Reference

Adversarial Discriminative Domain Adaptation

Data

images_train = mnist.train.images
labels_train = mnist.train.labels
images_test = mnist.test.images
labels_test = mnist.test.labels

import skimage.transform

# for train data
indices = np.random.choice(55000, 10000, replace = False)

images_train_0 = images_train[indices]
images_train_0_2d = np.reshape(images_train_0, (-1, 28, 28))
labels_train_0 = labels_train[indices]

# images_train_flip_2d = images_train_0_2d[:, :, ::-1]
# images_train_flip = np.reshape(images_train_0_2d[:, :, ::-1], (-1, 28*28))
# labels_train_flip = labels_train[indices]

# images_train_30_2d = []
# for i in range(len(images_train_0)):
#  images_train_30_2d.append(skimage.transform.rotate(images_train_0_2d[i], 30))
# images_train_30 = np.reshape(images_train_30_2d, (-1, 28*28))
# labels_train_30 = labels_train[indices] 

images_train_60_2d = []
for i in range(len(images_train_0)):
  images_train_60_2d.append(skimage.transform.rotate(images_train_0_2d[i], 60))
images_train_60 = np.reshape(images_train_60_2d, (-1, 28*28))
labels_train_60 = labels_train[indices] 

# images_train_90_2d = []
# for i in range(len(images_train_0)):
#  images_train_90_2d.append(skimage.transform.rotate(images_train_0_2d[i], 90))
# images_train_90 = np.reshape(images_train_90_2d, (-1, 28*28))
# labels_train_90 = labels_train[indices] 

# images_train_180_2d = []
# for i in range(len(images_train_0)):
#  images_train_180_2d.append(skimage.transform.rotate(images_train_0_2d[i], 180))
# images_train_180 = np.reshape(images_train_180_2d, (-1, 28*28))
# labels_train_180 = labels_train[indices] 

import skimage.transform

# for test data
indices = np.random.choice(10000, 10000, replace = False)

images_test_0 = images_test[indices]
images_test_0_2d = np.reshape(images_test_0, (-1, 28, 28))
labels_test_0 = labels_test[indices]

# images_test_flip_2d = images_test_0_2d[:, :, ::-1]
# images_test_flip = np.reshape(images_test_flip_2d, (-1, 28*28))
# labels_test_flip = labels_test[indices]

# images_test_30_2d = []
# for i in range(len(images_test_0)):
#  images_test_30_2d.append(skimage.transform.rotate(images_test_0_2d[i], 30))
# images_test_30 = np.reshape(images_test_30_2d, (-1, 28*28))
# labels_test_30 = labels_test[indices] 

images_test_60_2d = []
for i in range(len(images_test_0)):
  images_test_60_2d.append(skimage.transform.rotate(images_test_0_2d[i], 60))
images_test_60 = np.reshape(images_test_60_2d, (-1, 28*28))
labels_test_60 = labels_test[indices] 

# images_test_90_2d = []
# for i in range(len(images_test_0)):
#  images_test_90_2d.append(skimage.transform.rotate(images_test_0_2d[i], 90))
# images_test_90 = np.reshape(images_test_90_2d, (-1, 28*28))
# labels_test_90 = labels_test[indices] 

# images_test_180_2d = []
# for i in range(len(images_test_0)):
#  images_test_180_2d.append(skimage.transform.rotate(images_test_0_2d[i], 180))
# images_test_180 = np.reshape(images_test_180_2d, (-1, 28*28))
# labels_test_180 = labels_test[indices]

