def deserialize_image_record()

in machine_learning/ml_infrastructure/inference-server-performance/server/scripts/tensorrt-optimization.py [0:0]

• 12 lines of code
• 1 McCabe index (conditional complexity)

def deserialize_image_record(record):
  keys_to_features = {
    'image/encoded': tf.FixedLenFeature((), tf.string, ''),
    'image/class/label': tf.FixedLenFeature([], tf.int64, -1),
  }

  with tf.name_scope('deserialize_image_record'):
    parsed = tf.parse_single_example(record, keys_to_features)
    image_bytes = tf.reshape(parsed['image/encoded'], shape=[])
    label = tf.cast(tf.reshape(parsed['image/class/label'], shape=[]),
                    dtype=tf.int32)
        
    # 'image/class/label' is encoded as an integer from 1 to
    # num_label_classes in order to generate the correct one-hot label
    # vector from this number, we subtract the number by 1 to make it
    # in [0, num_label_classes).
    label -= 1
    return image_bytes, label