fix ref

.gitattributes

@@ -1,4 +1,3 @@
 llm/ext_server/* linguist-vendored
-llm/*.h linguist-vendored
 * text=auto
 *.go text eol=lf

app/ollama.iss

@@ -87,7 +87,7 @@ DialogFontSize=12
 
 [Files]
 Source: ".\app.exe"; DestDir: "{app}"; DestName: "{#MyAppExeName}" ; Flags: ignoreversion 64bit
-Source: "..\ollama.exe"; DestDir: "{app}"; Flags: ignoreversion 64bit
+Source: "..\ollama.exe"; DestDir: "{app}\bin"; Flags: ignoreversion 64bit
 Source: "..\dist\windows-{#ARCH}\lib\ollama\runners\*"; DestDir: "{app}\lib\ollama\runners"; Flags: ignoreversion 64bit recursesubdirs
 Source: "..\dist\ollama_welcome.ps1"; DestDir: "{app}"; Flags: ignoreversion
 Source: ".\assets\app.ico"; DestDir: "{app}"; Flags: ignoreversion
@@ -99,7 +99,7 @@ Name: "{userstartup}\{#MyAppName}"; Filename: "{app}\{#MyAppExeName}"; IconFilen
 Name: "{userprograms}\{#MyAppName}"; Filename: "{app}\{#MyAppExeName}"; IconFilename: "{app}\app.ico"
 
 [Run]
-Filename: "{cmd}"; Parameters: "/C set PATH={app};%PATH% & ""{app}\{#MyAppExeName}"""; Flags: postinstall nowait runhidden
+Filename: "{cmd}"; Parameters: "/C set PATH={app}\bin;%PATH% & ""{app}\{#MyAppExeName}"""; Flags: postinstall nowait runhidden
 
 [UninstallRun]
 ; Filename: "{cmd}"; Parameters: "/C ""taskkill /im ''{#MyAppExeName}'' /f /t"; Flags: runhidden
@@ -134,8 +134,8 @@ SetupAppRunningError=Another Ollama installer is running.%n%nPlease cancel or fi
 
 [Registry]
 Root: HKCU; Subkey: "Environment"; \
-    ValueType: expandsz; ValueName: "Path"; ValueData: "{olddata};{app}"; \
-    Check: NeedsAddPath('{app}')
+    ValueType: expandsz; ValueName: "Path"; ValueData: "{olddata};{app}\bin"; \
+    Check: NeedsAddPath('{app}\bin')
 
 [Code]
 

cmd/cmd.go

@@ -124,7 +124,6 @@ func CreateHandler(cmd *cobra.Command, args []string) error {
 	}
 
 	bars := make(map[string]*progress.Bar)
-	var quantizeSpin *progress.Spinner
 	fn := func(resp api.ProgressResponse) error {
 		if resp.Digest != "" {
 			spinner.Stop()
@@ -137,15 +136,6 @@ func CreateHandler(cmd *cobra.Command, args []string) error {
 			}
 
 			bar.Set(resp.Completed)
-		} else if strings.Contains(resp.Status, "quantizing") {
-			spinner.Stop()
-
-			if quantizeSpin != nil {
-				quantizeSpin.SetMessage(resp.Status)
-			} else {
-				quantizeSpin = progress.NewSpinner(resp.Status)
-				p.Add("quantize", quantizeSpin)
-			}
 		} else if status != resp.Status {
 			spinner.Stop()
 

docs/import.md

@@ -158,7 +158,7 @@ You can share any model you have created by pushing it to [ollama.com](https://o
 
 First, use your browser to go to the [Ollama Sign-Up](https://ollama.com/signup) page. If you already have an account, you can skip this step.
 
-<img src="images/signup.png" alt="Sign-Up" width="40%">
+![Sign-Up](images/signup.png)
 
 The `Username` field will be used as part of your model's name (e.g. `jmorganca/mymodel`), so make sure you are comfortable with the username that you have selected.
 
@@ -166,7 +166,7 @@ Now that you have created an account and are signed-in, go to the [Ollama Keys S
 
 Follow the directions on the page to determine where your Ollama Public Key is located.
 
-<img src="images/ollama-keys.png" alt="Ollama Keys" width="80%">
+![Ollama Key](images/ollama-keys.png)
 
 Click on the `Add Ollama Public Key` button, and copy and paste the contents of your Ollama Public Key into the text field.
 

docs/linux.md

@@ -28,11 +28,6 @@ Download and extract the Linux package:
 curl -fsSL https://ollama.com/download/ollama-linux-amd64.tgz | sudo tar zx -C /usr
 ```
 
-If you have an AMD GPU, also download and extract the ROCm package into the same location
-```bash
-curl -fsSL https://ollama.com/download/ollama-linux-amd64-rocm.tgz | sudo tar zx -C /usr
-```
-
 ### Adding Ollama as a startup service (recommended)
 
 Create a user for Ollama:

docs/modelfile.md

@@ -11,9 +11,8 @@ A model file is the blueprint to create and share models with Ollama.
 - [Examples](#examples)
 - [Instructions](#instructions)
   - [FROM (Required)](#from-required)
-    - [Build from llama3.1](#build-from-llama31)
-    - [Build from a Safetensors model](#build-from-a-safetensors-model)
-    - [Build from a GGUF file](#build-from-a-gguf-file)
+    - [Build from llama3](#build-from-llama3)
+    - [Build from a bin file](#build-from-a-bin-file)
   - [PARAMETER](#parameter)
     - [Valid Parameters and Values](#valid-parameters-and-values)
   - [TEMPLATE](#template)
@@ -100,39 +99,22 @@ The `FROM` instruction defines the base model to use when creating a model.
 FROM <model name>:<tag>
 ```
 
