about result[Reimplementation]

[1640477469]

Prerequisite

• I have searched Issues and Discussions but cannot get the expected help.
• I have read the FAQ documentation but cannot get the expected help.
• The bug has not been fixed in the latest version (master) or latest version (3.x).

💬 Describe the reimplementation questions

I want to ask a question：I use my own dataset, using one GPU to train faster_rcnn, lr=0.0025, 180epoch, AP50 on the training set can only reach about 45%, why is it so low?

Environment

no

Expected results

no

Additional information

change class to 5;
use coco datasets

[ZwwWayne]

The low AP can be caused by many reasons. I think the results is reasonable. To improve the performance, you can try larger batch size and stronger data augmentation.