Return model list

Signed-off-by: mudler <mudler@c3os.io>

Earthfile

@@ -14,10 +14,11 @@ go-deps:
 build:
     FROM +go-deps
     WORKDIR /build
-    RUN git clone https://github.com/go-skynet/llama
-    RUN cd llama && make libllama.a
+    RUN git clone --recurse-submodules https://github.com/go-skynet/go-llama.cpp
+    RUN cd go-llama.cpp && make libbinding.a
     COPY . .
-    RUN C_INCLUDE_PATH=/build/llama LIBRARY_PATH=/build/llama go build -o llama-cli ./
+    RUN go mod edit -replace github.com/go-skynet/go-llama.cpp=/build/go-llama.cpp
+    RUN C_INCLUDE_PATH=$GOPATH/src/github.com/go-skynet/go-llama.cpp LIBRARY_PATH=$GOPATH/src/github.com/go-skynet/go-llama.cpp go build -o llama-cli ./
     SAVE ARTIFACT llama-cli AS LOCAL llama-cli
 
 image:

api.go

@@ -2,24 +2,282 @@ package main
 
 import (
 	"embed"
+	"fmt"
 	"net/http"
 	"strconv"
+	"strings"
 	"sync"
 
-	llama "github.com/go-skynet/llama/go"
+	llama "github.com/go-skynet/go-llama.cpp"
 	"github.com/gofiber/fiber/v2"
+	"github.com/gofiber/fiber/v2/middleware/cors"
 	"github.com/gofiber/fiber/v2/middleware/filesystem"
+	"github.com/gofiber/fiber/v2/middleware/recover"
 )
 
+type OpenAIResponse struct {
+	Created int      `json:"created,omitempty"`
+	Object  string   `json:"chat.completion,omitempty"`
+	ID      string   `json:"id,omitempty"`
+	Model   string   `json:"model,omitempty"`
+	Choices []Choice `json:"choices,omitempty"`
+}
+
+type Choice struct {
+	Index        int     `json:"index,omitempty"`
+	FinishReason string  `json:"finish_reason,omitempty"`
+	Message      Message `json:"message,omitempty"`
+	Text         string  `json:"text,omitempty"`
+}
+
+type Message struct {
+	Role    string `json:"role,omitempty"`
+	Content string `json:"content,omitempty"`
+}
+
+type OpenAIModel struct {
+	ID     string `json:"id"`
+	Object string `json:"object"`
+}
+
 //go:embed index.html
 var indexHTML embed.FS
 
-func api(l *llama.LLama, listenAddr string, threads int) error {
+func completionEndpoint(defaultModel *llama.LLama, loader *ModelLoader, threads int, defaultMutex *sync.Mutex, mutexMap *sync.Mutex, mutexes map[string]*sync.Mutex) func(c *fiber.Ctx) error {
+	return func(c *fiber.Ctx) error {
+
+		var err error
+		var model *llama.LLama
+
+		// Get input data from the request body
+		input := new(struct {
+			Model  string `json:"model"`
+			Prompt string `json:"prompt"`
+		})
+		if err := c.BodyParser(input); err != nil {
+			return err
+		}
+
+		if input.Model == "" {
+			if defaultModel == nil {
+				return fmt.Errorf("no default model loaded, and no model specified")
+			}
+			model = defaultModel
+		} else {
+			model, err = loader.LoadModel(input.Model)
+			if err != nil {
+				return err
+			}
+		}
+
+		// This is still needed, see: https://github.com/ggerganov/llama.cpp/discussions/784
+		if input.Model != "" {
+			mutexMap.Lock()
+			l, ok := mutexes[input.Model]
+			if !ok {
+				m := &sync.Mutex{}
+				mutexes[input.Model] = m
+				l = m
+			}
+			mutexMap.Unlock()
+			l.Lock()
+			defer l.Unlock()
+		} else {
+			defaultMutex.Lock()
+			defer defaultMutex.Unlock()
+		}
+
+		// Set the parameters for the language model prediction
+		topP, err := strconv.ParseFloat(c.Query("topP", "0.9"), 64) // Default value of topP is 0.9
+		if err != nil {
+			return err
+		}
+
+		topK, err := strconv.Atoi(c.Query("topK", "40")) // Default value of topK is 40
+		if err != nil {
+			return err
+		}
+
+		temperature, err := strconv.ParseFloat(c.Query("temperature", "0.5"), 64) // Default value of temperature is 0.5
+		if err != nil {
+			return err
+		}
+
+		tokens, err := strconv.Atoi(c.Query("tokens", "128")) // Default value of tokens is 128
+		if err != nil {
+			return err
+		}
+
+		predInput := input.Prompt
+		// A model can have a "file.bin.tmpl" file associated with a prompt template prefix
+		templatedInput, err := loader.TemplatePrefix(input.Model, struct {
+			Input string
+		}{Input: input.Prompt})
+		if err == nil {
+			predInput = templatedInput
+		}
+
+		// Generate the prediction using the language model
+		prediction, err := model.Predict(
+			predInput,
+			llama.SetTemperature(temperature),
+			llama.SetTopP(topP),
+			llama.SetTopK(topK),
+			llama.SetTokens(tokens),
+			llama.SetThreads(threads),
+		)
+		if err != nil {
+			return err
+		}
+
+		// Return the prediction in the response body
+		return c.JSON(OpenAIResponse{
+			Model:   input.Model,
+			Choices: []Choice{{Text: prediction}},
+		})
+	}
+}
+
+func chatEndpoint(defaultModel *llama.LLama, loader *ModelLoader, threads int, defaultMutex *sync.Mutex, mutexMap *sync.Mutex, mutexes map[string]*sync.Mutex) func(c *fiber.Ctx) error {
+	return func(c *fiber.Ctx) error {
+
+		var err error
+		var model *llama.LLama
+
+		// Get input data from the request body
+		input := new(struct {
+			Messages []Message `json:"messages"`
+			Model    string    `json:"model"`
+		})
+		if err := c.BodyParser(input); err != nil {
+			return err
+		}
+
+		if input.Model == "" {
+			if defaultModel == nil {
+				return fmt.Errorf("no default model loaded, and no model specified")
+			}
+			model = defaultModel
+		} else {
+			model, err = loader.LoadModel(input.Model)
+			if err != nil {
+				return err
+			}
+		}
+
+		// This is still needed, see: https://github.com/ggerganov/llama.cpp/discussions/784
+		if input.Model != "" {
+			mutexMap.Lock()
+			l, ok := mutexes[input.Model]
+			if !ok {
+				m := &sync.Mutex{}
+				mutexes[input.Model] = m
+				l = m
+			}
+			mutexMap.Unlock()
+			l.Lock()
+			defer l.Unlock()
+		} else {
+			defaultMutex.Lock()
+			defer defaultMutex.Unlock()
+		}
+
+		// Set the parameters for the language model prediction
+		topP, err := strconv.ParseFloat(c.Query("topP", "0.9"), 64) // Default value of topP is 0.9
+		if err != nil {
+			return err
+		}
+
+		topK, err := strconv.Atoi(c.Query("topK", "40")) // Default value of topK is 40
+		if err != nil {
+			return err
+		}
+
+		temperature, err := strconv.ParseFloat(c.Query("temperature", "0.5"), 64) // Default value of temperature is 0.5
+		if err != nil {
+			return err
+		}
+
+		tokens, err := strconv.Atoi(c.Query("tokens", "128")) // Default value of tokens is 128
+		if err != nil {
+			return err
+		}
+
+		mess := []string{}
+		for _, i := range input.Messages {
+			mess = append(mess, i.Content)
+		}
+
+		predInput := strings.Join(mess, "\n")
+
+		// A model can have a "file.bin.tmpl" file associated with a prompt template prefix
+		templatedInput, err := loader.TemplatePrefix(input.Model, struct {
+			Input string
+		}{Input: predInput})
+		if err == nil {
+			predInput = templatedInput
+		}
+
+		// Generate the prediction using the language model
+		prediction, err := model.Predict(
+			predInput,
+			llama.SetTemperature(temperature),
+			llama.SetTopP(topP),
+			llama.SetTopK(topK),
+			llama.SetTokens(tokens),
+			llama.SetThreads(threads),
+		)
+		if err != nil {
+			return err
+		}
+
+		// Return the prediction in the response body
+		return c.JSON(OpenAIResponse{
+			Model:   input.Model,
+			Choices: []Choice{{Message: Message{Role: "assistant", Content: prediction}}},
+		})
+	}
+}
+
+func api(defaultModel *llama.LLama, loader *ModelLoader, listenAddr string, threads int) error {
 	app := fiber.New()
+
+	// Default middleware config
+	app.Use(recover.New())
+	app.Use(cors.New())
+
+	// This is still needed, see: https://github.com/ggerganov/llama.cpp/discussions/784
+	var mutex = &sync.Mutex{}
+	mu := map[string]*sync.Mutex{}
+	var mumutex = &sync.Mutex{}
+
+	// openAI compatible API endpoint
+	app.Post("/v1/chat/completions", chatEndpoint(defaultModel, loader, threads, mutex, mumutex, mu))
+	app.Post("/v1/completions", completionEndpoint(defaultModel, loader, threads, mutex, mumutex, mu))
+	app.Get("/v1/models", func(c *fiber.Ctx) error {
+		models, err := loader.ListModels()
+		if err != nil {
+			return err
+		}
+
+		dataModels := []OpenAIModel{}
+		for _, m := range models {
+			dataModels = append(dataModels, OpenAIModel{ID: m, Object: "model"})
+		}
+		return c.JSON(struct {
+			Object string        `json:"object"`
+			Data   []OpenAIModel `json:"data"`
+		}{
+			Object: "list",
+			Data:   dataModels,
+		})
+	})
+
 	app.Use("/", filesystem.New(filesystem.Config{
 		Root:         http.FS(indexHTML),
 		NotFoundFile: "index.html",
 	}))
+
 	/*
 		curl --location --request POST 'http://localhost:8080/predict' --header 'Content-Type: application/json' --data-raw '{
 		    "text": "What is an alpaca?",
@@ -29,8 +287,6 @@ func api(l *llama.LLama, listenAddr string, threads int) error {
 		    "tokens": 100
 		}'
 	*/
-	var mutex = &sync.Mutex{}
-
 	// Endpoint to generate the prediction
 	app.Post("/predict", func(c *fiber.Ctx) error {
 		mutex.Lock()
@@ -65,7 +321,7 @@ func api(l *llama.LLama, listenAddr string, threads int) error {
 		}
 