Sample Code

# Adversarial Discriminative Domain Adaptation

class ADDA():
  def __init__(self):
    pass

  def weight_variable(self, name, shape):
    initializer = tf.truncated_normal_initializer(mean = 0.0, stddev = 0.01, dtype = tf.float32)
    return tf.get_variable(name, shape, initializer = initializer)

  def bias_variable(self, name, shape):
    initializer = tf.constant_initializer(value = 0.0, dtype = tf.float32)
    return tf.get_variable(name, shape, initializer = initializer)
  
  def alpha_variable(self, name):
    initializer = tf.constant_initializer(value = 0.75, dtype = tf.float32)
    return tf.get_variable(name, shape = (), initializer = initializer)
  
  def generator_1(self, x, filter_size, n_filters_1, n_filters_2, n_units, keep_prob, reuse = False):
    x_reshaped = tf.reshape(x, [-1, 28, 28, 1])
    
    with tf.variable_scope('generator_1', reuse = reuse):
      w_1 = self.weight_variable('w_1', [filter_size, filter_size, 1, n_filters_1])
      b_1 = self.bias_variable('b_1', [n_filters_1])

      # conv
      conv = tf.nn.conv2d(x_reshaped, w_1, strides = [1, 2, 2, 1], padding = 'SAME') + b_1
    
      # batch norm
      #batch_mean, batch_var = tf.nn.moments(conv, [0, 1, 2])
      #conv = (conv - batch_mean) / (tf.sqrt(batch_var) + 1e-10)
    
      # relu
      conv = tf.nn.relu(conv)
    
      # max_pool
      #conv = tf.nn.max_pool(conv, ksize = [1, 2, 2, 1], strides = [1, 2, 2, 1], padding = 'SAME')
    
      w_2 = self.weight_variable('w_2', [filter_size, filter_size, n_filters_1, n_filters_2])
      b_2 = self.bias_variable('b_2', [n_filters_2])
      
      # conv
      conv = tf.nn.conv2d(conv, w_2, strides = [1, 2, 2, 1], padding = 'SAME') + b_2
    
      # batch norm
      #batch_mean, batch_var = tf.nn.moments(conv, [0, 1, 2])
      #conv = (conv - batch_mean) / (tf.sqrt(batch_var) + 1e-10)
    
      # relu
      conv = tf.nn.relu(conv)
    
      # max_pool
      #conv = tf.nn.max_pool(conv, ksize = [1, 2, 2, 1], strides = [1, 2, 2, 1], padding = 'SAME')
    
      conv_flat = tf.reshape(conv, [-1, 7 * 7 * n_filters_2])
    
      w_3 = self.weight_variable('w_3', [7 * 7 * n_filters_2, n_units])
      b_3 = self.bias_variable('b_3', [n_units])
    
      fc = tf.matmul(conv_flat, w_3) + b_3
      
      # batch norm
      #batch_mean, batch_var = tf.nn.moments(fc, [0])
    
      #fc = (fc - batch_mean) / (tf.sqrt(batch_var) + 1e-10)
    
      # dropout
      #fc = tf.nn.dropout(fc, keep_prob)
      
      # relu
      fc = tf.nn.relu(fc)
     
      # leaky relu
      #fc = tf.maximum(0.2 * fc, fc)
      
      feature = fc
      
    return feature

  def generator_t(self, x, filter_size, n_filters_1, n_filters_2, n_units, keep_prob, reuse = False):
    x_reshaped = tf.reshape(x, [-1, 28, 28, 1])
    
    with tf.variable_scope('generator_t', reuse = reuse):
      w_1 = self.weight_variable('w_1', [filter_size, filter_size, 1, n_filters_1])
      b_1 = self.bias_variable('b_1', [n_filters_1])

      # conv
      conv = tf.nn.conv2d(x_reshaped, w_1, strides = [1, 2, 2, 1], padding = 'SAME') + b_1
    
      # batch norm
      #batch_mean, batch_var = tf.nn.moments(conv, [0, 1, 2])
      #conv = (conv - batch_mean) / (tf.sqrt(batch_var) + 1e-10)
    
      # relu
      conv = tf.nn.relu(conv)

      # leaky relu
      #conv = tf.maximum(0.2 * conv, conv)

      # max_pool
      #conv = tf.nn.max_pool(conv, ksize = [1, 2, 2, 1], strides = [1, 2, 2, 1], padding = 'SAME')
    
      w_2 = self.weight_variable('w_2', [filter_size, filter_size, n_filters_1, n_filters_2])
      b_2 = self.bias_variable('b_2', [n_filters_2])
      