-#### Build from llama3.1
+#### Build from llama3
 
 ```modelfile
-FROM llama3.1
+FROM llama3
 ```
 
 A list of available base models:
 <https://github.com/ollama/ollama#model-library>
-Additional models can be found at:
-<https://ollama.com/library>
 
-#### Build from a Safetensors model
-
-```modelfile
-FROM <model directory>
-```
-
-The model directory should contain the Safetensors weights for a supported architecture.
-
-Currently supported model architectures:
-  * Llama (including Llama 2, Llama 3, and Llama 3.1)
-  * Mistral (including Mistral 1, Mistral 2, and Mixtral)
-  * Gemma (including Gemma 1 and Gemma 2)
-  * Phi3
-
-#### Build from a GGUF file
+#### Build from a `bin` file
 
 ```modelfile
 FROM ./ollama-model.bin
 ```
 
-The GGUF bin file location should be specified as an absolute path or relative to the `Modelfile` location.
-
+This bin file location should be specified as an absolute path or relative to the `Modelfile` location.
 
 ### PARAMETER
 
@@ -192,20 +174,7 @@ SYSTEM """<system message>"""
 
 ### ADAPTER
 
-The `ADAPTER` instruction specifies a fine tuned LoRA adapter that should apply to the base model. The value of the adapter should be an absolute path or a path relative to the Modelfile. The base model should be specified with a `FROM` instruction. If the base model is not the same as the base model that the adapter was tuned from the behaviour will be erratic.
-
-#### Safetensor adapter
-
-```modelfile
-ADAPTER <path to safetensor adapter>
-```
-
-Currently supported Safetensor adapters:
-  * Llama (including Llama 2, Llama 3, and Llama 3.1)
-  * Mistral (including Mistral 1, Mistral 2, and Mixtral)
-  * Gemma (including Gemma 1 and Gemma 2)
-
-#### GGUF adapter
+The `ADAPTER` instruction is an optional instruction that specifies any LoRA adapter that should apply to the base model. The value of this instruction should be an absolute path or a path relative to the Modelfile and the file must be in a GGML file format. The adapter should be tuned from the base model otherwise the behaviour is undefined.
 
 ```modelfile
 ADAPTER ./ollama-lora.bin

envconfig/config.go

@@ -190,7 +190,7 @@ func RunnersDir() (p string) {
 	}
 
 	var paths []string
-	for _, root := range []string{filepath.Dir(exe), filepath.Join(filepath.Dir(exe), LibRelativeToExe()), cwd} {
+	for _, root := range []string{filepath.Dir(exe), filepath.Join(filepath.Dir(exe), ".."), cwd} {
 		paths = append(paths,
 			root,
 			filepath.Join(root, runtime.GOOS+"-"+runtime.GOARCH),
@@ -282,12 +282,3 @@ func Values() map[string]string {
 func Var(key string) string {
 	return strings.Trim(strings.TrimSpace(os.Getenv(key)), "\"'")
 }
-
-// On windows, we keep the binary at the top directory, but
-// other platforms use a "bin" directory, so this returns ".."
-func LibRelativeToExe() string {
-	if runtime.GOOS == "windows" {
-		return "."
-	}
-	return ".."
-}

gpu/amd_common.go

@@ -9,8 +9,6 @@ import (
 	"path/filepath"
 	"runtime"
 	"strings"
-
-	"github.com/ollama/ollama/envconfig"
 )
 
 // Determine if the given ROCm lib directory is usable by checking for existence of some glob patterns
@@ -56,7 +54,7 @@ func commonAMDValidateLibDir() (string, error) {
 	// Installer payload location if we're running the installed binary
 	exe, err := os.Executable()
 	if err == nil {
-		rocmTargetDir := filepath.Join(filepath.Dir(exe), envconfig.LibRelativeToExe(), "lib", "ollama")
+		rocmTargetDir := filepath.Join(filepath.Dir(exe), "..", "lib", "ollama")
 		if rocmLibUsable(rocmTargetDir) {
 			slog.Debug("detected ROCM next to ollama executable " + rocmTargetDir)
 			return rocmTargetDir, nil

gpu/amd_windows.go

@@ -153,7 +153,7 @@ func AMDValidateLibDir() (string, error) {
 	// Installer payload (if we're running from some other location)
 	localAppData := os.Getenv("LOCALAPPDATA")
 	appDir := filepath.Join(localAppData, "Programs", "Ollama")
-	rocmTargetDir := filepath.Join(appDir, envconfig.LibRelativeToExe(), "lib", "ollama")
+	rocmTargetDir := filepath.Join(appDir, "..", "lib", "ollama")
 	if rocmLibUsable(rocmTargetDir) {
 		slog.Debug("detected ollama installed ROCm at " + rocmTargetDir)
 		return rocmTargetDir, nil

gpu/gpu.go

@@ -653,7 +653,7 @@ func LibraryDir() string {
 		slog.Warn("failed to lookup working directory", "error", err)
 	}
 	// Scan for any of our dependeices, and pick first match
-	for _, root := range []string{filepath.Dir(appExe), filepath.Join(filepath.Dir(appExe), envconfig.LibRelativeToExe()), cwd} {
+	for _, root := range []string{filepath.Dir(appExe), filepath.Join(filepath.Dir(appExe), ".."), cwd} {
 		libDep := filepath.Join("lib", "ollama")
 		if _, err := os.Stat(filepath.Join(root, libDep)); err == nil {
 			return filepath.Join(root, libDep)

llm/llm.go

@@ -1,6 +1,6 @@
 package llm
 
-// #cgo CPPFLAGS: -Illama.cpp/ggml/include
+// #cgo CFLAGS: -Illama.cpp -Illama.cpp/include -Illama.cpp/ggml/include
 // #cgo LDFLAGS: -lllama -lggml -lstdc++ -lpthread
 // #cgo darwin,arm64 LDFLAGS: -L${SRCDIR}/build/darwin/arm64_static -L${SRCDIR}/build/darwin/arm64_static/src -L${SRCDIR}/build/darwin/arm64_static/ggml/src -framework Accelerate -framework Metal
 // #cgo darwin,amd64 LDFLAGS: -L${SRCDIR}/build/darwin/x86_64_static -L${SRCDIR}/build/darwin/x86_64_static/src -L${SRCDIR}/build/darwin/x86_64_static/ggml/src
@@ -9,24 +9,12 @@ package llm
 // #cgo linux,amd64 LDFLAGS: -L${SRCDIR}/build/linux/x86_64_static -L${SRCDIR}/build/linux/x86_64_static/src -L${SRCDIR}/build/linux/x86_64_static/ggml/src
 // #cgo linux,arm64 LDFLAGS: -L${SRCDIR}/build/linux/arm64_static -L${SRCDIR}/build/linux/arm64_static/src -L${SRCDIR}/build/linux/arm64_static/ggml/src
 // #include <stdlib.h>
-// #include <stdatomic.h>
 // #include "llama.h"
-// bool update_quantize_progress(float progress, void* data) {
-//	atomic_int* atomicData = (atomic_int*)data;
-//  int intProgress = *((int*)&progress);
-//  atomic_store(atomicData, intProgress);
-//  return true;
-// }
 import "C"
 
 import (
 	"errors"
-	"fmt"
-	"sync/atomic"
-	"time"
 	"unsafe"
-
-	"github.com/ollama/ollama/api"
 )
 
 // SystemInfo is an unused example of calling llama.cpp functions using CGo
@@ -34,49 +22,17 @@ func SystemInfo() string {
 	return C.GoString(C.llama_print_system_info())
 }
 
-func Quantize(infile, outfile string, ftype fileType, fn func(resp api.ProgressResponse), tensorCount int) error {
+func Quantize(infile, outfile string, ftype fileType) error {
 	cinfile := C.CString(infile)
 	defer C.free(unsafe.Pointer(cinfile))
 
 	coutfile := C.CString(outfile)
 	defer C.free(unsafe.Pointer(coutfile))
+
 	params := C.llama_model_quantize_default_params()
 	params.nthread = -1
 	params.ftype = ftype.Value()
 
-	// Initialize "global" to store progress
-	store := (*int32)(C.malloc(C.sizeof_int))
-	defer C.free(unsafe.Pointer(store))
-
-	// Initialize store value, e.g., setting initial progress to 0
-	atomic.StoreInt32(store, 0)
-
-	params.quantize_callback_data = unsafe.Pointer(store)
-	params.quantize_callback = (C.llama_progress_callback)(C.update_quantize_progress)
-
-	ticker := time.NewTicker(30 * time.Millisecond)
-	done := make(chan struct{})
-	defer close(done)
-
-	go func() {
-		defer ticker.Stop()
-		for {
-			select {
-			case <-ticker.C:
-				progressInt := atomic.LoadInt32(store)
-				progress := *(*float32)(unsafe.Pointer(&progressInt))
-				fn(api.ProgressResponse{
-					Status: fmt.Sprintf("quantizing model %d%%", 100*int(progress)/tensorCount),
-				})
-			case <-done:
-				fn(api.ProgressResponse{
-					Status: fmt.Sprintf("quantizing model 100%%"),
-				})
-				return
-			}
-		}
-	}()
-
 	if rc := C.llama_model_quantize(cinfile, coutfile, &params); rc != 0 {
 		return errors.New("failed to quantize model. This model architecture may not be supported, or you may need to upgrade Ollama to the latest version")
 	}

llm/patches/10-quantize-callback.diff

@@ -1,52 +0,0 @@
-From ed941590d59fc07b1ad21d6aa458588e47d1e446 Mon Sep 17 00:00:00 2001
-From: Josh Yan <jyan00017@gmail.com>
-Date: Wed, 10 Jul 2024 13:39:39 -0700
-Subject: [PATCH] quantize progress
-
----
- include/llama.h | 3 +++
- src/llama.cpp   | 8 ++++++++
- 2 files changed, 11 insertions(+)
-
-diff --git a/include/llama.h b/include/llama.h
-index bb4b05ba..613db68e 100644
---- a/include/llama.h
-+++ b/include/llama.h
-@@ -349,6 +349,9 @@ extern "C" {
-         bool keep_split;                     // quantize to the same number of shards
-         void * imatrix;                      // pointer to importance matrix data
-         void * kv_overrides;                 // pointer to vector containing overrides
-+
-+        llama_progress_callback quantize_callback;   // callback to report quantization progress
-+        void * quantize_callback_data;               // user data for the callback
-     } llama_model_quantize_params;
- 
-     // grammar types
-diff --git a/src/llama.cpp b/src/llama.cpp
-index 2b9ace28..ac640c02 100644
---- a/src/llama.cpp
-+++ b/src/llama.cpp
-@@ -18252,6 +18252,12 @@ static void llama_model_quantize_internal(const std::string & fname_inp, const s
-     const auto tn = LLM_TN(model.arch);
-     new_ofstream(0);
-     for (int i = 0; i < ml.n_tensors; ++i) {
-+        if (params->quantize_callback){
-+            if (!params->quantize_callback(i, params->quantize_callback_data)) {
-+                return;
-+            }
-+        }
-+
-         auto weight = ml.get_weight(i);
-         struct ggml_tensor * tensor = weight->tensor;
-         if (weight->idx != cur_split && params->keep_split) {
-@@ -18789,6 +18795,8 @@ struct llama_model_quantize_params llama_model_quantize_default_params() {
-         /*.keep_split                  =*/ false,
-         /*.imatrix                     =*/ nullptr,
-         /*.kv_overrides                =*/ nullptr,
-+        /*.quantize_callback           =*/ nullptr,
-+        /*.quantize_callback_data      =*/ nullptr,
-     };
- 
-     return result;
--- 
-2.39.3 (Apple Git-146)

scripts/build_windows.ps1

@@ -122,8 +122,8 @@ function buildOllama() {
             /csp "Google Cloud KMS Provider" /kc ${env:KEY_CONTAINER} ollama.exe
         if ($LASTEXITCODE -ne 0) { exit($LASTEXITCODE)}
     }
-    New-Item -ItemType Directory -Path .\dist\windows-${script:TARGET_ARCH}\ -Force
-    cp .\ollama.exe .\dist\windows-${script:TARGET_ARCH}\
+    New-Item -ItemType Directory -Path .\dist\windows-${script:TARGET_ARCH}\bin\ -Force
+    cp .\ollama.exe .\dist\windows-${script:TARGET_ARCH}\bin\
 }
 
 function buildApp() {

server/images.go

@@ -435,14 +435,11 @@ func CreateModel(ctx context.Context, name model.Name, modelFileDir, quantizatio
 						return err
 					}
 
-					tensorCount := len(baseLayer.GGML.Tensors().Items)
 					ft := baseLayer.GGML.KV().FileType()
 					if !slices.Contains([]string{"F16", "F32"}, ft.String()) {
 						return errors.New("quantization is only supported for F16 and F32 models")
 					} else if want != ft {
-						fn(api.ProgressResponse{
-							Status: "quantizing model tensors",
-						})
+						fn(api.ProgressResponse{Status: fmt.Sprintf("quantizing %s model to %s", ft, quantization)})
 
 						blob, err := GetBlobsPath(baseLayer.Digest)
 						if err != nil {
@@ -456,7 +453,7 @@ func CreateModel(ctx context.Context, name model.Name, modelFileDir, quantizatio
 						defer temp.Close()
 						defer os.Remove(temp.Name())
 
-						if err := llm.Quantize(blob, temp.Name(), want, fn, tensorCount); err != nil {
+						if err := llm.Quantize(blob, temp.Name(), want); err != nil {
 							return err
 						}
 

server/routes.go

@@ -463,7 +463,7 @@ func (s *Server) EmbeddingsHandler(c *gin.Context) {
 	c.JSON(http.StatusOK, resp)
 }
 
-func (s *Server) PullHandler(c *gin.Context) {
+func (s *Server) PullModelHandler(c *gin.Context) {
 	var req api.PullRequest
 	err := c.ShouldBindJSON(&req)
 	switch {
@@ -513,7 +513,7 @@ func (s *Server) PullHandler(c *gin.Context) {
 	streamResponse(c, ch)
 }
 
-func (s *Server) PushHandler(c *gin.Context) {
+func (s *Server) PushModelHandler(c *gin.Context) {
 	var req api.PushRequest
 	err := c.ShouldBindJSON(&req)
 	switch {
@@ -577,7 +577,7 @@ func checkNameExists(name model.Name) error {
 	return nil
 }
 
-func (s *Server) CreateHandler(c *gin.Context) {
+func (s *Server) CreateModelHandler(c *gin.Context) {
 	var r api.CreateRequest
 	if err := c.ShouldBindJSON(&r); errors.Is(err, io.EOF) {
 		c.AbortWithStatusJSON(http.StatusBadRequest, gin.H{"error": "missing request body"})
@@ -647,7 +647,7 @@ func (s *Server) CreateHandler(c *gin.Context) {
 	streamResponse(c, ch)
 }
 
-func (s *Server) DeleteHandler(c *gin.Context) {
+func (s *Server) DeleteModelHandler(c *gin.Context) {
 	var r api.DeleteRequest
 	if err := c.ShouldBindJSON(&r); errors.Is(err, io.EOF) {
 		c.AbortWithStatusJSON(http.StatusBadRequest, gin.H{"error": "missing request body"})
@@ -680,7 +680,7 @@ func (s *Server) DeleteHandler(c *gin.Context) {
 	}
 }
 
-func (s *Server) ShowHandler(c *gin.Context) {
+func (s *Server) ShowModelHandler(c *gin.Context) {
