VSCODE Devcontainer for exporting bmodel from onnx models

This devcontainer is already configured for TPU-MLIR with all necessary dependencies to help you easily export bmodel files from onnx models.

Setup

1. Pull the docker image:

   docker pull sophgo/tpuc_dev:v3.4

   • If the pulling fails, you can download the required image file from the official website development materials here

2. Open this folder in VSCode as remote container:

   • Install the "Remote - Containers" extension in VSCode if you haven't already.
   • Open the command palette (Ctrl+Shift+P) and select "Remote-Containers: Open Folder in Container..."
   • Choose this folder to open it in the devcontainer.

Usage

• Once the devcontainer is set up and running, you can use the provided export_bmodel.py script and config.yaml in /workspace/model_transformer directory to convert your onnx model to bmodel.

Example:

Export bmodel from yolov5s.onnx

• Create a directory for the model and download the onnx file along with the required dataset and images:

cd /workspace
mkdir -p models/yolov5s && cd models/yolov5s
wget https://github.com/ultralytics/yolov5/releases/download/v6.0/yolov5s.onnx
cp -rf ../../tpu_mlir_resource/dataset/COCO2017 .
cp -rf ../../tpu_mlir_resource/image .

• modify the config.yaml file as per your requirements.
• Run the export script:

cd /workspace/model_transformer
python3 export_bmodel.py --config config.yaml

References:

• TPU-MLIR GitHub Repository
• TPU-MLIR Wiki
• Sophgo Developer Website

Developer Information

• Maintainer: W.M. Nipun Dhananjaya
• Email: nipun.dhananjaya@gmail.com