The TensorFlow Lite Support Library. Install the pip package: pip install tflite-support This package provides two major features: - Metadata writers: add metadata to TensorFlow Lite models. - Task Library: run TensorFlow Lite models of major machine learning tasks. To learn more about metadata, flatbuffers and TensorFlow Lite models, check out the [metadata section](https://www.tensorflow.org/lite/convert/metadata) of the TensorFlow Lite guide. To learn more about Task Library, check out the [documentation](https://www.tensorflow.org/lite/inference_with_metadata/task_library/overview) on the TensorFlow Lite website. ## Modules [`metadata`](https://developers.google.com/edge/api/tflite/python/tflite_support/metadata) module: TensorFlow Lite metadata tools. [`metadata_schema_py_generated`](https://developers.google.com/edge/api/tflite/python/tflite_support/metadata_schema_py_generated) module [`metadata_writers`](https://developers.google.com/edge/api/tflite/python/tflite_support/metadata_writers) module: TF Lite Metadata Writer API. [`task`](https://developers.google.com/edge/api/tflite/python/tflite_support/task) module: The TensorFlow Lite Task Library.