为什么vitis后端每层都量化？

[jeannotes]

对于vitis后端，如果当前是一个conv，其实已经对这个conv做了fake convolution，为什么在当前层后面还加一个act quantization，不是重复吗，意义在哪儿

[Tracin]

QConv一般只是做了weight的fake quant，输入的fake quant需要插入额外节点

[jeannotes]

能再解释详细一点吗，没怎么理解

[github-actions[bot]]

This issue has not received any updates in 120 days. Please reply to this issue if this still unresolved!