Hope1337

@AndreiMoraru123  ```w``` and ```h``` just size of prior box, I mean, the top-left point and the right-bottom point of rectangle need to be recalculated from ```cx, cy, w, h```...

@AndreiMoraru123 Yup, as I stated, we can avoid it by first convert ```(cx, cy, w, h)``` to ```(xmin, ymin, xmax, ymax)``` and then clamp it into range ```[0 .. 1]```...

@sgrvinod sorry for bothering you but I think it's a bug, beside that, pretrain link seem to be died

@sgrvinod I sincerely apologize for bothering you multiple times, but it seems that this is not correct : In ```class MultiBoxLoss``` : ```python self.smooth_l1 = nn.L1Loss() ``` ```nn.L1Loss()``` is not...

> Hi @AakiraOtok, it's not a bother at all. Yeah, I was aware of this. Thanks for reminding me. See #60. I'll make a note of this difference from the...

I'll try, thanks

@Jegp The link seems to be dead. I really, really want to join the Norse community. Could you please provide another link? 😊

It worked! Thanks

I encountered the same thing. Do you solve this problem ?

@nwojke I apologize for bothering you, but the purpose is to optimize for time and memory, isn't it?