Errors in using importONNXLayers to import my pytorch model into Matlab

I am trying to run my pytorch models in matlab (2018b) via onnx (importONNXLayers), but I have encounted several errors. I do not know how to fix them.

The error messages are as follow:

layers = importONNXLayers('super_resolution.onnx')

警告: Imported layers have no output layer because ONNX files do not specify the network's output layer type. The

layers will not be trainable until an output layer is added. Either add an output layer to the imported layers, or

specify the output layer type using 'OutputLayerType' in the call to importONNXLayers.

> In nnet.internal.cnn.onnx.importONNXLayers>iValidateOutputLayerType (line 36)

In nnet.internal.cnn.onnx.importONNXLayers>iValidateInputs (line 28)

In nnet.internal.cnn.onnx.importONNXLayers (line 5)

In importONNXLayers (line 45)

错误使用 nnet.internal.cnn.onnx.transformConstantNodeToInitializer (line 27)

断言失败。

出错 nnet.internal.cnn.onnx.translateONNX>iReplaceWeightNodes (line 270)

[tf, dim, rawData] = nnet.internal.cnn.onnx.transformConstantNodeToInitializer(node);

出错 nnet.internal.cnn.onnx.translateONNX (line 46)

[thisGraph, initializerNames, initializerDimMap, initializerRawDataMap] = iReplaceWeightNodes(...

出错 nnet.internal.cnn.onnx.importONNXLayers (line 8)

LayersOrGraph = nnet.internal.cnn.onnx.translateONNX(modelProto, OutputLayerType, UserImageInputSize, ImportWeights);

出错 importONNXLayers (line 45)

Layers = nnet.internal.cnn.onnx.importONNXLayers(ModelFile, varargin{:});

My pytorch onnx model (which is from an example in https://pytorch.org/tutorials/advanced/super_resolution_with_onnxruntime.html):

class SuperResolutionNet(nn.Module):

def __init__(self, upscale_factor, inplace=False):

super(SuperResolutionNet, self).__init__()

self.relu = nn.ReLU(inplace=inplace)

self.conv1 = nn.Conv2d(1, 64, (5, 5), (1, 1), (2, 2))

self.conv2 = nn.Conv2d(64, 64, (3, 3), (1, 1), (1, 1))

self.conv3 = nn.Conv2d(64, 32, (3, 3), (1, 1), (1, 1))

self.conv4 = nn.Conv2d(32, upscale_factor ** 2, (3, 3), (1, 1), (1, 1))

self.pixel_shuffle = nn.PixelShuffle(upscale_factor)

self._initialize_weights()

def forward(self, x):

x = self.relu(self.conv1(x))

x = self.relu(self.conv2(x))

x = self.relu(self.conv3(x))

x = self.pixel_shuffle(self.conv4(x))

return x

def _initialize_weights(self):

init.orthogonal_(self.conv1.weight, init.calculate_gain('relu'))

init.orthogonal_(self.conv2.weight, init.calculate_gain('relu'))

init.orthogonal_(self.conv3.weight, init.calculate_gain('relu'))

init.orthogonal_(self.conv4.weight)

# Create the super-resolution model by using the above model definition.

torch_model = SuperResolutionNet(upscale_factor=3)

batch_size = 1 # just a random number

# set the model to inference mode

torch_model.eval()

# Input to the model

x = torch.randn(batch_size, 1, 224, 224, requires_grad=True)

torch_out = torch_model(x)

# Export the model

torch.onnx.export(torch_model, # model being run

x, # model input (or a tuple for multiple inputs)

"super_resolution.onnx", # where to save the model (can be a file or file-like object)

export_params=True, # store the trained parameter weights inside the model file

opset_version=10, # the ONNX version to export the model to

do_constant_folding=True, # whether to execute constant folding for optimization

input_names = ['input'], # the model's input names

output_names = ['output'], # the model's output names

dynamic_axes={'input' : {0 : 'batch_size'}, # variable length axes

'output' : {0 : 'batch_size'}})

Many thanks for the help!

Answer 1

[Feiran Zhao]

I am also puzzled by a similar problem when importing .onnx into matlab. The key is that, I am also using Matlab 2018b! I guess this is the problem.

[quino]

请问问题解决了吗？