Richard-luckyc

> > /home/wuhao/djx/oneformer3d/MinkowskiEngine/src/3rdparty/concurrent_unordered_map.cuh(503): 错误：命名空间“thrust”没有成员“device”返回thrust::count_if(thrust::device, m_hashtbl_values, m_hashtbl_values + m_capacity, ^ > > 在编译“/home/wuhao/djx/oneformer3d/MinkowskiEngine/src/broadcast_gpu.cu”时检测到 1 个错误。错误：命令‘/usr/local/cuda-12.1/bin/nvcc’失败，退出代码为 2 > > [#543 ](https://github.com/NVIDIA/MinkowskiEngine/issues/543) [#596](https://github.com/NVIDIA/MinkowskiEngine/issues/596) 我遇到了类似的问题，在这里找到了一些有价值的信息。我现在已经在 cuda12.4 + g++11 环境中成功编译。 How did you solve this...