0
我的代碼中存在一些值錯誤：ValueError：變量d_conv2d1/weights/Adam /不存在，或者未使用tf.get_variable（）創建。你是否想在VarScope中設置重用=無？當我使用張量流實現GAN時，某些值錯誤
My tensorflow version is 1.2. The data set is mnist .The code is as follow:
# -*- coding: utf-8 -*-
import os
import numpy as np
import scipy.misc
import tensorflow as tf
from tensorflow.contrib.layers.python.layers import batch_norm as batch_norm
from tensorflow.examples.tutorials.mnist import input_data
import string
BATCH_SIZE = 64
def read_data():
#
mnist = input_data.read_data_sets("MNIST_data/", one_hot = True)
(x_train, y_train), (x_test, y_test) = tf.contrib.keras.datasets.mnist.load_data()
x_train = x_train.reshape((60000,784))
x_test = x_test.reshape((10000,784))
y_train_vec = np.zeros((len(y_train), 10), dtype=np.float)
for i, label in enumerate(y_train):
y_train_vec[i, int(y_train[i])] = 1.0
y_test_vec = np.zeros((len(y_test), 10), dtype=np.float)
for i, label in enumerate(y_test):
y_test_vec[i, int(y_test[i])] = 1.0
#
#
X = np.concatenate((x_train, x_test), axis=0)
y = np.concatenate((y_train_vec, y_test_vec), axis=0)
#
seed = 547
np.random.seed(seed)
np.random.shuffle(X)
np.random.seed(seed)
np.random.shuffle(y)
return X/255., y
# 常數偏置
def bias(name, shape, bias_start = 0.0, trainable = True):
dtype = tf.float32
var = tf.get_variable(name, shape, tf.float32, trainable = trainable,
initializer = tf.constant_initializer(
bias_start, dtype = dtype))
return var
# 隨機權重
def weight(name, shape, stddev = 0.02, trainable = True):
dtype = tf.float32
var = tf.get_variable(name, shape, tf.float32, trainable = trainable,
initializer = tf.random_normal_initializer(
stddev = stddev, dtype = dtype))
return var
# 全連接層
def fully_connected(value, output_shape, name = 'fully_connected', with_w = False):
shape = value.get_shape().as_list()
with tf.variable_scope(name):
weights = weight('weights', [shape[1], output_shape], 0.02)
biases = bias('biases', [output_shape], 0.0)
if with_w:
return tf.matmul(value, weights) + biases, weights, biases
else:
return tf.matmul(value, weights) + biases
# Leaky-ReLu 層
def lrelu(x, leak=0.2, name = 'lrelu'):
with tf.variable_scope(name):
return tf.maximum(x, leak*x, name = name)
# ReLu 層
def relu(value, name = 'relu'):
with tf.variable_scope(name):
return tf.nn.relu(value)
# 解卷積層
def deconv2d(value, output_shape, k_h = 5, k_w = 5, strides =[1, 2, 2, 1],
name = 'deconv2d', with_w = False):
with tf.variable_scope(name):
weights = weight('weights',
[k_h, k_w, output_shape[-1], value.get_shape()[-1]])
deconv = tf.nn.conv2d_transpose(value, weights,
output_shape, strides = strides)
biases = bias('biases', [output_shape[-1]])
deconv = tf.reshape(tf.nn.bias_add(deconv, biases), deconv.get_shape())
if with_w:
return deconv, weights, biases
else:
return deconv
# 卷積層
def conv2d(value, output_dim, k_h = 5, k_w = 5,
strides =[1, 2, 2, 1], name = 'conv2d'):
with tf.variable_scope(name):
weights = weight('weights',
[k_h, k_w, value.get_shape()[-1], output_dim])
conv = tf.nn.conv2d(value, weights, strides = strides, padding = 'SAME')
biases = bias('biases', [output_dim])
conv = tf.reshape(tf.nn.bias_add(conv, biases), conv.get_shape())
return conv
# 把約束條件串聯到 feature map
def conv_cond_concat(value, cond, name = 'concat'):