 		// Generate the prediction using the language model
-		prediction, err := l.Predict(
+		prediction, err := defaultModel.Predict(
 			input.Text,
 			llama.SetTemperature(temperature),
 			llama.SetTopP(topP),
@@ -86,6 +342,6 @@ func api(l *llama.LLama, listenAddr string, threads int) error {
 	})
 
 	// Start the server
-	app.Listen(":8080")
+	app.Listen(listenAddr)
 	return nil
 }

go.mod

@@ -17,6 +17,7 @@ require (
 	github.com/aymanbagabas/go-osc52/v2 v2.0.1 // indirect
 	github.com/containerd/console v1.0.3 // indirect
 	github.com/cpuguy83/go-md2man/v2 v2.0.2 // indirect
+	github.com/go-skynet/go-llama.cpp v0.0.0-20230405204601-5429d2339021 // indirect
 	github.com/google/uuid v1.3.0 // indirect
 	github.com/klauspost/compress v1.15.9 // indirect
 	github.com/lucasb-eyer/go-colorful v1.2.0 // indirect

go.sum

@@ -19,6 +19,10 @@ github.com/containerd/console v1.0.3 h1:lIr7SlA5PxZyMV30bDW0MGbiOPXwc63yRuCP0ARu
 github.com/containerd/console v1.0.3/go.mod h1:7LqA/THxQ86k76b8c/EMSiaJ3h1eZkMkXar0TQ1gf3U=
 github.com/cpuguy83/go-md2man/v2 v2.0.2 h1:p1EgwI/C7NhT0JmVkwCD2ZBK8j4aeHQX2pMHHBfMQ6w=
 github.com/cpuguy83/go-md2man/v2 v2.0.2/go.mod h1:tgQtvFlXSQOSOSIRvRPT7W67SCa46tRHOmNcaadrF8o=
+github.com/go-skynet/go-llama.cpp v0.0.0-20230404185816-24b85a924f09 h1:WPUWvw7DOv3WUuhtNfv+xJVE2CCTGa1op1PKGcNk2Bk=
+github.com/go-skynet/go-llama.cpp v0.0.0-20230404185816-24b85a924f09/go.mod h1:yD5HHNAHPReBlvWGWUr9OcMeE5BJH3xOUDtKCwjxdEQ=
+github.com/go-skynet/go-llama.cpp v0.0.0-20230405204601-5429d2339021 h1:SsUkTjdCCAJjULfspizf99Sfw8Fx9OAHF30kp3i6cxc=
+github.com/go-skynet/go-llama.cpp v0.0.0-20230405204601-5429d2339021/go.mod h1:yD5HHNAHPReBlvWGWUr9OcMeE5BJH3xOUDtKCwjxdEQ=
 github.com/go-skynet/llama v0.0.0-20230321172246-7be5326e18cc h1:NcmO8mA7iRZIX0Qy2SjcsSaV14+g87MiTey1neUJaFQ=
 github.com/go-skynet/llama v0.0.0-20230321172246-7be5326e18cc/go.mod h1:ZtYsAIud4cvP9VTTI9uhdgR1uCwaO/gGKnZZ95h9i7w=
 github.com/go-skynet/llama v0.0.0-20230325223742-a3563a2690ba h1:u6OhAqlWFHsTjfWKePdK2kP4/mTyXX5vsmKwrK5QX6o=

