TDA4VH-Q1: Error reporting issues when deploying onnx model on TDA4VH

chendy10

Part Number: TDA4VH-Q1

According GitHub - TexasInstruments/edgeai-tidl-tools: Edgeai TIDL Tools and Examples - This repository contains Tools and example developed for Deep learning runtime (DLRT) offering provided by TI’s edge AI solutions.

the official document was used for installation, but the following issues were encountered during the installation process:

Please refer to the attachment for installation and detailed execution steps: output_log.txt

The following is our operating environment:

Question 1：Under edgeai-tidl-tools$ when we run source ./scripts/run_python_examples.sh :

google.protobuf.message.DecodeError: Error parsing message with type 'onnx.ModelProto'

and onnxruntime.capi.onnxruntime_pybind11_state.InvalidProtobuf: [ONNXRuntimeError] : 7 : INVALID_PROTOBUF : Load model from ../../../models/public/resnet18_opset9.onnx failed:Protobuf parsing failed.

and Error : Error Code = <ERR_UNSUPPORTED_DATA_TYPE>

and Could not open /home/ubuntu/Documents/edgeai-tidl-tools/edgeai-tidl-tools/model-artifacts/cl-dlr-tflite_inceptionnetv3/tempDir/subgraph0_net/perfSimInfo.bin

Question 2：We copy:model-artifacts/models/to tda4VH-Q1 development board，when we run examples/osrt_python/ort/onnxrt_ep.py ,we encounter these mistake

output_log(1).txt

over 2 years ago

0 Pratik Kedar over 2 years ago

TI__Mastermind 24041 points

Hi,

Can you share which TIDL tools version you are using ?

You can check the git tag that you have cloned for the same.

Regards,

Pratik

0 chendy10 over 2 years ago in reply to Pratik Kedar

Prodigy 30 points

TIDL tools version : edgeai-tidl-tools - 08_06_00_03

0 chendy10 over 2 years ago

Prodigy 30 points

one more error when we run onnx : Unable to open network file ../../../model-artifacts//cl-ort-resnet18-v1//191_tidl_net.bin

0 chendy10 over 2 years ago

Prodigy 30 points

when we run ' python3 onnxrt_ep.py -c ' ,we meet this erroe, Is it because our model has unsupported layers?How can I solve this error?

0 Pratik Kedar over 2 years ago in reply to chendy10

TI__Mastermind 24041 points

Hi ,

Just wanted to check, is question 1 and 2 posted above are resolved ?

This thread caters to running standard supported model inside sdk, however the recent comment is specific to custom model compilation.

Regards,

Pratik

+1 chendy10 over 2 years ago in reply to Pratik Kedar

Prodigy 30 points

Problem 1 and 2 were resolved by downloading the model again, but when running the custom model, the error in the image above occurred

0 Pratik Kedar over 2 years ago in reply to chendy10

TI__Mastermind 24041 points

Great to hear that chendy !

We would recommend to create new thread for custom model compilation as this current thread caters to running standard model inside sdk, this will help us to track and assign issue to correct person if needed so.

Closing this thread.

chendy10 said:
when we run ' python3 onnxrt_ep.py -c ' ,we meet this erroe, Is it because our model has unsupported layers?How can I solve this error?

Please create new thread for the same.

Thank you.

Pratik

Processors

Processors forum

TDA4VH-Q1: Error reporting issues when deploying onnx model on TDA4VH