TF2 0 GAN
================ by Jawad Haider
# Install TensorFlow
# !pip install -q tensorflow-gpu==2.0.0-rc0
try:
%tensorflow_version 2.x # Colab only.
except Exception:
pass
import tensorflow as tf
print(tf.__version__)
`%tensorflow_version` only switches the major version: `1.x` or `2.x`.
You set: `2.x # Colab only.`. This will be interpreted as: `2.x`.
TensorFlow is already loaded. Please restart the runtime to change versions.
2.0.0-rc0
# More imports
from tensorflow.keras.layers import Input, Dense, LeakyReLU, Dropout, \
BatchNormalization
from tensorflow.keras.models import Model
from tensorflow.keras.optimizers import SGD, Adam
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import sys, os
# Load in the data
mnist = tf.keras.datasets.mnist
(x_train, y_train), (x_test, y_test) = mnist.load_data()
# map inputs to (-1, +1) for better training
x_train, x_test = x_train / 255.0 * 2 - 1, x_test / 255.0 * 2 - 1
print("x_train.shape:", x_train.shape)
x_train.shape: (60000, 28, 28)
# Flatten the data
N, H, W = x_train.shape
D = H * W
x_train = x_train.reshape(-1, D)
x_test = x_test.reshape(-1, D)
# Get the generator model
def build_generator(latent_dim):
i = Input(shape=(latent_dim,))
x = Dense(256, activation=LeakyReLU(alpha=0.2))(i)
x = BatchNormalization(momentum=0.7)(x)
x = Dense(512, activation=LeakyReLU(alpha=0.2))(x)
x = BatchNormalization(momentum=0.7)(x)
x = Dense(1024, activation=LeakyReLU(alpha=0.2))(x)
x = BatchNormalization(momentum=0.7)(x)
x = Dense(D, activation='tanh')(x)
model = Model(i, x)
return model
# Get the discriminator model
def build_discriminator(img_size):
i = Input(shape=(img_size,))
x = Dense(512, activation=LeakyReLU(alpha=0.2))(i)
x = Dense(256, activation=LeakyReLU(alpha=0.2))(x)
x = Dense(1, activation='sigmoid')(x)
model = Model(i, x)
return model
# Compile both models in preparation for training
# Build and compile the discriminator
discriminator = build_discriminator(D)
discriminator.compile(
loss='binary_crossentropy',
optimizer=Adam(0.0002, 0.5),
metrics=['accuracy'])
# Build and compile the combined model
generator = build_generator(latent_dim)
# Create an input to represent noise sample from latent space
z = Input(shape=(latent_dim,))
# Pass noise through generator to get an image
img = generator(z)
# Make sure only the generator is trained
discriminator.trainable = False
# The true output is fake, but we label them real!
fake_pred = discriminator(img)
# Create the combined model object
combined_model = Model(z, fake_pred)
# Compile the combined model
combined_model.compile(loss='binary_crossentropy', optimizer=Adam(0.0002, 0.5))
# Train the GAN
# Config
batch_size = 32
epochs = 30000
sample_period = 200 # every `sample_period` steps generate and save some data
# Create batch labels to use when calling train_on_batch
ones = np.ones(batch_size)
zeros = np.zeros(batch_size)
# Store the losses
d_losses = []
g_losses = []
# Create a folder to store generated images
if not os.path.exists('gan_images'):
os.makedirs('gan_images')
# A function to generate a grid of random samples from the generator
# and save them to a file
def sample_images(epoch):
rows, cols = 5, 5
noise = np.random.randn(rows * cols, latent_dim)
imgs = generator.predict(noise)
# Rescale images 0 - 1
imgs = 0.5 * imgs + 0.5
fig, axs = plt.subplots(rows, cols)
idx = 0
for i in range(rows):
for j in range(cols):
axs[i,j].imshow(imgs[idx].reshape(H, W), cmap='gray')
axs[i,j].axis('off')
idx += 1
fig.savefig("gan_images/%d.png" % epoch)
plt.close()
# Main training loop
for epoch in range(epochs):
###########################
### Train discriminator ###
###########################
# Select a random batch of images
idx = np.random.randint(0, x_train.shape[0], batch_size)
real_imgs = x_train[idx]
# Generate fake images
noise = np.random.randn(batch_size, latent_dim)
fake_imgs = generator.predict(noise)
# Train the discriminator
# both loss and accuracy are returned
d_loss_real, d_acc_real = discriminator.train_on_batch(real_imgs, ones)
d_loss_fake, d_acc_fake = discriminator.train_on_batch(fake_imgs, zeros)
d_loss = 0.5 * (d_loss_real + d_loss_fake)
d_acc = 0.5 * (d_acc_real + d_acc_fake)
#######################
### Train generator ###
#######################
noise = np.random.randn(batch_size, latent_dim)
g_loss = combined_model.train_on_batch(noise, ones)
# do it again!
noise = np.random.randn(batch_size, latent_dim)
g_loss = combined_model.train_on_batch(noise, ones)
# Save the losses
d_losses.append(d_loss)
g_losses.append(g_loss)
if epoch % 100 == 0:
print(f"epoch: {epoch+1}/{epochs}, d_loss: {d_loss:.2f}, \
d_acc: {d_acc:.2f}, g_loss: {g_loss:.2f}")
if epoch % sample_period == 0:
sample_images(epoch)
epoch: 1/30000, d_loss: 0.74, d_acc: 0.42, g_loss: 0.70
epoch: 101/30000, d_loss: 0.04, d_acc: 1.00, g_loss: 3.91
epoch: 201/30000, d_loss: 0.92, d_acc: 0.41, g_loss: 0.70
epoch: 301/30000, d_loss: 0.71, d_acc: 0.45, g_loss: 0.58
epoch: 401/30000, d_loss: 0.69, d_acc: 0.48, g_loss: 0.61
epoch: 501/30000, d_loss: 0.67, d_acc: 0.52, g_loss: 0.63
epoch: 601/30000, d_loss: 0.69, d_acc: 0.45, g_loss: 0.63
epoch: 701/30000, d_loss: 0.68, d_acc: 0.44, g_loss: 0.68
epoch: 801/30000, d_loss: 0.67, d_acc: 0.64, g_loss: 0.68
epoch: 901/30000, d_loss: 0.65, d_acc: 0.61, g_loss: 0.71
epoch: 1001/30000, d_loss: 0.66, d_acc: 0.58, g_loss: 0.71
epoch: 1101/30000, d_loss: 0.66, d_acc: 0.64, g_loss: 0.74
epoch: 1201/30000, d_loss: 0.63, d_acc: 0.72, g_loss: 0.76
epoch: 1301/30000, d_loss: 0.68, d_acc: 0.56, g_loss: 0.75
epoch: 1401/30000, d_loss: 0.66, d_acc: 0.69, g_loss: 0.72
epoch: 1501/30000, d_loss: 0.67, d_acc: 0.56, g_loss: 0.80
epoch: 1601/30000, d_loss: 0.67, d_acc: 0.59, g_loss: 0.80
epoch: 1701/30000, d_loss: 0.64, d_acc: 0.62, g_loss: 0.80
epoch: 1801/30000, d_loss: 0.64, d_acc: 0.78, g_loss: 0.75
epoch: 1901/30000, d_loss: 0.65, d_acc: 0.62, g_loss: 0.76
epoch: 2001/30000, d_loss: 0.66, d_acc: 0.55, g_loss: 0.76
epoch: 2101/30000, d_loss: 0.65, d_acc: 0.59, g_loss: 0.78
epoch: 2201/30000, d_loss: 0.69, d_acc: 0.48, g_loss: 0.75
epoch: 2301/30000, d_loss: 0.65, d_acc: 0.69, g_loss: 0.78
epoch: 2401/30000, d_loss: 0.64, d_acc: 0.67, g_loss: 0.78
epoch: 2501/30000, d_loss: 0.69, d_acc: 0.50, g_loss: 0.80
epoch: 2601/30000, d_loss: 0.68, d_acc: 0.62, g_loss: 0.78
epoch: 2701/30000, d_loss: 0.67, d_acc: 0.66, g_loss: 0.76
epoch: 2801/30000, d_loss: 0.68, d_acc: 0.61, g_loss: 0.79
epoch: 2901/30000, d_loss: 0.69, d_acc: 0.56, g_loss: 0.80
epoch: 3001/30000, d_loss: 0.68, d_acc: 0.52, g_loss: 0.76
epoch: 3101/30000, d_loss: 0.66, d_acc: 0.67, g_loss: 0.77
epoch: 3201/30000, d_loss: 0.71, d_acc: 0.52, g_loss: 0.77
epoch: 3301/30000, d_loss: 0.67, d_acc: 0.62, g_loss: 0.77
epoch: 3401/30000, d_loss: 0.64, d_acc: 0.64, g_loss: 0.78
epoch: 3501/30000, d_loss: 0.72, d_acc: 0.48, g_loss: 0.77
epoch: 3601/30000, d_loss: 0.67, d_acc: 0.59, g_loss: 0.82
epoch: 3701/30000, d_loss: 0.69, d_acc: 0.58, g_loss: 0.80
epoch: 3801/30000, d_loss: 0.68, d_acc: 0.52, g_loss: 0.73
epoch: 3901/30000, d_loss: 0.70, d_acc: 0.45, g_loss: 0.74
epoch: 4001/30000, d_loss: 0.67, d_acc: 0.58, g_loss: 0.79
epoch: 4101/30000, d_loss: 0.66, d_acc: 0.59, g_loss: 0.76
epoch: 4201/30000, d_loss: 0.69, d_acc: 0.52, g_loss: 0.74
