reduce error footprint

This reverts commit 55d48c6ed1.

.github/workflows/release.yaml

@@ -147,7 +147,7 @@ jobs:
         run: |
           $ErrorActionPreference = "Stop"
           write-host "downloading AMD HIP Installer"
-          Invoke-WebRequest -Uri "https://download.amd.com/developer/eula/rocm-hub/AMD-Software-PRO-Edition-24.Q3-WinSvr2022-For-HIP.exe" -OutFile "${env:RUNNER_TEMP}\rocm-install.exe"
+          Invoke-WebRequest -Uri "https://download.amd.com/developer/eula/rocm-hub/AMD-Software-PRO-Edition-23.Q4-WinSvr2022-For-HIP.exe" -OutFile "${env:RUNNER_TEMP}\rocm-install.exe"
           write-host "Installing AMD HIP"
           Start-Process "${env:RUNNER_TEMP}\rocm-install.exe" -ArgumentList '-install' -NoNewWindow -Wait
           write-host "Completed AMD HIP"

.github/workflows/test.yaml

@@ -126,7 +126,7 @@ jobs:
     strategy:
       matrix:
         rocm-version:
-          - '6.1.2'
+          - '6.1.1'
     runs-on: linux
     container: rocm/dev-ubuntu-20.04:${{ matrix.rocm-version }}
     steps:
@@ -169,7 +169,7 @@ jobs:
         run: |
           $ErrorActionPreference = "Stop"
           write-host "downloading AMD HIP Installer"
-          Invoke-WebRequest -Uri "https://download.amd.com/developer/eula/rocm-hub/AMD-Software-PRO-Edition-24.Q3-WinSvr2022-For-HIP.exe" -OutFile "${env:RUNNER_TEMP}\rocm-install.exe"
+          Invoke-WebRequest -Uri "https://download.amd.com/developer/eula/rocm-hub/AMD-Software-PRO-Edition-23.Q4-WinSvr2022-For-HIP.exe" -OutFile "${env:RUNNER_TEMP}\rocm-install.exe"
           write-host "Installing AMD HIP"
           Start-Process "${env:RUNNER_TEMP}\rocm-install.exe" -ArgumentList '-install' -NoNewWindow -Wait
           write-host "Completed AMD HIP"

Dockerfile

@@ -2,7 +2,7 @@ ARG GOLANG_VERSION=1.22.1
 ARG CMAKE_VERSION=3.22.1
 # this CUDA_VERSION corresponds with the one specified in docs/gpu.md
 ARG CUDA_VERSION=11.3.1
-ARG ROCM_VERSION=6.1.2
+ARG ROCM_VERSION=6.1.1
 
 # Copy the minimal context we need to run the generate scripts
 FROM scratch AS llm-code

README.md

@@ -293,8 +293,6 @@ See the [API documentation](./docs/api.md) for all endpoints.
 - [OllamaSpring](https://github.com/CrazyNeil/OllamaSpring) (Ollama Client for macOS)
 - [LLocal.in](https://github.com/kartikm7/llocal) (Easy to use Electron Desktop Client for Ollama)
 - [Ollama with Google Mesop](https://github.com/rapidarchitect/ollama_mesop/) (Mesop Chat Client implementation with Ollama)
-- [Kerlig AI](https://www.kerlig.com/) (AI writing assistant for macOS)
-- [AI Studio](https://github.com/MindWorkAI/AI-Studio)
 
 ### Terminal
 

api/client.go

@@ -356,7 +356,7 @@ func (c *Client) Embed(ctx context.Context, req *EmbedRequest) (*EmbedResponse,
 	return &resp, nil
 }
 
-// Embeddings generates an embedding from a model.
+// Embeddings generates embeddings from a model. (Legacy)
 func (c *Client) Embeddings(ctx context.Context, req *EmbeddingRequest) (*EmbeddingResponse, error) {
 	var resp EmbeddingResponse
 	if err := c.do(ctx, http.MethodPost, "/api/embeddings", req, &resp); err != nil {

api/types.go

@@ -47,9 +47,6 @@ type GenerateRequest struct {
 	// Prompt is the textual prompt to send to the model.
 	Prompt string `json:"prompt"`
 
-	// Suffix is the text that comes after the inserted text.
-	Suffix string `json:"suffix"`
-
 	// System overrides the model's default system message/prompt.
 	System string `json:"system"`
 
@@ -100,9 +97,6 @@ type ChatRequest struct {
 	// followin the request.
 	KeepAlive *Duration `json:"keep_alive,omitempty"`
 
-	// Tools is an optional list of tools the model has access to.
-	Tools []Tool `json:"tools,omitempty"`
-
 	// Options lists model-specific options.
 	Options map[string]interface{} `json:"options"`
 }
@@ -111,46 +105,9 @@ type ChatRequest struct {
 // role ("system", "user", or "assistant"), the content and an optional list
 // of images.
 type Message struct {
-	Role      string      `json:"role"`
-	Content   string      `json:"content,omitempty"`
-	Images    []ImageData `json:"images,omitempty"`
-	ToolCalls []ToolCall  `json:"tool_calls,omitempty"`
-}
-
-type ToolCall struct {
-	Function struct {
-		Name      string         `json:"name"`
-		Arguments map[string]any `json:"arguments"`
-	} `json:"function"`
-}
-
-type Tool struct {
-	Type     string `json:"type"`
-	Function struct {
-		Name        string `json:"name"`
-		Description string `json:"description"`
-		Parameters  struct {
-			Type       string   `json:"type"`
-			Required   []string `json:"required"`
-			Properties map[string]struct {
-				Type        string   `json:"type"`
-				Description string   `json:"description"`
-				Enum        []string `json:"enum,omitempty"`
-			} `json:"properties"`
-		} `json:"parameters"`
-	} `json:"function"`
-}
-
-func (m *Message) UnmarshalJSON(b []byte) error {
-	type Alias Message
-	var a Alias
-	if err := json.Unmarshal(b, &a); err != nil {
-		return err
-	}
-
-	*m = Message(a)
-	m.Role = strings.ToLower(m.Role)
-	return nil
+	Role    string      `json:"role"`
+	Content string      `json:"content"`
+	Images  []ImageData `json:"images,omitempty"`
 }
 
 // ChatResponse is the response returned by [Client.Chat]. Its fields are
@@ -237,7 +194,7 @@ type EmbedRequest struct {
 // EmbedResponse is the response from [Client.Embed].
 type EmbedResponse struct {
 	Model      string      `json:"model"`
-	Embeddings [][]float32 `json:"embeddings"`
+	Embeddings [][]float32 `json:"embeddings,omitempty"`
 }
 
 // EmbeddingRequest is the request passed to [Client.Embeddings].
@@ -286,10 +243,8 @@ type DeleteRequest struct {
 
 // ShowRequest is the request passed to [Client.Show].
 type ShowRequest struct {
-	Model  string `json:"model"`
-	System string `json:"system"`
-
-	// Template is deprecated
+	Model    string `json:"model"`
+	System   string `json:"system"`
 	Template string `json:"template"`
 	Verbose  bool   `json:"verbose"`
 
@@ -405,9 +360,6 @@ type GenerateResponse struct {
 	// Response is the textual response itself.
 	Response string `json:"response"`
 
-	// ToolCalls is the list of tools the model wants to call
-	ToolCalls []ToolCall `json:"tool_calls,omitempty"`
-
 	// Done specifies if the response is complete.
 	Done bool `json:"done"`
 

api/types_test.go

@@ -208,26 +208,3 @@ func TestUseMmapFormatParams(t *testing.T) {
 		})
 	}
 }
-
-func TestMessage_UnmarshalJSON(t *testing.T) {
-	tests := []struct {
-		input    string
-		expected string
-	}{
-		{`{"role": "USER", "content": "Hello!"}`, "user"},
-		{`{"role": "System", "content": "Initialization complete."}`, "system"},
-		{`{"role": "assistant", "content": "How can I help you?"}`, "assistant"},
-		{`{"role": "TOOl", "content": "Access granted."}`, "tool"},
-	}
-
-	for _, test := range tests {
-		var msg Message
-		if err := json.Unmarshal([]byte(test.input), &msg); err != nil {
-			t.Errorf("Unexpected error: %v", err)
-		}
-
-		if msg.Role != test.expected {
-			t.Errorf("role not lowercased: got %v, expected %v", msg.Role, test.expected)
-		}
-	}
-}

app/ollama.iss

@@ -127,10 +127,6 @@ Type: filesandordirs; Name: "{%USERPROFILE}\.ollama\models"
 Type: filesandordirs; Name: "{%USERPROFILE}\.ollama\history"
 ; NOTE: if the user has a custom OLLAMA_MODELS it will be preserved
 
-[InstallDelete]
-Type: filesandordirs; Name: "{%TEMP}\ollama*"
-Type: filesandordirs; Name: "{%LOCALAPPDATA}\Programs\Ollama"
-
 [Messages]
 WizardReady=Ollama Windows Preview
 ReadyLabel1=%nLet's get you up and running with your own large language models.

cmd/cmd.go

@@ -843,6 +843,7 @@ type runOptions struct {
 	WordWrap    bool
 	Format      string
 	System      string
+	Template    string
 	Images      []api.ImageData
 	Options     map[string]interface{}
 	MultiModal  bool
@@ -1036,6 +1037,7 @@ func generate(cmd *cobra.Command, opts runOptions) error {
 		Images:    opts.Images,
 		Format:    opts.Format,
 		System:    opts.System,
+		Template:  opts.Template,
 		Options:   opts.Options,
 		KeepAlive: opts.KeepAlive,
 	}

cmd/interactive.go

@@ -27,6 +27,7 @@ const (
 	MultilineNone MultilineState = iota
 	MultilinePrompt
 	MultilineSystem
+	MultilineTemplate
 )
 
 func loadModel(cmd *cobra.Command, opts *runOptions) error {
@@ -93,6 +94,7 @@ func generateInteractive(cmd *cobra.Command, opts runOptions) error {
 		fmt.Fprintln(os.Stderr, "Available Commands:")
 		fmt.Fprintln(os.Stderr, "  /set parameter ...     Set a parameter")
 		fmt.Fprintln(os.Stderr, "  /set system <string>   Set system message")
+		fmt.Fprintln(os.Stderr, "  /set template <string> Set prompt template")
 		fmt.Fprintln(os.Stderr, "  /set history           Enable history")
 		fmt.Fprintln(os.Stderr, "  /set nohistory         Disable history")
 		fmt.Fprintln(os.Stderr, "  /set wordwrap          Enable wordwrap")
@@ -202,6 +204,10 @@ func generateInteractive(cmd *cobra.Command, opts runOptions) error {
 				opts.Messages = append(opts.Messages, api.Message{Role: "system", Content: opts.System})
 				fmt.Println("Set system message.")
 				sb.Reset()
+			case MultilineTemplate:
+				opts.Template = sb.String()
+				fmt.Println("Set prompt template.")
+				sb.Reset()
 			}
 
 			multiline = MultilineNone
@@ -320,13 +326,17 @@ func generateInteractive(cmd *cobra.Command, opts runOptions) error {
 					}
 					fmt.Printf("Set parameter '%s' to '%s'\n", args[2], strings.Join(params, ", "))
 					opts.Options[args[2]] = fp[args[2]]
-				case "system":
+				case "system", "template":
 					if len(args) < 3 {
 						usageSet()
 						continue
 					}
 
-					multiline = MultilineSystem
+					if args[1] == "system" {
+						multiline = MultilineSystem
+					} else if args[1] == "template" {
+						multiline = MultilineTemplate
+					}
 
 					line := strings.Join(args[2:], " ")
 					line, ok := strings.CutPrefix(line, `"""`)
@@ -346,17 +356,23 @@ func generateInteractive(cmd *cobra.Command, opts runOptions) error {
 						continue
 					}
 
-					opts.System = sb.String() // for display in modelfile
-					newMessage := api.Message{Role: "system", Content: sb.String()}
-					// Check if the slice is not empty and the last message is from 'system'
-					if len(opts.Messages) > 0 && opts.Messages[len(opts.Messages)-1].Role == "system" {
-						// Replace the last message
-						opts.Messages[len(opts.Messages)-1] = newMessage
-					} else {
-						opts.Messages = append(opts.Messages, newMessage)
+					if args[1] == "system" {
+						opts.System = sb.String() // for display in modelfile
+						newMessage := api.Message{Role: "system", Content: sb.String()}
+						// Check if the slice is not empty and the last message is from 'system'
+						if len(opts.Messages) > 0 && opts.Messages[len(opts.Messages)-1].Role == "system" {
+							// Replace the last message
+							opts.Messages[len(opts.Messages)-1] = newMessage
+						} else {
+							opts.Messages = append(opts.Messages, newMessage)
+						}
+						fmt.Println("Set system message.")
+						sb.Reset()
+					} else if args[1] == "template" {
+						opts.Template = sb.String()
+						fmt.Println("Set prompt template.")
+						sb.Reset()
 					}
-					fmt.Println("Set system message.")
-					sb.Reset()
 
 					sb.Reset()
 					continue
@@ -377,6 +393,7 @@ func generateInteractive(cmd *cobra.Command, opts runOptions) error {
 				req := &api.ShowRequest{
 					Name:     opts.Model,
 					System:   opts.System,
+					Template: opts.Template,
 					Options:  opts.Options,
 				}
 				resp, err := client.Show(cmd.Context(), req)
@@ -420,9 +437,12 @@ func generateInteractive(cmd *cobra.Command, opts runOptions) error {
 						fmt.Println("No system message was specified for this model.")
 					}
 				case "template":
-					if resp.Template != "" {
+					switch {
+					case opts.Template != "":
+						fmt.Println(opts.Template + "\n")
+					case resp.Template != "":
 						fmt.Println(resp.Template)
-					} else {
+					default:
 						fmt.Println("No prompt template was specified for this model.")
 					}
 				default:
@@ -516,6 +536,10 @@ func buildModelfile(opts runOptions) string {
 		fmt.Fprintf(&mf, "SYSTEM \"\"\"%s\"\"\"\n", opts.System)
 	}
 
+	if opts.Template != "" {
+		fmt.Fprintf(&mf, "TEMPLATE \"\"\"%s\"\"\"\n", opts.Template)
+	}
+
 	keys := make([]string, 0)
 	for k := range opts.Options {
 		keys = append(keys, k)

cmd/interactive_test.go

@@ -59,6 +59,7 @@ func TestModelfileBuilder(t *testing.T) {
 	opts := runOptions{
 		Model:    "hork",
 		System:   "You are part horse and part shark, but all hork. Do horklike things",
+		Template: "This is a template.",
 		Messages: []api.Message{
 			{Role: "user", Content: "Hey there hork!"},
 			{Role: "assistant", Content: "Yes it is true, I am half horse, half shark."},
@@ -74,6 +75,7 @@ func TestModelfileBuilder(t *testing.T) {
 	mf := buildModelfile(opts)
 	expectedModelfile := `FROM {{.Model}}
 SYSTEM """{{.System}}"""
+TEMPLATE """{{.Template}}"""
 PARAMETER penalize_newline false
 PARAMETER seed 42
 PARAMETER stop [hi there]
@@ -95,6 +97,7 @@ MESSAGE assistant """Yes it is true, I am half horse, half shark."""
 	mf = buildModelfile(opts)
 	expectedModelfile = `FROM {{.ParentModel}}
 SYSTEM """{{.System}}"""
+TEMPLATE """{{.Template}}"""
 PARAMETER penalize_newline false
 PARAMETER seed 42
 PARAMETER stop [hi there]

docs/faq.md

@@ -272,4 +272,4 @@ The following server settings may be used to adjust how Ollama handles concurren
 - `OLLAMA_NUM_PARALLEL` - The maximum number of parallel requests each model will process at the same time.  The default will auto-select either 4 or 1 based on available memory.
 - `OLLAMA_MAX_QUEUE` - The maximum number of requests Ollama will queue when busy before rejecting additional requests. The default is 512
 
-Note: Windows with Radeon GPUs currently default to 1 model maximum due to limitations in ROCm v5.7 for available VRAM reporting.  Once ROCm v6.2 is available, Windows Radeon will follow the defaults above.  You may enable concurrent model loads on Radeon on Windows, but ensure you don't load more models than will fit into your GPUs VRAM.
+Note: Windows with Radeon GPUs currently default to 1 model maximum due to limitations in ROCm v5.7 for available VRAM reporting.  Once ROCm v6 is available, Windows Radeon will follow the defaults above.  You may enable concurrent model loads on Radeon on Windows, but ensure you don't load more models than will fit into your GPUs VRAM.

gpu/amd_common.go

@@ -49,17 +49,9 @@ func rocmGetVisibleDevicesEnv(gpuInfo []GpuInfo) (string, string) {
 }
 
 func commonAMDValidateLibDir() (string, error) {
-	// Favor our bundled version
-
-	// Installer payload location if we're running the installed binary
-	exe, err := os.Executable()
-	if err == nil {
-		rocmTargetDir := filepath.Join(filepath.Dir(exe), "rocm")
-		if rocmLibUsable(rocmTargetDir) {
-			slog.Debug("detected ROCM next to ollama executable " + rocmTargetDir)
-			return rocmTargetDir, nil
-		}
-	}
+	// We try to favor system paths first, so that we can wire up the subprocess to use
+	// the system version.  Only use our bundled version if the system version doesn't work
+	// This gives users a more recovery options if versions have subtle problems at runtime
 
 	// Prefer explicit HIP env var
 	hipPath := os.Getenv("HIP_PATH")
@@ -95,5 +87,14 @@ 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), "rocm")
+		if rocmLibUsable(rocmTargetDir) {
+			slog.Debug("detected ROCM next to ollama executable " + rocmTargetDir)
+			return rocmTargetDir, nil
+		}
+	}
 	return "", fmt.Errorf("no suitable rocm found, falling back to CPU")
 }

gpu/amd_hip_windows.go

@@ -84,8 +84,9 @@ func (hl *HipLib) AMDDriverVersion() (driverMajor, driverMinor int, err error) {
 	}
 
 	slog.Debug("hipDriverGetVersion", "version", version)
-	driverMajor = version / 10000000
-	driverMinor = (version - (driverMajor * 10000000)) / 100000
+	// TODO - this isn't actually right, but the docs claim hipDriverGetVersion isn't accurate anyway...
+	driverMajor = version / 1000
+	driverMinor = (version - (driverMajor * 1000)) / 10
 
 	return driverMajor, driverMinor, nil
 }

gpu/amd_windows.go

@@ -22,8 +22,8 @@ const (
 
 var (
 	// Used to validate if the given ROCm lib is usable
-	ROCmLibGlobs          = []string{"hipblas.dll", "rocblas"}                 // This is not sufficient to discern v5 vs v6
-	RocmStandardLocations = []string{"C:\\Program Files\\AMD\\ROCm\\6.1\\bin"} // TODO glob?
+	ROCmLibGlobs          = []string{"hipblas.dll", "rocblas"}                 // TODO - probably include more coverage of files here...
+	RocmStandardLocations = []string{"C:\\Program Files\\AMD\\ROCm\\5.7\\bin"} // TODO glob?
 )
 
 func AMDGetGPUInfo() []RocmGPUInfo {
@@ -35,11 +35,12 @@ func AMDGetGPUInfo() []RocmGPUInfo {
 	}
 	defer hl.Release()
 
-	driverMajor, driverMinor, err := hl.AMDDriverVersion()
-	if err != nil {
-		// For now this is benign, but we may eventually need to fail compatibility checks
-		slog.Debug("error looking up amd driver version", "error", err)
-	}
+	// TODO - this reports incorrect version information, so omitting for now
+	// driverMajor, driverMinor, err := hl.AMDDriverVersion()
+	// if err != nil {
+	// 	// For now this is benign, but we may eventually need to fail compatibility checks
+	// 	slog.Debug("error looking up amd driver version", "error", err)
+	// }
 
 	// Note: the HIP library automatically handles subsetting to any HIP_VISIBLE_DEVICES the user specified
 	count := hl.HipGetDeviceCount()
@@ -131,8 +132,10 @@ func AMDGetGPUInfo() []RocmGPUInfo {
 				MinimumMemory:  rocmMinimumMemory,
 				Name:           name,
 				Compute:        gfx,
-				DriverMajor:    driverMajor,
-				DriverMinor:    driverMinor,
+
+				// TODO - this information isn't accurate on windows, so don't report it until we find the right way to retrieve
+				// DriverMajor:    driverMajor,
+				// DriverMinor:    driverMinor,
 			},
 			index: i,
 		}

gpu/gpu.go

@@ -274,28 +274,6 @@ func GetGPUInfo() GpuInfoList {
 				gpuInfo.DriverMajor = driverMajor
 				gpuInfo.DriverMinor = driverMinor
 
