您好，我在量化网络时出现上采样层报错：

I0701 11:50:33.413280   277 net.cpp:269] Top shape: 1 18 9 11 (1782)
I0701 11:50:33.413285   277 net.cpp:277] Memory required for data: 50487480
I0701 11:50:33.413293   277 _factory.hpp:106] Creating 73-route
I0701 11:50:33.413300   277 net.cpp:209] Creating 73-route
I0701 11:50:33.413306   277 net.cpp:1007] 73-route <- 64-conv_inp_

I0701 11:50:33.413313   277 net.cpp:981] 73-route -> 73-route
I0701 11:50:33.413321   277 net.cpp:246] Setting up 73-route
I0701 11:50:33.413328   277 net.cpp:269] Top shape: 1 72 9 11 (7128)
I0701 11:50:33.413336   277 net.cpp:277] Memory required for data: 50515992
I0701 11:50:33.413341   277 _factory.hpp:106] Creating 74-upsample
I0701 11:50:33.413350   277 net.cpp:209] Creating 74-upsample
I0701 11:50:33.413357   277 net.cpp:1007] 74-upsample <- 73-route
I0701 11:50:33.413364   277 net.cpp:981] 74-upsample -> 74-upsample
F0701 11:50:33.413383   277 upsample_ .cpp:99] Check failed: 4 == bottom[1]->num_axes() (4 vs. 0) Input mask must have 4 axes, corresponding to (num, channels, height, width)
*** Check failure stack trace: ***
    @     0x7f52c3d4e0cd  google::LogMessage::Fail()
    @     0x7f52c3d4ff33  google::LogMessage::SendToLog()
    @     0x7f52c3d4dc28  google::LogMessage::Flush()
    @     0x7f52c3d50999  google::LogMessageFatal::~LogMessageFatal()
    @     0x7f52c4a92429  caffe::Upsample <>::Reshape()
    @     0x7f52c46a38fe  caffe:: <>::Reshape_tensor()
    @     0x7f52c4afac44  caffe::Net<>::Init()
    @     0x7f52c4afdbe0  caffe::Net<>::Net()
    @     0x5586b94eb32d  (unknown)
    @     0x5586b94d3df0  (unknown)
    @     0x7f52c1b30b97  __libc_start_main
    @     0x5586b94d3fea  (unknown)
Aborted (core dumped)

yoloface-500k-v2.zip

inp_inp