Unnormal Params In SK Attention

[99HU]

i Add SK Attention after the MobileNetV2 backbone, When i test the flops and params i find out an unnormal params nummber. Is there some mistake in my sittings or in the skAttention code?
channel=1280, reduction=8 when i use the skattention

+----------------------------------+----------------------+------------+--------------+
| module | #parameters or shape | #flops | #activations |
+----------------------------------+----------------------+------------+--------------+
| model | 0.141G | 9.218G | 9.055M |
| backbone | 2.224M | 0.409G | 8.722M |
| backbone.conv1 | 0.928K | 15.204M | 0.524M |
| backbone.conv1.conv | 0.864K | 14.156M | 0.524M |
| backbone.conv1.bn | 64 | 1.049M | 0 |
| backbone.layer1.0.conv | 0.896K | 14.68M | 0.786M |
| backbone.layer1.0.conv.0 | 0.352K | 5.767M | 0.524M |
| backbone.layer1.0.conv.1 | 0.544K | 8.913M | 0.262M |
| backbone.layer2 | 13.968K | 78.447M | 3.342M |
| backbone.layer2.0.conv | 5.136K | 42.271M | 2.064M |
| backbone.layer2.1.conv | 8.832K | 36.176M | 1.278M |
| backbone.layer3 | 39.696K | 52.15M | 1.622M |
| backbone.layer3.0.conv | 10K | 21.742M | 0.77M |
| backbone.layer3.1.conv | 14.848K | 15.204M | 0.426M |
| backbone.layer3.2.conv | 14.848K | 15.204M | 0.426M |
| backbone.layer4 | 0.184M | 52.085M | 0.901M |
| backbone.layer4.0.conv | 21.056K | 10.404M | 0.262M |
| backbone.layer4.1.conv | 54.272K | 13.894M | 0.213M |
| backbone.layer4.2.conv | 54.272K | 13.894M | 0.213M |
| backbone.layer4.3.conv | 54.272K | 13.894M | 0.213M |
| backbone.layer5 | 0.303M | 77.611M | 0.86M |
| backbone.layer5.0.conv | 66.624K | 17.056M | 0.221M |
| backbone.layer5.1.conv | 0.118M | 30.278M | 0.319M |
| backbone.layer5.2.conv | 0.118M | 30.278M | 0.319M |
| backbone.layer6 | 0.795M | 61.735M | 0.461M |
| backbone.layer6.0.conv | 0.155M | 20.775M | 0.195M |
| backbone.layer6.1.conv | 0.32M | 20.48M | 0.133M |
| backbone.layer6.2.conv | 0.32M | 20.48M | 0.133M |
| backbone.layer7.0.conv | 0.474M | 30.331M | 0.143M |
| backbone.layer7.0.conv.0 | 0.156M | 9.953M | 61.44K |
| backbone.layer7.0.conv.1 | 10.56K | 0.676M | 61.44K |
| backbone.layer7.0.conv.2 | 0.308M | 19.702M | 20.48K |
| backbone.conv2 | 0.412M | 26.378M | 81.92K |
| backbone.conv2.conv | 0.41M | 26.214M | 81.92K |
| backbone.conv2.bn | 2.56K | 0.164M | 0 |
| neck | 0.139G | 8.81G | 0.333M |
| neck.sk | 0.139G | 8.81G | 0.333M |
| neck.sk.convs | 0.138G | 8.809G | 0.328M |
| neck.sk.fc | 0.205M | 0.205M | 0.16K |
| neck.sk.fcs | 0.824M | 0.819M | 5.12K |
| neck.gap | | 81.92K | 0 |
| head | 74.28K | 20.48K | 16 |
| head.loss_module.flow_model | 53.784K | | |
| head.loss_module.flow_model.s | 26.892K | | |
| head.loss_module.flow_model.t | 26.892K | | |
| head.fc | 20.496K | 20.48K | 16 |
| head.fc.weight | (16, 1280) | | |
| head.fc.bias | (16,) | | |
+----------------------------------+------------
Here is the parameters I testd,Any help will be important to me,i am looking forward to your help