mcotton

How to install Tensorflow2 with GPU support in Ubuntu 20.04

Installation of Tensorflow2 with GPU support is easy and the only complication can be arisen from the CUDA compability which in turns depends on the Nvidia driver version. Before going farther, please check if your Nvidia Video Card is compatible with the required versions that are defined in this gist, use this link.

Tensorflow offers in its website a table of the compatibility between libraries for the target OS. You can visit that website in the following COMPATIBILITY TABLE that points to the Tensorflow installation from source for linux. For the time writing this gist, Tensorflow2 v2.5.0 requires CUDA v11.2 and CUDNN v8.2. It is really important to match the exact version, otherwise tensorflow will have problems loading the shared libraries as not finding the correct version.

CUDA version also requires for a minimum Nvidia driver version

mcotton

from the host system

tensorman run -p 8888:8888 --gpu --python3 --jupyter bash

from inside tensorman

jupyter notebook --ip=0.0.0.0 --no-browser

back on host system

open http://127.0.0.1:8888/?token=18d08c03e8ffd3bcc18fab37f6a4dc4cd146120979d62ed8

someone suggest running this instead of dealing with jupyter token

mcotton

https://www.reddit.com/r/MacOSBeta/comments/im0unc/smb_in_big_sur_beta_6/g3zevaf/

HI, these 2 commands in the terminal have fixed my SMB problem with the beta 6:
Open Terminal and type:
sudo sh -c 'echo "[default]\nsigning_required=no" > /etc/nsmb.conf'
defaults write com.apple.desktopservices DSDontWriteNetworkStores true
Quit terminal

mcotton

sudo du -a /home/mcotton/ | sort -n -r | head -n 20

mcotton

#
# Look at https://www.tensorflow.org/guide/gpu for an explaination
# Here are the two options they presented
#

# Option 1
# set_memory_growth before creatingt the CONV network

import tensorflow as tf
gpus = tf.config.experimental.list_physical_devices('GPU')

mcotton

ffplay -flags2 +export_mvs video.mp4 -vf codecview=mv=pf+bf+bb

mcotton

#.bashrc

alias disableGUI='sudo systemctl set-default multi-user.target'
alias enableGUI='sudo systemctl set-default graphical.target'
alias startGUI='sudo systemctl start gdm3.service'
alias stopGUI='sudo systemctl stop gdm3.service'

/usr/bin/setxkbmap -option "caps:escape"

if [ -f ~/dev/personal/dot_files/.docker_aliases ]; then

mcotton

def export_dict(self):
         return {
             "H1_id": self.key.id(),
             "gcs_filename": self.gcs_filename,
             "serving_url": self.serving_url,
             "file_name": self.file_name,
             "created_at": str(self.created_at),
             "company": self.company.id(),
             "project": self.project.id(),
             "mime_type": self.mime_type,

mcotton

execfile('remote.py')
set_current_namespace('mcotton')

from google.appengine.ext import blobstore 
from google.appengine.ext.webapp import blobstore_handlers
import cloudstorage as gcs

blob_key = blobstore.create_gs_key('/gs/' + s.GCS_BUCKET + '/152_592.zip')
billy = blobstore.BlobKey(blob_key)

mcotton

convert  DSC_5936.JPG DSC_5937.JPG DSC_5936.JPG -loop 0 -morph 18 -gravity South  -resize 600x400 -rotate 270 morph.gif