最新のTensorflowが対応しているバージョンに合わせてCUDA11.8とcuDNN8.7を入れる
https://www.tensorflow.org/install/source?hl=pl#gpu
ここを参考に
cuda_11.8_installation_on_Ubuntu_22.04
https://gist.github.com/MihailCosmin/affa6b1b71b43787e9228c25fe15aeba
以下で/usr/local/cuda-11.8/lib64/は/usr/local/cuda-11.8/libnvvp/に変更する必要がある
# installing CUDA-11.8 sudo apt install cuda-11-8 -y # setup your paths echo 'export PATH=/usr/local/cuda-11.8/bin:$PATH' >> ~/.bashrc echo 'export LD_LIBRARY_PATH=/usr/local/cuda-11.8/lib64:$LD_LIBRARY_PATH' >> ~/.bashrc source ~/.bashrc sudo ldconfig # install cuDNN v11.8 # First register here: https://developer.nvidia.com/developer-program/signup CUDNN_TAR_FILE="cudnn-linux-x86_64-8.7.0.84_cuda11-archive.tar.xz" sudo wget https://developer.download.nvidia.com/compute/redist/cudnn/v8.7.0/local_installers/11.8/cudnn-linux-x86_64-8.7.0.84_cuda11-archive.tar.xz sudo tar -xvf ${CUDNN_TAR_FILE} sudo mv cudnn-linux-x86_64-8.7.0.84_cuda11-archive cuda # copy the following files into the cuda toolkit directory. sudo cp -P cuda/include/cudnn.h /usr/local/cuda-11.8/include sudo cp -P cuda/lib/libcudnn* /usr/local/cuda-11.8/lib64/ sudo chmod a+r /usr/local/cuda-11.8/lib64/libcudnn* # Finally, to verify the installation, check nvidia-smi nvcc -V
nvccでエラーが出ていたので、以下を実行
sudo apt --fix-broken install -o Dpkg::Options::="--force-overwrite" sudo dpkg --configure -a sudo apt install nvidia-cuda-toolkit
インストールできた。
nvidia-smiではCUDAが12.3といっているが、気にしなくていいみたい（実際は11.8）
06:16:38 shohei@shohei-HP-Z2-Tower-G4-Workstation ~ → nvidia-smi Fri Nov 3 06:16:55 2023 +---------------------------------------------------------------------------------------+ | NVIDIA-SMI 545.23.06 Driver Version: 545.23.06 CUDA Version: 12.3 | |-----------------------------------------+----------------------+----------------------+ | GPU Name Persistence-M | Bus-Id Disp.A | Volatile Uncorr. ECC | | Fan Temp Perf Pwr:Usage/Cap | Memory-Usage | GPU-Util Compute M. | | | | MIG M. | |=========================================+======================+======================| | 0 NVIDIA GeForce GTX TITAN X On | 00000000:01:00.0 Off | N/A | | 49% 84C P2 190W / 250W | 11721MiB / 12288MiB | 93% Default | | | | N/A | +-----------------------------------------+----------------------+----------------------+ +---------------------------------------------------------------------------------------+ | Processes: | | GPU GI CI PID Type Process name GPU Memory | | ID ID Usage | |=======================================================================================| | 0 N/A N/A 20538 C /usr/bin/python3 11716MiB | +---------------------------------------------------------------------------------------+
$ nvcc -V nvcc: NVIDIA (R) Cuda compiler driver Copyright (c) 2005-2022 NVIDIA Corporation Built on Wed_Sep_21_10:33:58_PDT_2022 Cuda compilation tools, release 11.8, V11.8.89 Build cuda_11.8.r11.8/compiler.31833905_0
