My tflite is occupying an excessive amount of memory, please seek help!

I am trying to convert my onnx model into a tflite model. My model has very few parameters, but when I use it, it consumes a lot of memory (estimated to be 3gb or more) and even with fp16, the problem has not been alleviated. May I ask where I went wrong?

https://drive.google.com/file/d/1chq9Na76fSLQizOQmPsz5fgcY_u8XOd1/view?usp=drive_link, https://drive.google.com/file/d/10Oqd0dhK8Wi8wvGspr3XZCgsW8JKvyhQ/view?usp=drive_link, https://drive.google.com/file/d/1ZoaagBlrs76zKUsf-2LQwmIzNPrRRLr9/view?usp=drive_link

Hi @jun_yin, Could you please let us know how you have converted your onnx model to TFLite, like converting your onnx model to tensorFlow and then convert it to lite. If not could you please try doing that. Also I don’t have access to the links you have given to see your code. Thank You.