      # conv
      conv = tf.nn.conv2d(conv, w_2, strides = [1, 2, 2, 1], padding = 'SAME') + b_2
    
      # batch norm
      #batch_mean, batch_var = tf.nn.moments(conv, [0, 1, 2])
      #conv = (conv - batch_mean) / (tf.sqrt(batch_var) + 1e-10)
    
      # relu
      conv = tf.nn.relu(conv)
      
      # leaky relu
      #conv = tf.maximum(0.2 * conv, conv)

    
      # max_pool
      #conv = tf.nn.max_pool(conv, ksize = [1, 2, 2, 1], strides = [1, 2, 2, 1], padding = 'SAME')
    
      conv_flat = tf.reshape(conv, [-1, 7 * 7 * n_filters_2])
    
      w_3 = self.weight_variable('w_3', [7 * 7 * n_filters_2, n_units])
      b_3 = self.bias_variable('b_3', [n_units])
    
      fc = tf.matmul(conv_flat, w_3) + b_3
      
      # batch norm
      #batch_mean, batch_var = tf.nn.moments(fc, [0])
    
      #fc = (fc - batch_mean) / (tf.sqrt(batch_var) + 1e-10)
    
      # dropout
      #fc = tf.nn.dropout(fc, keep_prob)
      
      # relu
      fc = tf.nn.relu(fc)
      
      # leaky relu
      #fc = tf.maximum(0.2 * fc, fc)

      feature = fc
      
    return feature
  
  def classifier(self, x, n_units_1, n_units_2, keep_prob, reuse = False):
    
    with tf.variable_scope('classifier', reuse = reuse):
      w_1 = self.weight_variable('w_1', [n_units_1, n_units_2])
      b_1 = self.bias_variable('b_1', [n_units_2])
    
      fc = tf.matmul(x, w_1) + b_1
      
      # batch norm
      batch_mean, batch_var = tf.nn.moments(fc, [0])
      fc = (fc - batch_mean) / (tf.sqrt(batch_var) + 1e-10)
      
      # relu
      fc = tf.nn.relu(fc)
    
      # dropout
      #fc = tf.nn.dropout(fc, keep_prob)
    
      w_2 = self.weight_variable('w_2', [n_units_2, 10])
      b_2 = self.bias_variable('b_2', [10])

      fc = tf.matmul(fc, w_2) + b_2
      logits = fc
    
    return logits

  def discriminator(self, x, n_units_1, n_units_2, n_units_3, keep_prob, reuse = False):
    
    with tf.variable_scope('discriminator', reuse = reuse):
      w_1 = self.weight_variable('w_1', [n_units_1, n_units_2])
      b_1 = self.bias_variable('b_1', [n_units_2])
    
      fc = tf.matmul(x, w_1) + b_1
      
      # batch norm
      #batch_mean, batch_var = tf.nn.moments(fc, [0])
      #fc = (fc - batch_mean) / (tf.sqrt(batch_var) + 1e-10)
      
      # relu
      #fc = tf.nn.relu(fc)
      
      # leaky relu
      fc = tf.maximum(fc * 0.2, fc)
    
      w_2 = self.weight_variable('w_2', [n_units_2, n_units_3])
      b_2 = self.bias_variable('b_2', [n_units_3])

      fc = tf.matmul(fc, w_2) + b_2
      
      # batch norm
      #batch_mean, batch_var = tf.nn.moments(fc, [0])
      #fc = (fc - batch_mean) / (tf.sqrt(batch_var) + 1e-10)
      
