-1

I'm trying to get started with tensorflow using the python interface. I'm building an image classification system explained Here. But when running the code, epochs take too much time, almost 2 minutes for 1 epoch, and if number of steps are increased, epoch running time increases exponentially.
My system configurations are:

enter image description here

and software configurations are:

  • Python 3.7
  • Spyder 4
  • Tensorflow 2.2.0

I found similar thread Here but in my case, basic operations are fast enough.

How can I improve performance of tensorflow

Sachin Kumar
  • 781
  • 1
  • 9
  • 28
  • 2
    You need a Nvidia GPU to accelerate training. There's no way around it as of today. – Susmit Agrawal May 31 '20 at 06:03
  • Two minutes per epoch sounds reasonable for the platform,I would not call this "slow" – Dr. Snoopy May 31 '20 at 10:13
  • @Dr. Snoopy read the documentation from given link, There are only 15 steps per epochs... Accordingly it is very slow – Sachin Kumar May 31 '20 at 18:00
  • What is the comparison point? The article does not say which platform it is using to obtain 8 seconds per epoch, so claiming 15 steps per epoch must be fast is not correct. – Dr. Snoopy May 31 '20 at 18:12

1 Answers1

1

Sadly, both TensorFlow and PyTorch use only Cuda as a backend for GPU acceleration and they don't support any of the Mac'a GPUs - Intel's or AMD's. This means that your TensorFlow code would run only on CPU.

stan0
  • 11,549
  • 6
  • 42
  • 59