wip

README.md

@@ -29,8 +29,7 @@ curl https://ollama.ai/install.sh | sh
 
 ### Docker
 
-The official [Ollama Docker image `ollama/ollama`](https://hub.docker.com/r/ollama/ollama)
-is available on Docker Hub.
+See the official [Docker image](https://hub.docker.com/r/ollama/ollama).
 
 ## Quickstart
 
@@ -89,7 +88,7 @@ See the [guide](docs/import.md) on importing models for more information.
 
 ### Customize a prompt
 
-Models from the Ollama library can be customized with a prompt. For example, to customize the `llama2` model:
+Models from the Ollama library can be customized with a prompt. The example
 
 ```
 ollama pull llama2
@@ -179,7 +178,8 @@ ollama list
 Install `cmake` and `go`:
 
 ```
-brew install cmake go
+brew install cmake
+brew install go
 ```
 
 Then generate dependencies and build:
@@ -203,8 +203,9 @@ Finally, in a separate shell, run a model:
 
 ## REST API
 
-Ollama has a REST API for running and managing models.
-For example, to generate text from a model:
+See the [API documentation](docs/api.md) for all endpoints.
+
+Ollama has an API for running and managing models. For example to generate text from a model:
 
 ```
 curl -X POST http://localhost:11434/api/generate -d '{
@@ -213,8 +214,6 @@ curl -X POST http://localhost:11434/api/generate -d '{
 }'
 ```
 
-See the [API documentation](./docs/api.md) for all endpoints.
-
 ## Community Integrations
 
 - [LangChain](https://python.langchain.com/docs/integrations/llms/ollama) and [LangChain.js](https://js.langchain.com/docs/modules/model_io/models/llms/integrations/ollama) with [example](https://js.langchain.com/docs/use_cases/question_answering/local_retrieval_qa)
@@ -234,4 +233,3 @@ See the [API documentation](./docs/api.md) for all endpoints.
 - [oterm](https://github.com/ggozad/oterm)
 - [Ellama Emacs client](https://github.com/s-kostyaev/ellama)
 - [OllamaSharp for .NET](https://github.com/awaescher/OllamaSharp)
-- [Minimalistic React UI for Ollama Models](https://github.com/richawo/minimal-llm-ui)

api/client.go

@@ -18,6 +18,10 @@ import (
 	"github.com/jmorganca/ollama/version"
 )
 
+const DefaultHost = "127.0.0.1:11434"
+
+var envHost = os.Getenv("OLLAMA_HOST")
+
 type Client struct {
 	base *url.URL
 	http http.Client
@@ -40,24 +44,14 @@ func checkError(resp *http.Response, body []byte) error {
 }
 
 func ClientFromEnvironment() (*Client, error) {
-	defaultPort := "11434"
-
 	scheme, hostport, ok := strings.Cut(os.Getenv("OLLAMA_HOST"), "://")
-	switch {
-	case !ok:
+	if !ok {
 		scheme, hostport = "http", os.Getenv("OLLAMA_HOST")
-	case scheme == "http":
-		defaultPort = "80"
-	case scheme == "https":
-		defaultPort = "443"
 	}
 
-	// trim trailing slashes
-	hostport = strings.TrimRight(hostport, "/")
-
 	host, port, err := net.SplitHostPort(hostport)
 	if err != nil {
-		host, port = "127.0.0.1", defaultPort
+		host, port = "127.0.0.1", "11434"
 		if ip := net.ParseIP(strings.Trim(hostport, "[]")); ip != nil {
 			host = ip.String()
 		} else if hostport != "" {
@@ -72,7 +66,7 @@ func ClientFromEnvironment() (*Client, error) {
 		},
 	}
 
-	mockRequest, err := http.NewRequest(http.MethodHead, client.base.String(), nil)
+	mockRequest, err := http.NewRequest("HEAD", client.base.String(), nil)
 	if err != nil {
 		return nil, err
 	}

api/client_test.go

@@ -1,43 +0,0 @@
-package api
-
-import "testing"
-
-func TestClientFromEnvironment(t *testing.T) {
-	type testCase struct {
-		value  string
-		expect string
-		err    error
-	}
-
-	testCases := map[string]*testCase{
-		"empty":                      {value: "", expect: "http://127.0.0.1:11434"},
-		"only address":               {value: "1.2.3.4", expect: "http://1.2.3.4:11434"},
-		"only port":                  {value: ":1234", expect: "http://:1234"},
-		"address and port":           {value: "1.2.3.4:1234", expect: "http://1.2.3.4:1234"},
-		"scheme http and address":    {value: "http://1.2.3.4", expect: "http://1.2.3.4:80"},
-		"scheme https and address":   {value: "https://1.2.3.4", expect: "https://1.2.3.4:443"},
-		"scheme, address, and port":  {value: "https://1.2.3.4:1234", expect: "https://1.2.3.4:1234"},
-		"hostname":                   {value: "example.com", expect: "http://example.com:11434"},
-		"hostname and port":          {value: "example.com:1234", expect: "http://example.com:1234"},
-		"scheme http and hostname":   {value: "http://example.com", expect: "http://example.com:80"},
-		"scheme https and hostname":  {value: "https://example.com", expect: "https://example.com:443"},
-		"scheme, hostname, and port": {value: "https://example.com:1234", expect: "https://example.com:1234"},
-		"trailing slash":             {value: "example.com/", expect: "http://example.com:11434"},
-		"trailing slash port":        {value: "example.com:1234/", expect: "http://example.com:1234"},
-	}
-
-	for k, v := range testCases {
-		t.Run(k, func(t *testing.T) {
-			t.Setenv("OLLAMA_HOST", v.value)
-
-			client, err := ClientFromEnvironment()
-			if err != v.err {
-				t.Fatalf("expected %s, got %s", v.err, err)
-			}
-
-			if client.base.String() != v.expect {
-				t.Fatalf("expected %s, got %s", v.expect, client.base.String())
-			}
-		})
-	}
-}

api/types.go

@@ -293,7 +293,7 @@ func DefaultOptions() Options {
 	return Options{
 		// options set on request to runner
 		NumPredict:       -1,
-		NumKeep:          0,
+		NumKeep:          -1,
 		Temperature:      0.8,
 		TopK:             40,
 		TopP:             0.9,

cmd/cmd.go

@@ -11,7 +11,6 @@ import (
 	"io"
 	"log"
 	"net"
-	"net/http"
 	"os"
 	"os/exec"
 	"os/signal"
@@ -23,6 +22,7 @@ import (
 
 	"github.com/dustin/go-humanize"
 	"github.com/olekukonko/tablewriter"
+	"github.com/pdevine/readline"
 	"github.com/spf13/cobra"
 	"golang.org/x/crypto/ssh"
 	"golang.org/x/term"
@@ -30,11 +30,30 @@ import (
 	"github.com/jmorganca/ollama/api"
 	"github.com/jmorganca/ollama/format"
 	"github.com/jmorganca/ollama/progressbar"
-	"github.com/jmorganca/ollama/readline"
 	"github.com/jmorganca/ollama/server"
 	"github.com/jmorganca/ollama/version"
 )
 
+type Painter struct {
+	IsMultiLine bool
+}
+
+func (p Painter) Paint(line []rune, _ int) []rune {
+	termType := os.Getenv("TERM")
+	if termType == "xterm-256color" && len(line) == 0 {
+		var prompt string
+		if p.IsMultiLine {
+			prompt = "Use \"\"\" to end multi-line input"
+		} else {
+			prompt = "Send a message (/? for help)"
+		}
+		return []rune(fmt.Sprintf("\033[38;5;245m%s\033[%dD\033[0m", prompt, len(prompt)))
+	}
+	// add a space and a backspace to prevent the cursor from walking up the screen
+	line = append(line, []rune(" \b")...)
+	return line
+}
+
 func CreateHandler(cmd *cobra.Command, args []string) error {
 	filename, _ := cmd.Flags().GetString("file")
 	filename, err := filepath.Abs(filename)
@@ -99,16 +118,19 @@ func RunHandler(cmd *cobra.Command, args []string) error {
 		return err
 	}
 
-	name := args[0]
-	// check if the model exists on the server
-	_, err = client.Show(context.Background(), &api.ShowRequest{Name: name})
-	var statusError api.StatusError
-	switch {
-	case errors.As(err, &statusError) && statusError.StatusCode == http.StatusNotFound:
-		if err := PullHandler(cmd, args); err != nil {
-			return err
+	models, err := client.List(context.Background())
+	if err != nil {
+		return err
+	}
+
+	canonicalModelPath := server.ParseModelPath(args[0])
+	for _, model := range models.Models {
+		if model.Name == canonicalModelPath.GetShortTagname() {
+			return RunGenerate(cmd, args)
 		}
-	case err != nil:
+	}
+
+	if err := PullHandler(cmd, args); err != nil {
 		return err
 	}
 
@@ -486,11 +508,38 @@ func generate(cmd *cobra.Command, model, prompt string, wordWrap bool) error {
 }
 
 func generateInteractive(cmd *cobra.Command, model string) error {
+	home, err := os.UserHomeDir()
+	if err != nil {
+		return err
+	}
+
 	// load the model
 	if err := generate(cmd, model, "", false); err != nil {
 		return err
 	}
 
+	completer := readline.NewPrefixCompleter(
+		readline.PcItem("/help"),
+		readline.PcItem("/list"),
+		readline.PcItem("/set",
+			readline.PcItem("history"),
+			readline.PcItem("nohistory"),
+			readline.PcItem("wordwrap"),
+			readline.PcItem("nowordwrap"),
+			readline.PcItem("verbose"),
+			readline.PcItem("quiet"),
+		),
+		readline.PcItem("/show",
+			readline.PcItem("license"),
+			readline.PcItem("modelfile"),
+			readline.PcItem("parameters"),
+			readline.PcItem("system"),
+			readline.PcItem("template"),
+		),
+		readline.PcItem("/exit"),
+		readline.PcItem("/bye"),
+	)
+
 	usage := func() {
 		fmt.Fprintln(os.Stderr, "Available Commands:")
 		fmt.Fprintln(os.Stderr, "  /set         Set session variables")
@@ -523,17 +572,20 @@ func generateInteractive(cmd *cobra.Command, model string) error {
 		fmt.Fprintln(os.Stderr, "")
 	}
 
-	prompt := readline.Prompt{
-		Prompt:         ">>> ",
-		AltPrompt:      "... ",
-		Placeholder:    "Send a message (/? for help)",
-		AltPlaceholder: `Use """ to end multi-line input`,
+	var painter Painter
+
+	config := readline.Config{
+		Painter:      &painter,
+		Prompt:       ">>> ",
+		HistoryFile:  filepath.Join(home, ".ollama", "history"),
+		AutoComplete: completer,
 	}
 
-	scanner, err := readline.New(prompt)
+	scanner, err := readline.NewEx(&config)
 	if err != nil {
 		return err
 	}
+	defer scanner.Close()
 
 	var wordWrap bool
 	termType := os.Getenv("TERM")
@@ -550,20 +602,17 @@ func generateInteractive(cmd *cobra.Command, model string) error {
 		wordWrap = false
 	}
 
-	fmt.Print(readline.StartBracketedPaste)
-	defer fmt.Printf(readline.EndBracketedPaste)
-
 	var multiLineBuffer string
+	var isMultiLine bool
 
 	for {
 		line, err := scanner.Readline()
 		switch {
 		case errors.Is(err, io.EOF):
-			fmt.Println()
 			return nil
 		case errors.Is(err, readline.ErrInterrupt):
 			if line == "" {
-				fmt.Println("\nUse Ctrl-D or /bye to exit.")
+				fmt.Println("Use Ctrl-D or /bye to exit.")
 			}
 
 			continue
@@ -574,19 +623,23 @@ func generateInteractive(cmd *cobra.Command, model string) error {
 		line = strings.TrimSpace(line)
 
 		switch {
-		case scanner.Prompt.UseAlt:
+		case isMultiLine:
 			if strings.HasSuffix(line, `"""`) {
-				scanner.Prompt.UseAlt = false
+				isMultiLine = false
+				painter.IsMultiLine = isMultiLine
 				multiLineBuffer += strings.TrimSuffix(line, `"""`)
 				line = multiLineBuffer
 				multiLineBuffer = ""
+				scanner.SetPrompt(">>> ")
 			} else {
 				multiLineBuffer += line + " "
 				continue
 			}
 		case strings.HasPrefix(line, `"""`):
-			scanner.Prompt.UseAlt = true
+			isMultiLine = true
+			painter.IsMultiLine = isMultiLine
 			multiLineBuffer = strings.TrimPrefix(line, `"""`) + " "
+			scanner.SetPrompt("... ")
 			continue
 		case strings.HasPrefix(line, "/list"):
 			args := strings.Fields(line)
@@ -613,6 +666,19 @@ func generateInteractive(cmd *cobra.Command, model string) error {
 				case "quiet":
 					cmd.Flags().Set("verbose", "false")
 					fmt.Println("Set 'quiet' mode.")
+				case "mode":
+					if len(args) > 2 {
+						switch args[2] {
+						case "vim":
+							scanner.SetVimMode(true)
+						case "emacs", "default":
+							scanner.SetVimMode(false)
+						default:
+							usage()
+						}
+					} else {
+						usage()
+					}
 				default:
 					fmt.Printf("Unknown command '/set %s'. Type /? for help\n", args[1])
 				}
@@ -729,6 +795,21 @@ func RunServer(cmd *cobra.Command, _ []string) error {
 		origins = strings.Split(o, ",")
 	}
 
+	if noprune := os.Getenv("OLLAMA_NOPRUNE"); noprune == "" {
+		if err := server.PruneLayers(); err != nil {
+			return err
+		}
+
+		manifestsPath, err := server.GetManifestPath()
+		if err != nil {
+			return err
+		}
+
+		if err := server.PruneDirectory(manifestsPath); err != nil {
+			return err
+		}
+	}
+
 	return server.Serve(ln, origins)
 }
 

docs/api.md

@@ -45,11 +45,9 @@ Advanced parameters (optional):
 - `system`: system prompt to (overrides what is defined in the `Modelfile`)
 - `template`: the full prompt or prompt template (overrides what is defined in the `Modelfile`)
 - `context`: the context parameter returned from a previous request to `/generate`, this can be used to keep a short conversational memory
-- `stream`: if `false` the response will be returned as a single response object, rather than a stream of objects
+- `stream`: if `false` the response will be be returned as a single response object, rather than a stream of objects
 
-### Examples
-
-#### Request
+### Request
 
 ```shell
 curl -X POST http://localhost:11434/api/generate -d '{
@@ -58,9 +56,9 @@ curl -X POST http://localhost:11434/api/generate -d '{
 }'
 ```
 
-#### Response
+### Response
 
-A stream of JSON objects is returned:
+A stream of JSON objects:
 
 ```json
 {
@@ -104,38 +102,6 @@ To calculate how fast the response is generated in tokens per second (token/s),
 }
 ```
 
-#### Request
-
-```shell
-curl -X POST http://localhost:11434/api/generate -d '{
-  "model": "llama2:7b",
-  "prompt": "Why is the sky blue?",
-  "stream": false
-}'
-```
-
-#### Response
-
-If `stream` is set to `false`, the response will be a single JSON object:
-
-```json
-{
-  "model": "llama2:7b",
-  "created_at": "2023-08-04T19:22:45.499127Z",
-  "response": "The sky is blue because it is the color of the sky.",
-  "context": [1, 2, 3],
-  "done": true,
-  "total_duration": 5589157167,
-  "load_duration": 3013701500,
-  "sample_count": 114,
-  "sample_duration": 81442000,
-  "prompt_eval_count": 46,
-  "prompt_eval_duration": 1160282000,
-  "eval_count": 13,
-  "eval_duration": 1325948000
-}
-```
-
 ## Create a Model
 
 ```shell
@@ -148,11 +114,9 @@ Create a model from a [`Modelfile`](./modelfile.md)
 
 - `name`: name of the model to create
 - `path`: path to the Modelfile
-- `stream`: (optional) if `false` the response will be returned as a single response object, rather than a stream of objects
+- `stream`: (optional) if `false` the response will be be returned as a single response object, rather than a stream of objects
 
-### Examples
-
-#### Request
+### Request
 
 ```shell
 curl -X POST http://localhost:11434/api/create -d '{
@@ -161,7 +125,7 @@ curl -X POST http://localhost:11434/api/create -d '{
 }'
 ```
 
-#### Response
+### Response
 
 A stream of JSON objects. When finished, `status` is `success`.
 
@@ -179,17 +143,13 @@ GET /api/tags
 
 List models that are available locally.
 
-### Examples
-
-#### Request
+### Request
 
 ```shell
 curl http://localhost:11434/api/tags
 ```
 
-#### Response
-
-A single JSON object will be returned.
+### Response
 
 ```json
 {
@@ -220,9 +180,7 @@ Show details about a model including modelfile, template, parameters, license, a
 
 - `name`: name of the model to show
 
-### Examples
-
-#### Request
+### Request
 
 ```shell
 curl http://localhost:11434/api/show -d '{
@@ -230,7 +188,7 @@ curl http://localhost:11434/api/show -d '{
 }'
 ```
 
-#### Response
+### Response
 
 ```json
 {
@@ -249,9 +207,7 @@ POST /api/copy
 
 Copy a model. Creates a model with another name from an existing model.
 
-### Examples
-
-#### Request
+### Request
 
 ```shell
 curl http://localhost:11434/api/copy -d '{
@@ -260,10 +216,6 @@ curl http://localhost:11434/api/copy -d '{
 }'
 ```
 
-#### Response
-
-The only response is a 200 OK if successful.
-
 ## Delete a Model
 
 ```shell
@@ -274,11 +226,9 @@ Delete a model and its data.
 
 ### Parameters
 
-- `name`: model name to delete
+- `model`: model name to delete
 
-### Examples
-
-#### Request
+### Request
 
 ```shell
 curl -X DELETE http://localhost:11434/api/delete -d '{
@@ -286,10 +236,6 @@ curl -X DELETE http://localhost:11434/api/delete -d '{
 }'
 ```
 
-#### Response
-
-If successful, the only response is a 200 OK.
-
 ## Pull a Model
 
 ```shell
@@ -302,11 +248,9 @@ Download a model from the ollama library. Cancelled pulls are resumed from where
 
 - `name`: name of the model to pull
 - `insecure`: (optional) allow insecure connections to the library. Only use this if you are pulling from your own library during development.
-- `stream`: (optional) if `false` the response will be returned as a single response object, rather than a stream of objects
+- `stream`: (optional) if `false` the response will be be returned as a single response object, rather than a stream of objects
 
-### Examples
-
-#### Request
+### Request
 
 ```shell
 curl -X POST http://localhost:11434/api/pull -d '{
@@ -314,51 +258,13 @@ curl -X POST http://localhost:11434/api/pull -d '{
 }'
 ```
 
-#### Response
-
-If `stream` is not specified, or set to `true`, a stream of JSON objects is returned:
-
-The first object is the manifest:
-
-```json
-{
-  "status": "pulling manifest"
-}
-```
-
-Then there is a series of downloading responses. Until any of the download is completed, the `completed` key may not be included. The number of files to be downloaded depends on the number of layers specified in the manifest.
+### Response
 
 ```json
 {
   "status": "downloading digestname",
   "digest": "digestname",
-  "total": 2142590208,
-  "completed": 241970
-}
-```
-
-After all the files are downloaded, the final responses are:
-
-```json
-{
-    "status": "verifying sha256 digest"
-}
-{
-    "status": "writing manifest"
-}
-{
-    "status": "removing any unused layers"
-}
-{
-    "status": "success"
-}
-```
-
-if `stream` is set to false, then the response is a single JSON object:
-
-```json
-{
-  "status": "success"
+  "total": 2142590208
 }
 ```
 
@@ -374,11 +280,9 @@ Upload a model to a model library. Requires registering for ollama.ai and adding
 
 - `name`: name of the model to push in the form of `<namespace>/<model>:<tag>`
 - `insecure`: (optional) allow insecure connections to the library. Only use this if you are pushing to your library during development.
-- `stream`: (optional) if `false` the response will be returned as a single response object, rather than a stream of objects
+- `stream`: (optional) if `false` the response will be be returned as a single response object, rather than a stream of objects
 
-### Examples
-
-#### Request
+### Request
 
 ```shell
 curl -X POST http://localhost:11434/api/push -d '{
@@ -386,9 +290,9 @@ curl -X POST http://localhost:11434/api/push -d '{
 }'
 ```
 
-#### Response
+### Response
 
-If `stream` is not specified, or set to `true`, a stream of JSON objects is returned:
+Streaming response that starts with:
 
 ```json
 { "status": "retrieving manifest" }
@@ -421,12 +325,6 @@ Finally, when the upload is complete:
 {"status":"success"}
 ```
 
-If `stream` is set to `false`, then the response is a single JSON object:
-
-```json
-{ "status": "success" }
-```
-
 ## Generate Embeddings
 
 ```shell
@@ -444,9 +342,7 @@ Advanced parameters:
 
 - `options`: additional model parameters listed in the documentation for the [Modelfile](./modelfile.md#valid-parameters-and-values) such as `temperature`
 
-### Examples
-
-#### Request
+### Request
 
 ```shell
 curl -X POST http://localhost:11434/api/embeddings -d '{
@@ -455,7 +351,7 @@ curl -X POST http://localhost:11434/api/embeddings -d '{
 }'
 ```
 
-#### Response
+### Response
 
 ```json
 {

docs/faq.md

@@ -16,64 +16,19 @@ journalctl -u ollama
 
 If you're running `ollama serve` directly, the logs will be printed to the console.
 
-## How can I expose Ollama on my network?
-
-Ollama binds to 127.0.0.1 port 11434 by default. Change the bind address with the `OLLAMA_HOST` environment variable.
-
-On macOS:
+## How can I expose the Ollama server?
 
 ```bash
 OLLAMA_HOST=0.0.0.0:11435 ollama serve
 ```
 
-On Linux:
-
-Create a `systemd` drop-in directory and set `Environment=OLLAMA_HOST`
-
-```bash
-mkdir -p /etc/systemd/system/ollama.service.d
-echo "[Service]" >>/etc/systemd/system/ollama.service.d/environment.conf
-```
-
-```bash
-echo "Environment=OLLAMA_HOST=0.0.0.0:11434" >>/etc/systemd/system/ollama.service.d/environment.conf
-```
-
-Reload `systemd` and restart Ollama:
-
-```bash
-systemctl daemon-reload
-systemctl restart ollama
-```
-
-## How can I allow additional web origins to access Ollama?
-
-Ollama allows cross origin requests from `127.0.0.1` and `0.0.0.0` by default. Add additional origins with the `OLLAMA_ORIGINS` environment variable:
-
-On macOS:
+By default, Ollama allows cross origin requests from `127.0.0.1` and `0.0.0.0`. To support more origins, you can use the `OLLAMA_ORIGINS` environment variable:
 
 ```bash
 OLLAMA_ORIGINS=http://192.168.1.1:*,https://example.com ollama serve
 ```
 
-On Linux:
-
-```bash
-echo "Environment=OLLAMA_ORIGINS=http://129.168.1.1:*,https://example.com" >>/etc/systemd/system/ollama.service.d/environment.conf
-```
-
-Reload `systemd` and restart Ollama:
-
-```bash
-systemctl daemon-reload
-systemctl restart ollama
-```
-
 ## Where are models stored?
 
 - macOS: Raw model data is stored under `~/.ollama/models`.
 - Linux: Raw model data is stored under `/usr/share/ollama/.ollama/models`
-
-### How can I change where Ollama stores models?
-
-To modify where models are stored, you can use the `OLLAMA_MODELS` environment variable. Note that on Linux this means defining `OLLAMA_MODELS` in a drop-in `/etc/systemd/system/ollama.service.d` service file, reloading systemd, and restarting the ollama service.

docs/import.md

@@ -1,43 +1,8 @@
 # Import a model
 
-This guide walks through importing a GGUF, PyTorch or Safetensors model.
+This guide walks through importing a PyTorch, Safetensors or GGUF model.
 
-## Importing (GGUF)
-
-### Step 1: Write a `Modelfile`
-
-Start by creating a `Modelfile`. This file is the blueprint for your model, specifying weights, parameters, prompt templates and more.
-
-```
-FROM ./mistral-7b-v0.1.Q4_0.gguf
-```
-
-(Optional) many chat models require a prompt template in order to answer correctly. A default prompt template can be specified with the `TEMPLATE` instruction in the `Modelfile`:
-
-```
-FROM ./q4_0.bin
-TEMPLATE "[INST] {{ .Prompt }} [/INST]"
-```
-
-### Step 2: Create the Ollama model
-
-Finally, create a model from your `Modelfile`:
-
-```
-ollama create example -f Modelfile
-```
-
-### Step 3: Run your model
-
-Next, test the model with `ollama run`:
-
-```
-ollama run example "What is your favourite condiment?"
-```
-
-## Importing (PyTorch & Safetensors)
-
-### Supported models
+## Supported models
 
 Ollama supports a set of model architectures, with support for more coming soon:
 
@@ -48,6 +13,8 @@ Ollama supports a set of model architectures, with support for more coming soon:
 
 To view a model's architecture, check the `config.json` file in its HuggingFace repo. You should see an entry under `architectures` (e.g. `LlamaForCausalLM`).
 
+## Importing
+
 ### Step 1: Clone the HuggingFace repository (optional)
 
 If the model is currently hosted in a HuggingFace repository, first clone that repository to download the raw model.
@@ -77,7 +44,7 @@ This will output two files into the directory:
 
 ### Step 3: Write a `Modelfile`
 
-Next, create a `Modelfile` for your model:
+Next, create a `Modelfile` for your model. This file is the blueprint for your model, specifying weights, parameters, prompt templates and more.
 
 ```
 FROM ./q4_0.bin
@@ -98,15 +65,13 @@ Finally, create a model from your `Modelfile`:
 ollama create example -f Modelfile
 ```
 
-### Step 5: Run your model
-
 Next, test the model with `ollama run`:
 
 ```
 ollama run example "What is your favourite condiment?"
 ```
 
-## Publishing your model (optional – early alpha)
+### Step 5: Publish your model (optional – early alpha)
 
 Publishing models is in early alpha. If you'd like to publish your model to share with others, follow these steps:
 
@@ -185,7 +150,7 @@ python convert.py <path to model directory>
 python convert-falcon-hf-to-gguf.py <path to model directory>
 
 # GPTNeoXForCausalLM
-python convert-gptneox-hf-to-gguf.py <path to model directory>
+python convert-falcon-hf-to-gguf.py <path to model directory>
 
 # GPTBigCodeForCausalLM
 python convert-starcoder-hf-to-gguf.py <path to model directory>

docs/linux.md

@@ -1,16 +1,12 @@
-# Ollama on Linux
+# Installing Ollama on Linux
 
-## Install
-
-Install Ollama running this one-liner:
+> Note: A one line installer for Ollama is available by running:
 >
-```bash
-curl https://ollama.ai/install.sh | sh
-```
+> ```bash
+> curl https://ollama.ai/install.sh | sh
+> ```
 
-## Manual install
-
-### Download the `ollama` binary
+## Download the `ollama` binary
 
 Ollama is distributed as a self-contained binary. Download it to a directory in your PATH:
 
@@ -19,7 +15,31 @@ sudo curl -L https://ollama.ai/download/ollama-linux-amd64 -o /usr/bin/ollama
 sudo chmod +x /usr/bin/ollama
 ```
 
-### Adding Ollama as a startup service (recommended)
+## Start Ollama
+
+Start Ollama by running `ollama serve`:
+
+```bash
+ollama serve
+```
+
+Once Ollama is running, run a model in another terminal session:
+
+```bash
+ollama run llama2
+```
+
+## Install CUDA drivers (optional – for Nvidia GPUs)
+
+[Download and install](https://developer.nvidia.com/cuda-downloads) CUDA.
+
+Verify that the drivers are installed by running the following command, which should print details about your GPU:
+
+```bash
+nvidia-smi
+```
+
+## Adding Ollama as a startup service (optional)
 
 Create a user for Ollama:
 
@@ -40,6 +60,7 @@ User=ollama
 Group=ollama
 Restart=always
 RestartSec=3
+Environment="HOME=/usr/share/ollama"
 
 [Install]
 WantedBy=default.target
@@ -52,65 +73,10 @@ sudo systemctl daemon-reload
 sudo systemctl enable ollama
 ```
 
-### Install CUDA drivers (optional – for Nvidia GPUs)
-
-[Download and install](https://developer.nvidia.com/cuda-downloads) CUDA.
-
-Verify that the drivers are installed by running the following command, which should print details about your GPU:
-
-```bash
-nvidia-smi
-```
-
-### Start Ollama
-
-Start Ollama using `systemd`:
-
-```bash
-sudo systemctl start ollama
-```
-
-## Update
-
-Update ollama by running the install script again:
-
-```bash
-curl https://ollama.ai/install.sh | sh
-```
-
-Or by downloading the ollama binary:
-
-```bash
-sudo curl -L https://ollama.ai/download/ollama-linux-amd64 -o /usr/bin/ollama
-sudo chmod +x /usr/bin/ollama
-```
-
-## Viewing logs
+### Viewing logs
 
 To view logs of Ollama running as a startup service, run:
 
 ```bash
 journalctl -u ollama
 ```
-
-## Uninstall
-
-Remove the ollama service:
-
-```bash
-sudo systemctl stop ollama
-sudo systemctl disable ollama
-sudo rm /etc/systemd/system/ollama.service
-```
-
-Remove the ollama binary from your bin directory (either `/usr/local/bin`, `/usr/bin`, or `/bin`):
-
-```bash
-sudo rm $(which ollama)
-```
-
-Remove the downloaded models and Ollama service user:
-```bash
-sudo rm -r /usr/share/ollama
-sudo userdel ollama
-```

examples/langchain-python-rag-privategpt/constants.py

@@ -6,6 +6,7 @@ PERSIST_DIRECTORY = os.environ.get('PERSIST_DIRECTORY', 'db')
 
 # Define the Chroma settings
 CHROMA_SETTINGS = Settings(
+        chroma_db_impl='duckdb+parquet',
         persist_directory=PERSIST_DIRECTORY,
         anonymized_telemetry=False
 )

examples/langchain-python-rag-privategpt/ingest.py

@@ -150,7 +150,7 @@ def main():
         print("Creating new vectorstore")
         texts = process_documents()
         print(f"Creating embeddings. May take some minutes...")
-        db = Chroma.from_documents(texts, embeddings, persist_directory=persist_directory)
+        db = Chroma.from_documents(texts, embeddings, persist_directory=persist_directory, client_settings=CHROMA_SETTINGS)
     db.persist()
     db = None
 

examples/langchain-python-rag-privategpt/privateGPT.py

@@ -4,7 +4,6 @@ from langchain.embeddings import HuggingFaceEmbeddings
 from langchain.callbacks.streaming_stdout import StreamingStdOutCallbackHandler
 from langchain.vectorstores import Chroma
 from langchain.llms import Ollama
-import chromadb
 import os
 import argparse
 import time
@@ -23,9 +22,7 @@ def main():
     # Parse the command line arguments
     args = parse_arguments()
     embeddings = HuggingFaceEmbeddings(model_name=embeddings_model_name)
-
-    db = Chroma(persist_directory=persist_directory, embedding_function=embeddings)
-
+    db = Chroma(persist_directory=persist_directory, embedding_function=embeddings, client_settings=CHROMA_SETTINGS)
     retriever = db.as_retriever(search_kwargs={"k": target_source_chunks})
     # activate/deactivate the streaming StdOut callback for LLMs
     callbacks = [] if args.mute_stream else [StreamingStdOutCallbackHandler()]

go.mod

@@ -4,11 +4,11 @@ go 1.20
 
 require (
 	github.com/dustin/go-humanize v1.0.1
-	github.com/emirpasic/gods v1.18.1
 	github.com/gin-gonic/gin v1.9.1
 	github.com/mattn/go-runewidth v0.0.14
 	github.com/mitchellh/colorstring v0.0.0-20190213212951-d06e56a500db
 	github.com/olekukonko/tablewriter v0.0.5
+	github.com/pdevine/readline v1.5.2
 	github.com/spf13/cobra v1.7.0
 	golang.org/x/sync v0.3.0
 )
@@ -39,12 +39,12 @@ require (
 	github.com/twitchyliquid64/golang-asm v0.15.1 // indirect
 	github.com/ugorji/go/codec v1.2.11 // indirect
 	golang.org/x/arch v0.3.0 // indirect
-	golang.org/x/crypto v0.14.0
+	golang.org/x/crypto v0.10.0
 	golang.org/x/exp v0.0.0-20230817173708-d852ddb80c63
-	golang.org/x/net v0.17.0 // indirect
-	golang.org/x/sys v0.13.0 // indirect
-	golang.org/x/term v0.13.0
-	golang.org/x/text v0.13.0 // indirect
+	golang.org/x/net v0.10.0 // indirect
+	golang.org/x/sys v0.11.0 // indirect
+	golang.org/x/term v0.10.0
+	golang.org/x/text v0.10.0 // indirect
 	google.golang.org/protobuf v1.30.0 // indirect
 	gopkg.in/yaml.v3 v3.0.1 // indirect
 )

go.sum

@@ -4,6 +4,10 @@ github.com/bytedance/sonic v1.9.1/go.mod h1:i736AoUSYt75HyZLoJW9ERYxcy6eaN6h4BZX
 github.com/chenzhuoyu/base64x v0.0.0-20211019084208-fb5309c8db06/go.mod h1:DH46F32mSOjUmXrMHnKwZdA8wcEefY7UVqBKYGjpdQY=
 github.com/chenzhuoyu/base64x v0.0.0-20221115062448-fe3a3abad311 h1:qSGYFH7+jGhDF8vLC+iwCD4WpbV1EBDSzWkJODFLams=
 github.com/chenzhuoyu/base64x v0.0.0-20221115062448-fe3a3abad311/go.mod h1:b583jCggY9gE99b6G5LEC39OIiVsWj+R97kbl5odCEk=
+github.com/chzyer/logex v1.2.1 h1:XHDu3E6q+gdHgsdTPH6ImJMIp436vR6MPtH8gP05QzM=
+github.com/chzyer/logex v1.2.1/go.mod h1:JLbx6lG2kDbNRFnfkgvh4eRJRPX1QCoOIWomwysCBrQ=
+github.com/chzyer/test v1.0.0 h1:p3BQDXSxOhOG0P9z6/hGnII4LGiEPOYBhs8asl/fC04=
+github.com/chzyer/test v1.0.0/go.mod h1:2JlltgoNkt4TW/z9V/IzDdFaMTM2JPIi26O1pF38GC8=
 github.com/cpuguy83/go-md2man/v2 v2.0.2/go.mod h1:tgQtvFlXSQOSOSIRvRPT7W67SCa46tRHOmNcaadrF8o=
 github.com/creack/pty v1.1.9/go.mod h1:oKZEueFk5CKHvIhNR5MUki03XCEU+Q6VDXinZuGJ33E=
 github.com/davecgh/go-spew v1.1.0/go.mod h1:J7Y8YcW2NihsgmVo/mv3lAwl/skON4iLHjSsI+c5H38=
@@ -11,8 +15,6 @@ github.com/davecgh/go-spew v1.1.1 h1:vj9j/u1bqnvCEfJOwUhtlOARqs3+rkHYY13jYWTU97c
 github.com/davecgh/go-spew v1.1.1/go.mod h1:J7Y8YcW2NihsgmVo/mv3lAwl/skON4iLHjSsI+c5H38=
 github.com/dustin/go-humanize v1.0.1 h1:GzkhY7T5VNhEkwH0PVJgjz+fX1rhBrR7pRT3mDkpeCY=
 github.com/dustin/go-humanize v1.0.1/go.mod h1:Mu1zIs6XwVuF/gI1OepvI0qD18qycQx+mFykh5fBlto=
-github.com/emirpasic/gods v1.18.1 h1:FXtiHYKDGKCW2KzwZKx0iC0PQmdlorYgdFG9jPXJ1Bc=
-github.com/emirpasic/gods v1.18.1/go.mod h1:8tpGGwCnJ5H4r6BWwaV6OrWmMoPhUl5jm/FMNAnJvWQ=
 github.com/gabriel-vasile/mimetype v1.4.2 h1:w5qFW6JKBz9Y393Y4q372O9A7cUSequkh1Q7OhCmWKU=
 github.com/gabriel-vasile/mimetype v1.4.2/go.mod h1:zApsH/mKG4w07erKIaJPFiX0Tsq9BFQgN3qGY5GnNgA=
 github.com/gin-contrib/cors v1.4.0 h1:oJ6gwtUl3lqV0WEIwM/LxPF1QZ5qe2lGWdY2+bz7y0g=
@@ -76,6 +78,8 @@ github.com/olekukonko/tablewriter v0.0.5 h1:P2Ga83D34wi1o9J6Wh1mRuqd4mF/x/lgBS7N
 github.com/olekukonko/tablewriter v0.0.5/go.mod h1:hPp6KlRPjbx+hW8ykQs1w3UBbZlj6HuIJcUGPhkA7kY=
 github.com/pbnjay/memory v0.0.0-20210728143218-7b4eea64cf58 h1:onHthvaw9LFnH4t2DcNVpwGmV9E1BkGknEliJkfwQj0=
 github.com/pbnjay/memory v0.0.0-20210728143218-7b4eea64cf58/go.mod h1:DXv8WO4yhMYhSNPKjeNKa5WY9YCIEBRbNzFFPJbWO6Y=
+github.com/pdevine/readline v1.5.2 h1:oz6Y5GdTmhPG+08hhxcAvtHitSANWuA2100Sppb38xI=
+github.com/pdevine/readline v1.5.2/go.mod h1:na/LbuE5PYwxI7GyopWdIs3U8HVe89lYlNTFTXH3wOw=
 github.com/pelletier/go-toml/v2 v2.0.1/go.mod h1:r9LEWfGN8R5k0VXJ+0BkIe7MYkRdwZOjgMj2KwnJFUo=
 github.com/pelletier/go-toml/v2 v2.0.8 h1:0ctb6s9mE31h0/lhu+J6OPmVeDxJn+kYnJc2jZR9tGQ=
 github.com/pelletier/go-toml/v2 v2.0.8/go.mod h1:vuYfssBdrU2XDZ9bYydBu6t+6a6PYNcZljzZR9VXg+4=
@@ -114,30 +118,31 @@ golang.org/x/arch v0.0.0-20210923205945-b76863e36670/go.mod h1:5om86z9Hs0C8fWVUu
 golang.org/x/arch v0.3.0 h1:02VY4/ZcO/gBOH6PUaoiptASxtXU10jazRCP865E97k=
 golang.org/x/arch v0.3.0/go.mod h1:5om86z9Hs0C8fWVUuoMHwpExlXzs5Tkyp9hOrfG7pp8=
 golang.org/x/crypto v0.0.0-20210711020723-a769d52b0f97/go.mod h1:GvvjBRRGRdwPK5ydBHafDWAxML/pGHZbMvKqRZ5+Abc=
-golang.org/x/crypto v0.14.0 h1:wBqGXzWJW6m1XrIKlAH0Hs1JJ7+9KBwnIO8v66Q9cHc=
-golang.org/x/crypto v0.14.0/go.mod h1:MVFd36DqK4CsrnJYDkBA3VC4m2GkXAM0PvzMCn4JQf4=
+golang.org/x/crypto v0.10.0 h1:LKqV2xt9+kDzSTfOhx4FrkEBcMrAgHSYgzywV9zcGmM=
+golang.org/x/crypto v0.10.0/go.mod h1:o4eNf7Ede1fv+hwOwZsTHl9EsPFO6q6ZvYR8vYfY45I=
 golang.org/x/exp v0.0.0-20230817173708-d852ddb80c63 h1:m64FZMko/V45gv0bNmrNYoDEq8U5YUhetc9cBWKS1TQ=
 golang.org/x/exp v0.0.0-20230817173708-d852ddb80c63/go.mod h1:0v4NqG35kSWCMzLaMeX+IQrlSnVE/bqGSyC2cz/9Le8=
 golang.org/x/net v0.0.0-20210226172049-e18ecbb05110/go.mod h1:m0MpNAwzfU5UDzcl9v0D8zg8gWTRqZa9RBIspLL5mdg=
-golang.org/x/net v0.17.0 h1:pVaXccu2ozPjCXewfr1S7xza/zcXTity9cCdXQYSjIM=
-golang.org/x/net v0.17.0/go.mod h1:NxSsAGuq816PNPmqtQdLE42eU2Fs7NoRIZrHJAlaCOE=
+golang.org/x/net v0.10.0 h1:X2//UzNDwYmtCLn7To6G58Wr6f5ahEAQgKNzv9Y951M=
+golang.org/x/net v0.10.0/go.mod h1:0qNGK6F8kojg2nk9dLZ2mShWaEBan6FAoqfSigmmuDg=
 golang.org/x/sync v0.3.0 h1:ftCYgMx6zT/asHUrPw8BLLscYtGznsLAnjq5RH9P66E=
 golang.org/x/sync v0.3.0/go.mod h1:FU7BRWz2tNW+3quACPkgCx/L+uEAv1htQ0V83Z9Rj+Y=
 golang.org/x/sys v0.0.0-20201119102817-f84b799fce68/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
 golang.org/x/sys v0.0.0-20210615035016-665e8c7367d1/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
 golang.org/x/sys v0.0.0-20210630005230-0f9fa26af87c/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
 golang.org/x/sys v0.0.0-20210806184541-e5e7981a1069/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
+golang.org/x/sys v0.0.0-20220310020820-b874c991c1a5/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
 golang.org/x/sys v0.0.0-20220704084225-05e143d24a9e/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
 golang.org/x/sys v0.6.0/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
-golang.org/x/sys v0.13.0 h1:Af8nKPmuFypiUBjVoU9V20FiaFXOcuZI21p0ycVYYGE=
-golang.org/x/sys v0.13.0/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
+golang.org/x/sys v0.11.0 h1:eG7RXZHdqOJ1i+0lgLgCpSXAp6M3LYlAo6osgSi0xOM=
+golang.org/x/sys v0.11.0/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
 golang.org/x/term v0.0.0-20201126162022-7de9c90e9dd1/go.mod h1:bj7SfCRtBDWHUb9snDiAeCFNEtKQo2Wmx5Cou7ajbmo=
-golang.org/x/term v0.13.0 h1:bb+I9cTfFazGW51MZqBVmZy7+JEJMouUHTUSKVQLBek=
-golang.org/x/term v0.13.0/go.mod h1:LTmsnFJwVN6bCy1rVCoS+qHT1HhALEFxKncY3WNNh4U=
+golang.org/x/term v0.10.0 h1:3R7pNqamzBraeqj/Tj8qt1aQ2HpmlC+Cx/qL/7hn4/c=
+golang.org/x/term v0.10.0/go.mod h1:lpqdcUyK/oCiQxvxVrppt5ggO2KCZ5QblwqPnfZ6d5o=
 golang.org/x/text v0.3.3/go.mod h1:5Zoc/QRtKVWzQhOtBMvqHzDpF6irO9z98xDceosuGiQ=
 golang.org/x/text v0.3.6/go.mod h1:5Zoc/QRtKVWzQhOtBMvqHzDpF6irO9z98xDceosuGiQ=
-golang.org/x/text v0.13.0 h1:ablQoSUd0tRdKxZewP80B+BaqeKJuVhuRxj/dkrun3k=
-golang.org/x/text v0.13.0/go.mod h1:TvPlkZtksWOMsz7fbANvkp4WM8x/WCo/om8BMLbz+aE=
+golang.org/x/text v0.10.0 h1:UpjohKhiEgNc0CSauXmwYftY1+LlaC75SJwh0SgCX58=
+golang.org/x/text v0.10.0/go.mod h1:TvPlkZtksWOMsz7fbANvkp4WM8x/WCo/om8BMLbz+aE=
 golang.org/x/tools v0.0.0-20180917221912-90fa682c2a6e/go.mod h1:n7NCudcB/nEzxVGmLbDWY5pfWTLqBcC2KZ6jyYvM4mQ=
 golang.org/x/xerrors v0.0.0-20191204190536-9bdfabe68543/go.mod h1:I/5z698sn9Ka8TeJc9MKroUUfqBBauWjQqLJ2OPfmY0=
 google.golang.org/protobuf v1.26.0-rc.1/go.mod h1:jlhhOSvTdKEhbULTjvd4ARK9grFBp09yW+WbY/TyQbw=

llm/ggml.go

@@ -175,8 +175,7 @@ const (
 	// Magic constant for `ggla` files (LoRA adapter).
 	FILE_MAGIC_GGLA = 0x67676C61
 	// Magic constant for `gguf` files (versioned, gguf)
-	FILE_MAGIC_GGUF_LE = 0x46554747
-	FILE_MAGIC_GGUF_BE = 0x47475546
+	FILE_MAGIC_GGUF = 0x46554747
 )
 
 func DecodeGGML(r io.ReadSeeker) (*GGML, error) {
@@ -192,10 +191,8 @@ func DecodeGGML(r io.ReadSeeker) (*GGML, error) {
 		ggml.container = &containerGGJT{}
 	case FILE_MAGIC_GGLA:
 		ggml.container = &containerLORA{}
-	case FILE_MAGIC_GGUF_LE:
-		ggml.container = &containerGGUF{bo: binary.LittleEndian}
-	case FILE_MAGIC_GGUF_BE:
-		ggml.container = &containerGGUF{bo: binary.BigEndian}
+	case FILE_MAGIC_GGUF:
+		ggml.container = &containerGGUF{}
 	default:
 		return nil, errors.New("invalid file magic")
 	}

llm/gguf.go

@@ -3,13 +3,12 @@ package llm
 import (
 	"bytes"
 	"encoding/binary"
+	"errors"
 	"fmt"
 	"io"
 )
 
 type containerGGUF struct {
-	bo binary.ByteOrder
-
 	Version uint32
 
 	V1 struct {
@@ -28,13 +27,15 @@ func (c *containerGGUF) Name() string {
 }
 
 func (c *containerGGUF) Decode(r io.Reader) (model, error) {
-	binary.Read(r, c.bo, &c.Version)
+	binary.Read(r, binary.LittleEndian, &c.Version)
 
 	switch c.Version {
 	case 1:
-		binary.Read(r, c.bo, &c.V1)
+		binary.Read(r, binary.LittleEndian, &c.V1)
+	case 2:
+		binary.Read(r, binary.LittleEndian, &c.V2)
 	default:
-		binary.Read(r, c.bo, &c.V2)
+		return nil, errors.New("invalid version")
 	}
 
 	model := newGGUFModel(c)
@@ -208,75 +209,75 @@ func (llm *ggufModel) NumLayers() int64 {
 	return int64(v)
 }
 
