def _get_generated_data(num_images_generated, conditional_eval, num_classes): """Get generated images.""" noise = tf.random_normal([num_images_generated, 64]) # If conditional, generate class-specific images. if conditional_eval: conditioning = util.get_generator_conditioning(num_images_generated, num_classes) generator_inputs = (noise, conditioning) generator_fn = networks.conditional_generator else: generator_inputs = noise generator_fn = networks.generator # In order for variables to load, use the same variable scope as in the # train job. with tf.variable_scope('Generator'): data = generator_fn(generator_inputs, is_training=False) return data
def _get_generated_data(num_images_generated, conditional_eval, num_classes): """Get generated images.""" noise = tf.random_normal([num_images_generated, 64]) # If conditional, generate class-specific images. if conditional_eval: conditioning = util.get_generator_conditioning( num_images_generated, num_classes) generator_inputs = (noise, conditioning) generator_fn = networks.conditional_generator else: generator_inputs = noise generator_fn = networks.generator # In order for variables to load, use the same variable scope as in the # train job. with tf.variable_scope('Generator'): data = generator_fn(generator_inputs, is_training=False) return data
def test_get_generator_conditioning(self): conditioning = util.get_generator_conditioning(12, 4) self.assertEqual([12, 4], conditioning.shape.as_list())