# Overview

Source: [https://docs.qualcomm.com/doc/80-70014-54/topic/overview.html](https://docs.qualcomm.com/doc/80-70014-54/topic/overview.html)

TensorFlow Lite is an open-source deep learning framework for on-device inference.
        The TensorFlow framework provides tools and APIs to convert a typical pretrained TensorFlow
        model from the SavedModel or Keras format to a TensorFlow Lite format.

This document describes the available delegates and ways to execute TensorFlow Lite
            models for QCS6490 and QCS5430 applications using the Qualcomm software stack. It
            explains how to:

- Run TensorFlow Lite models on the Qualcomm® RB3 Gen 2 Vision kit.
- Convert TensorFlow models to TensorFlow Lite models and optimize them for on-device
                inference purposes.
- Run TensorFlow Lite models using a delegate on a hardware accelerator, such as CPU,
                GPU, and Qualcomm® Hexagon™ Tensor Processor.
- Perform TensorFlow Lite model benchmarking.

Last Published: Jul 12, 2024

[Next Topic
Getting started](https://docs.qualcomm.com/bundle/publicresource/80-70014-54/topics/getting-started.md)