Tensorflow Workshop setup

gistfile1.txt

## Installation Steps

For Mac and Linux users

1. Go to Anaconda page (link: https://www.anaconda.com/download). Download and install the Python 3.6 version. For more info, you can visit this webpage (Mac users: https://docs.anaconda.com/anaconda/install/mac-os, Linux users: https://docs.anaconda.com/anaconda/install/linux)

2. Create a FloydHub account (link: https://www.floydhub.com/signup). Make sure you create a FloydHub account with the same email address with which you have registered for DataHack Summit 2017. Only then we will be able to assign GPU credits.

3. Install FloydHub in your system by going to the terminal by executing 

    pip install -U floyd-cli

4. Now you have to login to FloydHub. Execute the command below in the terminal 

    floyd login 

5. The command above opens up a webpage that displays the authentication token. Copy this token, go to the terminal again and input this token
     
6. To check if your system is working fine, go through the tutorial mentioned in the webpage (link: https://docs.floydhub.com/getstarted/quick_start/)


For Windows users

1. Go to Anaconda page (link: https://www.anaconda.com/download/#windows) and download the appropriate package according to your system (32bit or 64bit). Install the Python 3.6 version. For more info, you can visit this webpage (link: https://docs.anaconda.com/anaconda/install/windows)

2. Create and FloydHub account (link: https://www.floydhub.com/signup). Make sure you create a FloydHub account with the same email address with which you have registered for DataHack Summit. Only then we will be able to assign GPU credits.

3. Install FloydHub in your system by going to the anaconda command prompt by executing 

    pip install -U floyd-cli

4. Now you have to login to FloydHub. Type the command below in the terminal 

    floyd login 

5. The command above opens up a webpage that displays the authentication token. Copy this token, go to the terminal again

6. In the command prompt, hit "backspace" a few times to remove the extra characters from the token login prompt request.

7. Right click on the menu bar, and select "Edit", and then "Paste"

8. Then press "Enter"

9. To check if your system is working fine, go through the tutorial mentioned in the webpage (link: https://docs.floydhub.com/getstarted/quick_start/)

## Steps for running the jupyter notebooks

1. Go to the webpage (link: https://www.floydhub.com/projects/create) and create a new project. Give the project name as tf_workshop

3. Open terminal/anaconda prompt and create a new directory by executing

	mkdir tf_workshop
	
4. **Change directory** to the created folder by executing

	cd tf_workshop

5. Execute command

    floyd init tf_workshop
	
6. Execute command 

	floyd clone analyticsvidhya/projects/tf_workshop/1

5. Execute the command given below

    floyd run --cpu --mode jupyter --env tensorflow-1.3 --data analyticsvidhya/datasets/height_n_weight/1:height_n_weight --data analyticsvidhya/datasets/cats_vs_dogs/1:cats_vs_dogs --data analyticsvidhya/datasets/mnist/2:mnist --data analyticsvidhya/datasets/imdb/1:imdb --tensorboard

5. Go to link provided in the terminal (link: https://www.floydhub.com/<user_name>/projects/<project_name>/<job_number>)