Is batch testing possible?

[chrystalquek]

Hi, I am new to retinanet and wanted to ask for some advice on batching. I understand that csv_eval.py passes a single image into the model at each time, which makes testing slow. Is it possible to modify the code such that a batch of images are passed into the model at the same time, or is that not possible because of how the model works?

[harrystuart]

I looked to do this also. It seems that the actual model script does not natively support this. Would would have to change the model file.