.html
+++ b/core/http/views/explorer.html
@@ -268,7 +268,7 @@
                                    Command to connect (click to copy): 
                                </p>
                                <code class="block bg-gray-700 text-yellow-300 p-4 rounded-lg break-words"  @click="copyToken($el.textContent)" >
-                                    docker run -d --restart=always -e ADDRESS=":80" -e LOCALAI_P2P_NETWORK_ID=<span class="token" x-text="cluster.NetworkID"></span> -e LOCALAI_P2P_LOGLEVEL=debug --name local-ai -e TOKEN="<span class="token" x-text="network.token"></span>" --net host -ti localai/localai:master-ffmpeg-core federated --debug
+                                    docker run -d --restart=always -e ADDRESS=":80" -e LOCALAI_P2P_NETWORK_ID=<span class="token" x-text="cluster.NetworkID"></span> -e LOCALAI_P2P_LOGLEVEL=debug --name local-ai -e TOKEN="<span class="token" x-text="network.token"></span>" --net host -ti localai/localai:master federated --debug
                                </code>
                                or via CLI:
                                <code class="block bg-gray-700 text-yellow-300 p-4 rounded-lg break-words"  @click="copyToken($el.textContent)" >
--- a/core/http/views/p2p.html
+++ b/core/http/views/p2p.html
@@ -49,11 +49,11 @@
            </div>
        
            <!-- Warning box if p2p token is empty and p2p is enabled -->
-            {{ if and .IsP2PEnabled (eq .P2PToken "") }}
+            {{ if eq .P2PToken "" }}
            <div class="bg-gradient-to-r from-red-800/70 to-red-700/70 border border-red-600/50 p-6 rounded-xl shadow-lg mb-10 text-left">
                <div class="flex items-center mb-2">
                    <i class="fa-solid fa-exclamation-triangle text-red-300 text-2xl mr-3"></i>
-                    <h3 class="text-xl font-bold text-white">Warning: P2P mode is disabled or no token was specified</h3>
+                    <h3 class="text-xl font-bold text-white">Warning: P2P token was not specified</h3>
                </div>
                <p class="mb-4 text-red-200">
                    You have to enable P2P mode by starting LocalAI with <code class="bg-red-900/50 px-2 py-0.5 rounded">--p2p</code>. Please restart the server with <code class="bg-red-900/50 px-2 py-0.5 rounded">--p2p</code> to generate a new token automatically that can be used to discover other nodes. If you already have a token, specify it with <code class="bg-red-900/50 px-2 py-0.5 rounded">export TOKEN=".."</code>
--- a/core/http/views/partials/navbar.html
+++ b/core/http/views/partials/navbar.html
@@ -40,11 +40,9 @@
                <a href="talk/" class="text-gray-300 hover:text-white px-3 py-2 rounded-lg transition duration-300 ease-in-out hover:bg-blue-900/30 flex items-center">
                    <i class="fa-solid fa-phone text-blue-400 mr-2"></i>Talk
                </a>
-                {{ if .IsP2PEnabled }}
                <a href="p2p/" class="text-gray-300 hover:text-white px-3 py-2 rounded-lg transition duration-300 ease-in-out hover:bg-blue-900/30 flex items-center">
                    <i class="fa-solid fa-circle-nodes text-blue-400 mr-2"></i>Swarm
                </a>
-                {{ end }}
                <a href="swagger/" class="text-gray-300 hover:text-white px-3 py-2 rounded-lg transition duration-300 ease-in-out hover:bg-blue-900/30 flex items-center">
                    <i class="fas fa-code text-blue-400 mr-2"></i>API
                </a>
@@ -75,11 +73,9 @@
                <a href="talk/" class="block text-gray-300 hover:text-white hover:bg-blue-900/30 px-3 py-2 rounded-lg transition duration-300 ease-in-out flex items-center">
                    <i class="fa-solid fa-phone text-blue-400 mr-3 w-5 text-center"></i>Talk
                </a>
-                {{ if .IsP2PEnabled }}
                <a href="p2p/" class="block text-gray-300 hover:text-white hover:bg-blue-900/30 px-3 py-2 rounded-lg transition duration-300 ease-in-out flex items-center">
                    <i class="fa-solid fa-circle-nodes text-blue-400 mr-3 w-5 text-center"></i>Swarm
                </a>
-                {{ end }}
                <a href="swagger/" class="block text-gray-300 hover:text-white hover:bg-blue-900/30 px-3 py-2 rounded-lg transition duration-300 ease-in-out flex items-center">
                    <i class="fas fa-code text-blue-400 mr-3 w-5 text-center"></i>API
                </a>
--- a/core/p2p/federated_server.go
+++ b/core/p2p/federated_server.go
@@ -1,6 +1,3 @@
-//go:build p2p
-// +build p2p
-
 package p2p