epoch: 4301/30000, d_loss: 0.70, d_acc: 0.50, g_loss: 0.76
epoch: 4401/30000, d_loss: 0.68, d_acc: 0.56, g_loss: 0.76
epoch: 4501/30000, d_loss: 0.70, d_acc: 0.55, g_loss: 0.78
epoch: 4601/30000, d_loss: 0.67, d_acc: 0.58, g_loss: 0.78
epoch: 4701/30000, d_loss: 0.66, d_acc: 0.59, g_loss: 0.72
epoch: 4801/30000, d_loss: 0.67, d_acc: 0.58, g_loss: 0.76
epoch: 4901/30000, d_loss: 0.72, d_acc: 0.52, g_loss: 0.73
epoch: 5001/30000, d_loss: 0.73, d_acc: 0.42, g_loss: 0.73
epoch: 5101/30000, d_loss: 0.66, d_acc: 0.67, g_loss: 0.78
epoch: 5201/30000, d_loss: 0.67, d_acc: 0.62, g_loss: 0.75
epoch: 5301/30000, d_loss: 0.67, d_acc: 0.53, g_loss: 0.77
epoch: 5401/30000, d_loss: 0.69, d_acc: 0.56, g_loss: 0.78
epoch: 5501/30000, d_loss: 0.68, d_acc: 0.61, g_loss: 0.75
epoch: 5601/30000, d_loss: 0.72, d_acc: 0.50, g_loss: 0.75
epoch: 5701/30000, d_loss: 0.69, d_acc: 0.55, g_loss: 0.76
epoch: 5801/30000, d_loss: 0.72, d_acc: 0.47, g_loss: 0.76
epoch: 5901/30000, d_loss: 0.70, d_acc: 0.47, g_loss: 0.75
epoch: 6001/30000, d_loss: 0.70, d_acc: 0.56, g_loss: 0.77
epoch: 6101/30000, d_loss: 0.67, d_acc: 0.59, g_loss: 0.80
epoch: 6201/30000, d_loss: 0.70, d_acc: 0.50, g_loss: 0.76
epoch: 6301/30000, d_loss: 0.73, d_acc: 0.48, g_loss: 0.75
epoch: 6401/30000, d_loss: 0.67, d_acc: 0.61, g_loss: 0.75
epoch: 6501/30000, d_loss: 0.70, d_acc: 0.55, g_loss: 0.79
epoch: 6601/30000, d_loss: 0.68, d_acc: 0.59, g_loss: 0.77
epoch: 6701/30000, d_loss: 0.69, d_acc: 0.53, g_loss: 0.76
epoch: 6801/30000, d_loss: 0.70, d_acc: 0.53, g_loss: 0.76
epoch: 6901/30000, d_loss: 0.70, d_acc: 0.53, g_loss: 0.77
epoch: 7001/30000, d_loss: 0.71, d_acc: 0.52, g_loss: 0.76
epoch: 7101/30000, d_loss: 0.66, d_acc: 0.59, g_loss: 0.77
epoch: 7201/30000, d_loss: 0.67, d_acc: 0.61, g_loss: 0.80
epoch: 7301/30000, d_loss: 0.70, d_acc: 0.52, g_loss: 0.77
epoch: 7401/30000, d_loss: 0.69, d_acc: 0.53, g_loss: 0.76
epoch: 7501/30000, d_loss: 0.70, d_acc: 0.52, g_loss: 0.76
epoch: 7601/30000, d_loss: 0.70, d_acc: 0.48, g_loss: 0.77
epoch: 7701/30000, d_loss: 0.70, d_acc: 0.45, g_loss: 0.76
epoch: 7801/30000, d_loss: 0.69, d_acc: 0.56, g_loss: 0.75
epoch: 7901/30000, d_loss: 0.68, d_acc: 0.58, g_loss: 0.76
epoch: 8001/30000, d_loss: 0.72, d_acc: 0.44, g_loss: 0.77
epoch: 8101/30000, d_loss: 0.67, d_acc: 0.55, g_loss: 0.78
epoch: 8201/30000, d_loss: 0.69, d_acc: 0.59, g_loss: 0.76
epoch: 8301/30000, d_loss: 0.71, d_acc: 0.47, g_loss: 0.77
epoch: 8401/30000, d_loss: 0.69, d_acc: 0.52, g_loss: 0.78
epoch: 8501/30000, d_loss: 0.64, d_acc: 0.69, g_loss: 0.78
epoch: 8601/30000, d_loss: 0.70, d_acc: 0.53, g_loss: 0.79
epoch: 8701/30000, d_loss: 0.67, d_acc: 0.53, g_loss: 0.79
epoch: 8801/30000, d_loss: 0.69, d_acc: 0.52, g_loss: 0.75
epoch: 8901/30000, d_loss: 0.70, d_acc: 0.52, g_loss: 0.76
epoch: 9001/30000, d_loss: 0.70, d_acc: 0.48, g_loss: 0.76
epoch: 9101/30000, d_loss: 0.67, d_acc: 0.55, g_loss: 0.74
epoch: 9201/30000, d_loss: 0.67, d_acc: 0.56, g_loss: 0.77
epoch: 9301/30000, d_loss: 0.71, d_acc: 0.53, g_loss: 0.77
epoch: 9401/30000, d_loss: 0.67, d_acc: 0.61, g_loss: 0.78
epoch: 9501/30000, d_loss: 0.68, d_acc: 0.61, g_loss: 0.77
epoch: 9601/30000, d_loss: 0.69, d_acc: 0.56, g_loss: 0.79
epoch: 9701/30000, d_loss: 0.70, d_acc: 0.52, g_loss: 0.71
epoch: 9801/30000, d_loss: 0.68, d_acc: 0.55, g_loss: 0.78
epoch: 9901/30000, d_loss: 0.66, d_acc: 0.64, g_loss: 0.76
epoch: 10001/30000, d_loss: 0.71, d_acc: 0.47, g_loss: 0.75
epoch: 10101/30000, d_loss: 0.71, d_acc: 0.48, g_loss: 0.79
epoch: 10201/30000, d_loss: 0.69, d_acc: 0.55, g_loss: 0.78
epoch: 10301/30000, d_loss: 0.70, d_acc: 0.58, g_loss: 0.80
epoch: 10401/30000, d_loss: 0.70, d_acc: 0.50, g_loss: 0.80
epoch: 10501/30000, d_loss: 0.69, d_acc: 0.59, g_loss: 0.76
epoch: 10601/30000, d_loss: 0.69, d_acc: 0.52, g_loss: 0.73
epoch: 10701/30000, d_loss: 0.70, d_acc: 0.56, g_loss: 0.75
epoch: 10801/30000, d_loss: 0.69, d_acc: 0.55, g_loss: 0.75
epoch: 10901/30000, d_loss: 0.67, d_acc: 0.61, g_loss: 0.80
epoch: 11001/30000, d_loss: 0.67, d_acc: 0.61, g_loss: 0.78
epoch: 11101/30000, d_loss: 0.69, d_acc: 0.55, g_loss: 0.77
epoch: 11201/30000, d_loss: 0.69, d_acc: 0.56, g_loss: 0.77
epoch: 11301/30000, d_loss: 0.72, d_acc: 0.52, g_loss: 0.76
epoch: 11401/30000, d_loss: 0.68, d_acc: 0.58, g_loss: 0.72
epoch: 11501/30000, d_loss: 0.70, d_acc: 0.50, g_loss: 0.76
epoch: 11601/30000, d_loss: 0.69, d_acc: 0.52, g_loss: 0.77
epoch: 11701/30000, d_loss: 0.69, d_acc: 0.50, g_loss: 0.77
epoch: 11801/30000, d_loss: 0.67, d_acc: 0.61, g_loss: 0.76
epoch: 11901/30000, d_loss: 0.71, d_acc: 0.48, g_loss: 0.79
epoch: 12001/30000, d_loss: 0.71, d_acc: 0.50, g_loss: 0.78
epoch: 12101/30000, d_loss: 0.72, d_acc: 0.52, g_loss: 0.75
epoch: 12201/30000, d_loss: 0.69, d_acc: 0.53, g_loss: 0.74
epoch: 12301/30000, d_loss: 0.68, d_acc: 0.53, g_loss: 0.72
epoch: 12401/30000, d_loss: 0.65, d_acc: 0.67, g_loss: 0.79
epoch: 12501/30000, d_loss: 0.66, d_acc: 0.62, g_loss: 0.73
epoch: 12601/30000, d_loss: 0.69, d_acc: 0.53, g_loss: 0.79
epoch: 12701/30000, d_loss: 0.65, d_acc: 0.69, g_loss: 0.73
epoch: 12801/30000, d_loss: 0.69, d_acc: 0.53, g_loss: 0.80
epoch: 12901/30000, d_loss: 0.70, d_acc: 0.52, g_loss: 0.78
epoch: 13001/30000, d_loss: 0.68, d_acc: 0.58, g_loss: 0.81
epoch: 13101/30000, d_loss: 0.67, d_acc: 0.64, g_loss: 0.73
epoch: 13201/30000, d_loss: 0.71, d_acc: 0.56, g_loss: 0.77
epoch: 13301/30000, d_loss: 0.70, d_acc: 0.53, g_loss: 0.80
epoch: 13401/30000, d_loss: 0.69, d_acc: 0.55, g_loss: 0.82
epoch: 13501/30000, d_loss: 0.70, d_acc: 0.52, g_loss: 0.75
epoch: 13601/30000, d_loss: 0.72, d_acc: 0.50, g_loss: 0.78
epoch: 13701/30000, d_loss: 0.67, d_acc: 0.53, g_loss: 0.75
epoch: 13801/30000, d_loss: 0.71, d_acc: 0.48, g_loss: 0.75
