MNIST (DCNN) in Keras

What is Keras

Library to make Theano and TensorFlow easier to use

Execution environment

EC2 (AWS) g2.2xlarge instance (Oregon = US West) Python 2.7.6 TensorFlow 0.8.0

The AWS instance was initialized using another person's AMI, but if you want to introduce it yourself, refer to the following

It looks like you don't need to build from source anymore to run TensorFlow on an EC2 GPU instance?

Install Keras

Do as per Documentation Assuming that TensorFlow is already installed, add'sudo' if necessary

pip install scipy
pip install scikit-learn
pip install pyyaml
apt-get install libhdf5-dev
pip install h5py
pip install keras

Once from python

import keras

Run Keras as and edit'~ / .keras / keras.json' as follows

"backend": "theano"

↓

"backend": "tensorflow"

code

`mnist.py`


from __future__ import absolute_import
from __future__ import division
from __future__ import print_function

from keras.datasets import mnist
from keras.utils import np_utils
from keras.models import Sequential
from keras.layers.convolutional import Convolution2D, MaxPooling2D
from keras.layers.core import Activation, Flatten, Dense
from keras.callbacks import EarlyStopping

(X_train, y_train), (X_test, y_test) = mnist.load_data()
nb_classes = 10

X_train = X_train.reshape(-1, 1, 28, 28).astype('float32')
X_test = X_test.reshape(-1, 1, 28, 28).astype('float32')
X_train /= 255
X_test /= 255
y_train = np_utils.to_categorical(y_train, nb_classes)
y_test = np_utils.to_categorical(y_test, nb_classes)

model = Sequential()

model.add(Convolution2D(nb_filter = 16, nb_row = 3, nb_col = 3, border_mode = 'same', input_shape = (1, 28, 28)))
model.add(Activation('relu'))

model.add(Convolution2D(nb_filter = 32, nb_row = 3, nb_col = 3, border_mode = 'same'))
model.add(Activation('relu'))

model.add(MaxPooling2D(pool_size = (2, 2), border_mode = 'same'))

model.add(Convolution2D(nb_filter = 64, nb_row = 3, nb_col = 3, border_mode = 'same'))
model.add(Activation('relu'))

model.add(Convolution2D(nb_filter = 128, nb_row = 3, nb_col = 3, border_mode = 'same'))
model.add(Activation('relu'))

model.add(MaxPooling2D(pool_size = (2, 2), border_mode = 'same'))

model.add(Flatten())

model.add(Dense(1024))
model.add(Activation('relu'))
model.add(Dense(1024))
model.add(Activation('relu'))
model.add(Dense(nb_classes))
model.add(Activation('softmax'))

early_stopping = EarlyStopping(monitor = 'val_loss', patience = 2)

model.compile(loss = 'categorical_crossentropy', optimizer = 'adam', metrics = ['accuracy'])
model.fit(X_train, y_train, nb_epoch = 5, batch_size = 100, callbacks = [early_stopping])
score = model.evaluate(X_test, y_test)
print('Test loss:', score[0])
print('Test accuracy:', score[1])

Referenced page

Deep Learning Library Keras Keras Documentation

MNIST (DCNN) with Keras (TensorFlow backend)