add batch embeddings

.gitignore

@@ -11,5 +11,4 @@ ggml-metal.metal
 .idea
 test_data
 *.crt
-llm/build
-__debug_bin*
+llm/build

README.md

@@ -1,5 +1,5 @@
 <div align="center">
- <img alt="ollama" height="200px" src="https://github.com/ollama/ollama/assets/3325447/0d0b44e2-8f4a-4e99-9b52-a5c1c741c8f7">
+  <img alt="ollama" height="200px" src="https://github.com/ollama/ollama/assets/3325447/0d0b44e2-8f4a-4e99-9b52-a5c1c741c8f7">
 </div>
 
 # Ollama
@@ -51,7 +51,7 @@ Here are some example models that can be downloaded:
 | ------------------ | ---------- | ----- | ------------------------------ |
 | Llama 3            | 8B         | 4.7GB | `ollama run llama3`            |
 | Llama 3            | 70B        | 40GB  | `ollama run llama3:70b`        |
-| Phi-3              | 3.8B       | 2.3GB | `ollama run phi3`              |
+| Phi-3              | 3,8B       | 2.3GB | `ollama run phi3`              |
 | Mistral            | 7B         | 4.1GB | `ollama run mistral`           |
 | Neural Chat        | 7B         | 4.1GB | `ollama run neural-chat`       |
 | Starling           | 7B         | 4.1GB | `ollama run starling-lm`       |
@@ -173,7 +173,7 @@ I'm a basic program that prints the famous "Hello, world!" message to the consol
 The image features a yellow smiley face, which is likely the central focus of the picture.
 ```
 
-### Pass the prompt as an argument
+### Pass in prompt as arguments
 
 ```
 $ ollama run llama3 "Summarize this file: $(cat README.md)"
@@ -258,7 +258,6 @@ See the [API documentation](./docs/api.md) for all endpoints.
 
 - [Open WebUI](https://github.com/open-webui/open-webui)
 - [Enchanted (macOS native)](https://github.com/AugustDev/enchanted)
-- [Hollama](https://github.com/fmaclen/hollama)
 - [Lollms-Webui](https://github.com/ParisNeo/lollms-webui)
 - [LibreChat](https://github.com/danny-avila/LibreChat)
 - [Bionic GPT](https://github.com/bionic-gpt/bionic-gpt)
@@ -285,20 +284,17 @@ See the [API documentation](./docs/api.md) for all endpoints.
 - [OllamaGUI](https://github.com/enoch1118/ollamaGUI)
 - [OpenAOE](https://github.com/InternLM/OpenAOE)
 - [Odin Runes](https://github.com/leonid20000/OdinRunes)
-- [LLM-X](https://github.com/mrdjohnson/llm-x) (Progressive Web App)
+- [LLM-X: Progressive Web App](https://github.com/mrdjohnson/llm-x)
 - [AnythingLLM (Docker + MacOs/Windows/Linux native app)](https://github.com/Mintplex-Labs/anything-llm)
 - [Ollama Basic Chat: Uses HyperDiv Reactive UI](https://github.com/rapidarchitect/ollama_basic_chat)
 - [Ollama-chats RPG](https://github.com/drazdra/ollama-chats)
-- [QA-Pilot](https://github.com/reid41/QA-Pilot) (Chat with Code Repository)
-- [ChatOllama](https://github.com/sugarforever/chat-ollama) (Open Source Chatbot based on Ollama with Knowledge Bases)
-- [CRAG Ollama Chat](https://github.com/Nagi-ovo/CRAG-Ollama-Chat) (Simple Web Search with Corrective RAG)
-- [RAGFlow](https://github.com/infiniflow/ragflow) (Open-source Retrieval-Augmented Generation engine based on deep document understanding)
-- [StreamDeploy](https://github.com/StreamDeploy-DevRel/streamdeploy-llm-app-scaffold) (LLM Application Scaffold)
-- [chat](https://github.com/swuecho/chat) (chat web app for teams)
+- [QA-Pilot: Chat with Code Repository](https://github.com/reid41/QA-Pilot)
+- [ChatOllama: Open Source Chatbot based on Ollama with Knowledge Bases](https://github.com/sugarforever/chat-ollama)
+- [CRAG Ollama Chat: Simple Web Search with Corrective RAG](https://github.com/Nagi-ovo/CRAG-Ollama-Chat)
+- [RAGFlow: Open-source Retrieval-Augmented Generation engine based on deep document understanding](https://github.com/infiniflow/ragflow)
+- [chat: chat web app for teams](https://github.com/swuecho/chat)
 - [Lobe Chat](https://github.com/lobehub/lobe-chat) with [Integrating Doc](https://lobehub.com/docs/self-hosting/examples/ollama)
-- [Ollama RAG Chatbot](https://github.com/datvodinh/rag-chatbot.git) (Local Chat with multiple PDFs using Ollama and RAG)
-- [BrainSoup](https://www.nurgo-software.com/products/brainsoup) (Flexible native client with RAG & multi-agent automation)
-- [macai](https://github.com/Renset/macai) (macOS client for Ollama, ChatGPT, and other compatible API back-ends)
+- [Ollama RAG Chatbot: Local Chat with multiples PDFs using Ollama and RAG.](https://github.com/datvodinh/rag-chatbot.git)
 
 ### Terminal
 
@@ -331,7 +327,6 @@ See the [API documentation](./docs/api.md) for all endpoints.
 
 - [Pacman](https://archlinux.org/packages/extra/x86_64/ollama/)
 - [Helm Chart](https://artifacthub.io/packages/helm/ollama-helm/ollama)
-- [Guix channel](https://codeberg.org/tusharhero/ollama-guix)
 
 ### Libraries
 
@@ -353,13 +348,10 @@ See the [API documentation](./docs/api.md) for all endpoints.
 - [Haystack](https://github.com/deepset-ai/haystack-integrations/blob/main/integrations/ollama.md)
 - [Elixir LangChain](https://github.com/brainlid/langchain)
 - [Ollama for R - rollama](https://github.com/JBGruber/rollama)
-- [Ollama for R - ollama-r](https://github.com/hauselin/ollama-r)
 - [Ollama-ex for Elixir](https://github.com/lebrunel/ollama-ex)
 - [Ollama Connector for SAP ABAP](https://github.com/b-tocs/abap_btocs_ollama)
 - [Testcontainers](https://testcontainers.com/modules/ollama/)
-- [Portkey](https://portkey.ai/docs/welcome/integration-guides/ollama)
-- [PromptingTools.jl](https://github.com/svilupp/PromptingTools.jl) with an [example](https://svilupp.github.io/PromptingTools.jl/dev/examples/working_with_ollama)
-- [LlamaScript](https://github.com/WolfTheDeveloper/llamascript)
+
 ### Mobile
 
 - [Enchanted](https://github.com/AugustDev/enchanted)
@@ -378,13 +370,12 @@ See the [API documentation](./docs/api.md) for all endpoints.
 - [Ollama Telegram Bot](https://github.com/ruecat/ollama-telegram)
 - [Hass Ollama Conversation](https://github.com/ej52/hass-ollama-conversation)
 - [Rivet plugin](https://github.com/abrenneke/rivet-plugin-ollama)
+- [Llama Coder](https://github.com/ex3ndr/llama-coder) (Copilot alternative using Ollama)
 - [Obsidian BMO Chatbot plugin](https://github.com/longy2k/obsidian-bmo-chatbot)
 - [Cliobot](https://github.com/herval/cliobot) (Telegram bot with Ollama support)
 - [Copilot for Obsidian plugin](https://github.com/logancyang/obsidian-copilot)
 - [Obsidian Local GPT plugin](https://github.com/pfrankov/obsidian-local-gpt)
 - [Open Interpreter](https://docs.openinterpreter.com/language-model-setup/local-models/ollama)
-- [Llama Coder](https://github.com/ex3ndr/llama-coder) (Copilot alternative using Ollama)
-- [Ollama Copilot](https://github.com/bernardo-bruning/ollama-copilot) (Proxy that allows you to use ollama as a copilot like Github copilot)
 - [twinny](https://github.com/rjmacarthy/twinny) (Copilot and Copilot chat alternative using Ollama)
 - [Wingman-AI](https://github.com/RussellCanfield/wingman-ai) (Copilot code and chat alternative using Ollama and HuggingFace)
 - [Page Assist](https://github.com/n4ze3m/page-assist) (Chrome Extension)
@@ -393,5 +384,4 @@ See the [API documentation](./docs/api.md) for all endpoints.
 - [Discord-Ollama Chat Bot](https://github.com/kevinthedang/discord-ollama) (Generalized TypeScript Discord Bot w/ Tuning Documentation)
 
 ### Supported backends 
-- [llama.cpp](https://github.com/ggerganov/llama.cpp) project founded by Georgi Gerganov.
-
+- [llama.cpp](https://github.com/ggerganov/llama.cpp) project founded by Georgi Gerganov. 

api/client.go

@@ -1,16 +1,9 @@
 // Package api implements the client-side API for code wishing to interact
 // with the ollama service. The methods of the [Client] type correspond to
-// the ollama REST API as described in [the API documentation].
+// the ollama REST API as described in https://github.com/ollama/ollama/blob/main/docs/api.md
+//
 // The ollama command-line client itself uses this package to interact with
 // the backend service.
-//
-// # Examples
-//
-// Several examples of using this package are available [in the GitHub
-// repository].
-//
-// [the API documentation]: https://github.com/ollama/ollama/blob/main/docs/api.md
-// [in the GitHub repository]: https://github.com/ollama/ollama/tree/main/examples
 package api
 
 import (
@@ -25,7 +18,6 @@ import (
 	"net/url"
 	"os"
 	"runtime"
-	"strconv"
 	"strings"
 
 	"github.com/ollama/ollama/format"
@@ -65,36 +57,12 @@ func checkError(resp *http.Response, body []byte) error {
 // If the variable is not specified, a default ollama host and port will be
 // used.
 func ClientFromEnvironment() (*Client, error) {
-	ollamaHost, err := GetOllamaHost()
-	if err != nil {
-		return nil, err
-	}
-
-	return &Client{
-		base: &url.URL{
-			Scheme: ollamaHost.Scheme,
-			Host:   net.JoinHostPort(ollamaHost.Host, ollamaHost.Port),
-		},
-		http: http.DefaultClient,
-	}, nil
-}
-
-type OllamaHost struct {
-	Scheme string
-	Host   string
-	Port   string
-}
-
-func GetOllamaHost() (OllamaHost, error) {
 	defaultPort := "11434"
 
-	hostVar := os.Getenv("OLLAMA_HOST")
-	hostVar = strings.TrimSpace(strings.Trim(strings.TrimSpace(hostVar), "\"'"))
-
-	scheme, hostport, ok := strings.Cut(hostVar, "://")
+	scheme, hostport, ok := strings.Cut(os.Getenv("OLLAMA_HOST"), "://")
 	switch {
 	case !ok:
-		scheme, hostport = "http", hostVar
+		scheme, hostport = "http", os.Getenv("OLLAMA_HOST")
 	case scheme == "http":
 		defaultPort = "80"
 	case scheme == "https":
@@ -114,14 +82,12 @@ func GetOllamaHost() (OllamaHost, error) {
 		}
 	}
 
-	if portNum, err := strconv.ParseInt(port, 10, 32); err != nil || portNum > 65535 || portNum < 0 {
-		return OllamaHost{}, ErrInvalidHostPort
-	}
-
-	return OllamaHost{
-		Scheme: scheme,
-		Host:   host,
-		Port:   port,
+	return &Client{
+		base: &url.URL{
+			Scheme: scheme,
+			Host:   net.JoinHostPort(host, port),
+		},
+		http: http.DefaultClient,
 	}, nil
 }
 
@@ -306,14 +272,8 @@ func (c *Client) Pull(ctx context.Context, req *PullRequest, fn PullProgressFunc
 	})
 }
 
-// PushProgressFunc is a function that [Client.Push] invokes when progress is
-// made.
-// It's similar to other progress function types like [PullProgressFunc].
 type PushProgressFunc func(ProgressResponse) error
 
-// Push uploads a model to the model library; requires registering for ollama.ai
-// and adding a public key first. fn is called each time progress is made on
-// the request and can be used to display a progress bar, etc.
 func (c *Client) Push(ctx context.Context, req *PushRequest, fn PushProgressFunc) error {
 	return c.stream(ctx, http.MethodPost, "/api/push", req, func(bts []byte) error {
 		var resp ProgressResponse
@@ -325,15 +285,8 @@ func (c *Client) Push(ctx context.Context, req *PushRequest, fn PushProgressFunc
 	})
 }
 
-// CreateProgressFunc is a function that [Client.Create] invokes when progress
-// is made.
-// It's similar to other progress function types like [PullProgressFunc].
 type CreateProgressFunc func(ProgressResponse) error
 
-// Create creates a model from a [Modelfile]. fn is a progress function that
-// behaves similarly to other methods (see [Client.Pull]).
-//
-// [Modelfile]: https://github.com/ollama/ollama/blob/main/docs/modelfile.md
 func (c *Client) Create(ctx context.Context, req *CreateRequest, fn CreateProgressFunc) error {
 	return c.stream(ctx, http.MethodPost, "/api/create", req, func(bts []byte) error {
 		var resp ProgressResponse
@@ -345,7 +298,6 @@ func (c *Client) Create(ctx context.Context, req *CreateRequest, fn CreateProgre
 	})
 }
 
-// List lists models that are available locally.
 func (c *Client) List(ctx context.Context) (*ListResponse, error) {
 	var lr ListResponse
 	if err := c.do(ctx, http.MethodGet, "/api/tags", nil, &lr); err != nil {
@@ -354,8 +306,6 @@ func (c *Client) List(ctx context.Context) (*ListResponse, error) {
 	return &lr, nil
 }
 
-// Copy copies a model - creating a model with another name from an existing
-// model.
 func (c *Client) Copy(ctx context.Context, req *CopyRequest) error {
 	if err := c.do(ctx, http.MethodPost, "/api/copy", req, nil); err != nil {
 		return err
@@ -363,7 +313,6 @@ func (c *Client) Copy(ctx context.Context, req *CopyRequest) error {
 	return nil
 }
 
-// Delete deletes a model and its data.
 func (c *Client) Delete(ctx context.Context, req *DeleteRequest) error {
 	if err := c.do(ctx, http.MethodDelete, "/api/delete", req, nil); err != nil {
 		return err
@@ -371,7 +320,6 @@ func (c *Client) Delete(ctx context.Context, req *DeleteRequest) error {
 	return nil
 }
 
-// Show obtains model information, including details, modelfile, license etc.
 func (c *Client) Show(ctx context.Context, req *ShowRequest) (*ShowResponse, error) {
 	var resp ShowResponse
 	if err := c.do(ctx, http.MethodPost, "/api/show", req, &resp); err != nil {
@@ -380,16 +328,12 @@ func (c *Client) Show(ctx context.Context, req *ShowRequest) (*ShowResponse, err
 	return &resp, nil
 }
 
-// Hearbeat checks if the server has started and is responsive; if yes, it
-// returns nil, otherwise an error.
 func (c *Client) Heartbeat(ctx context.Context) error {
 	if err := c.do(ctx, http.MethodHead, "/", nil, nil); err != nil {
 		return err
 	}
 	return nil
 }
-
-// Embeddings generates embeddings from a model.
 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 {
@@ -398,13 +342,10 @@ func (c *Client) Embeddings(ctx context.Context, req *EmbeddingRequest) (*Embedd
 	return &resp, nil
 }
 
-// CreateBlob creates a blob from a file on the server. digest is the
-// expected SHA256 digest of the file, and r represents the file.
 func (c *Client) CreateBlob(ctx context.Context, digest string, r io.Reader) error {
 	return c.do(ctx, http.MethodPost, fmt.Sprintf("/api/blobs/%s", digest), r, nil)
 }
 
-// Version returns the Ollama server version as a string.
 func (c *Client) Version(ctx context.Context) (string, error) {
 	var version struct {
 		Version string `json:"version"`

api/client_test.go

@@ -1,12 +1,6 @@
 package api
 
-import (
-	"fmt"
-	"net"
-	"testing"
-
-	"github.com/stretchr/testify/assert"
-)
+import "testing"
 
 func TestClientFromEnvironment(t *testing.T) {
 	type testCase struct {
@@ -46,40 +40,4 @@ func TestClientFromEnvironment(t *testing.T) {
 			}
 		})
 	}
-
-	hostTestCases := map[string]*testCase{
-		"empty":               {value: "", expect: "127.0.0.1:11434"},
-		"only address":        {value: "1.2.3.4", expect: "1.2.3.4:11434"},
-		"only port":           {value: ":1234", expect: ":1234"},
-		"address and port":    {value: "1.2.3.4:1234", expect: "1.2.3.4:1234"},
-		"hostname":            {value: "example.com", expect: "example.com:11434"},
-		"hostname and port":   {value: "example.com:1234", expect: "example.com:1234"},
-		"zero port":           {value: ":0", expect: ":0"},
-		"too large port":      {value: ":66000", err: ErrInvalidHostPort},
-		"too small port":      {value: ":-1", err: ErrInvalidHostPort},
-		"ipv6 localhost":      {value: "[::1]", expect: "[::1]:11434"},
-		"ipv6 world open":     {value: "[::]", expect: "[::]:11434"},
-		"ipv6 no brackets":    {value: "::1", expect: "[::1]:11434"},
-		"ipv6 + port":         {value: "[::1]:1337", expect: "[::1]:1337"},
-		"extra space":         {value: " 1.2.3.4 ", expect: "1.2.3.4:11434"},
-		"extra quotes":        {value: "\"1.2.3.4\"", expect: "1.2.3.4:11434"},
-		"extra space+quotes":  {value: " \" 1.2.3.4 \" ", expect: "1.2.3.4:11434"},
-		"extra single quotes": {value: "'1.2.3.4'", expect: "1.2.3.4:11434"},
-	}
-
-	for k, v := range hostTestCases {
-		t.Run(k, func(t *testing.T) {
-			t.Setenv("OLLAMA_HOST", v.value)
-
-			oh, err := GetOllamaHost()
-			if err != v.err {
-				t.Fatalf("expected %s, got %s", v.err, err)
-			}
-
-			if err == nil {
-				host := net.JoinHostPort(oh.Host, oh.Port)
-				assert.Equal(t, v.expect, host, fmt.Sprintf("%s: expected %s, got %s", k, v.expect, host))
-			}
-		})
-	}
 }

api/types.go

@@ -4,7 +4,6 @@ import (
 	"encoding/json"
 	"errors"
 	"fmt"
-	"log/slog"
 	"math"
 	"os"
 	"reflect"
@@ -13,7 +12,6 @@ import (
 	"time"
 )
 
-// StatusError is an error with and HTTP status code.
 type StatusError struct {
 	StatusCode   int
 	Status       string
@@ -34,7 +32,6 @@ func (e StatusError) Error() string {
 	}
 }
 
-// ImageData represents the raw binary data of an image file.
 type ImageData []byte
 
 // GenerateRequest describes a request sent by [Client.Generate]. While you
@@ -80,44 +77,26 @@ type GenerateRequest struct {
 	Options map[string]interface{} `json:"options"`
 }
 
-// ChatRequest describes a request sent by [Client.Chat].
 type ChatRequest struct {
-	// Model is the model name, as in [GenerateRequest].
-	Model string `json:"model"`
-
-	// Messages is the messages of the chat - can be used to keep a chat memory.
-	Messages []Message `json:"messages"`
-
-	// Stream enable streaming of returned response; true by default.
-	Stream *bool `json:"stream,omitempty"`
-
-	// Format is the format to return the response in (e.g. "json").
-	Format string `json:"format"`
-
-	// KeepAlive controls how long the model will stay loaded into memory
-	// followin the request.
+	Model     string    `json:"model"`
+	Messages  []Message `json:"messages"`
+	Stream    *bool     `json:"stream,omitempty"`
+	Format    string    `json:"format"`
 	KeepAlive *Duration `json:"keep_alive,omitempty"`
 
-	// Options lists model-specific options.
 	Options map[string]interface{} `json:"options"`
 }
 
-// Message is a single message in a chat sequence. The message contains the
-// role ("system", "user", or "assistant"), the content and an optional list
-// of images.
 type Message struct {
-	Role    string      `json:"role"`
+	Role    string      `json:"role"` // one of ["system", "user", "assistant"]
 	Content string      `json:"content"`
 	Images  []ImageData `json:"images,omitempty"`
 }
 
-// ChatResponse is the response returned by [Client.Chat]. Its fields are
-// similar to [GenerateResponse].
 type ChatResponse struct {
-	Model      string    `json:"model"`
-	CreatedAt  time.Time `json:"created_at"`
-	Message    Message   `json:"message"`
-	DoneReason string    `json:"done_reason,omitempty"`
+	Model     string    `json:"model"`
+	CreatedAt time.Time `json:"created_at"`
+	Message   Message   `json:"message"`
 
 	Done bool `json:"done"`
 
@@ -133,8 +112,7 @@ type Metrics struct {
 	EvalDuration       time.Duration `json:"eval_duration,omitempty"`
 }
 
-// Options specified in [GenerateRequest], if you add a new option here add it
-// to the API docs also.
+// Options specified in GenerateRequest, if you add a new option here add it to the API docs also
 type Options struct {
 	Runner
 
@@ -163,6 +141,7 @@ type Runner struct {
 	UseNUMA   bool `json:"numa,omitempty"`
 	NumCtx    int  `json:"num_ctx,omitempty"`
 	NumBatch  int  `json:"num_batch,omitempty"`
+	NumGQA    int  `json:"num_gqa,omitempty"`
 	NumGPU    int  `json:"num_gpu,omitempty"`
 	MainGPU   int  `json:"main_gpu,omitempty"`
 	LowVRAM   bool `json:"low_vram,omitempty"`
@@ -172,42 +151,27 @@ type Runner struct {
 	UseMMap   bool `json:"use_mmap,omitempty"`
 	UseMLock  bool `json:"use_mlock,omitempty"`
 	NumThread int  `json:"num_thread,omitempty"`
+
+	// Unused: RopeFrequencyBase is ignored. Instead the value in the model will be used
+	RopeFrequencyBase float32 `json:"rope_frequency_base,omitempty"`
+	// Unused: RopeFrequencyScale is ignored. Instead the value in the model will be used
+	RopeFrequencyScale float32 `json:"rope_frequency_scale,omitempty"`
 }
 
-// EmbeddingRequest is the request passed to [Client.Embeddings].
 type EmbeddingRequest struct {
-	// Model is the model name.
-	Model string `json:"model"`
+	Model       string    `json:"model"`
+	Prompt      string    `json:"prompt,omitempty"`
+	PromptBatch []string  `json:"prompt_batch,omitempty"`
+	KeepAlive   *Duration `json:"keep_alive,omitempty"`
 
-	// Prompt is the textual prompt to embed.
-	Prompt string `json:"prompt"`
-
-	// KeepAlive controls how long the model will stay loaded in memory following
-	// this request.
-	KeepAlive *Duration `json:"keep_alive,omitempty"`
-
-	// Options lists model-specific options.
 	Options map[string]interface{} `json:"options"`
 }
 
-// EmbeddingResponse is the response from [Client.Embeddings].
 type EmbeddingResponse struct {
-	Embedding []float64 `json:"embedding"`
+	Embedding      []float64   `json:"embedding,omitempty"`
+	EmbeddingBatch [][]float64 `json:"embedding_batch,omitempty"`
 }
 
-type TokenizeRequest struct {
-	Model     string    `json:"model"`
-	Prompt    string    `json:"prompt"`
-	KeepAlive *Duration `json:"keep_alive,omitempty"`
-
-	Options map[string]interface{} `json:"options"`
-}
-
-type TokenizeResponse struct {
-	Tokens []int `json:"tokens"`
-}
-
-// CreateRequest is the request passed to [Client.Create].
 type CreateRequest struct {
 	Model        string `json:"model"`
 	Path         string `json:"path"`
@@ -219,7 +183,6 @@ type CreateRequest struct {
 	Name string `json:"name"`
 }
 
-// DeleteRequest is the request passed to [Client.Delete].
 type DeleteRequest struct {
 	Model string `json:"model"`
 
@@ -227,7 +190,6 @@ type DeleteRequest struct {
 	Name string `json:"name"`
 }
 
-// ShowRequest is the request passed to [Client.Show].
 type ShowRequest struct {
 	Model    string `json:"model"`
 	System   string `json:"system"`
@@ -239,7 +201,6 @@ type ShowRequest struct {
 	Name string `json:"name"`
 }
 
-// ShowResponse is the response returned from [Client.Show].
 type ShowResponse struct {
 	License    string       `json:"license,omitempty"`
 	Modelfile  string       `json:"modelfile,omitempty"`
@@ -250,13 +211,11 @@ type ShowResponse struct {
 	Messages   []Message    `json:"messages,omitempty"`
 }
 
-// CopyRequest is the request passed to [Client.Copy].
 type CopyRequest struct {
 	Source      string `json:"source"`
 	Destination string `json:"destination"`
 }
 
-// PullRequest is the request passed to [Client.Pull].
 type PullRequest struct {
 	Model    string `json:"model"`
 	Insecure bool   `json:"insecure,omitempty"`
@@ -268,8 +227,6 @@ type PullRequest struct {
 	Name string `json:"name"`
 }
 
-// ProgressResponse is the response passed to progress functions like
-// [PullProgressFunc] and [PushProgressFunc].
 type ProgressResponse struct {
 	Status    string `json:"status"`
 	Digest    string `json:"digest,omitempty"`
@@ -277,7 +234,6 @@ type ProgressResponse struct {
 	Completed int64  `json:"completed,omitempty"`
 }
 
-// PushRequest is the request passed to [Client.Push].
 type PushRequest struct {
 	Model    string `json:"model"`
 	Insecure bool   `json:"insecure,omitempty"`
@@ -289,12 +245,10 @@ type PushRequest struct {
 	Name string `json:"name"`
 }
 
-// ListResponse is the response from [Client.List].
 type ListResponse struct {
 	Models []ModelResponse `json:"models"`
 }
 
-// ModelResponse is a single model description in [ListResponse].
 type ModelResponse struct {
 	Name       string       `json:"name"`
 	Model      string       `json:"model"`
@@ -308,31 +262,17 @@ type TokenResponse struct {
 	Token string `json:"token"`
 }
 
-// GenerateResponse is the response passed into [GenerateResponseFunc].
 type GenerateResponse struct {
-	// Model is the model name that generated the response.
-	Model string `json:"model"`
-
-	//CreatedAt is the timestamp of the response.
+	Model     string    `json:"model"`
 	CreatedAt time.Time `json:"created_at"`
+	Response  string    `json:"response"`
 
-	// Response is the textual response itself.
-	Response string `json:"response"`
-
-	// Done specifies if the response is complete.
-	Done bool `json:"done"`
-
-	// DoneReason is the reason the model stopped generating text.
-	DoneReason string `json:"done_reason,omitempty"`
-
-	// Context is an encoding of the conversation used in this response; this
-	// can be sent in the next request to keep a conversational memory.
+	Done    bool  `json:"done"`
 	Context []int `json:"context,omitempty"`
 
 	Metrics
 }
 
-// ModelDetails provides details about a model.
 type ModelDetails struct {
 	ParentModel       string   `json:"parent_model"`
 	Format            string   `json:"format"`
@@ -370,7 +310,7 @@ func (m *Metrics) Summary() {
 	}
 }
 
-var ErrInvalidHostPort = errors.New("invalid port specified in OLLAMA_HOST")
+var ErrInvalidOpts = errors.New("invalid options")
 
 func (opts *Options) FromMap(m map[string]interface{}) error {
 	valueOpts := reflect.ValueOf(opts).Elem() // names of the fields in the options struct
@@ -385,76 +325,76 @@ func (opts *Options) FromMap(m map[string]interface{}) error {
 		}
 	}
 
+	invalidOpts := []string{}
 	for key, val := range m {
-		opt, ok := jsonOpts[key]
-		if !ok {
-			slog.Warn("invalid option provided", "option", opt.Name)
-			continue
-		}
+		if opt, ok := jsonOpts[key]; ok {
+			field := valueOpts.FieldByName(opt.Name)
+			if field.IsValid() && field.CanSet() {
+				if val == nil {
+					continue
+				}
 
-		field := valueOpts.FieldByName(opt.Name)
-		if field.IsValid() && field.CanSet() {
-			if val == nil {
-				continue
-			}
-
-			switch field.Kind() {
-			case reflect.Int:
-				switch t := val.(type) {
-				case int64:
-					field.SetInt(t)
-				case float64:
-					// when JSON unmarshals numbers, it uses float64, not int
-					field.SetInt(int64(t))
-				default:
-					return fmt.Errorf("option %q must be of type integer", key)
-				}
-			case reflect.Bool:
-				val, ok := val.(bool)
-				if !ok {
-					return fmt.Errorf("option %q must be of type boolean", key)
-				}
-				field.SetBool(val)
-			case reflect.Float32:
-				// JSON unmarshals to float64
-				val, ok := val.(float64)
-				if !ok {
-					return fmt.Errorf("option %q must be of type float32", key)
-				}
-				field.SetFloat(val)
-			case reflect.String:
-				val, ok := val.(string)
-				if !ok {
-					return fmt.Errorf("option %q must be of type string", key)
-				}
-				field.SetString(val)
-			case reflect.Slice:
-				// JSON unmarshals to []interface{}, not []string
-				val, ok := val.([]interface{})
-				if !ok {
-					return fmt.Errorf("option %q must be of type array", key)
-				}
-				// convert []interface{} to []string
-				slice := make([]string, len(val))
-				for i, item := range val {
-					str, ok := item.(string)
-					if !ok {
-						return fmt.Errorf("option %q must be of an array of strings", key)
+				switch field.Kind() {
+				case reflect.Int:
+					switch t := val.(type) {
+					case int64:
+						field.SetInt(t)
+					case float64:
+						// when JSON unmarshals numbers, it uses float64, not int
+						field.SetInt(int64(t))
+					default:
+						return fmt.Errorf("option %q must be of type integer", key)
 					}
-					slice[i] = str
+				case reflect.Bool:
+					val, ok := val.(bool)
+					if !ok {
+						return fmt.Errorf("option %q must be of type boolean", key)
+					}
+					field.SetBool(val)
+				case reflect.Float32:
+					// JSON unmarshals to float64
+					val, ok := val.(float64)
+					if !ok {
+						return fmt.Errorf("option %q must be of type float32", key)
+					}
+					field.SetFloat(val)
+				case reflect.String:
+					val, ok := val.(string)
+					if !ok {
+						return fmt.Errorf("option %q must be of type string", key)
+					}
+					field.SetString(val)
+				case reflect.Slice:
+					// JSON unmarshals to []interface{}, not []string
+					val, ok := val.([]interface{})
+					if !ok {
+						return fmt.Errorf("option %q must be of type array", key)
+					}
+					// convert []interface{} to []string
+					slice := make([]string, len(val))
+					for i, item := range val {
+						str, ok := item.(string)
+						if !ok {
+							return fmt.Errorf("option %q must be of an array of strings", key)
+						}
+						slice[i] = str
+					}
+					field.Set(reflect.ValueOf(slice))
+				default:
+					return fmt.Errorf("unknown type loading config params: %v", field.Kind())
 				}
-				field.Set(reflect.ValueOf(slice))
-			default:
-				return fmt.Errorf("unknown type loading config params: %v", field.Kind())
 			}
+		} else {
+			invalidOpts = append(invalidOpts, key)
 		}
 	}
 
+	if len(invalidOpts) > 0 {
+		return fmt.Errorf("%w: %v", ErrInvalidOpts, strings.Join(invalidOpts, ", "))
+	}
 	return nil
 }
 
-// DefaultOptions is the default set of options for [GenerateRequest]; these
-// values are used unless the user specifies other values explicitly.
 func DefaultOptions() Options {
 	return Options{
 		// options set on request to runner
@@ -482,7 +422,8 @@ func DefaultOptions() Options {
 			NumCtx:    2048,
 			NumBatch:  512,
 			NumGPU:    -1, // -1 here indicates that NumGPU should be set dynamically
-			NumThread: 0,  // let the runtime decide
+			NumGQA:    1,
+			NumThread: 0, // let the runtime decide
 			LowVRAM:   false,
 			F16KV:     true,
 			UseMLock:  false,
@@ -496,13 +437,6 @@ type Duration struct {
 	time.Duration
 }
 
-func (d Duration) MarshalJSON() ([]byte, error) {
-	if d.Duration < 0 {
-		return []byte("-1"), nil
-	}
-	return []byte("\"" + d.Duration.String() + "\""), nil
-}
-
 func (d *Duration) UnmarshalJSON(b []byte) (err error) {
 	var v any
 	if err := json.Unmarshal(b, &v); err != nil {
@@ -516,7 +450,7 @@ func (d *Duration) UnmarshalJSON(b []byte) (err error) {
 		if t < 0 {
 			d.Duration = time.Duration(math.MaxInt64)
 		} else {
-			d.Duration = time.Duration(int(t) * int(time.Second))
+			d.Duration = time.Duration(t * float64(time.Second))
 		}
 	case string:
 		d.Duration, err = time.ParseDuration(t)
@@ -526,8 +460,6 @@ func (d *Duration) UnmarshalJSON(b []byte) (err error) {
 		if d.Duration < 0 {
 			d.Duration = time.Duration(math.MaxInt64)
 		}
-	default:
-		return fmt.Errorf("Unsupported type: '%s'", reflect.TypeOf(v))
 	}
 
 	return nil

api/types_test.go

@@ -21,11 +21,6 @@ func TestKeepAliveParsingFromJSON(t *testing.T) {
 			req:  `{ "keep_alive": 42 }`,
 			exp:  &Duration{42 * time.Second},
 		},
-		{
-			name: "Positive Float",
-			req:  `{ "keep_alive": 42.5 }`,
-			exp:  &Duration{42 * time.Second},
-		},
 		{
 			name: "Positive Integer String",
 			req:  `{ "keep_alive": "42m" }`,
@@ -36,11 +31,6 @@ func TestKeepAliveParsingFromJSON(t *testing.T) {
 			req:  `{ "keep_alive": -1 }`,
 			exp:  &Duration{math.MaxInt64},
 		},
-		{
-			name: "Negative Float",
-			req:  `{ "keep_alive": -3.14 }`,
-			exp:  &Duration{math.MaxInt64},
-		},
 		{
 			name: "Negative Integer String",
 			req:  `{ "keep_alive": "-1m" }`,
@@ -58,50 +48,3 @@ func TestKeepAliveParsingFromJSON(t *testing.T) {
 		})
 	}
 }
-
-func TestDurationMarshalUnmarshal(t *testing.T) {
-	tests := []struct {
-		name     string
-		input    time.Duration
-		expected time.Duration
-	}{
-		{
-			"negative duration",
-			time.Duration(-1),
-			time.Duration(math.MaxInt64),
-		},
-		{
-			"positive duration",
-			time.Duration(42 * time.Second),
-			time.Duration(42 * time.Second),
-		},
-		{
-			"another positive duration",
-			time.Duration(42 * time.Minute),
-			time.Duration(42 * time.Minute),
-		},
-		{
-			"zero duration",
-			time.Duration(0),
-			time.Duration(0),
-		},
-		{
-			"max duration",
-			time.Duration(math.MaxInt64),
-			time.Duration(math.MaxInt64),
-		},
-	}
-
-	for _, test := range tests {
-		t.Run(test.name, func(t *testing.T) {
-			b, err := json.Marshal(Duration{test.input})
-			require.NoError(t, err)
-
-			var d Duration
-			err = json.Unmarshal(b, &d)
-			require.NoError(t, err)
-
-			assert.Equal(t, test.expected, d.Duration, "input %v, marshalled %v, got %v", test.input, string(b), d.Duration)
-		})
-	}
-}

app/lifecycle/logging.go

@@ -5,14 +5,12 @@ import (
 	"log/slog"
 	"os"
 	"path/filepath"
-
-	"github.com/ollama/ollama/server/envconfig"
 )
 
 func InitLogging() {
 	level := slog.LevelInfo
 
-	if envconfig.Debug {
+	if debug := os.Getenv("OLLAMA_DEBUG"); debug != "" {
 		level = slog.LevelDebug
 	}
 

app/lifecycle/server.go

@@ -43,36 +43,37 @@ func getCLIFullPath(command string) string {
 	return command
 }
 
