run error with torch

[THHHomas]

I install all the package with the recommended version, cuda-11.1, torch=1.8.1, when I run "python train.py --config configs/config_shapenet.yaml", I encounter the following problem.
/home/magician/Workspace/anaconda3/envs/conv_onet/bin/python /snap/pycharm-professional/268/plugins/python/helpers/pydev/pydevd.py --multiproc --qt-support=auto --client 127.0.0.1 --port 45483 --file /disk2/POCO-main/train.py --config configs/config_shapenet.yaml
Connected to pydev debugger (build 212.5712.39)
INFO:root:Creating the network
INFO:root:InterpNet - Simple - K=64
INFO:root:Network -- backbone -- 12783956 parameters
INFO:root:Network -- projection -- 6498 parameters
INFO:root:Network -- Number of parameters 12790454
INFO:root:Getting the dataset
INFO:root:Dataset - ShapeNet- None
INFO:root:Dataset - len 22602
INFO:root:InterpNet - Simple - K=64
INFO:root:Network -- backbone -- 12783956 parameters
INFO:root:Network -- projection -- 6498 parameters
INFO:root:Dataset - ShapeNet- None
INFO:root:Dataset - len 3222
INFO:root:InterpNet - Simple - K=64
INFO:root:Network -- backbone -- 12783956 parameters
INFO:root:Network -- projection -- 6498 parameters
INFO:root:Creating the optimizer
INFO:root:Creating tensorboard summary writer
Traceback (most recent call last):
File "/snap/pycharm-professional/268/plugins/python/helpers/pydev/pydevd.py", line 1483, in _exec
pydev_imports.execfile(file, globals, locals) # execute the script
File "/snap/pycharm-professional/268/plugins/python/helpers/pydev/_pydev_imps/_pydev_execfile.py", line 18, in execfile
exec(compile(contents+"\n", file, 'exec'), glob, loc)
File "/disk2/POCO-main/train.py", line 331, in
main(config)
File "/disk2/POCO-main/train.py", line 194, in main
outputs = net(data, spectral_only=True)
File "/home/magician/Workspace/anaconda3/envs/conv_onet/lib/python3.7/site-packages/torch/nn/modules/module.py", line 889, in _call_impl
result = self.forward(*input, **kwargs)
File "/disk2/POCO-main/networks/network.py", line 45, in forward
latents = self.net(data, spectral_only=True)
File "/home/magician/Workspace/anaconda3/envs/conv_onet/lib/python3.7/site-packages/torch/nn/modules/module.py", line 889, in _call_impl
result = self.forward(*input, **kwargs)
File "/disk2/POCO-main/networks/backbone/fkaconv_network.py", line 194, in forward
x0 = self.resnetb01(x0, pos, pos, data["ids00"])
File "/home/magician/Workspace/anaconda3/envs/conv_onet/lib/python3.7/site-packages/torch/nn/modules/module.py", line 889, in _call_impl
result = self.forward(*input, **kwargs)
File "/disk2/POCO-main/networks/backbone/fkaconv_network.py", line 33, in forward
x = self.activation(self.bn1(self.cv1(x, pos, support_points, neighbors_indices)))
File "/home/magician/Workspace/anaconda3/envs/conv_onet/lib/python3.7/site-packages/torch/nn/modules/module.py", line 889, in _call_impl
result = self.forward(*input, **kwargs)
File "/disk2/POCO-main/lightconvpoint/nn/conv_fkaconv.py", line 140, in forward
features = torch.matmul(x.permute(0, 2, 1, 3), mat.permute(0, 2, 3, 1)).permute(0, 2, 1, 3)
RuntimeError: CUDA error: CUBLAS_STATUS_EXECUTION_FAILED when calling cublasSgemmStridedBatched( handle, opa, opb, m, n, k, &alpha, a, lda, stridea, b, ldb, strideb, &beta, c, ldc, stridec, num_batches)
python-BaseException

Process finished with exit code 1