interactive.go

@@ -1,142 +0,0 @@
-package main
-
-// A simple program demonstrating the text area component from the Bubbles
-// component library.
-
-import (
-	"fmt"
-	"strings"
-
-	"github.com/charmbracelet/bubbles/textarea"
-	"github.com/charmbracelet/bubbles/viewport"
-	tea "github.com/charmbracelet/bubbletea"
-	"github.com/charmbracelet/lipgloss"
-	llama "github.com/go-skynet/llama/go"
-)
-
-func startInteractive(l *llama.LLama, opts ...llama.PredictOption) error {
-	p := tea.NewProgram(initialModel(l, opts...))
-
-	_, err := p.Run()
-	return err
-}
-
-type (
-	errMsg error
-)
-
-type model struct {
-	viewport    viewport.Model
-	messages    *[]string
-	textarea    textarea.Model
-	senderStyle lipgloss.Style
-	err         error
-	l           *llama.LLama
-	opts        []llama.PredictOption
-
-	predictC chan string
-}
-
-func initialModel(l *llama.LLama, opts ...llama.PredictOption) model {
-	ta := textarea.New()
-	ta.Placeholder = "Send a message..."
-	ta.Focus()
-
-	ta.Prompt = "┃ "
-	ta.CharLimit = 280
-
-	ta.SetWidth(200)
-	ta.SetHeight(3)
-
-	// Remove cursor line styling
-	ta.FocusedStyle.CursorLine = lipgloss.NewStyle()
-
-	ta.ShowLineNumbers = false
-
-	vp := viewport.New(200, 5)
-	vp.SetContent(`Welcome to llama-cli. Type a message and press Enter to send. Alpaca doesn't keep context of the whole chat (yet).`)
-
-	ta.KeyMap.InsertNewline.SetEnabled(false)
-
-	predictChannel := make(chan string)
-	messages := []string{}
-	m := model{
-		textarea:    ta,
-		messages:    &messages,
-		viewport:    vp,
-		senderStyle: lipgloss.NewStyle().Foreground(lipgloss.Color("5")),
-		err:         nil,
-		l:           l,
-		opts:        opts,
-		predictC:    predictChannel,
-	}
-	go func() {
-		for p := range predictChannel {
-			str, _ := templateString(emptyInput, struct {
-				Instruction string
-				Input       string
-			}{Instruction: p})
-			res, _ := l.Predict(
-				str,
-				opts...,
-			)
-
-			mm := *m.messages
-			*m.messages = mm[:len(mm)-1]
-			*m.messages = append(*m.messages, m.senderStyle.Render("llama: ")+res)
-			m.viewport.SetContent(strings.Join(*m.messages, "\n"))
-			ta.Reset()
-			m.viewport.GotoBottom()
-		}
-	}()
-
-	return m
-}
-
-func (m model) Init() tea.Cmd {
-	return textarea.Blink
-}
-
-func (m model) Update(msg tea.Msg) (tea.Model, tea.Cmd) {
-	var (
-		tiCmd tea.Cmd
-		vpCmd tea.Cmd
-	)
-
-	m.textarea, tiCmd = m.textarea.Update(msg)
-	m.viewport, vpCmd = m.viewport.Update(msg)
-
-	switch msg := msg.(type) {
-	case tea.WindowSizeMsg:
-
-	//	m.viewport.Width = msg.Width
-	//	m.viewport.Height = msg.Height
-	case tea.KeyMsg:
-		switch msg.Type {
-		case tea.KeyCtrlC, tea.KeyEsc:
-			fmt.Println(m.textarea.Value())
-			return m, tea.Quit
-		case tea.KeyEnter:
-			*m.messages = append(*m.messages, m.senderStyle.Render("You: ")+m.textarea.Value(), m.senderStyle.Render("Loading response..."))
-			m.predictC <- m.textarea.Value()
-			m.viewport.SetContent(strings.Join(*m.messages, "\n"))
-			m.textarea.Reset()
-			m.viewport.GotoBottom()
-		}
-
-	// We handle errors just like any other message
-	case errMsg:
-		m.err = msg
-		return m, nil
-	}
-
-	return m, tea.Batch(tiCmd, vpCmd)
-}
-
-func (m model) View() string {
-	return fmt.Sprintf(
-		"%s\n\n%s",
-		m.viewport.View(),
-		m.textarea.View(),
-	) + "\n\n"
-}

