embedding and rag

Author: netdur

Diff for: example/apple_pie.dart

@@ -18,7 +18,7 @@ void main() async {
     samplerParams.penaltyRepeat = 1.1;
 
     Llama.libraryPath = "bin/MAC_ARM64/libllama.dylib";
-    String modelPath = "/Users/adel/Downloads/gemma-3-12b-it-Q4_K_M.gguf";
+    String modelPath = "/Users/adel/Workspace/gguf/gemma-3-12b-it-Q4_K_M.gguf";
     Llama llama = Llama(modelPath, ModelParams(), contextParams, samplerParams);
 
     llama.setPrompt(

Diff for: example/apple_pie.qwq.dart

@@ -0,0 +1,38 @@
+// ignore_for_file: avoid_print
+
+import 'dart:io';
+
+import 'package:llama_cpp_dart/llama_cpp_dart.dart';
+
+void main() async {
+  try {
+    ContextParams contextParams = ContextParams();
+    contextParams.nPredict = 512;
+    contextParams.nCtx = 512;
+    contextParams.nBatch = 512;
+
+    final samplerParams = SamplerParams();
+    samplerParams.temp = 0.6;
+    samplerParams.minP = 0;
+    samplerParams.topK = 20;
+    samplerParams.topP = 0.95;
+    // samplerParams.penaltyRepeat = 1.1;
+
+    Llama.libraryPath = "bin/MAC_ARM64/libllama.dylib";
+    String modelPath = "/Users/adel/Workspace/gguf/qwq-32b-q4_k_m.gguf";
+    Llama llama =
+        Llama(modelPath, ModelParams(), contextParams, samplerParams, true);
+
+    llama.setPrompt("apple pie recipe?");
+    while (true) {
+      var (token, done) = llama.getNext();
+      stdout.write(token);
+      if (done) break;
+    }
+    stdout.write("\n");
+
+    llama.dispose();
+  } catch (e) {
+    print("Error: ${e.toString()}");
+  }
+}

Diff for: example/chat_cli.dart

@@ -3,7 +3,6 @@
 import 'dart:io';
 
 import 'package:llama_cpp_dart/llama_cpp_dart.dart';
-import 'package:llama_cpp_dart/src/chat.dart';
 
 void main() async {
   try {
@@ -25,7 +24,7 @@ void main() async {
     // Load the LLM model
     print("Loading model, please wait...");
     Llama.libraryPath = "bin/MAC_ARM64/libllama.dylib";
-    String modelPath = "/Users/adel/Downloads/gemma-3-12b-it-Q4_K_M.gguf";
+    String modelPath = "/Users/adel/Workspace/gguf/gemma-3-12b-it-Q4_K_M.gguf";
     Llama llama =
         Llama(modelPath, ModelParams(), contextParams, samplerParams, false);
     print("Model loaded successfully! ${llama.status}");

Diff for: example/chat_cli_isolated.dart

@@ -4,7 +4,6 @@ import 'dart:io';
 import 'dart:async';
 
 import 'package:llama_cpp_dart/llama_cpp_dart.dart';
-import 'package:llama_cpp_dart/src/chat.dart';
 
 void main() async {
   print("Starting LLM CLI Chat App with Isolates...");
@@ -26,7 +25,7 @@ void main() async {
 
   // Initialize load command for the isolate
   final loadCommand = LlamaLoad(
-    path: "/Users/adel/Downloads/gemma-3-12b-it-Q4_K_M.gguf",
+    path: "/Users/adel/Workspace/gguf/gemma-3-12b-it-Q4_K_M.gguf",
     modelParams: ModelParams(),
     contextParams: contextParams,
     samplingParams: samplerParams,