[Tensorflow SENet support] Unknown layer type: Axpy

[ykamikawa]

Platform (like ubuntu 16.04/win10):ubuntu 16.04

Python version:2.7

Source framework with version (like Tensorflow 1.4.1 with GPU): tensorflow

Destination framework with version (like CNTK 2.3 with GPU): IR

Pre-trained model path (webpath or webdisk path): https://drive.google.com/open?id=1iq703NMBgp4YFgbCetjiZzZ5KGtB2o84 https://github.com/hujie-frank/SENet

Running scripts: -> # python -m mmdnn.conversion._script.convertToIR -f caffe -d se_inception_resnet_v2 -n SENet.prototxt -w SENet.caffemodel

[ykamikawa]

Check failed: registry.count(type) == 1 (0 vs. 1) Unknown layer type: Axpy (known types: AbsVal, Accum, Accuracy, ArgMax, BNLL, BatchNorm, BatchReindex, Bias, BlackAugmentation, ChannelNorm, Concat, ContrastiveLoss, Convolution, Correlation, Correlation1D, Crop, CustomData, Data, DataAugmentation, Deconvolution, DisparityData, Downsample, Dropout, DummyData, ELU, Eltwise, Embed, EuclideanLoss, Exp, FLOWriter, Filter, Flatten, FloatReader, FloatWriter, FlowAugmentation, FlowWarp, GenerateAugmentationParameters, HDF5Data, HDF5Output, HingeLoss, Im2col, ImageData, ImgReader, ImgWriter, InfogainLoss, InnerProduct, Input, L1Loss, LRN, LSTM, LSTMUnit, Log, LpqLoss, MVN, Mean, MemoryData, MultinomialLogisticLoss, NegReLU, PFMWriter, PReLU, Parameter, Pooling, Power, Python, RNN, ReLU, Reduction, Resample, Reshape, SPP, Scale, Sigmoid, SigmoidCrossEntropyLoss, Silence, Slice, Softmax, SoftmaxWithLoss, Split, TanH, Threshold, Tile, WindowData)

[kitstar]

Hi @ykamikawa , the model contains the custom layer Axpy. MMdnn can't handle custom layer currently. Thanks.