# 把張量的維度形狀轉化成 Python 的 list
value_shapes = value.get_shape().as_list()
cond_shapes = cond.get_shape().as_list()
# 在第三個維度上（feature map 維度上）把條件和輸入串聯起來，
# 條件會被預先設爲四維張量的形式，假設輸入爲 [64, 32, 32, 32] 維的張量，
# 條件爲 [64, 32, 32, 10] 維的張量，那麼輸出就是一個 [64, 32, 32, 42] 維張量
with tf.variable_scope(name):
return tf.concat([value, cond * tf.ones(value_shapes[0:3] + cond_shapes[3:])],3)
# Batch Normalization 層
def batch_norm_layer(value, is_train = True, name = 'batch_norm'):
with tf.variable_scope(name) as scope:
if is_train:
return batch_norm(value, decay = 0.9, epsilon = 1e-5, scale = True,
is_training = is_train,
updates_collections = None, scope = scope)
else:
return batch_norm(value, decay = 0.9, epsilon = 1e-5, scale = True,
is_training = is_train, reuse = True,
updates_collections = None, scope = scope)
# 保存圖片函數
def save_images(images, size, path):
"""
Save the samples images
The best size number is
int(max(sqrt(image.shape[0]),sqrt(image.shape[1]))) + 1
example:
The batch_size is 64, then the size is recommended [8, 8]
The batch_size is 32, then the size is recommended [6, 6]
"""
# 圖片歸一化，主要用於生成器輸出是 tanh 形式的歸一化
img = (images + 1.0)/2.0
h, w = img.shape[1], img.shape[2]
# 產生一個大畫布，用來保存生成的 batch_size 個圖像
merge_img = np.zeros((h * size[0], w * size[1], 3))
# 循環使得畫布特定地方值爲某一幅圖像的值
for idx, image in enumerate(images):
i = idx % size[1]
j = idx // size[1]
merge_img[j*h:j*h+h, i*w:i*w+w, :] = image
# 保存畫布
return scipy.misc.imsave(path, merge_img)
# 定義生成器
def generator(z, y, train = True):
# y 是一個 [BATCH_SIZE, 10] 維的向量，把 y 轉成四維張量
yb = tf.reshape(y, [BATCH_SIZE, 1, 1, 10], name = 'yb')
# 把 y 作爲約束條件和 z 拼接起來
z = tf.concat([z, y], 1, name = 'z_concat_y')
# 經過一個全連接，BN 和激活層 ReLu
h1 = tf.nn.relu(batch_norm_layer(fully_connected(z, 1024, 'g_fully_connected1'),
is_train = train, name = 'g_bn1'))
# 把約束條件和上一層拼接起來
h1 = tf.concat([h1, y], 1, name = 'active1_concat_y')
h2 = tf.nn.relu(batch_norm_layer(fully_connected(h1, 128 * 49, 'g_fully_connected2'),
is_train = train, name = 'g_bn2'))
h2 = tf.reshape(h2, [64, 7, 7, 128], name = 'h2_reshape')
# 把約束條件和上一層拼接起來
h2 = conv_cond_concat(h2, yb, name = 'active2_concat_y')
h3 = tf.nn.relu(batch_norm_layer(deconv2d(h2, [64,14,14,128],
name = 'g_deconv2d3'),
is_train = train, name = 'g_bn3'))
h3 = conv_cond_concat(h3, yb, name = 'active3_concat_y')
# 經過一個 sigmoid 函數把值歸一化爲 0~1 之間，
h4 = tf.nn.sigmoid(deconv2d(h3, [64, 28, 28, 1],
name = 'g_deconv2d4'), name = 'generate_image')
return h4
# 定義判別器
def discriminator(image, y, reuse = False):
# 因爲真實數據和生成數據都要經過判別器，所以需要指定 reuse 是否可用
if reuse:
tf.get_variable_scope().reuse_variables()
# 同生成器一樣，判別器也需要把約束條件串聯進來
yb = tf.reshape(y, [BATCH_SIZE, 1, 1, 10], name = 'yb')
x = conv_cond_concat(image, yb, name = 'image_concat_y')
# 卷積，激活，串聯條件。
h1 = lrelu(conv2d(x, 11, name = 'd_conv2d1'), name = 'lrelu1')
h1 = conv_cond_concat(h1, yb, name = 'h1_concat_yb')
h2 = lrelu(batch_norm_layer(conv2d(h1, 74, name = 'd_conv2d2'),
name = 'd_bn2'), name = 'lrelu2')
h2 = tf.reshape(h2, [BATCH_SIZE, -1], name = 'reshape_lrelu2_to_2d')
h2 = tf.concat([h2, y],1, name = 'lrelu2_concat_y')
h3 = lrelu(batch_norm_layer(fully_connected(h2, 1024, name = 'd_fully_connected3'),
name = 'd_bn3'), name = 'lrelu3')
h3 = tf.concat([h3, y], 1,name = 'lrelu3_concat_y')
