Page-1

# 架构 Source: [https://docs.qualcomm.com/doc/80-70014-54Y/topic/arch.html](https://docs.qualcomm.com/doc/80-70014-54Y/topic/arch.html) TensorFlow Lite 框架通过对延迟、模型大小和功耗进行优化，在移动、嵌入式和边缘平台等低功耗需求的设备上运行模型。该框架借助于 delegate 来运行模型。Delegate 是一些软件层，它们使用特定硬件上高效执行神经网络模型的库。 Figure : TensorFlow Lite Runtime 架构 - **[TensorFlow Lite Runtime](https://docs.qualcomm.com/doc/80-70014-54Y/topic/tensorflow-lite-runtime.html)** TensorFlow Lite 设备推理将模型加载到解析器中，解析器解析模型并使用 delegate 来运行它。 - **[Delegate](https://docs.qualcomm.com/doc/80-70014-54Y/topic/delegates.html)** Delegate 帮助您将 TensorFlow Lite 图处理分配到 CPU、GPU 和 Hexagon Tensor Processor 硬件加速器上。 Last Published: Aug 06, 2024 [Previous Topic 使用原生 TensorFlow Lite 示例程序运行 TensorFlow Lite 模型](https://docs.qualcomm.com/bundle/publicresource/80-70014-54Y/topics/run-a-tensorflow-lite-model-using-a-native-tensorflow-lite-sample-application.md) [Next Topic TensorFlow Lite Runtime](https://docs.qualcomm.com/bundle/publicresource/80-70014-54Y/topics/tensorflow-lite-runtime.md)