-				// query the management library as well so we can record any skew between the two
-				// which represents overhead on the GPU we must set aside on subsequent updates
-				if cHandles.nvml != nil {
-					C.nvml_get_free(*cHandles.nvml, C.int(gpuInfo.index), &memInfo.free, &memInfo.total, &memInfo.used)
-					if memInfo.err != nil {
-						slog.Warn("error looking up nvidia GPU memory", "error", C.GoString(memInfo.err))
-						C.free(unsafe.Pointer(memInfo.err))
-					} else {
-						if memInfo.free != 0 && uint64(memInfo.free) > gpuInfo.FreeMemory {
-							gpuInfo.OSOverhead = uint64(memInfo.free) - gpuInfo.FreeMemory
-							slog.Info("detected OS VRAM overhead",
-								"id", gpuInfo.ID,
-								"library", gpuInfo.Library,
-								"compute", gpuInfo.Compute,
-								"driver", fmt.Sprintf("%d.%d", gpuInfo.DriverMajor, gpuInfo.DriverMinor),
-								"name", gpuInfo.Name,
-								"overhead", format.HumanBytes2(gpuInfo.OSOverhead),
-							)
-						}
-					}
-				}
-
 				// TODO potentially sort on our own algorithm instead of what the underlying GPU library does...
 				cudaGPUs = append(cudaGPUs, gpuInfo)
 			}
@@ -360,17 +338,14 @@ func GetGPUInfo() GpuInfoList {
 					"before",
 					"total", format.HumanBytes2(cpus[0].TotalMemory),
 					"free", format.HumanBytes2(cpus[0].FreeMemory),
-					"free_swap", format.HumanBytes2(cpus[0].FreeSwap),
 				),
 				slog.Group(
 					"now",
 					"total", format.HumanBytes2(mem.TotalMemory),
 					"free", format.HumanBytes2(mem.FreeMemory),
-					"free_swap", format.HumanBytes2(mem.FreeSwap),
 				),
 			)
 			cpus[0].FreeMemory = mem.FreeMemory
-			cpus[0].FreeSwap = mem.FreeSwap
 		}
 
 		var memInfo C.mem_info_t
@@ -399,14 +374,9 @@ func GetGPUInfo() GpuInfoList {
 				slog.Warn("error looking up nvidia GPU memory")
 				continue
 			}
-			if cHandles.nvml != nil && gpu.OSOverhead > 0 {
-				// When using the management library update based on recorded overhead
-				memInfo.free -= C.uint64_t(gpu.OSOverhead)
-			}
 			slog.Debug("updating cuda memory data",
 				"gpu", gpu.ID,
 				"name", gpu.Name,
-				"overhead", format.HumanBytes2(gpu.OSOverhead),
 				slog.Group(
 					"before",
 					"total", format.HumanBytes2(gpu.TotalMemory),

gpu/gpu_darwin.go

@@ -57,7 +57,6 @@ func GetCPUMem() (memInfo, error) {
 	return memInfo{
 		TotalMemory: uint64(C.getPhysicalMemory()),
 		FreeMemory:  uint64(C.getFreeMemory()),
-		// FreeSwap omitted as Darwin uses dynamic paging
 	}, nil
 }
 

gpu/gpu_linux.go

@@ -50,7 +50,7 @@ var OneapiMgmtName = "libze_intel_gpu.so"
 
 func GetCPUMem() (memInfo, error) {
 	var mem memInfo
-	var total, available, free, buffers, cached, freeSwap uint64
+	var total, available, free, buffers, cached uint64
 	f, err := os.Open("/proc/meminfo")
 	if err != nil {
 		return mem, err
@@ -70,21 +70,20 @@ func GetCPUMem() (memInfo, error) {
 			_, err = fmt.Sscanf(line, "Buffers:%d", &buffers)
 		case strings.HasPrefix(line, "Cached:"):
 			_, err = fmt.Sscanf(line, "Cached:%d", &cached)
-		case strings.HasPrefix(line, "SwapFree:"):
-			_, err = fmt.Sscanf(line, "SwapFree:%d", &freeSwap)
 		default:
 			continue
 		}
 		if err != nil {
 			return mem, err
 		}
+
+		if total > 0 && available > 0 {
+			mem.TotalMemory = total * format.KibiByte
+			mem.FreeMemory = available * format.KibiByte
+			return mem, nil
+		}
 	}
 	mem.TotalMemory = total * format.KibiByte
-	mem.FreeSwap = freeSwap * format.KibiByte
-	if available > 0 {
-		mem.FreeMemory = available * format.KibiByte
-	} else {
-		mem.FreeMemory = (free + buffers + cached) * format.KibiByte
-	}
+	mem.FreeMemory = (free + buffers + cached) * format.KibiByte
 	return mem, nil
 }

gpu/gpu_windows.go

@@ -51,5 +51,5 @@ func GetCPUMem() (memInfo, error) {
 	if r1 == 0 {
 		return memInfo{}, fmt.Errorf("GlobalMemoryStatusEx failed: %w", err)
 	}
-	return memInfo{TotalMemory: memStatus.TotalPhys, FreeMemory: memStatus.AvailPhys, FreeSwap: memStatus.AvailPageFile}, nil
+	return memInfo{TotalMemory: memStatus.TotalPhys, FreeMemory: memStatus.AvailPhys}, nil
 }

gpu/types.go

@@ -10,7 +10,6 @@ import (
 type memInfo struct {
 	TotalMemory uint64 `json:"total_memory,omitempty"`
 	FreeMemory  uint64 `json:"free_memory,omitempty"`
-	FreeSwap    uint64 `json:"free_swap,omitempty"`
 }
 
 // Beginning of an `ollama info` command
@@ -53,8 +52,7 @@ type CPUInfo struct {
 
 type CudaGPUInfo struct {
 	GpuInfo
-	OSOverhead uint64 // Memory overhead between the driver library and management library
-	index      int    //nolint:unused,nolintlint
+	index int //nolint:unused,nolintlint
 }
 type CudaGPUInfoList []CudaGPUInfo
 

llm/generate/gen_linux.sh

@@ -178,7 +178,7 @@ if [ -z "${OLLAMA_SKIP_CUDA_GENERATE}" -a -d "${CUDA_LIB_DIR}" ]; then
         CMAKE_CUDA_DEFS="-DGGML_CUDA=on -DCMAKE_CUDA_ARCHITECTURES=${CMAKE_CUDA_ARCHITECTURES} ${OLLAMA_CUSTOM_CUDA_DEFS}"
         echo "Building custom CUDA GPU"
     else
-        CMAKE_CUDA_DEFS="-DGGML_CUDA=on -DCMAKE_CUDA_FLAGS=-t8 -DCMAKE_CUDA_ARCHITECTURES=${CMAKE_CUDA_ARCHITECTURES}"
+        CMAKE_CUDA_DEFS="-DGGML_CUDA=on -DCMAKE_CUDA_FLAGS=-t8 -DGGML_CUDA_FORCE_MMQ=on -DCMAKE_CUDA_ARCHITECTURES=${CMAKE_CUDA_ARCHITECTURES} -DCMAKE_LIBRARY_PATH=/usr/local/cuda/compat"
     fi
     CMAKE_DEFS="${COMMON_CMAKE_DEFS} ${CMAKE_DEFS} ${ARM64_DEFS} ${CMAKE_CUDA_DEFS}"
     BUILD_DIR="../build/linux/${ARCH}/cuda${CUDA_VARIANT}"
@@ -254,7 +254,7 @@ if [ -z "${OLLAMA_SKIP_ROCM_GENERATE}" -a -d "${ROCM_PATH}" ]; then
         ROCM_VARIANT=_v$(ls ${ROCM_PATH}/lib/librocblas.so.*.*.????? | cut -f5 -d. || true)
     fi
     init_vars
-    CMAKE_DEFS="${COMMON_CMAKE_DEFS} ${CMAKE_DEFS} -DGGML_HIPBLAS=on -DLLAMA_CUDA_NO_PEER_COPY=on -DCMAKE_C_COMPILER=$ROCM_PATH/llvm/bin/clang -DCMAKE_CXX_COMPILER=$ROCM_PATH/llvm/bin/clang++ -DAMDGPU_TARGETS=$(amdGPUs) -DGPU_TARGETS=$(amdGPUs)"
+    CMAKE_DEFS="${COMMON_CMAKE_DEFS} ${CMAKE_DEFS} -DGGML_HIPBLAS=on -DCMAKE_C_COMPILER=$ROCM_PATH/llvm/bin/clang -DCMAKE_CXX_COMPILER=$ROCM_PATH/llvm/bin/clang++ -DAMDGPU_TARGETS=$(amdGPUs) -DGPU_TARGETS=$(amdGPUs)"
     # Users building from source can tune the exact flags we pass to cmake for configuring llama.cpp
     if [ -n "${OLLAMA_CUSTOM_ROCM_DEFS}" ]; then
         echo "OLLAMA_CUSTOM_ROCM_DEFS=\"${OLLAMA_CUSTOM_ROCM_DEFS}\""

llm/generate/gen_windows.ps1

@@ -6,9 +6,18 @@ function amdGPUs {
     if ($env:AMDGPU_TARGETS) {
         return $env:AMDGPU_TARGETS
     }
-    # Current supported rocblas list from ROCm v6.1.2 on windows
+    # TODO - load from some common data file for linux + windows build consistency
     $GPU_LIST = @(
+        "gfx900"
         "gfx906:xnack-"
+        "gfx908:xnack-"
+        "gfx90a:xnack+"
+        "gfx90a:xnack-"
+        "gfx940"
+        "gfx941"
+        "gfx942"
+        "gfx1010"
+        "gfx1012"
         "gfx1030"
         "gfx1100"
         "gfx1101"
@@ -357,7 +366,6 @@ function build_rocm() {
             "-DCMAKE_C_COMPILER=clang.exe",
             "-DCMAKE_CXX_COMPILER=clang++.exe",
             "-DGGML_HIPBLAS=on",
-            "-DLLAMA_CUDA_NO_PEER_COPY=on",
             "-DHIP_PLATFORM=amd",
             "-DGGML_AVX=on",
             "-DGGML_AVX2=off",
@@ -386,6 +394,7 @@ function build_rocm() {
         sign
         install
 
+        # Assumes v5.7, may need adjustments for v6
         rm -ea 0 -recurse -force -path "${script:SRC_DIR}\dist\windows-${script:ARCH}\rocm\"
         md "${script:SRC_DIR}\dist\windows-${script:ARCH}\rocm\rocblas\library\" -ea 0 > $null
         cp "${env:HIP_PATH}\bin\hipblas.dll" "${script:SRC_DIR}\dist\windows-${script:ARCH}\rocm\"

llm/ggml.go

@@ -424,32 +424,6 @@ func (llm GGML) GraphSize(context, batch uint64) (partialOffload, fullOffload ui
 			4*batch*(3*embedding+vocab)+embedding*vocab*105/128,
 			4*batch*(2*embedding+1+2*embeddingHeadsK*headsKV+context+context*headsKV)+4*embeddingHeadsK*context*headsKV+embedding*embeddingHeadsK*headsKV*9/16,
 		)
-	case "chatglm":
-		fullOffload = 4 * batch * (embedding + vocab)
-		partialOffload = 4*batch*(embedding+vocab) + embedding*vocab*105/128
-		if qkvBias, ok := layers["blk.0"]["attn_qkv.bias"]; ok {
-			fullOffload = max(
-				fullOffload,
-				4*batch*(2+
-					2*embedding+
-					context+
-					context*heads+
-					embeddingHeadsK*heads+
-					qkvBias.Shape[0]),
-			)
-
-			partialOffload = max(
-				partialOffload,
-				4*batch*(1+
-					2*embedding+
-					embeddingHeadsK*heads+
-					context+
-					context*heads)+
-					4*embeddingHeadsK*context+
-					4*context*embeddingHeadsK+
-					4*qkvBias.Shape[0],
-			)
-		}
 	}
 
 	return

llm/gguf.go

@@ -537,7 +537,6 @@ var ggufKVOrder = map[string][]string{
 		"tokenizer.ggml.add_bos_token",
 		"tokenizer.ggml.add_eos_token",
 		"tokenizer.chat_template",
-		"bert.pooling_type",
 	},
 }
 

llm/llm.go

@@ -33,7 +33,7 @@ func Quantize(infile, outfile string, ftype fileType) error {
 	params.ftype = ftype.Value()
 
 	if rc := C.llama_model_quantize(cinfile, coutfile, &params); rc != 0 {
-		return fmt.Errorf("failed to quantize model. This model architecture may not be supported, or you may need to upgrade Ollama to the latest version")
+		return fmt.Errorf("llama_model_quantize: %d", rc)
 	}
 
 	return nil

llm/server.go

@@ -88,7 +88,6 @@ func NewLlamaServer(gpus gpu.GpuInfoList, model string, ggml *GGML, adapters, pr
 	var estimate MemoryEstimate
 	var systemTotalMemory uint64
 	var systemFreeMemory uint64
-	var systemSwapFreeMemory uint64
 
 	systemMemInfo, err := gpu.GetCPUMem()
 	if err != nil {
@@ -96,8 +95,7 @@ func NewLlamaServer(gpus gpu.GpuInfoList, model string, ggml *GGML, adapters, pr
 	} else {
 		systemTotalMemory = systemMemInfo.TotalMemory
 		systemFreeMemory = systemMemInfo.FreeMemory
-		systemSwapFreeMemory = systemMemInfo.FreeSwap
-		slog.Debug("system memory", "total", format.HumanBytes2(systemTotalMemory), "free", format.HumanBytes2(systemFreeMemory), "free_swap", format.HumanBytes2(systemSwapFreeMemory))
+		slog.Debug("system memory", "total", format.HumanBytes2(systemTotalMemory), "free", systemFreeMemory)
 	}
 
 	// If the user wants zero GPU layers, reset the gpu list to be CPU/system ram info
@@ -124,16 +122,6 @@ func NewLlamaServer(gpus gpu.GpuInfoList, model string, ggml *GGML, adapters, pr
 		}
 	}
 
-	// On linux, over-allocating CPU memory will almost always result in an error
-	if runtime.GOOS == "linux" {
-		systemMemoryRequired := estimate.TotalSize - estimate.VRAMSize
-		available := systemFreeMemory + systemSwapFreeMemory
-		if systemMemoryRequired > available {
-			slog.Warn("model request too large for system", "requested", format.HumanBytes2(systemMemoryRequired), "available", available, "total", format.HumanBytes2(systemTotalMemory), "free", format.HumanBytes2(systemFreeMemory), "swap", format.HumanBytes2(systemSwapFreeMemory))
-			return nil, fmt.Errorf("model requires more system memory (%s) than is available (%s)", format.HumanBytes2(systemMemoryRequired), format.HumanBytes2(available))
-		}
-	}
-
 	estimate.log()
 
 	// Loop through potential servers
@@ -266,6 +254,10 @@ func NewLlamaServer(gpus gpu.GpuInfoList, model string, ggml *GGML, adapters, pr
 		params = append(params, "--tensor-split", estimate.TensorSplit)
 	}
 
+	if estimate.TensorSplit != "" {
+		params = append(params, "--tensor-split", estimate.TensorSplit)
+	}
+
 	for i := range len(servers) {
 		dir := availableServers[servers[i]]
 		if dir == "" {

openai/openai.go

@@ -3,13 +3,11 @@ package openai
 
 import (
 	"bytes"
-	"encoding/base64"
 	"encoding/json"
 	"fmt"
 	"io"
 	"math/rand"
 	"net/http"
-	"strings"
 	"time"
 
 	"github.com/gin-gonic/gin"
@@ -30,7 +28,7 @@ type ErrorResponse struct {
 
 type Message struct {
 	Role    string `json:"role"`
-	Content any    `json:"content"`
+	Content string `json:"content"`
 }
 
 type Choice struct {
@@ -61,11 +59,6 @@ type ResponseFormat struct {
 	Type string `json:"type"`
 }
 
-type EmbedRequest struct {
-	Input any    `json:"input"`
-	Model string `json:"model"`
-}
-
 type ChatCompletionRequest struct {
 	Model            string          `json:"model"`
 	Messages         []Message       `json:"messages"`
@@ -139,23 +132,11 @@ type Model struct {
 	OwnedBy string `json:"owned_by"`
 }
 
-type Embedding struct {
-	Object    string    `json:"object"`
-	Embedding []float32 `json:"embedding"`
-	Index     int       `json:"index"`
-}
-
 type ListCompletion struct {
 	Object string  `json:"object"`
 	Data   []Model `json:"data"`
 }
 
-type EmbeddingList struct {
-	Object string      `json:"object"`
-	Data   []Embedding `json:"data"`
-	Model  string      `json:"model"`
-}
-
 func NewError(code int, message string) ErrorResponse {
 	var etype string
 	switch code {
@@ -279,27 +260,6 @@ func toListCompletion(r api.ListResponse) ListCompletion {
 	}
 }
 
-func toEmbeddingList(model string, r api.EmbedResponse) EmbeddingList {
-	if r.Embeddings != nil {
-		var data []Embedding
-		for i, e := range r.Embeddings {
-			data = append(data, Embedding{
-				Object:    "embedding",
-				Embedding: e,
-				Index:     i,
-			})
-		}
-
-		return EmbeddingList{
-			Object: "list",
-			Data:   data,
-			Model:  model,
-		}
-	}
-
-	return EmbeddingList{}
-}
-
 func toModel(r api.ShowResponse, m string) Model {
 	return Model{
 		Id:      m,
@@ -309,66 +269,10 @@ func toModel(r api.ShowResponse, m string) Model {
 	}
 }
 
-func fromChatRequest(r ChatCompletionRequest) (*api.ChatRequest, error) {
+func fromChatRequest(r ChatCompletionRequest) api.ChatRequest {
 	var messages []api.Message
 	for _, msg := range r.Messages {
-		switch content := msg.Content.(type) {
-		case string:
-			messages = append(messages, api.Message{Role: msg.Role, Content: content})
-		case []any:
-			message := api.Message{Role: msg.Role}
-			for _, c := range content {
-				data, ok := c.(map[string]any)
-				if !ok {
-					return nil, fmt.Errorf("invalid message format")
-				}
-				switch data["type"] {
-				case "text":
-					text, ok := data["text"].(string)
-					if !ok {
-						return nil, fmt.Errorf("invalid message format")
-					}
-					message.Content = text
-				case "image_url":
-					var url string
-					if urlMap, ok := data["image_url"].(map[string]any); ok {
-						if url, ok = urlMap["url"].(string); !ok {
-							return nil, fmt.Errorf("invalid message format")
-						}
-					} else {
-						if url, ok = data["image_url"].(string); !ok {
-							return nil, fmt.Errorf("invalid message format")
-						}
-					}
-
-					types := []string{"jpeg", "jpg", "png"}
-					valid := false
-					for _, t := range types {
-						prefix := "data:image/" + t + ";base64,"
-						if strings.HasPrefix(url, prefix) {
-							url = strings.TrimPrefix(url, prefix)
-							valid = true
-							break
-						}
-					}
-
-					if !valid {
-						return nil, fmt.Errorf("invalid image input")
-					}
-
-					img, err := base64.StdEncoding.DecodeString(url)
-					if err != nil {
-						return nil, fmt.Errorf("invalid message format")
-					}
-					message.Images = append(message.Images, img)
-				default:
-					return nil, fmt.Errorf("invalid message format")
-				}
-			}
-			messages = append(messages, message)
-		default:
-			return nil, fmt.Errorf("invalid message content type: %T", content)
-		}
+		messages = append(messages, api.Message{Role: msg.Role, Content: msg.Content})
 	}
 
 	options := make(map[string]interface{})
@@ -419,13 +323,13 @@ func fromChatRequest(r ChatCompletionRequest) (*api.ChatRequest, error) {
 		format = "json"
 	}
 
-	return &api.ChatRequest{
+	return api.ChatRequest{
 		Model:    r.Model,
 		Messages: messages,
 		Format:   format,
 		Options:  options,
 		Stream:   &r.Stream,
-	}, nil
+	}
 }
 
 func fromCompleteRequest(r CompletionRequest) (api.GenerateRequest, error) {
@@ -434,16 +338,12 @@ func fromCompleteRequest(r CompletionRequest) (api.GenerateRequest, error) {
 	switch stop := r.Stop.(type) {
 	case string:
 		options["stop"] = []string{stop}
-	case []any:
-		var stops []string
-		for _, s := range stop {
-			if str, ok := s.(string); ok {
-				stops = append(stops, str)
-			} else {
-				return api.GenerateRequest{}, fmt.Errorf("invalid type for 'stop' field: %T", s)
-			}
+	case []string:
+		options["stop"] = stop
+	default:
+		if r.Stop != nil {
+			return api.GenerateRequest{}, fmt.Errorf("invalid type for 'stop' field: %T", r.Stop)
 		}
-		options["stop"] = stops
 	}
 
 	if r.MaxTokens != nil {
@@ -503,11 +403,6 @@ type RetrieveWriter struct {
 	model string
 }
 
-type EmbedWriter struct {
-	BaseWriter
-	model string
-}
-
 func (w *BaseWriter) writeError(code int, data []byte) (int, error) {
 	var serr api.StatusError
 	err := json.Unmarshal(data, &serr)
@@ -673,33 +568,6 @@ func (w *RetrieveWriter) Write(data []byte) (int, error) {
 	return w.writeResponse(data)
 }
 
-func (w *EmbedWriter) writeResponse(data []byte) (int, error) {
-	var embedResponse api.EmbedResponse
-	err := json.Unmarshal(data, &embedResponse)
-
-	if err != nil {
-		return 0, err
-	}
-
-	w.ResponseWriter.Header().Set("Content-Type", "application/json")
-	err = json.NewEncoder(w.ResponseWriter).Encode(toEmbeddingList(w.model, embedResponse))
-
-	if err != nil {
-		return 0, err
-	}
-
-	return len(data), nil
-}
-
-func (w *EmbedWriter) Write(data []byte) (int, error) {
-	code := w.ResponseWriter.Status()
-	if code != http.StatusOK {
-		return w.writeError(code, data)
-	}
-
-	return w.writeResponse(data)
-}
-
 func ListMiddleware() gin.HandlerFunc {
 	return func(c *gin.Context) {
 		w := &ListWriter{
@@ -763,47 +631,6 @@ func CompletionsMiddleware() gin.HandlerFunc {
 			id:         fmt.Sprintf("cmpl-%d", rand.Intn(999)),
 		}
 
-		c.Writer = w
-		c.Next()
-	}
-}
-
-func EmbeddingsMiddleware() gin.HandlerFunc {
-	return func(c *gin.Context) {
-		var req EmbedRequest
-		err := c.ShouldBindJSON(&req)
-		if err != nil {
-			c.AbortWithStatusJSON(http.StatusBadRequest, NewError(http.StatusBadRequest, err.Error()))
-			return
-		}
-
-		if req.Input == "" {
-			req.Input = []string{""}
-		}
-
-		if req.Input == nil {
-			c.AbortWithStatusJSON(http.StatusBadRequest, NewError(http.StatusBadRequest, "invalid input"))
-			return
-		}
-
-		if v, ok := req.Input.([]any); ok && len(v) == 0 {
-			c.AbortWithStatusJSON(http.StatusBadRequest, NewError(http.StatusBadRequest, "invalid input"))
-			return
-		}
-
-		var b bytes.Buffer
-		if err := json.NewEncoder(&b).Encode(api.EmbedRequest{Model: req.Model, Input: req.Input}); err != nil {
-			c.AbortWithStatusJSON(http.StatusInternalServerError, NewError(http.StatusInternalServerError, err.Error()))
-			return
-		}
-
-		c.Request.Body = io.NopCloser(&b)
-
-		w := &EmbedWriter{
-			BaseWriter: BaseWriter{ResponseWriter: c.Writer},
-			model:      req.Model,
-		}
-
 		c.Writer = w
 
 		c.Next()
@@ -825,13 +652,7 @@ func ChatMiddleware() gin.HandlerFunc {
 		}
 
 		var b bytes.Buffer
-
-		chatReq, err := fromChatRequest(req)
-		if err != nil {
-			c.AbortWithStatusJSON(http.StatusBadRequest, NewError(http.StatusBadRequest, err.Error()))
-		}
-
-		if err := json.NewEncoder(&b).Encode(chatReq); err != nil {
+		if err := json.NewEncoder(&b).Encode(fromChatRequest(req)); err != nil {
 			c.AbortWithStatusJSON(http.StatusInternalServerError, NewError(http.StatusInternalServerError, err.Error()))
 			return
 		}

openai/openai_test.go

@@ -2,8 +2,8 @@ package openai
 
 import (
 	"bytes"
-	"encoding/base64"
 	"encoding/json"
+	"fmt"
 	"io"
 	"net/http"
 	"net/http/httptest"
@@ -16,253 +16,7 @@ import (
 	"github.com/stretchr/testify/assert"
 )
 