# 全連接層，輸出以爲 loss 值
h4 = fully_connected(h3, 1, name = 'd_result_withouts_sigmoid')
return tf.nn.sigmoid(h4, name = 'discriminator_result_with_sigmoid'), h4
# 定義訓練過程中的採樣函數
def sampler(z, y, train = True):
tf.get_variable_scope().reuse_variables()
return generator(z, y, train = train)
def train():
# 設置 global_step ，用來記錄訓練過程中的 step
global_step = tf.Variable(0, name = 'global_step', trainable = False)
# 訓練過程中的日誌保存文件
train_dir = '/home/Zhoupinyi/DCGAN/logs'
# 放置三個 placeholder，y 表示約束條件，images 表示送入判別器的圖片，
# z 表示隨機噪聲
y= tf.placeholder(tf.float32, [BATCH_SIZE, 10], name='y')
images = tf.placeholder(tf.float32, [64, 28, 28, 1], name='real_images')
z = tf.placeholder(tf.float32, [None, 100], name='z')
# 由生成器生成圖像 G
G = generator(z, y)
# 真實圖像送入判別器
D, D_logits = discriminator(images, y)
# 採樣器採樣圖像
samples = sampler(z, y)
# 生成圖像送入判別器
D_, D_logits_ = discriminator(G, y, reuse = True)
# 損失計算
d_loss_real = tf.reduce_mean(tf.nn.sigmoid_cross_entropy_with_logits(labels = tf.ones_like(D),logits = D_logits))
d_loss_fake = tf.reduce_mean(tf.nn.sigmoid_cross_entropy_with_logits(labels = tf.zeros_like(D_), logits = D_logits_))
d_loss = d_loss_real + d_loss_fake
g_loss = tf.reduce_mean(tf.nn.sigmoid_cross_entropy_with_logits(labels = tf.ones_like(D_), logits = D_logits_))
# 總結操作
z_sum = tf.summary.histogram("z", z)
d_sum = tf.summary.histogram("d", D)
d__sum = tf.summary.histogram("d_", D_)
G_sum = tf.summary.image("G", G)
d_loss_real_sum = tf.summary.scalar("d_loss_real", d_loss_real)
d_loss_fake_sum = tf.summary.scalar("d_loss_fake", d_loss_fake)
d_loss_sum = tf.summary.scalar("d_loss", d_loss)
g_loss_sum = tf.summary.scalar("g_loss", g_loss)
# 合併各自的總結
g_sum = tf.summary.merge([z_sum, d__sum, G_sum, d_loss_fake_sum, g_loss_sum])
d_sum = tf.summary.merge([z_sum, d_sum, d_loss_real_sum, d_loss_sum])
# 生成器和判別器要更新的變量，用於 tf.train.Optimizer 的 var_list
t_vars = tf.trainable_variables()
d_vars = [var for var in t_vars if 'd_' in var.name]
g_vars = [var for var in t_vars if 'g_' in var.name]
saver = tf.train.Saver()
# 優化算法採用 Adam
optimizer = tf.train.AdamOptimizer(learning_rate = 0.0002, beta1 = 0.5)
d_optim = optimizer.minimize(d_loss, global_step = global_step, var_list = d_vars)
g_optim = optimizer.minimize(g_loss, global_step = global_step, var_list = g_vars)
os.environ['CUDA_VISIBLE_DEVICES'] = str(0)
config = tf.ConfigProto()
config.gpu_options.per_process_gpu_memory_fraction = 0.2
sess = tf.InteractiveSession(config=config)
init = tf.initialize_all_variables()
writer = tf.train.SummaryWriter(train_dir, sess.graph)
# 這個自己理解吧
data_x, data_y = read_data()
sample_z = np.random.uniform(-1, 1, size=(BATCH_SIZE, 100))
# sample_images = data_x[0: 64]
sample_labels = data_y[0: 64]
sess.run(init)
# 循環 25 個 epoch 訓練網絡
for epoch in range(25):
batch_idxs = 1093
for idx in range(batch_idxs):
batch_images = data_x[idx*64: (idx+1)*64]
batch_labels = data_y[idx*64: (idx+1)*64]
batch_z = np.random.uniform(-1, 1, size=(BATCH_SIZE, 100))
# 更新 D 的參數
_, summary_str = sess.run([d_optim, d_sum],
feed_dict = {images: batch_images,
z: batch_z,
y: batch_labels})
writer.add_summary(summary_str, idx+1)
# 更新 G 的參數
_, summary_str = sess.run([g_optim, g_sum],
feed_dict = {z: batch_z,
y: batch_labels})
writer.add_summary(summary_str, idx+1)
# 更新兩次 G 的參數確保網絡的穩定
_, summary_str = sess.run([g_optim, g_sum],
