Tensorflow Object Detection API MS-COCO Evaluation Multiple Categories Result Gathering

I am using tensorflow object detection api with coco evaluation

I made a custom data set of roads, and has 2 categories, which is 'vehicle' and 'heavyVehicle'.

Now the results shows like this

enter image description here

What I want to show is the evaluation results of each category: vehicle and heavyVehicle

I first thought the file 'coco_tools.py' and 'coco_evaluation.py' would make this work. These are on 'models/research/object_detection/metrics' folder.

There are some lines that seems to be affect the evaluation results. Such as 'include_metrics_per_category' or 'all_metrics_per_category'.

However it seems to be related with original categories of COCO metric. My dataset has my own customized category: vehicle and heavyVehicle.

If I change 'include_metrics_per_category' and 'all_metrics_per_category' in 'coco_tools.py' and 'coco_evaluation.py' to True, the error shows as below

enter image description here

The actual error is much longer, but it is mostly saying the same thing. 'Category stats do not exist' is the key word.

How can I get a evaluation result for each category? and if it is possible, how can I make a txt file which includes those metric results.

Thank you!