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...