feed_dict = {z: batch_z,
y: batch_labels})
writer.add_summary(summary_str, idx+1)
# 計算訓練過程中的損失，打印出來
errD_fake = d_loss_fake.eval({z: batch_z, y: batch_labels})
errD_real = d_loss_real.eval({images: batch_images, y: batch_labels})
errG = g_loss.eval({z: batch_z, y: batch_labels})
if idx % 20 == 0:
print("Epoch: [%2d] [%4d/%4d] d_loss: %.8f, g_loss: %.8f" \
% (epoch, idx, batch_idxs, errD_fake+errD_real, errG))
# 訓練過程中，用採樣器採樣，並且保存採樣的圖片到
# /home/your_name/TensorFlow/DCGAN/samples/
if idx % 100 == 1:
sample = sess.run(samples, feed_dict = {z: sample_z, y: sample_labels})
samples_path = '/home/your_name/TensorFlow/DCGAN/samples/'
save_images(sample, [8, 8],
samples_path + 'test_%d_epoch_%d.png' % (epoch, idx))
print 'save down'
# 每過 500 次迭代，保存一次模型
if idx % 500 == 2:
checkpoint_path = os.path.join(train_dir, 'DCGAN_model.ckpt')
saver.save(sess, checkpoint_path, global_step = idx+1)
sess.close()
if __name__ == '__main__':
train()
The error information is as follow:
Traceback (most recent call last):
File "/home/zhoupinbyi/DCGAN/test_gan.py", line 370, in <module>
train()
File "/home/zhoupinbyi/DCGAN/test_gan.py", line 297, in train
d_optim = optimizer.minimize(d_loss, global_step = global_step, var_list = d_vars)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/training/optimizer.py", line 325, in minimize
name=name)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/training/optimizer.py", line 446, in apply_gradients
self._create_slots([_get_variable_for(v) for v in var_list])
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/training/adam.py", line 128, in _create_slots
self._zeros_slot(v, "m", self._name)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/training/optimizer.py", line 766, in _zeros_slot
named_slots[_var_key(var)] = slot_creator.create_zeros_slot(var, op_name)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/training/slot_creator.py", line 174, in create_zeros_slot
colocate_with_primary=colocate_with_primary)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/training/slot_creator.py", line 146, in create_slot_with_initializer
dtype)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/training/slot_creator.py", line 66, in _create_slot_var
validate_shape=validate_shape)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/ops/variable_scope.py", line 1065, in get_variable
use_resource=use_resource, custom_getter=custom_getter)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/ops/variable_scope.py", line 962, in get_variable
use_resource=use_resource, custom_getter=custom_getter)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/ops/variable_scope.py", line 367, in get_variable
validate_shape=validate_shape, use_resource=use_resource)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/ops/variable_scope.py", line 352, in _true_getter
use_resource=use_resource)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/ops/variable_scope.py", line 682, in _get_single_variable
"VarScope?" % name)
ValueError: Variable d_conv2d1/weights/Adam/ does not exist, or was not created with tf.get_variable(). Did you mean to set reuse=None in VarScope?
這真是一個_minimal_工作的例子嗎？ https://stackoverflow.com/help/mcve你可以嘗試刪除儘可能多的代碼，但仍會產生這個問題嗎？ –