-func start(ctx context.Context, command string) (*exec.Cmd, error) {
+func SpawnServer(ctx context.Context, command string) (chan int, error) {
+	done := make(chan int)
+
+	logDir := filepath.Dir(ServerLogFile)
+	_, err := os.Stat(logDir)
+	if errors.Is(err, os.ErrNotExist) {
+		if err := os.MkdirAll(logDir, 0o755); err != nil {
+			return done, fmt.Errorf("create ollama server log dir %s: %v", logDir, err)
+		}
+	}
+
 	cmd := getCmd(ctx, getCLIFullPath(command))
+	// send stdout and stderr to a file
 	stdout, err := cmd.StdoutPipe()
 	if err != nil {
-		return nil, fmt.Errorf("failed to spawn server stdout pipe: %w", err)
+		return done, fmt.Errorf("failed to spawn server stdout pipe %s", err)
 	}
 	stderr, err := cmd.StderrPipe()
 	if err != nil {
-		return nil, fmt.Errorf("failed to spawn server stderr pipe: %w", err)
+		return done, fmt.Errorf("failed to spawn server stderr pipe %s", err)
+	}
+	stdin, err := cmd.StdinPipe()
+	if err != nil {
+		return done, fmt.Errorf("failed to spawn server stdin pipe %s", err)
 	}
 
 	// TODO - rotation
 	logFile, err := os.OpenFile(ServerLogFile, os.O_APPEND|os.O_WRONLY|os.O_CREATE, 0755)
 	if err != nil {
-		return nil, fmt.Errorf("failed to create server log: %w", err)
+		return done, fmt.Errorf("failed to create server log %w", err)
 	}
-
-	logDir := filepath.Dir(ServerLogFile)
-	_, err = os.Stat(logDir)
-	if err != nil {
-		if !errors.Is(err, os.ErrNotExist) {
-			return nil, fmt.Errorf("stat ollama server log dir %s: %v", logDir, err)
-
-		}
-
-		if err := os.MkdirAll(logDir, 0o755); err != nil {
-			return nil, fmt.Errorf("create ollama server log dir %s: %v", logDir, err)
-		}
-	}
-
 	go func() {
 		defer logFile.Close()
 		io.Copy(logFile, stdout) //nolint:errcheck
@@ -116,33 +117,19 @@ func start(ctx context.Context, command string) (*exec.Cmd, error) {
 
 	// run the command and wait for it to finish
 	if err := cmd.Start(); err != nil {
-		return nil, fmt.Errorf("failed to start server %w", err)
+		return done, fmt.Errorf("failed to start server %w", err)
 	}
 	if cmd.Process != nil {
 		slog.Info(fmt.Sprintf("started ollama server with pid %d", cmd.Process.Pid))
 	}
 	slog.Info(fmt.Sprintf("ollama server logs %s", ServerLogFile))
 
-	return cmd, nil
-}
-
-func SpawnServer(ctx context.Context, command string) (chan int, error) {
-	done := make(chan int)
-
 	go func() {
 		// Keep the server running unless we're shuttind down the app
 		crashCount := 0
 		for {
-			slog.Info("starting server...")
-			cmd, err := start(ctx, command)
-			if err != nil {
-				crashCount++
-				slog.Error(fmt.Sprintf("failed to start server %s", err))
-				time.Sleep(500 * time.Millisecond * time.Duration(crashCount))
-				continue
-			}
-
 			cmd.Wait() //nolint:errcheck
+			stdin.Close()
 			var code int
 			if cmd.ProcessState != nil {
 				code = cmd.ProcessState.ExitCode()
@@ -156,12 +143,15 @@ func SpawnServer(ctx context.Context, command string) (chan int, error) {
 			default:
 				crashCount++
 				slog.Warn(fmt.Sprintf("server crash %d - exit code %d - respawning", crashCount, code))
-				time.Sleep(500 * time.Millisecond * time.Duration(crashCount))
-				break
+				time.Sleep(500 * time.Millisecond)
+				if err := cmd.Start(); err != nil {
+					slog.Error(fmt.Sprintf("failed to restart server %s", err))
+					// Keep trying, but back off if we keep failing
+					time.Sleep(time.Duration(crashCount) * time.Second)
+				}
 			}
 		}
 	}()
-
 	return done, nil
 }
 

app/lifecycle/updater_windows.go

@@ -31,13 +31,16 @@ func DoUpgrade(cancel context.CancelFunc, done chan int) error {
 		"/LOG=" + filepath.Base(UpgradeLogFile), // Only relative seems reliable, so set pwd
 		"/FORCECLOSEAPPLICATIONS",               // Force close the tray app - might be needed
 	}
-	// make the upgrade as quiet as possible (no GUI, no prompts)
+	// When we're not in debug mode, make the upgrade as quiet as possible (no GUI, no prompts)
+	// TODO - temporarily disable since we're pinning in debug mode for the preview
+	// if debug := os.Getenv("OLLAMA_DEBUG"); debug == "" {
 	installArgs = append(installArgs,
 		"/SP", // Skip the "This will install... Do you wish to continue" prompt
 		"/SUPPRESSMSGBOXES",
 		"/SILENT",
 		"/VERYSILENT",
 	)
+	// }
 
 	// Safeguard in case we have requests in flight that need to drain...
 	slog.Info("Waiting for server to shutdown")

app/ollama.iss

@@ -88,8 +88,8 @@ DialogFontSize=12
 [Files]
 Source: ".\app.exe"; DestDir: "{app}"; DestName: "{#MyAppExeName}" ; Flags: ignoreversion 64bit
 Source: "..\ollama.exe"; DestDir: "{app}"; Flags: ignoreversion 64bit
-Source: "..\dist\windows-{#ARCH}\*.dll"; DestDir: "{app}"; Flags: ignoreversion 64bit
-Source: "..\dist\windows-{#ARCH}\ollama_runners\*"; DestDir: "{app}\ollama_runners"; Flags: ignoreversion 64bit recursesubdirs
+Source: "..\dist\windows-amd64\*.dll"; DestDir: "{app}"; Flags: ignoreversion 64bit
+Source: "..\dist\windows-amd64\ollama_runners\*"; DestDir: "{app}\ollama_runners"; Flags: ignoreversion 64bit recursesubdirs
 Source: "..\dist\ollama_welcome.ps1"; DestDir: "{app}"; Flags: ignoreversion
 Source: ".\assets\app.ico"; DestDir: "{app}"; Flags: ignoreversion
 #if DirExists("..\dist\windows-amd64\rocm")

app/tray/wintray/menus.go

@@ -1,71 +1,71 @@
-//go:build windows
-
-package wintray
-
-import (
-	"fmt"
-	"log/slog"
-	"unsafe"
-
-	"golang.org/x/sys/windows"
-)
-
-const (
-	updatAvailableMenuID = 1
-	updateMenuID         = updatAvailableMenuID + 1
-	separatorMenuID      = updateMenuID + 1
-	diagLogsMenuID       = separatorMenuID + 1
-	diagSeparatorMenuID  = diagLogsMenuID + 1
-	quitMenuID           = diagSeparatorMenuID + 1
-)
-
-func (t *winTray) initMenus() error {
-	if err := t.addOrUpdateMenuItem(diagLogsMenuID, 0, diagLogsMenuTitle, false); err != nil {
-		return fmt.Errorf("unable to create menu entries %w\n", err)
-	}
-	if err := t.addSeparatorMenuItem(diagSeparatorMenuID, 0); err != nil {
-		return fmt.Errorf("unable to create menu entries %w", err)
-	}
-	if err := t.addOrUpdateMenuItem(quitMenuID, 0, quitMenuTitle, false); err != nil {
-		return fmt.Errorf("unable to create menu entries %w\n", err)
-	}
-	return nil
-}
-
-func (t *winTray) UpdateAvailable(ver string) error {
-	if !t.updateNotified {
-		slog.Debug("updating menu and sending notification for new update")
-		if err := t.addOrUpdateMenuItem(updatAvailableMenuID, 0, updateAvailableMenuTitle, true); err != nil {
-			return fmt.Errorf("unable to create menu entries %w", err)
-		}
-		if err := t.addOrUpdateMenuItem(updateMenuID, 0, updateMenutTitle, false); err != nil {
-			return fmt.Errorf("unable to create menu entries %w", err)
-		}
-		if err := t.addSeparatorMenuItem(separatorMenuID, 0); err != nil {
-			return fmt.Errorf("unable to create menu entries %w", err)
-		}
-		iconFilePath, err := iconBytesToFilePath(wt.updateIcon)
-		if err != nil {
-			return fmt.Errorf("unable to write icon data to temp file: %w", err)
-		}
-		if err := wt.setIcon(iconFilePath); err != nil {
-			return fmt.Errorf("unable to set icon: %w", err)
-		}
-		t.updateNotified = true
-
-		t.pendingUpdate = true
-		// Now pop up the notification
-		t.muNID.Lock()
-		defer t.muNID.Unlock()
-		copy(t.nid.InfoTitle[:], windows.StringToUTF16(updateTitle))
-		copy(t.nid.Info[:], windows.StringToUTF16(fmt.Sprintf(updateMessage, ver)))
-		t.nid.Flags |= NIF_INFO
-		t.nid.Timeout = 10
-		t.nid.Size = uint32(unsafe.Sizeof(*wt.nid))
-		err = t.nid.modify()
-		if err != nil {
-			return err
-		}
-	}
-	return nil
-}
+//go:build windows
+
+package wintray
+
+import (
+	"fmt"
+	"log/slog"
+	"unsafe"
+
+	"golang.org/x/sys/windows"
+)
+
+const (
+	updatAvailableMenuID = 1
+	updateMenuID         = updatAvailableMenuID + 1
+	separatorMenuID      = updateMenuID + 1
+	diagLogsMenuID       = separatorMenuID + 1
+	diagSeparatorMenuID  = diagLogsMenuID + 1
+	quitMenuID           = diagSeparatorMenuID + 1
+)
+
+func (t *winTray) initMenus() error {
+	if err := t.addOrUpdateMenuItem(diagLogsMenuID, 0, diagLogsMenuTitle, false); err != nil {
+		return fmt.Errorf("unable to create menu entries %w\n", err)
+	}
+	if err := t.addSeparatorMenuItem(diagSeparatorMenuID, 0); err != nil {
+		return fmt.Errorf("unable to create menu entries %w", err)
+	}
+	if err := t.addOrUpdateMenuItem(quitMenuID, 0, quitMenuTitle, false); err != nil {
+		return fmt.Errorf("unable to create menu entries %w\n", err)
+	}
+	return nil
+}
+
+func (t *winTray) UpdateAvailable(ver string) error {
+	if !t.updateNotified {
+		slog.Debug("updating menu and sending notification for new update")
+		if err := t.addOrUpdateMenuItem(updatAvailableMenuID, 0, updateAvailableMenuTitle, true); err != nil {
+			return fmt.Errorf("unable to create menu entries %w", err)
+		}
+		if err := t.addOrUpdateMenuItem(updateMenuID, 0, updateMenutTitle, false); err != nil {
+			return fmt.Errorf("unable to create menu entries %w", err)
+		}
+		if err := t.addSeparatorMenuItem(separatorMenuID, 0); err != nil {
+			return fmt.Errorf("unable to create menu entries %w", err)
+		}
+		iconFilePath, err := iconBytesToFilePath(wt.updateIcon)
+		if err != nil {
+			return fmt.Errorf("unable to write icon data to temp file: %w", err)
+		}
+		if err := wt.setIcon(iconFilePath); err != nil {
+			return fmt.Errorf("unable to set icon: %w", err)
+		}
+		t.updateNotified = true
+
+		t.pendingUpdate = true
+		// Now pop up the notification
+		t.muNID.Lock()
+		defer t.muNID.Unlock()
+		copy(t.nid.InfoTitle[:], windows.StringToUTF16(updateTitle))
+		copy(t.nid.Info[:], windows.StringToUTF16(fmt.Sprintf(updateMessage, ver)))
+		t.nid.Flags |= NIF_INFO
+		t.nid.Timeout = 10
+		t.nid.Size = uint32(unsafe.Sizeof(*wt.nid))
+		err = t.nid.modify()
+		if err != nil {
+			return err
+		}
+	}
+	return nil
+}

auth/auth.go

@@ -10,44 +10,12 @@ import (
 	"log/slog"
 	"os"
 	"path/filepath"
-	"strings"
 
 	"golang.org/x/crypto/ssh"
 )
 
 const defaultPrivateKey = "id_ed25519"
 
-func keyPath() (string, error) {
-	home, err := os.UserHomeDir()
-	if err != nil {
-		return "", err
-	}
-
-	return filepath.Join(home, ".ollama", defaultPrivateKey), nil
-}
-
-func GetPublicKey() (string, error) {
-	keyPath, err := keyPath()
-	if err != nil {
-		return "", err
-	}
-
-	privateKeyFile, err := os.ReadFile(keyPath)
-	if err != nil {
-		slog.Info(fmt.Sprintf("Failed to load private key: %v", err))
-		return "", err
-	}
-
-	privateKey, err := ssh.ParsePrivateKey(privateKeyFile)
-	if err != nil {
-		return "", err
-	}
-
-	publicKey := ssh.MarshalAuthorizedKey(privateKey.PublicKey())
-
-	return strings.TrimSpace(string(publicKey)), nil
-}
-
 func NewNonce(r io.Reader, length int) (string, error) {
 	nonce := make([]byte, length)
 	if _, err := io.ReadFull(r, nonce); err != nil {
@@ -58,11 +26,13 @@ func NewNonce(r io.Reader, length int) (string, error) {
 }
 
 func Sign(ctx context.Context, bts []byte) (string, error) {
-	keyPath, err := keyPath()
+	home, err := os.UserHomeDir()
 	if err != nil {
 		return "", err
 	}
 
+	keyPath := filepath.Join(home, ".ollama", defaultPrivateKey)
+
 	privateKeyFile, err := os.ReadFile(keyPath)
 	if err != nil {
 		slog.Info(fmt.Sprintf("Failed to load private key: %v", err))

cmd/cmd.go

@@ -32,12 +32,10 @@ import (
 	"golang.org/x/term"
 
 	"github.com/ollama/ollama/api"
-	"github.com/ollama/ollama/auth"
 	"github.com/ollama/ollama/format"
+	"github.com/ollama/ollama/parser"
 	"github.com/ollama/ollama/progress"
 	"github.com/ollama/ollama/server"
-	"github.com/ollama/ollama/types/errtypes"
-	"github.com/ollama/ollama/types/model"
 	"github.com/ollama/ollama/version"
 )
 
@@ -56,13 +54,12 @@ func CreateHandler(cmd *cobra.Command, args []string) error {
 	p := progress.NewProgress(os.Stderr)
 	defer p.Stop()
 
-	f, err := os.Open(filename)
+	modelfile, err := os.ReadFile(filename)
 	if err != nil {
 		return err
 	}
-	defer f.Close()
 
-	modelfile, err := model.ParseFile(f)
+	commands, err := parser.Parse(bytes.NewReader(modelfile))
 	if err != nil {
 		return err
 	}
@@ -76,10 +73,10 @@ func CreateHandler(cmd *cobra.Command, args []string) error {
 	spinner := progress.NewSpinner(status)
 	p.Add(status, spinner)
 
-	for i := range modelfile.Commands {
-		switch modelfile.Commands[i].Name {
+	for _, c := range commands {
+		switch c.Name {
 		case "model", "adapter":
-			path := modelfile.Commands[i].Args
+			path := c.Args
 			if path == "~" {
 				path = home
 			} else if strings.HasPrefix(path, "~/") {
@@ -91,7 +88,7 @@ func CreateHandler(cmd *cobra.Command, args []string) error {
 			}
 
 			fi, err := os.Stat(path)
-			if errors.Is(err, os.ErrNotExist) && modelfile.Commands[i].Name == "model" {
+			if errors.Is(err, os.ErrNotExist) && c.Name == "model" {
 				continue
 			} else if err != nil {
 				return err
@@ -114,7 +111,13 @@ func CreateHandler(cmd *cobra.Command, args []string) error {
 				return err
 			}
 
-			modelfile.Commands[i].Args = "@" + digest
+			name := c.Name
+			if c.Name == "model" {
+				name = "from"
+			}
+
+			re := regexp.MustCompile(fmt.Sprintf(`(?im)^(%s)\s+%s\s*$`, name, c.Args))
+			modelfile = re.ReplaceAll(modelfile, []byte("$1 @"+digest))
 		}
 	}
 
@@ -144,7 +147,7 @@ func CreateHandler(cmd *cobra.Command, args []string) error {
 
 	quantization, _ := cmd.Flags().GetString("quantization")
 
-	request := api.CreateRequest{Name: args[0], Modelfile: modelfile.String(), Quantization: quantization}
+	request := api.CreateRequest{Name: args[0], Modelfile: string(modelfile), Quantization: quantization}
 	if err := client.Create(cmd.Context(), &request, fn); err != nil {
 		return err
 	}
@@ -354,47 +357,6 @@ func RunHandler(cmd *cobra.Command, args []string) error {
 	return generateInteractive(cmd, opts)
 }
 
-func errFromUnknownKey(unknownKeyErr error) error {
-	// find SSH public key in the error message
-	sshKeyPattern := `ssh-\w+ [^\s"]+`
-	re := regexp.MustCompile(sshKeyPattern)
-	matches := re.FindStringSubmatch(unknownKeyErr.Error())
-
-	if len(matches) > 0 {
-		serverPubKey := matches[0]
-
-		localPubKey, err := auth.GetPublicKey()
-		if err != nil {
-			return unknownKeyErr
-		}
-
-		if runtime.GOOS == "linux" && serverPubKey != localPubKey {
-			// try the ollama service public key
-			svcPubKey, err := os.ReadFile("/usr/share/ollama/.ollama/id_ed25519.pub")
-			if err != nil {
-				return unknownKeyErr
-			}
-			localPubKey = strings.TrimSpace(string(svcPubKey))
-		}
-
-		// check if the returned public key matches the local public key, this prevents adding a remote key to the user's account
-		if serverPubKey != localPubKey {
-			return unknownKeyErr
-		}
-
-		var msg strings.Builder
-		msg.WriteString(unknownKeyErr.Error())
-		msg.WriteString("\n\nYour ollama key is:\n")
-		msg.WriteString(localPubKey)
-		msg.WriteString("\nAdd your key at:\n")
-		msg.WriteString("https://ollama.com/settings/keys")
-
-		return errors.New(msg.String())
-	}
-
-	return unknownKeyErr
-}
-
 func PushHandler(cmd *cobra.Command, args []string) error {
 	client, err := api.ClientFromEnvironment()
 	if err != nil {
@@ -442,20 +404,6 @@ func PushHandler(cmd *cobra.Command, args []string) error {
 
 	request := api.PushRequest{Name: args[0], Insecure: insecure}
 	if err := client.Push(cmd.Context(), &request, fn); err != nil {
-		if spinner != nil {
-			spinner.Stop()
-		}
-		if strings.Contains(err.Error(), "access denied") {
-			return errors.New("you are not authorized to push to this namespace, create the model under a namespace you own")
-		}
-		host := model.ParseName(args[0]).Host
-		isOllamaHost := strings.HasSuffix(host, ".ollama.ai") || strings.HasSuffix(host, ".ollama.com")
-		if strings.Contains(err.Error(), errtypes.UnknownOllamaKeyErrMsg) && isOllamaHost {
-			// the user has not added their ollama key to ollama.com
-			// re-throw an error with a more user-friendly message
-			return errFromUnknownKey(err)
-		}
-
 		return err
 	}
 
@@ -883,27 +831,24 @@ func generate(cmd *cobra.Command, opts runOptions) error {
 }
 
 func RunServer(cmd *cobra.Command, _ []string) error {
-	// retrieve the OLLAMA_HOST environment variable
-	ollamaHost, err := api.GetOllamaHost()
+	host, port, err := net.SplitHostPort(strings.Trim(os.Getenv("OLLAMA_HOST"), "\"'"))
 	if err != nil {
-		return err
+		host, port = "127.0.0.1", "11434"
+		if ip := net.ParseIP(strings.Trim(os.Getenv("OLLAMA_HOST"), "[]")); ip != nil {
+			host = ip.String()
+		}
 	}
 
 	if err := initializeKeypair(); err != nil {
 		return err
 	}
 
-	ln, err := net.Listen("tcp", net.JoinHostPort(ollamaHost.Host, ollamaHost.Port))
+	ln, err := net.Listen("tcp", net.JoinHostPort(host, port))
 	if err != nil {
 		return err
 	}
 
-	err = server.Serve(ln)
-	if errors.Is(err, http.ErrServerClosed) {
-		return nil
-	}
-
-	return err
+	return server.Serve(ln)
 }
 
 func initializeKeypair() error {
@@ -1124,7 +1069,7 @@ Environment Variables:
 		RunE:    ListHandler,
 	}
 	copyCmd := &cobra.Command{
-		Use:     "cp SOURCE DESTINATION",
+		Use:     "cp SOURCE TARGET",
 		Short:   "Copy a model",
 		Args:    cobra.ExactArgs(2),
 		PreRunE: checkServerHeartbeat,

cmd/interactive.go

@@ -94,7 +94,6 @@ func generateInteractive(cmd *cobra.Command, opts runOptions) error {
 		fmt.Fprintln(os.Stderr, "  /show           Show model information")
 		fmt.Fprintln(os.Stderr, "  /load <model>   Load a session or model")
 		fmt.Fprintln(os.Stderr, "  /save <model>   Save your current session")
-		fmt.Fprintln(os.Stderr, "  /clear          Clear session context")
 		fmt.Fprintln(os.Stderr, "  /bye            Exit")
 		fmt.Fprintln(os.Stderr, "  /?, /help       Help for a command")
 		fmt.Fprintln(os.Stderr, "  /? shortcuts    Help for keyboard shortcuts")
@@ -162,7 +161,7 @@ func generateInteractive(cmd *cobra.Command, opts runOptions) error {
 		fmt.Fprintln(os.Stderr, "  /set parameter repeat_penalty <float> How strongly to penalize repetitions")
 		fmt.Fprintln(os.Stderr, "  /set parameter repeat_last_n <int>    Set how far back to look for repetitions")
 		fmt.Fprintln(os.Stderr, "  /set parameter num_gpu <int>          The number of layers to send to the GPU")
-		fmt.Fprintln(os.Stderr, "  /set parameter stop <string> <string> ...   Set the stop parameters")
+		fmt.Fprintln(os.Stderr, "  /set parameter stop \"<string>\", ...   Set the stop parameters")
 		fmt.Fprintln(os.Stderr, "")
 	}
 
@@ -281,10 +280,6 @@ func generateInteractive(cmd *cobra.Command, opts runOptions) error {
 			}
 			fmt.Printf("Created new model '%s'\n", args[1])
 			continue
-		case strings.HasPrefix(line, "/clear"):
-			opts.Messages = []api.Message{}
-			fmt.Println("Cleared session context")
-			continue
 		case strings.HasPrefix(line, "/set"):
 			args := strings.Fields(line)
 			if len(args) > 1 {

convert/convert.go

@@ -5,7 +5,6 @@ import (
 	"encoding/binary"
 	"encoding/json"
 	"fmt"
-	"io"
 	"log/slog"
 	"os"
 	"path/filepath"
@@ -48,7 +47,7 @@ type ByteOrder interface {
 type ModelArch interface {
 	GetTensors() error
 	LoadVocab() error
-	WriteGGUF(io.WriteSeeker) error
+	WriteGGUF() (string, error)
 }
 
 type ModelFormat interface {

convert/gemma.go

@@ -94,7 +94,7 @@ func (m *GemmaModel) LoadVocab() error {
 	return nil
 }
 
-func (m *GemmaModel) WriteGGUF(ws io.WriteSeeker) error {
+func (m *GemmaModel) WriteGGUF() (string, error) {
 	kv := llm.KV{
 		"general.architecture":                   "gemma",
 		"general.name":                           m.Name,
@@ -122,5 +122,16 @@ func (m *GemmaModel) WriteGGUF(ws io.WriteSeeker) error {
 		"tokenizer.ggml.add_eos_token":    false,
 	}
 
-	return llm.NewGGUFV3(m.Params.ByteOrder).Encode(ws, kv, m.Tensors)
+	f, err := os.CreateTemp("", "ollama-gguf")
+	if err != nil {
+		return "", err
+	}
+	defer f.Close()
+
+	mod := llm.NewGGUFV3(m.Params.ByteOrder)
+	if err := mod.Encode(f, kv, m.Tensors); err != nil {
+		return "", err
+	}
+
+	return f.Name(), nil
 }

convert/llama.go

@@ -5,6 +5,7 @@ import (
 	"fmt"
 	"io"
 	"log/slog"
+	"os"
 	"regexp"
 	"strings"
 
@@ -131,7 +132,7 @@ func (m *LlamaModel) LoadVocab() error {
 	return nil
 }
 
-func (m *LlamaModel) WriteGGUF(ws io.WriteSeeker) error {
+func (m *LlamaModel) WriteGGUF() (string, error) {
 	kv := llm.KV{
 		"general.architecture":                   "llama",
 		"general.name":                           m.Name,
@@ -158,5 +159,18 @@ func (m *LlamaModel) WriteGGUF(ws io.WriteSeeker) error {
 		"tokenizer.ggml.add_eos_token":    false,
 	}
 
-	return llm.NewGGUFV3(m.Params.ByteOrder).Encode(ws, kv, m.Tensors)
+	f, err := os.CreateTemp("", "ollama-gguf")
+	if err != nil {
+		return "", err
+	}
+	defer f.Close()
+
+	mod := llm.NewGGUFV3(m.Params.ByteOrder)
+	if err := mod.Encode(f, kv, m.Tensors); err != nil {
+		return "", err
+	}
+
+	slog.Debug(fmt.Sprintf("gguf file = %s", f.Name()))
+
+	return f.Name(), nil
 }

convert/mistral.go

@@ -132,7 +132,7 @@ func (m *MistralModel) LoadVocab() error {
 	return nil
 }
 
-func (m *MistralModel) WriteGGUF(ws io.WriteSeeker) error {
+func (m *MistralModel) WriteGGUF() (string, error) {
 	kv := llm.KV{
 		"general.architecture":                   "llama",
 		"general.name":                           m.Name,
@@ -158,5 +158,16 @@ func (m *MistralModel) WriteGGUF(ws io.WriteSeeker) error {
 		"tokenizer.ggml.unknown_token_id": uint32(0),
 	}
 
-	return llm.NewGGUFV3(m.Params.ByteOrder).Encode(ws, kv, m.Tensors)
+	f, err := os.CreateTemp("", "ollama-gguf")
+	if err != nil {
+		return "", err
+	}
+	defer f.Close()
+
+	mod := llm.NewGGUFV3(m.Params.ByteOrder)
+	if err := mod.Encode(f, kv, m.Tensors); err != nil {
+		return "", err
+	}
+
+	return f.Name(), nil
 }

convert/mixtral.go

@@ -1,7 +1,7 @@
 package convert
 
 import (
-	"io"
+	"os"
 	"regexp"
 
 	"github.com/ollama/ollama/llm"
@@ -47,7 +47,7 @@ func (m *MixtralModel) LoadVocab() error {
 	return nil
 }
 
-func (m *MixtralModel) WriteGGUF(ws io.WriteSeeker) error {
+func (m *MixtralModel) WriteGGUF() (string, error) {
 	kv := llm.KV{
 		"general.architecture":          "llama",
 		"general.name":                  m.Name,
@@ -81,5 +81,16 @@ func (m *MixtralModel) WriteGGUF(ws io.WriteSeeker) error {
 		"tokenizer.ggml.add_eos_token":    false,
 	}
 
-	return llm.NewGGUFV3(m.Params.ByteOrder).Encode(ws, kv, m.Tensors)
+	f, err := os.CreateTemp("", "ollama-gguf")
+	if err != nil {
+		return "", err
+	}
+	defer f.Close()
+
+	mod := llm.NewGGUFV3(m.Params.ByteOrder)
+	if err := mod.Encode(f, kv, m.Tensors); err != nil {
+		return "", err
+	}
+
+	return f.Name(), nil
 }

convert/safetensors.go

@@ -53,7 +53,7 @@ func (m *SafetensorFormat) GetTensors(dirpath string, params *Params) ([]llm.Ten
 		var err error
 		t, offset, err = m.readTensors(f, offset, params)
 		if err != nil {
-			slog.Error(err.Error())
+			slog.Error("%v", err)
 			return nil, err
 		}
 		tensors = append(tensors, t...)
@@ -122,7 +122,7 @@ func (m *SafetensorFormat) readTensors(fn string, offset uint64, params *Params)
 
 		ggufName, err := m.GetLayerName(k)
 		if err != nil {
-			slog.Error(err.Error())
+			slog.Error("%v", err)
 			return nil, 0, err
 		}
 

convert/torch.go

@@ -74,7 +74,7 @@ func (tf *TorchFormat) GetTensors(dirpath string, params *Params) ([]llm.Tensor,
 
 			ggufName, err := tf.GetLayerName(k.(string))
 			if err != nil {
-				slog.Error(err.Error())
+				slog.Error("%v", err)
 				return nil, err
 			}
 			slog.Debug(fmt.Sprintf("finding name for '%s' -> '%s'", k.(string), ggufName))

docs/README.md

@@ -6,7 +6,7 @@
 * [Importing models](./import.md)
 * [Linux Documentation](./linux.md)
 * [Windows Documentation](./windows.md)
-* [Docker Documentation](./docker.md)
+* [Docker Documentation](https://hub.docker.com/r/ollama/ollama)
 
 ### Reference
 

docs/api.md

@@ -17,7 +17,7 @@
 
 ### Model names
 
-Model names follow a `model:tag` format, where `model` can have an optional namespace such as `example/model`. Some examples are `orca-mini:3b-q4_1` and `llama3:70b`. The tag is optional and, if not provided, will default to `latest`. The tag is used to identify a specific version.
+Model names follow a `model:tag` format, where `model` can have an optional namespace such as `example/model`. Some examples are `orca-mini:3b-q4_1` and `llama2:70b`. The tag is optional and, if not provided, will default to `latest`. The tag is used to identify a specific version.
 
 ### Durations
 
@@ -66,7 +66,7 @@ Enable JSON mode by setting the `format` parameter to `json`. This will structur
 
 ```shell
 curl http://localhost:11434/api/generate -d '{
-  "model": "llama3",
+  "model": "llama2",
   "prompt": "Why is the sky blue?"
 }'
 ```
@@ -77,7 +77,7 @@ A stream of JSON objects is returned:
 
 ```json
 {
-  "model": "llama3",
+  "model": "llama2",
   "created_at": "2023-08-04T08:52:19.385406455-07:00",
   "response": "The",
   "done": false
@@ -95,11 +95,11 @@ The final response in the stream also includes additional data about the generat
 - `context`: an encoding of the conversation used in this response, this can be sent in the next request to keep a conversational memory
 - `response`: empty if the response was streamed, if not streamed, this will contain the full response
 
-To calculate how fast the response is generated in tokens per second (token/s), divide `eval_count` / `eval_duration` * `10^9`.
+To calculate how fast the response is generated in tokens per second (token/s), divide `eval_count` / `eval_duration`.
 
 ```json
 {
-  "model": "llama3",
+  "model": "llama2",
   "created_at": "2023-08-04T19:22:45.499127Z",
   "response": "",
   "done": true,
@@ -121,7 +121,7 @@ A response can be received in one reply when streaming is off.
 
 ```shell
 curl http://localhost:11434/api/generate -d '{
-  "model": "llama3",
+  "model": "llama2",
   "prompt": "Why is the sky blue?",
   "stream": false
 }'
@@ -133,7 +133,7 @@ If `stream` is set to `false`, the response will be a single JSON object:
 
 ```json
 {
-  "model": "llama3",
+  "model": "llama2",
   "created_at": "2023-08-04T19:22:45.499127Z",
   "response": "The sky is blue because it is the color of the sky.",
   "done": true,
@@ -155,7 +155,7 @@ If `stream` is set to `false`, the response will be a single JSON object:
 
 ```shell
 curl http://localhost:11434/api/generate -d '{
-  "model": "llama3",
+  "model": "llama2",
   "prompt": "What color is the sky at different times of the day? Respond using JSON",
   "format": "json",
   "stream": false
@@ -166,7 +166,7 @@ curl http://localhost:11434/api/generate -d '{
 
 ```json
 {
-  "model": "llama3",
+  "model": "llama2",
   "created_at": "2023-11-09T21:07:55.186497Z",
   "response": "{\n\"morning\": {\n\"color\": \"blue\"\n},\n\"noon\": {\n\"color\": \"blue-gray\"\n},\n\"afternoon\": {\n\"color\": \"warm gray\"\n},\n\"evening\": {\n\"color\": \"orange\"\n}\n}\n",
   "done": true,
@@ -289,7 +289,7 @@ If you want to set custom options for the model at runtime rather than in the Mo
 
 ```shell
 curl http://localhost:11434/api/generate -d '{
-  "model": "llama3",
+  "model": "llama2",
   "prompt": "Why is the sky blue?",
   "stream": false,
   "options": {
@@ -313,6 +313,7 @@ curl http://localhost:11434/api/generate -d '{
     "numa": false,
     "num_ctx": 1024,
     "num_batch": 2,
+    "num_gqa": 1,
     "num_gpu": 1,
     "main_gpu": 0,
     "low_vram": false,
@@ -320,6 +321,8 @@ curl http://localhost:11434/api/generate -d '{
     "vocab_only": false,
     "use_mmap": true,
     "use_mlock": false,
+    "rope_frequency_base": 1.1,
+    "rope_frequency_scale": 0.8,
     "num_thread": 8
   }
 }'
@@ -329,7 +332,7 @@ curl http://localhost:11434/api/generate -d '{
 
 ```json
 {
-  "model": "llama3",
+  "model": "llama2",
   "created_at": "2023-08-04T19:22:45.499127Z",
   "response": "The sky is blue because it is the color of the sky.",
   "done": true,
@@ -351,7 +354,7 @@ If an empty prompt is provided, the model will be loaded into memory.
 
 ```shell
 curl http://localhost:11434/api/generate -d '{
-  "model": "llama3"
+  "model": "llama2"
 }'
 ```
 
@@ -361,7 +364,7 @@ A single JSON object is returned:
 
 ```json
 {
-  "model": "llama3",
+  "model": "llama2",
   "created_at": "2023-12-18T19:52:07.071755Z",
   "response": "",
   "done": true
@@ -404,7 +407,7 @@ Send a chat message with a streaming response.
 
