About version of numpy

[xioatian1]

Hi, Thanks for your open-source code, I'm trying to reappear your code.
But I encountered some problems, I successfully installed the environment according to#Issues 1, and also processed the data. But when I run it, it shows_**ValueError: numpy.ndarray size changed, may indicate binary incompatibility. Expected 96 from C header, got 80 from PyObject**_
I searched for related solutions, then improved the numpy version. But mmdet3d 0.11.0 requires numpy<=1.20.0. I tried to improve the numpy version and run it, but there was a problem with the dataset.

Original Traceback (most recent call last):

I think it is the reason for the numpy version. Could you please provide some advice? Thanks a lot!

ys.platform: linux
Python: 3.8.16 (default, Mar 2 2023, 03:21:46) [GCC 11.2.0]
CUDA available: True
GPU 0,1: NVIDIA GeForce RTX 3090
CUDA_HOME: /home/lxt/cuda-11.1
NVCC: Build cuda_11.1.TC455_06.29190527_0
GCC: gcc (Ubuntu 5.5.0-12ubuntu1) 5.5.0 20171010
PyTorch: 1.9.0+cu111
PyTorch compiling details: PyTorch built with:

• GCC 7.3
• C++ Version: 201402
• Intel(R) Math Kernel Library Version 2020.0.0 Product Build 20191122 for Intel(R) 64 architecture applications
• Intel(R) MKL-DNN v2.1.2 (Git Hash 98be7e8afa711dc9b66c8ff3504129cb82013cdb)
• OpenMP 201511 (a.k.a. OpenMP 4.5)
• NNPACK is enabled
• CPU capability usage: AVX2
• CUDA Runtime 11.1
• NVCC architecture flags: -gencode;arch=compute_37,code=sm_37;-gencode;arch=compute_50,code=sm_50;-gencode;arch=compute_60,code=sm_60;-gencode;arch=compute_70,code=sm_70;-gencode;arch=compute_75,code=sm_75;-gencode;arch=compute_80,code=sm_80;-gencode;arch=compute_86,code=sm_86
• CuDNN 8.0.5
• Magma 2.5.2
• Build settings: BLAS_INFO=mkl, BUILD_TYPE=Release, CUDA_VERSION=11.1, CUDNN_VERSION=8.0.5, CXX_COMPILER=/opt/rh/devtoolset-7/root/usr/bin/c++, CXX_FLAGS= -Wno-deprecated -fvisibility-inlines-hidden -DUSE_PTHREADPOOL -fopenmp -DNDEBUG -DUSE_KINETO -DUSE_FBGEMM -DUSE_QNNPACK -DUSE_PYTORCH_QNNPACK -DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE -O2 -fPIC -Wno-narrowing -Wall -Wextra -Werror=return-type -Wno-missing-field-initializers -Wno-type-limits -Wno-array-bounds -Wno-unknown-pragmas -Wno-sign-compare -Wno-unused-parameter -Wno-unused-variable -Wno-unused-function -Wno-unused-result -Wno-unused-local-typedefs -Wno-strict-overflow -Wno-strict-aliasing -Wno-error=deprecated-declarations -Wno-stringop-overflow -Wno-psabi -Wno-error=pedantic -Wno-error=redundant-decls -Wno-error=old-style-cast -fdiagnostics-color=always -faligned-new -Wno-unused-but-set-variable -Wno-maybe-uninitialized -fno-math-errno -fno-trapping-math -Werror=format -Wno-stringop-overflow, LAPACK_INFO=mkl, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, TORCH_VERSION=1.9.0, USE_CUDA=ON, USE_CUDNN=ON, USE_EXCEPTION_PTR=1, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_MKL=ON, USE_MKLDNN=ON, USE_MPI=OFF, USE_NCCL=ON, USE_NNPACK=ON, USE_OPENMP=ON,

TorchVision: 0.10.0+cu111
OpenCV: 4.7.0
MMCV: 1.3.18
MMCV Compiler: GCC 7.3
MMCV CUDA Compiler: 11.1
MMDetection: 2.10.0
MMDetection3D: 0.11.0+

[SxJyJay]

From your snapshot, it seems that a file of virtual points doesn't exist. Maybe you should check the completeness of your downloaded virtual points.

[xioatian1]

From your snapshot, it seems that a file of virtual points doesn't exist. Maybe you should check the completeness of your downloaded virtual points.

I am sure the file is complete, and the virtual point file displayed by the error can also be searched, but the code can only start when I set numpy==1.23.x, other versions of numpy will report an error at the beginning, it may be numpy problem?

[SxJyJay]

I use numpy==1.19.2 as shown here. You can use ipdb.set_trace() to insert a breakpoint right before the line of code that makes error so as to see what happened.