In [1]:
from __future__ import print_function
In [2]:
# Download helper from google tutorial
from __future__ import print_function
import gzip
import os
import urllib
import numpy
SOURCE_URL = 'http://yann.lecun.com/exdb/mnist/'
def maybe_download(filename, work_directory):
"""Download the data from Yann's website, unless it's already here."""
if not os.path.exists(work_directory):
os.mkdir(work_directory)
filepath = os.path.join(work_directory, filename)
if not os.path.exists(filepath):
filepath, _ = urllib.urlretrieve(SOURCE_URL + filename, filepath)
statinfo = os.stat(filepath)
print('Succesfully downloaded', filename, statinfo.st_size, 'bytes.')
return filepath
def _read32(bytestream):
dt = numpy.dtype(numpy.uint32).newbyteorder('>')
return numpy.frombuffer(bytestream.read(4), dtype=dt)
def extract_images(filename):
"""Extract the images into a 4D uint8 numpy array [index, y, x, depth]."""
print('Extracting', filename)
with gzip.open(filename) as bytestream:
magic = _read32(bytestream)
if magic != 2051:
raise ValueError(
'Invalid magic number %d in MNIST image file: %s' %
(magic, filename))
num_images = _read32(bytestream)
rows = _read32(bytestream)
cols = _read32(bytestream)
buf = bytestream.read(rows * cols * num_images)
data = numpy.frombuffer(buf, dtype=numpy.uint8)
data = data.reshape(num_images, rows, cols, 1)
return data
def dense_to_one_hot(labels_dense, num_classes=10):
"""Convert class labels from scalars to one-hot vectors."""
num_labels = labels_dense.shape[0]
index_offset = numpy.arange(num_labels) * num_classes
labels_one_hot = numpy.zeros((num_labels, num_classes))
labels_one_hot.flat[index_offset + labels_dense.ravel()] = 1
return labels_one_hot
def extract_labels(filename, one_hot=False):
"""Extract the labels into a 1D uint8 numpy array [index]."""
print('Extracting', filename)
with gzip.open(filename) as bytestream:
magic = _read32(bytestream)
if magic != 2049:
raise ValueError(
'Invalid magic number %d in MNIST label file: %s' %
(magic, filename))
num_items = _read32(bytestream)
buf = bytestream.read(num_items)
labels = numpy.frombuffer(buf, dtype=numpy.uint8)
if one_hot:
return dense_to_one_hot(labels)
return labels
class DataSet(object):
def __init__(self, images, labels, fake_data=False):
if fake_data:
self._num_examples = 10000
else:
assert images.shape[0] == labels.shape[0], (
"images.shape: %s labels.shape: %s" % (images.shape,
labels.shape))
self._num_examples = images.shape[0]
# Convert shape from [num examples, rows, columns, depth]
# to [num examples, rows*columns] (assuming depth == 1)
assert images.shape[3] == 1
images = images.reshape(images.shape[0],
images.shape[1] * images.shape[2])
# Convert from [0, 255] -> [0.0, 1.0].
images = images.astype(numpy.float32)
images = numpy.multiply(images, 1.0 / 255.0)
self._images = images
self._labels = labels
self._epochs_completed = 0
self._index_in_epoch = 0
@property
def images(self):
return self._images
@property
def labels(self):
return self._labels
@property
def num_examples(self):
return self._num_examples
@property
def epochs_completed(self):
return self._epochs_completed
def next_batch(self, batch_size, fake_data=False):
"""Return the next `batch_size` examples from this data set."""
if fake_data:
fake_image = [1.0 for _ in xrange(784)]
fake_label = 0
return [fake_image for _ in xrange(batch_size)], [
fake_label for _ in xrange(batch_size)]
start = self._index_in_epoch
self._index_in_epoch += batch_size
if self._index_in_epoch > self._num_examples:
# Finished epoch
self._epochs_completed += 1
# Shuffle the data
perm = numpy.arange(self._num_examples)
numpy.random.shuffle(perm)
self._images = self._images[perm]
self._labels = self._labels[perm]
# Start next epoch
start = 0
self._index_in_epoch = batch_size
assert batch_size <= self._num_examples
end = self._index_in_epoch
return self._images[start:end], self._labels[start:end]
def read_data_sets(train_dir, fake_data=False, one_hot=False):
