Skip to content

[Bug] window10在部署运行cpu版本onnx推理模型的时候报错 #2880

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.

Sign up for GitHub

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

Jump to bottom
Open
2 of 3 tasks
fritest opened this issue Feb 12, 2025 · 2 comments
Open
2 of 3 tasks

[Bug] window10在部署运行cpu版本onnx推理模型的时候报错 #2880

fritest opened this issue Feb 12, 2025 · 2 comments

Comments

@fritest
Copy link

fritest commented Feb 12, 2025

Checklist

  • I have searched related issues but cannot get the expected help.
  • 2. I have read the FAQ documentation but cannot get the expected help.
  • 3. The bug has not been fixed in the latest version.

Describe the bug

按照这个教程https://github.com/open-mmlab/mmdeploy/blob/main/docs/zh_cn/02-how-to-run/prebuilt_package_windows.md
在运行到使用sdk推理时候运行出现问题，有时候没结果有时候报错。
.\image_classification.exe cpu C:\workspace\work_dir\onnx\resnet\ C:\workspace\mmpretrain\demo\demo.JPEG

有时候报错
调用加载模型时候报了这个错误
(mmdeploy) E:\zhp\code\python\mmdeployCode>python .\mmdeploy\demo\python\image_classification.py cpu .\work_dir\onnx\resnet\ .\mmpretrain\demo\demo.JPEG
loading mmdeploy_ort_net.dll ...
[2025-02-12 13:23:19.388] [mmdeploy] [info] [model.cpp:35] [DirectoryModel] Load model: ".\work_dir\onnx\resnet"
2025-02-12 13:23:19.5217686 [E:onnxruntime:, sequential_executor.cc:339 onnxruntime::SequentialExecutor::Execute] Non-zero status code returned while running Conv node. Name:'/backbone/layer2/layer2.0/conv2/Conv_output_0_nchwc' Status
Message: nchwc_ops.cc:194 onnxruntime::contrib::NchwcConv::Compute output and sum shape must match

下降了onnxruntime版本到教程上1.8.1，没有出现提示，但是也没有返回结果

Reproduction

python .\mmdeploy\demo\python\image_classification.py cpu .\work_dir\onnx\resnet\ .\mmpretrain\demo\demo.JPEG
按照教程出现这个问题，#2491
按照里面升级了onnxruntime版本，升级和不升级，转换出来的onnx模型在运行这条命令
python .\mmdeploy\demo\python\image_classification.py cpu .\work_dir\onnx\resnet\ .\mmpretrain\demo\demo.JPEG
一个报错，一个没有结果。

Environment

(mmdeploy) E:\zhp\code\python\mmdeployCode>python mmdeploy/tools/check_env.py
02/12 17:22:58 - mmengine - INFO -

02/12 17:22:58 - mmengine - INFO - **********Environmental information**********
02/12 17:23:00 - mmengine - INFO - sys.platform: win32
02/12 17:23:00 - mmengine - INFO - Python: 3.8.20 (default, Oct  3 2024, 15:19:54) [MSC v.1929 64 bit (AMD64)]
02/12 17:23:00 - mmengine - INFO - CUDA available: True
02/12 17:23:00 - mmengine - INFO - MUSA available: False
02/12 17:23:00 - mmengine - INFO - numpy_random_seed: 2147483648
02/12 17:23:00 - mmengine - INFO - GPU 0,1: NVIDIA GeForce RTX 4070
02/12 17:23:00 - mmengine - INFO - CUDA_HOME: C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.6
02/12 17:23:00 - mmengine - INFO - NVCC: Cuda compilation tools, release 11.6, V11.6.55
02/12 17:23:00 - mmengine - INFO - MSVC: 用于 x64 的 Microsoft (R) C/C++ 优化编译器 19.29.30152 版
02/12 17:23:01 - mmengine - INFO - GCC: n/a
02/12 17:23:01 - mmengine - INFO - PyTorch: 1.13.1+cu116
02/12 17:23:01 - mmengine - INFO - PyTorch compiling details: PyTorch built with:
  - C++ Version: 199711
  - MSVC 192829337
  - Intel(R) Math Kernel Library Version 2020.0.2 Product Build 20200624 for Intel(R) 64 architecture applications
  - Intel(R) MKL-DNN v2.6.0 (Git Hash 52b5f107dd9cf10910aaa19cb47f3abf9b349815)
  - OpenMP 2019
  - LAPACK is enabled (usually provided by MKL)
  - CPU capability usage: AVX2
  - CUDA Runtime 11.6
  - 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_61,code=sm_61;-gencode;arch=compute_70,code=sm_70;-gencode;arch=compute_75,c
ode=sm_75;-gencode;arch=compute_80,code=sm_80;-gencode;arch=compute_86,code=sm_86;-gencode;arch=compute_37,code=compute_37
  - CuDNN 8.3.2  (built against CUDA 11.5)
  - Magma 2.5.4
  - Build settings: BLAS_INFO=mkl, BUILD_TYPE=Release, CUDA_VERSION=11.6, CUDNN_VERSION=8.3.2, CXX_COMPILER=C:/actions-runner/_work/pytorch/pytorch/builder/windows/tmp_bin/sccache-cl.exe, CXX_FLAGS=/DWIN32 /D_WINDOWS /GR /EHsc /w /bigo
bj -DUSE_PTHREADPOOL -openmp:experimental -IC:/actions-runner/_work/pytorch/pytorch/builder/windows/mkl/include -DNDEBUG -DUSE_KINETO -DLIBKINETO_NOCUPTI -DUSE_FBGEMM -DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE -DEDGE_PROFILER_USE_
KINETO, LAPACK_INFO=mkl, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, TORCH_VERSION=1.13.1, 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=OFF, U
SE_NNPACK=OFF, USE_OPENMP=ON, USE_ROCM=OFF,

02/12 17:23:01 - mmengine - INFO - TorchVision: 0.14.1+cu116
02/12 17:23:01 - mmengine - INFO - OpenCV: 4.11.0
02/12 17:23:01 - mmengine - INFO - MMEngine: 0.10.6
02/12 17:23:01 - mmengine - INFO - MMCV: 2.0.0
02/12 17:23:01 - mmengine - INFO - MMCV Compiler: MSVC 192829924
02/12 17:23:01 - mmengine - INFO - MMCV CUDA Compiler: 11.6
02/12 17:23:01 - mmengine - INFO - MMDeploy: 1.3.1+
02/12 17:23:01 - mmengine - INFO -

02/12 17:23:01 - mmengine - INFO - **********Backend information**********
02/12 17:23:01 - mmengine - INFO - tensorrt:    None
02/12 17:23:01 - mmengine - INFO - ONNXRuntime: 1.8.1
02/12 17:23:01 - mmengine - INFO - ONNXRuntime-gpu:     None
02/12 17:23:01 - mmengine - INFO - ONNXRuntime custom ops:      Available
02/12 17:23:01 - mmengine - INFO - pplnn:       None
02/12 17:23:01 - mmengine - INFO - ncnn:        None
02/12 17:23:01 - mmengine - INFO - snpe:        None
02/12 17:23:01 - mmengine - INFO - openvino:    None
02/12 17:23:01 - mmengine - INFO - torchscript: 1.13.1+cu116
02/12 17:23:01 - mmengine - INFO - torchscript custom ops:      NotAvailable
02/12 17:23:01 - mmengine - INFO - rknn-toolkit:        None
02/12 17:23:01 - mmengine - INFO - rknn-toolkit2:       None
02/12 17:23:01 - mmengine - INFO - ascend:      None
02/12 17:23:01 - mmengine - INFO - coreml:      None
02/12 17:23:01 - mmengine - INFO - tvm: None
02/12 17:23:01 - mmengine - INFO - vacc:        None
02/12 17:23:01 - mmengine - INFO -

02/12 17:23:01 - mmengine - INFO - **********Codebase information**********
02/12 17:23:01 - mmengine - INFO - mmdet:       3.3.0
02/12 17:23:01 - mmengine - INFO - mmseg:       None
02/12 17:23:01 - mmengine - INFO - mmpretrain:  1.2.0
02/12 17:23:01 - mmengine - INFO - mmocr:       None
02/12 17:23:01 - mmengine - INFO - mmagic:      None
02/12 17:23:01 - mmengine - INFO - mmdet3d:     None
02/12 17:23:01 - mmengine - INFO - mmpose:      None
02/12 17:23:01 - mmengine - INFO - mmrotate:    None
02/12 17:23:01 - mmengine - INFO - mmaction:    None
02/12 17:23:01 - mmengine - INFO - mmrazor:     None
02/12 17:23:01 - mmengine - INFO - mmyolo:      None

(mmdeploy) E:\zhp\code\python\mmdeployCode>

Error traceback

(mmdeploy) E:\zhp\code\python\mmdeployCode>python .\mmdeploy\demo\python\image_classification.py cpu .\work_dir\onnx\resnet\ .\mmpretrain\demo\demo.JPEG
loading mmdeploy_ort_net.dll ...
[2025-02-12 13:23:19.388] [mmdeploy] [info] [model.cpp:35] [DirectoryModel] Load model: ".\work_dir\onnx\resnet\"
2025-02-12 13:23:19.5217686 [E:onnxruntime:, sequential_executor.cc:339 onnxruntime::SequentialExecutor::Execute] Non-zero status code returned while running Conv node. Name:'/backbone/layer2/layer2.0/conv2/Conv_output_0_nchwc' Status
Message: nchwc_ops.cc:194 onnxruntime::contrib::NchwcConv::Compute output and sum shape must match
@fritest fritest marked this as a duplicate of #2881 Feb 13, 2025
@fritest
Copy link
Author

fritest commented Feb 13, 2025

多次运行，有时候出现报错有时候没有报错。
Microsoft Windows [版本 10.0.19045.5487]
(c) Microsoft Corporation。保留所有权利。

