Deep learning is getting popular with the data scientists. Here, Im sharing my first-hand experience with the TensorFlow and Theano. It was started to help the MTech Students from MG university. All Credits to goes to Noufal Salim who helped me to understand the configuration of different python deep learning environments. Let us explore deep learning !!!
Before starting the installation guide, let us familiar with the technical terms and tools.
Keras is an open source neural network library written in Python
TensorFlow is an open source software library for machine learning developed by Google
Theano is a numerical computation library for Python
Configuration Prerequisite: Python 2.7 , 64Bit
Step 1 : Download Anaconda 64 bit installer for Python 2.7 from continuum website - Link
Step 2 : Bash Anaconda - Goto corresponding download directory in Terminal and In your terminal window type the below instruction:
bash Anaconda2-4.3.1-Linux-x86_64.sh
Iti will ask your permission for installing and preferred installation directory . It will create default environment . It will ask
" Do you wish the installer to prepend the Anaconda2 install location to PATH in your /home/prem/.bashrc ? [yes|no] "
>>> yes
Prepending PATH=/home/prem/anaconda2/bin to PATH in /home/prem/.bashrc
A backup will be made to: /home/prem/.bashrc-anaconda2.bak [default]
For this change to become active, you have to open a new terminal.
Step 3 : Add PATH
Add this line to the file .bashrc in your home directory:
export PATH="/home/username/anaconda/bin:$PATH"
Type the following in new terminal
Step4: >>conda update conda
Step5 : >>conda install Theano
Step6 : >>conda install tensorflow
Step7 : >>conda install keras
Step 8 : Activate the Anaconda Python by activating the corresponding path.
>>source activate PATH
In this , PATH is /home/prem/Anaconda2/
Step 9 :Run the program python filename.py
>>source activate PATH
In this , PATH is /home/prem/Anaconda2/
Step 9 :Run the program python filename.py
You can see lot of sample deep learning programs in Internet sites. Try it !!!
Few resources are
Few resources are
Comments
Post a Comment