class DataSets(object):
pass
data_sets = DataSets()
if fake_data:
data_sets.train = DataSet([], [], fake_data=True)
data_sets.validation = DataSet([], [], fake_data=True)
data_sets.test = DataSet([], [], fake_data=True)
return data_sets
TRAIN_IMAGES = 'train-images-idx3-ubyte.gz'
TRAIN_LABELS = 'train-labels-idx1-ubyte.gz'
TEST_IMAGES = 't10k-images-idx3-ubyte.gz'
TEST_LABELS = 't10k-labels-idx1-ubyte.gz'
VALIDATION_SIZE = 5000
local_file = maybe_download(TRAIN_IMAGES, train_dir)
train_images = extract_images(local_file)
local_file = maybe_download(TRAIN_LABELS, train_dir)
train_labels = extract_labels(local_file, one_hot=one_hot)
local_file = maybe_download(TEST_IMAGES, train_dir)
test_images = extract_images(local_file)
local_file = maybe_download(TEST_LABELS, train_dir)
test_labels = extract_labels(local_file, one_hot=one_hot)
validation_images = train_images[:VALIDATION_SIZE]
validation_labels = train_labels[:VALIDATION_SIZE]
train_images = train_images[VALIDATION_SIZE:]
train_labels = train_labels[VALIDATION_SIZE:]
data_sets.train = DataSet(train_images, train_labels)
data_sets.validation = DataSet(validation_images, validation_labels)
data_sets.test = DataSet(test_images, test_labels)
return data_sets
In [3]:
mnist = read_data_sets("MNIST_data/", one_hot=True)
In [4]:
mnist.test.images
Out[4]:
In [5]:
mnist.test.images.shape
Out[5]:
In [6]:
import matplotlib.pyplot as plt
%matplotlib inline
plt.imshow(mnist.test.images[0].reshape([28,28]), cmap='gray')
Out[6]:
In [7]:
from io import BytesIO
import PIL
import numpy as np
from IPython.display import display, Image
def display_img_array(ima, **kwargs):
if ima.dtype == np.float32 or ima.dtype == np.float64:
ima = (ima*255).astype(np.uint8)
im = PIL.Image.fromarray(ima)
bio = BytesIO()
im.save(bio, format='png')
display(Image(bio.getvalue(), format='png', **kwargs))
print(mnist.train.num_examples)
for i in range(10):
print(mnist.train.labels[i])
display_img_array(mnist.train.images[i].reshape([28,28]), width=200)
In [8]:
import tensorflow as tf
# Interactive session (aka default session)
sess = tf.InteractiveSession()
In [9]:
from tfdot import tfdot
In [10]:
x = tf.placeholder("float", shape=[None, 784], name="x")
y_ = tf.placeholder("float", shape=[None, 10], name="y_")
In [11]:
W = tf.Variable(tf.zeros([784, 10]), name='W')
b = tf.Variable(tf.zeros([10]), name='b')
tf.initialize_all_variables().run()
In [12]:
tfdot()
Out[12]:
In [13]:
y = tf.nn.softmax(tf.matmul(x,W)+b, name="y_softmax")
cross_entropy = -tf.reduce_sum(y_*tf.log(y))
In [14]:
tfdot()
Out[14]:
In [15]:
train_step = tf.train.GradientDescentOptimizer(0.01).minimize(cross_entropy)
In [16]:
tfdot()
Out[16]:
In [17]:
for i in range(1000):
batch = mnist.train.next_batch(50)
train_step.run(feed_dict={x:batch[0], y_:batch[1]})
In [18]:
y.eval(feed_dict={x: mnist.test.images[:10]})
Out[18]:
In [19]:
prediction = tf.argmax(y, 1)
In [20]:
# print predictions
prediction.eval(feed_dict={x: mnist.test.images[:10]})
Out[20]:
In [21]:
# print labels
np.argmax(mnist.test.labels[:10],1)
Out[21]:
In [22]:
display_img_array(mnist.test.images[8].reshape(28,28), width=100)
In [23]:
correct_prediction = tf.equal(tf.argmax(y,1), tf.argmax(y_, 1))
In [24]:
correct_prediction.eval(feed_dict={x: mnist.test.images[:10] , y_: mnist.test.labels[:10]})
Out[24]:
In [25]:
accuracy = tf.reduce_mean(tf.cast(correct_prediction, "float"))
In [26]:
accuracy.eval(feed_dict={x: mnist.test.images[:10] , y_: mnist.test.labels[:10]})
Out[26]:
In [27]:
accuracy.eval(feed_dict={x: mnist.test.images , y_: mnist.test.labels})
Out[27]:
In [28]:
for t in range(10):
for i in range(1000):
batch = mnist.train.next_batch(200)
train_step.run(feed_dict={x:batch[0], y_:batch[1]})
a = accuracy.eval(feed_dict={x: mnist.validation.images , y_: mnist.validation.labels})
print (t, a)
In [29]:
accuracy.eval(feed_dict={x: mnist.test.images , y_: mnist.test.labels})
Out[29]:
91% accuracy on MNIST is bad. It's almost embarrassingly bad.
http://rodrigob.github.io/are_we_there_yet/build/classification_datasets_results.html
In [30]:
# reset session and graph
import tensorflow as tf
from tfdot import tfdot
tf.reset_default_graph()
if 'sess' in globals():
sess.close()
sess = tf.InteractiveSession()
x = tf.placeholder("float", shape=[None, 784], name="x")
y_ = tf.placeholder("float", shape=[None, 10], name="y_")
In [31]:
def weight_variable(shape):
initial = tf.truncated_normal(shape, stddev=0.1)
return tf.Variable(initial, name ='W')
def bias_variable(shape):
initial = tf.constant(0.1, shape=shape)
return tf.Variable(initial, name = 'b')
In [32]:
def conv2d(x, W):
return tf.nn.conv2d(x, W, strides=[1,1,1,1], padding='SAME')
def max_pool_2x2(x):
return tf.nn.max_pool(x, ksize=[1,2,2,1], strides=[1,2,2,1], padding='SAME')
In [33]:
# fisrt layer
with tf.name_scope('conv1'):
## variables
W_conv1 = weight_variable([5,5,1,32])
b_conv1 = bias_variable([32])
## build the layer
x_image = tf.reshape(x, [-1, 28, 28, 1])
h_conv1 = tf.nn.relu(conv2d(x_image, W_conv1) + b_conv1)
h_pool1 = max_pool_2x2(h_conv1)
In [34]:
tfdot()
Out[34]:
In [35]:
# second layer
with tf.name_scope('conv2'):
## variables
W_conv2 = weight_variable([5,5,32,64])
b_conv2 = bias_variable([64])
## build the layer
h_conv2 = tf.nn.relu(conv2d(h_pool1, W_conv2) + b_conv2)
h_pool2 = max_pool_2x2(h_conv2)
In [36]:
tfdot()
Out[36]:
In [37]:
# fully-connected layer
with tf.name_scope('full'):
W_fc1 = weight_variable([7*7*64, 1024])
b_fc1 = bias_variable([1024])
h_pool2_flat = tf.reshape(h_pool2, [-1, 7*7*64])
h_fc1 = tf.nn.relu(tf.matmul(h_pool2_flat, W_fc1)+b_fc1)
In [38]:
tfdot()
Out[38]:
In [39]:
# Dropout: A Simple Way to Prevent Neural Networks from Over fitting
# https://www.cs.toronto.edu/~hinton/absps/JMLRdropout.pdf
with tf.name_scope('dropout'):
keep_prob = tf.placeholder("float")
h_fc1_drop = tf.nn.dropout(h_fc1, keep_prob)
In [40]:
# Readout
with tf.name_scope('readout'):
W_fc2 = weight_variable([1024,10])
b_fc2 = bias_variable([10])
y_conv = tf.nn.softmax(tf.matmul(h_fc1_drop, W_fc2)+b_fc2)
In [41]:
tfdot()
Out[41]:
In [42]:
cross_entropy = - tf.reduce_sum(y_*tf.log(y_conv))
train_step = tf.train.AdamOptimizer(1e-4).minimize(cross_entropy)
prediction = tf.argmax(y_conv, 1)
correct_prediction = tf.equal(prediction, tf.argmax(y_, 1))
accuracy = tf.reduce_mean(tf.cast(correct_prediction, "float"))
In [43]:
%%timeit -r 1 -n 1
tf.initialize_all_variables().run()
for i in range(20000):
batch = mnist.train.next_batch(50)
if i%100 == 0:
train_accuracy = accuracy.eval(feed_dict = {
x: batch[0], y_: batch[1], keep_prob: 1.0 })
print("step %d, training accuracy %g"%(i, train_accuracy))
train_step.run(feed_dict= {x: batch[0], y_: batch[1], keep_prob: 0.5 })
for i in range(0, mnist.test.num_examples, 1000):
print("test accuracy %g"%accuracy.eval(feed_dict={x: mnist.test.images[i:i+1000],
y_: mnist.test.labels[i:i+1000],
keep_prob: 1.0}))
In [44]:
np.mean([accuracy.eval(feed_dict={x: mnist.test.images[i:i+1000],
y_: mnist.test.labels[i:i+1000],
keep_prob: 1.0})
for i in range(0, mnist.test.num_examples, 1000)]
)
Out[44]:
In [45]:
tf.scalar_summary(accuracy.op.name, accuracy)
summary_op = tf.merge_all_summaries()
summary_writer = tf.train.SummaryWriter("log1", graph=sess.graph)
In [ ]:
%%timeit -r 1 -n 1
tf.initialize_all_variables().run()
for i in range(20000):
batch = mnist.train.next_batch(50)
if i%100 == 0:
train_accuracy = accuracy.eval(feed_dict = {
x: batch[0], y_: batch[1], keep_prob: 1.0 })
print("step %d, training accuracy %g"%(i, train_accuracy))
summary_str = sess.run(summary_op, feed_dict={x: batch[0], y_: batch[1], keep_prob: 1.0 })
summary_writer.add_summary(summary_str, i)
train_step.run(feed_dict= {x: batch[0], y_: batch[1], keep_prob: 0.5 })
for i in range(0, mnist.test.num_examples, 1000):
print("test accuracy %g"%accuracy.eval(feed_dict={x: mnist.test.images[i:i+1000],
y_: mnist.test.labels[i:i+1000],
keep_prob: 1.0}))
In [ ]:
np.mean([accuracy.eval(feed_dict={x: mnist.test.images[i:i+1000],
y_: mnist.test.labels[i:i+1000],
keep_prob: 1.0})
for i in range(0, mnist.test.num_examples, 1000)]
)
In [ ]: