zimenglan
hi @liangzimei, sorry about that. i just finish the code. i will test it in the next few days. btw, can u give how do define the `focal loss` in...
hi @liangzimei, i have fixed the bugs: the loss and the gradient. now it can work. please have a try. for more details, u can see #2
hi @liangzimei, If i try small dataset using pvanet, it's ok, but it encounters the `NAN`, when i use larger dataset. I still try to fix it.
hi @liangzimei, i fix the `NaN` problem. i guess that, the op `(1 - p_t)^gamma / (1 - p_t)` causes the problem, when `1 - p_t` is so small. now...
hi @bharatsingh430 do u still have the plan to release the gpu version of soft-nms? thanks
> Yes, we used bicubic interpolation. hi how load modify the pretrained model of input size 384 to finetune the input 448 when load model ?
hi @xvjiarui have u tried `GN` instead of `Sync BN`?
hi @xvjiarui have u tried putting `context_block` to `FPN`, `bbox_head` and `mask_head`?
hi @thangvubk i use bu-path, ada-pool, and ms-train, without `gn`, and test use `MIN_SIZE_TEST: 1000`, only get 38.1 mAP. is there something i miss?
hi @thangvubk do u try soft-nms instead of standard-nms? if use, can u share the results here?