Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Cuda Illegal Memory 700 Error #716

Open
rangganast opened this issue Aug 19, 2024 · 0 comments
Open

Cuda Illegal Memory 700 Error #716

rangganast opened this issue Aug 19, 2024 · 0 comments

Comments

@rangganast
Copy link

rangganast commented Aug 19, 2024

Hi, I was trying voxel_gen.py included in the example of this repository

import numpy as np
import spconv.pytorch as spconv
from spconv.utils import Point2VoxelCPU3d
from spconv.pytorch.utils import PointToVoxel
import torch


pc = np.random.uniform(-2, 8, size=[100, 3]).astype(np.float32)

device = torch.device("cuda")

gen = PointToVoxel(vsize_xyz=[0.25, 0.25, 0.25],
                   coors_range_xyz=[0, 0, 0, 10, 10, 10],
                   num_point_features=3,
                   max_num_voxels=500,
                   max_num_points_per_voxel=5,
                   device=device)

pc_th = torch.from_numpy(pc).to(device)

voxels_th, indices_th, num_p_in_vx_th = gen(pc_th)

batch_idx = torch.zeros((indices_th.shape[0], 1), dtype=torch.int32).to(device)
indices_th = torch.cat((batch_idx, indices_th), dim=1)
voxels_th = torch.mean(voxels_th, dim=1)

voxel_feature = spconv.SparseConvTensor(features=voxels_th, indices=indices_th, 
                                        spatial_shape=[10, 10, 10], 
                                        batch_size=1)

simple_net = spconv.SubMConv3d(in_channels=3, out_channels=6,
                                 kernel_size=3, padding=1, stride=2)
output = simple_net(voxel_feature)

I got error:

mask_width, tune_res_cpp = ConvGemmOps.implicit_gemm(
RuntimeError: /io/build/temp.linux-x86_64-cpython-310/spconv/build/core_cc/src/cumm/conv/main/ConvMainUnitTest/ConvMainUnitTest_matmul_split_Simt_f32f32f32_0.cu:1047
cuda execution failed with error 700 an illegal memory access was encountered
Simt_f32f32f32f32f32tnt_m32n128k16m32n32k8A1_200_C301LLL_SK error with params [48, 3] [6, 27, 3] [48, 6] [27, 48] [48, 1] [48] [] -1

Please help me. Thanks.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant