Hailo Model Conversion - up-division/up-ai GitHub Wiki

Linux

  1. Set up a python virtual environment 
    python3 -m venv <env_name>
  2. Activate virtual environment 
    source <env_name>/bin/activate
  3. Install Hailo Dataflow Compiler and HailoRT-Python package 
    pip install hailort-4.21.0-cp310-cp310-linux_x86_64.whl hailo_dataflow_compiler-3.31.0-py3-none-linux_x86_64.whl
  1. Clone the Hailo Model Zoo 
    git clone https://github.com/hailo-ai/hailo_model_zoo.git
  2. Run the setup script 
    cd hailo_model_zoo; pip install -e .
  1. Prepare calibration data 
    python3 hailo_model_zoo/datasets/create_coco_tfrecord.py calib2017
  1. Parse the model
    For Hailo-8L, use --hw-arch hailo8l 
    hailomz parse --hw-arch hailo8 --ckpt ./<model_name>.onnx <filename>
  1. Compile the model 
    hailomz compile <filename> --hw-arch hailo8 --har ./<model.name>.har --calib-path ~/.hailomz/data/models_files/coco/2023-08-03/coco_calib2017.tfrecord
⚠️ **GitHub.com Fallback** ⚠️