search

dvlab-research / spconv-plus

Apache License 2.0
153 stars 7 forks source link

usage of spconv-plus #3

Closed liulin813 closed 1 year ago

liulin813 commented 1 year ago

Hello, I noticed that there were issues during the use of spconvplus. For example, in the example code, I did not find any corresponding files in the corresponding folders for many referenced libraries. image

Thank you very much for your reply

liulin813 commented 1 year ago

Especially the package referenced by the red line below, I did not find this file

yukang2017 commented 1 year ago

Hi,

This line can be directly removed. It is unneeded. I have updated this line. https://github.com/dvlab-research/FocalsConv/blob/master/OpenPCDet/pcdet/models/backbones_3d/focal_sparse_conv/focal_sparse_conv_cuda.py

Regards, Yukang Chen

yukang2017 commented 1 year ago

For more details, I tried to write a CUDA-version check-repeat operation to replace the pytorch-version checkrepeat for better efficiency. But the actual speed-up is little. Thus, I gave up this operation.

The attached is my trials before. You can try it by yourself if you are interested in it. check_repeat.zip

Regards, Yukang Chen

liulin813 commented 1 year ago

Thank you very much

------------------ 原始邮件 ------------------ 发件人: @.>; 发送时间: 2023年4月5日(星期三) 晚上10:58 收件人: @.>; 抄送: @.>; @.>; 主题: Re: [dvlab-research/spconv-plus] usage of spconv-plus (Issue #3)

Hi,

This line can be directly removed. It is unneeded. I have updated this line. https://github.com/dvlab-research/FocalsConv/blob/master/OpenPCDet/pcdet/models/backbones_3d/focal_sparse_conv/focal_sparse_conv_cuda.py

Regards, Yukang Chen

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>