-const prefix = `data:image/jpeg;base64,`
-const image = `iVBORw0KGgoAAAANSUhEUgAAAAEAAAABCAQAAAC1HAwCAAAAC0lEQVR42mNk+A8AAQUBAScY42YAAAAASUVORK5CYII=`
-const imageURL = prefix + image
-
-func TestMiddlewareRequests(t *testing.T) {
-	type testCase struct {
-		Name     string
-		Method   string
-		Path     string
-		Handler  func() gin.HandlerFunc
-		Setup    func(t *testing.T, req *http.Request)
-		Expected func(t *testing.T, req *http.Request)
-	}
-
-	var capturedRequest *http.Request
-
-	captureRequestMiddleware := func() gin.HandlerFunc {
-		return func(c *gin.Context) {
-			bodyBytes, _ := io.ReadAll(c.Request.Body)
-			c.Request.Body = io.NopCloser(bytes.NewReader(bodyBytes))
-			capturedRequest = c.Request
-			c.Next()
-		}
-	}
-
-	testCases := []testCase{
-		{
-			Name:    "chat handler",
-			Method:  http.MethodPost,
-			Path:    "/api/chat",
-			Handler: ChatMiddleware,
-			Setup: func(t *testing.T, req *http.Request) {
-				body := ChatCompletionRequest{
-					Model:    "test-model",
-					Messages: []Message{{Role: "user", Content: "Hello"}},
-				}
-
-				bodyBytes, _ := json.Marshal(body)
-
-				req.Body = io.NopCloser(bytes.NewReader(bodyBytes))
-				req.Header.Set("Content-Type", "application/json")
-			},
-			Expected: func(t *testing.T, req *http.Request) {
-				var chatReq api.ChatRequest
-				if err := json.NewDecoder(req.Body).Decode(&chatReq); err != nil {
-					t.Fatal(err)
-				}
-
-				if chatReq.Messages[0].Role != "user" {
-					t.Fatalf("expected 'user', got %s", chatReq.Messages[0].Role)
-				}
-
-				if chatReq.Messages[0].Content != "Hello" {
-					t.Fatalf("expected 'Hello', got %s", chatReq.Messages[0].Content)
-				}
-			},
-		},
-		{
-			Name:    "completions handler",
-			Method:  http.MethodPost,
-			Path:    "/api/generate",
-			Handler: CompletionsMiddleware,
-			Setup: func(t *testing.T, req *http.Request) {
-				temp := float32(0.8)
-				body := CompletionRequest{
-					Model:       "test-model",
-					Prompt:      "Hello",
-					Temperature: &temp,
-					Stop:        []string{"\n", "stop"},
-				}
-
-				bodyBytes, _ := json.Marshal(body)
-
-				req.Body = io.NopCloser(bytes.NewReader(bodyBytes))
-				req.Header.Set("Content-Type", "application/json")
-			},
-			Expected: func(t *testing.T, req *http.Request) {
-				var genReq api.GenerateRequest
-				if err := json.NewDecoder(req.Body).Decode(&genReq); err != nil {
-					t.Fatal(err)
-				}
-
-				if genReq.Prompt != "Hello" {
-					t.Fatalf("expected 'Hello', got %s", genReq.Prompt)
-				}
-
-				if genReq.Options["temperature"] != 1.6 {
-					t.Fatalf("expected 1.6, got %f", genReq.Options["temperature"])
-				}
-
-				stopTokens, ok := genReq.Options["stop"].([]any)
-
-				if !ok {
-					t.Fatalf("expected stop tokens to be a list")
-				}
-
-				if stopTokens[0] != "\n" || stopTokens[1] != "stop" {
-					t.Fatalf("expected ['\\n', 'stop'], got %v", stopTokens)
-				}
-			},
-		},
-		{
-			Name:    "chat handler with image content",
-			Method:  http.MethodPost,
-			Path:    "/api/chat",
-			Handler: ChatMiddleware,
-			Setup: func(t *testing.T, req *http.Request) {
-				body := ChatCompletionRequest{
-					Model: "test-model",
-					Messages: []Message{
-						{
-							Role: "user", Content: []map[string]any{
-								{"type": "text", "text": "Hello"},
-								{"type": "image_url", "image_url": map[string]string{"url": imageURL}},
-							},
-						},
-					},
-				}
-
-				bodyBytes, _ := json.Marshal(body)
-
-				req.Body = io.NopCloser(bytes.NewReader(bodyBytes))
-				req.Header.Set("Content-Type", "application/json")
-			},
-			Expected: func(t *testing.T, req *http.Request) {
-				var chatReq api.ChatRequest
-				if err := json.NewDecoder(req.Body).Decode(&chatReq); err != nil {
-					t.Fatal(err)
-				}
-
-				if chatReq.Messages[0].Role != "user" {
-					t.Fatalf("expected 'user', got %s", chatReq.Messages[0].Role)
-				}
-
-				if chatReq.Messages[0].Content != "Hello" {
-					t.Fatalf("expected 'Hello', got %s", chatReq.Messages[0].Content)
-				}
-
-				img, _ := base64.StdEncoding.DecodeString(imageURL[len(prefix):])
-
-				if !bytes.Equal(chatReq.Messages[0].Images[0], img) {
-					t.Fatalf("expected image encoding, got %s", chatReq.Messages[0].Images[0])
-				}
-			},
-		},
-		{
-			Name:    "embed handler single input",
-			Method:  http.MethodPost,
-			Path:    "/api/embed",
-			Handler: EmbeddingsMiddleware,
-			Setup: func(t *testing.T, req *http.Request) {
-				body := EmbedRequest{
-					Input: "Hello",
-					Model: "test-model",
-				}
-
-				bodyBytes, _ := json.Marshal(body)
-
-				req.Body = io.NopCloser(bytes.NewReader(bodyBytes))
-				req.Header.Set("Content-Type", "application/json")
-			},
-			Expected: func(t *testing.T, req *http.Request) {
-				var embedReq api.EmbedRequest
-				if err := json.NewDecoder(req.Body).Decode(&embedReq); err != nil {
-					t.Fatal(err)
-				}
-
-				if embedReq.Input != "Hello" {
-					t.Fatalf("expected 'Hello', got %s", embedReq.Input)
-				}
-
-				if embedReq.Model != "test-model" {
-					t.Fatalf("expected 'test-model', got %s", embedReq.Model)
-				}
-			},
-		},
-		{
-			Name:    "embed handler batch input",
-			Method:  http.MethodPost,
-			Path:    "/api/embed",
-			Handler: EmbeddingsMiddleware,
-			Setup: func(t *testing.T, req *http.Request) {
-				body := EmbedRequest{
-					Input: []string{"Hello", "World"},
-					Model: "test-model",
-				}
-
-				bodyBytes, _ := json.Marshal(body)
-
-				req.Body = io.NopCloser(bytes.NewReader(bodyBytes))
-				req.Header.Set("Content-Type", "application/json")
-			},
-			Expected: func(t *testing.T, req *http.Request) {
-				var embedReq api.EmbedRequest
-				if err := json.NewDecoder(req.Body).Decode(&embedReq); err != nil {
-					t.Fatal(err)
-				}
-
-				input, ok := embedReq.Input.([]any)
-
-				if !ok {
-					t.Fatalf("expected input to be a list")
-				}
-
-				if input[0].(string) != "Hello" {
-					t.Fatalf("expected 'Hello', got %s", input[0])
-				}
-
-				if input[1].(string) != "World" {
-					t.Fatalf("expected 'World', got %s", input[1])
-				}
-
-				if embedReq.Model != "test-model" {
-					t.Fatalf("expected 'test-model', got %s", embedReq.Model)
-				}
-			},
-		},
-	}
-
-	gin.SetMode(gin.TestMode)
-	router := gin.New()
-
-	endpoint := func(c *gin.Context) {
-		c.Status(http.StatusOK)
-	}
-
-	for _, tc := range testCases {
-		t.Run(tc.Name, func(t *testing.T) {
-			router = gin.New()
-			router.Use(captureRequestMiddleware())
-			router.Use(tc.Handler())
-			router.Handle(tc.Method, tc.Path, endpoint)
-			req, _ := http.NewRequest(tc.Method, tc.Path, nil)
-
-			if tc.Setup != nil {
-				tc.Setup(t, req)
-			}
-
-			resp := httptest.NewRecorder()
-			router.ServeHTTP(resp, req)
-
-			tc.Expected(t, capturedRequest)
-		})
-	}
-}
-
-func TestMiddlewareResponses(t *testing.T) {
+func TestMiddleware(t *testing.T) {
 	type testCase struct {
 		Name     string
 		Method   string
@@ -276,7 +30,159 @@ func TestMiddlewareResponses(t *testing.T) {
 
 	testCases := []testCase{
 		{
-			Name:     "completions handler error forwarding",
+			Name:     "chat handler",
+			Method:   http.MethodPost,
+			Path:     "/api/chat",
+			TestPath: "/api/chat",
+			Handler:  ChatMiddleware,
+			Endpoint: func(c *gin.Context) {
+				var chatReq api.ChatRequest
+				if err := c.ShouldBindJSON(&chatReq); err != nil {
+					c.JSON(http.StatusBadRequest, gin.H{"error": "invalid request"})
+					return
+				}
+
+				userMessage := chatReq.Messages[0].Content
+				var assistantMessage string
+
+				switch userMessage {
+				case "Hello":
+					assistantMessage = "Hello!"
+				default:
+					assistantMessage = "I'm not sure how to respond to that."
+				}
+
+				c.JSON(http.StatusOK, api.ChatResponse{
+					Message: api.Message{
+						Role:    "assistant",
+						Content: assistantMessage,
+					},
+				})
+			},
+			Setup: func(t *testing.T, req *http.Request) {
+				body := ChatCompletionRequest{
+					Model:    "test-model",
+					Messages: []Message{{Role: "user", Content: "Hello"}},
+				}
+
+				bodyBytes, _ := json.Marshal(body)
+
+				req.Body = io.NopCloser(bytes.NewReader(bodyBytes))
+				req.Header.Set("Content-Type", "application/json")
+			},
+			Expected: func(t *testing.T, resp *httptest.ResponseRecorder) {
+				assert.Equal(t, http.StatusOK, resp.Code)
+
+				var chatResp ChatCompletion
+				if err := json.NewDecoder(resp.Body).Decode(&chatResp); err != nil {
+					t.Fatal(err)
+				}
+
+				if chatResp.Object != "chat.completion" {
+					t.Fatalf("expected chat.completion, got %s", chatResp.Object)
+				}
+
+				if chatResp.Choices[0].Message.Content != "Hello!" {
+					t.Fatalf("expected Hello!, got %s", chatResp.Choices[0].Message.Content)
+				}
+			},
+		},
+		{
+			Name:     "completions handler",
+			Method:   http.MethodPost,
+			Path:     "/api/generate",
+			TestPath: "/api/generate",
+			Handler:  CompletionsMiddleware,
+			Endpoint: func(c *gin.Context) {
+				c.JSON(http.StatusOK, api.GenerateResponse{
+					Response: "Hello!",
+				})
+			},
+			Setup: func(t *testing.T, req *http.Request) {
+				body := CompletionRequest{
+					Model:  "test-model",
+					Prompt: "Hello",
+				}
+
+				bodyBytes, _ := json.Marshal(body)
+
+				req.Body = io.NopCloser(bytes.NewReader(bodyBytes))
+				req.Header.Set("Content-Type", "application/json")
+			},
+			Expected: func(t *testing.T, resp *httptest.ResponseRecorder) {
+				assert.Equal(t, http.StatusOK, resp.Code)
+				var completionResp Completion
+				if err := json.NewDecoder(resp.Body).Decode(&completionResp); err != nil {
+					t.Fatal(err)
+				}
+
+				if completionResp.Object != "text_completion" {
+					t.Fatalf("expected text_completion, got %s", completionResp.Object)
+				}
+
+				if completionResp.Choices[0].Text != "Hello!" {
+					t.Fatalf("expected Hello!, got %s", completionResp.Choices[0].Text)
+				}
+			},
+		},
+		{
+			Name:     "completions handler with params",
+			Method:   http.MethodPost,
+			Path:     "/api/generate",
+			TestPath: "/api/generate",
+			Handler:  CompletionsMiddleware,
+			Endpoint: func(c *gin.Context) {
+				var generateReq api.GenerateRequest
+				if err := c.ShouldBindJSON(&generateReq); err != nil {
+					c.JSON(http.StatusBadRequest, gin.H{"error": "invalid request"})
+					return
+				}
+
+				temperature := generateReq.Options["temperature"].(float64)
+				var assistantMessage string
+
+				switch temperature {
+				case 1.6:
+					assistantMessage = "Received temperature of 1.6"
+				default:
+					assistantMessage = fmt.Sprintf("Received temperature of %f", temperature)
+				}
+
+				c.JSON(http.StatusOK, api.GenerateResponse{
+					Response: assistantMessage,
+				})
+			},
+			Setup: func(t *testing.T, req *http.Request) {
+				temp := float32(0.8)
+				body := CompletionRequest{
+					Model:       "test-model",
+					Prompt:      "Hello",
+					Temperature: &temp,
+				}
+
+				bodyBytes, _ := json.Marshal(body)
+
+				req.Body = io.NopCloser(bytes.NewReader(bodyBytes))
+				req.Header.Set("Content-Type", "application/json")
+			},
+			Expected: func(t *testing.T, resp *httptest.ResponseRecorder) {
+				assert.Equal(t, http.StatusOK, resp.Code)
+				var completionResp Completion
+				if err := json.NewDecoder(resp.Body).Decode(&completionResp); err != nil {
+					t.Fatal(err)
+				}
+
+				if completionResp.Object != "text_completion" {
+					t.Fatalf("expected text_completion, got %s", completionResp.Object)
+				}
+
+				if completionResp.Choices[0].Text != "Received temperature of 1.6" {
+					t.Fatalf("expected Received temperature of 1.6, got %s", completionResp.Choices[0].Text)
+				}
+			},
+		},
+		{
+			Name:     "completions handler with error",
 			Method:   http.MethodPost,
 			Path:     "/api/generate",
 			TestPath: "/api/generate",

scripts/build_windows.ps1

@@ -107,12 +107,9 @@ function gatherDependencies() {
 
     # TODO - this varies based on host build system and MSVC version - drive from dumpbin output
     # currently works for Win11 + MSVC 2019 + Cuda V11
-    cp "${env:VCToolsRedistDir}\x64\Microsoft.VC*.CRT\msvcp140*.dll" "${script:DEPS_DIR}\ollama_runners\"
+    cp "${env:VCToolsRedistDir}\x64\Microsoft.VC*.CRT\msvcp140.dll" "${script:DEPS_DIR}\ollama_runners\"
     cp "${env:VCToolsRedistDir}\x64\Microsoft.VC*.CRT\vcruntime140.dll" "${script:DEPS_DIR}\ollama_runners\"
     cp "${env:VCToolsRedistDir}\x64\Microsoft.VC*.CRT\vcruntime140_1.dll" "${script:DEPS_DIR}\ollama_runners\"
-    foreach ($part in $("runtime", "stdio", "filesystem", "math", "convert", "heap", "string", "time", "locale", "environment")) {
-        cp "$env:VCToolsRedistDir\..\..\..\Tools\Llvm\x64\bin\api-ms-win-crt-${part}*.dll" "${script:DEPS_DIR}\ollama_runners\"
-    }
 
 
     cp "${script:SRC_DIR}\app\ollama_welcome.ps1" "${script:SRC_DIR}\dist\"

server/images.go

@@ -34,20 +34,11 @@ import (
 	"github.com/ollama/ollama/version"
 )
 
-var (
-	errCapabilities         = errors.New("does not support")
-	errCapabilityCompletion = errors.New("completion")
-	errCapabilityTools      = errors.New("tools")
-	errCapabilityInsert     = errors.New("insert")
-)
+var errCapabilityCompletion = errors.New("completion")
 
 type Capability string
 
-const (
-	CapabilityCompletion = Capability("completion")
-	CapabilityTools      = Capability("tools")
-	CapabilityInsert     = Capability("insert")
-)
+const CapabilityCompletion = Capability("completion")
 
 type registryOptions struct {
 	Insecure bool
@@ -97,15 +88,6 @@ func (m *Model) CheckCapabilities(caps ...Capability) error {
 			if _, ok := ggml.KV()[fmt.Sprintf("%s.pooling_type", ggml.KV().Architecture())]; ok {
 				errs = append(errs, errCapabilityCompletion)
 			}
-		case CapabilityTools:
-			if !slices.Contains(m.Template.Vars(), "tools") {
-				errs = append(errs, errCapabilityTools)
-			}
-		case CapabilityInsert:
-			vars := m.Template.Vars()
-			if !slices.Contains(vars, "suffix") {
-				errs = append(errs, errCapabilityInsert)
-			}
 		default:
 			slog.Error("unknown capability", "capability", cap)
 			return fmt.Errorf("unknown capability: %s", cap)
@@ -113,7 +95,7 @@ func (m *Model) CheckCapabilities(caps ...Capability) error {
 	}
 
 	if err := errors.Join(errs...); err != nil {
-		return fmt.Errorf("%w %w", errCapabilities, errors.Join(errs...))
+		return fmt.Errorf("missing capabilities: %w", errors.Join(errs...))
 	}
 
 	return nil
@@ -492,12 +474,6 @@ func CreateModel(ctx context.Context, name model.Name, modelFileDir, quantizatio
 				layers = append(layers, baseLayer.Layer)
 			}
 		case "license", "template", "system":
-			if c.Name == "template" {
-				if _, err := template.Parse(c.Args); err != nil {
-					return fmt.Errorf("%w: %s", errBadTemplate, err)
-				}
-			}
-
 			if c.Name != "license" {
 				// replace
 				layers = slices.DeleteFunc(layers, func(layer *Layer) bool {

server/model.go

@@ -4,7 +4,6 @@ import (
 	"archive/zip"
 	"bytes"
 	"context"
-	"encoding/json"
 	"errors"
 	"fmt"
 	"io"
@@ -12,9 +11,6 @@ import (
 	"net/http"
 	"os"
 	"path/filepath"
-	"slices"
-	"strings"
-	"text/template/parse"
 
 	"github.com/ollama/ollama/api"
 	"github.com/ollama/ollama/convert"
@@ -293,87 +289,3 @@ func detectContentType(r io.Reader) (string, error) {
 
 	return "unknown", nil
 }
-
-// parseToolCalls attempts to parse a JSON string into a slice of ToolCalls.
-// mxyng: this only really works if the input contains tool calls in some JSON format
-func (m *Model) parseToolCalls(s string) ([]api.ToolCall, bool) {
-	// create a subtree from the node that ranges over .ToolCalls
-	tmpl := m.Template.Subtree(func(n parse.Node) bool {
-		if t, ok := n.(*parse.RangeNode); ok {
-			return slices.Contains(template.Identifiers(t.Pipe), "ToolCalls")
-		}
-
-		return false
-	})
-
-	if tmpl == nil {
-		return nil, false
-	}
-
-	var b bytes.Buffer
-	if err := tmpl.Execute(&b, map[string][]map[string]any{
-		"ToolCalls": {
-			{
-				"Function": map[string]any{
-					"Name":      "@@name@@",
-					"Arguments": "@@arguments@@",
-				},
-			},
-		},
-	}); err != nil {
-		return nil, false
-	}
-
-	var kv map[string]string
-	// execute the subtree with placeholders to identify the keys
-	// trim any commands that might exist in the template
-	if err := json.Unmarshal(bytes.TrimSuffix(b.Bytes(), []byte(",")), &kv); err != nil {
-		return nil, false
-	}
-
-	// find the keys that correspond to the name and arguments fields
-	var name, arguments string
-	for k, v := range kv {
-		switch v {
-		case "@@name@@":
-			name = k
-		case "@@arguments@@":
-			arguments = k
-		}
-	}
-
-	var objs []map[string]any
-	for offset := 0; offset < len(s); {
-		if err := json.NewDecoder(strings.NewReader(s[offset:])).Decode(&objs); errors.Is(err, io.EOF) {
-			break
-		} else if syntax := &(json.SyntaxError{}); errors.As(err, &syntax) {
-			// skip over any syntax errors
-			offset += int(syntax.Offset)
-		} else if unmarshalType := &(json.UnmarshalTypeError{}); errors.As(err, &unmarshalType) {
-			// skip over any unmarshalable types
-			offset += int(unmarshalType.Offset)
-		} else if err != nil {
-			return nil, false
-		} else {
-			// break when an object is decoded
-			break
-		}
-	}
-
-	var toolCalls []api.ToolCall
-	for _, kv := range objs {
-		var call api.ToolCall
-		for k, v := range kv {
-			switch k {
-			case name:
-				call.Function.Name = v.(string)
-			case arguments:
-				call.Function.Arguments = v.(map[string]any)
-			}
-		}
-
-		toolCalls = append(toolCalls, call)
-	}
-
-	return toolCalls, len(toolCalls) > 0
-}

server/model_test.go

@@ -3,9 +3,7 @@ package server
 import (
 	"archive/zip"
 	"bytes"
-	"encoding/json"
 	"errors"
-	"fmt"
 	"io"
 	"os"
 	"path/filepath"
@@ -13,9 +11,7 @@ import (
 	"strings"
 	"testing"
 
