学习DCGAN网络的时候遇到的错误代码
在用keras学习DCGAN网络的时候遇到如下的错误代码：
tensorflow.python.framework.errors_impl.FailedPreconditionError: Error while reading resource variable _AnonymousVar33 from Container: localhost. This could mean that the variable was uninitialized. Not found: Resource localhost/_AnonymousVar33/N10tensorflow3VarE does not exist.
[[node mul_1/ReadVariableOp (defined at /Users/xxx/anaconda3/envs/tensorflow/lib/python3.6/site-packages/tensorflow_core/python/framework/ops.py:1751) ]] [Op:__inference_keras_scratch_graph_2262]
Function call stack:
keras_scratch_graph
这是因为作者的代码是用tensorflow 1.x的版本写的，而我们本地的环境是tensorflow2.0及以上，出现了不兼容问题，可以解决的一种方法是在头部添加以下代码：
-
import tensorflow.compat.v1 as tf #使用1.0版本的方法
-
tf.disable_v2_behavior() #禁用2.0版本的方法
通过对tensorflow2.0降级的方式来运行代码。
当然也可以通过对旧代码更改，调用tf.Session.run（）方法的方式来使旧代码适配新的tensorflow版本，相关资料较多此处不做详细介绍。
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
问题描述
在学习DCGAN时，遇到如下警告:
keras UserWarning: Discrepancy between trainable weights and collected trainable weigh...
报错位置：[line 138] d_loss_real = self.discriminator.train_on_batch(imgs, valid)
问题的官网描述：在实例化之后将网络层的 trainable 属性设置为 True 或 False。为了使之生效，在修改 trainable 属性之后，需要在模型上调用 compile()。
解决方法
构造一个新的frozen_D 替代 combined 中的 discriminator 。
参考keras DCGAN中的代码。
代码基于 eriklindernoren/Keras-GAN ，并修改了trainable与compile 易于混淆的代码。
from keras.datasets import mnist
from keras.layers import Input, Dense, Reshape, Flatten, Dropout
from keras.layers import BatchNormalization, Activation, ZeroPadding2D
from keras.layers.advanced_activations import LeakyReLU
from keras.layers.convolutional import UpSampling2D, Conv2D
from keras.models import Sequential, Model
from keras.optimizers import Adam
import matplotlib.pyplot as plt
import numpy as np
class DCGAN():
def __init__(self):
# Input shape
self.img_rows = 28
self.img_cols = 28
self.channels = 1
self.img_shape = (self.img_rows, self.img_cols, self.channels)
self.latent_dim = 100
optimizer = Adam(0.0002, 0.5)
base_generator = self.build_generator()
base_discriminator = self.build_discriminator()
########
self.generator = Model(
inputs=base_generator.inputs,
outputs=base_generator.outputs)
self.discriminator = Model(
inputs=base_discriminator.inputs,
outputs=base_discriminator.outputs)
self.discriminator.compile(loss='binary_crossentropy',
optimizer=optimizer,
metrics=['accuracy'])
frozen_D = Model(
inputs=base_discriminator.inputs,
outputs=base_discriminator.outputs)
frozen_D.trainable = False
z = Input(shape=(self.latent_dim,))
img = self.generator(z)
valid = frozen_D(img)
self.combined = Model(z, valid)
self.combined.compile(loss='binary_crossentropy', optimizer=optimizer)
def build_generator(self):
model = Sequential()
model.add(
Dense(
128 * 7 * 7,
activation="relu",
input_dim=self.latent_dim))
model.add(Reshape((7, 7, 128)))
model.add(UpSampling2D())
model.add(Conv2D(128, kernel_size=3, padding="same"))
model.add(BatchNormalization(momentum=0.8))
model.add(Activation("relu"))
model.add(UpSampling2D())
model.add(Conv2D(64, kernel_size=3, padding="same"))
model.add(BatchNormalization(momentum=0.8))
model.add(Activation("relu"))
