Tutorial: Anaconda + TensorFlow 2.0 @ Ubuntu 18.04

Installation of Anaconda and TensorFlow 2.0 at Ubuntu 18.04

In this story, I will install Anaconda and TensorFlow 2.0 at Ubuntu 18.04. It is supposed that, at Ubuntu, you have installed the updated GPU driver, CUDA and cuDNN. If not, please refer to my another story: Tutorial: CUDA v10.2 + CUDNN v7.6.5 Installation @ Ubuntu 18.04. (Sik-Ho Tsang @ Medium)

Outline

1. Anaconda Installation
2. TensorFlow Installation
3. Import TensorFlow

1. Anaconda Installation

• Go to https://www.anaconda.com or https://www.anaconda.com/distribution/#download-section to download the Anaconda for Linux.
• I downloaded Python 3.7 64-Bit version as shown above.
• Then install it:

bash Anaconda3-2019.10-Linux-x86_64.sh

2. TensorFlow Installation

• For me, I create a new virtual environment for TensorFlow installation to keep the base clean.
• As my computer got GPU, I install the GPU version with tf-gpu as the name of the new virtual environment.

conda create -n tf-gpu tensorflow-gpu

• After that, activate it:

conda activate tf-gpu

3. Import TensorFlow

• Go into python:

python

• If TensorFlow is successfully installed, we can import it.

import tensorflow as tf

• And we can print the version

print(tf.version)

• It should be version 2.0.

Linux [Ubuntu Installation] [NVIDIA Driver Installation (2018) (Old)] [OpenSSH Installation] [Hard Drive Partitioning/Formatting/Mounting] [Add 1 More GPU] [TeamViewer Installation][NVIDIA Driver Installation (2019)] [CUDA v10.2 + CUDNN v7.6.5 Installation @ Ubuntu 18.04] [Anaconda + TensorFlow 2.0 @ Ubuntu 18.04]

Docker [Docker Installation] [Pulling Image] [Running Image] [Exporting/Saving Image] [Nvidia-Docker 2.0 Installation]

Caffe [Image Classification] [Handwritten Digit Classification] [Style Recognition]

Video Coding [Generate VTM Solution (2019) (Old)] [Generate HM+360 Solution] [Generate VTM+360 Solution] [Generate VTM Solution (2020)]