 	var req api.ShowRequest
 	err := c.ShouldBindJSON(&req)
 	switch {
@@ -829,7 +829,7 @@ func getKVData(digest string, verbose bool) (llm.KV, error) {
 	return kv, nil
 }
 
-func (s *Server) ListHandler(c *gin.Context) {
+func (s *Server) ListModelsHandler(c *gin.Context) {
 	ms, err := Manifests()
 	if err != nil {
 		c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
@@ -879,7 +879,7 @@ func (s *Server) ListHandler(c *gin.Context) {
 	c.JSON(http.StatusOK, api.ListResponse{Models: models})
 }
 
-func (s *Server) CopyHandler(c *gin.Context) {
+func (s *Server) CopyModelHandler(c *gin.Context) {
 	var r api.CopyRequest
 	if err := c.ShouldBindJSON(&r); errors.Is(err, io.EOF) {
 		c.AbortWithStatusJSON(http.StatusBadRequest, gin.H{"error": "missing request body"})
@@ -1081,33 +1081,33 @@ func (s *Server) GenerateRoutes() http.Handler {
 		allowedHostsMiddleware(s.addr),
 	)
 
-	r.POST("/api/pull", s.PullHandler)
+	r.POST("/api/pull", s.PullModelHandler)
 	r.POST("/api/generate", s.GenerateHandler)
 	r.POST("/api/chat", s.ChatHandler)
 	r.POST("/api/embed", s.EmbedHandler)
 	r.POST("/api/embeddings", s.EmbeddingsHandler)
-	r.POST("/api/create", s.CreateHandler)
-	r.POST("/api/push", s.PushHandler)
-	r.POST("/api/copy", s.CopyHandler)
-	r.DELETE("/api/delete", s.DeleteHandler)
-	r.POST("/api/show", s.ShowHandler)
+	r.POST("/api/create", s.CreateModelHandler)
+	r.POST("/api/push", s.PushModelHandler)
+	r.POST("/api/copy", s.CopyModelHandler)
+	r.DELETE("/api/delete", s.DeleteModelHandler)
+	r.POST("/api/show", s.ShowModelHandler)
 	r.POST("/api/blobs/:digest", s.CreateBlobHandler)
 	r.HEAD("/api/blobs/:digest", s.HeadBlobHandler)
-	r.GET("/api/ps", s.PsHandler)
+	r.GET("/api/ps", s.ProcessHandler)
 
 	// Compatibility endpoints
 	r.POST("/v1/chat/completions", openai.ChatMiddleware(), s.ChatHandler)
 	r.POST("/v1/completions", openai.CompletionsMiddleware(), s.GenerateHandler)
 	r.POST("/v1/embeddings", openai.EmbeddingsMiddleware(), s.EmbedHandler)
-	r.GET("/v1/models", openai.ListMiddleware(), s.ListHandler)
-	r.GET("/v1/models/:model", openai.RetrieveMiddleware(), s.ShowHandler)
+	r.GET("/v1/models", openai.ListMiddleware(), s.ListModelsHandler)
+	r.GET("/v1/models/:model", openai.RetrieveMiddleware(), s.ShowModelHandler)
 
 	for _, method := range []string{http.MethodGet, http.MethodHead} {
 		r.Handle(method, "/", func(c *gin.Context) {
 			c.String(http.StatusOK, "Ollama is running")
 		})
 
-		r.Handle(method, "/api/tags", s.ListHandler)
+		r.Handle(method, "/api/tags", s.ListModelsHandler)
 		r.Handle(method, "/api/version", func(c *gin.Context) {
 			c.JSON(http.StatusOK, gin.H{"version": version.Version})
 		})
@@ -1269,7 +1269,7 @@ func streamResponse(c *gin.Context, ch chan any) {
 	})
 }
 
-func (s *Server) PsHandler(c *gin.Context) {
+func (s *Server) ProcessHandler(c *gin.Context) {
 	models := []api.ProcessModelResponse{}
 
 	for _, v := range s.sched.loaded {

server/routes_create_test.go

@@ -93,7 +93,7 @@ func TestCreateFromBin(t *testing.T) {
 	t.Setenv("OLLAMA_MODELS", p)
 
 	var s Server
-	w := createRequest(t, s.CreateHandler, api.CreateRequest{
+	w := createRequest(t, s.CreateModelHandler, api.CreateRequest{
 		Name:      "test",
 		Modelfile: fmt.Sprintf("FROM %s", createBinFile(t, nil, nil)),
 		Stream:    &stream,
@@ -120,7 +120,7 @@ func TestCreateFromModel(t *testing.T) {
 	t.Setenv("OLLAMA_MODELS", p)
 	var s Server
 
-	w := createRequest(t, s.CreateHandler, api.CreateRequest{
+	w := createRequest(t, s.CreateModelHandler, api.CreateRequest{
 		Name:      "test",
 		Modelfile: fmt.Sprintf("FROM %s", createBinFile(t, nil, nil)),
 		Stream:    &stream,
@@ -134,7 +134,7 @@ func TestCreateFromModel(t *testing.T) {
 		filepath.Join(p, "manifests", "registry.ollama.ai", "library", "test", "latest"),
 	})
 
-	w = createRequest(t, s.CreateHandler, api.CreateRequest{
+	w = createRequest(t, s.CreateModelHandler, api.CreateRequest{
 		Name:      "test2",
 		Modelfile: "FROM test",
 		Stream:    &stream,
@@ -162,7 +162,7 @@ func TestCreateRemovesLayers(t *testing.T) {
 	t.Setenv("OLLAMA_MODELS", p)
 	var s Server
 
-	w := createRequest(t, s.CreateHandler, api.CreateRequest{
+	w := createRequest(t, s.CreateModelHandler, api.CreateRequest{
 		Name:      "test",
 		Modelfile: fmt.Sprintf("FROM %s\nTEMPLATE {{ .Prompt }}", createBinFile(t, nil, nil)),
 		Stream:    &stream,
@@ -182,7 +182,7 @@ func TestCreateRemovesLayers(t *testing.T) {
 		filepath.Join(p, "blobs", "sha256-bc80b03733773e0728011b2f4adf34c458b400e1aad48cb28d61170f3a2ad2d6"),
 	})
 
-	w = createRequest(t, s.CreateHandler, api.CreateRequest{
+	w = createRequest(t, s.CreateModelHandler, api.CreateRequest{
 		Name:      "test",
 		Modelfile: fmt.Sprintf("FROM %s\nTEMPLATE {{ .System }} {{ .Prompt }}", createBinFile(t, nil, nil)),
 		Stream:    &stream,
@@ -210,7 +210,7 @@ func TestCreateUnsetsSystem(t *testing.T) {
 	t.Setenv("OLLAMA_MODELS", p)
 	var s Server
 
-	w := createRequest(t, s.CreateHandler, api.CreateRequest{
+	w := createRequest(t, s.CreateModelHandler, api.CreateRequest{
 		Name:      "test",
 		Modelfile: fmt.Sprintf("FROM %s\nSYSTEM Say hi!", createBinFile(t, nil, nil)),
 		Stream:    &stream,
@@ -230,7 +230,7 @@ func TestCreateUnsetsSystem(t *testing.T) {
 		filepath.Join(p, "blobs", "sha256-f29e82a8284dbdf5910b1555580ff60b04238b8da9d5e51159ada67a4d0d5851"),
 	})
 
-	w = createRequest(t, s.CreateHandler, api.CreateRequest{
+	w = createRequest(t, s.CreateModelHandler, api.CreateRequest{
 		Name:      "test",
 		Modelfile: fmt.Sprintf("FROM %s\nSYSTEM \"\"", createBinFile(t, nil, nil)),
 		Stream:    &stream,
@@ -267,7 +267,7 @@ func TestCreateMergeParameters(t *testing.T) {
 	t.Setenv("OLLAMA_MODELS", p)
 	var s Server
 
-	w := createRequest(t, s.CreateHandler, api.CreateRequest{
+	w := createRequest(t, s.CreateModelHandler, api.CreateRequest{
 		Name:      "test",
 		Modelfile: fmt.Sprintf("FROM %s\nPARAMETER temperature 1\nPARAMETER top_k 10\nPARAMETER stop USER:\nPARAMETER stop ASSISTANT:", createBinFile(t, nil, nil)),
 		Stream:    &stream,
@@ -288,7 +288,7 @@ func TestCreateMergeParameters(t *testing.T) {
 	})
 
 	// in order to merge parameters, the second model must be created FROM the first
-	w = createRequest(t, s.CreateHandler, api.CreateRequest{
+	w = createRequest(t, s.CreateModelHandler, api.CreateRequest{
 		Name:      "test2",
 		Modelfile: "FROM test\nPARAMETER temperature 0.6\nPARAMETER top_p 0.7",
 		Stream:    &stream,
@@ -326,7 +326,7 @@ func TestCreateMergeParameters(t *testing.T) {
 	}
 
 	// slices are replaced
-	w = createRequest(t, s.CreateHandler, api.CreateRequest{
+	w = createRequest(t, s.CreateModelHandler, api.CreateRequest{
 		Name:      "test2",
 		Modelfile: "FROM test\nPARAMETER temperature 0.6\nPARAMETER top_p 0.7\nPARAMETER stop <|endoftext|>",
 		Stream:    &stream,
@@ -371,7 +371,7 @@ func TestCreateReplacesMessages(t *testing.T) {
 	t.Setenv("OLLAMA_MODELS", p)
 	var s Server
 
-	w := createRequest(t, s.CreateHandler, api.CreateRequest{
+	w := createRequest(t, s.CreateModelHandler, api.CreateRequest{
 		Name:      "test",
 		Modelfile: fmt.Sprintf("FROM %s\nMESSAGE assistant \"What is my purpose?\"\nMESSAGE user \"You run tests.\"\nMESSAGE assistant \"Oh, my god.\"", createBinFile(t, nil, nil)),
 		Stream:    &stream,
@@ -391,7 +391,7 @@ func TestCreateReplacesMessages(t *testing.T) {
 		filepath.Join(p, "blobs", "sha256-e0e27d47045063ccb167ae852c51d49a98eab33fabaee4633fdddf97213e40b5"),
 	})
 
-	w = createRequest(t, s.CreateHandler, api.CreateRequest{
+	w = createRequest(t, s.CreateModelHandler, api.CreateRequest{
 		Name:      "test2",
 		Modelfile: "FROM test\nMESSAGE assistant \"You're a test, Harry.\"\nMESSAGE user \"I-I'm a what?\"\nMESSAGE assistant \"A test. And a thumping good one at that, I'd wager.\"",
 		Stream:    &stream,
@@ -448,7 +448,7 @@ func TestCreateTemplateSystem(t *testing.T) {
 	t.Setenv("OLLAMA_MODELS", p)
 	var s Server
 
-	w := createRequest(t, s.CreateHandler, api.CreateRequest{
+	w := createRequest(t, s.CreateModelHandler, api.CreateRequest{
 		Name:      "test",
 		Modelfile: fmt.Sprintf("FROM %s\nTEMPLATE {{ .Prompt }}\nSYSTEM Say hello!\nTEMPLATE {{ .System }} {{ .Prompt }}\nSYSTEM Say bye!", createBinFile(t, nil, nil)),
 		Stream:    &stream,
@@ -488,7 +488,7 @@ func TestCreateTemplateSystem(t *testing.T) {
 	}
 
 	t.Run("incomplete template", func(t *testing.T) {
-		w := createRequest(t, s.CreateHandler, api.CreateRequest{
+		w := createRequest(t, s.CreateModelHandler, api.CreateRequest{
 			Name:      "test",
 			Modelfile: fmt.Sprintf("FROM %s\nTEMPLATE {{ .Prompt", createBinFile(t, nil, nil)),
 			Stream:    &stream,