model.add(Conv2D(self.channels, kernel_size=3, padding="same"))
model.add(Activation("tanh"))
model.summary()
return model
def build_discriminator(self):
model = Sequential()
model.add(
Conv2D(
32,
kernel_size=3,
strides=2,
input_shape=self.img_shape,
padding="same"))
model.add(LeakyReLU(alpha=0.2))
model.add(Dropout(0.25))
model.add(Conv2D(64, kernel_size=3, strides=2, padding="same"))
model.add(ZeroPadding2D(padding=((0, 1), (0, 1))))
model.add(BatchNormalization(momentum=0.8))
model.add(LeakyReLU(alpha=0.2))
model.add(Dropout(0.25))
model.add(Conv2D(128, kernel_size=3, strides=2, padding="same"))
model.add(BatchNormalization(momentum=0.8))
model.add(LeakyReLU(alpha=0.2))
model.add(Dropout(0.25))
model.add(Conv2D(256, kernel_size=3, strides=1, padding="same"))
model.add(BatchNormalization(momentum=0.8))
model.add(LeakyReLU(alpha=0.2))
model.add(Dropout(0.25))
model.add(Flatten())
model.add(Dense(1, activation='sigmoid'))
model.summary()
return model
def train(self, epochs, batch_size, save_interval, log_interval):
# Load the dataset
(X_train, _), (_, _) = mnist.load_data()
# Rescale -1 to 1
X_train = X_train / 127.5 - 1.
X_train = np.expand_dims(X_train, axis=3)
# Adversarial ground truths
valid = np.ones((batch_size, 1))
fake = np.zeros((batch_size, 1))
logs = []
for epoch in range(epochs):
# ---------------------
# Train Discriminator
# ---------------------
# Select a random half of images
idx = np.random.randint(0, X_train.shape[0], batch_size)
imgs = X_train[idx]
# Sample noise and generate a batch of new images
noise = np.random.normal(0, 1, (batch_size, self.latent_dim))
gen_imgs = self.generator.predict(noise)
# Train the discriminator (real classified as ones and generated as
# zeros)
d_loss_real = self.discriminator.train_on_batch(imgs, valid)
d_loss_fake = self.discriminator.train_on_batch(gen_imgs, fake)
d_loss = 0.5 * np.add(d_loss_real, d_loss_fake)
# ---------------------
# Train Generator
# ---------------------
# Train the generator (wants discriminator to mistake images as
# real)
g_loss = self.combined.train_on_batch(noise, valid)
if epoch % log_interval == 0:
logs.append([epoch, d_loss[0], d_loss[1], g_loss])
if epoch % save_interval == 0:
print("%d [D loss: %f, acc.: %.2f%%] [G loss: %f]" %
(epoch, d_loss[0], 100 * d_loss[1], g_loss))
self.save_imgs(epoch)
self.showlogs(logs)
def showlogs(self, logs):
logs = np.array(logs)
names = ["d_loss", "d_acc", "g_loss"]
for i in range(3):
plt.subplot(2, 2, i + 1)
plt.plot(logs[:, 0], logs[:, i + 1])
plt.xlabel("epoch")
plt.ylabel(names[i])
plt.tight_layout()
plt.show()
def save_imgs(self, epoch):
r, c = 5, 5
noise = np.random.normal(0, 1, (r * c, self.latent_dim))
gen_imgs = self.generator.predict(noise)
# Rescale images 0 - 1
gen_imgs = 0.5 * gen_imgs + 0.5
fig, axs = plt.subplots(r, c)
cnt = 0
for i in range(r):
for j in range(c):
axs[i, j].imshow(gen_imgs[cnt, :, :, 0], cmap='gray')
axs[i, j].axis('off')
cnt += 1
fig.savefig("images/mnist_%d.png" % epoch)
plt.close()
if __name__ == '__main__':
dcgan = DCGAN()
dcgan.train(epochs=4000, batch_size=32, save_interval=50, log_interval=10)