-func (llm ggufModel) readU8(r io.Reader) uint8 {
+func (ggufModel) readU8(r io.Reader) uint8 {
 	var u8 uint8
-	binary.Read(r, llm.bo, &u8)
+	binary.Read(r, binary.LittleEndian, &u8)
 	return u8
 }
 
-func (llm ggufModel) readI8(r io.Reader) int8 {
+func (ggufModel) readI8(r io.Reader) int8 {
 	var i8 int8
-	binary.Read(r, llm.bo, &i8)
+	binary.Read(r, binary.LittleEndian, &i8)
 	return i8
 }
 
-func (llm ggufModel) readU16(r io.Reader) uint16 {
+func (ggufModel) readU16(r io.Reader) uint16 {
 	var u16 uint16
-	binary.Read(r, llm.bo, &u16)
+	binary.Read(r, binary.LittleEndian, &u16)
 	return u16
 }
 
-func (llm ggufModel) readI16(r io.Reader) int16 {
+func (ggufModel) readI16(r io.Reader) int16 {
 	var i16 int16
-	binary.Read(r, llm.bo, &i16)
+	binary.Read(r, binary.LittleEndian, &i16)
 	return i16
 }
 
-func (llm ggufModel) readU32(r io.Reader) uint32 {
+func (ggufModel) readU32(r io.Reader) uint32 {
 	var u32 uint32
-	binary.Read(r, llm.bo, &u32)
+	binary.Read(r, binary.LittleEndian, &u32)
 	return u32
 }
 
-func (llm ggufModel) readI32(r io.Reader) int32 {
+func (ggufModel) readI32(r io.Reader) int32 {
 	var i32 int32
-	binary.Read(r, llm.bo, &i32)
+	binary.Read(r, binary.LittleEndian, &i32)
 	return i32
 }
 
-func (llm ggufModel) readU64(r io.Reader) uint64 {
+func (ggufModel) readU64(r io.Reader) uint64 {
 	var u64 uint64
-	binary.Read(r, llm.bo, &u64)
+	binary.Read(r, binary.LittleEndian, &u64)
 	return u64
 }
 
-func (llm ggufModel) readI64(r io.Reader) int64 {
+func (ggufModel) readI64(r io.Reader) int64 {
 	var i64 int64
-	binary.Read(r, llm.bo, &i64)
+	binary.Read(r, binary.LittleEndian, &i64)
 	return i64
 }
 
-func (llm ggufModel) readF32(r io.Reader) float32 {
+func (ggufModel) readF32(r io.Reader) float32 {
 	var f32 float32
-	binary.Read(r, llm.bo, &f32)
+	binary.Read(r, binary.LittleEndian, &f32)
 	return f32
 }
 
-func (llm ggufModel) readF64(r io.Reader) float64 {
+func (ggufModel) readF64(r io.Reader) float64 {
 	var f64 float64
-	binary.Read(r, llm.bo, &f64)
+	binary.Read(r, binary.LittleEndian, &f64)
 	return f64
 }
 
-func (llm ggufModel) readBool(r io.Reader) bool {
+func (ggufModel) readBool(r io.Reader) bool {
 	var b bool
-	binary.Read(r, llm.bo, &b)
+	binary.Read(r, binary.LittleEndian, &b)
 	return b
 }
 
-func (llm ggufModel) readStringV1(r io.Reader) (string, error) {
+func (ggufModel) readStringV1(r io.Reader) (string, error) {
 	var nameLength uint32
-	binary.Read(r, llm.bo, &nameLength)
+	binary.Read(r, binary.LittleEndian, &nameLength)
 
 	var b bytes.Buffer
 	if _, err := io.CopyN(&b, r, int64(nameLength)); err != nil {
@@ -291,7 +292,7 @@ func (llm ggufModel) readStringV1(r io.Reader) (string, error) {
 
 func (llm ggufModel) readString(r io.Reader) (string, error) {
 	var nameLength uint64
-	binary.Read(r, llm.bo, &nameLength)
+	binary.Read(r, binary.LittleEndian, &nameLength)
 
 	var b bytes.Buffer
 	if _, err := io.CopyN(&b, r, int64(nameLength)); err != nil {

llm/llama.cpp/generate_darwin_amd64.go

@@ -12,8 +12,7 @@ package llm
 //go:generate mv ggml/build/cpu/bin/server ggml/build/cpu/bin/ollama-runner
 
 //go:generate git submodule update --force gguf
-//go:generate git -C gguf apply ../patches/0001-update-default-log-target.patch
-//go:generate git -C gguf apply ../patches/0001-metal-handle-ggml_scale-for-n-4-0-close-3754.patch
+//go:generate git -C gguf apply ../patches/0001-remove-warm-up-logging.patch
 //go:generate cmake -S gguf -B gguf/build/cpu -DLLAMA_ACCELERATE=on -DLLAMA_K_QUANTS=on -DCMAKE_SYSTEM_PROCESSOR=x86_64 -DCMAKE_OSX_ARCHITECTURES=x86_64 -DCMAKE_OSX_DEPLOYMENT_TARGET=11.0
 //go:generate cmake --build gguf/build/cpu --target server --config Release
 //go:generate mv gguf/build/cpu/bin/server gguf/build/cpu/bin/ollama-runner

llm/llama.cpp/generate_darwin_arm64.go

@@ -12,8 +12,7 @@ package llm
 //go:generate mv ggml/build/metal/bin/server ggml/build/metal/bin/ollama-runner
 
 //go:generate git submodule update --force gguf
-//go:generate git -C gguf apply ../patches/0001-update-default-log-target.patch
-//go:generate git -C gguf apply ../patches/0001-metal-handle-ggml_scale-for-n-4-0-close-3754.patch
+//go:generate git -C gguf apply ../patches/0001-remove-warm-up-logging.patch
 //go:generate cmake -S gguf -B gguf/build/metal -DLLAMA_METAL=on -DLLAMA_ACCELERATE=on -DLLAMA_K_QUANTS=on -DCMAKE_SYSTEM_PROCESSOR=arm64 -DCMAKE_OSX_ARCHITECTURES=arm64 -DCMAKE_OSX_DEPLOYMENT_TARGET=11.0
 //go:generate cmake --build gguf/build/metal --target server --config Release
 //go:generate mv gguf/build/metal/bin/server gguf/build/metal/bin/ollama-runner

llm/llama.cpp/generate_linux.go

@@ -13,14 +13,14 @@ package llm
 
 //go:generate git submodule update --force gguf
 //go:generate git -C gguf apply ../patches/0001-copy-cuda-runtime-libraries.patch
-//go:generate git -C gguf apply ../patches/0001-update-default-log-target.patch
-//go:generate cmake -S gguf -B gguf/build/cpu -DLLAMA_K_QUANTS=on -DLLAMA_NATIVE=off -DLLAMA_AVX=on -DLLAMA_AVX2=off -DLLAMA_AVX512=off -DLLAMA_FMA=off -DLLAMA_F16C=off
+//go:generate git -C gguf apply ../patches/0001-remove-warm-up-logging.patch
+//go:generate cmake -S gguf -B gguf/build/cpu -DLLAMA_K_QUANTS=on
 //go:generate cmake --build gguf/build/cpu --target server --config Release
 //go:generate mv gguf/build/cpu/bin/server gguf/build/cpu/bin/ollama-runner
 
 //go:generate cmake -S ggml -B ggml/build/cuda -DLLAMA_CUBLAS=on -DLLAMA_ACCELERATE=on -DLLAMA_K_QUANTS=on
 //go:generate cmake --build ggml/build/cuda --target server --config Release
 //go:generate mv ggml/build/cuda/bin/server ggml/build/cuda/bin/ollama-runner
-//go:generate cmake -S gguf -B gguf/build/cuda -DLLAMA_CUBLAS=on -DLLAMA_ACCELERATE=on -DLLAMA_K_QUANTS=on -DLLAMA_NATIVE=off -DLLAMA_AVX=on -DLLAMA_AVX2=off -DLLAMA_AVX512=off -DLLAMA_FMA=off -DLLAMA_F16C=off
+//go:generate cmake -S gguf -B gguf/build/cuda -DLLAMA_CUBLAS=on -DLLAMA_ACCELERATE=on -DLLAMA_K_QUANTS=on
 //go:generate cmake --build gguf/build/cuda --target server --config Release
 //go:generate mv gguf/build/cuda/bin/server gguf/build/cuda/bin/ollama-runner

llm/llama.cpp/generate_windows.go

@@ -10,7 +10,7 @@ package llm
 //go:generate cmd /c move ggml\build\cpu\bin\Release\server.exe ggml\build\cpu\bin\Release\ollama-runner.exe
 
 //go:generate git submodule update --force gguf
-//go:generate git -C gguf apply ../patches/0001-update-default-log-target.patch
-//go:generate cmake -S gguf -B gguf/build/cpu -DLLAMA_K_QUANTS=on -DLLAMA_NATIVE=off -DLLAMA_AVX=on -DLLAMA_AVX2=off -DLLAMA_AVX512=off -DLLAMA_FMA=off -DLLAMA_F16C=off
+//go:generate git -C gguf apply ../patches/0001-remove-warm-up-logging.patch
+//go:generate cmake -S gguf -B gguf/build/cpu -DLLAMA_K_QUANTS=on
 //go:generate cmake --build gguf/build/cpu --target server --config Release
 //go:generate cmd /c move gguf\build\cpu\bin\Release\server.exe gguf\build\cpu\bin\Release\ollama-runner.exe

llm/llama.cpp/patches/0001-metal-handle-ggml_scale-for-n-4-0-close-3754.patch

@@ -1,91 +0,0 @@
-From 469c9addef75893e6be12edda852d12e840bf064 Mon Sep 17 00:00:00 2001
-From: Georgi Gerganov <ggerganov@gmail.com>
-Date: Tue, 24 Oct 2023 09:46:50 +0300
-Subject: [PATCH 1/2] metal : handle ggml_scale for n%4 != 0 (close #3754)
-
-ggml-ci
----
- ggml-metal.m     | 18 +++++++++++++-----
- ggml-metal.metal | 10 +++++++++-
- 2 files changed, 22 insertions(+), 6 deletions(-)
-
-diff --git a/ggml-metal.m b/ggml-metal.m
-index c908106..c1901dc 100644
---- a/ggml-metal.m
-+++ b/ggml-metal.m
-@@ -62,6 +62,7 @@
-     GGML_METAL_DECL_KERNEL(mul);
-     GGML_METAL_DECL_KERNEL(mul_row); // TODO: avoid this extra kernel, instead extend the "mul" kernel to support broadcast
-     GGML_METAL_DECL_KERNEL(scale);
-+    GGML_METAL_DECL_KERNEL(scale_4);
-     GGML_METAL_DECL_KERNEL(silu);
-     GGML_METAL_DECL_KERNEL(relu);
-     GGML_METAL_DECL_KERNEL(gelu);
-@@ -249,6 +250,7 @@ static void ggml_metal_log(enum ggml_log_level level, const char* format, ...){
-         GGML_METAL_ADD_KERNEL(mul);
-         GGML_METAL_ADD_KERNEL(mul_row);
-         GGML_METAL_ADD_KERNEL(scale);
-+        GGML_METAL_ADD_KERNEL(scale_4);
-         GGML_METAL_ADD_KERNEL(silu);
-         GGML_METAL_ADD_KERNEL(relu);
-         GGML_METAL_ADD_KERNEL(gelu);
-@@ -347,6 +349,7 @@ void ggml_metal_free(struct ggml_metal_context * ctx) {
-     GGML_METAL_DEL_KERNEL(mul);
-     GGML_METAL_DEL_KERNEL(mul_row);
-     GGML_METAL_DEL_KERNEL(scale);
-+    GGML_METAL_DEL_KERNEL(scale_4);
-     GGML_METAL_DEL_KERNEL(silu);
-     GGML_METAL_DEL_KERNEL(relu);
-     GGML_METAL_DEL_KERNEL(gelu);
-@@ -923,15 +926,20 @@ void ggml_metal_graph_compute(
- 
-                             const float scale = *(const float *) src1->data;
- 
--                            [encoder setComputePipelineState:ctx->pipeline_scale];
-+                            int64_t n = ggml_nelements(dst);
-+
-+                            if (n % 4 == 0) {
-+                                n /= 4;
-+                                [encoder setComputePipelineState:ctx->pipeline_scale_4];
-+                            } else {
-+                                [encoder setComputePipelineState:ctx->pipeline_scale];
-+                            }
-+
-                             [encoder setBuffer:id_src0 offset:offs_src0 atIndex:0];
-                             [encoder setBuffer:id_dst  offset:offs_dst  atIndex:1];
-                             [encoder setBytes:&scale length:sizeof(scale) atIndex:2];
- 
--                            const int64_t n = ggml_nelements(dst);
--                            GGML_ASSERT(n % 4 == 0);
--
--                            [encoder dispatchThreadgroups:MTLSizeMake(n/4, 1, 1) threadsPerThreadgroup:MTLSizeMake(1, 1, 1)];
-+                            [encoder dispatchThreadgroups:MTLSizeMake(n, 1, 1) threadsPerThreadgroup:MTLSizeMake(1, 1, 1)];
-                         } break;
-                     case GGML_OP_UNARY:
-                         switch (ggml_get_unary_op(gf->nodes[i])) {
-diff --git a/ggml-metal.metal b/ggml-metal.metal
-index 69fc713..f4b4605 100644
---- a/ggml-metal.metal
-+++ b/ggml-metal.metal
-@@ -125,9 +125,17 @@ kernel void kernel_mul_row(
- }
- 
- kernel void kernel_scale(
-+        device const float * src0,
-+        device       float * dst,
-+        constant     float & scale,
-+        uint tpig[[thread_position_in_grid]]) {
-+    dst[tpig] = src0[tpig] * scale;
-+}
-+
-+kernel void kernel_scale_4(
-         device const float4 * src0,
-         device       float4 * dst,
--        constant     float & scale,
-+        constant     float  & scale,
-         uint tpig[[thread_position_in_grid]]) {
-     dst[tpig] = src0[tpig] * scale;
- }
--- 
-2.39.3 (Apple Git-145)
-

llm/llama.cpp/patches/0001-remove-warm-up-logging.patch

@@ -0,0 +1,25 @@
+From 8dbb5449db259a9c24796e7927d89bee98b6c8f5 Mon Sep 17 00:00:00 2001
+From: Bruce MacDonald <brucewmacdonald@gmail.com>
+Date: Thu, 5 Oct 2023 11:21:12 -0400
+Subject: [PATCH] remove warm up logging
+
+---
+ common/common.cpp | 2 --
+ 1 file changed, 2 deletions(-)
+
+diff --git a/common/common.cpp b/common/common.cpp
+index 7370017..c4433fe 100644
+--- a/common/common.cpp
++++ b/common/common.cpp
+@@ -839,8 +839,6 @@ std::tuple<struct llama_model *, struct llama_context *> llama_init_from_gpt_par
+     }
+ 
+     {
+-        LOG("warming up the model with an empty run\n");
+-
+         std::vector<llama_token> tmp = { llama_token_bos(lctx), llama_token_eos(lctx), };
+         llama_decode(lctx, llama_batch_get_one(tmp.data(), std::min(tmp.size(), (size_t) params.n_batch), 0, 0));
+         llama_kv_cache_tokens_rm(lctx, -1, -1);
+-- 
+2.39.2 (Apple Git-143)
+

llm/llama.cpp/patches/0001-update-default-log-target.patch

@@ -1,25 +0,0 @@
-From 6465fec6290f0a7f5d4d0fbe6bcf634e4810dde6 Mon Sep 17 00:00:00 2001
-From: Michael Yang <mxyng@pm.me>
-Date: Mon, 23 Oct 2023 10:39:34 -0700
-Subject: [PATCH] default log stderr
-
----
- common/log.h | 2 +-
- 1 file changed, 1 insertion(+), 1 deletion(-)
-
-diff --git a/common/log.h b/common/log.h
-index b8953fd..25522cd 100644
---- a/common/log.h
-+++ b/common/log.h
-@@ -90,7 +90,7 @@
- //  }
- //
- #ifndef LOG_TARGET
--    #define LOG_TARGET log_handler()
-+    #define LOG_TARGET nullptr
- #endif
- 
- #ifndef LOG_TEE_TARGET
--- 
-2.42.0
-

llm/llama.go

@@ -212,10 +212,6 @@ func CheckVRAM() (int64, error) {
 	scanner := bufio.NewScanner(&stdout)
 	for scanner.Scan() {
 		line := scanner.Text()
-		if strings.Contains(line, "[Insufficient Permissions]") {
-			return 0, fmt.Errorf("GPU support may not enabled, check you have installed GPU drivers and have the necessary permissions to run nvidia-smi")
-		}
-
 		vram, err := strconv.ParseInt(strings.TrimSpace(line), 10, 64)
 		if err != nil {
 			return 0, fmt.Errorf("failed to parse available VRAM: %v", err)
@@ -247,15 +243,12 @@ func NumGPU(numLayer, fileSizeBytes int64, opts api.Options) int {
 			return 0
 		}
 
-		/*
-		 Calculate bytes per layer, this will roughly be the size of the model file divided by the number of layers.
-		 We can store the model weights and the kv cache in vram,
-		 to enable kv chache vram storage add two additional layers to the number of layers retrieved from the model file.
-		*/
+		// Calculate bytes per layer
+		// TODO: this is a rough heuristic, better would be to calculate this based on number of layers and context size
 		bytesPerLayer := fileSizeBytes / numLayer
 
-		// 75% of the absolute max number of layers we can fit in available VRAM, off-loading too many layers to the GPU can cause OOM errors
-		layers := int(freeBytes/bytesPerLayer) * 3 / 4
+		// max number of layers we can fit in VRAM, subtract 8% to prevent consuming all available VRAM and running out of memory
+		layers := int(freeBytes/bytesPerLayer) * 92 / 100
 		log.Printf("%d MB VRAM available, loading up to %d GPU layers", freeBytes/(1024*1024), layers)
 
 		return layers
@@ -306,19 +299,13 @@ func newLlama(model string, adapters []string, runners []ModelRunner, numLayers
 	params := []string{
 		"--model", model,
 		"--ctx-size", fmt.Sprintf("%d", opts.NumCtx),
+		"--rope-freq-base", fmt.Sprintf("%f", opts.RopeFrequencyBase),
+		"--rope-freq-scale", fmt.Sprintf("%f", opts.RopeFrequencyScale),
 		"--batch-size", fmt.Sprintf("%d", opts.NumBatch),
 		"--n-gpu-layers", fmt.Sprintf("%d", numGPU),
 		"--embedding",
 	}
 
-	if opts.RopeFrequencyBase > 0 {
-		params = append(params, "--rope-freq-base", fmt.Sprintf("%f", opts.RopeFrequencyBase))
-	}
-
-	if opts.RopeFrequencyScale > 0 {
-		params = append(params, "--rope-freq-scale", fmt.Sprintf("%f", opts.RopeFrequencyScale))
-	}
-
 	if opts.NumGQA > 0 {
 		params = append(params, "--gqa", fmt.Sprintf("%d", opts.NumGQA))
 	}
@@ -366,15 +353,7 @@ func newLlama(model string, adapters []string, runners []ModelRunner, numLayers
 			runner.Path,
 			append(params, "--port", strconv.Itoa(port))...,
 		)
-
-		var libraryPaths []string
-		if libraryPath, ok := os.LookupEnv("LD_LIBRARY_PATH"); ok {
-			libraryPaths = append(libraryPaths, libraryPath)
-		}
-
-		libraryPaths = append(libraryPaths, filepath.Dir(runner.Path))
-
-		cmd.Env = append(os.Environ(), fmt.Sprintf("LD_LIBRARY_PATH=%s", strings.Join(libraryPaths, ":")))
+		cmd.Env = append(os.Environ(), fmt.Sprintf("LD_LIBRARY_PATH=%s", filepath.Dir(runner.Path)))
 		cmd.Stdout = os.Stderr
 		statusWriter := NewStatusWriter()
 		cmd.Stderr = statusWriter

llm/llm.go

@@ -85,10 +85,7 @@ func New(workDir, model string, adapters []string, opts api.Options) (LLM, error
 
 	switch ggml.Name() {
 	case "gguf":
-		// TODO: gguf will load these options automatically from the model binary
-		opts.NumGQA = 0
-		opts.RopeFrequencyBase = 0.0
-		opts.RopeFrequencyScale = 0.0
+		opts.NumGQA = 0 // TODO: remove this when llama.cpp runners differ enough to need separate newLlama functions
 		return newLlama(model, adapters, chooseRunners(workDir, "gguf"), ggml.NumLayers(), opts)
 	case "ggml", "ggmf", "ggjt", "ggla":
 		return newLlama(model, adapters, chooseRunners(workDir, "ggml"), ggml.NumLayers(), opts)

readline/buffer.go

@@ -1,372 +0,0 @@
-package readline
-
-import (
-	"fmt"
-	"os"
-
-	"github.com/emirpasic/gods/lists/arraylist"
-	"golang.org/x/term"
-)
-
-type Buffer struct {
-	Pos       int
-	Buf       *arraylist.List
-	Prompt    *Prompt
-	LineWidth int
-	Width     int
-	Height    int
-}
-
-func NewBuffer(prompt *Prompt) (*Buffer, error) {
-	fd := int(os.Stdout.Fd())
-	width, height, err := term.GetSize(fd)
-	if err != nil {
-		fmt.Println("Error getting size:", err)
-		return nil, err
-	}
-
-	lwidth := width - len(prompt.Prompt)
-	if prompt.UseAlt {
-		lwidth = width - len(prompt.AltPrompt)
-	}
-
-	b := &Buffer{
-		Pos:       0,
-		Buf:       arraylist.New(),
-		Prompt:    prompt,
-		Width:     width,
-		Height:    height,
-		LineWidth: lwidth,
-	}
-
-	return b, nil
-}
-
-func (b *Buffer) MoveLeft() {
-	if b.Pos > 0 {
-		if b.Pos%b.LineWidth == 0 {
-			fmt.Printf(CursorUp + CursorBOL + cursorRightN(b.Width))
-		} else {
-			fmt.Print(CursorLeft)
-		}
-		b.Pos -= 1
-	}
-}
-
-func (b *Buffer) MoveLeftWord() {
-	if b.Pos > 0 {
-		var foundNonspace bool
-		for {
-			v, _ := b.Buf.Get(b.Pos - 1)
-			if v == ' ' {
-				if foundNonspace {
-					break
-				}
-			} else {
-				foundNonspace = true
-			}
-			b.MoveLeft()
-
-			if b.Pos == 0 {
-				break
-			}
-		}
-	}
-}
-
-func (b *Buffer) MoveRight() {
-	if b.Pos < b.Size() {
-		b.Pos += 1
-		if b.Pos%b.LineWidth == 0 {
-			fmt.Printf(CursorDown + CursorBOL + cursorRightN(b.PromptSize()))
-		} else {
-			fmt.Print(CursorRight)
-		}
-	}
-}
-
-func (b *Buffer) MoveRightWord() {
-	if b.Pos < b.Size() {
-		for {
-			b.MoveRight()
-			v, _ := b.Buf.Get(b.Pos)
-			if v == ' ' {
-				break
-			}
-
-			if b.Pos == b.Size() {
-				break
-			}
-		}
-	}
-}
-
-func (b *Buffer) MoveToStart() {
-	if b.Pos > 0 {
-		currLine := b.Pos / b.LineWidth
-		if currLine > 0 {
-			for cnt := 0; cnt < currLine; cnt++ {
-				fmt.Print(CursorUp)
-			}
-		}
-		fmt.Printf(CursorBOL + cursorRightN(b.PromptSize()))
-		b.Pos = 0
-	}
-}
-
-func (b *Buffer) MoveToEnd() {
-	if b.Pos < b.Size() {
-		currLine := b.Pos / b.LineWidth
-		totalLines := b.Size() / b.LineWidth
-		if currLine < totalLines {
-			for cnt := 0; cnt < totalLines-currLine; cnt++ {
-				fmt.Print(CursorDown)
-			}
-			remainder := b.Size() % b.LineWidth
-			fmt.Printf(CursorBOL + cursorRightN(b.PromptSize()+remainder))
-		} else {
-			fmt.Print(cursorRightN(b.Size() - b.Pos))
-		}
-
-		b.Pos = b.Size()
-	}
-}
-
-func (b *Buffer) Size() int {
-	return b.Buf.Size()
-}
-
-func min(n, m int) int {
-	if n > m {
-		return m
-	}
-	return n
-}
-
-func (b *Buffer) PromptSize() int {
-	if b.Prompt.UseAlt {
-		return len(b.Prompt.AltPrompt)
-	}
-	return len(b.Prompt.Prompt)
-}
-
-func (b *Buffer) Add(r rune) {
-	if b.Pos == b.Buf.Size() {
-		fmt.Printf("%c", r)
-		b.Buf.Add(r)
-		b.Pos += 1
-		if b.Pos > 0 && b.Pos%b.LineWidth == 0 {
-			fmt.Printf("\n%s", b.Prompt.AltPrompt)
-		}
-	} else {
-		fmt.Printf("%c", r)
-		b.Buf.Insert(b.Pos, r)
-		b.Pos += 1
-		if b.Pos > 0 && b.Pos%b.LineWidth == 0 {
-			fmt.Printf("\n%s", b.Prompt.AltPrompt)
-		}
-		b.drawRemaining()
-	}
-}
-
-func (b *Buffer) drawRemaining() {
-	var place int
-	remainingText := b.StringN(b.Pos)
-	if b.Pos > 0 {
-		place = b.Pos % b.LineWidth
-	}
-	fmt.Print(CursorHide)
-
-	// render the rest of the current line
-	currLine := remainingText[:min(b.LineWidth-place, len(remainingText))]
-	if len(currLine) > 0 {
-		fmt.Printf(ClearToEOL + currLine)
-		fmt.Print(cursorLeftN(len(currLine)))
-	} else {
-		fmt.Print(ClearToEOL)
-	}
-
-	// render the other lines
-	if len(remainingText) > len(currLine) {
-		remaining := []rune(remainingText[len(currLine):])
-		var totalLines int
-		for i, c := range remaining {
-			if i%b.LineWidth == 0 {
-				fmt.Printf("\n%s", b.Prompt.AltPrompt)
-				totalLines += 1
-			}
-			fmt.Printf("%c", c)
-		}
-		fmt.Print(ClearToEOL)
-		fmt.Print(cursorUpN(totalLines))
-		fmt.Printf(CursorBOL + cursorRightN(b.Width-len(currLine)))
-	}
-
-	fmt.Print(CursorShow)
-}
-
-func (b *Buffer) Remove() {
-	if b.Buf.Size() > 0 && b.Pos > 0 {
-		if b.Pos%b.LineWidth == 0 {
-			// if the user backspaces over the word boundary, do this magic to clear the line
-			// and move to the end of the previous line
-			fmt.Printf(CursorBOL + ClearToEOL)
-			fmt.Printf(CursorUp + CursorBOL + cursorRightN(b.Width) + " " + CursorLeft)
-		} else {
-			fmt.Printf(CursorLeft + " " + CursorLeft)
-		}
-
-		var eraseExtraLine bool
-		if (b.Size()-1)%b.LineWidth == 0 {
-			eraseExtraLine = true
-		}
-
-		b.Pos -= 1
-		b.Buf.Remove(b.Pos)
-
-		if b.Pos < b.Size() {
-			b.drawRemaining()
-			// this erases a line which is left over when backspacing in the middle of a line and there
-			// are trailing characters which go over the line width boundary
-			if eraseExtraLine {
-				remainingLines := (b.Size() - b.Pos) / b.LineWidth
-				fmt.Printf(cursorDownN(remainingLines+1) + CursorBOL + ClearToEOL)
-				place := b.Pos % b.LineWidth
-				fmt.Printf(cursorUpN(remainingLines+1) + cursorRightN(place+len(b.Prompt.Prompt)))
-			}
-		}
-	}
-}
-
-func (b *Buffer) Delete() {
-	if b.Size() > 0 && b.Pos < b.Size() {
-		b.Buf.Remove(b.Pos)
-		b.drawRemaining()
-		if b.Size()%b.LineWidth == 0 {
-			if b.Pos != b.Size() {
-				remainingLines := (b.Size() - b.Pos) / b.LineWidth
-				fmt.Printf(cursorDownN(remainingLines) + CursorBOL + ClearToEOL)
-				place := b.Pos % b.LineWidth
-				fmt.Printf(cursorUpN(remainingLines) + cursorRightN(place+len(b.Prompt.Prompt)))
-			}
-		}
-	}
-}
-
-func (b *Buffer) DeleteBefore() {
-	if b.Pos > 0 {
-		for cnt := b.Pos - 1; cnt >= 0; cnt-- {
-			b.Remove()
-		}
-	}
-}
-
-func (b *Buffer) DeleteRemaining() {
-	if b.Size() > 0 && b.Pos < b.Size() {
-		charsToDel := b.Size() - b.Pos
-		for cnt := 0; cnt < charsToDel; cnt++ {
-			b.Delete()
-		}
-	}
-}
-
-func (b *Buffer) DeleteWord() {
-	if b.Buf.Size() > 0 && b.Pos > 0 {
-		var foundNonspace bool
-		for {
-			v, _ := b.Buf.Get(b.Pos - 1)
-			if v == ' ' {
-				if !foundNonspace {
-					b.Remove()
-				} else {
-					break
-				}
-			} else {
-				foundNonspace = true
-				b.Remove()
-			}
-
-			if b.Pos == 0 {
-				break
-			}
-		}
-	}
-}
-
-func (b *Buffer) ClearScreen() {
-	fmt.Printf(ClearScreen + CursorReset + b.Prompt.Prompt)
-	if b.IsEmpty() {
-		ph := b.Prompt.Placeholder
-		fmt.Printf(ColorGrey + ph + cursorLeftN(len(ph)) + ColorDefault)
-	} else {
-		currPos := b.Pos
-		b.Pos = 0
-		b.drawRemaining()
-		fmt.Printf(CursorReset + cursorRightN(len(b.Prompt.Prompt)))
-		if currPos > 0 {
-			targetLine := currPos / b.LineWidth
-			if targetLine > 0 {
-				for cnt := 0; cnt < targetLine; cnt++ {
-					fmt.Print(CursorDown)
-				}
-			}
-			remainder := currPos % b.LineWidth
-			if remainder > 0 {
-				fmt.Print(cursorRightN(remainder))
-			}
-			if currPos%b.LineWidth == 0 {
-				fmt.Printf(CursorBOL + b.Prompt.AltPrompt)
-			}
-		}
-		b.Pos = currPos
-	}
-}
-
-func (b *Buffer) IsEmpty() bool {
-	return b.Buf.Empty()
-}
-
-func (b *Buffer) Replace(r []rune) {
-	b.Pos = 0
-	b.Buf.Clear()
-	fmt.Printf(ClearLine + CursorBOL + b.Prompt.Prompt)
-	for _, c := range r {
-		b.Add(c)
-	}
-}
-
-func (b *Buffer) String() string {
-	return b.StringN(0)
-}
-
-func (b *Buffer) StringN(n int) string {
-	return b.StringNM(n, 0)
-}
-
-func (b *Buffer) StringNM(n, m int) string {
-	var s string
-	if m == 0 {
-		m = b.Size()
-	}
-	for cnt := n; cnt < m; cnt++ {
-		c, _ := b.Buf.Get(cnt)
-		s += string(c.(rune))
-	}
-	return s
-}
-
-func cursorLeftN(n int) string {
-	return fmt.Sprintf(CursorLeftN, n)
-}
-
-func cursorRightN(n int) string {
-	return fmt.Sprintf(CursorRightN, n)
-}
-
-func cursorUpN(n int) string {
-	return fmt.Sprintf(CursorUpN, n)
-}
-
-func cursorDownN(n int) string {
-	return fmt.Sprintf(CursorDownN, n)
-}

readline/errors.go

@@ -1,17 +0,0 @@
-package readline
-
-import (
-	"errors"
-)
-
-var (
-	ErrInterrupt = errors.New("Interrupt")
-)
-
-type InterruptError struct {
-	Line []rune
-}
-
-func (*InterruptError) Error() string {
-	return "Interrupted"
-}

readline/history.go

@@ -1,152 +0,0 @@
-package readline
-
-import (
-	"bufio"
-	"errors"
-	"io"
-	"os"
-	"path/filepath"
-	"strings"
-
-	"github.com/emirpasic/gods/lists/arraylist"
-)
-
-type History struct {
-	Buf      *arraylist.List
-	Autosave bool
-	Pos      int
-	Limit    int
-	Filename string
-	Enabled  bool
-}
-
-func NewHistory() (*History, error) {
-	h := &History{
-		Buf:      arraylist.New(),
-		Limit:    100, //resizeme
-		Autosave: true,
-		Enabled:  true,
-	}
-
-	err := h.Init()
-	if err != nil {
-		return nil, err
-	}
-
-	return h, nil
-}
-
-func (h *History) Init() error {
-	home, err := os.UserHomeDir()
-	if err != nil {
-		return err
-	}
-
-	path := filepath.Join(home, ".ollama", "history")
-	h.Filename = path
-
-	//todo check if the file exists
-	f, err := os.OpenFile(path, os.O_CREATE|os.O_RDONLY, 0600)
-	if err != nil {
-		if errors.Is(err, os.ErrNotExist) {
-			return nil
-		}
-		return err
-	}
-	defer f.Close()
-
-	r := bufio.NewReader(f)
-	for {
-		line, err := r.ReadString('\n')
-		if err != nil {
-			if err == io.EOF {
-				break
-			}
-			return err
-		}
-
-		line = strings.TrimSpace(line)
-		if len(line) == 0 {
-			continue
-		}
-
-		h.Add([]rune(line))
-	}
-
-	return nil
-}
-
-func (h *History) Add(l []rune) {
-	h.Buf.Add(l)
-	h.Compact()
-	h.Pos = h.Size()
-	if h.Autosave {
-		h.Save()
-	}
-}
-
-func (h *History) Compact() {
-	s := h.Buf.Size()
-	if s > h.Limit {
-		for cnt := 0; cnt < s-h.Limit; cnt++ {
-			h.Buf.Remove(0)
-		}
-	}
-}
-
-func (h *History) Clear() {
-	h.Buf.Clear()
-}
-
-func (h *History) Prev() []rune {
-	var line []rune
-	if h.Pos > 0 {
-		h.Pos -= 1
-	}
-	v, _ := h.Buf.Get(h.Pos)
-	line, _ = v.([]rune)
-	return line
-}
-
-func (h *History) Next() []rune {
-	var line []rune
-	if h.Pos < h.Buf.Size() {
-		h.Pos += 1
-		v, _ := h.Buf.Get(h.Pos)
-		line, _ = v.([]rune)
-	}
-	return line
-}
-
-func (h *History) Size() int {
-	return h.Buf.Size()
-}
-
-func (h *History) Save() error {
-	if !h.Enabled {
-		return nil
-	}
-
-	tmpFile := h.Filename + ".tmp"
-
-	f, err := os.OpenFile(tmpFile, os.O_CREATE|os.O_WRONLY|os.O_TRUNC|os.O_APPEND, 0666)
-	if err != nil {
-		return err
-	}
-	defer f.Close()
-
-	buf := bufio.NewWriter(f)
-	for cnt := 0; cnt < h.Size(); cnt++ {
-		v, _ := h.Buf.Get(cnt)
-		line, _ := v.([]rune)
-		buf.WriteString(string(line) + "\n")
-	}
-	buf.Flush()
-	f.Close()
-
-	if err = os.Rename(tmpFile, h.Filename); err != nil {
-		return err
-	}
-
-	return nil
-}

readline/readline.go

@@ -1,254 +0,0 @@
-package readline
-
-import (
-	"bufio"
-	"fmt"
-	"io"
-	"os"
-	"syscall"
-)
-
-type Prompt struct {
-	Prompt         string
-	AltPrompt      string
-	Placeholder    string
-	AltPlaceholder string
-	UseAlt         bool
-}
-
-type Terminal struct {
-	outchan chan rune
-}
-
-type Instance struct {
-	Prompt   *Prompt
-	Terminal *Terminal
-	History  *History
-}
-
-func New(prompt Prompt) (*Instance, error) {
-	term, err := NewTerminal()
-	if err != nil {
-		return nil, err
-	}
-
-	history, err := NewHistory()
-	if err != nil {
-		return nil, err
-	}
-
-	return &Instance{
-		Prompt:   &prompt,
-		Terminal: term,
-		History:  history,
-	}, nil
-}
-
-func (i *Instance) Readline() (string, error) {
-	prompt := i.Prompt.Prompt
-	if i.Prompt.UseAlt {
-		prompt = i.Prompt.AltPrompt
-	}
-	fmt.Print(prompt)
-
-	fd := int(syscall.Stdin)
-	termios, err := SetRawMode(fd)
-	if err != nil {
-		return "", err
-	}
-	defer UnsetRawMode(fd, termios)
-
-	buf, _ := NewBuffer(i.Prompt)
-
-	var esc bool
-	var escex bool
-	var metaDel bool
-	var pasteMode PasteMode
-
-	var currentLineBuf []rune
-
-	for {
-		if buf.IsEmpty() {
-			ph := i.Prompt.Placeholder
-			if i.Prompt.UseAlt {
-				ph = i.Prompt.AltPlaceholder
-			}
-			fmt.Printf(ColorGrey + ph + fmt.Sprintf(CursorLeftN, len(ph)) + ColorDefault)
-		}
-
-		r, err := i.Terminal.Read()
-
-		if buf.IsEmpty() {
-			fmt.Print(ClearToEOL)
-		}
-
-		if err != nil {
-			return "", io.EOF
-		}
-
-		if escex {
-			escex = false
-
-			switch r {
-			case KeyUp:
-				if i.History.Pos > 0 {
-					if i.History.Pos == i.History.Size() {
-						currentLineBuf = []rune(buf.String())
-					}
-					buf.Replace(i.History.Prev())
-				}
-			case KeyDown:
-				if i.History.Pos < i.History.Size() {
-					buf.Replace(i.History.Next())
-					if i.History.Pos == i.History.Size() {
-						buf.Replace(currentLineBuf)
-					}
-				}
-			case KeyLeft:
-				buf.MoveLeft()
-			case KeyRight:
-				buf.MoveRight()
-			case CharBracketedPaste:
-				var code string
-				for cnt := 0; cnt < 3; cnt++ {
-					r, err = i.Terminal.Read()
-					if err != nil {
-						return "", io.EOF
-					}
-
-					code += string(r)
-				}
-				if code == CharBracketedPasteStart {
-					pasteMode = PasteModeStart
-				} else if code == CharBracketedPasteEnd {
-					pasteMode = PasteModeEnd
-				}
-			case KeyDel:
-				if buf.Size() > 0 {
-					buf.Delete()
-				}
-				metaDel = true
-			case MetaStart:
-				buf.MoveToStart()
-			case MetaEnd:
-				buf.MoveToEnd()
-			default:
-				// skip any keys we don't know about
-				continue
-			}
-			continue
-		} else if esc {
-			esc = false
-
-			switch r {
-			case 'b':
-				buf.MoveLeftWord()
-			case 'f':
-				buf.MoveRightWord()
-			case CharEscapeEx:
-				escex = true
-			}
-			continue
-		}
-
-		switch r {
-		case CharNull:
-			continue
-		case CharEsc:
-			esc = true
-		case CharInterrupt:
-			return "", ErrInterrupt
-		case CharLineStart:
-			buf.MoveToStart()
-		case CharLineEnd:
-			buf.MoveToEnd()
-		case CharBackward:
-			buf.MoveLeft()
-		case CharForward:
-			buf.MoveRight()
-		case CharBackspace, CharCtrlH:
-			buf.Remove()
-		case CharTab:
-			// todo: convert back to real tabs
-			for cnt := 0; cnt < 8; cnt++ {
-				buf.Add(' ')
-			}
-		case CharDelete:
-			if buf.Size() > 0 {
-				buf.Delete()
-			} else {
-				return "", io.EOF
-			}
-		case CharKill:
-			buf.DeleteRemaining()
-		case CharCtrlU:
-			buf.DeleteBefore()
-		case CharCtrlL:
-			buf.ClearScreen()
-		case CharCtrlW:
-			buf.DeleteWord()
-		case CharEnter:
-			output := buf.String()
-			if output != "" {
-				i.History.Add([]rune(output))
-			}
-			buf.MoveToEnd()
-			fmt.Println()
-			switch pasteMode {
-			case PasteModeStart:
-				output = `"""` + output
-			case PasteModeEnd:
-				output = output + `"""`
-			}
-			return output, nil
-		default:
-			if metaDel {
-				metaDel = false
-				continue
-			}
-			if r >= CharSpace || r == CharEnter {
-				buf.Add(r)
-			}
-		}
-	}
-}
-
-func (i *Instance) HistoryEnable() {
-	i.History.Enabled = true
-}
-
-func (i *Instance) HistoryDisable() {
-	i.History.Enabled = false
-}
-
-func NewTerminal() (*Terminal, error) {
-	t := &Terminal{
-		outchan: make(chan rune),
-	}
-
-	go t.ioloop()
-
-	return t, nil
-}
-
-func (t *Terminal) ioloop() {
-	buf := bufio.NewReader(os.Stdin)
-
-	for {
-		r, _, err := buf.ReadRune()
-		if err != nil {
-			close(t.outchan)
-			break
-		}
-		t.outchan <- r
-	}
-}
-
-func (t *Terminal) Read() (rune, error) {
-	r, ok := <-t.outchan
-	if !ok {
-		return 0, io.EOF
-	}
-
-	return r, nil
-}

