Step 8/15 : RUN apt-get update & apt-get install -y -no-install-recommends cuda-nvrtc- $CUDA_PKG_VERSION cuda-nvgraph- $CUDA_PKG_VERSION cuda-cusolver- $CUDA_PKG_VERSION cuda-cublas-8-0 =8.0.61.2-1 cuda-cufft- $CUDA_PKG_VERSION cuda-curand- $CUDA_PKG_VERSION cuda-cusparse- $CUDA_PKG_VERSION cuda-npp- $CUDA_PKG_VERSION cuda-cudart- $CUDA_PKG_VERSION & ln -s cuda-8.0 /usr/local/cuda & rm -rf /var/lib/apt/lists/* Removing intermediate container 11358e055e46 Step 7/15 : ENV CUDA_PKG_VERSION 8-0 = $CUDA_VERSION-1 Removing intermediate container 1c2adf7e726b Step 6/15 : ENV NVIDIA_CUDA_VERSION $CUDA_VERSION Removing intermediate container 055b6e0f8b18 NVIDIA_GPGKEY_FPR =ae09fe4bbd223a84b2ccfce3f60f4b3d7fa2af80 & \Īpt-key adv -export -no-emit-version -a $NVIDIA_GPGKEY_FPR | tail -n +5 > cudasign.pub & \Įcho " $NVIDIA_GPGKEY_SUM cudasign.pub" | sha256sum -c -strict - & rm cudasign.pub & \Įcho "deb /" > /etc/apt//cuda.list ENV CUDA_VERSION 8.0.61 LABEL = " $ " sudo cp cuda/include/cudnn.h /usr/local/cuda/include sudo cp cuda/lib64/libcudnn* /usr/local/cuda/lib64 sudo chmod a+r /usr/local/cuda/include/cudnn.# FROM ubuntu:16.04 LABEL maintainer "Gemfield " RUN NVIDIA_GPGKEY_SUM =d1be581509378368edeec8c1eb2958702feedf3bc3d17011adbf24efacce4ab5 & \ Unzip the cuDNN package tar -xzvf cudnn-10.2-linux-圆4-v7.6.5.32.tgzĬopy the following files into the CUDA Toolkit directory, and change the file permissions. Reboot system: sudo reboot Check CUDA installation #1: nvidia-smi Check CUDA installation #2: nvcc -V cuDNN Installation Install GCC sudo apt-get install gcc CUDA Toolkit 10.1 Download Use this command to check your GPU lspci | grep -i nvidia Handle conflicting installation methods.įirst and foremost, your GPU must be CUDA compatible.Verify the system has the correct kernel headers and development packages installed.Verify the system is running a supported version of Linux.Verify the system has a CUDA-capable GPU.Some actions must be taken before the CUDA Toolkit and Driver can be installed on Linux: So how to install CUDA 10.1? Follow me here. As of today (Feb 2020) pytorch on GPU requires CUDA 10.1 but CUDA 10.2 is the latest version.
0 kommentar(er)