(mmdeploy) E:\zhp\code\python\mmdeployCode>python .\mmdeploy\demo\python\image_classification.py cpu .\work_dir\onnx\resnet2\ .\mmpretrain\demo\demo.JPEG
loading mmdeploy_ort_net.dll ...
[2025-02-13 11:32:42.521] [mmdeploy] [info] [model.cpp:35] [DirectoryModel] Load model: ".\work_dir\onnx\resnet2"

(mmdeploy) E:\zhp\code\python\mmdeployCode>python .\mmdeploy\demo\python\image_classification.py cpu .\work_dir\onnx\resnet2\ .\mmpretrain\demo\demo.JPEG
loading mmdeploy_ort_net.dll ...
[2025-02-13 11:32:53.310] [mmdeploy] [info] [model.cpp:35] [DirectoryModel] Load model: ".\work_dir\onnx\resnet2"

(mmdeploy) E:\zhp\code\python\mmdeployCode>python .\mmdeploy\demo\python\image_classification.py cpu .\work_dir\onnx\resnet2\ .\mmpretrain\demo\demo.JPEG
loading mmdeploy_ort_net.dll ...
[2025-02-13 11:32:57.676] [mmdeploy] [info] [model.cpp:35] [DirectoryModel] Load model: ".\work_dir\onnx\resnet2"

(mmdeploy) E:\zhp\code\python\mmdeployCode>python .\mmdeploy\demo\python\image_classification.py cpu .\work_dir\onnx\resnet\ .\mmpretrain\demo\demo.JPEG
loading mmdeploy_ort_net.dll ...
[2025-02-13 11:33:07.555] [mmdeploy] [info] [model.cpp:35] [DirectoryModel] Load model: ".\work_dir\onnx\resnet"

(mmdeploy) E:\zhp\code\python\mmdeployCode>python .\mmdeploy\demo\python\image_classification.py cpu .\work_dir\onnx\resnet2\ .\mmpretrain\demo\demo.JPEG
loading mmdeploy_ort_net.dll ...
[2025-02-13 11:33:17.841] [mmdeploy] [info] [model.cpp:35] [DirectoryModel] Load model: ".\work_dir\onnx\resnet2"

(mmdeploy) E:\zhp\code\python\mmdeployCode>python .\mmdeploy\demo\python\image_classification.py cpu .\work_dir\onnx\resnet2\ .\mmpretrain\demo\demo.JPEG
loading mmdeploy_ort_net.dll ...
[2025-02-13 11:33:24.508] [mmdeploy] [info] [model.cpp:35] [DirectoryModel] Load model: ".\work_dir\onnx\resnet2"
2025-02-13 11:33:24.6331722 [E:onnxruntime:, sequential_executor.cc:339 onnxruntime::SequentialExecutor::Execute] Non-zero status code returned while running Conv node. Name:'/backbone/layer2/layer2.0/conv2/Conv_output_0_nchwc' Status
Message: nchwc_ops.cc:194 onnxruntime::contrib::NchwcConv::Compute output and sum shape must match

@fritest
Copy link
Author

fritest commented Feb 13, 2025

(mmdeploy) E:\zhp\code\python\mmdeployCode>python .\mmdeploy\demo\python\image_classification.py cpu .\work_dir\onnx\resnet2\ .\mmpretrain\demo\demo.JPEG
loading mmdeploy_ort_net.dll ...
[2025-02-13 14:27:28.343] [mmdeploy] [info] [model.cpp:35] [DirectoryModel] Load model: ".\work_dir\onnx\resnet2"
2025-02-13 14:27:28.4704312 [E:onnxruntime:, sequential_executor.cc:339 onnxruntime::SequentialExecutor::Execute] Non-zero status code returned while running Conv node. Name:'/backbone/layer1/layer1.0/conv2/Conv_output_0_nchwc' Status
Message: nchwc_ops.cc:194 onnxruntime::contrib::NchwcConv::Compute output and sum shape must match

(mmdeploy) E:\zhp\code\python\mmdeployCode>python .\mmdeploy\demo\python\image_classification.py cpu .\work_dir\onnx\resnet2\ .\mmpretrain\demo\demo.JPEG
loading mmdeploy_ort_net.dll ...
[2025-02-13 14:27:31.865] [mmdeploy] [info] [model.cpp:35] [DirectoryModel] Load model: ".\work_dir\onnx\resnet2"

(mmdeploy) E:\zhp\code\python\mmdeployCode>python .\mmdeploy\demo\python\image_classification.py cpu .\work_dir\onnx\resnet2\ .\mmpretrain\demo\demo.JPEG
loading mmdeploy_ort_net.dll ...
[2025-02-13 14:27:37.058] [mmdeploy] [info] [model.cpp:35] [DirectoryModel] Load model: ".\work_dir\onnx\resnet2"
58 0.3177702724933624
62 0.2017456740140915
65 0.12394838780164719
54 0.10745801031589508
49 0.10204800963401794

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
1 participant
@fritest