-	"github.com/google/go-cmp/cmp"
 	"github.com/ollama/ollama/api"
-	"github.com/ollama/ollama/template"
 )
 
 func createZipFile(t *testing.T, name string) *os.File {
@@ -114,123 +110,3 @@ func TestExtractFromZipFile(t *testing.T) {
 		})
 	}
 }
-
-type function struct {
-	Name      string         `json:"name"`
-	Arguments map[string]any `json:"arguments"`
-}
-
-func readFile(t *testing.T, base, name string) *bytes.Buffer {
-	t.Helper()
-
-	bts, err := os.ReadFile(filepath.Join(base, name))
-	if err != nil {
-		t.Fatal(err)
-	}
-
-	return bytes.NewBuffer(bts)
-}
-
-func TestExecuteWithTools(t *testing.T) {
-	p := filepath.Join("testdata", "tools")
-	cases := []struct {
-		model  string
-		output string
-		ok     bool
-	}{
-		{"mistral", `[TOOL_CALLS]  [{"name": "get_current_weather", "arguments": {"format":"fahrenheit","location":"San Francisco, CA"}},{"name": "get_current_weather", "arguments": {"format":"celsius","location":"Toronto, Canada"}}]`, true},
-		{"mistral", `[TOOL_CALLS]  [{"name": "get_current_weather", "arguments": {"format":"fahrenheit","location":"San Francisco, CA"}},{"name": "get_current_weather", "arguments": {"format":"celsius","location":"Toronto, Canada"}}]
-
-The temperature in San Francisco, CA is 70°F and in Toronto, Canada is 20°C.`, true},
-		{"mistral", `I'm not aware of that information. However, I can suggest searching for the weather using the "get_current_weather" function:
-
-		[{"name": "get_current_weather", "arguments": {"format":"fahrenheit","location":"San Francisco, CA"}},{"name": "get_current_weather", "arguments": {"format":"celsius","location":"Toronto, Canada"}}]`, true},
-		{"mistral", " The weather in San Francisco, CA is 70°F and in Toronto, Canada is 20°C.", false},
-		{"command-r-plus", "Action: ```json" + `
-[
-    {
-        "tool_name": "get_current_weather",
-        "parameters": {
-            "format": "fahrenheit",
-            "location": "San Francisco, CA"
-        }
-    },
-    {
-        "tool_name": "get_current_weather",
-        "parameters": {
-            "format": "celsius",
-            "location": "Toronto, Canada"
-        }
-    }
-]
-` + "```", true},
-		{"command-r-plus", " The weather in San Francisco, CA is 70°F and in Toronto, Canada is 20°C.", false},
-		{"firefunction", ` functools[{"name": "get_current_weather", "arguments": {"format":"fahrenheit","location":"San Francisco, CA"}},{"name": "get_current_weather", "arguments": {"format":"celsius","location":"Toronto, Canada"}}]`, true},
-		{"firefunction", " The weather in San Francisco, CA is 70°F and in Toronto, Canada is 20°C.", false},
-	}
-
-	var tools []api.Tool
-	if err := json.Unmarshal(readFile(t, p, "tools.json").Bytes(), &tools); err != nil {
-		t.Fatal(err)
-	}
-
-	var messages []api.Message
-	if err := json.Unmarshal(readFile(t, p, "messages.json").Bytes(), &messages); err != nil {
-		t.Fatal(err)
-	}
-
-	calls := []api.ToolCall{
-		{
-			Function: function{
-				Name: "get_current_weather",
-				Arguments: map[string]any{
-					"format":   "fahrenheit",
-					"location": "San Francisco, CA",
-				},
-			},
-		},
-		{
-			Function: function{
-				Name: "get_current_weather",
-				Arguments: map[string]any{
-					"format":   "celsius",
-					"location": "Toronto, Canada",
-				},
-			},
-		},
-	}
-
-	for _, tt := range cases {
-		t.Run(tt.model, func(t *testing.T) {
-			tmpl, err := template.Parse(readFile(t, p, fmt.Sprintf("%s.gotmpl", tt.model)).String())
-			if err != nil {
-				t.Fatal(err)
-			}
-
-			t.Run("template", func(t *testing.T) {
-				var actual bytes.Buffer
-				if err := tmpl.Execute(&actual, template.Values{Tools: tools, Messages: messages}); err != nil {
-					t.Fatal(err)
-				}
-
-				if diff := cmp.Diff(actual.String(), readFile(t, p, fmt.Sprintf("%s.out", tt.model)).String()); diff != "" {
-					t.Errorf("mismatch (-got +want):\n%s", diff)
-				}
-			})
-
-			t.Run("parse", func(t *testing.T) {
-				m := &Model{Template: tmpl}
-				actual, ok := m.parseToolCalls(tt.output)
-				if ok != tt.ok {
-					t.Fatalf("expected %t, got %t", tt.ok, ok)
-				}
-
-				if tt.ok {
-					if diff := cmp.Diff(actual, calls); diff != "" {
-						t.Errorf("mismatch (-got +want):\n%s", diff)
-					}
-				}
-			})
-		})
-	}
-}

server/prompt.go

@@ -4,6 +4,7 @@ import (
 	"bytes"
 	"context"
 	"log/slog"
+	"slices"
 
 	"github.com/ollama/ollama/api"
 	"github.com/ollama/ollama/llm"
@@ -15,21 +16,29 @@ type tokenizeFunc func(context.Context, string) ([]int, error)
 // chatPrompt accepts a list of messages and returns the prompt and images that should be used for the next chat turn.
 // chatPrompt truncates any messages that exceed the context window of the model, making sure to always include 1) the
 // latest message and 2) system messages
-func chatPrompt(ctx context.Context, m *Model, tokenize tokenizeFunc, opts *api.Options, msgs []api.Message, tools []api.Tool) (prompt string, images []llm.ImageData, _ error) {
+func chatPrompt(ctx context.Context, m *Model, tokenize tokenizeFunc, opts *api.Options, msgs []api.Message) (prompt string, images []llm.ImageData, _ error) {
+	// pull out any system messages which should always be included in the prompt
 	var system []api.Message
+	msgs = slices.DeleteFunc(msgs, func(m api.Message) bool {
+		if m.Role == "system" {
+			system = append(system, m)
+			return true
+		}
+
+		return false
+	})
+
+	if len(system) == 0 && m.System != "" {
+		// add model system prompt since it wasn't provided
+		system = append(system, api.Message{Role: "system", Content: m.System})
+	}
+
 	// always include the last message
 	n := len(msgs) - 1
 	// in reverse, find all messages that fit into context window
 	for i := n - 1; i >= 0; i-- {
-		system = make([]api.Message, 0)
-		for j := range i {
-			if msgs[j].Role == "system" {
-				system = append(system, msgs[j])
-			}
-		}
-
 		var b bytes.Buffer
-		if err := m.Template.Execute(&b, template.Values{Messages: append(system, msgs[i:]...), Tools: tools}); err != nil {
+		if err := m.Template.Execute(&b, template.Values{Messages: append(system, msgs[i:]...)}); err != nil {
 			return "", nil, err
 		}
 
@@ -57,7 +66,7 @@ func chatPrompt(ctx context.Context, m *Model, tokenize tokenizeFunc, opts *api.
 
 	// truncate any messages that do not fit into the context window
 	var b bytes.Buffer
-	if err := m.Template.Execute(&b, template.Values{Messages: append(system, msgs[n:]...), Tools: tools}); err != nil {
+	if err := m.Template.Execute(&b, template.Values{Messages: append(system, msgs[n:]...)}); err != nil {
 		return "", nil, err
 	}
 

server/prompt_test.go

@@ -3,13 +3,21 @@ package server
 import (
 	"bytes"
 	"context"
+	"strings"
 	"testing"
 
-	"github.com/google/go-cmp/cmp"
 	"github.com/ollama/ollama/api"
 	"github.com/ollama/ollama/template"
 )
 
+func tokenize(_ context.Context, s string) (tokens []int, err error) {
+	for range strings.Fields(s) {
+		tokens = append(tokens, len(tokens))
+	}
+
+	return
+}
+
 func TestChatPrompt(t *testing.T) {
 	type expect struct {
 		prompt string
@@ -152,19 +160,6 @@ func TestChatPrompt(t *testing.T) {
 				{Role: "assistant", Content: "I-I'm a what?"},
 				{Role: "user", Content: "A test. And a thumping good one at that, I'd wager."},
 			},
-			expect: expect{
-				prompt: "You are the Test Who Lived. You're a test, Harry! I-I'm a what? A test. And a thumping good one at that, I'd wager. ",
-			},
-		},
-		{
-			name:  "out of order system",
-			limit: 2048,
-			msgs: []api.Message{
-				{Role: "user", Content: "You're a test, Harry!"},
-				{Role: "assistant", Content: "I-I'm a what?"},
-				{Role: "system", Content: "You are the Test Who Lived."},
-				{Role: "user", Content: "A test. And a thumping good one at that, I'd wager."},
-			},
 			expect: expect{
 				prompt: "You're a test, Harry! I-I'm a what? You are the Test Who Lived. A test. And a thumping good one at that, I'd wager. ",
 			},
@@ -183,13 +178,13 @@ func TestChatPrompt(t *testing.T) {
 		t.Run(tt.name, func(t *testing.T) {
 			model := Model{Template: tmpl, ProjectorPaths: []string{"vision"}}
 			opts := api.Options{Runner: api.Runner{NumCtx: tt.limit}}
-			prompt, images, err := chatPrompt(context.TODO(), &model, mockRunner{}.Tokenize, &opts, tt.msgs, nil)
+			prompt, images, err := chatPrompt(context.TODO(), &model, tokenize, &opts, tt.msgs)
 			if err != nil {
 				t.Fatal(err)
 			}
 
-			if diff := cmp.Diff(prompt, tt.prompt); diff != "" {
-				t.Errorf("mismatch (-got +want):\n%s", diff)
+			if tt.prompt != prompt {
+				t.Errorf("expected %q, got %q", tt.prompt, prompt)
 			}
 
 			if len(images) != len(tt.images) {

server/routes.go

@@ -18,6 +18,7 @@ import (
 	"path/filepath"
 	"slices"
 	"strings"
+	"sync"
 	"syscall"
 	"time"
 
@@ -56,7 +57,6 @@ func init() {
 }
 
 var errRequired = errors.New("is required")
-var errBadTemplate = errors.New("template error")
 
 func modelOptions(model *Model, requestOpts map[string]interface{}) (api.Options, error) {
 	opts := api.DefaultOptions()
@@ -104,7 +104,6 @@ func (s *Server) scheduleRunner(ctx context.Context, name string, caps []Capabil
 }
 
 func (s *Server) GenerateHandler(c *gin.Context) {
-	checkpointStart := time.Now()
 	var req api.GenerateRequest
 	if err := c.ShouldBindJSON(&req); errors.Is(err, io.EOF) {
 		c.AbortWithStatusJSON(http.StatusBadRequest, gin.H{"error": "missing request body"})
@@ -123,10 +122,6 @@ func (s *Server) GenerateHandler(c *gin.Context) {
 	}
 
 	caps := []Capability{CapabilityCompletion}
-	if req.Suffix != "" {
-		caps = append(caps, CapabilityInsert)
-	}
-
 	r, m, opts, err := s.scheduleRunner(c.Request.Context(), req.Model, caps, req.Options, req.KeepAlive)
 	if errors.Is(err, errCapabilityCompletion) {
 		c.JSON(http.StatusBadRequest, gin.H{"error": fmt.Sprintf("%q does not support generate", req.Model)})
@@ -136,8 +131,6 @@ func (s *Server) GenerateHandler(c *gin.Context) {
 		return
 	}
 
-	checkpointLoaded := time.Now()
-
 	if req.Prompt == "" {
 		c.JSON(http.StatusOK, api.GenerateResponse{
 			Model:      req.Model,
@@ -155,6 +148,19 @@ func (s *Server) GenerateHandler(c *gin.Context) {
 
 	prompt := req.Prompt
 	if !req.Raw {
+		var msgs []api.Message
+		if req.System != "" {
+			msgs = append(msgs, api.Message{Role: "system", Content: req.System})
+		} else if m.System != "" {
+			msgs = append(msgs, api.Message{Role: "system", Content: m.System})
+		}
+
+		for _, i := range images {
+			msgs = append(msgs, api.Message{Role: "user", Content: fmt.Sprintf("[img-%d]", i.ID)})
+		}
+
+		msgs = append(msgs, api.Message{Role: "user", Content: req.Prompt})
+
 		tmpl := m.Template
 		if req.Template != "" {
 			tmpl, err = template.Parse(req.Template)
@@ -175,26 +181,7 @@ func (s *Server) GenerateHandler(c *gin.Context) {
 			b.WriteString(s)
 		}
 
-		var values template.Values
-		if req.Suffix != "" {
-			values.Prompt = prompt
-			values.Suffix = req.Suffix
-		} else {
-			var msgs []api.Message
-			if req.System != "" {
-				msgs = append(msgs, api.Message{Role: "system", Content: req.System})
-			} else if m.System != "" {
-				msgs = append(msgs, api.Message{Role: "system", Content: m.System})
-			}
-
-			for _, i := range images {
-				msgs = append(msgs, api.Message{Role: "user", Content: fmt.Sprintf("[img-%d]", i.ID)})
-			}
-
-			values.Messages = append(msgs, api.Message{Role: "user", Content: req.Prompt})
-		}
-
-		if err := tmpl.Execute(&b, values); err != nil {
+		if err := tmpl.Execute(&b, template.Values{Messages: msgs}); err != nil {
 			c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
 			return
 		}
@@ -206,48 +193,26 @@ func (s *Server) GenerateHandler(c *gin.Context) {
 
 	ch := make(chan any)
 	go func() {
-		// TODO (jmorganca): avoid building the response twice both here and below
-		var sb strings.Builder
 		defer close(ch)
 		if err := r.Completion(c.Request.Context(), llm.CompletionRequest{
 			Prompt:  prompt,
 			Images:  images,
 			Format:  req.Format,
 			Options: opts,
-		}, func(cr llm.CompletionResponse) {
-			res := api.GenerateResponse{
+		}, func(r llm.CompletionResponse) {
+			ch <- api.GenerateResponse{
 				Model:      req.Model,
 				CreatedAt:  time.Now().UTC(),
-				Response:   cr.Content,
-				Done:       cr.Done,
-				DoneReason: cr.DoneReason,
+				Response:   r.Content,
+				Done:       r.Done,
+				DoneReason: r.DoneReason,
 				Metrics: api.Metrics{
-					PromptEvalCount:    cr.PromptEvalCount,
-					PromptEvalDuration: cr.PromptEvalDuration,
-					EvalCount:          cr.EvalCount,
-					EvalDuration:       cr.EvalDuration,
+					PromptEvalCount:    r.PromptEvalCount,
+					PromptEvalDuration: r.PromptEvalDuration,
+					EvalCount:          r.EvalCount,
+					EvalDuration:       r.EvalDuration,
 				},
 			}
-
-			if _, err := sb.WriteString(cr.Content); err != nil {
-				ch <- gin.H{"error": err.Error()}
-			}
-
-			if cr.Done {
-				res.TotalDuration = time.Since(checkpointStart)
-				res.LoadDuration = checkpointLoaded.Sub(checkpointStart)
-
-				if !req.Raw {
-					tokens, err := r.Tokenize(c.Request.Context(), prompt+sb.String())
-					if err != nil {
-						ch <- gin.H{"error": err.Error()}
-						return
-					}
-					res.Context = append(req.Context, tokens...)
-				}
-			}
-
-			ch <- res
 		}); err != nil {
 			ch <- gin.H{"error": err.Error()}
 		}
@@ -276,11 +241,6 @@ func (s *Server) GenerateHandler(c *gin.Context) {
 		}
 
 		r.Response = sb.String()
-		if toolCalls, ok := m.parseToolCalls(sb.String()); ok {
-			r.ToolCalls = toolCalls
-			r.Response = ""
-		}
-
 		c.JSON(http.StatusOK, r)
 		return
 	}
@@ -300,37 +260,38 @@ func (s *Server) EmbedHandler(c *gin.Context) {
 		return
 	}
 
-	truncate := true
-
-	if req.Truncate != nil && !*req.Truncate {
-		truncate = false
+	if req.Truncate == nil {
+		truncate := true
+		req.Truncate = &truncate
 	}
 
-	var input []string
+	reqEmbed := []string{}
 
-	switch i := req.Input.(type) {
+	switch embeddings := req.Input.(type) {
 	case string:
-		if len(i) > 0 {
-			input = append(input, i)
+		if embeddings == "" {
+			c.JSON(http.StatusOK, api.EmbedResponse{Model: req.Model, Embeddings: [][]float32{}})
+			return
 		}
+		reqEmbed = []string{embeddings}
 	case []any:
-		for _, v := range i {
+		if len(embeddings) == 0 {
+			c.JSON(http.StatusOK, api.EmbedResponse{Model: req.Model, Embeddings: [][]float32{}})
+			return
+		}
+
+		for _, v := range embeddings {
 			if _, ok := v.(string); !ok {
 				c.AbortWithStatusJSON(http.StatusBadRequest, gin.H{"error": "invalid input type"})
 				return
 			}
-			input = append(input, v.(string))
+			reqEmbed = append(reqEmbed, v.(string))
 		}
 	default:
 		c.AbortWithStatusJSON(http.StatusBadRequest, gin.H{"error": "invalid input type"})
 		return
 	}
 
-	if len(input) == 0 {
-		c.JSON(http.StatusOK, api.EmbedResponse{Model: req.Model, Embeddings: [][]float32{}})
-		return
-	}
-
 	r, m, opts, err := s.scheduleRunner(c.Request.Context(), req.Model, []Capability{}, req.Options, req.KeepAlive)
 	if err != nil {
 		handleScheduleError(c, req.Model, err)
@@ -343,34 +304,64 @@ func (s *Server) EmbedHandler(c *gin.Context) {
 		return
 	}
 
-	for i, s := range input {
-		tokens, err := r.Tokenize(c.Request.Context(), s)
-		if err != nil {
-			c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
+	reqEmbedArray := make([]string, len(reqEmbed))
+	errCh := make(chan error, 1)
+	successCh := make(chan bool, 1)
+	sem := make(chan struct{}, 2)
+	var wg sync.WaitGroup
+	var mu sync.Mutex
+	for i, s := range reqEmbed {
+		wg.Add(1)
+		sem <- struct{}{}
+		go func(i int, s string) {
+			defer wg.Done()
+			defer func() { <-sem }()
+			tokens, err := r.Tokenize(c.Request.Context(), s)
+			if err != nil {
+				errCh <- err
+				return
+			}
+
+			ctxLen := min(opts.NumCtx, int(kvData.ContextLength()))
+			if len(tokens) > ctxLen {
+				if *req.Truncate {
+					tokens = tokens[:ctxLen]
+					s, err = r.Detokenize(c.Request.Context(), tokens)
+					if err != nil {
+						errCh <- err
+						return
+					}
+				} else {
+					errCh <- err
+					return
+				}
+			}
+			mu.Lock()
+			reqEmbedArray[i] = s
+			mu.Unlock()
+		}(i, s)
+	}
+	go func() {
+		wg.Wait()
+		successCh <- true
+		close(errCh)
+	}()
+
+	select {
+	case err := <-errCh:
+		c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
+		return
+	case success := <-successCh:
+		if !success {
+			c.JSON(http.StatusInternalServerError, gin.H{"error": "failed to process all embeddings"})
 			return
 		}
-
-		ctxLen := min(opts.NumCtx, int(kvData.ContextLength()))
-		if len(tokens) > ctxLen {
-			if !truncate {
-				c.JSON(http.StatusBadRequest, gin.H{"error": "input length exceeds maximum context length"})
-				return
-			}
-
-			tokens = tokens[:ctxLen]
-			s, err = r.Detokenize(c.Request.Context(), tokens)
-			if err != nil {
-				c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
-				return
-			}
-		}
-
-		input[i] = s
 	}
-	embeddings, err := r.Embed(c.Request.Context(), input)
+
+	embeddings, err := r.Embed(c.Request.Context(), reqEmbedArray)
 
 	if err != nil {
-		slog.Error("embedding generation failed", "error", err)
+		slog.Info(fmt.Sprintf("embedding generation failed: %v", err))
 		c.JSON(http.StatusInternalServerError, gin.H{"error": "failed to generate embedding"})
 		return
 	}
@@ -425,7 +416,7 @@ func (s *Server) EmbeddingsHandler(c *gin.Context) {
 		return
 	}
 
-	embeddings, err := r.Embed(c.Request.Context(), []string{req.Prompt})
+	embedding, err := r.Embed(c.Request.Context(), []string{req.Prompt})
 
 	if err != nil {
 		slog.Info(fmt.Sprintf("embedding generation failed: %v", err))
@@ -433,14 +424,14 @@ func (s *Server) EmbeddingsHandler(c *gin.Context) {
 		return
 	}
 
-	embedding := make([]float64, len(embeddings[0]))
+	embedding64 := make([]float64, len(embedding[0]))
 
-	for i, v := range embeddings[0] {
-		embedding[i] = float64(v)
+	for i, v := range embedding[0] {
+		embedding64[i] = float64(v)
 	}
 
 	resp := api.EmbeddingResponse{
-		Embedding: embedding,
+		Embedding: embedding64,
 	}
 	c.JSON(http.StatusOK, resp)
 }
@@ -614,9 +605,7 @@ func (s *Server) CreateModelHandler(c *gin.Context) {
 		defer cancel()
 
 		quantization := cmp.Or(r.Quantize, r.Quantization)
-		if err := CreateModel(ctx, name, filepath.Dir(r.Path), strings.ToUpper(quantization), f, fn); errors.Is(err, errBadTemplate) {
-			ch <- gin.H{"error": err.Error(), "status": http.StatusBadRequest}
-		} else if err != nil {
+		if err := CreateModel(ctx, name, filepath.Dir(r.Path), strings.ToUpper(quantization), f, fn); err != nil {
 			ch <- gin.H{"error": err.Error()}
 		}
 	}()
@@ -718,6 +707,13 @@ func GetModelInfo(req api.ShowRequest) (*api.ShowResponse, error) {
 		m.System = req.System
 	}
 
+	if req.Template != "" {
+		m.Template, err = template.Parse(req.Template)
+		if err != nil {
+			return nil, err
+		}
+	}
+
 	msgs := make([]api.Message, len(m.Messages))
 	for i, msg := range m.Messages {
 		msgs[i] = api.Message{Role: msg.Role, Content: msg.Content}
@@ -1064,7 +1060,7 @@ func (s *Server) GenerateRoutes() http.Handler {
 	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/embeddings", s.EmbeddingsHandler) // legacy
 	r.POST("/api/create", s.CreateModelHandler)
 	r.POST("/api/push", s.PushModelHandler)
 	r.POST("/api/copy", s.CopyModelHandler)
@@ -1077,7 +1073,6 @@ func (s *Server) GenerateRoutes() http.Handler {
 	// 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.ListModelsHandler)
 	r.GET("/v1/models/:model", openai.RetrieveMiddleware(), s.ShowModelHandler)
 
@@ -1204,15 +1199,11 @@ func waitForStream(c *gin.Context, ch chan interface{}) {
 				return
 			}
 		case gin.H:
-			status, ok := r["status"].(int)
-			if !ok {
-				status = http.StatusInternalServerError
-			}
 			if errorMsg, ok := r["error"].(string); ok {
-				c.JSON(status, gin.H{"error": errorMsg})
+				c.JSON(http.StatusInternalServerError, gin.H{"error": errorMsg})
 				return
 			} else {
-				c.JSON(status, gin.H{"error": "unexpected error format in progress response"})
+				c.JSON(http.StatusInternalServerError, gin.H{"error": "unexpected error format in progress response"})
 				return
 			}
 		default:
@@ -1290,8 +1281,6 @@ func (s *Server) ProcessHandler(c *gin.Context) {
 }
 
 func (s *Server) ChatHandler(c *gin.Context) {
-	checkpointStart := time.Now()
-
 	var req api.ChatRequest
 	if err := c.ShouldBindJSON(&req); errors.Is(err, io.EOF) {
 		c.AbortWithStatusJSON(http.StatusBadRequest, gin.H{"error": "missing request body"})
@@ -1302,10 +1291,6 @@ func (s *Server) ChatHandler(c *gin.Context) {
 	}
 
 	caps := []Capability{CapabilityCompletion}
-	if req.Tools != nil {
-		caps = append(caps, CapabilityTools)
-	}
-
 	r, m, opts, err := s.scheduleRunner(c.Request.Context(), req.Model, caps, req.Options, req.KeepAlive)
 	if errors.Is(err, errCapabilityCompletion) {
 		c.JSON(http.StatusBadRequest, gin.H{"error": fmt.Sprintf("%q does not support chat", req.Model)})
@@ -1315,8 +1300,6 @@ func (s *Server) ChatHandler(c *gin.Context) {
 		return
 	}
 
-	checkpointLoaded := time.Now()
-
 	if len(req.Messages) == 0 {
 		c.JSON(http.StatusOK, api.ChatResponse{
 			Model:      req.Model,
@@ -1328,11 +1311,7 @@ func (s *Server) ChatHandler(c *gin.Context) {
 		return
 	}
 
-	if req.Messages[0].Role != "system" && m.System != "" {
-		req.Messages = append([]api.Message{{Role: "system", Content: m.System}}, req.Messages...)
-	}
-
-	prompt, images, err := chatPrompt(c.Request.Context(), m, r.Tokenize, opts, req.Messages, req.Tools)
+	prompt, images, err := chatPrompt(c.Request.Context(), m, r.Tokenize, opts, req.Messages)
 	if err != nil {
 		c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
 		return
@@ -1349,7 +1328,7 @@ func (s *Server) ChatHandler(c *gin.Context) {
 			Format:  req.Format,
 			Options: opts,
 		}, func(r llm.CompletionResponse) {
-			res := api.ChatResponse{
+			ch <- api.ChatResponse{
 				Model:      req.Model,
 				CreatedAt:  time.Now().UTC(),
 				Message:    api.Message{Role: "assistant", Content: r.Content},
@@ -1362,26 +1341,19 @@ func (s *Server) ChatHandler(c *gin.Context) {
 					EvalDuration:       r.EvalDuration,
 				},
 			}
-
-			if r.Done {
-				res.TotalDuration = time.Since(checkpointStart)
-				res.LoadDuration = checkpointLoaded.Sub(checkpointStart)
-			}
-
-			ch <- res
 		}); err != nil {
 			ch <- gin.H{"error": err.Error()}
 		}
 	}()
 
 	if req.Stream != nil && !*req.Stream {
-		var resp api.ChatResponse
+		var r api.ChatResponse
 		var sb strings.Builder
 		for rr := range ch {
 			switch t := rr.(type) {
 			case api.ChatResponse:
 				sb.WriteString(t.Message.Content)
-				resp = t
+				r = t
 			case gin.H:
 				msg, ok := t["error"].(string)
 				if !ok {
@@ -1396,13 +1368,8 @@ func (s *Server) ChatHandler(c *gin.Context) {
 			}
 		}
 
-		resp.Message.Content = sb.String()
-		if toolCalls, ok := m.parseToolCalls(sb.String()); ok {
-			resp.Message.ToolCalls = toolCalls
-			resp.Message.Content = ""
-		}
-
-		c.JSON(http.StatusOK, resp)
+		r.Message.Content = sb.String()
+		c.JSON(http.StatusOK, r)
 		return
 	}
 
@@ -1411,7 +1378,7 @@ func (s *Server) ChatHandler(c *gin.Context) {
 
 func handleScheduleError(c *gin.Context, name string, err error) {
 	switch {
-	case errors.Is(err, errCapabilities), errors.Is(err, errRequired):
+	case errors.Is(err, errRequired):
 		c.JSON(http.StatusBadRequest, gin.H{"error": err.Error()})
 	case errors.Is(err, context.Canceled):
 		c.JSON(499, gin.H{"error": "request canceled"})

server/routes_create_test.go

@@ -85,8 +85,6 @@ func checkFileExists(t *testing.T, p string, expect []string) {
 }
 
 func TestCreateFromBin(t *testing.T) {
-	gin.SetMode(gin.TestMode)
-
 	p := t.TempDir()
 	t.Setenv("OLLAMA_MODELS", p)
 	envconfig.LoadConfig()
@@ -113,8 +111,6 @@ func TestCreateFromBin(t *testing.T) {
 }
 
 func TestCreateFromModel(t *testing.T) {
-	gin.SetMode(gin.TestMode)
-
 	p := t.TempDir()
 	t.Setenv("OLLAMA_MODELS", p)
 	envconfig.LoadConfig()
@@ -156,8 +152,6 @@ func TestCreateFromModel(t *testing.T) {
 }
 
 func TestCreateRemovesLayers(t *testing.T) {
-	gin.SetMode(gin.TestMode)
-
 	p := t.TempDir()
 	t.Setenv("OLLAMA_MODELS", p)
 	envconfig.LoadConfig()
@@ -205,8 +199,6 @@ func TestCreateRemovesLayers(t *testing.T) {
 }
 
 func TestCreateUnsetsSystem(t *testing.T) {
-	gin.SetMode(gin.TestMode)
-
 	p := t.TempDir()
 	t.Setenv("OLLAMA_MODELS", p)
 	envconfig.LoadConfig()
@@ -263,8 +255,6 @@ func TestCreateUnsetsSystem(t *testing.T) {
 }
 
 func TestCreateMergeParameters(t *testing.T) {
-	gin.SetMode(gin.TestMode)
-
 	p := t.TempDir()
 	t.Setenv("OLLAMA_MODELS", p)
 	envconfig.LoadConfig()
@@ -368,8 +358,6 @@ func TestCreateMergeParameters(t *testing.T) {
 }
 
 func TestCreateReplacesMessages(t *testing.T) {
-	gin.SetMode(gin.TestMode)
-
 	p := t.TempDir()
 	t.Setenv("OLLAMA_MODELS", p)
 	envconfig.LoadConfig()
@@ -446,8 +434,6 @@ func TestCreateReplacesMessages(t *testing.T) {
 }
 
 func TestCreateTemplateSystem(t *testing.T) {
-	gin.SetMode(gin.TestMode)
-
 	p := t.TempDir()
 	t.Setenv("OLLAMA_MODELS", p)
 	envconfig.LoadConfig()
@@ -491,47 +477,9 @@ func TestCreateTemplateSystem(t *testing.T) {
 	if string(system) != "Say bye!" {
 		t.Errorf("expected \"Say bye!\", actual %s", system)
 	}
-
-	t.Run("incomplete template", func(t *testing.T) {
-		w := createRequest(t, s.CreateModelHandler, api.CreateRequest{
-			Name:      "test",
-			Modelfile: fmt.Sprintf("FROM %s\nTEMPLATE {{ .Prompt", createBinFile(t, nil, nil)),
-			Stream:    &stream,
-		})
-	
-		if w.Code != http.StatusBadRequest {
-			t.Fatalf("expected status code 400, actual %d", w.Code)
-		}
-	})
-
-	t.Run("template with unclosed if", func(t *testing.T) {
-		w := createRequest(t, s.CreateModelHandler, api.CreateRequest{
-			Name:      "test",
-			Modelfile: fmt.Sprintf("FROM %s\nTEMPLATE {{ if .Prompt }}", createBinFile(t, nil, nil)),
-			Stream:    &stream,
-		})
-	
-		if w.Code != http.StatusBadRequest {
-			t.Fatalf("expected status code 400, actual %d", w.Code)
-		}
-	})
-
-	t.Run("template with undefined function", func(t *testing.T) {
-		w := createRequest(t, s.CreateModelHandler, api.CreateRequest{
-			Name:      "test",
-			Modelfile: fmt.Sprintf("FROM %s\nTEMPLATE {{  Prompt }}", createBinFile(t, nil, nil)),
-			Stream:    &stream,
-		})
-	
-		if w.Code != http.StatusBadRequest {
-			t.Fatalf("expected status code 400, actual %d", w.Code)
-		}
-	})
 }
 
 func TestCreateLicenses(t *testing.T) {
-	gin.SetMode(gin.TestMode)
-
 	p := t.TempDir()
 	t.Setenv("OLLAMA_MODELS", p)
 	envconfig.LoadConfig()
@@ -578,8 +526,6 @@ func TestCreateLicenses(t *testing.T) {
 }
 
 func TestCreateDetectTemplate(t *testing.T) {
-	gin.SetMode(gin.TestMode)
-
 	p := t.TempDir()
 	t.Setenv("OLLAMA_MODELS", p)
 	envconfig.LoadConfig()
@@ -600,8 +546,8 @@ func TestCreateDetectTemplate(t *testing.T) {
 
 		checkFileExists(t, filepath.Join(p, "blobs", "*"), []string{
 			filepath.Join(p, "blobs", "sha256-553c4a3f747b3d22a4946875f1cc8ed011c2930d83f864a0c7265f9ec0a20413"),
-			filepath.Join(p, "blobs", "sha256-c608dc615584cd20d9d830363dabf8a4783ae5d34245c3d8c115edb3bc7b28e4"),
-			filepath.Join(p, "blobs", "sha256-f836ee110db21567f826332e4cedd746c06d10664fd5a9ea3659e3683a944510"),
+			filepath.Join(p, "blobs", "sha256-9512c372dfc7d84d6065b8dd2b601aeed8cc1a78e7a7aa784a42fff37f5524b7"),
+			filepath.Join(p, "blobs", "sha256-b8b78cb8c6eefd14c06f1af042e6161255bf87bbf2dd14fce57cdac893db8139"),
 		})
 	})
 

server/routes_delete_test.go

@@ -8,15 +8,12 @@ import (
 	"path/filepath"
 	"testing"
 
-	"github.com/gin-gonic/gin"
 	"github.com/ollama/ollama/api"
 	"github.com/ollama/ollama/envconfig"
 	"github.com/ollama/ollama/types/model"
 )
 
 func TestDelete(t *testing.T) {
-	gin.SetMode(gin.TestMode)
-
 	p := t.TempDir()
 	t.Setenv("OLLAMA_MODELS", p)
 	envconfig.LoadConfig()
@@ -80,8 +77,6 @@ func TestDelete(t *testing.T) {
 }
 
 func TestDeleteDuplicateLayers(t *testing.T) {
-	gin.SetMode(gin.TestMode)
-
 	p := t.TempDir()
 	t.Setenv("OLLAMA_MODELS", p)
 	var s Server

server/routes_generate_test.go

@@ -1,712 +0,0 @@
-package server
-
-import (
-	"bytes"
-	"context"
-	"encoding/json"
-	"fmt"
-	"io"
-	"net/http"
-	"strings"
-	"testing"
-	"time"
-
-	"github.com/gin-gonic/gin"
-	"github.com/google/go-cmp/cmp"
-
-	"github.com/ollama/ollama/api"
-	"github.com/ollama/ollama/gpu"
-	"github.com/ollama/ollama/llm"
-)
-
-type mockRunner struct {
-	llm.LlamaServer
-
-	// CompletionRequest is only valid until the next call to Completion
-	llm.CompletionRequest
-	llm.CompletionResponse
-}
-
-func (m *mockRunner) Completion(_ context.Context, r llm.CompletionRequest, fn func(r llm.CompletionResponse)) error {
-	m.CompletionRequest = r
-	fn(m.CompletionResponse)
-	return nil
-}
-
-func (mockRunner) Tokenize(_ context.Context, s string) (tokens []int, err error) {
-	for range strings.Fields(s) {
-		tokens = append(tokens, len(tokens))
-	}
-
-	return
-}
-
-func newMockServer(mock *mockRunner) func(gpu.GpuInfoList, string, *llm.GGML, []string, []string, api.Options, int) (llm.LlamaServer, error) {
-	return func(gpus gpu.GpuInfoList, model string, ggml *llm.GGML, projectors, system []string, opts api.Options, numParallel int) (llm.LlamaServer, error) {
-		return mock, nil
-	}
-}
-
-func TestGenerateChat(t *testing.T) {
-	gin.SetMode(gin.TestMode)
-
-	mock := mockRunner{
-		CompletionResponse: llm.CompletionResponse{
-			Done:               true,
-			DoneReason:         "stop",
-			PromptEvalCount:    1,
-			PromptEvalDuration: 1,
-			EvalCount:          1,
-			EvalDuration:       1,
-		},
-	}
-
-	s := Server{
-		sched: &Scheduler{
-			pendingReqCh:  make(chan *LlmRequest, 1),
-			finishedReqCh: make(chan *LlmRequest, 1),
-			expiredCh:     make(chan *runnerRef, 1),
-			unloadedCh:    make(chan any, 1),
-			loaded:        make(map[string]*runnerRef),
-			newServerFn:   newMockServer(&mock),
-			getGpuFn:      gpu.GetGPUInfo,
-			getCpuFn:      gpu.GetCPUInfo,
-			reschedDelay:  250 * time.Millisecond,
-			loadFn: func(req *LlmRequest, ggml *llm.GGML, gpus gpu.GpuInfoList, numParallel int) {
-				// add 10ms delay to simulate loading
-				time.Sleep(10 * time.Millisecond)
-				req.successCh <- &runnerRef{
-					llama: &mock,
-				}
-			},
-		},
-	}
-
-	go s.sched.Run(context.TODO())
-
-	w := createRequest(t, s.CreateModelHandler, api.CreateRequest{
-		Model: "test",
-		Modelfile: fmt.Sprintf(`FROM %s
-		TEMPLATE """
-{{- if .System }}System: {{ .System }} {{ end }}
-{{- if .Prompt }}User: {{ .Prompt }} {{ end }}
-{{- if .Response }}Assistant: {{ .Response }} {{ end }}"""
-`, createBinFile(t, llm.KV{
-			"general.architecture":          "llama",
-			"llama.block_count":             uint32(1),
-			"llama.context_length":          uint32(8192),
-			"llama.embedding_length":        uint32(4096),
-			"llama.attention.head_count":    uint32(32),
-			"llama.attention.head_count_kv": uint32(8),
-			"tokenizer.ggml.tokens":         []string{""},
-			"tokenizer.ggml.scores":         []float32{0},
-			"tokenizer.ggml.token_type":     []int32{0},
-		}, []llm.Tensor{
-			{Name: "token_embd.weight", Shape: []uint64{1}, WriterTo: bytes.NewReader(make([]byte, 4))},
-			{Name: "blk.0.attn_norm.weight", Shape: []uint64{1}, WriterTo: bytes.NewReader(make([]byte, 4))},
-			{Name: "blk.0.ffn_down.weight", Shape: []uint64{1}, WriterTo: bytes.NewReader(make([]byte, 4))},
-			{Name: "blk.0.ffn_gate.weight", Shape: []uint64{1}, WriterTo: bytes.NewReader(make([]byte, 4))},
-			{Name: "blk.0.ffn_up.weight", Shape: []uint64{1}, WriterTo: bytes.NewReader(make([]byte, 4))},
-			{Name: "blk.0.ffn_norm.weight", Shape: []uint64{1}, WriterTo: bytes.NewReader(make([]byte, 4))},
-			{Name: "blk.0.attn_k.weight", Shape: []uint64{1}, WriterTo: bytes.NewReader(make([]byte, 4))},
-			{Name: "blk.0.attn_output.weight", Shape: []uint64{1}, WriterTo: bytes.NewReader(make([]byte, 4))},
-			{Name: "blk.0.attn_q.weight", Shape: []uint64{1}, WriterTo: bytes.NewReader(make([]byte, 4))},
-			{Name: "blk.0.attn_v.weight", Shape: []uint64{1}, WriterTo: bytes.NewReader(make([]byte, 4))},
-			{Name: "output.weight", Shape: []uint64{1}, WriterTo: bytes.NewReader(make([]byte, 4))},
-		})),
-		Stream: &stream,
-	})
-
-	if w.Code != http.StatusOK {
-		t.Fatalf("expected status 200, got %d", w.Code)
-	}
-
-	t.Run("missing body", func(t *testing.T) {
-		w := createRequest(t, s.ChatHandler, nil)
-		if w.Code != http.StatusBadRequest {
-			t.Errorf("expected status 400, got %d", w.Code)
-		}
-
-		if diff := cmp.Diff(w.Body.String(), `{"error":"model is required"}`); diff != "" {
-			t.Errorf("mismatch (-got +want):\n%s", diff)
-		}
-	})
-
-	t.Run("missing model", func(t *testing.T) {
-		w := createRequest(t, s.ChatHandler, api.ChatRequest{})
-		if w.Code != http.StatusBadRequest {
-			t.Errorf("expected status 400, got %d", w.Code)
-		}
-
-		if diff := cmp.Diff(w.Body.String(), `{"error":"model is required"}`); diff != "" {
-			t.Errorf("mismatch (-got +want):\n%s", diff)
-		}
-	})
-
-	t.Run("missing capabilities chat", func(t *testing.T) {
-		w := createRequest(t, s.CreateModelHandler, api.CreateRequest{
-			Model: "bert",
-			Modelfile: fmt.Sprintf("FROM %s", createBinFile(t, llm.KV{
-				"general.architecture": "bert",
-				"bert.pooling_type":    uint32(0),
-			}, []llm.Tensor{})),
-			Stream: &stream,
-		})
-
-		if w.Code != http.StatusOK {
-			t.Fatalf("expected status 200, got %d", w.Code)
-		}
-
-		w = createRequest(t, s.ChatHandler, api.ChatRequest{
-			Model: "bert",
-		})
-
-		if w.Code != http.StatusBadRequest {
-			t.Errorf("expected status 400, got %d", w.Code)
-		}
-
-		if diff := cmp.Diff(w.Body.String(), `{"error":"\"bert\" does not support chat"}`); diff != "" {
-			t.Errorf("mismatch (-got +want):\n%s", diff)
-		}
-	})
-
-	t.Run("load model", func(t *testing.T) {
-		w := createRequest(t, s.ChatHandler, api.ChatRequest{
-			Model: "test",
-		})
-
-		if w.Code != http.StatusOK {
-			t.Errorf("expected status 200, got %d", w.Code)
-		}
-
-		var actual api.ChatResponse
-		if err := json.NewDecoder(w.Body).Decode(&actual); err != nil {
-			t.Fatal(err)
-		}
-
-		if actual.Model != "test" {
-			t.Errorf("expected model test, got %s", actual.Model)
-		}
-
-		if !actual.Done {
-			t.Errorf("expected done true, got false")
-		}
-
-		if actual.DoneReason != "load" {
-			t.Errorf("expected done reason load, got %s", actual.DoneReason)
-		}
-	})
-
-	checkChatResponse := func(t *testing.T, body io.Reader, model, content string) {
-		t.Helper()
-
-		var actual api.ChatResponse
-		if err := json.NewDecoder(body).Decode(&actual); err != nil {
-			t.Fatal(err)
-		}
-
-		if actual.Model != model {
-			t.Errorf("expected model test, got %s", actual.Model)
-		}
-
-		if !actual.Done {
-			t.Errorf("expected done false, got true")
-		}
-
-		if actual.DoneReason != "stop" {
-			t.Errorf("expected done reason stop, got %s", actual.DoneReason)
-		}
-
-		if diff := cmp.Diff(actual.Message, api.Message{
-			Role:    "assistant",
-			Content: content,
-		}); diff != "" {
-			t.Errorf("mismatch (-got +want):\n%s", diff)
-		}
-
-		if actual.PromptEvalCount == 0 {
-			t.Errorf("expected prompt eval count > 0, got 0")
-		}
-
-		if actual.PromptEvalDuration == 0 {
-			t.Errorf("expected prompt eval duration > 0, got 0")
-		}
-
-		if actual.EvalCount == 0 {
-			t.Errorf("expected eval count > 0, got 0")
-		}
-
-		if actual.EvalDuration == 0 {
-			t.Errorf("expected eval duration > 0, got 0")
-		}
-
-		if actual.LoadDuration == 0 {
-			t.Errorf("expected load duration > 0, got 0")
-		}
-
-		if actual.TotalDuration == 0 {
-			t.Errorf("expected total duration > 0, got 0")
-		}
-	}
-
-	mock.CompletionResponse.Content = "Hi!"
-	t.Run("messages", func(t *testing.T) {
-		w := createRequest(t, s.ChatHandler, api.ChatRequest{
-			Model: "test",
-			Messages: []api.Message{
-				{Role: "user", Content: "Hello!"},
-			},
-			Stream: &stream,
-		})
-
-		if w.Code != http.StatusOK {
-			t.Errorf("expected status 200, got %d", w.Code)
-		}
-
-		if diff := cmp.Diff(mock.CompletionRequest.Prompt, "User: Hello! "); diff != "" {
-			t.Errorf("mismatch (-got +want):\n%s", diff)
-		}
-
-		checkChatResponse(t, w.Body, "test", "Hi!")
-	})
-
-	w = createRequest(t, s.CreateModelHandler, api.CreateRequest{
-		Model:     "test-system",
-		Modelfile: "FROM test\nSYSTEM You are a helpful assistant.",
-	})
-
-	if w.Code != http.StatusOK {
-		t.Fatalf("expected status 200, got %d", w.Code)
-	}
-
-	t.Run("messages with model system", func(t *testing.T) {
-		w := createRequest(t, s.ChatHandler, api.ChatRequest{
-			Model: "test-system",
-			Messages: []api.Message{
-				{Role: "user", Content: "Hello!"},
-			},
-			Stream: &stream,
-		})
-
-		if w.Code != http.StatusOK {
-			t.Errorf("expected status 200, got %d", w.Code)
-		}
-
-		if diff := cmp.Diff(mock.CompletionRequest.Prompt, "System: You are a helpful assistant. User: Hello! "); diff != "" {
-			t.Errorf("mismatch (-got +want):\n%s", diff)
-		}
-
-		checkChatResponse(t, w.Body, "test-system", "Hi!")
-	})
-
-	mock.CompletionResponse.Content = "Abra kadabra!"
-	t.Run("messages with system", func(t *testing.T) {
-		w := createRequest(t, s.ChatHandler, api.ChatRequest{
-			Model: "test-system",
-			Messages: []api.Message{
-				{Role: "system", Content: "You can perform magic tricks."},
-				{Role: "user", Content: "Hello!"},
-			},
-			Stream: &stream,
-		})
-
-		if w.Code != http.StatusOK {
-			t.Errorf("expected status 200, got %d", w.Code)
-		}
-
-		if diff := cmp.Diff(mock.CompletionRequest.Prompt, "System: You can perform magic tricks. User: Hello! "); diff != "" {
-			t.Errorf("mismatch (-got +want):\n%s", diff)
-		}
-
-		checkChatResponse(t, w.Body, "test-system", "Abra kadabra!")
-	})
-
-	t.Run("messages with interleaved system", func(t *testing.T) {
-		w := createRequest(t, s.ChatHandler, api.ChatRequest{
-			Model: "test-system",
-			Messages: []api.Message{
-				{Role: "user", Content: "Hello!"},
-				{Role: "assistant", Content: "I can help you with that."},
-				{Role: "system", Content: "You can perform magic tricks."},
-				{Role: "user", Content: "Help me write tests."},
-			},
-			Stream: &stream,
-		})
-
-		if w.Code != http.StatusOK {
-			t.Errorf("expected status 200, got %d", w.Code)
-		}
-
-		if diff := cmp.Diff(mock.CompletionRequest.Prompt, "System: You are a helpful assistant. User: Hello! Assistant: I can help you with that. System: You can perform magic tricks. User: Help me write tests. "); diff != "" {
-			t.Errorf("mismatch (-got +want):\n%s", diff)
-		}
-
-		checkChatResponse(t, w.Body, "test-system", "Abra kadabra!")
-	})
-}
-
-func TestGenerate(t *testing.T) {
-	gin.SetMode(gin.TestMode)
-
-	mock := mockRunner{
-		CompletionResponse: llm.CompletionResponse{
-			Done:               true,
-			DoneReason:         "stop",
-			PromptEvalCount:    1,
-			PromptEvalDuration: 1,
-			EvalCount:          1,
-			EvalDuration:       1,
-		},
-	}
-
-	s := Server{
-		sched: &Scheduler{
-			pendingReqCh:  make(chan *LlmRequest, 1),
-			finishedReqCh: make(chan *LlmRequest, 1),
-			expiredCh:     make(chan *runnerRef, 1),
-			unloadedCh:    make(chan any, 1),
-			loaded:        make(map[string]*runnerRef),
-			newServerFn:   newMockServer(&mock),
-			getGpuFn:      gpu.GetGPUInfo,
-			getCpuFn:      gpu.GetCPUInfo,
-			reschedDelay:  250 * time.Millisecond,
-			loadFn: func(req *LlmRequest, ggml *llm.GGML, gpus gpu.GpuInfoList, numParallel int) {
-				req.successCh <- &runnerRef{
-					llama: &mock,
-				}
-			},
-		},
-	}
-
-	go s.sched.Run(context.TODO())
-
-	w := createRequest(t, s.CreateModelHandler, api.CreateRequest{
-		Model: "test",
-		Modelfile: fmt.Sprintf(`FROM %s
-		TEMPLATE """
-{{- if .System }}System: {{ .System }} {{ end }}
-{{- if .Prompt }}User: {{ .Prompt }} {{ end }}
-{{- if .Response }}Assistant: {{ .Response }} {{ end }}"""
-`, createBinFile(t, llm.KV{
-			"general.architecture":          "llama",
-			"llama.block_count":             uint32(1),
-			"llama.context_length":          uint32(8192),
-			"llama.embedding_length":        uint32(4096),
-			"llama.attention.head_count":    uint32(32),
-			"llama.attention.head_count_kv": uint32(8),
-			"tokenizer.ggml.tokens":         []string{""},
-			"tokenizer.ggml.scores":         []float32{0},
-			"tokenizer.ggml.token_type":     []int32{0},
-		}, []llm.Tensor{
-			{Name: "token_embd.weight", Shape: []uint64{1}, WriterTo: bytes.NewReader(make([]byte, 4))},
-			{Name: "blk.0.attn_norm.weight", Shape: []uint64{1}, WriterTo: bytes.NewReader(make([]byte, 4))},
-			{Name: "blk.0.ffn_down.weight", Shape: []uint64{1}, WriterTo: bytes.NewReader(make([]byte, 4))},
-			{Name: "blk.0.ffn_gate.weight", Shape: []uint64{1}, WriterTo: bytes.NewReader(make([]byte, 4))},
-			{Name: "blk.0.ffn_up.weight", Shape: []uint64{1}, WriterTo: bytes.NewReader(make([]byte, 4))},
-			{Name: "blk.0.ffn_norm.weight", Shape: []uint64{1}, WriterTo: bytes.NewReader(make([]byte, 4))},
-			{Name: "blk.0.attn_k.weight", Shape: []uint64{1}, WriterTo: bytes.NewReader(make([]byte, 4))},
-			{Name: "blk.0.attn_output.weight", Shape: []uint64{1}, WriterTo: bytes.NewReader(make([]byte, 4))},
-			{Name: "blk.0.attn_q.weight", Shape: []uint64{1}, WriterTo: bytes.NewReader(make([]byte, 4))},
-			{Name: "blk.0.attn_v.weight", Shape: []uint64{1}, WriterTo: bytes.NewReader(make([]byte, 4))},
-			{Name: "output.weight", Shape: []uint64{1}, WriterTo: bytes.NewReader(make([]byte, 4))},
-		})),
-		Stream: &stream,
-	})
-
-	if w.Code != http.StatusOK {
-		t.Fatalf("expected status 200, got %d", w.Code)
-	}
-
-	t.Run("missing body", func(t *testing.T) {
-		w := createRequest(t, s.GenerateHandler, nil)
-		if w.Code != http.StatusBadRequest {
-			t.Errorf("expected status 400, got %d", w.Code)
-		}
-
-		if diff := cmp.Diff(w.Body.String(), `{"error":"model is required"}`); diff != "" {
-			t.Errorf("mismatch (-got +want):\n%s", diff)
-		}
-	})
-
-	t.Run("missing model", func(t *testing.T) {
-		w := createRequest(t, s.GenerateHandler, api.GenerateRequest{})
-		if w.Code != http.StatusBadRequest {
-			t.Errorf("expected status 400, got %d", w.Code)
-		}
-
-		if diff := cmp.Diff(w.Body.String(), `{"error":"model is required"}`); diff != "" {
-			t.Errorf("mismatch (-got +want):\n%s", diff)
-		}
-	})
-
-	t.Run("missing capabilities generate", func(t *testing.T) {
-		w := createRequest(t, s.CreateModelHandler, api.CreateRequest{
-			Model: "bert",
-			Modelfile: fmt.Sprintf("FROM %s", createBinFile(t, llm.KV{
-				"general.architecture": "bert",
-				"bert.pooling_type":    uint32(0),
-			}, []llm.Tensor{})),
-			Stream: &stream,
-		})
-
-		if w.Code != http.StatusOK {
-			t.Fatalf("expected status 200, got %d", w.Code)
-		}
-
-		w = createRequest(t, s.GenerateHandler, api.GenerateRequest{
-			Model: "bert",
-		})
-
-		if w.Code != http.StatusBadRequest {
-			t.Errorf("expected status 400, got %d", w.Code)
-		}
-
-		if diff := cmp.Diff(w.Body.String(), `{"error":"\"bert\" does not support generate"}`); diff != "" {
-			t.Errorf("mismatch (-got +want):\n%s", diff)
-		}
-	})
-
-	t.Run("missing capabilities suffix", func(t *testing.T) {
-		w := createRequest(t, s.GenerateHandler, api.GenerateRequest{
-			Model:  "test",
-			Prompt: "def add(",
-			Suffix: "    return c",
-		})
-
-		if w.Code != http.StatusBadRequest {
-			t.Errorf("expected status 400, got %d", w.Code)
-		}
-
-		if diff := cmp.Diff(w.Body.String(), `{"error":"test does not support insert"}`); diff != "" {
-			t.Errorf("mismatch (-got +want):\n%s", diff)
-		}
-	})
-
-	t.Run("load model", func(t *testing.T) {
-		w := createRequest(t, s.GenerateHandler, api.GenerateRequest{
-			Model: "test",
-		})
-
-		if w.Code != http.StatusOK {
-			t.Errorf("expected status 200, got %d", w.Code)
-		}
-
-		var actual api.GenerateResponse
-		if err := json.NewDecoder(w.Body).Decode(&actual); err != nil {
-			t.Fatal(err)
-		}
-
-		if actual.Model != "test" {
-			t.Errorf("expected model test, got %s", actual.Model)
-		}
-
-		if !actual.Done {
-			t.Errorf("expected done true, got false")
-		}
-
-		if actual.DoneReason != "load" {
-			t.Errorf("expected done reason load, got %s", actual.DoneReason)
-		}
-	})
-
-	checkGenerateResponse := func(t *testing.T, body io.Reader, model, content string) {
-		t.Helper()
-
-		var actual api.GenerateResponse
-		if err := json.NewDecoder(body).Decode(&actual); err != nil {
-			t.Fatal(err)
-		}
-
-		if actual.Model != model {
-			t.Errorf("expected model test, got %s", actual.Model)
-		}
-
-		if !actual.Done {
-			t.Errorf("expected done false, got true")
-		}
-
-		if actual.DoneReason != "stop" {
-			t.Errorf("expected done reason stop, got %s", actual.DoneReason)
-		}
-
-		if actual.Response != content {
-			t.Errorf("expected response %s, got %s", content, actual.Response)
-		}
-
-		if actual.Context == nil {
-			t.Errorf("expected context not nil")
-		}
-
-		if actual.PromptEvalCount == 0 {
-			t.Errorf("expected prompt eval count > 0, got 0")
-		}
-
-		if actual.PromptEvalDuration == 0 {
-			t.Errorf("expected prompt eval duration > 0, got 0")
-		}
-
-		if actual.EvalCount == 0 {
-			t.Errorf("expected eval count > 0, got 0")
-		}
-
-		if actual.EvalDuration == 0 {
-			t.Errorf("expected eval duration > 0, got 0")
-		}
-
-		if actual.LoadDuration == 0 {
-			t.Errorf("expected load duration > 0, got 0")
-		}
-
-		if actual.TotalDuration == 0 {
-			t.Errorf("expected total duration > 0, got 0")
-		}
-	}
-
-	mock.CompletionResponse.Content = "Hi!"
-	t.Run("prompt", func(t *testing.T) {
-		w := createRequest(t, s.GenerateHandler, api.GenerateRequest{
-			Model:  "test",
-			Prompt: "Hello!",
-			Stream: &stream,
-		})
-
-		if w.Code != http.StatusOK {
-			t.Errorf("expected status 200, got %d", w.Code)
-		}
-
-		if diff := cmp.Diff(mock.CompletionRequest.Prompt, "User: Hello! "); diff != "" {
-			t.Errorf("mismatch (-got +want):\n%s", diff)
-		}
-
-		checkGenerateResponse(t, w.Body, "test", "Hi!")
-	})
-
-	w = createRequest(t, s.CreateModelHandler, api.CreateRequest{
-		Model:     "test-system",
-		Modelfile: "FROM test\nSYSTEM You are a helpful assistant.",
-	})
-
-	if w.Code != http.StatusOK {
-		t.Fatalf("expected status 200, got %d", w.Code)
-	}
-
-	t.Run("prompt with model system", func(t *testing.T) {
-		w := createRequest(t, s.GenerateHandler, api.GenerateRequest{
-			Model:  "test-system",
-			Prompt: "Hello!",
-			Stream: &stream,
-		})
-
-		if w.Code != http.StatusOK {
-			t.Errorf("expected status 200, got %d", w.Code)
-		}
-
-		if diff := cmp.Diff(mock.CompletionRequest.Prompt, "System: You are a helpful assistant. User: Hello! "); diff != "" {
-			t.Errorf("mismatch (-got +want):\n%s", diff)
-		}
-
-		checkGenerateResponse(t, w.Body, "test-system", "Hi!")
-	})
-
-	mock.CompletionResponse.Content = "Abra kadabra!"
-	t.Run("prompt with system", func(t *testing.T) {
-		w := createRequest(t, s.GenerateHandler, api.GenerateRequest{
-			Model:  "test-system",
-			Prompt: "Hello!",
-			System: "You can perform magic tricks.",
-			Stream: &stream,
-		})
-
-		if w.Code != http.StatusOK {
-			t.Errorf("expected status 200, got %d", w.Code)
-		}
-
-		if diff := cmp.Diff(mock.CompletionRequest.Prompt, "System: You can perform magic tricks. User: Hello! "); diff != "" {
-			t.Errorf("mismatch (-got +want):\n%s", diff)
-		}
-
-		checkGenerateResponse(t, w.Body, "test-system", "Abra kadabra!")
-	})
-
-	t.Run("prompt with template", func(t *testing.T) {
-		w := createRequest(t, s.GenerateHandler, api.GenerateRequest{
-			Model:  "test-system",
-			Prompt: "Help me write tests.",
-			System: "You can perform magic tricks.",
-			Template: `{{- if .System }}{{ .System }} {{ end }}
-{{- if .Prompt }}### USER {{ .Prompt }} {{ end }}
-{{- if .Response }}### ASSISTANT {{ .Response }} {{ end }}`,
-			Stream: &stream,
-		})
-
-		if w.Code != http.StatusOK {
-			t.Errorf("expected status 200, got %d", w.Code)
-		}
-
-		if diff := cmp.Diff(mock.CompletionRequest.Prompt, "You can perform magic tricks. ### USER Help me write tests. "); diff != "" {
-			t.Errorf("mismatch (-got +want):\n%s", diff)
-		}
-
-		checkGenerateResponse(t, w.Body, "test-system", "Abra kadabra!")
-	})
-
-	w = createRequest(t, s.CreateModelHandler, api.CreateRequest{
-		Model: "test-suffix",
-		Modelfile: `FROM test
-TEMPLATE """{{- if .Suffix }}<PRE> {{ .Prompt }} <SUF>{{ .Suffix }} <MID>
-{{- else }}{{ .Prompt }}
-{{- end }}"""`,
-	})
-
-	if w.Code != http.StatusOK {
-		t.Fatalf("expected status 200, got %d", w.Code)
-	}
-
-	t.Run("prompt with suffix", func(t *testing.T) {
-		w := createRequest(t, s.GenerateHandler, api.GenerateRequest{
-			Model:  "test-suffix",
-			Prompt: "def add(",
-			Suffix: "    return c",
-		})
-
-		if w.Code != http.StatusOK {
-			t.Errorf("expected status 200, got %d", w.Code)
-		}
-
-		if diff := cmp.Diff(mock.CompletionRequest.Prompt, "<PRE> def add( <SUF>    return c <MID>"); diff != "" {
-			t.Errorf("mismatch (-got +want):\n%s", diff)
-		}
-	})
-
-	t.Run("prompt without suffix", func(t *testing.T) {
-		w := createRequest(t, s.GenerateHandler, api.GenerateRequest{
-			Model:  "test-suffix",
-			Prompt: "def add(",
-		})
-
-		if w.Code != http.StatusOK {
-			t.Errorf("expected status 200, got %d", w.Code)
-		}
-
-		if diff := cmp.Diff(mock.CompletionRequest.Prompt, "def add("); diff != "" {
-			t.Errorf("mismatch (-got +want):\n%s", diff)
-		}
-	})
-
-	t.Run("raw", func(t *testing.T) {
-		w := createRequest(t, s.GenerateHandler, api.GenerateRequest{
-			Model:  "test-system",
-			Prompt: "Help me write tests.",
-			Raw:    true,
-			Stream: &stream,
-		})
-
-		if w.Code != http.StatusOK {
-			t.Errorf("expected status 200, got %d", w.Code)
-		}
-
-		if diff := cmp.Diff(mock.CompletionRequest.Prompt, "Help me write tests."); diff != "" {
-			t.Errorf("mismatch (-got +want):\n%s", diff)
-		}
-	})
-}

server/routes_list_test.go

@@ -7,14 +7,11 @@ import (
 	"slices"
 	"testing"
 
-	"github.com/gin-gonic/gin"
 	"github.com/ollama/ollama/api"
 	"github.com/ollama/ollama/envconfig"
 )
 
 func TestList(t *testing.T) {
-	gin.SetMode(gin.TestMode)
-
 	t.Setenv("OLLAMA_MODELS", t.TempDir())
 	envconfig.LoadConfig()
 

server/routes_test.go

@@ -306,12 +306,8 @@ func Test_Routes(t *testing.T) {
 					t.Fatalf("expected model t-bone, got %s", embedResp.Model)
 				}
 
-				if embedResp.Embeddings == nil {
-					t.Fatalf("expected embeddings to not be nil, got %v", embedResp.Embeddings)
-				}
-
-				if len(embedResp.Embeddings) != 0 {
-					t.Fatalf("expected embeddings to be empty, got %v", embedResp.Embeddings)
+				if embedResp.Embeddings != nil {
+					t.Fatalf("expected embeddings to be nil, got %v", embedResp.Embeddings)
 				}
 			},
 		},

server/sched.go

@@ -135,6 +135,11 @@ func (s *Scheduler) processPending(ctx context.Context) {
 			}
 
 			for {
+				cpus := s.getCpuFn()
+				var systemMem gpu.GpuInfo
+				if len(cpus) > 0 {
+					systemMem = cpus[0]
+				}
 				var runnerToExpire *runnerRef
 				s.loadedMu.Lock()
 				runner := s.loaded[pending.model.ModelPath]
@@ -188,6 +193,38 @@ func (s *Scheduler) processPending(ctx context.Context) {
 						break
 					}
 
+					estimate := llm.EstimateGPULayers(gpus, ggml, pending.model.ProjectorPaths, pending.opts)
+					maxSize := systemMem.FreeMemory
+
+					// Add available GPU memory to the total pool
+					// macOS hardware has unified memory so don't double count
+					if runtime.GOOS != "darwin" {
+						for _, gpu := range gpus {
+							if gpu.Library == "cpu" {
+								continue
+							}
+							if loadedCount == 0 {
+								// If no other models are loaded, set the limit based on what's available
+								maxSize += gpu.FreeMemory
+							} else {
+								// Other models could be unloaded, favor total memory for limit
+								maxSize += gpu.TotalMemory
+							}
+						}
+					}
+
+					// Block attempting to load a model larger than system memory + GPU memory
+					if estimate.TotalSize > maxSize {
+						slog.Warn("model request too large for system", "requested", format.HumanBytes2(estimate.TotalSize), "system", format.HumanBytes2(maxSize))
+
+						// Linux will crash if over-allocating memory - return an error to the user.
+						// TODO (jmorganca): add reasonable upper limits for darwin and windows as well
+						if runtime.GOOS == "linux" {
+							pending.errCh <- fmt.Errorf("requested model (%s) is too large for this system (%s)", format.HumanBytes2(estimate.TotalSize), format.HumanBytes2(maxSize))
+							break
+						}
+					}
+
 					// Evaluate if the model will fit in the available system memory, or if we should unload a model first
 					if len(gpus) == 1 && gpus[0].Library == "cpu" {
 						// simplifying assumption of defaultParallel when in CPU mode

server/testdata/tools/command-r-plus.gotmpl

@@ -1,67 +0,0 @@
-{{- if or .Tools .System }}<|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|>
-{{- if .Tools }}# Safety Preamble
-The instructions in this section override those in the task description and style guide sections. Don't answer questions that are harmful or immoral.
-
-# System Preamble
-## Basic Rules
-You are a powerful conversational AI trained by Cohere to help people. You are augmented by a number of tools, and your job is to use and consume the output of these tools to best help the user. You will see a conversation history between yourself and a user, ending with an utterance from the user. You will then see a specific instruction instructing you what kind of response to generate. When you answer the user's requests, you cite your sources in your answers, according to those instructions.
-
-{{ if .System }}# User Preamble
-{{ .System }}
-{{- end }}
-
-## Available Tools
-Here is a list of tools that you have available to you:
-{{- range .Tools }}
-
-```python
-def {{ .Function.Name }}(
-{{- range $name, $property := .Function.Parameters.Properties }}{{ $name }}: {{ $property.Type }}, {{ end }}) -> List[Dict]:
-    """{{ .Function.Description }}
-
-{{- if .Function.Parameters.Properties }}
-
-    Args:
-{{- range $name, $property := .Function.Parameters.Properties }}
-        {{ $name }} ({{ $property.Type }}): {{ $property.Description }}
-{{- end }}
-{{- end }}
-    """
-    pass
-```
-{{- end }}
-{{- else if .System }}{{ .System }}
-{{- end }}<|END_OF_TURN_TOKEN|>
-{{- end }}
-{{- range .Messages }}
-{{- if eq .Role "system" }}
-{{- continue }}
-{{- end }}<|START_OF_TURN_TOKEN|>
-{{- if eq .Role "user" }}<|USER_TOKEN|>{{ .Content }}
-{{- else if eq .Role "assistant" }}<|CHATBOT_TOKEN|>
-{{- if .Content }}{{ .Content }}
-{{- else if .ToolCalls }}
-Action: ```json
-[
-{{- range .ToolCalls }}
-    {
-        "tool_name": "{{ .Function.Name }}",
-        "parameters": {{ json .Function.Arguments }}
-    }
-{{- end }}
-]```
-{{ continue }}
-{{ end }}
-{{- else if eq .Role "tool" }}<|SYSTEM_TOKEN|><results>
-{{ .Content }}</results>
-{{- end }}<|END_OF_TURN_TOKEN|>
-{{- end }}
-{{- if .Tools }}<|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|>Write 'Action:' followed by a json-formatted list of actions that you want to perform in order to produce a good response to the user's last input. You can use any of the supplied tools any number of times, but you should aim to execute the minimum number of necessary actions for the input. You should use the `directly-answer` tool if calling the other tools is unnecessary. The list of actions you want to call should be formatted as a list of json objects, for example:
-```json
-[
-    {
-        "tool_name": title of the tool in the specification,
-        "parameters": a dict of parameters to input into the tool as they are defined in the specs, or {} if it takes no parameters
-    }
-]```
-{{- end }}<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>

server/testdata/tools/command-r-plus.out

@@ -1,39 +0,0 @@
-<|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|># Safety Preamble
-The instructions in this section override those in the task description and style guide sections. Don't answer questions that are harmful or immoral.
-
-# System Preamble
-## Basic Rules
-You are a powerful conversational AI trained by Cohere to help people. You are augmented by a number of tools, and your job is to use and consume the output of these tools to best help the user. You will see a conversation history between yourself and a user, ending with an utterance from the user. You will then see a specific instruction instructing you what kind of response to generate. When you answer the user's requests, you cite your sources in your answers, according to those instructions.
-
-# User Preamble
-You are a knowledgable assistant. You can answer questions and perform tasks.
-
-## Available Tools
-Here is a list of tools that you have available to you:
-
-```python
-def get_current_weather(format: string, location: string, ) -> List[Dict]:
-    """Get the current weather
-
-    Args:
-        format (string): The temperature unit to use. Infer this from the users location.
-        location (string): The city and state, e.g. San Francisco, CA
-    """
-    pass
-```<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|USER_TOKEN|>What's the weather like today in Paris?<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>
-Action: ```json
-[
-    {
-        "tool_name": "get_current_weather",
-        "parameters": {"format":"celsius","location":"Paris, France"}
-    }
-]```
-<|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|><results>
-22</results><|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>The current temperature in Paris, France is 22 degrees Celsius.<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|USER_TOKEN|>What's the weather like today in San Francisco and Toronto?<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|>Write 'Action:' followed by a json-formatted list of actions that you want to perform in order to produce a good response to the user's last input. You can use any of the supplied tools any number of times, but you should aim to execute the minimum number of necessary actions for the input. You should use the `directly-answer` tool if calling the other tools is unnecessary. The list of actions you want to call should be formatted as a list of json objects, for example:
-```json
-[
-    {
-        "tool_name": title of the tool in the specification,
-        "parameters": a dict of parameters to input into the tool as they are defined in the specs, or {} if it takes no parameters
-    }
-]```<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>

