Inference VGG16 caffe model can not find the output

Robotics & Edge Computing Jetson & Embedded Systems Jetson Xavier NX
5

Hi,

We do test VGG on Jetson in our benchmark.
The model we used can be downloaded with the below command:

$ wget https://www.dropbox.com/s/t4qq079g5q4jibx/vgg19_N2.prototxt

The problem with this error is that we only support the original prototxt definition.
Please convert the file into the supported version to fix this issue.

For example, please convert

layers {
  bottom: "data"
  top: "conv1_1"
  name: "conv1_1"
  type: CONVOLUTION
  convolution_param {
    num_output: 64
    pad: 1
    kernel_size: 3
  }
}

into

layer {
  bottom: "data"
  top: "conv1_1"
  name: "conv1_1"
  type: "Convolution"
  convolution_param {
    num_output: 64
    pad: 1
    kernel_size: 3
  }
}

Thanks