def loadFullFashion_MNSIT(shift):
imgs, y, num_train_data = data.mnist_load('Fashion_MNIST', shift=shift)
imgs_t, y_t, num_train_data_t = data.mnist_load('Fashion_MNIST',
dataset='test',
shift=shift)
imgs.shape = imgs.shape + (1, )
imgs_t.shape = imgs_t.shape + (1, )
return np.concatenate((imgs, imgs_t)), np.concatenate((y, y_t))
def getFashion_MNISTDatapool(batch_size, shift, keep=None):
if keep is None:
imgs, y, num_train_data = data.mnist_load('fmnist',shift=shift)
else:
imgs, y, num_train_data = data.mnist_load('fmnist', keep=keep, shift=shift)
print "Total number of training dataset: " + str(num_train_data)
imgs.shape = imgs.shape + (1,)
data_pool = utils.MemoryData({'img': imgs, 'label':y}, batch_size)
return data_pool
def getMNISTDatapool(batch_size, keep=None, shift=True):
if keep is None:
imgs, _, num_train_data = data.mnist_load('MNIST_data', shift=shift)
else:
imgs, _, num_train_data = data.mnist_load('MNIST_data',
keep=keep,
shift=shift)
print "Total number of training data: " + str(num_train_data)
imgs.shape = imgs.shape + (1, )
data_pool = utils.MemoryData({'img': imgs}, batch_size)
return data_pool
def getFullFashion_MNISTDatapool(batch_size, shift,keep=None):
if keep is None:
imgs, y, num_train_data = data.mnist_load('fmnist',shift=shift)
imgs_t, y_t, num_train_data_t = data.mnist_load('fmnist',dataset='test',shift=shift)
else:
imgs, y, num_train_data = data.mnist_load('fmnist', keep=keep, shift=shift)
imgs_t, y_t, num_train_data_t = data.mnist_load('fmnist', keep=keep,dataset='test',shift = shift)
print "Total number of training dataset: " + str(num_train_data + num_train_data_t)
imgs.shape = imgs.shape + (1,)
imgs_t.shape = imgs_t.shape + (1,)
data_pool = utils.MemoryData({'img': np.concatenate((imgs,imgs_t)), 'label':np.concatenate((y,y_t))}, batch_size)
return data_pool
import tensorflow as tf
import data_mnist as data
import models_mnist as models
""" param """
epoch = 50
batch_size = 64
lr = 0.0002
z_dim = 100
gpu_id = 3
''' data '''
utils.mkdir('./data/mnist/')
data.mnist_download('./data/mnist')
imgs, _, _ = data.mnist_load('./data/mnist')
imgs.shape = imgs.shape + (1,)
data_pool = utils.MemoryData({'img': imgs}, batch_size)
""" graphs """
with tf.device('/gpu:%d' % gpu_id):
''' models '''
generator = models.generator
discriminator = models.discriminator
''' graph '''
# inputs
real = tf.placeholder(tf.float32, shape=[None, 28, 28, 1])
z = tf.placeholder(tf.float32, shape=[None, z_dim])
#test code snippet here
import data_mnist as data
import numpy as np
from matplotlib import pyplot as plt
import tensorflow as tf
import my_utils
#extracting subset of mnist
if 0:
imgs, labels, _ = data.mnist_load('MNIST_data') #one-hot = False
for l in range(10):
print labels[l]
keep = [0, 9]
X, Y = [], []
for x, y in zip(imgs, labels):
if y in keep:
X.append(x)
Y.append(y)
X = np.array(X)
Y = np.array(Y)
for l in range(10):
img = np.reshape(X[l], [28, 28])
plt.imshow(img, cmap='gray')
print Y[l]
plt.show()
#convert labels to one-hot
if 0:
relu = tf.nn.relu
fc_relu = partial(fc, activation_fn=relu)
real = tf.placeholder(tf.float32, shape=[None, 784])
with tf.variable_scope("encoder", reuse=False):
y = fc_relu(real, 500, scope='layer1')
with tf.variable_scope("encoder", reuse=True):
weights = tf.get_variable('layer1/weights')
biases = tf.get_variable('layer1/biases')
print(weights.shape)
print(biases.shape)
a = 0
if 0:
imgs, _, _ = data.mnist_load('MNIST_data')
a = 0
if 0:
batch_size = 100
z_dim = 10
n_cen = 5
z = tf.get_variable('z', shape=(batch_size, z_dim))
theta_p = tf.get_variable('theta_p', shape=(n_cen))
u_p = tf.get_variable('u_p', shape=(n_cen, z_dim))
lambda_p = tf.get_variable('lambda_p', shape=(n_cen, z_dim))
theta_p_tensor = tf.expand_dims(theta_p, 0)
theta_p_tensor = tf.expand_dims(theta_p_tensor, 2)
theta_p_tensor = tf.tile(theta_p_tensor, [batch_size, 1, 1])
u_p_tensor = tf.tile(tf.expand_dims(u_p, 0), [batch_size, 1, 1])
