# Easiest Cuda, cuDNN installation tutorial for Deep learning(Keras/TensorFlow)

## Nvidia Drivers

$sudo add-apt-repository ppa:graphics-drivers
$sudo apt-get update

then intsall the driver in Ubuntu settings -> Software and update-> Additional driver.

## Cuda
[CUDA 9 DOWNLOAD Link](https://developer.nvidia.com/cuda-90-download-archive?target_os=Linux) (Recommend)

1. choose the OSversion and click **deb(network)**

2. then run the instructions below

$ sudo dpkg -i cuda-repo-ubuntu1604_9.0.176-1_amd64.deb`
$ sudo apt-key adv --fetch-keys
http://developer.download.nvidia.com/compute/cuda/repos/ubuntu1604/x86_64/7fa2af80.pub`
$sudo apt-get update`
$sudo apt-get install cuda=9.0.176-1

`cuda=9.0.176-1` need to be corrsponded the downloaded deb file
`cuda-repo-ubuntu1604_9.0.176-1_amd64.deb`.

<font size=3 color=red>DO NOT USE</font> `sudo apt-get install cuda`, it maybe install other version of cuda.

3. edit `~/.bashrc` and add the text below

    #CUDA settings
    export PATH=/usr/local/cuda-9.0/bin:$PATH
    export LD_LIBRARY_PATH=/usr/local/cuda-9.0/lib64:$LD_LIBRARY_PATH

## cuDNN

[cuDNN DOWNLOAD Link](https://developer.nvidia.com/rdp/form/cudnn-download-survey)

Download the cudnn with corresponding <font size=3 color=red>cuda version supported!</font>

* For example if you downloaded CUDA 9.0, then click

Download cuDNN v7.3.1 (Sept 28, 2018), for CUDA 9.0

1. sign in and download the 3 files

* cuDNN vx.x.x Runtime Library for Ubuntuxx.xx (Deb)

* cuDNN vx.x.x Developer Library for Ubuntuxx.xx (Deb)

* cuDNN vx.x.x Code Samples and User Guide for Ubuntuxx.xx (Deb)

2. run the instructions below

$ sudo dpkg -i libcudnnx_x.x.x.xx-x+cudax.x_amd64.deb

$ sudo dpkg -i libcudnnx-dev_x.x.x.xx-x+cudax.x_amd64.deb

$ sudo dpkg -i libcudnnx-doc_x.x.x.xx-x+cudax.x_amd64.deb

## Python installations
#### Anaconda [Download](https://www.anaconda.com/download/#linux) (Recommend)
1. install Python 3.6 is recommend

$ bash ~/Downloads/Anaconda-versionsxxxx-Linux-x86_64.sh
$ conda install python=3.6
2. create a enviroment..

$ conda create --name my-env python=3.6
3. activate the enviroment

$ conda activate my-env
4. install Tensorflow

$ pip install tensorflow-gpu
or
$ pip install tensorflow

5. install Keras

$ pip install keras

5. edit `~/.bashrc` and add this(optional)

        alias foo='conda activate my-env'

4. test if it work
* For tensorflow

        import tensorflow as tf
        hello = tf.constant('Hello, TensorFlow!')
        sess = tf.Session()
        print (sess.run(hello))

* For Keras

        import keras  
        from keras.models import Sequential
        from keras.layers import Dense, Dropout

        def model(input_size=50):
            model = Sequential()
            model.add(Dense(50, input_dim=input_size, activation='relu'))
            model.add(Dropout(0.2))
            model.add(Dense(40, activation='relu'))
            model.add(Dropout(0.2))
            model.add(Dense(2,activation='linear')) 
            model.add(Dense(1))
            # Compile model
            model.compile(loss='mse', optimizer='adam')
        
            return model
        
        model = model(50)
        print(model.summary())