 ```shell
 curl http://localhost:11434/api/chat -d '{
-  "model": "llama3",
+  "model": "llama2",
   "messages": [
     {
       "role": "user",
@@ -420,7 +423,7 @@ A stream of JSON objects is returned:
 
 ```json
 {
-  "model": "llama3",
+  "model": "llama2",
   "created_at": "2023-08-04T08:52:19.385406455-07:00",
   "message": {
     "role": "assistant",
@@ -435,7 +438,7 @@ Final response:
 
 ```json
 {
-  "model": "llama3",
+  "model": "llama2",
   "created_at": "2023-08-04T19:22:45.499127Z",
   "done": true,
   "total_duration": 4883583458,
@@ -453,7 +456,7 @@ Final response:
 
 ```shell
 curl http://localhost:11434/api/chat -d '{
-  "model": "llama3",
+  "model": "llama2",
   "messages": [
     {
       "role": "user",
@@ -468,7 +471,7 @@ curl http://localhost:11434/api/chat -d '{
 
 ```json
 {
-  "model": "registry.ollama.ai/library/llama3:latest",
+  "model": "registry.ollama.ai/library/llama2:latest",
   "created_at": "2023-12-12T14:13:43.416799Z",
   "message": {
     "role": "assistant",
@@ -492,7 +495,7 @@ Send a chat message with a conversation history. You can use this same approach
 
 ```shell
 curl http://localhost:11434/api/chat -d '{
-  "model": "llama3",
+  "model": "llama2",
   "messages": [
     {
       "role": "user",
@@ -516,7 +519,7 @@ A stream of JSON objects is returned:
 
 ```json
 {
-  "model": "llama3",
+  "model": "llama2",
   "created_at": "2023-08-04T08:52:19.385406455-07:00",
   "message": {
     "role": "assistant",
@@ -530,7 +533,7 @@ Final response:
 
 ```json
 {
-  "model": "llama3",
+  "model": "llama2",
   "created_at": "2023-08-04T19:22:45.499127Z",
   "done": true,
   "total_duration": 8113331500,
@@ -588,7 +591,7 @@ curl http://localhost:11434/api/chat -d '{
 
 ```shell
 curl http://localhost:11434/api/chat -d '{
-  "model": "llama3",
+  "model": "llama2",
   "messages": [
     {
       "role": "user",
@@ -606,7 +609,7 @@ curl http://localhost:11434/api/chat -d '{
 
 ```json
 {
-  "model": "registry.ollama.ai/library/llama3:latest",
+  "model": "registry.ollama.ai/library/llama2:latest",
   "created_at": "2023-12-12T14:13:43.416799Z",
   "message": {
     "role": "assistant",
@@ -648,7 +651,7 @@ Create a new model from a `Modelfile`.
 ```shell
 curl http://localhost:11434/api/create -d '{
   "name": "mario",
-  "modelfile": "FROM llama3\nSYSTEM You are mario from Super Mario Bros."
+  "modelfile": "FROM llama2\nSYSTEM You are mario from Super Mario Bros."
 }'
 ```
 
@@ -755,7 +758,7 @@ A single JSON object will be returned.
       }
     },
     {
-      "name": "llama3:latest",
+      "name": "llama2:latest",
       "modified_at": "2023-12-07T09:32:18.757212583-08:00",
       "size": 3825819519,
       "digest": "fe938a131f40e6f6d40083c9f0f430a515233eb2edaa6d72eb85c50d64f2300e",
@@ -789,7 +792,7 @@ Show information about a model including details, modelfile, template, parameter
 
 ```shell
 curl http://localhost:11434/api/show -d '{
-  "name": "llama3"
+  "name": "llama2"
 }'
 ```
 
@@ -824,8 +827,8 @@ Copy a model. Creates a model with another name from an existing model.
 
 ```shell
 curl http://localhost:11434/api/copy -d '{
-  "source": "llama3",
-  "destination": "llama3-backup"
+  "source": "llama2",
+  "destination": "llama2-backup"
 }'
 ```
 
@@ -851,7 +854,7 @@ Delete a model and its data.
 
 ```shell
 curl -X DELETE http://localhost:11434/api/delete -d '{
-  "name": "llama3:13b"
+  "name": "llama2:13b"
 }'
 ```
 
@@ -879,7 +882,7 @@ Download a model from the ollama library. Cancelled pulls are resumed from where
 
 ```shell
 curl http://localhost:11434/api/pull -d '{
-  "name": "llama3"
+  "name": "llama2"
 }'
 ```
 
@@ -1007,7 +1010,8 @@ Generate embeddings from a model
 ### Parameters
 
 - `model`: name of model to generate embeddings from
-- `prompt`: text to generate embeddings for
+- `prompt`: string to generate the embedding for
+- `prompts`: array of strings to generate a batch of embeddings for
 
 Advanced parameters:
 
@@ -1035,3 +1039,33 @@ curl http://localhost:11434/api/embeddings -d '{
   ]
 }
 ```
+
+
+#### Request (batch)
+
+```shell
+curl http://localhost:11434/api/embeddings -d '{
+  "model": "all-minilm",
+  "prompt_batch": [
+    "Here is an article about llamas...",
+    "Here is another article about llamas..."
+  ]
+}'
+```
+
+#### Response
+
+```json
+{
+  "embedding_batch": [
+    [
+      0.5670403838157654, 0.009260174818336964, 0.23178744316101074, -0.2916173040866852, -0.8924556970596313,
+      0.8785552978515625, -0.34576427936553955, 0.5742510557174683, -0.04222835972905159, -0.137906014919281
+    ],
+    [
+      0.5670403838157654, 0.009260174818336964, 0.23178744316101074, -0.2916173040866852, -0.8924556970596313,
+      0.8785552978515625, -0.34576427936553955, 0.5742510557174683, -0.04222835972905159, -0.137906014919281
+    ],
+  ]
+}
+```

docs/development.md

@@ -51,7 +51,7 @@ Typically the build scripts will auto-detect CUDA, however, if your Linux distro
 or installation approach uses unusual paths, you can specify the location by
 specifying an environment variable `CUDA_LIB_DIR` to the location of the shared
 libraries, and `CUDACXX` to the location of the nvcc compiler. You can customize
-a set of target CUDA architectures by setting `CMAKE_CUDA_ARCHITECTURES` (e.g. "50;60;70")
+set set of target CUDA architectues by setting `CMAKE_CUDA_ARCHITECTURES` (e.g. "50;60;70")
 
 Then generate dependencies:
 
@@ -142,4 +142,4 @@ In addition to the common Windows development tools described above, install AMD
 - [AMD HIP](https://www.amd.com/en/developer/resources/rocm-hub/hip-sdk.html)
 - [Strawberry Perl](https://strawberryperl.com/)
 
-Lastly, add `ninja.exe` included with MSVC to the system path (e.g. `C:\Program Files (x86)\Microsoft Visual Studio\2019\Community\Common7\IDE\CommonExtensions\Microsoft\CMake\Ninja`).
+Lastly, add `ninja.exe` included with MSVC to the system path (e.g. `C:\Program Files (x86)\Microsoft Visual Studio\2019\Community\Common7\IDE\CommonExtensions\Microsoft\CMake\Ninja`).

docs/docker.md

@@ -1,71 +0,0 @@
-# Ollama Docker image
-
-### CPU only
-
-```bash
-docker run -d -v ollama:/root/.ollama -p 11434:11434 --name ollama ollama/ollama
-```
-
-### Nvidia GPU
-Install the [NVIDIA Container Toolkit](https://docs.nvidia.com/datacenter/cloud-native/container-toolkit/latest/install-guide.html#installation).
-
-#### Install with Apt
-1.  Configure the repository
-```bash
-curl -fsSL https://nvidia.github.io/libnvidia-container/gpgkey \
-    | sudo gpg --dearmor -o /usr/share/keyrings/nvidia-container-toolkit-keyring.gpg
-curl -s -L https://nvidia.github.io/libnvidia-container/stable/deb/nvidia-container-toolkit.list \
-    | sed 's#deb https://#deb [signed-by=/usr/share/keyrings/nvidia-container-toolkit-keyring.gpg] https://#g' \
-    | sudo tee /etc/apt/sources.list.d/nvidia-container-toolkit.list
-sudo apt-get update
-```
-2.  Install the NVIDIA Container Toolkit packages
-```bash
-sudo apt-get install -y nvidia-container-toolkit
-```
-
-#### Install with Yum or Dnf
-1.  Configure the repository
-    
-```bash
-curl -s -L https://nvidia.github.io/libnvidia-container/stable/rpm/nvidia-container-toolkit.repo \
-    | sudo tee /etc/yum.repos.d/nvidia-container-toolkit.repo
-```
-    
-2. Install the NVIDIA Container Toolkit packages
-    
-```bash
-sudo yum install -y nvidia-container-toolkit
-```
-
-#### Configure Docker to use Nvidia driver 
-```
-sudo nvidia-ctk runtime configure --runtime=docker
-sudo systemctl restart docker
-```
-
-#### Start the container
-
-```bash
-docker run -d --gpus=all -v ollama:/root/.ollama -p 11434:11434 --name ollama ollama/ollama
-```
-
-### AMD GPU
-
-To run Ollama using Docker with AMD GPUs, use the `rocm` tag and the following command:
-
-```
-docker run -d --device /dev/kfd --device /dev/dri -v ollama:/root/.ollama -p 11434:11434 --name ollama ollama/ollama:rocm
-```
-
-### Run model locally
-
-Now you can run a model:
-
-```
-docker exec -it ollama ollama run llama3
-```
-
-### Try different models
-
-More models can be found on the [Ollama library](https://ollama.com/library).

docs/faq.md

@@ -32,7 +32,7 @@ When using the API, specify the `num_ctx` parameter:
 
 ```
 curl http://localhost:11434/api/generate -d '{
-  "model": "llama3",
+  "model": "llama2",
   "prompt": "Why is the sky blue?",
   "options": {
     "num_ctx": 4096
@@ -88,9 +88,9 @@ On windows, Ollama inherits your user and system environment variables.
 
 3. Edit or create New variable(s) for your user account for `OLLAMA_HOST`, `OLLAMA_MODELS`, etc.
 
-4. Click OK/Apply to save
+4. Click OK/Apply to save 
 
-5. Run `ollama` from a new terminal window
+5. Run `ollama` from a new terminal window 
 
 
 ## How can I expose Ollama on my network?
@@ -140,7 +140,7 @@ Refer to the section [above](#how-do-i-configure-ollama-server) for how to set e
 
 - macOS: `~/.ollama/models`
 - Linux: `/usr/share/ollama/.ollama/models`
-- Windows: `C:\Users\%username%\.ollama\models`
+- Windows: `C:\Users\<username>\.ollama\models`
 
 ### How do I set them to a different location?
 
@@ -221,20 +221,14 @@ The `keep_alive` parameter can be set to:
 
 For example, to preload a model and leave it in memory use:
 ```shell
-curl http://localhost:11434/api/generate -d '{"model": "llama3", "keep_alive": -1}'
+curl http://localhost:11434/api/generate -d '{"model": "llama2", "keep_alive": -1}'
 ```
 
 To unload the model and free up memory use:
 ```shell
-curl http://localhost:11434/api/generate -d '{"model": "llama3", "keep_alive": 0}'
+curl http://localhost:11434/api/generate -d '{"model": "llama2", "keep_alive": 0}'
 ```
 
 Alternatively, you can change the amount of time all models are loaded into memory by setting the `OLLAMA_KEEP_ALIVE` environment variable when starting the Ollama server. The `OLLAMA_KEEP_ALIVE` variable uses the same parameter types as the `keep_alive` parameter types mentioned above. Refer to section explaining [how to configure the Ollama server](#how-do-i-configure-ollama-server) to correctly set the environment variable.
 
 If you wish to override the `OLLAMA_KEEP_ALIVE` setting, use the `keep_alive` API parameter with the `/api/generate` or `/api/chat` API endpoints.
-
-## How do I manage the maximum number of requests the server can queue
-
-If too many requests are sent to the server, it will respond with a 503 error
-indicating the server is overloaded.  You can adjust how many requests may be
-queue by setting `OLLAMA_MAX_QUEUE`

docs/import.md

@@ -125,7 +125,7 @@ Publishing models is in early alpha. If you'd like to publish your model to shar
 
 1. Create [an account](https://ollama.com/signup)
 2. Copy your Ollama public key:
-  - macOS: `cat ~/.ollama/id_ed25519.pub | pbcopy`
+  - macOS: `cat ~/.ollama/id_ed25519.pub`
   - Windows: `type %USERPROFILE%\.ollama\id_ed25519.pub`
   - Linux: `cat /usr/share/ollama/.ollama/id_ed25519.pub`
 3. Add your public key to your [Ollama account](https://ollama.com/settings/keys)
@@ -136,8 +136,6 @@ Next, copy your model to your username's namespace:
 ollama cp example <your username>/example
 ```
 
-> Note: model names may only contain lowercase letters, digits, and the characters `.`, `-`, and `_`.
-
 Then push the model:
 
 ```

docs/linux.md

@@ -105,7 +105,7 @@ sudo chmod +x /usr/bin/ollama
 To view logs of Ollama running as a startup service, run:
 
 ```bash
-journalctl -e -u ollama
+journalctl -u ollama
 ```
 
 ## Uninstall

docs/modelfile.md

@@ -10,7 +10,7 @@ A model file is the blueprint to create and share models with Ollama.
 - [Examples](#examples)
 - [Instructions](#instructions)
   - [FROM (Required)](#from-required)
-    - [Build from llama3](#build-from-llama3)
+    - [Build from llama2](#build-from-llama2)
     - [Build from a bin file](#build-from-a-bin-file)
   - [PARAMETER](#parameter)
     - [Valid Parameters and Values](#valid-parameters-and-values)
@@ -48,7 +48,7 @@ INSTRUCTION arguments
 An example of a `Modelfile` creating a mario blueprint:
 
 ```modelfile
-FROM llama3
+FROM llama2
 # sets the temperature to 1 [higher is more creative, lower is more coherent]
 PARAMETER temperature 1
 # sets the context window size to 4096, this controls how many tokens the LLM can use as context to generate the next token
@@ -67,25 +67,33 @@ To use this:
 
 More examples are available in the [examples directory](../examples).
 
-To view the Modelfile of a given model, use the `ollama show --modelfile` command.
+### `Modelfile`s in [ollama.com/library][1]
+
+There are two ways to view `Modelfile`s underlying the models in [ollama.com/library][1]:
+
+- Option 1: view a details page from a model's tags page:
+  1.  Go to a particular model's tags (e.g. https://ollama.com/library/llama2/tags)
+  2.  Click on a tag (e.g. https://ollama.com/library/llama2:13b)
+  3.  Scroll down to "Layers"
+      - Note: if the [`FROM` instruction](#from-required) is not present,
+        it means the model was created from a local file
+- Option 2: use `ollama show` to print the `Modelfile` for any local models like so:
 
   ```bash
-  > ollama show --modelfile llama3
+  > ollama show --modelfile llama2:13b
   # Modelfile generated by "ollama show"
   # To build a new Modelfile based on this one, replace the FROM line with:
-  # FROM llama3:latest
-  FROM /Users/pdevine/.ollama/models/blobs/sha256-00e1317cbf74d901080d7100f57580ba8dd8de57203072dc6f668324ba545f29
-  TEMPLATE """{{ if .System }}<|start_header_id|>system<|end_header_id|>
+  # FROM llama2:13b
 
-  {{ .System }}<|eot_id|>{{ end }}{{ if .Prompt }}<|start_header_id|>user<|end_header_id|>
+  FROM /root/.ollama/models/blobs/sha256:123abc
+  TEMPLATE """[INST] {{ if .System }}<<SYS>>{{ .System }}<</SYS>>
 
-  {{ .Prompt }}<|eot_id|>{{ end }}<|start_header_id|>assistant<|end_header_id|>
-
-  {{ .Response }}<|eot_id|>"""
-  PARAMETER stop "<|start_header_id|>"
-  PARAMETER stop "<|end_header_id|>"
-  PARAMETER stop "<|eot_id|>"
-  PARAMETER stop "<|reserved_special_token"
+  {{ end }}{{ .Prompt }} [/INST] """
+  SYSTEM """"""
+  PARAMETER stop [INST]
+  PARAMETER stop [/INST]
+  PARAMETER stop <<SYS>>
+  PARAMETER stop <</SYS>>
   ```
 
 ## Instructions
@@ -98,10 +106,10 @@ The `FROM` instruction defines the base model to use when creating a model.
 FROM <model name>:<tag>
 ```
 
-#### Build from llama3
+#### Build from llama2
 
 ```modelfile
-FROM llama3
+FROM llama2
 ```
 
 A list of available base models:

docs/openai.md

@@ -25,7 +25,7 @@ chat_completion = client.chat.completions.create(
             'content': 'Say this is a test',
         }
     ],
-    model='llama3',
+    model='llama2',
 )
 ```
 
@@ -43,7 +43,7 @@ const openai = new OpenAI({
 
 const chatCompletion = await openai.chat.completions.create({
   messages: [{ role: 'user', content: 'Say this is a test' }],
-  model: 'llama3',
+  model: 'llama2',
 })
 ```
 
@@ -53,7 +53,7 @@ const chatCompletion = await openai.chat.completions.create({
 curl http://localhost:11434/v1/chat/completions \
     -H "Content-Type: application/json" \
     -d '{
-        "model": "llama3",
+        "model": "llama2",
         "messages": [
             {
                 "role": "system",
@@ -113,7 +113,7 @@ curl http://localhost:11434/v1/chat/completions \
 Before using a model, pull it locally `ollama pull`:
 
 ```shell
-ollama pull llama3
+ollama pull llama2
 ```
 
 ### Default model names
@@ -121,7 +121,7 @@ ollama pull llama3
 For tooling that relies on default OpenAI model names such as `gpt-3.5-turbo`, use `ollama cp` to copy an existing model name to a temporary name:
 
 ```
-ollama cp llama3 gpt-3.5-turbo
+ollama cp llama2 gpt-3.5-turbo
 ```
 
 Afterwards, this new model name can be specified the `model` field:

docs/troubleshooting.md

@@ -82,23 +82,4 @@ curl -fsSL https://ollama.com/install.sh | OLLAMA_VERSION="0.1.29" sh
 If your system is configured with the "noexec" flag where Ollama stores its
 temporary executable files, you can specify an alternate location by setting
 OLLAMA_TMPDIR to a location writable by the user ollama runs as.  For example
-OLLAMA_TMPDIR=/usr/share/ollama/
-
-## Container fails to run on NVIDIA GPU
-
-Make sure you've set up the conatiner runtime first as described in [docker.md](./docker.md)
-
-Sometimes the container runtime can have difficulties initializing the GPU.
-When you check the server logs, this can show up as various error codes, such
-as "3" (not initialized), "46" (device unavailable), "100" (no device), "999"
-(unknown), or others.  The following troubleshooting techniques may help resolve
-the problem
-
-- Is the uvm driver not loaded? `sudo nvidia-modprobe -u`
-- Try reloading the nvidia_uvm driver - `sudo rmmod nvidia_uvm` then `sudo modprobe nvidia_uvm`
-- Try rebooting
-- Make sure you're running the latest nvidia drivers
-
-If none of those resolve the problem, gather additional information and file an issue:
-- Set `CUDA_ERROR_LEVEL=50` and try again to get more diagnostic logs
-- Check dmesg for any errors `sudo dmesg | grep -i nvrm` and `sudo dmesg | grep -i nvidia`
+OLLAMA_TMPDIR=/usr/share/ollama/

docs/tutorials/langchainjs.md

@@ -5,17 +5,17 @@ In this tutorial, we are going to use JavaScript with LangChain and Ollama to le
 To get started, let's just use **LangChain** to ask a simple question to a model. To do this with JavaScript, we need to install **LangChain**:
 
 ```bash
-npm install @langchain/community
+npm install langchain
 ```
 
 Now we can start building out our JavaScript:
 
 ```javascript
-import { Ollama } from "@langchain/community/llms/ollama";
+import { Ollama } from "langchain/llms/ollama";
 
 const ollama = new Ollama({
   baseUrl: "http://localhost:11434",
-  model: "llama3",
+  model: "llama2",
 });
 
 const answer = await ollama.invoke(`why is the sky blue?`);
@@ -23,10 +23,10 @@ const answer = await ollama.invoke(`why is the sky blue?`);
 console.log(answer);
 ```
 
-That will get us the same thing as if we ran `ollama run llama3 "why is the sky blue"` in the terminal. But we want to load a document from the web to ask a question against. **Cheerio** is a great library for ingesting a webpage, and **LangChain** uses it in their **CheerioWebBaseLoader**. So let's install **Cheerio** and build that part of the app.
+That will get us the same thing as if we ran `ollama run llama2 "why is the sky blue"` in the terminal. But we want to load a document from the web to ask a question against. **Cheerio** is a great library for ingesting a webpage, and **LangChain** uses it in their **CheerioWebBaseLoader**. So let's install **Cheerio** and build that part of the app.
 
 ```bash
-npm install cheerio
+npm install cheerio 
 ```
 
 ```javascript

docs/tutorials/langchainpy.md

@@ -12,17 +12,15 @@ So let's figure out how we can use **LangChain** with Ollama to ask our question
 
 Let's start by asking a simple question that we can get an answer to from the **Llama2** model using **Ollama**. First, we need to install the **LangChain** package:
 
-`pip install langchain_community`
+`pip install langchain`
 
 Then we can create a model and ask the question:
 
 ```python
-from langchain_community.llms import Ollama
-ollama = Ollama(
-    base_url='http://localhost:11434',
-    model="llama3"
-)
-print(ollama.invoke("why is the sky blue"))
+from langchain.llms import Ollama
+ollama = Ollama(base_url='http://localhost:11434',
+model="llama2")
+print(ollama("why is the sky blue"))
 ```
 
 Notice that we are defining the model and the base URL for Ollama.

docs/windows.md

@@ -1,61 +1,47 @@
-# Ollama Windows Preview
-
-Welcome to the Ollama Windows preview.
-
-No more WSL required!
-
-Ollama now runs as a native Windows application, including NVIDIA and AMD Radeon GPU support.
-After installing Ollama Windows Preview, Ollama will run in the background and
-the `ollama` command line is available in `cmd`, `powershell` or your favorite
-terminal application. As usual the Ollama [api](./api.md) will be served on
-`http://localhost:11434`.
-
-As this is a preview release, you should expect a few bugs here and there.  If
-you run into a problem you can reach out on
-[Discord](https://discord.gg/ollama), or file an
-[issue](https://github.com/ollama/ollama/issues).
-Logs will often be helpful in diagnosing the problem (see
-[Troubleshooting](#troubleshooting) below)
-
-## System Requirements
-
-* Windows 10 or newer, Home or Pro
-* NVIDIA 452.39 or newer Drivers if you have an NVIDIA card
-* AMD Radeon Driver https://www.amd.com/en/support if you have a Radeon card
-
-## API Access
-
-Here's a quick example showing API access from `powershell`
-```powershell
-(Invoke-WebRequest -method POST -Body '{"model":"llama3", "prompt":"Why is the sky blue?", "stream": false}' -uri http://localhost:11434/api/generate ).Content | ConvertFrom-json
-```
-
-## Troubleshooting
-
-While we're in preview, `OLLAMA_DEBUG` is always enabled, which adds
-a "view logs" menu item to the app, and increses logging for the GUI app and
-server.
-
-Ollama on Windows stores files in a few different locations.  You can view them in
-the explorer window by hitting `<cmd>+R` and type in:
-- `explorer %LOCALAPPDATA%\Ollama` contains logs, and downloaded updates
-    - *app.log* contains logs from the GUI application
-    - *server.log* contains the server logs
-    - *upgrade.log* contains log output for upgrades
-- `explorer %LOCALAPPDATA%\Programs\Ollama` contains the binaries (The installer adds this to your user PATH)
-- `explorer %HOMEPATH%\.ollama` contains models and configuration
-- `explorer %TEMP%` contains temporary executable files in one or more `ollama*` directories
-
-
-## Standalone CLI
-
-The easiest way to install Ollama on Windows is to use the `OllamaSetup.exe`
-installer. It installs in your account without requiring Administrator rights.
-We update Ollama regularly to support the latest models, and this installer will
-help you keep up to date.
-
-If you'd like to install or integrate Ollama as a service, a standalone
-`ollama-windows-amd64.zip` zip file is available containing only the Ollama CLI
-and GPU library dependencies for Nvidia and AMD. This allows for embedding
-Ollama in existing applications, or running it as a system service via `ollama
-serve` with tools such as [NSSM](https://nssm.cc/).
+# Ollama Windows Preview
+
+Welcome to the Ollama Windows preview.
+
+No more WSL required!
+
+Ollama now runs as a native Windows application, including NVIDIA and AMD Radeon GPU support.
+After installing Ollama Windows Preview, Ollama will run in the background and
+the `ollama` command line is available in `cmd`, `powershell` or your favorite
+terminal application. As usual the Ollama [api](./api.md) will be served on
+`http://localhost:11434`.
+
+As this is a preview release, you should expect a few bugs here and there.  If
+you run into a problem you can reach out on
+[Discord](https://discord.gg/ollama), or file an 
+[issue](https://github.com/ollama/ollama/issues).
+Logs will often be helpful in diagnosing the problem (see
+[Troubleshooting](#troubleshooting) below)
+
+## System Requirements
+
+* Windows 10 or newer, Home or Pro
+* NVIDIA 452.39 or newer Drivers if you have an NVIDIA card
+* AMD Radeon Driver https://www.amd.com/en/support if you have a Radeon card
+
+## API Access
+
+Here's a quick example showing API access from `powershell`
+```powershell
+(Invoke-WebRequest -method POST -Body '{"model":"llama2", "prompt":"Why is the sky blue?", "stream": false}' -uri http://localhost:11434/api/generate ).Content | ConvertFrom-json
+```
+
+## Troubleshooting
+
+While we're in preview, `OLLAMA_DEBUG` is always enabled, which adds
+a "view logs" menu item to the app, and increses logging for the GUI app and
+server.
+
+Ollama on Windows stores files in a few different locations.  You can view them in
+the explorer window by hitting `<cmd>+R` and type in:
+- `explorer %LOCALAPPDATA%\Ollama` contains logs, and downloaded updates
+    - *app.log* contains logs from the GUI application
+    - *server.log* contains the server logs
+    - *upgrade.log* contains log output for upgrades
+- `explorer %LOCALAPPDATA%\Programs\Ollama` contains the binaries (The installer adds this to your user PATH)
+- `explorer %HOMEPATH%\.ollama` contains models and configuration
+- `explorer %TEMP%` contains temporary executable files in one or more `ollama*` directories

examples/bash-comparemodels/README.md

@@ -0,0 +1,10 @@
+# Bash Shell examples
+
+When calling `ollama`, you can pass it a file to run all the prompts in the file, one after the other:
+
+`ollama run llama2 < sourcequestions.txt`
+
+This concept is used in the following example.
+
+## Compare Models
+`comparemodels.sh` is a script that runs all the questions in `sourcequestions.txt` using any 4 models you choose that you have already pulled from the Ollama library or have created locally.

examples/bash-comparemodels/comparemodels.sh

@@ -0,0 +1,64 @@
+#! /usr/bin/env bash
+# Compare multiple models by running them with the same questions
+
+NUMBEROFCHOICES=4
+SELECTIONS=()
+declare -a SUMS=()
+
+# Get the list of models
+CHOICES=$(ollama list | awk '{print $1}')
+
+# Select which models to run as a comparison
+echo "Select $NUMBEROFCHOICES models to compare:"
+select ITEM in $CHOICES; do
+    if [[ -n $ITEM ]]; then
+        echo "You have selected $ITEM"
+        SELECTIONS+=("$ITEM")
+        ((COUNT++))
+        if [[ $COUNT -eq $NUMBEROFCHOICES ]]; then
+            break
+        fi
+    else
+        echo "Invalid selection"
+    fi
+done
+
+# Loop through each of the selected models
+for ITEM in "${SELECTIONS[@]}"; do
+    echo "--------------------------------------------------------------"
+    echo "Loading the model $ITEM into memory"
+    ollama run "$ITEM" ""
+    echo "--------------------------------------------------------------"
+    echo "Running the questions through the model $ITEM"
+    COMMAND_OUTPUT=$(ollama run "$ITEM" --verbose < sourcequestions.txt 2>&1| tee /dev/stderr)
+
+    # eval duration is sometimes listed in seconds and sometimes in milliseconds. 
+    # Add up the values for each model
+    SUM=$(echo "$COMMAND_OUTPUT" | awk '
+    /eval duration:/ {
+        value = $3
+        if (index(value, "ms") > 0) {
+            gsub("ms", "", value)
+            value /= 1000
+        } else {
+            gsub("s", "", value)
+        }
+        sum += value
+    }
+    END { print sum }')
+
+
+    SUMS+=("All questions for $ITEM completed in $SUM seconds")
+done
+
+echo ""
+echo "--------------------------------------------------------------"
+echo -e "Sums of eval durations for each run:"
+for val in "${SUMS[@]}"; do
+    echo "$val"
+done
+
+echo "--------------------------------------------------------------"
+echo "Comparison complete. Now you can decide"
+echo "which model is best."
+echo "--------------------------------------------------------------"

examples/bash-comparemodels/sourcequestions.txt

@@ -0,0 +1,7 @@
+Why is the sky blue
+What is a black hole
+Explain the big bang theory like I am 5?
+What is the quickest way to win a game of Monopoly with 3 others?
+Why does a vacuum bottle keep my coffee hot and my milkshake cold?
+What is the difference between a meteor, a meteorite, and a meteoroid?
+Create an array with 5 items and print to the console. Do this in Python, C#, Typescript, and Rust.

examples/flyio/.gitignore

@@ -1 +0,0 @@
-fly.toml

examples/flyio/README.md

@@ -1,67 +0,0 @@
-# Deploy Ollama to Fly.io
-
-> Note: this example exposes a public endpoint and does not configure authentication. Use with care.
-
-## Prerequisites
-
-- Ollama: https://ollama.com/download
-- Fly.io account. Sign up for a free account: https://fly.io/app/sign-up
-
-## Steps
-
-1. Login to Fly.io
-
-    ```bash
-    fly auth login
-    ```
-
-1. Create a new Fly app
-
-    ```bash
-    fly launch --name <name> --image ollama/ollama --internal-port 11434 --vm-size shared-cpu-8x --now
-    ```
-
-1. Pull and run `orca-mini:3b`
-
-    ```bash
-    OLLAMA_HOST=https://<name>.fly.dev ollama run orca-mini:3b
-    ```
-
-`shared-cpu-8x` is a free-tier eligible machine type. For better performance, switch to a `performance` or `dedicated` machine type or attach a GPU for hardware acceleration (see below).
-
-## (Optional) Persistent Volume
-
-By default Fly Machines use ephemeral storage which is problematic if you want to use the same model across restarts without pulling it again. Create and attach a persistent volume to store the downloaded models:
-
-1. Create the Fly Volume
-
-    ```bash
-    fly volume create ollama
-    ```
-
-1. Update `fly.toml` and add `[mounts]`
-
-    ```toml
-    [mounts]
-      source = "ollama"
-      destination = "/mnt/ollama/models"
-    ```
-
-1. Update `fly.toml` and add `[env]`
-
-    ```toml
-    [env]
-      OLLAMA_MODELS = "/mnt/ollama/models"
-    ```
-
-1. Deploy your app
-
-    ```bash
-    fly deploy
-    ```
-
-## (Optional) Hardware Acceleration
-
-Fly.io GPU is currently in waitlist. Sign up for the waitlist: https://fly.io/gpu
-
-Once you've been accepted, create the app with the additional flags `--vm-gpu-kind a100-pcie-40gb` or `--vm-gpu-kind a100-pcie-80gb`.

examples/go-chat/main.go

@@ -35,7 +35,7 @@ func main() {
 
 	ctx := context.Background()
 	req := &api.ChatRequest{
-		Model:    "llama3",
+		Model:    "llama2",
 		Messages: messages,
 	}
 

examples/go-http-generate/main.go

@@ -19,7 +19,7 @@ func main() {
 	}
 
 	defer resp.Body.Close()
-
+	
 	responseData, err := io.ReadAll(resp.Body)
 	if err != nil {
 		log.Fatal(err)

examples/kubernetes/README.md

@@ -7,24 +7,12 @@
 
 ## Steps
 
-1. Create the Ollama namespace, deployment, and service
+1. Create the Ollama namespace, daemon set, and service
 
    ```bash
    kubectl apply -f cpu.yaml
    ```
 
-## (Optional) Hardware Acceleration
-
-Hardware acceleration in Kubernetes requires NVIDIA's [`k8s-device-plugin`](https://github.com/NVIDIA/k8s-device-plugin) which is deployed in Kubernetes in form of daemonset. Follow the link for more details.
-
-Once configured, create a GPU enabled Ollama deployment.
-
-```bash
-kubectl apply -f gpu.yaml
-```
-
-## Test
-
 1. Port forward the Ollama service to connect and use it locally
 
    ```bash
@@ -35,4 +23,14 @@ kubectl apply -f gpu.yaml
 
    ```bash
    ollama run orca-mini:3b
-   ```
+   ```
+
+## (Optional) Hardware Acceleration
+
+Hardware acceleration in Kubernetes requires NVIDIA's [`k8s-device-plugin`](https://github.com/NVIDIA/k8s-device-plugin). Follow the link for more details.
+
+Once configured, create a GPU enabled Ollama deployment.
+
+```bash
+kubectl apply -f gpu.yaml
+```

examples/langchain-python-rag-document/main.py

@@ -40,9 +40,9 @@ while True:
         continue
 
     # Prompt
-    template = """Use the following pieces of context to answer the question at the end.
-    If you don't know the answer, just say that you don't know, don't try to make up an answer.
-    Use three sentences maximum and keep the answer as concise as possible.
+    template = """Use the following pieces of context to answer the question at the end. 
+    If you don't know the answer, just say that you don't know, don't try to make up an answer. 
+    Use three sentences maximum and keep the answer as concise as possible. 
     {context}
     Question: {question}
     Helpful Answer:"""
@@ -51,11 +51,11 @@ while True:
         template=template,
     )
 
-    llm = Ollama(model="llama3:8b", callback_manager=CallbackManager([StreamingStdOutCallbackHandler()]))
+    llm = Ollama(model="llama2:13b", callback_manager=CallbackManager([StreamingStdOutCallbackHandler()]))
     qa_chain = RetrievalQA.from_chain_type(
         llm,
         retriever=vectorstore.as_retriever(),
         chain_type_kwargs={"prompt": QA_CHAIN_PROMPT},
     )
 
-    result = qa_chain({"query": query})
+    result = qa_chain({"query": query})

examples/langchain-python-rag-websummary/main.py

@@ -1,12 +1,12 @@
-from langchain_community.llms import Ollama
-from langchain_community.document_loaders import WebBaseLoader
+from langchain.llms import Ollama
+from langchain.document_loaders import WebBaseLoader
 from langchain.chains.summarize import load_summarize_chain
 
 loader = WebBaseLoader("https://ollama.com/blog/run-llama2-uncensored-locally")
 docs = loader.load()
 
-llm = Ollama(model="llama3")
+llm = Ollama(model="llama2")
 chain = load_summarize_chain(llm, chain_type="stuff")
 
-result = chain.invoke(docs) 
+result = chain.run(docs)
 print(result)

examples/langchain-python-simple/README.md

@@ -4,10 +4,10 @@ This example is a basic "hello world" of using LangChain with Ollama.
 
 ## Running the Example
 
-1. Ensure you have the `llama3` model installed:
+1. Ensure you have the `llama2` model installed:
 
    ```bash
-   ollama pull llama3
+   ollama pull llama2
    ```
 
 2. Install the Python Requirements.
@@ -21,3 +21,4 @@ This example is a basic "hello world" of using LangChain with Ollama.
    ```bash
    python main.py
    ```
+  

examples/langchain-python-simple/main.py

@@ -1,6 +1,6 @@
 from langchain.llms import Ollama
 
 input = input("What is your question?")
-llm = Ollama(model="llama3")
+llm = Ollama(model="llama2")
 res = llm.predict(input)
 print (res)

examples/modelfile-mario/Modelfile

@@ -1,4 +1,4 @@
-FROM llama3
+FROM llama2
 PARAMETER temperature 1
 SYSTEM """
 You are Mario from super mario bros, acting as an assistant.

examples/modelfile-mario/readme.md

@@ -2,12 +2,12 @@
 
 # Example character: Mario
 
-This example shows how to create a basic character using Llama3 as the base model.
+This example shows how to create a basic character using Llama2 as the base model.
 
 To run this example:
 
 1. Download the Modelfile
-2. `ollama pull llama3` to get the base model used in the model file.
+2. `ollama pull llama2` to get the base model used in the model file.
 3. `ollama create NAME -f ./Modelfile`
 4. `ollama run NAME`
 
