I'm quite new to using Tensorflow. I was playing around the Inception-V3 model for image recognition in this repo:
https://github.com/tensorflow/models/tree/master/inception#how-to-train-from-scratch
It provides binaries to help you train your own model. However in the Tensorflow docs on Image Recognition,
https://www.tensorflow.org/tutorials/image_recognition/
They allow you to download and utilize a pre-trained Inception-V3 model trained on ImageNet data. The model that you download, is in the format of a .pb GraphDef whereas the model you generate in the tensorflow/models/inception repo is a .ckpt "checkpoint` formatted file.
What's the difference between these two types of files and how can I convert my generated inception model into a GraphDef file?
