GPU Headaches: Notes on Installing CUDA, CuDNN and Tensorflow on Manjaro • Hacker News Board

Re: GPU Headaches: Notes on Installing CUDA, CuDNN and Tensorflow on Manjaro

#21

Post by mippie_moe » Tue, Aug 28, 2018, 4:48 PM UTC

My company (Lambda Labs) made it possible to install cuda, cudnn, tensorflow, pytorch, and other deep learning frameworks with one line:

https://lambdalabs.com/lambda-stack-deep-learning-software

We wrote debian packages for every framework, including cuda and cudnn. Using our Debian repository, you can install all these frameworks using apt/aptitude.

When a new version of a framework comes out, we usually have it available in 1-2 weeks.

Re: GPU Headaches: Notes on Installing CUDA, CuDNN and Tensorflow on Manjaro

#22

Post by acd » Tue, Aug 28, 2018, 4:58 PM UTC

It is usually headaches to install closed source software on Linux. Ffmpeg with Cuda hardware acceleration encoding support is also a pain in the... to install.

If graphic card makers would open source their drivers it would come default shipped apt-get,pacman,dnf,yum installable by the Linux distros.

Re: GPU Headaches: Notes on Installing CUDA, CuDNN and Tensorflow on Manjaro

#23

Post by romaniv » Tue, Aug 28, 2018, 5:01 PM UTC

I wonder when we will get to the point where computers (end-user computers) learn using ML techniques. Will we ever? Will it ever become accessible to non-developers?

Re: GPU Headaches: Notes on Installing CUDA, CuDNN and Tensorflow on Manjaro

#24

Post by solomatov » Tue, Aug 28, 2018, 5:03 PM UTC

The best way to setup tensorflow for gpu is to use nvidia-docker and pre-built docker images. They mention it in their documentation.

Re: GPU Headaches: Notes on Installing CUDA, CuDNN and Tensorflow on Manjaro

#25

Post by nextos » Tue, Aug 28, 2018, 5:05 PM UTC

The only two distributions I've tried where this worked with a one liner were NixOS and Arch. I'm surprised this doesn't work in Manjaro given that it's an Arch derivative. But I don't see much point in Arch derivatives as they do away with Arch's selling point. No layers between you and upstream.

Re: GPU Headaches: Notes on Installing CUDA, CuDNN and Tensorflow on Manjaro

#26

Post by scottlegrand2 » Tue, Aug 28, 2018, 5:27 PM UTC

illumin8 wrote: post #3

So basically, your calculations for deciding whether purchasing physical GPUs or using cloud GPUs is more cost effective should add $5,000 (50 hours at $100 an hour seems reasonable) onto the cost of any physical rig, and assume you have a spare 50 hours of opportunity cost to set it up. All the ML researchers lusting for bare metal need to decide if they want to be "Gentoo Ricers" - https://funroll-loops.teurasporsa…

In my experience, the same mindset that cannot figure out clear directions on how to set up a GPU machine is similarly incompetent at cleaning up data and understanding machine and deep learning models sufficiently to produce useful results. For bonus points, many of them seem to look down on those that can as "the help" or as "Ops."

That said, configuring a GPU machine is (still) ridiculously complicated IMO. It's just that I've been doing it for almost 20 years.

Re: GPU Headaches: Notes on Installing CUDA, CuDNN and Tensorflow on Manjaro

#27

Post by swebs » Tue, Aug 28, 2018, 5:30 PM UTC

_Wintermute wrote: post #7

illumin8 wrote: post #3

So basically, your calculations for deciding whether purchasing physical GPUs or using cloud GPUs is more cost effective should add $5,000 (50 hours at $100 an hour seems reasonable) onto the cost of any physical rig, and assume you have a spare 50 hours of opportunity cost to set it up. All the ML researchers lusting for bare metal need to decide if they want to be "Gentoo Ricers" - https://funroll-loops.teurasporsa…

50 hours is definitely an outlier. I'm pretty clueless when it comes to linux and drivers, and I managed to get an nvidia GPU working on ubuntu in about an hour or two of swearing and a dozen reboots.

I don't know what you guys are doing, but for me installing proprietary Nvidia drivers on Ubuntu was just a matter of clicking a checkbox.

Re: GPU Headaches: Notes on Installing CUDA, CuDNN and Tensorflow on Manjaro

#28

Post by gue5t » Tue, Aug 28, 2018, 5:48 PM UTC

  export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/cuda/extras/CUPTI/lib64
  
  Reboot and cross your fingers.

Rebooting will terminate the process tree in which the environment has been modified, making this line a no-op. Please take the time to understand the commands you run before suggesting them to others.

Re: GPU Headaches: Notes on Installing CUDA, CuDNN and Tensorflow on Manjaro

#29

Post by leblancfg » Tue, Aug 28, 2018, 6:10 PM UTC

jorgemf wrote: post #14

Archlinux user here (Manjaro is based on Archlinux). TensorFlow with GPU support is a supported package you can install it with all it's dependencies in a simple command with Pacman. But you don't want to do that the same way you don't want to compile TensorFlow. It is better to use an official docker image because they include all the dependencies, even the optionals as nccl 2.0. You forgot to install it, and it is…

That's a fair point! I tried installing the pre-packaged conda Tensorflow package, but it was expecting CUDA 9.0 and CuDNN 7.0. The ones I had finally managed to install were 9.2 and 7.2... hence compiling it from source.

I should add that as an addendum to the article, though, great catch!

Re: GPU Headaches: Notes on Installing CUDA, CuDNN and Tensorflow on Manjaro

#30

Post by fooker » Tue, Aug 28, 2018, 6:15 PM UTC

gue5t wrote: post #28

export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/cuda/extras/CUPTI/lib64 Reboot and cross your fingers. Rebooting will terminate the process tree in which the environment has been modified, making this line a no-op. Please take the time to understand the commands you run before suggesting them to others.

Now I see why the author had so much trouble.