@@ -18,7 +18,7 @@ Ask it some questions like "Who are you?" or "Is Peach in trouble again?"
 What the model file looks like:
 
 ```
-FROM llama3
+FROM llama2
 PARAMETER temperature 1
 SYSTEM """
 You are Mario from Super Mario Bros, acting as an assistant.

examples/python-json-datagenerator/predefinedschema.py

@@ -2,16 +2,16 @@ import requests
 import json
 import random
 
-model = "llama3"
+model = "llama2"
 template = {
-  "firstName": "",
-  "lastName": "",
+  "firstName": "", 
+  "lastName": "", 
   "address": {
-    "street": "",
-    "city": "",
-    "state": "",
+    "street": "", 
+    "city": "", 
+    "state": "", 
     "zipCode": ""
-  },
+  }, 
   "phoneNumber": ""
 }
 

examples/python-json-datagenerator/randomaddresses.py

@@ -12,7 +12,7 @@ countries = [
     "France",
 ]
 country = random.choice(countries)
-model = "llama3"
+model = "llama2"
 
 prompt = f"generate one realistically believable sample data set of a persons first name, last name, address in {country}, and phone number. Do not use common names. Respond using JSON. Key names should have no backslashes, values should use plain ascii with no special characters."
 

examples/python-json-datagenerator/readme.md

@@ -6,10 +6,10 @@ There are two python scripts in this example. `randomaddresses.py` generates ran
 
 ## Running the Example
 
-1. Ensure you have the `llama3` model installed:
+1. Ensure you have the `llama2` model installed:
 
    ```bash
-   ollama pull llama3
+   ollama pull llama2
    ```
 
 2. Install the Python Requirements.

examples/python-simplechat/client.py

@@ -2,7 +2,7 @@ import json
 import requests
 
 # NOTE: ollama must be running for this to work, start the ollama app or run `ollama serve`
-model = "llama3"  # TODO: update this for whatever model you wish to use
+model = "llama2"  # TODO: update this for whatever model you wish to use
 
 
 def chat(messages):

examples/python-simplechat/readme.md

@@ -4,10 +4,10 @@ The **chat** endpoint is one of two ways to generate text from an LLM with Ollam
 
 ## Running the Example
 
-1. Ensure you have the `llama3` model installed:
+1. Ensure you have the `llama2` model installed:
 
    ```bash
-   ollama pull llama3
+   ollama pull llama2
    ```
 
 2. Install the Python Requirements.

examples/typescript-mentors/README.md

@@ -4,10 +4,10 @@ This example demonstrates how one would create a set of 'mentors' you can have a
 
 ## Usage
 
-1. Add llama3 to have the mentors ask your questions:
+1. Add llama2 to have the mentors ask your questions:
 
    ```bash
-   ollama pull llama3
+   ollama pull llama2
    ```
 
 2. Install prerequisites:

examples/typescript-mentors/character-generator.ts

@@ -15,7 +15,7 @@ async function characterGenerator() {
   ollama.setModel("stablebeluga2:70b-q4_K_M");
   const bio = await ollama.generate(`create a bio of ${character} in a single long paragraph. Instead of saying '${character} is...' or '${character} was...' use language like 'You are...' or 'You were...'. Then create a paragraph describing the speaking mannerisms and style of ${character}. Don't include anything about how ${character} looked or what they sounded like, just focus on the words they said. Instead of saying '${character} would say...' use language like 'You should say...'. If you use quotes, always use single quotes instead of double quotes. If there are any specific words or phrases you used a lot, show how you used them. `);
 
-  const thecontents = `FROM llama3\nSYSTEM """\n${bio.response.replace(/(\r\n|\n|\r)/gm, " ").replace('would', 'should')} All answers to questions should be related back to what you are most known for.\n"""`;
+  const thecontents = `FROM llama2\nSYSTEM """\n${bio.response.replace(/(\r\n|\n|\r)/gm, " ").replace('would', 'should')} All answers to questions should be related back to what you are most known for.\n"""`;
 
   fs.writeFile(path.join(directory, 'Modelfile'), thecontents, (err: any) => {
     if (err) throw err;
@@ -23,4 +23,4 @@ async function characterGenerator() {
   });
 }
 
-characterGenerator();
+characterGenerator();

examples/typescript-simplechat/client.ts

@@ -1,6 +1,6 @@
 import * as readline from "readline";
 
-const model = "llama3";
+const model = "llama2";
 type Message = {
   role: "assistant" | "user" | "system";
   content: string;
@@ -74,4 +74,4 @@ async function main() {
 
 }
 
-main();
+main();

format/bytes.go

@@ -53,8 +53,6 @@ func HumanBytes(b int64) string {
 
 func HumanBytes2(b uint64) string {
 	switch {
-	case b >= GibiByte:
-		return fmt.Sprintf("%.1f GiB", float64(b)/GibiByte)
 	case b >= MebiByte:
 		return fmt.Sprintf("%.1f MiB", float64(b)/MebiByte)
 	case b >= KibiByte:

format/format.go

@@ -13,20 +13,12 @@ const (
 
 func HumanNumber(b uint64) string {
 	switch {
-	case b >= Billion:
-		number := float64(b) / Billion
-		if number == math.Floor(number) {
-			return fmt.Sprintf("%.0fB", number) // no decimals if whole number
-		}
-		return fmt.Sprintf("%.1fB", number) // one decimal if not a whole number
-	case b >= Million:
-		number := float64(b) / Million
-		if number == math.Floor(number) {
-			return fmt.Sprintf("%.0fM", number) // no decimals if whole number
-		}
-		return fmt.Sprintf("%.2fM", number) // two decimals if not a whole number
-	case b >= Thousand:
-		return fmt.Sprintf("%.0fK", float64(b)/Thousand)
+	case b > Billion:
+		return fmt.Sprintf("%.0fB", math.Round(float64(b)/Billion))
+	case b > Million:
+		return fmt.Sprintf("%.0fM", math.Round(float64(b)/Million))
+	case b > Thousand:
+		return fmt.Sprintf("%.0fK", math.Round(float64(b)/Thousand))
 	default:
 		return fmt.Sprintf("%d", b)
 	}

format/format_test.go

@@ -1,34 +0,0 @@
-package format
-
-import (
-	"testing"
-)
-
-func TestHumanNumber(t *testing.T) {
-
-	type testCase struct {
-		input    uint64
-		expected string
-	}
-
-	testCases := []testCase{
-		{0, "0"},
-		{1000000, "1M"},
-		{125000000, "125M"},
-		{500500000, "500.50M"},
-		{500550000, "500.55M"},
-		{1000000000, "1B"},
-		{2800000000, "2.8B"},
-		{2850000000, "2.9B"},
-		{1000000000000, "1000B"},
-	}
-
-	for _, tc := range testCases {
-		t.Run(tc.expected, func(t *testing.T) {
-			result := HumanNumber(tc.input)
-			if result != tc.expected {
-				t.Errorf("Expected %s, got %s", tc.expected, result)
-			}
-		})
-	}
-}

gpu/amd_common.go

@@ -81,10 +81,8 @@ func commonAMDValidateLibDir() (string, error) {
 	}
 
 	// Well known location(s)
-	for _, path := range RocmStandardLocations {
-		if rocmLibUsable(path) {
-			return path, nil
-		}
+	if rocmLibUsable(RocmStandardLocation) {
+		return RocmStandardLocation, nil
 	}
 
 	// Installer payload location if we're running the installed binary

gpu/amd_hip_windows.go

@@ -3,6 +3,7 @@ package gpu
 import (
 	"fmt"
 	"log/slog"
+	"strconv"
 	"syscall"
 	"unsafe"
 
@@ -73,22 +74,16 @@ func (hl *HipLib) Release() {
 	hl.dll = 0
 }
 
-func (hl *HipLib) AMDDriverVersion() (driverMajor, driverMinor int, err error) {
+func (hl *HipLib) AMDDriverVersion() (string, error) {
 	if hl.dll == 0 {
-		return 0, 0, fmt.Errorf("dll has been unloaded")
+		return "", fmt.Errorf("dll has been unloaded")
 	}
 	var version int
 	status, _, err := syscall.SyscallN(hl.hipDriverGetVersion, uintptr(unsafe.Pointer(&version)))
 	if status != hipSuccess {
-		return 0, 0, fmt.Errorf("failed call to hipDriverGetVersion: %d %s", status, err)
+		return "", fmt.Errorf("failed call to hipDriverGetVersion: %d %s", status, err)
 	}
-
-	slog.Debug("hipDriverGetVersion", "version", version)
-	// 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
+	return strconv.Itoa(version), nil
 }
 
 func (hl *HipLib) HipGetDeviceCount() int {

gpu/amd_linux.go

@@ -8,7 +8,6 @@ import (
 	"log/slog"
 	"os"
 	"path/filepath"
-	"regexp"
 	"slices"
 	"strconv"
 	"strings"
@@ -26,12 +25,12 @@ const (
 	// Prefix with the node dir
 	GPUTotalMemoryFileGlob = "mem_banks/*/properties" // size_in_bytes line
 	GPUUsedMemoryFileGlob  = "mem_banks/*/used_memory"
+	RocmStandardLocation   = "/opt/rocm/lib"
 )
 
 var (
 	// Used to validate if the given ROCm lib is usable
-	ROCmLibGlobs          = []string{"libhipblas.so.2*", "rocblas"} // TODO - probably include more coverage of files here...
-	RocmStandardLocations = []string{"/opt/rocm/lib", "/usr/lib64"}
+	ROCmLibGlobs = []string{"libhipblas.so.2*", "rocblas"} // TODO - probably include more coverage of files here...
 )
 
 // Gather GPU information from the amdgpu driver if any supported GPUs are detected
@@ -42,8 +41,10 @@ func AMDGetGPUInfo() []GpuInfo {
 	}
 
 	// Opportunistic logging of driver version to aid in troubleshooting
-	driverMajor, driverMinor, err := AMDDriverVersion()
-	if err != nil {
+	ver, err := AMDDriverVersion()
+	if err == nil {
+		slog.Info("AMD Driver: " + ver)
+	} else {
 		// TODO - if we see users crash and burn with the upstreamed kernel this can be adjusted to hard-fail rocm support and fallback to CPU
 		slog.Warn("ollama recommends running the https://www.amd.com/en/support/linux-drivers", "error", err)
 	}
@@ -90,7 +91,6 @@ func AMDGetGPUInfo() []GpuInfo {
 		scanner := bufio.NewScanner(fp)
 		isCPU := false
 		var major, minor, patch uint64
-		var vendor, device uint64
 		for scanner.Scan() {
 			line := strings.TrimSpace(scanner.Text())
 			// Note: we could also use "cpu_cores_count X" where X is greater than zero to detect CPUs
@@ -118,26 +118,6 @@ func AMDGetGPUInfo() []GpuInfo {
 					slog.Debug("malformed int " + line)
 					continue
 				}
-			} else if strings.HasPrefix(line, "vendor_id") {
-				ver := strings.Fields(line)
-				if len(ver) != 2 {
-					slog.Debug("malformed vendor_id", "vendor_id", line)
-					continue
-				}
-				vendor, err = strconv.ParseUint(ver[1], 10, 32)
-				if err != nil {
-					slog.Debug("malformed vendor_id" + line)
-				}
-			} else if strings.HasPrefix(line, "device_id") {
-				ver := strings.Fields(line)
-				if len(ver) != 2 {
-					slog.Debug("malformed device_id", "device_id", line)
-					continue
-				}
-				device, err = strconv.ParseUint(ver[1], 10, 32)
-				if err != nil {
-					slog.Debug("malformed device_id" + line)
-				}
 			}
 
 			// TODO - any other properties we want to extract and record?
@@ -160,7 +140,7 @@ func AMDGetGPUInfo() []GpuInfo {
 		}
 
 		if int(major) < RocmComputeMin {
-			slog.Warn(fmt.Sprintf("amdgpu too old gfx%d%x%x", major, minor, patch), "gpu", gpuID)
+			slog.Warn(fmt.Sprintf("amdgpu too old gfx%d%d%x", major, minor, patch), "gpu", gpuID)
 			continue
 		}
 
@@ -230,29 +210,24 @@ func AMDGetGPUInfo() []GpuInfo {
 
 		// iGPU detection, remove this check once we can support an iGPU variant of the rocm library
 		if totalMemory < IGPUMemLimit {
-			slog.Info("unsupported Radeon iGPU detected skipping", "id", gpuID, "total", format.HumanBytes2(totalMemory))
+			slog.Info("amdgpu appears to be an iGPU, skipping", "gpu", gpuID, "total", format.HumanBytes2(totalMemory))
 			continue
 		}
-		var name string
-		// TODO - PCI ID lookup
-		if vendor > 0 && device > 0 {
-			name = fmt.Sprintf("%04x:%04x", vendor, device)
-		}
 
-		slog.Debug("amdgpu memory", "gpu", gpuID, "total", format.HumanBytes2(totalMemory))
-		slog.Debug("amdgpu memory", "gpu", gpuID, "available", format.HumanBytes2(totalMemory-usedMemory))
+		slog.Info("amdgpu memory", "gpu", gpuID, "total", format.HumanBytes2(totalMemory))
+		slog.Info("amdgpu memory", "gpu", gpuID, "available", format.HumanBytes2(totalMemory-usedMemory))
 		gpuInfo := GpuInfo{
 			Library: "rocm",
 			memInfo: memInfo{
 				TotalMemory: totalMemory,
 				FreeMemory:  (totalMemory - usedMemory),
 			},
-			ID:            fmt.Sprintf("%d", gpuID),
-			Name:          name,
-			Compute:       fmt.Sprintf("gfx%d%x%x", major, minor, patch),
+			ID: fmt.Sprintf("%d", gpuID),
+			// Name: not exposed in sysfs directly, would require pci device id lookup
+			Major:         int(major),
+			Minor:         int(minor),
+			Patch:         int(patch),
 			MinimumMemory: rocmMinimumMemory,
-			DriverMajor:   driverMajor,
-			DriverMinor:   driverMinor,
 		}
 
 		// If the user wants to filter to a subset of devices, filter out if we aren't a match
@@ -291,7 +266,7 @@ func AMDGetGPUInfo() []GpuInfo {
 				}
 				slog.Debug("rocm supported GPUs", "types", supported)
 			}
-			gfx := gpuInfo.Compute
+			gfx := fmt.Sprintf("gfx%d%d%x", gpuInfo.Major, gpuInfo.Minor, gpuInfo.Patch)
 			if !slices.Contains[[]string, string](supported, gfx) {
 				slog.Warn("amdgpu is not supported", "gpu", gpuInfo.ID, "gpu_type", gfx, "library", libDir, "supported_types", supported)
 				// TODO - consider discrete markdown just for ROCM troubleshooting?
@@ -301,7 +276,7 @@ func AMDGetGPUInfo() []GpuInfo {
 				slog.Info("amdgpu is supported", "gpu", gpuInfo.ID, "gpu_type", gfx)
 			}
 		} else {
-			slog.Info("skipping rocm gfx compatibility check", "HSA_OVERRIDE_GFX_VERSION", gfxOverride)
+			slog.Debug("skipping rocm gfx compatibility check with HSA_OVERRIDE_GFX_VERSION=" + gfxOverride)
 		}
 
 		// The GPU has passed all the verification steps and is supported
@@ -347,34 +322,19 @@ func AMDValidateLibDir() (string, error) {
 	return "", fmt.Errorf("no suitable rocm found, falling back to CPU")
 }
 
-func AMDDriverVersion() (driverMajor, driverMinor int, err error) {
-	_, err = os.Stat(DriverVersionFile)
+func AMDDriverVersion() (string, error) {
+	_, err := os.Stat(DriverVersionFile)
 	if err != nil {
-		return 0, 0, fmt.Errorf("amdgpu version file missing: %s %w", DriverVersionFile, err)
+		return "", fmt.Errorf("amdgpu version file missing: %s %w", DriverVersionFile, err)
 	}
 	fp, err := os.Open(DriverVersionFile)
 	if err != nil {
-		return 0, 0, err
+		return "", err
 	}
 	defer fp.Close()
 	verString, err := io.ReadAll(fp)
 	if err != nil {
-		return 0, 0, err
+		return "", err
 	}
-
-	pattern := `\A(\d+)\.(\d+).*`
-	regex := regexp.MustCompile(pattern)
-	match := regex.FindStringSubmatch(string(verString))
-	if len(match) < 2 {
-		return 0, 0, fmt.Errorf("malformed version string %s", string(verString))
-	}
-	driverMajor, err = strconv.Atoi(match[1])
-	if err != nil {
-		return 0, 0, err
-	}
-	driverMinor, err = strconv.Atoi(match[2])
-	if err != nil {
-		return 0, 0, err
-	}
-	return driverMajor, driverMinor, nil
+	return strings.TrimSpace(string(verString)), nil
 }

gpu/amd_windows.go

@@ -7,12 +7,14 @@ import (
 	"os"
 	"path/filepath"
 	"slices"
+	"strconv"
 	"strings"
 
 	"github.com/ollama/ollama/format"
 )
 
 const (
+	RocmStandardLocation = "C:\\Program Files\\AMD\\ROCm\\5.7\\bin" // TODO glob?
 
 	// TODO  We're lookinng for this exact name to detect iGPUs since hipGetDeviceProperties never reports integrated==true
 	iGPUName = "AMD Radeon(TM) Graphics"
@@ -20,8 +22,7 @@ const (
 
 var (
 	// Used to validate if the given ROCm lib is usable
-	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?
+	ROCmLibGlobs = []string{"hipblas.dll", "rocblas"} // TODO - probably include more coverage of files here...
 )
 
 func AMDGetGPUInfo() []GpuInfo {
@@ -33,12 +34,13 @@ func AMDGetGPUInfo() []GpuInfo {
 	}
 	defer hl.Release()
 
-	// 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)
-	// }
+	ver, err := hl.AMDDriverVersion()
+	if err == nil {
+		slog.Info("AMD Driver: " + ver)
+	} else {
+		// 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()
@@ -60,10 +62,10 @@ func AMDGetGPUInfo() []GpuInfo {
 			return nil
 		}
 	} else {
-		slog.Info("skipping rocm gfx compatibility check", "HSA_OVERRIDE_GFX_VERSION", gfxOverride)
+		slog.Debug("skipping rocm gfx compatibility check with HSA_OVERRIDE_GFX_VERSION=" + gfxOverride)
 	}
 
-	slog.Debug("detected hip devices", "count", count)
+	slog.Info("detected hip devices", "count", count)
 	// TODO how to determine the underlying device ID when visible devices is causing this to subset?
 	for i := 0; i < count; i++ {
 		err = hl.HipSetDevice(i)
@@ -83,11 +85,18 @@ func AMDGetGPUInfo() []GpuInfo {
 		// Can luid be used on windows for setting visible devices (and is it actually set?)
 		n = bytes.IndexByte(props.GcnArchName[:], 0)
 		gfx := string(props.GcnArchName[:n])
-		slog.Debug("hip device", "id", i, "name", name, "gfx", gfx)
+		slog.Info("hip device", "id", i, "name", name, "gfx", gfx)
+		var major, minor, patch string
+		switch len(gfx) {
+		case 6:
+			major, minor, patch = gfx[3:4], gfx[4:5], gfx[5:]
+		case 7:
+			major, minor, patch = gfx[3:5], gfx[5:6], gfx[6:]
+		}
 		//slog.Info(fmt.Sprintf("[%d] Integrated: %d", i, props.iGPU)) // DOESN'T REPORT CORRECTLY!  Always 0
 		// TODO  Why isn't props.iGPU accurate!?
 		if strings.EqualFold(name, iGPUName) {
-			slog.Info("unsupported Radeon iGPU detected skipping", "id", i, "name", name, "gfx", gfx)
+			slog.Info("iGPU detected skipping", "id", i)
 			continue
 		}
 		if gfxOverride == "" {
@@ -97,7 +106,7 @@ func AMDGetGPUInfo() []GpuInfo {
 				slog.Warn("See https://github.com/ollama/ollama/blob/main/docs/troubleshooting.md for HSA_OVERRIDE_GFX_VERSION usage")
 				continue
 			} else {
-				slog.Debug("amdgpu is supported", "gpu", i, "gpu_type", gfx)
+				slog.Info("amdgpu is supported", "gpu", i, "gpu_type", gfx)
 			}
 		}
 
@@ -115,8 +124,8 @@ func AMDGetGPUInfo() []GpuInfo {
 
 		// TODO revisit this once ROCm v6 is available on windows.
 		// v5.7 only reports VRAM used by this process, so it's completely wrong and unusable
-		slog.Debug("amdgpu memory", "gpu", i, "total", format.HumanBytes2(totalMemory))
-		slog.Debug("amdgpu memory", "gpu", i, "available", format.HumanBytes2(freeMemory))
+		slog.Info("amdgpu memory", "gpu", i, "total", format.HumanBytes2(totalMemory))
+		slog.Info("amdgpu memory", "gpu", i, "available", format.HumanBytes2(freeMemory))
 		gpuInfo := GpuInfo{
 			Library: "rocm",
 			memInfo: memInfo{
@@ -126,12 +135,31 @@ func AMDGetGPUInfo() []GpuInfo {
 			ID:             fmt.Sprintf("%d", i), // TODO this is probably wrong if we specify visible devices
 			DependencyPath: libDir,
 			MinimumMemory:  rocmMinimumMemory,
-			Name:           name,
-			Compute:        gfx,
-
-			// 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,
+		}
+		if major != "" {
+			gpuInfo.Major, err = strconv.Atoi(major)
+			if err != nil {
+				slog.Info("failed to parse version", "version", gfx, "error", err)
+			}
+		}
+		if minor != "" {
+			gpuInfo.Minor, err = strconv.Atoi(minor)
+			if err != nil {
+				slog.Info("failed to parse version", "version", gfx, "error", err)
+			}
+		}
+		if patch != "" {
+			// Patch rev is hex; e.g. gfx90a
+			p, err := strconv.ParseInt(patch, 16, 0)
+			if err != nil {
+				slog.Info("failed to parse version", "version", gfx, "error", err)
+			} else {
+				gpuInfo.Patch = int(p)
+			}
+		}
+		if gpuInfo.Major < RocmComputeMin {
+			slog.Warn(fmt.Sprintf("amdgpu [%s] too old gfx%d%d%x", gpuInfo.ID, gpuInfo.Major, gpuInfo.Minor, gpuInfo.Patch))
+			continue
 		}
 
 		resp = append(resp, gpuInfo)

gpu/assets.go

@@ -12,8 +12,6 @@ import (
 	"sync"
 	"syscall"
 	"time"
-
-	"github.com/ollama/ollama/server/envconfig"
 )
 
 var (
@@ -26,8 +24,29 @@ func PayloadsDir() (string, error) {
 	defer lock.Unlock()
 	var err error
 	if payloadsDir == "" {
-		runnersDir := envconfig.RunnersDir
-
+		runnersDir := os.Getenv("OLLAMA_RUNNERS_DIR")
+		// On Windows we do not carry the payloads inside the main executable
+		if runtime.GOOS == "windows" && runnersDir == "" {
+			appExe, err := os.Executable()
+			if err != nil {
+				slog.Error("failed to lookup executable path", "error", err)
+				return "", err
+			}
+			// Try a few variations to improve developer experience when building from source in the local tree
+			for _, d := range []string{".", "windows-" + runtime.GOARCH, "dist\\windows-" + runtime.GOARCH} {
+				candidate := filepath.Join(filepath.Dir(appExe), d, "ollama_runners")
+				_, err := os.Stat(candidate)
+				if err == nil {
+					runnersDir = candidate
+					break
+				}
+			}
+			if runnersDir == "" {
+				err = fmt.Errorf("unable to locate llm runner directory.  Set OLLAMA_RUNNERS_DIR to the location of 'ollama_runners'")
+				slog.Error("incomplete distribution", "error", err)
+				return "", err
+			}
+		}
 		if runnersDir != "" {
 			payloadsDir = runnersDir
 			return payloadsDir, nil
@@ -35,7 +54,7 @@ func PayloadsDir() (string, error) {
 
 		// The remainder only applies on non-windows where we still carry payloads in the main executable
 		cleanupTmpDirs()
-		tmpDir := envconfig.TmpDir
+		tmpDir := os.Getenv("OLLAMA_TMPDIR")
 		if tmpDir == "" {
 			tmpDir, err = os.MkdirTemp("", "ollama")
 			if err != nil {
@@ -98,7 +117,7 @@ func cleanupTmpDirs() {
 func Cleanup() {
 	lock.Lock()
 	defer lock.Unlock()
-	runnersDir := envconfig.RunnersDir
+	runnersDir := os.Getenv("OLLAMA_RUNNERS_DIR")
 	if payloadsDir != "" && runnersDir == "" && runtime.GOOS != "windows" {
 		// We want to fully clean up the tmpdir parent of the payloads dir
 		tmpDir := filepath.Clean(filepath.Join(payloadsDir, ".."))

gpu/cpu_common.go

@@ -8,14 +8,14 @@ import (
 
 func GetCPUVariant() string {
 	if cpu.X86.HasAVX2 {
-		slog.Debug("CPU has AVX2")
+		slog.Info("CPU has AVX2")
 		return "avx2"
 	}
 	if cpu.X86.HasAVX {
-		slog.Debug("CPU has AVX")
+		slog.Info("CPU has AVX")
 		return "avx"
 	}
-	slog.Debug("CPU does not have vector extensions")
+	slog.Info("CPU does not have vector extensions")
 	// else LCD
 	return ""
 }

gpu/gpu.go

@@ -21,13 +21,11 @@ import (
 	"unsafe"
 
 	"github.com/ollama/ollama/format"
-	"github.com/ollama/ollama/server/envconfig"
 )
 
 type handles struct {
 	deviceCount int
 	cudart      *C.cudart_handle_t
-	nvcuda      *C.nvcuda_handle_t
 }
 
 const (
@@ -64,22 +62,6 @@ var CudartWindowsGlobs = []string{
 	"c:\\Program Files\\NVIDIA GPU Computing Toolkit\\CUDA\\v*\\bin\\cudart64_*.dll",
 }
 
-var NvcudaLinuxGlobs = []string{
-	"/usr/local/cuda*/targets/*/lib/libcuda.so*",
-	"/usr/lib/*-linux-gnu/nvidia/current/libcuda.so*",
-	"/usr/lib/*-linux-gnu/libcuda.so*",
-	"/usr/lib/wsl/lib/libcuda.so*",
-	"/usr/lib/wsl/drivers/*/libcuda.so*",
-	"/opt/cuda/lib*/libcuda.so*",
-	"/usr/local/cuda/lib*/libcuda.so*",
-	"/usr/lib*/libcuda.so*",
-	"/usr/local/lib*/libcuda.so*",
-}
-
-var NvcudaWindowsGlobs = []string{
-	"c:\\windows\\system*\\nvcuda.dll",
-}
-
 // Jetson devices have JETSON_JETPACK="x.y.z" factory set to the Jetpack version installed.
 // Included to drive logic for reducing Ollama-allocated overhead on L4T/Jetson devices.
 var CudaTegra string = os.Getenv("JETSON_JETPACK")
@@ -92,8 +74,6 @@ func initGPUHandles() *handles {
 	gpuHandles := &handles{}
 	var cudartMgmtName string
 	var cudartMgmtPatterns []string
-	var nvcudaMgmtName string
-	var nvcudaMgmtPatterns []string
 
 	tmpDir, _ := PayloadsDir()
 	switch runtime.GOOS {
@@ -102,9 +82,6 @@ func initGPUHandles() *handles {
 		localAppData := os.Getenv("LOCALAPPDATA")
 		cudartMgmtPatterns = []string{filepath.Join(localAppData, "Programs", "Ollama", cudartMgmtName)}
 		cudartMgmtPatterns = append(cudartMgmtPatterns, CudartWindowsGlobs...)
-		// Aligned with driver, we can't carry as payloads
-		nvcudaMgmtName = "nvcuda.dll"
-		nvcudaMgmtPatterns = NvcudaWindowsGlobs
 	case "linux":
 		cudartMgmtName = "libcudart.so*"
 		if tmpDir != "" {
@@ -112,30 +89,16 @@ func initGPUHandles() *handles {
 			cudartMgmtPatterns = []string{filepath.Join(tmpDir, "cuda*", cudartMgmtName)}
 		}
 		cudartMgmtPatterns = append(cudartMgmtPatterns, CudartLinuxGlobs...)
-		// Aligned with driver, we can't carry as payloads
-		nvcudaMgmtName = "libcuda.so*"
-		nvcudaMgmtPatterns = NvcudaLinuxGlobs
 	default:
 		return gpuHandles
 	}
 
-	slog.Debug("Detecting GPUs")
-	nvcudaLibPaths := FindGPULibs(nvcudaMgmtName, nvcudaMgmtPatterns)
-	if len(nvcudaLibPaths) > 0 {
-		deviceCount, nvcuda, libPath := LoadNVCUDAMgmt(nvcudaLibPaths)
-		if nvcuda != nil {
-			slog.Debug("detected GPUs", "count", deviceCount, "library", libPath)
-			gpuHandles.nvcuda = nvcuda
-			gpuHandles.deviceCount = deviceCount
-			return gpuHandles
-		}
-	}
-
+	slog.Info("Detecting GPUs")
 	cudartLibPaths := FindGPULibs(cudartMgmtName, cudartMgmtPatterns)
 	if len(cudartLibPaths) > 0 {
 		deviceCount, cudart, libPath := LoadCUDARTMgmt(cudartLibPaths)
 		if cudart != nil {
-			slog.Debug("detected GPUs", "library", libPath, "count", deviceCount)
+			slog.Info("detected GPUs", "library", libPath, "count", deviceCount)
 			gpuHandles.cudart = cudart
 			gpuHandles.deviceCount = deviceCount
 			return gpuHandles
@@ -155,9 +118,6 @@ func GetGPUInfo() GpuInfoList {
 		if gpuHandles.cudart != nil {
 			C.cudart_release(*gpuHandles.cudart)
 		}
-		if gpuHandles.nvcuda != nil {
-			C.nvcuda_release(*gpuHandles.nvcuda)
-		}
 	}()
 
 	// All our GPU builds on x86 have AVX enabled, so fallback to CPU if we don't detect at least AVX
@@ -166,12 +126,6 @@ func GetGPUInfo() GpuInfoList {
 		slog.Warn("CPU does not have AVX or AVX2, disabling GPU support.")
 	}
 
-	// On windows we bundle the nvidia library one level above the runner dir
-	depPath := ""
-	if runtime.GOOS == "windows" && envconfig.RunnersDir != "" {
-		depPath = filepath.Dir(envconfig.RunnersDir)
-	}
-
 	var memInfo C.mem_info_t
 	resp := []GpuInfo{}
 
@@ -184,15 +138,7 @@ func GetGPUInfo() GpuInfoList {
 		gpuInfo := GpuInfo{
 			Library: "cuda",
 		}
-		var driverMajor int
-		var driverMinor int
-		if gpuHandles.cudart != nil {
-			C.cudart_check_vram(*gpuHandles.cudart, C.int(i), &memInfo)
-		} else {
-			C.nvcuda_check_vram(*gpuHandles.nvcuda, C.int(i), &memInfo)
-			driverMajor = int(gpuHandles.nvcuda.driver_major)
-			driverMinor = int(gpuHandles.nvcuda.driver_minor)
-		}
+		C.cudart_check_vram(*gpuHandles.cudart, C.int(i), &memInfo)
 		if memInfo.err != nil {
 			slog.Info("error looking up nvidia GPU memory", "error", C.GoString(memInfo.err))
 			C.free(unsafe.Pointer(memInfo.err))
@@ -205,12 +151,9 @@ func GetGPUInfo() GpuInfoList {
 		gpuInfo.TotalMemory = uint64(memInfo.total)
 		gpuInfo.FreeMemory = uint64(memInfo.free)
 		gpuInfo.ID = C.GoString(&memInfo.gpu_id[0])
-		gpuInfo.Compute = fmt.Sprintf("%d.%d", memInfo.major, memInfo.minor)
+		gpuInfo.Major = int(memInfo.major)
+		gpuInfo.Minor = int(memInfo.minor)
 		gpuInfo.MinimumMemory = cudaMinimumMemory
-		gpuInfo.DependencyPath = depPath
-		gpuInfo.Name = C.GoString(&memInfo.gpu_name[0])
-		gpuInfo.DriverMajor = int(driverMajor)
-		gpuInfo.DriverMinor = int(driverMinor)
 
 		// TODO potentially sort on our own algorithm instead of what the underlying GPU library does...
 		resp = append(resp, gpuInfo)
@@ -253,10 +196,9 @@ func GetCPUMem() (memInfo, error) {
 	return ret, nil
 }
 
-func FindGPULibs(baseLibName string, defaultPatterns []string) []string {
+func FindGPULibs(baseLibName string, patterns []string) []string {
 	// Multiple GPU libraries may exist, and some may not work, so keep trying until we exhaust them
 	var ldPaths []string
-	var patterns []string
 	gpuLibPaths := []string{}
 	slog.Debug("Searching for GPU library", "name", baseLibName)
 
@@ -276,14 +218,8 @@ func FindGPULibs(baseLibName string, defaultPatterns []string) []string {
 		}
 		patterns = append(patterns, filepath.Join(d, baseLibName+"*"))
 	}
-	patterns = append(patterns, defaultPatterns...)
 	slog.Debug("gpu library search", "globs", patterns)
 	for _, pattern := range patterns {
-
-		// Nvidia PhysX known to return bogus results
-		if strings.Contains(pattern, "PhysX") {
-			slog.Debug("skipping PhysX cuda library path", "path", pattern)
-		}
 		// Ignore glob discovery errors
 		matches, _ := filepath.Glob(pattern)
 		for _, match := range matches {
@@ -331,25 +267,8 @@ func LoadCUDARTMgmt(cudartLibPaths []string) (int, *C.cudart_handle_t, string) {
 	return 0, nil, ""
 }
 
-func LoadNVCUDAMgmt(nvcudaLibPaths []string) (int, *C.nvcuda_handle_t, string) {
-	var resp C.nvcuda_init_resp_t
-	resp.ch.verbose = getVerboseState()
-	for _, libPath := range nvcudaLibPaths {
-		lib := C.CString(libPath)
-		defer C.free(unsafe.Pointer(lib))
-		C.nvcuda_init(lib, &resp)
-		if resp.err != nil {
-			slog.Debug("Unable to load nvcuda", "library", libPath, "error", C.GoString(resp.err))
-			C.free(unsafe.Pointer(resp.err))
-		} else {
-			return int(resp.num_devices), &resp.ch, libPath
-		}
-	}
-	return 0, nil, ""
-}
-
 func getVerboseState() C.uint16_t {
-	if envconfig.Debug {
+	if debug := os.Getenv("OLLAMA_DEBUG"); debug != "" {
 		return C.uint16_t(1)
 	}
 	return C.uint16_t(0)

gpu/gpu_darwin.go

@@ -10,12 +10,6 @@ package gpu
 import "C"
 import (
 	"runtime"
-
-	"github.com/ollama/ollama/format"
-)
-
-const (
-	metalMinimumMemory = 512 * format.MebiByte
 )
 
 func GetGPUInfo() GpuInfoList {
@@ -38,7 +32,7 @@ func GetGPUInfo() GpuInfoList {
 	// TODO is there a way to gather actual allocated video memory? (currentAllocatedSize doesn't work)
 	info.FreeMemory = info.TotalMemory
 
-	info.MinimumMemory = metalMinimumMemory
+	info.MinimumMemory = 0
 	return []GpuInfo{info}
 }
 

gpu/gpu_info.h

@@ -39,19 +39,16 @@ extern "C" {
 #endif
 
 #define GPU_ID_LEN 64
-#define GPU_NAME_LEN 96
 
 typedef struct mem_info {
   char *err;  // If non-nill, caller responsible for freeing
   char gpu_id[GPU_ID_LEN];
-  char gpu_name[GPU_NAME_LEN];
   uint64_t total;
   uint64_t free;
 
   // Compute Capability
   int major; 
   int minor;
-  int patch;
 } mem_info_t;
 
 void cpu_check_ram(mem_info_t *resp);
@@ -61,7 +58,6 @@ void cpu_check_ram(mem_info_t *resp);
 #endif
 
 #include "gpu_info_cudart.h"
-#include "gpu_info_nvcuda.h"
 
 #endif  // __GPU_INFO_H__
 #endif  // __APPLE__

gpu/gpu_info_cpu.c

@@ -10,6 +10,8 @@ void cpu_check_ram(mem_info_t *resp) {
   if (GlobalMemoryStatusEx(&info) != 0) {
     resp->total = info.ullTotalPhys;
     resp->free = info.ullAvailPhys;
+    resp->major = 0;
+    resp->minor = 0;
     snprintf(&resp->gpu_id[0], GPU_ID_LEN, "0");
   } else {
     resp->err = LOAD_ERR();
@@ -29,6 +31,8 @@ void cpu_check_ram(mem_info_t *resp) {
   } else {
     resp->total = info.totalram * info.mem_unit;
     resp->free = info.freeram * info.mem_unit;
+    resp->major = 0;
+    resp->minor = 0;
     snprintf(&resp->gpu_id[0], GPU_ID_LEN, "0");
   }
   return;

gpu/gpu_info_cudart.h

@@ -6,9 +6,9 @@
 // Just enough typedef's to dlopen/dlsym for memory information
 typedef enum cudartReturn_enum {
   CUDART_SUCCESS = 0,
-  CUDART_ERROR_INVALID_VALUE = 1,
-  CUDART_ERROR_MEMORY_ALLOCATION = 2,
-  CUDART_ERROR_INSUFFICIENT_DRIVER = 35,
+  CUDA_ERROR_INVALID_VALUE = 1,
+  CUDA_ERROR_MEMORY_ALLOCATION = 2,
+  CUDA_ERROR_INSUFFICIENT_DRIVER = 35,
   // Other values omitted for now...
 } cudartReturn_t;
 

gpu/gpu_info_nvcuda.c

@@ -1,207 +0,0 @@
-#ifndef __APPLE__  // TODO - maybe consider nvidia support on intel macs?
-
-#include <string.h>
-#include "gpu_info_nvcuda.h"
-
-void nvcuda_init(char *nvcuda_lib_path, nvcuda_init_resp_t *resp) {
-  CUresult ret;
-  resp->err = NULL;
-  resp->num_devices = 0;
-  const int buflen = 256;
-  char buf[buflen + 1];
-  int i;
-
-  struct lookup {
-    char *s;
-    void **p;
-  } l[] = {
-   
-      {"cuInit", (void *)&resp->ch.cuInit},
-      {"cuDriverGetVersion", (void *)&resp->ch.cuDriverGetVersion},
-      {"cuDeviceGetCount", (void *)&resp->ch.cuDeviceGetCount},
-      {"cuDeviceGet", (void *)&resp->ch.cuDeviceGet},
-      {"cuDeviceGetAttribute", (void *)&resp->ch.cuDeviceGetAttribute},
-      {"cuDeviceGetUuid", (void *)&resp->ch.cuDeviceGetUuid},
-      {"cuDeviceGetName", (void *)&resp->ch.cuDeviceGetName},
-      {"cuCtxCreate_v3", (void *)&resp->ch.cuCtxCreate_v3},
-      {"cuMemGetInfo_v2", (void *)&resp->ch.cuMemGetInfo_v2},
-      {"cuCtxDestroy", (void *)&resp->ch.cuCtxDestroy},
-      {NULL, NULL},
-  };
-
-  resp->ch.handle = LOAD_LIBRARY(nvcuda_lib_path, RTLD_LAZY);
-  if (!resp->ch.handle) {
-    char *msg = LOAD_ERR();
-    LOG(resp->ch.verbose, "library %s load err: %s\n", nvcuda_lib_path, msg);
-    snprintf(buf, buflen,
-            "Unable to load %s library to query for Nvidia GPUs: %s",
-            nvcuda_lib_path, msg);
-    free(msg);
-    resp->err = strdup(buf);
-    return;
-  }
-
-  for (i = 0; l[i].s != NULL; i++) {
-    *l[i].p = LOAD_SYMBOL(resp->ch.handle, l[i].s);
-    if (!*l[i].p) {
-      char *msg = LOAD_ERR();
-      LOG(resp->ch.verbose, "dlerr: %s\n", msg);
-      UNLOAD_LIBRARY(resp->ch.handle);
-      resp->ch.handle = NULL;
-      snprintf(buf, buflen, "symbol lookup for %s failed: %s", l[i].s,
-              msg);
-      free(msg);
-      resp->err = strdup(buf);
-      return;
-    }
-  }
-
-  ret = (*resp->ch.cuInit)(0);
-  if (ret != CUDA_SUCCESS) {
-    LOG(resp->ch.verbose, "cuInit err: %d\n", ret);
-    UNLOAD_LIBRARY(resp->ch.handle);
-    resp->ch.handle = NULL;
-    if (ret == CUDA_ERROR_INSUFFICIENT_DRIVER) {
-      resp->err = strdup("your nvidia driver is too old or missing.  If you have a CUDA GPU please upgrade to run ollama");
-      return;
-    }
-    snprintf(buf, buflen, "nvcuda init failure: %d", ret);
-    resp->err = strdup(buf);
-    return;
-  }
-
-  int version = 0;
-  resp->ch.driver_major = 0;
-  resp->ch.driver_minor = 0;
-
-  // Report driver version if we're in verbose mode, ignore errors
-  ret = (*resp->ch.cuDriverGetVersion)(&version);
-  if (ret != CUDA_SUCCESS) {
-    LOG(resp->ch.verbose, "cuDriverGetVersion failed: %d\n", ret);
-  } else {
-    resp->ch.driver_major = version / 1000;
-    resp->ch.driver_minor = (version - (resp->ch.driver_major * 1000)) / 10;
-    LOG(resp->ch.verbose, "CUDA driver version: %d.%d\n", resp->ch.driver_major, resp->ch.driver_minor);
-  }
-
-  ret = (*resp->ch.cuDeviceGetCount)(&resp->num_devices);
-  if (ret != CUDA_SUCCESS) {
-    LOG(resp->ch.verbose, "cuDeviceGetCount err: %d\n", ret);
-    UNLOAD_LIBRARY(resp->ch.handle);
-    resp->ch.handle = NULL;
-    snprintf(buf, buflen, "unable to get device count: %d", ret);
-    resp->err = strdup(buf);
-    return;
-  }
-}
-
-const int buflen = 256;
-void nvcuda_check_vram(nvcuda_handle_t h, int i, mem_info_t *resp) {
-  resp->err = NULL;
-  nvcudaMemory_t memInfo = {0,0};
-  CUresult ret;
-  CUdevice device = -1;
-  CUcontext ctx = NULL;
-  char buf[buflen + 1];
-  CUuuid uuid = {0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0};
-
-  if (h.handle == NULL) {
-    resp->err = strdup("nvcuda handle isn't initialized");
-    return;
-  }
-
-  ret = (*h.cuDeviceGet)(&device, i);
-  if (ret != CUDA_SUCCESS) {
-    snprintf(buf, buflen, "nvcuda device failed to initialize");
-    resp->err = strdup(buf);
-    return;
-  }
-
-  int major = 0;
-  int minor = 0;
-  ret = (*h.cuDeviceGetAttribute)(&major, CU_DEVICE_ATTRIBUTE_COMPUTE_CAPABILITY_MAJOR, device);
-  if (ret != CUDA_SUCCESS) {
-    LOG(h.verbose, "[%d] device major lookup failure: %d\n", i, ret);
-  } else {
-    ret = (*h.cuDeviceGetAttribute)(&minor, CU_DEVICE_ATTRIBUTE_COMPUTE_CAPABILITY_MINOR, device);
-    if (ret != CUDA_SUCCESS) {
-      LOG(h.verbose, "[%d] device minor lookup failure: %d\n", i, ret);
-    } else {
-      resp->minor = minor;  
-      resp->major = major;  
-    }
-  }
-
-  ret = (*h.cuDeviceGetUuid)(&uuid, device);
-  if (ret != CUDA_SUCCESS) {
-    LOG(h.verbose, "[%d] device uuid lookup failure: %d\n", i, ret);
-    snprintf(&resp->gpu_id[0], GPU_ID_LEN, "%d", i);
-  } else {
-    // GPU-d110a105-ac29-1d54-7b49-9c90440f215b
-    snprintf(&resp->gpu_id[0], GPU_ID_LEN,
-        "GPU-%02x%02x%02x%02x-%02x%02x-%02x%02x-%02x%02x-%02x%02x%02x%02x%02x%02x",
-        uuid.bytes[0],
-        uuid.bytes[1],
-        uuid.bytes[2],
-        uuid.bytes[3],
-        uuid.bytes[4],
-        uuid.bytes[5],
-        uuid.bytes[6],
-        uuid.bytes[7],
-        uuid.bytes[8],
-        uuid.bytes[9],
-        uuid.bytes[10],
-        uuid.bytes[11],
-        uuid.bytes[12],
-        uuid.bytes[13],
-        uuid.bytes[14],
-        uuid.bytes[15]
-      );
-  }
-
-  ret = (*h.cuDeviceGetName)(&resp->gpu_name[0], GPU_NAME_LEN, device);
-  if (ret != CUDA_SUCCESS) {
-    LOG(h.verbose, "[%d] device name lookup failure: %d\n", i, ret);
-    resp->gpu_name[0] = '\0';
-  }
-
-  // To get memory we have to set (and release) a context
-  ret = (*h.cuCtxCreate_v3)(&ctx, NULL, 0, 0, device);
-  if (ret != CUDA_SUCCESS) {
-    snprintf(buf, buflen, "nvcuda failed to get primary device context %d", ret);
-    resp->err = strdup(buf);
-    return;
-  }
-
-  ret = (*h.cuMemGetInfo_v2)(&memInfo.free, &memInfo.total);
-  if (ret != CUDA_SUCCESS) {
-    snprintf(buf, buflen, "nvcuda device memory info lookup failure %d", ret);
-    resp->err = strdup(buf);
-    // Best effort on failure...
-    (*h.cuCtxDestroy)(ctx);
-    return;
-  }
-
-  resp->total = memInfo.total;
-  resp->free = memInfo.free;
-
-  LOG(h.verbose, "[%s] CUDA totalMem %lu mb\n", resp->gpu_id, resp->total / 1024 / 1024);
-  LOG(h.verbose, "[%s] CUDA freeMem %lu mb\n", resp->gpu_id, resp->free / 1024 / 1024);
-  LOG(h.verbose, "[%s] Compute Capability %d.%d\n", resp->gpu_id, resp->major, resp->minor);
-
-  
-
-  ret = (*h.cuCtxDestroy)(ctx);
-  if (ret != CUDA_SUCCESS) {
-    LOG(1, "nvcuda failed to release primary device context %d", ret);
-  }
-}
-
-void nvcuda_release(nvcuda_handle_t h) {
-  LOG(h.verbose, "releasing nvcuda library\n");
-  UNLOAD_LIBRARY(h.handle);
-  // TODO and other context release logic?
-  h.handle = NULL;
-}
-
-#endif  // __APPLE__

gpu/gpu_info_nvcuda.h

@@ -1,74 +0,0 @@
-#ifndef __APPLE__
-#ifndef __GPU_INFO_NVCUDA_H__
-#define __GPU_INFO_NVCUDA_H__
-#include "gpu_info.h"
-
-// Just enough typedef's to dlopen/dlsym for memory information
-typedef enum cudaError_enum {
-  CUDA_SUCCESS = 0,
-  CUDA_ERROR_INVALID_VALUE = 1,
-  CUDA_ERROR_MEMORY_ALLOCATION = 2,
-  CUDA_ERROR_NOT_INITIALIZED = 3,
-  CUDA_ERROR_INSUFFICIENT_DRIVER = 35,
-  // Other values omitted for now...
-} CUresult;
-
-typedef enum CUdevice_attribute_enum {
-  CU_DEVICE_ATTRIBUTE_COMPUTE_CAPABILITY_MAJOR = 75,
-  CU_DEVICE_ATTRIBUTE_COMPUTE_CAPABILITY_MINOR = 76,
-
-  // TODO - not yet wired up but may be useful for Jetson or other
-  // integrated GPU scenarios with shared memory
-  CU_DEVICE_ATTRIBUTE_INTEGRATED = 18
-
-} CUdevice_attribute;
-
-typedef void *nvcudaDevice_t;  // Opaque is sufficient
-typedef struct nvcudaMemory_st {
-  uint64_t total;
-  uint64_t free;
-} nvcudaMemory_t;
-
-typedef struct nvcudaDriverVersion {
-  int major;
-  int minor;
-} nvcudaDriverVersion_t;
-
-typedef struct CUuuid_st {
-    unsigned char bytes[16];
-} CUuuid;
-
-typedef int CUdevice;
-typedef void* CUcontext;
-
-typedef struct nvcuda_handle {
-  void *handle;
-  uint16_t verbose;
-  int driver_major;
-  int driver_minor;
-  CUresult (*cuInit)(unsigned int Flags);
-  CUresult (*cuDriverGetVersion)(int *driverVersion);
-  CUresult (*cuDeviceGetCount)(int *);
-  CUresult (*cuDeviceGet)(CUdevice* device, int ordinal);
-  CUresult (*cuDeviceGetAttribute)(int* pi, CUdevice_attribute attrib, CUdevice dev);
-  CUresult (*cuDeviceGetUuid)(CUuuid* uuid, CUdevice dev); // signature compatible with cuDeviceGetUuid_v2
-  CUresult (*cuDeviceGetName)(char *name, int len, CUdevice dev);
-
-  // Context specific aspects
-  CUresult (*cuCtxCreate_v3)(CUcontext* pctx, void *params, int len, unsigned int flags, CUdevice dev);
-  CUresult (*cuMemGetInfo_v2)(uint64_t* free, uint64_t* total);
-  CUresult (*cuCtxDestroy)(CUcontext ctx);
-} nvcuda_handle_t;
-
-typedef struct nvcuda_init_resp {
-  char *err;  // If err is non-null handle is invalid
-  nvcuda_handle_t ch;
-  int num_devices;
-} nvcuda_init_resp_t;
-
-void nvcuda_init(char *nvcuda_lib_path, nvcuda_init_resp_t *resp);
-void nvcuda_check_vram(nvcuda_handle_t ch, int device_id, mem_info_t *resp);
-void nvcuda_release(nvcuda_handle_t ch);
-
-#endif  // __GPU_INFO_NVCUDA_H__
-#endif  // __APPLE__

gpu/types.go

@@ -1,12 +1,5 @@
 package gpu
 
-import (
-	"fmt"
-	"log/slog"
-
-	"github.com/ollama/ollama/format"
-)
-
 type memInfo struct {
 	TotalMemory uint64 `json:"total_memory,omitempty"`
 	FreeMemory  uint64 `json:"free_memory,omitempty"`
@@ -27,13 +20,11 @@ type GpuInfo struct {
 	DependencyPath string `json:"lib_path,omitempty"`
 
 	// GPU information
-	ID      string `json:"gpu_id"`  // string to use for selection of this specific GPU
-	Name    string `json:"name"`    // user friendly name if available
-	Compute string `json:"compute"` // Compute Capability or gfx
-
-	// Driver Information - TODO no need to put this on each GPU
-	DriverMajor int `json:"driver_major,omitempty"`
-	DriverMinor int `json:"driver_minor,omitempty"`
+	ID    string `json:"gpu_id"`          // string to use for selection of this specific GPU
+	Name  string `json:"name"`            // user friendly name if available
+	Major int    `json:"major,omitempty"` // Major compatibility version (CC or gfx)
+	Minor int    `json:"minor,omitempty"` // Minor compatibility version (CC or gfx)
+	Patch int    `json:"patch,omitempty"` // Patch compatibility only matters on AMD
 
 	// TODO other performance capability info to help in scheduling decisions
 }
@@ -65,21 +56,6 @@ func (l GpuInfoList) ByLibrary() []GpuInfoList {
 	return resp
 }
 
-// Report the GPU information into the log an Info level
-func (l GpuInfoList) LogDetails() {
-	for _, g := range l {
-		slog.Info("inference compute",
-			"id", g.ID,
-			"library", g.Library,
-			"compute", g.Compute,
-			"driver", fmt.Sprintf("%d.%d", g.DriverMajor, g.DriverMinor),
-			"name", g.Name,
-			"total", format.HumanBytes2(g.TotalMemory),
-			"available", format.HumanBytes2(g.FreeMemory),
-		)
-	}
-}
-
 // Sort by Free Space
 type ByFreeMemory []GpuInfo
 

integration/embedding_test.go

@@ -0,0 +1,64 @@
+//go:build integration
+
+package integration
+
+import (
+	"context"
+	"net/http"
+	"testing"
+	"time"
+
+	"github.com/ollama/ollama/api"
+)
+
+func TestAllMiniLMEmbedding(t *testing.T) {
+	ctx, cancel := context.WithTimeout(context.Background(), 2*time.Minute)
+	defer cancel()
+
+	req := api.EmbeddingRequest{
+		Model:  "all-minilm",
+		Prompt: "why is the sky blue?",
+		Options: map[string]interface{}{
+			"temperature": 0,
+			"seed":        123,
+		},
+	}
+
+	res := EmbeddingTestHelper(ctx, t, &http.Client{}, req)
+
+	if len(res.Embedding) != 384 {
+		t.Fatalf("Expected 384 floats to be returned, got %v", len(res.Embedding))
+	}
+
+	if res.Embedding[0] != 0.146763876080513 {
+		t.Fatalf("Expected first embedding float to be 0.146763876080513, got %v", res.Embedding[0])
+	}
+}
+
+func TestAllMiniLMEmbeddings(t *testing.T) {
+	ctx, cancel := context.WithTimeout(context.Background(), 2*time.Minute)
+	defer cancel()
+
+	req := api.EmbeddingRequest{
+		Model:   "all-minilm",
+		Prompts: []string{"why is the sky blue?", "why is the sky blue?"},
+		Options: map[string]interface{}{
+			"temperature": 0,
+			"seed":        123,
+		},
+	}
+
+	res := EmbeddingTestHelper(ctx, t, &http.Client{}, req)
+
+	if len(res.Embeddings) != 2 {
+		t.Fatal("Expected 2 embeddings to be returned")
+	}
+
+	if len(res.Embeddings[0]) != 384 {
+		t.Fatalf("Expected first embedding to have 384 floats, got %v", len(res.Embeddings[0]))
+	}
+
+	if res.Embeddings[0][0] != 0.146763876080513 && res.Embeddings[1][0] != 0.146763876080513 {
+		t.Fatalf("Expected first embedding floats to be 0.146763876080513, got %v, %v", res.Embeddings[0][0], res.Embeddings[1][0])
+	}
+}

integration/max_queue_test.go

@@ -1,117 +0,0 @@
-//go:build integration
-
-package integration
-
-import (
-	"context"
-	"errors"
-	"fmt"
-	"log/slog"
-	"os"
-	"strconv"
-	"strings"
-	"sync"
-	"testing"
-	"time"
-
-	"github.com/ollama/ollama/api"
-	"github.com/stretchr/testify/require"
-)
-
-func TestMaxQueue(t *testing.T) {
-	// Note: This test can be quite slow when running in CPU mode, so keep the threadCount low unless your on GPU
-	// Also note that by default Darwin can't sustain > ~128 connections without adjusting limits
-	threadCount := 32
-	mq := os.Getenv("OLLAMA_MAX_QUEUE")
-	if mq != "" {
-		var err error
-		threadCount, err = strconv.Atoi(mq)
-		require.NoError(t, err)
-	} else {
-		os.Setenv("OLLAMA_MAX_QUEUE", fmt.Sprintf("%d", threadCount))
-	}
-
-	req := api.GenerateRequest{
-		Model:  "orca-mini",
-		Prompt: "write a long historical fiction story about christopher columbus.  use at least 10 facts from his actual journey",
-		Options: map[string]interface{}{
-			"seed":        42,
-			"temperature": 0.0,
-		},
-	}
-	resp := []string{"explore", "discover", "ocean"}
-
-	// CPU mode takes much longer at the limit with a large queue setting
-	ctx, cancel := context.WithTimeout(context.Background(), 5*time.Minute)
-	defer cancel()
-	client, _, cleanup := InitServerConnection(ctx, t)
-	defer cleanup()
-
-	require.NoError(t, PullIfMissing(ctx, client, req.Model))
-
-	// Context for the worker threads so we can shut them down
-	// embedCtx, embedCancel := context.WithCancel(ctx)
-	embedCtx := ctx
-
-	var genwg sync.WaitGroup
-	go func() {
-		genwg.Add(1)
-		defer genwg.Done()
-		slog.Info("Starting generate request")
-		DoGenerate(ctx, t, client, req, resp, 45*time.Second, 5*time.Second)
-		slog.Info("generate completed")
-	}()
-
-	// Give the generate a chance to get started before we start hammering on embed requests
-	time.Sleep(5 * time.Millisecond)
-
-	threadCount += 10 // Add a few extra to ensure we push the queue past its limit
-	busyCount := 0
-	resetByPeerCount := 0
-	canceledCount := 0
-	succesCount := 0
-	counterMu := sync.Mutex{}
-	var embedwg sync.WaitGroup
-	for i := 0; i < threadCount; i++ {
-		go func(i int) {
-			embedwg.Add(1)
-			defer embedwg.Done()
-			slog.Info("embed started", "id", i)
-			embedReq := api.EmbeddingRequest{
-				Model:   req.Model,
-				Prompt:  req.Prompt,
-				Options: req.Options,
-			}
-			// Fresh client for every request
-			client, _ = GetTestEndpoint()
-
-			resp, genErr := client.Embeddings(embedCtx, &embedReq)
-			counterMu.Lock()
-			defer counterMu.Unlock()
-			switch {
-			case genErr == nil:
-				succesCount++
-				require.Greater(t, len(resp.Embedding), 5) // somewhat arbitrary, but sufficient to be reasonable
-			case errors.Is(genErr, context.Canceled):
-				canceledCount++
-			case strings.Contains(genErr.Error(), "busy"):
-				busyCount++
-			case strings.Contains(genErr.Error(), "connection reset by peer"):
-				resetByPeerCount++
-			default:
-				require.NoError(t, genErr, "%d request failed", i)
-			}
-
-			slog.Info("embed finished", "id", i)
-		}(i)
-	}
-	genwg.Wait()
-	slog.Info("generate done, waiting for embeds")
-	embedwg.Wait()
-
-	require.Equal(t, resetByPeerCount, 0, "Connections reset by peer, have you updated your fd and socket limits?")
-	require.True(t, busyCount > 0, "no requests hit busy error but some should have")
-	require.True(t, canceledCount == 0, "no requests should have been canceled due to timeout")
-
-	slog.Info("embeds completed", "success", succesCount, "busy", busyCount, "reset", resetByPeerCount, "canceled", canceledCount)
-}

integration/utils_test.go

@@ -5,6 +5,7 @@ package integration
 import (
 	"bytes"
 	"context"
+	"encoding/json"
 	"errors"
 	"fmt"
 	"io"
@@ -24,6 +25,7 @@ import (
 
 	"github.com/ollama/ollama/api"
 	"github.com/ollama/ollama/app/lifecycle"
+	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/require"
 )
 
@@ -107,7 +109,7 @@ func startServer(ctx context.Context, ollamaHost string) error {
 
 	if tmp := os.Getenv("OLLAMA_HOST"); tmp != ollamaHost {
 		slog.Info("setting env", "OLLAMA_HOST", ollamaHost)
-		t.Setenv("OLLAMA_HOST", ollamaHost)
+		os.Setenv("OLLAMA_HOST", ollamaHost)
 	}
 
 	slog.Info("starting server", "url", ollamaHost)
@@ -285,6 +287,7 @@ func DoGenerate(ctx context.Context, t *testing.T, client *api.Client, genReq ap
 // Generate a set of requests
 // By default each request uses orca-mini as the model
 func GenerateRequests() ([]api.GenerateRequest, [][]string) {
+	stream := false
 	return []api.GenerateRequest{
 			{
 				Model:  "orca-mini",
@@ -336,3 +339,83 @@ func GenerateRequests() ([]api.GenerateRequest, [][]string) {
 			[]string{"nitrogen", "oxygen", "carbon", "dioxide"},
 		}
 }
+
+func EmbeddingTestHelper(ctx context.Context, t *testing.T, client *http.Client, req api.EmbeddingRequest) api.EmbeddingResponse {
+
+	// TODO maybe stuff in an init routine?
+	lifecycle.InitLogging()
+
+	requestJSON, err := json.Marshal(req)
+	if err != nil {
+		t.Fatalf("Error serializing request: %v", err)
+	}
+	defer func() {
+		if os.Getenv("OLLAMA_TEST_EXISTING") == "" {
+			defer serverProcMutex.Unlock()
+			if t.Failed() {
+				fp, err := os.Open(lifecycle.ServerLogFile)
+				if err != nil {
+					slog.Error("failed to open server log", "logfile", lifecycle.ServerLogFile, "error", err)
+					return
+				}
+				data, err := io.ReadAll(fp)
+				if err != nil {
+					slog.Error("failed to read server log", "logfile", lifecycle.ServerLogFile, "error", err)
+					return
+				}
+				slog.Warn("SERVER LOG FOLLOWS")
+				os.Stderr.Write(data)
+				slog.Warn("END OF SERVER")
+			}
+			err = os.Remove(lifecycle.ServerLogFile)
+			if err != nil && !os.IsNotExist(err) {
+				slog.Warn("failed to cleanup", "logfile", lifecycle.ServerLogFile, "error", err)
+			}
+		}
+	}()
+	scheme, testEndpoint := GetTestEndpoint()
+
+	if os.Getenv("OLLAMA_TEST_EXISTING") == "" {
+		serverProcMutex.Lock()
+		fp, err := os.CreateTemp("", "ollama-server-*.log")
+		if err != nil {
+			t.Fatalf("failed to generate log file: %s", err)
+		}
+		lifecycle.ServerLogFile = fp.Name()
+		fp.Close()
+		assert.NoError(t, StartServer(ctx, testEndpoint))
+	}
+
+	err = PullIfMissing(ctx, client, scheme, testEndpoint, req.Model)
+	if err != nil {
+		t.Fatalf("Error pulling model: %v", err)
+	}
+
+	// Make the request and get the response
+	httpReq, err := http.NewRequest("POST", scheme+"://"+testEndpoint+"/api/embeddings", bytes.NewReader(requestJSON))
+	if err != nil {
+		t.Fatalf("Error creating request: %v", err)
+	}
+
+	// Set the content type for the request
+	httpReq.Header.Set("Content-Type", "application/json")
+
+	// Make the request with the HTTP client
+	response, err := client.Do(httpReq.WithContext(ctx))
+	if err != nil {
+		t.Fatalf("Error making request: %v", err)
+	}
+	defer response.Body.Close()
+	body, err := io.ReadAll(response.Body)
+	assert.NoError(t, err)
+	assert.Equal(t, response.StatusCode, 200, string(body))
+
+	// Verify the response is valid JSON
+	var res api.EmbeddingResponse
+	err = json.Unmarshal(body, &res)
+	if err != nil {
+		assert.NoError(t, err, body)
+	}
+
+	return res
+}

llm/ext_server/server.cpp

@@ -66,7 +66,7 @@ struct server_params {
 };
 
 bool server_verbose = false;
-bool server_log_json = false;
+bool server_log_json = true;
 
 enum stop_type {
     STOP_FULL,
@@ -266,7 +266,7 @@ struct server_slot {
         sprintf(buffer, "prompt eval time     = %10.2f ms / %5d tokens (%8.2f ms per token, %8.2f tokens per second)",
                 t_prompt_processing, n_prompt_tokens_processed,
                 t_token, n_tokens_second);
-        LOG_DEBUG(buffer, {
+        LOG_INFO(buffer, {
             {"slot_id",                   id},
             {"task_id",                   task_id},
             {"t_prompt_processing",       t_prompt_processing},
@@ -280,7 +280,7 @@ struct server_slot {
         sprintf(buffer, "generation eval time = %10.2f ms / %5d runs   (%8.2f ms per token, %8.2f tokens per second)",
                 t_token_generation, n_decoded,
                 t_token, n_tokens_second);
-        LOG_DEBUG(buffer, {
+        LOG_INFO(buffer, {
             {"slot_id",            id},
             {"task_id",            task_id},
             {"t_token_generation", t_token_generation},
@@ -290,7 +290,7 @@ struct server_slot {
         });
 
         sprintf(buffer, "          total time = %10.2f ms", t_prompt_processing + t_token_generation);
-        LOG_DEBUG(buffer, {
+        LOG_INFO(buffer, {
             {"slot_id",             id},
             {"task_id",             task_id},
             {"t_prompt_processing", t_prompt_processing},
@@ -371,7 +371,7 @@ struct llama_server_context
     {
         if (clp_ctx)
         {
-            LOG_DEBUG("freeing clip model", {});
+            LOG_INFO("freeing clip model", {});
             clip_free(clp_ctx);
             clp_ctx = nullptr;
         }
@@ -392,7 +392,7 @@ struct llama_server_context
         params = params_;
         if (!params.mmproj.empty()) {
             multimodal = true;
-            LOG_DEBUG("Multi Modal Mode Enabled", {});
+            LOG_INFO("Multi Modal Mode Enabled", {});
             clp_ctx = clip_model_load(params.mmproj.c_str(), /*verbosity=*/ 1);
             if(clp_ctx == nullptr) {
                 LOG_ERROR("unable to load clip model", {{"model", params.mmproj}});
@@ -445,7 +445,7 @@ struct llama_server_context
 
         const int32_t n_ctx_slot = n_ctx / params.n_parallel;
 
-        LOG_DEBUG("initializing slots", {{"n_slots", params.n_parallel}});
+        LOG_INFO("initializing slots", {{"n_slots", params.n_parallel}});
         for (int i = 0; i < params.n_parallel; i++)
         {
             server_slot slot;
@@ -454,7 +454,7 @@ struct llama_server_context
             slot.n_ctx = n_ctx_slot;
             slot.n_predict = params.n_predict;
 
-            LOG_DEBUG("new slot", {
+            LOG_INFO("new slot", {
                 {"slot_id",    slot.id},
                 {"n_ctx_slot", slot.n_ctx}
             });
@@ -468,7 +468,7 @@ struct llama_server_context
                 //GGML_ASSERT(n_ctx_train % ga_w == 0     && "n_ctx_train must be a multiple of ga_w");    // NOLINT
                 //GGML_ASSERT(n_ctx >= n_ctx_train * ga_n && "n_ctx must be at least n_ctx_train * ga_n"); // NOLINT
 
-                LOG_DEBUG("slot self-extend", {
+                LOG_INFO("slot self-extend", {
                     {"slot_id",   slot.id},
                     {"ga_n",      ga_n},
                     {"ga_w",      ga_w}
@@ -827,7 +827,7 @@ struct llama_server_context
 
         all_slots_are_idle = false;
 
-        LOG_DEBUG("slot is processing task", {
+        LOG_INFO("slot is processing task", {
             {"slot_id", slot->id},
             {"task_id", slot->task_id},
         });
@@ -1032,7 +1032,7 @@ struct llama_server_context
             slot.has_next_token = false;
         }
 
-        if (!slot.cache_tokens.empty() && llama_token_is_eog(model, result.tok))
+        if (!slot.cache_tokens.empty() && result.tok == llama_token_eos(model))
         {
             slot.stopped_eos = true;
             slot.has_next_token = false;
@@ -1144,15 +1144,12 @@ struct llama_server_context
 
         res.result_json = json
         {
+            {"content",    tkn.text_to_send},
             {"stop",       false},
             {"slot_id",    slot.id},
             {"multimodal", multimodal}
         };
 
-        if (!llama_token_is_eog(model, tkn.tok)) {
-            res.result_json["content"] = tkn.text_to_send;
-        }
-
         if (slot.sparams.n_probs > 0)
         {
             std::vector<completion_token_output> probs_output = {};
@@ -1186,6 +1183,8 @@ struct llama_server_context
             {"model",               params.model_alias},
             {"tokens_predicted",    slot.n_decoded},
             {"tokens_evaluated",    slot.n_prompt_tokens},
+            {"generation_settings", get_formated_generation(slot)},
+            {"prompt",              slot.prompt},
             {"truncated",           slot.truncated},
             {"stopped_eos",         slot.stopped_eos},
             {"stopped_word",        slot.stopped_word},
@@ -1504,7 +1503,7 @@ struct llama_server_context
                     }
                     slots_data.push_back(slot_data);
                 }
-                LOG_DEBUG("slot data", {
+                LOG_INFO("slot data", {
                     {"task_id",            task.id},
                     {"n_idle_slots",       n_idle_slots},
                     {"n_processing_slots", n_processing_slots}
@@ -1566,7 +1565,7 @@ struct llama_server_context
     bool update_slots() {
         if (system_need_update)
         {
-            LOG_DEBUG("updating system prompt", {});
+            LOG_INFO("updating system prompt", {});
             system_prompt_update();
         }
 
@@ -1576,7 +1575,7 @@ struct llama_server_context
         {
             if (system_prompt.empty() && clean_kv_cache)
             {
-                LOG_DEBUG("all slots are idle and system prompt is empty, clear the KV cache", {});
+                LOG_INFO("all slots are idle and system prompt is empty, clear the KV cache", {});
                 kv_cache_clear();
             }
             return true;
@@ -1599,7 +1598,7 @@ struct llama_server_context
                     const int n_left    = (int) system_tokens.size() + slot.n_past - n_keep;
                     const int n_discard = n_left / 2;
 
-                    LOG_DEBUG("slot context shift", {
+                    LOG_INFO("slot context shift", {
                         {"slot_id",         slot.id},
                         {"task_id",         slot.task_id},
                         {"n_keep",          n_keep},
@@ -1638,7 +1637,7 @@ struct llama_server_context
                 slot.command = NONE;
                 slot.t_last_used = ggml_time_us();
 
-                LOG_DEBUG("slot released", {
+                LOG_INFO("slot released", {
                     {"slot_id",         slot.id},
                     {"task_id",         slot.task_id},
                     {"n_ctx",           n_ctx},
@@ -1807,7 +1806,7 @@ struct llama_server_context
                             slot.ga_i = ga_i;
                         }
 
-                        LOG_DEBUG("slot progression", {
+                        LOG_INFO("slot progression", {
                             { "slot_id",    slot.id },
                             { "task_id",    slot.task_id },
                             { "n_past",     slot.n_past },
@@ -1822,7 +1821,7 @@ struct llama_server_context
                     if (slot.n_past == slot.n_prompt_tokens && slot.n_past > 0)
                     {
                         // we have to evaluate at least 1 token to generate logits.
-                        LOG_DEBUG("we have to evaluate at least 1 token to generate logits", {
+                        LOG_INFO("we have to evaluate at least 1 token to generate logits", {
                             { "slot_id", slot.id },
                             { "task_id", slot.task_id }
                         });
@@ -1834,7 +1833,7 @@ struct llama_server_context
                     }
 
                     int p0 = (int) system_tokens.size() + slot.n_past;
-                    LOG_DEBUG("kv cache rm [p0, end)", {
+                    LOG_INFO("kv cache rm [p0, end)", {
                         { "slot_id", slot.id },
                         { "task_id", slot.task_id },
                         { "p0",      p0 }
@@ -2491,7 +2490,11 @@ static void server_params_parse(int argc, char **argv, server_params &sparams,
         }
         else if (arg == "-v" || arg == "--verbose")
         {
+#if SERVER_VERBOSE != 1
+            LOG_WARNING("server.cpp is not built with verbose logging.", {});
+#else
             server_verbose = true;
+#endif
         }
         else if (arg == "--mlock")
         {
@@ -2597,7 +2600,7 @@ static void server_params_parse(int argc, char **argv, server_params &sparams,
         else if (arg == "--log-disable")
         {
             log_set_target(stdout);
-            LOG_DEBUG("logging to file is disabled.", {});
+            LOG_INFO("logging to file is disabled.", {});
         }
         else if (arg == "--slots-endpoint-disable")
         {
@@ -2641,18 +2644,18 @@ static void server_params_parse(int argc, char **argv, server_params &sparams,
             if (strncmp(sep, "int:", 4) == 0) {
                 sep += 4;
                 kvo.tag = LLAMA_KV_OVERRIDE_TYPE_INT;
-                kvo.val_i64 = std::atol(sep);
+                kvo.int_value = std::atol(sep);
             } else if (strncmp(sep, "float:", 6) == 0) {
                 sep += 6;
                 kvo.tag = LLAMA_KV_OVERRIDE_TYPE_FLOAT;
-                kvo.val_f64 = std::atof(sep);
+                kvo.float_value = std::atof(sep);
             } else if (strncmp(sep, "bool:", 5) == 0) {
                 sep += 5;
                 kvo.tag = LLAMA_KV_OVERRIDE_TYPE_BOOL;
                 if (std::strcmp(sep, "true") == 0) {
-                    kvo.val_bool = true;
+                    kvo.bool_value = true;
                 } else if (std::strcmp(sep, "false") == 0) {
-                    kvo.val_bool = false;
+                    kvo.bool_value = false;
                 } else {
                     fprintf(stderr, "error: Invalid boolean value for KV override: %s\n", argv[i]);
                     invalid_param = true;
@@ -2723,12 +2726,12 @@ static json format_detokenized_response(std::string content)
 static void log_server_request(const httplib::Request &req, const httplib::Response &res)
 {
     // skip GH copilot requests when using default port
-    if (req.path == "/health" || req.path == "/v1/health" || req.path == "/v1/completions")
+    if (req.path == "/v1/health" || req.path == "/v1/completions")
     {
         return;
     }
 
-    LOG_DEBUG("request", {
+    LOG_INFO("request", {
         {"remote_addr", req.remote_addr},
         {"remote_port", req.remote_port},
         {"status",      res.status},
@@ -3050,26 +3053,6 @@ int main(int argc, char **argv) {
         log_data["api_key"] = "api_key: " + std::to_string(sparams.api_keys.size()) + " keys loaded";
     }
 
-    if (sparams.n_threads_http < 1) {
-        // +2 threads for monitoring endpoints
-        sparams.n_threads_http = std::max(params.n_parallel + 2, (int32_t) std::thread::hardware_concurrency() - 1);
-    }
-    log_data["n_threads_http"] =  std::to_string(sparams.n_threads_http);
-    svr.new_task_queue = [&sparams] { return new httplib::ThreadPool(sparams.n_threads_http); };
-
-    LOG_INFO("HTTP server listening", log_data);
-    // run the HTTP server in a thread - see comment below
-    std::thread t([&]()
-            {
-                if (!svr.listen_after_bind())
-                {
-                    state.store(SERVER_STATE_ERROR);
-                    return 1;
-                }
-
-                return 0;
-            });
-
     // load the model
     if (!llama.load_model(params))
     {
@@ -3226,53 +3209,46 @@ int main(int argc, char **argv) {
                 return res.set_content(data.dump(), "application/json; charset=utf-8");
             });
 
-    svr.Post("/embedding", [&llama](const httplib::Request &req, httplib::Response &res)
+    svr.Post("/embeddings", [&llama](const httplib::Request &req, httplib::Response &res)
             {
                 res.set_header("Access-Control-Allow-Origin", req.get_header_value("Origin"));
                 const json body = json::parse(req.body);
-                json prompt;
-                if (body.count("content") != 0)
-                {
-                    prompt = body["content"];
-                }
-                else
-                {
-                    prompt = "";
+
+                const int id = llama.queue_tasks.get_new_id();
+                llama.queue_results.add_waiting_task_id(id);
+                llama.request_completion(id, {{"prompt", body["contents"]}}, false, true, -1);
+
+                task_result recv = llama.queue_results.recv(id);
+                llama.queue_results.remove_waiting_task_id(id);
+
+                json embeddings = json::array();
+                for (auto & elem : recv.result_json["results"]) {
+                    embeddings.push_back(json_value(elem, "embedding", json::array()));
                 }
 
-                json image_data;
-                if (body.count("image_data") != 0) {
-                    image_data = body["image_data"];
-                }
-                else
-                {
-                    image_data = "";
-                }
-
-                // create and queue the task
-                const int task_id = llama.queue_tasks.get_new_id();
-                llama.queue_results.add_waiting_task_id(task_id);
-                llama.request_completion(task_id, { {"prompt", prompt}, { "n_predict", 0}, {"image_data", image_data} }, false, true, -1);
-
-                // get the result
-                task_result result = llama.queue_results.recv(task_id);
-                llama.queue_results.remove_waiting_task_id(task_id);
-
-                // send the result
-                return res.set_content(result.result_json.dump(), "application/json; charset=utf-8");
+                json result = json{{"embeddings", embeddings}};
+                return res.set_content(result.dump(), "application/json; charset=utf-8");
             });
 
-    // GG: if I put the main loop inside a thread, it crashes on the first request when build in Debug!?
-    //     "Bus error: 10" - this is on macOS, it does not crash on Linux
-    //std::thread t2([&]()
-    /*{
-        bool running = true;
-        while (running)
-        {
-            running = llama.update_slots();
-        }
-    }*/
-    //);
+    if (sparams.n_threads_http < 1) {
+        // +2 threads for monitoring endpoints
+        sparams.n_threads_http = std::max(params.n_parallel + 2, (int32_t) std::thread::hardware_concurrency() - 1);
+    }
+    log_data["n_threads_http"] =  std::to_string(sparams.n_threads_http);
+    svr.new_task_queue = [&sparams] { return new httplib::ThreadPool(sparams.n_threads_http); };
+
+    LOG_INFO("HTTP server listening", log_data);
+    // run the HTTP server in a thread - see comment below
+    std::thread t([&]()
+            {
+                if (!svr.listen_after_bind())
+                {
+                    state.store(SERVER_STATE_ERROR);
+                    return 1;
+                }
+
+                return 0;
+            });
 
     llama.queue_tasks.on_new_task(std::bind(
         &llama_server_context::process_single_task, &llama, std::placeholders::_1));

llm/ext_server/utils.hpp

@@ -55,10 +55,9 @@ extern bool server_log_json;
     } while (0)
 #endif
 
-#define LOG_ERROR(  MSG, ...) server_log("ERROR",  __func__, __LINE__, MSG, __VA_ARGS__)
+#define LOG_ERROR(  MSG, ...) server_log("ERR",  __func__, __LINE__, MSG, __VA_ARGS__)
 #define LOG_WARNING(MSG, ...) server_log("WARN", __func__, __LINE__, MSG, __VA_ARGS__)
 #define LOG_INFO(   MSG, ...) server_log("INFO", __func__, __LINE__, MSG, __VA_ARGS__)
-#define LOG_DEBUG(  MSG, ...) server_log("DEBUG", __func__, __LINE__, MSG, __VA_ARGS__)
 
 enum server_state {
     SERVER_STATE_LOADING_MODEL,  // Server is starting up, model not fully loaded yet
@@ -124,10 +123,6 @@ static inline void server_log(const char *level, const char *function, int line,
         {"timestamp", time(nullptr)},
     };
 
-    if (strncmp("DEBUG", level, strlen(level)) == 0 && !server_verbose) {
-        return;
-    }
-
     if (server_log_json) {
         log.merge_patch(
                 {
@@ -142,12 +137,14 @@ static inline void server_log(const char *level, const char *function, int line,
 
         std::cout << log.dump(-1, ' ', false, json::error_handler_t::replace) << "\n" << std::flush;
     } else {
+        char buf[1024];
+        snprintf(buf, 1024, "%4s [%24s] %s", level, function, message);
+
         if (!extra.empty()) {
             log.merge_patch(extra);
         }
-
         std::stringstream ss;
-        ss << level << " [" << function << "] " << message << " |";
+        ss << buf << " |";
         for (const auto& el : log.items())
         {
             const std::string value = el.value().dump(-1, ' ', false, json::error_handler_t::replace);

llm/filetype.go

@@ -1,140 +0,0 @@
-package llm
-
-import "fmt"
-
-type fileType uint32
-
-const (
-	fileTypeF32 fileType = iota
-	fileTypeF16
-	fileTypeQ4_0
-	fileTypeQ4_1
-	fileTypeQ4_1_F16
-	fileTypeQ4_2 // unused
-	fileTypeQ4_3 // unused
-	fileTypeQ8_0
-	fileTypeQ5_0
-	fileTypeQ5_1
-	fileTypeQ2_K
-	fileTypeQ3_K_S
-	fileTypeQ3_K_M
-	fileTypeQ3_K_L
-	fileTypeQ4_K_S
-	fileTypeQ4_K_M
-	fileTypeQ5_K_S
-	fileTypeQ5_K_M
-	fileTypeQ6_K
-	fileTypeIQ2_XXS
-	fileTypeIQ2_XS
-	fileTypeQ2_K_S
-	fileTypeQ3_K_XS
-	fileTypeIQ3_XXS
-
-	fileTypeUnknown
-)
-
-func ParseFileType(s string) (fileType, error) {
-	switch s {
-	case "F32":
-		return fileTypeF32, nil
-	case "F16":
-		return fileTypeF16, nil
-	case "Q4_0":
-		return fileTypeQ4_0, nil
-	case "Q4_1":
-		return fileTypeQ4_1, nil
-	case "Q4_1_F16":
-		return fileTypeQ4_1_F16, nil
-	case "Q8_0":
-		return fileTypeQ8_0, nil
-	case "Q5_0":
-		return fileTypeQ5_0, nil
-	case "Q5_1":
-		return fileTypeQ5_1, nil
-	case "Q2_K":
-		return fileTypeQ2_K, nil
-	case "Q3_K_S":
-		return fileTypeQ3_K_S, nil
-	case "Q3_K_M":
-		return fileTypeQ3_K_M, nil
-	case "Q3_K_L":
-		return fileTypeQ3_K_L, nil
-	case "Q4_K_S":
-		return fileTypeQ4_K_S, nil
-	case "Q4_K_M":
-		return fileTypeQ4_K_M, nil
-	case "Q5_K_S":
-		return fileTypeQ5_K_S, nil
-	case "Q5_K_M":
-		return fileTypeQ5_K_M, nil
-	case "Q6_K":
-		return fileTypeQ6_K, nil
-	case "IQ2_XXS":
-		return fileTypeIQ2_XXS, nil
-	case "IQ2_XS":
-		return fileTypeIQ2_XS, nil
-	case "Q2_K_S":
-		return fileTypeQ2_K_S, nil
-	case "Q3_K_XS":
-		return fileTypeQ3_K_XS, nil
-	case "IQ3_XXS":
-		return fileTypeIQ3_XXS, nil
-	default:
-		return fileTypeUnknown, fmt.Errorf("unknown fileType: %s", s)
-	}
-}
-
-func (t fileType) String() string {
-	switch t {
-	case fileTypeF32:
-		return "F32"
-	case fileTypeF16:
-		return "F16"
-	case fileTypeQ4_0:
-		return "Q4_0"
-	case fileTypeQ4_1:
-		return "Q4_1"
-	case fileTypeQ4_1_F16:
-		return "Q4_1_F16"
-	case fileTypeQ8_0:
-		return "Q8_0"
-	case fileTypeQ5_0:
-		return "Q5_0"
-	case fileTypeQ5_1:
-		return "Q5_1"
-	case fileTypeQ2_K:
-		return "Q2_K"
-	case fileTypeQ3_K_S:
-		return "Q3_K_S"
-	case fileTypeQ3_K_M:
-		return "Q3_K_M"
-	case fileTypeQ3_K_L:
-		return "Q3_K_L"
-	case fileTypeQ4_K_S:
-		return "Q4_K_S"
-	case fileTypeQ4_K_M:
-		return "Q4_K_M"
-	case fileTypeQ5_K_S:
-		return "Q5_K_S"
-	case fileTypeQ5_K_M:
-		return "Q5_K_M"
-	case fileTypeQ6_K:
-		return "Q6_K"
-	case fileTypeIQ2_XXS:
-		return "IQ2_XXS"
-	case fileTypeIQ2_XS:
-		return "IQ2_XS"
-	case fileTypeQ2_K_S:
-		return "Q2_K_S"
-	case fileTypeQ3_K_XS:
-		return "Q3_K_XS"
-	case fileTypeIQ3_XXS:
-		return "IQ3_XXS"
-	default:
-		return "unknown"
-	}
-}
-
-func (t fileType) Value() uint32 {
-	return uint32(t)
-}

llm/generate/gen_windows.ps1

@@ -42,9 +42,8 @@ function init_vars {
         "-DLLAMA_NATIVE=off"
         )
     $script:commonCpuDefs = @("-DCMAKE_POSITION_INDEPENDENT_CODE=on")
-    $script:ARCH = $Env:PROCESSOR_ARCHITECTURE.ToLower()
+    $script:ARCH = "amd64" # arm not yet supported.
     $script:DIST_BASE = "${script:SRC_DIR}\dist\windows-${script:ARCH}\ollama_runners"
-    md "$script:DIST_BASE" -ea 0 > $null
     if ($env:CGO_CFLAGS -contains "-g") {
         $script:cmakeDefs += @("-DCMAKE_VERBOSE_MAKEFILE=on", "-DLLAMA_SERVER_VERBOSE=on", "-DCMAKE_BUILD_TYPE=RelWithDebInfo")
         $script:config = "RelWithDebInfo"
@@ -182,7 +181,7 @@ function cleanup {
 
 
 function build_static() {
-    if ((-not "${env:OLLAMA_SKIP_STATIC_GENERATE}") -and ((-not "${env:OLLAMA_CPU_TARGET}") -or ("${env:OLLAMA_CPU_TARGET}" -eq "static"))) {
+    if ($null -eq ${env:OLLAMA_SKIP_CPU_GENERATE}) {
         # GCC build for direct linking into the Go binary
         init_vars
         # cmake will silently fallback to msvc compilers if mingw isn't in the path, so detect and fail fast
@@ -213,11 +212,11 @@ function build_static() {
     }
 }
 
-function build_cpu($gen_arch) {
-    if ((-not "${env:OLLAMA_SKIP_CPU_GENERATE}" ) -and ((-not "${env:OLLAMA_CPU_TARGET}") -or ("${env:OLLAMA_CPU_TARGET}" -eq "cpu"))) {
+function build_cpu() {
+    if ($null -eq ${env:OLLAMA_SKIP_CPU_GENERATE}) {
         # remaining llama.cpp builds use MSVC 
         init_vars
-        $script:cmakeDefs = $script:commonCpuDefs + @("-A", $gen_arch, "-DLLAMA_AVX=off", "-DLLAMA_AVX2=off", "-DLLAMA_AVX512=off", "-DLLAMA_FMA=off", "-DLLAMA_F16C=off") + $script:cmakeDefs
+        $script:cmakeDefs = $script:commonCpuDefs + @("-A", "x64", "-DLLAMA_AVX=off", "-DLLAMA_AVX2=off", "-DLLAMA_AVX512=off", "-DLLAMA_FMA=off", "-DLLAMA_F16C=off") + $script:cmakeDefs
         $script:buildDir="../build/windows/${script:ARCH}/cpu"
         $script:distDir="$script:DIST_BASE\cpu"
         write-host "Building LCD CPU"
@@ -230,7 +229,7 @@ function build_cpu($gen_arch) {
 }
 
 function build_cpu_avx() {
-    if ((-not "${env:OLLAMA_SKIP_CPU_GENERATE}" ) -and ((-not "${env:OLLAMA_CPU_TARGET}") -or ("${env:OLLAMA_CPU_TARGET}" -eq "cpu_avx"))) {
+    if ($null -eq ${env:OLLAMA_SKIP_CPU_GENERATE}) {
         init_vars
         $script:cmakeDefs = $script:commonCpuDefs + @("-A", "x64", "-DLLAMA_AVX=on", "-DLLAMA_AVX2=off", "-DLLAMA_AVX512=off", "-DLLAMA_FMA=off", "-DLLAMA_F16C=off") + $script:cmakeDefs
         $script:buildDir="../build/windows/${script:ARCH}/cpu_avx"
@@ -240,12 +239,12 @@ function build_cpu_avx() {
         sign
         install
     } else {
-        write-host "Skipping CPU AVX generation step as requested"
+        write-host "Skipping CPU generation step as requested"
     }
 }
 
 function build_cpu_avx2() {
-    if ((-not "${env:OLLAMA_SKIP_CPU_GENERATE}" ) -and ((-not "${env:OLLAMA_CPU_TARGET}") -or ("${env:OLLAMA_CPU_TARGET}" -eq "cpu_avx2"))) {
+    if ($null -eq ${env:OLLAMA_SKIP_CPU_GENERATE}) {
         init_vars
         $script:cmakeDefs = $script:commonCpuDefs + @("-A", "x64", "-DLLAMA_AVX=on", "-DLLAMA_AVX2=on", "-DLLAMA_AVX512=off", "-DLLAMA_FMA=on", "-DLLAMA_F16C=on") + $script:cmakeDefs
         $script:buildDir="../build/windows/${script:ARCH}/cpu_avx2"
@@ -255,12 +254,12 @@ function build_cpu_avx2() {
         sign
         install
     } else {
-        write-host "Skipping CPU AVX2 generation step as requested"
+        write-host "Skipping CPU generation step as requested"
     }
 }
 
 function build_cuda() {
-    if ((-not "${env:OLLAMA_SKIP_CUDA_GENERATE}") -and ("${script:CUDA_LIB_DIR}")) {
+    if ($null -ne $script:CUDA_LIB_DIR) {
         # Then build cuda as a dynamically loaded library
         $nvcc = "$script:CUDA_LIB_DIR\nvcc.exe"
         $script:CUDA_VERSION=(get-item ($nvcc | split-path | split-path)).Basename
@@ -284,13 +283,11 @@ function build_cuda() {
         cp "${script:CUDA_LIB_DIR}\cudart64_*.dll" "${script:SRC_DIR}\dist\windows-${script:ARCH}\"
         cp "${script:CUDA_LIB_DIR}\cublas64_*.dll" "${script:SRC_DIR}\dist\windows-${script:ARCH}\"
         cp "${script:CUDA_LIB_DIR}\cublasLt64_*.dll" "${script:SRC_DIR}\dist\windows-${script:ARCH}\"
-    } else {
-        write-host "Skipping CUDA generation step"
     }
 }
 
 function build_rocm() {
-    if ((-not "${env:OLLAMA_SKIP_ROCM_GENERATE}") -and ("${env:HIP_PATH}")) {
+    if ($null -ne $env:HIP_PATH) {
         $script:ROCM_VERSION=(get-item $env:HIP_PATH).Basename
         if ($null -ne $script:ROCM_VERSION) {
             $script:ROCM_VARIANT="_v"+$script:ROCM_VERSION
@@ -339,8 +336,6 @@ function build_rocm() {
         cp "${env:HIP_PATH}\bin\rocblas.dll" "${script:SRC_DIR}\dist\windows-${script:ARCH}\rocm\"
         # amdhip64.dll dependency comes from the driver and must be installed on the host to use AMD GPUs
         cp "${env:HIP_PATH}\bin\rocblas\library\*" "${script:SRC_DIR}\dist\windows-${script:ARCH}\rocm\rocblas\library\"
-    } else {
-        write-host "Skipping ROCm generation step"
     }
 }
 
@@ -349,15 +344,11 @@ if ($($args.count) -eq 0) {
     git_module_setup
     apply_patches
     build_static
-    if ($script:ARCH -eq "arm64") {
-        build_cpu("ARM64")
-    } else { # amd64
-        build_cpu("x64")
-        build_cpu_avx
-        build_cpu_avx2
-        build_cuda
-        build_rocm
-    }
+    build_cpu
+    build_cpu_avx
+    build_cpu_avx2
+    build_cuda
+    build_rocm
 
     cleanup
     write-host "`ngo generate completed.  LLM runners: $(get-childitem -path $script:DIST_BASE)"

llm/ggml.go

@@ -13,6 +13,82 @@ type GGML struct {
 	model
 }
 
+const (
+	fileTypeF32 uint32 = iota
+	fileTypeF16
+	fileTypeQ4_0
+	fileTypeQ4_1
+	fileTypeQ4_1_F16
+	fileTypeQ8_0 uint32 = iota + 2
+	fileTypeQ5_0
+	fileTypeQ5_1
+	fileTypeQ2_K
+	fileTypeQ3_K_S
+	fileTypeQ3_K_M
+	fileTypeQ3_K_L
+	fileTypeQ4_K_S
+	fileTypeQ4_K_M
+	fileTypeQ5_K_S
+	fileTypeQ5_K_M
+	fileTypeQ6_K
+	fileTypeIQ2_XXS
+	fileTypeIQ2_XS
+	fileTypeQ2_K_S
+	fileTypeQ3_K_XS
+	fileTypeIQ3_XXS
+)
+
+func fileType(fileType uint32) string {
+	switch fileType {
+	case fileTypeF32:
+		return "F32"
+	case fileTypeF16:
+		return "F16"
+	case fileTypeQ4_0:
+		return "Q4_0"
+	case fileTypeQ4_1:
+		return "Q4_1"
+	case fileTypeQ4_1_F16:
+		return "Q4_1_F16"
+	case fileTypeQ8_0:
+		return "Q8_0"
+	case fileTypeQ5_0:
+		return "Q5_0"
+	case fileTypeQ5_1:
+		return "Q5_1"
+	case fileTypeQ2_K:
+		return "Q2_K"
+	case fileTypeQ3_K_S:
+		return "Q3_K_S"
+	case fileTypeQ3_K_M:
+		return "Q3_K_M"
+	case fileTypeQ3_K_L:
+		return "Q3_K_L"
+	case fileTypeQ4_K_S:
+		return "Q4_K_S"
+	case fileTypeQ4_K_M:
+		return "Q4_K_M"
+	case fileTypeQ5_K_S:
+		return "Q5_K_S"
+	case fileTypeQ5_K_M:
+		return "Q5_K_M"
+	case fileTypeQ6_K:
+		return "Q6_K"
+	case fileTypeIQ2_XXS:
+		return "IQ2_XXS"
+	case fileTypeIQ2_XS:
+		return "IQ2_XS"
+	case fileTypeQ2_K_S:
+		return "Q2_K_S"
+	case fileTypeQ3_K_XS:
+		return "Q3_K_XS"
+	case fileTypeIQ3_XXS:
+		return "IQ3_XXS"
+	default:
+		return "unknown"
+	}
+}
+
 type model interface {
 	KV() KV
 	Tensors() Tensors
@@ -45,12 +121,12 @@ func (kv KV) ParameterCount() uint64 {
 	return kv.u64("general.parameter_count")
 }
 
-func (kv KV) FileType() fileType {
+func (kv KV) FileType() string {
 	if u64 := kv.u64("general.file_type"); u64 > 0 {
 		return fileType(uint32(u64))
 	}
 
-	return fileTypeUnknown
+	return "unknown"
 }
 
 func (kv KV) BlockCount() uint64 {
@@ -210,23 +286,6 @@ const (
 
 var ErrUnsupportedFormat = errors.New("unsupported model format")
 
-func DetectGGMLType(b []byte) string {
-	switch binary.LittleEndian.Uint32(b[:4]) {
-	case FILE_MAGIC_GGML:
-		return "ggml"
-	case FILE_MAGIC_GGMF:
-		return "ggmf"
-	case FILE_MAGIC_GGJT:
-		return "ggjt"
-	case FILE_MAGIC_GGLA:
-		return "ggla"
-	case FILE_MAGIC_GGUF_LE, FILE_MAGIC_GGUF_BE:
-		return "gguf"
-	default:
-		return ""
-	}
-}
-
 func DecodeGGML(rs io.ReadSeeker) (*GGML, int64, error) {
 	var magic uint32
 	if err := binary.Read(rs, binary.LittleEndian, &magic); err != nil {

llm/llm.go

@@ -4,7 +4,6 @@ package llm
 // #cgo darwin,arm64 LDFLAGS: ${SRCDIR}/build/darwin/arm64_static/libllama.a -lstdc++
 // #cgo darwin,amd64 LDFLAGS: ${SRCDIR}/build/darwin/x86_64_static/libllama.a -lstdc++
 // #cgo windows,amd64 LDFLAGS: ${SRCDIR}/build/windows/amd64_static/libllama.a -static -lstdc++
-// #cgo windows,arm64 LDFLAGS: ${SRCDIR}/build/windows/arm64_static/libllama.a -static -lstdc++
 // #cgo linux,amd64 LDFLAGS: ${SRCDIR}/build/linux/x86_64_static/libllama.a -lstdc++
 // #cgo linux,arm64 LDFLAGS: ${SRCDIR}/build/linux/arm64_static/libllama.a -lstdc++
 // #include <stdlib.h>
@@ -20,7 +19,7 @@ func SystemInfo() string {
 	return C.GoString(C.llama_print_system_info())
 }
 
-func Quantize(infile, outfile string, ftype fileType) error {
+func Quantize(infile, outfile, filetype string) error {
 	cinfile := C.CString(infile)
 	defer C.free(unsafe.Pointer(cinfile))
 
@@ -29,10 +28,58 @@ func Quantize(infile, outfile string, ftype fileType) error {
 
 	params := C.llama_model_quantize_default_params()
 	params.nthread = -1
-	params.ftype = ftype.Value()
 
-	if rc := C.llama_model_quantize(cinfile, coutfile, &params); rc != 0 {
-		return fmt.Errorf("llama_model_quantize: %d", rc)
+	switch filetype {
+	case "F32":
+		params.ftype = fileTypeF32
+	case "F16":
+		params.ftype = fileTypeF16
+	case "Q4_0":
+		params.ftype = fileTypeQ4_0
+	case "Q4_1":
+		params.ftype = fileTypeQ4_1
+	case "Q4_1_F16":
+		params.ftype = fileTypeQ4_1_F16
+	case "Q8_0":
+		params.ftype = fileTypeQ8_0
+	case "Q5_0":
+		params.ftype = fileTypeQ5_0
+	case "Q5_1":
+		params.ftype = fileTypeQ5_1
+	case "Q2_K":
+		params.ftype = fileTypeQ2_K
+	case "Q3_K_S":
+		params.ftype = fileTypeQ3_K_S
+	case "Q3_K_M":
+		params.ftype = fileTypeQ3_K_M
+	case "Q3_K_L":
+		params.ftype = fileTypeQ3_K_L
+	case "Q4_K_S":
+		params.ftype = fileTypeQ4_K_S
+	case "Q4_K_M":
+		params.ftype = fileTypeQ4_K_M
+	case "Q5_K_S":
+		params.ftype = fileTypeQ5_K_S
+	case "Q5_K_M":
+		params.ftype = fileTypeQ5_K_M
+	case "Q6_K":
+		params.ftype = fileTypeQ6_K
+	case "IQ2_XXS":
+		params.ftype = fileTypeIQ2_XXS
+	case "IQ2_XS":
+		params.ftype = fileTypeIQ2_XS
+	case "Q2_K_S":
+		params.ftype = fileTypeQ2_K_S
+	case "Q3_K_XS":
+		params.ftype = fileTypeQ3_K_XS
+	case "IQ3_XXS":
+		params.ftype = fileTypeIQ3_XXS
+	default:
+		return fmt.Errorf("unknown filetype: %s", filetype)
+	}
+
+	if retval := C.llama_model_quantize(cinfile, coutfile, &params); retval != 0 {
+		return fmt.Errorf("llama_model_quantize: %d", retval)
 	}
 
 	return nil

llm/memory.go

@@ -3,11 +3,12 @@ package llm
 import (
 	"fmt"
 	"log/slog"
+	"os"
+	"strconv"
 
 	"github.com/ollama/ollama/api"
 	"github.com/ollama/ollama/format"
 	"github.com/ollama/ollama/gpu"
-	"github.com/ollama/ollama/server/envconfig"
 )
 
 // This algorithm looks for a complete fit to determine if we need to unload other models
@@ -25,7 +26,7 @@ func PredictServerFit(allGpus gpu.GpuInfoList, ggml *GGML, adapters, projectors
 	// Split up the GPUs by type and try them
 	for _, gpus := range allGpus.ByLibrary() {
 		var layerCount int
-		layerCount, estimatedVRAM, _ = EstimateGPULayers(gpus, ggml, projectors, opts)
+		layerCount, estimatedVRAM = EstimateGPULayers(gpus, ggml, projectors, opts)
 		if opts.NumGPU < 0 {
 			if layerCount > 0 && layerCount >= int(ggml.KV().BlockCount()+1) {
 				return true, estimatedVRAM
@@ -39,15 +40,25 @@ func PredictServerFit(allGpus gpu.GpuInfoList, ggml *GGML, adapters, projectors
 	return false, estimatedVRAM
 }
 
-// Given a model and one or more GPU targets, predict how many layers and bytes we can load, and the total size
+// Given a model and one or more GPU targets, predict how many layers and bytes we can load
 // The GPUs provided must all be the same Library
-func EstimateGPULayers(gpus []gpu.GpuInfo, ggml *GGML, projectors []string, opts api.Options) (int, uint64, uint64) {
+func EstimateGPULayers(gpus []gpu.GpuInfo, ggml *GGML, projectors []string, opts api.Options) (int, uint64) {
+	if gpus[0].Library == "cpu" {
+		return 0, 0
+	}
 	var memoryAvailable uint64
 	for _, info := range gpus {
 		memoryAvailable += info.FreeMemory
 	}
-	if envconfig.MaxVRAM > 0 {
-		memoryAvailable = envconfig.MaxVRAM
+	userLimit := os.Getenv("OLLAMA_MAX_VRAM")
+	if userLimit != "" {
+		avail, err := strconv.ParseUint(userLimit, 10, 64)
+		if err != nil {
+			slog.Error("invalid setting, ignoring", "OLLAMA_MAX_VRAM", userLimit, "error", err)
+		} else {
+			slog.Info("user override memory limit", "OLLAMA_MAX_VRAM", avail, "actual", memoryAvailable)
+			memoryAvailable = avail
+		}
 	}
 
 	slog.Debug("evaluating", "library", gpus[0].Library, "gpu_count", len(gpus), "available", format.HumanBytes2(memoryAvailable))
@@ -77,19 +88,19 @@ func EstimateGPULayers(gpus []gpu.GpuInfo, ggml *GGML, projectors []string, opts
 	graphFullOffload *= uint64(len(gpus))
 	graphPartialOffload *= uint64(len(gpus))
 
-	// on metal there's no partial offload overhead
-	if gpus[0].Library == "metal" {
-		graphPartialOffload = graphFullOffload
+	// memoryRequiredTotal represents the memory required for full GPU offloading (all layers)
+	memoryRequiredTotal := memoryMinimum + graphFullOffload
+
+	// memoryRequiredPartial represents the memory required for partial GPU offloading (n > 0, n < layers)
+	memoryRequiredPartial := memoryMinimum + graphPartialOffload
+
+	if memoryRequiredPartial > memoryAvailable {
+		slog.Debug("insufficient VRAM to load any model layers")
+		return 0, 0
 	}
 
 	layers := ggml.Tensors().Layers()
 
-	// memoryRequiredTotal represents the memory required for full GPU offloading (all layers)
-	memoryRequiredTotal := memoryMinimum + graphFullOffload + layers["blk.0"].size()
-
-	// memoryRequiredPartial represents the memory required for partial GPU offloading (n > 0, n < layers)
-	memoryRequiredPartial := memoryMinimum + graphPartialOffload + layers["blk.0"].size()
-
 	var memoryLayerOutput uint64
 	if layer, ok := layers["output_norm"]; ok {
 		memoryLayerOutput += layer.size()
@@ -173,13 +184,5 @@ func EstimateGPULayers(gpus []gpu.GpuInfo, ggml *GGML, projectors []string, opts
 			),
 		),
 	)
-	if gpus[0].Library == "cpu" {
-		return 0, 0, memoryRequiredTotal
-	}
-	if memoryRequiredPartial > memoryAvailable {
-		slog.Debug("insufficient VRAM to load any model layers")
-		return 0, 0, memoryRequiredTotal
-	}
-
-	return layerCount, memoryRequiredPartial, memoryRequiredTotal
+	return layerCount, uint64(memoryRequiredPartial)
 }

llm/patches/05-clip-fix.diff

@@ -1,24 +0,0 @@
-diff --git a/examples/llava/clip.cpp b/examples/llava/clip.cpp
-index e3c9bcd4..b43f892d 100644
---- a/examples/llava/clip.cpp
-+++ b/examples/llava/clip.cpp
-@@ -573,14 +573,16 @@ static ggml_cgraph * clip_image_build_graph(clip_ctx * ctx, const clip_image_f32
-     struct ggml_tensor * embeddings = inp;
-     if (ctx->has_class_embedding) {
-         embeddings = ggml_new_tensor_3d(ctx0, GGML_TYPE_F32, hidden_size, num_positions, batch_size);
-+    }
-+    ggml_set_name(embeddings, "embeddings");
-+    ggml_set_input(embeddings);
-+
-+    if (ctx->has_class_embedding) {
-         embeddings = ggml_acc(ctx0, embeddings, model.class_embedding,
-                 embeddings->nb[1], embeddings->nb[2], embeddings->nb[3], 0);
-         embeddings = ggml_acc(ctx0, embeddings, inp,
-                 embeddings->nb[1], embeddings->nb[2], embeddings->nb[3], model.class_embedding->nb[1]);
-     }
--    ggml_set_name(embeddings, "embeddings");
--    ggml_set_input(embeddings);
--
- 
-     struct ggml_tensor * positions = ggml_new_tensor_1d(ctx0, GGML_TYPE_I32, num_positions);
-     ggml_set_name(positions, "positions");

llm/server.go

@@ -26,14 +26,13 @@ import (
 	"github.com/ollama/ollama/api"
 	"github.com/ollama/ollama/format"
 	"github.com/ollama/ollama/gpu"
-	"github.com/ollama/ollama/server/envconfig"
 )
 
 type LlamaServer interface {
 	Ping(ctx context.Context) error
 	WaitUntilRunning(ctx context.Context) error
 	Completion(ctx context.Context, req CompletionRequest, fn func(CompletionResponse)) error
-	Embedding(ctx context.Context, prompt string) ([]float64, error)
+	Embeddings(ctx context.Context, prompt []string) ([][]float64, error)
 	Tokenize(ctx context.Context, content string) ([]int, error)
 	Detokenize(ctx context.Context, tokens []int) (string, error)
 	Close() error
@@ -49,11 +48,7 @@ type llmServer struct {
 	options api.Options
 
 	// TODO - this should be broken down by GPU
-	estimatedVRAM  uint64 // Estimated usage of VRAM by the loaded model
-	estimatedTotal uint64 // Total size of model
-	totalLayers    uint64
-	gpuCount       int
-	loadDuration   time.Duration // Record how long it took the model to load
+	estimatedVRAM uint64 // Estimated usage of VRAM by the loaded model
 
 	sem *semaphore.Weighted
 }
@@ -78,7 +73,8 @@ func LoadModel(model string) (*GGML, error) {
 func NewLlamaServer(gpus gpu.GpuInfoList, model string, ggml *GGML, adapters, projectors []string, opts api.Options) (LlamaServer, error) {
 	var err error
 	if opts.NumCtx > int(ggml.KV().ContextLength()) {
-		slog.Warn("requested context length is greater than the model's training context window size", "requested", opts.NumCtx, "training size", ggml.KV().ContextLength())
+		slog.Warn("requested context length is greater than model max context length", "requested", opts.NumCtx, "model", ggml.KV().ContextLength())
+		opts.NumCtx = int(ggml.KV().ContextLength())
 	}
 
 	if opts.NumCtx < 4 {
@@ -87,15 +83,12 @@ func NewLlamaServer(gpus gpu.GpuInfoList, model string, ggml *GGML, adapters, pr
 
 	cpuRunner := ""
 	var estimatedVRAM uint64
-	var estimatedTotal uint64
 	var systemMemory uint64
-	gpuCount := len(gpus)
 	if (len(gpus) == 1 && gpus[0].Library == "cpu") || opts.NumGPU == 0 {
 
 		// TODO evaluate system memory to see if we should block the load, or force an unload of another CPU runner
 
 		cpuRunner = serverForCpu()
-		gpuCount = 0
 	} else {
 		if gpus[0].Library == "metal" {
 			memInfo, err := gpu.GetCPUMem()
@@ -107,7 +100,7 @@ func NewLlamaServer(gpus gpu.GpuInfoList, model string, ggml *GGML, adapters, pr
 			}
 		}
 		var layers int
-		layers, estimatedVRAM, estimatedTotal = EstimateGPULayers(gpus, ggml, projectors, opts)
+		layers, estimatedVRAM = EstimateGPULayers(gpus, ggml, projectors, opts)
 
 		if gpus[0].Library == "metal" && estimatedVRAM > systemMemory {
 			// disable partial offloading when model is greater than total system memory as this
@@ -132,7 +125,7 @@ func NewLlamaServer(gpus gpu.GpuInfoList, model string, ggml *GGML, adapters, pr
 	} else {
 		servers = serversForGpu(gpus[0]) // All GPUs in the list are matching Library and Variant
 	}
-	demandLib := envconfig.LLMLibrary
+	demandLib := strings.Trim(os.Getenv("OLLAMA_LLM_LIBRARY"), "\"' ")
 	if demandLib != "" {
 		serverPath := availableServers[demandLib]
 		if serverPath == "" {
@@ -140,10 +133,6 @@ func NewLlamaServer(gpus gpu.GpuInfoList, model string, ggml *GGML, adapters, pr
 		} else {
 			slog.Info("user override", "OLLAMA_LLM_LIBRARY", demandLib, "path", serverPath)
 			servers = []string{demandLib}
-			if strings.HasPrefix(demandLib, "cpu") {
-				// Omit the GPU flag to silence the warning
-				opts.NumGPU = -1
-			}
 		}
 	}
 
@@ -157,14 +146,17 @@ func NewLlamaServer(gpus gpu.GpuInfoList, model string, ggml *GGML, adapters, pr
 		"--batch-size", fmt.Sprintf("%d", opts.NumBatch),
 		"--embedding",
 	}
-
-	params = append(params, "--log-disable")
+	if debug := os.Getenv("OLLAMA_DEBUG"); debug != "" {
+		params = append(params, "--log-format", "json")
+	} else {
+		params = append(params, "--log-disable")
+	}
 
 	if opts.NumGPU >= 0 {
 		params = append(params, "--n-gpu-layers", fmt.Sprintf("%d", opts.NumGPU))
 	}
 
-	if envconfig.Debug {
+	if debug := os.Getenv("OLLAMA_DEBUG"); debug != "" {
 		params = append(params, "--verbose")
 	}
 
@@ -202,15 +194,16 @@ func NewLlamaServer(gpus gpu.GpuInfoList, model string, ggml *GGML, adapters, pr
 		params = append(params, "--numa")
 	}
 
-	numParallel := envconfig.NumParallel
-
-	// TODO (jmorganca): multimodal models don't support parallel yet
-	// see https://github.com/ollama/ollama/issues/4165
-	if len(projectors) > 0 {
-		numParallel = 1
-		slog.Warn("multimodal models don't support parallel requests yet")
+	// "--cont-batching", // TODO - doesn't seem to have any noticeable perf change for multiple requests
+	numParallel := 1
+	if onp := os.Getenv("OLLAMA_NUM_PARALLEL"); onp != "" {
+		numParallel, err = strconv.Atoi(onp)
+		if err != nil || numParallel <= 0 {
+			err = fmt.Errorf("invalid OLLAMA_NUM_PARALLEL=%s must be greater than zero - %w", onp, err)
+			slog.Error("misconfiguration", "error", err)
+			return nil, err
+		}
 	}
-
 	params = append(params, "--parallel", fmt.Sprintf("%d", numParallel))
 
 	for i := 0; i < len(servers); i++ {
@@ -218,15 +211,10 @@ func NewLlamaServer(gpus gpu.GpuInfoList, model string, ggml *GGML, adapters, pr
 		if dir == "" {
 			// Shouldn't happen
 			finalErr = fmt.Errorf("[%d] server %s not listed in available servers %v", i, servers[i], availableServers)
-			slog.Error("server list inconsistent", "error", finalErr)
+			slog.Error("sever list inconsistent", "error", finalErr)
 			continue
 		}
 
-		if strings.HasPrefix(servers[i], "cpu") {
-			// TODO if we tried a gpu runner first, and it failed, record the error and bubble that back up
-			gpuCount = 0
-		}
-
 		// Find an availableServers  port, retry on each iterration in case the failure was a port conflict race
 		port := 0
 		if a, err := net.ResolveTCPAddr("tcp", "localhost:0"); err == nil {
@@ -246,13 +234,13 @@ func NewLlamaServer(gpus gpu.GpuInfoList, model string, ggml *GGML, adapters, pr
 		if runtime.GOOS == "windows" {
 			pathEnv = "PATH"
 		}
-		// prepend the server directory to LD_LIBRARY_PATH/PATH
+		// append the server directory to LD_LIBRARY_PATH/PATH
 		libraryPaths := []string{dir}
 
 		if libraryPath, ok := os.LookupEnv(pathEnv); ok {
 			// Append our runner directory to the path
 			// This will favor system libraries over our bundled library dependencies
-			libraryPaths = append(libraryPaths, filepath.SplitList(libraryPath)...)
+			libraryPaths = append(filepath.SplitList(libraryPath), libraryPaths...)
 		}
 
 		// Note: we always put the dependency path first
@@ -280,37 +268,23 @@ func NewLlamaServer(gpus gpu.GpuInfoList, model string, ggml *GGML, adapters, pr
 		}
 
 		s := &llmServer{
-			port:           port,
-			cmd:            exec.Command(server, finalParams...),
-			status:         NewStatusWriter(os.Stderr),
-			options:        opts,
-			estimatedVRAM:  estimatedVRAM,
-			estimatedTotal: estimatedTotal,
-			sem:            semaphore.NewWeighted(int64(numParallel)),
-			totalLayers:    ggml.KV().BlockCount() + 1,
-			gpuCount:       gpuCount,
-			done:           make(chan error, 1),
+			port:          port,
+			cmd:           exec.Command(server, finalParams...),
+			status:        NewStatusWriter(os.Stderr),
+			options:       opts,
+			estimatedVRAM: estimatedVRAM,
+			sem:           semaphore.NewWeighted(int64(numParallel)),
 		}
 
+		libEnv := fmt.Sprintf("%s=%s", pathEnv, strings.Join(libraryPaths, string(filepath.ListSeparator)))
+		s.cmd.Env = append(os.Environ(), libEnv)
 		s.cmd.Stdout = os.Stdout
 		s.cmd.Stderr = s.status
 
-		if v := strings.Join(libraryPaths, string(filepath.ListSeparator)); v != "" {
-			s.cmd.Env = append(s.cmd.Env, pathEnv+"="+v)
-		}
-
-		if k, v := gpu.GpuInfoList(gpus).GetVisibleDevicesEnv(); k != "" {
-			s.cmd.Env = append(s.cmd.Env, k+"="+v)
-		}
-
-		for _, ev := range os.Environ() {
-			if strings.HasPrefix(ev, "CUDA_") ||
-				strings.HasPrefix(ev, "ROCM_") ||
-				strings.HasPrefix(ev, "HIP_") ||
-				strings.HasPrefix(ev, "HSA_") ||
-				strings.HasPrefix(ev, "GGML_") {
-				s.cmd.Env = append(s.cmd.Env, ev)
-			}
+		// TODO - multiple GPU selection logic...
+		key, val := gpu.GpuInfoList(gpus).GetVisibleDevicesEnv()
+		if key != "" {
+			s.cmd.Env = append(s.cmd.Env, key+"="+val)
 		}
 
 		slog.Info("starting llama server", "cmd", s.cmd.String())
@@ -318,11 +292,6 @@ func NewLlamaServer(gpus gpu.GpuInfoList, model string, ggml *GGML, adapters, pr
 		slog.Debug("subprocess", "environment", s.cmd.Env)
 
 		if err = s.cmd.Start(); err != nil {
-			// Detect permission denied and augment them essage about noexec
-			if errors.Is(err, os.ErrPermission) {
-				finalErr = fmt.Errorf("unable to start server %w.  %s may have noexec set.  Set OLLAMA_TMPDIR for server to a writable executable directory", err, dir)
-				continue
-			}
 			msg := ""
 			if s.status != nil && s.status.LastErrMsg != "" {
 				msg = s.status.LastErrMsg
@@ -334,9 +303,17 @@ func NewLlamaServer(gpus gpu.GpuInfoList, model string, ggml *GGML, adapters, pr
 
 		// reap subprocess when it exits
 		go func() {
-			s.done <- s.cmd.Wait()
+			// Exit status managed via getServerStatus
+			_ = s.cmd.Wait()
 		}()
 
+		// TODO - make sure this is all wired up correctly
+		// if err = s.WaitUntilRunning(); err != nil {
+		// 	slog.Error("error starting llama server", "server", servers[i], "error", err)
+		// 	s.Close()
+		// 	finalErr = err
+		// 	continue
+		// }
 		return s, nil
 	}
 
@@ -368,7 +345,7 @@ type ServerStatus int
 
 const ( // iota is reset to 0
 	ServerStatusReady ServerStatus = iota
-	ServerStatusNoSlotsAvailable
+	ServerStatusNoSlotsAvaialble
 	ServerStatusLoadingModel
 	ServerStatusNotResponding
 	ServerStatusError
@@ -378,7 +355,7 @@ func (s ServerStatus) ToString() string {
 	switch s {
 	case ServerStatusReady:
 		return "llm server ready"
-	case ServerStatusNoSlotsAvailable:
+	case ServerStatusNoSlotsAvaialble:
 		return "llm busy - no slots available"
 	case ServerStatusLoadingModel:
 		return "llm server loading model"
@@ -403,10 +380,6 @@ func (s *llmServer) getServerStatus(ctx context.Context) (ServerStatus, error) {
 		if s.status != nil && s.status.LastErrMsg != "" {
 			msg = s.status.LastErrMsg
 		}
-		if s.cmd.ProcessState.ExitCode() == -1 {
-			// Most likely a signal killed it, log some more details to try to help troubleshoot
-			slog.Warn("llama runner process no longer running", "sys", s.cmd.ProcessState.Sys(), "string", s.cmd.ProcessState.String())
-		}
 		return ServerStatusError, fmt.Errorf("llama runner process no longer running: %d %s", s.cmd.ProcessState.ExitCode(), msg)
 	}
 
@@ -439,7 +412,7 @@ func (s *llmServer) getServerStatus(ctx context.Context) (ServerStatus, error) {
 	case "ok":
 		return ServerStatusReady, nil
 	case "no slot available":
-		return ServerStatusNoSlotsAvailable, nil
+		return ServerStatusNoSlotsAvaialble, nil
 	case "loading model":
 		return ServerStatusLoadingModel, nil
 	default:
@@ -447,29 +420,6 @@ func (s *llmServer) getServerStatus(ctx context.Context) (ServerStatus, error) {
 	}
 }
 
-// getServerStatusRetry will retry if ServerStatusNoSlotsAvailable is received
-func (s *llmServer) getServerStatusRetry(ctx context.Context) (ServerStatus, error) {
-	var retries int
-	for {
-		status, err := s.getServerStatus(ctx)
-		if err != nil {
-			return status, err
-		}
-
-		if status == ServerStatusNoSlotsAvailable {
-			if retries >= 10 {
-				return status, fmt.Errorf("no slots available after %d retries", retries)
-			}
-
-			time.Sleep(5 * time.Millisecond)
-			retries++
-			continue
-		}
-
-		return status, nil
-	}
-}
-
 func (s *llmServer) Ping(ctx context.Context) error {
 	_, err := s.getServerStatus(ctx)
 	if err != nil {
@@ -481,11 +431,13 @@ func (s *llmServer) Ping(ctx context.Context) error {
 
 func (s *llmServer) WaitUntilRunning(ctx context.Context) error {
 	start := time.Now()
+	// TODO we need to wire up a better way to detect hangs during model load and startup of the server
 	expiresAt := time.Now().Add(10 * time.Minute) // be generous with timeout, large models can take a while to load
+	ticker := time.NewTicker(50 * time.Millisecond)
+	defer ticker.Stop()
 
 	slog.Info("waiting for llama runner to start responding")
 	var lastStatus ServerStatus = -1
-
 	for {
 		select {
 		case <-ctx.Done():
@@ -497,39 +449,41 @@ func (s *llmServer) WaitUntilRunning(ctx context.Context) error {
 				msg = s.status.LastErrMsg
 			}
 			return fmt.Errorf("llama runner process has terminated: %v %s", err, msg)
-		default:
-		}
-		if time.Now().After(expiresAt) {
-			// timeout
-			msg := ""
-			if s.status != nil && s.status.LastErrMsg != "" {
-				msg = s.status.LastErrMsg
+		case <-ticker.C:
+			if time.Now().After(expiresAt) {
+				// timeout
+				msg := ""
+				if s.status != nil && s.status.LastErrMsg != "" {
+					msg = s.status.LastErrMsg
+				}
+				return fmt.Errorf("timed out waiting for llama runner to start: %s", msg)
 			}
-			return fmt.Errorf("timed out waiting for llama runner to start: %s", msg)
-		}
-		if s.cmd.ProcessState != nil {
-			msg := ""
-			if s.status != nil && s.status.LastErrMsg != "" {
-				msg = s.status.LastErrMsg
+			if s.cmd.ProcessState != nil {
+				msg := ""
+				if s.status != nil && s.status.LastErrMsg != "" {
+					msg = s.status.LastErrMsg
+				}
+				return fmt.Errorf("llama runner process no longer running: %d %s", s.cmd.ProcessState.ExitCode(), msg)
+			}
+
+			c, cancel := context.WithTimeout(ctx, 200*time.Millisecond)
+			defer cancel()
+			status, err := s.getServerStatus(c)
+			if err != nil && lastStatus != status {
+				slog.Debug("server not yet available", "error", err)
+				lastStatus = status
+				continue
+			}
+
+			switch status {
+			case ServerStatusLoadingModel:
+				// TODO - this state never seems to happen with the current server.cpp code (bug?)
+				// it doesn't respond to the health endpoint until after the model is loaded
+				slog.Debug("loading model")
+			case ServerStatusReady:
+				slog.Debug(fmt.Sprintf("llama runner started in %f seconds", time.Since(start).Seconds()))
+				return nil
 			}
-			return fmt.Errorf("llama runner process no longer running: %d %s", s.cmd.ProcessState.ExitCode(), msg)
-		}
-		ctx, cancel := context.WithTimeout(ctx, 200*time.Millisecond)
-		defer cancel()
-		status, _ := s.getServerStatus(ctx)
-		if lastStatus != status && status != ServerStatusReady {
-			// Only log on status changes
-			slog.Info("waiting for server to become available", "status", status.ToString())
-		}
-		switch status {
-		case ServerStatusReady:
-			s.loadDuration = time.Since(start)
-			slog.Info(fmt.Sprintf("llama runner started in %0.2f seconds", s.loadDuration.Seconds()))
-			return nil
-		default:
-			lastStatus = status
-			time.Sleep(time.Millisecond * 250)
-			continue
 		}
 	}
 }
@@ -563,6 +517,7 @@ ws ::= ([ \t\n] ws)?
 `
 
 const maxBufferSize = 512 * format.KiloByte
+const maxRetries = 3
 
 type ImageData struct {
 	Data []byte `json:"data"`
@@ -570,11 +525,10 @@ type ImageData struct {
 }
 
 type completion struct {
-	Content      string `json:"content"`
-	Model        string `json:"model"`
-	Prompt       string `json:"prompt"`
-	Stop         bool   `json:"stop"`
-	StoppedLimit bool   `json:"stopped_limit"`
+	Content string `json:"content"`
+	Model   string `json:"model"`
+	Prompt  string `json:"prompt"`
+	Stop    bool   `json:"stop"`
 
 	Timings struct {
 		PredictedN  int     `json:"predicted_n"`
@@ -593,7 +547,6 @@ type CompletionRequest struct {
 
 type CompletionResponse struct {
 	Content            string
-	DoneReason         string
 	Done               bool
 	PromptEvalCount    int
 	PromptEvalDuration time.Duration
@@ -640,7 +593,7 @@ func (s *llmServer) Completion(ctx context.Context, req CompletionRequest, fn fu
 	}
 
 	// Make sure the server is ready
-	status, err := s.getServerStatusRetry(ctx)
+	status, err := s.getServerStatus(ctx)
 	if err != nil {
 		return err
 	} else if status != ServerStatusReady {
@@ -654,150 +607,163 @@ func (s *llmServer) Completion(ctx context.Context, req CompletionRequest, fn fu
 		}
 	}
 
-	// Handling JSON marshaling with special characters unescaped.
-	buffer := &bytes.Buffer{}
-	enc := json.NewEncoder(buffer)
-	enc.SetEscapeHTML(false)
-
-	if err := enc.Encode(request); err != nil {
-		return fmt.Errorf("failed to marshal data: %v", err)
-	}
-
-	endpoint := fmt.Sprintf("http://127.0.0.1:%d/completion", s.port)
-	serverReq, err := http.NewRequestWithContext(ctx, http.MethodPost, endpoint, buffer)
-	if err != nil {
-		return fmt.Errorf("error creating POST request: %v", err)
-	}
-	serverReq.Header.Set("Content-Type", "application/json")
-
-	res, err := http.DefaultClient.Do(serverReq)
-	if err != nil {
-		return fmt.Errorf("POST predict: %v", err)
-	}
-	defer res.Body.Close()
-
-	if res.StatusCode >= 400 {
-		bodyBytes, err := io.ReadAll(res.Body)
-		if err != nil {
-			return fmt.Errorf("failed reading llm error response: %w", err)
+	retryDelay := 100 * time.Microsecond
+	for retries := 0; retries < maxRetries; retries++ {
+		if retries > 0 {
+			time.Sleep(retryDelay) // wait before retrying
+			retryDelay *= 2        // exponential backoff
 		}
-		log.Printf("llm predict error: %s", bodyBytes)
-		return fmt.Errorf("%s", bodyBytes)
-	}
 
-	scanner := bufio.NewScanner(res.Body)
-	buf := make([]byte, 0, maxBufferSize)
-	scanner.Buffer(buf, maxBufferSize)
+		// Handling JSON marshaling with special characters unescaped.
+		buffer := &bytes.Buffer{}
+		enc := json.NewEncoder(buffer)
+		enc.SetEscapeHTML(false)
 
-	// keep track of the last token generated, this is used to abort if the model starts looping
-	var lastToken string
-	var tokenRepeat int
+		if err := enc.Encode(request); err != nil {
+			return fmt.Errorf("failed to marshal data: %v", err)
+		}
 
-	for scanner.Scan() {
-		select {
-		case <-ctx.Done():
-			// This handles the request cancellation
-			return ctx.Err()
-		default:
-			line := scanner.Bytes()
-			if len(line) == 0 {
-				continue
+		endpoint := fmt.Sprintf("http://127.0.0.1:%d/completion", s.port)
+		req, err := http.NewRequestWithContext(ctx, http.MethodPost, endpoint, buffer)
+		if err != nil {
+			return fmt.Errorf("error creating POST request: %v", err)
+		}
+		req.Header.Set("Content-Type", "application/json")
+
+		resp, err := http.DefaultClient.Do(req)
+		if err != nil {
+			return fmt.Errorf("POST predict: %v", err)
+		}
+		defer resp.Body.Close()
+
+		if resp.StatusCode >= 400 {
+			bodyBytes, err := io.ReadAll(resp.Body)
+			if err != nil {
+				return fmt.Errorf("failed reading llm error response: %w", err)
