When the cubeai tool is used to analyze the model, the unquantized model can be analyzed normally, but the static quantization analysis can make mistakes. I use the onnx model. The following is my quantified model. I hope someone can help me.
Thank you very much for your answer, I found the problem. I did not perform the same transform as in the training process when generating npy data, nor did I perform the same data transform when generating calibration data in the static quantization ...