readline/term.go

@@ -1,36 +0,0 @@
-//go:build aix || darwin || dragonfly || freebsd || (linux && !appengine) || netbsd || openbsd || os400 || solaris
-
-package readline
-
-import (
-	"syscall"
-)
-
-type Termios syscall.Termios
-
-func SetRawMode(fd int) (*Termios, error) {
-	termios, err := getTermios(fd)
-	if err != nil {
-		return nil, err
-	}
-
-	newTermios := *termios
-	newTermios.Iflag &^= syscall.IGNBRK | syscall.BRKINT | syscall.PARMRK | syscall.ISTRIP | syscall.INLCR | syscall.IGNCR | syscall.ICRNL | syscall.IXON
-	newTermios.Lflag &^= syscall.ECHO | syscall.ECHONL | syscall.ICANON | syscall.ISIG | syscall.IEXTEN
-	newTermios.Cflag &^= syscall.CSIZE | syscall.PARENB
-	newTermios.Cflag |= syscall.CS8
-	newTermios.Cc[syscall.VMIN] = 1
-	newTermios.Cc[syscall.VTIME] = 0
-
-	return termios, setTermios(fd, &newTermios)
-}
-
-func UnsetRawMode(fd int, termios *Termios) error {
-	return setTermios(fd, termios)
-}
-
-// IsTerminal returns true if the given file descriptor is a terminal.
-func IsTerminal(fd int) bool {
-	_, err := getTermios(fd)
-	return err == nil
-}

readline/term_bsd.go

@@ -1,25 +0,0 @@
-//go:build darwin || freebsd || netbsd || openbsd
-
-package readline
-
-import (
-	"syscall"
-	"unsafe"
-)
-
-func getTermios(fd int) (*Termios, error) {
-	termios := new(Termios)
-	_, _, err := syscall.Syscall6(syscall.SYS_IOCTL, uintptr(fd), syscall.TIOCGETA, uintptr(unsafe.Pointer(termios)), 0, 0, 0)
-	if err != 0 {
-		return nil, err
-	}
-	return termios, nil
-}
-
-func setTermios(fd int, termios *Termios) error {
-	_, _, err := syscall.Syscall6(syscall.SYS_IOCTL, uintptr(fd), syscall.TIOCSETA, uintptr(unsafe.Pointer(termios)), 0, 0, 0)
-	if err != 0 {
-		return err
-	}
-	return nil
-}

readline/term_linux.go

@@ -1,28 +0,0 @@
-//go:build linux || solaris
-
-package readline
-
-import (
-	"syscall"
-	"unsafe"
-)
-
-const tcgets = 0x5401
-const tcsets = 0x5402
-
-func getTermios(fd int) (*Termios, error) {
-	termios := new(Termios)
-	_, _, err := syscall.Syscall6(syscall.SYS_IOCTL, uintptr(fd), tcgets, uintptr(unsafe.Pointer(termios)), 0, 0, 0)
-	if err != 0 {
-		return nil, err
-	}
-	return termios, nil
-}
-
-func setTermios(fd int, termios *Termios) error {
-	_, _, err := syscall.Syscall6(syscall.SYS_IOCTL, uintptr(fd), tcsets, uintptr(unsafe.Pointer(termios)), 0, 0, 0)
-	if err != 0 {
-		return err
-	}
-	return nil
-}

readline/term_windows.go

@@ -1,62 +0,0 @@
-package readline
-
-import (
-	"syscall"
-	"unsafe"
-)
-
-const (
-	enableLineInput       = 2
-	enableWindowInput     = 8
-	enableMouseInput      = 16
-	enableInsertMode      = 32
-	enableQuickEditMode   = 64
-	enableExtendedFlags   = 128
-	enableProcessedOutput = 1
-	enableWrapAtEolOutput = 2
-	enableAutoPosition    = 256 // Cursor position is not affected by writing data to the console.
-	enableEchoInput       = 4   // Characters are written to the console as they're read.
-	enableProcessedInput  = 1   // Enables input processing (like recognizing Ctrl+C).
-)
-
-var kernel32 = syscall.NewLazyDLL("kernel32.dll")
-
-var (
-	procGetConsoleMode = kernel32.NewProc("GetConsoleMode")
-	procSetConsoleMode = kernel32.NewProc("SetConsoleMode")
-)
-
-type State struct {
-	mode uint32
-}
-
-// IsTerminal checks if the given file descriptor is associated with a terminal
-func IsTerminal(fd int) bool {
-	var st uint32
-	r, _, e := syscall.SyscallN(procGetConsoleMode.Addr(), uintptr(fd), uintptr(unsafe.Pointer(&st)), 0)
-	// if the call succeeds and doesn't produce an error, it's a terminal
-	return r != 0 && e == 0
-}
-
-func SetRawMode(fd int) (*State, error) {
-	var st uint32
-	// retrieve the current mode of the terminal
-	_, _, e := syscall.SyscallN(procGetConsoleMode.Addr(), uintptr(fd), uintptr(unsafe.Pointer(&st)), 0)
-	if e != 0 {
-		return nil, error(e)
-	}
-	// modify the mode to set it to raw
-	raw := st &^ (enableEchoInput | enableProcessedInput | enableLineInput | enableProcessedOutput)
-	// apply the new mode to the terminal
-	_, _, e = syscall.SyscallN(procSetConsoleMode.Addr(), uintptr(fd), uintptr(raw), 0)
-	if e != 0 {
-		return nil, error(e)
-	}
-	// return the original state so that it can be restored later
-	return &State{st}, nil
-}
-
-func UnsetRawMode(fd int, state *State) error {
-	_, _, err := syscall.SyscallN(procSetConsoleMode.Addr(), uintptr(fd), uintptr(state.mode), 0)
-	return err
-}

readline/types.go

@@ -1,86 +0,0 @@
-package readline
-
-const (
-	CharNull      = 0
-	CharLineStart = 1
-	CharBackward  = 2
-	CharInterrupt = 3
-	CharDelete    = 4
-	CharLineEnd   = 5
-	CharForward   = 6
-	CharBell      = 7
-	CharCtrlH     = 8
-	CharTab       = 9
-	CharCtrlJ     = 10
-	CharKill      = 11
-	CharCtrlL     = 12
-	CharEnter     = 13
-	CharNext      = 14
-	CharPrev      = 16
-	CharBckSearch = 18
-	CharFwdSearch = 19
-	CharTranspose = 20
-	CharCtrlU     = 21
-	CharCtrlW     = 23
-	CharCtrlY     = 25
-	CharCtrlZ     = 26
-	CharEsc       = 27
-	CharSpace     = 32
-	CharEscapeEx  = 91
-	CharBackspace = 127
-)
-
-const (
-	KeyDel    = 51
-	KeyUp     = 65
-	KeyDown   = 66
-	KeyRight  = 67
-	KeyLeft   = 68
-	MetaEnd   = 70
-	MetaStart = 72
-)
-
-const (
-	CursorUp    = "\033[1A"
-	CursorDown  = "\033[1B"
-	CursorRight = "\033[1C"
-	CursorLeft  = "\033[1D"
-
-	CursorSave    = "\033[s"
-	CursorRestore = "\033[u"
-
-	CursorUpN    = "\033[%dA"
-	CursorDownN  = "\033[%dB"
-	CursorRightN = "\033[%dC"
-	CursorLeftN  = "\033[%dD"
-
-	CursorEOL  = "\033[E"
-	CursorBOL  = "\033[1G"
-	CursorHide = "\033[?25l"
-	CursorShow = "\033[?25h"
-
-	ClearToEOL  = "\033[K"
-	ClearLine   = "\033[2K"
-	ClearScreen = "\033[2J"
-	CursorReset = "\033[0;0f"
-
-	ColorGrey    = "\033[38;5;245m"
-	ColorDefault = "\033[0m"
-
-	StartBracketedPaste = "\033[?2004h"
-	EndBracketedPaste   = "\033[?2004l"
-)
-
-const (
-	CharBracketedPaste      = 50
-	CharBracketedPasteStart = "00~"
-	CharBracketedPasteEnd   = "01~"
-)
-
-type PasteMode int
-
-const (
-	PastModeOff = iota
-	PasteModeStart
-	PasteModeEnd
-)

runner/.gitignore

@@ -0,0 +1,2 @@
+model.bin
+runner

runner/darwin.go

@@ -0,0 +1,39 @@
+package main
+
+import (
+	"embed"
+	"io"
+	"os"
+	"path/filepath"
+)
+
+//go:embed ggml-metal.metal
+var fs embed.FS
+
+func init() {
+	exec, err := os.Executable()
+	if err != nil {
+		return
+	}
+
+	exec, err = filepath.EvalSymlinks(exec)
+	if err != nil {
+		return
+	}
+
+	dst, err := os.Create(filepath.Join(filepath.Dir(exec), "ggml-metal.metal"))
+	if err != nil {
+		return
+	}
+	defer dst.Close()
+
+	src, err := fs.Open("ggml-metal.metal")
+	if err != nil {
+		return
+	}
+	defer src.Close()
+
+	if _, err := io.Copy(dst, src); err != nil {
+		return
+	}
+}

runner/ggml-alloc.c

@@ -0,0 +1,620 @@
+/**
+ * llama.cpp - git 465219b9143ac01db0990bbcb0a081ef72ec2008
+ *
+ * MIT License
+ *
+ * Copyright (c) 2023 Georgi Gerganov
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to deal
+ * in the Software without restriction, including without limitation the rights
+ * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+ * copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in all
+ * copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#include "ggml-alloc.h"
+#include "ggml-backend.h"
+#include "ggml.h"
+#include <assert.h>
+#include <stdarg.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+
+
+#define UNUSED(x) (void)(x)
+#define MAX(a, b) ((a) > (b) ? (a) : (b))
+#define GGML_MAX_CONCUR (2*GGML_MAX_NODES)
+
+//#define GGML_ALLOCATOR_DEBUG
+
+//#define AT_PRINTF printf
+#define AT_PRINTF(...) ((void)0)
+
+struct hash_node {
+    struct ggml_tensor * t;
+    int n_children;
+    int n_views;
+};
+
+static size_t hash(void * p) {
+    return (size_t)p % GGML_GRAPH_HASHTABLE_SIZE;
+}
+
+static struct hash_node * hash_get(struct hash_node hash_table[], struct ggml_tensor * t) {
+    size_t h = hash(t);
+
+    // linear probing
+    size_t i = h;
+    while (hash_table[i].t != NULL) {
+        if (hash_table[i].t == t) {
+            return &hash_table[i];
+        }
+        i = (i + 1) % GGML_GRAPH_HASHTABLE_SIZE;
+        if (i == h) {
+            // hash table is full
+            GGML_ASSERT(false);
+        }
+    }
+
+    hash_table[i].t = t;
+    return &hash_table[i];
+}
+
+// TODO: GGML_PAD ?
+static size_t aligned_offset(const void * buffer, size_t offset, size_t alignment) {
+    assert(alignment && !(alignment & (alignment - 1))); // power of 2
+    size_t align = (alignment - (((uintptr_t)buffer + offset) % alignment)) % alignment;
+    return offset + align;
+}
+
+struct free_block {
+    void * addr;
+    size_t size;
+};
+
+#define MAX_FREE_BLOCKS 256
+
+struct ggml_allocr {
+    struct ggml_backend_buffer * buffer;
+    bool buffer_owned;
+    void * data;
+    size_t alignment;
+    int n_free_blocks;
+    struct free_block free_blocks[MAX_FREE_BLOCKS];
+    struct hash_node hash_table[GGML_GRAPH_HASHTABLE_SIZE];
+    size_t max_size;
+    bool measure;
+    int parse_seq[GGML_MAX_CONCUR];
+    int parse_seq_len;
+
+#ifdef GGML_ALLOCATOR_DEBUG
+    struct ggml_tensor * allocated_tensors[1024];
+#endif
+};
+
+#ifdef GGML_ALLOCATOR_DEBUG
+static void add_allocated_tensor(struct ggml_allocr * alloc, struct ggml_tensor * tensor) {
+    for (int i = 0; i < 1024; i++) {
+        if (alloc->allocated_tensors[i] == NULL) {
+            alloc->allocated_tensors[i] = tensor;
+            return;
+        }
+    }
+    GGML_ASSERT(!"out of allocated_tensors");
+}
+static void remove_allocated_tensor(struct ggml_allocr * alloc, struct ggml_tensor * tensor) {
+    for (int i = 0; i < 1024; i++) {
+        if (alloc->allocated_tensors[i] == tensor ||
+            (alloc->allocated_tensors[i] != NULL && alloc->allocated_tensors[i]->data == tensor->data)) {
+            alloc->allocated_tensors[i] = NULL;
+            return;
+        }
+    }
+    printf("tried to free tensor %s not found\n", tensor->name);
+    GGML_ASSERT(!"tensor not found");
+}
+#endif
+
+// check if a tensor is allocated by this buffer
+static bool ggml_allocr_is_own(struct ggml_allocr * alloc, const struct ggml_tensor * tensor) {
+    return tensor->buffer == alloc->buffer;
+}
+
+static bool ggml_is_view(struct ggml_tensor * t) {
+    return t->view_src != NULL;
+}
+
+void ggml_allocr_alloc(struct ggml_allocr * alloc, struct ggml_tensor * tensor) {
+    GGML_ASSERT(!ggml_is_view(tensor)); // views generally get data pointer from one of their sources
+    GGML_ASSERT(tensor->data == NULL); // avoid allocating tensor which already has memory allocated
+
+    size_t size = ggml_backend_buffer_get_alloc_size(alloc->buffer, tensor);
+    size = aligned_offset(NULL, size, alloc->alignment);
+
+    AT_PRINTF("%s: allocating %s (%zu bytes) - ", __func__, tensor->name, size);
+
+    size_t max_avail = 0;
+
+    // find the best fitting free block besides the last block
+    int best_fit_block = -1;
+    size_t best_fit_size = SIZE_MAX;
+    for (int i = 0; i < alloc->n_free_blocks - 1; i++) {
+        struct free_block * block = &alloc->free_blocks[i];
+        max_avail = MAX(max_avail, block->size);
+        if (block->size >= size && block->size <= best_fit_size) {
+            best_fit_block = i;
+            best_fit_size = block->size;
+        }
+    }
+
+    AT_PRINTF("block %d\n", best_fit_block);
+
+    if (best_fit_block == -1) {
+        // the last block is our last resort
+        struct free_block * block = &alloc->free_blocks[alloc->n_free_blocks - 1];
+        max_avail = MAX(max_avail, block->size);
+        if (block->size >= size) {
+            best_fit_block = alloc->n_free_blocks - 1;
+        } else {
+            fprintf(stderr, "%s: not enough space in the buffer (needed %zu, largest block available %zu)\n",
+                    __func__, size, max_avail);
+            GGML_ASSERT(!"not enough space in the buffer");
+            return;
+        }
+    }
+    struct free_block * block = &alloc->free_blocks[best_fit_block];
+    void * addr = block->addr;
+    block->addr = (char*)block->addr + size;
+    block->size -= size;
+    if (block->size == 0) {
+        // remove block if empty
+        alloc->n_free_blocks--;
+        for (int j = best_fit_block; j < alloc->n_free_blocks; j++) {
+            alloc->free_blocks[j] = alloc->free_blocks[j+1];
+        }
+    }
+
+    tensor->data = addr;
+    AT_PRINTF("%s: allocated data at %p\n", __func__, tensor->data);
+    tensor->buffer = alloc->buffer;
+    ggml_backend_buffer_init_tensor(alloc->buffer, tensor);
+
+#ifdef GGML_ALLOCATOR_DEBUG
+    add_allocated_tensor(alloc, tensor);
+    size_t cur_max = (char*)addr - (char*)alloc->data + size;
+    if (cur_max > alloc->max_size) {
+        printf("max_size = %.2f MB: tensors: ", cur_max / 1024.0 / 1024.0);
+        for (int i = 0; i < 1024; i++) {
+            if (alloc->allocated_tensors[i]) {
+                printf("%s (%.2f MB) ", alloc->allocated_tensors[i]->name, ggml_nbytes(alloc->allocated_tensors[i]) / 1024.0 / 1024.0);
+            }
+        }
+        printf("\n");
+    }
+#endif
+
+    alloc->max_size = MAX(alloc->max_size, (char*)addr - (char*)alloc->data + size);
+}
+
+// this is a very naive implementation, but for our case the number of free blocks should be very small
+static void ggml_allocr_free_tensor(struct ggml_allocr * alloc, struct ggml_tensor * tensor) {
+    if (ggml_allocr_is_own(alloc, tensor) == false) {
+        // the tensor was not allocated in this buffer
+        // this can happen because the graph allocator will try to free weights and other tensors from different buffers
+        // the easiest way to deal with this is just to ignore it
+        AT_PRINTF("ignoring %s (their buffer: %p, our buffer: %p)\n", tensor->name, (void *)tensor->buffer, (void *)alloc->buffer);
+        return;
+    }
+
+    void * ptr = tensor->data;
+
+    size_t size = ggml_backend_buffer_get_alloc_size(alloc->buffer, tensor);
+    size = aligned_offset(NULL, size, alloc->alignment);
+    AT_PRINTF("%s: freeing %s at %p (%zu bytes) - n_free_blocks = %d\n", __func__, tensor->name, ptr, size, alloc->n_free_blocks);
+
+    ggml_backend_buffer_free_tensor(alloc->buffer, tensor);
+
+#ifdef GGML_ALLOCATOR_DEBUG
+    remove_allocated_tensor(alloc, tensor);
+#endif
+
+    // see if we can merge with an existing block
+    for (int i = 0; i < alloc->n_free_blocks; i++) {
+        struct free_block * block = &alloc->free_blocks[i];
+        // check if ptr is at the end of the block
+        if ((char*)block->addr + block->size == ptr) {
+            block->size += size;
+            // check if we can merge with the next block
+            if (i < alloc->n_free_blocks - 1 && (char*)block->addr + block->size == alloc->free_blocks[i+1].addr) {
+                block->size += alloc->free_blocks[i+1].size;
+                alloc->n_free_blocks--;
+                for (int j = i+1; j < alloc->n_free_blocks; j++) {
+                    alloc->free_blocks[j] = alloc->free_blocks[j+1];
+                }
+            }
+            return;
+        }
+        // check if ptr is at the beginning of the block
+        if ((char*)ptr + size == block->addr) {
+            block->addr = ptr;
+            block->size += size;
+            // check if we can merge with the previous block
+            if (i > 0 && (char*)alloc->free_blocks[i-1].addr + alloc->free_blocks[i-1].size == block->addr) {
+                alloc->free_blocks[i-1].size += block->size;
+                alloc->n_free_blocks--;
+                for (int j = i; j < alloc->n_free_blocks; j++) {
+                    alloc->free_blocks[j] = alloc->free_blocks[j+1];
+                }
+            }
+            return;
+        }
+    }
+    // otherwise, add a new block
+    GGML_ASSERT(alloc->n_free_blocks < MAX_FREE_BLOCKS && "out of free blocks");
+    // insert the new block in the correct position to keep the array sorted by address (to make merging blocks faster)
+    int insert_pos = 0;
+    while (insert_pos < alloc->n_free_blocks && alloc->free_blocks[insert_pos].addr < ptr) {
+        insert_pos++;
+    }
+    // shift all blocks from insert_pos onward to make room for the new block
+    for (int i = alloc->n_free_blocks; i > insert_pos; i--) {
+        alloc->free_blocks[i] = alloc->free_blocks[i-1];
+    }
+    // insert the new block
+    alloc->free_blocks[insert_pos].addr = ptr;
+    alloc->free_blocks[insert_pos].size = size;
+    alloc->n_free_blocks++;
+}
+
+void ggml_allocr_set_parse_seq(struct ggml_allocr * alloc, const int * list, int n) {
+    for (int i = 0; i < n; i++) {
+        alloc->parse_seq[i] = list[i];
+    }
+    alloc->parse_seq_len = n;
+}
+
+void ggml_allocr_reset(struct ggml_allocr * alloc) {
+    alloc->n_free_blocks = 1;
+    size_t align_offset = aligned_offset(alloc->data, 0, alloc->alignment);
+    alloc->free_blocks[0].addr = (char *)alloc->data + align_offset;
+    alloc->free_blocks[0].size = ggml_backend_buffer_get_size(alloc->buffer) - align_offset;
+}
+
+struct ggml_allocr * ggml_allocr_new(void * data, size_t size, size_t alignment) {
+    struct ggml_backend_buffer * buffer = ggml_backend_cpu_buffer_from_ptr(NULL, data, size);
+
+    struct ggml_allocr * alloc = (struct ggml_allocr *)malloc(sizeof(struct ggml_allocr));
+
+    *alloc = (struct ggml_allocr){
+        /*.buffer        = */ buffer,
+        /*.buffer_owned  = */ true,
+        /*.base          = */ ggml_backend_buffer_get_base(buffer),
+        /*.alignment     = */ alignment,
+        /*.n_free_blocks = */ 0,
+        /*.free_blocks   = */ {{0}},
+        /*.hash_table    = */ {{0}},
+        /*.max_size      = */ 0,
+        /*.measure       = */ false,
+        /*.parse_seq     = */ {0},
+        /*.parse_seq_len = */ 0,
+#ifdef GGML_ALLOCATOR_DEBUG
+        /*.allocated_tensors = */ {0},
+#endif
+    };
+
+    ggml_allocr_reset(alloc);
+
+    return alloc;
+}
+
+struct ggml_allocr * ggml_allocr_new_measure(size_t alignment) {
+    struct ggml_allocr * alloc = ggml_allocr_new((void *)0x1000, (size_t)-0x1001, alignment);
+    alloc->measure = true;
+
+    return alloc;
+}
+
+struct ggml_allocr * ggml_allocr_new_from_buffer(struct ggml_backend_buffer * buffer) {
+    struct ggml_allocr * alloc = (struct ggml_allocr *)malloc(sizeof(struct ggml_allocr));
+
+    *alloc = (struct ggml_allocr){
+        /*.buffer        = */ buffer,
+        /*.buffer_owned  = */ false,
+        /*.base          = */ ggml_backend_buffer_get_base(buffer),
+        /*.alignment     = */ ggml_backend_buffer_get_alignment(buffer),
+        /*.n_free_blocks = */ 0,
+        /*.free_blocks   = */ {{0}},
+        /*.hash_table    = */ {{0}},
+        /*.max_size      = */ 0,
+        /*.measure       = */ false,
+        /*.parse_seq     = */ {0},
+        /*.parse_seq_len = */ 0,
+#ifdef GGML_ALLOCATOR_DEBUG
+        /*.allocated_tensors = */ {0},
+#endif
+    };
+
+    ggml_allocr_reset(alloc);
+
+    return alloc;
+}
+
+void ggml_allocr_free(struct ggml_allocr * alloc) {
+    if (alloc->buffer_owned) {
+        ggml_backend_buffer_free(alloc->buffer);
+    }
+    free(alloc);
+}
+
+bool ggml_allocr_is_measure(struct ggml_allocr * alloc) {
+    return alloc->measure;
+}
+
+//////////// compute graph allocator
+
+static bool ggml_are_same_layout(const struct ggml_tensor * a, const struct ggml_tensor * b) {
+    if (a->type != b->type) {
+        return false;
+    }
+    for (int i = 0; i < GGML_MAX_DIMS; i++) {
+        if (a->ne[i] != b->ne[i]) {
+            return false;
+        }
+        if (a->nb[i] != b->nb[i]) {
+            return false;
+        }
+    }
+    return true;
+}
+
+static bool ggml_op_can_inplace(enum ggml_op op) {
+    switch (op) {
+        case GGML_OP_SCALE:
+        case GGML_OP_DIAG_MASK_ZERO:
+        case GGML_OP_DIAG_MASK_INF:
+        case GGML_OP_ADD:
+        case GGML_OP_ADD1:
+        case GGML_OP_SUB:
+        case GGML_OP_MUL:
+        case GGML_OP_DIV:
+        case GGML_OP_SQR:
+        case GGML_OP_SQRT:
+        case GGML_OP_LOG:
+        case GGML_OP_UNARY:
+        case GGML_OP_ROPE:
+        case GGML_OP_RMS_NORM:
+        case GGML_OP_SOFT_MAX:
+            return true;
+
+        default:
+            return false;
+    }
+}
+
+static void init_view(struct ggml_allocr * alloc, struct ggml_tensor * view) {
+    assert(view->view_src != NULL && view->view_src->data != NULL);
+    view->backend = view->view_src->backend;
+    view->buffer  = view->view_src->buffer;
+    view->data    = (char *)view->view_src->data + view->view_offs;
+
+    // FIXME: the view should be initialized by the owning buffer, but currently this breaks the CUDA backend
+    // due to the ggml_tensor_extra_gpu ring buffer overwriting the KV cache extras
+    assert(ggml_allocr_is_measure(alloc) || !view->buffer || view->buffer->backend == alloc->buffer->backend);
+    ggml_backend_buffer_init_tensor(alloc->buffer, view);
+}
+
+static void allocate_node(struct ggml_allocr * alloc, struct ggml_tensor * node) {
+    struct hash_node * ht = alloc->hash_table;
+    if (node->data == NULL) {
+        if (ggml_is_view(node)) {
+            init_view(alloc, node);
+        } else {
+            // see if we can reuse a parent's buffer (inplace)
+            if (ggml_op_can_inplace(node->op)) {
+                for (int i = 0; i < GGML_MAX_SRC; i++) {
+                    struct ggml_tensor * parent = node->src[i];
+                    if (parent == NULL) {
+                        break;
+                    }
+
+                    // if the node's data is external, then we cannot re-use it
+                    if (ggml_allocr_is_own(alloc, parent) == false) {
+                        AT_PRINTF("not reusing parent %s for %s as %p is external\n", parent->name, node->name, parent->data);
+                        continue;
+                    }
+
+                    struct hash_node * p_hn = hash_get(ht, parent);
+                    if (parent->data != NULL && p_hn->n_children == 1 && p_hn->n_views == 0 && ggml_are_same_layout(node, parent)) {
+                        if (ggml_is_view(parent)) {
+                            struct ggml_tensor * view_src = parent->view_src;
+                            struct hash_node * view_src_hn = hash_get(ht, view_src);
+                            if (view_src_hn->n_views == 1 && view_src_hn->n_children == 0 && view_src->data == parent->data) {
+                                // TODO: the offset of the view parent must be kept to ensure that the op doesn't overwrite
+                                // the parent's data that it will need later (same layout requirement). the problem is that then
+                                // we cannot free the tensor because the original address of the allocation is lost.
+                                // adding a view_src pointer to the tensor would solve this and simplify the code dealing with views
+                                // for now, we only reuse the parent's data if the offset is zero (view_src->data == parent->data)
+                                AT_PRINTF("reusing view parent %s (%s) for %s\n", parent->name, view_src->name, node->name);
+                                node->view_src = view_src;
+                                view_src_hn->n_views += 1;
+                                init_view(alloc, node);
+                                return;
+                            }
+                        }
+                        else {
+                            AT_PRINTF("reusing parent %s for %s\n", parent->name, node->name);
+                            node->view_src = parent;
+                            p_hn->n_views += 1;
+                            init_view(alloc, node);
+                            return;
+                        }
+                    }
+                }
+            }
+            ggml_allocr_alloc(alloc, node);
+        }
+    }
+}
+
+size_t ggml_allocr_alloc_graph_n(
+    struct ggml_allocr * alloc,
+    struct ggml_cgraph ** graphs, int n_graphs,
+    struct ggml_tensor *** inputs, struct ggml_tensor *** outputs) {
+
+    // reset hash table
+    struct hash_node * ht = alloc->hash_table;
+    memset(ht, 0, sizeof(struct hash_node) * GGML_GRAPH_HASHTABLE_SIZE);
+
+    // count number of children and views
+    for (int g = 0; g < n_graphs; g++) {
+        struct ggml_cgraph * gf = graphs[g];
+        for (int i = 0; i < gf->n_nodes; i++) {
+            struct ggml_tensor * node = gf->nodes[i];
+
+            if (ggml_is_view(node)) {
+                struct ggml_tensor * view_src = node->view_src;
+                hash_get(ht, view_src)->n_views += 1;
+                if (node->buffer == NULL && node->data != NULL) {
+                    // view of a pre-allocated tensor, didn't call init_view() yet
+                    init_view(alloc, node);
+                }
+            }
+
+            for (int j = 0; j < GGML_MAX_SRC; j++) {
+                struct ggml_tensor * parent = node->src[j];
+                if (parent == NULL) {
+                    break;
+                }
+                hash_get(ht, parent)->n_children += 1;
+                if (ggml_is_view(parent) && parent->buffer == NULL && parent->data != NULL) {
+                    init_view(alloc, parent);
+                }
+            }
+        }
+    }
+
+    // allocate tensors
+    for (int g = 0; g < n_graphs; g++) {
+        struct ggml_cgraph * gf = graphs[g];
+        AT_PRINTF("####### graph %d/%d\n", g, n_graphs);
+        // graph inputs are allocated first to ensure that they are not overwritten by each other
+        if (inputs != NULL && inputs[g] != NULL) {
+            for (int i = 0; inputs[g][i] != NULL; i++) {
+                struct ggml_tensor * input = inputs[g][i];
+                AT_PRINTF("input: %s\n", input->name);
+                allocate_node(alloc, input);
+            }
+        }
+        // if we have parse_seq then we allocate nodes following the list, and we only free nodes at barriers
+        int last_barrier_pos = 0;
+        int n_nodes = alloc->parse_seq_len ? alloc->parse_seq_len : gf->n_nodes;
+
+        for (int ind = 0; ind < n_nodes; ind++) {
+            // allocate a node if there is no parse_seq or this is not a barrier
+            if ((alloc->parse_seq_len==0) || alloc->parse_seq[ind] != -1) {
+                int i = alloc->parse_seq_len ? alloc->parse_seq[ind] : ind;
+                struct ggml_tensor * node = gf->nodes[i];
+
+                // allocate parents (leafs)
+                for (int j = 0; j < GGML_MAX_SRC; j++) {
+                    struct ggml_tensor * parent = node->src[j];
+                    if (parent == NULL) {
+                        break;
+                    }
+                    allocate_node(alloc, parent);
+                }
+
+                // allocate node
+                allocate_node(alloc, node);
+
+                AT_PRINTF("exec: %s (%s) <= ", ggml_op_name(node->op), node->name);
+                for (int j = 0; j < GGML_MAX_SRC; j++) {
+                    struct ggml_tensor * parent = node->src[j];
+                    if (parent == NULL) {
+                        break;
+                    }
+                    AT_PRINTF("%s", parent->name);
+                    if (j < GGML_MAX_SRC - 1 && node->src[j + 1] != NULL) {
+                        AT_PRINTF(", ");
+                    }
+                }
+                AT_PRINTF("\n");
+            }
+
+            // update parents
+            // update immediately if there is no parse_seq
+            // update only at barriers if there is parse_seq
+            if ((alloc->parse_seq_len == 0) || alloc->parse_seq[ind] == -1) {
+                int update_start = alloc->parse_seq_len ? last_barrier_pos : ind;
+                int update_end   = alloc->parse_seq_len ? ind              : ind + 1;
+                for (int i = update_start; i < update_end; i++) {
+                    int node_i = alloc->parse_seq_len ? alloc->parse_seq[i] : i;
+                    struct ggml_tensor * node = gf->nodes[node_i];
+
+                    for (int j = 0; j < GGML_MAX_SRC; j++) {
+                        struct ggml_tensor * parent = node->src[j];
+                        if (parent == NULL) {
+                            break;
+                        }
+                        struct hash_node * p_hn = hash_get(ht, parent);
+                        p_hn->n_children -= 1;
+
+                        //AT_PRINTF("parent %s: %d children, %d views\n", parent->name, parent->n_children, parent->n_views);
+
+                        if (p_hn->n_children == 0 && p_hn->n_views == 0) {
+                            if (ggml_is_view(parent)) {
+                                struct ggml_tensor * view_src = parent->view_src;
+                                struct hash_node * view_src_hn = hash_get(ht, view_src);
+                                view_src_hn->n_views -= 1;
+                                AT_PRINTF("view_src %s: %d children, %d views\n", view_src->name, view_src_hn->n_children, view_src_hn->n_views);
+                                if (view_src_hn->n_views == 0 && view_src_hn->n_children == 0 && view_src->data != node->data) {
+                                    ggml_allocr_free_tensor(alloc, view_src);
+                                }
+                            }
+                            else {
+                                if (parent->data != node->data) {
+                                    ggml_allocr_free_tensor(alloc, parent);
+                                }
+                            }
+                        }
+                    }
+                }
+                AT_PRINTF("\n");
+                if (alloc->parse_seq_len) {
+                    last_barrier_pos = ind + 1;
+                }
+            }
+        }
+        // free graph outputs here that wouldn't be freed otherwise because they have no children
+        if (outputs != NULL && outputs[g] != NULL) {
+            for (int i = 0; outputs[g][i] != NULL; i++) {
+                struct ggml_tensor * output = outputs[g][i];
+                AT_PRINTF("output: %s\n", output->name);
+                ggml_allocr_free_tensor(alloc, output);
+            }
+        }
+    }
+
+    return alloc->max_size;
+}
+
+size_t ggml_allocr_alloc_graph(struct ggml_allocr * alloc, struct ggml_cgraph * graph) {
+    return ggml_allocr_alloc_graph_n(alloc, &graph, 1, NULL, NULL);
+}
+
+size_t ggml_allocr_max_size(struct ggml_allocr * alloc) {
+    return alloc->max_size;
+}

runner/ggml-alloc.h

@@ -0,0 +1,59 @@
+/**
+ * llama.cpp - git 465219b9143ac01db0990bbcb0a081ef72ec2008
+ *
+ * MIT License
+ *
+ * Copyright (c) 2023 Georgi Gerganov
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to deal
+ * in the Software without restriction, including without limitation the rights
+ * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+ * copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in all
+ * copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#pragma once
+
+#include "ggml.h"
+
+#ifdef  __cplusplus
+extern "C" {
+#endif
+
+struct ggml_backend_buffer;
+
+GGML_API struct ggml_allocr * ggml_allocr_new(void * data, size_t size, size_t alignment);
+GGML_API struct ggml_allocr * ggml_allocr_new_measure(size_t alignment);
+GGML_API struct ggml_allocr * ggml_allocr_new_from_buffer(struct ggml_backend_buffer * buffer);
+
+// tell the allocator to parse nodes following the order described in the list
+// you should call this if your graph are optimized to execute out-of-order
+GGML_API void   ggml_allocr_set_parse_seq(struct ggml_allocr * alloc, const int * list, int n);
+
+GGML_API void   ggml_allocr_free       (struct ggml_allocr * alloc);
+GGML_API bool   ggml_allocr_is_measure (struct ggml_allocr * alloc);
+GGML_API void   ggml_allocr_reset      (struct ggml_allocr * alloc);
+GGML_API void   ggml_allocr_alloc      (struct ggml_allocr * alloc, struct ggml_tensor * tensor);
+GGML_API size_t ggml_allocr_alloc_graph(struct ggml_allocr * alloc, struct ggml_cgraph * graph);
+GGML_API size_t ggml_allocr_max_size   (struct ggml_allocr * alloc);
+
+GGML_API size_t ggml_allocr_alloc_graph_n(
+                    struct ggml_allocr * alloc,
+                    struct ggml_cgraph ** graphs, int n_graphs,
+                    struct ggml_tensor *** inputs, struct ggml_tensor *** outputs);
+
+#ifdef  __cplusplus
+}
+#endif

