Installation guide

August 11, 2022 ยท View on GitHub

Requirements

  • Python 3.7.0
  • Pytorch 1.7.1
  • CUDA 10.2

1) Clone this repository

git clone https://github.com/VinAIResearch/GeoFormer
cd GeoFormer

2) Install pytorch (version 1.7.1), cudatoolkit (version 10.2) and other dependencies

conda install pytorch==1.7.1 cudatoolkit=10.2 -c pytorch

conda install -c bioconda google-sparsehash 

pip install -r requirements.txt

We do not recommend to use newer version of Pytorch due to the lack of THC library.

3) For the SparseConv, we use spconv1.0 from PointGroup

  • To compile spconv, firstly install the dependent libraries.
conda install libboost
conda install -c daleydeng gcc-5 # need gcc-5.4 for sparseconv

Add the $INCLUDE_PATH$ that contains boost in lib/spconv/CMakeLists.txt. (Not necessary if it could be found.)

include_directories($INCLUDE_PATH$)
  • Clone and compile the spconv library.
cd lib/
git clone https://github.com/llijiang/spconv.git --recursive
cd spconv/
python setup.py bdist_wheel
  • Run cd dist and pip install the generated .whl file.

Currently, there are some bugs with spconv2.0. We are planning to refactor and optimize our model to run with spconv2.0.

4) Compile the pointgroup_ops library.

cd lib/pointgroup_ops
python setup.py develop

If any header files could not be found, run the following commands.

python setup.py build_ext --include-dirs=$INCLUDE_PATH$
python setup.py develop

$INCLUDE_PATH$ is the path to the folder containing the header files that could not be found.

5) Compile the pointnet2 library.

cd lib/pointnet2
python setup.py install

6) Install FAISS:

conda install -c faiss-gpu cudatoolkit=10.2 # for CUDA 10.2