[deeplab]when set label_weight raise error as details?

== tensorflow installed from info ==================
Name: tensorflow
Version: 1.13.2
Summary: TensorFlow is an open source machine learning framework for everyone.
Home-page: https://www.tensorflow.org/
License: Apache 2.0
Location: /Users/mlinxiang/anaconda3/envs/mpy/lib/python3.6/site-packages
Requires: protobuf, six, keras-preprocessing, gast, tensorflow-estimator, astor, tensorboard, termcolor, absl-py, numpy, keras-applications, grpcio, wheel
Required-by: imagededup

== python version ==============================================
(major, minor, micro, releaselevel, serial)
(3, 6, 9, 'final', 0)

== bazel version ===============================================
Build label: 0.29.0-homebrew
Build time: Wed Aug 28 18:07:41 2019 (1567015661)
Build timestamp: 1567015661
Build timestamp as int: 1567015661

@aquariusjay

when use label_weights flags, run error as follow, cause by code core/utils.py line 147,

how to fix this
File "/Users/mlinxiang/DeepLearning/ml/hand/mobile-segmentation/slim/deployment/model_deploy.py", line 193, in create_clones
outputs = model_fn(*args, **kwargs)
File "/Users/mlinxiang/DeepLearning/ml/hand/deeplab/deeplab/train.py", line 270, in _build_deeplab
scope=output)
File "/Users/mlinxiang/DeepLearning/ml/hand/deeplab/deeplab/utils/train_utils.py", line 113, in add_softmax_cross_entropy_loss_for_each_scale
scaled_labels, ignore_label, num_classes, label_weights=loss_weight)
File "/Users/mlinxiang/DeepLearning/ml/hand/deeplab/deeplab/core/utils.py", line 149, in get_label_weight_mask
label_weights)
File "/Users/mlinxiang/anaconda3/envs/mpy/lib/python3.6/site-packages/tensorflow/python/ops/special_math_ops.py", line 212, in einsum
raise ValueError('Subscripts with ellipses are not yet supported.')
ValueError: Subscripts with ellipses are not yet supported.