ncnn icon indicating copy to clipboard operation
ncnn copied to clipboard

onnx2ncnn error

Open taxuezcy opened this issue 3 years ago • 3 comments

error log | 日志或报错信息 | [libprotobuf FATAL google/protobuf/generated_message_util.cc:818] CHECK failed: (scc->visit_status.load(std::memory_order_relaxed)) == (SCCInfoBase::kRunning):

terminate called after throwing an instance of 'google::protobuf::FatalException' what(): CHECK failed: (scc->visit_status.load(std::memory_order_relaxed)) == (SCCInfoBase::kRunning):

context | 编译/运行环境 protobuf 3.9.2

how to reproduce | 复现步骤 | 再現方法

more | 其他 | その他

@nihui ,

taxuezcy avatar May 10 '22 07:05 taxuezcy

@taxuezcy Hi, the reported error message is not so informative. We just see it is an error print by the protobuf library, can't make sure if this caused by onnx2ncnn.

e.g. If we search the key word visit_status in ncnn source project, we get nothing.

Please consider add more informations, including:

  • Complete call stack for the error message
  • The C++ code or command line arguments that produce this error
  • Necessary asset files such as the model file
  • Version info about the onn2ncnn
  • Build info, such as your system, etc.

zchrissirhcz avatar May 11 '22 06:05 zchrissirhcz

@taxuezcy Hi, the reported error message is not so informative. We just see it is an error print by the protobuf library, can't make sure if this caused by onnx2ncnn.

e.g. If we search the key word visit_status in ncnn source project, we get nothing.

Please consider add more informations, including:

  • Complete call stack for the error message
  • The C++ code or command line arguments that produce this error
  • Necessary asset files such as the model file
  • Version info about the onn2ncnn
  • Build info, such as your system, etc.

this error I have solved ,thank you for your reply!

taxuezcy avatar May 12 '22 06:05 taxuezcy

针对onnx模型转换的各种问题,推荐使用最新的pnnx工具转换到ncnn In view of various problems in onnx model conversion, it is recommended to use the latest pnnx tool to convert your model to ncnn

pip install pnnx
pnnx model.onnx inputshape=[1,3,224,224]

详细参考文档 Detailed reference documentation https://github.com/pnnx/pnnx https://github.com/Tencent/ncnn/wiki/use-ncnn-with-pytorch-or-onnx#how-to-use-pnnx

nihui avatar Aug 19 '24 03:08 nihui