I really wanted to know which version of Tensorflow Object Detection API did you use?
Unfortunately on training with the current version of Tensorflow Object Detection API, the trained models do not work with Tensorflow 1.3.0, which is now a requirement for Udacity Nanodegree.
We basically get the following error on running the trained model on TF1.3.0
ValueError: No op named NonMaxSuppressionV3 in defined operations
Please help!!