main.go

@@ -8,7 +8,7 @@ import (
 	"runtime"
 	"text/template"
 
-	llama "github.com/go-skynet/llama/go"
+	llama "github.com/go-skynet/go-llama.cpp"
 	"github.com/urfave/cli/v2"
 )
 
@@ -33,12 +33,6 @@ var nonEmptyInput string = `Below is an instruction that describes a task, paire
 
 func llamaFromOptions(ctx *cli.Context) (*llama.LLama, error) {
 	opts := []llama.ModelOption{llama.SetContext(ctx.Int("context-size"))}
-	if ctx.Bool("alpaca") {
-		opts = append(opts, llama.EnableAlpaca)
-	}
-	if ctx.Bool("gpt4all") {
-		opts = append(opts, llama.EnableGPT4All)
-	}
 	return llama.New(ctx.String("model"), opts...)
 }
 
@@ -92,16 +86,6 @@ var modelFlags = []cli.Flag{
 		EnvVars: []string{"TOP_K"},
 		Value:   20,
 	},
-	&cli.BoolFlag{
-		Name:    "alpaca",
-		EnvVars: []string{"ALPACA"},
-		Value:   true,
-	},
-	&cli.BoolFlag{
-		Name:    "gpt4all",
-		EnvVars: []string{"GPT4ALL"},
-		Value:   false,
-	},
 }
 
 func main() {
@@ -134,24 +118,6 @@ echo "An Alpaca (Vicugna pacos) is a domesticated species of South American came
 `,
 		Copyright: "go-skynet authors",
 		Commands: []*cli.Command{
-			{
-				Flags: modelFlags,
-				Name:  "interactive",
-				Action: func(ctx *cli.Context) error {
-
-					l, err := llamaFromOptions(ctx)
-					if err != nil {
-						fmt.Println("Loading the model failed:", err.Error())
-						os.Exit(1)
-					}
-
-					return startInteractive(l, llama.SetTemperature(ctx.Float64("temperature")),
-						llama.SetTopP(ctx.Float64("topp")),
-						llama.SetTopK(ctx.Int("topk")),
-						llama.SetTokens(ctx.Int("tokens")),
-						llama.SetThreads(ctx.Int("threads")))
-				},
-			},
 			{
 
 				Name: "api",
@@ -162,24 +128,18 @@ echo "An Alpaca (Vicugna pacos) is a domesticated species of South American came
 						Value:   runtime.NumCPU(),
 					},
 					&cli.StringFlag{
-						Name:    "model",
-						EnvVars: []string{"MODEL_PATH"},
+						Name:    "models-path",
+						EnvVars: []string{"MODELS_PATH"},
+					},
+					&cli.StringFlag{
+						Name:    "default-model",
+						EnvVars: []string{"default-model"},
 					},
 					&cli.StringFlag{
 						Name:    "address",
 						EnvVars: []string{"ADDRESS"},
 						Value:   ":8080",
 					},
-					&cli.BoolFlag{
-						Name:    "alpaca",
-						EnvVars: []string{"ALPACA"},
-						Value:   true,
-					},
-					&cli.BoolFlag{
-						Name:    "gpt4all",
-						EnvVars: []string{"GPT4ALL"},
-						Value:   false,
-					},
 					&cli.IntFlag{
 						Name:    "context-size",
 						EnvVars: []string{"CONTEXT_SIZE"},
@@ -187,13 +147,19 @@ echo "An Alpaca (Vicugna pacos) is a domesticated species of South American came
 					},
 				},
 				Action: func(ctx *cli.Context) error {
-					l, err := llamaFromOptions(ctx)
-					if err != nil {
-						fmt.Println("Loading the model failed:", err.Error())
-						os.Exit(1)
+
+					var defaultModel *llama.LLama
+					defModel := ctx.String("default-model")
+					if defModel != "" {
+						opts := []llama.ModelOption{llama.SetContext(ctx.Int("context-size"))}
+						var err error
+						defaultModel, err = llama.New(ctx.String("default-model"), opts...)
+						if err != nil {
+							return err
+						}
 					}
 
-					return api(l, ctx.String("address"), ctx.Int("threads"))
+					return api(defaultModel, NewModelLoader(ctx.String("models-path")), ctx.String("address"), ctx.Int("threads"))
 				},
 			},
 		},

model_loader.go

@@ -0,0 +1,114 @@
+package main
+
+import (
+	"bytes"
+	"fmt"
+	"io/ioutil"
+	"os"
+	"path/filepath"
+	"strings"
+	"sync"
+	"text/template"
+
+	llama "github.com/go-skynet/go-llama.cpp"
+)
+
+type ModelLoader struct {
+	modelPath        string
+	mu               sync.Mutex
+	models           map[string]*llama.LLama
+	promptsTemplates map[string]*template.Template
+}
+
+func NewModelLoader(modelPath string) *ModelLoader {
+	return &ModelLoader{modelPath: modelPath, models: make(map[string]*llama.LLama), promptsTemplates: make(map[string]*template.Template)}
+}
+
+func (ml *ModelLoader) ListModels() ([]string, error) {
+	files, err := ioutil.ReadDir(ml.modelPath)
+	if err != nil {
+		return []string{}, err
+	}
+
+	models := []string{}
+	for _, file := range files {
+		if strings.HasSuffix(file.Name(), ".bin") {
+			models = append(models, strings.TrimRight(file.Name(), ".bin"))
+		}
+	}
+
+	return models, nil
+}
+
+func (ml *ModelLoader) TemplatePrefix(modelName string, in interface{}) (string, error) {
+	ml.mu.Lock()
+	defer ml.mu.Unlock()
+
+	m, ok := ml.promptsTemplates[modelName]
+	if !ok {
+		// try to find a s.bin
+		modelBin := fmt.Sprintf("%s.bin", modelName)
+		m, ok = ml.promptsTemplates[modelBin]
+		if !ok {
+			return "", fmt.Errorf("no prompt template available")
+		}
+	}
+
+	var buf bytes.Buffer
+
+	if err := m.Execute(&buf, in); err != nil {
+		return "", err
+	}
+	return buf.String(), nil
+}
+
+func (ml *ModelLoader) LoadModel(modelName string, opts ...llama.ModelOption) (*llama.LLama, error) {
+	ml.mu.Lock()
+	defer ml.mu.Unlock()
+
+	// Check if we already have a loaded model
+	modelFile := filepath.Join(ml.modelPath, modelName)
+
+	if m, ok := ml.models[modelFile]; ok {
+		return m, nil
+	}
+
+	// Check if the model path exists
+	if _, err := os.Stat(modelFile); os.IsNotExist(err) {
+		// try to find a s.bin
+		modelBin := fmt.Sprintf("%s.bin", modelFile)
+		if _, err := os.Stat(modelBin); os.IsNotExist(err) {
+			return nil, err
+		} else {
+			modelName = fmt.Sprintf("%s.bin", modelName)
+			modelFile = modelBin
+		}
+	}
+
+	// Load the model and keep it in memory for later use
+	model, err := llama.New(modelFile, opts...)
+	if err != nil {
+		return nil, err
+	}
+
+	// If there is a prompt template, load it
+
+	modelTemplateFile := fmt.Sprintf("%s.tmpl", modelFile)
+	// Check if the model path exists
+	if _, err := os.Stat(modelTemplateFile); err == nil {
+		dat, err := os.ReadFile(modelTemplateFile)
+		if err != nil {
+			return nil, err
+		}
+
+		// Parse the template
+		tmpl, err := template.New("prompt").Parse(string(dat))
+		if err != nil {
+			return nil, err
+		}
+		ml.promptsTemplates[modelName] = tmpl
+	}
+
+	ml.models[modelFile] = model
+	return model, err
+}