      # relu
      #fc = tf.nn.relu(fc)
      
      # leaky relu
      fc = tf.maximum(fc * 0.2, fc)
      
      w_3 = self.weight_variable('w_3', [n_units_3, 1])
      b_3 = self.bias_variable('b_3', [1])

      fc = tf.matmul(fc, w_3) + b_3
          
      # dropout
      #fc = tf.nn.dropout(fc, keep_prob)
      
      fc = tf.nn.sigmoid(fc)
    
    return fc

  def loss_cross_entropy(self, y, t):
    cross_entropy = - tf.reduce_mean(tf.reduce_sum(t * tf.log(tf.clip_by_value(y, 1e-10, 1.0)), axis = 1))
    return cross_entropy
  
  def loss_discriminator(self, prob_1, prob_2):
    return - tf.reduce_mean(tf.log(tf.clip_by_value(prob_1, 1e-10, 1.0)) \
                                +tf.log(tf.clip_by_value(1.0 - prob_2, 1e-10, 1.0))) 
    #return - tf.reduce_mean(0.5 * tf.log(tf.clip_by_value(prob_1, 1e-10, 1.0)) \
    #                            + 0.5 * tf.log(tf.clip_by_value(1.0 - prob_2, 1e-10, 1.0))) 
    #return tf.reduce_mean(tf.square(prob_1 - 1.0)) + tf.reduce_mean(tf.square(prob_2))
  
  def loss_generator(self, prob):
    return - tf.reduce_mean(tf.log(tf.clip_by_value(prob, 1e-10, 1.0)))
    #return tf.reduce_mean(tf.square(prob - 1.0))

  def loss_entropy(self, p):
    entropy = - tf.reduce_mean(tf.reduce_sum(p * tf.log(tf.clip_by_value(p, 1e-10, 1.0)), axis = 1))
    return entropy

  def loss_mutual_information(self, p):
    p_ave = tf.reduce_mean(p, axis = 0)
    h_y = -tf.reduce_sum(p_ave * tf.log(p_ave + 1e-16))
    h_y_x = - tf.reduce_mean(tf.reduce_sum(p * tf.log(tf.clip_by_value(p, 1e-10, 1.0)), axis = 1))
    mutual_info = h_y - h_y_x
    return -mutual_info

  def accuracy(self, y, t):
    correct_preds = tf.equal(tf.argmax(y, axis = 1), tf.argmax(t, axis = 1))
    accuracy = tf.reduce_mean(tf.cast(correct_preds, tf.float32))
    return accuracy

  def training(self, loss, learning_rate, var_list):
    optimizer = tf.train.AdamOptimizer(learning_rate = learning_rate)
    #optimizer = tf.train.GradientDescentOptimizer(learning_rate = learning_rate)
    train_step = optimizer.minimize(loss, var_list = var_list)
    return train_step

  def training_clipped(self, loss, learning_rate, clip_norm, var_list):
    optimizer = tf.train.AdamOptimizer(learning_rate = learning_rate)

    grads_and_vars = optimizer.compute_gradients(loss, var_list = var_list)
    clipped_grads_and_vars = [(tf.clip_by_norm(grad, clip_norm = clip_norm), \
                             var) for grad, var in grads_and_vars]
    train_step = optimizer.apply_gradients(clipped_grads_and_vars)

    return train_step

  def fit(self, images_train_1, labels_train_1, images_test_1, labels_test_1, \
          images_train_t, labels_train_t, images_test_t, labels_test_t, \
          filter_size, n_filters_1, n_filters_2, n_units_g, n_units_c, \
          n_units_d_1, n_units_d_2, learning_rate, n_iter_1, n_iter_2, \
          batch_size, show_step_1, show_step_2, is_saving, model_path):

    tf.reset_default_graph()

    x_1 = tf.placeholder(shape = [None, 28 * 28], dtype = tf.float32)
    y_1 = tf.placeholder(shape = [None, 10], dtype = tf.float32)
    x_t = tf.placeholder(shape = [None, 28 * 28], dtype = tf.float32)
    y_t = tf.placeholder(shape = [None, 10], dtype = tf.float32)
    keep_prob = tf.placeholder(shape = (), dtype = tf.float32)

    feat_1 = self.generator_1(x_1, filter_size, n_filters_1, n_filters_2, n_units_g, \
                            keep_prob, reuse = False)
    feat_t = self.generator_t(x_t, filter_size, n_filters_1, n_filters_2, n_units_g, \
                            keep_prob, reuse = False)
    
    logits_1 = self.classifier(feat_1, n_units_g, n_units_c, keep_prob, reuse = False)
    probs_l_1 = tf.nn.softmax(logits_1)
    loss_1 = self.loss_cross_entropy(probs_l_1, y_1)

    logits_t = self.classifier(feat_t, n_units_g, n_units_c, keep_prob, reuse = True)
    probs_l_t = tf.nn.softmax(logits_t)
    loss_t = self.loss_cross_entropy(probs_l_t, y_t)
    
    prob_d_1 = self.discriminator(feat_1, n_units_g, n_units_d_1, n_units_d_2, \
                                    keep_prob, reuse = False)
    prob_d_t = self.discriminator(feat_t, n_units_g, n_units_d_1, n_units_d_2, \
                                    keep_prob, reuse = True)
    prob_d_1_ave = tf.reduce_mean(prob_d_1)
    prob_d_t_ave = tf.reduce_mean(prob_d_t)
    loss_d = self.loss_discriminator(prob_d_1, prob_d_t)
    loss_g_t = self.loss_generator(prob_d_t)
    
    var_list_g_1 = tf.trainable_variables('generator_1')
    var_list_g_t = tf.trainable_variables('generator_t')
    var_list_c = tf.trainable_variables('classifier')
    var_list_d = tf.trainable_variables('discriminator')
    
    var_list_g_1_c = var_list_g_1 + var_list_c
    var_list_g_t_c = var_list_g_t + var_list_c
    
    # Without Gradient Clipping
    train_step_g_1_c = self.training(loss_1, learning_rate, var_list_g_1_c)
    train_step_g_t_c = self.training(loss_t, learning_rate, var_list_g_t_c)
    train_step_d = self.training(loss_d, learning_rate, var_list_d)
    train_step_g_t = self.training(loss_g_t, learning_rate, var_list_g_t)
    
    # With Gradient Clipping
    #train_step_1 = self.training_clipped(loss_1, learning_rate, 0.1, var_list_0_1)
    #train_step_t = self.training_clipped(loss_t, learning_rate, 0.1, var_list_0_t)
    
    acc_1 =  self.accuracy(probs_l_1, y_1)
    acc_t =  self.accuracy(probs_l_t, y_t)
    
    init = tf.global_variables_initializer()
    saver = tf.train.Saver()

    with tf.Session() as sess:

      sess.run(init)
      
      print ('-' * 45)
      print ('Pre-training for generator_1, generator_t and classifier') 
      print ('-' * 45)

      # for 1
      history_loss_train_1 = []
      history_loss_test_1 = []
      history_acc_train_1 = []
      history_acc_test_1 = []

      for i in range(n_iter_1):
        # Train
        rand_index = np.random.choice(len(images_train_1), size = batch_size)
        x_batch = images_train_1[rand_index]
        y_batch = labels_train_1[rand_index]

        feed_dict = {x_1: x_batch, y_1: y_batch, keep_prob: 1.0}

        sess.run(train_step_g_1_c, feed_dict = feed_dict)

        temp_loss = sess.run(loss_1, feed_dict = feed_dict)
        temp_acc = sess.run(acc_1, feed_dict = feed_dict)

        history_loss_train_1.append(temp_loss)
        history_acc_train_1.append(temp_acc)

        if (i + 1) % show_step_1 == 0:
          print ('-' * 15)
          print ('Iteration: ' + str(i + 1) + '  Loss_1: ' + str(temp_loss) + \
                '  Accuracy_1: ' + str(temp_acc))