epoch: 13901/30000, d_loss: 0.72, d_acc: 0.47, g_loss: 0.76
epoch: 14001/30000, d_loss: 0.67, d_acc: 0.55, g_loss: 0.76
epoch: 14101/30000, d_loss: 0.67, d_acc: 0.61, g_loss: 0.75
epoch: 14201/30000, d_loss: 0.70, d_acc: 0.58, g_loss: 0.75
epoch: 14301/30000, d_loss: 0.71, d_acc: 0.48, g_loss: 0.74
epoch: 14401/30000, d_loss: 0.68, d_acc: 0.59, g_loss: 0.71
epoch: 14501/30000, d_loss: 0.68, d_acc: 0.56, g_loss: 0.81
epoch: 14601/30000, d_loss: 0.71, d_acc: 0.48, g_loss: 0.76
epoch: 14701/30000, d_loss: 0.72, d_acc: 0.50, g_loss: 0.73
epoch: 14801/30000, d_loss: 0.69, d_acc: 0.45, g_loss: 0.78
epoch: 14901/30000, d_loss: 0.70, d_acc: 0.48, g_loss: 0.77
epoch: 15001/30000, d_loss: 0.64, d_acc: 0.67, g_loss: 0.80
epoch: 15101/30000, d_loss: 0.70, d_acc: 0.48, g_loss: 0.75
epoch: 15201/30000, d_loss: 0.65, d_acc: 0.66, g_loss: 0.73
epoch: 15301/30000, d_loss: 0.73, d_acc: 0.44, g_loss: 0.72
epoch: 15401/30000, d_loss: 0.71, d_acc: 0.48, g_loss: 0.76
epoch: 15501/30000, d_loss: 0.69, d_acc: 0.53, g_loss: 0.75
epoch: 15601/30000, d_loss: 0.65, d_acc: 0.66, g_loss: 0.77
epoch: 15701/30000, d_loss: 0.68, d_acc: 0.58, g_loss: 0.78
epoch: 15801/30000, d_loss: 0.70, d_acc: 0.50, g_loss: 0.71
epoch: 15901/30000, d_loss: 0.68, d_acc: 0.59, g_loss: 0.77
epoch: 16001/30000, d_loss: 0.69, d_acc: 0.55, g_loss: 0.81
epoch: 16101/30000, d_loss: 0.66, d_acc: 0.66, g_loss: 0.77
epoch: 16201/30000, d_loss: 0.69, d_acc: 0.56, g_loss: 0.74
epoch: 16301/30000, d_loss: 0.68, d_acc: 0.61, g_loss: 0.74
epoch: 16401/30000, d_loss: 0.69, d_acc: 0.58, g_loss: 0.79
epoch: 16501/30000, d_loss: 0.69, d_acc: 0.56, g_loss: 0.80
epoch: 16601/30000, d_loss: 0.68, d_acc: 0.56, g_loss: 0.80
epoch: 16701/30000, d_loss: 0.71, d_acc: 0.50, g_loss: 0.77
epoch: 16801/30000, d_loss: 0.68, d_acc: 0.56, g_loss: 0.75
epoch: 16901/30000, d_loss: 0.67, d_acc: 0.58, g_loss: 0.78
epoch: 17001/30000, d_loss: 0.65, d_acc: 0.64, g_loss: 0.76
epoch: 17101/30000, d_loss: 0.72, d_acc: 0.45, g_loss: 0.77
epoch: 17201/30000, d_loss: 0.68, d_acc: 0.53, g_loss: 0.77
epoch: 17301/30000, d_loss: 0.67, d_acc: 0.59, g_loss: 0.81
epoch: 17401/30000, d_loss: 0.70, d_acc: 0.52, g_loss: 0.83
epoch: 17501/30000, d_loss: 0.70, d_acc: 0.47, g_loss: 0.78
epoch: 17601/30000, d_loss: 0.67, d_acc: 0.64, g_loss: 0.83
epoch: 17701/30000, d_loss: 0.68, d_acc: 0.55, g_loss: 0.78
epoch: 17801/30000, d_loss: 0.72, d_acc: 0.48, g_loss: 0.76
epoch: 17901/30000, d_loss: 0.69, d_acc: 0.59, g_loss: 0.81
epoch: 18001/30000, d_loss: 0.70, d_acc: 0.52, g_loss: 0.74
epoch: 18101/30000, d_loss: 0.67, d_acc: 0.66, g_loss: 0.76
epoch: 18201/30000, d_loss: 0.71, d_acc: 0.48, g_loss: 0.82
epoch: 18301/30000, d_loss: 0.72, d_acc: 0.45, g_loss: 0.77
epoch: 18401/30000, d_loss: 0.70, d_acc: 0.47, g_loss: 0.74
epoch: 18501/30000, d_loss: 0.69, d_acc: 0.58, g_loss: 0.78
epoch: 18601/30000, d_loss: 0.68, d_acc: 0.58, g_loss: 0.77
