mobileFaceNet.log

I0510 16:30:42.442382 39557 caffe.cpp:218] Using GPUs 0, 1
I0510 16:30:42.455554 39557 caffe.cpp:223] GPU 0: GeForce GTX 1080 Ti
I0510 16:30:42.456099 39557 caffe.cpp:223] GPU 1: GeForce GTX 1080 Ti
I0510 16:30:43.443825 39557 solver.cpp:44] Initializing solver from parameters: 
base_lr: 0.1
display: 100
max_iter: 160000
lr_policy: "multistep"
gamma: 0.1
momentum: 0.9
weight_decay: 0.0005
snapshot: 5000
snapshot_prefix: "face_snapshot/MobileFaceNet_16w_0pt3_model"
solver_mode: GPU
device_id: 0
net: "MobileFaceNet_1.prototxt"
train_state {
  level: 0
  stage: ""
}
stepvalue: 80000
stepvalue: 120000
stepvalue: 140000
iter_size: 1
I0510 16:30:43.448559 39557 solver.cpp:87] Creating training net from net file: MobileFaceNet_1.prototxt
I0510 16:30:43.449743 39557 upgrade_proto.cpp:77] Attempting to upgrade batch norm layers using deprecated params: MobileFaceNet_1.prototxt
I0510 16:30:43.449759 39557 upgrade_proto.cpp:80] Successfully upgraded batch norm layers using deprecated params.
I0510 16:30:43.451156 39557 net.cpp:51] Initializing net from parameters: 
name: "Caffe_MobileFaceNet"
state {
  phase: TRAIN
  level: 0
  stage: ""
}
layer {
  name: "data"
  type: "ImageData"
  top: "data"
  top: "label"
  transform_param {
    scale: 0.0078125
    mirror: true
    mean_value: 127.5
    mean_value: 127.5
    mean_value: 127.5
  }
  image_data_param {
    source: "/data/vggface2/VGGFACE_train_list-112X96-add.txt"
    batch_size: 64
    shuffle: true
  }
}
layer {
  name: "conv1"
  type: "Convolution"
  bottom: "data"
  top: "conv1"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  convolution_param {
    num_output: 64
    bias_term: false
    pad: 1
    kernel_size: 3
    stride: 2
    weight_filler {
      type: "xavier"
    }
  }
}
layer {
  name: "conv1/bn"
  type: "BatchNorm"
  bottom: "conv1"
  top: "conv1"
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
}
layer {
  name: "conv1/scale"
  type: "Scale"
  bottom: "conv1"
  top: "conv1"
  param {
    lr_mult: 1
    decay_mult: 0
  }
  param {
    lr_mult: 1
    decay_mult: 0
  }
  scale_param {
    bias_term: true
  }
}
layer {
  name: "relu1"
  type: "PReLU"
  bottom: "conv1"
  top: "conv1"
}
layer {
  name: "conv1_dw"
  type: "DepthwiseConvolution"
  bottom: "conv1"
  top: "conv1_dw"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  convolution_param {
    num_output: 64
    bias_term: false
    pad: 1
    kernel_size: 3
    group: 64
    stride: 1
    weight_filler {
      type: "xavier"
    }
  }
}
layer {
  name: "conv1_dw/bn"
  type: "BatchNorm"
  bottom: "conv1_dw"
  top: "conv1_dw"
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
}
layer {
  name: "conv1_dw/scale"
  type: "Scale"
  bottom: "conv1_dw"
  top: "conv1_dw"
  param {
    lr_mult: 1
    decay_mult: 0
  }
  param {
    lr_mult: 1
    decay_mult: 0
  }
  scale_param {
    bias_term: true
  }
}
layer {
  name: "relu1_dw"
  type: "PReLU"
  bottom: "conv1_dw"
  top: "conv1_dw"
}
layer {
  name: "conv2_ex"
  type: "Convolution"
  bottom: "conv1_dw"
  top: "conv2_ex"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  convolution_param {
    num_output: 128
    bias_term: false
    pad: 0
    kernel_size: 1
    stride: 1
    weight_filler {
      type: "xavier"
    }
  }
}
layer {
  name: "conv2_ex/bn"
  type: "BatchNorm"
  bottom: "conv2_ex"
  top: "conv2_ex"
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
}
layer {
  name: "conv2_ex/scale"
  type: "Scale"
  bottom: "conv2_ex"
  top: "conv2_ex"
  param {
    lr_mult: 1
    decay_mult: 0
  }
  param {
    lr_mult: 1
    decay_mult: 0
  }
  scale_param {
    bias_term: true
  }
}
layer {
  name: "relu2_ex"
  type: "PReLU"
  bottom: "conv2_ex"
  top: "conv2_ex"
}
layer {
  name: "conv2_dw"
  type: "DepthwiseConvolution"
  bottom: "conv2_ex"
  top: "conv2_dw"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  convolution_param {
    num_output: 128
    bias_term: false
    pad: 1
    kernel_size: 3
    group: 128
    stride: 2
    weight_filler {
      type: "xavier"
    }
    engine: CAFFE
  }
}
layer {
  name: "conv2_dw/bn"
  type: "BatchNorm"
  bottom: "conv2_dw"
  top: "conv2_dw"
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
}
layer {
  name: "conv2_dw/scale"
  type: "Scale"
  bottom: "conv2_dw"
  top: "conv2_dw"
  param {
    lr_mult: 1
    decay_mult: 0
  }
  param {
    lr_mult: 1
    decay_mult: 0
  }
  scale_param {
    bias_term: true
  }
}
layer {
  name: "relu2_dw"
  type: "PReLU"
  bottom: "conv2_dw"
  top: "conv2_dw"
}
layer {
  name: "conv2_em"
  type: "Convolution"
  bottom: "conv2_dw"
  top: "conv2_em"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  convolution_param {
    num_output: 64
    bias_term: false
    pad: 0
    kernel_size: 1
    stride: 1
    weight_filler {
      type: "xavier"
    }
  }
}
layer {
  name: "conv2_em/bn"
  type: "BatchNorm"
  bottom: "conv2_em"
  top: "conv2_em"
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
}
layer {
  name: "conv2_em/scale"
  type: "Scale"
  bottom: "conv2_em"
  top: "conv2_em"
  param {
    lr_mult: 1
    decay_mult: 0
  }
  param {
    lr_mult: 1
    decay_mult: 0
  }
  scale_param {
    bias_term: true
  }
}
layer {
  name: "conv2_1_ex"
  type: "Convolution"
  bottom: "conv2_em"
  top: "conv2_1_ex"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  convolution_param {
    num_output: 128
    bias_term: false
    pad: 0
    kernel_size: 1
    stride: 1
    weight_filler {
      type: "xavier"
    }
  }
}
layer {
  name: "conv2_1_ex/bn"
  type: "BatchNorm"
  bottom: "conv2_1_ex"
  top: "conv2_1_ex"
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
}
layer {
  name: "conv2_1_ex/scale"
  type: "Scale"
  bottom: "conv2_1_ex"
  top: "conv2_1_ex"
  param {
    lr_mult: 1
    decay_mult: 0
  }
  param {
    lr_mult: 1
    decay_mult: 0
  }
  scale_param {
    bias_term: true
  }
}
layer {
  name: "relu2_1_ex"
  type: "PReLU"
  bottom: "conv2_1_ex"
  top: "conv2_1_ex"
}
layer {
  name: "conv2_1_dw"
  type: "DepthwiseConvolution"
  bottom: "conv2_1_ex"
  top: "conv2_1_dw"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  convolution_param {
    num_output: 128
    bias_term: false
    pad: 1
    kernel_size: 3
    group: 128
    stride: 1
    weight_filler {
      type: "xavier"
    }
    engine: CAFFE
  }
}
layer {
  name: "conv2_1_dw/bn"
  type: "BatchNorm"
  bottom: "conv2_1_dw"
  top: "conv2_1_dw"
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
}
layer {
  name: "conv2_1_dw/scale"
  type: "Scale"
  bottom: "conv2_1_dw"
  top: "conv2_1_dw"
  param {
    lr_mult: 1
    decay_mult: 0
  }
  param {
    lr_mult: 1
    decay_mult: 0
  }
  scale_param {
    bias_term: true
  }
}
layer {
  name: "relu2_1_dw"
  type: "PReLU"
  bottom: "conv2_1_dw"
  top: "conv2_1_dw"
}
layer {
  name: "conv2_1_em"
  type: "Convolution"
  bottom: "conv2_1_dw"
  top: "conv2_1_em"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  convolution_param {
    num_output: 64
    bias_term: false
    pad: 0
    kernel_size: 1
    stride: 1
    weight_filler {
      type: "xavier"
    }
  }
}
layer {
  name: "conv2_1_em/bn"
  type: "BatchNorm"
  bottom: "conv2_1_em"
  top: "conv2_1_em"
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
}
layer {
  name: "conv2_1_em/scale"
  type: "Scale"
  bottom: "conv2_1_em"
  top: "conv2_1_em"
  param {
    lr_mult: 1
    decay_mult: 0
  }
  param {
    lr_mult: 1
    decay_mult: 0
  }
  scale_param {
    bias_term: true
  }
}
layer {
  name: "res2_1"
  type: "Eltwise"
  bottom: "conv2_em"
  bottom: "conv2_1_em"
  top: "res2_1"
}
layer {
  name: "conv2_2_ex"
  type: "Convolution"
  bottom: "res2_1"
  top: "conv2_2_ex"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  convolution_param {
    num_output: 128
    bias_term: false
    pad: 0
    kernel_size: 1
    stride: 1
    weight_filler {
      type: "xavier"
    }
  }
}
layer {
  name: "conv2_2_ex/bn"
  type: "BatchNorm"
  bottom: "conv2_2_ex"
  top: "conv2_2_ex"
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
}
layer {
  name: "conv2_2_ex/scale"
  type: "Scale"
  bottom: "conv2_2_ex"
  top: "conv2_2_ex"
  param {
    lr_mult: 1
    decay_mult: 0
  }
  param {
    lr_mult: 1
    decay_mult: 0
  }
  scale_param {
    bias_term: true
  }
}
layer {
  name: "relu2_2_ex"
  type: "PReLU"
  bottom: "conv2_2_ex"
  top: "conv2_2_ex"
}
layer {
  name: "conv2_2_dw"
  type: "DepthwiseConvolution"
  bottom: "conv2_2_ex"
  top: "conv2_2_dw"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  convolution_param {
    num_output: 128
    bias_term: false
    pad: 1
    kernel_size: 3
    group: 128
    stride: 1
    weight_filler {
      type: "xavier"
    }
    engine: CAFFE
  }
}
layer {
  name: "conv2_2_dw/bn"
  type: "BatchNorm"
  bottom: "conv2_2_dw"
  top: "conv2_2_dw"
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
}
layer {
  name: "conv2_2_dw/scale"
  type: "Scale"
  bottom: "conv2_2_dw"
  top: "conv2_2_dw"
  param {
    lr_mult: 1
    decay_mult: 0
  }
  param {
    lr_mult: 1
    decay_mult: 0
  }
  scale_param {
    bias_term: true
  }
}
layer {
  name: "relu2_2_dw"
  type: "PReLU"
  bottom: "conv2_2_dw"
  top: "conv2_2_dw"
}
layer {
  name: "conv2_2_em"
  type: "Convolution"
  bottom: "conv2_2_dw"
  top: "conv2_2_em"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  convolution_param {
    num_output: 64
    bias_term: false
    pad: 0
    kernel_size: 1
    stride: 1
    weight_filler {
      type: "xavier"
    }
  }
}
layer {
  name: "conv2_2_em/bn"
  type: "BatchNorm"
  bottom: "conv2_2_em"
  top: "conv2_2_em"
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
}
layer {
  name: "conv2_2_em/scale"
  type: "Scale"
  bottom: "conv2_2_em"
  top: "conv2_2_em"
  param {
    lr_mult: 1
    decay_mult: 0
  }
  param {
    lr_mult: 1
    decay_mult: 0
  }
  scale_param {
    bias_term: true
  }
}
layer {
  name: "res2_2"
  type: "Eltwise"
  bottom: "res2_1"
  bottom: "conv2_2_em"
  top: "res2_2"
}
layer {
  name: "conv2_3_ex"
  type: "Convolution"
  bottom: "res2_2"
  top: "conv2_3_ex"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  convolution_param {
    num_output: 128
    bias_term: false
    pad: 0
    kernel_size: 1
    stride: 1
    weight_filler {
      type: "xavier"
    }
  }
}
layer {
  name: "conv2_3_ex/bn"
  type: "BatchNorm"
  bottom: "conv2_3_ex"
  top: "conv2_3_ex"
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
}
layer {
  name: "conv2_3_ex/scale"
  type: "Scale"
  bottom: "conv2_3_ex"
  top: "conv2_3_ex"
  param {
    lr_mult: 1
    decay_mult: 0
  }
  param {
    lr_mult: 1
    decay_mult: 0
  }
  scale_param {
    bias_term: true
  }
}
layer {
  name: "relu2_3_ex"
  type: "PReLU"
  bottom: "conv2_3_ex"
  top: "conv2_3_ex"
}
layer {
  name: "conv2_3_dw"
  type: "DepthwiseConvolution"
  bottom: "conv2_3_ex"
  top: "conv2_3_dw"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  convolution_param {
    num_output: 128
    bias_term: false
    pad: 1
    kernel_size: 3
    group: 128
    stride: 1
    weight_filler {
      type: "xavier"
    }
    engine: CAFFE
  }
}
layer {
  name: "conv2_3_dw/bn"
  type: "BatchNorm"
  bottom: "conv2_3_dw"
  top: "conv2_3_dw"
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
}
layer {
  name: "conv2_3_dw/scale"
  type: "Scale"
  bottom: "conv2_3_dw"
  top: "conv2_3_dw"
  param {
    lr_mult: 1
    decay_mult: 0
  }
  param {
    lr_mult: 1
    decay_mult: 0
  }
  scale_param {
    bias_term: true
  }
}
layer {
  name: "relu2_3_dw"
  type: "PReLU"
  bottom: "conv2_3_dw"
  top: "conv2_3_dw"
}
layer {
  name: "conv2_3_em"
  type: "Convolution"
  bottom: "conv2_3_dw"
  top: "conv2_3_em"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  convolution_param {
    num_output: 64
    bias_term: false
    pad: 0
    kernel_size: 1
    stride: 1
    weight_filler {
      type: "xavier"
    }
  }
}
layer {
  name: "conv2_3_em/bn"
  type: "BatchNorm"
  bottom: "conv2_3_em"
  top: "conv2_3_em"
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
}
layer {
  name: "conv2_3_em/scale"
  type: "Scale"
  bottom: "conv2_3_em"
  top: "conv2_3_em"
  param {
    lr_mult: 1
    decay_mult: 0
  }
  param {
    lr_mult: 1
    decay_mult: 0
  }
  scale_param {
    bias_term: true
  }
}
layer {
  name: "res2_3"
  type: "Eltwise"
  bottom: "res2_2"
  bottom: "conv2_3_em"
  top: "res2_3"
}
layer {
  name: "conv2_4_ex"
  type: "Convolution"
  bottom: "res2_3"
  top: "conv2_4_ex"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  convolution_param {
    num_output: 128
    bias_term: false
    pad: 0
    kernel_size: 1
    stride: 1
    weight_filler {
      type: "xavier"
    }
  }
}
layer {
  name: "conv2_4_ex/bn"
  type: "BatchNorm"
  bottom: "conv2_4_ex"
  top: "conv2_4_ex"
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
}
layer {
  name: "conv2_4_ex/scale"
  type: "Scale"
  bottom: "conv2_4_ex"
  top: "conv2_4_ex"
  param {
    lr_mult: 1
    decay_mult: 0
  }
  param {
    lr_mult: 1
    decay_mult: 0
  }
  scale_param {
    bias_term: true
  }
}
layer {
  name: "relu2_4_ex"
  type: "PReLU"
  bottom: "conv2_4_ex"
  top: "conv2_4_ex"
}
layer {
  name: "conv2_4_dw"
  type: "DepthwiseConvolution"
  bottom: "conv2_4_ex"
  top: "conv2_4_dw"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  convolution_param {
    num_output: 128
    bias_term: false
    pad: 1
    kernel_size: 3
    group: 128
    stride: 1
    weight_filler {
      type: "xavier"
    }
    engine: CAFFE
  }
}
layer {
  name: "conv2_4_dw/bn"
  type: "BatchNorm"
  bottom: "conv2_4_dw"
  top: "conv2_4_dw"
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
}
layer {
  name: "conv2_4_dw/scale"
  type: "Scale"
  bottom: "conv2_4_dw"
  top: "conv2_4_dw"
  param {
    lr_mult: 1
    decay_mult: 0
  }
  param {
    lr_mult: 1
    decay_mult: 0
  }
  scale_param {
    bias_term: true
  }
}
layer {
  name: "relu2_4_dw"
  type: "PReLU"
  bottom: "conv2_4_dw"
  top: "conv2_4_dw"
}
layer {
  name: "conv2_4_em"
  type: "Convolution"
  bottom: "conv2_4_dw"
  top: "conv2_4_em"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  convolution_param {
    num_output: 64
    bias_term: false
    pad: 0
    kernel_size: 1
    stride: 1
    weight_filler {
      type: "xavier"
    }
  }
}
layer {
  name: "conv2_4_em/bn"
  type: "BatchNorm"
  bottom: "conv2_4_em"
  top: "conv2_4_em"
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
}
layer {
  name: "conv2_4_em/scale"
  type: "Scale"
  bottom: "conv2_4_em"
  top: "conv2_4_em"
  param {
    lr_mult: 1
    decay_mult: 0
  }
  param {
    lr_mult: 1
    decay_mult: 0
  }
  scale_param {
    bias_term: true
  }
}
layer {
  name: "res2_4"
  type: "Eltwise"
  bottom: "res2_3"
  bottom: "conv2_4_em"
  top: "res2_4"
}
layer {
  name: "conv3_ex"
  type: "Convolution"
  bottom: "res2_4"
  top: "conv3_ex"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  convolution_param {
    num_output: 256
    bias_term: false
    pad: 0
    kernel_size: 1
    stride: 1
    weight_filler {
      type: "xavier"
    }
  }
}
layer {
  name: "conv3_ex/bn"
  type: "BatchNorm"
  bottom: "conv3_ex"
  top: "conv3_ex"
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
}
layer {
  name: "conv3_ex/scale"
  type: "Scale"
  bottom: "conv3_ex"
  top: "conv3_ex"
  param {
    lr_mult: 1
    decay_mult: 0
  }
  param {
    lr_mult: 1
    decay_mult: 0
  }
  scale_param {
    bias_term: true
  }
}
layer {
  name: "relu3_ex"
  type: "PReLU"
  bottom: "conv3_ex"
  top: "conv3_ex"
}
layer {
  name: "conv3_dw"
  type: "DepthwiseConvolution"
  bottom: "conv3_ex"
  top: "conv3_dw"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  convolution_param {
    num_output: 256
    bias_term: false
    pad: 1
    kernel_size: 3
    group: 256
    stride: 2
    weight_filler {
      type: "xavier"
    }
    engine: CAFFE
  }
}
layer {
  name: "conv3_dw/bn"
  type: "BatchNorm"
  bottom: "conv3_dw"
  top: "conv3_dw"
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
}
layer {
  name: "conv3_dw/scale"
  type: "Scale"
  bottom: "conv3_dw"
  top: "conv3_dw"
  param {
    lr_mult: 1
    decay_mult: 0
  }
  param {
    lr_mult: 1
    decay_mult: 0
  }
  scale_param {
    bias_term: true
  }
}
layer {
  name: "relu3_dw"
  type: "PReLU"
  bottom: "conv3_dw"
  top: "conv3_dw"
}
layer {
  name: "conv3_em"
  type: "Convolution"
  bottom: "conv3_dw"
  top: "conv3_em"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  convolution_param {
    num_output: 128
    bias_term: false
    pad: 0
    kernel_size: 1
    stride: 1
    weight_filler {
      type: "xavier"
    }
  }
}
layer {
  name: "conv3_em/bn"
  type: "BatchNorm"
  bottom: "conv3_em"
  top: "conv3_em"
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
}
layer {
  name: "conv3_em/scale"
  type: "Scale"
  bottom: "conv3_em"
  top: "conv3_em"
  param {
    lr_mult: 1
    decay_mult: 0
  }
  param {
    lr_mult: 1
    decay_mult: 0
  }
  scale_param {
    bias_term: true
  }
}
layer {
  name: "conv3_1_ex"
  type: "Convolution"
  bottom: "conv3_em"
  top: "conv3_1_ex"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  convolution_param {
    num_output: 256
    bias_term: false
    pad: 0
    kernel_size: 1
    stride: 1
    weight_filler {
      type: "xavier"
    }
  }
}
layer {
  name: "conv3_1_ex/bn"
  type: "BatchNorm"
  bottom: "conv3_1_ex"
  top: "conv3_1_ex"
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
}
layer {
