Add simple user interface

llama.log

@@ -1 +1 @@
-[1751004438] warming up the model with an empty run
+[1751008457] warming up the model with an empty run

pom.xml

@@ -1,46 +1,61 @@
 <?xml version="1.0" encoding="UTF-8"?>
-<project xmlns="http://maven.apache.org/POM/4.0.0" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
+<project xmlns="http://maven.apache.org/POM/4.0.0"
-	xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 https://maven.apache.org/xsd/maven-4.0.0.xsd">
+         xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
-	<modelVersion>4.0.0</modelVersion>
+         xsi:schemaLocation="http://maven.apache.org/POM/4.0.0
-	<parent>
+                             https://maven.apache.org/xsd/maven-4.0.0.xsd">
-		<groupId>org.springframework.boot</groupId>
+    <modelVersion>4.0.0</modelVersion>
-		<artifactId>spring-boot-starter-parent</artifactId>
-		<version>3.5.3</version>
-		<relativePath/> <!-- lookup parent from repository -->
-	</parent>
-	<groupId>com.pablotj</groupId>
-	<artifactId>ia-chat-boot</artifactId>
-	<version>0.0.1-SNAPSHOT</version>
-	<name>ia-chat-boot</name>
-	<description>Project IA chat boot</description>
 
-	<properties>
+    <parent>
-		<java.version>17</java.version>
+        <groupId>org.springframework.boot</groupId>
-	</properties>
+        <artifactId>spring-boot-starter-parent</artifactId>
+        <version>3.5.3</version>
+        <relativePath/>
+    </parent>
 
-	<dependencies>
+    <groupId>com.pablotj</groupId>
-		<dependency>
+    <artifactId>ia-chat-boot</artifactId>
-			<groupId>de.kherud</groupId>
+    <version>0.0.1-SNAPSHOT</version>
-			<artifactId>llama</artifactId>
+    <name>ia-chat-boot</name>
-			<version>3.4.1</version>
+    <description>Project IA chat boot</description>
-		</dependency>
-		<dependency>
-			<groupId>org.springframework.boot</groupId>
-			<artifactId>spring-boot-starter-web</artifactId>
-		</dependency>
-		<dependency>
-			<groupId>org.springframework.boot</groupId>
-			<artifactId>spring-boot-starter-test</artifactId>
-			<scope>test</scope>
-		</dependency>
-	</dependencies>
 
-	<build>
+    <properties>
-		<plugins>
+        <java.version>17</java.version>
-			<plugin>
+    </properties>
-				<groupId>org.springframework.boot</groupId>
+
-				<artifactId>spring-boot-maven-plugin</artifactId>
+    <dependencies>
-			</plugin>
+        <!-- Spring Boot Web -->
-		</plugins>
+        <dependency>
-	</build>
+            <groupId>org.springframework.boot</groupId>
+            <artifactId>spring-boot-starter-web</artifactId>
+        </dependency>
+
+        <!-- Thymeleaf template engine -->
+        <dependency>
+            <groupId>org.springframework.boot</groupId>
+            <artifactId>spring-boot-starter-thymeleaf</artifactId>
+        </dependency>
+
+        <!-- llama-java (IA offline) -->
+        <dependency>
+            <groupId>de.kherud</groupId>
+            <artifactId>llama</artifactId>
+            <version>3.4.1</version>
+        </dependency>
+
+        <!-- Test support -->
+        <dependency>
+            <groupId>org.springframework.boot</groupId>
+            <artifactId>spring-boot-starter-test</artifactId>
+            <scope>test</scope>
+        </dependency>
+    </dependencies>
+
+    <build>
+        <plugins>
+            <plugin>
+                <groupId>org.springframework.boot</groupId>
+                <artifactId>spring-boot-maven-plugin</artifactId>
+            </plugin>
+        </plugins>
+    </build>
 </project>

src/main/java/com/pablotj/ia/chat/boot/ChatController.java

@@ -1,21 +1,35 @@
 package com.pablotj.ia.chat.boot;
 
-import org.springframework.web.bind.annotation.GetMapping;
+import org.springframework.stereotype.Controller;
-import org.springframework.web.bind.annotation.RequestMapping;
+import org.springframework.ui.Model;
-import org.springframework.web.bind.annotation.RequestParam;
+import org.springframework.web.bind.annotation.*;
-import org.springframework.web.bind.annotation.RestController;
 
-@RestController
+import java.util.ArrayList;
-@RequestMapping("/chat")
+import java.util.List;
+
+@Controller
+@RequestMapping("/")
 public class ChatController {
+
+    private final List<ChatMessage> messages = new ArrayList<>();
     private final LlamaService llamaService;
 
     public ChatController(LlamaService llamaService) {
         this.llamaService = llamaService;
     }
 
