Tuesday, May 25, 2021

Generator for DCGAN

 Following is the code for generator for MNIST DCGAN 


import tensorflow as tf 
from tensorflow import keras 
from keras import layers 
import matplotlib.pyplot as plt 



img_height = 40 
img_width  = 80 

model = keras.models.Sequential()

model.add(layers.Dense(img_height/4 * img_width /4 * 256 , use_bias = False
                       input_shape = (100,)))
model.add(layers.BatchNormalization())
model.add(layers.LeakyReLU())

model.add(layers.Reshape(( int(img_height/4) , int(img_width/4) , 256)))


model.add(layers.Conv2DTranspose(128 , 5, strides = 1 , padding = "same" , 
                                 use_bias = False))
model.add(layers.BatchNormalization())
model.add(layers.LeakyReLU())


model.add(layers.Conv2DTranspose(645, strides = 2 , use_bias = False
                                 padding = "same"))
model.add(layers.BatchNormalization())
model.add(layers.LeakyReLU())


model.add(layers.Conv2DTranspose(15, strides = 2 , use_bias = False , 
                                 padding = "same"))


noise = tf.random.normal([1,100])
gen_image = model(noise , training = False)

plt.imshow(gen_image[0, :, :, 0])


No comments:

Post a Comment

Loop Engineering: Designing the Systems That Prompt Your Agents

Loop Engineering For the last couple of years, the core skill in working with AI was writing a good prompt. You'd craft careful instru...