Environment Setup

January 8, 2026 ยท View on GitHub

This project uses CUDA 11.8 and recommends venv for environment management.

a. Create a virtual environment

cd SparseWorld
python3.9 -m venv env
source env/bin/activate

Conda is also completely feasible, with only the difference being in the creation of the environment.

b. Download and compile some core packages

pip install torch==2.0.1 torchvision==0.15.2 torchaudio==2.0.2 --index-url https://download.pytorch.org/whl/cu118
pip install -v -e .
pip install -r requirements.txt
cd mmdet3d/models/sparsedetectors/csrc
python setup.py build_ext --inplace