This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

TDA4VM: Edgeai : ioBuf tensor for onnx tensor with matching name on customer model

Part Number: TDA4VM


Tool/software:

Hello Champs , 

we are trying to run our model with edgeai 9.20.7 https://github.com/TexasInstruments/edgeai-tidl-tools/tree/09_02_07_00?tab=readme-ov-file, and getting above error. Because of this the graph output doesn't get recognized by edgeai.

n TIDL_runtimesPostProcessNet 4
************ in TIDL_subgraphRtCreate ************ 
************ TIDL_subgraphRtCreate done ************ 
Warning : Couldn't find corresponding ioBuf tensor for onnx tensor with matching name 
Warning : Couldn't find corresponding ioBuf tensor for onnx tensor with matching name

log_e2e_c.txt

  • Hello,

    Thank you for providing the log. Before those warnings, I'm also seeing that you are getting the warning and error which could be causing the later warnings you mentioned and then then the issues with the graph output. 

    Warning : Requested Input Data Convert Layer is not Added to the network, It is currently not Optimal
    Error: Layer 18, Transpose_n_1:rout_tensor_1 is missing inputs in the network and cannot be topologically sorted
    Input 0: , dataId=38

    Did you verify that this model works without C7x offload (running with the -d option)?

    If so can you share the model, so I can inspect the specific layer that the model compilation is erroring on?

    Best,

    Asha