dc13eb7049
* 对RKNPU2后端进行修改，当模型为非量化模型时，不在NPU执行normalize操作，当模型为量化模型时，在NUP上执行normalize操作 * 更新RKNPU2框架，输出数据的数据类型统一返回fp32类型 * 更新scrfd，拆分disable_normalize和disable_permute * 更新scrfd代码，支持量化 * 更新scrfd python example代码 * 更新模型转换代码，支持量化模型 * 更新文档 * 按照要求修改 * 按照要求修改 * 修正模型转换文档 * 更新一下转换脚本
FastDeploy Toolkit
FastDeploy provides a series of efficient and easy-to-use tools to optimize the deployment experience and improve inference performance.
One-Click Model Auto Compression Tool
Based on PaddleSlim's Auto Compression Toolkit (ACT), FastDeploy provides users with a one-click model automation compression tool that allows users to easily compress the model with a single command. This document will take FastDeploy's one-click model automation compression tool as an example, introduce how to install the tool, and provide the corresponding documentation for usage.
Environmental Preparation
1.Install PaddlePaddle 2.4 version
https://www.paddlepaddle.org.cn/install/quick?docurl=/documentation/docs/zh/develop/install/pip/linux-pip.html
2.Install PaddleSlim 2.4 version
pip install paddleslim==2.4.0
3.Install fastdeploy-tools package
# Installing fastdeploy-tools via pip
# This tool is included in the python installer of FastDeploy, so you don't need to install it again.
pip install fastdeploy-tools==0.0.1
The Usage of One-Click Model Auto Compression Tool
After the above steps are successfully installed, you can use FastDeploy one-click model automation compression tool, as shown in the following example.
fastdeploy compress --config_path=./configs/detection/yolov5s_quant.yaml --method='PTQ' --save_dir='./yolov5s_ptq_model/'
For detailed documentation, please refer to FastDeploy One-Click Model Auto Compression Tool
Model Conversion Tool
Based on X2Paddle, FastDeploy provides users with a model conversion tool. Users can easily migrate external framework models to the Paddle framework with one line of commands. Currently, ONNX, TensorFlow and Caffe are supported, and most mainstream CV and NLP model conversions are supported.
Environmental Preparation
- Install PaddlePaddle, refer to the following documents for quick installation
https://www.paddlepaddle.org.cn/install/quick?docurl=/documentation/docs/zh/develop/install/pip/linux-pip.html
- Install X2Paddle
To use the stable version, install X2Paddle via pip:
pip install x2paddle
To experience the latest features, you can use the source installation method:
git clone https://github.com/PaddlePaddle/X2Paddle.git
cd X2Paddle
python setup.py install
How to use
After successful installation according to the above steps, you can use the FastDeploy one-click conversion tool. The example is as follows:
fastdeploy convert --framework onnx --model yolov5s.onnx --save_dir pd_model
For more details, please refer toX2Paddle