HaoHe

I might reproduce the result. The best performace is as blow. But still a litter lower than the paper declared. Just use the cfg tools/cfgs/nuscenes_models/3dssd_sasa.yaml. It seems the batchsize matters....

> Hi @pangsu0613 One small question. Is the CLOCS code setup to train with just batch_size=1? Because even if I am increasing the batch_size=8, it is returning just 1 IOU...

> Sorry, there is no open source for libspconv. Thanks for your great work! But Will the source code of libspconv be avaliable in the furure? Or there's no plan...

> @ihaohe Could you please provide the version information for your CUDA and TensorRT? @liuanqi-libra7 CUDA 11.3 TensorRT-8.5.1.7 pytorch-quantization 2.1.2 torch 1.10.1+cu113 mmcv 1.4.0 mmdet 2.20.0 My BEVFusion model is...

My BEVFusion model Int8 result [onnx_int8](https://drive.google.com/file/d/1cb7zDyk512AJbwDRTbc9XI4MXQTfKiMc/view?usp=drive_link). Hope it can help you to locate the problem.

@hopef @liuanqi-libra7 I'm sorry to bother you. Is there any progress about this issue？

> Multiline is currently not supported in nvim-dap, I have submitted a PR that adds that functionality here [mfussenegger/nvim-dap#773](https://github.com/mfussenegger/nvim-dap/pull/773) Got it. Thanks

> 看看我的工程，吊打目前所有的3D目标检测 哈哈 https://github.com/rubbish001/Co-Fix3d @rubbish001 你好，请问下你这个项目里面的贴图是在nuScenes的榜单上吗？我在官方网站上好像没找到你的方法 [nuScenes](https://www.nuscenes.org/object-detection?externalData=all&mapData=all&modalities=Lidar)

> 里面有个提交的页面，上面展示各种算法的，我这个是无测试增强搞出来的 @rubbish001 谢谢，我找到提交的那个榜单了, 看到你的Co-系列算法了。但是那个榜单好像不能区分用的哪个模态的数据吧（lidar or lidar+camera）? 另外，请问下你知道怎么从那个提交的榜单转到正式榜单吗？

> 里面的排行很混乱，各种数据增强，测试增强，只要点数够高就可以排在前面，我的这个是没有测试增强的，最重要的是我还一篇文章都没有中，中了就开源，我感觉我这个想法估计还可以提升 好的好的，谢谢老哥，祝早中顶会