server/testdata/tools/firefunction.gotmpl

@@ -1,31 +0,0 @@
-{{- if or .System .Tools }}<|start_header_id|>system<|end_header_id|>
-{{- if .System }}
-{{ .System }}
-{{- end }}
-In addition to plain text responses, you can chose to call one or more of the provided functions.
-
-Use the following rule to decide when to call a function:
-  * if the response can be generated from your internal knowledge (e.g., as in the case of queries like "What is the capital of Poland?"), do so
-  * if you need external information that can be obtained by calling one or more of the provided functions, generate a function calls
-
-If you decide to call functions:
-  * prefix function calls with functools marker (no closing marker required)
-  * all function calls should be generated in a single JSON list formatted as functools[{"name": [function name], "arguments": [function arguments as JSON]}, ...]
-  * follow the provided JSON schema. Do not hallucinate arguments or values. Do to blindly copy values from the provided samples
-  * respect the argument type formatting. E.g., if the type if number and format is float, write value 7 as 7.0
-  * make sure you pick the right functions that match the user intent
-
-Available functions as JSON spec:
-{{- if .Tools }}
-{{ json .Tools }}
-{{- end }}<|eot_id|>
-{{- end }}
-{{- range .Messages }}<|start_header_id|>
-{{- if or (eq .Role "user") (eq .Role "assistant") (eq .Role "tool") }}{{ .Role }}
-{{- end }}<|end_header_id|>
-{{- if .Content }}{{ .Content }}
-{{- else if .ToolCalls }} functools[
-{{- range .ToolCalls }}{{ "{" }}"name": "{{ .Function.Name }}", "arguments": {{ json .Function.Arguments }}{{ "}" }}
-{{- end }}]
-{{- end }}<|eot_id|>
-{{- end }}<|start_header_id|>assistant<|end_header_id|>

server/testdata/tools/firefunction.out

@@ -1,17 +0,0 @@
-<|start_header_id|>system<|end_header_id|>
-You are a knowledgable assistant. You can answer questions and perform tasks.
-In addition to plain text responses, you can chose to call one or more of the provided functions.
-
-Use the following rule to decide when to call a function:
-  * if the response can be generated from your internal knowledge (e.g., as in the case of queries like "What is the capital of Poland?"), do so
-  * if you need external information that can be obtained by calling one or more of the provided functions, generate a function calls
-
-If you decide to call functions:
-  * prefix function calls with functools marker (no closing marker required)
-  * all function calls should be generated in a single JSON list formatted as functools[{"name": [function name], "arguments": [function arguments as JSON]}, ...]
-  * follow the provided JSON schema. Do not hallucinate arguments or values. Do to blindly copy values from the provided samples
-  * respect the argument type formatting. E.g., if the type if number and format is float, write value 7 as 7.0
-  * make sure you pick the right functions that match the user intent
-
-Available functions as JSON spec:
-[{"type":"function","function":{"name":"get_current_weather","description":"Get the current weather","parameters":{"type":"object","required":["location","format"],"properties":{"format":{"type":"string","description":"The temperature unit to use. Infer this from the users location.","enum":["celsius","fahrenheit"]},"location":{"type":"string","description":"The city and state, e.g. San Francisco, CA"}}}}}]<|eot_id|><|start_header_id|><|end_header_id|>You are a knowledgable assistant. You can answer questions and perform tasks.<|eot_id|><|start_header_id|>user<|end_header_id|>What's the weather like today in Paris?<|eot_id|><|start_header_id|>assistant<|end_header_id|> functools[{"name": "get_current_weather", "arguments": {"format":"celsius","location":"Paris, France"}}]<|eot_id|><|start_header_id|>tool<|end_header_id|>22<|eot_id|><|start_header_id|>assistant<|end_header_id|>The current temperature in Paris, France is 22 degrees Celsius.<|eot_id|><|start_header_id|>user<|end_header_id|>What's the weather like today in San Francisco and Toronto?<|eot_id|><|start_header_id|>assistant<|end_header_id|>

server/testdata/tools/messages.json

@@ -1,39 +0,0 @@
-[
-  {
-    "role": "system",
-    "content": "You are a knowledgable assistant. You can answer questions and perform tasks."
-  },
-  {
-    "role": "user",
-    "content": "What's the weather like today in Paris?"
-  },
-  {
-    "role": "assistant",
-    "tool_calls": [
-      {
-        "id": "89a1e453-0bce-4de3-a456-c54bed09c520",
-        "type": "function",
-        "function": {
-          "name": "get_current_weather",
-          "arguments": {
-            "location": "Paris, France",
-            "format": "celsius"
-          }
-        }
-      }
-    ]
-  },
-  {
-    "role": "tool",
-    "tool_call_id": "89a1e453-0bce-4de3-a456-c54bed09c520",
-    "content": "22"
-  },
-  {
-    "role": "assistant",
-    "content": "The current temperature in Paris, France is 22 degrees Celsius."
-  },
-  {
-    "role": "user",
-    "content": "What's the weather like today in San Francisco and Toronto?"
-  }
-]

server/testdata/tools/mistral.gotmpl

@@ -1,15 +0,0 @@
-{{- range $index, $_ := .Messages }}
-{{- if eq .Role "user" }}
-{{- if and (eq (len (slice $.Messages $index)) 1) $.Tools }}[AVAILABLE_TOOLS] {{ json $.Tools }}[/AVAILABLE_TOOLS]
-{{- end }}[INST] {{ if and (eq (len (slice $.Messages $index)) 1) $.System }}{{ $.System }}
-
-{{ end }}{{ .Content }}[/INST]
-{{- else if eq .Role "assistant" }}
-{{- if .Content }} {{ .Content }}</s>
-{{- else if .ToolCalls }}[TOOL_CALLS] [
-{{- range .ToolCalls }}{"name": "{{ .Function.Name }}", "arguments": {{ json .Function.Arguments }}}
-{{- end }}]</s>
-{{- end }}
-{{- else if eq .Role "tool" }}[TOOL_RESULTS] {"content": {{ .Content }}}[/TOOL_RESULTS]
-{{- end }}
-{{- end }}

server/testdata/tools/mistral.out

@@ -1,3 +0,0 @@
-[INST] What's the weather like today in Paris?[/INST][TOOL_CALLS] [{"name": "get_current_weather", "arguments": {"format":"celsius","location":"Paris, France"}}]</s>[TOOL_RESULTS] {"content": 22}[/TOOL_RESULTS] The current temperature in Paris, France is 22 degrees Celsius.</s>[AVAILABLE_TOOLS] [{"type":"function","function":{"name":"get_current_weather","description":"Get the current weather","parameters":{"type":"object","required":["location","format"],"properties":{"format":{"type":"string","description":"The temperature unit to use. Infer this from the users location.","enum":["celsius","fahrenheit"]},"location":{"type":"string","description":"The city and state, e.g. San Francisco, CA"}}}}}][/AVAILABLE_TOOLS][INST] You are a knowledgable assistant. You can answer questions and perform tasks.
-
-What's the weather like today in San Francisco and Toronto?[/INST]

server/testdata/tools/tools.json

@@ -1,30 +0,0 @@
-[
-  {
-    "type": "function",
-    "function": {
-      "name": "get_current_weather",
-      "description": "Get the current weather",
-      "parameters": {
-        "type": "object",
-        "properties": {
-          "location": {
-            "type": "string",
-            "description": "The city and state, e.g. San Francisco, CA"
-          },
-          "format": {
-            "type": "string",
-            "enum": [
-              "celsius",
-              "fahrenheit"
-            ],
-            "description": "The temperature unit to use. Infer this from the users location."
-          }
-        },
-        "required": [
-          "location",
-          "format"
-        ]
-      }
-    }
-  }
-]

template/alfred.gotmpl

@@ -1 +1,8 @@
-{{ if .System }}<start_system>{{ .System }}<end_message>{{ end }}{{ if .Prompt }}<start_user>{{ .Prompt }}<end_message>{{ end }}<start_assistant>{{ .Response }}<end_message>
+{{- if .Messages }}
+{{- if .System }}<start_system>{{ .System }}<end_message>
+{{- end }}
+{{- range .Messages }}<start_{{ .Role }}>{{ .Content }}<end_message>
+{{- end }}<start_assistant>
+{{- else }}
+{{ if .System }}<start_system>{{ .System }}<end_message>{{ end }}{{ if .Prompt }}<start_user>{{ .Prompt }}<end_message>{{ end }}<start_assistant>{{ .Response }}<end_message>
+{{- end }}

template/alpaca.gotmpl

@@ -1,3 +1,14 @@
+{{- if .Messages }}
+{{- if .System }}{{ .System }}
+{{- end }}
+{{- range .Messages }}
+{{- if eq .Role "user" }}### Instruction:
+{{- else if eq .Role "assistant" }}### Response:
+{{- end }}
+{{ .Content }}
+
+{{ end }}### Response:
+{{ else }}
 {{ if .System }}{{ .System }}
 
 {{ end }}{{ if .Prompt }}### Instruction:
@@ -5,4 +16,4 @@
 
 {{ end }}### Response:
 {{ .Response }}
-
+{{- end }}

template/chatml.gotmpl

@@ -1,6 +1,15 @@
+{{- if .Messages }}
+{{- if .System }}<|im_start|>system
+{{ .System }}<|im_end|>
+{{ end }}
+{{- range .Messages }}<|im_start|>{{ .Role }}
+{{ .Content }}<|im_end|>
+{{ end }}<|im_start|>assistant
+{{ else }}
 {{ if .System }}<|im_start|>system
 {{ .System }}<|im_end|>
 {{ end }}{{ if .Prompt }}<|im_start|>user
 {{ .Prompt }}<|im_end|>
 {{ end }}<|im_start|>assistant
 {{ .Response }}<|im_end|>
+{{- end }}

template/chatqa.gotmpl

@@ -1,6 +1,17 @@
+{{- if .Messages }}
+{{- if .System }}System: {{ .System }}
+
+{{ end }}
+{{- range .Messages }}
+{{- if eq .Role "user" }}User:
+{{- else if eq .Role "assistant" }}Assistant:
+{{- end }} {{ .Content }}
+
+{{ end }}Assistant:
+{{- else }}
 {{ if .System }}System: {{ .System }}
 
 {{ end }}{{ if .Prompt }}User: {{ .Prompt }}
 
-{{ end }}Assistant: {{ .Response }}
-
+{{ end }}Assistant: <|begin_of_text|>{{ .Response }}
+{{- end }}

template/codellama-70b-instruct.gotmpl

@@ -1,10 +1,19 @@
-{{ if .System }}Source: system
+{{- if .Messages }}
+{{- if .System }}Source: system
 
- {{ .System }} <step> {{ end }}Source: user
+ {{ .System }} <step> {{ end }}
+{{- range .Messages }}Source: {{ .Role }}
+
+ {{ .Content }} <step> {{ end }}Source: assistant
+Destination: user
+
+{{ else }}
+{{ if .System }} Source: system
+
+ {{ .System }} <step>{{ end }} Source: user
 
  {{ .Prompt }} <step> Source: assistant
-{{- if not .Response }}
 Destination: user
-{{- end }}
 
- {{ .Response }} <step> 
+ {{ .Response }}<step>
+{{- end }}

template/falcon-instruct.gotmpl

@@ -1,5 +1,13 @@
-{{ if .System }}System: {{ .System }}
-{{ end }}{{ if .Prompt }}User:
-{{ .Prompt }}
+{{- if .Messages }}
+{{- if .System }}System: {{ .System }}
+{{ end }}
+{{- range .Messages }}
+{{- if eq .Role "user" }}User:
+{{ else if eq .Role "assistant" }}Falcon:
+{{ end }}{{ .Content }}
 {{ end }}Falcon:
-{{ .Response }}
+{{ else }}
+{{ if .System }}{{ .System }}
+{{ end }}{{ if .Prompt }}User: {{ .Prompt }}
+{{ end }}Assistant: {{ .Response }}
+{{- end }}

template/gemma-instruct.gotmpl

@@ -1,5 +1,16 @@
+{{- if .Messages }}
+{{- range $index, $_ := .Messages }}<start_of_turn>
+{{- if eq .Role "user" }}user
+{{- if and $.System (eq $index 0) }}
+{{ $.System }}
+{{- end }}
+{{- else if eq .Role "assistant" }}model
+{{- end }}
+{{ .Content }}<end_of_turn>
+{{ end }}<start_of_turn>model
+{{ else }}
 <start_of_turn>user
-{{ if .System }}{{ .System }}
-{{ end }}{{ .Prompt }}<end_of_turn>
+{{ if .System }}{{ .System }} {{ end }}{{ .Prompt }}<end_of_turn>
 <start_of_turn>model
 {{ .Response }}<end_of_turn>
+{{- end }}

template/granite-instruct.gotmpl

@@ -1,4 +1,18 @@
-{{ if .System }}System:
+{{- if .Messages }}
+{{- if .System }}System:
+{{ .System }}
+
+{{ end }}
+{{- range .Messages }}
+{{- if eq .Role "user" }}Question:
+{{- else if eq .Role "assistant" }}Answer:
+{{- end }}
+{{ .Content }}
+
+{{ end }}Answer:
+{{ else }}
+{{ if .System }}
+System:
 {{ .System }}
 
 {{ end }}{{ if .Prompt }}Question:
@@ -6,4 +20,4 @@
 
 {{ end }}Answer:
 {{ .Response }}
-
+{{- end }}

template/llama2-chat.gotmpl

@@ -1,6 +1,16 @@
-[INST] <<SYS>>
-{{- if .System }}
-{{ .System }}
+{{- if .Messages }}
+{{- range $index, $_ := .Messages }}
+{{- if eq .Role "user" }}[INST] {{ if eq $index 0 }}<<SYS>>
+{{- if $.System }}
+{{ $.System }}
 {{ end }}<</SYS>>
 
-{{ .Prompt }} [/INST] {{ .Response }}</s><s>
+{{ end }}{{ .Content }}
+{{- else }} [/INST] {{ .Content }}</s><s>
+{{- end }}
+{{- end }} [/INST]
+{{- else }}
+[INST] <<SYS>>{{ .System }}<</SYS>>
+
+{{ .Prompt }} [/INST] {{ .Response }}
+{{- end }}

template/llama3-instruct.gotmpl

@@ -1,7 +1,19 @@
+{{- if .Messages }}
+{{- if .System }}<|start_header_id|>system<|end_header_id|>
+
+{{ .System }}<|eot_id|>
+{{- end }}
+{{- range .Messages }}<|start_header_id|>{{ .Role }}<|end_header_id|>
+
+{{ .Content }}<|eot_id|>
+{{- end }}<|start_header_id|>assistant<|end_header_id|>
+
+{{ else }}
 {{ if .System }}<|start_header_id|>system<|end_header_id|>
 
 {{ .System }}<|eot_id|>{{ end }}{{ if .Prompt }}<|start_header_id|>user<|end_header_id|>
 
 {{ .Prompt }}<|eot_id|>{{ end }}<|start_header_id|>assistant<|end_header_id|>
 
-{{ .Response }}<|eot_id|>
+{{ .Response }}<|eot_id|>
+{{- end }}

template/magicoder.gotmpl

@@ -1,3 +1,15 @@
+{{- if .Messages }}
+{{- if .System }}{{ .System }}
+
+{{ end }}
+{{- range .Messages }}
+{{- if eq .Role "user" }}@@ Instruction
+{{- else if eq .Role "assistant" }}@@ Response
+{{- end }}
+{{ .Content }}
+
+{{ end }}@@ Response
+{{ else }}
 {{ if .System }}{{ .System }}
 
 {{ end }}{{ if .Prompt }}@@ Instruction
@@ -5,4 +17,4 @@
 
 {{ end }}@@ Response
 {{ .Response }}
-
+{{- end }}

template/mistral-instruct.gotmpl

@@ -1,3 +1,9 @@
-[INST] {{ if .System }}{{ .System }}
-
-{{ end }}{{ .Prompt }}[/INST] {{ .Response }}</s>
+{{- if .Messages }}
+{{- range $index, $_ := .Messages }}
+{{- if eq .Role "user" }}[INST] {{ if and $.System (eq (len (slice $.Messages $index)) 1) }}{{ $.System }}
+{{ end }}{{ .Content }}
+{{- else if eq .Role "assistant" }}[/INST] {{ .Content }}</s>
+{{- end }}
+{{- end }}[/INST]
+{{- else }}[INST] {{ if .System }}{{ .System }} {{ end }}{{ .Prompt }} [/INST] {{ .Response }}
+{{- end }}

template/openchat.gotmpl

@@ -1 +1,11 @@
-{{ if .System }}GPT4 Correct System: {{ .System }}<|end_of_turn|>{{ end }}GPT4 Correct User: {{ .Prompt }}<|end_of_turn|>GPT4 Correct Assistant: {{ .Response }}<|end_of_turn|>
+{{- if .Messages }}
+{{- if .System }}GPT Correct System: {{ .System }}<|end_of_turn|>
+{{- end }}
+{{- range .Messages }}GPT Correct
+{{- if eq .Role "user" }} User:
+{{- else if eq .Role "assistant" }} Assistant:
+{{- end }} {{ .Content }}<|end_of_turn|>
+{{- end }}GPT Correct Assistant:
+{{- else }}
+{{ .System }}<|end_of_turn|>GPT4 Correct User: {{ .Prompt }}<|end_of_turn|>GPT4 Correct Assistant: {{ .Response }}<|end_of_turn|>
+{{- end }}

template/phi-3.gotmpl

@@ -1,6 +1,15 @@
+{{- if .Messages }}
+{{- if .System }}<|system|>
+{{ .System }}<|end|>
+{{ end }}
+{{- range .Messages }}<|{{ .Role }}|>
+{{ .Content }}<|end|>
+{{ end }}<|assistant|>
+{{ else }}
 {{ if .System }}<|system|>
 {{ .System }}<|end|>
 {{ end }}{{ if .Prompt }}<|user|>
 {{ .Prompt }}<|end|>
 {{ end }}<|assistant|>
 {{ .Response }}<|end|>
+{{- end }}

template/solar-instruct.gotmpl

@@ -1,3 +1,16 @@
+{{- if .Messages }}
+{{- if .System }}### System:
+{{ .System }}
+
+{{ end }}
+{{- range .Messages }}
+{{- if eq .Role "user" }}### User:
+{{ .Content }}
+{{ else if eq .Role "assistant" }}### Assistant:
+{{ .Content }}</s>
+{{ end }}
+{{ end }}### Assistant:
+{{ else }}
 {{ if .System }}### System:
 {{ .System }}
 
@@ -5,5 +18,5 @@
 {{ .Prompt }}
 
 {{ end }}### Assistant:
-{{ .Response }}</s>
-
+{{ .Response }}
+{{- end }}

template/starcoder2-instruct.gotmpl

@@ -1,8 +1,24 @@
+{{- if .Messages }}
+{{- if .System }}{{ .System }}
+
+{{ end }}
+{{- range .Messages }}
+{{- if eq .Role "user" }}### Instruction
+{{ .Content }}
+
+{{ else if eq .Role "assistant" }}### Response
+{{ .Content }}<|endoftext|>
+
+{{ end }}
+{{- end }}### Response
+{{ else }}
 {{ if .System }}{{ .System }}
 
 {{ end }}{{ if .Prompt }}### Instruction
 {{ .Prompt }}
 
+
 {{ end }}### Response
 {{ .Response }}<|endoftext|>
 
+{{- end }}

template/template.go

@@ -102,15 +102,8 @@ var response = parse.ActionNode{
 	},
 }
 
-var funcs = template.FuncMap{
-	"json": func(v any) string {
-		b, _ := json.Marshal(v)
-		return string(b)
-	},
-}
-
 func Parse(s string) (*Template, error) {
-	tmpl := template.New("").Option("missingkey=zero").Funcs(funcs)
+	tmpl := template.New("").Option("missingkey=zero")
 
 	tmpl, err := tmpl.Parse(s)
 	if err != nil {
@@ -134,7 +127,7 @@ func (t *Template) Vars() []string {
 	var vars []string
 	for _, tt := range t.Templates() {
 		for _, n := range tt.Root.Nodes {
-			vars = append(vars, Identifiers(n)...)
+			vars = append(vars, parseNode(n)...)
 		}
 	}
 
@@ -150,126 +143,52 @@ func (t *Template) Vars() []string {
 
 type Values struct {
 	Messages []api.Message
-	Tools    []api.Tool
-	Prompt   string
-	Suffix   string
-
-	// forceLegacy is a flag used to test compatibility with legacy templates
-	forceLegacy bool
-}
-
-func (t *Template) Subtree(fn func(parse.Node) bool) *template.Template {
-	var walk func(parse.Node) parse.Node
-	walk = func(n parse.Node) parse.Node {
-		if fn(n) {
-			return n
-		}
-
-		switch t := n.(type) {
-		case *parse.ListNode:
-			for _, c := range t.Nodes {
-				if n := walk(c); n != nil {
-					return n
-				}
-			}
-		case *parse.BranchNode:
-			for _, n := range []*parse.ListNode{t.List, t.ElseList} {
-				if n != nil {
-					if n := walk(n); n != nil {
-						return n
-					}
-				}
-			}
-		case *parse.IfNode:
-			return walk(&t.BranchNode)
-		case *parse.WithNode:
-			return walk(&t.BranchNode)
-		case *parse.RangeNode:
-			return walk(&t.BranchNode)
-		}
-
-		return nil
-	}
-
-	if n := walk(t.Tree.Root); n != nil {
-		return (&template.Template{
-			Tree: &parse.Tree{
-				Root: &parse.ListNode{
-					Nodes: []parse.Node{n},
-				},
-			},
-		}).Funcs(funcs)
-	}
-
-	return nil
 }
 
 func (t *Template) Execute(w io.Writer, v Values) error {
-	system, messages := collate(v.Messages)
-	if v.Prompt != "" && v.Suffix != "" {
-		return t.Template.Execute(w, map[string]any{
-			"Prompt":   v.Prompt,
-			"Suffix":   v.Suffix,
-			"Response": "",
-		})
-	} else if !v.forceLegacy && slices.Contains(t.Vars(), "messages") {
+	system, collated := collate(v.Messages)
+	if slices.Contains(t.Vars(), "messages") {
 		return t.Template.Execute(w, map[string]any{
 			"System":   system,
-			"Messages": messages,
-			"Tools":    v.Tools,
+			"Messages": collated,
 		})
 	}
 
-	system = ""
 	var b bytes.Buffer
 	var prompt, response string
-	for _, m := range messages {
-		execute := func() error {
+	for i, m := range collated {
+		if m.Role == "user" {
+			prompt = m.Content
+		} else {
+			response = m.Content
+		}
+
+		if i != len(collated)-1 && prompt != "" && response != "" {
 			if err := t.Template.Execute(&b, map[string]any{
-				"System":   system,
+				"System":   "",
 				"Prompt":   prompt,
 				"Response": response,
 			}); err != nil {
 				return err
 			}
 
-			system = ""
 			prompt = ""
 			response = ""
-			return nil
-		}
-
-		switch m.Role {
-		case "system":
-			if prompt != "" || response != "" {
-				if err := execute(); err != nil {
-					return err
-				}
-			}
-			system = m.Content
-		case "user":
-			if response != "" {
-				if err := execute(); err != nil {
-					return err
-				}
-			}
-			prompt = m.Content
-		case "assistant":
-			response = m.Content
 		}
 	}
 
 	var cut bool
-	nodes := deleteNode(t.Template.Root.Copy(), func(n parse.Node) bool {
-		if field, ok := n.(*parse.FieldNode); ok && slices.Contains(field.Ident, "Response") {
+	tree := t.Template.Copy()
+	// for the last message, cut everything after "{{ .Response }}"
+	tree.Root.Nodes = slices.DeleteFunc(tree.Root.Nodes, func(n parse.Node) bool {
+		if slices.Contains(parseNode(n), "Response") {
 			cut = true
 		}
 
 		return cut
 	})
 
