Skip to content

This project is ready for custom deploy, its extracted from tflite package documentation and examples but it's put all together specific for image classification models with quantization to float16.

Notifications You must be signed in to change notification settings

MOo207/starter_tflite_classification_model

Repository files navigation

starter_tflite_classification_model

This project is an example of integration between tflite model and flutter mobile application framework. Resourcecs:

  1. tflite model maker
  2. Official package for tflite
  3. My answer on some issues related to tflite package

How to use this project?

  1. Clone it
git clone https://github.com/MOo207/starter_tflite_classification_model.git
  1. Get dependencies
flutter pub get
  1. When you export your model, jsut copy it in assets/model and update pubspec if required
  2. Change tf_model_config.dart but make sure that your model is quantized to float16
class TFModelConfig {
  // TODO: change the model here
  static const assetsModel = "assets/model/model.tflite";
  // TODO: Change the labels here
  static const assetsLabels = "assets/model/labels.txt";
  // TODO: change your assets also
}
  1. Run the project
flutter run
  1. You will have all use cases for using tflite with flutter (run on image, run on frame)

About

This project is ready for custom deploy, its extracted from tflite package documentation and examples but it's put all together specific for image classification models with quantization to float16.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published