Welcome to acuitylite’s documentation!

Acuitylite is an end-to-end neural-network deployment tool for embedded systems. Acuitylite support converting caffe/darknet/onnx/tensorflow/tflite model to TIM-VX/TFLite case. In addition, Acuitylite support asysmetric uint8 and sysmetric int8 quantization.

Attention: We have introduced some important changes and updated the APIs that are not compatible with the version before Acuitylite6.20.0(include). Please read the document and demos carefully.

Indices and tables

• Index

• Search Page

Guide

• A brief guide to Acuitylite
  • System Requirement
  • Install
  • Document
  • Framework Support
  • How to generate nbg and TIM-VX case
  • How to run TIM-VX case
  • How to generate nbg by Ovxlib
  • Support

API Reference

• API
  • Importer
  • Quantization
  • Inference
  • Exporter

Demo

• demo.caffe
• demo.daraknet
• demo.onnx
• demo.tensorflow
• demo.tflite

Support

create issue on github or email to ML_Support@verisilicon.com