createonnxmodel

a tool to convert mmlab model to onnx

requirement

• python >= 3.7

build createOnnx

git clone https://github.com.cnpmjs.org/tangyanf/createonnxmodel.git
cd createonnxmodel
python setup.py install

before use we need install mmcv-full

• mmcv-full

  to install mmcv-full with custom ops, we need to download onnxruntime library, And add it into environment

  1. download onnxruntime-1.5.1

  wget https://github.com.cnpmjs.org/microsoft/onnxruntime/releases/download/v1.5.1/onnxruntime-linux-x64-1.5.1.tgz
  tar -zxvf onnxruntime-linux-x64-1.5.1.tgz

  2. add environment variable

  cd onnxruntime-linux-x64-1.5.1
  export ONNXRUNTIME_DIR=$(pwd)
  export LD_LIBRARY_PATH=$ONNXRUNTIME_DIR/lib:$LD_LIBRARY_PATH

  3. build on linux

  git clone https://github.com.cnpmjs.org/open-mmlab/mmcv.git
  cd mmcv
  MMCV_WITH_OPS=1 MMCV_WITH_ORT=1 pip install -e .

usage

createOnnx --class-name detection \
           --config xxx.py \
           --checkpoint xxx.pth\
           --dynamic-shape \
           --onnx-name xxx.onnx \
           --simplify \
           --save-input \
           --save-output \
           --veirfy

Parameter Description

• --class-name:         which class to convert(detection/classification/segmentation/editin)
• --config:                  config file
• --checkpoint:          checkpoint file
• --dynamic-shape:   whether to export onnx with dynamic shape
• --onnx-name:         output onnx model file's name
• --simplify:               whether to simplify onnx model
• --save-input:           whether to save model's input
• --save-output:         whether to save onnxruntime's output
• --verify :                  whether compare the outputs between Pytorch and ONNX