Tensorflow Model Garden always outputs validation loss equals zero for Instance / Semantic Segmentation

Question

I'm running the notebook of the Instance Segmentation tutorial of Tensorflow Model Garden. During the training, the output of the 'Train and Evaluate' periodically displays the metrics, but the value of the validation loss is always zero. For example:

eval | step:   1200 | steps/sec:    3.3 | eval time:   60.0 sec | output: 
{'AP': 0.0814979,
 'AP50': 0.16509584,
 'AP75': 0.07145937,
  ...
 'mask_ARs': 0.0015765766,
 'steps_per_second': 3.3340747415649394,
 'validation_loss': 0.0}

I noticed the same behavior in the Semantic Segmentation tutorial. For this tutorial, the outputs are shown on the tutorial page: https://www.tensorflow.org/tfmodels/vision/semantic_segmentation#train_and_evaluate

Is this a TensorFlow Models bug? Do I need to set any training configuration variable to calculate the validation loss?