open-mmlab / mmyolo

OpenMMLab YOLO series toolbox and benchmark. Implemented RTMDet, RTMDet-Rotated,YOLOv5, YOLOv6, YOLOv7, YOLOv8,YOLOX, PPYOLOE, etc.

Home Page:https://mmyolo.readthedocs.io/zh_CN/dev/

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

cspnext imagenet pretrained model in RTMDET

zugofn opened this issue · comments

What is the problem this feature will solve?

What time will the cspnext pretrained model be relased in the page?
Like backbone in RTMDet-m、RTMDet-x、RTMDet-l?

What is the feature you are proposing to solve the problem?

backbone pretrained model in RTMDet.

What alternatives have you considered?

No response

@zugofn Large models do not require pre-trained weights, so there is no development plan for the time being.

@zugofn Large models do not require pre-trained weights, so there is no development plan for the time being.

Thanks for your reply!
So can I train RTMDet-m/x/l directly from scratch without using imagenet pretrained model?

Yes

Thanks a lot!
Have a nice day😀