flutter_gemma 0.1.4

The plugin allows running the Gemma AI model locally on a device from a Flutter application.

Flutter Gemma

Gemma is a family of lightweight, state-of-the art open models built from the same research and technology used to create the Gemini models

Bring the power of Google's lightweight Gemma language models directly to your Flutter applications. With Flutter Gemma, you can seamlessly incorporate advanced AI capabilities into your iOS and Android apps, all without relying on external servers.

There is an example of using:

Features

• Local Execution: Run Gemma models directly on user devices for enhanced privacy and offline functionality.
• Platform Support: Compatible with both iOS and Android platforms.
• Ease of Use: Simple interface for integrating Gemma models into your Flutter projects.

Installation

1. Add flutter_gemma to your pubspec.yaml:

   dependencies:
     flutter_gemma: latest_version
   

2. Run flutter pub get to install.

Setup

1. Download Model: Obtain a pre-trained Gemma model (recommended: 2b or 2b-it) from Kaggle
   • Optionally, fine-tune a model for your specific use case
2. Rename Model: Rename the downloaded file to model.bin.
3. Integrate Model into Your App:

iOS

• Enable file sharing in info.plist:

<key>UIFileSharingEnabled</key>
<true/>

• Change the linking type of pods to static, replace use_frameworks! in Podfile with use_frameworks! :linkage => :static
• Transfer model.bin to your device
  1. Connect your iPhone
  2. Open Finder, your iPhone should appear in the Finder's sidebar under "Locations." Click on it.
  3. Access Files. In the button bar, click on "Files" to see apps that can transfer files between your iPhone and Mac.
  4. Drag and Drop or Add Files. You can drag model.bin directly to an app under the "Files" section to transfer them. Alternatively, click the "Add" button to browse and select model.bin to upload.

Android

• Transfer model.bin to your device (for testing purposes only, uploading by network will be implemented in next versions)
  1. Install adb tool, if you didn't install it before
  2. Connect your Android device
  3. Copy model.bin to the output_path folder
  4. Push the content of the output_path folder to the Android device

 adb shell rm -r /data/local/tmp/llm/ # Remove any previously loaded models
 adb shell mkdir -p /data/local/tmp/llm/
 adb push output_path /data/local/tmp/llm/model.bin

• If you want to use a GPU to work with the model, you need to add OpenCL support in the manifest.xml. If you plan to use only the CPU, you can skip this step.
  1. Add to 'AndroidManifest.xml' above tag </application>

 <uses-native-library
     android:name="libOpenCL.so"
     android:required="false"/>
 <uses-native-library android:name="libOpenCL-car.so" android:required="false"/>
 <uses-native-library android:name="libOpenCL-pixel.so" android:required="false"/>

Web

• Add dependencies to index.html file in web folder

  <script type="module">
  import { FilesetResolver, LlmInference } from 'https://cdn.jsdelivr.net/npm/@mediapipe/tasks-genai';
  window.FilesetResolver = FilesetResolver;
  window.LlmInference = LlmInference;
  </script>

• Copy model.bin to your web folder

Usage

1. Initialize:

void main() async {
  WidgetsFlutterBinding.ensureInitialized();
  await FlutterGemmaPlugin.instance.init(
    maxTokens: 512,  /// maxTokens is optional, by default the value is 1024
    temperature: 1.0,   /// temperature is optional, by default the value is 1.0
    topK: 1,   /// topK is optional, by default the value is 1
    randomSeed: 1,   /// randomSeed is optional, by default the value is 1
  );  
  
  runApp(const MyApp());
}

2. Generate response

final flutterGemma = FlutterGemmaPlugin.instance;
String response = await flutterGemma.getResponse(prompt: 'Tell me something interesting');
print(response);

2. Generate response as a stream

final flutterGemma = FlutterGemmaPlugin.instance;
flutterGemma.getAsyncResponse(prompt: 'Tell me something interesting').listen((String? token) => print(token));

Important Considerations

• Currently, models must be manually transferred to devices for testing. Network download functionality will be included in future versions.
• Larger models (like 7b and 7b-it) may be too resource-intensive for on-device use.

Coming Soon

• Network-based model using/downloading for seamless updates.