如何控制由在MNIST上训练的GAN生成的数字？

要控制由在MNIST上训练的GAN生成的数字，你可以使用条件生成对抗网络（Conditional GANs）。条件GAN允许你通过向生成器提供额外的输入来控制生成的图像。

下面是一个示例代码，展示如何使用条件GAN生成特定数字的MNIST图像：

import tensorflow as tf
from tensorflow.keras.datasets import mnist
from tensorflow.keras.layers import Input, Dense, Reshape, Flatten, Embedding, multiply
from tensorflow.keras.layers import BatchNormalization, Activation, LeakyReLU
from tensorflow.keras.models import Sequential, Model
from tensorflow.keras.optimizers import Adam

# 加载MNIST数据集
(x_train, y_train), (_, _) = mnist.load_data()

# 数据预处理
x_train = (x_train.astype('float32') - 127.5) / 127.5
y_train = y_train.reshape(-1, 1)

# 定义生成器
def build_generator():
    model = Sequential()
    model.add(Dense(256, input_dim=100))
    model.add(LeakyReLU(alpha=0.2))
    model.add(BatchNormalization(momentum=0.8))
    model.add(Dense(512))
    model.add(LeakyReLU(alpha=0.2))
    model.add(BatchNormalization(momentum=0.8))
    model.add(Dense(784, activation='tanh'))
    model.add(Reshape((28, 28, 1)))
    noise = Input(shape=(100,))
    label = Input(shape=(1,), dtype='int32')
    label_embedding = Flatten()(Embedding(10, 100)(label))
    model_input = multiply([noise, label_embedding])
    img = model(model_input)
    return Model([noise, label], img)

# 定义判别器
def build_discriminator():
    model = Sequential()
    model.add(Dense(512, input_dim=784))
    model.add(LeakyReLU(alpha=0.2))
    model.add(Dense(256))
    model.add(LeakyReLU(alpha=0.2))
    model.add(Dense(1, activation='sigmoid'))
    img = Input(shape=(28, 28, 1))
    label = Input(shape=(1,), dtype='int32')
    label_embedding = Flatten()(Embedding(10, 784)(label))
    flat_img = Flatten()(img)
    model_input = multiply([flat_img, label_embedding])
    validity = model(model_input)
    return Model([img, label], validity)

# 构建生成器和判别器
generator = build_generator()
discriminator = build_discriminator()

# 编译和训练判别器
discriminator.compile(loss='binary_crossentropy',
                      optimizer=Adam(0.0002, 0.5),
                      metrics=['accuracy'])

# 冻结判别器的权重
discriminator.trainable = False

# 定义GAN模型
noise = Input(shape=(100,))
label = Input(shape=(1,))
img = generator([noise, label])
valid = discriminator([img, label])
combined = Model([noise, label], valid)
combined.compile(loss='binary_crossentropy', optimizer=Adam(0.0002, 0.5))

# 训练GAN模型
epochs = 30000
batch_size = 32

for epoch in range(epochs):
    # 训练判别器
    idx = np.random.randint(0, x_train.shape[0], batch_size)
    imgs, labels = x_train[idx], y_train[idx]

    noise = np.random.normal(0, 1, (batch_size, 100))
    gen_imgs = generator.predict([noise, labels])

    d_loss_real = discriminator.train_on_batch([imgs, labels], np.ones((batch_size, 1)))
    d_loss_fake = discriminator.train_on_batch([gen_imgs, labels], np.zeros((batch_size, 1)))
    d_loss = 0.5 * np.add(d_loss_real, d_loss_fake)

    # 训练生成器
    noise = np.random.normal(0, 1, (batch_size, 100))
    valid_y = np.ones((batch_size, 1))

    g_loss = combined.train_on_batch([noise, labels], valid_y)

    # 每1000个epoch打印一次进度
    if epoch % 1000 == 0:
        print(f"Epoch: {epoch}, 判别器损失: {d_loss[0]}, 生成器损失: {g_loss}")

# 生成指定数字的图像
def generate_digit(digit