Install CUDA Toolkit and cuDNN Library on Windows 10 #cuda #deeplearning #nvidia

Video Statistics and Information

Video
Captions Word Cloud
Reddit Comments
Captions
this is shubham and welcome to my YouTube channel so in this video we are going to learn or we are going to see how to install a Cuda the Nvidia Cuda toolkit and CU DNN that is convolutional neural network Cuda neural network sorry for neural network library it is offered by Nvidia okay so in this video we are going to learn all this for deep learning we are installing all these dependencies for utilizing your GPU so you may have a GeForce Nvidia like GTX or RTX whatever it is but you want to use that GPU for train your machine learning or deep learning model to how to utilize that GPU for that you need some dependencies like Cuda toolkit you have to install and and Cuda deep neural network that is c u d n n this both dependencies you must have in your laptop or your system then only you can utilize your GPU for train your neural network or deep learning model got it so let's see how to do that so before started you must have to ensure that you must have uh downloaded this Cuda toolkit from official website you can see here the www.animedia.com Cuda toolkit you just have to write Google in the Google you must you just have to write Coda toolkit it will give you this page go to download now select your parallel version and all this then you can able to select it okay it's pretty much simple after that you select the Windows machine after that 64 bytes whatever the operation system operating system you have mine is 10 make sure you want what exe local network or exe local whatever you want to choose you can choose I have choose PLC local ex local you have to use and you can download look this is 33.4 GB file okay after downloading I already downloaded it if I'm going to show you it should be look like this okay where is my software folder now here it is Nvidia Graphics look show it is here okay so it's look like this so this is your uh Coda 11.5 version okay this is a exe file that I have downloaded it is around 2.44 GB so now we have latest version but this file I had downloaded previously now next thing we have to download see you again and see you DNN see you deep neural network that file so again you have to go to Googles and search for studn and neural network and then you have to then you will come on this page then simply go here download CU DNN it will give you one zip file or archive file that file you must have download and before downloading all this thing you can see a different uh Co DNR networks are there archives you do not release different different release are there so make sure while downloading your Cuda toolkit and CU DNN both have the same version okay both have the same version mine is our mind both file are same network same version as you can see here okay it is 11.5 here it is also Cuda 11.5 version okay for downloading if you are facing any issue make sure you must have your login with your Nvidia developer account okay you have to register it it's free of cost it will not take that much cost or it will free only okay after logging or registering to this developer Nvidia or Nvidia then only you can able to download both files okay now after that it's pretty much simple now you understand how to download both file just simply uh double click on this file Cuda toolkit okay just double click on it so I have written some stage over here if you can see here for installation of any media Quran Network there are certain about seven steps are there for installation of nvdr toolkit and CU DNN for deep learning the first step we have completed download the Cuda toolkit and cuden archive file from internet so these steps are done now second step what we'll do we'll install Cuda okay for that I have just double click over it it will take some time which are asking permission so I will say yes so it will show you the exact where it is going to extract their local files so it is going to show app data somewhere else okay if you want some other directory you can install it in my program files also but my is using Shivam that is a one of the user of my shoe that is one of user of my operating system so I am going to extract there only okay so it's going to extract now it will take some time just wait for it or I you can skip the stage let me try I can pause the video or not I will try uh it's only stop there okay [Music] yeah now it is 100 done Now One dashboard will come come in front of you look like this Coulda deep neural network wow looks amazing it will take some time to reopen this so you can see here it's checking system capability is checking that your system or your laptop will have the GeForce Nvidia graphics card or not make sure after 2008 or 2000 to 16 whatever release is there I forgot the exact date they are having the coda capabilities okay after 2018 I think so make sure that all are there so you terms and condition you must have to agree and continue so it will provide you two option of installation was in Express and one is custom so you can see in a custom if you go and click on the next button it will show you what he's going to install it is installing Cuda and media as you for experience app other component if you expand it you can see different different libraries development for development purpose we are using this if you want some sample code this is also installing that sample code General sample code so all the things are here okay just go back again but we are not going to with this Advanced option we are selecting this Express option here okay now go to express and select that and click on next button okay now it's going to install it will take server time 2-3 minutes okay to that I am going to pause my video [Music] so yeah as you can see here my Cuda toolkit Nvidia Cuda till kit version 11.5 inch installed here properly okay and some of the depends here install some are not installed that you can take because it's using 2019 version of Visual Studio 2017. I have already installed that if you don't know about that please go to Google search for video Visual Studio code for the older version you must have to download it make sure you have to download Community version of Visual Studio code 2017 and install that and while installing make sure that you it's already I mentioned in the same six step the download visual studio Community 2017 version and everything related to C plus plus just check the click checkbox and download that dependency okay because this Nvidia graphics card is working on C plus plus okay so after that you can see here my installation part is upgraded now click on the next button so if you want to launch the sample you can launch but we are not going to launch sample now so make sure you have to close this okay after this installation of Nvidia Cuda we need to install CU DNN that is deep neural network that is used for a deep learning if you want to utilize your this is a home screen after installation of Cuda okay I'm minimizing it after that next step what is there we have to follow so we have installed our Kura next to it next step that is the third step in that we have to do extract the CV data and archive file into your local Drive see right okay that thing we have to do so I'm I will go to my location where I have that archive file what I will do I will copy an extra call so it will ask me the browser location where I want to extract back make sure you have to extract that archive file or zip file in the C drive so I am going to my C drive here all program files everything are there but I am going to extract this file here only outside so in C drive I have given the location I will select the folder okay so it will install in my C drive extract so my this Cuda save file or you can say archive file is extracting over there in my C drive okay so wait for it yeah it's done okay so you can see if I go inside my C drive there is one folder is created that is coda archive file okay this is here uh if you want to rename it you can name it I will make it simple uh I will simply say c u d n n okay you can rename it but uh this is how I can simply understand it after that what is the next step we have to follow okay so we have started our file after that open the location where Windows have Coda installed so this is my location when you see Drive program file is NVIDIA toolkit this is not my location my location is different one so I am going to open that location so one window I will open uh this uh Coda file codnn okay uh it have certain files like bin you can see and also I will open new window uh C drive only so here here I am inside my C drive you must have to go to program files in the program file you have to search for NVIDIA GPU Computing toolkit in that Cuda in that go to version 11.5 in that pin okay so open the this both windows side by side and copy all this codnn files present in the see dry location only Ctrl C from this side in bin location and Ctrl V okay uh it's showing the adventure permission of copy this so make sure you have to yes continue with it administer your permission required for all the current item you want okay please do that so I am copying all the my bin libraries into that folder you can see here I already mentioned that so it's copied there okay after that what to do uh same thing what we have to do for for include also and for lib also but for leave we have to copy all file from leap and paste it into x 46 folder is there so I will show you show you how to do that so this is our file this is our Coda files okay now in Cuban folder there is include folder also copy that Ctrl C folder [Music] come here and Ctrl V all current file you want to copy yes I want to copy all the files now I just copied look on the final after that what you have to do the lib Library okay Ctrl a Ctrl C come into 11.5 folder again Row 2 this time you have to go to uh lip Okay lib is there inside leave you have one folder x46 so we have to go here and Ctrl V and continue so all the files will copy from here to here okay so we have done with this step which Step that we have copied all the flight from c u DNN folder to whatever mentioned files I have shown you or whatever the mentioned folder I have shown you just copy it okay after that we have to set mention the environmental path okay we have to mention the environment path how to do that for this simply what we have to do I will minimize this search here for a nvi environment variable okay go to Advanced go to Advanced then go to the environment variable okay inside path is there okay just double click it or open it so make sure you have to mention that path whatever the file where you copied not mention that path particularly here so you can see here all the bean path okay all the bin and live V nvp path to this folder so I'm going to my CLI folder so this is not far for Cuda I am asking for the another path for x246 here I will come here I will go to bin for Puda bin I'm going to copy this path Ctrl C now I will go to my environment variable I will create new one I will paste this path here okay and same thing I will do for same thing I will do for V what what is name divyan vvp for that file so go to leave 1vb folder copy the path Ctrl C come to the environment variable create new One controls V again the both paths are covered into your environment variable after that what you have to do okay okay and okay so then we are done with our setup for Cuda Network and all so this is all guys make sure you have to download your videos to the community version download it and create a new environment variable and tensorflow that I will show you into the next video okay so that's all if you want to taste it you can taste it otherwise the water or the setup requires you must have to see that into your Nvidia GPS title for all the drivers required for it you can download it from here what are the things required you can download it from here so I'm going to close this okay see you then thank you thank you for watching this video please make sure like do subscribe my channel thanks a lot thank you
Info
Channel: S-Qube
Views: 16,519
Rating: undefined out of 5
Keywords:
Id: ctQi9mU7t9o
Channel Id: undefined
Length: 15min 56sec (956 seconds)
Published: Mon Dec 26 2022
Related Videos
Note
Please note that this website is currently a work in progress! Lots of interesting data and statistics to come.