  name: "conv3_1_ex/scale"
  type: "Scale"
  bottom: "conv3_1_ex"
  top: "conv3_1_ex"
  param {
    lr_mult: 1
    decay_mult: 0
  }
  param {
    lr_mult: 1
    decay_mult: 0
  }
  scale_param {
    bias_term: true
  }
}
layer {
  name: "relu3_1_ex"
  type: "PReLU"
  bottom: "conv3_1_ex"
  top: "conv3_1_ex"
}
layer {
  name: "conv3_1_dw"
  type: "DepthwiseConvolution"
  bottom: "conv3_1_ex"
  top: "conv3_1_dw"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  convolution_param {
    num_output: 256
    bias_term: false
    pad: 1
    kernel_size: 3
    group: 256
    stride: 1
    weight_filler {
      type: "xavier"
    }
    engine: CAFFE
  }
}
layer {
  name: "conv3_1_dw/bn"
  type: "BatchNorm"
  bottom: "conv3_1_dw"
  top: "conv3_1_dw"
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
}
layer {
  name: "conv3_1_dw/scale"
  type: "Scale"
  bottom: "conv3_1_dw"
  top: "conv3_1_dw"
  param {
    lr_mult: 1
    decay_mult: 0
  }
  param {
    lr_mult: 1
    decay_mult: 0
  }
  scale_param {
    bias_term: true
  }
}
layer {
  name: "relu3_1_dw"
  type: "PReLU"
  bottom: "conv3_1_dw"
  top: "conv3_1_dw"
}
layer {
  name: "conv3_1_em"
  type: "Convolution"
  bottom: "conv3_1_dw"
  top: "conv3_1_em"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  convolution_param {
    num_output: 128
    bias_term: false
    pad: 0
    kernel_size: 1
    stride: 1
    weight_filler {
      type: "xavier"
    }
  }
}
layer {
  name: "conv3_1_em/bn"
  type: "BatchNorm"
  bottom: "conv3_1_em"
  top: "conv3_1_em"
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
}
layer {
  name: "conv3_1_em/scale"
  type: "Scale"
  bottom: "conv3_1_em"
  top: "conv3_1_em"
  param {
    lr_mult: 1
    decay_mult: 0
  }
  param {
    lr_mult: 1
    decay_mult: 0
  }
  scale_param {
    bias_term: true
  }
}
layer {
  name: "res3_1"
  type: "Eltwise"
  bottom: "conv3_em"
  bottom: "conv3_1_em"
  top: "res3_1"
}
layer {
  name: "conv3_2_ex"
  type: "Convolution"
  bottom: "res3_1"
  top: "conv3_2_ex"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  convolution_param {
    num_output: 256
    bias_term: false
    pad: 0
    kernel_size: 1
    stride: 1
    weight_filler {
      type: "xavier"
    }
  }
}
layer {
  name: "conv3_2_ex/bn"
  type: "BatchNorm"
  bottom: "conv3_2_ex"
  top: "conv3_2_ex"
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
}
layer {
  name: "conv3_2_ex/scale"
  type: "Scale"
  bottom: "conv3_2_ex"
  top: "conv3_2_ex"
  param {
    lr_mult: 1
    decay_mult: 0
  }
  param {
    lr_mult: 1
    decay_mult: 0
  }
  scale_param {
    bias_term: true
  }
}
layer {
  name: "relu3_2_ex"
  type: "PReLU"
  bottom: "conv3_2_ex"
  top: "conv3_2_ex"
}
layer {
  name: "conv3_2_dw"
  type: "DepthwiseConvolution"
  bottom: "conv3_2_ex"
  top: "conv3_2_dw"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  convolution_param {
    num_output: 256
    bias_term: false
    pad: 1
    kernel_size: 3
    group: 256
    stride: 1
    weight_filler {
      type: "xavier"
    }
    engine: CAFFE
  }
}
layer {
  name: "conv3_2_dw/bn"
  type: "BatchNorm"
  bottom: "conv3_2_dw"
  top: "conv3_2_dw"
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
}
layer {
  name: "conv3_2_dw/scale"
  type: "Scale"
  bottom: "conv3_2_dw"
  top: "conv3_2_dw"
  param {
    lr_mult: 1
    decay_mult: 0
  }
  param {
    lr_mult: 1
    decay_mult: 0
  }
  scale_param {
    bias_term: true
  }
}
layer {
  name: "relu3_2_dw"
  type: "PReLU"
  bottom: "conv3_2_dw"
  top: "conv3_2_dw"
}
layer {
  name: "conv3_2_em"
  type: "Convolution"
  bottom: "conv3_2_dw"
  top: "conv3_2_em"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  convolution_param {
    num_output: 128
    bias_term: false
    pad: 0
    kernel_size: 1
    stride: 1
    weight_filler {
      type: "xavier"
    }
  }
}
layer {
  name: "conv3_2_em/bn"
  type: "BatchNorm"
  bottom: "conv3_2_em"
  top: "conv3_2_em"
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
}
layer {
  name: "conv3_2_em/scale"
  type: "Scale"
  bottom: "conv3_2_em"
  top: "conv3_2_em"
  param {
    lr_mult: 1
    decay_mult: 0
  }
  param {
    lr_mult: 1
    decay_mult: 0
  }
  scale_param {
    bias_term: true
  }
}
layer {
  name: "res3_2"
  type: "Eltwise"
  bottom: "res3_1"
  bottom: "conv3_2_em"
  top: "res3_2"
}
layer {
  name: "conv3_3_ex"
  type: "Convolution"
  bottom: "res3_2"
  top: "conv3_3_ex"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  convolution_param {
    num_output: 256
    bias_term: false
    pad: 0
    kernel_size: 1
    stride: 1
    weight_filler {
      type: "xavier"
    }
  }
}
layer {
  name: "conv3_3_ex/bn"
  type: "BatchNorm"
  bottom: "conv3_3_ex"
  top: "conv3_3_ex"
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
}
layer {
  name: "conv3_3_ex/scale"
  type: "Scale"
  bottom: "conv3_3_ex"
  top: "conv3_3_ex"
  param {
    lr_mult: 1
    decay_mult: 0
  }
  param {
    lr_mult: 1
    decay_mult: 0
  }
  scale_param {
    bias_term: true
  }
}
layer {
  name: "relu3_3_ex"
  type: "PReLU"
  bottom: "conv3_3_ex"
  top: "conv3_3_ex"
}
layer {
  name: "conv3_3_dw"
  type: "DepthwiseConvolution"
  bottom: "conv3_3_ex"
  top: "conv3_3_dw"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  convolution_param {
    num_output: 256
    bias_term: false
    pad: 1
    kernel_size: 3
    group: 256
    stride: 1
    weight_filler {
      type: "xavier"
    }
    engine: CAFFE
  }
}
layer {
  name: "conv3_3_dw/bn"
  type: "BatchNorm"
  bottom: "conv3_3_dw"
  top: "conv3_3_dw"
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
}
layer {
  name: "conv3_3_dw/scale"
  type: "Scale"
  bottom: "conv3_3_dw"
  top: "conv3_3_dw"
  param {
    lr_mult: 1
    decay_mult: 0
  }
  param {
    lr_mult: 1
    decay_mult: 0
  }
  scale_param {
    bias_term: true
  }
}
layer {
  name: "relu3_3_dw"
  type: "PReLU"
  bottom: "conv3_3_dw"
  top: "conv3_3_dw"
}
layer {
  name: "conv3_3_em"
  type: "Convolution"
  bottom: "conv3_3_dw"
  top: "conv3_3_em"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  convolution_param {
    num_output: 128
    bias_term: false
    pad: 0
    kernel_size: 1
    stride: 1
    weight_filler {
      type: "xavier"
    }
  }
}
layer {
  name: "conv3_3_em/bn"
  type: "BatchNorm"
  bottom: "conv3_3_em"
  top: "conv3_3_em"
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
}
layer {
  name: "conv3_3_em/scale"
  type: "Scale"
  bottom: "conv3_3_em"
  top: "conv3_3_em"
  param {
    lr_mult: 1
    decay_mult: 0
  }
  param {
    lr_mult: 1
    decay_mult: 0
  }
  scale_param {
    bias_term: true
  }
}
layer {
  name: "res3_3"
  type: "Eltwise"
  bottom: "res3_2"
  bottom: "conv3_3_em"
  top: "res3_3"
}
layer {
  name: "conv3_4_ex"
  type: "Convolution"
  bottom: "res3_3"
  top: "conv3_4_ex"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  convolution_param {
    num_output: 256
    bias_term: false
    pad: 0
    kernel_size: 1
    stride: 1
    weight_filler {
      type: "xavier"
    }
  }
}
layer {
  name: "conv3_4_ex/bn"
  type: "BatchNorm"
  bottom: "conv3_4_ex"
  top: "conv3_4_ex"
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
}
layer {
  name: "conv3_4_ex/scale"
  type: "Scale"
  bottom: "conv3_4_ex"
  top: "conv3_4_ex"
  param {
    lr_mult: 1
    decay_mult: 0
  }
  param {
    lr_mult: 1
    decay_mult: 0
  }
  scale_param {
    bias_term: true
  }
}
layer {
  name: "relu3_4_ex"
  type: "PReLU"
  bottom: "conv3_4_ex"
  top: "conv3_4_ex"
}
layer {
  name: "conv3_4_dw"
  type: "DepthwiseConvolution"
  bottom: "conv3_4_ex"
  top: "conv3_4_dw"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  convolution_param {
    num_output: 256
    bias_term: false
    pad: 1
    kernel_size: 3
    group: 256
    stride: 1
    weight_filler {
      type: "xavier"
    }
    engine: CAFFE
  }
}
layer {
  name: "conv3_4_dw/bn"
  type: "BatchNorm"
  bottom: "conv3_4_dw"
  top: "conv3_4_dw"
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
}
layer {
  name: "conv3_4_dw/scale"
  type: "Scale"
  bottom: "conv3_4_dw"
  top: "conv3_4_dw"
  param {
    lr_mult: 1
    decay_mult: 0
  }
  param {
    lr_mult: 1
    decay_mult: 0
  }
  scale_param {
    bias_term: true
  }
}
layer {
  name: "relu3_4_dw"
  type: "PReLU"
  bottom: "conv3_4_dw"
  top: "conv3_4_dw"
}
layer {
  name: "conv3_4_em"
  type: "Convolution"
  bottom: "conv3_4_dw"
  top: "conv3_4_em"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  convolution_param {
    num_output: 128
    bias_term: false
    pad: 0
    kernel_size: 1
    stride: 1
    weight_filler {
      type: "xavier"
    }
  }
}
layer {
  name: "conv3_4_em/bn"
  type: "BatchNorm"
  bottom: "conv3_4_em"
  top: "conv3_4_em"
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
}
layer {
  name: "conv3_4_em/scale"
  type: "Scale"
  bottom: "conv3_4_em"
  top: "conv3_4_em"
  param {
    lr_mult: 1
    decay_mult: 0
  }
  param {
    lr_mult: 1
    decay_mult: 0
  }
  scale_param {
    bias_term: true
  }
}
layer {
  name: "res3_4"
  type: "Eltwise"
  bottom: "res3_3"
  bottom: "conv3_4_em"
  top: "res3_4"
}
layer {
  name: "conv3_5_ex"
  type: "Convolution"
  bottom: "res3_4"
  top: "conv3_5_ex"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  convolution_param {
    num_output: 256
    bias_term: false
    pad: 0
    kernel_size: 1
    stride: 1
    weight_filler {
      type: "xavier"
    }
  }
}
layer {
  name: "conv3_5_ex/bn"
  type: "BatchNorm"
  bottom: "conv3_5_ex"
  top: "conv3_5_ex"
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
}
layer {
  name: "conv3_5_ex/scale"
  type: "Scale"
  bottom: "conv3_5_ex"
  top: "conv3_5_ex"
  param {
    lr_mult: 1
    decay_mult: 0
  }
  param {
    lr_mult: 1
    decay_mult: 0
  }
  scale_param {
    bias_term: true
  }
}
layer {
  name: "relu3_5_ex"
  type: "PReLU"
  bottom: "conv3_5_ex"
  top: "conv3_5_ex"
}
layer {
  name: "conv3_5_dw"
  type: "DepthwiseConvolution"
  bottom: "conv3_5_ex"
  top: "conv3_5_dw"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  convolution_param {
    num_output: 256
    bias_term: false
    pad: 1
    kernel_size: 3
    group: 256
    stride: 1
    weight_filler {
      type: "xavier"
    }
    engine: CAFFE
  }
}
layer {
  name: "conv3_5_dw/bn"
  type: "BatchNorm"
  bottom: "conv3_5_dw"
  top: "conv3_5_dw"
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
  param {
    lr_mult: 0
    decay_mult: 0
  }
}
layer {
  name: "conv3_5_dw/scale"
  type: "Scale"
  bottom: "conv3_5_dw"
  top: "conv3_5_dw"
  param {
    lr_mult: 1
    decay_mult: 0
  }
  param {
    lr_mult: 1
    decay_mult: 0
  }
  scale_param {
    bias_term: true
  }
}
layer {
  name: "relu3_5_dw"
  type: "PReLU"
  bottom: "conv3_5_dw"
  top: "conv3_5_dw"
}
layer {
  name: "conv3_5_em"
  type: "Convolution"
  bottom: "conv3_5_dw"
  top: "conv3_5_em"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  convolution_param {
I0510 16:30:43.452024 39557 layer_factory.hpp:77] Creating layer data
I0510 16:30:43.452060 39557 net.cpp:84] Creating Layer data
I0510 16:30:43.452069 39557 net.cpp:380] data -> data
I0510 16:30:43.452102 39557 net.cpp:380] data -> label
I0510 16:30:43.452167 39557 image_data_layer.cpp:38] Opening file /data/vggface2/VGGFACE_train_list-112X96-add.txt
I0510 16:30:44.392499 39557 image_data_layer.cpp:53] Shuffling data
I0510 16:30:44.540964 39557 image_data_layer.cpp:63] A total of 3306982 images.
I0510 16:30:44.543790 39557 image_data_layer.cpp:90] output data size: 64,3,112,96
I0510 16:30:44.564759 39557 net.cpp:122] Setting up data
I0510 16:30:44.564785 39557 net.cpp:129] Top shape: 64 3 112 96 (2064384)
I0510 16:30:44.564790 39557 net.cpp:129] Top shape: 64 (64)
I0510 16:30:44.564793 39557 net.cpp:137] Memory required for data: 8257792
I0510 16:30:44.564822 39557 layer_factory.hpp:77] Creating layer label_data_1_split
I0510 16:30:44.564913 39557 net.cpp:84] Creating Layer label_data_1_split
I0510 16:30:44.564920 39557 net.cpp:406] label_data_1_split <- label
I0510 16:30:44.564946 39557 net.cpp:380] label_data_1_split -> label_data_1_split_0
I0510 16:30:44.564960 39557 net.cpp:380] label_data_1_split -> label_data_1_split_1
I0510 16:30:44.565265 39557 net.cpp:122] Setting up label_data_1_split
I0510 16:30:44.565274 39557 net.cpp:129] Top shape: 64 (64)
I0510 16:30:44.565289 39557 net.cpp:129] Top shape: 64 (64)
I0510 16:30:44.565291 39557 net.cpp:137] Memory required for data: 8258304
I0510 16:30:44.565296 39557 layer_factory.hpp:77] Creating layer conv1
I0510 16:30:44.565323 39557 net.cpp:84] Creating Layer conv1
I0510 16:30:44.565328 39557 net.cpp:406] conv1 <- data
I0510 16:30:44.565335 39557 net.cpp:380] conv1 -> conv1
I0510 16:30:45.123051 39557 net.cpp:122] Setting up conv1
I0510 16:30:45.123095 39557 net.cpp:129] Top shape: 64 64 56 48 (11010048)
I0510 16:30:45.123142 39557 net.cpp:137] Memory required for data: 52298496
I0510 16:30:45.123170 39557 layer_factory.hpp:77] Creating layer conv1/bn
I0510 16:30:45.123185 39557 net.cpp:84] Creating Layer conv1/bn
I0510 16:30:45.123190 39557 net.cpp:406] conv1/bn <- conv1
I0510 16:30:45.123198 39557 net.cpp:367] conv1/bn -> conv1 (in-place)
I0510 16:30:45.123412 39557 net.cpp:122] Setting up conv1/bn
I0510 16:30:45.123422 39557 net.cpp:129] Top shape: 64 64 56 48 (11010048)
I0510 16:30:45.123426 39557 net.cpp:137] Memory required for data: 96338688
I0510 16:30:45.123437 39557 layer_factory.hpp:77] Creating layer conv1/scale
I0510 16:30:45.123448 39557 net.cpp:84] Creating Layer conv1/scale
I0510 16:30:45.123456 39557 net.cpp:406] conv1/scale <- conv1
I0510 16:30:45.123461 39557 net.cpp:367] conv1/scale -> conv1 (in-place)
I0510 16:30:45.123517 39557 layer_factory.hpp:77] Creating layer conv1/scale
I0510 16:30:45.123633 39557 net.cpp:122] Setting up conv1/scale
I0510 16:30:45.123642 39557 net.cpp:129] Top shape: 64 64 56 48 (11010048)
I0510 16:30:45.123646 39557 net.cpp:137] Memory required for data: 140378880
I0510 16:30:45.123652 39557 layer_factory.hpp:77] Creating layer relu1
I0510 16:30:45.123664 39557 net.cpp:84] Creating Layer relu1
I0510 16:30:45.123669 39557 net.cpp:406] relu1 <- conv1
I0510 16:30:45.123674 39557 net.cpp:367] relu1 -> conv1 (in-place)
I0510 16:30:45.124999 39557 net.cpp:122] Setting up relu1
I0510 16:30:45.125012 39557 net.cpp:129] Top shape: 64 64 56 48 (11010048)
I0510 16:30:45.125016 39557 net.cpp:137] Memory required for data: 184419072
I0510 16:30:45.125032 39557 layer_factory.hpp:77] Creating layer conv1_dw
I0510 16:30:45.125043 39557 net.cpp:84] Creating Layer conv1_dw
I0510 16:30:45.125048 39557 net.cpp:406] conv1_dw <- conv1
I0510 16:30:45.125054 39557 net.cpp:380] conv1_dw -> conv1_dw
I0510 16:30:45.126323 39557 net.cpp:122] Setting up conv1_dw
I0510 16:30:45.126338 39557 net.cpp:129] Top shape: 64 64 56 48 (11010048)
I0510 16:30:45.126341 39557 net.cpp:137] Memory required for data: 228459264
I0510 16:30:45.126358 39557 layer_factory.hpp:77] Creating layer conv1_dw/bn
I0510 16:30:45.126368 39557 net.cpp:84] Creating Layer conv1_dw/bn
I0510 16:30:45.126372 39557 net.cpp:406] conv1_dw/bn <- conv1_dw
I0510 16:30:45.126377 39557 net.cpp:367] conv1_dw/bn -> conv1_dw (in-place)
I0510 16:30:45.126581 39557 net.cpp:122] Setting up conv1_dw/bn
I0510 16:30:45.126590 39557 net.cpp:129] Top shape: 64 64 56 48 (11010048)
I0510 16:30:45.126592 39557 net.cpp:137] Memory required for data: 272499456
I0510 16:30:45.126605 39557 layer_factory.hpp:77] Creating layer conv1_dw/scale
I0510 16:30:45.126613 39557 net.cpp:84] Creating Layer conv1_dw/scale
I0510 16:30:45.126617 39557 net.cpp:406] conv1_dw/scale <- conv1_dw
I0510 16:30:45.126622 39557 net.cpp:367] conv1_dw/scale -> conv1_dw (in-place)
I0510 16:30:45.126662 39557 layer_factory.hpp:77] Creating layer conv1_dw/scale
I0510 16:30:45.126766 39557 net.cpp:122] Setting up conv1_dw/scale
I0510 16:30:45.126775 39557 net.cpp:129] Top shape: 64 64 56 48 (11010048)
I0510 16:30:45.126778 39557 net.cpp:137] Memory required for data: 316539648
I0510 16:30:45.126783 39557 layer_factory.hpp:77] Creating layer relu1_dw
I0510 16:30:45.126790 39557 net.cpp:84] Creating Layer relu1_dw
I0510 16:30:45.126792 39557 net.cpp:406] relu1_dw <- conv1_dw
I0510 16:30:45.126796 39557 net.cpp:367] relu1_dw -> conv1_dw (in-place)
I0510 16:30:45.126977 39557 net.cpp:122] Setting up relu1_dw
I0510 16:30:45.126986 39557 net.cpp:129] Top shape: 64 64 56 48 (11010048)
I0510 16:30:45.126988 39557 net.cpp:137] Memory required for data: 360579840
I0510 16:30:45.127005 39557 layer_factory.hpp:77] Creating layer conv2_ex
I0510 16:30:45.127025 39557 net.cpp:84] Creating Layer conv2_ex
I0510 16:30:45.127029 39557 net.cpp:406] conv2_ex <- conv1_dw
I0510 16:30:45.127035 39557 net.cpp:380] conv2_ex -> conv2_ex
I0510 16:30:45.129163 39557 net.cpp:122] Setting up conv2_ex
I0510 16:30:45.129178 39557 net.cpp:129] Top shape: 64 128 56 48 (22020096)
I0510 16:30:45.129191 39557 net.cpp:137] Memory required for data: 448660224
I0510 16:30:45.129199 39557 layer_factory.hpp:77] Creating layer conv2_ex/bn
I0510 16:30:45.129207 39557 net.cpp:84] Creating Layer conv2_ex/bn
I0510 16:30:45.129222 39557 net.cpp:406] conv2_ex/bn <- conv2_ex
I0510 16:30:45.129230 39557 net.cpp:367] conv2_ex/bn -> conv2_ex (in-place)
I0510 16:30:45.129423 39557 net.cpp:122] Setting up conv2_ex/bn
I0510 16:30:45.129432 39557 net.cpp:129] Top shape: 64 128 56 48 (22020096)
I0510 16:30:45.129436 39557 net.cpp:137] Memory required for data: 536740608
