Installation¶
The installation goes as follows.
Required: Conda installation¶
wget https://repo.anaconda.com/archive/Anaconda3-2023.03-Linux-x86_64.sh
Then execute the installer with bash filename.sh
Install environment¶
Remark: To avoid No space left on device problem with conda or pip, set the temporary path first
to point to a path where there is enough space, for example:
mkdir $HOME/tmp
export TMPDIR=$HOME/tmp
Then execute:
git clone git@github.com:mieskolainen/hypertrack.git && cd hypertrack
# Create the environment
conda env create -f environment.yml
conda activate hypertrack
# Install pip packages
pip install -r requirements.txt && pip install -r requirements-aux.txt
Initialize environment¶
Always start with
conda activate hypertrack
source setenv.sh
GPU-support commands¶
Show the graphics card status
nvidia-smi
Show CUDA-compiler tools status
nvcc --version
Show Pytorch GPU support in Python
import torch
torch.cuda.is_available()
print(torch.cuda.get_device_name(0))
Conda virtual environment commands¶
conda activate hypertrack
...[install dependencies with pip, do your work]...
conda deactivate
conda info --envs
conda list --name hypertrack
# Remove environment completely
conda env remove --name hypertrack
C-library versions¶
ldd --version