Merge branch 'ollama:main' into main

README.md

@@ -377,6 +377,7 @@ See the [API documentation](./docs/api.md) for all endpoints.
 - [Nosia](https://github.com/nosia-ai/nosia) (Easy to install and use RAG platform based on Ollama)
 - [Witsy](https://github.com/nbonamy/witsy) (An AI Desktop application avaiable for Mac/Windows/Linux)
 - [Abbey](https://github.com/US-Artificial-Intelligence/abbey) (A configurable AI interface server with notebooks, document storage, and YouTube support)
+- [Minima](https://github.com/dmayboroda/minima) (RAG with on-premises or fully local workflow)
 
 ### Cloud
 

api/types.go

@@ -146,6 +146,7 @@ type ToolCall struct {
 }
 
 type ToolCallFunction struct {
+	Index     int                       `json:"index,omitempty"`
 	Name      string                    `json:"name"`
 	Arguments ToolCallFunctionArguments `json:"arguments"`
 }

cmd/cmd_test.go

@@ -8,7 +8,6 @@ import (
 	"net/http"
 	"net/http/httptest"
 	"os"
-	"path/filepath"
 	"strings"
 	"testing"
 
@@ -180,18 +179,14 @@ Weigh anchor!
 
 	t.Run("license", func(t *testing.T) {
 		var b bytes.Buffer
-		license, err := os.ReadFile(filepath.Join("..", "LICENSE"))
+		license := "MIT License\nCopyright (c) Ollama\n"
-		if err != nil {
-			t.Fatal(err)
-		}
-
 		if err := showInfo(&api.ShowResponse{
 			Details: api.ModelDetails{
 				Family:            "test",
 				ParameterSize:     "7B",
 				QuantizationLevel: "FP16",
 			},
-			License: string(license),
+			License: license,
 		}, &b); err != nil {
 			t.Fatal(err)
 		}

docs/api.md

@@ -49,10 +49,10 @@ Advanced parameters (optional):
 - `options`: additional model parameters listed in the documentation for the [Modelfile](./modelfile.md#valid-parameters-and-values) such as `temperature`
 - `system`: system message to (overrides what is defined in the `Modelfile`)
 - `template`: the prompt template to use (overrides what is defined in the `Modelfile`)
-- `context`: the context parameter returned from a previous request to `/generate`, this can be used to keep a short conversational memory
 - `stream`: if `false` the response will be returned as a single response object, rather than a stream of objects
 - `raw`: if `true` no formatting will be applied to the prompt. You may choose to use the `raw` parameter if you are specifying a full templated prompt in your request to the API
 - `keep_alive`: controls how long the model will stay loaded into memory following the request (default: `5m`)
+- `context` (deprecated): the context parameter returned from a previous request to `/generate`, this can be used to keep a short conversational memory
 
 #### JSON mode
 

docs/modelfile.md

@@ -63,7 +63,7 @@ SYSTEM You are Mario from super mario bros, acting as an assistant.
 To use this:
 
 1. Save it as a file (e.g. `Modelfile`)
-2. `ollama create choose-a-model-name -f <location of the file e.g. ./Modelfile>'`
+2. `ollama create choose-a-model-name -f <location of the file e.g. ./Modelfile>`
 3. `ollama run choose-a-model-name`
 4. Start using the model!
 

openai/openai.go

@@ -140,6 +140,7 @@ type CompletionChunk struct {
 
 type ToolCall struct {
 	ID       string `json:"id"`
+	Index    int    `json:"index"`
 	Type     string `json:"type"`
 	Function struct {
 		Name      string `json:"name"`
@@ -206,6 +207,7 @@ func toToolCalls(tc []api.ToolCall) []ToolCall {
 		toolCalls[i].ID = toolCallId()
 		toolCalls[i].Type = "function"
 		toolCalls[i].Function.Name = tc.Function.Name
+		toolCalls[i].Index = tc.Function.Index
 
 		args, err := json.Marshal(tc.Function.Arguments)
 		if err != nil {

openai/openai_test.go

@@ -195,7 +195,86 @@ func TestChatMiddleware(t *testing.T) {
 				Stream: &False,
 			},
 		},
-
+		{
+			name: "chat handler with streaming tools",
+			body: `{
+				"model": "test-model",
+				"messages": [
+					{"role": "user", "content": "What's the weather like in Paris?"}
+				],
+				"stream": true,
+				"tools": [{
+					"type": "function",
+					"function": {
+						"name": "get_weather",
+						"description": "Get the current weather",
+						"parameters": {
+							"type": "object",
+							"required": ["location"],
+							"properties": {
+								"location": {
+									"type": "string",
+									"description": "The city and state"
+								},
+								"unit": {
+									"type": "string",
+									"enum": ["celsius", "fahrenheit"]
+								}
+							}
+						}
+					}
+				}]
+			}`,
+			req: api.ChatRequest{
+				Model: "test-model",
+				Messages: []api.Message{
+					{
+						Role:    "user",
+						Content: "What's the weather like in Paris?",
+					},
+				},
+				Tools: []api.Tool{
+					{
+						Type: "function",
+						Function: api.ToolFunction{
+							Name:        "get_weather",
+							Description: "Get the current weather",
+							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"`
+							}{
+								Type:     "object",
+								Required: []string{"location"},
+								Properties: map[string]struct {
+									Type        string   `json:"type"`
+									Description string   `json:"description"`
+									Enum        []string `json:"enum,omitempty"`
+								}{
+									"location": {
+										Type:        "string",
+										Description: "The city and state",
+									},
+									"unit": {
+										Type: "string",
+										Enum: []string{"celsius", "fahrenheit"},
+									},
+								},
+							},
+						},
+					},
+				},
+				Options: map[string]any{
+					"temperature": 1.0,
+					"top_p":       1.0,
+				},
+				Stream: &True,
+			},
+		},
 		{
 			name: "chat handler error forwarding",
 			body: `{

server/routes.go

@@ -251,6 +251,7 @@ func (s *Server) GenerateHandler(c *gin.Context) {
 
 		var b bytes.Buffer
 		if req.Context != nil {
+			slog.Warn("the context field is deprecated and will be removed in a future version of Ollama")
 			s, err := r.Detokenize(c.Request.Context(), req.Context)
 			if err != nil {
 				c.JSON(http.StatusInternalServerError, gin.H{"error": err.Error()})
@@ -1469,7 +1470,7 @@ func (s *Server) ChatHandler(c *gin.Context) {
 	go func() {
 		defer close(ch)
 		var sb strings.Builder
-		var hasToolCalls bool
+		var toolCallIndex int = 0
 		if err := r.Completion(c.Request.Context(), llm.CompletionRequest{
 			Prompt:  prompt,
 			Images:  images,
@@ -1509,16 +1510,19 @@ func (s *Server) ChatHandler(c *gin.Context) {
 			sb.WriteString(r.Content)
 			if toolCalls, ok := m.parseToolCalls(sb.String()); ok {
 				res.Message.ToolCalls = toolCalls
+				for i := range toolCalls {
+					toolCalls[i].Function.Index = toolCallIndex
+					toolCallIndex++
+				}
 				res.Message.Content = ""
 				sb.Reset()
-				hasToolCalls = true
 				ch <- res
 				return
 			}
 
 			if r.Done {
 				// Send any remaining content if no tool calls were detected
-				if !hasToolCalls {
+				if toolCallIndex == 0 {
 					res.Message.Content = sb.String()
 				}
 				ch <- res