 import (
--- a/core/p2p/p2p.go
+++ b/core/p2p/p2p.go
@@ -1,6 +1,3 @@
-//go:build p2p
-// +build p2p
-
 package p2p

 import (
@@ -65,10 +62,6 @@ func GenerateToken(DHTInterval, OTPInterval int) string {
 	return generateNewConnectionData(DHTInterval, OTPInterval).Base64()
 }

-func IsP2PEnabled() bool {
-	return true
-}
-
 func nodeID(s string) string {
 	hostname, _ := os.Hostname()
 	return fmt.Sprintf("%s-%s", hostname, s)
--- a/core/p2p/p2p_disabled.go
+++ b/core/p2p/p2p_disabled.go
@@ -1,35 +0,0 @@
-//go:build !p2p
-// +build !p2p
-
-package p2p
-
-import (
-	"context"
-	"fmt"
-
-	"github.com/mudler/edgevpn/pkg/node"
-)
-
-func GenerateToken(DHTInterval, OTPInterval int) string {
-	return "not implemented"
-}
-
-func (f *FederatedServer) Start(ctx context.Context) error {
-	return fmt.Errorf("not implemented")
-}
-
-func ServiceDiscoverer(ctx context.Context, node *node.Node, token, servicesID string, fn func(string, NodeData), allocate bool) error {
-	return fmt.Errorf("not implemented")
-}
-
-func ExposeService(ctx context.Context, host, port, token, servicesID string) (*node.Node, error) {
-	return nil, fmt.Errorf("not implemented")
-}
-
-func IsP2PEnabled() bool {
-	return false
-}
-
-func NewNode(token string) (*node.Node, error) {
-	return nil, fmt.Errorf("not implemented")
-}
--- a/core/schema/localai.go
+++ b/core/schema/localai.go
@@ -63,23 +63,29 @@ type VADResponse struct {
 	Segments []VADSegment `json:"segments" yaml:"segments"`
 }

+type StoreCommon struct {
+	Backend string `json:"backend,omitempty" yaml:"backend,omitempty"`
+}
 type StoresSet struct {
 	Store string `json:"store,omitempty" yaml:"store,omitempty"`

 	Keys   [][]float32 `json:"keys" yaml:"keys"`
 	Values []string    `json:"values" yaml:"values"`
+	StoreCommon
 }

 type StoresDelete struct {
 	Store string `json:"store,omitempty" yaml:"store,omitempty"`

 	Keys [][]float32 `json:"keys"`
+	StoreCommon
 }

 type StoresGet struct {
 	Store string `json:"store,omitempty" yaml:"store,omitempty"`

 	Keys [][]float32 `json:"keys" yaml:"keys"`
+	StoreCommon
 }

 type StoresGetResponse struct {
@@ -92,6 +98,7 @@ type StoresFind struct {

 	Key  []float32 `json:"key" yaml:"key"`
 	Topk int       `json:"topk" yaml:"topk"`
+	StoreCommon
 }

 type StoresFindResponse struct {