wuhen777
wuhen777
Have you ever encountered this problem when executing training statements ./tools/dist_ Train.sh./projects/configs/bevformer/bevformer_ Base.py 1 When an error occurs, what should be done to resolve it Using one GPU Traceback (most...
The same as BEVFormerV1, see DD3D. https://github.com/TRI-ML/dd3d _Originally posted by @htian01 in https://github.com/fundamentalvision/BEVFormer/issues/245#issuecomment-2092776234_ After changing the network to vovnet99, I used the weight you mentioned. The map values for the...
如果你只是想要跑通整个模型，可以使用nuscnes-mini 替代nuscene. 但是如果只是想训练前两个压缩包的话，可以通过修改data convert 来生成只包含前两个压缩包数据的json. _Originally posted by @zhiqi-li in https://github.com/fundamentalvision/BEVFormer/issues/54#issuecomment-1174635992_ 感谢你的建议，请问您能给出更具体的一点建议吗，是修改mmdetections3d/tools文件夹下的create data.py文件吗，期待您的回复
> For most simple case, hard code on mmdet3d. https://github.com/open-mmlab/mmdetection3d/blob/f1107977dfd26155fc1f83779ee6535d2468f449/mmdet3d/datasets/custom_3d.py#L64C12-L64C12 for example, sub sample to 1/5. From self.data_infos = self.load_annotations(self.ann_file) to self.data_infos = self.load_annotations(self.ann_file)[::5] > > For more reasonable way,...