python、tensorflowにてDCGAN動かしたい

Question

前提・実現したいこと

tensorflowでDCGANを組む。
データセットにはfashion_mnistを用いる。

発生している問題・エラーメッセージ

-----------------------------------------------------------
ValueError                                Traceback (most recent call last)
<ipython-input-27-49aad1ceb5b1> in <module>
      1 dataset = Dataset(trainset, testset, scale_func=scale)
      2 
----> 3 losses, samples = train(net, dataset, epochs, batch_size, figsize=(8,8))

<ipython-input-25-2f229ab850d5> in train(net, dataset, epochs, batch_size, print_every, show_every, figsize)
     15                 batch_z = np.random.uniform(-1, 1, size=(batch_size, z_size))
     16 
---> 17                 _ = sess.run(net.d_opt, feed_dict={net.input_real: x, net.input_z: batch_z})
     18                 _ = sess.run(net.g_opt, feed_dict={net.input_z: batch_z, net.input_real: x})
     19

------------------略------------------

ValueError: Cannot feed value of shape (96, 28, 28) for Tensor 'input_real:0', which has shape '(?, 28, 28, 3)'

該当のソースコード

ソースコードが長いため、該当箇所と思われる部分を抜粋しています。

#トレーニング実行関数の定義

from __future__ import absolute_import, division, print_function, unicode_literals
 
# TensorFlow and tf.keras
import tensorflow as tf
from tensorflow import keras
 
import numpy as np
import matplotlib.pyplot as plt

print(tf.__version__)

fashion_mnist = keras.datasets.fashion_mnist
 
(train_images, train_labels), (test_images, test_labels) = fashion_mnist.load_data()

class_names = ['T-shirt/top', 'Trouser', 'Pullover', 'Dress', 'Coat',
               'Sandal', 'Shirt', 'Sneaker', 'Bag', 'Ankle boot']

trainset = train_images
testset = test_images

def train(net, dataset, epochs, batch_size, print_every=10, show_every=100, figsize=(8,8)):
    saver = tf.train.Saver()
    sample_z = np.random.uniform(-1, 1, size=(72, z_size))
    
    samples, losses = [], []
    steps = 0

    with tf.Session() as sess:
        sess.run(tf.global_variables_initializer())
        for e in range(epochs):
            for x,y in dataset.batches(batch_size):
                steps += 1
                batch_z = np.random.uniform(-1, 1, size=(batch_size, z_size))
                
                _ = sess.run(net.d_opt, feed_dict={net.input_real: x, net.input_z: batch_z})
                _ = sess.run(net.g_opt, feed_dict={net.input_z: batch_z, net.input_real: x})


#net.input_real, net.input_z : 入力次元(ノイズの次元）をプレイスホルダーに入れたもの

                if(steps % print_every  == 0):
                    train_loss_d = net.d_loss.eval({net.input_z: batch_z, net_input_real: x})
                    train_loss_g = net.g_loss.eval({net.input_z: batch_z})

                    print("epoch {}/{}:".format(e+1, epochs),
                        "D_loss: {:.4f}".format(train_loss_d),
                        "G_loss: {:.4f}".format(train_loss_g))

                    losses.append(train_loss_d, train_loss_g)

                if(steps % show_every == 0):
                    gen_samples = sess.run(generator(net.input_z, 1, reuse=True, training=False),
                                           feed_dict={net.input_z: sample_z})

                    samples.append(gen_samples)
                    _ = view_samples(-1, samples, 6, 12, figsize=figsize)

                    plt.show()

        saver.save(sess, 'generator.ckpt')

    with open('samples.pkl', 'wb') as f:
        pkl.dump(samples, f)

    return losses, samples 

def model_inputs(real_dim, z_dim):#real_dim=(28,28,3), z_dim=100(z_size)
    inputs_real = tf.placeholder(tf.float32, (None, *real_dim), name='input_real')
    inputs_z = tf.placeholder(tf.float32, (None, z_dim), name='input_z')

    return inputs_real, inputs_z

#損失関数の定義

def model_loss(input_real, input_z, output_dim, alpha=0.2):
    g_model = generator(input_z, output_dim, alpha=alpha)
    d_model_real, d_logits_real = discriminator(input_real, alpha=alpha)
    d_model_fake, d_logits_fake = discriminator(g_model, reuse=True, alpha=alpha)

    d_loss_real = tf.reduce_mean(tf.nn.sigmoid_cross_entropy_with_logits(logits=d_logits_real, labels=tf.ones_like(d_model_real)))
    d_loss_fake = tf.reduce_mean(tf.nn.sigmoid_cross_entropy_with_logits(logits=d_logits_fake, labels=tf.zeros_like(d_model_fake)))
    g_loss = tf.reduce_mean(tf.nn.sigmoid_cross_entropy_with_logits(logits=d_logits_fake, labels=tf.ones_like(d_model_fake)))

    d_loss = d_loss_real + d_loss_fake

    return d_loss, g_loss


#最適化関数の定義

def model_opt(d_loss, g_loss, learning_rate, beta1):
    t_vars = tf.trainable_variables()
    d_vars = [var for var in t_vars if var.name.startswith('discriminator')]
    g_vars = [var for var in t_vars if var.name.startswith('generator')]

    with tf.control_dependencies(tf.get_collection(tf.GraphKeys.UPDATE_OPS)):
        d_train_opt = tf.train.AdamOptimizer(learning_rate, beta1=beta1).minimize(d_loss, var_list=d_vars)
        g_train_opt = tf.train.AdamOptimizer(learning_rate, beta1=beta1).minimize(g_loss, var_list=g_vars)
        
    return d_train_opt, g_train_opt

class GAN:
    def __init__(self, real_size, z_size, learning_rate, alpha=0.2, beta1=0.5):
        tf.reset_default_graph()

        self.input_real, self.input_z = model_inputs(real_size, z_size)
        self.d_loss, self.g_loss = model_loss(self.input_real, self.input_z, real_size[2], alpha=alpha)#リアルサイズ[2]ここでoutput_dimを定義している
        self.d_opt, self.g_opt = model_opt(self.d_loss, self.g_loss, learning_rate, beta1)

#ハイパーパラメータの初期化とトレーニングの実行


real_size = (28, 28, 3)
# real_size = (32, 32, 3)
z_size = 100
learning_rate = 0.0002
batch_size = 128
epochs = 25
alpha = 0.2
beta1 = 0.5

net = GAN(real_size, z_size, learning_rate, alpha=alpha, beta1=beta1)
dataset = Dataset(trainset, testset, scale_func=scale)

losses, samples = train(net, dataset, epochs, batch_size, figsize=(8,8))

試したこと

placeholderがおかしいのかと思い、いろいろ調整してみましたが解決できませんでした。
また、エラーにあるshape '(?, 28, 28, 3)'は定義しているreal_sizeを指していると思うのですが、
shape (96, 28, 28)が何を指しているかわかりませんでした。

解決策でなくても、情報をいただけるだけでも幸いです。
よろしくお願いします。