pyfunc_1 returns 2 values, but expects to see 1 values.pyfunc_1 returns 2 values, but expects to see 1 values.

Using Tensorflow Object Detection API, I tried training the pre-trained (Resnet101) model using model_main.py . However, when I use the deprecated file train.py, the training is perfect but with model_main.py, I'm getting the following error:
InvalidArgumentError: pyfunc_1 returns 2 values, but expects to see 1 values.
[[node map_3/while/PyFunc (defined at /etc/anaconda3/envs/tensorflow/lib/python3.6/site-packages/tensorflow/models/research/object_detection/utils/visualization_utils.py:440) = PyFunc[Tin=[DT_UINT8, DT_FLOAT, DT_INT64, DT_FLOAT], Tout=[DT_UINT8], _class=["loc:@map_3/while/TensorArrayWrite/TensorArrayWriteV3"], token="pyfunc_1", _device="/job:localhost/replica:0/task:0/device:CPU:0"](map_3/while/Squeeze/_3841, map_3/while/TensorArrayReadV3_3/_3843, map_3/while/TensorArrayReadV3_4/_3845, map_3/while/TensorArrayReadV3_5/_3847)]]

What is the top-level directory of the model you are using: Object Detection
Have I written custom code: No
OS Platform and Distribution: Ubuntu
TensorFlow installed from: Pip
TensorFlow version: 1.12.0
CUDA/cuDNN version: V9.0.176
GPU model and memory: GeForce GTX 1080 - 8GB
Exact command to reproduce: python model_main.py (arguments are predefined within the code)

P.S: Although there's an open issue #5661, no exact solution has been given for this error.