eIQ TensorFlow Lite for Micro Library User Guide

Table of Contents

• Overview
• Deployment
• Example applications
• Model Conversion to TensorFlow Lite Format
• Model Conversion for NXP eIQ Neutron NPU
• Example: Converting a Quantized TensorFlow Lite Model for Neutron NPU
• Run and debug eIQ HiFi4 and HiFi1 DSP examples using Xplorer IDE
  • Prepare CM33 Core for the examples
  • Prepare DSP core for the examples
• Running an inference
  • Converting a model to a C language header file {#EXAMPLE_4 .section}
  • NPU inference {#npu_infer .section}
  • Adjusting the tensor arena size {#adjust_arena .section}
• Code size optimization
  • Register only used operators in TensorFlow Lite Micro {#SECTION_SS1_DJQ_QPB .section}
• Note about the source code in the document