I0510 16:30:45.129447 39557 layer_factory.hpp:77] Creating layer conv2_ex/scale
I0510 16:30:45.129456 39557 net.cpp:84] Creating Layer conv2_ex/scale
I0510 16:30:45.129458 39557 net.cpp:406] conv2_ex/scale <- conv2_ex
I0510 16:30:45.129463 39557 net.cpp:367] conv2_ex/scale -> conv2_ex (in-place)
I0510 16:30:45.129506 39557 layer_factory.hpp:77] Creating layer conv2_ex/scale
I0510 16:30:45.129618 39557 net.cpp:122] Setting up conv2_ex/scale
I0510 16:30:45.129626 39557 net.cpp:129] Top shape: 64 128 56 48 (22020096)
I0510 16:30:45.129628 39557 net.cpp:137] Memory required for data: 624820992
I0510 16:30:45.129634 39557 layer_factory.hpp:77] Creating layer relu2_ex
I0510 16:30:45.129652 39557 net.cpp:84] Creating Layer relu2_ex
I0510 16:30:45.129657 39557 net.cpp:406] relu2_ex <- conv2_ex
I0510 16:30:45.129662 39557 net.cpp:367] relu2_ex -> conv2_ex (in-place)
I0510 16:30:45.131059 39557 net.cpp:122] Setting up relu2_ex
I0510 16:30:45.131074 39557 net.cpp:129] Top shape: 64 128 56 48 (22020096)
I0510 16:30:45.131078 39557 net.cpp:137] Memory required for data: 712901376
I0510 16:30:45.131098 39557 layer_factory.hpp:77] Creating layer conv2_dw
I0510 16:30:45.131106 39557 net.cpp:84] Creating Layer conv2_dw
I0510 16:30:45.131111 39557 net.cpp:406] conv2_dw <- conv2_ex
I0510 16:30:45.131119 39557 net.cpp:380] conv2_dw -> conv2_dw
I0510 16:30:45.131312 39557 net.cpp:122] Setting up conv2_dw
I0510 16:30:45.131322 39557 net.cpp:129] Top shape: 64 128 28 24 (5505024)
I0510 16:30:45.131325 39557 net.cpp:137] Memory required for data: 734921472
I0510 16:30:45.131348 39557 layer_factory.hpp:77] Creating layer conv2_dw/bn
I0510 16:30:45.131357 39557 net.cpp:84] Creating Layer conv2_dw/bn
I0510 16:30:45.131361 39557 net.cpp:406] conv2_dw/bn <- conv2_dw
I0510 16:30:45.131366 39557 net.cpp:367] conv2_dw/bn -> conv2_dw (in-place)
I0510 16:30:45.131552 39557 net.cpp:122] Setting up conv2_dw/bn
I0510 16:30:45.131561 39557 net.cpp:129] Top shape: 64 128 28 24 (5505024)
I0510 16:30:45.131563 39557 net.cpp:137] Memory required for data: 756941568
I0510 16:30:45.131572 39557 layer_factory.hpp:77] Creating layer conv2_dw/scale
I0510 16:30:45.131582 39557 net.cpp:84] Creating Layer conv2_dw/scale
I0510 16:30:45.131587 39557 net.cpp:406] conv2_dw/scale <- conv2_dw
I0510 16:30:45.131592 39557 net.cpp:367] conv2_dw/scale -> conv2_dw (in-place)
I0510 16:30:45.131635 39557 layer_factory.hpp:77] Creating layer conv2_dw/scale
I0510 16:30:45.131759 39557 net.cpp:122] Setting up conv2_dw/scale
I0510 16:30:45.131767 39557 net.cpp:129] Top shape: 64 128 28 24 (5505024)
I0510 16:30:45.131769 39557 net.cpp:137] Memory required for data: 778961664
I0510 16:30:45.131775 39557 layer_factory.hpp:77] Creating layer relu2_dw
I0510 16:30:45.131784 39557 net.cpp:84] Creating Layer relu2_dw
I0510 16:30:45.131789 39557 net.cpp:406] relu2_dw <- conv2_dw
I0510 16:30:45.131794 39557 net.cpp:367] relu2_dw -> conv2_dw (in-place)
I0510 16:30:45.131920 39557 net.cpp:122] Setting up relu2_dw
I0510 16:30:45.131928 39557 net.cpp:129] Top shape: 64 128 28 24 (5505024)
I0510 16:30:45.131932 39557 net.cpp:137] Memory required for data: 800981760
I0510 16:30:45.131937 39557 layer_factory.hpp:77] Creating layer conv2_em
I0510 16:30:45.131947 39557 net.cpp:84] Creating Layer conv2_em
I0510 16:30:45.131950 39557 net.cpp:406] conv2_em <- conv2_dw
I0510 16:30:45.131956 39557 net.cpp:380] conv2_em -> conv2_em
I0510 16:30:45.133080 39557 net.cpp:122] Setting up conv2_em
I0510 16:30:45.133095 39557 net.cpp:129] Top shape: 64 64 28 24 (2752512)
I0510 16:30:45.133118 39557 net.cpp:137] Memory required for data: 811991808
I0510 16:30:45.133134 39557 layer_factory.hpp:77] Creating layer conv2_em/bn
I0510 16:30:45.133143 39557 net.cpp:84] Creating Layer conv2_em/bn
I0510 16:30:45.133160 39557 net.cpp:406] conv2_em/bn <- conv2_em
I0510 16:30:45.133168 39557 net.cpp:367] conv2_em/bn -> conv2_em (in-place)
I0510 16:30:45.133375 39557 net.cpp:122] Setting up conv2_em/bn
I0510 16:30:45.133384 39557 net.cpp:129] Top shape: 64 64 28 24 (2752512)
I0510 16:30:45.133388 39557 net.cpp:137] Memory required for data: 823001856
I0510 16:30:45.133399 39557 layer_factory.hpp:77] Creating layer conv2_em/scale
I0510 16:30:45.133409 39557 net.cpp:84] Creating Layer conv2_em/scale
I0510 16:30:45.133412 39557 net.cpp:406] conv2_em/scale <- conv2_em
I0510 16:30:45.133416 39557 net.cpp:367] conv2_em/scale -> conv2_em (in-place)
I0510 16:30:45.133464 39557 layer_factory.hpp:77] Creating layer conv2_em/scale
I0510 16:30:45.133584 39557 net.cpp:122] Setting up conv2_em/scale
I0510 16:30:45.133594 39557 net.cpp:129] Top shape: 64 64 28 24 (2752512)
I0510 16:30:45.133597 39557 net.cpp:137] Memory required for data: 834011904
I0510 16:30:45.133607 39557 layer_factory.hpp:77] Creating layer conv2_em_conv2_em/scale_0_split
I0510 16:30:45.133615 39557 net.cpp:84] Creating Layer conv2_em_conv2_em/scale_0_split
I0510 16:30:45.133618 39557 net.cpp:406] conv2_em_conv2_em/scale_0_split <- conv2_em
I0510 16:30:45.133623 39557 net.cpp:380] conv2_em_conv2_em/scale_0_split -> conv2_em_conv2_em/scale_0_split_0
I0510 16:30:45.133631 39557 net.cpp:380] conv2_em_conv2_em/scale_0_split -> conv2_em_conv2_em/scale_0_split_1
I0510 16:30:45.133669 39557 net.cpp:122] Setting up conv2_em_conv2_em/scale_0_split
I0510 16:30:45.133677 39557 net.cpp:129] Top shape: 64 64 28 24 (2752512)
I0510 16:30:45.133682 39557 net.cpp:129] Top shape: 64 64 28 24 (2752512)
I0510 16:30:45.133684 39557 net.cpp:137] Memory required for data: 856032000
I0510 16:30:45.133687 39557 layer_factory.hpp:77] Creating layer conv2_1_ex
I0510 16:30:45.133697 39557 net.cpp:84] Creating Layer conv2_1_ex
I0510 16:30:45.133702 39557 net.cpp:406] conv2_1_ex <- conv2_em_conv2_em/scale_0_split_0
I0510 16:30:45.133708 39557 net.cpp:380] conv2_1_ex -> conv2_1_ex
I0510 16:30:45.134884 39557 net.cpp:122] Setting up conv2_1_ex
I0510 16:30:45.134899 39557 net.cpp:129] Top shape: 64 128 28 24 (5505024)
I0510 16:30:45.134903 39557 net.cpp:137] Memory required for data: 878052096
I0510 16:30:45.134910 39557 layer_factory.hpp:77] Creating layer conv2_1_ex/bn
I0510 16:30:45.134932 39557 net.cpp:84] Creating Layer conv2_1_ex/bn
I0510 16:30:45.134949 39557 net.cpp:406] conv2_1_ex/bn <- conv2_1_ex
I0510 16:30:45.134958 39557 net.cpp:367] conv2_1_ex/bn -> conv2_1_ex (in-place)
I0510 16:30:45.135148 39557 net.cpp:122] Setting up conv2_1_ex/bn
I0510 16:30:45.135156 39557 net.cpp:129] Top shape: 64 128 28 24 (5505024)
I0510 16:30:45.135159 39557 net.cpp:137] Memory required for data: 900072192
I0510 16:30:45.135166 39557 layer_factory.hpp:77] Creating layer conv2_1_ex/scale
I0510 16:30:45.135174 39557 net.cpp:84] Creating Layer conv2_1_ex/scale
I0510 16:30:45.135176 39557 net.cpp:406] conv2_1_ex/scale <- conv2_1_ex
I0510 16:30:45.135181 39557 net.cpp:367] conv2_1_ex/scale -> conv2_1_ex (in-place)
I0510 16:30:45.135223 39557 layer_factory.hpp:77] Creating layer conv2_1_ex/scale
I0510 16:30:45.135342 39557 net.cpp:122] Setting up conv2_1_ex/scale
I0510 16:30:45.135354 39557 net.cpp:129] Top shape: 64 128 28 24 (5505024)
I0510 16:30:45.135356 39557 net.cpp:137] Memory required for data: 922092288
I0510 16:30:45.135362 39557 layer_factory.hpp:77] Creating layer relu2_1_ex
I0510 16:30:45.135367 39557 net.cpp:84] Creating Layer relu2_1_ex
I0510 16:30:45.135371 39557 net.cpp:406] relu2_1_ex <- conv2_1_ex
I0510 16:30:45.135375 39557 net.cpp:367] relu2_1_ex -> conv2_1_ex (in-place)
I0510 16:30:45.135527 39557 net.cpp:122] Setting up relu2_1_ex
I0510 16:30:45.135536 39557 net.cpp:129] Top shape: 64 128 28 24 (5505024)
I0510 16:30:45.135540 39557 net.cpp:137] Memory required for data: 944112384
I0510 16:30:45.135560 39557 layer_factory.hpp:77] Creating layer conv2_1_dw
I0510 16:30:45.135568 39557 net.cpp:84] Creating Layer conv2_1_dw
I0510 16:30:45.135573 39557 net.cpp:406] conv2_1_dw <- conv2_1_ex
I0510 16:30:45.135581 39557 net.cpp:380] conv2_1_dw -> conv2_1_dw
I0510 16:30:45.135795 39557 net.cpp:122] Setting up conv2_1_dw
I0510 16:30:45.135804 39557 net.cpp:129] Top shape: 64 128 28 24 (5505024)
I0510 16:30:45.135807 39557 net.cpp:137] Memory required for data: 966132480
I0510 16:30:45.135823 39557 layer_factory.hpp:77] Creating layer conv2_1_dw/bn
I0510 16:30:45.135839 39557 net.cpp:84] Creating Layer conv2_1_dw/bn
I0510 16:30:45.135843 39557 net.cpp:406] conv2_1_dw/bn <- conv2_1_dw
I0510 16:30:45.135850 39557 net.cpp:367] conv2_1_dw/bn -> conv2_1_dw (in-place)
I0510 16:30:45.136029 39557 net.cpp:122] Setting up conv2_1_dw/bn
I0510 16:30:45.136037 39557 net.cpp:129] Top shape: 64 128 28 24 (5505024)
I0510 16:30:45.136040 39557 net.cpp:137] Memory required for data: 988152576
I0510 16:30:45.136047 39557 layer_factory.hpp:77] Creating layer conv2_1_dw/scale
I0510 16:30:45.136056 39557 net.cpp:84] Creating Layer conv2_1_dw/scale
I0510 16:30:45.136060 39557 net.cpp:406] conv2_1_dw/scale <- conv2_1_dw
I0510 16:30:45.136065 39557 net.cpp:367] conv2_1_dw/scale -> conv2_1_dw (in-place)
I0510 16:30:45.136106 39557 layer_factory.hpp:77] Creating layer conv2_1_dw/scale
I0510 16:30:45.136215 39557 net.cpp:122] Setting up conv2_1_dw/scale
I0510 16:30:45.136224 39557 net.cpp:129] Top shape: 64 128 28 24 (5505024)
I0510 16:30:45.136225 39557 net.cpp:137] Memory required for data: 1010172672
I0510 16:30:45.136232 39557 layer_factory.hpp:77] Creating layer relu2_1_dw
I0510 16:30:45.136240 39557 net.cpp:84] Creating Layer relu2_1_dw
I0510 16:30:45.136245 39557 net.cpp:406] relu2_1_dw <- conv2_1_dw
I0510 16:30:45.136250 39557 net.cpp:367] relu2_1_dw -> conv2_1_dw (in-place)
I0510 16:30:45.136392 39557 net.cpp:122] Setting up relu2_1_dw
I0510 16:30:45.136400 39557 net.cpp:129] Top shape: 64 128 28 24 (5505024)
I0510 16:30:45.136404 39557 net.cpp:137] Memory required for data: 1032192768
I0510 16:30:45.136409 39557 layer_factory.hpp:77] Creating layer conv2_1_em
I0510 16:30:45.136418 39557 net.cpp:84] Creating Layer conv2_1_em
I0510 16:30:45.136425 39557 net.cpp:406] conv2_1_em <- conv2_1_dw
I0510 16:30:45.136430 39557 net.cpp:380] conv2_1_em -> conv2_1_em
I0510 16:30:45.137631 39557 net.cpp:122] Setting up conv2_1_em
I0510 16:30:45.137662 39557 net.cpp:129] Top shape: 64 64 28 24 (2752512)
I0510 16:30:45.137666 39557 net.cpp:137] Memory required for data: 1043202816
I0510 16:30:45.137672 39557 layer_factory.hpp:77] Creating layer conv2_1_em/bn
I0510 16:30:45.137679 39557 net.cpp:84] Creating Layer conv2_1_em/bn
I0510 16:30:45.137683 39557 net.cpp:406] conv2_1_em/bn <- conv2_1_em
I0510 16:30:45.137691 39557 net.cpp:367] conv2_1_em/bn -> conv2_1_em (in-place)
I0510 16:30:45.137913 39557 net.cpp:122] Setting up conv2_1_em/bn
I0510 16:30:45.137923 39557 net.cpp:129] Top shape: 64 64 28 24 (2752512)
I0510 16:30:45.137926 39557 net.cpp:137] Memory required for data: 1054212864
I0510 16:30:45.137936 39557 layer_factory.hpp:77] Creating layer conv2_1_em/scale
I0510 16:30:45.137944 39557 net.cpp:84] Creating Layer conv2_1_em/scale
I0510 16:30:45.137948 39557 net.cpp:406] conv2_1_em/scale <- conv2_1_em
I0510 16:30:45.137953 39557 net.cpp:367] conv2_1_em/scale -> conv2_1_em (in-place)
I0510 16:30:45.138010 39557 layer_factory.hpp:77] Creating layer conv2_1_em/scale
I0510 16:30:45.138128 39557 net.cpp:122] Setting up conv2_1_em/scale
I0510 16:30:45.138137 39557 net.cpp:129] Top shape: 64 64 28 24 (2752512)
I0510 16:30:45.138140 39557 net.cpp:137] Memory required for data: 1065222912
I0510 16:30:45.138146 39557 layer_factory.hpp:77] Creating layer res2_1
I0510 16:30:45.138152 39557 net.cpp:84] Creating Layer res2_1
I0510 16:30:45.138155 39557 net.cpp:406] res2_1 <- conv2_em_conv2_em/scale_0_split_1
I0510 16:30:45.138159 39557 net.cpp:406] res2_1 <- conv2_1_em
I0510 16:30:45.138164 39557 net.cpp:380] res2_1 -> res2_1
I0510 16:30:45.138202 39557 net.cpp:122] Setting up res2_1
I0510 16:30:45.138211 39557 net.cpp:129] Top shape: 64 64 28 24 (2752512)
I0510 16:30:45.138214 39557 net.cpp:137] Memory required for data: 1076232960
I0510 16:30:45.138217 39557 layer_factory.hpp:77] Creating layer res2_1_res2_1_0_split
I0510 16:30:45.138223 39557 net.cpp:84] Creating Layer res2_1_res2_1_0_split
I0510 16:30:45.138226 39557 net.cpp:406] res2_1_res2_1_0_split <- res2_1
I0510 16:30:45.138231 39557 net.cpp:380] res2_1_res2_1_0_split -> res2_1_res2_1_0_split_0
I0510 16:30:45.138243 39557 net.cpp:380] res2_1_res2_1_0_split -> res2_1_res2_1_0_split_1
I0510 16:30:45.138283 39557 net.cpp:122] Setting up res2_1_res2_1_0_split
I0510 16:30:45.138290 39557 net.cpp:129] Top shape: 64 64 28 24 (2752512)
I0510 16:30:45.138294 39557 net.cpp:129] Top shape: 64 64 28 24 (2752512)
I0510 16:30:45.138296 39557 net.cpp:137] Memory required for data: 1098253056
I0510 16:30:45.138301 39557 layer_factory.hpp:77] Creating layer conv2_2_ex
I0510 16:30:45.138327 39557 net.cpp:84] Creating Layer conv2_2_ex
I0510 16:30:45.138332 39557 net.cpp:406] conv2_2_ex <- res2_1_res2_1_0_split_0
I0510 16:30:45.138340 39557 net.cpp:380] conv2_2_ex -> conv2_2_ex
I0510 16:30:45.139557 39557 net.cpp:122] Setting up conv2_2_ex
I0510 16:30:45.139573 39557 net.cpp:129] Top shape: 64 128 28 24 (5505024)
I0510 16:30:45.139588 39557 net.cpp:137] Memory required for data: 1120273152
I0510 16:30:45.139593 39557 layer_factory.hpp:77] Creating layer conv2_2_ex/bn
I0510 16:30:45.139602 39557 net.cpp:84] Creating Layer conv2_2_ex/bn
I0510 16:30:45.139606 39557 net.cpp:406] conv2_2_ex/bn <- conv2_2_ex
I0510 16:30:45.139614 39557 net.cpp:367] conv2_2_ex/bn -> conv2_2_ex (in-place)
I0510 16:30:45.139822 39557 net.cpp:122] Setting up conv2_2_ex/bn
I0510 16:30:45.139830 39557 net.cpp:129] Top shape: 64 128 28 24 (5505024)
I0510 16:30:45.139833 39557 net.cpp:137] Memory required for data: 1142293248
I0510 16:30:45.139840 39557 layer_factory.hpp:77] Creating layer conv2_2_ex/scale
I0510 16:30:45.139848 39557 net.cpp:84] Creating Layer conv2_2_ex/scale
I0510 16:30:45.139853 39557 net.cpp:406] conv2_2_ex/scale <- conv2_2_ex
I0510 16:30:45.139858 39557 net.cpp:367] conv2_2_ex/scale -> conv2_2_ex (in-place)
I0510 16:30:45.139904 39557 layer_factory.hpp:77] Creating layer conv2_2_ex/scale
I0510 16:30:45.140020 39557 net.cpp:122] Setting up conv2_2_ex/scale
I0510 16:30:45.140029 39557 net.cpp:129] Top shape: 64 128 28 24 (5505024)
I0510 16:30:45.140033 39557 net.cpp:137] Memory required for data: 1164313344
I0510 16:30:45.140041 39557 layer_factory.hpp:77] Creating layer relu2_2_ex
I0510 16:30:45.140048 39557 net.cpp:84] Creating Layer relu2_2_ex
I0510 16:30:45.140053 39557 net.cpp:406] relu2_2_ex <- conv2_2_ex
I0510 16:30:45.140058 39557 net.cpp:367] relu2_2_ex -> conv2_2_ex (in-place)
I0510 16:30:45.140203 39557 net.cpp:122] Setting up relu2_2_ex
I0510 16:30:45.140211 39557 net.cpp:129] Top shape: 64 128 28 24 (5505024)
I0510 16:30:45.140214 39557 net.cpp:137] Memory required for data: 1186333440
I0510 16:30:45.140219 39557 layer_factory.hpp:77] Creating layer conv2_2_dw
I0510 16:30:45.140228 39557 net.cpp:84] Creating Layer conv2_2_dw
I0510 16:30:45.140244 39557 net.cpp:406] conv2_2_dw <- conv2_2_ex
I0510 16:30:45.140249 39557 net.cpp:380] conv2_2_dw -> conv2_2_dw
I0510 16:30:45.140449 39557 net.cpp:122] Setting up conv2_2_dw
I0510 16:30:45.140458 39557 net.cpp:129] Top shape: 64 128 28 24 (5505024)
I0510 16:30:45.140461 39557 net.cpp:137] Memory required for data: 1208353536
I0510 16:30:45.140466 39557 layer_factory.hpp:77] Creating layer conv2_2_dw/bn
I0510 16:30:45.140472 39557 net.cpp:84] Creating Layer conv2_2_dw/bn
I0510 16:30:45.140475 39557 net.cpp:406] conv2_2_dw/bn <- conv2_2_dw
I0510 16:30:45.140482 39557 net.cpp:367] conv2_2_dw/bn -> conv2_2_dw (in-place)
I0510 16:30:45.140672 39557 net.cpp:122] Setting up conv2_2_dw/bn
I0510 16:30:45.140681 39557 net.cpp:129] Top shape: 64 128 28 24 (5505024)
I0510 16:30:45.140683 39557 net.cpp:137] Memory required for data: 1230373632
I0510 16:30:45.140697 39557 layer_factory.hpp:77] Creating layer conv2_2_dw/scale
I0510 16:30:45.140719 39557 net.cpp:84] Creating Layer conv2_2_dw/scale
I0510 16:30:45.140724 39557 net.cpp:406] conv2_2_dw/scale <- conv2_2_dw
I0510 16:30:45.140729 39557 net.cpp:367] conv2_2_dw/scale -> conv2_2_dw (in-place)
I0510 16:30:45.140789 39557 layer_factory.hpp:77] Creating layer conv2_2_dw/scale
I0510 16:30:45.140903 39557 net.cpp:122] Setting up conv2_2_dw/scale
I0510 16:30:45.140911 39557 net.cpp:129] Top shape: 64 128 28 24 (5505024)
I0510 16:30:45.140914 39557 net.cpp:137] Memory required for data: 1252393728
I0510 16:30:45.140920 39557 layer_factory.hpp:77] Creating layer relu2_2_dw
I0510 16:30:45.140925 39557 net.cpp:84] Creating Layer relu2_2_dw
I0510 16:30:45.140928 39557 net.cpp:406] relu2_2_dw <- conv2_2_dw
I0510 16:30:45.140938 39557 net.cpp:367] relu2_2_dw -> conv2_2_dw (in-place)
I0510 16:30:45.142144 39557 net.cpp:122] Setting up relu2_2_dw
I0510 16:30:45.142168 39557 net.cpp:129] Top shape: 64 128 28 24 (5505024)
I0510 16:30:45.142171 39557 net.cpp:137] Memory required for data: 1274413824
I0510 16:30:45.142177 39557 layer_factory.hpp:77] Creating layer conv2_2_em
I0510 16:30:45.142186 39557 net.cpp:84] Creating Layer conv2_2_em
I0510 16:30:45.142189 39557 net.cpp:406] conv2_2_em <- conv2_2_dw
I0510 16:30:45.142199 39557 net.cpp:380] conv2_2_em -> conv2_2_em
I0510 16:30:45.143365 39557 net.cpp:122] Setting up conv2_2_em
I0510 16:30:45.143379 39557 net.cpp:129] Top shape: 64 64 28 24 (2752512)
I0510 16:30:45.143393 39557 net.cpp:137] Memory required for data: 1285423872
I0510 16:30:45.143400 39557 layer_factory.hpp:77] Creating layer conv2_2_em/bn
I0510 16:30:45.143409 39557 net.cpp:84] Creating Layer conv2_2_em/bn
I0510 16:30:45.143414 39557 net.cpp:406] conv2_2_em/bn <- conv2_2_em
I0510 16:30:45.143420 39557 net.cpp:367] conv2_2_em/bn -> conv2_2_em (in-place)
I0510 16:30:45.143625 39557 net.cpp:122] Setting up conv2_2_em/bn
I0510 16:30:45.143633 39557 net.cpp:129] Top shape: 64 64 28 24 (2752512)
I0510 16:30:45.143636 39557 net.cpp:137] Memory required for data: 1296433920
I0510 16:30:45.143643 39557 layer_factory.hpp:77] Creating layer conv2_2_em/scale
I0510 16:30:45.143651 39557 net.cpp:84] Creating Layer conv2_2_em/scale
I0510 16:30:45.143656 39557 net.cpp:406] conv2_2_em/scale <- conv2_2_em
I0510 16:30:45.143659 39557 net.cpp:367] conv2_2_em/scale -> conv2_2_em (in-place)
I0510 16:30:45.143705 39557 layer_factory.hpp:77] Creating layer conv2_2_em/scale
I0510 16:30:45.143827 39557 net.cpp:122] Setting up conv2_2_em/scale
I0510 16:30:45.143837 39557 net.cpp:129] Top shape: 64 64 28 24 (2752512)
I0510 16:30:45.143841 39557 net.cpp:137] Memory required for data: 1307443968
I0510 16:30:45.143847 39557 layer_factory.hpp:77] Creating layer res2_2
I0510 16:30:45.143852 39557 net.cpp:84] Creating Layer res2_2
I0510 16:30:45.143857 39557 net.cpp:406] res2_2 <- res2_1_res2_1_0_split_1
I0510 16:30:45.143860 39557 net.cpp:406] res2_2 <- conv2_2_em
I0510 16:30:45.143864 39557 net.cpp:380] res2_2 -> res2_2
I0510 16:30:45.143892 39557 net.cpp:122] Setting up res2_2
I0510 16:30:45.143899 39557 net.cpp:129] Top shape: 64 64 28 24 (2752512)
I0510 16:30:45.143903 39557 net.cpp:137] Memory required for data: 1318454016
I0510 16:30:45.143906 39557 layer_factory.hpp:77] Creating layer res2_2_res2_2_0_split
I0510 16:30:45.143913 39557 net.cpp:84] Creating Layer res2_2_res2_2_0_split
I0510 16:30:45.143915 39557 net.cpp:406] res2_2_res2_2_0_split <- res2_2
I0510 16:30:45.143920 39557 net.cpp:380] res2_2_res2_2_0_split -> res2_2_res2_2_0_split_0
I0510 16:30:45.143929 39557 net.cpp:380] res2_2_res2_2_0_split -> res2_2_res2_2_0_split_1
I0510 16:30:45.143963 39557 net.cpp:122] Setting up res2_2_res2_2_0_split
I0510 16:30:45.143970 39557 net.cpp:129] Top shape: 64 64 28 24 (2752512)
I0510 16:30:45.143975 39557 net.cpp:129] Top shape: 64 64 28 24 (2752512)
I0510 16:30:45.143976 39557 net.cpp:137] Memory required for data: 1340474112
I0510 16:30:45.143980 39557 layer_factory.hpp:77] Creating layer conv2_3_ex
I0510 16:30:45.143990 39557 net.cpp:84] Creating Layer conv2_3_ex
I0510 16:30:45.144004 39557 net.cpp:406] conv2_3_ex <- res2_2_res2_2_0_split_0
I0510 16:30:45.144011 39557 net.cpp:380] conv2_3_ex -> conv2_3_ex
I0510 16:30:45.145143 39557 net.cpp:122] Setting up conv2_3_ex
I0510 16:30:45.145159 39557 net.cpp:129] Top shape: 64 128 28 24 (5505024)
I0510 16:30:45.145164 39557 net.cpp:137] Memory required for data: 1362494208
I0510 16:30:45.145169 39557 layer_factory.hpp:77] Creating layer conv2_3_ex/bn
I0510 16:30:45.145175 39557 net.cpp:84] Creating Layer conv2_3_ex/bn
I0510 16:30:45.145179 39557 net.cpp:406] conv2_3_ex/bn <- conv2_3_ex
I0510 16:30:45.145186 39557 net.cpp:367] conv2_3_ex/bn -> conv2_3_ex (in-place)
I0510 16:30:45.145383 39557 net.cpp:122] Setting up conv2_3_ex/bn
I0510 16:30:45.145392 39557 net.cpp:129] Top shape: 64 128 28 24 (5505024)
I0510 16:30:45.145395 39557 net.cpp:137] Memory required for data: 1384514304
I0510 16:30:45.145406 39557 layer_factory.hpp:77] Creating layer conv2_3_ex/scale
I0510 16:30:45.145412 39557 net.cpp:84] Creating Layer conv2_3_ex/scale
I0510 16:30:45.145416 39557 net.cpp:406] conv2_3_ex/scale <- conv2_3_ex
I0510 16:30:45.145422 39557 net.cpp:367] conv2_3_ex/scale -> conv2_3_ex (in-place)
I0510 16:30:45.145462 39557 layer_factory.hpp:77] Creating layer conv2_3_ex/scale
I0510 16:30:45.145579 39557 net.cpp:122] Setting up conv2_3_ex/scale
I0510 16:30:45.145587 39557 net.cpp:129] Top shape: 64 128 28 24 (5505024)
I0510 16:30:45.145591 39557 net.cpp:137] Memory required for data: 1406534400
I0510 16:30:45.145596 39557 layer_factory.hpp:77] Creating layer relu2_3_ex
I0510 16:30:45.145601 39557 net.cpp:84] Creating Layer relu2_3_ex
I0510 16:30:45.145606 39557 net.cpp:406] relu2_3_ex <- conv2_3_ex
I0510 16:30:45.145613 39557 net.cpp:367] relu2_3_ex -> conv2_3_ex (in-place)
I0510 16:30:45.145746 39557 net.cpp:122] Setting up relu2_3_ex
I0510 16:30:45.145756 39557 net.cpp:129] Top shape: 64 128 28 24 (5505024)
I0510 16:30:45.145758 39557 net.cpp:137] Memory required for data: 1428554496
I0510 16:30:45.145762 39557 layer_factory.hpp:77] Creating layer conv2_3_dw
I0510 16:30:45.145769 39557 net.cpp:84] Creating Layer conv2_3_dw
I0510 16:30:45.145772 39557 net.cpp:406] conv2_3_dw <- conv2_3_ex
I0510 16:30:45.145781 39557 net.cpp:380] conv2_3_dw -> conv2_3_dw
I0510 16:30:45.145977 39557 net.cpp:122] Setting up conv2_3_dw
I0510 16:30:45.145987 39557 net.cpp:129] Top shape: 64 128 28 24 (5505024)
I0510 16:30:45.145989 39557 net.cpp:137] Memory required for data: 1450574592
I0510 16:30:45.145995 39557 layer_factory.hpp:77] Creating layer conv2_3_dw/bn
I0510 16:30:45.146001 39557 net.cpp:84] Creating Layer conv2_3_dw/bn
I0510 16:30:45.146005 39557 net.cpp:406] conv2_3_dw/bn <- conv2_3_dw
I0510 16:30:45.146010 39557 net.cpp:367] conv2_3_dw/bn -> conv2_3_dw (in-place)
I0510 16:30:45.146199 39557 net.cpp:122] Setting up conv2_3_dw/bn
I0510 16:30:45.146209 39557 net.cpp:129] Top shape: 64 128 28 24 (5505024)
I0510 16:30:45.146211 39557 net.cpp:137] Memory required for data: 1472594688
I0510 16:30:45.146217 39557 layer_factory.hpp:77] Creating layer conv2_3_dw/scale
I0510 16:30:45.146224 39557 net.cpp:84] Creating Layer conv2_3_dw/scale
I0510 16:30:45.146227 39557 net.cpp:406] conv2_3_dw/scale <- conv2_3_dw
I0510 16:30:45.146231 39557 net.cpp:367] conv2_3_dw/scale -> conv2_3_dw (in-place)
I0510 16:30:45.146276 39557 layer_factory.hpp:77] Creating layer conv2_3_dw/scale
I0510 16:30:45.146389 39557 net.cpp:122] Setting up conv2_3_dw/scale
I0510 16:30:45.146397 39557 net.cpp:129] Top shape: 64 128 28 24 (5505024)
I0510 16:30:45.146400 39557 net.cpp:137] Memory required for data: 1494614784
I0510 16:30:45.146405 39557 layer_factory.hpp:77] Creating layer relu2_3_dw
I0510 16:30:45.146412 39557 net.cpp:84] Creating Layer relu2_3_dw
I0510 16:30:45.146415 39557 net.cpp:406] relu2_3_dw <- conv2_3_dw
I0510 16:30:45.146420 39557 net.cpp:367] relu2_3_dw -> conv2_3_dw (in-place)
I0510 16:30:45.146549 39557 net.cpp:122] Setting up relu2_3_dw
I0510 16:30:45.146558 39557 net.cpp:129] Top shape: 64 128 28 24 (5505024)
I0510 16:30:45.146560 39557 net.cpp:137] Memory required for data: 1516634880
I0510 16:30:45.146579 39557 layer_factory.hpp:77] Creating layer conv2_3_em
I0510 16:30:45.146589 39557 net.cpp:84] Creating Layer conv2_3_em
I0510 16:30:45.146594 39557 net.cpp:406] conv2_3_em <- conv2_3_dw
I0510 16:30:45.146598 39557 net.cpp:380] conv2_3_em -> conv2_3_em
I0510 16:30:45.147781 39557 net.cpp:122] Setting up conv2_3_em
I0510 16:30:45.147795 39557 net.cpp:129] Top shape: 64 64 28 24 (2752512)
I0510 16:30:45.147799 39557 net.cpp:137] Memory required for data: 1527644928
I0510 16:30:45.147805 39557 layer_factory.hpp:77] Creating layer conv2_3_em/bn
I0510 16:30:45.147812 39557 net.cpp:84] Creating Layer conv2_3_em/bn
I0510 16:30:45.147815 39557 net.cpp:406] conv2_3_em/bn <- conv2_3_em
I0510 16:30:45.147821 39557 net.cpp:367] conv2_3_em/bn -> conv2_3_em (in-place)
I0510 16:30:45.148047 39557 net.cpp:122] Setting up conv2_3_em/bn
I0510 16:30:45.148056 39557 net.cpp:129] Top shape: 64 64 28 24 (2752512)
I0510 16:30:45.148059 39557 net.cpp:137] Memory required for data: 1538654976
I0510 16:30:45.148069 39557 layer_factory.hpp:77] Creating layer conv2_3_em/scale
I0510 16:30:45.148078 39557 net.cpp:84] Creating Layer conv2_3_em/scale
I0510 16:30:45.148082 39557 net.cpp:406] conv2_3_em/scale <- conv2_3_em
I0510 16:30:45.148088 39557 net.cpp:367] conv2_3_em/scale -> conv2_3_em (in-place)
I0510 16:30:45.148133 39557 layer_factory.hpp:77] Creating layer conv2_3_em/scale
I0510 16:30:45.148262 39557 net.cpp:122] Setting up conv2_3_em/scale
I0510 16:30:45.148272 39557 net.cpp:129] Top shape: 64 64 28 24 (2752512)
I0510 16:30:45.148275 39557 net.cpp:137] Memory required for data: 1549665024
I0510 16:30:45.148283 39557 layer_factory.hpp:77] Creating layer res2_3
I0510 16:30:45.148288 39557 net.cpp:84] Creating Layer res2_3
I0510 16:30:45.148293 39557 net.cpp:406] res2_3 <- res2_2_res2_2_0_split_1
I0510 16:30:45.148296 39557 net.cpp:406] res2_3 <- conv2_3_em
I0510 16:30:45.148301 39557 net.cpp:380] res2_3 -> res2_3
I0510 16:30:45.148330 39557 net.cpp:122] Setting up res2_3
I0510 16:30:45.148339 39557 net.cpp:129] Top shape: 64 64 28 24 (2752512)
I0510 16:30:45.148353 39557 net.cpp:137] Memory required for data: 1560675072
I0510 16:30:45.148356 39557 layer_factory.hpp:77] Creating layer res2_3_res2_3_0_split
I0510 16:30:45.148362 39557 net.cpp:84] Creating Layer res2_3_res2_3_0_split
I0510 16:30:45.148365 39557 net.cpp:406] res2_3_res2_3_0_split <- res2_3
I0510 16:30:45.148370 39557 net.cpp:380] res2_3_res2_3_0_split -> res2_3_res2_3_0_split_0
I0510 16:30:45.148376 39557 net.cpp:380] res2_3_res2_3_0_split -> res2_3_res2_3_0_split_1
I0510 16:30:45.148414 39557 net.cpp:122] Setting up res2_3_res2_3_0_split
I0510 16:30:45.148422 39557 net.cpp:129] Top shape: 64 64 28 24 (2752512)
I0510 16:30:45.148425 39557 net.cpp:129] Top shape: 64 64 28 24 (2752512)
I0510 16:30:45.148428 39557 net.cpp:137] Memory required for data: 1582695168
I0510 16:30:45.148432 39557 layer_factory.hpp:77] Creating layer conv2_4_ex
I0510 16:30:45.148442 39557 net.cpp:84] Creating Layer conv2_4_ex
I0510 16:30:45.148445 39557 net.cpp:406] conv2_4_ex <- res2_3_res2_3_0_split_0
I0510 16:30:45.148452 39557 net.cpp:380] conv2_4_ex -> conv2_4_ex
I0510 16:30:45.149595 39557 net.cpp:122] Setting up conv2_4_ex
I0510 16:30:45.149608 39557 net.cpp:129] Top shape: 64 128 28 24 (5505024)
I0510 16:30:45.149622 39557 net.cpp:137] Memory required for data: 1604715264
I0510 16:30:45.149629 39557 layer_factory.hpp:77] Creating layer conv2_4_ex/bn
I0510 16:30:45.149638 39557 net.cpp:84] Creating Layer conv2_4_ex/bn
I0510 16:30:45.149642 39557 net.cpp:406] conv2_4_ex/bn <- conv2_4_ex
I0510 16:30:45.149652 39557 net.cpp:367] conv2_4_ex/bn -> conv2_4_ex (in-place)
I0510 16:30:45.149852 39557 net.cpp:122] Setting up conv2_4_ex/bn
I0510 16:30:45.149860 39557 net.cpp:129] Top shape: 64 128 28 24 (5505024)
I0510 16:30:45.149863 39557 net.cpp:137] Memory required for data: 1626735360
I0510 16:30:45.149870 39557 layer_factory.hpp:77] Creating layer conv2_4_ex/scale
I0510 16:30:45.149876 39557 net.cpp:84] Creating Layer conv2_4_ex/scale
I0510 16:30:45.149880 39557 net.cpp:406] conv2_4_ex/scale <- conv2_4_ex
I0510 16:30:45.149899 39557 net.cpp:367] conv2_4_ex/scale -> conv2_4_ex (in-place)
I0510 16:30:45.149943 39557 layer_factory.hpp:77] Creating layer conv2_4_ex/scale
I0510 16:30:45.150063 39557 net.cpp:122] Setting up conv2_4_ex/scale
I0510 16:30:45.150071 39557 net.cpp:129] Top shape: 64 128 28 24 (5505024)
I0510 16:30:45.150074 39557 net.cpp:137] Memory required for data: 1648755456
I0510 16:30:45.150080 39557 layer_factory.hpp:77] Creating layer relu2_4_ex
I0510 16:30:45.150091 39557 net.cpp:84] Creating Layer relu2_4_ex
I0510 16:30:45.150094 39557 net.cpp:406] relu2_4_ex <- conv2_4_ex
I0510 16:30:45.150100 39557 net.cpp:367] relu2_4_ex -> conv2_4_ex (in-place)
I0510 16:30:45.150234 39557 net.cpp:122] Setting up relu2_4_ex
I0510 16:30:45.150243 39557 net.cpp:129] Top shape: 64 128 28 24 (5505024)
I0510 16:30:45.150246 39557 net.cpp:137] Memory required for data: 1670775552
I0510 16:30:45.150254 39557 layer_factory.hpp:77] Creating layer conv2_4_dw
I0510 16:30:45.150264 39557 net.cpp:84] Creating Layer conv2_4_dw
I0510 16:30:45.150267 39557 net.cpp:406] conv2_4_dw <- conv2_4_ex
I0510 16:30:45.150274 39557 net.cpp:380] conv2_4_dw -> conv2_4_dw
I0510 16:30:45.150473 39557 net.cpp:122] Setting up conv2_4_dw
I0510 16:30:45.150482 39557 net.cpp:129] Top shape: 64 128 28 24 (5505024)
I0510 16:30:45.150486 39557 net.cpp:137] Memory required for data: 1692795648
I0510 16:30:45.150490 39557 layer_factory.hpp:77] Creating layer conv2_4_dw/bn
I0510 16:30:45.150498 39557 net.cpp:84] Creating Layer conv2_4_dw/bn
I0510 16:30:45.150502 39557 net.cpp:406] conv2_4_dw/bn <- conv2_4_dw
I0510 16:30:45.150506 39557 net.cpp:367] conv2_4_dw/bn -> conv2_4_dw (in-place)
I0510 16:30:45.150699 39557 net.cpp:122] Setting up conv2_4_dw/bn
I0510 16:30:45.150707 39557 net.cpp:129] Top shape: 64 128 28 24 (5505024)
I0510 16:30:45.150710 39557 net.cpp:137] Memory required for data: 1714815744
I0510 16:30:45.150717 39557 layer_factory.hpp:77] Creating layer conv2_4_dw/scale
I0510 16:30:45.150725 39557 net.cpp:84] Creating Layer conv2_4_dw/scale
I0510 16:30:45.150728 39557 net.cpp:406] conv2_4_dw/scale <- conv2_4_dw
I0510 16:30:45.150733 39557 net.cpp:367] conv2_4_dw/scale -> conv2_4_dw (in-place)
I0510 16:30:45.150774 39557 layer_factory.hpp:77] Creating layer conv2_4_dw/scale
I0510 16:30:45.150892 39557 net.cpp:122] Setting up conv2_4_dw/scale
I0510 16:30:45.150899 39557 net.cpp:129] Top shape: 64 128 28 24 (5505024)
I0510 16:30:45.150902 39557 net.cpp:137] Memory required for data: 1736835840
I0510 16:30:45.150909 39557 layer_factory.hpp:77] Creating layer relu2_4_dw
I0510 16:30:45.150916 39557 net.cpp:84] Creating Layer relu2_4_dw
I0510 16:30:45.150920 39557 net.cpp:406] relu2_4_dw <- conv2_4_dw
I0510 16:30:45.150924 39557 net.cpp:367] relu2_4_dw -> conv2_4_dw (in-place)
I0510 16:30:45.151057 39557 net.cpp:122] Setting up relu2_4_dw
I0510 16:30:45.151065 39557 net.cpp:129] Top shape: 64 128 28 24 (5505024)
I0510 16:30:45.151068 39557 net.cpp:137] Memory required for data: 1758855936
I0510 16:30:45.151073 39557 layer_factory.hpp:77] Creating layer conv2_4_em
I0510 16:30:45.151082 39557 net.cpp:84] Creating Layer conv2_4_em
I0510 16:30:45.151087 39557 net.cpp:406] conv2_4_em <- conv2_4_dw
I0510 16:30:45.151091 39557 net.cpp:380] conv2_4_em -> conv2_4_em
I0510 16:30:45.152259 39557 net.cpp:122] Setting up conv2_4_em
I0510 16:30:45.152283 39557 net.cpp:129] Top shape: 64 64 28 24 (2752512)
I0510 16:30:45.152287 39557 net.cpp:137] Memory required for data: 1769865984
I0510 16:30:45.152292 39557 layer_factory.hpp:77] Creating layer conv2_4_em/bn
I0510 16:30:45.152302 39557 net.cpp:84] Creating Layer conv2_4_em/bn
I0510 16:30:45.152305 39557 net.cpp:406] conv2_4_em/bn <- conv2_4_em
I0510 16:30:45.152312 39557 net.cpp:367] conv2_4_em/bn -> conv2_4_em (in-place)
I0510 16:30:45.152521 39557 net.cpp:122] Setting up conv2_4_em/bn
I0510 16:30:45.152530 39557 net.cpp:129] Top shape: 64 64 28 24 (2752512)
I0510 16:30:45.152534 39557 net.cpp:137] Memory required for data: 1780876032
I0510 16:30:45.152542 39557 layer_factory.hpp:77] Creating layer conv2_4_em/scale
I0510 16:30:45.152561 39557 net.cpp:84] Creating Layer conv2_4_em/scale
I0510 16:30:45.152566 39557 net.cpp:406] conv2_4_em/scale <- conv2_4_em
I0510 16:30:45.152573 39557 net.cpp:367] conv2_4_em/scale -> conv2_4_em (in-place)
I0510 16:30:45.152631 39557 layer_factory.hpp:77] Creating layer conv2_4_em/scale
I0510 16:30:45.152760 39557 net.cpp:122] Setting up conv2_4_em/scale
I0510 16:30:45.152768 39557 net.cpp:129] Top shape: 64 64 28 24 (2752512)
I0510 16:30:45.152771 39557 net.cpp:137] Memory required for data: 1791886080
I0510 16:30:45.152777 39557 layer_factory.hpp:77] Creating layer res2_4
I0510 16:30:45.152786 39557 net.cpp:84] Creating Layer res2_4
I0510 16:30:45.152789 39557 net.cpp:406] res2_4 <- res2_3_res2_3_0_split_1
I0510 16:30:45.152794 39557 net.cpp:406] res2_4 <- conv2_4_em
I0510 16:30:45.152802 39557 net.cpp:380] res2_4 -> res2_4
I0510 16:30:45.152833 39557 net.cpp:122] Setting up res2_4
I0510 16:30:45.152842 39557 net.cpp:129] Top shape: 64 64 28 24 (2752512)
I0510 16:30:45.152845 39557 net.cpp:137] Memory required for data: 1802896128
I0510 16:30:45.152848 39557 layer_factory.hpp:77] Creating layer conv3_ex
I0510 16:30:45.152856 39557 net.cpp:84] Creating Layer conv3_ex
I0510 16:30:45.152859 39557 net.cpp:406] conv3_ex <- res2_4
I0510 16:30:45.152868 39557 net.cpp:380] conv3_ex -> conv3_ex
I0510 16:30:45.154090 39557 net.cpp:122] Setting up conv3_ex
I0510 16:30:45.154105 39557 net.cpp:129] Top shape: 64 256 28 24 (11010048)
I0510 16:30:45.154120 39557 net.cpp:137] Memory required for data: 1846936320
I0510 16:30:45.154129 39557 layer_factory.hpp:77] Creating layer conv3_ex/bn
I0510 16:30:45.154137 39557 net.cpp:84] Creating Layer conv3_ex/bn
I0510 16:30:45.154141 39557 net.cpp:406] conv3_ex/bn <- conv3_ex
I0510 16:30:45.154148 39557 net.cpp:367] conv3_ex/bn -> conv3_ex (in-place)
I0510 16:30:45.154356 39557 net.cpp:122] Setting up conv3_ex/bn
I0510 16:30:45.154364 39557 net.cpp:129] Top shape: 64 256 28 24 (11010048)
I0510 16:30:45.154367 39557 net.cpp:137] Memory required for data: 1890976512
I0510 16:30:45.154376 39557 layer_factory.hpp:77] Creating layer conv3_ex/scale
I0510 16:30:45.154383 39557 net.cpp:84] Creating Layer conv3_ex/scale
I0510 16:30:45.154386 39557 net.cpp:406] conv3_ex/scale <- conv3_ex
I0510 16:30:45.154392 39557 net.cpp:367] conv3_ex/scale -> conv3_ex (in-place)
I0510 16:30:45.154438 39557 layer_factory.hpp:77] Creating layer conv3_ex/scale
I0510 16:30:45.154561 39557 net.cpp:122] Setting up conv3_ex/scale
I0510 16:30:45.154569 39557 net.cpp:129] Top shape: 64 256 28 24 (11010048)
I0510 16:30:45.154572 39557 net.cpp:137] Memory required for data: 1935016704
I0510 16:30:45.154578 39557 layer_factory.hpp:77] Creating layer relu3_ex
I0510 16:30:45.154583 39557 net.cpp:84] Creating Layer relu3_ex
I0510 16:30:45.154587 39557 net.cpp:406] relu3_ex <- conv3_ex
I0510 16:30:45.154593 39557 net.cpp:367] relu3_ex -> conv3_ex (in-place)
I0510 16:30:45.155886 39557 net.cpp:122] Setting up relu3_ex
I0510 16:30:45.155901 39557 net.cpp:129] Top shape: 64 256 28 24 (11010048)
I0510 16:30:45.155915 39557 net.cpp:137] Memory required for data: 1979056896
I0510 16:30:45.155923 39557 layer_factory.hpp:77] Creating layer conv3_dw
I0510 16:30:45.155932 39557 net.cpp:84] Creating Layer conv3_dw
I0510 16:30:45.155936 39557 net.cpp:406] conv3_dw <- conv3_ex
I0510 16:30:45.155943 39557 net.cpp:380] conv3_dw -> conv3_dw
I0510 16:30:45.156169 39557 net.cpp:122] Setting up conv3_dw
I0510 16:30:45.156180 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.156183 39557 net.cpp:137] Memory required for data: 1990066944
I0510 16:30:45.156188 39557 layer_factory.hpp:77] Creating layer conv3_dw/bn
I0510 16:30:45.156194 39557 net.cpp:84] Creating Layer conv3_dw/bn
I0510 16:30:45.156198 39557 net.cpp:406] conv3_dw/bn <- conv3_dw
I0510 16:30:45.156203 39557 net.cpp:367] conv3_dw/bn -> conv3_dw (in-place)
I0510 16:30:45.156404 39557 net.cpp:122] Setting up conv3_dw/bn
I0510 16:30:45.156412 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.156415 39557 net.cpp:137] Memory required for data: 2001076992
I0510 16:30:45.156442 39557 layer_factory.hpp:77] Creating layer conv3_dw/scale
I0510 16:30:45.156450 39557 net.cpp:84] Creating Layer conv3_dw/scale
I0510 16:30:45.156453 39557 net.cpp:406] conv3_dw/scale <- conv3_dw
I0510 16:30:45.156458 39557 net.cpp:367] conv3_dw/scale -> conv3_dw (in-place)
I0510 16:30:45.156507 39557 layer_factory.hpp:77] Creating layer conv3_dw/scale
I0510 16:30:45.156628 39557 net.cpp:122] Setting up conv3_dw/scale
I0510 16:30:45.156636 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.156639 39557 net.cpp:137] Memory required for data: 2012087040
I0510 16:30:45.156647 39557 layer_factory.hpp:77] Creating layer relu3_dw
I0510 16:30:45.156656 39557 net.cpp:84] Creating Layer relu3_dw
I0510 16:30:45.156658 39557 net.cpp:406] relu3_dw <- conv3_dw
I0510 16:30:45.156662 39557 net.cpp:367] relu3_dw -> conv3_dw (in-place)
I0510 16:30:45.157883 39557 net.cpp:122] Setting up relu3_dw
I0510 16:30:45.157897 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.157912 39557 net.cpp:137] Memory required for data: 2023097088
I0510 16:30:45.157917 39557 layer_factory.hpp:77] Creating layer conv3_em
I0510 16:30:45.157939 39557 net.cpp:84] Creating Layer conv3_em
I0510 16:30:45.157943 39557 net.cpp:406] conv3_em <- conv3_dw
I0510 16:30:45.157951 39557 net.cpp:380] conv3_em -> conv3_em
I0510 16:30:45.159610 39557 net.cpp:122] Setting up conv3_em
I0510 16:30:45.159624 39557 net.cpp:129] Top shape: 64 128 14 12 (1376256)
I0510 16:30:45.159639 39557 net.cpp:137] Memory required for data: 2028602112
I0510 16:30:45.159658 39557 layer_factory.hpp:77] Creating layer conv3_em/bn
I0510 16:30:45.159669 39557 net.cpp:84] Creating Layer conv3_em/bn
I0510 16:30:45.159674 39557 net.cpp:406] conv3_em/bn <- conv3_em
I0510 16:30:45.159682 39557 net.cpp:367] conv3_em/bn -> conv3_em (in-place)
I0510 16:30:45.159895 39557 net.cpp:122] Setting up conv3_em/bn
I0510 16:30:45.159905 39557 net.cpp:129] Top shape: 64 128 14 12 (1376256)
I0510 16:30:45.159909 39557 net.cpp:137] Memory required for data: 2034107136
I0510 16:30:45.159917 39557 layer_factory.hpp:77] Creating layer conv3_em/scale
I0510 16:30:45.159924 39557 net.cpp:84] Creating Layer conv3_em/scale
I0510 16:30:45.159927 39557 net.cpp:406] conv3_em/scale <- conv3_em
I0510 16:30:45.159935 39557 net.cpp:367] conv3_em/scale -> conv3_em (in-place)
I0510 16:30:45.159976 39557 layer_factory.hpp:77] Creating layer conv3_em/scale
I0510 16:30:45.160095 39557 net.cpp:122] Setting up conv3_em/scale
I0510 16:30:45.160104 39557 net.cpp:129] Top shape: 64 128 14 12 (1376256)
I0510 16:30:45.160106 39557 net.cpp:137] Memory required for data: 2039612160
I0510 16:30:45.160112 39557 layer_factory.hpp:77] Creating layer conv3_em_conv3_em/scale_0_split
I0510 16:30:45.160118 39557 net.cpp:84] Creating Layer conv3_em_conv3_em/scale_0_split
I0510 16:30:45.160121 39557 net.cpp:406] conv3_em_conv3_em/scale_0_split <- conv3_em
I0510 16:30:45.160130 39557 net.cpp:380] conv3_em_conv3_em/scale_0_split -> conv3_em_conv3_em/scale_0_split_0
I0510 16:30:45.160137 39557 net.cpp:380] conv3_em_conv3_em/scale_0_split -> conv3_em_conv3_em/scale_0_split_1
I0510 16:30:45.160177 39557 net.cpp:122] Setting up conv3_em_conv3_em/scale_0_split
I0510 16:30:45.160185 39557 net.cpp:129] Top shape: 64 128 14 12 (1376256)
I0510 16:30:45.160189 39557 net.cpp:129] Top shape: 64 128 14 12 (1376256)
I0510 16:30:45.160192 39557 net.cpp:137] Memory required for data: 2050622208
I0510 16:30:45.160195 39557 layer_factory.hpp:77] Creating layer conv3_1_ex
I0510 16:30:45.160207 39557 net.cpp:84] Creating Layer conv3_1_ex
I0510 16:30:45.160212 39557 net.cpp:406] conv3_1_ex <- conv3_em_conv3_em/scale_0_split_0
I0510 16:30:45.160218 39557 net.cpp:380] conv3_1_ex -> conv3_1_ex
I0510 16:30:45.161572 39557 net.cpp:122] Setting up conv3_1_ex
I0510 16:30:45.161598 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.161602 39557 net.cpp:137] Memory required for data: 2061632256
I0510 16:30:45.161610 39557 layer_factory.hpp:77] Creating layer conv3_1_ex/bn
I0510 16:30:45.161630 39557 net.cpp:84] Creating Layer conv3_1_ex/bn
I0510 16:30:45.161635 39557 net.cpp:406] conv3_1_ex/bn <- conv3_1_ex
I0510 16:30:45.161641 39557 net.cpp:367] conv3_1_ex/bn -> conv3_1_ex (in-place)
I0510 16:30:45.161850 39557 net.cpp:122] Setting up conv3_1_ex/bn
I0510 16:30:45.161860 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.161864 39557 net.cpp:137] Memory required for data: 2072642304
I0510 16:30:45.161871 39557 layer_factory.hpp:77] Creating layer conv3_1_ex/scale
I0510 16:30:45.161880 39557 net.cpp:84] Creating Layer conv3_1_ex/scale
I0510 16:30:45.161883 39557 net.cpp:406] conv3_1_ex/scale <- conv3_1_ex
I0510 16:30:45.161888 39557 net.cpp:367] conv3_1_ex/scale -> conv3_1_ex (in-place)
I0510 16:30:45.161933 39557 layer_factory.hpp:77] Creating layer conv3_1_ex/scale
I0510 16:30:45.162056 39557 net.cpp:122] Setting up conv3_1_ex/scale
I0510 16:30:45.162065 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.162067 39557 net.cpp:137] Memory required for data: 2083652352
I0510 16:30:45.162073 39557 layer_factory.hpp:77] Creating layer relu3_1_ex
I0510 16:30:45.162086 39557 net.cpp:84] Creating Layer relu3_1_ex
I0510 16:30:45.162088 39557 net.cpp:406] relu3_1_ex <- conv3_1_ex
I0510 16:30:45.162093 39557 net.cpp:367] relu3_1_ex -> conv3_1_ex (in-place)
I0510 16:30:45.162205 39557 net.cpp:122] Setting up relu3_1_ex
I0510 16:30:45.162214 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.162217 39557 net.cpp:137] Memory required for data: 2094662400
I0510 16:30:45.162223 39557 layer_factory.hpp:77] Creating layer conv3_1_dw
I0510 16:30:45.162233 39557 net.cpp:84] Creating Layer conv3_1_dw
I0510 16:30:45.162237 39557 net.cpp:406] conv3_1_dw <- conv3_1_ex
I0510 16:30:45.162243 39557 net.cpp:380] conv3_1_dw -> conv3_1_dw
I0510 16:30:45.162456 39557 net.cpp:122] Setting up conv3_1_dw
I0510 16:30:45.162467 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.162469 39557 net.cpp:137] Memory required for data: 2105672448
I0510 16:30:45.162474 39557 layer_factory.hpp:77] Creating layer conv3_1_dw/bn
I0510 16:30:45.162479 39557 net.cpp:84] Creating Layer conv3_1_dw/bn
I0510 16:30:45.162483 39557 net.cpp:406] conv3_1_dw/bn <- conv3_1_dw
I0510 16:30:45.162490 39557 net.cpp:367] conv3_1_dw/bn -> conv3_1_dw (in-place)
I0510 16:30:45.162683 39557 net.cpp:122] Setting up conv3_1_dw/bn
I0510 16:30:45.162693 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.162695 39557 net.cpp:137] Memory required for data: 2116682496
I0510 16:30:45.162703 39557 layer_factory.hpp:77] Creating layer conv3_1_dw/scale
I0510 16:30:45.162708 39557 net.cpp:84] Creating Layer conv3_1_dw/scale
I0510 16:30:45.162711 39557 net.cpp:406] conv3_1_dw/scale <- conv3_1_dw
I0510 16:30:45.162716 39557 net.cpp:367] conv3_1_dw/scale -> conv3_1_dw (in-place)
I0510 16:30:45.162760 39557 layer_factory.hpp:77] Creating layer conv3_1_dw/scale
I0510 16:30:45.162876 39557 net.cpp:122] Setting up conv3_1_dw/scale
I0510 16:30:45.162886 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.162890 39557 net.cpp:137] Memory required for data: 2127692544
I0510 16:30:45.162897 39557 layer_factory.hpp:77] Creating layer relu3_1_dw
I0510 16:30:45.162902 39557 net.cpp:84] Creating Layer relu3_1_dw
I0510 16:30:45.162906 39557 net.cpp:406] relu3_1_dw <- conv3_1_dw
I0510 16:30:45.162911 39557 net.cpp:367] relu3_1_dw -> conv3_1_dw (in-place)
I0510 16:30:45.163024 39557 net.cpp:122] Setting up relu3_1_dw
I0510 16:30:45.163033 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.163036 39557 net.cpp:137] Memory required for data: 2138702592
I0510 16:30:45.163040 39557 layer_factory.hpp:77] Creating layer conv3_1_em
I0510 16:30:45.163048 39557 net.cpp:84] Creating Layer conv3_1_em
I0510 16:30:45.163053 39557 net.cpp:406] conv3_1_em <- conv3_1_dw
I0510 16:30:45.163059 39557 net.cpp:380] conv3_1_em -> conv3_1_em
I0510 16:30:45.164418 39557 net.cpp:122] Setting up conv3_1_em
I0510 16:30:45.164444 39557 net.cpp:129] Top shape: 64 128 14 12 (1376256)
I0510 16:30:45.164459 39557 net.cpp:137] Memory required for data: 2144207616
I0510 16:30:45.164466 39557 layer_factory.hpp:77] Creating layer conv3_1_em/bn
I0510 16:30:45.164475 39557 net.cpp:84] Creating Layer conv3_1_em/bn
I0510 16:30:45.164479 39557 net.cpp:406] conv3_1_em/bn <- conv3_1_em
I0510 16:30:45.164486 39557 net.cpp:367] conv3_1_em/bn -> conv3_1_em (in-place)
I0510 16:30:45.164690 39557 net.cpp:122] Setting up conv3_1_em/bn
I0510 16:30:45.164700 39557 net.cpp:129] Top shape: 64 128 14 12 (1376256)
I0510 16:30:45.164703 39557 net.cpp:137] Memory required for data: 2149712640
I0510 16:30:45.164712 39557 layer_factory.hpp:77] Creating layer conv3_1_em/scale
I0510 16:30:45.164721 39557 net.cpp:84] Creating Layer conv3_1_em/scale
I0510 16:30:45.164724 39557 net.cpp:406] conv3_1_em/scale <- conv3_1_em
I0510 16:30:45.164729 39557 net.cpp:367] conv3_1_em/scale -> conv3_1_em (in-place)
I0510 16:30:45.164775 39557 layer_factory.hpp:77] Creating layer conv3_1_em/scale
I0510 16:30:45.164896 39557 net.cpp:122] Setting up conv3_1_em/scale
I0510 16:30:45.164904 39557 net.cpp:129] Top shape: 64 128 14 12 (1376256)
I0510 16:30:45.164907 39557 net.cpp:137] Memory required for data: 2155217664
I0510 16:30:45.164912 39557 layer_factory.hpp:77] Creating layer res3_1
I0510 16:30:45.164921 39557 net.cpp:84] Creating Layer res3_1
I0510 16:30:45.164924 39557 net.cpp:406] res3_1 <- conv3_em_conv3_em/scale_0_split_1
I0510 16:30:45.164929 39557 net.cpp:406] res3_1 <- conv3_1_em
I0510 16:30:45.164935 39557 net.cpp:380] res3_1 -> res3_1
I0510 16:30:45.164963 39557 net.cpp:122] Setting up res3_1
I0510 16:30:45.164969 39557 net.cpp:129] Top shape: 64 128 14 12 (1376256)
I0510 16:30:45.164973 39557 net.cpp:137] Memory required for data: 2160722688
I0510 16:30:45.164975 39557 layer_factory.hpp:77] Creating layer res3_1_res3_1_0_split
I0510 16:30:45.164985 39557 net.cpp:84] Creating Layer res3_1_res3_1_0_split
I0510 16:30:45.164989 39557 net.cpp:406] res3_1_res3_1_0_split <- res3_1
I0510 16:30:45.164994 39557 net.cpp:380] res3_1_res3_1_0_split -> res3_1_res3_1_0_split_0
I0510 16:30:45.165001 39557 net.cpp:380] res3_1_res3_1_0_split -> res3_1_res3_1_0_split_1
I0510 16:30:45.165040 39557 net.cpp:122] Setting up res3_1_res3_1_0_split
I0510 16:30:45.165048 39557 net.cpp:129] Top shape: 64 128 14 12 (1376256)
I0510 16:30:45.165052 39557 net.cpp:129] Top shape: 64 128 14 12 (1376256)
I0510 16:30:45.165055 39557 net.cpp:137] Memory required for data: 2171732736
I0510 16:30:45.165058 39557 layer_factory.hpp:77] Creating layer conv3_2_ex
I0510 16:30:45.165066 39557 net.cpp:84] Creating Layer conv3_2_ex
I0510 16:30:45.165069 39557 net.cpp:406] conv3_2_ex <- res3_1_res3_1_0_split_0
I0510 16:30:45.165077 39557 net.cpp:380] conv3_2_ex -> conv3_2_ex
I0510 16:30:45.166404 39557 net.cpp:122] Setting up conv3_2_ex
I0510 16:30:45.166429 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.166434 39557 net.cpp:137] Memory required for data: 2182742784
I0510 16:30:45.166440 39557 layer_factory.hpp:77] Creating layer conv3_2_ex/bn
I0510 16:30:45.166450 39557 net.cpp:84] Creating Layer conv3_2_ex/bn
I0510 16:30:45.166455 39557 net.cpp:406] conv3_2_ex/bn <- conv3_2_ex
I0510 16:30:45.166461 39557 net.cpp:367] conv3_2_ex/bn -> conv3_2_ex (in-place)
I0510 16:30:45.166668 39557 net.cpp:122] Setting up conv3_2_ex/bn
I0510 16:30:45.166677 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.166682 39557 net.cpp:137] Memory required for data: 2193752832
I0510 16:30:45.166688 39557 layer_factory.hpp:77] Creating layer conv3_2_ex/scale
I0510 16:30:45.166698 39557 net.cpp:84] Creating Layer conv3_2_ex/scale
I0510 16:30:45.166702 39557 net.cpp:406] conv3_2_ex/scale <- conv3_2_ex
I0510 16:30:45.166707 39557 net.cpp:367] conv3_2_ex/scale -> conv3_2_ex (in-place)
I0510 16:30:45.166755 39557 layer_factory.hpp:77] Creating layer conv3_2_ex/scale
I0510 16:30:45.166874 39557 net.cpp:122] Setting up conv3_2_ex/scale
I0510 16:30:45.166883 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.166887 39557 net.cpp:137] Memory required for data: 2204762880
I0510 16:30:45.166903 39557 layer_factory.hpp:77] Creating layer relu3_2_ex
I0510 16:30:45.166909 39557 net.cpp:84] Creating Layer relu3_2_ex
I0510 16:30:45.166913 39557 net.cpp:406] relu3_2_ex <- conv3_2_ex
I0510 16:30:45.166919 39557 net.cpp:367] relu3_2_ex -> conv3_2_ex (in-place)
I0510 16:30:45.167034 39557 net.cpp:122] Setting up relu3_2_ex
I0510 16:30:45.167043 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.167047 39557 net.cpp:137] Memory required for data: 2215772928
I0510 16:30:45.167052 39557 layer_factory.hpp:77] Creating layer conv3_2_dw
I0510 16:30:45.167060 39557 net.cpp:84] Creating Layer conv3_2_dw
I0510 16:30:45.167063 39557 net.cpp:406] conv3_2_dw <- conv3_2_ex
I0510 16:30:45.167069 39557 net.cpp:380] conv3_2_dw -> conv3_2_dw
I0510 16:30:45.167281 39557 net.cpp:122] Setting up conv3_2_dw
I0510 16:30:45.167290 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.167294 39557 net.cpp:137] Memory required for data: 2226782976
I0510 16:30:45.167299 39557 layer_factory.hpp:77] Creating layer conv3_2_dw/bn
I0510 16:30:45.167306 39557 net.cpp:84] Creating Layer conv3_2_dw/bn
I0510 16:30:45.167310 39557 net.cpp:406] conv3_2_dw/bn <- conv3_2_dw
I0510 16:30:45.167315 39557 net.cpp:367] conv3_2_dw/bn -> conv3_2_dw (in-place)
I0510 16:30:45.167522 39557 net.cpp:122] Setting up conv3_2_dw/bn
I0510 16:30:45.167531 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.167534 39557 net.cpp:137] Memory required for data: 2237793024
I0510 16:30:45.167543 39557 layer_factory.hpp:77] Creating layer conv3_2_dw/scale
I0510 16:30:45.167551 39557 net.cpp:84] Creating Layer conv3_2_dw/scale
I0510 16:30:45.167553 39557 net.cpp:406] conv3_2_dw/scale <- conv3_2_dw
I0510 16:30:45.167560 39557 net.cpp:367] conv3_2_dw/scale -> conv3_2_dw (in-place)
I0510 16:30:45.167605 39557 layer_factory.hpp:77] Creating layer conv3_2_dw/scale
I0510 16:30:45.167726 39557 net.cpp:122] Setting up conv3_2_dw/scale
I0510 16:30:45.167734 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.167737 39557 net.cpp:137] Memory required for data: 2248803072
I0510 16:30:45.167743 39557 layer_factory.hpp:77] Creating layer relu3_2_dw
I0510 16:30:45.167750 39557 net.cpp:84] Creating Layer relu3_2_dw
I0510 16:30:45.167753 39557 net.cpp:406] relu3_2_dw <- conv3_2_dw
I0510 16:30:45.167758 39557 net.cpp:367] relu3_2_dw -> conv3_2_dw (in-place)
I0510 16:30:45.167881 39557 net.cpp:122] Setting up relu3_2_dw
I0510 16:30:45.167891 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.167893 39557 net.cpp:137] Memory required for data: 2259813120
I0510 16:30:45.167897 39557 layer_factory.hpp:77] Creating layer conv3_2_em
I0510 16:30:45.167907 39557 net.cpp:84] Creating Layer conv3_2_em
I0510 16:30:45.167910 39557 net.cpp:406] conv3_2_em <- conv3_2_dw
I0510 16:30:45.167917 39557 net.cpp:380] conv3_2_em -> conv3_2_em
I0510 16:30:45.169239 39557 net.cpp:122] Setting up conv3_2_em
I0510 16:30:45.169263 39557 net.cpp:129] Top shape: 64 128 14 12 (1376256)
I0510 16:30:45.169266 39557 net.cpp:137] Memory required for data: 2265318144
I0510 16:30:45.169275 39557 layer_factory.hpp:77] Creating layer conv3_2_em/bn
I0510 16:30:45.169284 39557 net.cpp:84] Creating Layer conv3_2_em/bn
I0510 16:30:45.169288 39557 net.cpp:406] conv3_2_em/bn <- conv3_2_em
I0510 16:30:45.169293 39557 net.cpp:367] conv3_2_em/bn -> conv3_2_em (in-place)
I0510 16:30:45.169497 39557 net.cpp:122] Setting up conv3_2_em/bn
I0510 16:30:45.169505 39557 net.cpp:129] Top shape: 64 128 14 12 (1376256)
I0510 16:30:45.169508 39557 net.cpp:137] Memory required for data: 2270823168
I0510 16:30:45.169517 39557 layer_factory.hpp:77] Creating layer conv3_2_em/scale
I0510 16:30:45.169523 39557 net.cpp:84] Creating Layer conv3_2_em/scale
I0510 16:30:45.169525 39557 net.cpp:406] conv3_2_em/scale <- conv3_2_em
I0510 16:30:45.169530 39557 net.cpp:367] conv3_2_em/scale -> conv3_2_em (in-place)
I0510 16:30:45.169577 39557 layer_factory.hpp:77] Creating layer conv3_2_em/scale
I0510 16:30:45.169693 39557 net.cpp:122] Setting up conv3_2_em/scale
I0510 16:30:45.169716 39557 net.cpp:129] Top shape: 64 128 14 12 (1376256)
I0510 16:30:45.169720 39557 net.cpp:137] Memory required for data: 2276328192
I0510 16:30:45.169726 39557 layer_factory.hpp:77] Creating layer res3_2
I0510 16:30:45.169735 39557 net.cpp:84] Creating Layer res3_2
I0510 16:30:45.169739 39557 net.cpp:406] res3_2 <- res3_1_res3_1_0_split_1
I0510 16:30:45.169744 39557 net.cpp:406] res3_2 <- conv3_2_em
I0510 16:30:45.169749 39557 net.cpp:380] res3_2 -> res3_2
I0510 16:30:45.169780 39557 net.cpp:122] Setting up res3_2
I0510 16:30:45.169788 39557 net.cpp:129] Top shape: 64 128 14 12 (1376256)
I0510 16:30:45.169791 39557 net.cpp:137] Memory required for data: 2281833216
I0510 16:30:45.169795 39557 layer_factory.hpp:77] Creating layer res3_2_res3_2_0_split
I0510 16:30:45.169800 39557 net.cpp:84] Creating Layer res3_2_res3_2_0_split
I0510 16:30:45.169803 39557 net.cpp:406] res3_2_res3_2_0_split <- res3_2
I0510 16:30:45.169808 39557 net.cpp:380] res3_2_res3_2_0_split -> res3_2_res3_2_0_split_0
I0510 16:30:45.169816 39557 net.cpp:380] res3_2_res3_2_0_split -> res3_2_res3_2_0_split_1
I0510 16:30:45.169853 39557 net.cpp:122] Setting up res3_2_res3_2_0_split
I0510 16:30:45.169862 39557 net.cpp:129] Top shape: 64 128 14 12 (1376256)
I0510 16:30:45.169865 39557 net.cpp:129] Top shape: 64 128 14 12 (1376256)
I0510 16:30:45.169867 39557 net.cpp:137] Memory required for data: 2292843264
I0510 16:30:45.169870 39557 layer_factory.hpp:77] Creating layer conv3_3_ex
I0510 16:30:45.169880 39557 net.cpp:84] Creating Layer conv3_3_ex
I0510 16:30:45.169884 39557 net.cpp:406] conv3_3_ex <- res3_2_res3_2_0_split_0
I0510 16:30:45.169890 39557 net.cpp:380] conv3_3_ex -> conv3_3_ex
I0510 16:30:45.171545 39557 net.cpp:122] Setting up conv3_3_ex
I0510 16:30:45.171561 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.171564 39557 net.cpp:137] Memory required for data: 2303853312
I0510 16:30:45.171571 39557 layer_factory.hpp:77] Creating layer conv3_3_ex/bn
I0510 16:30:45.171581 39557 net.cpp:84] Creating Layer conv3_3_ex/bn
I0510 16:30:45.171586 39557 net.cpp:406] conv3_3_ex/bn <- conv3_3_ex
I0510 16:30:45.171593 39557 net.cpp:367] conv3_3_ex/bn -> conv3_3_ex (in-place)
I0510 16:30:45.171808 39557 net.cpp:122] Setting up conv3_3_ex/bn
I0510 16:30:45.171816 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.171819 39557 net.cpp:137] Memory required for data: 2314863360
I0510 16:30:45.171828 39557 layer_factory.hpp:77] Creating layer conv3_3_ex/scale
I0510 16:30:45.171835 39557 net.cpp:84] Creating Layer conv3_3_ex/scale
I0510 16:30:45.171839 39557 net.cpp:406] conv3_3_ex/scale <- conv3_3_ex
I0510 16:30:45.171844 39557 net.cpp:367] conv3_3_ex/scale -> conv3_3_ex (in-place)
I0510 16:30:45.171891 39557 layer_factory.hpp:77] Creating layer conv3_3_ex/scale
I0510 16:30:45.172019 39557 net.cpp:122] Setting up conv3_3_ex/scale
I0510 16:30:45.172027 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.172030 39557 net.cpp:137] Memory required for data: 2325873408
I0510 16:30:45.172036 39557 layer_factory.hpp:77] Creating layer relu3_3_ex
I0510 16:30:45.172041 39557 net.cpp:84] Creating Layer relu3_3_ex
I0510 16:30:45.172044 39557 net.cpp:406] relu3_3_ex <- conv3_3_ex
I0510 16:30:45.172051 39557 net.cpp:367] relu3_3_ex -> conv3_3_ex (in-place)
I0510 16:30:45.172165 39557 net.cpp:122] Setting up relu3_3_ex
I0510 16:30:45.172175 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.172178 39557 net.cpp:137] Memory required for data: 2336883456
I0510 16:30:45.172183 39557 layer_factory.hpp:77] Creating layer conv3_3_dw
I0510 16:30:45.172194 39557 net.cpp:84] Creating Layer conv3_3_dw
I0510 16:30:45.172196 39557 net.cpp:406] conv3_3_dw <- conv3_3_ex
I0510 16:30:45.172202 39557 net.cpp:380] conv3_3_dw -> conv3_3_dw
I0510 16:30:45.172423 39557 net.cpp:122] Setting up conv3_3_dw
I0510 16:30:45.172433 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.172436 39557 net.cpp:137] Memory required for data: 2347893504
I0510 16:30:45.172441 39557 layer_factory.hpp:77] Creating layer conv3_3_dw/bn
I0510 16:30:45.172458 39557 net.cpp:84] Creating Layer conv3_3_dw/bn
I0510 16:30:45.172462 39557 net.cpp:406] conv3_3_dw/bn <- conv3_3_dw
I0510 16:30:45.172469 39557 net.cpp:367] conv3_3_dw/bn -> conv3_3_dw (in-place)
I0510 16:30:45.172678 39557 net.cpp:122] Setting up conv3_3_dw/bn
I0510 16:30:45.172688 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.172690 39557 net.cpp:137] Memory required for data: 2358903552
I0510 16:30:45.172698 39557 layer_factory.hpp:77] Creating layer conv3_3_dw/scale
I0510 16:30:45.172703 39557 net.cpp:84] Creating Layer conv3_3_dw/scale
I0510 16:30:45.172708 39557 net.cpp:406] conv3_3_dw/scale <- conv3_3_dw
I0510 16:30:45.172714 39557 net.cpp:367] conv3_3_dw/scale -> conv3_3_dw (in-place)
I0510 16:30:45.172757 39557 layer_factory.hpp:77] Creating layer conv3_3_dw/scale
I0510 16:30:45.172879 39557 net.cpp:122] Setting up conv3_3_dw/scale
I0510 16:30:45.172888 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.172891 39557 net.cpp:137] Memory required for data: 2369913600
I0510 16:30:45.172896 39557 layer_factory.hpp:77] Creating layer relu3_3_dw
I0510 16:30:45.172902 39557 net.cpp:84] Creating Layer relu3_3_dw
I0510 16:30:45.172904 39557 net.cpp:406] relu3_3_dw <- conv3_3_dw
I0510 16:30:45.172911 39557 net.cpp:367] relu3_3_dw -> conv3_3_dw (in-place)
I0510 16:30:45.173027 39557 net.cpp:122] Setting up relu3_3_dw
I0510 16:30:45.173036 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.173039 39557 net.cpp:137] Memory required for data: 2380923648
I0510 16:30:45.173045 39557 layer_factory.hpp:77] Creating layer conv3_3_em
I0510 16:30:45.173055 39557 net.cpp:84] Creating Layer conv3_3_em
I0510 16:30:45.173059 39557 net.cpp:406] conv3_3_em <- conv3_3_dw
I0510 16:30:45.173065 39557 net.cpp:380] conv3_3_em -> conv3_3_em
I0510 16:30:45.175480 39557 net.cpp:122] Setting up conv3_3_em
I0510 16:30:45.175494 39557 net.cpp:129] Top shape: 64 128 14 12 (1376256)
I0510 16:30:45.175508 39557 net.cpp:137] Memory required for data: 2386428672
I0510 16:30:45.175513 39557 layer_factory.hpp:77] Creating layer conv3_3_em/bn
I0510 16:30:45.175523 39557 net.cpp:84] Creating Layer conv3_3_em/bn
I0510 16:30:45.175526 39557 net.cpp:406] conv3_3_em/bn <- conv3_3_em
I0510 16:30:45.175534 39557 net.cpp:367] conv3_3_em/bn -> conv3_3_em (in-place)
I0510 16:30:45.175732 39557 net.cpp:122] Setting up conv3_3_em/bn
I0510 16:30:45.175741 39557 net.cpp:129] Top shape: 64 128 14 12 (1376256)
I0510 16:30:45.175745 39557 net.cpp:137] Memory required for data: 2391933696
I0510 16:30:45.175751 39557 layer_factory.hpp:77] Creating layer conv3_3_em/scale
I0510 16:30:45.175762 39557 net.cpp:84] Creating Layer conv3_3_em/scale
I0510 16:30:45.175766 39557 net.cpp:406] conv3_3_em/scale <- conv3_3_em
I0510 16:30:45.175771 39557 net.cpp:367] conv3_3_em/scale -> conv3_3_em (in-place)
I0510 16:30:45.175815 39557 layer_factory.hpp:77] Creating layer conv3_3_em/scale
I0510 16:30:45.175935 39557 net.cpp:122] Setting up conv3_3_em/scale
I0510 16:30:45.175942 39557 net.cpp:129] Top shape: 64 128 14 12 (1376256)
I0510 16:30:45.175945 39557 net.cpp:137] Memory required for data: 2397438720
I0510 16:30:45.175951 39557 layer_factory.hpp:77] Creating layer res3_3
I0510 16:30:45.175958 39557 net.cpp:84] Creating Layer res3_3
I0510 16:30:45.175961 39557 net.cpp:406] res3_3 <- res3_2_res3_2_0_split_1
I0510 16:30:45.175966 39557 net.cpp:406] res3_3 <- conv3_3_em
I0510 16:30:45.175971 39557 net.cpp:380] res3_3 -> res3_3
I0510 16:30:45.175998 39557 net.cpp:122] Setting up res3_3
I0510 16:30:45.176007 39557 net.cpp:129] Top shape: 64 128 14 12 (1376256)
I0510 16:30:45.176009 39557 net.cpp:137] Memory required for data: 2402943744
I0510 16:30:45.176012 39557 layer_factory.hpp:77] Creating layer res3_3_res3_3_0_split
I0510 16:30:45.176017 39557 net.cpp:84] Creating Layer res3_3_res3_3_0_split
I0510 16:30:45.176019 39557 net.cpp:406] res3_3_res3_3_0_split <- res3_3
I0510 16:30:45.176026 39557 net.cpp:380] res3_3_res3_3_0_split -> res3_3_res3_3_0_split_0
I0510 16:30:45.176033 39557 net.cpp:380] res3_3_res3_3_0_split -> res3_3_res3_3_0_split_1
I0510 16:30:45.176085 39557 net.cpp:122] Setting up res3_3_res3_3_0_split
I0510 16:30:45.176095 39557 net.cpp:129] Top shape: 64 128 14 12 (1376256)
I0510 16:30:45.176098 39557 net.cpp:129] Top shape: 64 128 14 12 (1376256)
I0510 16:30:45.176100 39557 net.cpp:137] Memory required for data: 2413953792
I0510 16:30:45.176103 39557 layer_factory.hpp:77] Creating layer conv3_4_ex
I0510 16:30:45.176113 39557 net.cpp:84] Creating Layer conv3_4_ex
I0510 16:30:45.176117 39557 net.cpp:406] conv3_4_ex <- res3_3_res3_3_0_split_0
I0510 16:30:45.176127 39557 net.cpp:380] conv3_4_ex -> conv3_4_ex
I0510 16:30:45.177351 39557 net.cpp:122] Setting up conv3_4_ex
I0510 16:30:45.177366 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.177378 39557 net.cpp:137] Memory required for data: 2424963840
I0510 16:30:45.177383 39557 layer_factory.hpp:77] Creating layer conv3_4_ex/bn
I0510 16:30:45.177392 39557 net.cpp:84] Creating Layer conv3_4_ex/bn
I0510 16:30:45.177397 39557 net.cpp:406] conv3_4_ex/bn <- conv3_4_ex
I0510 16:30:45.177415 39557 net.cpp:367] conv3_4_ex/bn -> conv3_4_ex (in-place)
I0510 16:30:45.177624 39557 net.cpp:122] Setting up conv3_4_ex/bn
I0510 16:30:45.177633 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.177636 39557 net.cpp:137] Memory required for data: 2435973888
I0510 16:30:45.177644 39557 layer_factory.hpp:77] Creating layer conv3_4_ex/scale
I0510 16:30:45.177651 39557 net.cpp:84] Creating Layer conv3_4_ex/scale
I0510 16:30:45.177655 39557 net.cpp:406] conv3_4_ex/scale <- conv3_4_ex
I0510 16:30:45.177660 39557 net.cpp:367] conv3_4_ex/scale -> conv3_4_ex (in-place)
I0510 16:30:45.177706 39557 layer_factory.hpp:77] Creating layer conv3_4_ex/scale
I0510 16:30:45.177827 39557 net.cpp:122] Setting up conv3_4_ex/scale
I0510 16:30:45.177835 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.177839 39557 net.cpp:137] Memory required for data: 2446983936
I0510 16:30:45.177844 39557 layer_factory.hpp:77] Creating layer relu3_4_ex
I0510 16:30:45.177851 39557 net.cpp:84] Creating Layer relu3_4_ex
I0510 16:30:45.177855 39557 net.cpp:406] relu3_4_ex <- conv3_4_ex
I0510 16:30:45.177861 39557 net.cpp:367] relu3_4_ex -> conv3_4_ex (in-place)
I0510 16:30:45.177973 39557 net.cpp:122] Setting up relu3_4_ex
I0510 16:30:45.177983 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.177985 39557 net.cpp:137] Memory required for data: 2457993984
I0510 16:30:45.177992 39557 layer_factory.hpp:77] Creating layer conv3_4_dw
I0510 16:30:45.178007 39557 net.cpp:84] Creating Layer conv3_4_dw
I0510 16:30:45.178012 39557 net.cpp:406] conv3_4_dw <- conv3_4_ex
I0510 16:30:45.178019 39557 net.cpp:380] conv3_4_dw -> conv3_4_dw
I0510 16:30:45.178233 39557 net.cpp:122] Setting up conv3_4_dw
I0510 16:30:45.178243 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.178246 39557 net.cpp:137] Memory required for data: 2469004032
I0510 16:30:45.178251 39557 layer_factory.hpp:77] Creating layer conv3_4_dw/bn
I0510 16:30:45.178259 39557 net.cpp:84] Creating Layer conv3_4_dw/bn
I0510 16:30:45.178263 39557 net.cpp:406] conv3_4_dw/bn <- conv3_4_dw
I0510 16:30:45.178269 39557 net.cpp:367] conv3_4_dw/bn -> conv3_4_dw (in-place)
I0510 16:30:45.178472 39557 net.cpp:122] Setting up conv3_4_dw/bn
I0510 16:30:45.178481 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.178484 39557 net.cpp:137] Memory required for data: 2480014080
I0510 16:30:45.178491 39557 layer_factory.hpp:77] Creating layer conv3_4_dw/scale
I0510 16:30:45.178500 39557 net.cpp:84] Creating Layer conv3_4_dw/scale
I0510 16:30:45.178504 39557 net.cpp:406] conv3_4_dw/scale <- conv3_4_dw
I0510 16:30:45.178508 39557 net.cpp:367] conv3_4_dw/scale -> conv3_4_dw (in-place)
I0510 16:30:45.178555 39557 layer_factory.hpp:77] Creating layer conv3_4_dw/scale
I0510 16:30:45.178674 39557 net.cpp:122] Setting up conv3_4_dw/scale
I0510 16:30:45.178683 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.178685 39557 net.cpp:137] Memory required for data: 2491024128
I0510 16:30:45.178701 39557 layer_factory.hpp:77] Creating layer relu3_4_dw
I0510 16:30:45.178707 39557 net.cpp:84] Creating Layer relu3_4_dw
I0510 16:30:45.178711 39557 net.cpp:406] relu3_4_dw <- conv3_4_dw
I0510 16:30:45.178717 39557 net.cpp:367] relu3_4_dw -> conv3_4_dw (in-place)
I0510 16:30:45.178831 39557 net.cpp:122] Setting up relu3_4_dw
I0510 16:30:45.178840 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.178843 39557 net.cpp:137] Memory required for data: 2502034176
I0510 16:30:45.178848 39557 layer_factory.hpp:77] Creating layer conv3_4_em
I0510 16:30:45.178858 39557 net.cpp:84] Creating Layer conv3_4_em
I0510 16:30:45.178861 39557 net.cpp:406] conv3_4_em <- conv3_4_dw
I0510 16:30:45.178866 39557 net.cpp:380] conv3_4_em -> conv3_4_em
I0510 16:30:45.180209 39557 net.cpp:122] Setting up conv3_4_em
I0510 16:30:45.180223 39557 net.cpp:129] Top shape: 64 128 14 12 (1376256)
I0510 16:30:45.180238 39557 net.cpp:137] Memory required for data: 2507539200
I0510 16:30:45.180243 39557 layer_factory.hpp:77] Creating layer conv3_4_em/bn
I0510 16:30:45.180251 39557 net.cpp:84] Creating Layer conv3_4_em/bn
I0510 16:30:45.180255 39557 net.cpp:406] conv3_4_em/bn <- conv3_4_em
I0510 16:30:45.180263 39557 net.cpp:367] conv3_4_em/bn -> conv3_4_em (in-place)
I0510 16:30:45.180474 39557 net.cpp:122] Setting up conv3_4_em/bn
I0510 16:30:45.180482 39557 net.cpp:129] Top shape: 64 128 14 12 (1376256)
I0510 16:30:45.180485 39557 net.cpp:137] Memory required for data: 2513044224
I0510 16:30:45.180492 39557 layer_factory.hpp:77] Creating layer conv3_4_em/scale
I0510 16:30:45.180498 39557 net.cpp:84] Creating Layer conv3_4_em/scale
I0510 16:30:45.180501 39557 net.cpp:406] conv3_4_em/scale <- conv3_4_em
I0510 16:30:45.180508 39557 net.cpp:367] conv3_4_em/scale -> conv3_4_em (in-place)
I0510 16:30:45.180557 39557 layer_factory.hpp:77] Creating layer conv3_4_em/scale
I0510 16:30:45.180678 39557 net.cpp:122] Setting up conv3_4_em/scale
I0510 16:30:45.180686 39557 net.cpp:129] Top shape: 64 128 14 12 (1376256)
I0510 16:30:45.180690 39557 net.cpp:137] Memory required for data: 2518549248
I0510 16:30:45.180696 39557 layer_factory.hpp:77] Creating layer res3_4
I0510 16:30:45.180701 39557 net.cpp:84] Creating Layer res3_4
I0510 16:30:45.180704 39557 net.cpp:406] res3_4 <- res3_3_res3_3_0_split_1
I0510 16:30:45.180708 39557 net.cpp:406] res3_4 <- conv3_4_em
I0510 16:30:45.180716 39557 net.cpp:380] res3_4 -> res3_4
I0510 16:30:45.180744 39557 net.cpp:122] Setting up res3_4
I0510 16:30:45.180752 39557 net.cpp:129] Top shape: 64 128 14 12 (1376256)
I0510 16:30:45.180754 39557 net.cpp:137] Memory required for data: 2524054272
I0510 16:30:45.180757 39557 layer_factory.hpp:77] Creating layer res3_4_res3_4_0_split
I0510 16:30:45.180764 39557 net.cpp:84] Creating Layer res3_4_res3_4_0_split
I0510 16:30:45.180768 39557 net.cpp:406] res3_4_res3_4_0_split <- res3_4
I0510 16:30:45.180775 39557 net.cpp:380] res3_4_res3_4_0_split -> res3_4_res3_4_0_split_0
I0510 16:30:45.180781 39557 net.cpp:380] res3_4_res3_4_0_split -> res3_4_res3_4_0_split_1
I0510 16:30:45.180822 39557 net.cpp:122] Setting up res3_4_res3_4_0_split
I0510 16:30:45.180830 39557 net.cpp:129] Top shape: 64 128 14 12 (1376256)
I0510 16:30:45.180835 39557 net.cpp:129] Top shape: 64 128 14 12 (1376256)
I0510 16:30:45.180836 39557 net.cpp:137] Memory required for data: 2535064320
I0510 16:30:45.180840 39557 layer_factory.hpp:77] Creating layer conv3_5_ex
I0510 16:30:45.180847 39557 net.cpp:84] Creating Layer conv3_5_ex
I0510 16:30:45.180850 39557 net.cpp:406] conv3_5_ex <- res3_4_res3_4_0_split_0
I0510 16:30:45.180860 39557 net.cpp:380] conv3_5_ex -> conv3_5_ex
I0510 16:30:45.182209 39557 net.cpp:122] Setting up conv3_5_ex
I0510 16:30:45.182224 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.182226 39557 net.cpp:137] Memory required for data: 2546074368
I0510 16:30:45.182232 39557 layer_factory.hpp:77] Creating layer conv3_5_ex/bn
I0510 16:30:45.182253 39557 net.cpp:84] Creating Layer conv3_5_ex/bn
I0510 16:30:45.182258 39557 net.cpp:406] conv3_5_ex/bn <- conv3_5_ex
I0510 16:30:45.182278 39557 net.cpp:367] conv3_5_ex/bn -> conv3_5_ex (in-place)
I0510 16:30:45.182499 39557 net.cpp:122] Setting up conv3_5_ex/bn
I0510 16:30:45.182509 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.182525 39557 net.cpp:137] Memory required for data: 2557084416
I0510 16:30:45.182533 39557 layer_factory.hpp:77] Creating layer conv3_5_ex/scale
I0510 16:30:45.182540 39557 net.cpp:84] Creating Layer conv3_5_ex/scale
I0510 16:30:45.182544 39557 net.cpp:406] conv3_5_ex/scale <- conv3_5_ex
I0510 16:30:45.182549 39557 net.cpp:367] conv3_5_ex/scale -> conv3_5_ex (in-place)
I0510 16:30:45.182595 39557 layer_factory.hpp:77] Creating layer conv3_5_ex/scale
I0510 16:30:45.182718 39557 net.cpp:122] Setting up conv3_5_ex/scale
I0510 16:30:45.182725 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.182729 39557 net.cpp:137] Memory required for data: 2568094464
I0510 16:30:45.182736 39557 layer_factory.hpp:77] Creating layer relu3_5_ex
I0510 16:30:45.182744 39557 net.cpp:84] Creating Layer relu3_5_ex
I0510 16:30:45.182747 39557 net.cpp:406] relu3_5_ex <- conv3_5_ex