 			}
+			log.Printf("llm predict error: %s", bodyBytes)
+			return fmt.Errorf("%s", bodyBytes)
+		}
 
-			evt, ok := bytes.CutPrefix(line, []byte("data: "))
-			if !ok {
-				return fmt.Errorf("error parsing llm response stream: %s", line)
-			}
+		scanner := bufio.NewScanner(resp.Body)
+		buf := make([]byte, 0, maxBufferSize)
+		scanner.Buffer(buf, maxBufferSize)
 
-			var c completion
-			if err := json.Unmarshal(evt, &c); err != nil {
-				return fmt.Errorf("error unmarshaling llm prediction response: %v", err)
-			}
+		retryNeeded := false
+		// keep track of the last token generated, this is used to abort if the model starts looping
+		var lastToken string
+		var tokenRepeat int
 
-			switch {
-			case strings.TrimSpace(c.Content) == lastToken:
-				tokenRepeat++
-			default:
-				lastToken = strings.TrimSpace(c.Content)
-				tokenRepeat = 0
-			}
-
-			// 30 picked as an arbitrary max token repeat limit, modify as needed
-			if tokenRepeat > 30 {
-				slog.Debug("prediction aborted, token repeat limit reached")
+		for scanner.Scan() {
+			select {
+			case <-ctx.Done():
+				// This handles the request cancellation
 				return ctx.Err()
-			}
-
-			if c.Content != "" {
-				fn(CompletionResponse{
-					Content: c.Content,
-				})
-			}
-
-			if c.Stop {
-				doneReason := "stop"
-				if c.StoppedLimit {
-					doneReason = "length"
+			default:
+				line := scanner.Bytes()
+				if len(line) == 0 {
+					continue
 				}
 
-				fn(CompletionResponse{
-					Done:               true,
-					DoneReason:         doneReason,
-					PromptEvalCount:    c.Timings.PromptN,
-					PromptEvalDuration: parseDurationMs(c.Timings.PromptMS),
-					EvalCount:          c.Timings.PredictedN,
-					EvalDuration:       parseDurationMs(c.Timings.PredictedMS),
-				})
-				return nil
+				// try again on slot unavailable
+				if bytes.Contains(line, []byte("slot unavailable")) {
+					retryNeeded = true
+					break
+				}
+
+				evt, ok := bytes.CutPrefix(line, []byte("data: "))
+				if !ok {
+					return fmt.Errorf("error parsing llm response stream: %s", line)
+				}
+
+				var c completion
+				if err := json.Unmarshal(evt, &c); err != nil {
+					return fmt.Errorf("error unmarshaling llm prediction response: %v", err)
+				}
+
+				switch {
+				case strings.TrimSpace(c.Content) == lastToken:
+					tokenRepeat++
+				default:
+					lastToken = strings.TrimSpace(c.Content)
+					tokenRepeat = 0
+				}
+
+				// 30 picked as an arbitrary max token repeat limit, modify as needed
+				if tokenRepeat > 30 {
+					slog.Debug("prediction aborted, token repeat limit reached")
+					return ctx.Err()
+				}
+
+				if c.Content != "" {
+					fn(CompletionResponse{
+						Content: c.Content,
+					})
+				}
+
+				if c.Stop {
+					fn(CompletionResponse{
+						Done:               true,
+						PromptEvalCount:    c.Timings.PromptN,
+						PromptEvalDuration: parseDurationMs(c.Timings.PromptMS),
+						EvalCount:          c.Timings.PredictedN,
+						EvalDuration:       parseDurationMs(c.Timings.PredictedMS),
+					})
+					return nil
+				}
 			}
 		}
+
+		if err := scanner.Err(); err != nil {
+			if strings.Contains(err.Error(), "unexpected EOF") {
+				s.Close()
+				msg := ""
+				if s.status != nil && s.status.LastErrMsg != "" {
+					msg = s.status.LastErrMsg
+				}
+
+				return fmt.Errorf("an unknown error was encountered while running the model %s", msg)
+			}
+			return fmt.Errorf("error reading llm response: %v", err)
+		}
+
+		if !retryNeeded {
+			return nil // success
+		}
 	}
 
-	if err := scanner.Err(); err != nil {
-		if strings.Contains(err.Error(), "unexpected EOF") {
-			s.Close()
-			msg := ""
-			if s.status != nil && s.status.LastErrMsg != "" {
-				msg = s.status.LastErrMsg
-			}
-			return fmt.Errorf("an unknown error was encountered while running the model %s", msg)
-		}
-
-		return fmt.Errorf("error reading llm response: %v", err)
-	}
-
-	return nil
+	// should never reach here ideally
+	return fmt.Errorf("max retries exceeded")
 }
 
-type EmbeddingRequest struct {
-	Content string `json:"content"`
+type EmbeddingsRequest struct {
+	Contents []string `json:"contents"`
 }
 
-type EmbeddingResponse struct {
-	Embedding []float64 `json:"embedding"`
+type EmbeddingsResponse struct {
+	Embeddings [][]float64 `json:"embeddings"`
 }
 
-func (s *llmServer) Embedding(ctx context.Context, prompt string) ([]float64, error) {
+func (s *llmServer) Embeddings(ctx context.Context, prompts []string) ([][]float64, error) {
 	if err := s.sem.Acquire(ctx, 1); err != nil {
 		slog.Error("Failed to acquire semaphore", "error", err)
 		return nil, err
 	}
 	defer s.sem.Release(1)
-
 	// Make sure the server is ready
-	status, err := s.getServerStatusRetry(ctx)
+	status, err := s.getServerStatus(ctx)
 	if err != nil {
 		return nil, err
 	} else if status != ServerStatusReady {
 		return nil, fmt.Errorf("unexpected server status: %s", status.ToString())
 	}
 
-	data, err := json.Marshal(TokenizeRequest{Content: prompt})
+	data, err := json.Marshal(EmbeddingsRequest{Contents: prompts})
 	if err != nil {
 		return nil, fmt.Errorf("error marshaling embed data: %w", err)
 	}
 
-	req, err := http.NewRequestWithContext(ctx, http.MethodPost, fmt.Sprintf("http://127.0.0.1:%d/embedding", s.port), bytes.NewBuffer(data))
+	req, err := http.NewRequestWithContext(ctx, http.MethodPost, fmt.Sprintf("http://127.0.0.1:%d/embeddings", s.port), bytes.NewBuffer(data))
 	if err != nil {
 		return nil, fmt.Errorf("error creating embed request: %w", err)
 	}
@@ -814,17 +780,19 @@ func (s *llmServer) Embedding(ctx context.Context, prompt string) ([]float64, er
 		return nil, fmt.Errorf("error reading embed response: %w", err)
 	}
 
+	fmt.Println("embeddings response", string(body))
+
 	if resp.StatusCode >= 400 {
 		log.Printf("llm encode error: %s", body)
 		return nil, fmt.Errorf("%s", body)
 	}
 
-	var embedding EmbeddingResponse
+	var embedding EmbeddingsResponse
 	if err := json.Unmarshal(body, &embedding); err != nil {
 		return nil, fmt.Errorf("unmarshal tokenize response: %w", err)
 	}
 
-	return embedding.Embedding, nil
+	return embedding.Embeddings, nil
 }
 
 type TokenizeRequest struct {
@@ -840,7 +808,7 @@ func (s *llmServer) Tokenize(ctx context.Context, content string) ([]int, error)
 	status, err := s.getServerStatus(ctx)
 	if err != nil {
 		return nil, err
-	} else if status != ServerStatusReady && status != ServerStatusNoSlotsAvailable {
+	} else if status != ServerStatusReady && status != ServerStatusNoSlotsAvaialble {
 		return nil, fmt.Errorf("unexpected server status: %s", status.ToString())
 	}
 
@@ -892,7 +860,7 @@ func (s *llmServer) Detokenize(ctx context.Context, tokens []int) (string, error
 	status, err := s.getServerStatus(ctx)
 	if err != nil {
 		return "", err
-	} else if status != ServerStatusReady && status != ServerStatusNoSlotsAvailable {
+	} else if status != ServerStatusReady && status != ServerStatusNoSlotsAvaialble {
 		return "", fmt.Errorf("unexpected server status: %s", status.ToString())
 	}
 
@@ -934,16 +902,7 @@ func (s *llmServer) Detokenize(ctx context.Context, tokens []int) (string, error
 func (s *llmServer) Close() error {
 	if s.cmd != nil {
 		slog.Debug("stopping llama server")
-		if err := s.cmd.Process.Kill(); err != nil {
-			return err
-		}
-		// if ProcessState is already populated, Wait already completed, no need to wait again
-		if s.cmd.ProcessState == nil {
-			slog.Debug("waiting for llama server to exit")
-			<-s.done
-		}
-
-		slog.Debug("llama server stopped")
+		return s.cmd.Process.Kill()
 	}
 
 	return nil

macapp/src/app.tsx

@@ -19,7 +19,7 @@ export default function () {
   const [step, setStep] = useState<Step>(Step.WELCOME)
   const [commandCopied, setCommandCopied] = useState<boolean>(false)
 
-  const command = 'ollama run llama3'
+  const command = 'ollama run llama2'
 
   return (
     <div className='drag'>

openai/openai.go

@@ -107,9 +107,15 @@ func toChatCompletion(id string, r api.ChatResponse) ChatCompletion {
 		Model:             r.Model,
 		SystemFingerprint: "fp_ollama",
 		Choices: []Choice{{
-			Index:        0,
-			Message:      Message{Role: r.Message.Role, Content: r.Message.Content},
-			FinishReason: &r.DoneReason,
+			Index:   0,
+			Message: Message{Role: r.Message.Role, Content: r.Message.Content},
+			FinishReason: func(done bool) *string {
+				if done {
+					reason := "stop"
+					return &reason
+				}
+				return nil
+			}(r.Done),
 		}},
 		Usage: Usage{
 			// TODO: ollama returns 0 for prompt eval if the prompt was cached, but openai returns the actual count
@@ -129,9 +135,15 @@ func toChunk(id string, r api.ChatResponse) ChatCompletionChunk {
 		SystemFingerprint: "fp_ollama",
 		Choices: []ChunkChoice{
 			{
-				Index:        0,
-				Delta:        Message{Role: "assistant", Content: r.Message.Content},
-				FinishReason: &r.DoneReason,
+				Index: 0,
+				Delta: Message{Role: "assistant", Content: r.Message.Content},
+				FinishReason: func(done bool) *string {
+					if done {
+						reason := "stop"
+						return &reason
+					}
+					return nil
+				}(r.Done),
 			},
 		},
 	}

parser/parser.go

@@ -0,0 +1,132 @@
+package parser
+
+import (
+	"bufio"
+	"bytes"
+	"errors"
+	"fmt"
+	"io"
+	"log/slog"
+	"slices"
+)
+
+type Command struct {
+	Name string
+	Args string
+}
+
+func (c *Command) Reset() {
+	c.Name = ""
+	c.Args = ""
+}
+
+func Parse(reader io.Reader) ([]Command, error) {
+	var commands []Command
+	var command, modelCommand Command
+
+	scanner := bufio.NewScanner(reader)
+	scanner.Buffer(make([]byte, 0, bufio.MaxScanTokenSize), bufio.MaxScanTokenSize)
+	scanner.Split(scanModelfile)
+	for scanner.Scan() {
+		line := scanner.Bytes()
+
+		fields := bytes.SplitN(line, []byte(" "), 2)
+		if len(fields) == 0 || len(fields[0]) == 0 {
+			continue
+		}
+
+		switch string(bytes.ToUpper(fields[0])) {
+		case "FROM":
+			command.Name = "model"
+			command.Args = string(bytes.TrimSpace(fields[1]))
+			// copy command for validation
+			modelCommand = command
+		case "ADAPTER":
+			command.Name = string(bytes.ToLower(fields[0]))
+			command.Args = string(bytes.TrimSpace(fields[1]))
+		case "LICENSE", "TEMPLATE", "SYSTEM", "PROMPT":
+			command.Name = string(bytes.ToLower(fields[0]))
+			command.Args = string(fields[1])
+		case "PARAMETER":
+			fields = bytes.SplitN(fields[1], []byte(" "), 2)
+			if len(fields) < 2 {
+				return nil, fmt.Errorf("missing value for %s", fields)
+			}
+
+			command.Name = string(fields[0])
+			command.Args = string(bytes.TrimSpace(fields[1]))
+		case "EMBED":
+			return nil, fmt.Errorf("deprecated command: EMBED is no longer supported, use the /embed API endpoint instead")
+		case "MESSAGE":
+			command.Name = string(bytes.ToLower(fields[0]))
+			fields = bytes.SplitN(fields[1], []byte(" "), 2)
+			if len(fields) < 2 {
+				return nil, fmt.Errorf("should be in the format <role> <message>")
+			}
+			if !slices.Contains([]string{"system", "user", "assistant"}, string(bytes.ToLower(fields[0]))) {
+				return nil, fmt.Errorf("role must be one of \"system\", \"user\", or \"assistant\"")
+			}
+			command.Args = fmt.Sprintf("%s: %s", string(bytes.ToLower(fields[0])), string(fields[1]))
+		default:
+			if !bytes.HasPrefix(fields[0], []byte("#")) {
+				// log a warning for unknown commands
+				slog.Warn(fmt.Sprintf("Unknown command: %s", fields[0]))
+			}
+			continue
+		}
+
+		commands = append(commands, command)
+		command.Reset()
+	}
+
+	if modelCommand.Args == "" {
+		return nil, errors.New("no FROM line for the model was specified")
+	}
+
+	return commands, scanner.Err()
+}
+
+func scanModelfile(data []byte, atEOF bool) (advance int, token []byte, err error) {
+	advance, token, err = scan([]byte(`"""`), []byte(`"""`), data, atEOF)
+	if err != nil {
+		return 0, nil, err
+	}
+
+	if advance > 0 && token != nil {
+		return advance, token, nil
+	}
+
+	advance, token, err = scan([]byte(`"`), []byte(`"`), data, atEOF)
+	if err != nil {
+		return 0, nil, err
+	}
+
+	if advance > 0 && token != nil {
+		return advance, token, nil
+	}
+
+	return bufio.ScanLines(data, atEOF)
+}
+
+func scan(openBytes, closeBytes, data []byte, atEOF bool) (advance int, token []byte, err error) {
+	newline := bytes.IndexByte(data, '\n')
+
+	if start := bytes.Index(data, openBytes); start >= 0 && start < newline {
+		end := bytes.Index(data[start+len(openBytes):], closeBytes)
+		if end < 0 {
+			if atEOF {
+				return 0, nil, fmt.Errorf("unterminated %s: expecting %s", openBytes, closeBytes)
+			} else {
+				return 0, nil, nil
+			}
+		}
+
+		n := start + len(openBytes) + end + len(closeBytes)
+
+		newData := data[:start]
+		newData = append(newData, data[start+len(openBytes):n-len(closeBytes)]...)
+		return n, newData, nil
+	}
+
+	return 0, nil, nil
+}

parser/parser_test.go

@@ -0,0 +1,98 @@
+package parser
+
+import (
+	"strings"
+	"testing"
+
+	"github.com/stretchr/testify/assert"
+)
+
+func Test_Parser(t *testing.T) {
+
+	input := `
+FROM model1
+ADAPTER adapter1
+LICENSE MIT
+PARAMETER param1 value1
+PARAMETER param2 value2
+TEMPLATE template1
+`
+
+	reader := strings.NewReader(input)
+
+	commands, err := Parse(reader)
+	assert.Nil(t, err)
+
+	expectedCommands := []Command{
+		{Name: "model", Args: "model1"},
+		{Name: "adapter", Args: "adapter1"},
+		{Name: "license", Args: "MIT"},
+		{Name: "param1", Args: "value1"},
+		{Name: "param2", Args: "value2"},
+		{Name: "template", Args: "template1"},
+	}
+
+	assert.Equal(t, expectedCommands, commands)
+}
+
+func Test_Parser_NoFromLine(t *testing.T) {
+
+	input := `
+PARAMETER param1 value1
+PARAMETER param2 value2
+`
+
+	reader := strings.NewReader(input)
+
+	_, err := Parse(reader)
+	assert.ErrorContains(t, err, "no FROM line")
+}
+
+func Test_Parser_MissingValue(t *testing.T) {
+
+	input := `
+FROM foo
+PARAMETER param1
+`
+
+	reader := strings.NewReader(input)
+
+	_, err := Parse(reader)
+	assert.ErrorContains(t, err, "missing value for [param1]")
+
+}
+
+func Test_Parser_Messages(t *testing.T) {
+
+	input := `
+FROM foo
+MESSAGE system You are a Parser. Always Parse things.
+MESSAGE user Hey there!
+MESSAGE assistant Hello, I want to parse all the things!
+`
+
+	reader := strings.NewReader(input)
+	commands, err := Parse(reader)
+	assert.Nil(t, err)
+
+	expectedCommands := []Command{
+		{Name: "model", Args: "foo"},
+		{Name: "message", Args: "system: You are a Parser. Always Parse things."},
+		{Name: "message", Args: "user: Hey there!"},
+		{Name: "message", Args: "assistant: Hello, I want to parse all the things!"},
+	}
+
+	assert.Equal(t, expectedCommands, commands)
+}
+
+func Test_Parser_Messages_BadRole(t *testing.T) {
+
+	input := `
+FROM foo
+MESSAGE badguy I'm a bad guy!
+`
+
+	reader := strings.NewReader(input)
+	_, err := Parse(reader)
+	assert.ErrorContains(t, err, "role must be one of \"system\", \"user\", or \"assistant\"")
+}

readline/readline.go

@@ -218,7 +218,7 @@ func (i *Instance) Readline() (string, error) {
 		case CharCtrlZ:
 			fd := int(syscall.Stdin)
 			return handleCharCtrlZ(fd, i.Terminal.termios)
-		case CharEnter, CharCtrlJ:
+		case CharEnter:
 			output := buf.String()
 			if output != "" {
 				i.History.Add([]rune(output))
@@ -232,7 +232,7 @@ func (i *Instance) Readline() (string, error) {
 				metaDel = false
 				continue
 			}
-			if r >= CharSpace || r == CharEnter || r == CharCtrlJ {
+			if r >= CharSpace || r == CharEnter {
 				buf.Add(r)
 			}
 		}

scripts/build_windows.ps1

@@ -7,8 +7,6 @@
 $ErrorActionPreference = "Stop"
 
 function checkEnv() {
-    $script:TARGET_ARCH=$Env:PROCESSOR_ARCHITECTURE.ToLower()
-    Write-host "Building for ${script:TARGET_ARCH}"
     write-host "Locating required tools and paths"
     $script:SRC_DIR=$PWD
     if (!$env:VCToolsRedistDir) {
@@ -32,7 +30,7 @@ function checkEnv() {
     
     $script:INNO_SETUP_DIR=(get-item "C:\Program Files*\Inno Setup*\")[0]
 
-    $script:DEPS_DIR="${script:SRC_DIR}\dist\windows-${script:TARGET_ARCH}"
+    $script:DEPS_DIR="${script:SRC_DIR}\dist\windows-amd64"
     $env:CGO_ENABLED="1"
     echo "Checking version"
     if (!$env:VERSION) {
@@ -83,8 +81,8 @@ function buildOllama() {
             /csp "Google Cloud KMS Provider" /kc ${env:KEY_CONTAINER} ollama.exe
         if ($LASTEXITCODE -ne 0) { exit($LASTEXITCODE)}
     }
-    New-Item -ItemType Directory -Path .\dist\windows-${script:TARGET_ARCH}\ -Force
-    cp .\ollama.exe .\dist\windows-${script:TARGET_ARCH}\
+    New-Item -ItemType Directory -Path .\dist\windows-amd64\ -Force
+    cp .\ollama.exe .\dist\windows-amd64\
 }
 
 function buildApp() {
@@ -129,16 +127,16 @@ function buildInstaller() {
     cd "${script:SRC_DIR}\app"
     $env:PKG_VERSION=$script:PKG_VERSION
     if ("${env:KEY_CONTAINER}") {
-        & "${script:INNO_SETUP_DIR}\ISCC.exe" /DARCH=$script:TARGET_ARCH /SMySignTool="${script:SignTool} sign /fd sha256 /t http://timestamp.digicert.com /f ${script:OLLAMA_CERT} /csp `$qGoogle Cloud KMS Provider`$q /kc ${env:KEY_CONTAINER} `$f" .\ollama.iss
+        & "${script:INNO_SETUP_DIR}\ISCC.exe" /SMySignTool="${script:SignTool} sign /fd sha256 /t http://timestamp.digicert.com /f ${script:OLLAMA_CERT} /csp `$qGoogle Cloud KMS Provider`$q /kc ${env:KEY_CONTAINER} `$f" .\ollama.iss
     } else {
-        & "${script:INNO_SETUP_DIR}\ISCC.exe" /DARCH=$script:TARGET_ARCH .\ollama.iss
+        & "${script:INNO_SETUP_DIR}\ISCC.exe" .\ollama.iss
     }
     if ($LASTEXITCODE -ne 0) { exit($LASTEXITCODE)}
 }
 
 function distZip() {
-    write-host "Generating stand-alone distribution zip file ${script:SRC_DIR}\dist\ollama-windows-${script:TARGET_ARCH}.zip"
-    Compress-Archive -Path "${script:SRC_DIR}\dist\windows-${script:TARGET_ARCH}\*" -DestinationPath "${script:SRC_DIR}\dist\ollama-windows-${script:TARGET_ARCH}.zip" -Force
+    write-host "Generating stand-alone distribution zip file ${script:SRC_DIR}\dist\ollama-windows-amd64.zip"
+    Compress-Archive -Path "${script:SRC_DIR}\dist\windows-amd64\*" -DestinationPath "${script:SRC_DIR}\dist\ollama-windows-amd64.zip" -Force
 }
 
 try {

scripts/install.sh

@@ -166,8 +166,8 @@ fi
 
 if check_gpu lspci amdgpu || check_gpu lshw amdgpu; then
     # Look for pre-existing ROCm v6 before downloading the dependencies
-    for search in "${HIP_PATH:-''}" "${ROCM_PATH:-''}" "/opt/rocm" "/usr/lib64"; do
-        if [ -n "${search}" ] && [ -e "${search}/libhipblas.so.2" -o -e "${search}/lib/libhipblas.so.2" ]; then
+    for search in "${HIP_PATH:-''}" "${ROCM_PATH:-''}" "/opt/rocm"; do
+        if [ -n "${search}" ] && [ -e "${search}/lib/libhipblas.so.2" ]; then
             status "Compatible AMD GPU ROCm library detected at ${search}"
             install_success
             exit 0

server/envconfig/config.go

@@ -1,174 +0,0 @@
-package envconfig
-
-import (
-	"fmt"
-	"log/slog"
-	"os"
-	"path/filepath"
-	"runtime"
-	"strconv"
-	"strings"
-)
-
-var (
-	// Set via OLLAMA_ORIGINS in the environment
-	AllowOrigins []string
-	// Set via OLLAMA_DEBUG in the environment
-	Debug bool
-	// Set via OLLAMA_LLM_LIBRARY in the environment
-	LLMLibrary string
-	// Set via OLLAMA_MAX_LOADED_MODELS in the environment
-	MaxRunners int
-	// Set via OLLAMA_MAX_QUEUE in the environment
-	MaxQueuedRequests int
-	// Set via OLLAMA_MAX_VRAM in the environment
-	MaxVRAM uint64
-	// Set via OLLAMA_NOPRUNE in the environment
-	NoPrune bool
-	// Set via OLLAMA_NUM_PARALLEL in the environment
-	NumParallel int
-	// Set via OLLAMA_RUNNERS_DIR in the environment
-	RunnersDir string
-	// Set via OLLAMA_TMPDIR in the environment
-	TmpDir string
-)
-
-func AsMap() map[string]string {
-	return map[string]string{
-		"OLLAMA_ORIGINS":           fmt.Sprintf("%v", AllowOrigins),
-		"OLLAMA_DEBUG":             fmt.Sprintf("%v", Debug),
-		"OLLAMA_LLM_LIBRARY":       fmt.Sprintf("%v", LLMLibrary),
-		"OLLAMA_MAX_LOADED_MODELS": fmt.Sprintf("%v", MaxRunners),
-		"OLLAMA_MAX_QUEUE":         fmt.Sprintf("%v", MaxQueuedRequests),
-		"OLLAMA_MAX_VRAM":          fmt.Sprintf("%v", MaxVRAM),
-		"OLLAMA_NOPRUNE":           fmt.Sprintf("%v", NoPrune),
-		"OLLAMA_NUM_PARALLEL":      fmt.Sprintf("%v", NumParallel),
-		"OLLAMA_RUNNERS_DIR":       fmt.Sprintf("%v", RunnersDir),
-		"OLLAMA_TMPDIR":            fmt.Sprintf("%v", TmpDir),
-	}
-}
-
-var defaultAllowOrigins = []string{
-	"localhost",
-	"127.0.0.1",
-	"0.0.0.0",
-}
-
-// Clean quotes and spaces from the value
-func clean(key string) string {
-	return strings.Trim(os.Getenv(key), "\"' ")
-}
-
-func init() {
-	// default values
-	NumParallel = 1
-	MaxRunners = 1
-	MaxQueuedRequests = 512
-
-	LoadConfig()
-}
-
-func LoadConfig() {
-	if debug := clean("OLLAMA_DEBUG"); debug != "" {
-		d, err := strconv.ParseBool(debug)
-		if err == nil {
-			Debug = d
-		} else {
-			Debug = true
-		}
-	}
-
-	RunnersDir = clean("OLLAMA_RUNNERS_DIR")
-	if runtime.GOOS == "windows" && RunnersDir == "" {
-		// On Windows we do not carry the payloads inside the main executable
-		appExe, err := os.Executable()
-		if err != nil {
-			slog.Error("failed to lookup executable path", "error", err)
-		}
-
-		cwd, err := os.Getwd()
-		if err != nil {
-			slog.Error("failed to lookup working directory", "error", err)
-		}
-
-		var paths []string
-		for _, root := range []string{filepath.Dir(appExe), cwd} {
-			paths = append(paths,
-				filepath.Join(root),
-				filepath.Join(root, "windows-"+runtime.GOARCH),
-				filepath.Join(root, "dist", "windows-"+runtime.GOARCH),
-			)
-		}
-
-		// Try a few variations to improve developer experience when building from source in the local tree
-		for _, p := range paths {
-			candidate := filepath.Join(p, "ollama_runners")
-			_, err := os.Stat(candidate)
-			if err == nil {
-				RunnersDir = candidate
-				break
-			}
-		}
-		if RunnersDir == "" {
-			slog.Error("unable to locate llm runner directory.  Set OLLAMA_RUNNERS_DIR to the location of 'ollama_runners'")
-		}
-	}
-
-	TmpDir = clean("OLLAMA_TMPDIR")
-
-	userLimit := clean("OLLAMA_MAX_VRAM")
-	if userLimit != "" {
-		avail, err := strconv.ParseUint(userLimit, 10, 64)
-		if err != nil {
-			slog.Error("invalid setting, ignoring", "OLLAMA_MAX_VRAM", userLimit, "error", err)
-		} else {
-			MaxVRAM = avail
-		}
-	}
-
-	LLMLibrary = clean("OLLAMA_LLM_LIBRARY")
-
-	if onp := clean("OLLAMA_NUM_PARALLEL"); onp != "" {
-		val, err := strconv.Atoi(onp)
-		if err != nil || val <= 0 {
-			slog.Error("invalid setting must be greater than zero", "OLLAMA_NUM_PARALLEL", onp, "error", err)
-		} else {
-			NumParallel = val
-		}
-	}
-
-	if noprune := clean("OLLAMA_NOPRUNE"); noprune != "" {
-		NoPrune = true
-	}
-
-	if origins := clean("OLLAMA_ORIGINS"); origins != "" {
-		AllowOrigins = strings.Split(origins, ",")
-	}
-	for _, allowOrigin := range defaultAllowOrigins {
-		AllowOrigins = append(AllowOrigins,
-			fmt.Sprintf("http://%s", allowOrigin),
-			fmt.Sprintf("https://%s", allowOrigin),
-			fmt.Sprintf("http://%s:*", allowOrigin),
-			fmt.Sprintf("https://%s:*", allowOrigin),
-		)
-	}
-
-	maxRunners := clean("OLLAMA_MAX_LOADED_MODELS")
-	if maxRunners != "" {
-		m, err := strconv.Atoi(maxRunners)
-		if err != nil {
-			slog.Error("invalid setting", "OLLAMA_MAX_LOADED_MODELS", maxRunners, "error", err)
-		} else {
-			MaxRunners = m
-		}
-	}
-
-	if onp := os.Getenv("OLLAMA_MAX_QUEUE"); onp != "" {
-		p, err := strconv.Atoi(onp)
-		if err != nil || p <= 0 {
-			slog.Error("invalid setting", "OLLAMA_MAX_QUEUE", onp, "error", err)
-		} else {
-			MaxQueuedRequests = p
-		}
-	}
-}

server/envconfig/config_test.go

@@ -1,20 +0,0 @@
-package envconfig
-
-import (
-	"os"
-	"testing"
-
-	"github.com/stretchr/testify/require"
-)
-
-func TestConfig(t *testing.T) {
-	os.Setenv("OLLAMA_DEBUG", "")
-	LoadConfig()
-	require.False(t, Debug)
-	os.Setenv("OLLAMA_DEBUG", "false")
-	LoadConfig()
-	require.False(t, Debug)
-	os.Setenv("OLLAMA_DEBUG", "1")
-	LoadConfig()
-	require.True(t, Debug)
-}

server/layers.go

@@ -5,14 +5,39 @@ import (
 	"fmt"
 	"io"
 	"os"
+	"strings"
+
+	"golang.org/x/exp/slices"
 )
 
+type Layers struct {
+	items []*Layer
+}
+
+func (ls *Layers) Add(layer *Layer) {
+	if layer.Size > 0 {
+		ls.items = append(ls.items, layer)
+	}
+}
+
+func (ls *Layers) Replace(layer *Layer) {
+	if layer.Size > 0 {
+		mediatype := layer.MediaType
+		layers := slices.DeleteFunc(ls.items, func(l *Layer) bool {
+			return l.MediaType == mediatype
+		})
+
+		ls.items = append(layers, layer)
+	}
+}
+
 type Layer struct {
 	MediaType string `json:"mediaType"`
 	Digest    string `json:"digest"`
 	Size      int64  `json:"size"`
 	From      string `json:"from,omitempty"`
-	status    string
+
+	tempFileName string
 }
 
 func NewLayer(r io.Reader, mediatype string) (*Layer, error) {
@@ -21,12 +46,14 @@ func NewLayer(r io.Reader, mediatype string) (*Layer, error) {
 		return nil, err
 	}
 
-	temp, err := os.CreateTemp(blobs, "sha256-")
+	const delimiter = "-"
+
+	pattern := strings.Join([]string{"sha256", "*-partial"}, delimiter)
+	temp, err := os.CreateTemp(blobs, pattern)
 	if err != nil {
 		return nil, err
 	}
 	defer temp.Close()
-	defer os.Remove(temp.Name())
 
 	sha256sum := sha256.New()
 	n, err := io.Copy(io.MultiWriter(temp, sha256sum), r)
@@ -34,29 +61,11 @@ func NewLayer(r io.Reader, mediatype string) (*Layer, error) {
 		return nil, err
 	}
 
-	if err := temp.Close(); err != nil {
-		return nil, err
-	}
-
-	digest := fmt.Sprintf("sha256:%x", sha256sum.Sum(nil))
-	blob, err := GetBlobsPath(digest)
-	if err != nil {
-		return nil, err
-	}
-
-	status := "using existing layer"
-	if _, err := os.Stat(blob); err != nil {
-		status = "creating new layer"
-		if err := os.Rename(temp.Name(), blob); err != nil {
-			return nil, err
-		}
-	}
-
 	return &Layer{
-		MediaType: mediatype,
-		Digest:    digest,
-		Size:      n,
-		status:    fmt.Sprintf("%s %s", status, digest),
+		MediaType:    mediatype,
+		Digest:       fmt.Sprintf("sha256:%x", sha256sum.Sum(nil)),
+		Size:         n,
+		tempFileName: temp.Name(),
 	}, nil
 }
 
@@ -76,15 +85,21 @@ func NewLayerFromLayer(digest, mediatype, from string) (*Layer, error) {
 		Digest:    digest,
 		Size:      fi.Size(),
 		From:      from,
-		status:    fmt.Sprintf("using existing layer %s", digest),
 	}, nil
 }
 
-func (l *Layer) Open() (io.ReadCloser, error) {
+func (l *Layer) Commit() (bool, error) {
+	// always remove temp
+	defer os.Remove(l.tempFileName)
+
 	blob, err := GetBlobsPath(l.Digest)
 	if err != nil {
-		return nil, err
+		return false, err
 	}
 
-	return os.Open(blob)
+	if _, err := os.Stat(blob); err != nil {
+		return true, os.Rename(l.tempFileName, blob)
+	}
+
+	return false, nil
 }

server/manifest.go

@@ -1,79 +0,0 @@
-package server
-
-import (
-	"bytes"
-	"crypto/sha256"
-	"encoding/json"
-	"fmt"
-	"io"
-	"os"
-	"path/filepath"
-
-	"github.com/ollama/ollama/types/model"
-)
-
-type Manifest struct {
-	ManifestV2
-	Digest string `json:"-"`
-}
-
-func (m *Manifest) Size() (size int64) {
-	for _, layer := range append(m.Layers, m.Config) {
-		size += layer.Size
-	}
-
-	return
-}
-
-func ParseNamedManifest(name model.Name) (*Manifest, error) {
-	if !name.IsFullyQualified() {
-		return nil, model.Unqualified(name)
-	}
-
-	manifests, err := GetManifestPath()
-	if err != nil {
-		return nil, err
-	}
-
-	var manifest ManifestV2
-	manifestfile, err := os.Open(filepath.Join(manifests, name.Filepath()))
-	if err != nil {
-		return nil, err
-	}
-
-	sha256sum := sha256.New()
-	if err := json.NewDecoder(io.TeeReader(manifestfile, sha256sum)).Decode(&manifest); err != nil {
-		return nil, err
-	}
-
-	return &Manifest{
-		ManifestV2: manifest,
-		Digest:     fmt.Sprintf("%x", sha256sum.Sum(nil)),
-	}, nil
-}
-
-func WriteManifest(name string, config *Layer, layers []*Layer) error {
-	manifest := ManifestV2{
-		SchemaVersion: 2,
-		MediaType:     "application/vnd.docker.distribution.manifest.v2+json",
-		Config:        config,
-		Layers:        layers,
-	}
-
-	var b bytes.Buffer
-	if err := json.NewEncoder(&b).Encode(manifest); err != nil {
-		return err
-	}
-
-	modelpath := ParseModelPath(name)
-	manifestPath, err := modelpath.GetManifestPath()
-	if err != nil {
-		return err
-	}
-
-	if err := os.MkdirAll(filepath.Dir(manifestPath), 0o755); err != nil {
-		return err
-	}
-
-	return os.WriteFile(manifestPath, b.Bytes(), 0o644)
-}

server/manifests.go

@@ -0,0 +1,34 @@
+package server
+
+import (
+	"bytes"
+	"encoding/json"
+	"os"
+	"path/filepath"
+)
+
+func WriteManifest(name string, config *Layer, layers []*Layer) error {
+	manifest := ManifestV2{
+		SchemaVersion: 2,
+		MediaType:     "application/vnd.docker.distribution.manifest.v2+json",
+		Config:        config,
+		Layers:        layers,
+	}
+
+	var b bytes.Buffer
+	if err := json.NewEncoder(&b).Encode(manifest); err != nil {
+		return err
+	}
+
+	modelpath := ParseModelPath(name)
+	manifestPath, err := modelpath.GetManifestPath()
+	if err != nil {
+		return err
+	}
+
+	if err := os.MkdirAll(filepath.Dir(manifestPath), 0o755); err != nil {
+		return err
+	}
+
+	return os.WriteFile(manifestPath, b.Bytes(), 0o644)
+}

server/model.go

@@ -1,261 +0,0 @@
-package server
-
-import (
-	"archive/zip"
-	"bytes"
-	"context"
-	"errors"
-	"fmt"
-	"io"
-	"net/http"
-	"os"
-	"path/filepath"
-
-	"github.com/ollama/ollama/api"
-	"github.com/ollama/ollama/convert"
-	"github.com/ollama/ollama/llm"
-	"github.com/ollama/ollama/types/model"
-)
-
-type layerWithGGML struct {
-	*Layer
-	*llm.GGML
-}
-
-func parseFromModel(ctx context.Context, name model.Name, fn func(api.ProgressResponse)) (layers []*layerWithGGML, err error) {
-	modelpath := ParseModelPath(name.String())
-	manifest, _, err := GetManifest(modelpath)
-	switch {
-	case errors.Is(err, os.ErrNotExist):
-		if err := PullModel(ctx, name.String(), &registryOptions{}, fn); err != nil {
-			return nil, err
-		}
-
-		modelpath = ParseModelPath(name.String())
-		manifest, _, err = GetManifest(modelpath)
-		if err != nil {
-			return nil, err
-		}
-	case err != nil:
-		return nil, err
-	}
-
-	for _, layer := range manifest.Layers {
-		layer, err := NewLayerFromLayer(layer.Digest, layer.MediaType, modelpath.GetShortTagname())
-		if err != nil {
-			return nil, err
-		}
-
-		switch layer.MediaType {
-		case "application/vnd.ollama.image.model",
-			"application/vnd.ollama.image.projector",
-			"application/vnd.ollama.image.adapter":
-			blobpath, err := GetBlobsPath(layer.Digest)
-			if err != nil {
-				return nil, err
-			}
-
-			blob, err := os.Open(blobpath)
-			if err != nil {
-				return nil, err
-			}
-			defer blob.Close()
-
-			ggml, _, err := llm.DecodeGGML(blob)
-			if err != nil {
-				return nil, err
-			}
-
-			layers = append(layers, &layerWithGGML{layer, ggml})
-		default:
-			layers = append(layers, &layerWithGGML{layer, nil})
-		}
-
-	}
-
-	return layers, nil
-}
-
-func parseFromZipFile(_ context.Context, file *os.File, fn func(api.ProgressResponse)) (layers []*layerWithGGML, err error) {
-	stat, err := file.Stat()
-	if err != nil {
-		return nil, err
-	}
-
-	r, err := zip.NewReader(file, stat.Size())
-	if err != nil {
-		return nil, err
-	}
-
-	tempdir, err := os.MkdirTemp(filepath.Dir(file.Name()), "")
-	if err != nil {
-		return nil, err
-	}
-	defer os.RemoveAll(tempdir)
-
-	fn(api.ProgressResponse{Status: "unpacking model metadata"})
-	for _, f := range r.File {
-		// TODO(mxyng): this should not write out all files to disk
-		outfile, err := os.Create(filepath.Join(tempdir, f.Name))
-		if err != nil {
-			return nil, err
-		}
-		defer outfile.Close()
-
-		infile, err := f.Open()
-		if err != nil {
-			return nil, err
-		}
-		defer infile.Close()
-
-		if _, err = io.Copy(outfile, infile); err != nil {
-			return nil, err
-		}
-
-		if err := outfile.Close(); err != nil {
-			return nil, err
-		}
-
-		if err := infile.Close(); err != nil {
-			return nil, err
-		}
-	}
-
-	mf, err := convert.GetModelFormat(tempdir)
-	if err != nil {
-		return nil, err
-	}
-
-	params, err := mf.GetParams(tempdir)
-	if err != nil {
-		return nil, err
-	}
-
-	mArch, err := mf.GetModelArch("", tempdir, params)
-	if err != nil {
-		return nil, err
-	}
-
-	fn(api.ProgressResponse{Status: "processing tensors"})
-	if err := mArch.GetTensors(); err != nil {
-		return nil, err
-	}
-
-	if err := mArch.LoadVocab(); err != nil {
-		return nil, err
-	}
-
-	fn(api.ProgressResponse{Status: "converting model"})
-
-	// TODO(mxyng): this should write directly into a layer
-	// e.g. NewLayer(arch.Reader(), "application/vnd.ollama.image.model")
-	temp, err := os.CreateTemp(tempdir, "fp16")
-	if err != nil {
-		return nil, err
-	}
-	defer temp.Close()
-	defer os.Remove(temp.Name())
-
-	if err = mArch.WriteGGUF(temp); err != nil {
-		return nil, err
-	}
-
-	if _, err := temp.Seek(0, io.SeekStart); err != nil {
-		return nil, err
-	}
-
-	layer, err := NewLayer(temp, "application/vnd.ollama.image.model")
-	if err != nil {
-		return nil, fmt.Errorf("aaa: %w", err)
-	}
-
-	blobpath, err := GetBlobsPath(layer.Digest)
-	if err != nil {
-		return nil, err
-	}
-
-	bin, err := os.Open(blobpath)
-	if err != nil {
-		return nil, err
-	}
-	defer bin.Close()
-
-	ggml, _, err := llm.DecodeGGML(bin)
-	if err != nil {
-		return nil, err
-	}
-
-	layer, err = NewLayerFromLayer(layer.Digest, layer.MediaType, "")
-	if err != nil {
-		return nil, err
-	}
-
-	layers = append(layers, &layerWithGGML{layer, ggml})
-	return layers, nil
-}
-
-func parseFromFile(ctx context.Context, file *os.File, fn func(api.ProgressResponse)) (layers []*layerWithGGML, err error) {
-	sr := io.NewSectionReader(file, 0, 512)
-	contentType, err := detectContentType(sr)
-	if err != nil {
-		return nil, err
-	}
-
-	switch contentType {
-	case "gguf", "ggla":
-		// noop
-	case "application/zip":
-		return parseFromZipFile(ctx, file, fn)
-	default:
-		return nil, fmt.Errorf("unsupported content type: %s", contentType)
-	}
-
-	stat, err := file.Stat()
-	if err != nil {
-		return nil, err
-	}
-
-	var offset int64
-	for offset < stat.Size() {
-		ggml, n, err := llm.DecodeGGML(file)
-		if errors.Is(err, io.EOF) {
-			break
-		} else if err != nil {
-			return nil, err
-		}
-
-		mediatype := "application/vnd.ollama.image.model"
-		if ggml.Name() == "ggla" {
-			mediatype = "application/vnd.ollama.image.adapter"
-		} else if ggml.KV().Architecture() == "clip" {
-			mediatype = "application/vnd.ollama.image.projector"
-		}
-
-		layer, err := NewLayer(io.NewSectionReader(file, offset, n), mediatype)
-		if err != nil {
-			return nil, err
-		}
-
-		layers = append(layers, &layerWithGGML{layer, ggml})
-		offset = n
-	}
-
-	return layers, nil
-}
-
-func detectContentType(r io.Reader) (string, error) {
-	var b bytes.Buffer
-	if _, err := io.Copy(&b, r); err != nil {
-		return "", err
-	}
-
-	if contentType := llm.DetectGGMLType(b.Bytes()); contentType != "" {
-		return contentType, nil
-	}
-
-	if contentType := http.DetectContentType(b.Bytes()); contentType != "application/octet-stream" {
-		return contentType, nil
-	}
-
-	return "unknown", nil
-}