ChenT1994

Results 1 comments of ChenT1994

> I have tried many training tricks, such as different batch_size, different learning rate. But, almost each loss is 2.x. I checked other issues, their mAP are 7x.x%, but mine...