-    @GetMapping
+    @GetMapping("/chat")
-    public String chat(@RequestParam String prompt) {
+    public String showChat(Model model) {
-        return llamaService.chat(prompt);
+        model.addAttribute("messages", messages);
+        return "chat";
+    }
+
+    @PostMapping("/chat")
+    public String handleChat(@RequestParam("prompt") String prompt, Model model) {
+        messages.add(new ChatMessage("user", prompt));
+        String reply = llamaService.chat(prompt);
+        messages.add(new ChatMessage("bot", reply));
+        model.addAttribute("messages", messages);
+        return "chat";
     }
 }

src/main/java/com/pablotj/ia/chat/boot/ChatMessage.java

@@ -0,0 +1,19 @@
+package com.pablotj.ia.chat.boot;
+
+public class ChatMessage {
+    private String role; // "user" o "bot"
+    private String text;
+
+    public ChatMessage(String role, String text) {
+        this.role = role;
+        this.text = text;
+    }
+
+    public String getRole() {
+        return role;
+    }
+
+    public String getText() {
+        return text;
+    }
+}

src/main/java/com/pablotj/ia/chat/boot/LlamaService.java

@@ -15,7 +15,14 @@ public class LlamaService implements AutoCloseable {
     public void init() {
         try {
             ModelParameters params = new ModelParameters()
-                    .setModelFilePath("models/openchat-3.5-0106.Q4_K_M.gguf");
+                    .setModelFilePath("models/ggml-model-q3_k.gguf")
+                    .setSeed(42)
+                    .setNThreads(8)             // usa 8 hilos CPU (ajusta según tu CPU)
+                    .setNGpuLayers(0)           // no usar GPU
+                    .setMainGpu(-1)             // deshabilitar GPU principal
+                    .setNoKvOffload(true)       // no descargar KV, evitar errores GPU
+                    .setUseMmap(true)           // mejorar gestión memoria
+                    .setNPredict(128);
             model = new LlamaModel(params);
         } catch (Exception e) {
             throw new RuntimeException("Error cargando el modelo", e);
@@ -33,9 +40,11 @@ public class LlamaService implements AutoCloseable {
         String finalPrompt = chat.build();
 
         InferenceParameters inf = new InferenceParameters(finalPrompt)
-                .setTemperature(0.7f)
+            .setNPredict(200)
-                .setTopP(0.9f)
+            .setTemperature(0.7f)
-                .setTopK(40);
+            .setTopP(0.9f)
+            .setTopK(40)
+            .setUseChatTemplate(true);
 
 
         StringBuilder sb = new StringBuilder();

src/main/java/com/pablotj/ia/chat/boot/PromptBuilder.java

@@ -13,21 +13,23 @@ public class PromptBuilder {
     }
 
     public void user(String message) {
-        turns.add("GPT4 Correct User: " + message + "<|end_of_turn|>");
+        turns.add("<|im_start|>user\n" + message + "\n<|im_end|>");
     }
 
     public void assistant(String message) {
-        turns.add("GPT4 Correct Assistant: " + message + "<|end_of_turn|>");
+        turns.add("<|im_start|>assistant\n" + message + "\n<|im_end|>");
     }
 
     public String build() {
         StringBuilder sb = new StringBuilder();
-        sb.append(systemPrompt).append("<|end_of_turn|>\n");
+        sb.append("<|im_start|>system\n")
+          .append(systemPrompt)
+          .append("\n<|im_end|>\n");
         for (String turn : turns) {
             sb.append(turn).append("\n");
         }
-        // Deja el último turno preparado para que el modelo continúe como "Assistant"
+        // Deja listo para que el modelo continúe como assistant generando respuesta:
-        sb.append("GPT4 Correct Assistant: ");
+        sb.append("<|im_start|>assistant\n");
         return sb.toString();
     }
 }

src/main/resources/templates/chat.html

@@ -0,0 +1,28 @@
+<!DOCTYPE html>
+<html xmlns:th="http://www.thymeleaf.org" lang="es">
+<head>
+    <meta charset="UTF-8">
+    <title>Chat con IA Offline</title>
+    <style>
+        body { font-family: sans-serif; max-width: 800px; margin: 2rem auto; }
+        .bubble { margin-bottom: 1rem; }
+        .user { color: #003366; font-weight: bold; }
+        .bot { color: #006600; }
+        textarea { width: 100%; height: 4em; margin-top: 1rem; }
+        button { padding: 0.5rem 1rem; margin-top: 0.5rem; }
+    </style>
+</head>
+<body>
+    <h1>Chat con IA Offline</h1>
+
+    <div th:each="msg : ${messages}" class="bubble">
+        <div th:if="${msg.role} == 'user'" class="user">🧑‍💻: <span th:text="${msg.text}"></span></div>
+        <div th:if="${msg.role} == 'bot'" class="bot">🤖: <span th:text="${msg.text}"></span></div>
+    </div>
+
+    <form method="post" th:action="@{/chat}">
+        <textarea name="prompt" placeholder="Escribe tu mensaje aquí..."></textarea><br>
+        <button type="submit">Enviar</button>
+    </form>
+</body>
+</html>