TDA4VM: I try to use edgeai-tidl-tools to compile the model, but I have a lot of problems

Hello Gangqiang,

Can you make sure that you ran "source ./setup.sh" before running the demo: https://github.com/TexasInstruments/edgeai-tidl-tools/blob/master/README.md#setup

This downloads the runtime libraries from TI repository and sets up some environment variables to use those libraries. I am able to run tfl runtime demos on the PC without having to manually install any add-ons for PC emulation mode. 

On the other hand, I am seeing some different errors running onnx runtime demos:

TypeError: initialize_session(): incompatible function arguments. The following argument types are supported:
    1. (self: onnxruntime.capi.onnxruntime_pybind11_state.InferenceSession, arg0: List[str], arg1: List[Dict[str, str]], arg2: Set[str]) -> None

Let me know if the initial error is resolved by running the setup script, and whether you see additional issues.

Regards,

Takuma

We have validated the release with Github action. Please refer below for the complete test log.

github.com/.../5830738221

Hello Takuma，

I have run "source ./setup.sh" before running the demo. I have successfully run tfl runtime demos on the PC using “python3 tflrt_delegate.py -c”.

But when I run “python3 onnxrt_ep.py -c”, I have a error, as follows：

sess.initialize_session(providers, provider_options, disabled_optimizers)
RuntimeError: Unknown Provider Type: TIDLCompilationProvider

And I run “python3 onnxrt_ep.py -d” on the PC, An error occurred：

benchmark_dict = interpreter.get_TI_benchmark_data()
AttributeError: 'InferenceSession' object has no attribute 'get_TI_benchmark_data'

I didn't get your error，and I don't know why.

Regards,

Gangqiang

I have solved my problem according to your method, thanks