onnx qnn 部署模型问题
wangshankun opened this issue · 7 comments
wangshankun commented
Tracin commented
这里应该是DequantizeLinear
wangshankun commented
嗯,不过maxpool也能支持uint8/int8吧,我把maxpool放到了qlinear_op_type里面,算是暂时绕开这个错误;
另外,onnx已经支持FC了,量化算子是QLinearMatMul; onnx_qnn应该要把这个算子加上了吧?
Tracin commented
嗯,不过maxpool也能支持uint8/int8吧,我把maxpool放到了qlinear_op_type里面,算是暂时绕开这个错误;
另外,onnx已经支持FC了,量化算子是QLinearMatMul; onnx_qnn应该要把这个算子加上了吧?
QLinearMatMul不完全是FC, 因为缺少bias add操作.
wangshankun commented
wangshankun commented
是的, Pytorch实现的难点是无法统计MatMul和Add中间Tensor的数值范围, 给ONNX提了QLinearFC的issue.
Add要求bias 8bit,这个有点不符合部署实情,那我还是在deploy_onnx_qnn里面用1x1卷积代替FC吧;
别的不说,至少resnet18和50得支持了吧
github-actions commented
This issue has not received any updates in 120 days. Please reply to this issue if this still unresolved!