Still unfinished

Implement VectorNet using the NuPlan Dataset

This guide outlines the steps to implement VectorNet using the NuPlan dataset.

Requirements

In addition to the dependencies required by the NuPlan-devkit, the only additional library needed is pyG. It is recommended to use torch_geometric version 2.0.3.

Installation Steps:

1. Download the .whl files for pyG from this link.
2. Choose the files tagged with cp39-linux and the appropriate version for your setup.
3. Install the downloaded .whl files using pip:

   pip install <path_to_downloaded_file>.whl

4. Install torch_geometric:

   pip install torch_geometric==2.0.3

Workflow

Step 1: Process Data for Training

Run the following command to prepare the data for training:

python3 process_data.py

Step 2: Train VectorNet

To start training VectorNet, execute:

python3 train.py

Step 3: Visualize the Prediction Result

Follow the instructions in visual.ipynb to visualize the prediction results. The visualization notebook allows you to see the outcomes interactively within a Jupyter environment.

Additional Information

• Configuration settings can be modified in utils/config.py.
• Example of a prediction result: