VGG caffe is not working

Autonomous Machines Jetson & Embedded Systems Jetson TX2
1

Hello everyone.
I’m doing facial recognition using Caffe framework and VGG pre-trained model.
When I run the train_val.prototxt I got this error:

I0125 19:34:31.965451  5184 net.cpp:746] Ignoring source layer fc8
I0125 19:34:31.965508  5184 net.cpp:746] Ignoring source layer prob
I0125 19:34:32.007084  5184 solver.cpp:57] Solver scaffolding done.
I0125 19:34:32.011937  5184 caffe.cpp:239] Starting Optimization
I0125 19:34:32.011982  5184 solver.cpp:289] Solving Face_Recognition_VGG_NET
I0125 19:34:32.011994  5184 solver.cpp:290] Learning Rate Policy: step
I0125 19:34:32.246888  5184 solver.cpp:347] Iteration 0, Testing net (#0)
nvidia@tegra-ubuntu:~$

I’m using this train_val.prototxt:

name: "Face_Recognition_VGG_NET"
layer {
  name: "data"
  type: "Data"
  include {
    phase: TRAIN
  }
 transform_param {
    crop_size: 224
    mean_file: "/home/nvidia/Face-Recognition-using-VGG_FaceNet/mean_train.binaryproto"
    mirror: true
 }
 data_param {
    source: "/home/nvidia/caffe/examples/imagenet/ilsvrc12_train_lmdb"
    batch_size: 256
    backend: LMDB
  }
  top: "data"
  top: "label"
}
layer {
  name: "data"
  type: "Data"
  include {
    phase: TEST
  }
 transform_param {
    crop_size: 224
    mean_file: "/home/nvidia/Face-Recognition-using-VGG_FaceNet/mean_val.binaryproto"
    mirror: false
 }
 data_param {
    source: "/home/nvidia/caffe/examples/imagenet/ilsvrc12_val_lmdb"
    batch_size: 50
    backend: LMDB
  }
  top: "data"
  top: "label"
}
layer {
  bottom: "data"
  top: "conv1_1"
  name: "conv1_1"
  type: "Convolution"
  convolution_param {
    num_output: 64
    pad: 1
    kernel_size: 3
    stride: 1
  }
}
layer {
  bottom: "conv1_1"
  top: "conv1_1"
  name: "relu1_1"
  type: "ReLU"
}
layer {
  bottom: "conv1_1"
  top: "conv1_2"
  name: "conv1_2"
  type: "Convolution"
  convolution_param {
    num_output: 64
    pad: 1
    kernel_size: 3
    stride: 1
  }
}
layer {
  bottom: "conv1_2"
  top: "conv1_2"
  name: "relu1_2"
  type: "ReLU"
}
layer {
  bottom: "conv1_2"
  top: "pool1"
  name: "pool1"
  type: "Pooling"
  pooling_param {
    pool: MAX
    kernel_size: 2
    stride: 2
  }
}
layer {
  bottom: "pool1"
  top: "conv2_1"
  name: "conv2_1"
  type: "Convolution"
  convolution_param {
    num_output: 128
    pad: 1
    kernel_size: 3
    stride: 1
  }
}
layer {
  bottom: "conv2_1"
  top: "conv2_1"
  name: "relu2_1"
  type: "ReLU"
}
layer {
  bottom: "conv2_1"
  top: "conv2_2"
  name: "conv2_2"
  type: "Convolution"
  convolution_param {
    num_output: 128
    pad: 1
    kernel_size: 3
    stride: 1
  }
}
layer {
  bottom: "conv2_2"
  top: "conv2_2"
  name: "relu2_2"
  type: "ReLU"
}
layer {
  bottom: "conv2_2"
  top: "pool2"
  name: "pool2"
  type: "Pooling"
  pooling_param {
    pool: MAX
    kernel_size: 2
    stride: 2
  }
}
layer {
  bottom: "pool2"
  top: "conv3_1"
  name: "conv3_1"
  type: "Convolution"
  convolution_param {
    num_output: 256
    pad: 1
    kernel_size: 3
    stride: 1
  }
}
layer {
  bottom: "conv3_1"
  top: "conv3_1"
  name: "relu3_1"
  type: "ReLU"
}
layer {
  bottom: "conv3_1"
  top: "conv3_2"
  name: "conv3_2"
  type: "Convolution"
  convolution_param {
    num_output: 256
    pad: 1
    kernel_size: 3
    stride: 1
  }
}
layer {
  bottom: "conv3_2"
  top: "conv3_2"
  name: "relu3_2"
  type: "ReLU"
}
layer {
  bottom: "conv3_2"
  top: "conv3_3"
  name: "conv3_3"
  type: "Convolution"
  convolution_param {
    num_output: 256
    pad: 1
    kernel_size: 3
    stride: 1
  }
}
layer {
  bottom: "conv3_3"
  top: "conv3_3"
  name: "relu3_3"
  type: "ReLU"
}
layer {
  bottom: "conv3_3"
  top: "pool3"
  name: "pool3"
  type: "Pooling"
  pooling_param {
    pool: MAX
    kernel_size: 2
    stride: 2
  }
}
layer {
  bottom: "pool3"
  top: "conv4_1"
  name: "conv4_1"
  type: "Convolution"
  convolution_param {
    num_output: 512
    pad: 1
    kernel_size: 3
    stride: 1
  }
}
layer {
  bottom: "conv4_1"
  top: "conv4_1"
  name: "relu4_1"
  type: "ReLU"
}
layer {
  bottom: "conv4_1"
  top: "conv4_2"
  name: "conv4_2"
  type: "Convolution"
  convolution_param {
    num_output: 512
    pad: 1
    kernel_size: 3
    stride: 1
  }
}
layer {
  bottom: "conv4_2"
  top: "conv4_2"
  name: "relu4_2"
  type: "ReLU"
}
layer {
  bottom: "conv4_2"
  top: "conv4_3"
  name: "conv4_3"
  type: "Convolution"
  convolution_param {
    num_output: 512
    pad: 1
    kernel_size: 3
    stride: 1
  }
}
layer {
  bottom: "conv4_3"
  top: "conv4_3"
  name: "relu4_3"
  type: "ReLU"
}
layer {
  bottom: "conv4_3"
  top: "pool4"
  name: "pool4"
  type: "Pooling"
  pooling_param {
    pool: MAX
    kernel_size: 2
    stride: 2
  }
}
layer {
  bottom: "pool4"
  top: "conv5_1"
  name: "conv5_1"
  type: "Convolution"
  convolution_param {
    num_output: 512
    pad: 1
    kernel_size: 3
    stride: 1
  }
}
layer {
  bottom: "conv5_1"
  top: "conv5_1"
  name: "relu5_1"
  type: "ReLU"
}
layer {
  bottom: "conv5_1"
  top: "conv5_2"
  name: "conv5_2"
  type: "Convolution"
  convolution_param {
    num_output: 512
    pad: 1
    kernel_size: 3
    stride: 1
  }
}
layer {
  bottom: "conv5_2"
  top: "conv5_2"
  name: "relu5_2"
  type: "ReLU"
}
layer {
  bottom: "conv5_2"
  top: "conv5_3"
  name: "conv5_3"
  type: "Convolution"
  convolution_param {
    num_output: 512
    pad: 1
    kernel_size: 3
    stride: 1
  }
}
layer {
  bottom: "conv5_3"
  top: "conv5_3"
  name: "relu5_3"
  type: "ReLU"
}