I0510 16:30:45.182751 39557 net.cpp:367] relu3_5_ex -> conv3_5_ex (in-place)
I0510 16:30:45.182868 39557 net.cpp:122] Setting up relu3_5_ex
I0510 16:30:45.182878 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.182880 39557 net.cpp:137] Memory required for data: 2579104512
I0510 16:30:45.182884 39557 layer_factory.hpp:77] Creating layer conv3_5_dw
I0510 16:30:45.182893 39557 net.cpp:84] Creating Layer conv3_5_dw
I0510 16:30:45.182896 39557 net.cpp:406] conv3_5_dw <- conv3_5_ex
I0510 16:30:45.182902 39557 net.cpp:380] conv3_5_dw -> conv3_5_dw
I0510 16:30:45.183132 39557 net.cpp:122] Setting up conv3_5_dw
I0510 16:30:45.183142 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.183145 39557 net.cpp:137] Memory required for data: 2590114560
I0510 16:30:45.183151 39557 layer_factory.hpp:77] Creating layer conv3_5_dw/bn
I0510 16:30:45.183158 39557 net.cpp:84] Creating Layer conv3_5_dw/bn
I0510 16:30:45.183161 39557 net.cpp:406] conv3_5_dw/bn <- conv3_5_dw
I0510 16:30:45.183168 39557 net.cpp:367] conv3_5_dw/bn -> conv3_5_dw (in-place)
I0510 16:30:45.183404 39557 net.cpp:122] Setting up conv3_5_dw/bn
I0510 16:30:45.183414 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.183418 39557 net.cpp:137] Memory required for data: 2601124608
I0510 16:30:45.183424 39557 layer_factory.hpp:77] Creating layer conv3_5_dw/scale
I0510 16:30:45.183430 39557 net.cpp:84] Creating Layer conv3_5_dw/scale
I0510 16:30:45.183434 39557 net.cpp:406] conv3_5_dw/scale <- conv3_5_dw
I0510 16:30:45.183440 39557 net.cpp:367] conv3_5_dw/scale -> conv3_5_dw (in-place)
I0510 16:30:45.183483 39557 layer_factory.hpp:77] Creating layer conv3_5_dw/scale
I0510 16:30:45.183601 39557 net.cpp:122] Setting up conv3_5_dw/scale
I0510 16:30:45.183611 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.183614 39557 net.cpp:137] Memory required for data: 2612134656
I0510 16:30:45.183620 39557 layer_factory.hpp:77] Creating layer relu3_5_dw
I0510 16:30:45.183625 39557 net.cpp:84] Creating Layer relu3_5_dw
I0510 16:30:45.183629 39557 net.cpp:406] relu3_5_dw <- conv3_5_dw
I0510 16:30:45.183632 39557 net.cpp:367] relu3_5_dw -> conv3_5_dw (in-place)
I0510 16:30:45.183745 39557 net.cpp:122] Setting up relu3_5_dw
I0510 16:30:45.183754 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.183758 39557 net.cpp:137] Memory required for data: 2623144704
I0510 16:30:45.183761 39557 layer_factory.hpp:77] Creating layer conv3_5_em
I0510 16:30:45.183771 39557 net.cpp:84] Creating Layer conv3_5_em
I0510 16:30:45.183774 39557 net.cpp:406] conv3_5_em <- conv3_5_dw
I0510 16:30:45.183781 39557 net.cpp:380] conv3_5_em -> conv3_5_em
I0510 16:30:45.185156 39557 net.cpp:122] Setting up conv3_5_em
I0510 16:30:45.185179 39557 net.cpp:129] Top shape: 64 128 14 12 (1376256)
I0510 16:30:45.185184 39557 net.cpp:137] Memory required for data: 2628649728
I0510 16:30:45.185189 39557 layer_factory.hpp:77] Creating layer conv3_5_em/bn
I0510 16:30:45.185212 39557 net.cpp:84] Creating Layer conv3_5_em/bn
I0510 16:30:45.185217 39557 net.cpp:406] conv3_5_em/bn <- conv3_5_em
I0510 16:30:45.185225 39557 net.cpp:367] conv3_5_em/bn -> conv3_5_em (in-place)
I0510 16:30:45.185432 39557 net.cpp:122] Setting up conv3_5_em/bn
I0510 16:30:45.185441 39557 net.cpp:129] Top shape: 64 128 14 12 (1376256)
I0510 16:30:45.185444 39557 net.cpp:137] Memory required for data: 2634154752
I0510 16:30:45.185451 39557 layer_factory.hpp:77] Creating layer conv3_5_em/scale
I0510 16:30:45.185461 39557 net.cpp:84] Creating Layer conv3_5_em/scale
I0510 16:30:45.185463 39557 net.cpp:406] conv3_5_em/scale <- conv3_5_em
I0510 16:30:45.185468 39557 net.cpp:367] conv3_5_em/scale -> conv3_5_em (in-place)
I0510 16:30:45.185514 39557 layer_factory.hpp:77] Creating layer conv3_5_em/scale
I0510 16:30:45.185637 39557 net.cpp:122] Setting up conv3_5_em/scale
I0510 16:30:45.185645 39557 net.cpp:129] Top shape: 64 128 14 12 (1376256)
I0510 16:30:45.185648 39557 net.cpp:137] Memory required for data: 2639659776
I0510 16:30:45.185653 39557 layer_factory.hpp:77] Creating layer res3_5
I0510 16:30:45.185663 39557 net.cpp:84] Creating Layer res3_5
I0510 16:30:45.185667 39557 net.cpp:406] res3_5 <- res3_4_res3_4_0_split_1
I0510 16:30:45.185672 39557 net.cpp:406] res3_5 <- conv3_5_em
I0510 16:30:45.185676 39557 net.cpp:380] res3_5 -> res3_5
I0510 16:30:45.185705 39557 net.cpp:122] Setting up res3_5
I0510 16:30:45.185714 39557 net.cpp:129] Top shape: 64 128 14 12 (1376256)
I0510 16:30:45.185716 39557 net.cpp:137] Memory required for data: 2645164800
I0510 16:30:45.185719 39557 layer_factory.hpp:77] Creating layer res3_5_res3_5_0_split
I0510 16:30:45.185724 39557 net.cpp:84] Creating Layer res3_5_res3_5_0_split
I0510 16:30:45.185727 39557 net.cpp:406] res3_5_res3_5_0_split <- res3_5
I0510 16:30:45.185735 39557 net.cpp:380] res3_5_res3_5_0_split -> res3_5_res3_5_0_split_0
I0510 16:30:45.185742 39557 net.cpp:380] res3_5_res3_5_0_split -> res3_5_res3_5_0_split_1
I0510 16:30:45.185781 39557 net.cpp:122] Setting up res3_5_res3_5_0_split
I0510 16:30:45.185788 39557 net.cpp:129] Top shape: 64 128 14 12 (1376256)
I0510 16:30:45.185792 39557 net.cpp:129] Top shape: 64 128 14 12 (1376256)
I0510 16:30:45.185796 39557 net.cpp:137] Memory required for data: 2656174848
I0510 16:30:45.185798 39557 layer_factory.hpp:77] Creating layer conv3_6_ex
I0510 16:30:45.185806 39557 net.cpp:84] Creating Layer conv3_6_ex
I0510 16:30:45.185809 39557 net.cpp:406] conv3_6_ex <- res3_5_res3_5_0_split_0
I0510 16:30:45.185817 39557 net.cpp:380] conv3_6_ex -> conv3_6_ex
I0510 16:30:45.187150 39557 net.cpp:122] Setting up conv3_6_ex
I0510 16:30:45.187165 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.187178 39557 net.cpp:137] Memory required for data: 2667184896
I0510 16:30:45.187183 39557 layer_factory.hpp:77] Creating layer conv3_6_ex/bn
I0510 16:30:45.187196 39557 net.cpp:84] Creating Layer conv3_6_ex/bn
I0510 16:30:45.187201 39557 net.cpp:406] conv3_6_ex/bn <- conv3_6_ex
I0510 16:30:45.187207 39557 net.cpp:367] conv3_6_ex/bn -> conv3_6_ex (in-place)
I0510 16:30:45.187427 39557 net.cpp:122] Setting up conv3_6_ex/bn
I0510 16:30:45.187436 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.187439 39557 net.cpp:137] Memory required for data: 2678194944
I0510 16:30:45.187446 39557 layer_factory.hpp:77] Creating layer conv3_6_ex/scale
I0510 16:30:45.187455 39557 net.cpp:84] Creating Layer conv3_6_ex/scale
I0510 16:30:45.187458 39557 net.cpp:406] conv3_6_ex/scale <- conv3_6_ex
I0510 16:30:45.187463 39557 net.cpp:367] conv3_6_ex/scale -> conv3_6_ex (in-place)
I0510 16:30:45.187511 39557 layer_factory.hpp:77] Creating layer conv3_6_ex/scale
I0510 16:30:45.187647 39557 net.cpp:122] Setting up conv3_6_ex/scale
I0510 16:30:45.187656 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.187659 39557 net.cpp:137] Memory required for data: 2689204992
I0510 16:30:45.187664 39557 layer_factory.hpp:77] Creating layer relu3_6_ex
I0510 16:30:45.187672 39557 net.cpp:84] Creating Layer relu3_6_ex
I0510 16:30:45.187688 39557 net.cpp:406] relu3_6_ex <- conv3_6_ex
I0510 16:30:45.187696 39557 net.cpp:367] relu3_6_ex -> conv3_6_ex (in-place)
I0510 16:30:45.187815 39557 net.cpp:122] Setting up relu3_6_ex
I0510 16:30:45.187825 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.187829 39557 net.cpp:137] Memory required for data: 2700215040
I0510 16:30:45.187834 39557 layer_factory.hpp:77] Creating layer conv3_6_dw
I0510 16:30:45.187844 39557 net.cpp:84] Creating Layer conv3_6_dw
I0510 16:30:45.187847 39557 net.cpp:406] conv3_6_dw <- conv3_6_ex
I0510 16:30:45.187852 39557 net.cpp:380] conv3_6_dw -> conv3_6_dw
I0510 16:30:45.188077 39557 net.cpp:122] Setting up conv3_6_dw
I0510 16:30:45.188087 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.188091 39557 net.cpp:137] Memory required for data: 2711225088
I0510 16:30:45.188096 39557 layer_factory.hpp:77] Creating layer conv3_6_dw/bn
I0510 16:30:45.188100 39557 net.cpp:84] Creating Layer conv3_6_dw/bn
I0510 16:30:45.188103 39557 net.cpp:406] conv3_6_dw/bn <- conv3_6_dw
I0510 16:30:45.188110 39557 net.cpp:367] conv3_6_dw/bn -> conv3_6_dw (in-place)
I0510 16:30:45.188318 39557 net.cpp:122] Setting up conv3_6_dw/bn
I0510 16:30:45.188328 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.188330 39557 net.cpp:137] Memory required for data: 2722235136
I0510 16:30:45.188338 39557 layer_factory.hpp:77] Creating layer conv3_6_dw/scale
I0510 16:30:45.188344 39557 net.cpp:84] Creating Layer conv3_6_dw/scale
I0510 16:30:45.188346 39557 net.cpp:406] conv3_6_dw/scale <- conv3_6_dw
I0510 16:30:45.188354 39557 net.cpp:367] conv3_6_dw/scale -> conv3_6_dw (in-place)
I0510 16:30:45.188396 39557 layer_factory.hpp:77] Creating layer conv3_6_dw/scale
I0510 16:30:45.188520 39557 net.cpp:122] Setting up conv3_6_dw/scale
I0510 16:30:45.188529 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.188531 39557 net.cpp:137] Memory required for data: 2733245184
I0510 16:30:45.188537 39557 layer_factory.hpp:77] Creating layer relu3_6_dw
I0510 16:30:45.188544 39557 net.cpp:84] Creating Layer relu3_6_dw
I0510 16:30:45.188547 39557 net.cpp:406] relu3_6_dw <- conv3_6_dw
I0510 16:30:45.188554 39557 net.cpp:367] relu3_6_dw -> conv3_6_dw (in-place)
I0510 16:30:45.188670 39557 net.cpp:122] Setting up relu3_6_dw
I0510 16:30:45.188680 39557 net.cpp:129] Top shape: 64 256 14 12 (2752512)
I0510 16:30:45.188683 39557 net.cpp:137] Memory required for data: 2744255232
I0510 16:30:45.188688 39557 layer_factory.hpp:77] Creating layer conv3_6_em
I0510 16:30:45.188695 39557 net.cpp:84] Creating Layer conv3_6_em
I0510 16:30:45.188699 39557 net.cpp:406] conv3_6_em <- conv3_6_dw
I0510 16:30:45.188709 39557 net.cpp:380] conv3_6_em -> conv3_6_em
I0510 16:30:45.190098 39557 net.cpp:122] Setting up conv3_6_em
I0510 16:30:45.190112 39557 net.cpp:129] Top shape: 64 128 14 12 (1376256)
I0510 16:30:45.190115 39557 net.cpp:137] Memory required for data: 2749760256
I0510 16:30:45.190120 39557 layer_factory.hpp:77] Creating layer conv3_6_em/bn
I0510 16:30:45.190131 39557 net.cpp:84] Creating Layer conv3_6_em/bn
I0510 16:30:45.190135 39557 net.cpp:406] conv3_6_em/bn <- conv3_6_em
I0510 16:30:45.190140 39557 net.cpp:367] conv3_6_em/bn -> conv3_6_em (in-place)
I0510 16:30:45.190356 39557 net.cpp:122] Setting up conv3_6_em/bn
I0510 16:30:45.190366 39557 net.cpp:129] Top shape: 64 128 14 12 (1376256)
I0510 16:30:45.190369 39557 net.cpp:137] Memory required for data: 2755265280
I0510 16:30:45.190376 39557 layer_factory.hpp:77] Creating layer conv3_6_em/scale
I0510 16:30:45.190387 39557 net.cpp:84] Creating Layer conv3_6_em/scale
I0510 16:30:45.190392 39557 net.cpp:406] conv3_6_em/scale <- conv3_6_em
I0510 16:30:45.190397 39557 net.cpp:367] conv3_6_em/scale -> conv3_6_em (in-place)
I0510 16:30:45.190445 39557 layer_factory.hpp:77] Creating layer conv3_6_em/scale
I0510 16:30:45.190569 39557 net.cpp:122] Setting up conv3_6_em/scale
I0510 16:30:45.190578 39557 net.cpp:129] Top shape: 64 128 14 12 (1376256)
I0510 16:30:45.190582 39557 net.cpp:137] Memory required for data: 2760770304
I0510 16:30:45.190598 39557 layer_factory.hpp:77] Creating layer res3_6
I0510 16:30:45.190604 39557 net.cpp:84] Creating Layer res3_6
I0510 16:30:45.190610 39557 net.cpp:406] res3_6 <- res3_5_res3_5_0_split_1
I0510 16:30:45.190615 39557 net.cpp:406] res3_6 <- conv3_6_em
I0510 16:30:45.190619 39557 net.cpp:380] res3_6 -> res3_6
I0510 16:30:45.190649 39557 net.cpp:122] Setting up res3_6
I0510 16:30:45.190657 39557 net.cpp:129] Top shape: 64 128 14 12 (1376256)
I0510 16:30:45.190660 39557 net.cpp:137] Memory required for data: 2766275328
I0510 16:30:45.190663 39557 layer_factory.hpp:77] Creating layer conv4_ex
I0510 16:30:45.190675 39557 net.cpp:84] Creating Layer conv4_ex
I0510 16:30:45.190678 39557 net.cpp:406] conv4_ex <- res3_6
I0510 16:30:45.190685 39557 net.cpp:380] conv4_ex -> conv4_ex
I0510 16:30:45.193234 39557 net.cpp:122] Setting up conv4_ex
I0510 16:30:45.193259 39557 net.cpp:129] Top shape: 64 512 14 12 (5505024)
I0510 16:30:45.193262 39557 net.cpp:137] Memory required for data: 2788295424
I0510 16:30:45.193270 39557 layer_factory.hpp:77] Creating layer conv4_ex/bn
I0510 16:30:45.193279 39557 net.cpp:84] Creating Layer conv4_ex/bn
I0510 16:30:45.193282 39557 net.cpp:406] conv4_ex/bn <- conv4_ex
I0510 16:30:45.193289 39557 net.cpp:367] conv4_ex/bn -> conv4_ex (in-place)
I0510 16:30:45.193503 39557 net.cpp:122] Setting up conv4_ex/bn
I0510 16:30:45.193513 39557 net.cpp:129] Top shape: 64 512 14 12 (5505024)
I0510 16:30:45.193516 39557 net.cpp:137] Memory required for data: 2810315520
I0510 16:30:45.193539 39557 layer_factory.hpp:77] Creating layer conv4_ex/scale
I0510 16:30:45.193547 39557 net.cpp:84] Creating Layer conv4_ex/scale
I0510 16:30:45.193550 39557 net.cpp:406] conv4_ex/scale <- conv4_ex
I0510 16:30:45.193555 39557 net.cpp:367] conv4_ex/scale -> conv4_ex (in-place)
I0510 16:30:45.193596 39557 layer_factory.hpp:77] Creating layer conv4_ex/scale
I0510 16:30:45.193735 39557 net.cpp:122] Setting up conv4_ex/scale
I0510 16:30:45.193744 39557 net.cpp:129] Top shape: 64 512 14 12 (5505024)
I0510 16:30:45.193747 39557 net.cpp:137] Memory required for data: 2832335616
I0510 16:30:45.193753 39557 layer_factory.hpp:77] Creating layer relu4_ex
I0510 16:30:45.193758 39557 net.cpp:84] Creating Layer relu4_ex
I0510 16:30:45.193761 39557 net.cpp:406] relu4_ex <- conv4_ex
I0510 16:30:45.193768 39557 net.cpp:367] relu4_ex -> conv4_ex (in-place)
I0510 16:30:45.193923 39557 net.cpp:122] Setting up relu4_ex
I0510 16:30:45.193933 39557 net.cpp:129] Top shape: 64 512 14 12 (5505024)
I0510 16:30:45.193935 39557 net.cpp:137] Memory required for data: 2854355712
I0510 16:30:45.193939 39557 layer_factory.hpp:77] Creating layer conv4_dw
I0510 16:30:45.193948 39557 net.cpp:84] Creating Layer conv4_dw
I0510 16:30:45.193951 39557 net.cpp:406] conv4_dw <- conv4_ex
I0510 16:30:45.193958 39557 net.cpp:380] conv4_dw -> conv4_dw
I0510 16:30:45.194185 39557 net.cpp:122] Setting up conv4_dw
I0510 16:30:45.194195 39557 net.cpp:129] Top shape: 64 512 7 6 (1376256)
I0510 16:30:45.194198 39557 net.cpp:137] Memory required for data: 2859860736
I0510 16:30:45.194202 39557 layer_factory.hpp:77] Creating layer conv4_dw/bn
I0510 16:30:45.194208 39557 net.cpp:84] Creating Layer conv4_dw/bn
I0510 16:30:45.194211 39557 net.cpp:406] conv4_dw/bn <- conv4_dw
I0510 16:30:45.194216 39557 net.cpp:367] conv4_dw/bn -> conv4_dw (in-place)
I0510 16:30:45.194416 39557 net.cpp:122] Setting up conv4_dw/bn
I0510 16:30:45.194424 39557 net.cpp:129] Top shape: 64 512 7 6 (1376256)
I0510 16:30:45.194427 39557 net.cpp:137] Memory required for data: 2865365760
I0510 16:30:45.194445 39557 layer_factory.hpp:77] Creating layer conv4_dw/scale
I0510 16:30:45.194453 39557 net.cpp:84] Creating Layer conv4_dw/scale
I0510 16:30:45.194456 39557 net.cpp:406] conv4_dw/scale <- conv4_dw
I0510 16:30:45.194461 39557 net.cpp:367] conv4_dw/scale -> conv4_dw (in-place)
I0510 16:30:45.194502 39557 layer_factory.hpp:77] Creating layer conv4_dw/scale
I0510 16:30:45.194631 39557 net.cpp:122] Setting up conv4_dw/scale
I0510 16:30:45.194640 39557 net.cpp:129] Top shape: 64 512 7 6 (1376256)
I0510 16:30:45.194654 39557 net.cpp:137] Memory required for data: 2870870784
I0510 16:30:45.194663 39557 layer_factory.hpp:77] Creating layer relu4_dw
I0510 16:30:45.194667 39557 net.cpp:84] Creating Layer relu4_dw
I0510 16:30:45.194684 39557 net.cpp:406] relu4_dw <- conv4_dw
I0510 16:30:45.194690 39557 net.cpp:367] relu4_dw -> conv4_dw (in-place)
I0510 16:30:45.194803 39557 net.cpp:122] Setting up relu4_dw
I0510 16:30:45.194813 39557 net.cpp:129] Top shape: 64 512 7 6 (1376256)
I0510 16:30:45.194814 39557 net.cpp:137] Memory required for data: 2876375808
I0510 16:30:45.194819 39557 layer_factory.hpp:77] Creating layer conv4_em
I0510 16:30:45.194829 39557 net.cpp:84] Creating Layer conv4_em
I0510 16:30:45.194833 39557 net.cpp:406] conv4_em <- conv4_dw
I0510 16:30:45.194839 39557 net.cpp:380] conv4_em -> conv4_em
I0510 16:30:45.196355 39557 net.cpp:122] Setting up conv4_em
I0510 16:30:45.196368 39557 net.cpp:129] Top shape: 64 128 7 6 (344064)
I0510 16:30:45.196382 39557 net.cpp:137] Memory required for data: 2877752064
I0510 16:30:45.196388 39557 layer_factory.hpp:77] Creating layer conv4_em/bn
I0510 16:30:45.196393 39557 net.cpp:84] Creating Layer conv4_em/bn
I0510 16:30:45.196398 39557 net.cpp:406] conv4_em/bn <- conv4_em
I0510 16:30:45.196415 39557 net.cpp:367] conv4_em/bn -> conv4_em (in-place)
I0510 16:30:45.196632 39557 net.cpp:122] Setting up conv4_em/bn
I0510 16:30:45.196641 39557 net.cpp:129] Top shape: 64 128 7 6 (344064)
I0510 16:30:45.196645 39557 net.cpp:137] Memory required for data: 2879128320
I0510 16:30:45.196653 39557 layer_factory.hpp:77] Creating layer conv4_em/scale
I0510 16:30:45.196660 39557 net.cpp:84] Creating Layer conv4_em/scale
I0510 16:30:45.196662 39557 net.cpp:406] conv4_em/scale <- conv4_em
I0510 16:30:45.196669 39557 net.cpp:367] conv4_em/scale -> conv4_em (in-place)
I0510 16:30:45.196724 39557 layer_factory.hpp:77] Creating layer conv4_em/scale
I0510 16:30:45.196846 39557 net.cpp:122] Setting up conv4_em/scale
I0510 16:30:45.196854 39557 net.cpp:129] Top shape: 64 128 7 6 (344064)
I0510 16:30:45.196856 39557 net.cpp:137] Memory required for data: 2880504576
I0510 16:30:45.196862 39557 layer_factory.hpp:77] Creating layer conv4_em_conv4_em/scale_0_split
I0510 16:30:45.196874 39557 net.cpp:84] Creating Layer conv4_em_conv4_em/scale_0_split
I0510 16:30:45.196877 39557 net.cpp:406] conv4_em_conv4_em/scale_0_split <- conv4_em