        # Test
        rand_index = np.random.choice(len(images_test_1), size = batch_size)
        x_batch = images_test_1[rand_index]
        y_batch = labels_test_1[rand_index]

        feed_dict = {x_1: x_batch, y_1: y_batch, keep_prob: 1.0}

        temp_loss = sess.run(loss_1, feed_dict = feed_dict)
        temp_acc = sess.run(acc_1, feed_dict = feed_dict)

        history_loss_test_1.append(temp_loss)
        history_acc_test_1.append(temp_acc)

      print ('-'* 15)    
      fig = plt.figure(figsize = (10, 3))
      ax1 = fig.add_subplot(1, 2, 1)
      ax1.plot(range(n_iter_1), history_loss_train_1, 'b-', label = 'Train')
      ax1.plot(range(n_iter_1), history_loss_test_1, 'r--', label = 'Test')
      ax1.set_title('Loss_1')
      ax1.legend(loc = 'upper right')

      ax2 = fig.add_subplot(1, 2, 2)
      ax2.plot(range(n_iter_1), history_acc_train_1, 'b-', label = 'Train')
      ax2.plot(range(n_iter_1), history_acc_test_1, 'r--', label = 'Test')
      ax2.set_ylim(0.0, 1.0)
      ax2.set_title('Accuracy_1')
      ax2.legend(loc = 'lower right')

      plt.show()
      
      # for t  
      history_loss_train_t = []
      history_loss_test_t = []
      history_acc_train_t = []
      history_acc_test_t = []

      for i in range(n_iter_1):
        # Train
        rand_index = np.random.choice(len(images_train_t), size = batch_size)
        x_batch = images_train_t[rand_index]
        y_batch = labels_train_t[rand_index]

        feed_dict = {x_t: x_batch, y_t: y_batch, keep_prob: 1.0}

        sess.run(train_step_g_t_c, feed_dict = feed_dict)

        temp_loss = sess.run(loss_t, feed_dict = feed_dict)
        temp_acc = sess.run(acc_t, feed_dict = feed_dict)

        history_loss_train_t.append(temp_loss)
        history_acc_train_t.append(temp_acc)

        if (i + 1) % show_step_1 == 0:
          print ('-' * 15)
          print ('Iteration: ' + str(i + 1) + '  Loss_t: ' + str(temp_loss) + \
                '  Accuracy_t: ' + str(temp_acc))

        # Test
        rand_index = np.random.choice(len(images_test_t), size = batch_size)
        x_batch = images_test_t[rand_index]
        y_batch = labels_test_t[rand_index]

        feed_dict = {x_t: x_batch, y_t: y_batch, keep_prob: 1.0}

        temp_loss = sess.run(loss_t, feed_dict = feed_dict)
        temp_acc = sess.run(acc_t, feed_dict = feed_dict)

        history_loss_test_t.append(temp_loss)
        history_acc_test_t.append(temp_acc)
        
      print ('-'* 15)    
      fig = plt.figure(figsize = (10, 3))
      ax1 = fig.add_subplot(1, 2, 1)
      ax1.plot(range(n_iter_1), history_loss_train_t, 'b-', label = 'Train')
      ax1.plot(range(n_iter_1), history_loss_test_t, 'r--', label = 'Test')
      ax1.set_title('Loss_t')
      ax1.legend(loc = 'upper right')

      ax2 = fig.add_subplot(1, 2, 2)
      ax2.plot(range(n_iter_1), history_acc_train_t, 'b-', label = 'Train')
      ax2.plot(range(n_iter_1), history_acc_test_t, 'r--', label = 'Test')
      ax2.set_ylim(0.0, 1.0)
      ax2.set_title('Accuracy_t')
      ax2.legend(loc = 'lower right')

      plt.show()
      
