RuntimeError: step!=1 is currently not supported

[btwbtw01]

Hi, everyone
I'm trying to turn resnet20 brevitas network into ONNX graph. However, it shows RuntimeError: step!=1 is currently not supported

Here is my resnet20 brevitas network.

class LambdaLayer(nn.Module):
def init(self, lambd):
super(LambdaLayer, self).init()
self.lambd = lambd

def forward(self, x):
    return self.lambd(x)

class BasicBlock(nn.Module):
expansion = 1

def __init__(self, in_planes, planes, stride=1, option='A', **kwargs):
    super(BasicBlock, self).__init__()
    self.conv1 = QuantConv2d(in_planes, planes, kernel_size=3, 
                                   stride=stride, padding=1, bias=False, 
                                   weight_bit_width=8)
    self.bn1 = nn.BatchNorm2d(planes)
    self.conv2 = QuantConv2d(planes, planes, kernel_size=3, 
                                   stride=1, padding=1, bias=False, 
                                   weight_bit_width=8)
    self.bn2 = nn.BatchNorm2d(planes)
    self.relu = QuantReLU(bit_width=8)        

    self.shortcut = nn.Sequential()
    if stride != 1 or in_planes != planes:
        if option == 'A':
            """
            For CIFAR10 ResNet paper uses option A.
            """
            self.shortcut = LambdaLayer(lambda x:
                                        F.pad(x[:, :, ::2, ::2], (0, 0, 0, 0, planes//4, planes//4), "constant", 0))
        '''
        elif option == 'B':
            self.shortcut = nn.Sequential(
                 q.QuantizedConv2d(in_planes, self.expansion * planes, 
                                   kernel_size=1, stride=stride, bias=False, 
                                   wbits=kwargs['wbits'], abits=kwargs['abits']),
                 nn.BatchNorm2d(self.expansion * planes)
            )
        '''
def forward(self, x):
    out = self.relu(self.bn1(self.conv1(x)))
    out = self.bn2(self.conv2(out))
    out += self.shortcut(x)
    out = self.relu(out)
    return out

class ResNet(nn.Module):
def init(self, block, num_blocks, num_classes=10, **kwargs):
super(ResNet, self).init()
self.in_planes = 16

    self.conv1 = QuantConv2d(3, 16, kernel_size=3, stride=1, padding=1, bias=False, weight_bit_width=8)
    self.bn1 = nn.BatchNorm2d(16)
    self.layer1 = self._make_layer(block, 16, num_blocks[0], stride=1, **kwargs)
    self.layer2 = self._make_layer(block, 32, num_blocks[1], stride=2, **kwargs)
    self.layer3 = self._make_layer(block, 64, num_blocks[2], stride=2, **kwargs)
    self.linear = QuantLinear(64, num_classes, bias=False, weight_bit_width=8)
    self.relu = QuantReLU(bit_width=8)       
    self.finalpool2d = QuantMaxPool2d(kernel_size=8, stride=1)
    self.apply(_weights_init)

def _make_layer(self, block, planes, num_blocks, stride, **kwargs):
    strides = [stride] + [1]*(num_blocks-1)
    layers = []
    for stride in strides:
        layers.append(block(self.in_planes, planes, stride, **kwargs))
        self.in_planes = planes * block.expansion

    return nn.Sequential(*layers)

def forward(self, x):
    out = self.relu(self.bn1(self.conv1(x)))
    out = self.layer1(out)
    out = self.layer2(out)
    out = self.layer3(out)
    #print(out.shape)
    #out = F.avg_pool2d(out, out.size()[3])
    out = self.finalpool2d(out)
    #print(out.shape)
    out = out.view(out.size(0), -1)
    out = self.linear(out)
    return out

def resnet20(num_classes=10, **kwargs):
return ResNet(BasicBlock, [3, 3, 3], num_classes=num_classes, **kwargs)

Can anyone help me? Thanks!!!