We have detected your current browser version is not the latest one. Xilinx.com uses the latest web technologies to bring you the best online experience possible. Please upgrade to a Xilinx.com supported browser:Chrome, Firefox, Internet Explorer 11, Safari. Thank you!

Showing results for 
Search instead for 
Did you mean: 
Visitor mghodsi
Registered: ‎11-16-2018

Error with DNNC for a custom network

I have a network that I trained and tested with Caffe. I used DECENT to quantize and successfully generated the output from DECENT "deploy.caffemodel" and "deploy.prototxt". Then I used DNNC to compile but I get the following error:

[DNNC][Fatal] Check failed for condition [fixed() && fixinfo_.size() > 3] in [/tmp/DNNC_V010_Package/dnnc/dnnc_impl/core/parameter.h:200] :Invalid fix info for [conv1], current fix info is [ 8 8 8 8 8 12 ]


and this is the fixed_point conv1 layer grabed from fix_train_test.prototxt:

layer {
  name: "conv1"
  type: "ConvolutionFixed"
  bottom: "data"
  top: "conv1"
  param {
    lr_mult: 1
  param {
    lr_mult: 0.1
  convolution_param {
    num_output: 64
    pad: 1
    kernel_size: 3
    stride: 1
    weight_filler {
      type: "msra"
      variance_norm: AVERAGE
    bias_filler {
      type: "constant"
      value: 0
  fixed_param {
    fixed_method: DIFF_S
    bit_width: 8


Can anybody help me fix this issue?

Tags (3)
0 Kudos