Finetuning an EfficientDet Model for use with MLKit

[cjohn001]

Hello together,
I am new to keras-cv and I am currently trying to train an efficientDet model linked below. Once trained, I would like to deploy it with MLKit, hence need to convert it to TF light considering the following constraints

https://developers.google.com/ml-kit/custom-models#model-compatibility

Unfortuantely, I have not found any information yet, how I could train a pretrained model which is not part of keras-cv. I assume I must wrap the model in some way? Would be great if someone could provide directions on how to train the model and how to export it in a way that it is executable by MLKit. Thanks for your help!

https://www.kaggle.com/models/tensorflow/efficientdet
https://github.com/tensorflow/models/blob/master/research/object_detection/g3doc/tf2_detection_zoo.md