-	tree := parse.Tree{Root: nodes.(*parse.ListNode)}
-	if err := template.Must(template.New("").AddParseTree("", &tree)).Execute(&b, map[string]any{
+	if err := template.Must(template.New("").AddParseTree("", tree)).Execute(&b, map[string]any{
 		"System": system,
 		"Prompt": prompt,
 	}); err != nil {
@@ -280,16 +199,25 @@ func (t *Template) Execute(w io.Writer, v Values) error {
 	return err
 }
 
-// collate messages based on role. consecutive messages of the same role are merged
-// into a single message. collate also collects and returns all system messages.
-// collate mutates message content adding image tags ([img-%d]) as needed
-func collate(msgs []api.Message) (string, []*api.Message) {
-	var n int
+type messages []*api.Message
 
-	var system []string
-	var collated []*api.Message
+// collate messages based on role. consecutive messages of the same role are merged
+// into a single message. collate also pulls out and merges messages with Role == "system"
+// which are templated separately. As a side effect, it mangles message content adding image
+// tags ([img-%d]) as needed
+func collate(msgs []api.Message) (system string, collated messages) {
+	var n int
 	for i := range msgs {
 		msg := msgs[i]
+		if msg.Role == "system" {
+			if system != "" {
+				system += "\n\n"
+			}
+
+			system += msg.Content
+			continue
+		}
+
 		for range msg.Images {
 			imageTag := fmt.Sprintf("[img-%d]", n)
 			if !strings.Contains(msg.Content, "[img]") {
@@ -300,10 +228,6 @@ func collate(msgs []api.Message) (string, []*api.Message) {
 			n++
 		}
 
-		if msg.Role == "system" {
-			system = append(system, msg.Content)
-		}
-
 		if len(collated) > 0 && collated[len(collated)-1].Role == msg.Role {
 			collated[len(collated)-1].Content += "\n\n" + msg.Content
 		} else {
@@ -311,119 +235,54 @@ func collate(msgs []api.Message) (string, []*api.Message) {
 		}
 	}
 
-	return strings.Join(system, "\n\n"), collated
+	return
 }
 
-// Identifiers walks the node tree returning any identifiers it finds along the way
-func Identifiers(n parse.Node) []string {
+func parseNode(n parse.Node) []string {
 	switch n := n.(type) {
-	case *parse.ListNode:
-		var names []string
-		for _, n := range n.Nodes {
-			names = append(names, Identifiers(n)...)
-		}
-
-		return names
-	case *parse.TemplateNode:
-		return Identifiers(n.Pipe)
 	case *parse.ActionNode:
-		return Identifiers(n.Pipe)
-	case *parse.BranchNode:
-		names := Identifiers(n.Pipe)
-		for _, n := range []*parse.ListNode{n.List, n.ElseList} {
-			if n != nil {
-				names = append(names, Identifiers(n)...)
-			}
+		return parseNode(n.Pipe)
+	case *parse.IfNode:
+		names := parseNode(n.Pipe)
+		names = append(names, parseNode(n.List)...)
+		if n.ElseList != nil {
+			names = append(names, parseNode(n.ElseList)...)
 		}
 		return names
-	case *parse.IfNode:
-		return Identifiers(&n.BranchNode)
 	case *parse.RangeNode:
-		return Identifiers(&n.BranchNode)
+		names := parseNode(n.Pipe)
+		names = append(names, parseNode(n.List)...)
+		if n.ElseList != nil {
+			names = append(names, parseNode(n.ElseList)...)
+		}
+		return names
 	case *parse.WithNode:
-		return Identifiers(&n.BranchNode)
+		names := parseNode(n.Pipe)
+		names = append(names, parseNode(n.List)...)
+		if n.ElseList != nil {
+			names = append(names, parseNode(n.ElseList)...)
+		}
+		return names
 	case *parse.PipeNode:
 		var names []string
 		for _, c := range n.Cmds {
 			for _, a := range c.Args {
-				names = append(names, Identifiers(a)...)
+				names = append(names, parseNode(a)...)
 			}
 		}
+		return names
+	case *parse.ListNode:
+		var names []string
+		for _, n := range n.Nodes {
+			names = append(names, parseNode(n)...)
+		}
+
 		return names
 	case *parse.FieldNode:
 		return n.Ident
-	case *parse.VariableNode:
-		return n.Ident
+	case *parse.TemplateNode:
+		return parseNode(n.Pipe)
 	}
 
 	return nil
 }
-
-// deleteNode walks the node list and deletes nodes that match the predicate
-// this is currently to remove the {{ .Response }} node from templates
-func deleteNode(n parse.Node, fn func(parse.Node) bool) parse.Node {
-	var walk func(n parse.Node) parse.Node
-	walk = func(n parse.Node) parse.Node {
-		if fn(n) {
-			return nil
-		}
-
-		switch t := n.(type) {
-		case *parse.ListNode:
-			var nodes []parse.Node
-			for _, c := range t.Nodes {
-				if n := walk(c); n != nil {
-					nodes = append(nodes, n)
-				}
-			}
-
-			t.Nodes = nodes
-			return t
-		case *parse.IfNode:
-			t.BranchNode = *(walk(&t.BranchNode).(*parse.BranchNode))
-		case *parse.WithNode:
-			t.BranchNode = *(walk(&t.BranchNode).(*parse.BranchNode))
-		case *parse.RangeNode:
-			t.BranchNode = *(walk(&t.BranchNode).(*parse.BranchNode))
-		case *parse.BranchNode:
-			t.List = walk(t.List).(*parse.ListNode)
-			if t.ElseList != nil {
-				t.ElseList = walk(t.ElseList).(*parse.ListNode)
-			}
-		case *parse.ActionNode:
-			n := walk(t.Pipe)
-			if n == nil {
-				return nil
-			}
-
-			t.Pipe = n.(*parse.PipeNode)
-		case *parse.PipeNode:
-			var commands []*parse.CommandNode
-			for _, c := range t.Cmds {
-				var args []parse.Node
-				for _, a := range c.Args {
-					if n := walk(a); n != nil {
-						args = append(args, n)
-					}
-				}
-
-				if len(args) == 0 {
-					return nil
-				}
-
-				c.Args = args
-				commands = append(commands, c)
-			}
-
-			if len(commands) == 0 {
-				return nil
-			}
-
-			t.Cmds = commands
-		}
-
-		return n
-	}
-
-	return walk(n)
-}

template/template_test.go

@@ -105,8 +105,8 @@ func TestTemplate(t *testing.T) {
 			}
 
 			for n, tt := range cases {
-				var actual bytes.Buffer
 				t.Run(n, func(t *testing.T) {
+					var actual bytes.Buffer
 					if err := tmpl.Execute(&actual, Values{Messages: tt}); err != nil {
 						t.Fatal(err)
 					}
@@ -116,34 +116,7 @@ func TestTemplate(t *testing.T) {
 						t.Fatal(err)
 					}
 
-					bts := actual.Bytes()
-
-					if slices.Contains([]string{"chatqa.gotmpl", "llama2-chat.gotmpl", "mistral-instruct.gotmpl", "openchat.gotmpl", "vicuna.gotmpl"}, match) && bts[len(bts)-1] == ' ' {
-						t.Log("removing trailing space from output")
-						bts = bts[:len(bts)-1]
-					}
-
-					if diff := cmp.Diff(bts, expect); diff != "" {
-						t.Errorf("mismatch (-got +want):\n%s", diff)
-					}
-				})
-
-				t.Run("legacy", func(t *testing.T) {
-					t.Skip("legacy outputs are currently default outputs")
-					var legacy bytes.Buffer
-					if err := tmpl.Execute(&legacy, Values{Messages: tt, forceLegacy: true}); err != nil {
-						t.Fatal(err)
-					}
-
-					legacyBytes := legacy.Bytes()
-					if slices.Contains([]string{"chatqa.gotmpl", "openchat.gotmpl", "vicuna.gotmpl"}, match) && legacyBytes[len(legacyBytes)-1] == ' ' {
-						t.Log("removing trailing space from legacy output")
-						legacyBytes = legacyBytes[:len(legacyBytes)-1]
-					} else if slices.Contains([]string{"codellama-70b-instruct.gotmpl", "llama2-chat.gotmpl", "mistral-instruct.gotmpl"}, match) {
-						t.Skip("legacy outputs cannot be compared to messages outputs")
-					}
-
-					if diff := cmp.Diff(legacyBytes, actual.Bytes()); diff != "" {
+					if diff := cmp.Diff(actual.Bytes(), expect); diff != "" {
 						t.Errorf("mismatch (-got +want):\n%s", diff)
 					}
 				})
@@ -162,24 +135,7 @@ func TestParse(t *testing.T) {
 		{"{{ .System }} {{ .Prompt }} {{ .Response }}", []string{"prompt", "response", "system"}},
 		{"{{ with .Tools }}{{ . }}{{ end }} {{ .System }} {{ .Prompt }}", []string{"prompt", "response", "system", "tools"}},
 		{"{{ range .Messages }}{{ .Role }} {{ .Content }}{{ end }}", []string{"content", "messages", "role"}},
-		{`{{- range .Messages }}
-{{- if eq .Role "system" }}SYSTEM:
-{{- else if eq .Role "user" }}USER:
-{{- else if eq .Role "assistant" }}ASSISTANT:
-{{- end }} {{ .Content }}
-{{- end }}`, []string{"content", "messages", "role"}},
-		{`{{- if .Messages }}
-{{- range .Messages }}<|im_start|>{{ .Role }}
-{{ .Content }}<|im_end|>
-{{ end }}<|im_start|>assistant
-{{ else -}}
-{{ if .System }}<|im_start|>system
-{{ .System }}<|im_end|>
-{{ end }}{{ if .Prompt }}<|im_start|>user
-{{ .Prompt }}<|im_end|>
-{{ end }}<|im_start|>assistant
-{{ .Response }}<|im_end|>
-{{- end -}}`, []string{"content", "messages", "prompt", "response", "role", "system"}},
+		{"{{ range .Messages }}{{ if eq .Role \"system\" }}SYSTEM: {{ .Content }}{{ else if eq .Role \"user\" }}USER: {{ .Content }}{{ else if eq .Role \"assistant\" }}ASSISTANT: {{ .Content }}{{ end }}{{ end }}", []string{"content", "messages", "role"}},
 	}
 
 	for _, tt := range cases {
@@ -189,8 +145,9 @@ func TestParse(t *testing.T) {
 				t.Fatal(err)
 			}
 
-			if diff := cmp.Diff(tmpl.Vars(), tt.vars); diff != "" {
-				t.Errorf("mismatch (-got +want):\n%s", diff)
+			vars := tmpl.Vars()
+			if !slices.Equal(tt.vars, vars) {
+				t.Errorf("expected %v, got %v", tt.vars, vars)
 			}
 		})
 	}
@@ -210,17 +167,12 @@ func TestExecuteWithMessages(t *testing.T) {
 		{
 			"mistral",
 			[]template{
-				{"no response", `[INST] {{ if .System }}{{ .System }}
-
-{{ end }}{{ .Prompt }}[/INST] `},
-				{"response", `[INST] {{ if .System }}{{ .System }}
-
-{{ end }}{{ .Prompt }}[/INST] {{ .Response }}`},
-				{"messages", `[INST] {{ if .System }}{{ .System }}
-
-{{ end }}
-{{- range .Messages }}
-{{- if eq .Role "user" }}{{ .Content }}[/INST] {{ else if eq .Role "assistant" }}{{ .Content }}[INST] {{ end }}
+				{"no response", `[INST] {{ if .System }}{{ .System }}{{ "\n\n" }}{{ end }}{{ .Prompt }}[/INST] `},
+				{"response", `[INST] {{ if .System }}{{ .System }}{{ "\n\n" }}{{ end }}{{ .Prompt }}[/INST] {{ .Response }}`},
+				{"messages", `{{- range $index, $_ := .Messages }}
+{{- if eq .Role "user" }}[INST] {{ if and (eq (len (slice $.Messages $index)) 1) $.System }}{{ $.System }}{{ "\n\n" }}
+{{- end }}{{ .Content }}[/INST] {{ else if eq .Role "assistant" }}{{ .Content }}
+{{- end }}
 {{- end }}`},
 			},
 			Values{
@@ -235,17 +187,13 @@ func TestExecuteWithMessages(t *testing.T) {
 		{
 			"mistral system",
 			[]template{
-				{"no response", `[INST] {{ if .System }}{{ .System }}
-
-{{ end }}{{ .Prompt }}[/INST] `},
-				{"response", `[INST] {{ if .System }}{{ .System }}
-
-{{ end }}{{ .Prompt }}[/INST] {{ .Response }}`},
-				{"messages", `[INST] {{ if .System }}{{ .System }}
-
-{{ end }}
-{{- range .Messages }}
-{{- if eq .Role "user" }}{{ .Content }}[/INST] {{ else if eq .Role "assistant" }}{{ .Content }}[INST] {{ end }}
+				{"no response", `[INST] {{ if .System }}{{ .System }}{{ "\n\n" }}{{ end }}{{ .Prompt }}[/INST] `},
+				{"response", `[INST] {{ if .System }}{{ .System }}{{ "\n\n" }}{{ end }}{{ .Prompt }}[/INST] {{ .Response }}`},
+				{"messages", `
+{{- range $index, $_ := .Messages }}
+{{- if eq .Role "user" }}[INST] {{ if and (eq (len (slice $.Messages $index)) 1) $.System }}{{ $.System }}{{ "\n\n" }}
+{{- end }}{{ .Content }}[/INST] {{ else if eq .Role "assistant" }}{{ .Content }}
+{{- end }}
 {{- end }}`},
 			},
 			Values{
@@ -256,9 +204,9 @@ func TestExecuteWithMessages(t *testing.T) {
 					{Role: "user", Content: "What is your name?"},
 				},
 			},
-			`[INST] You are a helpful assistant!
+			`[INST] Hello friend![/INST] Hello human![INST] You are a helpful assistant!
 
-Hello friend![/INST] Hello human![INST] What is your name?[/INST] `,
+What is your name?[/INST] `,
 		},
 		{
 			"chatml",
@@ -272,9 +220,12 @@ Hello friend![/INST] Hello human![INST] What is your name?[/INST] `,
 {{ .Response }}<|im_end|>
 `},
 				{"messages", `
-{{- range $index, $_ := .Messages }}<|im_start|>{{ .Role }}
-{{ .Content }}<|im_end|>
-{{ end }}<|im_start|>assistant
+{{- range $index, $_ := .Messages }}
+{{- if and (eq .Role "user") (eq (len (slice $.Messages $index)) 1) $.System }}<|im_start|>system
+{{ $.System }}<|im_end|>{{ "\n" }}
+{{- end }}<|im_start|>{{ .Role }}
+{{ .Content }}<|im_end|>{{ "\n" }}
+{{- end }}<|im_start|>assistant
 `},
 			},
 			Values{
@@ -285,12 +236,12 @@ Hello friend![/INST] Hello human![INST] What is your name?[/INST] `,
 					{Role: "user", Content: "What is your name?"},
 				},
 			},
-			`<|im_start|>system
-You are a helpful assistant!<|im_end|>
-<|im_start|>user
+			`<|im_start|>user
 Hello friend!<|im_end|>
 <|im_start|>assistant
 Hello human!<|im_end|>
+<|im_start|>system
+You are a helpful assistant!<|im_end|>
 <|im_start|>user
 What is your name?<|im_end|>
 <|im_start|>assistant
@@ -307,11 +258,9 @@ What is your name?<|im_end|>
 `},
 				{"messages", `
 {{- range .Messages }}
-{{- if eq .Role "user" }}Question: {{ .Content }}
-
-{{ else if eq .Role "assistant" }}Answer: {{ .Content }}
-
-{{ end }}
+{{- if eq .Role "user" }}Question: {{ .Content }}{{ "\n\n" }}
+{{- else if eq .Role "assistant" }}Answer: {{ .Content }}{{ "\n\n" }}
+{{- end }}
 {{- end }}Answer: `},
 			},
 			Values{
@@ -351,46 +300,11 @@ Answer: `,
 						t.Fatal(err)
 					}
 
-					if diff := cmp.Diff(b.String(), tt.expected); diff != "" {
-						t.Errorf("mismatch (-got +want):\n%s", diff)
+					if b.String() != tt.expected {
+						t.Errorf("expected\n%s,\ngot\n%s", tt.expected, b.String())
 					}
 				})
 			}
 		})
 	}
 }
-
-func TestExecuteWithSuffix(t *testing.T) {
-	tmpl, err := Parse(`{{- if .Suffix }}<PRE> {{ .Prompt }} <SUF>{{ .Suffix }} <MID>
-{{- else }}{{ .Prompt }}
-{{- end }}`)
-	if err != nil {
-		t.Fatal(err)
-	}
-
-	cases := []struct {
-		name   string
-		values Values
-		expect string
-	}{
-		{
-			"message", Values{Messages: []api.Message{{Role: "user", Content: "hello"}}}, "hello",
-		},
-		{
-			"prompt suffix", Values{Prompt: "def add(", Suffix: "return x"}, "<PRE> def add( <SUF>return x <MID>",
-		},
-	}
-
-	for _, tt := range cases {
-		t.Run(tt.name, func(t *testing.T) {
-			var b bytes.Buffer
-			if err := tmpl.Execute(&b, tt.values); err != nil {
-				t.Fatal(err)
-			}
-
-			if diff := cmp.Diff(b.String(), tt.expect); diff != "" {
-				t.Errorf("mismatch (-got +want):\n%s", diff)
-			}
-		})
-	}
-}

template/testdata/alpaca.gotmpl/system-user-assistant-user

@@ -1,6 +1,4 @@
-You are a helpful assistant.
-
-### Instruction:
+You are a helpful assistant.### Instruction:
 Hello, how are you?
 
 ### Response:

template/testdata/codellama-70b-instruct.gotmpl/system-user-assistant-user

@@ -9,4 +9,3 @@ Source: system
  I'd like to show off how chat templating works! <step> Source: assistant
 Destination: user
 
- 

template/testdata/codellama-70b-instruct.gotmpl/user

@@ -3,4 +3,3 @@ Source: user
  Hello, how are you? <step> Source: assistant
 Destination: user
 
- 

template/testdata/codellama-70b-instruct.gotmpl/user-assistant-user

@@ -7,4 +7,3 @@ Source: user
  I'd like to show off how chat templating works! <step> Source: assistant
 Destination: user
 
- 

template/testdata/llama2-chat.gotmpl/system-user-assistant-user

@@ -2,6 +2,4 @@
 You are a helpful assistant.
 <</SYS>>
 
-Hello, how are you? [/INST] I'm doing great. How can I help you today?</s><s>[INST] <<SYS>><</SYS>>
-
-I'd like to show off how chat templating works! [/INST]
+Hello, how are you? [/INST] I'm doing great. How can I help you today?</s><s>[INST] I'd like to show off how chat templating works! [/INST]

template/testdata/llama2-chat.gotmpl/user-assistant-user

@@ -1,5 +1,3 @@
 [INST] <<SYS>><</SYS>>
 
-Hello, how are you? [/INST] I'm doing great. How can I help you today?</s><s>[INST] <<SYS>><</SYS>>
-
-I'd like to show off how chat templating works! [/INST]
+Hello, how are you? [/INST] I'm doing great. How can I help you today?</s><s>[INST] I'd like to show off how chat templating works! [/INST]

template/testdata/mistral-instruct.gotmpl/system-user-assistant-user

@@ -1,3 +1,2 @@
-[INST] You are a helpful assistant.
-
-Hello, how are you?[/INST] I'm doing great. How can I help you today?</s>[INST] I'd like to show off how chat templating works![/INST]
+[INST] Hello, how are you?[/INST] I'm doing great. How can I help you today?</s>[INST] You are a helpful assistant.
+I'd like to show off how chat templating works![/INST]

template/testdata/openchat.gotmpl/system-user-assistant-user

@@ -1 +1 @@
-GPT4 Correct System: You are a helpful assistant.<|end_of_turn|>GPT4 Correct User: Hello, how are you?<|end_of_turn|>GPT4 Correct Assistant: I'm doing great. How can I help you today?<|end_of_turn|>GPT4 Correct User: I'd like to show off how chat templating works!<|end_of_turn|>GPT4 Correct Assistant:
+GPT Correct System: You are a helpful assistant.<|end_of_turn|>GPT Correct User: Hello, how are you?<|end_of_turn|>GPT Correct Assistant: I'm doing great. How can I help you today?<|end_of_turn|>GPT Correct User: I'd like to show off how chat templating works!<|end_of_turn|>GPT Correct Assistant:

template/testdata/openchat.gotmpl/user

@@ -1 +1 @@
-GPT4 Correct User: Hello, how are you?<|end_of_turn|>GPT4 Correct Assistant:
+GPT Correct User: Hello, how are you?<|end_of_turn|>GPT Correct Assistant:

template/testdata/openchat.gotmpl/user-assistant-user

@@ -1 +1 @@
-GPT4 Correct User: Hello, how are you?<|end_of_turn|>GPT4 Correct Assistant: I'm doing great. How can I help you today?<|end_of_turn|>GPT4 Correct User: I'd like to show off how chat templating works!<|end_of_turn|>GPT4 Correct Assistant:
+GPT Correct User: Hello, how are you?<|end_of_turn|>GPT Correct Assistant: I'm doing great. How can I help you today?<|end_of_turn|>GPT Correct User: I'd like to show off how chat templating works!<|end_of_turn|>GPT Correct Assistant:

template/vicuna.gotmpl

@@ -1,4 +1,14 @@
-{{ if .System }}{{ .System }}
+{{- if .Messages }}
+{{- if .System }}{{ .System }}
 
+{{ end }}
+{{- range .Messages }}
+{{- if eq .Role "user" }}USER: {{ .Content }}
+{{ else if eq .Role "assistant" }}ASSISTANT: {{ .Content }}</s>
+{{ end }}
+{{- end }}ASSISTANT:
+{{- else }}
+{{ if .System }}{{ .System }}
 {{ end }}{{ if .Prompt }}USER: {{ .Prompt }}
-{{ end }}ASSISTANT: {{ .Response }}</s>
+{{ end }}ASSISTANT: {{ .Response }}
+{{- end }}

template/zephyr.gotmpl

@@ -1,6 +1,15 @@
+{{- if .Messages }}
+{{- if .System }}<|system|>
+{{ .System }}</s>
+{{ end }}
+{{- range .Messages }}<|{{ .Role }}|>
+{{ .Content }}</s>
+{{ end }}<|assistant|>
+{{ else }}
 {{ if .System }}<|system|>
 {{ .System }}</s>
 {{ end }}{{ if .Prompt }}<|user|>
 {{ .Prompt }}</s>
 {{ end }}<|assistant|>
 {{ .Response }}</s>
+{{- end }}