376feb9f02920deffd10ca2030056d5f5e5e3287,cleverhans/attacks.py,,fgm,#Any#Any#Any#Any#Any#Any#Any#Any#,386

Before Change

  adv_x = x + scaled_grad

  // If clipping is needed, reset all values outside of [clip_min, clip_max]
  if (clip_min is not None) and (clip_max is not None):
    adv_x = tf.clip_by_value(adv_x, clip_min, clip_max)

  return adv_x

class ProjectedGradientDescent(Attack):
  

After Change

  if (clip_min is not None) or (clip_max is not None):
    // We don"t currently support one-sided clipping
    assert clip_min is not None and clip_max is not None
    if adv_x.dtype == tf.float32 and clip_min.dtype != tf.float32:
      clip_min = tf.cast(clip_min, tf.float32)
    if adv_x.dtype == tf.float32 and clip_max.dtype != tf.float32:
      clip_max = tf.cast(clip_max, tf.float32)
    adv_x = tf.clip_by_value(adv_x, clip_min, clip_max)

  return adv_x

Instances