@@ -500,7 +500,7 @@ func TestCreateTemplateSystem(t *testing.T) {
 	})
 
 	t.Run("template with unclosed if", func(t *testing.T) {
-		w := createRequest(t, s.CreateHandler, api.CreateRequest{
+		w := createRequest(t, s.CreateModelHandler, api.CreateRequest{
 			Name:      "test",
 			Modelfile: fmt.Sprintf("FROM %s\nTEMPLATE {{ if .Prompt }}", createBinFile(t, nil, nil)),
 			Stream:    &stream,
@@ -512,7 +512,7 @@ func TestCreateTemplateSystem(t *testing.T) {
 	})
 
 	t.Run("template with undefined function", func(t *testing.T) {
-		w := createRequest(t, s.CreateHandler, api.CreateRequest{
+		w := createRequest(t, s.CreateModelHandler, api.CreateRequest{
 			Name:      "test",
 			Modelfile: fmt.Sprintf("FROM %s\nTEMPLATE {{  Prompt }}", createBinFile(t, nil, nil)),
 			Stream:    &stream,
@@ -531,7 +531,7 @@ func TestCreateLicenses(t *testing.T) {
 	t.Setenv("OLLAMA_MODELS", p)
 	var s Server
 
-	w := createRequest(t, s.CreateHandler, api.CreateRequest{
+	w := createRequest(t, s.CreateModelHandler, api.CreateRequest{
 		Name:      "test",
 		Modelfile: fmt.Sprintf("FROM %s\nLICENSE MIT\nLICENSE Apache-2.0", createBinFile(t, nil, nil)),
 		Stream:    &stream,
@@ -579,7 +579,7 @@ func TestCreateDetectTemplate(t *testing.T) {
 	var s Server
 
 	t.Run("matched", func(t *testing.T) {
-		w := createRequest(t, s.CreateHandler, api.CreateRequest{
+		w := createRequest(t, s.CreateModelHandler, api.CreateRequest{
 			Name: "test",
 			Modelfile: fmt.Sprintf("FROM %s", createBinFile(t, llm.KV{
 				"tokenizer.chat_template": "{{ bos_token }}{% for message in messages %}{{'<|' + message['role'] + '|>' + '\n' + message['content'] + '<|end|>\n' }}{% endfor %}{% if add_generation_prompt %}{{ '<|assistant|>\n' }}{% else %}{{ eos_token }}{% endif %}",
@@ -600,7 +600,7 @@ func TestCreateDetectTemplate(t *testing.T) {
 	})
 
 	t.Run("unmatched", func(t *testing.T) {
-		w := createRequest(t, s.CreateHandler, api.CreateRequest{
+		w := createRequest(t, s.CreateModelHandler, api.CreateRequest{
 			Name:      "test",
 			Modelfile: fmt.Sprintf("FROM %s", createBinFile(t, nil, nil)),
 			Stream:    &stream,

server/routes_delete_test.go

@@ -22,7 +22,7 @@ func TestDelete(t *testing.T) {
 
 	var s Server
 
-	w := createRequest(t, s.CreateHandler, api.CreateRequest{
+	w := createRequest(t, s.CreateModelHandler, api.CreateRequest{
 		Name:      "test",
 		Modelfile: fmt.Sprintf("FROM %s", createBinFile(t, nil, nil)),
 	})
@@ -31,7 +31,7 @@ func TestDelete(t *testing.T) {
 		t.Fatalf("expected status code 200, actual %d", w.Code)
 	}
 
-	w = createRequest(t, s.CreateHandler, api.CreateRequest{
+	w = createRequest(t, s.CreateModelHandler, api.CreateRequest{
 		Name:      "test2",
 		Modelfile: fmt.Sprintf("FROM %s\nTEMPLATE {{ .System }} {{ .Prompt }}", createBinFile(t, nil, nil)),
 	})
@@ -52,7 +52,7 @@ func TestDelete(t *testing.T) {
 		filepath.Join(p, "blobs", "sha256-fe7ac77b725cda2ccad03f88a880ecdfd7a33192d6cae08fce2c0ee1455991ed"),
 	})
 
-	w = createRequest(t, s.DeleteHandler, api.DeleteRequest{Name: "test"})
+	w = createRequest(t, s.DeleteModelHandler, api.DeleteRequest{Name: "test"})
 
 	if w.Code != http.StatusOK {
 		t.Fatalf("expected status code 200, actual %d", w.Code)
@@ -68,7 +68,7 @@ func TestDelete(t *testing.T) {
 		filepath.Join(p, "blobs", "sha256-fe7ac77b725cda2ccad03f88a880ecdfd7a33192d6cae08fce2c0ee1455991ed"),
 	})
 
-	w = createRequest(t, s.DeleteHandler, api.DeleteRequest{Name: "test2"})
+	w = createRequest(t, s.DeleteModelHandler, api.DeleteRequest{Name: "test2"})
 
 	if w.Code != http.StatusOK {
 		t.Fatalf("expected status code 200, actual %d", w.Code)
@@ -102,7 +102,7 @@ func TestDeleteDuplicateLayers(t *testing.T) {
 		t.Fatal(err)
 	}
 
-	w := createRequest(t, s.DeleteHandler, api.DeleteRequest{Name: "test"})
+	w := createRequest(t, s.DeleteModelHandler, api.DeleteRequest{Name: "test"})
 	if w.Code != http.StatusOK {
 		t.Errorf("expected status code 200, actual %d", w.Code)
 	}

server/routes_generate_test.go

@@ -84,7 +84,7 @@ func TestGenerateChat(t *testing.T) {
 
 	go s.sched.Run(context.TODO())
 
-	w := createRequest(t, s.CreateHandler, api.CreateRequest{
+	w := createRequest(t, s.CreateModelHandler, api.CreateRequest{
 		Model: "test",
 		Modelfile: fmt.Sprintf(`FROM %s
 		TEMPLATE """
@@ -144,7 +144,7 @@ func TestGenerateChat(t *testing.T) {
 	})
 
