- https://www.nvidia.co.kr/Download/index.aspx <- conda
- https://developer.nvidia.com/cuda-11-7-0-download-archive <- cuda
- https://developer.nvidia.com/rdp/cudnn-archive <-cudnn
- https://www.anaconda.com/products/individual
- Sh 파일로 받았다면,
$./Anaconda3-2021.11-Linux-x86_64.sh
conda create -n pytorch python=3.9
conda install pytorch torchvision torchaudio pytorch-cuda=11.7 -c pytorch -c nvidia
conda install captum –c pyotorch
pip install wandb
- https://www.jetbrains.com/ko-kr/pycharm/download/, 학교 계정이 있고, 연구실 서버를 외부에서 이용할 계획이 있다면 Professional 버전 추
- https://cosmosproject.tistory.com/480, interpreter 설정 / professional 버전이라면 ssh 외부 interpreter 설정가능
conda install -c conda-forge tqdm
Pip install numpy
Pip install matplotlib