Xiaoyang Wu
Xiaoyang Wu
Hi, thanks for pointing out the issue. I will fix it in the next version
What about tomorrow?
Hi, I am still on holiday and I will try my best to release all left configs before Feb 2024. But I have already noted your request, I will release...
Hi, Pointcept currently does not support gradient accumulation but I do plan to add it in the next version. You can consider modifying the code here (https://github.com/Pointcept/Pointcept/blob/main/pointcept/engines/train.py#L183). You can access...
> We also want to deploy the model. We were checking the dependencies and the most problematic issue seems like spconv dependency. There is a comment for substituting SubMConv3D, which...
> Hi, I want to test spunet and pt series on my own data. What inference framework dose your code support? spconv seems not support jit. Any advise to convert...
Hi, I will gradually release model weights for PTv3 and PTv2 in the future.
Hi all, one choice is to replace SpConv with OCNN. Pengshuai rewrites OCNN with Pytorch and it is easy to deploy.
Fundamentally, they are the same thing
Hi, the example config might not be the best setting, tuning augmentation (whether distort, scale level, crop point number), learning rate, and weight decay are needed. Also maybe adding a...