layer {
  bottom: "conv5_3"
  top: "conv5_4"
  name: "conv5_4"
  type: "Convolution"
  convolution_param {
    num_output: 512
    pad: 1
    kernel_size: 3
  }
}
layer {
  bottom: "conv5_4"
  top: "conv5_4"
  name: "relu5_4"
  type: "ReLU"
}
layer {
  bottom: "conv5_4"
  top: "pool5"
  name: "pool5"
  type: "Pooling"
  pooling_param {
    pool: MAX
    kernel_size: 2
    stride: 2
  }
}
layer {
  bottom: "pool5"
  top: "fc6"
  name: "fc6"
  type: "InnerProduct"
  inner_product_param {
    num_output: 4096
  }
}
layer {
  bottom: "fc6"
  top: "fc6"
  name: "relu6"
  type: "ReLU"
}
layer {
  bottom: "fc6"
  top: "fc6"
  name: "drop6"
  type: "Dropout"
  dropout_param {
    dropout_ratio: 0.5
  }
}
layer {
  name: "fc7"
  type: "InnerProduct"
  bottom: "fc6"
  top: "fc7"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  param {
    lr_mult: 2
    decay_mult: 0
  }
  inner_product_param {
    num_output: 4096
    weight_filler {
      type: "gaussian"
      std: 0.005
    }
    bias_filler {
      type: "constant"
      value: 0.1
    }
  }

}
layer {
  name: "relu7"
  type: "ReLU"
  bottom: "fc7"
  top: "fc7"
}
layer {
  name: "drop7"
  type: "Dropout"
  bottom: "fc7"
  top: "fc7"
  dropout_param {
    dropout_ratio: 0.5
  }
}
layer {
  name: "my-fc8"
  type: "InnerProduct"
  bottom: "fc7"
  top: "fc8"
  inner_product_param {
    num_output: 28
    weight_filler {
        type: "xavier"
    }
    bias_filler {
        type: "constant"
        value: 0.1
    }
  }

}
layer {
    name: "accuracy"
    type: "Accuracy"
    bottom: "fc8"
    bottom: "label"
    top: "accuracy"
    include: { phase: TEST }
}
layer {
    name: "loss"
    type: "SoftmaxWithLoss"
    bottom: "fc8"
    bottom: "label"
    top: "loss"
}

Please help me.
I’m using JETSON TX2

2

Hi,

What kind of use case do you want to do?
Re-train the model or try the inference?

By the way, could you help us point out the error message?
We cannot find an obvious error inside your log.

I0125 19:34:31.965451  5184 net.cpp:746] Ignoring source layer fc8
I0125 19:34:31.965508  5184 net.cpp:746] Ignoring source layer prob
I0125 19:34:32.007084  5184 solver.cpp:57] Solver scaffolding done.
I0125 19:34:32.011937  5184 caffe.cpp:239] Starting Optimization
I0125 19:34:32.011982  5184 solver.cpp:289] Solving Face_Recognition_VGG_NET
I0125 19:34:32.011994  5184 solver.cpp:290] Learning Rate Policy: step
I0125 19:34:32.246888  5184 solver.cpp:347] Iteration 0, Testing net (#0)
nvidia@tegra-ubuntu:~$

Thanks.

3

Hello. Thanks for your reply and sorry for my late reply.
I was looking information about it.
I was trying to do fine-tuning using VGG 19 layers.
During the training process after Iteration 0, Testing net (#0) As you can see, the training process ends abruptly.
I think is because Jetson TX2 doesn’t has enough memory.

4

Hi,

YES. TX2 is designed for inference and is not suitable for training due to hardware limitation.
It’s recommended to apply the training job in a desktop environment.

Thanks.