Problems about how to process coco's test dataset

Question

In coco, we know that both the training set and validation set correspond to particular annotation files, i.e., instances_train2017.json and instances_val2017.json. However, the coco's test set does not have such a annotation file. Maybe the train/val annotation is the alternative but I don't know what to do. So, I don't know how to generate the test.txt file and make further processing. Thank you very much for your help!