facebookresearch / PointContrast

Code for paper <PointContrast: Unsupervised Pretraining for 3D Point Cloud Understanding>
MIT License
328 stars 33 forks source link

pytorch version #31

Closed JudyYe closed 2 years ago

JudyYe commented 2 years ago

Hi friends,

I'm aware that the repo is out of maintenance but if you happen to run it on V100 and encounter the 'sample_gpu.cu' error

CUDA kernel failed : no kernel image is available for execution on the device
void furthest_point_sampling_kernel_wrapper(int, int, int, const float*, float*, int*) at L:233 in /private/home/yufeiy2/ssl3d/PointContrast/downstream/votenet_det_new/models/backbone/pointnet2/_ext_src/src/sampling_gpu.cu

consider to change your cuda version other than 9.0 / 10.2 suggested by this code. It took me lots of time to figure it out :/