runner/ggml-backend.c

@@ -0,0 +1,411 @@
+/**
+ * llama.cpp - git 465219b9143ac01db0990bbcb0a081ef72ec2008
+ *
+ * MIT License
+ *
+ * Copyright (c) 2023 Georgi Gerganov
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to deal
+ * in the Software without restriction, including without limitation the rights
+ * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+ * copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in all
+ * copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#include "ggml-backend.h"
+#include "ggml-alloc.h"
+
+#include <assert.h>
+#include <stdarg.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+
+#define UNUSED GGML_UNUSED
+
+#define MAX(a, b) ((a) > (b) ? (a) : (b))
+
+// backend buffer
+
+ggml_backend_buffer_t ggml_backend_buffer_init(
+        struct ggml_backend                  * backend,
+        struct ggml_backend_buffer_i           iface,
+               ggml_backend_buffer_context_t   context,
+               size_t                          size) {
+    ggml_backend_buffer_t buffer = malloc(sizeof(struct ggml_backend_buffer));
+
+    GGML_ASSERT(iface.get_base != NULL);
+
+    (*buffer) = (struct ggml_backend_buffer) {
+        /* .interface = */ iface,
+        /* .backend   = */ backend,
+        /* .context   = */ context,
+        /* .size      = */ size,
+    };
+
+    return buffer;
+}
+
+void ggml_backend_buffer_free(ggml_backend_buffer_t buffer) {
+    if (buffer->iface.free_buffer != NULL) {
+        buffer->iface.free_buffer(buffer);
+    }
+    free(buffer);
+}
+
+size_t ggml_backend_buffer_get_alignment(ggml_backend_buffer_t buffer) {
+    return ggml_backend_get_alignment(buffer->backend);
+}
+
+void * ggml_backend_buffer_get_base(ggml_backend_buffer_t buffer) {
+    return buffer->iface.get_base(buffer);
+}
+
+size_t ggml_backend_buffer_get_size(ggml_backend_buffer_t buffer) {
+    return buffer->size;
+}
+
+size_t ggml_backend_buffer_get_alloc_size(ggml_backend_buffer_t buffer, struct ggml_tensor * tensor) {
+    if (buffer->iface.get_alloc_size) {
+        return buffer->iface.get_alloc_size(buffer, tensor);
+    }
+    return ggml_nbytes(tensor);
+}
+
+void ggml_backend_buffer_init_tensor(ggml_backend_buffer_t buffer, struct ggml_tensor * tensor) {
+    if (buffer->iface.init_tensor) {
+        buffer->iface.init_tensor(buffer, tensor);
+    }
+}
+
+void ggml_backend_buffer_free_tensor(ggml_backend_buffer_t buffer, struct ggml_tensor * tensor) {
+    if (buffer->iface.free_tensor) {
+        buffer->iface.free_tensor(buffer, tensor);
+    }
+}
+
+// backend
+
+ggml_backend_t ggml_get_backend(const struct ggml_tensor * tensor) {
+    return tensor->buffer->backend;
+}
+
+const char * ggml_backend_name(ggml_backend_t backend) {
+    return backend->iface.get_name(backend);
+}
+
+void ggml_backend_free(ggml_backend_t backend) {
+    backend->iface.free(backend);
+}
+
+ggml_backend_buffer_t ggml_backend_alloc_buffer(ggml_backend_t backend, size_t size) {
+    return backend->iface.alloc_buffer(backend, size);
+}
+
+size_t ggml_backend_get_alignment(ggml_backend_t backend) {
+    return backend->iface.get_alignment(backend);
+}
+
+void ggml_backend_tensor_set_async(struct ggml_tensor * tensor, const void * data, size_t offset, size_t size) {
+    ggml_get_backend(tensor)->iface.set_tensor_async(ggml_get_backend(tensor), tensor, data, offset, size);
+}
+
+void ggml_backend_tensor_get_async(const struct ggml_tensor * tensor, void * data, size_t offset, size_t size) {
+    ggml_get_backend(tensor)->iface.get_tensor_async(ggml_get_backend(tensor), tensor, data, offset, size);
+}
+
+void ggml_backend_tensor_set(struct ggml_tensor * tensor, const void * data, size_t offset, size_t size) {
+    ggml_get_backend(tensor)->iface.set_tensor_async(ggml_get_backend(tensor), tensor, data, offset, size);
+    ggml_get_backend(tensor)->iface.synchronize(ggml_get_backend(tensor));
+}
+
+void ggml_backend_tensor_get(const struct ggml_tensor * tensor, void * data, size_t offset, size_t size) {
+    ggml_get_backend(tensor)->iface.get_tensor_async(ggml_get_backend(tensor), tensor, data, offset, size);
+    ggml_get_backend(tensor)->iface.synchronize(ggml_get_backend(tensor));
+}
+
+void ggml_backend_synchronize(ggml_backend_t backend) {
+    backend->iface.synchronize(backend);
+}
+
+ggml_backend_graph_plan_t ggml_backend_graph_plan_create(ggml_backend_t backend, struct ggml_cgraph * cgraph) {
+    return backend->iface.graph_plan_create(backend, cgraph);
+}
+
+void ggml_backend_graph_plan_free(ggml_backend_t backend, ggml_backend_graph_plan_t plan) {
+    backend->iface.graph_plan_free(backend, plan);
+}
+
+void ggml_backend_graph_plan_compute(ggml_backend_t backend, ggml_backend_graph_plan_t plan) {
+    backend->iface.graph_plan_compute(backend, plan);
+}
+
+void ggml_backend_graph_compute(ggml_backend_t backend, struct ggml_cgraph * cgraph) {
+    backend->iface.graph_compute(backend, cgraph);
+}
+
+bool ggml_backend_supports_op(ggml_backend_t backend, const struct ggml_tensor * op) {
+    return backend->iface.supports_op(backend, op);
+}
+
+// backend copy
+
+static bool ggml_are_same_layout(const struct ggml_tensor * a, const struct ggml_tensor * b) {
+    if (a->type != b->type) {
+        return false;
+    }
+    for (int i = 0; i < GGML_MAX_DIMS; i++) {
+        if (a->ne[i] != b->ne[i]) {
+            return false;
+        }
+        if (a->nb[i] != b->nb[i]) {
+            return false;
+        }
+    }
+    return true;
+}
+
+void ggml_backend_tensor_copy(struct ggml_tensor * src, struct ggml_tensor * dst) {
+    //printf("src: %s ne: [%d %d %d %d] nb: [%d %d %d %d]\n", src->name, (int)src->ne[0], (int)src->ne[1], (int)src->ne[2], (int)src->ne[3], (int)src->nb[0], (int)src->nb[1], (int)src->nb[2], (int)src->nb[3]);
+    //printf("dst: %s ne: [%d %d %d %d] nb: [%d %d %d %d]\n", dst->name, (int)dst->ne[0], (int)dst->ne[1], (int)dst->ne[2], (int)dst->ne[3], (int)dst->nb[0], (int)dst->nb[1], (int)dst->nb[2], (int)dst->nb[3]);
+    GGML_ASSERT(ggml_are_same_layout(src, dst) && "cannot copy tensors with different layouts");
+
+    // printf("cpy tensor %s from %s to %s (%lu bytes)\n", src->name, ggml_backend_name(src->backend), ggml_backend_name(dst->backend), ggml_nbytes(src));
+
+    if (src == dst) {
+        return;
+    }
+
+    // TODO: allow backends to support copy to/from same backend
+
+    if (ggml_get_backend(dst)->iface.cpy_tensor_from != NULL) {
+        ggml_get_backend(dst)->iface.cpy_tensor_from(ggml_get_backend(dst)->context, src, dst);
+    } else if (ggml_get_backend(src)->iface.cpy_tensor_to != NULL) {
+        ggml_get_backend(src)->iface.cpy_tensor_to(ggml_get_backend(src)->context, src, dst);
+    } else {
+        // shouldn't be hit when copying from/to CPU
+        #ifndef NDEBUG
+        fprintf(stderr, "ggml_backend_tensor_copy: neither cpy_tensor_from nor cpy_tensor_to are implemented for backends %s and %s, falling back to get/set\n", ggml_backend_name(src->buffer->backend), ggml_backend_name(dst->buffer->backend));
+        #endif
+        size_t nbytes = ggml_nbytes(src);
+        void * data = malloc(nbytes);
+        ggml_backend_tensor_get(src, data, 0, nbytes);
+        ggml_backend_tensor_set(dst, data, 0, nbytes);
+        free(data);
+    }
+}
+
+// backend CPU
+
+struct ggml_backend_cpu_context {
+    int n_threads;
+    void * work_data;
+    size_t work_size;
+};
+
+static const char * ggml_backend_cpu_name(ggml_backend_t backend) {
+    return "CPU";
+
+    UNUSED(backend);
+}
+
+static void ggml_backend_cpu_free(ggml_backend_t backend) {
+    struct ggml_backend_cpu_context * cpu_ctx = (struct ggml_backend_cpu_context *)backend->context;
+    free(cpu_ctx->work_data);
+    free(cpu_ctx);
+    free(backend);
+}
+
+static void * ggml_backend_cpu_buffer_get_base(ggml_backend_buffer_t buffer) {
+    return (void *)buffer->context;
+}
+
+static void ggml_backend_cpu_buffer_free_buffer(ggml_backend_buffer_t buffer) {
+    free(buffer->context);
+    UNUSED(buffer);
+}
+
+static struct ggml_backend_buffer_i cpu_backend_buffer_i = {
+    /* .free_buffer    = */ ggml_backend_cpu_buffer_free_buffer,
+    /* .get_base       = */ ggml_backend_cpu_buffer_get_base,
+    /* .get_alloc_size = */ NULL, // defaults to ggml_nbytes
+    /* .init_tensor    = */ NULL, // no initialization required
+    /* .free_tensor    = */ NULL, // no cleanup required
+};
+
+// for buffers from ptr, free is not called
+static struct ggml_backend_buffer_i cpu_backend_buffer_i_from_ptr = {
+    /* .free_buffer    = */ NULL, // ptr is not owned by the buffer, so it does not need to be freed
+    /* .get_base       = */ ggml_backend_cpu_buffer_get_base,
+    /* .get_alloc_size = */ NULL, // defaults to ggml_nbytes
+    /* .init_tensor    = */ NULL,
+    /* .free_tensor    = */ NULL,
+};
+
+static const size_t TENSOR_ALIGNMENT = 64; // should be enough for AVX 512
+
+static ggml_backend_buffer_t ggml_backend_cpu_alloc_buffer(ggml_backend_t backend, size_t size) {
+    size += TENSOR_ALIGNMENT;   // malloc may return an address that is not aligned
+    void * data = malloc(size); // TODO: maybe use GGML_ALIGNED_MALLOC?
+
+    return ggml_backend_buffer_init(backend, cpu_backend_buffer_i, data, size);
+}
+
+static size_t ggml_backend_cpu_get_alignment(ggml_backend_t backend) {
+    return TENSOR_ALIGNMENT;
+    UNUSED(backend);
+}
+
+static void ggml_backend_cpu_set_tensor_async(ggml_backend_t backend, struct ggml_tensor * tensor, const void * data, size_t offset, size_t size) {
+    GGML_ASSERT(offset + size <= ggml_nbytes(tensor) && "tensor write out of bounds");
+    GGML_ASSERT(tensor->data != NULL && "tensor not allocated");
+
+    memcpy((char *)tensor->data + offset, data, size);
+
+    UNUSED(backend);
+}
+
+static void ggml_backend_cpu_get_tensor_async(ggml_backend_t backend, const struct ggml_tensor * tensor, void * data, size_t offset, size_t size) {
+    GGML_ASSERT(offset + size <= ggml_nbytes(tensor) && "tensor read out of bounds");
+    GGML_ASSERT(tensor->data != NULL && "tensor not allocated");
+
+    memcpy(data, (const char *)tensor->data + offset, size);
+
+    UNUSED(backend);
+}
+
+static void ggml_backend_cpu_synchronize(ggml_backend_t backend) {
+    UNUSED(backend);
+}
+
+static void ggml_backend_cpu_cpy_tensor_from(ggml_backend_t backend, struct ggml_tensor * src, struct ggml_tensor * dst) {
+    ggml_backend_tensor_get(src, dst->data, 0, ggml_nbytes(src));
+
+    UNUSED(backend);
+}
+
+static void ggml_backend_cpu_cpy_tensor_to(ggml_backend_t backend, struct ggml_tensor * src, struct ggml_tensor * dst) {
+    // for a backend such as CUDA that can queue async calls, it is ok to do this asynchronously, but it may not be the case for other backends
+    ggml_backend_tensor_set_async(dst, src->data, 0, ggml_nbytes(src));
+
+    UNUSED(backend);
+}
+
+struct ggml_backend_plan_cpu {
+    struct ggml_cplan cplan;
+    struct ggml_cgraph cgraph;
+};
+
+static ggml_backend_graph_plan_t ggml_backend_cpu_graph_plan_create(ggml_backend_t backend, struct ggml_cgraph * cgraph) {
+    struct ggml_backend_cpu_context * cpu_ctx = (struct ggml_backend_cpu_context *)backend->context;
+
+    struct ggml_backend_plan_cpu * cpu_plan = malloc(sizeof(struct ggml_backend_plan_cpu));
+
+    cpu_plan->cplan = ggml_graph_plan(cgraph, cpu_ctx->n_threads);
+    cpu_plan->cgraph = *cgraph;
+
+    if (cpu_plan->cplan.work_size > 0) {
+        cpu_plan->cplan.work_data = malloc(cpu_plan->cplan.work_size);
+    }
+
+    return cpu_plan;
+}
+
+static void ggml_backend_cpu_graph_plan_free(ggml_backend_t backend, ggml_backend_graph_plan_t plan) {
+    struct ggml_backend_plan_cpu * cpu_plan = (struct ggml_backend_plan_cpu *)plan;
+
+    free(cpu_plan->cplan.work_data);
+    free(cpu_plan);
+
+    UNUSED(backend);
+}
+
+static void ggml_backend_cpu_graph_plan_compute(ggml_backend_t backend, ggml_backend_graph_plan_t plan) {
+    struct ggml_backend_plan_cpu * cpu_plan = (struct ggml_backend_plan_cpu *)plan;
+
+    ggml_graph_compute(&cpu_plan->cgraph, &cpu_plan->cplan);
+
+    UNUSED(backend);
+}
+
+static void ggml_backend_cpu_graph_compute(ggml_backend_t backend, struct ggml_cgraph * cgraph) {
+    struct ggml_backend_cpu_context * cpu_ctx = (struct ggml_backend_cpu_context *)backend->context;
+
+    struct ggml_cplan cplan = ggml_graph_plan(cgraph, cpu_ctx->n_threads);
+
+    if (cpu_ctx->work_size < cplan.work_size) {
+        // TODO: may be faster to free and use malloc to avoid the copy
+        cpu_ctx->work_data = realloc(cpu_ctx->work_data, cplan.work_size);
+        cpu_ctx->work_size = cplan.work_size;
+    }
+
+    cplan.work_data = cpu_ctx->work_data;
+
+    ggml_graph_compute(cgraph, &cplan);
+}
+
+static bool ggml_backend_cpu_supports_op(ggml_backend_t backend, const struct ggml_tensor * op) {
+    return true;
+    UNUSED(backend);
+    UNUSED(op);
+}
+
+static struct ggml_backend_i cpu_backend_i = {
+    /* .get_name            = */ ggml_backend_cpu_name,
+    /* .free                = */ ggml_backend_cpu_free,
+    /* .alloc_buffer        = */ ggml_backend_cpu_alloc_buffer,
+    /* .get_alignment       = */ ggml_backend_cpu_get_alignment,
+    /* .set_tensor_async    = */ ggml_backend_cpu_set_tensor_async,
+    /* .get_tensor_async    = */ ggml_backend_cpu_get_tensor_async,
+    /* .synchronize         = */ ggml_backend_cpu_synchronize,
+    /* .cpy_tensor_from     = */ ggml_backend_cpu_cpy_tensor_from,
+    /* .cpy_tensor_to       = */ ggml_backend_cpu_cpy_tensor_to,
+    /* .graph_plan_create   = */ ggml_backend_cpu_graph_plan_create,
+    /* .graph_plan_free     = */ ggml_backend_cpu_graph_plan_free,
+    /* .graph_plan_compute  = */ ggml_backend_cpu_graph_plan_compute,
+    /* .graph_compute       = */ ggml_backend_cpu_graph_compute,
+    /* .supports_op         = */ ggml_backend_cpu_supports_op,
+};
+
+ggml_backend_t ggml_backend_cpu_init(void) {
+    struct ggml_backend_cpu_context * ctx = malloc(sizeof(struct ggml_backend_cpu_context));
+
+    ctx->n_threads = GGML_DEFAULT_N_THREADS;
+    ctx->work_data = NULL;
+    ctx->work_size = 0;
+
+    ggml_backend_t cpu_backend = malloc(sizeof(struct ggml_backend));
+
+    *cpu_backend = (struct ggml_backend) {
+        /* .interface = */ cpu_backend_i,
+        /* .context   = */ ctx
+    };
+    return cpu_backend;
+}
+
+bool ggml_backend_is_cpu(ggml_backend_t backend) {
+    return backend->iface.get_name == ggml_backend_cpu_name;
+}
+
+void ggml_backend_cpu_set_n_threads(ggml_backend_t backend_cpu, int n_threads) {
+    GGML_ASSERT(ggml_backend_is_cpu(backend_cpu));
+
+    struct ggml_backend_cpu_context * ctx = (struct ggml_backend_cpu_context *)backend_cpu->context;
+    ctx->n_threads = n_threads;
+}
+
+ggml_backend_buffer_t ggml_backend_cpu_buffer_from_ptr(ggml_backend_t backend_cpu, void * ptr, size_t size) {
+    return ggml_backend_buffer_init(backend_cpu, cpu_backend_buffer_i_from_ptr, ptr, size);
+}

runner/ggml-backend.h

@@ -0,0 +1,169 @@
+/**
+ * llama.cpp - git 465219b9143ac01db0990bbcb0a081ef72ec2008
+ *
+ * MIT License
+ *
+ * Copyright (c) 2023 Georgi Gerganov
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to deal
+ * in the Software without restriction, including without limitation the rights
+ * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+ * copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in all
+ * copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#pragma once
+
+#include "ggml.h"
+
+#ifdef  __cplusplus
+extern "C" {
+#endif
+    struct ggml_backend;
+    struct ggml_backend_buffer;
+
+    // type-erased backend-specific types / wrappers
+    typedef void * ggml_backend_context_t;
+    typedef void * ggml_backend_graph_plan_t;
+    typedef void * ggml_backend_buffer_context_t;
+
+    // avoid accessing internals of these types
+    typedef struct ggml_backend        * ggml_backend_t;
+    typedef struct ggml_backend_buffer * ggml_backend_buffer_t;
+
+    //
+    // backend buffer
+    //
+
+    struct ggml_backend_buffer_i {
+        void   (*free_buffer)   (ggml_backend_buffer_t buffer);
+        void * (*get_base)      (ggml_backend_buffer_t buffer); // get base pointer
+        size_t (*get_alloc_size)(ggml_backend_buffer_t buffer, struct ggml_tensor * tensor); // pre-allocation callback
+        void   (*init_tensor)   (ggml_backend_buffer_t buffer, struct ggml_tensor * tensor); // post-allocation callback
+        void   (*free_tensor)   (ggml_backend_buffer_t buffer, struct ggml_tensor * tensor); // pre-free callback
+    };
+
+    // TODO: hide behind API
+    struct ggml_backend_buffer {
+        struct ggml_backend_buffer_i iface;
+
+        ggml_backend_t                backend;
+        ggml_backend_buffer_context_t context;
+
+        size_t size;
+    };
+
+    // backend buffer functions
+    GGML_API ggml_backend_buffer_t ggml_backend_buffer_init(
+            struct ggml_backend                  * backend,
+            struct ggml_backend_buffer_i           iface,
+                   ggml_backend_buffer_context_t   context,
+                   size_t                          size);
+
+    GGML_API void   ggml_backend_buffer_free          (ggml_backend_buffer_t buffer);
+    GGML_API size_t ggml_backend_buffer_get_alignment (ggml_backend_buffer_t buffer);
+    GGML_API void * ggml_backend_buffer_get_base      (ggml_backend_buffer_t buffer);
+    GGML_API size_t ggml_backend_buffer_get_size      (ggml_backend_buffer_t buffer);
+    GGML_API size_t ggml_backend_buffer_get_alloc_size(ggml_backend_buffer_t buffer, struct ggml_tensor * tensor);
+    GGML_API void   ggml_backend_buffer_init_tensor   (ggml_backend_buffer_t buffer, struct ggml_tensor * tensor);
+    GGML_API void   ggml_backend_buffer_free_tensor   (ggml_backend_buffer_t buffer, struct ggml_tensor * tensor);
+
+    //
+    // backend
+    //
+
+    struct ggml_backend_i {
+        const char * (*get_name)(ggml_backend_t backend);
+
+        void (*free)(ggml_backend_t backend);
+
+        // buffer allocation
+        ggml_backend_buffer_t (*alloc_buffer)(ggml_backend_t backend, size_t size);
+
+        // get buffer alignment
+        size_t (*get_alignment)(ggml_backend_t backend);
+
+        // tensor data access
+        // these functions can be asynchronous, helper functions are provided for synchronous access that automatically call synchronize
+        void (*set_tensor_async)(ggml_backend_t backend,       struct ggml_tensor * tensor, const void * data, size_t offset, size_t size);
+        void (*get_tensor_async)(ggml_backend_t backend, const struct ggml_tensor * tensor,       void * data, size_t offset, size_t size);
+        void (*synchronize)     (ggml_backend_t backend);
+
+        // (optional) copy tensor between different backends, allow for single-copy tranfers
+        void (*cpy_tensor_from)(ggml_backend_t backend, struct ggml_tensor * src, struct ggml_tensor * dst);
+        void (*cpy_tensor_to)  (ggml_backend_t backend, struct ggml_tensor * src, struct ggml_tensor * dst);
+
+        // compute graph with a plan
+        ggml_backend_graph_plan_t (*graph_plan_create) (ggml_backend_t backend, struct ggml_cgraph * cgraph);
+        void                      (*graph_plan_free)   (ggml_backend_t backend, ggml_backend_graph_plan_t plan);
+        void                      (*graph_plan_compute)(ggml_backend_t backend, ggml_backend_graph_plan_t plan);
+
+        // compute graph without a plan
+        void (*graph_compute)(ggml_backend_t backend, struct ggml_cgraph * cgraph);
+
+        // check if the backend supports an operation
+        bool (*supports_op)(ggml_backend_t backend, const struct ggml_tensor * op);
+    };
+
+    // TODO: hide behind API
+    struct ggml_backend {
+        struct ggml_backend_i iface;
+
+        ggml_backend_context_t context;
+    };
+
+    // backend helper functions
+    GGML_API ggml_backend_t ggml_get_backend(const struct ggml_tensor * tensor);
+
+    GGML_API const char * ggml_backend_name(ggml_backend_t backend);
+    GGML_API void         ggml_backend_free(ggml_backend_t backend);
+
+    GGML_API ggml_backend_buffer_t ggml_backend_alloc_buffer(ggml_backend_t backend, size_t size);
+
+    GGML_API size_t ggml_backend_get_alignment(ggml_backend_t backend);
+
+    GGML_API void ggml_backend_tensor_set_async(      struct ggml_tensor * tensor, const void * data, size_t offset, size_t size);
+    GGML_API void ggml_backend_tensor_get_async(const struct ggml_tensor * tensor,       void * data, size_t offset, size_t size);
+
+    GGML_API void ggml_backend_tensor_set(      struct ggml_tensor * tensor, const void * data, size_t offset, size_t size);
+    GGML_API void ggml_backend_tensor_get(const struct ggml_tensor * tensor,       void * data, size_t offset, size_t size);
+
+    GGML_API void ggml_backend_synchronize(ggml_backend_t backend);
+
+    GGML_API ggml_backend_graph_plan_t ggml_backend_graph_plan_create (ggml_backend_t backend, struct ggml_cgraph * cgraph);
+
+    GGML_API void ggml_backend_graph_plan_free   (ggml_backend_t backend, ggml_backend_graph_plan_t plan);
+    GGML_API void ggml_backend_graph_plan_compute(ggml_backend_t backend, ggml_backend_graph_plan_t plan);
+    GGML_API void ggml_backend_graph_compute     (ggml_backend_t backend, struct ggml_cgraph * cgraph);
+    GGML_API bool ggml_backend_supports_op       (ggml_backend_t backend, const struct ggml_tensor * op);
+
+    // tensor copy between different backends
+    GGML_API void ggml_backend_tensor_copy(struct ggml_tensor * src, struct ggml_tensor * dst);
+
+    //
+    // CPU backend
+    //
+
+    GGML_API ggml_backend_t ggml_backend_cpu_init(void);
+
+    GGML_API bool ggml_backend_is_cpu(ggml_backend_t backend);
+
+    GGML_API void ggml_backend_cpu_set_n_threads(ggml_backend_t backend_cpu, int n_threads);
+
+    GGML_API ggml_backend_buffer_t ggml_backend_cpu_buffer_from_ptr(ggml_backend_t backend_cpu, void * ptr, size_t size);
+
+#ifdef  __cplusplus
+}
+#endif

runner/ggml-cuda.h

@@ -0,0 +1,77 @@
+/**
+ * llama.cpp - git 465219b9143ac01db0990bbcb0a081ef72ec2008
+ *
+ * MIT License
+ *
+ * Copyright (c) 2023 Georgi Gerganov
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to deal
+ * in the Software without restriction, including without limitation the rights
+ * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+ * copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in all
+ * copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#pragma once
+
+#include "ggml.h"
+#include "ggml-backend.h"
+
+#ifdef GGML_USE_HIPBLAS
+#define GGML_CUDA_NAME "ROCm"
+#define GGML_CUBLAS_NAME "hipBLAS"
+#else
+#define GGML_CUDA_NAME "CUDA"
+#define GGML_CUBLAS_NAME "cuBLAS"
+#endif
+
+#ifdef  __cplusplus
+extern "C" {
+#endif
+
+#define GGML_CUDA_MAX_DEVICES       16
+
+GGML_API void   ggml_init_cublas(void);
+GGML_API void * ggml_cuda_host_malloc(size_t size);
+GGML_API void   ggml_cuda_host_free(void * ptr);
+
+GGML_API bool   ggml_cuda_can_mul_mat(const struct ggml_tensor * src0, const struct ggml_tensor * src1, struct ggml_tensor * dst);
+GGML_API void   ggml_cuda_set_tensor_split(const float * tensor_split);
+GGML_API void   ggml_cuda_transform_tensor(void * data, struct ggml_tensor * tensor);
+GGML_API void   ggml_cuda_free_data(struct ggml_tensor * tensor);
+
+GGML_API void   ggml_cuda_assign_buffers(struct ggml_tensor * tensor);
+GGML_API void   ggml_cuda_assign_buffers_no_scratch(struct ggml_tensor * tensor);
+GGML_API void   ggml_cuda_assign_buffers_force_inplace(struct ggml_tensor * tensor);
+
+GGML_API void   ggml_cuda_assign_buffers_no_alloc(struct ggml_tensor * tensor);
+GGML_API void   ggml_cuda_assign_scratch_offset(struct ggml_tensor * tensor, size_t offset);
+GGML_API void   ggml_cuda_copy_to_device(struct ggml_tensor * tensor);
+
+GGML_API void   ggml_cuda_set_main_device(int main_device);
+GGML_API void   ggml_cuda_set_mul_mat_q(bool mul_mat_q);
+GGML_API void   ggml_cuda_set_scratch_size(size_t scratch_size);
+GGML_API void   ggml_cuda_free_scratch(void);
+GGML_API bool   ggml_cuda_compute_forward(struct ggml_compute_params * params, struct ggml_tensor * tensor);
+
+GGML_API int    ggml_cuda_get_device_count(void);
+GGML_API void   ggml_cuda_get_device_description(int device, char * description, size_t description_size);
+
+// backend API
+GGML_API ggml_backend_t ggml_backend_cuda_init(void); // TODO: take a list of devices to use
+
+#ifdef  __cplusplus
+}
+#endif

runner/ggml-metal.h

@@ -0,0 +1,134 @@
+//go:build darwin
+
+/**
+ * llama.cpp - git 465219b9143ac01db0990bbcb0a081ef72ec2008
+ *
+ * MIT License
+ *
+ * Copyright (c) 2023 Georgi Gerganov
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to deal
+ * in the Software without restriction, including without limitation the rights
+ * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+ * copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in all
+ * copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+// An interface allowing to compute ggml_cgraph with Metal
+//
+// This is a fully functional interface that extends ggml with GPU support for Apple devices.
+// A similar interface can be created for other GPU backends (e.g. Vulkan, CUDA, OpenCL, etc.)
+//
+// How it works?
+//
+// As long as your program can create and evaluate a ggml_cgraph on the CPU, you can use this
+// interface to evaluate the same graph on the GPU. Instead of using ggml_graph_compute(), you
+// use ggml_metal_graph_compute() (or ggml_vulkan_graph_compute(), etc.)
+//
+// You only need to make sure that all memory buffers that you used during the graph creation
+// are mapped to the device memory with the ggml_metal_add_buffer() function. This mapping is
+// used during the graph evaluation to determine the arguments of the compute kernels.
+//
+// Synchronization between device and host memory (for example for input and output tensors)
+// is done with the ggml_metal_set_tensor() and ggml_metal_get_tensor() functions.
+//
+
+#pragma once
+
+#include "ggml.h"
+#include "ggml-backend.h"
+
+#include <stddef.h>
+#include <stdbool.h>
+
+// max memory buffers that can be mapped to the device
+#define GGML_METAL_MAX_BUFFERS 16
+#define GGML_METAL_MAX_COMMAND_BUFFERS 32
+
+struct ggml_tensor;
+struct ggml_cgraph;
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+//
+// internal API
+// temporary exposed to user-code
+//
+
+struct ggml_metal_context;
+
+void ggml_metal_log_set_callback(ggml_log_callback log_callback, void * user_data);
+
+// number of command buffers to use
+struct ggml_metal_context * ggml_metal_init(int n_cb);
+void ggml_metal_free(struct ggml_metal_context * ctx);
+
+void * ggml_metal_host_malloc(size_t n);
+void   ggml_metal_host_free  (void * data);
+
+// set the number of command buffers to use
+void ggml_metal_set_n_cb(struct ggml_metal_context * ctx, int n_cb);
+
+// creates a mapping between a host memory buffer and a device memory buffer
+// - make sure to map all buffers used in the graph before calling ggml_metal_graph_compute
+// - the mapping is used during computation to determine the arguments of the compute kernels
+// - you don't need to keep the host memory buffer allocated as it is never accessed by Metal
+// - max_size specifies the maximum size of a tensor and is used to create shared views such
+//   that it is guaranteed that the tensor will fit in at least one of the views
+//
+bool ggml_metal_add_buffer(
+        struct ggml_metal_context * ctx,
+                       const char * name,
+                             void * data,
+                           size_t   size,
+                           size_t   max_size);
+
+// set data from host memory into the device
+void ggml_metal_set_tensor(struct ggml_metal_context * ctx, struct ggml_tensor * t);
+
+// get data from the device into host memory
+void ggml_metal_get_tensor(struct ggml_metal_context * ctx, struct ggml_tensor * t);
+
+// try to find operations that can be run concurrently in the graph
+// you should run it again if the topology of your graph changes
+void ggml_metal_graph_find_concurrency(struct ggml_metal_context * ctx, struct ggml_cgraph * gf, bool check_mem);
+
+// if the graph has been optimized for concurrently dispatch, return length of the concur_list if optimized
+int ggml_metal_if_optimized(struct ggml_metal_context * ctx);
+
+// output the concur_list for ggml_alloc
+int * ggml_metal_get_concur_list(struct ggml_metal_context * ctx);
+
+// same as ggml_graph_compute but uses Metal
+// creates gf->n_threads command buffers in parallel
+void ggml_metal_graph_compute(struct ggml_metal_context * ctx, struct ggml_cgraph * gf);
+
+//
+// backend API
+// user-code should use only these functions
+//
+
+GGML_API ggml_backend_t ggml_backend_metal_init(void);
+
+GGML_API bool ggml_backend_is_metal(ggml_backend_t backend);
+
+GGML_API void ggml_backend_metal_set_n_cb(ggml_backend_t backend, int n_cb);
+
+#ifdef __cplusplus
+}
+#endif
+

runner/ggml-mpi.c

@@ -0,0 +1,244 @@
+//go:build mpi
+
+/**
+ * llama.cpp - git 465219b9143ac01db0990bbcb0a081ef72ec2008
+ *
+ * MIT License
+ *
+ * Copyright (c) 2023 Georgi Gerganov
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to deal
+ * in the Software without restriction, including without limitation the rights
+ * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+ * copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in all
+ * copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#include "ggml-mpi.h"
+
+#include "ggml.h"
+
+#include <mpi.h>
+
+#include <stdio.h>
+#include <stdlib.h>
+
+#define MIN(a, b) ((a) < (b) ? (a) : (b))
+
+#define UNUSED GGML_UNUSED
+
+struct ggml_mpi_context {
+    int rank;
+    int size;
+};
+
+void ggml_mpi_backend_init(void) {
+    MPI_Init(NULL, NULL);
+}
+
+void ggml_mpi_backend_free(void) {
+    MPI_Finalize();
+}
+
+struct ggml_mpi_context * ggml_mpi_init(void) {
+    struct ggml_mpi_context * ctx = calloc(1, sizeof(struct ggml_mpi_context));
+
+    MPI_Comm_rank(MPI_COMM_WORLD, &ctx->rank);
+    MPI_Comm_size(MPI_COMM_WORLD, &ctx->size);
+
+    return ctx;
+}
+
+void ggml_mpi_free(struct ggml_mpi_context * ctx) {
+    free(ctx);
+}
+
+int ggml_mpi_rank(struct ggml_mpi_context * ctx) {
+    return ctx->rank;
+}
+
+void ggml_mpi_eval_init(
+        struct ggml_mpi_context * ctx_mpi,
+                            int * n_tokens,
+                            int * n_past,
+                            int * n_threads) {
+    UNUSED(ctx_mpi);
+
+    // synchronize the worker node parameters with the root node
+    MPI_Barrier(MPI_COMM_WORLD);
+
+    MPI_Bcast(n_tokens,  1, MPI_INT, 0, MPI_COMM_WORLD);
+    MPI_Bcast(n_past,    1, MPI_INT, 0, MPI_COMM_WORLD);
+    MPI_Bcast(n_threads, 1, MPI_INT, 0, MPI_COMM_WORLD);
+}
+
+static int ggml_graph_get_node_idx(struct ggml_cgraph * gf, const char * name) {
+    struct ggml_tensor * t = ggml_graph_get_tensor(gf, name);
+    if (t == NULL) {
+        fprintf(stderr, "%s: tensor %s not found\n", __func__, name);
+        return -1;
+    }
+
+    for (int i = 0; i < gf->n_nodes; i++) {
+        if (gf->nodes[i] == t) {
+            return i;
+        }
+    }
+
+    fprintf(stderr, "%s: tensor %s not found in graph (should not happen)\n", __func__, name);
+    return -1;
+}
+
+static void ggml_mpi_tensor_send(struct ggml_tensor * t, int mpi_rank_dst) {
+    MPI_Datatype mpi_type;
+
+    switch (t->type) {
+        case GGML_TYPE_I32: mpi_type = MPI_INT32_T; break;
+        case GGML_TYPE_F32: mpi_type = MPI_FLOAT;   break;
+        default: GGML_ASSERT(false && "not implemented");
+    }
+
+    const int retval = MPI_Send(t->data, ggml_nelements(t), mpi_type, mpi_rank_dst, 0, MPI_COMM_WORLD);
+    GGML_ASSERT(retval == MPI_SUCCESS);
+}
+
+static void ggml_mpi_tensor_recv(struct ggml_tensor * t, int mpi_rank_src) {
+    MPI_Datatype mpi_type;
+
+    switch (t->type) {
+        case GGML_TYPE_I32: mpi_type = MPI_INT32_T; break;
+        case GGML_TYPE_F32: mpi_type = MPI_FLOAT;   break;
+        default: GGML_ASSERT(false && "not implemented");
+    }
+
+    MPI_Status status; UNUSED(status);
+
+    const int retval = MPI_Recv(t->data, ggml_nelements(t), mpi_type, mpi_rank_src, MPI_ANY_TAG, MPI_COMM_WORLD, &status);
+    GGML_ASSERT(retval == MPI_SUCCESS);
+}
+
+// TODO: there are many improvements that can be done to this implementation
+void ggml_mpi_graph_compute_pre(
+        struct ggml_mpi_context * ctx_mpi,
+             struct ggml_cgraph * gf,
+                            int   n_layers) {
+    const int mpi_rank = ctx_mpi->rank;
+    const int mpi_size = ctx_mpi->size;
+
+    struct ggml_tensor * inp_tokens = ggml_graph_get_tensor(gf, "inp_tokens");
+    if (inp_tokens == NULL) {
+        fprintf(stderr, "%s: tensor 'inp_tokens' not found\n", __func__);
+        return;
+    }
+
+    struct ggml_tensor * inp0 = ggml_graph_get_tensor(gf, "layer_inp_0");
+    if (inp0 == NULL) {
+        fprintf(stderr, "%s: tensor 'inp0' not found\n", __func__);
+        return;
+    }
+
+    GGML_ASSERT(inp0 == gf->nodes[0]);
+
+    // distribute the compute graph into slices across the MPI nodes
+    //
+    // the main node (0) processes the last layers + the remainder of the compute graph
+    // and is responsible to pass the input tokens to the first node (1)
+    //
+    // node 1:   [(  0) * n_per_node, (  1) * n_per_node)
+    // node 2:   [(  1) * n_per_node, (  2) * n_per_node)
+    // ...
+    // node n-1: [(n-2) * n_per_node, (n-1) * n_per_node)
+    // node 0:   [(n-1) * n_per_node,            n_nodes)
+    //
+    if (mpi_rank > 0) {
+        if (mpi_rank == 1) {
+            // the first node (1) receives the input tokens from the main node (0)
+            ggml_mpi_tensor_recv(inp_tokens, 0);
+        } else {
+            // recv input data for each node into the "inp0" tensor (i.e. the first node in the compute graph)
+            ggml_mpi_tensor_recv(inp0, mpi_rank - 1);
+        }
+    } else if (mpi_size > 1) {
+        // node 0 sends the input tokens to node 1
+        ggml_mpi_tensor_send(inp_tokens, 1);
+
+        // recv the output data from the last node
+        ggml_mpi_tensor_recv(inp0, mpi_size - 1);
+    }
+
+    {
+        const int n_per_node = (n_layers + (mpi_size - 1)) / mpi_size;
+
+        const int mpi_idx = mpi_rank > 0 ? mpi_rank - 1 : mpi_size - 1;
+
+        const int il0 =               (mpi_idx + 0) * n_per_node;
+        const int il1 = MIN(n_layers, (mpi_idx + 1) * n_per_node);
+
+        char name_l0[GGML_MAX_NAME];
+        char name_l1[GGML_MAX_NAME];
+
+        snprintf(name_l0, sizeof(name_l0), "layer_inp_%d", il0);
+        snprintf(name_l1, sizeof(name_l1), "layer_inp_%d", il1);
+
+        const int idx_l0 =                ggml_graph_get_node_idx(gf, name_l0);
+        const int idx_l1 = mpi_rank > 0 ? ggml_graph_get_node_idx(gf, name_l1) + 1 : gf->n_nodes;
+
+        if (idx_l0 < 0 || idx_l1 < 0) {
+            fprintf(stderr, "%s: layer input nodes not found\n", __func__);
+            return;
+        }
+
+        // attach the input data to all nodes that need it
+        // TODO: not great - should be able to do this without modifying the compute graph (see next TODO below)
+        for (int i = idx_l0; i < idx_l1; i++) {
+            if (gf->nodes[i]->src[0] == gf->nodes[idx_l0]) {
+                gf->nodes[i]->src[0] =  inp0;
+            }
+            if (gf->nodes[i]->src[1] == gf->nodes[idx_l0]) {
+                gf->nodes[i]->src[1] =  inp0;
+            }
+        }
+
+        // TODO: instead of rearranging the nodes, we should be able to execute a subset of the compute graph
+        for (int i = 1; i < idx_l1 - idx_l0; i++) {
+            gf->nodes[i] = gf->nodes[idx_l0 + i];
+            gf->grads[i] = gf->grads[idx_l0 + i];
+        }
+
+        // the first node performs the "get_rows" operation, the rest of the nodes get the data from the previous node
+        if (mpi_idx != 0) {
+            gf->nodes[0]->op = GGML_OP_NONE;
+        }
+
+        gf->n_nodes = idx_l1 - idx_l0;
+
+        //fprintf(stderr, "%s: node %d: processing %d nodes [%d, %d)\n", __func__, mpi_rank, gf->n_nodes, il0, il1);
+    }
+}
+
+void ggml_mpi_graph_compute_post(
+        struct ggml_mpi_context * ctx_mpi,
+             struct ggml_cgraph * gf,
+                            int   n_layers) {
+    UNUSED(n_layers);
+
+    const int mpi_rank = ctx_mpi->rank;
+    const int mpi_size = ctx_mpi->size;
+
+    // send the output data to the next node
+    if (mpi_rank > 0) {
+        ggml_mpi_tensor_send(gf->nodes[gf->n_nodes - 1], (mpi_rank + 1) % mpi_size);
+    }
+}

runner/ggml-mpi.h

@@ -0,0 +1,67 @@
+//go:build mpi
+
+/**
+ * llama.cpp - git 465219b9143ac01db0990bbcb0a081ef72ec2008
+ *
+ * MIT License
+ *
+ * Copyright (c) 2023 Georgi Gerganov
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to deal
+ * in the Software without restriction, including without limitation the rights
+ * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+ * copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in all
+ * copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#pragma once
+
+struct ggml_context;
+struct ggml_tensor;
+struct ggml_cgraph;
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+struct ggml_mpi_context;
+
+void ggml_mpi_backend_init(void);
+void ggml_mpi_backend_free(void);
+
+struct ggml_mpi_context * ggml_mpi_init(void);
+void ggml_mpi_free(struct ggml_mpi_context * ctx);
+
+int ggml_mpi_rank(struct ggml_mpi_context * ctx);
+
+void ggml_mpi_eval_init(
+        struct ggml_mpi_context * ctx_mpi,
+                            int * n_tokens,
+                            int * n_past,
+                            int * n_threads);
+
+void ggml_mpi_graph_compute_pre(
+        struct ggml_mpi_context * ctx_mpi,
+             struct ggml_cgraph * gf,
+                            int   n_layers);
+
+void ggml_mpi_graph_compute_post(
+        struct ggml_mpi_context * ctx_mpi,
+             struct ggml_cgraph * gf,
+                            int   n_layers);
+
+#ifdef __cplusplus
+}
+#endif

runner/ggml-opencl.h

@@ -0,0 +1,53 @@
+//go:build opencl
+
+/**
+ * llama.cpp - git 465219b9143ac01db0990bbcb0a081ef72ec2008
+ *
+ * MIT License
+ *
+ * Copyright (c) 2023 Georgi Gerganov
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to deal
+ * in the Software without restriction, including without limitation the rights
+ * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+ * copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in all
+ * copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#pragma once
+
+#include "ggml.h"
+
+#ifdef  __cplusplus
+extern "C" {
+#endif
+
+void ggml_cl_init(void);
+
+void   ggml_cl_mul(const struct ggml_tensor * src0, const struct ggml_tensor * src1, struct ggml_tensor * dst);
+bool   ggml_cl_can_mul_mat(const struct ggml_tensor * src0, const struct ggml_tensor * src1, struct ggml_tensor * dst);
+size_t ggml_cl_mul_mat_get_wsize(const struct ggml_tensor * src0, const struct ggml_tensor * src1, struct ggml_tensor * dst);
+void   ggml_cl_mul_mat(const struct ggml_tensor * src0, const struct ggml_tensor * src1, struct ggml_tensor * dst, void * wdata, size_t wsize);
+
+void * ggml_cl_host_malloc(size_t size);
+void   ggml_cl_host_free(void * ptr);
+
+void ggml_cl_free_data(const struct ggml_tensor* tensor);
+
+void ggml_cl_transform_tensor(void * data, struct ggml_tensor * tensor);
+
+#ifdef  __cplusplus
+}
+#endif

runner/k_quants.h

@@ -0,0 +1,191 @@
+/**
+ * llama.cpp - git 465219b9143ac01db0990bbcb0a081ef72ec2008
+ *
+ * MIT License
+ *
+ * Copyright (c) 2023 Georgi Gerganov
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to deal
+ * in the Software without restriction, including without limitation the rights
+ * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+ * copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in all
+ * copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#pragma once
+
+#include "ggml.h"
+
+#include <stdint.h>
+#include <assert.h>
+#include <stddef.h>
+
+// Super-block size
+#ifdef GGML_QKK_64
+#define QK_K 64
+#define K_SCALE_SIZE 4
+#else
+#define QK_K 256
+#define K_SCALE_SIZE 12
+#endif
+
+#ifndef static_assert
+#if defined(__STDC_VERSION__) && (__STDC_VERSION__ >= 201100L)
+#define static_assert(cond, msg) _Static_assert(cond, msg)
+#else
+#define static_assert(cond, msg) struct global_scope_noop_trick
+#endif
+#endif
+
+//
+// Super-block quantization structures
+//
+
+// 2-bit quantization
+// weight is represented as x = a * q + b
+// 16 blocks of 16 elements each
+// Effectively 2.5625 bits per weight
+typedef struct {
+    uint8_t scales[QK_K/16]; // scales and mins, quantized with 4 bits
+    uint8_t qs[QK_K/4];      // quants
+    ggml_fp16_t d;           // super-block scale for quantized scales
+    ggml_fp16_t dmin;        // super-block scale for quantized mins
+} block_q2_K;
+static_assert(sizeof(block_q2_K) == 2*sizeof(ggml_fp16_t) + QK_K/16 + QK_K/4, "wrong q2_K block size/padding");
+
+// 3-bit quantization
+// weight is represented as x = a * q
+// 16 blocks of 16 elements each
+// Effectively 3.4375 bits per weight
+#ifdef GGML_QKK_64
+typedef struct {
+    uint8_t hmask[QK_K/8];     // quants - high bit
+    uint8_t qs[QK_K/4];        // quants - low 2 bits
+    uint8_t scales[2];
+    ggml_fp16_t d;             // super-block scale
+} block_q3_K;
+static_assert(sizeof(block_q3_K) == sizeof(ggml_fp16_t) + QK_K / 4 + QK_K / 8 + 2, "wrong q3_K block size/padding");
+#else
+typedef struct {
+    uint8_t hmask[QK_K/8];     // quants - high bit
+    uint8_t qs[QK_K/4];        // quants - low 2 bits
+    uint8_t scales[12];        // scales, quantized with 6 bits
+    ggml_fp16_t d;             // super-block scale
+} block_q3_K;
+static_assert(sizeof(block_q3_K) == sizeof(ggml_fp16_t) + QK_K / 4 + QK_K / 8 + 12, "wrong q3_K block size/padding");
+#endif
+
+// 4-bit quantization
+// 8 blocks of 32 elements each
+// weight is represented as x = a * q + b
+// Effectively 4.5 bits per weight
+#ifdef GGML_QKK_64
+typedef struct {
+    ggml_fp16_t d[2];          // super-block scales/mins
+    uint8_t scales[2];         // 4-bit block scales/mins
+    uint8_t qs[QK_K/2];        // 4--bit quants
+} block_q4_K;
+static_assert(sizeof(block_q4_K) == 2*sizeof(ggml_fp16_t) + QK_K/2 + 2, "wrong q4_K block size/padding");
+#else
+typedef struct {
+    ggml_fp16_t d;             // super-block scale for quantized scales
+    ggml_fp16_t dmin;          // super-block scale for quantized mins
+    uint8_t scales[K_SCALE_SIZE]; // scales and mins, quantized with 6 bits
+    uint8_t qs[QK_K/2];        // 4--bit quants
+} block_q4_K;
+static_assert(sizeof(block_q4_K) == 2*sizeof(ggml_fp16_t) + K_SCALE_SIZE + QK_K/2, "wrong q4_K block size/padding");
+#endif
+
+// 5-bit quantization
+// 8 blocks of 32 elements each
+// weight is represented as x = a * q + b
+// Effectively 5.5 bits per weight
+#ifdef GGML_QKK_64
+typedef struct {
+    ggml_fp16_t d;               // super-block scale
+    int8_t  scales[QK_K/16];     // 8-bit block scales
+    uint8_t qh[QK_K/8];          // quants, high bit
+    uint8_t qs[QK_K/2];          // quants, low 4 bits
+} block_q5_K;
+static_assert(sizeof(block_q5_K) == sizeof(ggml_fp16_t) + QK_K/2 + QK_K/8 + QK_K/16, "wrong q5_K block size/padding");
+#else
+typedef struct {
+    ggml_fp16_t d;               // super-block scale for quantized scales
+    ggml_fp16_t dmin;            // super-block scale for quantized mins
+    uint8_t scales[K_SCALE_SIZE];   // scales and mins, quantized with 6 bits
+    uint8_t qh[QK_K/8];          // quants, high bit
+    uint8_t qs[QK_K/2];          // quants, low 4 bits
+} block_q5_K;
+static_assert(sizeof(block_q5_K) == 2*sizeof(ggml_fp16_t) + K_SCALE_SIZE + QK_K/2 + QK_K/8, "wrong q5_K block size/padding");
+#endif
+
+// 6-bit quantization
+// weight is represented as x = a * q
+// 16 blocks of 16 elements each
+// Effectively 6.5625 bits per weight
+typedef struct {
+    uint8_t ql[QK_K/2];      // quants, lower 4 bits
+    uint8_t qh[QK_K/4];      // quants, upper 2 bits
+    int8_t  scales[QK_K/16]; // scales, quantized with 8 bits
+    ggml_fp16_t d;           // super-block scale
+} block_q6_K;
+static_assert(sizeof(block_q6_K) == sizeof(ggml_fp16_t) + QK_K / 16 + 3*QK_K/4, "wrong q6_K block size/padding");
+
+// This is only used for intermediate quantization and dot products
+typedef struct {
+    float   d;              // delta
+    int8_t  qs[QK_K];       // quants
+    int16_t bsums[QK_K/16]; // sum of quants in groups of 16
+} block_q8_K;
+static_assert(sizeof(block_q8_K) == sizeof(float) + QK_K + QK_K/16*sizeof(int16_t), "wrong q8_K block size/padding");
+
+
+// Quantization
+void quantize_row_q2_K_reference(const float * restrict x, block_q2_K * restrict y, int k);
+void quantize_row_q3_K_reference(const float * restrict x, block_q3_K * restrict y, int k);
+void quantize_row_q4_K_reference(const float * restrict x, block_q4_K * restrict y, int k);
+void quantize_row_q5_K_reference(const float * restrict x, block_q5_K * restrict y, int k);
+void quantize_row_q6_K_reference(const float * restrict x, block_q6_K * restrict y, int k);
+void quantize_row_q8_K_reference(const float * restrict x, block_q8_K * restrict y, int k);
+
+void quantize_row_q2_K(const float * restrict x, void * restrict y, int k);
+void quantize_row_q3_K(const float * restrict x, void * restrict y, int k);
+void quantize_row_q4_K(const float * restrict x, void * restrict y, int k);
+void quantize_row_q5_K(const float * restrict x, void * restrict y, int k);
+void quantize_row_q6_K(const float * restrict x, void * restrict y, int k);
+void quantize_row_q8_K(const float * restrict x, void * restrict y, int k);
+
+// Dequantization
+void dequantize_row_q2_K(const block_q2_K * restrict x, float * restrict y, int k);
+void dequantize_row_q3_K(const block_q3_K * restrict x, float * restrict y, int k);
+void dequantize_row_q4_K(const block_q4_K * restrict x, float * restrict y, int k);
+void dequantize_row_q5_K(const block_q5_K * restrict x, float * restrict y, int k);
+void dequantize_row_q6_K(const block_q6_K * restrict x, float * restrict y, int k);
+void dequantize_row_q8_K(const block_q8_K * restrict x, float * restrict y, int k);
+
+// Dot product
+void ggml_vec_dot_q2_K_q8_K(int n, float * restrict s, const void * restrict vx, const void * restrict vy);
+void ggml_vec_dot_q3_K_q8_K(int n, float * restrict s, const void * restrict vx, const void * restrict vy);
+void ggml_vec_dot_q4_K_q8_K(int n, float * restrict s, const void * restrict vx, const void * restrict vy);
+void ggml_vec_dot_q5_K_q8_K(int n, float * restrict s, const void * restrict vx, const void * restrict vy);
+void ggml_vec_dot_q6_K_q8_K(int n, float * restrict s, const void * restrict vx, const void * restrict vy);
+
+// Quantization with histogram collection
+size_t ggml_quantize_q2_K(const float * src, void * dst, int n, int k, int64_t * hist);
+size_t ggml_quantize_q3_K(const float * src, void * dst, int n, int k, int64_t * hist);
+size_t ggml_quantize_q4_K(const float * src, void * dst, int n, int k, int64_t * hist);
+size_t ggml_quantize_q5_K(const float * src, void * dst, int n, int k, int64_t * hist);
+size_t ggml_quantize_q6_K(const float * src, void * dst, int n, int k, int64_t * hist);
+

runner/llama.h

@@ -0,0 +1,778 @@
+/**
+ * llama.cpp - git 465219b9143ac01db0990bbcb0a081ef72ec2008
+ *
+ * MIT License
+ *
+ * Copyright (c) 2023 Georgi Gerganov
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to deal
+ * in the Software without restriction, including without limitation the rights
+ * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+ * copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in all
+ * copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#ifndef LLAMA_H
+#define LLAMA_H
+
+#include "ggml.h"
+#ifdef GGML_USE_CUBLAS
+#include "ggml-cuda.h"
+#define LLAMA_MAX_DEVICES GGML_CUDA_MAX_DEVICES
+#else
+#define LLAMA_MAX_DEVICES 1
+#endif // GGML_USE_CUBLAS
+#include <stddef.h>
+#include <stdint.h>
+#include <stdio.h>
+#include <stdbool.h>
+
+#ifdef LLAMA_SHARED
+#    if defined(_WIN32) && !defined(__MINGW32__)
+#        ifdef LLAMA_BUILD
+#            define LLAMA_API __declspec(dllexport)
+#        else
+#            define LLAMA_API __declspec(dllimport)
+#        endif
+#    else
+#        define LLAMA_API __attribute__ ((visibility ("default")))
+#    endif
+#else
+#    define LLAMA_API
+#endif
+
+#ifdef __GNUC__
+#    define DEPRECATED(func, hint) func __attribute__((deprecated(hint)))
+#elif defined(_MSC_VER)
+#    define DEPRECATED(func, hint) __declspec(deprecated(hint)) func
+#else
+#    define DEPRECATED(func, hint) func
+#endif
+
+#define LLAMA_DEFAULT_SEED 0xFFFFFFFF
+
+#define LLAMA_MAX_RNG_STATE (64*1024)
+
+#define LLAMA_FILE_MAGIC_GGSN 0x6767736eu // 'ggsn'
+
+#define LLAMA_SESSION_MAGIC   LLAMA_FILE_MAGIC_GGSN
+#define LLAMA_SESSION_VERSION 2
+
+#if defined(GGML_USE_CUBLAS) || defined(GGML_USE_CLBLAST) || defined(GGML_USE_METAL)
+// Defined when llama.cpp is compiled with support for offloading model layers to GPU.
+#define LLAMA_SUPPORTS_GPU_OFFLOAD
+#endif
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+    //
+    // C interface
+    //
+    // TODO: show sample usage
+    //
+
+    struct llama_model;
+    struct llama_context;
+
+    typedef int32_t llama_pos;
+    typedef int32_t llama_token;
+    typedef int32_t llama_seq_id;
+
+    enum llama_vocab_type {
+        LLAMA_VOCAB_TYPE_SPM = 0, // SentencePiece
+        LLAMA_VOCAB_TYPE_BPE = 1, // Byte Pair Encoding
+    };
+
+    enum llama_token_type {
+        LLAMA_TOKEN_TYPE_UNDEFINED    = 0,
+        LLAMA_TOKEN_TYPE_NORMAL       = 1,
+        LLAMA_TOKEN_TYPE_UNKNOWN      = 2,
+        LLAMA_TOKEN_TYPE_CONTROL      = 3,
+        LLAMA_TOKEN_TYPE_USER_DEFINED = 4,
+        LLAMA_TOKEN_TYPE_UNUSED       = 5,
+        LLAMA_TOKEN_TYPE_BYTE         = 6,
+    };
+
+    // model file types
+    enum llama_ftype {
+        LLAMA_FTYPE_ALL_F32              = 0,
+        LLAMA_FTYPE_MOSTLY_F16           = 1,  // except 1d tensors
+        LLAMA_FTYPE_MOSTLY_Q4_0          = 2,  // except 1d tensors
+        LLAMA_FTYPE_MOSTLY_Q4_1          = 3,  // except 1d tensors
+        LLAMA_FTYPE_MOSTLY_Q4_1_SOME_F16 = 4,  // tok_embeddings.weight and output.weight are F16
+        // LLAMA_FTYPE_MOSTLY_Q4_2       = 5,  // support has been removed
+        // LLAMA_FTYPE_MOSTLY_Q4_3       = 6,  // support has been removed
+        LLAMA_FTYPE_MOSTLY_Q8_0          = 7,  // except 1d tensors
+        LLAMA_FTYPE_MOSTLY_Q5_0          = 8,  // except 1d tensors
+        LLAMA_FTYPE_MOSTLY_Q5_1          = 9,  // except 1d tensors
+        LLAMA_FTYPE_MOSTLY_Q2_K          = 10, // except 1d tensors
+        LLAMA_FTYPE_MOSTLY_Q3_K_S        = 11, // except 1d tensors
+        LLAMA_FTYPE_MOSTLY_Q3_K_M        = 12, // except 1d tensors
+        LLAMA_FTYPE_MOSTLY_Q3_K_L        = 13, // except 1d tensors
+        LLAMA_FTYPE_MOSTLY_Q4_K_S        = 14, // except 1d tensors
+        LLAMA_FTYPE_MOSTLY_Q4_K_M        = 15, // except 1d tensors
+        LLAMA_FTYPE_MOSTLY_Q5_K_S        = 16, // except 1d tensors
+        LLAMA_FTYPE_MOSTLY_Q5_K_M        = 17, // except 1d tensors
+        LLAMA_FTYPE_MOSTLY_Q6_K          = 18, // except 1d tensors
+
+        LLAMA_FTYPE_GUESSED = 1024, // not specified in the model file
+    };
+
+    typedef struct llama_token_data {
+        llama_token id; // token id
+        float logit;    // log-odds of the token
+        float p;        // probability of the token
+    } llama_token_data;
+
+    typedef struct llama_token_data_array {
+        llama_token_data * data;
+        size_t size;
+        bool sorted;
+    } llama_token_data_array;
+
+    typedef void (*llama_progress_callback)(float progress, void *ctx);
+
+    // Input data for llama_decode
+    // A llama_batch object can contain input about one or many sequences
+    // The provided arrays (i.e. token, embd, pos, etc.) must have size of n_tokens
+    //
+    // - token  : the token ids of the input (used when embd is NULL)
+    // - embd   : token embeddings (i.e. float vector of size n_embd) (used when token is NULL)
+    // - pos    : the positions of the respective token in the sequence
+    // - seq_id : the sequence to which the respective token belongs
+    // - logits : if zero, the logits for the respective token will not be output
+    //
+    typedef struct llama_batch {
+        int32_t n_tokens;
+
+        llama_token  *  token;
+        float        *  embd;
+        llama_pos    *  pos;
+        int32_t      *  n_seq_id;
+        llama_seq_id ** seq_id;
+        int8_t       *  logits;
+
+        // NOTE: helpers for smooth API transition - can be deprecated in the future
+        //       for future-proof code, use the above fields instead and ignore everything below
+        //
+        // pos[i] = all_pos_0 + i*all_pos_1
+        //
+        llama_pos    all_pos_0;  // used if pos == NULL
+        llama_pos    all_pos_1;  // used if pos == NULL
+        llama_seq_id all_seq_id; // used if seq_id == NULL
+    } llama_batch;
+
+    struct llama_model_params {
+        int32_t n_gpu_layers; // number of layers to store in VRAM
+        int32_t main_gpu;     // the GPU that is used for scratch and small tensors
+        const float * tensor_split; // how to split layers across multiple GPUs (size: LLAMA_MAX_DEVICES)
+
+        // called with a progress value between 0 and 1, pass NULL to disable
+        llama_progress_callback progress_callback;
+        // context pointer passed to the progress callback
+        void * progress_callback_user_data;
+
+        // Keep the booleans together to avoid misalignment during copy-by-value.
+        bool vocab_only; // only load the vocabulary, no weights
+        bool use_mmap;   // use mmap if possible
+        bool use_mlock;  // force system to keep model in RAM
+    };
+
+    struct llama_context_params {
+        uint32_t seed;            // RNG seed, -1 for random
+        uint32_t n_ctx;           // text context, 0 = from model
+        uint32_t n_batch;         // prompt processing maximum batch size
+        uint32_t n_threads;       // number of threads to use for generation
+        uint32_t n_threads_batch; // number of threads to use for batch processing
+
+        // ref: https://github.com/ggerganov/llama.cpp/pull/2054
+        float rope_freq_base;  // RoPE base frequency, 0 = from model
+        float rope_freq_scale; // RoPE frequency scaling factor, 0 = from model
+
+        // Keep the booleans together to avoid misalignment during copy-by-value.
+        bool mul_mat_q;  // if true, use experimental mul_mat_q kernels
+        bool f16_kv;     // use fp16 for KV cache, fp32 otherwise
+        bool logits_all; // the llama_eval() call computes all logits, not just the last one
+        bool embedding;  // embedding mode only
+    };
+
+    // model quantization parameters
+    typedef struct llama_model_quantize_params {
+        int nthread;                 // number of threads to use for quantizing, if <=0 will use std::thread::hardware_concurrency()
+        enum llama_ftype ftype;      // quantize to this llama_ftype
+        bool allow_requantize;       // allow quantizing non-f32/f16 tensors
+        bool quantize_output_tensor; // quantize output.weight
+        bool only_copy;              // only copy tensors - ftype, allow_requantize and quantize_output_tensor are ignored
+    } llama_model_quantize_params;
+
+    // grammar types
+    struct llama_grammar;
+
+    // grammar element type
+    enum llama_gretype {
+        // end of rule definition
+        LLAMA_GRETYPE_END            = 0,
+
+        // start of alternate definition for rule
+        LLAMA_GRETYPE_ALT            = 1,
+
+        // non-terminal element: reference to rule
+        LLAMA_GRETYPE_RULE_REF       = 2,
+
+        // terminal element: character (code point)
+        LLAMA_GRETYPE_CHAR           = 3,
+
+        // inverse char(s) ([^a], [^a-b] [^abc])
+        LLAMA_GRETYPE_CHAR_NOT       = 4,
+
+        // modifies a preceding LLAMA_GRETYPE_CHAR or LLAMA_GRETYPE_CHAR_ALT to
+        // be an inclusive range ([a-z])
+        LLAMA_GRETYPE_CHAR_RNG_UPPER = 5,
+
+        // modifies a preceding LLAMA_GRETYPE_CHAR or
+        // LLAMA_GRETYPE_CHAR_RNG_UPPER to add an alternate char to match ([ab], [a-zA])
+        LLAMA_GRETYPE_CHAR_ALT       = 6,
+    };
+
+    typedef struct llama_grammar_element {
+        enum llama_gretype type;
+        uint32_t           value; // Unicode code point or rule ID
+    } llama_grammar_element;
+
+    // performance timing information
+    struct llama_timings {
+        double t_start_ms;
+        double t_end_ms;
+        double t_load_ms;
+        double t_sample_ms;
+        double t_p_eval_ms;
+        double t_eval_ms;
+
+        int32_t n_sample;
+        int32_t n_p_eval;
+        int32_t n_eval;
+    };
+
+    // Helpers for getting default parameters
+    LLAMA_API struct llama_model_params llama_model_default_params(void);
+    LLAMA_API struct llama_context_params llama_context_default_params(void);
+    LLAMA_API struct llama_model_quantize_params llama_model_quantize_default_params(void);
+
+    // Initialize the llama + ggml backend
+    // If numa is true, use NUMA optimizations
+    // Call once at the start of the program
+    LLAMA_API void llama_backend_init(bool numa);
+
+    // Call once at the end of the program - currently only used for MPI
+    LLAMA_API void llama_backend_free(void);
+
+    LLAMA_API struct llama_model * llama_load_model_from_file(
+                             const char * path_model,
+            struct llama_model_params     params);
+
+    LLAMA_API void llama_free_model(struct llama_model * model);
+
+    LLAMA_API struct llama_context * llama_new_context_with_model(
+                     struct llama_model * model,
+            struct llama_context_params   params);
+
+    // Frees all allocated memory
+    LLAMA_API void llama_free(struct llama_context * ctx);
+
+    LLAMA_API int64_t llama_time_us(void);
+
+    LLAMA_API int  llama_max_devices    (void);
+    LLAMA_API bool llama_mmap_supported (void);
+    LLAMA_API bool llama_mlock_supported(void);
+
+    LLAMA_API const struct llama_model * llama_get_model(const struct llama_context * ctx);
+
+    LLAMA_API int llama_n_ctx      (const struct llama_context * ctx);
+
+    LLAMA_API enum llama_vocab_type llama_vocab_type(const struct llama_model * model);
+
+    LLAMA_API int llama_n_vocab    (const struct llama_model * model);
+    LLAMA_API int llama_n_ctx_train(const struct llama_model * model);
+    LLAMA_API int llama_n_embd     (const struct llama_model * model);
+
+    // Get the model's RoPE frequency scaling factor
+    LLAMA_API float llama_rope_freq_scale_train(const struct llama_model * model);
+
+    // Get a string describing the model type
+    LLAMA_API int llama_model_desc(const struct llama_model * model, char * buf, size_t buf_size);
+
+    // Returns the total size of all the tensors in the model in bytes
+    LLAMA_API uint64_t llama_model_size(const struct llama_model * model);
+
+    // Returns the total number of parameters in the model
+    LLAMA_API uint64_t llama_model_n_params(const struct llama_model * model);
+
+    // Get a llama model tensor
+    LLAMA_API struct ggml_tensor * llama_get_model_tensor(struct llama_model * model, const char * name);
+
+    // Returns 0 on success
+    LLAMA_API int llama_model_quantize(
+            const char * fname_inp,
+            const char * fname_out,
+            const llama_model_quantize_params * params);
+
+    // Apply a LoRA adapter to a loaded model
+    // path_base_model is the path to a higher quality model to use as a base for
+    // the layers modified by the adapter. Can be NULL to use the current loaded model.
+    // The model needs to be reloaded before applying a new adapter, otherwise the adapter
+    // will be applied on top of the previous one
+    // Returns 0 on success
+    LLAMA_API DEPRECATED(int llama_apply_lora_from_file(
+            struct llama_context * ctx,
+                      const char * path_lora,
+                           float   scale,
+                      const char * path_base_model,
+                             int   n_threads),
+            "use llama_model_apply_lora_from_file instead");
+
+    LLAMA_API int llama_model_apply_lora_from_file(
+            const struct llama_model * model,
+                      const char * path_lora,
+                           float   scale,
+                      const char * path_base_model,
+                             int   n_threads);
+
+    //
+    // KV cache
+    //
+
+    // Returns the number of tokens in the KV cache
+    LLAMA_API DEPRECATED(int llama_get_kv_cache_token_count(const struct llama_context * ctx),
+            "avoid using this, it will be removed in the future, instead - count the tokens in user code");
+
+    // Remove all tokens data of cells in [c0, c1)
+    // c0 < 0 : [0,  c1]
+    // c1 < 0 : [c0, inf)
+    LLAMA_API void llama_kv_cache_tokens_rm(
+            struct llama_context * ctx,
+                         int32_t   c0,
+                         int32_t   c1);
+
+    // Removes all tokens that belong to the specified sequence and have positions in [p0, p1)
+    // p0 < 0 : [0,  p1]
+    // p1 < 0 : [p0, inf)
+    LLAMA_API void llama_kv_cache_seq_rm(
+            struct llama_context * ctx,
+                    llama_seq_id   seq_id,
+                       llama_pos   p0,
+                       llama_pos   p1);
+
+    // Copy all tokens that belong to the specified sequence to another sequence
+    // Note that this does not allocate extra KV cache memory - it simply assigns the tokens to the new sequence
+    // p0 < 0 : [0,  p1]
+    // p1 < 0 : [p0, inf)
+    LLAMA_API void llama_kv_cache_seq_cp(
+            struct llama_context * ctx,
+                    llama_seq_id   seq_id_src,
+                    llama_seq_id   seq_id_dst,
+                       llama_pos   p0,
+                       llama_pos   p1);
+
+    // Removes all tokens that do not belong to the specified sequence
+    LLAMA_API void llama_kv_cache_seq_keep(
+            struct llama_context * ctx,
+                    llama_seq_id   seq_id);
+
+    // Adds relative position "delta" to all tokens that belong to the specified sequence and have positions in [p0, p1)
+    // If the KV cache is RoPEd, the KV data is updated accordingly
+    // p0 < 0 : [0,  p1]
+    // p1 < 0 : [p0, inf)
+    LLAMA_API void llama_kv_cache_seq_shift(
+            struct llama_context * ctx,
+                    llama_seq_id   seq_id,
+                       llama_pos   p0,
+                       llama_pos   p1,
+                       llama_pos   delta);
+
+    //
+    // State / sessions
+    //
+
+    // Returns the maximum size in bytes of the state (rng, logits, embedding
+    // and kv_cache) - will often be smaller after compacting tokens
+    LLAMA_API size_t llama_get_state_size(const struct llama_context * ctx);
+
+    // Copies the state to the specified destination address.
+    // Destination needs to have allocated enough memory.
+    // Returns the number of bytes copied
+    LLAMA_API size_t llama_copy_state_data(
+            struct llama_context * ctx,
+                         uint8_t * dst);
+
+    // Set the state reading from the specified address
+    // Returns the number of bytes read
+    LLAMA_API size_t llama_set_state_data(
+            struct llama_context * ctx,
+                         uint8_t * src);
+
+    // Save/load session file
+    LLAMA_API bool llama_load_session_file(
+            struct llama_context * ctx,
+                      const char * path_session,
+                     llama_token * tokens_out,
+                          size_t   n_token_capacity,
+                          size_t * n_token_count_out);
+
+    LLAMA_API bool llama_save_session_file(
+            struct llama_context * ctx,
+                      const char * path_session,
+               const llama_token * tokens,
+                          size_t   n_token_count);
+
+    //
+    // Decoding
+    //
+
+    // Run the llama inference to obtain the logits and probabilities for the next token(s).
+    // tokens + n_tokens is the provided batch of new tokens to process
+    // n_past is the number of tokens to use from previous eval calls
+    // Returns 0 on success
+    // DEPRECATED: use llama_decode() instead
+    LLAMA_API DEPRECATED(int llama_eval(
+            struct llama_context * ctx,
+                     llama_token * tokens,
+                         int32_t   n_tokens,
+                             int   n_past),
+            "use llama_decode() instead");
+
+    // Same as llama_eval, but use float matrix input directly.
+    // DEPRECATED: use llama_decode() instead
+    LLAMA_API DEPRECATED(int llama_eval_embd(
+            struct llama_context * ctx,
+                           float * embd,
+                         int32_t   n_tokens,
+                             int   n_past),
+            "use llama_decode() instead");
+
+    // Return batch for single sequence of tokens starting at pos_0
+    //
+    // NOTE: this is a helper function to facilitate transition to the new batch API - avoid using it
+    //
+    LLAMA_API struct llama_batch llama_batch_get_one(
+                  llama_token * tokens,
+                      int32_t   n_tokens,
+                    llama_pos   pos_0,
+                 llama_seq_id   seq_id);
+
+    // Allocates a batch of tokens on the heap that can hold a maximum of n_tokens
+    // Each token can be assigned up to n_seq_max sequence ids
+    // The batch has to be freed with llama_batch_free()
+    // If embd != 0, llama_batch.embd will be allocated with size of n_tokens * embd * sizeof(float)
+    // Otherwise, llama_batch.token will be allocated to store n_tokens llama_token
+    // The rest of the llama_batch members are allocated with size n_tokens
+    // All members are left uninitialized
+    LLAMA_API struct llama_batch llama_batch_init(
+            int32_t n_tokens,
+            int32_t embd,
+            int32_t n_seq_max);
+
+    // Frees a batch of tokens allocated with llama_batch_init()
+    LLAMA_API void llama_batch_free(struct llama_batch batch);
+
+    // Positive return values does not mean a fatal error, but rather a warning.
+    //   0 - success
+    //   1 - could not find a KV slot for the batch (try reducing the size of the batch or increase the context)
+    // < 0 - error
+    LLAMA_API int llama_decode(
+            struct llama_context * ctx,
+              struct llama_batch   batch);
+
+    // Set the number of threads used for decoding
+    // n_threads is the number of threads used for generation (single token)
+    // n_threads_batch is the number of threads used for prompt and batch processing (multiple tokens)
+    LLAMA_API void llama_set_n_threads(struct llama_context * ctx, uint32_t n_threads, uint32_t n_threads_batch);
+
+    // Token logits obtained from the last call to llama_eval()
+    // The logits for the last token are stored in the last row
+    // Logits for which llama_batch.logits[i] == 0 are undefined
+    // Rows: n_tokens provided with llama_batch
+    // Cols: n_vocab
+    LLAMA_API float * llama_get_logits(struct llama_context * ctx);
+
+    // Logits for the ith token. Equivalent to:
+    // llama_get_logits(ctx) + i*n_vocab
+    LLAMA_API float * llama_get_logits_ith(struct llama_context * ctx, int32_t i);
+
+    // Get the embeddings for the input
+    // shape: [n_embd] (1-dimensional)
+    LLAMA_API float * llama_get_embeddings(struct llama_context * ctx);
+
+    //
+    // Vocab
+    //
+
+    LLAMA_API const char * llama_token_get_text(const struct llama_context * ctx, llama_token token);
+
+    LLAMA_API float llama_token_get_score(const struct llama_context * ctx, llama_token token);
+
+    LLAMA_API enum llama_token_type llama_token_get_type(const struct llama_context * ctx, llama_token token);
+
+    // Special tokens
+    LLAMA_API llama_token llama_token_bos(const struct llama_context * ctx);  // beginning-of-sentence
+    LLAMA_API llama_token llama_token_eos(const struct llama_context * ctx);  // end-of-sentence
+    LLAMA_API llama_token llama_token_nl (const struct llama_context * ctx);  // next-line
+    // codellama infill tokens
+    LLAMA_API llama_token llama_token_prefix(const struct llama_context * ctx); // Beginning of infill prefix
+    LLAMA_API llama_token llama_token_middle(const struct llama_context * ctx); // Beginning of infill middle
+    LLAMA_API llama_token llama_token_suffix(const struct llama_context * ctx); // Beginning of infill suffix
+    LLAMA_API llama_token llama_token_eot   (const struct llama_context * ctx); // End of infill middle
+
+    //
+    // Tokenization
+    //
+
+    /// @details Convert the provided text into tokens.
+    /// @param tokens The tokens pointer must be large enough to hold the resulting tokens.
+    /// @return Returns the number of tokens on success, no more than n_max_tokens
+    /// @return Returns a negative number on failure - the number of tokens that would have been returned
+    /// @param special Allow tokenizing special and/or control tokens which otherwise are not exposed and treated as plaintext.
+    ///                Does not insert a leading space.
+    LLAMA_API int llama_tokenize(
+        const struct llama_model * model,
+                      const char * text,
+                             int   text_len,
+                     llama_token * tokens,
+                             int   n_max_tokens,
+                            bool   add_bos,
+                            bool   special);
+
+    // Token Id -> Piece.
+    // Uses the vocabulary in the provided context.
+    // Does not write null terminator to the buffer.
+    // User code is responsible to remove the leading whitespace of the first non-BOS token when decoding multiple tokens.
+    LLAMA_API int llama_token_to_piece(
+              const struct llama_model * model,
+                           llama_token   token,
+                                  char * buf,
+                                  int    length);
+
+    //
+    // Grammar
+    //
+
+    LLAMA_API struct llama_grammar * llama_grammar_init(
+            const llama_grammar_element ** rules,
+                                 size_t    n_rules,
+                                 size_t    start_rule_index);
+
+    LLAMA_API void llama_grammar_free(struct llama_grammar * grammar);
+
+    LLAMA_API struct llama_grammar * llama_grammar_copy(const struct llama_grammar * grammar);
+
+    //
+    // Sampling functions
+    //
+
+    // Sets the current rng seed.
+    LLAMA_API void llama_set_rng_seed(struct llama_context * ctx, uint32_t seed);
+
+    /// @details Repetition penalty described in CTRL academic paper https://arxiv.org/abs/1909.05858, with negative logit fix.
+    /// @details Frequency and presence penalties described in OpenAI API https://platform.openai.com/docs/api-reference/parameter-details.
+    LLAMA_API void llama_sample_repetition_penalties(
+            struct llama_context * ctx,
+          llama_token_data_array * candidates,
+               const llama_token * last_tokens,
+                          size_t   penalty_last_n,
+                           float   penalty_repeat,
+                           float   penalty_freq,
+                           float   penalty_present);
+
+    /// @details Apply classifier-free guidance to the logits as described in academic paper "Stay on topic with Classifier-Free Guidance" https://arxiv.org/abs/2306.17806
+    /// @param candidates A vector of `llama_token_data` containing the candidate tokens, the logits must be directly extracted from the original generation context without being sorted.
+    /// @params guidance_ctx A separate context from the same model. Other than a negative prompt at the beginning, it should have all generated and user input tokens copied from the main context.
+    /// @params scale Guidance strength. 1.0f means no guidance. Higher values mean stronger guidance.
+    LLAMA_API void llama_sample_classifier_free_guidance(
+              struct llama_context * ctx,
+            llama_token_data_array * candidates,
+              struct llama_context * guidance_ctx,
+                             float   scale);
+
+    /// @details Sorts candidate tokens by their logits in descending order and calculate probabilities based on logits.
+    LLAMA_API void llama_sample_softmax(
+            struct llama_context * ctx,
+          llama_token_data_array * candidates);
+
+    /// @details Top-K sampling described in academic paper "The Curious Case of Neural Text Degeneration" https://arxiv.org/abs/1904.09751
+    LLAMA_API void llama_sample_top_k(
+            struct llama_context * ctx,
+          llama_token_data_array * candidates,
+                             int   k,
+                          size_t   min_keep);
+
+    /// @details Nucleus sampling described in academic paper "The Curious Case of Neural Text Degeneration" https://arxiv.org/abs/1904.09751
+    LLAMA_API void llama_sample_top_p(
+            struct llama_context * ctx,
+          llama_token_data_array * candidates,
+                           float   p,
+                          size_t   min_keep);
+
+    /// @details Tail Free Sampling described in https://www.trentonbricken.com/Tail-Free-Sampling/.
+    LLAMA_API void llama_sample_tail_free(
+            struct llama_context * ctx,
+          llama_token_data_array * candidates,
+                           float   z,
+                          size_t   min_keep);
+
+    /// @details Locally Typical Sampling implementation described in the paper https://arxiv.org/abs/2202.00666.
+    LLAMA_API void llama_sample_typical(
+            struct llama_context * ctx,
+          llama_token_data_array * candidates,
+                           float   p,
+                          size_t   min_keep);
+
+    LLAMA_API void llama_sample_temp(
+            struct llama_context * ctx,
+          llama_token_data_array * candidates,
+                           float   temp);
+
+    LLAMA_API DEPRECATED(void llama_sample_temperature(
+                struct llama_context * ctx,
+              llama_token_data_array * candidates,
+                               float   temp),
+            "use llama_sample_temp instead");
+
+    /// @details Apply constraints from grammar
+    LLAMA_API void llama_sample_grammar(
+            struct llama_context * ctx,
+          llama_token_data_array * candidates,
+      const struct llama_grammar * grammar);
+
+    /// @details Mirostat 1.0 algorithm described in the paper https://arxiv.org/abs/2007.14966. Uses tokens instead of words.
+    /// @param candidates A vector of `llama_token_data` containing the candidate tokens, their probabilities (p), and log-odds (logit) for the current position in the generated text.
+    /// @param tau  The target cross-entropy (or surprise) value you want to achieve for the generated text. A higher value corresponds to more surprising or less predictable text, while a lower value corresponds to less surprising or more predictable text.
+    /// @param eta The learning rate used to update `mu` based on the error between the target and observed surprisal of the sampled word. A larger learning rate will cause `mu` to be updated more quickly, while a smaller learning rate will result in slower updates.
+    /// @param m The number of tokens considered in the estimation of `s_hat`. This is an arbitrary value that is used to calculate `s_hat`, which in turn helps to calculate the value of `k`. In the paper, they use `m = 100`, but you can experiment with different values to see how it affects the performance of the algorithm.
+    /// @param mu Maximum cross-entropy. This value is initialized to be twice the target cross-entropy (`2 * tau`) and is updated in the algorithm based on the error between the target and observed surprisal.
+    LLAMA_API llama_token llama_sample_token_mirostat(
+            struct llama_context * ctx,
+          llama_token_data_array * candidates,
+                           float   tau,
+                           float   eta,
+                             int   m,
+                           float * mu);
+
+    /// @details Mirostat 2.0 algorithm described in the paper https://arxiv.org/abs/2007.14966. Uses tokens instead of words.
+    /// @param candidates A vector of `llama_token_data` containing the candidate tokens, their probabilities (p), and log-odds (logit) for the current position in the generated text.
+    /// @param tau  The target cross-entropy (or surprise) value you want to achieve for the generated text. A higher value corresponds to more surprising or less predictable text, while a lower value corresponds to less surprising or more predictable text.
+    /// @param eta The learning rate used to update `mu` based on the error between the target and observed surprisal of the sampled word. A larger learning rate will cause `mu` to be updated more quickly, while a smaller learning rate will result in slower updates.
+    /// @param mu Maximum cross-entropy. This value is initialized to be twice the target cross-entropy (`2 * tau`) and is updated in the algorithm based on the error between the target and observed surprisal.
+    LLAMA_API llama_token llama_sample_token_mirostat_v2(
+            struct llama_context * ctx,
+          llama_token_data_array * candidates,
+                           float   tau,
+                           float   eta,
+                           float * mu);
+
+    /// @details Selects the token with the highest probability.
+    LLAMA_API llama_token llama_sample_token_greedy(
+            struct llama_context * ctx,
+          llama_token_data_array * candidates);
+
+    /// @details Randomly selects a token from the candidates based on their probabilities.
+    LLAMA_API llama_token llama_sample_token(
+            struct llama_context * ctx,
+          llama_token_data_array * candidates);
+
+    /// @details Accepts the sampled token into the grammar
+    LLAMA_API void llama_grammar_accept_token(
+            struct llama_context * ctx,
+            struct llama_grammar * grammar,
+                     llama_token   token);
+
+    //
+    // Beam search
+    //
+
+    struct llama_beam_view {
+        const llama_token * tokens;
+
+        size_t n_tokens;
+        float  p;        // Cumulative beam probability (renormalized relative to all beams)
+        bool   eob;      // Callback should set this to true when a beam is at end-of-beam.
+    };
+
+    // Passed to beam_search_callback function.
+    // Whenever 0 < common_prefix_length, this number of tokens should be copied from any of the beams
+    // (e.g. beams[0]) as they will be removed (shifted) from all beams in all subsequent callbacks.
+    // These pointers are valid only during the synchronous callback, so should not be saved.
+    struct llama_beams_state {
+        struct llama_beam_view * beam_views;
+
+        size_t n_beams;               // Number of elements in beam_views[].
+        size_t common_prefix_length;  // Current max length of prefix tokens shared by all beams.
+        bool   last_call;             // True iff this is the last callback invocation.
+    };
+
+    // Type of pointer to the beam_search_callback function.
+    // void* callback_data is any custom data passed to llama_beam_search, that is subsequently
+    // passed back to beam_search_callback. This avoids having to use global variables in the callback.
+    typedef void (*llama_beam_search_callback_fn_t)(void * callback_data, struct llama_beams_state);
+
+    /// @details Deterministically returns entire sentence constructed by a beam search.
+    /// @param ctx Pointer to the llama_context.
+    /// @param callback Invoked for each iteration of the beam_search loop, passing in beams_state.
+    /// @param callback_data A pointer that is simply passed back to callback.
+    /// @param n_beams Number of beams to use.
+    /// @param n_past Number of tokens already evaluated.
+    /// @param n_predict Maximum number of tokens to predict. EOS may occur earlier.
+    LLAMA_API void llama_beam_search(
+                   struct llama_context * ctx,
+        llama_beam_search_callback_fn_t   callback,
+                                   void * callback_data,
+                                 size_t   n_beams,
+                                    int   n_past,
+                                    int   n_predict);
+
+    // Performance information
+    LLAMA_API struct llama_timings llama_get_timings(struct llama_context * ctx);
+
+    LLAMA_API void llama_print_timings(struct llama_context * ctx);
+    LLAMA_API void llama_reset_timings(struct llama_context * ctx);
+
+    // Print system information
+    LLAMA_API const char * llama_print_system_info(void);
+
+    // Set callback for all future logging events.
+    // If this is not called, or NULL is supplied, everything is output on stderr.
+    LLAMA_API void llama_log_set(ggml_log_callback log_callback, void * user_data);
+
+    LLAMA_API void llama_dump_timing_info_yaml(FILE * stream, const struct llama_context * ctx);
+
+#ifdef __cplusplus
+}
+#endif
+
+// Internal API to be implemented by llama.cpp and used by tests/benchmarks only
+#ifdef LLAMA_API_INTERNAL
+
+#include <vector>
+#include <string>
+
+struct ggml_tensor;
+
+const std::vector<std::pair<std::string, struct ggml_tensor *>> & llama_internal_get_tensor_map(
+    struct llama_context * ctx
+);
+
+#endif // LLAMA_API_INTERNAL
+
+#endif // LLAMA_H

runner/main.cpp

@@ -0,0 +1,272 @@
+#include <cmath>
+#include <cstdio>
+#include <string>
+#include <vector>
+#include <thread>
+
+#include "llama.h"
+#include "main.h"
+
+std::vector<llama_token> tokenize(const struct llama_model * model, const std::string & text, bool add_bos, bool special = false) {
+    // upper limit for the number of tokens
+    int n_tokens = text.length() + add_bos;
+    std::vector<llama_token> result(n_tokens);
+    n_tokens = llama_tokenize(model, text.data(), text.length(), result.data(), result.size(), add_bos, special);
+    if (n_tokens < 0) {
+        result.resize(-n_tokens);
+        int check = llama_tokenize(model, text.data(), text.length(), result.data(), result.size(), add_bos, special);
+        GGML_ASSERT(check == -n_tokens);
+    } else {
+        result.resize(n_tokens);
+    }
+    return result;
+}
+
+std::string token_to_piece(const struct llama_context * ctx, llama_token token) {
+    std::vector<char> result(8, 0);
+    const int n_tokens = llama_token_to_piece(llama_get_model(ctx), token, result.data(), result.size());
+    if (n_tokens < 0) {
+        result.resize(-n_tokens);
+        int check = llama_token_to_piece(llama_get_model(ctx), token, result.data(), result.size());
+        GGML_ASSERT(check == -n_tokens);
+    } else {
+        result.resize(n_tokens);
+    }
+
+    return std::string(result.data(), result.size());
+}
+
+void batch_add(
+                 struct llama_batch & batch,
+                        llama_token   id,
+                          llama_pos   pos,
+    const std::vector<llama_seq_id> & seq_ids,
+                               bool   logits) {
+    batch.token   [batch.n_tokens] = id;
+    batch.pos     [batch.n_tokens] = pos,
+    batch.n_seq_id[batch.n_tokens] = seq_ids.size();
+    for (size_t i = 0; i < seq_ids.size(); ++i) {
+        batch.seq_id[batch.n_tokens][i] = seq_ids[i];
+    }
+    batch.logits  [batch.n_tokens] = logits;
+
+    batch.n_tokens++;
+}
+
+void batch_clear(struct llama_batch & batch) {
+    batch.n_tokens = 0;
+}
+
+int generate(const char *model_path, const char *prompt) {
+    // number of parallel batches
+    int n_parallel = 1;
+
+    // total length of the sequences including the prompt
+    int n_len = 32;
+
+    // init LLM
+    llama_backend_init(true);
+
+    // initialize the model
+
+    llama_model_params model_params = llama_model_default_params();
+
+    // model_params.n_gpu_layers = 99; // offload all layers to the GPU
+
+    llama_model * model = llama_load_model_from_file(model_path, model_params);
+
+    if (model == NULL) {
+        fprintf(stderr , "%s: error: unable to load model\n" , __func__);
+        return 1;
+    }
+
+    // tokenize the prompt
+
+    std::vector<llama_token> tokens_list;
+    tokens_list = tokenize(model, std::string(prompt), true);
+    const int n_kv_req = tokens_list.size() + (n_len - tokens_list.size())*n_parallel;
+
+    // initialize the context
+
+    llama_context_params ctx_params = llama_context_default_params();
+
+    ctx_params.seed  = 1234;
+    ctx_params.n_ctx = n_kv_req;
+    ctx_params.n_batch = std::max(n_len, n_parallel);
+    ctx_params.n_threads       = std::thread::hardware_concurrency();
+    ctx_params.n_threads_batch = ctx_params.n_threads;
+
+    llama_context * ctx = llama_new_context_with_model(model, ctx_params);
+
+    if (ctx == NULL) {
+        fprintf(stderr , "%s: error: failed to create the llama_context\n" , __func__);
+        return 1;
+    }
+
+    const int n_ctx    = llama_n_ctx(ctx);
+
+    printf("\n%s: n_len = %d, n_ctx = %d, n_batch = %d, n_parallel = %d, n_kv_req = %d\n", __func__, n_len, n_ctx, ctx_params.n_batch, n_parallel, n_kv_req);
+
+    // make sure the KV cache is big enough to hold all the prompt and generated tokens
+    if (n_kv_req > n_ctx) {
+        printf("%s: error: n_kv_req (%d) > n_ctx, the required KV cache size is not big enough\n", __func__,  n_kv_req);
+        printf("%s:        either reduce n_parallel or increase n_ctx\n", __func__);
+        return 1;
+    }
+
+    // print the prompt token-by-token
+
+    fprintf(stderr, "\n");
+
+    for (auto id : tokens_list) {
+        fprintf(stderr, "%s", token_to_piece(ctx, id).c_str());
+    }
+
+    fflush(stderr);
+
+    // create a llama_batch
+    // we use this object to submit token data for decoding
+    llama_batch batch = llama_batch_init(std::max(tokens_list.size(), (size_t)n_parallel), 0, 1);
+
+    // evaluate the initial prompt
+    for (size_t i = 0; i < tokens_list.size(); ++i) {
+        batch_add(batch, tokens_list[i], i, { 0 }, false);
+    }
+    GGML_ASSERT(batch.n_tokens == (int) tokens_list.size());
+
+    // llama_decode will output logits only for the last token of the prompt
+    batch.logits[batch.n_tokens - 1] = true;
+
+    if (llama_decode(ctx, batch) != 0) {
+        printf("%s: llama_decode() failed\n", __func__);
+        return 1;
+    }
+
+    // assign the system KV cache to all parallel sequences
+    // this way, the parallel sequences will "reuse" the prompt tokens without having to copy them
+    for (int32_t i = 1; i < n_parallel; ++i) {
+        llama_kv_cache_seq_cp(ctx, 0, i, 0, batch.n_tokens);
+    }
+
+    if (n_parallel > 1) {
+        printf("\n\n%s: generating %d sequences ...\n", __func__, n_parallel);
+    }
+
+    // main loop
+
+    // we will store the parallel decoded sequences in this vector
+    std::vector<std::string> streams(n_parallel);
+
+    // remember the batch index of the last token for each parallel sequence
+    // we need this to determine which logits to sample from
+    std::vector<int32_t> i_batch(n_parallel, batch.n_tokens - 1);
+
+    int n_cur    = batch.n_tokens;
+    int n_decode = 0;
+
+    const auto t_main_start = ggml_time_us();
+
+    while (n_cur <= n_len) {
+        // prepare the next batch
+        batch_clear(batch);
+
+        // sample the next token for each parallel sequence / stream
+        for (int32_t i = 0; i < n_parallel; ++i) {
+            if (i_batch[i] < 0) {
+                // the stream has already finished
+                continue;
+            }
+
+            auto   n_vocab = llama_n_vocab(model);
+            auto * logits  = llama_get_logits_ith(ctx, i_batch[i]);
+
+            std::vector<llama_token_data> candidates;
+            candidates.reserve(n_vocab);
+
+            for (llama_token token_id = 0; token_id < n_vocab; token_id++) {
+                candidates.emplace_back(llama_token_data{ token_id, logits[token_id], 0.0f });
+            }
+
+            llama_token_data_array candidates_p = { candidates.data(), candidates.size(), false };
+
+            const int   top_k = 40;
+            const float top_p = 0.9f;
+            const float temp  = 0.4f;
+
+            llama_sample_top_k(ctx, &candidates_p, top_k, 1);
+            llama_sample_top_p(ctx, &candidates_p, top_p, 1);
+            llama_sample_temp (ctx, &candidates_p, temp);
+
+            const llama_token new_token_id = llama_sample_token(ctx, &candidates_p);
+
+            //const llama_token new_token_id = llama_sample_token_greedy(ctx, &candidates_p);
+
+            // is it an end of stream? -> mark the stream as finished
+            if (new_token_id == llama_token_eos(ctx) || n_cur == n_len) {
+                i_batch[i] = -1;
+                printf("\n");
+                if (n_parallel > 1) {
+                    printf("%s: stream %d finished at n_cur = %d", __func__, i, n_cur);
+                }
+
+                continue;
+            }
+
+            // if there is only one stream, we print immediately to stdout
+            if (n_parallel == 1) {
+                printf("%s", token_to_piece(ctx, new_token_id).c_str());
+                fflush(stdout);
+            }
+
+            streams[i] += token_to_piece(ctx, new_token_id);
+
+            i_batch[i] = batch.n_tokens;
+
+            // push this new token for next evaluation
+            batch_add(batch, new_token_id, n_cur, { i }, true);
+
+            n_decode += 1;
+        }
+
+        // all streams are finished
+        if (batch.n_tokens == 0) {
+            break;
+        }
+
+        n_cur += 1;
+
+        // evaluate the current batch with the transformer model
+        if (llama_decode(ctx, batch)) {
+            fprintf(stderr, "%s : failed to eval, return code %d\n", __func__, 1);
+            return 1;
+        }
+    }
+
+    printf("\n");
+
+    if (n_parallel > 1) {
+        printf("\n");
+
+        for (int32_t i = 0; i < n_parallel; ++i) {
+            printf("sequence %d:\n\n%s%s\n\n", i, prompt, streams[i].c_str());
+        }
+    }
+
+    const auto t_main_end = ggml_time_us();
+
+    printf("%s: decoded %d tokens in %.2f s, speed: %.2f t/s\n",
+            __func__, n_decode, (t_main_end - t_main_start) / 1000000.0f, n_decode / ((t_main_end - t_main_start) / 1000000.0f));
+
+    llama_print_timings(ctx);
+
+    fprintf(stderr, "\n");
+
+    llama_batch_free(batch);
+
+    llama_free(ctx);
+    llama_free_model(model);
+
+    llama_backend_free();
+
+    return 0;
+}

runner/main.go

@@ -0,0 +1,32 @@
+package main
+
+/*
+#cgo CFLAGS: -Ofast -std=c11 -fPIC -Wno-deprecated-declarations -Wno-unused-but-set-variable
+#cgo CPPFLAGS: -Ofast -Wall -Wextra -Wno-unused-function -Wno-unused-variable -Wno-deprecated-declarations -Wno-unused-but-set-variable -DNDEBUG -DGGML_USE_K_QUANTS
+#cgo CXXFLAGS: -std=c++11 -fPIC
+#cgo darwin CPPFLAGS:  -DGGML_USE_ACCELERATE
+#cgo darwin,arm64 CPPFLAGS: -DGGML_USE_METAL
+#cgo darwin LDFLAGS: -framework Accelerate -framework Foundation -framework Metal -framework MetalKit -framework MetalPerformanceShaders
+#include <stdlib.h>
+#include "main.h"
+*/
+import "C"
+import (
+	"fmt"
+	"os"
+	"unsafe"
+)
+
+func main() {
+	if len(os.Args) < 2 {
+		fmt.Println("No prompt provided")
+		return
+	}
+
+	prompt := C.CString(os.Args[1])
+	defer C.free(unsafe.Pointer(prompt))
+	model := C.CString("./model.bin")
+	defer C.free(unsafe.Pointer(model))
+
+	C.generate(model, prompt)
+}

runner/main.h

@@ -0,0 +1,14 @@
+#ifndef MAIN_H
+#define MAIN_H
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+int generate(const char *model_path, const char *prompt);
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif // MAIN_H

