[Quantization]Is there any prequantized model autotune tutorial or script?

I using prequantized tutorial on Deploy a Framework-prequantized Model with TVM — tvm 0.8.dev0 documentation. And tune it with kernel tuning, and seems graph tuning is not available for qnn op? The model after tuning perform a worse performance, is there something wrong with my script?

My script address:GitHub - CeasarLee/tvm_quant

Thank you for helping me!!!