 	t.Run("missing capabilities chat", func(t *testing.T) {
-		w := createRequest(t, s.CreateHandler, api.CreateRequest{
+		w := createRequest(t, s.CreateModelHandler, api.CreateRequest{
 			Model: "bert",
 			Modelfile: fmt.Sprintf("FROM %s", createBinFile(t, llm.KV{
 				"general.architecture": "bert",
@@ -270,7 +270,7 @@ func TestGenerateChat(t *testing.T) {
 		checkChatResponse(t, w.Body, "test", "Hi!")
 	})
 
-	w = createRequest(t, s.CreateHandler, api.CreateRequest{
+	w = createRequest(t, s.CreateModelHandler, api.CreateRequest{
 		Model:     "test-system",
 		Modelfile: "FROM test\nSYSTEM You are a helpful assistant.",
 	})
@@ -382,7 +382,7 @@ func TestGenerate(t *testing.T) {
 
 	go s.sched.Run(context.TODO())
 
-	w := createRequest(t, s.CreateHandler, api.CreateRequest{
+	w := createRequest(t, s.CreateModelHandler, api.CreateRequest{
 		Model: "test",
 		Modelfile: fmt.Sprintf(`FROM %s
 		TEMPLATE """
@@ -442,7 +442,7 @@ func TestGenerate(t *testing.T) {
 	})
 
 	t.Run("missing capabilities generate", func(t *testing.T) {
-		w := createRequest(t, s.CreateHandler, api.CreateRequest{
+		w := createRequest(t, s.CreateModelHandler, api.CreateRequest{
 			Model: "bert",
 			Modelfile: fmt.Sprintf("FROM %s", createBinFile(t, llm.KV{
 				"general.architecture": "bert",
@@ -583,7 +583,7 @@ func TestGenerate(t *testing.T) {
 		checkGenerateResponse(t, w.Body, "test", "Hi!")
 	})
 
-	w = createRequest(t, s.CreateHandler, api.CreateRequest{
+	w = createRequest(t, s.CreateModelHandler, api.CreateRequest{
 		Model:     "test-system",
 		Modelfile: "FROM test\nSYSTEM You are a helpful assistant.",
 	})
@@ -652,7 +652,7 @@ func TestGenerate(t *testing.T) {
 		checkGenerateResponse(t, w.Body, "test-system", "Abra kadabra!")
 	})
 
-	w = createRequest(t, s.CreateHandler, api.CreateRequest{
+	w = createRequest(t, s.CreateModelHandler, api.CreateRequest{
 		Model: "test-suffix",
 		Modelfile: `FROM test
 TEMPLATE """{{- if .Suffix }}<PRE> {{ .Prompt }} <SUF>{{ .Suffix }} <MID>

server/routes_list_test.go

@@ -31,13 +31,13 @@ func TestList(t *testing.T) {
 
 	var s Server
 	for _, n := range expectNames {
-		createRequest(t, s.CreateHandler, api.CreateRequest{
+		createRequest(t, s.CreateModelHandler, api.CreateRequest{
 			Name:      n,
 			Modelfile: fmt.Sprintf("FROM %s", createBinFile(t, nil, nil)),
 		})
 	}
 
-	w := createRequest(t, s.ListHandler, nil)
+	w := createRequest(t, s.ListModelsHandler, nil)
 	if w.Code != http.StatusOK {
 		t.Fatalf("expected status code 200, actual %d", w.Code)
 	}

server/routes_test.go

@@ -318,7 +318,7 @@ func TestCase(t *testing.T) {
 	var s Server
 	for _, tt := range cases {
 		t.Run(tt, func(t *testing.T) {
-			w := createRequest(t, s.CreateHandler, api.CreateRequest{
+			w := createRequest(t, s.CreateModelHandler, api.CreateRequest{
 				Name:      tt,
 				Modelfile: fmt.Sprintf("FROM %s", createBinFile(t, nil, nil)),
 				Stream:    &stream,
@@ -334,7 +334,7 @@ func TestCase(t *testing.T) {
 			}
 
 			t.Run("create", func(t *testing.T) {
-				w = createRequest(t, s.CreateHandler, api.CreateRequest{
+				w = createRequest(t, s.CreateModelHandler, api.CreateRequest{
 					Name:      strings.ToUpper(tt),
 					Modelfile: fmt.Sprintf("FROM %s", createBinFile(t, nil, nil)),
 					Stream:    &stream,
@@ -350,7 +350,7 @@ func TestCase(t *testing.T) {
 			})
 
 			t.Run("pull", func(t *testing.T) {
-				w := createRequest(t, s.PullHandler, api.PullRequest{
+				w := createRequest(t, s.PullModelHandler, api.PullRequest{
 					Name:   strings.ToUpper(tt),
 					Stream: &stream,
 				})
@@ -365,7 +365,7 @@ func TestCase(t *testing.T) {
 			})
 
 			t.Run("copy", func(t *testing.T) {
-				w := createRequest(t, s.CopyHandler, api.CopyRequest{
+				w := createRequest(t, s.CopyModelHandler, api.CopyRequest{
 					Source:      tt,
 					Destination: strings.ToUpper(tt),
 				})
@@ -387,7 +387,7 @@ func TestShow(t *testing.T) {
 
 	var s Server
 
-	createRequest(t, s.CreateHandler, api.CreateRequest{
+	createRequest(t, s.CreateModelHandler, api.CreateRequest{
 		Name: "show-model",
 		Modelfile: fmt.Sprintf(
 			"FROM %s\nFROM %s",
@@ -396,7 +396,7 @@ func TestShow(t *testing.T) {
 		),
 	})
 
-	w := createRequest(t, s.ShowHandler, api.ShowRequest{
+	w := createRequest(t, s.ShowModelHandler, api.ShowRequest{
 		Name: "show-model",
 	})