![]() Install and setup a virtual environment with Python libraries (including Tensorflow and Keras).You guys should check it out too (especially, if you want to gain expertise in Computer Vision), if you already haven’t. I am just trying to present how I used Adrian’s guide to setup my personal laptop.Īdrian is one of the best experts on Computer Vision and Deep Learning, and I follow his blogs on. The only difference being that Adrian’s guide is focused on setting up a remote system (AWS or another system). This guide heavily follows Adrian Rosebrock’s guide on Setting up Ubuntu 16.04 + CUDA + GPU for deep learning with Python. Create a bootable USB drive with Ubuntu 16 installer.You can follow my earlier blogs to get that done. The prerequisite to this guide is to have Ubuntu ready on your system. Now, I am using a friend’s system, remotely via SSH, which has a 11 GB Nividia 1080i GPU. Addition: I have surpassed my laptop’s system specifications, in terms of the hardware requirements for my projects, esp.Now, Ubuntu is the only OS on my system, utilizing all of RAM (16 GB), SSD (256 GB), and GPU (Nvidia GeForce 940MX 2 GB).Again, there were limitations to it - the read/write speeds were slow, I couldn’t accommodate large datasets, and occasional “Bus Error”,to name a few. Setup Linux and everything else mentioned in this blog on a USB drive and use it to directly boot my system.I could not utilise 100% of my system hardware and GPU wasn’t accessible at all. This was a good way to get started with coding. Setup a Virtual Machine on Windows and then install Ubuntu and required Python packages.The earlier two types of attempts (I setup the system multiple times in both types) were with reluctance to do away with Windows on my system. I realized the need of it after the 3rd time I was trying to setup my machine. I am writing this guide so that I can refer back to it whenever I am setting up a Deep learning machine for Computer Vision. Sudo ln -s libcudnn.so.7 libcudnn.so 5.How to Setup Ubuntu 16.04 with CUDA, GPU, and other requirements for Deep Learning Sudo ln -s libcudnn.so.7.0.4 libcudnn.so.7 Regenerate the soft link (the specific content is modified according to the actual situation) sudo rm -rf libcudnn.so libcudnn.so.7 ![]() Sudo cp cuda/lib64/libcudnn* /usr/local/cuda/lib64/Ĥ. Move files sudo cp cuda/include/cudnn.h /usr/local/cuda/include/ You need to register an account first, and then select the corresponding version to downloadģ. (2) Terminal input sudo gedit /etc/profileĪdd to: export PATH=/usr/local/cuda/bin:$PATHĮxport PATH=/usr/local/cuda/bin:$PATH (3) Terminal inputĪdd to: /usr/local/cuda/lib64 (4) Terminal input sudo ldconfig Note that when you encounter an option asking whether to install the driver, select NO and no longer install the driver.Īdd to: export PATH=/usr/local/cuda-9.0/bin$ Enter the official website and select the corresponding version to download: Check NVIDIA, terminal input nvidia-smiġ. –no-opengl-files means not to install opengl, this is necessary, otherwise there may be a problem of circular login, –no-x-check and –no-nouveau-check is to not check the X service and nouvea when installing the driver, These two sentences are not necessary.ħ. NVIDIA-Linux-x86_n -no-x-check -no-nouveau-check -no-opengl-files Press Ctrl+Alt+F1 to switch to tty1, you need to enter the Ubuntu username and password again. Type in the terminal sudo service lightdm stop Select the corresponding version to downloadĤ. Reboot the system after the update is complete sudo reboot After reboot enter: Options new modeset=0 Be sure to update: sudo update-initramfs -u Type in the terminal: sudo gedit /etc/modprobe.d/nf First uninstall the original NVIDIA driver: sudo apt-get purge nvidia* Install the NVIDIA graphics card driverġ. Ubuntu 16.04 install NVIDIA graphics driver and CUDA, CuDnnġ.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |