Tool/software:
We quantized the 3dOD model smoke in edgeai , we use the script 'onnxrt_ep.py' under the edgeai tool to quantized the model and infered the bin model。the tensor_bit is 8 . We visualize the result , and it looks fine . But, when we run it on the TDA4VME board , the result is not right , we visualized the result , we can hardly find the 3d box in the picture , the result is terrible.
Wen I quantized the model again ,and set the quantized model to 16bit , we run ti on PC , the result is fine too. But when we try to run ti on the TDA4VE , it occurs to error.
How can we find the reason and solve it ?
Our edgeai is edgeai-tidl-tools-09_01_08_00 , Our sdk that run on the TDA4VE board is SDK09_01 。Any details can ask me , if you need .
the int16 model infer error
our onnx model
our int16 and int8 bin model