epoch: 18701/30000, d_loss: 0.69, d_acc: 0.52, g_loss: 0.81
epoch: 18801/30000, d_loss: 0.69, d_acc: 0.52, g_loss: 0.76
epoch: 18901/30000, d_loss: 0.69, d_acc: 0.53, g_loss: 0.77
epoch: 19001/30000, d_loss: 0.68, d_acc: 0.64, g_loss: 0.77
epoch: 19101/30000, d_loss: 0.70, d_acc: 0.50, g_loss: 0.73
epoch: 19201/30000, d_loss: 0.70, d_acc: 0.47, g_loss: 0.78
epoch: 19301/30000, d_loss: 0.69, d_acc: 0.55, g_loss: 0.71
epoch: 19401/30000, d_loss: 0.69, d_acc: 0.55, g_loss: 0.77
epoch: 19501/30000, d_loss: 0.67, d_acc: 0.52, g_loss: 0.76
epoch: 19601/30000, d_loss: 0.69, d_acc: 0.55, g_loss: 0.81
epoch: 19701/30000, d_loss: 0.66, d_acc: 0.55, g_loss: 0.76
epoch: 19801/30000, d_loss: 0.69, d_acc: 0.53, g_loss: 0.78
epoch: 19901/30000, d_loss: 0.68, d_acc: 0.52, g_loss: 0.79
epoch: 20001/30000, d_loss: 0.71, d_acc: 0.50, g_loss: 0.77
epoch: 20101/30000, d_loss: 0.70, d_acc: 0.50, g_loss: 0.76
epoch: 20201/30000, d_loss: 0.64, d_acc: 0.67, g_loss: 0.78
epoch: 20301/30000, d_loss: 0.68, d_acc: 0.58, g_loss: 0.75
epoch: 20401/30000, d_loss: 0.66, d_acc: 0.59, g_loss: 0.77
epoch: 20501/30000, d_loss: 0.67, d_acc: 0.55, g_loss: 0.75
epoch: 20601/30000, d_loss: 0.68, d_acc: 0.53, g_loss: 0.77
epoch: 20701/30000, d_loss: 0.67, d_acc: 0.59, g_loss: 0.80
epoch: 20801/30000, d_loss: 0.68, d_acc: 0.58, g_loss: 0.77
epoch: 20901/30000, d_loss: 0.70, d_acc: 0.47, g_loss: 0.77
epoch: 21001/30000, d_loss: 0.75, d_acc: 0.42, g_loss: 0.76
epoch: 21101/30000, d_loss: 0.69, d_acc: 0.53, g_loss: 0.74
epoch: 21201/30000, d_loss: 0.67, d_acc: 0.61, g_loss: 0.79
epoch: 21301/30000, d_loss: 0.67, d_acc: 0.55, g_loss: 0.78
epoch: 21401/30000, d_loss: 0.67, d_acc: 0.55, g_loss: 0.81
epoch: 21501/30000, d_loss: 0.69, d_acc: 0.48, g_loss: 0.79
epoch: 21601/30000, d_loss: 0.70, d_acc: 0.53, g_loss: 0.76
epoch: 21701/30000, d_loss: 0.66, d_acc: 0.64, g_loss: 0.76
epoch: 21801/30000, d_loss: 0.66, d_acc: 0.66, g_loss: 0.78
epoch: 21901/30000, d_loss: 0.68, d_acc: 0.56, g_loss: 0.78
epoch: 22001/30000, d_loss: 0.68, d_acc: 0.59, g_loss: 0.80
epoch: 22101/30000, d_loss: 0.71, d_acc: 0.39, g_loss: 0.79
epoch: 22201/30000, d_loss: 0.69, d_acc: 0.50, g_loss: 0.73
epoch: 22301/30000, d_loss: 0.69, d_acc: 0.47, g_loss: 0.75
epoch: 22401/30000, d_loss: 0.67, d_acc: 0.53, g_loss: 0.80
epoch: 22501/30000, d_loss: 0.67, d_acc: 0.53, g_loss: 0.77
epoch: 22601/30000, d_loss: 0.70, d_acc: 0.44, g_loss: 0.77
epoch: 22701/30000, d_loss: 0.66, d_acc: 0.64, g_loss: 0.80
epoch: 22801/30000, d_loss: 0.69, d_acc: 0.48, g_loss: 0.74
epoch: 22901/30000, d_loss: 0.69, d_acc: 0.53, g_loss: 0.76
epoch: 23001/30000, d_loss: 0.69, d_acc: 0.56, g_loss: 0.79
epoch: 23101/30000, d_loss: 0.72, d_acc: 0.39, g_loss: 0.81
epoch: 23201/30000, d_loss: 0.68, d_acc: 0.58, g_loss: 0.76
epoch: 23301/30000, d_loss: 0.71, d_acc: 0.53, g_loss: 0.76
epoch: 23401/30000, d_loss: 0.69, d_acc: 0.50, g_loss: 0.81
epoch: 23501/30000, d_loss: 0.69, d_acc: 0.62, g_loss: 0.79
epoch: 23601/30000, d_loss: 0.69, d_acc: 0.56, g_loss: 0.75
epoch: 23701/30000, d_loss: 0.73, d_acc: 0.39, g_loss: 0.82
epoch: 23801/30000, d_loss: 0.68, d_acc: 0.58, g_loss: 0.76
epoch: 23901/30000, d_loss: 0.68, d_acc: 0.62, g_loss: 0.76
epoch: 24001/30000, d_loss: 0.69, d_acc: 0.42, g_loss: 0.79
epoch: 24101/30000, d_loss: 0.66, d_acc: 0.62, g_loss: 0.79
epoch: 24201/30000, d_loss: 0.66, d_acc: 0.59, g_loss: 0.77
epoch: 24301/30000, d_loss: 0.66, d_acc: 0.64, g_loss: 0.81
epoch: 24401/30000, d_loss: 0.67, d_acc: 0.56, g_loss: 0.81
epoch: 24501/30000, d_loss: 0.71, d_acc: 0.44, g_loss: 0.78
epoch: 24601/30000, d_loss: 0.65, d_acc: 0.64, g_loss: 0.84
epoch: 24701/30000, d_loss: 0.68, d_acc: 0.50, g_loss: 0.82
epoch: 24801/30000, d_loss: 0.68, d_acc: 0.58, g_loss: 0.75
epoch: 24901/30000, d_loss: 0.72, d_acc: 0.47, g_loss: 0.75
epoch: 25001/30000, d_loss: 0.67, d_acc: 0.58, g_loss: 0.77
epoch: 25101/30000, d_loss: 0.67, d_acc: 0.61, g_loss: 0.76
epoch: 25201/30000, d_loss: 0.67, d_acc: 0.56, g_loss: 0.81
epoch: 25301/30000, d_loss: 0.69, d_acc: 0.42, g_loss: 0.74
epoch: 25401/30000, d_loss: 0.69, d_acc: 0.58, g_loss: 0.80
epoch: 25501/30000, d_loss: 0.69, d_acc: 0.55, g_loss: 0.84
epoch: 25601/30000, d_loss: 0.71, d_acc: 0.52, g_loss: 0.74
epoch: 25701/30000, d_loss: 0.69, d_acc: 0.50, g_loss: 0.79
epoch: 25801/30000, d_loss: 0.68, d_acc: 0.56, g_loss: 0.79
epoch: 25901/30000, d_loss: 0.67, d_acc: 0.59, g_loss: 0.74
epoch: 26001/30000, d_loss: 0.67, d_acc: 0.67, g_loss: 0.76
epoch: 26101/30000, d_loss: 0.70, d_acc: 0.55, g_loss: 0.76
epoch: 26201/30000, d_loss: 0.68, d_acc: 0.50, g_loss: 0.80
epoch: 26301/30000, d_loss: 0.68, d_acc: 0.59, g_loss: 0.80
epoch: 26401/30000, d_loss: 0.70, d_acc: 0.48, g_loss: 0.76
epoch: 26501/30000, d_loss: 0.70, d_acc: 0.44, g_loss: 0.81
epoch: 26601/30000, d_loss: 0.67, d_acc: 0.61, g_loss: 0.78
epoch: 26701/30000, d_loss: 0.68, d_acc: 0.53, g_loss: 0.75
epoch: 26801/30000, d_loss: 0.69, d_acc: 0.64, g_loss: 0.78
epoch: 26901/30000, d_loss: 0.67, d_acc: 0.58, g_loss: 0.82
epoch: 27001/30000, d_loss: 0.70, d_acc: 0.52, g_loss: 0.77
epoch: 27101/30000, d_loss: 0.72, d_acc: 0.53, g_loss: 0.79
epoch: 27201/30000, d_loss: 0.67, d_acc: 0.56, g_loss: 0.73
epoch: 27301/30000, d_loss: 0.69, d_acc: 0.58, g_loss: 0.77
epoch: 27401/30000, d_loss: 0.70, d_acc: 0.52, g_loss: 0.78
epoch: 27501/30000, d_loss: 0.70, d_acc: 0.48, g_loss: 0.73
epoch: 27601/30000, d_loss: 0.65, d_acc: 0.62, g_loss: 0.76
epoch: 27701/30000, d_loss: 0.70, d_acc: 0.56, g_loss: 0.78
epoch: 27801/30000, d_loss: 0.70, d_acc: 0.55, g_loss: 0.79
epoch: 27901/30000, d_loss: 0.68, d_acc: 0.52, g_loss: 0.74
epoch: 28001/30000, d_loss: 0.66, d_acc: 0.58, g_loss: 0.76
epoch: 28101/30000, d_loss: 0.69, d_acc: 0.55, g_loss: 0.79
epoch: 28201/30000, d_loss: 0.72, d_acc: 0.44, g_loss: 0.78
epoch: 28301/30000, d_loss: 0.68, d_acc: 0.55, g_loss: 0.74
epoch: 28401/30000, d_loss: 0.66, d_acc: 0.59, g_loss: 0.75
epoch: 28501/30000, d_loss: 0.70, d_acc: 0.50, g_loss: 0.78
epoch: 28601/30000, d_loss: 0.68, d_acc: 0.58, g_loss: 0.77
epoch: 28701/30000, d_loss: 0.68, d_acc: 0.62, g_loss: 0.76
epoch: 28801/30000, d_loss: 0.71, d_acc: 0.53, g_loss: 0.79
epoch: 28901/30000, d_loss: 0.70, d_acc: 0.45, g_loss: 0.79
epoch: 29001/30000, d_loss: 0.70, d_acc: 0.52, g_loss: 0.76
epoch: 29101/30000, d_loss: 0.69, d_acc: 0.50, g_loss: 0.81
epoch: 29201/30000, d_loss: 0.70, d_acc: 0.50, g_loss: 0.74
epoch: 29301/30000, d_loss: 0.69, d_acc: 0.45, g_loss: 0.80
epoch: 29401/30000, d_loss: 0.68, d_acc: 0.59, g_loss: 0.76
epoch: 29501/30000, d_loss: 0.69, d_acc: 0.61, g_loss: 0.74
epoch: 29601/30000, d_loss: 0.68, d_acc: 0.55, g_loss: 0.75
epoch: 29701/30000, d_loss: 0.65, d_acc: 0.61, g_loss: 0.78
epoch: 29801/30000, d_loss: 0.66, d_acc: 0.61, g_loss: 0.75
epoch: 29901/30000, d_loss: 0.66, d_acc: 0.64, g_loss: 0.78
<matplotlib.legend.Legend at 0x7f46a05ca748>
0.png 13800.png 17800.png 21600.png 25600.png 29600.png 6600.png
10000.png 14000.png 18000.png 21800.png 25800.png 29800.png 6800.png
1000.png 1400.png 1800.png 22000.png 26000.png 3000.png 7000.png
10200.png 14200.png 18200.png 2200.png 2600.png 3200.png 7200.png
10400.png 14400.png 18400.png 22200.png 26200.png 3400.png 7400.png
10600.png 14600.png 18600.png 22400.png 26400.png 3600.png 7600.png
10800.png 14800.png 18800.png 22600.png 26600.png 3800.png 7800.png
11000.png 15000.png 19000.png 22800.png 26800.png 4000.png 8000.png
11200.png 15200.png 19200.png 23000.png 27000.png 400.png 800.png
11400.png 15400.png 19400.png 23200.png 27200.png 4200.png 8200.png
11600.png 15600.png 19600.png 23400.png 27400.png 4400.png 8400.png
11800.png 15800.png 19800.png 23600.png 27600.png 4600.png 8600.png
12000.png 16000.png 20000.png 23800.png 27800.png 4800.png 8800.png
1200.png 1600.png 2000.png 24000.png 28000.png 5000.png 9000.png
12200.png 16200.png 200.png 2400.png 2800.png 5200.png 9200.png
12400.png 16400.png 20200.png 24200.png 28200.png 5400.png 9400.png
12600.png 16600.png 20400.png 24400.png 28400.png 5600.png 9600.png
12800.png 16800.png 20600.png 24600.png 28600.png 5800.png 9800.png
13000.png 17000.png 20800.png 24800.png 28800.png 6000.png
13200.png 17200.png 21000.png 25000.png 29000.png 600.png
13400.png 17400.png 21200.png 25200.png 29200.png 6200.png
13600.png 17600.png 21400.png 25400.png 29400.png 6400.png
<matplotlib.image.AxesImage at 0x7f46a05a69b0>
<matplotlib.image.AxesImage at 0x7f46a0484eb8>
<matplotlib.image.AxesImage at 0x7f46a0467a58>
<matplotlib.image.AxesImage at 0x7f46a03cd588>
<matplotlib.image.AxesImage at 0x7f46a03720f0>
<matplotlib.image.AxesImage at 0x7f46a024ec18>
