Setup TensorFlow, Keras, Theano, Pytorch/torchvision on the CentOS VM

We need to use Keras with TensorFlow as backend for some deep learning projects in the Spark platform we practice data science. The platform we have been using is a CentOS VM running on software based hypervisor called VirtualBox.

Additionally, for future deep learning projects that may use the dataset with Apache Spark, we need to install Theano, Pytorch and Torchvision.

We already have Python 3.6 under virtual environment on Anaconda. Here are the steps:

Enter virtual environment spark

conda activate spark

Upgrade pip

 pip install pip --upgrade

#Collecting pip
#  Downloading pip-20.1.1-py2.py3-none-any.whl (1.5 MB)
#     |████████████████████████████████| 1.5 MB 2.5 MB/s
#Installing collected packages: pip
#  Attempting uninstall: pip
#    Found existing installation: pip 20.0.2
#    Uninstalling pip-20.0.2:
#      Successfully uninstalled pip-20.0.2
#Successfully installed pip-20.1.1

Install TensorFlow

conda install -c conda-forge tensorflow

#Collecting package metadata (current_repodata.json): done
#Solving environment: done
#
## Package Plan ##
#.....
#  environment location: /opt/hadoop/anaconda3/envs/spark

#  added / updated specs:
#    - tensorflow
#....
#Preparing transaction: done
#Verifying transaction: done
#Executing transaction: done

Verify TensorFlow installed successfully

python -c "import tensorflow"

Install Keras

conda install -c conda-forge keras
#Collecting package metadata (current_repodata.json): done
#Solving environment: done

## Package Plan ##

 # environment location: /opt/hadoop/anaconda3/envs/spark

 # added / updated specs:
 #   - keras


# The following packages will be downloaded:

#    package                    |            build
#    ---------------------------|-----------------
#    keras-2.4.3                |             py_0          30 KB  conda-forge
#    ------------------------------------------------------------
#                                           Total:          30 KB

#The following NEW packages will be INSTALLED:

#  keras              conda-forge/noarch::keras-2.4.3-py_0


# Proceed ([y]/n)? y


#Downloading and Extracting Packages
#keras-2.4.3          | 30 KB     | ################################################################################################################## | 100%
#Preparing transaction: done
#Verifying transaction: done
#Executing transaction: done

Verify Keras is installed successfully

python -c "import keras"

Install Theano

conda install -c anaconda theano
#Collecting package metadata (current_repodata.json): done
#Solving environment: done

## Package Plan ##

#  environment location: /opt/hadoop/anaconda3/envs/spark

#  added / updated specs:
#    - theano

#...

#Preparing transaction: done
#Verifying transaction: done
#Executing transaction: done

Install Pytorch

conda install -c anaconda pytorch
#Collecting package metadata (current_repodata.json): done
#Solving environment: done

## Package Plan ##

#  environment location: /opt/hadoop/anaconda3/envs/spark

#  added / updated specs:
#    - pytorch

#...
#Preparing transaction: done
#Verifying transaction: done
#Executing transaction: done

Install Torchvision

 conda install -c  torchvision
#Collecting package metadata (current_repodata.json): done

## Package Plan ##

#  environment location: /opt/hadoop/anaconda3/envs/spark

#  added / updated specs:
#    - torchvision

#...
#Preparing transaction: done
#Verifying transaction: done
#Executing transaction: done

Verify all installation

python
Python 3.6.10 |Anaconda, Inc.| (default, Mar 25 2020, 23:51:54)
[GCC 7.3.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import tensorflow
>>> import keras
>>> import theano
>>> import torch
>>> import torchvision
>>> quit()