onnxruntime.capi.onnxruntime_pybind11_state.InvalidArgument: [ONNXRuntimeError] : 2 : INVALID_ARGUMENT : Invalid Output Name:score_8

[jinzaz]

I have downloaded the onnx model according to #99 issue , but now I still encounter a problem. I don't know how to change it

[jinzaz]

oh i fined this bug source,is that i was not exeis centerface.onnx in models directory, i search source code find it.
class DetectorInfer:

def __init__(self, model='/models/onnx/centerface/centerface.onnx',
             output_order=None, **kwargs):

[jinzaz]

@felixdollack This problem seems to have not been solved ，retinaface_r50_v1.onnx and centerface.onnx I have both of these documents

[felixdollack]

Hey @jinzaz, not sure why you tagged me. I am not affiliated with the project.
The error says something about an invalid argument given to a C/C++ function through PyBind.
And at the end python is missing the package cupy.
Are you trying to run the GPU or CPU version?

[SthPhoenix]

Hey @jinzaz, not sure why you tagged me. I am not affiliated with the project. The error says something about an invalid argument given to a C/C++ function through PyBind. And at the end python is missing the package cupy. Are you trying to run the GPU or CPU version?

Message about missing cupy fires in CPU version, though it's not required there, maybe I should remove this message.

I think most relevant error is from line 7 at screenshot, ONNX hash mismatch, it seems model was broken during download and script haven't succeeded in redownloading it.

@jinzaz, which models you are trying to use? default model from deploy_cpu.sh?

[jinzaz]

@SthPhoenix i use the deploy_cpu.sh , Because my region can't access drive.google.com , So I download the corresponding onnx file in the directory in advance, and then execute deploy_ cpu.sh

[jinzaz]

Hey @jinzaz, not sure why you tagged me. I am not affiliated with the project. The error says something about an invalid argument given to a C/C++ function through PyBind. And at the end python is missing the package cupy. Are you trying to run the GPU or CPU version?

sorry I'm @ wrong person

[SthPhoenix]

@SthPhoenix i use the deploy_cpu.sh , Because my region can't access drive.google.com , So I download the corresponding onnx file in the directory in advance, and then execute deploy_ cpu.sh

Can you calculate md5 sums of downloaded models and post them here?

[jinzaz]

@SthPhoenix i use the deploy_cpu.sh , Because my region can't access drive.google.com , So I download the corresponding onnx file in the directory in advance, and then execute deploy_ cpu.sh

Can you calculate md5 sums of downloaded models and post them here?

what? I dont't know what your mean, this is my project directory of models ,i downloaded all noox file in this directory
Do you mean that the downloaded model can only be used after md5 verification? :thinking

[SthPhoenix]

Just run for example md5sum glintr100.onnx on downloaded models.
It should be:

50febd32caa699ef7a47cf7422c56bbd  scrfd_2.5g_gnkps.onnx
3b366b98f786426f79629ddb2e56629c  glintr100.onnx

If it's different - you need to re-download the models

[kuanyshbakytuly]

Did you solve?