python - Tensorflow extremely slow after switching to python3 -


i developed cnn using tensorflow , python2.7. switching code python3.5. have both python versions on machine , have 2 tensorflow versions installed (one through pip , 1 through pip3). using linux 16.04.

when try run code using python3 command, takes long time load , doesn't start training (where used take 3 seconds). slows down entire machine, it's memory issue. coworker gets memory error when running same code on windows (his machine has 128gb of memory).

my cnn has 2 convolutional layers , 1 connected layer, , i'm loading less 100mb of data.

why tensorflow acting differently when change python version?


Comments

Popular posts from this blog

android - InAppBilling registering BroadcastReceiver in AndroidManifest -

python Tkinter Capturing keyboard events save as one single string -

sql server - Why does Linq-to-SQL add unnecessary COUNT()? -