      ###################################################
      print ('-' * 45)
      print ('Adversarial adaptation for generator_t and discriminator') 
      print ('-' * 45)

      history_loss_g_t_train = []
      history_loss_d_train = []
      history_loss_g_t_test = []
      history_loss_d_test = []
      history_prob_d_1_train = []
      history_prob_d_t_train = []
      history_prob_d_1_test = []
      history_prob_d_t_test = []

      for i in range(n_iter_2):
        # Train
        rand_index = np.random.choice(len(images_train_1), size = batch_size)
        x_batch_1 = images_train_1[rand_index]
        y_batch_1 = labels_train_1[rand_index]

        rand_index = np.random.choice(len(images_train_t), size = batch_size)
        x_batch_t = images_train_t[rand_index]
        y_batch_t = labels_train_t[rand_index]

        feed_dict = {x_1: x_batch_1, y_1: y_batch_1, x_t: x_batch_t, y_t: y_batch_t, keep_prob: 1.0}

        train_g = True
        train_d = True
        
        temp_loss_g_t = sess.run(loss_g_t, feed_dict = feed_dict)
        temp_loss_d = sess.run(loss_d, feed_dict = feed_dict)

        #if temp_loss_g_t * 1.5 < temp_loss_d:
        #  train_g = False
        if temp_loss_d * 0.1 < temp_loss_g_t:
          train_d = False
          
        #if i < 100:
        #  train_d = False
          
        if train_g:
          sess.run(train_step_g_t, feed_dict = feed_dict)
          sess.run(train_step_g_t, feed_dict = feed_dict)
        if train_d:
          sess.run(train_step_d, feed_dict = feed_dict)
          
        #sess.run(train_step_g_t, feed_dict = feed_dict)
        #sess.run(train_step_g_t, feed_dict = feed_dict)
        #sess.run(train_step_d, feed_dict = feed_dict)

        temp_prob_d_1 = sess.run(prob_d_1_ave, feed_dict = feed_dict)
        temp_prob_d_t = sess.run(prob_d_t_ave, feed_dict = feed_dict)
        temp_loss_g_t = sess.run(loss_g_t, feed_dict = feed_dict)
        temp_loss_d = sess.run(loss_d, feed_dict = feed_dict)

        history_prob_d_1_train.append(temp_prob_d_1)
        history_prob_d_t_train.append(temp_prob_d_t)
        history_loss_g_t_train.append(temp_loss_g_t)
        history_loss_d_train.append(temp_loss_d)

        if (i + 1) % show_step_2 == 0:
          print ('-' * 15)
          print ('Iteration: ' + str(i + 1) + '  Loss_train_g_t: ' + str(temp_loss_g_t) + \
                '  Loss_train_d: ' + str(temp_loss_d))

        # Test
        rand_index = np.random.choice(len(images_test_1), size = batch_size)
        x_batch_1 = images_test_1[rand_index]
        y_batch_1 = labels_test_1[rand_index]

        rand_index = np.random.choice(len(images_test_t), size = batch_size)
        x_batch_t = images_test_t[rand_index]
        y_batch_t = labels_test_t[rand_index]

        feed_dict = {x_1: x_batch_1, y_1: y_batch_1, x_t: x_batch_t, y_t: y_batch_t, keep_prob: 1.0}

        temp_prob_d_1 = sess.run(prob_d_1_ave, feed_dict = feed_dict)
        temp_prob_d_t = sess.run(prob_d_t_ave, feed_dict = feed_dict)
        temp_loss_g_t = sess.run(loss_g_t, feed_dict = feed_dict)
        temp_loss_d = sess.run(loss_d, feed_dict = feed_dict)
        
        history_prob_d_1_test.append(temp_prob_d_1)
        history_prob_d_t_test.append(temp_prob_d_t)
        history_loss_g_t_test.append(temp_loss_g_t)
        history_loss_d_test.append(temp_loss_d)
        
      print ('-'* 15)    
      fig = plt.figure(figsize = (10, 3))
      ax1 = fig.add_subplot(1, 2, 1)
      ax1.plot(range(n_iter_2), history_loss_g_t_train, 'b-', label = 'Train')
      ax1.plot(range(n_iter_2), history_loss_g_t_test, 'r--', label = 'Test')
      ax1.set_title('Loss_generator_t')
      ax1.legend(loc = 'upper right')

      ax2 = fig.add_subplot(1, 2, 2)
      ax2.plot(range(n_iter_2), history_loss_d_train, 'b-', label = 'Train')
      ax2.plot(range(n_iter_2), history_loss_d_test, 'r--', label = 'Test')
      ax2.set_title('Loss_discriminator')
      ax2.legend(loc = 'upper right')
      
      plt.show()

      print ('-'* 15)    
      fig = plt.figure(figsize = (10, 3))
      ax3 = fig.add_subplot(1, 2, 1)
      ax3.plot(range(n_iter_2), history_prob_d_1_train, 'b-', label = 'Train')
      ax3.plot(range(n_iter_2), history_prob_d_1_test, 'r--', label = 'Test')
      ax3.set_title('Probability_d_1')
      ax3.legend(loc = 'lower right')

      ax4 = fig.add_subplot(1, 2, 2)
      ax4.plot(range(n_iter_2), history_prob_d_t_train, 'b-', label = 'Train')
      ax4.plot(range(n_iter_2), history_prob_d_t_test, 'r--', label = 'Test')
      ax4.set_title('Probability_d_t')
      ax4.legend(loc = 'lower right')

      plt.show()
      
      ###################################################
      print ('-' * 45)
      print ('Testing') 
      print ('-' * 45)
      
      # for t  
      history_loss_test_t_2 = []
      history_acc_test_t_2 = []

      for i in range(n_iter_1):
        # Test
        rand_index = np.random.choice(len(images_test_t), size = batch_size)
        x_batch = images_test_t[rand_index]
        y_batch = labels_test_t[rand_index]

        feed_dict = {x_t: x_batch, y_t: y_batch, keep_prob: 1.0}

        temp_loss = sess.run(loss_t, feed_dict = feed_dict)
        temp_acc = sess.run(acc_t, feed_dict = feed_dict)

        history_loss_test_t_2.append(temp_loss)
        history_acc_test_t_2.append(temp_acc)
        
      print ('-'* 15)    
      fig = plt.figure(figsize = (10, 3))
      ax1 = fig.add_subplot(1, 2, 1)
      #ax1.plot(range(n_iter_1), history_loss_train_t, 'b-', label = 'Train')
      ax1.plot(range(n_iter_1), history_loss_test_t_2, 'r--', label = 'Test')
      ax1.set_title('Loss_t')
      ax1.legend(loc = 'upper right')

      ax2 = fig.add_subplot(1, 2, 2)
      #ax2.plot(range(n_iter_1), history_acc_train_t, 'b-', label = 'Train')
      ax2.plot(range(n_iter_1), history_acc_test_t_2, 'r--', label = 'Test')
      ax2.set_ylim(0.0, 1.0)
      ax2.set_title('Accuracy_t')
      ax2.legend(loc = 'lower right')

      plt.show()
      
      if is_saving:
        model_path = saver.save(sess, model_path)
        print ('done saving at ', model_path)

Parameters

filter_size = 3
n_filters_1 = 32
n_filters_2 = 32
n_units_g = 256
n_units_c = 128
n_units_d_1 = 64
n_units_d_2 = 64
learning_rate = 0.001
n_iter_1 = 100
n_iter_2 = 300
batch_size = 64
show_step_1 = 50
show_step_2 = 100
model_path = 'datalab/model'

Output

images_train_1 = images_train_0
labels_train_1 = labels_train_0
images_test_1 = images_test_0
labels_test_1 = labels_test_0

images_train_t = images_train_60
labels_train_t = labels_train_60
images_test_t = images_test_60
labels_test_t = labels_test_60

is_saving = False

adda.fit(images_train_1, labels_train_1, images_test_1, labels_test_1, \
        images_train_t, labels_train_t, images_test_t, labels_test_t, \
        filter_size, n_filters_1, n_filters_2, n_units_g, n_units_c, n_units_d_1, n_units_d_2, \
        learning_rate, n_iter_1, n_iter_2, batch_size, show_step_1, show_step_2, is_saving, model_path)