Install DriveVLA

1. Install LLaVA

conda create -n drivevla python=3.10 -y
conda activate drivevla
pip install --upgrade pip  # Enable PEP 660 support.
pip install torch==2.1.2
# pip install -e ".[train]"

2. Install mmcv-full from source code

(1) GCC: Make sure gcc>=5 in conda env

# If gcc is not installed:
# conda install -c omgarcia gcc-6 # gcc-6.2

export PATH=YOUR_GCC_PATH/bin:$PATH
# Eg: export PATH=/mnt/gcc-5.4/bin:$PATH

(2) CUDA: Before installing MMCV family, you need to set up the CUDA_HOME (for compiling some operators on the gpu)

export CUDA_HOME=YOUR_CUDA_PATH/
# Eg: export CUDA_HOME=/mnt/cuda-11.1/

(3) Install mmcv-full from source code

cd DriveVLA # cd to the root of this repo
cd third_party/mmcv_1_7_2
pip install -r requirements/optional.txt
MMCV_WITH_OPS=1 pip install .

3. Install mmdet and mmseg

pip install mmdet==2.26.0 mmsegmentation==0.29.1 mmengine==0.9.0 motmetrics==1.4.0 casadi==3.6.0

4. Install mmdet3d from source code

cd DriveVLA # cd to the root of this repo
cd third_party/mmdetection3d_1_0_0rc6
pip install scipy==1.10.1 scikit-image==0.19.3 fsspec
pip install .

5. Troubleshoot

If you encounter the error: ImportError: libGL.so.1: cannot open shared object file: No such file or directory

apt-get update && apt-get install libgl1

If you encounter the error: Error code: libgfortran.so.5: cannot open shared object file: No such file or directory

sudo apt-get install libgfortran5
# or
conda install libgfortran