pasteruu.blogg.se - Ubuntu 18.04 cuda 10.0 tensorflow

#UBUNTU 18.04 CUDA 10.0 TENSORFLOW INSTALL#
#UBUNTU 18.04 CUDA 10.0 TENSORFLOW UPDATE#
#UBUNTU 18.04 CUDA 10.0 TENSORFLOW PATCH#
#UBUNTU 18.04 CUDA 10.0 TENSORFLOW DOWNLOAD#

Virtualenv -p python3.7 ~/tensorflow_1-15 Create a virtual environment and add the desired python as the default interpreter.This just installs python3.7 to the system it does not make it default. Sudo add-apt-repository ppa:deadsnakes/ppa

#UBUNTU 18.04 CUDA 10.0 TENSORFLOW INSTALL#

One can install python 3.7 for example using this repository which contains old (and new versions of python): This means obligatory a virtual environment (since messing with system python is never not a good idea). The supported versions of python for cuda 10.0 ends with 3.7 so an older version should be installed.Anyway, this way the system will search for the relevant so files in:Ī) /usr/local/cuda/lib64 (the default symbolic link) and it will failī) to the virtually same as the latter /usr/local/cuda-11.3/lib64 and also fail BUT it will search alsoĬ) /usr/local/cuda-10.0/lib64 which will be successful.

This requires a restart though to be evaluated I think.

This command would normally do the work: $ sudo sh -c ‘echo export LD_LIBRARY_PATH=/usr/local/cuda/lib64:/usr/local/cuda-11.3/lib64:/usr/local/cuda-10.0/lib64:\$LD_LIBRARY_PATH > /etc/profile.d/cuda.sh’ It should contain something like:Įxport LD_LIBRARY_PATH=/usr/local/cuda/lib64:/usr/local/cuda-11.3/lib64:/usr/local/cuda-10.0/lib64:$LD_LIBRARY_PATH

#UBUNTU 18.04 CUDA 10.0 TENSORFLOW UPDATE#

One way (there are others) system-wide is to create (in not existent) a /etc/profile.d/cuda.sh which will contain the update to the LD_LIBRARY_PATH variable. We should update the relevant system environment for cuda 10.0. So, all call to it will fail as if non existent.

Upto this point although installed the system is unaware of the presence of cuda 10.0.

$ sudo chmod a+r /usr/local/cuda-10.0/include/cudnn.h /usr/local/cuda-10.0/lib64/libcudnn* $ sudo cp cuda/lib64/libcudnn* /usr/local/cuda-10.0/lib64 $ sudo cp cuda/include/cudnn.h /usr/local/cuda-10.0/include So, an extraction of the compressed file and copy to the folder would suffice: The version I used is cudnn-10.0-linux-圆4-v7.6.5.32.tgz.Ĭudnn basically just copies files in the right places (do not actually install anything that is).

#UBUNTU 18.04 CUDA 10.0 TENSORFLOW DOWNLOAD#

I had this one from a previous installation but its shouldn't be hard to download it also. The toolkit can be safely installed using the instructions provided with no risk since each version allocates a different folder in the system (typically this would be /usr/local/cuda-10.0/).

#UBUNTU 18.04 CUDA 10.0 TENSORFLOW PATCH#

I chose runfile as method which resulted into 1 main runfile and 1 patch runfile being available: cuda_10.0.130_410.48_n I resolved to the Ubuntu 18.04 version though which works fine.įinal toolkit for Ubuntu here (as it's obvious not 20.04 version is available). I have some trouble finding this version because it's not officially available for Ubuntu 20.04. To make a working solution for Ubuntu 20.04 and TensorFlow 1.15 one needs: The main idea is from here (not to claim credit from others). Given the situation is there a way to run tensorflow 1.15 with gpu support on Ubuntu 20.04.1?Īs this also bothered me I found a working solution that I think is more versatile than using docker containers.

Should I install and older version of Ubuntu and hope I don't go obsolete too quickly.

Should I attempt to compile tensorflow from source against CUDA 10.1 (heh heh heh).

Should I download the install for CUDA 10.0 for a the older Ubuntu version and see if it will install anyway.

Should I install CUDA 10.1 and cross my fingers 1.15 will like it.

Uninstalling things isn't always straightforward. I also saw a random comment that tensorflow-gpu-1.15 will just-work with tf 1.15, but I am not wanting to make a miss-step installing things until I have a signal that is the direction to go. I see some possible clues that it might be possible to build tensorflow-gpu-1.15 from source with cuda 10.1 support. However experimenting with reinforcement learning was the motive for purchasing this PC. I would highly prefer not to have to reinstall the OS with an older version of Ubuntu. $ sudo apt-cache policy nvidia-cuda-toolkit Searching the apt-cache does not result in CUDA 10.0 either. I am wanting to use baselines-stable which isn't tensorflow 2.0 compatible yet.Īccording to here and here, tensorflow-gpu-1.15 is only listed as compatible with CUDA 10.0, not CUDA 10.1.Īttempting to download CUDA from Nvidia, the option for Ubuntu 20.04 is not available for CUDA 10.0. I am building a Deep Learning rig with a GeForce RTX 2060.