runner/unicode.h

@@ -0,0 +1,488 @@
+/**
+ * llama.cpp - git 465219b9143ac01db0990bbcb0a081ef72ec2008
+ *
+ * MIT License
+ *
+ * Copyright (c) 2023 Georgi Gerganov
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to deal
+ * in the Software without restriction, including without limitation the rights
+ * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+ * copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in all
+ * copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#pragma once
+
+#include <cassert>
+#include <stdexcept>
+#include <vector>
+#include <unordered_map>
+
+static const std::vector<std::pair<uint32_t, uint32_t>> digit_ranges = {
+{0x30, 0x39}, {0xB2, 0xB3}, {0xB9, 0xB9}, {0x660, 0x669}, {0x6F0, 0x6F9}, {0x7C0, 0x7C9}, {0x966, 0x96F}, {0x9E6, 0x9EF}, {0xA66, 0xA6F}, {0xAE6, 0xAEF}, {0xB66, 0xB6F}, {0xBE6, 0xBEF}, {0xC66, 0xC6F},
+{0xCE6, 0xCEF}, {0xD66, 0xD6F}, {0xDE6, 0xDEF}, {0xE50, 0xE59}, {0xED0, 0xED9}, {0xF20, 0xF29}, {0x1040, 0x1049}, {0x1090, 0x1099}, {0x1369, 0x1371}, {0x17E0, 0x17E9}, {0x1810, 0x1819}, {0x1946, 0x194F},
+{0x19D0, 0x19DA}, {0x1A80, 0x1A89}, {0x1A90, 0x1A99}, {0x1B50, 0x1B59}, {0x1BB0, 0x1BB9}, {0x1C40, 0x1C49}, {0x1C50, 0x1C59}, {0x2070, 0x2070}, {0x2074, 0x2079}, {0x2080, 0x2089}, {0x2460, 0x2468},
+{0x2474, 0x247C}, {0x2488, 0x2490}, {0x24EA, 0x24EA}, {0x24F5, 0x24FD}, {0x24FF, 0x24FF}, {0x2776, 0x277E}, {0x2780, 0x2788}, {0x278A, 0x2792}, {0xA620, 0xA629}, {0xA8D0, 0xA8D9}, {0xA900, 0xA909},
+{0xA9D0, 0xA9D9}, {0xA9F0, 0xA9F9}, {0xAA50, 0xAA59}, {0xABF0, 0xABF9}, {0xFF10, 0xFF19}, {0x104A0, 0x104A9}, {0x10A40, 0x10A43}, {0x10D30, 0x10D39}, {0x10E60, 0x10E68}, {0x11052, 0x1105A},
+{0x11066, 0x1106F}, {0x110F0, 0x110F9}, {0x11136, 0x1113F}, {0x111D0, 0x111D9}, {0x112F0, 0x112F9}, {0x11450, 0x11459}, {0x114D0, 0x114D9}, {0x11650, 0x11659}, {0x116C0, 0x116C9}, {0x11730, 0x11739},
+{0x118E0, 0x118E9}, {0x11950, 0x11959}, {0x11C50, 0x11C59}, {0x11D50, 0x11D59}, {0x11DA0, 0x11DA9}, {0x16A60, 0x16A69}, {0x16B50, 0x16B59}, {0x1D7CE, 0x1D7FF}, {0x1E140, 0x1E149}, {0x1E2F0, 0x1E2F9},
+{0x1E950, 0x1E959}, {0x1F100, 0x1F10A}, {0x1FBF0, 0x1FBF9},
+};
+
+static const std::vector<std::pair<uint32_t, uint32_t>> letter_ranges = {
+{0x41, 0x5A}, {0x61, 0x7A}, {0xAA, 0xAA}, {0xB5, 0xB5}, {0xBA, 0xBA}, {0xC0, 0xD6}, {0xD8, 0xF6}, {0xF8, 0x2C1}, {0x2C6, 0x2D1}, {0x2E0, 0x2E4}, {0x2EC, 0x2EC}, {0x2EE, 0x2EE}, {0x370, 0x374},
+{0x376, 0x377}, {0x37A, 0x37D}, {0x37F, 0x37F}, {0x386, 0x386}, {0x388, 0x38A}, {0x38C, 0x38C}, {0x38E, 0x3A1}, {0x3A3, 0x3F5}, {0x3F7, 0x481}, {0x48A, 0x52F}, {0x531, 0x556}, {0x559, 0x559},
+{0x560, 0x588}, {0x5D0, 0x5EA}, {0x5EF, 0x5F2}, {0x620, 0x64A}, {0x66E, 0x66F}, {0x671, 0x6D3}, {0x6D5, 0x6D5}, {0x6E5, 0x6E6}, {0x6EE, 0x6EF}, {0x6FA, 0x6FC}, {0x6FF, 0x6FF}, {0x710, 0x710},
+{0x712, 0x72F}, {0x74D, 0x7A5}, {0x7B1, 0x7B1}, {0x7CA, 0x7EA}, {0x7F4, 0x7F5}, {0x7FA, 0x7FA}, {0x800, 0x815}, {0x81A, 0x81A}, {0x824, 0x824}, {0x828, 0x828}, {0x840, 0x858}, {0x860, 0x86A},
+{0x8A0, 0x8B4}, {0x8B6, 0x8C7}, {0x904, 0x939}, {0x93D, 0x93D}, {0x950, 0x950}, {0x958, 0x961}, {0x971, 0x980}, {0x985, 0x98C}, {0x98F, 0x990}, {0x993, 0x9A8}, {0x9AA, 0x9B0}, {0x9B2, 0x9B2},
+{0x9B6, 0x9B9}, {0x9BD, 0x9BD}, {0x9CE, 0x9CE}, {0x9DC, 0x9DD}, {0x9DF, 0x9E1}, {0x9F0, 0x9F1}, {0x9FC, 0x9FC}, {0xA05, 0xA0A}, {0xA0F, 0xA10}, {0xA13, 0xA28}, {0xA2A, 0xA30}, {0xA32, 0xA33},
+{0xA35, 0xA36}, {0xA38, 0xA39}, {0xA59, 0xA5C}, {0xA5E, 0xA5E}, {0xA72, 0xA74}, {0xA85, 0xA8D}, {0xA8F, 0xA91}, {0xA93, 0xAA8}, {0xAAA, 0xAB0}, {0xAB2, 0xAB3}, {0xAB5, 0xAB9}, {0xABD, 0xABD},
+{0xAD0, 0xAD0}, {0xAE0, 0xAE1}, {0xAF9, 0xAF9}, {0xB05, 0xB0C}, {0xB0F, 0xB10}, {0xB13, 0xB28}, {0xB2A, 0xB30}, {0xB32, 0xB33}, {0xB35, 0xB39}, {0xB3D, 0xB3D}, {0xB5C, 0xB5D}, {0xB5F, 0xB61},
+{0xB71, 0xB71}, {0xB83, 0xB83}, {0xB85, 0xB8A}, {0xB8E, 0xB90}, {0xB92, 0xB95}, {0xB99, 0xB9A}, {0xB9C, 0xB9C}, {0xB9E, 0xB9F}, {0xBA3, 0xBA4}, {0xBA8, 0xBAA}, {0xBAE, 0xBB9}, {0xBD0, 0xBD0},
+{0xC05, 0xC0C}, {0xC0E, 0xC10}, {0xC12, 0xC28}, {0xC2A, 0xC39}, {0xC3D, 0xC3D}, {0xC58, 0xC5A}, {0xC60, 0xC61}, {0xC80, 0xC80}, {0xC85, 0xC8C}, {0xC8E, 0xC90}, {0xC92, 0xCA8}, {0xCAA, 0xCB3},
+{0xCB5, 0xCB9}, {0xCBD, 0xCBD}, {0xCDE, 0xCDE}, {0xCE0, 0xCE1}, {0xCF1, 0xCF2}, {0xD04, 0xD0C}, {0xD0E, 0xD10}, {0xD12, 0xD3A}, {0xD3D, 0xD3D}, {0xD4E, 0xD4E}, {0xD54, 0xD56}, {0xD5F, 0xD61},
+{0xD7A, 0xD7F}, {0xD85, 0xD96}, {0xD9A, 0xDB1}, {0xDB3, 0xDBB}, {0xDBD, 0xDBD}, {0xDC0, 0xDC6}, {0xE01, 0xE30}, {0xE32, 0xE33}, {0xE40, 0xE46}, {0xE81, 0xE82}, {0xE84, 0xE84}, {0xE86, 0xE8A},
+{0xE8C, 0xEA3}, {0xEA5, 0xEA5}, {0xEA7, 0xEB0}, {0xEB2, 0xEB3}, {0xEBD, 0xEBD}, {0xEC0, 0xEC4}, {0xEC6, 0xEC6}, {0xEDC, 0xEDF}, {0xF00, 0xF00}, {0xF40, 0xF47}, {0xF49, 0xF6C}, {0xF88, 0xF8C},
+{0x1000, 0x102A}, {0x103F, 0x103F}, {0x1050, 0x1055}, {0x105A, 0x105D}, {0x1061, 0x1061}, {0x1065, 0x1066}, {0x106E, 0x1070}, {0x1075, 0x1081}, {0x108E, 0x108E}, {0x10A0, 0x10C5}, {0x10C7, 0x10C7},
+{0x10CD, 0x10CD}, {0x10D0, 0x10FA}, {0x10FC, 0x1248}, {0x124A, 0x124D}, {0x1250, 0x1256}, {0x1258, 0x1258}, {0x125A, 0x125D}, {0x1260, 0x1288}, {0x128A, 0x128D}, {0x1290, 0x12B0}, {0x12B2, 0x12B5},
+{0x12B8, 0x12BE}, {0x12C0, 0x12C0}, {0x12C2, 0x12C5}, {0x12C8, 0x12D6}, {0x12D8, 0x1310}, {0x1312, 0x1315}, {0x1318, 0x135A}, {0x1380, 0x138F}, {0x13A0, 0x13F5}, {0x13F8, 0x13FD}, {0x1401, 0x166C},
+{0x166F, 0x167F}, {0x1681, 0x169A}, {0x16A0, 0x16EA}, {0x16F1, 0x16F8}, {0x1700, 0x170C}, {0x170E, 0x1711}, {0x1720, 0x1731}, {0x1740, 0x1751}, {0x1760, 0x176C}, {0x176E, 0x1770}, {0x1780, 0x17B3},
+{0x17D7, 0x17D7}, {0x17DC, 0x17DC}, {0x1820, 0x1878}, {0x1880, 0x1884}, {0x1887, 0x18A8}, {0x18AA, 0x18AA}, {0x18B0, 0x18F5}, {0x1900, 0x191E}, {0x1950, 0x196D}, {0x1970, 0x1974}, {0x1980, 0x19AB},
+{0x19B0, 0x19C9}, {0x1A00, 0x1A16}, {0x1A20, 0x1A54}, {0x1AA7, 0x1AA7}, {0x1B05, 0x1B33}, {0x1B45, 0x1B4B}, {0x1B83, 0x1BA0}, {0x1BAE, 0x1BAF}, {0x1BBA, 0x1BE5}, {0x1C00, 0x1C23}, {0x1C4D, 0x1C4F},
+{0x1C5A, 0x1C7D}, {0x1C80, 0x1C88}, {0x1C90, 0x1CBA}, {0x1CBD, 0x1CBF}, {0x1CE9, 0x1CEC}, {0x1CEE, 0x1CF3}, {0x1CF5, 0x1CF6}, {0x1CFA, 0x1CFA}, {0x1D00, 0x1DBF}, {0x1E00, 0x1F15}, {0x1F18, 0x1F1D},
+{0x1F20, 0x1F45}, {0x1F48, 0x1F4D}, {0x1F50, 0x1F57}, {0x1F59, 0x1F59}, {0x1F5B, 0x1F5B}, {0x1F5D, 0x1F5D}, {0x1F5F, 0x1F7D}, {0x1F80, 0x1FB4}, {0x1FB6, 0x1FBC}, {0x1FBE, 0x1FBE}, {0x1FC2, 0x1FC4},
+{0x1FC6, 0x1FCC}, {0x1FD0, 0x1FD3}, {0x1FD6, 0x1FDB}, {0x1FE0, 0x1FEC}, {0x1FF2, 0x1FF4}, {0x1FF6, 0x1FFC}, {0x2071, 0x2071}, {0x207F, 0x207F}, {0x2090, 0x209C}, {0x2102, 0x2102}, {0x2107, 0x2107},
+{0x210A, 0x2113}, {0x2115, 0x2115}, {0x2119, 0x211D}, {0x2124, 0x2124}, {0x2126, 0x2126}, {0x2128, 0x2128}, {0x212A, 0x212D}, {0x212F, 0x2139}, {0x213C, 0x213F}, {0x2145, 0x2149}, {0x214E, 0x214E},
+{0x2183, 0x2184}, {0x2C00, 0x2C2E}, {0x2C30, 0x2C5E}, {0x2C60, 0x2CE4}, {0x2CEB, 0x2CEE}, {0x2CF2, 0x2CF3}, {0x2D00, 0x2D25}, {0x2D27, 0x2D27}, {0x2D2D, 0x2D2D}, {0x2D30, 0x2D67}, {0x2D6F, 0x2D6F},
+{0x2D80, 0x2D96}, {0x2DA0, 0x2DA6}, {0x2DA8, 0x2DAE}, {0x2DB0, 0x2DB6}, {0x2DB8, 0x2DBE}, {0x2DC0, 0x2DC6}, {0x2DC8, 0x2DCE}, {0x2DD0, 0x2DD6}, {0x2DD8, 0x2DDE}, {0x2E2F, 0x2E2F}, {0x3005, 0x3006},
+{0x3031, 0x3035}, {0x303B, 0x303C}, {0x3041, 0x3096}, {0x309D, 0x309F}, {0x30A1, 0x30FA}, {0x30FC, 0x30FF}, {0x3105, 0x312F}, {0x3131, 0x318E}, {0x31A0, 0x31BF}, {0x31F0, 0x31FF}, {0x3400, 0x4DBF},
+{0x4E00, 0x9FFC}, {0xA000, 0xA48C}, {0xA4D0, 0xA4FD}, {0xA500, 0xA60C}, {0xA610, 0xA61F}, {0xA62A, 0xA62B}, {0xA640, 0xA66E}, {0xA67F, 0xA69D}, {0xA6A0, 0xA6E5}, {0xA717, 0xA71F}, {0xA722, 0xA788},
+{0xA78B, 0xA7BF}, {0xA7C2, 0xA7CA}, {0xA7F5, 0xA801}, {0xA803, 0xA805}, {0xA807, 0xA80A}, {0xA80C, 0xA822}, {0xA840, 0xA873}, {0xA882, 0xA8B3}, {0xA8F2, 0xA8F7}, {0xA8FB, 0xA8FB}, {0xA8FD, 0xA8FE},
+{0xA90A, 0xA925}, {0xA930, 0xA946}, {0xA960, 0xA97C}, {0xA984, 0xA9B2}, {0xA9CF, 0xA9CF}, {0xA9E0, 0xA9E4}, {0xA9E6, 0xA9EF}, {0xA9FA, 0xA9FE}, {0xAA00, 0xAA28}, {0xAA40, 0xAA42}, {0xAA44, 0xAA4B},
+{0xAA60, 0xAA76}, {0xAA7A, 0xAA7A}, {0xAA7E, 0xAAAF}, {0xAAB1, 0xAAB1}, {0xAAB5, 0xAAB6}, {0xAAB9, 0xAABD}, {0xAAC0, 0xAAC0}, {0xAAC2, 0xAAC2}, {0xAADB, 0xAADD}, {0xAAE0, 0xAAEA}, {0xAAF2, 0xAAF4},
+{0xAB01, 0xAB06}, {0xAB09, 0xAB0E}, {0xAB11, 0xAB16}, {0xAB20, 0xAB26}, {0xAB28, 0xAB2E}, {0xAB30, 0xAB5A}, {0xAB5C, 0xAB69}, {0xAB70, 0xABE2}, {0xAC00, 0xD7A3}, {0xD7B0, 0xD7C6}, {0xD7CB, 0xD7FB},
+{0xF900, 0xFA6D}, {0xFA70, 0xFAD9}, {0xFB00, 0xFB06}, {0xFB13, 0xFB17}, {0xFB1D, 0xFB1D}, {0xFB1F, 0xFB28}, {0xFB2A, 0xFB36}, {0xFB38, 0xFB3C}, {0xFB3E, 0xFB3E}, {0xFB40, 0xFB41}, {0xFB43, 0xFB44},
+{0xFB46, 0xFBB1}, {0xFBD3, 0xFD3D}, {0xFD50, 0xFD8F}, {0xFD92, 0xFDC7}, {0xFDF0, 0xFDFB}, {0xFE70, 0xFE74}, {0xFE76, 0xFEFC}, {0xFF21, 0xFF3A}, {0xFF41, 0xFF5A}, {0xFF66, 0xFFBE}, {0xFFC2, 0xFFC7},
+{0xFFCA, 0xFFCF}, {0xFFD2, 0xFFD7}, {0xFFDA, 0xFFDC}, {0x10000, 0x1000B}, {0x1000D, 0x10026}, {0x10028, 0x1003A}, {0x1003C, 0x1003D}, {0x1003F, 0x1004D}, {0x10050, 0x1005D}, {0x10080, 0x100FA},
+{0x10280, 0x1029C}, {0x102A0, 0x102D0}, {0x10300, 0x1031F}, {0x1032D, 0x10340}, {0x10342, 0x10349}, {0x10350, 0x10375}, {0x10380, 0x1039D}, {0x103A0, 0x103C3}, {0x103C8, 0x103CF}, {0x10400, 0x1049D},
+{0x104B0, 0x104D3}, {0x104D8, 0x104FB}, {0x10500, 0x10527}, {0x10530, 0x10563}, {0x10600, 0x10736}, {0x10740, 0x10755}, {0x10760, 0x10767}, {0x10800, 0x10805}, {0x10808, 0x10808}, {0x1080A, 0x10835},
+{0x10837, 0x10838}, {0x1083C, 0x1083C}, {0x1083F, 0x10855}, {0x10860, 0x10876}, {0x10880, 0x1089E}, {0x108E0, 0x108F2}, {0x108F4, 0x108F5}, {0x10900, 0x10915}, {0x10920, 0x10939}, {0x10980, 0x109B7},
+{0x109BE, 0x109BF}, {0x10A00, 0x10A00}, {0x10A10, 0x10A13}, {0x10A15, 0x10A17}, {0x10A19, 0x10A35}, {0x10A60, 0x10A7C}, {0x10A80, 0x10A9C}, {0x10AC0, 0x10AC7}, {0x10AC9, 0x10AE4}, {0x10B00, 0x10B35},
+{0x10B40, 0x10B55}, {0x10B60, 0x10B72}, {0x10B80, 0x10B91}, {0x10C00, 0x10C48}, {0x10C80, 0x10CB2}, {0x10CC0, 0x10CF2}, {0x10D00, 0x10D23}, {0x10E80, 0x10EA9}, {0x10EB0, 0x10EB1}, {0x10F00, 0x10F1C},
+{0x10F27, 0x10F27}, {0x10F30, 0x10F45}, {0x10FB0, 0x10FC4}, {0x10FE0, 0x10FF6}, {0x11003, 0x11037}, {0x11083, 0x110AF}, {0x110D0, 0x110E8}, {0x11103, 0x11126}, {0x11144, 0x11144}, {0x11147, 0x11147},
+{0x11150, 0x11172}, {0x11176, 0x11176}, {0x11183, 0x111B2}, {0x111C1, 0x111C4}, {0x111DA, 0x111DA}, {0x111DC, 0x111DC}, {0x11200, 0x11211}, {0x11213, 0x1122B}, {0x11280, 0x11286}, {0x11288, 0x11288},
+{0x1128A, 0x1128D}, {0x1128F, 0x1129D}, {0x1129F, 0x112A8}, {0x112B0, 0x112DE}, {0x11305, 0x1130C}, {0x1130F, 0x11310}, {0x11313, 0x11328}, {0x1132A, 0x11330}, {0x11332, 0x11333}, {0x11335, 0x11339},
+{0x1133D, 0x1133D}, {0x11350, 0x11350}, {0x1135D, 0x11361}, {0x11400, 0x11434}, {0x11447, 0x1144A}, {0x1145F, 0x11461}, {0x11480, 0x114AF}, {0x114C4, 0x114C5}, {0x114C7, 0x114C7}, {0x11580, 0x115AE},
+{0x115D8, 0x115DB}, {0x11600, 0x1162F}, {0x11644, 0x11644}, {0x11680, 0x116AA}, {0x116B8, 0x116B8}, {0x11700, 0x1171A}, {0x11800, 0x1182B}, {0x118A0, 0x118DF}, {0x118FF, 0x11906}, {0x11909, 0x11909},
+{0x1190C, 0x11913}, {0x11915, 0x11916}, {0x11918, 0x1192F}, {0x1193F, 0x1193F}, {0x11941, 0x11941}, {0x119A0, 0x119A7}, {0x119AA, 0x119D0}, {0x119E1, 0x119E1}, {0x119E3, 0x119E3}, {0x11A00, 0x11A00},
+{0x11A0B, 0x11A32}, {0x11A3A, 0x11A3A}, {0x11A50, 0x11A50}, {0x11A5C, 0x11A89}, {0x11A9D, 0x11A9D}, {0x11AC0, 0x11AF8}, {0x11C00, 0x11C08}, {0x11C0A, 0x11C2E}, {0x11C40, 0x11C40}, {0x11C72, 0x11C8F},
+{0x11D00, 0x11D06}, {0x11D08, 0x11D09}, {0x11D0B, 0x11D30}, {0x11D46, 0x11D46}, {0x11D60, 0x11D65}, {0x11D67, 0x11D68}, {0x11D6A, 0x11D89}, {0x11D98, 0x11D98}, {0x11EE0, 0x11EF2}, {0x11FB0, 0x11FB0},
+{0x12000, 0x12399}, {0x12480, 0x12543}, {0x13000, 0x1342E}, {0x14400, 0x14646}, {0x16800, 0x16A38}, {0x16A40, 0x16A5E}, {0x16AD0, 0x16AED}, {0x16B00, 0x16B2F}, {0x16B40, 0x16B43}, {0x16B63, 0x16B77},
+{0x16B7D, 0x16B8F}, {0x16E40, 0x16E7F}, {0x16F00, 0x16F4A}, {0x16F50, 0x16F50}, {0x16F93, 0x16F9F}, {0x16FE0, 0x16FE1}, {0x16FE3, 0x16FE3}, {0x17000, 0x187F7}, {0x18800, 0x18CD5}, {0x18D00, 0x18D08},
+{0x1B000, 0x1B11E}, {0x1B150, 0x1B152}, {0x1B164, 0x1B167}, {0x1B170, 0x1B2FB}, {0x1BC00, 0x1BC6A}, {0x1BC70, 0x1BC7C}, {0x1BC80, 0x1BC88}, {0x1BC90, 0x1BC99}, {0x1D400, 0x1D454}, {0x1D456, 0x1D49C},
+{0x1D49E, 0x1D49F}, {0x1D4A2, 0x1D4A2}, {0x1D4A5, 0x1D4A6}, {0x1D4A9, 0x1D4AC}, {0x1D4AE, 0x1D4B9}, {0x1D4BB, 0x1D4BB}, {0x1D4BD, 0x1D4C3}, {0x1D4C5, 0x1D505}, {0x1D507, 0x1D50A}, {0x1D50D, 0x1D514},
+{0x1D516, 0x1D51C}, {0x1D51E, 0x1D539}, {0x1D53B, 0x1D53E}, {0x1D540, 0x1D544}, {0x1D546, 0x1D546}, {0x1D54A, 0x1D550}, {0x1D552, 0x1D6A5}, {0x1D6A8, 0x1D6C0}, {0x1D6C2, 0x1D6DA}, {0x1D6DC, 0x1D6FA},
+{0x1D6FC, 0x1D714}, {0x1D716, 0x1D734}, {0x1D736, 0x1D74E}, {0x1D750, 0x1D76E}, {0x1D770, 0x1D788}, {0x1D78A, 0x1D7A8}, {0x1D7AA, 0x1D7C2}, {0x1D7C4, 0x1D7CB}, {0x1E100, 0x1E12C}, {0x1E137, 0x1E13D},
+{0x1E14E, 0x1E14E}, {0x1E2C0, 0x1E2EB}, {0x1E800, 0x1E8C4}, {0x1E900, 0x1E943}, {0x1E94B, 0x1E94B}, {0x1EE00, 0x1EE03}, {0x1EE05, 0x1EE1F}, {0x1EE21, 0x1EE22}, {0x1EE24, 0x1EE24}, {0x1EE27, 0x1EE27},
+{0x1EE29, 0x1EE32}, {0x1EE34, 0x1EE37}, {0x1EE39, 0x1EE39}, {0x1EE3B, 0x1EE3B}, {0x1EE42, 0x1EE42}, {0x1EE47, 0x1EE47}, {0x1EE49, 0x1EE49}, {0x1EE4B, 0x1EE4B}, {0x1EE4D, 0x1EE4F}, {0x1EE51, 0x1EE52},
+{0x1EE54, 0x1EE54}, {0x1EE57, 0x1EE57}, {0x1EE59, 0x1EE59}, {0x1EE5B, 0x1EE5B}, {0x1EE5D, 0x1EE5D}, {0x1EE5F, 0x1EE5F}, {0x1EE61, 0x1EE62}, {0x1EE64, 0x1EE64}, {0x1EE67, 0x1EE6A}, {0x1EE6C, 0x1EE72},
+{0x1EE74, 0x1EE77}, {0x1EE79, 0x1EE7C}, {0x1EE7E, 0x1EE7E}, {0x1EE80, 0x1EE89}, {0x1EE8B, 0x1EE9B}, {0x1EEA1, 0x1EEA3}, {0x1EEA5, 0x1EEA9}, {0x1EEAB, 0x1EEBB}, {0x20000, 0x2A6DD}, {0x2A700, 0x2B734},
+{0x2B740, 0x2B81D}, {0x2B820, 0x2CEA1}, {0x2CEB0, 0x2EBE0}, {0x2F800, 0x2FA1D}, {0x30000, 0x3134A},
+};
+
+static const std::vector<std::pair<uint32_t, uint32_t>> whitespace_ranges = {
+{0x9, 0xD}, {0x1C, 0x20}, {0x85, 0x85}, {0xA0, 0xA0}, {0x1680, 0x1680}, {0x2000, 0x200A}, {0x2028, 0x2029}, {0x202F, 0x202F}, {0x205F, 0x205F}, {0x3000, 0x3000},
+};
+
+static const std::vector<std::pair<uint32_t, uint32_t>> accent_mark_ranges = {
+{0x300, 0x36F}, {0x483, 0x489}, {0x591, 0x5BD}, {0x5BF, 0x5BF}, {0x5C1, 0x5C2}, {0x5C4, 0x5C5}, {0x5C7, 0x5C7}, {0x610, 0x61A}, {0x64B, 0x65F}, {0x670, 0x670}, {0x6D6, 0x6DC}, {0x6DF, 0x6E4},
+{0x6E7, 0x6E8}, {0x6EA, 0x6ED}, {0x711, 0x711}, {0x730, 0x74A}, {0x7A6, 0x7B0}, {0x7EB, 0x7F3}, {0x7FD, 0x7FD}, {0x816, 0x819}, {0x81B, 0x823}, {0x825, 0x827}, {0x829, 0x82D}, {0x859, 0x85B},
+{0x8D3, 0x8E1}, {0x8E3, 0x903}, {0x93A, 0x93C}, {0x93E, 0x94F}, {0x951, 0x957}, {0x962, 0x963}, {0x981, 0x983}, {0x9BC, 0x9BC}, {0x9BE, 0x9C4}, {0x9C7, 0x9C8}, {0x9CB, 0x9CD}, {0x9D7, 0x9D7},
+{0x9E2, 0x9E3}, {0x9FE, 0x9FE}, {0xA01, 0xA03}, {0xA3C, 0xA3C}, {0xA3E, 0xA42}, {0xA47, 0xA48}, {0xA4B, 0xA4D}, {0xA51, 0xA51}, {0xA70, 0xA71}, {0xA75, 0xA75}, {0xA81, 0xA83}, {0xABC, 0xABC},
+{0xABE, 0xAC5}, {0xAC7, 0xAC9}, {0xACB, 0xACD}, {0xAE2, 0xAE3}, {0xAFA, 0xAFF}, {0xB01, 0xB03}, {0xB3C, 0xB3C}, {0xB3E, 0xB44}, {0xB47, 0xB48}, {0xB4B, 0xB4D}, {0xB55, 0xB57}, {0xB62, 0xB63},
+{0xB82, 0xB82}, {0xBBE, 0xBC2}, {0xBC6, 0xBC8}, {0xBCA, 0xBCD}, {0xBD7, 0xBD7}, {0xC00, 0xC04}, {0xC3E, 0xC44}, {0xC46, 0xC48}, {0xC4A, 0xC4D}, {0xC55, 0xC56}, {0xC62, 0xC63}, {0xC81, 0xC83},
+{0xCBC, 0xCBC}, {0xCBE, 0xCC4}, {0xCC6, 0xCC8}, {0xCCA, 0xCCD}, {0xCD5, 0xCD6}, {0xCE2, 0xCE3}, {0xD00, 0xD03}, {0xD3B, 0xD3C}, {0xD3E, 0xD44}, {0xD46, 0xD48}, {0xD4A, 0xD4D}, {0xD57, 0xD57},
+{0xD62, 0xD63}, {0xD81, 0xD83}, {0xDCA, 0xDCA}, {0xDCF, 0xDD4}, {0xDD6, 0xDD6}, {0xDD8, 0xDDF}, {0xDF2, 0xDF3}, {0xE31, 0xE31}, {0xE34, 0xE3A}, {0xE47, 0xE4E}, {0xEB1, 0xEB1}, {0xEB4, 0xEBC},
+{0xEC8, 0xECD}, {0xF18, 0xF19}, {0xF35, 0xF35}, {0xF37, 0xF37}, {0xF39, 0xF39}, {0xF3E, 0xF3F}, {0xF71, 0xF84}, {0xF86, 0xF87}, {0xF8D, 0xF97}, {0xF99, 0xFBC}, {0xFC6, 0xFC6}, {0x102B, 0x103E},
+{0x1056, 0x1059}, {0x105E, 0x1060}, {0x1062, 0x1064}, {0x1067, 0x106D}, {0x1071, 0x1074}, {0x1082, 0x108D}, {0x108F, 0x108F}, {0x109A, 0x109D}, {0x135D, 0x135F}, {0x1712, 0x1714}, {0x1732, 0x1734},
+{0x1752, 0x1753}, {0x1772, 0x1773}, {0x17B4, 0x17D3}, {0x17DD, 0x17DD}, {0x180B, 0x180D}, {0x1885, 0x1886}, {0x18A9, 0x18A9}, {0x1920, 0x192B}, {0x1930, 0x193B}, {0x1A17, 0x1A1B}, {0x1A55, 0x1A5E},
+{0x1A60, 0x1A7C}, {0x1A7F, 0x1A7F}, {0x1AB0, 0x1AC0}, {0x1B00, 0x1B04}, {0x1B34, 0x1B44}, {0x1B6B, 0x1B73}, {0x1B80, 0x1B82}, {0x1BA1, 0x1BAD}, {0x1BE6, 0x1BF3}, {0x1C24, 0x1C37}, {0x1CD0, 0x1CD2},
+{0x1CD4, 0x1CE8}, {0x1CED, 0x1CED}, {0x1CF4, 0x1CF4}, {0x1CF7, 0x1CF9}, {0x1DC0, 0x1DF9}, {0x1DFB, 0x1DFF}, {0x20D0, 0x20F0}, {0x2CEF, 0x2CF1}, {0x2D7F, 0x2D7F}, {0x2DE0, 0x2DFF}, {0x302A, 0x302F},
+{0x3099, 0x309A}, {0xA66F, 0xA672}, {0xA674, 0xA67D}, {0xA69E, 0xA69F}, {0xA6F0, 0xA6F1}, {0xA802, 0xA802}, {0xA806, 0xA806}, {0xA80B, 0xA80B}, {0xA823, 0xA827}, {0xA82C, 0xA82C}, {0xA880, 0xA881},
+{0xA8B4, 0xA8C5}, {0xA8E0, 0xA8F1}, {0xA8FF, 0xA8FF}, {0xA926, 0xA92D}, {0xA947, 0xA953}, {0xA980, 0xA983}, {0xA9B3, 0xA9C0}, {0xA9E5, 0xA9E5}, {0xAA29, 0xAA36}, {0xAA43, 0xAA43}, {0xAA4C, 0xAA4D},
+{0xAA7B, 0xAA7D}, {0xAAB0, 0xAAB0}, {0xAAB2, 0xAAB4}, {0xAAB7, 0xAAB8}, {0xAABE, 0xAABF}, {0xAAC1, 0xAAC1}, {0xAAEB, 0xAAEF}, {0xAAF5, 0xAAF6}, {0xABE3, 0xABEA}, {0xABEC, 0xABED}, {0xFB1E, 0xFB1E},
+{0xFE00, 0xFE0F}, {0xFE20, 0xFE2F}, {0x101FD, 0x101FD}, {0x102E0, 0x102E0}, {0x10376, 0x1037A}, {0x10A01, 0x10A03}, {0x10A05, 0x10A06}, {0x10A0C, 0x10A0F}, {0x10A38, 0x10A3A}, {0x10A3F, 0x10A3F},
+{0x10AE5, 0x10AE6}, {0x10D24, 0x10D27}, {0x10EAB, 0x10EAC}, {0x10F46, 0x10F50}, {0x11000, 0x11002}, {0x11038, 0x11046}, {0x1107F, 0x11082}, {0x110B0, 0x110BA}, {0x11100, 0x11102}, {0x11127, 0x11134},
+{0x11145, 0x11146}, {0x11173, 0x11173}, {0x11180, 0x11182}, {0x111B3, 0x111C0}, {0x111C9, 0x111CC}, {0x111CE, 0x111CF}, {0x1122C, 0x11237}, {0x1123E, 0x1123E}, {0x112DF, 0x112EA}, {0x11300, 0x11303},
+{0x1133B, 0x1133C}, {0x1133E, 0x11344}, {0x11347, 0x11348}, {0x1134B, 0x1134D}, {0x11357, 0x11357}, {0x11362, 0x11363}, {0x11366, 0x1136C}, {0x11370, 0x11374}, {0x11435, 0x11446}, {0x1145E, 0x1145E},
+{0x114B0, 0x114C3}, {0x115AF, 0x115B5}, {0x115B8, 0x115C0}, {0x115DC, 0x115DD}, {0x11630, 0x11640}, {0x116AB, 0x116B7}, {0x1171D, 0x1172B}, {0x1182C, 0x1183A}, {0x11930, 0x11935}, {0x11937, 0x11938},
+{0x1193B, 0x1193E}, {0x11940, 0x11940}, {0x11942, 0x11943}, {0x119D1, 0x119D7}, {0x119DA, 0x119E0}, {0x119E4, 0x119E4}, {0x11A01, 0x11A0A}, {0x11A33, 0x11A39}, {0x11A3B, 0x11A3E}, {0x11A47, 0x11A47},
+{0x11A51, 0x11A5B}, {0x11A8A, 0x11A99}, {0x11C2F, 0x11C36}, {0x11C38, 0x11C3F}, {0x11C92, 0x11CA7}, {0x11CA9, 0x11CB6}, {0x11D31, 0x11D36}, {0x11D3A, 0x11D3A}, {0x11D3C, 0x11D3D}, {0x11D3F, 0x11D45},
+{0x11D47, 0x11D47}, {0x11D8A, 0x11D8E}, {0x11D90, 0x11D91}, {0x11D93, 0x11D97}, {0x11EF3, 0x11EF6}, {0x16AF0, 0x16AF4}, {0x16B30, 0x16B36}, {0x16F4F, 0x16F4F}, {0x16F51, 0x16F87}, {0x16F8F, 0x16F92},
+{0x16FE4, 0x16FE4}, {0x16FF0, 0x16FF1}, {0x1BC9D, 0x1BC9E}, {0x1D165, 0x1D169}, {0x1D16D, 0x1D172}, {0x1D17B, 0x1D182}, {0x1D185, 0x1D18B}, {0x1D1AA, 0x1D1AD}, {0x1D242, 0x1D244}, {0x1DA00, 0x1DA36},
+{0x1DA3B, 0x1DA6C}, {0x1DA75, 0x1DA75}, {0x1DA84, 0x1DA84}, {0x1DA9B, 0x1DA9F}, {0x1DAA1, 0x1DAAF}, {0x1E000, 0x1E006}, {0x1E008, 0x1E018}, {0x1E01B, 0x1E021}, {0x1E023, 0x1E024}, {0x1E026, 0x1E02A},
+{0x1E130, 0x1E136}, {0x1E2EC, 0x1E2EF}, {0x1E8D0, 0x1E8D6}, {0x1E944, 0x1E94A}, {0xE0100, 0xE01EF},
+};
+
+static const std::vector<std::pair<uint32_t, uint32_t>> punctuation_ranges = {
+{0x21, 0x23}, {0x25, 0x2A}, {0x2C, 0x2F}, {0x3A, 0x3B}, {0x3F, 0x40}, {0x5B, 0x5D}, {0x5F, 0x5F}, {0x7B, 0x7B}, {0x7D, 0x7D}, {0xA1, 0xA1}, {0xA7, 0xA7}, {0xAB, 0xAB}, {0xB6, 0xB7}, {0xBB, 0xBB},
+{0xBF, 0xBF}, {0x37E, 0x37E}, {0x387, 0x387}, {0x55A, 0x55F}, {0x589, 0x58A}, {0x5BE, 0x5BE}, {0x5C0, 0x5C0}, {0x5C3, 0x5C3}, {0x5C6, 0x5C6}, {0x5F3, 0x5F4}, {0x609, 0x60A}, {0x60C, 0x60D},
+{0x61B, 0x61B}, {0x61E, 0x61F}, {0x66A, 0x66D}, {0x6D4, 0x6D4}, {0x700, 0x70D}, {0x7F7, 0x7F9}, {0x830, 0x83E}, {0x85E, 0x85E}, {0x964, 0x965}, {0x970, 0x970}, {0x9FD, 0x9FD}, {0xA76, 0xA76},
+{0xAF0, 0xAF0}, {0xC77, 0xC77}, {0xC84, 0xC84}, {0xDF4, 0xDF4}, {0xE4F, 0xE4F}, {0xE5A, 0xE5B}, {0xF04, 0xF12}, {0xF14, 0xF14}, {0xF3A, 0xF3D}, {0xF85, 0xF85}, {0xFD0, 0xFD4}, {0xFD9, 0xFDA},
+{0x104A, 0x104F}, {0x10FB, 0x10FB}, {0x1360, 0x1368}, {0x1400, 0x1400}, {0x166E, 0x166E}, {0x169B, 0x169C}, {0x16EB, 0x16ED}, {0x1735, 0x1736}, {0x17D4, 0x17D6}, {0x17D8, 0x17DA}, {0x1800, 0x180A},
+{0x1944, 0x1945}, {0x1A1E, 0x1A1F}, {0x1AA0, 0x1AA6}, {0x1AA8, 0x1AAD}, {0x1B5A, 0x1B60}, {0x1BFC, 0x1BFF}, {0x1C3B, 0x1C3F}, {0x1C7E, 0x1C7F}, {0x1CC0, 0x1CC7}, {0x1CD3, 0x1CD3}, {0x2010, 0x2027},
+{0x2030, 0x2043}, {0x2045, 0x2051}, {0x2053, 0x205E}, {0x207D, 0x207E}, {0x208D, 0x208E}, {0x2308, 0x230B}, {0x2329, 0x232A}, {0x2768, 0x2775}, {0x27C5, 0x27C6}, {0x27E6, 0x27EF}, {0x2983, 0x2998},
+{0x29D8, 0x29DB}, {0x29FC, 0x29FD}, {0x2CF9, 0x2CFC}, {0x2CFE, 0x2CFF}, {0x2D70, 0x2D70}, {0x2E00, 0x2E2E}, {0x2E30, 0x2E4F}, {0x2E52, 0x2E52}, {0x3001, 0x3003}, {0x3008, 0x3011}, {0x3014, 0x301F},
+{0x3030, 0x3030}, {0x303D, 0x303D}, {0x30A0, 0x30A0}, {0x30FB, 0x30FB}, {0xA4FE, 0xA4FF}, {0xA60D, 0xA60F}, {0xA673, 0xA673}, {0xA67E, 0xA67E}, {0xA6F2, 0xA6F7}, {0xA874, 0xA877}, {0xA8CE, 0xA8CF},
+{0xA8F8, 0xA8FA}, {0xA8FC, 0xA8FC}, {0xA92E, 0xA92F}, {0xA95F, 0xA95F}, {0xA9C1, 0xA9CD}, {0xA9DE, 0xA9DF}, {0xAA5C, 0xAA5F}, {0xAADE, 0xAADF}, {0xAAF0, 0xAAF1}, {0xABEB, 0xABEB}, {0xFD3E, 0xFD3F},
+{0xFE10, 0xFE19}, {0xFE30, 0xFE52}, {0xFE54, 0xFE61}, {0xFE63, 0xFE63}, {0xFE68, 0xFE68}, {0xFE6A, 0xFE6B}, {0xFF01, 0xFF03}, {0xFF05, 0xFF0A}, {0xFF0C, 0xFF0F}, {0xFF1A, 0xFF1B}, {0xFF1F, 0xFF20},
+{0xFF3B, 0xFF3D}, {0xFF3F, 0xFF3F}, {0xFF5B, 0xFF5B}, {0xFF5D, 0xFF5D}, {0xFF5F, 0xFF65}, {0x10100, 0x10102}, {0x1039F, 0x1039F}, {0x103D0, 0x103D0}, {0x1056F, 0x1056F}, {0x10857, 0x10857},
+{0x1091F, 0x1091F}, {0x1093F, 0x1093F}, {0x10A50, 0x10A58}, {0x10A7F, 0x10A7F}, {0x10AF0, 0x10AF6}, {0x10B39, 0x10B3F}, {0x10B99, 0x10B9C}, {0x10EAD, 0x10EAD}, {0x10F55, 0x10F59}, {0x11047, 0x1104D},
+{0x110BB, 0x110BC}, {0x110BE, 0x110C1}, {0x11140, 0x11143}, {0x11174, 0x11175}, {0x111C5, 0x111C8}, {0x111CD, 0x111CD}, {0x111DB, 0x111DB}, {0x111DD, 0x111DF}, {0x11238, 0x1123D}, {0x112A9, 0x112A9},
+{0x1144B, 0x1144F}, {0x1145A, 0x1145B}, {0x1145D, 0x1145D}, {0x114C6, 0x114C6}, {0x115C1, 0x115D7}, {0x11641, 0x11643}, {0x11660, 0x1166C}, {0x1173C, 0x1173E}, {0x1183B, 0x1183B}, {0x11944, 0x11946},
+{0x119E2, 0x119E2}, {0x11A3F, 0x11A46}, {0x11A9A, 0x11A9C}, {0x11A9E, 0x11AA2}, {0x11C41, 0x11C45}, {0x11C70, 0x11C71}, {0x11EF7, 0x11EF8}, {0x11FFF, 0x11FFF}, {0x12470, 0x12474}, {0x16A6E, 0x16A6F},
+{0x16AF5, 0x16AF5}, {0x16B37, 0x16B3B}, {0x16B44, 0x16B44}, {0x16E97, 0x16E9A}, {0x16FE2, 0x16FE2}, {0x1BC9F, 0x1BC9F}, {0x1DA87, 0x1DA8B}, {0x1E95E, 0x1E95F},
+};
+
+static const std::vector<std::pair<uint32_t, uint32_t>> symbol_ranges = {
+{0x24, 0x24}, {0x2B, 0x2B}, {0x3C, 0x3E}, {0x5E, 0x5E}, {0x60, 0x60}, {0x7C, 0x7C}, {0x7E, 0x7E}, {0xA2, 0xA6}, {0xA8, 0xA9}, {0xAC, 0xAC}, {0xAE, 0xB1}, {0xB4, 0xB4}, {0xB8, 0xB8}, {0xD7, 0xD7},
+{0xF7, 0xF7}, {0x2C2, 0x2C5}, {0x2D2, 0x2DF}, {0x2E5, 0x2EB}, {0x2ED, 0x2ED}, {0x2EF, 0x2FF}, {0x375, 0x375}, {0x384, 0x385}, {0x3F6, 0x3F6}, {0x482, 0x482}, {0x58D, 0x58F}, {0x606, 0x608},
+{0x60B, 0x60B}, {0x60E, 0x60F}, {0x6DE, 0x6DE}, {0x6E9, 0x6E9}, {0x6FD, 0x6FE}, {0x7F6, 0x7F6}, {0x7FE, 0x7FF}, {0x9F2, 0x9F3}, {0x9FA, 0x9FB}, {0xAF1, 0xAF1}, {0xB70, 0xB70}, {0xBF3, 0xBFA},
+{0xC7F, 0xC7F}, {0xD4F, 0xD4F}, {0xD79, 0xD79}, {0xE3F, 0xE3F}, {0xF01, 0xF03}, {0xF13, 0xF13}, {0xF15, 0xF17}, {0xF1A, 0xF1F}, {0xF34, 0xF34}, {0xF36, 0xF36}, {0xF38, 0xF38}, {0xFBE, 0xFC5},
+{0xFC7, 0xFCC}, {0xFCE, 0xFCF}, {0xFD5, 0xFD8}, {0x109E, 0x109F}, {0x1390, 0x1399}, {0x166D, 0x166D}, {0x17DB, 0x17DB}, {0x1940, 0x1940}, {0x19DE, 0x19FF}, {0x1B61, 0x1B6A}, {0x1B74, 0x1B7C},
+{0x1FBD, 0x1FBD}, {0x1FBF, 0x1FC1}, {0x1FCD, 0x1FCF}, {0x1FDD, 0x1FDF}, {0x1FED, 0x1FEF}, {0x1FFD, 0x1FFE}, {0x2044, 0x2044}, {0x2052, 0x2052}, {0x207A, 0x207C}, {0x208A, 0x208C}, {0x20A0, 0x20BF},
+{0x2100, 0x2101}, {0x2103, 0x2106}, {0x2108, 0x2109}, {0x2114, 0x2114}, {0x2116, 0x2118}, {0x211E, 0x2123}, {0x2125, 0x2125}, {0x2127, 0x2127}, {0x2129, 0x2129}, {0x212E, 0x212E}, {0x213A, 0x213B},
+{0x2140, 0x2144}, {0x214A, 0x214D}, {0x214F, 0x214F}, {0x218A, 0x218B}, {0x2190, 0x2307}, {0x230C, 0x2328}, {0x232B, 0x2426}, {0x2440, 0x244A}, {0x249C, 0x24E9}, {0x2500, 0x2767}, {0x2794, 0x27C4},
+{0x27C7, 0x27E5}, {0x27F0, 0x2982}, {0x2999, 0x29D7}, {0x29DC, 0x29FB}, {0x29FE, 0x2B73}, {0x2B76, 0x2B95}, {0x2B97, 0x2BFF}, {0x2CE5, 0x2CEA}, {0x2E50, 0x2E51}, {0x2E80, 0x2E99}, {0x2E9B, 0x2EF3},
+{0x2F00, 0x2FD5}, {0x2FF0, 0x2FFB}, {0x3004, 0x3004}, {0x3012, 0x3013}, {0x3020, 0x3020}, {0x3036, 0x3037}, {0x303E, 0x303F}, {0x309B, 0x309C}, {0x3190, 0x3191}, {0x3196, 0x319F}, {0x31C0, 0x31E3},
+{0x3200, 0x321E}, {0x322A, 0x3247}, {0x3250, 0x3250}, {0x3260, 0x327F}, {0x328A, 0x32B0}, {0x32C0, 0x33FF}, {0x4DC0, 0x4DFF}, {0xA490, 0xA4C6}, {0xA700, 0xA716}, {0xA720, 0xA721}, {0xA789, 0xA78A},
+{0xA828, 0xA82B}, {0xA836, 0xA839}, {0xAA77, 0xAA79}, {0xAB5B, 0xAB5B}, {0xAB6A, 0xAB6B}, {0xFB29, 0xFB29}, {0xFBB2, 0xFBC1}, {0xFDFC, 0xFDFD}, {0xFE62, 0xFE62}, {0xFE64, 0xFE66}, {0xFE69, 0xFE69},
+{0xFF04, 0xFF04}, {0xFF0B, 0xFF0B}, {0xFF1C, 0xFF1E}, {0xFF3E, 0xFF3E}, {0xFF40, 0xFF40}, {0xFF5C, 0xFF5C}, {0xFF5E, 0xFF5E}, {0xFFE0, 0xFFE6}, {0xFFE8, 0xFFEE}, {0xFFFC, 0xFFFD}, {0x10137, 0x1013F},
+{0x10179, 0x10189}, {0x1018C, 0x1018E}, {0x10190, 0x1019C}, {0x101A0, 0x101A0}, {0x101D0, 0x101FC}, {0x10877, 0x10878}, {0x10AC8, 0x10AC8}, {0x1173F, 0x1173F}, {0x11FD5, 0x11FF1}, {0x16B3C, 0x16B3F},
+{0x16B45, 0x16B45}, {0x1BC9C, 0x1BC9C}, {0x1D000, 0x1D0F5}, {0x1D100, 0x1D126}, {0x1D129, 0x1D164}, {0x1D16A, 0x1D16C}, {0x1D183, 0x1D184}, {0x1D18C, 0x1D1A9}, {0x1D1AE, 0x1D1E8}, {0x1D200, 0x1D241},
+{0x1D245, 0x1D245}, {0x1D300, 0x1D356}, {0x1D6C1, 0x1D6C1}, {0x1D6DB, 0x1D6DB}, {0x1D6FB, 0x1D6FB}, {0x1D715, 0x1D715}, {0x1D735, 0x1D735}, {0x1D74F, 0x1D74F}, {0x1D76F, 0x1D76F}, {0x1D789, 0x1D789},
+{0x1D7A9, 0x1D7A9}, {0x1D7C3, 0x1D7C3}, {0x1D800, 0x1D9FF}, {0x1DA37, 0x1DA3A}, {0x1DA6D, 0x1DA74}, {0x1DA76, 0x1DA83}, {0x1DA85, 0x1DA86}, {0x1E14F, 0x1E14F}, {0x1E2FF, 0x1E2FF}, {0x1ECAC, 0x1ECAC},
+{0x1ECB0, 0x1ECB0}, {0x1ED2E, 0x1ED2E}, {0x1EEF0, 0x1EEF1}, {0x1F000, 0x1F02B}, {0x1F030, 0x1F093}, {0x1F0A0, 0x1F0AE}, {0x1F0B1, 0x1F0BF}, {0x1F0C1, 0x1F0CF}, {0x1F0D1, 0x1F0F5}, {0x1F10D, 0x1F1AD},
+{0x1F1E6, 0x1F202}, {0x1F210, 0x1F23B}, {0x1F240, 0x1F248}, {0x1F250, 0x1F251}, {0x1F260, 0x1F265}, {0x1F300, 0x1F6D7}, {0x1F6E0, 0x1F6EC}, {0x1F6F0, 0x1F6FC}, {0x1F700, 0x1F773}, {0x1F780, 0x1F7D8},
+{0x1F7E0, 0x1F7EB}, {0x1F800, 0x1F80B}, {0x1F810, 0x1F847}, {0x1F850, 0x1F859}, {0x1F860, 0x1F887}, {0x1F890, 0x1F8AD}, {0x1F8B0, 0x1F8B1}, {0x1F900, 0x1F978}, {0x1F97A, 0x1F9CB}, {0x1F9CD, 0x1FA53},
+{0x1FA60, 0x1FA6D}, {0x1FA70, 0x1FA74}, {0x1FA78, 0x1FA7A}, {0x1FA80, 0x1FA86}, {0x1FA90, 0x1FAA8}, {0x1FAB0, 0x1FAB6}, {0x1FAC0, 0x1FAC2}, {0x1FAD0, 0x1FAD6}, {0x1FB00, 0x1FB92}, {0x1FB94, 0x1FBCA},
+};
+
+static const std::vector<std::pair<uint32_t, uint32_t>> control_ranges = {
+{0x0, 0x8}, {0xE, 0x1B}, {0x7F, 0x84}, {0x86, 0x9F}, {0xAD, 0xAD}, {0x378, 0x379}, {0x380, 0x383}, {0x38B, 0x38B}, {0x38D, 0x38D}, {0x3A2, 0x3A2}, {0x530, 0x530}, {0x557, 0x558}, {0x58B, 0x58C},
+{0x590, 0x590}, {0x5C8, 0x5CF}, {0x5EB, 0x5EE}, {0x5F5, 0x605}, {0x61C, 0x61D}, {0x6DD, 0x6DD}, {0x70E, 0x70F}, {0x74B, 0x74C}, {0x7B2, 0x7BF}, {0x7FB, 0x7FC}, {0x82E, 0x82F}, {0x83F, 0x83F},
+{0x85C, 0x85D}, {0x85F, 0x85F}, {0x86B, 0x89F}, {0x8B5, 0x8B5}, {0x8C8, 0x8D2}, {0x8E2, 0x8E2}, {0x984, 0x984}, {0x98D, 0x98E}, {0x991, 0x992}, {0x9A9, 0x9A9}, {0x9B1, 0x9B1}, {0x9B3, 0x9B5},
+{0x9BA, 0x9BB}, {0x9C5, 0x9C6}, {0x9C9, 0x9CA}, {0x9CF, 0x9D6}, {0x9D8, 0x9DB}, {0x9DE, 0x9DE}, {0x9E4, 0x9E5}, {0x9FF, 0xA00}, {0xA04, 0xA04}, {0xA0B, 0xA0E}, {0xA11, 0xA12}, {0xA29, 0xA29},
+{0xA31, 0xA31}, {0xA34, 0xA34}, {0xA37, 0xA37}, {0xA3A, 0xA3B}, {0xA3D, 0xA3D}, {0xA43, 0xA46}, {0xA49, 0xA4A}, {0xA4E, 0xA50}, {0xA52, 0xA58}, {0xA5D, 0xA5D}, {0xA5F, 0xA65}, {0xA77, 0xA80},
+{0xA84, 0xA84}, {0xA8E, 0xA8E}, {0xA92, 0xA92}, {0xAA9, 0xAA9}, {0xAB1, 0xAB1}, {0xAB4, 0xAB4}, {0xABA, 0xABB}, {0xAC6, 0xAC6}, {0xACA, 0xACA}, {0xACE, 0xACF}, {0xAD1, 0xADF}, {0xAE4, 0xAE5},
+{0xAF2, 0xAF8}, {0xB00, 0xB00}, {0xB04, 0xB04}, {0xB0D, 0xB0E}, {0xB11, 0xB12}, {0xB29, 0xB29}, {0xB31, 0xB31}, {0xB34, 0xB34}, {0xB3A, 0xB3B}, {0xB45, 0xB46}, {0xB49, 0xB4A}, {0xB4E, 0xB54},
+{0xB58, 0xB5B}, {0xB5E, 0xB5E}, {0xB64, 0xB65}, {0xB78, 0xB81}, {0xB84, 0xB84}, {0xB8B, 0xB8D}, {0xB91, 0xB91}, {0xB96, 0xB98}, {0xB9B, 0xB9B}, {0xB9D, 0xB9D}, {0xBA0, 0xBA2}, {0xBA5, 0xBA7},
+{0xBAB, 0xBAD}, {0xBBA, 0xBBD}, {0xBC3, 0xBC5}, {0xBC9, 0xBC9}, {0xBCE, 0xBCF}, {0xBD1, 0xBD6}, {0xBD8, 0xBE5}, {0xBFB, 0xBFF}, {0xC0D, 0xC0D}, {0xC11, 0xC11}, {0xC29, 0xC29}, {0xC3A, 0xC3C},
+{0xC45, 0xC45}, {0xC49, 0xC49}, {0xC4E, 0xC54}, {0xC57, 0xC57}, {0xC5B, 0xC5F}, {0xC64, 0xC65}, {0xC70, 0xC76}, {0xC8D, 0xC8D}, {0xC91, 0xC91}, {0xCA9, 0xCA9}, {0xCB4, 0xCB4}, {0xCBA, 0xCBB},
+{0xCC5, 0xCC5}, {0xCC9, 0xCC9}, {0xCCE, 0xCD4}, {0xCD7, 0xCDD}, {0xCDF, 0xCDF}, {0xCE4, 0xCE5}, {0xCF0, 0xCF0}, {0xCF3, 0xCFF}, {0xD0D, 0xD0D}, {0xD11, 0xD11}, {0xD45, 0xD45}, {0xD49, 0xD49},
+{0xD50, 0xD53}, {0xD64, 0xD65}, {0xD80, 0xD80}, {0xD84, 0xD84}, {0xD97, 0xD99}, {0xDB2, 0xDB2}, {0xDBC, 0xDBC}, {0xDBE, 0xDBF}, {0xDC7, 0xDC9}, {0xDCB, 0xDCE}, {0xDD5, 0xDD5}, {0xDD7, 0xDD7},
+{0xDE0, 0xDE5}, {0xDF0, 0xDF1}, {0xDF5, 0xE00}, {0xE3B, 0xE3E}, {0xE5C, 0xE80}, {0xE83, 0xE83}, {0xE85, 0xE85}, {0xE8B, 0xE8B}, {0xEA4, 0xEA4}, {0xEA6, 0xEA6}, {0xEBE, 0xEBF}, {0xEC5, 0xEC5},
+{0xEC7, 0xEC7}, {0xECE, 0xECF}, {0xEDA, 0xEDB}, {0xEE0, 0xEFF}, {0xF48, 0xF48}, {0xF6D, 0xF70}, {0xF98, 0xF98}, {0xFBD, 0xFBD}, {0xFCD, 0xFCD}, {0xFDB, 0xFFF}, {0x10C6, 0x10C6}, {0x10C8, 0x10CC},
+{0x10CE, 0x10CF}, {0x1249, 0x1249}, {0x124E, 0x124F}, {0x1257, 0x1257}, {0x1259, 0x1259}, {0x125E, 0x125F}, {0x1289, 0x1289}, {0x128E, 0x128F}, {0x12B1, 0x12B1}, {0x12B6, 0x12B7}, {0x12BF, 0x12BF},
+{0x12C1, 0x12C1}, {0x12C6, 0x12C7}, {0x12D7, 0x12D7}, {0x1311, 0x1311}, {0x1316, 0x1317}, {0x135B, 0x135C}, {0x137D, 0x137F}, {0x139A, 0x139F}, {0x13F6, 0x13F7}, {0x13FE, 0x13FF}, {0x169D, 0x169F},
+{0x16F9, 0x16FF}, {0x170D, 0x170D}, {0x1715, 0x171F}, {0x1737, 0x173F}, {0x1754, 0x175F}, {0x176D, 0x176D}, {0x1771, 0x1771}, {0x1774, 0x177F}, {0x17DE, 0x17DF}, {0x17EA, 0x17EF}, {0x17FA, 0x17FF},
+{0x180E, 0x180F}, {0x181A, 0x181F}, {0x1879, 0x187F}, {0x18AB, 0x18AF}, {0x18F6, 0x18FF}, {0x191F, 0x191F}, {0x192C, 0x192F}, {0x193C, 0x193F}, {0x1941, 0x1943}, {0x196E, 0x196F}, {0x1975, 0x197F},
+{0x19AC, 0x19AF}, {0x19CA, 0x19CF}, {0x19DB, 0x19DD}, {0x1A1C, 0x1A1D}, {0x1A5F, 0x1A5F}, {0x1A7D, 0x1A7E}, {0x1A8A, 0x1A8F}, {0x1A9A, 0x1A9F}, {0x1AAE, 0x1AAF}, {0x1AC1, 0x1AFF}, {0x1B4C, 0x1B4F},
+{0x1B7D, 0x1B7F}, {0x1BF4, 0x1BFB}, {0x1C38, 0x1C3A}, {0x1C4A, 0x1C4C}, {0x1C89, 0x1C8F}, {0x1CBB, 0x1CBC}, {0x1CC8, 0x1CCF}, {0x1CFB, 0x1CFF}, {0x1DFA, 0x1DFA}, {0x1F16, 0x1F17}, {0x1F1E, 0x1F1F},
+{0x1F46, 0x1F47}, {0x1F4E, 0x1F4F}, {0x1F58, 0x1F58}, {0x1F5A, 0x1F5A}, {0x1F5C, 0x1F5C}, {0x1F5E, 0x1F5E}, {0x1F7E, 0x1F7F}, {0x1FB5, 0x1FB5}, {0x1FC5, 0x1FC5}, {0x1FD4, 0x1FD5}, {0x1FDC, 0x1FDC},
+{0x1FF0, 0x1FF1}, {0x1FF5, 0x1FF5}, {0x1FFF, 0x1FFF}, {0x200B, 0x200F}, {0x202A, 0x202E}, {0x2060, 0x206F}, {0x2072, 0x2073}, {0x208F, 0x208F}, {0x209D, 0x209F}, {0x20C0, 0x20CF}, {0x20F1, 0x20FF},
+{0x218C, 0x218F}, {0x2427, 0x243F}, {0x244B, 0x245F}, {0x2B74, 0x2B75}, {0x2B96, 0x2B96}, {0x2C2F, 0x2C2F}, {0x2C5F, 0x2C5F}, {0x2CF4, 0x2CF8}, {0x2D26, 0x2D26}, {0x2D28, 0x2D2C}, {0x2D2E, 0x2D2F},
+{0x2D68, 0x2D6E}, {0x2D71, 0x2D7E}, {0x2D97, 0x2D9F}, {0x2DA7, 0x2DA7}, {0x2DAF, 0x2DAF}, {0x2DB7, 0x2DB7}, {0x2DBF, 0x2DBF}, {0x2DC7, 0x2DC7}, {0x2DCF, 0x2DCF}, {0x2DD7, 0x2DD7}, {0x2DDF, 0x2DDF},
+{0x2E53, 0x2E7F}, {0x2E9A, 0x2E9A}, {0x2EF4, 0x2EFF}, {0x2FD6, 0x2FEF}, {0x2FFC, 0x2FFF}, {0x3040, 0x3040}, {0x3097, 0x3098}, {0x3100, 0x3104}, {0x3130, 0x3130}, {0x318F, 0x318F}, {0x31E4, 0x31EF},
+{0x321F, 0x321F}, {0x9FFD, 0x9FFF}, {0xA48D, 0xA48F}, {0xA4C7, 0xA4CF}, {0xA62C, 0xA63F}, {0xA6F8, 0xA6FF}, {0xA7C0, 0xA7C1}, {0xA7CB, 0xA7F4}, {0xA82D, 0xA82F}, {0xA83A, 0xA83F}, {0xA878, 0xA87F},
+{0xA8C6, 0xA8CD}, {0xA8DA, 0xA8DF}, {0xA954, 0xA95E}, {0xA97D, 0xA97F}, {0xA9CE, 0xA9CE}, {0xA9DA, 0xA9DD}, {0xA9FF, 0xA9FF}, {0xAA37, 0xAA3F}, {0xAA4E, 0xAA4F}, {0xAA5A, 0xAA5B}, {0xAAC3, 0xAADA},
+{0xAAF7, 0xAB00}, {0xAB07, 0xAB08}, {0xAB0F, 0xAB10}, {0xAB17, 0xAB1F}, {0xAB27, 0xAB27}, {0xAB2F, 0xAB2F}, {0xAB6C, 0xAB6F}, {0xABEE, 0xABEF}, {0xABFA, 0xABFF}, {0xD7A4, 0xD7AF}, {0xD7C7, 0xD7CA},
+{0xD7FC, 0xF8FF}, {0xFA6E, 0xFA6F}, {0xFADA, 0xFAFF}, {0xFB07, 0xFB12}, {0xFB18, 0xFB1C}, {0xFB37, 0xFB37}, {0xFB3D, 0xFB3D}, {0xFB3F, 0xFB3F}, {0xFB42, 0xFB42}, {0xFB45, 0xFB45}, {0xFBC2, 0xFBD2},
+{0xFD40, 0xFD4F}, {0xFD90, 0xFD91}, {0xFDC8, 0xFDEF}, {0xFDFE, 0xFDFF}, {0xFE1A, 0xFE1F}, {0xFE53, 0xFE53}, {0xFE67, 0xFE67}, {0xFE6C, 0xFE6F}, {0xFE75, 0xFE75}, {0xFEFD, 0xFF00}, {0xFFBF, 0xFFC1},
+{0xFFC8, 0xFFC9}, {0xFFD0, 0xFFD1}, {0xFFD8, 0xFFD9}, {0xFFDD, 0xFFDF}, {0xFFE7, 0xFFE7}, {0xFFEF, 0xFFFB}, {0xFFFE, 0xFFFF}, {0x1000C, 0x1000C}, {0x10027, 0x10027}, {0x1003B, 0x1003B},
+{0x1003E, 0x1003E}, {0x1004E, 0x1004F}, {0x1005E, 0x1007F}, {0x100FB, 0x100FF}, {0x10103, 0x10106}, {0x10134, 0x10136}, {0x1018F, 0x1018F}, {0x1019D, 0x1019F}, {0x101A1, 0x101CF}, {0x101FE, 0x1027F},
+{0x1029D, 0x1029F}, {0x102D1, 0x102DF}, {0x102FC, 0x102FF}, {0x10324, 0x1032C}, {0x1034B, 0x1034F}, {0x1037B, 0x1037F}, {0x1039E, 0x1039E}, {0x103C4, 0x103C7}, {0x103D6, 0x103FF}, {0x1049E, 0x1049F},
+{0x104AA, 0x104AF}, {0x104D4, 0x104D7}, {0x104FC, 0x104FF}, {0x10528, 0x1052F}, {0x10564, 0x1056E}, {0x10570, 0x105FF}, {0x10737, 0x1073F}, {0x10756, 0x1075F}, {0x10768, 0x107FF}, {0x10806, 0x10807},
+{0x10809, 0x10809}, {0x10836, 0x10836}, {0x10839, 0x1083B}, {0x1083D, 0x1083E}, {0x10856, 0x10856}, {0x1089F, 0x108A6}, {0x108B0, 0x108DF}, {0x108F3, 0x108F3}, {0x108F6, 0x108FA}, {0x1091C, 0x1091E},
+{0x1093A, 0x1093E}, {0x10940, 0x1097F}, {0x109B8, 0x109BB}, {0x109D0, 0x109D1}, {0x10A04, 0x10A04}, {0x10A07, 0x10A0B}, {0x10A14, 0x10A14}, {0x10A18, 0x10A18}, {0x10A36, 0x10A37}, {0x10A3B, 0x10A3E},
+{0x10A49, 0x10A4F}, {0x10A59, 0x10A5F}, {0x10AA0, 0x10ABF}, {0x10AE7, 0x10AEA}, {0x10AF7, 0x10AFF}, {0x10B36, 0x10B38}, {0x10B56, 0x10B57}, {0x10B73, 0x10B77}, {0x10B92, 0x10B98}, {0x10B9D, 0x10BA8},
+{0x10BB0, 0x10BFF}, {0x10C49, 0x10C7F}, {0x10CB3, 0x10CBF}, {0x10CF3, 0x10CF9}, {0x10D28, 0x10D2F}, {0x10D3A, 0x10E5F}, {0x10E7F, 0x10E7F}, {0x10EAA, 0x10EAA}, {0x10EAE, 0x10EAF}, {0x10EB2, 0x10EFF},
+{0x10F28, 0x10F2F}, {0x10F5A, 0x10FAF}, {0x10FCC, 0x10FDF}, {0x10FF7, 0x10FFF}, {0x1104E, 0x11051}, {0x11070, 0x1107E}, {0x110BD, 0x110BD}, {0x110C2, 0x110CF}, {0x110E9, 0x110EF}, {0x110FA, 0x110FF},
+{0x11135, 0x11135}, {0x11148, 0x1114F}, {0x11177, 0x1117F}, {0x111E0, 0x111E0}, {0x111F5, 0x111FF}, {0x11212, 0x11212}, {0x1123F, 0x1127F}, {0x11287, 0x11287}, {0x11289, 0x11289}, {0x1128E, 0x1128E},
+{0x1129E, 0x1129E}, {0x112AA, 0x112AF}, {0x112EB, 0x112EF}, {0x112FA, 0x112FF}, {0x11304, 0x11304}, {0x1130D, 0x1130E}, {0x11311, 0x11312}, {0x11329, 0x11329}, {0x11331, 0x11331}, {0x11334, 0x11334},
+{0x1133A, 0x1133A}, {0x11345, 0x11346}, {0x11349, 0x1134A}, {0x1134E, 0x1134F}, {0x11351, 0x11356}, {0x11358, 0x1135C}, {0x11364, 0x11365}, {0x1136D, 0x1136F}, {0x11375, 0x113FF}, {0x1145C, 0x1145C},
+{0x11462, 0x1147F}, {0x114C8, 0x114CF}, {0x114DA, 0x1157F}, {0x115B6, 0x115B7}, {0x115DE, 0x115FF}, {0x11645, 0x1164F}, {0x1165A, 0x1165F}, {0x1166D, 0x1167F}, {0x116B9, 0x116BF}, {0x116CA, 0x116FF},
+{0x1171B, 0x1171C}, {0x1172C, 0x1172F}, {0x11740, 0x117FF}, {0x1183C, 0x1189F}, {0x118F3, 0x118FE}, {0x11907, 0x11908}, {0x1190A, 0x1190B}, {0x11914, 0x11914}, {0x11917, 0x11917}, {0x11936, 0x11936},
+{0x11939, 0x1193A}, {0x11947, 0x1194F}, {0x1195A, 0x1199F}, {0x119A8, 0x119A9}, {0x119D8, 0x119D9}, {0x119E5, 0x119FF}, {0x11A48, 0x11A4F}, {0x11AA3, 0x11ABF}, {0x11AF9, 0x11BFF}, {0x11C09, 0x11C09},
+{0x11C37, 0x11C37}, {0x11C46, 0x11C4F}, {0x11C6D, 0x11C6F}, {0x11C90, 0x11C91}, {0x11CA8, 0x11CA8}, {0x11CB7, 0x11CFF}, {0x11D07, 0x11D07}, {0x11D0A, 0x11D0A}, {0x11D37, 0x11D39}, {0x11D3B, 0x11D3B},
+{0x11D3E, 0x11D3E}, {0x11D48, 0x11D4F}, {0x11D5A, 0x11D5F}, {0x11D66, 0x11D66}, {0x11D69, 0x11D69}, {0x11D8F, 0x11D8F}, {0x11D92, 0x11D92}, {0x11D99, 0x11D9F}, {0x11DAA, 0x11EDF}, {0x11EF9, 0x11FAF},
+{0x11FB1, 0x11FBF}, {0x11FF2, 0x11FFE}, {0x1239A, 0x123FF}, {0x1246F, 0x1246F}, {0x12475, 0x1247F}, {0x12544, 0x12FFF}, {0x1342F, 0x143FF}, {0x14647, 0x167FF}, {0x16A39, 0x16A3F}, {0x16A5F, 0x16A5F},
+{0x16A6A, 0x16A6D}, {0x16A70, 0x16ACF}, {0x16AEE, 0x16AEF}, {0x16AF6, 0x16AFF}, {0x16B46, 0x16B4F}, {0x16B5A, 0x16B5A}, {0x16B62, 0x16B62}, {0x16B78, 0x16B7C}, {0x16B90, 0x16E3F}, {0x16E9B, 0x16EFF},
+{0x16F4B, 0x16F4E}, {0x16F88, 0x16F8E}, {0x16FA0, 0x16FDF}, {0x16FE5, 0x16FEF}, {0x16FF2, 0x16FFF}, {0x187F8, 0x187FF}, {0x18CD6, 0x18CFF}, {0x18D09, 0x1AFFF}, {0x1B11F, 0x1B14F}, {0x1B153, 0x1B163},
+{0x1B168, 0x1B16F}, {0x1B2FC, 0x1BBFF}, {0x1BC6B, 0x1BC6F}, {0x1BC7D, 0x1BC7F}, {0x1BC89, 0x1BC8F}, {0x1BC9A, 0x1BC9B}, {0x1BCA0, 0x1CFFF}, {0x1D0F6, 0x1D0FF}, {0x1D127, 0x1D128}, {0x1D173, 0x1D17A},
+{0x1D1E9, 0x1D1FF}, {0x1D246, 0x1D2DF}, {0x1D2F4, 0x1D2FF}, {0x1D357, 0x1D35F}, {0x1D379, 0x1D3FF}, {0x1D455, 0x1D455}, {0x1D49D, 0x1D49D}, {0x1D4A0, 0x1D4A1}, {0x1D4A3, 0x1D4A4}, {0x1D4A7, 0x1D4A8},
+{0x1D4AD, 0x1D4AD}, {0x1D4BA, 0x1D4BA}, {0x1D4BC, 0x1D4BC}, {0x1D4C4, 0x1D4C4}, {0x1D506, 0x1D506}, {0x1D50B, 0x1D50C}, {0x1D515, 0x1D515}, {0x1D51D, 0x1D51D}, {0x1D53A, 0x1D53A}, {0x1D53F, 0x1D53F},
+{0x1D545, 0x1D545}, {0x1D547, 0x1D549}, {0x1D551, 0x1D551}, {0x1D6A6, 0x1D6A7}, {0x1D7CC, 0x1D7CD}, {0x1DA8C, 0x1DA9A}, {0x1DAA0, 0x1DAA0}, {0x1DAB0, 0x1DFFF}, {0x1E007, 0x1E007}, {0x1E019, 0x1E01A},
+{0x1E022, 0x1E022}, {0x1E025, 0x1E025}, {0x1E02B, 0x1E0FF}, {0x1E12D, 0x1E12F}, {0x1E13E, 0x1E13F}, {0x1E14A, 0x1E14D}, {0x1E150, 0x1E2BF}, {0x1E2FA, 0x1E2FE}, {0x1E300, 0x1E7FF}, {0x1E8C5, 0x1E8C6},
+{0x1E8D7, 0x1E8FF}, {0x1E94C, 0x1E94F}, {0x1E95A, 0x1E95D}, {0x1E960, 0x1EC70}, {0x1ECB5, 0x1ED00}, {0x1ED3E, 0x1EDFF}, {0x1EE04, 0x1EE04}, {0x1EE20, 0x1EE20}, {0x1EE23, 0x1EE23}, {0x1EE25, 0x1EE26},
+{0x1EE28, 0x1EE28}, {0x1EE33, 0x1EE33}, {0x1EE38, 0x1EE38}, {0x1EE3A, 0x1EE3A}, {0x1EE3C, 0x1EE41}, {0x1EE43, 0x1EE46}, {0x1EE48, 0x1EE48}, {0x1EE4A, 0x1EE4A}, {0x1EE4C, 0x1EE4C}, {0x1EE50, 0x1EE50},
+{0x1EE53, 0x1EE53}, {0x1EE55, 0x1EE56}, {0x1EE58, 0x1EE58}, {0x1EE5A, 0x1EE5A}, {0x1EE5C, 0x1EE5C}, {0x1EE5E, 0x1EE5E}, {0x1EE60, 0x1EE60}, {0x1EE63, 0x1EE63}, {0x1EE65, 0x1EE66}, {0x1EE6B, 0x1EE6B},
+{0x1EE73, 0x1EE73}, {0x1EE78, 0x1EE78}, {0x1EE7D, 0x1EE7D}, {0x1EE7F, 0x1EE7F}, {0x1EE8A, 0x1EE8A}, {0x1EE9C, 0x1EEA0}, {0x1EEA4, 0x1EEA4}, {0x1EEAA, 0x1EEAA}, {0x1EEBC, 0x1EEEF}, {0x1EEF2, 0x1EFFF},
+{0x1F02C, 0x1F02F}, {0x1F094, 0x1F09F}, {0x1F0AF, 0x1F0B0}, {0x1F0C0, 0x1F0C0}, {0x1F0D0, 0x1F0D0}, {0x1F0F6, 0x1F0FF}, {0x1F1AE, 0x1F1E5}, {0x1F203, 0x1F20F}, {0x1F23C, 0x1F23F}, {0x1F249, 0x1F24F},
+{0x1F252, 0x1F25F}, {0x1F266, 0x1F2FF}, {0x1F6D8, 0x1F6DF}, {0x1F6ED, 0x1F6EF}, {0x1F6FD, 0x1F6FF}, {0x1F774, 0x1F77F}, {0x1F7D9, 0x1F7DF}, {0x1F7EC, 0x1F7FF}, {0x1F80C, 0x1F80F}, {0x1F848, 0x1F84F},
+{0x1F85A, 0x1F85F}, {0x1F888, 0x1F88F}, {0x1F8AE, 0x1F8AF}, {0x1F8B2, 0x1F8FF}, {0x1F979, 0x1F979}, {0x1F9CC, 0x1F9CC}, {0x1FA54, 0x1FA5F}, {0x1FA6E, 0x1FA6F}, {0x1FA75, 0x1FA77}, {0x1FA7B, 0x1FA7F},
+{0x1FA87, 0x1FA8F}, {0x1FAA9, 0x1FAAF}, {0x1FAB7, 0x1FABF}, {0x1FAC3, 0x1FACF}, {0x1FAD7, 0x1FAFF}, {0x1FB93, 0x1FB93}, {0x1FBCB, 0x1FBEF}, {0x1FBFA, 0x1FFFF}, {0x2A6DE, 0x2A6FF}, {0x2B735, 0x2B73F},
+{0x2B81E, 0x2B81F}, {0x2CEA2, 0x2CEAF}, {0x2EBE1, 0x2F7FF}, {0x2FA1E, 0x2FFFF}, {0x3134B, 0xE00FF}, {0xE01F0, 0x10FFFF},
+};
+
+static std::string codepoint_to_utf8(uint32_t cp) {
+    std::string result;
+    if (/* 0x00 <= cp && */ cp <= 0x7f) {
+        result.push_back(cp);
+    }
+    else if (0x80 <= cp && cp <= 0x7ff) {
+        result.push_back(0xc0 | ((cp >> 6) & 0x1f));
+        result.push_back(0x80 | (cp & 0x3f));
+    }
+    else if (0x800 <= cp && cp <= 0xffff) {
+        result.push_back(0xe0 | ((cp >> 12) & 0x0f));
+        result.push_back(0x80 | ((cp >> 6) & 0x3f));
+        result.push_back(0x80 | (cp & 0x3f));
+    }
+    else if (0x10000 <= cp && cp <= 0x10ffff) {
+        result.push_back(0xf0 | ((cp >> 18) & 0x07));
+        result.push_back(0x80 | ((cp >> 12) & 0x3f));
+        result.push_back(0x80 | ((cp >> 6) & 0x3f));
+        result.push_back(0x80 | (cp & 0x3f));
+    }
+    else {
+        throw std::invalid_argument("invalid codepoint");
+    }
+    return result;
+}
+
+static std::string codepoints_to_utf8(const std::vector<uint32_t> & cps) {
+    std::string result;
+    for (size_t i = 0; i < cps.size(); ++i) {
+        result.append(codepoint_to_utf8(cps[i]));
+    }
+    return result;
+}
+
+static uint32_t codepoint_from_utf8(const std::string & utf8, size_t & offset) {
+    assert(offset < utf8.size());
+    if (!(utf8[offset + 0] & 0x80)) {
+        auto result = utf8[offset + 0];
+        offset += 1;
+        return result;
+    }
+    else if (!(utf8[offset + 0] & 0x40)) {
+        throw std::invalid_argument("invalid character");
+    }
+    else if (!(utf8[offset + 0] & 0x20)) {
+        if (offset + 1 >= utf8.size() || ! ((utf8[offset + 1] & 0xc0) == 0x80))
+            throw std::invalid_argument("invalid character");
+        auto result = ((utf8[offset + 0] & 0x1f) << 6) | (utf8[offset + 1] & 0x3f);
+        offset += 2;
+        return result;
+    }
+    else if (!(utf8[offset + 0] & 0x10)) {
+        if (offset + 2 >= utf8.size() || ! ((utf8[offset + 1] & 0xc0) == 0x80) || ! ((utf8[offset + 2] & 0xc0) == 0x80))
+            throw std::invalid_argument("invalid character");
+        auto result = ((utf8[offset + 0] & 0x0f) << 12) | ((utf8[offset + 1] & 0x3f) << 6) | (utf8[offset + 2] & 0x3f);
+        offset += 3;
+        return result;
+    }
+    else if (!(utf8[offset + 0] & 0x08)) {
+        if (offset + 3 >= utf8.size() || ! ((utf8[offset + 1] & 0xc0) == 0x80) || ! ((utf8[offset + 2] & 0xc0) == 0x80) || !((utf8[offset + 3] & 0xc0) == 0x80))
+            throw std::invalid_argument("invalid character");
+        auto result = ((utf8[offset + 0] & 0x07) << 18) | ((utf8[offset + 1] & 0x3f) << 12) | ((utf8[offset + 2] & 0x3f) << 6) | (utf8[offset + 3] & 0x3f);
+        offset += 4;
+        return result;
+    }
+    throw std::invalid_argument("invalid string");
+}
+
+static std::vector<uint32_t> codepoints_from_utf8(const std::string & utf8) {
+    std::vector<uint32_t> result;
+    size_t offset = 0;
+    while (offset < utf8.size()) {
+        result.push_back(codepoint_from_utf8(utf8, offset));
+    }
+    return result;
+}
+
+static std::vector<uint16_t> codepoint_to_utf16(uint32_t cp) {
+    std::vector<uint16_t> result;
+    if (/* 0x0000 <= cp && */ cp <= 0xffff) {
+        result.emplace_back(cp);
+    }
+    else if (0x10000 <= cp && cp <= 0x10ffff) {
+        result.emplace_back(0xd800 | ((cp - 0x10000) >> 10));
+        result.emplace_back(0xdc00 | ((cp - 0x10000) & 0x03ff));
+    }
+    else {
+        throw std::invalid_argument("invalid codepoint");
+    }
+    return result;
+}
+
+static std::vector<uint16_t> codepoints_to_utf16(const std::vector<uint32_t> & cps) {
+    std::vector<uint16_t> result;
+    for (size_t i = 0; i < cps.size(); ++i) {
+        auto temp = codepoint_to_utf16(cps[i]);
+        result.insert(result.end(), temp.begin(), temp.end());
+    }
+    return result;
+}
+
+static uint32_t codepoint_from_utf16(const std::vector<uint16_t> & utf16, size_t & offset) {
+    assert(offset < utf16.size());
+    if (((utf16[0] >> 10) << 10) != 0xd800) {
+        auto result = utf16[offset + 0];
+        offset += 1;
+        return result;
+    }
+    else {
+        if (offset + 1 >= utf16.size() || !((utf16[1] & 0xdc00) == 0xdc00))
+            throw std::invalid_argument("invalid character");
+        auto result = 0x10000 + (((utf16[0] & 0x03ff) << 10) | (utf16[1] & 0x03ff));
+        offset += 2;
+        return result;
+    }
+    throw std::invalid_argument("invalid string");
+}
+
+static std::vector<uint32_t> codepoints_from_utf16(const std::vector<uint16_t> & utf16) {
+    std::vector<uint32_t> result;
+    size_t offset = 0;
+    while (offset < utf16.size())
+        result.push_back(codepoint_from_utf16(utf16, offset));
+    return result;
+}
+
+#define CODEPOINT_TYPE_UNIDENTIFIED 0
+#define CODEPOINT_TYPE_DIGIT 1
+#define CODEPOINT_TYPE_LETTER 2
+#define CODEPOINT_TYPE_WHITESPACE 3
+#define CODEPOINT_TYPE_ACCENT_MARK 4
+#define CODEPOINT_TYPE_PUNCTUATION 5
+#define CODEPOINT_TYPE_SYMBOL 6
+#define CODEPOINT_TYPE_CONTROL 7
+
+static std::unordered_map<uint32_t, int> codepoint_type_map() {
+    std::unordered_map<uint32_t, int> codepoint_types;
+    for (auto p : digit_ranges) {
+        for(auto i = p.first; i <= p.second; ++ i)
+            codepoint_types[i] = CODEPOINT_TYPE_DIGIT;
+    }
+    for(auto p : letter_ranges) {
+        for(auto i = p.first; i <= p.second; ++ i)
+            codepoint_types[i] = CODEPOINT_TYPE_LETTER;
+    }
+    for(auto p : whitespace_ranges) {
+        for(auto i = p.first; i <= p.second; ++ i)
+            codepoint_types[i] = CODEPOINT_TYPE_WHITESPACE;
+    }
+    for(auto p : accent_mark_ranges) {
+        for(auto i = p.first; i <= p.second; ++ i)
+            codepoint_types[i] = CODEPOINT_TYPE_ACCENT_MARK;
+    }
+    for(auto p : punctuation_ranges) {
+        for(auto i = p.first; i <= p.second; ++ i)
+            codepoint_types[i] = CODEPOINT_TYPE_PUNCTUATION;
+    }
+    for (auto p : symbol_ranges) {
+        for (auto i = p.first; i <= p.second; ++i)
+            codepoint_types[i] = CODEPOINT_TYPE_SYMBOL;
+    }
+    for(auto p : control_ranges) {
+        for(auto i = p.first; i <= p.second; ++ i)
+            codepoint_types[i] = CODEPOINT_TYPE_CONTROL;
+    }
+    return codepoint_types;
+}
+
+static int codepoint_type(uint32_t cp) {
+    static std::unordered_map<uint32_t, int> codepoint_types = codepoint_type_map();
+    return codepoint_types[cp];
+}
+
+static int codepoint_type(const std::string & utf8) {
+    if (utf8.length() == 0)
+        return CODEPOINT_TYPE_UNIDENTIFIED;
+    size_t offset = 0;
+    return codepoint_type(codepoint_from_utf8(utf8, offset));
+}
+
+static std::unordered_map<uint8_t, std::string> bytes_to_unicode_map_bpe() {
+    std::unordered_map<uint8_t, std::string> map;
+    for (int ch = u'!'; ch <= u'~'; ++ch) {
+        assert(0 <= ch && ch < 256);
+        map[ch] = codepoint_to_utf8(ch);
+    }
+    for (int ch = u'¡'; ch <= u'¬'; ++ch) {
+        assert(0 <= ch && ch < 256);
+        map[ch] = codepoint_to_utf8(ch);
+    }
+    for (int ch = u'®'; ch <= u'ÿ'; ++ch) {
+        assert(0 <= ch && ch < 256);
+        map[ch] = codepoint_to_utf8(ch);
+    }
+    auto n = 0;
+    for (int ch = 0; ch < 256; ++ch) {
+        if (map.find(ch) == map.end()) {
+            map[ch] = codepoint_to_utf8(256 + n);
+            ++n;
+        }
+    }
+    return map;
+}
+
+static std::string bytes_to_unicode_bpe(uint8_t byte) {
+    static std::unordered_map<uint8_t, std::string> map = bytes_to_unicode_map_bpe();
+    return map.at(byte);
+}
+
+static std::unordered_map<std::string, uint8_t> unicode_to_bytes_map_bpe() {
+    std::unordered_map<std::string, uint8_t> map;
+    for (int ch = u'!'; ch <= u'~'; ++ch) {
+        assert(0 <= ch && ch < 256);
+        map[codepoint_to_utf8(ch)] = ch;
+    }
+    for (int ch = u'¡'; ch <= u'¬'; ++ch) {
+        assert(0 <= ch && ch < 256);
+        map[codepoint_to_utf8(ch)] = ch;
+    }
+    for (int ch = u'®'; ch <= u'ÿ'; ++ch) {
+        assert(0 <= ch && ch < 256);
+        map[codepoint_to_utf8(ch)] = ch;
+    }
+    auto n = 0;
+    for (int ch = 0; ch < 256; ++ch) {
+        if (map.find(codepoint_to_utf8(ch)) == map.end()) {
+            map[codepoint_to_utf8(256 + n)] = ch;
+            ++n;
+        }
+    }
+    return map;
+}
+
+static uint8_t unicode_to_bytes_bpe(const std::string & utf8) {
+    static std::unordered_map<std::string, uint8_t> map = unicode_to_bytes_map_bpe();
+    return map.at(utf8);
+}
+

runner/update.sh

@@ -0,0 +1,70 @@
+#!/bin/sh
+
+set -eu
+
+
+status() { echo >&2 ">>> $*"; }
+error() { status "ERROR $*"; }
+usage() {
+    echo "usage: $(basename $0) /path/to/repo"
+    exit 1
+}
+
+OUT=$(dirname $0)
+while getopts "hC:" OPTION; do
+    case $OPTION in
+        C) OUT=$OPTARG ;;
+        *) usage ;;
+    esac
+done
+
+shift $(( $OPTIND - 1 ))
+[ $# -eq 1 ] || usage
+
+status "updating source..."
+cp -a "$1"/*.{c,h,cpp,m,metal,cu} "$OUT"
+
+status "removing incompatible files..."
+rm -f "$OUT"/build-info.h
+
+SHA1=$(git -C $1 rev-parse @)
+
+LICENSE=$(mktemp)
+cleanup() {
+    rm -f $LICENSE
+}
+trap cleanup 0
+
+cat <<EOF | sed 's/ *$//' >$LICENSE
+/**
+ * llama.cpp - git $SHA1
+ *
+$(sed 's/^/ * /' <$1/LICENSE)
+ */
+
+EOF
+
+for IN in $OUT/*.{c,h,cpp,m,metal,cu}; do
+    TMP=$(mktemp)
+    status "updating license $IN"
+    cat $LICENSE $IN >$TMP
+    mv $TMP $IN
+done
+
+touchup() {
+    local CONSTRAINT=$1 && shift
+
+    for IN in $*; do
+        status "touching up $IN..."
+        TMP=$(mktemp)
+        {
+            echo "//go:build $CONSTRAINT"
+            echo
+        } | cat - $IN >$TMP
+        mv $TMP $IN
+    done
+}
+
+touchup darwin $OUT/ggml-metal.*
+touchup mpi $OUT/ggml-mpi.*
+touchup opencl $OUT/ggml-opencl.*

scripts/install.sh

@@ -63,10 +63,7 @@ status "Installing ollama to $BINDIR..."
 $SUDO install -o0 -g0 -m755 -d $BINDIR
 $SUDO install -o0 -g0 -m755 $TEMP_DIR/ollama $BINDIR/ollama
 
-install_success() { 
-    status 'The Ollama API is now available at 0.0.0.0:11434.'
-    status 'Install complete. Run "ollama" from the command line.'
-}
+install_success() { status 'Install complete. Run "ollama" from the command line.'; }
 trap install_success EXIT
 
 # Everything from this point onwards is optional.
@@ -77,9 +74,6 @@ configure_systemd() {
         $SUDO useradd -r -s /bin/false -m -d /usr/share/ollama ollama
     fi
 
-    status "Adding current user to ollama group..."
-    $SUDO usermod -a -G ollama $(whoami)
-
     status "Creating ollama systemd service..."
     cat <<EOF | $SUDO tee /etc/systemd/system/ollama.service >/dev/null
 [Unit]
@@ -92,6 +86,7 @@ User=ollama
 Group=ollama
 Restart=always
 RestartSec=3
+Environment="HOME=/usr/share/ollama"
 Environment="PATH=$PATH"
 
 [Install]
@@ -133,7 +128,6 @@ if check_gpu nvidia-smi; then
 fi
 
 if ! check_gpu lspci && ! check_gpu lshw; then
-    install_success
     warning "No NVIDIA GPU detected. Ollama will run in CPU-only mode."
     exit 0
 fi

server/auth.go

@@ -91,7 +91,7 @@ func getAuthToken(ctx context.Context, redirData AuthRedirect) (string, error) {
 	}
 
 	s := SignatureData{
-		Method: http.MethodGet,
+		Method: "GET",
 		Path:   redirectURL.String(),
 		Data:   nil,
 	}
@@ -103,7 +103,7 @@ func getAuthToken(ctx context.Context, redirData AuthRedirect) (string, error) {
 
 	headers := make(http.Header)
 	headers.Set("Authorization", sig)
-	resp, err := makeRequest(ctx, http.MethodGet, redirectURL, headers, nil, nil)
+	resp, err := makeRequest(ctx, "GET", redirectURL, headers, nil, nil)
 	if err != nil {
 		log.Printf("couldn't get token: %q", err)
 		return "", err

server/download.go

@@ -15,7 +15,6 @@ import (
 	"strings"
 	"sync"
 	"sync/atomic"
-	"syscall"
 	"time"
 
 	"golang.org/x/sync/errgroup"
@@ -89,12 +88,17 @@ func (b *blobDownload) Prepare(ctx context.Context, requestURL *url.URL, opts *R
 	}
 
 	if len(b.Parts) == 0 {
-		resp, err := makeRequestWithRetry(ctx, http.MethodHead, requestURL, nil, nil, opts)
+		resp, err := makeRequest(ctx, "HEAD", requestURL, nil, nil, opts)
 		if err != nil {
 			return err
 		}
 		defer resp.Body.Close()
 
+		if resp.StatusCode >= http.StatusBadRequest {
+			body, _ := io.ReadAll(resp.Body)
+			return fmt.Errorf("registry responded with code %d: %v", resp.StatusCode, string(body))
+		}
+
 		b.Total, _ = strconv.ParseInt(resp.Header.Get("Content-Length"), 10, 64)
 
 		var size = b.Total / numDownloadParts
@@ -129,6 +133,7 @@ func (b *blobDownload) Run(ctx context.Context, requestURL *url.URL, opts *Regis
 
 func (b *blobDownload) run(ctx context.Context, requestURL *url.URL, opts *RegistryOptions) error {
 	defer blobDownloadManager.Delete(b.Digest)
+
 	ctx, b.CancelFunc = context.WithCancel(ctx)
 
 	file, err := os.OpenFile(b.Name+"-partial", os.O_CREATE|os.O_RDWR, 0644)
@@ -153,8 +158,7 @@ func (b *blobDownload) run(ctx context.Context, requestURL *url.URL, opts *Regis
 				w := io.NewOffsetWriter(file, part.StartsAt())
 				err := b.downloadChunk(inner, requestURL, w, part, opts)
 				switch {
-				case errors.Is(err, context.Canceled), errors.Is(err, syscall.ENOSPC):
-					// return immediately if the context is canceled or the device is out of space
+				case errors.Is(err, context.Canceled):
 					return err
 				case err != nil:
 					log.Printf("%s part %d attempt %d failed: %v, retrying", b.Digest[7:19], i, try, err)
@@ -164,7 +168,7 @@ func (b *blobDownload) run(ctx context.Context, requestURL *url.URL, opts *Regis
 				}
 			}
 
-			return errMaxRetriesExceeded
+			return errors.New("max retries exceeded")
 		})
 	}
 
@@ -194,7 +198,7 @@ func (b *blobDownload) run(ctx context.Context, requestURL *url.URL, opts *Regis
 func (b *blobDownload) downloadChunk(ctx context.Context, requestURL *url.URL, w io.Writer, part *blobDownloadPart, opts *RegistryOptions) error {
 	headers := make(http.Header)
 	headers.Set("Range", fmt.Sprintf("bytes=%d-%d", part.StartsAt(), part.StopsAt()-1))
-	resp, err := makeRequestWithRetry(ctx, http.MethodGet, requestURL, headers, nil, opts)
+	resp, err := makeRequest(ctx, "GET", requestURL, headers, nil, opts)
 	if err != nil {
 		return err
 	}
@@ -302,8 +306,6 @@ type downloadOpts struct {
 
 const maxRetries = 3
 
-var errMaxRetriesExceeded = errors.New("max retries exceeded")
-
 // downloadBlob downloads a blob from the registry and stores it in the blobs directory
 func downloadBlob(ctx context.Context, opts downloadOpts) error {
 	fp, err := GetBlobsPath(opts.digest)

server/images.go

@@ -63,11 +63,15 @@ func (m *Model) Prompt(request api.GenerateRequest) (string, error) {
 		First  bool
 		System string
 		Prompt string
+
+		// deprecated: versions <= 0.0.7 used this to omit the system prompt
+		Context []int
 	}
 
 	vars.First = len(request.Context) == 0
 	vars.System = m.System
 	vars.Prompt = request.Prompt
+	vars.Context = request.Context
 
 	if request.System != "" {
 		vars.System = request.System
@@ -127,7 +131,7 @@ func (m *ManifestV2) GetTotalSize() (total int64) {
 }
 
 func GetManifest(mp ModelPath) (*ManifestV2, string, error) {
-	fp, err := mp.GetManifestPath()
+	fp, err := mp.GetManifestPath(false)
 	if err != nil {
 		return nil, "", err
 	}
@@ -591,13 +595,10 @@ func CreateManifest(name string, cfg *LayerReader, layers []*Layer) error {
 		return err
 	}
 
-	fp, err := mp.GetManifestPath()
+	fp, err := mp.GetManifestPath(true)
 	if err != nil {
 		return err
 	}
-	if err := os.MkdirAll(filepath.Dir(fp), 0o755); err != nil {
-		return err
-	}
 	return os.WriteFile(fp, manifestJSON, 0o644)
 }
 
@@ -709,19 +710,16 @@ func CreateLayer(f io.ReadSeeker) (*LayerReader, error) {
 
 func CopyModel(src, dest string) error {
 	srcModelPath := ParseModelPath(src)
-	srcPath, err := srcModelPath.GetManifestPath()
+	srcPath, err := srcModelPath.GetManifestPath(false)
 	if err != nil {
 		return err
 	}
 
 	destModelPath := ParseModelPath(dest)
-	destPath, err := destModelPath.GetManifestPath()
+	destPath, err := destModelPath.GetManifestPath(true)
 	if err != nil {
 		return err
 	}
-	if err := os.MkdirAll(filepath.Dir(destPath), 0o755); err != nil {
-		return err
-	}
 
 	// copy the file
 	input, err := os.ReadFile(srcPath)
@@ -884,7 +882,7 @@ func DeleteModel(name string) error {
 		return err
 	}
 
-	fp, err := mp.GetManifestPath()
+	fp, err := mp.GetManifestPath(false)
 	if err != nil {
 		return err
 	}
@@ -977,7 +975,46 @@ func PushModel(ctx context.Context, name string, regOpts *RegistryOptions, fn fu
 	layers = append(layers, &manifest.Config)
 
 	for _, layer := range layers {
-		if err := uploadBlob(ctx, mp, layer, regOpts, fn); err != nil {
+		exists, err := checkBlobExistence(ctx, mp, layer.Digest, regOpts)
+		if err != nil {
+			return err
+		}
+
+		if exists {
+			fn(api.ProgressResponse{
+				Status:    "using existing layer",
+				Digest:    layer.Digest,
+				Total:     layer.Size,
+				Completed: layer.Size,
+			})
+			log.Printf("Layer %s already exists", layer.Digest)
+			continue
+		}
+
+		fn(api.ProgressResponse{
+			Status: "starting upload",
+			Digest: layer.Digest,
+			Total:  layer.Size,
+		})
+
+		location, chunkSize, err := startUpload(ctx, mp, layer, regOpts)
+		if err != nil {
+			log.Printf("couldn't start upload: %v", err)
+			return err
+		}
+
+		if strings.HasPrefix(filepath.Base(location.Path), "sha256:") {
+			layer.Digest = filepath.Base(location.Path)
+			fn(api.ProgressResponse{
+				Status:    "using existing layer",
+				Digest:    layer.Digest,
+				Total:     layer.Size,
+				Completed: layer.Size,
+			})
+			continue
+		}
+
+		if err := uploadBlob(ctx, location, layer, chunkSize, regOpts, fn); err != nil {
 			log.Printf("error uploading blob: %v", err)
 			return err
 		}
@@ -994,7 +1031,7 @@ func PushModel(ctx context.Context, name string, regOpts *RegistryOptions, fn fu
 
 	headers := make(http.Header)
 	headers.Set("Content-Type", "application/vnd.docker.distribution.manifest.v2+json")
-	resp, err := makeRequestWithRetry(ctx, http.MethodPut, requestURL, headers, bytes.NewReader(manifestJSON), regOpts)
+	resp, err := makeRequestWithRetry(ctx, "PUT", requestURL, headers, bytes.NewReader(manifestJSON), regOpts)
 	if err != nil {
 		return err
 	}
@@ -1084,13 +1121,10 @@ func PullModel(ctx context.Context, name string, regOpts *RegistryOptions, fn fu
 		return err
 	}
 
-	fp, err := mp.GetManifestPath()
+	fp, err := mp.GetManifestPath(true)
 	if err != nil {
 		return err
 	}
-	if err := os.MkdirAll(filepath.Dir(fp), 0o755); err != nil {
-		return err
-	}
 
 	err = os.WriteFile(fp, manifestJSON, 0o644)
 	if err != nil {
@@ -1116,12 +1150,22 @@ func pullModelManifest(ctx context.Context, mp ModelPath, regOpts *RegistryOptio
 
 	headers := make(http.Header)
 	headers.Set("Accept", "application/vnd.docker.distribution.manifest.v2+json")
-	resp, err := makeRequestWithRetry(ctx, http.MethodGet, requestURL, headers, nil, regOpts)
+	resp, err := makeRequest(ctx, "GET", requestURL, headers, nil, regOpts)
 	if err != nil {
+		log.Printf("couldn't get manifest: %v", err)
 		return nil, err
 	}
 	defer resp.Body.Close()
 
+	if resp.StatusCode >= http.StatusBadRequest {
+		if resp.StatusCode == http.StatusNotFound {
+			return nil, fmt.Errorf("model not found")
+		}
+
+		body, _ := io.ReadAll(resp.Body)
+		return nil, fmt.Errorf("on pull registry responded with code %d: %s", resp.StatusCode, body)
+	}
+
 	var m *ManifestV2
 	if err := json.NewDecoder(resp.Body).Decode(&m); err != nil {
 		return nil, err
@@ -1165,7 +1209,24 @@ func GetSHA256Digest(r io.Reader) (string, int64) {
 	return fmt.Sprintf("sha256:%x", h.Sum(nil)), n
 }
 
+// Function to check if a blob already exists in the Docker registry
+func checkBlobExistence(ctx context.Context, mp ModelPath, digest string, regOpts *RegistryOptions) (bool, error) {
+	requestURL := mp.BaseURL()
+	requestURL = requestURL.JoinPath("v2", mp.GetNamespaceRepository(), "blobs", digest)
+
+	resp, err := makeRequest(ctx, "HEAD", requestURL, nil, nil, regOpts)
+	if err != nil {
+		log.Printf("couldn't check for blob: %v", err)
+		return false, err
+	}
+	defer resp.Body.Close()
+
+	// Check for success: If the blob exists, the Docker registry will respond with a 200 OK
+	return resp.StatusCode < http.StatusBadRequest, nil
+}
+
 func makeRequestWithRetry(ctx context.Context, method string, requestURL *url.URL, headers http.Header, body io.ReadSeeker, regOpts *RegistryOptions) (*http.Response, error) {
+	var status string
 	for try := 0; try < maxRetries; try++ {
 		resp, err := makeRequest(ctx, method, requestURL, headers, body, regOpts)
 		if err != nil {
@@ -1173,6 +1234,8 @@ func makeRequestWithRetry(ctx context.Context, method string, requestURL *url.UR
 			return nil, err
 		}
 
+		status = resp.Status
+
 		switch {
 		case resp.StatusCode == http.StatusUnauthorized:
 			auth := resp.Header.Get("www-authenticate")
@@ -1184,25 +1247,21 @@ func makeRequestWithRetry(ctx context.Context, method string, requestURL *url.UR
 
 			regOpts.Token = token
 			if body != nil {
-				body.Seek(0, io.SeekStart)
+				if _, err := body.Seek(0, io.SeekStart); err != nil {
+					return nil, err
+				}
 			}
 
 			continue
-		case resp.StatusCode == http.StatusNotFound:
-			return nil, os.ErrNotExist
 		case resp.StatusCode >= http.StatusBadRequest:
-			body, err := io.ReadAll(resp.Body)
-			if err != nil {
-				return nil, fmt.Errorf("%d: %s", resp.StatusCode, err)
-			}
-
-			return nil, fmt.Errorf("%d: %s", resp.StatusCode, body)
+			body, _ := io.ReadAll(resp.Body)
+			return nil, fmt.Errorf("on upload registry responded with code %d: %s", resp.StatusCode, body)
 		default:
 			return resp, nil
 		}
 	}
 
-	return nil, errMaxRetriesExceeded
+	return nil, fmt.Errorf("max retry exceeded: %v", status)
 }
 
 func makeRequest(ctx context.Context, method string, requestURL *url.URL, headers http.Header, body io.Reader, regOpts *RegistryOptions) (*http.Response, error) {

server/modelpath.go

@@ -85,27 +85,20 @@ func (mp ModelPath) GetShortTagname() string {
 	return fmt.Sprintf("%s/%s/%s:%s", mp.Registry, mp.Namespace, mp.Repository, mp.Tag)
 }
 
-// modelsDir returns the value of the OLLAMA_MODELS environment variable or the user's home directory if OLLAMA_MODELS is not set.
-// The models directory is where Ollama stores its model files and manifests.
-func modelsDir() (string, error) {
-	if models, exists := os.LookupEnv("OLLAMA_MODELS"); exists {
-		return models, nil
-	}
+func (mp ModelPath) GetManifestPath(createDir bool) (string, error) {
 	home, err := os.UserHomeDir()
 	if err != nil {
 		return "", err
 	}
-	return filepath.Join(home, ".ollama", "models"), nil
-}
 
-// GetManifestPath returns the path to the manifest file for the given model path, it is up to the caller to create the directory if it does not exist.
-func (mp ModelPath) GetManifestPath() (string, error) {
-	dir, err := modelsDir()
-	if err != nil {
-		return "", err
+	path := filepath.Join(home, ".ollama", "models", "manifests", mp.Registry, mp.Namespace, mp.Repository, mp.Tag)
+	if createDir {
+		if err := os.MkdirAll(filepath.Dir(path), 0o755); err != nil {
+			return "", err
+		}
 	}
 
-	return filepath.Join(dir, "manifests", mp.Registry, mp.Namespace, mp.Repository, mp.Tag), nil
+	return path, nil
 }
 
 func (mp ModelPath) BaseURL() *url.URL {
@@ -116,12 +109,12 @@ func (mp ModelPath) BaseURL() *url.URL {
 }
 
 func GetManifestPath() (string, error) {
-	dir, err := modelsDir()
+	home, err := os.UserHomeDir()
 	if err != nil {
 		return "", err
 	}
 
-	path := filepath.Join(dir, "manifests")
+	path := filepath.Join(home, ".ollama", "models", "manifests")
 	if err := os.MkdirAll(path, 0o755); err != nil {
 		return "", err
 	}
@@ -130,7 +123,7 @@ func GetManifestPath() (string, error) {
 }
 
 func GetBlobsPath(digest string) (string, error) {
-	dir, err := modelsDir()
+	home, err := os.UserHomeDir()
 	if err != nil {
 		return "", err
 	}
@@ -139,7 +132,7 @@ func GetBlobsPath(digest string) (string, error) {
 		digest = strings.ReplaceAll(digest, ":", "-")
 	}
 
-	path := filepath.Join(dir, "blobs", digest)
+	path := filepath.Join(home, ".ollama", "models", "blobs", digest)
 	dirPath := filepath.Dir(path)
 	if digest == "" {
 		dirPath = path

server/routes.go

@@ -365,9 +365,7 @@ func PushModelHandler(c *gin.Context) {
 			Insecure: req.Insecure,
 		}
 
-		ctx, cancel := context.WithCancel(c.Request.Context())
-		defer cancel()
-
+		ctx := context.Background()
 		if err := PushModel(ctx, req.Name, regOpts, fn); err != nil {
 			ch <- gin.H{"error": err.Error()}
 		}
@@ -616,22 +614,6 @@ var defaultAllowOrigins = []string{
 }
 
 func Serve(ln net.Listener, allowOrigins []string) error {
-	if noprune := os.Getenv("OLLAMA_NOPRUNE"); noprune == "" {
-		// clean up unused layers and manifests
-		if err := PruneLayers(); err != nil {
-			return err
-		}
-
-		manifestsPath, err := GetManifestPath()
-		if err != nil {
-			return err
-		}
-
-		if err := PruneDirectory(manifestsPath); err != nil {
-			return err
-		}
-	}
-
 	config := cors.DefaultConfig()
 	config.AllowWildcard = true
 
@@ -697,7 +679,7 @@ func Serve(ln net.Listener, allowOrigins []string) error {
 	if runtime.GOOS == "linux" {
 		// check compatibility to log warnings
 		if _, err := llm.CheckVRAM(); err != nil {
-			log.Printf("Warning: GPU support may not be enabled, check you have installed GPU drivers: %v", err)
+			log.Printf("Warning: GPU support may not enabled, check you have installed install GPU drivers: %v", err)
 		}
 	}
 

server/upload.go

@@ -2,367 +2,218 @@ package server
 
 import (
 	"context"
-	"crypto/md5"
 	"errors"
 	"fmt"
-	"hash"
 	"io"
 	"log"
 	"net/http"
 	"net/url"
 	"os"
-	"strings"
+	"strconv"
 	"sync"
-	"sync/atomic"
-	"time"
 
 	"github.com/jmorganca/ollama/api"
-	"github.com/jmorganca/ollama/format"
-	"golang.org/x/sync/errgroup"
 )
 
-var blobUploadManager sync.Map
-
-type blobUpload struct {
-	*Layer
-
-	Total     int64
-	Completed atomic.Int64
-
-	Parts []blobUploadPart
-
-	nextURL chan *url.URL
-
-	context.CancelFunc
-
-	done       bool
-	err        error
-	references atomic.Int32
-}
-
-type blobUploadPart struct {
-	// N is the part number
-	N      int
-	Offset int64
-	Size   int64
-	hash.Hash
-}
-
 const (
-	numUploadParts          = 64
-	minUploadPartSize int64 = 95 * 1000 * 1000
-	maxUploadPartSize int64 = 1000 * 1000 * 1000
+	redirectChunkSize int64 = 1024 * 1024 * 1024
+	regularChunkSize  int64 = 95 * 1024 * 1024
 )
 
-func (b *blobUpload) Prepare(ctx context.Context, requestURL *url.URL, opts *RegistryOptions) error {
-	p, err := GetBlobsPath(b.Digest)
-	if err != nil {
-		return err
-	}
-
-	if b.From != "" {
+func startUpload(ctx context.Context, mp ModelPath, layer *Layer, regOpts *RegistryOptions) (*url.URL, int64, error) {
+	requestURL := mp.BaseURL()
+	requestURL = requestURL.JoinPath("v2", mp.GetNamespaceRepository(), "blobs/uploads/")
+	if layer.From != "" {
 		values := requestURL.Query()
-		values.Add("mount", b.Digest)
-		values.Add("from", b.From)
+		values.Add("mount", layer.Digest)
+		values.Add("from", layer.From)
 		requestURL.RawQuery = values.Encode()
 	}
 
-	resp, err := makeRequestWithRetry(ctx, http.MethodPost, requestURL, nil, nil, opts)
+	resp, err := makeRequestWithRetry(ctx, "POST", requestURL, nil, nil, regOpts)
 	if err != nil {
-		return err
+		log.Printf("couldn't start upload: %v", err)
+		return nil, 0, err
 	}
 	defer resp.Body.Close()
 
 	location := resp.Header.Get("Docker-Upload-Location")
+	chunkSize := redirectChunkSize
 	if location == "" {
 		location = resp.Header.Get("Location")
+		chunkSize = regularChunkSize
 	}
 
-	fi, err := os.Stat(p)
+	locationURL, err := url.Parse(location)
 	if err != nil {
-		return err
+		return nil, 0, err
 	}
 
-	b.Total = fi.Size()
-
-	var size = b.Total / numUploadParts
-	switch {
-	case size < minUploadPartSize:
-		size = minUploadPartSize
-	case size > maxUploadPartSize:
-		size = maxUploadPartSize
-	}
-
-	var offset int64
-	for offset < fi.Size() {
-		if offset+size > fi.Size() {
-			size = fi.Size() - offset
-		}
-
-		// set part.N to the current number of parts
-		b.Parts = append(b.Parts, blobUploadPart{N: len(b.Parts), Offset: offset, Size: size, Hash: md5.New()})
-		offset += size
-	}
-
-	log.Printf("uploading %s in %d %s part(s)", b.Digest[7:19], len(b.Parts), format.HumanBytes(b.Parts[0].Size))
-
-	requestURL, err = url.Parse(location)
-	if err != nil {
-		return err
-	}
-
-	b.nextURL = make(chan *url.URL, 1)
-	b.nextURL <- requestURL
-	return nil
+	return locationURL, chunkSize, nil
 }
 
-// Run uploads blob parts to the upstream. If the upstream supports redirection, parts will be uploaded
-// in parallel as defined by Prepare. Otherwise, parts will be uploaded serially. Run sets b.err on error.
-func (b *blobUpload) Run(ctx context.Context, opts *RegistryOptions) {
-	defer blobUploadManager.Delete(b.Digest)
-	ctx, b.CancelFunc = context.WithCancel(ctx)
+func uploadBlob(ctx context.Context, requestURL *url.URL, layer *Layer, chunkSize int64, regOpts *RegistryOptions, fn func(api.ProgressResponse)) error {
+	// TODO allow resumability
+	// TODO allow canceling uploads via DELETE
 
-	p, err := GetBlobsPath(b.Digest)
+	fp, err := GetBlobsPath(layer.Digest)
 	if err != nil {
-		b.err = err
-		return
+		return err
 	}
 
-	f, err := os.Open(p)
+	f, err := os.Open(fp)
 	if err != nil {
-		b.err = err
-		return
+		return err
 	}
 	defer f.Close()
 
-	g, inner := errgroup.WithContext(ctx)
-	g.SetLimit(numUploadParts)
-	for i := range b.Parts {
-		part := &b.Parts[i]
-		select {
-		case <-inner.Done():
-		case requestURL := <-b.nextURL:
-			g.Go(func() error {
-				for try := 0; try < maxRetries; try++ {
-					r := io.NewSectionReader(f, part.Offset, part.Size)
-					err := b.uploadChunk(inner, http.MethodPatch, requestURL, r, part, opts)
-					switch {
-					case errors.Is(err, context.Canceled):
-						return err
-					case errors.Is(err, errMaxRetriesExceeded):
-						return err
-					case err != nil:
-						log.Printf("%s part %d attempt %d failed: %v, retrying", b.Digest[7:19], part.N, try, err)
-						continue
-					}
+	pw := ProgressWriter{
+		status: fmt.Sprintf("uploading %s", layer.Digest),
+		digest: layer.Digest,
+		total:  layer.Size,
+		fn:     fn,
+	}
 
-					return nil
-				}
+	for offset := int64(0); offset < layer.Size; {
+		chunk := layer.Size - offset
+		if chunk > chunkSize {
+			chunk = chunkSize
+		}
 
-				return errMaxRetriesExceeded
+		resp, err := uploadBlobChunk(ctx, http.MethodPatch, requestURL, f, offset, chunk, regOpts, &pw)
+		if err != nil {
+			fn(api.ProgressResponse{
+				Status:    fmt.Sprintf("error uploading chunk: %v", err),
+				Digest:    layer.Digest,
+				Total:     layer.Size,
+				Completed: offset,
 			})
+
+			return err
+		}
+
+		offset += chunk
+		location := resp.Header.Get("Docker-Upload-Location")
+		if location == "" {
+			location = resp.Header.Get("Location")
+		}
+
+		requestURL, err = url.Parse(location)
+		if err != nil {
+			return err
 		}
 	}
 
-	if err := g.Wait(); err != nil {
-		b.err = err
-		return
-	}
-
-	requestURL := <-b.nextURL
-
-	var sb strings.Builder
-	for _, part := range b.Parts {
-		sb.Write(part.Sum(nil))
-	}
-
-	md5sum := md5.Sum([]byte(sb.String()))
-
 	values := requestURL.Query()
-	values.Add("digest", b.Digest)
-	values.Add("etag", fmt.Sprintf("%x-%d", md5sum, len(b.Parts)))
+	values.Add("digest", layer.Digest)
 	requestURL.RawQuery = values.Encode()
 
 	headers := make(http.Header)
 	headers.Set("Content-Type", "application/octet-stream")
 	headers.Set("Content-Length", "0")
 
-	resp, err := makeRequestWithRetry(ctx, http.MethodPut, requestURL, headers, nil, opts)
-	if err != nil {
-		b.err = err
-		return
-	}
-	defer resp.Body.Close()
-
-	b.done = true
-}
-
-func (b *blobUpload) uploadChunk(ctx context.Context, method string, requestURL *url.URL, rs io.ReadSeeker, part *blobUploadPart, opts *RegistryOptions) error {
-	headers := make(http.Header)
-	headers.Set("Content-Type", "application/octet-stream")
-	headers.Set("Content-Length", fmt.Sprintf("%d", part.Size))
-	headers.Set("X-Redirect-Uploads", "1")
-
-	if method == http.MethodPatch {
-		headers.Set("Content-Range", fmt.Sprintf("%d-%d", part.Offset, part.Offset+part.Size-1))
-	}
-
-	buw := blobUploadWriter{blobUpload: b}
-	resp, err := makeRequest(ctx, method, requestURL, headers, io.TeeReader(rs, io.MultiWriter(&buw, part.Hash)), opts)
+	// finish the upload
+	resp, err := makeRequest(ctx, "PUT", requestURL, headers, nil, regOpts)
 	if err != nil {
+		log.Printf("couldn't finish upload: %v", err)
 		return err
 	}
 	defer resp.Body.Close()
 
-	location := resp.Header.Get("Docker-Upload-Location")
-	if location == "" {
-		location = resp.Header.Get("Location")
+	if resp.StatusCode >= http.StatusBadRequest {
+		body, _ := io.ReadAll(resp.Body)
+		return fmt.Errorf("on finish upload registry responded with code %d: %v", resp.StatusCode, string(body))
 	}
-
-	nextURL, err := url.Parse(location)
-	if err != nil {
-		return err
-	}
-
-	switch {
-	case resp.StatusCode == http.StatusTemporaryRedirect:
-		b.nextURL <- nextURL
-
-		redirectURL, err := resp.Location()
-		if err != nil {
-			return err
-		}
-
-		for try := 0; try < maxRetries; try++ {
-			rs.Seek(0, io.SeekStart)
-			b.Completed.Add(-buw.written)
-			buw.written = 0
-			part.Hash = md5.New()
-			err := b.uploadChunk(ctx, http.MethodPut, redirectURL, rs, part, nil)
-			switch {
-			case errors.Is(err, context.Canceled):
-				return err
-			case errors.Is(err, errMaxRetriesExceeded):
-				return err
-			case err != nil:
-				log.Printf("%s part %d attempt %d failed: %v, retrying", b.Digest[7:19], part.N, try, err)
-				continue
-			}
-
-			return nil
-		}
-
-		return errMaxRetriesExceeded
-
-	case resp.StatusCode == http.StatusUnauthorized:
-		auth := resp.Header.Get("www-authenticate")
-		authRedir := ParseAuthRedirectString(auth)
-		token, err := getAuthToken(ctx, authRedir)
-		if err != nil {
-			return err
-		}
-
-		opts.Token = token
-		fallthrough
-	case resp.StatusCode >= http.StatusBadRequest:
-		body, err := io.ReadAll(resp.Body)
-		if err != nil {
-			return err
-		}
-
-		rs.Seek(0, io.SeekStart)
-		b.Completed.Add(-buw.written)
-		buw.written = 0
-		return fmt.Errorf("http status %d %s: %s", resp.StatusCode, resp.Status, body)
-	}
-
-	if method == http.MethodPatch {
-		b.nextURL <- nextURL
-	}
-
 	return nil
 }
 
-func (b *blobUpload) acquire() {
-	b.references.Add(1)
-}
+func uploadBlobChunk(ctx context.Context, method string, requestURL *url.URL, r io.ReaderAt, offset, limit int64, opts *RegistryOptions, pw *ProgressWriter) (*http.Response, error) {
+	sectionReader := io.NewSectionReader(r, offset, limit)
 
-func (b *blobUpload) release() {
-	if b.references.Add(-1) == 0 {
-		b.CancelFunc()
+	headers := make(http.Header)
+	headers.Set("Content-Type", "application/octet-stream")
+	headers.Set("Content-Length", strconv.Itoa(int(limit)))
+	headers.Set("X-Redirect-Uploads", "1")
+
+	if method == http.MethodPatch {
+		headers.Set("Content-Range", fmt.Sprintf("%d-%d", offset, offset+sectionReader.Size()-1))
 	}
-}
 
-func (b *blobUpload) Wait(ctx context.Context, fn func(api.ProgressResponse)) error {
-	b.acquire()
-	defer b.release()
+	for try := 0; try < maxRetries; try++ {
+		resp, err := makeRequest(ctx, method, requestURL, headers, io.TeeReader(sectionReader, pw), opts)
+		if err != nil && !errors.Is(err, io.EOF) {
+			return nil, err
+		}
+		defer resp.Body.Close()
 
-	ticker := time.NewTicker(60 * time.Millisecond)
-	for {
-		select {
-		case <-ticker.C:
-		case <-ctx.Done():
-			return ctx.Err()
+		switch {
+		case resp.StatusCode == http.StatusTemporaryRedirect:
+			location, err := resp.Location()
+			if err != nil {
+				return nil, err
+			}
+
+			pw.completed = offset
+			if _, err := uploadBlobChunk(ctx, http.MethodPut, location, r, offset, limit, nil, pw); err != nil {
+				// retry
+				log.Printf("retrying redirected upload: %v", err)
+				continue
+			}
+
+			return resp, nil
+		case resp.StatusCode == http.StatusUnauthorized:
+			auth := resp.Header.Get("www-authenticate")
+			authRedir := ParseAuthRedirectString(auth)
+			token, err := getAuthToken(ctx, authRedir)
+			if err != nil {
+				return nil, err
+			}
+
+			opts.Token = token
+
+			pw.completed = offset
+			sectionReader = io.NewSectionReader(r, offset, limit)
+			continue
+		case resp.StatusCode >= http.StatusBadRequest:
+			body, _ := io.ReadAll(resp.Body)
+			return nil, fmt.Errorf("on upload registry responded with code %d: %s", resp.StatusCode, body)
 		}
 
-		fn(api.ProgressResponse{
-			Status:    fmt.Sprintf("uploading %s", b.Digest),
-			Digest:    b.Digest,
-			Total:     b.Total,
-			Completed: b.Completed.Load(),
+		return resp, nil
+	}
+
+	return nil, fmt.Errorf("max retries exceeded")
+}
+
+type ProgressWriter struct {
+	status    string
+	digest    string
+	bucket    int64
+	completed int64
+	total     int64
+	fn        func(api.ProgressResponse)
+	mu        sync.Mutex
+}
+
+func (pw *ProgressWriter) Write(b []byte) (int, error) {
+	pw.mu.Lock()
+	defer pw.mu.Unlock()
+
+	n := len(b)
+	pw.bucket += int64(n)
+
+	// throttle status updates to not spam the client
+	if pw.bucket >= 1024*1024 || pw.completed+pw.bucket >= pw.total {
+		pw.completed += pw.bucket
+		pw.fn(api.ProgressResponse{
+			Status:    pw.status,
+			Digest:    pw.digest,
+			Total:     pw.total,
+			Completed: pw.completed,
 		})
 
-		if b.done || b.err != nil {
-			return b.err
-		}
+		pw.bucket = 0
 	}
-}
 
-type blobUploadWriter struct {
-	written int64
-	*blobUpload
-}
-
-func (b *blobUploadWriter) Write(p []byte) (n int, err error) {
-	n = len(p)
-	b.written += int64(n)
-	b.Completed.Add(int64(n))
 	return n, nil
 }
-
-func uploadBlob(ctx context.Context, mp ModelPath, layer *Layer, opts *RegistryOptions, fn func(api.ProgressResponse)) error {
-	requestURL := mp.BaseURL()
-	requestURL = requestURL.JoinPath("v2", mp.GetNamespaceRepository(), "blobs", layer.Digest)
-
-	resp, err := makeRequestWithRetry(ctx, http.MethodHead, requestURL, nil, nil, opts)
-	switch {
-	case errors.Is(err, os.ErrNotExist):
-	case err != nil:
-		return err
-	default:
-		defer resp.Body.Close()
-		fn(api.ProgressResponse{
-			Status:    fmt.Sprintf("uploading %s", layer.Digest),
-			Digest:    layer.Digest,
-			Total:     layer.Size,
-			Completed: layer.Size,
-		})
-
-		return nil
-	}
-
-	data, ok := blobUploadManager.LoadOrStore(layer.Digest, &blobUpload{Layer: layer})
-	upload := data.(*blobUpload)
-	if !ok {
-		requestURL := mp.BaseURL()
-		requestURL = requestURL.JoinPath("v2", mp.GetNamespaceRepository(), "blobs/uploads/")
-		if err := upload.Prepare(ctx, requestURL, opts); err != nil {
-			blobUploadManager.Delete(layer.Digest)
-			return err
-		}
-
-		go upload.Run(context.Background(), opts)
-	}
-
-	return upload.Wait(ctx, fn)
-}