I0510 16:30:45.196885 39557 net.cpp:380] conv4_em_conv4_em/scale_0_split -> conv4_em_conv4_em/scale_0_split_0
I0510 16:30:45.196893 39557 net.cpp:380] conv4_em_conv4_em/scale_0_split -> conv4_em_conv4_em/scale_0_split_1
I0510 16:30:45.196936 39557 net.cpp:122] Setting up conv4_em_conv4_em/scale_0_split
I0510 16:30:45.196944 39557 net.cpp:129] Top shape: 64 128 7 6 (344064)
I0510 16:30:45.196949 39557 net.cpp:129] Top shape: 64 128 7 6 (344064)
I0510 16:30:45.196950 39557 net.cpp:137] Memory required for data: 2883257088
I0510 16:30:45.196954 39557 layer_factory.hpp:77] Creating layer conv4_1_ex
I0510 16:30:45.196961 39557 net.cpp:84] Creating Layer conv4_1_ex
I0510 16:30:45.196964 39557 net.cpp:406] conv4_1_ex <- conv4_em_conv4_em/scale_0_split_0
I0510 16:30:45.196972 39557 net.cpp:380] conv4_1_ex -> conv4_1_ex
I0510 16:30:45.198271 39557 net.cpp:122] Setting up conv4_1_ex
I0510 16:30:45.198284 39557 net.cpp:129] Top shape: 64 256 7 6 (688128)
I0510 16:30:45.198297 39557 net.cpp:137] Memory required for data: 2886009600
I0510 16:30:45.198304 39557 layer_factory.hpp:77] Creating layer conv4_1_ex/bn
I0510 16:30:45.198313 39557 net.cpp:84] Creating Layer conv4_1_ex/bn
I0510 16:30:45.198318 39557 net.cpp:406] conv4_1_ex/bn <- conv4_1_ex
I0510 16:30:45.198324 39557 net.cpp:367] conv4_1_ex/bn -> conv4_1_ex (in-place)
I0510 16:30:45.198537 39557 net.cpp:122] Setting up conv4_1_ex/bn
I0510 16:30:45.198546 39557 net.cpp:129] Top shape: 64 256 7 6 (688128)
I0510 16:30:45.198549 39557 net.cpp:137] Memory required for data: 2888762112
I0510 16:30:45.198556 39557 layer_factory.hpp:77] Creating layer conv4_1_ex/scale
I0510 16:30:45.198565 39557 net.cpp:84] Creating Layer conv4_1_ex/scale
I0510 16:30:45.198580 39557 net.cpp:406] conv4_1_ex/scale <- conv4_1_ex
I0510 16:30:45.198586 39557 net.cpp:367] conv4_1_ex/scale -> conv4_1_ex (in-place)
I0510 16:30:45.198635 39557 layer_factory.hpp:77] Creating layer conv4_1_ex/scale
I0510 16:30:45.198760 39557 net.cpp:122] Setting up conv4_1_ex/scale
I0510 16:30:45.198770 39557 net.cpp:129] Top shape: 64 256 7 6 (688128)
I0510 16:30:45.198771 39557 net.cpp:137] Memory required for data: 2891514624
I0510 16:30:45.198777 39557 layer_factory.hpp:77] Creating layer relu4_1_ex
I0510 16:30:45.198784 39557 net.cpp:84] Creating Layer relu4_1_ex
I0510 16:30:45.198787 39557 net.cpp:406] relu4_1_ex <- conv4_1_ex
I0510 16:30:45.198793 39557 net.cpp:367] relu4_1_ex -> conv4_1_ex (in-place)
I0510 16:30:45.198889 39557 net.cpp:122] Setting up relu4_1_ex
I0510 16:30:45.198896 39557 net.cpp:129] Top shape: 64 256 7 6 (688128)
I0510 16:30:45.198899 39557 net.cpp:137] Memory required for data: 2894267136
I0510 16:30:45.198904 39557 layer_factory.hpp:77] Creating layer conv4_1_dw
I0510 16:30:45.198915 39557 net.cpp:84] Creating Layer conv4_1_dw
I0510 16:30:45.198918 39557 net.cpp:406] conv4_1_dw <- conv4_1_ex
I0510 16:30:45.198923 39557 net.cpp:380] conv4_1_dw -> conv4_1_dw
I0510 16:30:45.199154 39557 net.cpp:122] Setting up conv4_1_dw
I0510 16:30:45.199164 39557 net.cpp:129] Top shape: 64 256 7 6 (688128)
I0510 16:30:45.199167 39557 net.cpp:137] Memory required for data: 2897019648
I0510 16:30:45.199172 39557 layer_factory.hpp:77] Creating layer conv4_1_dw/bn
I0510 16:30:45.199177 39557 net.cpp:84] Creating Layer conv4_1_dw/bn
I0510 16:30:45.199180 39557 net.cpp:406] conv4_1_dw/bn <- conv4_1_dw
I0510 16:30:45.199187 39557 net.cpp:367] conv4_1_dw/bn -> conv4_1_dw (in-place)
I0510 16:30:45.199405 39557 net.cpp:122] Setting up conv4_1_dw/bn
I0510 16:30:45.199414 39557 net.cpp:129] Top shape: 64 256 7 6 (688128)
I0510 16:30:45.199417 39557 net.cpp:137] Memory required for data: 2899772160
I0510 16:30:45.199435 39557 layer_factory.hpp:77] Creating layer conv4_1_dw/scale
I0510 16:30:45.199442 39557 net.cpp:84] Creating Layer conv4_1_dw/scale
I0510 16:30:45.199446 39557 net.cpp:406] conv4_1_dw/scale <- conv4_1_dw
I0510 16:30:45.199452 39557 net.cpp:367] conv4_1_dw/scale -> conv4_1_dw (in-place)
I0510 16:30:45.199496 39557 layer_factory.hpp:77] Creating layer conv4_1_dw/scale
I0510 16:30:45.199623 39557 net.cpp:122] Setting up conv4_1_dw/scale
I0510 16:30:45.199632 39557 net.cpp:129] Top shape: 64 256 7 6 (688128)
I0510 16:30:45.199635 39557 net.cpp:137] Memory required for data: 2902524672
I0510 16:30:45.199641 39557 layer_factory.hpp:77] Creating layer relu4_1_dw
I0510 16:30:45.199646 39557 net.cpp:84] Creating Layer relu4_1_dw
I0510 16:30:45.199651 39557 net.cpp:406] relu4_1_dw <- conv4_1_dw
I0510 16:30:45.199656 39557 net.cpp:367] relu4_1_dw -> conv4_1_dw (in-place)
I0510 16:30:45.199753 39557 net.cpp:122] Setting up relu4_1_dw
I0510 16:30:45.199760 39557 net.cpp:129] Top shape: 64 256 7 6 (688128)
I0510 16:30:45.199764 39557 net.cpp:137] Memory required for data: 2905277184
I0510 16:30:45.199769 39557 layer_factory.hpp:77] Creating layer conv4_1_em
I0510 16:30:45.199775 39557 net.cpp:84] Creating Layer conv4_1_em
I0510 16:30:45.199779 39557 net.cpp:406] conv4_1_em <- conv4_1_dw
I0510 16:30:45.199786 39557 net.cpp:380] conv4_1_em -> conv4_1_em
I0510 16:30:45.201117 39557 net.cpp:122] Setting up conv4_1_em
I0510 16:30:45.201131 39557 net.cpp:129] Top shape: 64 128 7 6 (344064)
I0510 16:30:45.201144 39557 net.cpp:137] Memory required for data: 2906653440
I0510 16:30:45.201150 39557 layer_factory.hpp:77] Creating layer conv4_1_em/bn
I0510 16:30:45.201159 39557 net.cpp:84] Creating Layer conv4_1_em/bn
I0510 16:30:45.201164 39557 net.cpp:406] conv4_1_em/bn <- conv4_1_em
I0510 16:30:45.201169 39557 net.cpp:367] conv4_1_em/bn -> conv4_1_em (in-place)
I0510 16:30:45.201381 39557 net.cpp:122] Setting up conv4_1_em/bn
I0510 16:30:45.201390 39557 net.cpp:129] Top shape: 64 128 7 6 (344064)
I0510 16:30:45.201393 39557 net.cpp:137] Memory required for data: 2908029696
I0510 16:30:45.201401 39557 layer_factory.hpp:77] Creating layer conv4_1_em/scale
I0510 16:30:45.201422 39557 net.cpp:84] Creating Layer conv4_1_em/scale
I0510 16:30:45.201429 39557 net.cpp:406] conv4_1_em/scale <- conv4_1_em
I0510 16:30:45.201434 39557 net.cpp:367] conv4_1_em/scale -> conv4_1_em (in-place)
I0510 16:30:45.201485 39557 layer_factory.hpp:77] Creating layer conv4_1_em/scale
I0510 16:30:45.201608 39557 net.cpp:122] Setting up conv4_1_em/scale
I0510 16:30:45.201617 39557 net.cpp:129] Top shape: 64 128 7 6 (344064)
I0510 16:30:45.201619 39557 net.cpp:137] Memory required for data: 2909405952
I0510 16:30:45.201625 39557 layer_factory.hpp:77] Creating layer res4_1
I0510 16:30:45.201630 39557 net.cpp:84] Creating Layer res4_1
I0510 16:30:45.201634 39557 net.cpp:406] res4_1 <- conv4_em_conv4_em/scale_0_split_1
I0510 16:30:45.201638 39557 net.cpp:406] res4_1 <- conv4_1_em
I0510 16:30:45.201645 39557 net.cpp:380] res4_1 -> res4_1
I0510 16:30:45.201671 39557 net.cpp:122] Setting up res4_1
I0510 16:30:45.201678 39557 net.cpp:129] Top shape: 64 128 7 6 (344064)
I0510 16:30:45.201681 39557 net.cpp:137] Memory required for data: 2910782208
I0510 16:30:45.201684 39557 layer_factory.hpp:77] Creating layer res4_1_res4_1_0_split
I0510 16:30:45.201692 39557 net.cpp:84] Creating Layer res4_1_res4_1_0_split
I0510 16:30:45.201696 39557 net.cpp:406] res4_1_res4_1_0_split <- res4_1
I0510 16:30:45.201701 39557 net.cpp:380] res4_1_res4_1_0_split -> res4_1_res4_1_0_split_0
I0510 16:30:45.201707 39557 net.cpp:380] res4_1_res4_1_0_split -> res4_1_res4_1_0_split_1
I0510 16:30:45.201746 39557 net.cpp:122] Setting up res4_1_res4_1_0_split
I0510 16:30:45.201755 39557 net.cpp:129] Top shape: 64 128 7 6 (344064)
I0510 16:30:45.201757 39557 net.cpp:129] Top shape: 64 128 7 6 (344064)
I0510 16:30:45.201761 39557 net.cpp:137] Memory required for data: 2913534720
I0510 16:30:45.201764 39557 layer_factory.hpp:77] Creating layer conv4_2_ex
I0510 16:30:45.201776 39557 net.cpp:84] Creating Layer conv4_2_ex
I0510 16:30:45.201781 39557 net.cpp:406] conv4_2_ex <- res4_1_res4_1_0_split_0
I0510 16:30:45.201787 39557 net.cpp:380] conv4_2_ex -> conv4_2_ex
I0510 16:30:45.203121 39557 net.cpp:122] Setting up conv4_2_ex
I0510 16:30:45.203145 39557 net.cpp:129] Top shape: 64 256 7 6 (688128)
I0510 16:30:45.203148 39557 net.cpp:137] Memory required for data: 2916287232
I0510 16:30:45.203155 39557 layer_factory.hpp:77] Creating layer conv4_2_ex/bn
I0510 16:30:45.203163 39557 net.cpp:84] Creating Layer conv4_2_ex/bn
I0510 16:30:45.203167 39557 net.cpp:406] conv4_2_ex/bn <- conv4_2_ex
I0510 16:30:45.203172 39557 net.cpp:367] conv4_2_ex/bn -> conv4_2_ex (in-place)
I0510 16:30:45.203416 39557 net.cpp:122] Setting up conv4_2_ex/bn
I0510 16:30:45.203426 39557 net.cpp:129] Top shape: 64 256 7 6 (688128)
I0510 16:30:45.203429 39557 net.cpp:137] Memory required for data: 2919039744
I0510 16:30:45.203438 39557 layer_factory.hpp:77] Creating layer conv4_2_ex/scale
I0510 16:30:45.203444 39557 net.cpp:84] Creating Layer conv4_2_ex/scale
I0510 16:30:45.203446 39557 net.cpp:406] conv4_2_ex/scale <- conv4_2_ex
I0510 16:30:45.203451 39557 net.cpp:367] conv4_2_ex/scale -> conv4_2_ex (in-place)
I0510 16:30:45.203500 39557 layer_factory.hpp:77] Creating layer conv4_2_ex/scale
I0510 16:30:45.203626 39557 net.cpp:122] Setting up conv4_2_ex/scale
I0510 16:30:45.203634 39557 net.cpp:129] Top shape: 64 256 7 6 (688128)
I0510 16:30:45.203637 39557 net.cpp:137] Memory required for data: 2921792256
I0510 16:30:45.203642 39557 layer_factory.hpp:77] Creating layer relu4_2_ex
I0510 16:30:45.203651 39557 net.cpp:84] Creating Layer relu4_2_ex
I0510 16:30:45.203655 39557 net.cpp:406] relu4_2_ex <- conv4_2_ex
I0510 16:30:45.203660 39557 net.cpp:367] relu4_2_ex -> conv4_2_ex (in-place)
I0510 16:30:45.203769 39557 net.cpp:122] Setting up relu4_2_ex
I0510 16:30:45.203778 39557 net.cpp:129] Top shape: 64 256 7 6 (688128)
I0510 16:30:45.203780 39557 net.cpp:137] Memory required for data: 2924544768
I0510 16:30:45.203785 39557 layer_factory.hpp:77] Creating layer conv4_2_dw
I0510 16:30:45.203793 39557 net.cpp:84] Creating Layer conv4_2_dw
I0510 16:30:45.203807 39557 net.cpp:406] conv4_2_dw <- conv4_2_ex
I0510 16:30:45.203816 39557 net.cpp:380] conv4_2_dw -> conv4_2_dw
I0510 16:30:45.204054 39557 net.cpp:122] Setting up conv4_2_dw
I0510 16:30:45.204064 39557 net.cpp:129] Top shape: 64 256 7 6 (688128)
I0510 16:30:45.204067 39557 net.cpp:137] Memory required for data: 2927297280
I0510 16:30:45.204072 39557 layer_factory.hpp:77] Creating layer conv4_2_dw/bn
I0510 16:30:45.204080 39557 net.cpp:84] Creating Layer conv4_2_dw/bn
I0510 16:30:45.204084 39557 net.cpp:406] conv4_2_dw/bn <- conv4_2_dw
I0510 16:30:45.204089 39557 net.cpp:367] conv4_2_dw/bn -> conv4_2_dw (in-place)
I0510 16:30:45.204293 39557 net.cpp:122] Setting up conv4_2_dw/bn
I0510 16:30:45.204300 39557 net.cpp:129] Top shape: 64 256 7 6 (688128)
I0510 16:30:45.204308 39557 net.cpp:137] Memory required for data: 2930049792
I0510 16:30:45.204315 39557 layer_factory.hpp:77] Creating layer conv4_2_dw/scale
I0510 16:30:45.204324 39557 net.cpp:84] Creating Layer conv4_2_dw/scale
I0510 16:30:45.204326 39557 net.cpp:406] conv4_2_dw/scale <- conv4_2_dw
I0510 16:30:45.204331 39557 net.cpp:367] conv4_2_dw/scale -> conv4_2_dw (in-place)
I0510 16:30:45.204376 39557 layer_factory.hpp:77] Creating layer conv4_2_dw/scale
I0510 16:30:45.204499 39557 net.cpp:122] Setting up conv4_2_dw/scale
I0510 16:30:45.204507 39557 net.cpp:129] Top shape: 64 256 7 6 (688128)
I0510 16:30:45.204510 39557 net.cpp:137] Memory required for data: 2932802304
I0510 16:30:45.204515 39557 layer_factory.hpp:77] Creating layer relu4_2_dw
I0510 16:30:45.204535 39557 net.cpp:84] Creating Layer relu4_2_dw
I0510 16:30:45.204538 39557 net.cpp:406] relu4_2_dw <- conv4_2_dw
I0510 16:30:45.204546 39557 net.cpp:367] relu4_2_dw -> conv4_2_dw (in-place)
I0510 16:30:45.204656 39557 net.cpp:122] Setting up relu4_2_dw
I0510 16:30:45.204665 39557 net.cpp:129] Top shape: 64 256 7 6 (688128)
I0510 16:30:45.204668 39557 net.cpp:137] Memory required for data: 2935554816
I0510 16:30:45.204672 39557 layer_factory.hpp:77] Creating layer conv4_2_em
I0510 16:30:45.204682 39557 net.cpp:84] Creating Layer conv4_2_em
I0510 16:30:45.204685 39557 net.cpp:406] conv4_2_em <- conv4_2_dw
I0510 16:30:45.204691 39557 net.cpp:380] conv4_2_em -> conv4_2_em
I0510 16:30:45.206064 39557 net.cpp:122] Setting up conv4_2_em
I0510 16:30:45.206090 39557 net.cpp:129] Top shape: 64 128 7 6 (344064)
I0510 16:30:45.206094 39557 net.cpp:137] Memory required for data: 2936931072
I0510 16:30:45.206100 39557 layer_factory.hpp:77] Creating layer conv4_2_em/bn
I0510 16:30:45.206109 39557 net.cpp:84] Creating Layer conv4_2_em/bn
I0510 16:30:45.206112 39557 net.cpp:406] conv4_2_em/bn <- conv4_2_em
I0510 16:30:45.206120 39557 net.cpp:367] conv4_2_em/bn -> conv4_2_em (in-place)
I0510 16:30:45.206346 39557 net.cpp:122] Setting up conv4_2_em/bn
I0510 16:30:45.206356 39557 net.cpp:129] Top shape: 64 128 7 6 (344064)
I0510 16:30:45.206358 39557 net.cpp:137] Memory required for data: 2938307328
I0510 16:30:45.206365 39557 layer_factory.hpp:77] Creating layer conv4_2_em/scale
I0510 16:30:45.206372 39557 net.cpp:84] Creating Layer conv4_2_em/scale
I0510 16:30:45.206375 39557 net.cpp:406] conv4_2_em/scale <- conv4_2_em
I0510 16:30:45.206382 39557 net.cpp:367] conv4_2_em/scale -> conv4_2_em (in-place)
I0510 16:30:45.206427 39557 layer_factory.hpp:77] Creating layer conv4_2_em/scale
I0510 16:30:45.206573 39557 net.cpp:122] Setting up conv4_2_em/scale
I0510 16:30:45.206583 39557 net.cpp:129] Top shape: 64 128 7 6 (344064)
I0510 16:30:45.206585 39557 net.cpp:137] Memory required for data: 2939683584
I0510 16:30:45.206591 39557 layer_factory.hpp:77] Creating layer res4_2
I0510 16:30:45.206596 39557 net.cpp:84] Creating Layer res4_2
I0510 16:30:45.206600 39557 net.cpp:406] res4_2 <- res4_1_res4_1_0_split_1
I0510 16:30:45.206604 39557 net.cpp:406] res4_2 <- conv4_2_em
I0510 16:30:45.206611 39557 net.cpp:380] res4_2 -> res4_2
I0510 16:30:45.206638 39557 net.cpp:122] Setting up res4_2
I0510 16:30:45.206646 39557 net.cpp:129] Top shape: 64 128 7 6 (344064)
I0510 16:30:45.206650 39557 net.cpp:137] Memory required for data: 2941059840
I0510 16:30:45.206665 39557 layer_factory.hpp:77] Creating layer conv5_ex
I0510 16:30:45.206676 39557 net.cpp:84] Creating Layer conv5_ex
I0510 16:30:45.206679 39557 net.cpp:406] conv5_ex <- res4_2
I0510 16:30:45.206688 39557 net.cpp:380] conv5_ex -> conv5_ex
I0510 16:30:45.208248 39557 net.cpp:122] Setting up conv5_ex
I0510 16:30:45.208262 39557 net.cpp:129] Top shape: 64 512 7 6 (1376256)
I0510 16:30:45.208276 39557 net.cpp:137] Memory required for data: 2946564864
I0510 16:30:45.208281 39557 layer_factory.hpp:77] Creating layer conv5_ex/bn
I0510 16:30:45.208290 39557 net.cpp:84] Creating Layer conv5_ex/bn
I0510 16:30:45.208294 39557 net.cpp:406] conv5_ex/bn <- conv5_ex
I0510 16:30:45.208302 39557 net.cpp:367] conv5_ex/bn -> conv5_ex (in-place)
I0510 16:30:45.208521 39557 net.cpp:122] Setting up conv5_ex/bn
I0510 16:30:45.208530 39557 net.cpp:129] Top shape: 64 512 7 6 (1376256)
I0510 16:30:45.208534 39557 net.cpp:137] Memory required for data: 2952069888
I0510 16:30:45.208542 39557 layer_factory.hpp:77] Creating layer conv5_ex/scale
I0510 16:30:45.208551 39557 net.cpp:84] Creating Layer conv5_ex/scale
I0510 16:30:45.208559 39557 net.cpp:406] conv5_ex/scale <- conv5_ex
I0510 16:30:45.208564 39557 net.cpp:367] conv5_ex/scale -> conv5_ex (in-place)
I0510 16:30:45.208607 39557 layer_factory.hpp:77] Creating layer conv5_ex/scale
I0510 16:30:45.208741 39557 net.cpp:122] Setting up conv5_ex/scale
I0510 16:30:45.208750 39557 net.cpp:129] Top shape: 64 512 7 6 (1376256)
I0510 16:30:45.208753 39557 net.cpp:137] Memory required for data: 2957574912
I0510 16:30:45.208760 39557 layer_factory.hpp:77] Creating layer relu5_ex
I0510 16:30:45.208765 39557 net.cpp:84] Creating Layer relu5_ex
I0510 16:30:45.208767 39557 net.cpp:406] relu5_ex <- conv5_ex
I0510 16:30:45.208773 39557 net.cpp:367] relu5_ex -> conv5_ex (in-place)
I0510 16:30:45.208874 39557 net.cpp:122] Setting up relu5_ex
I0510 16:30:45.208883 39557 net.cpp:129] Top shape: 64 512 7 6 (1376256)
I0510 16:30:45.208885 39557 net.cpp:137] Memory required for data: 2963079936
I0510 16:30:45.208890 39557 layer_factory.hpp:77] Creating layer conv5_dw
I0510 16:30:45.208899 39557 net.cpp:84] Creating Layer conv5_dw
I0510 16:30:45.208902 39557 net.cpp:406] conv5_dw <- conv5_ex
I0510 16:30:45.208907 39557 net.cpp:380] conv5_dw -> conv5_dw
I0510 16:30:45.209229 39557 net.cpp:122] Setting up conv5_dw
I0510 16:30:45.209239 39557 net.cpp:129] Top shape: 64 512 1 1 (32768)
I0510 16:30:45.209242 39557 net.cpp:137] Memory required for data: 2963211008
I0510 16:30:45.209249 39557 layer_factory.hpp:77] Creating layer conv5_dw/bn
I0510 16:30:45.209256 39557 net.cpp:84] Creating Layer conv5_dw/bn
I0510 16:30:45.209259 39557 net.cpp:406] conv5_dw/bn <- conv5_dw
I0510 16:30:45.209264 39557 net.cpp:367] conv5_dw/bn -> conv5_dw (in-place)
I0510 16:30:45.209478 39557 net.cpp:122] Setting up conv5_dw/bn
I0510 16:30:45.209486 39557 net.cpp:129] Top shape: 64 512 1 1 (32768)
I0510 16:30:45.209489 39557 net.cpp:137] Memory required for data: 2963342080
I0510 16:30:45.209496 39557 layer_factory.hpp:77] Creating layer conv5_dw/scale
I0510 16:30:45.209501 39557 net.cpp:84] Creating Layer conv5_dw/scale
I0510 16:30:45.209504 39557 net.cpp:406] conv5_dw/scale <- conv5_dw
I0510 16:30:45.209511 39557 net.cpp:367] conv5_dw/scale -> conv5_dw (in-place)
I0510 16:30:45.209549 39557 layer_factory.hpp:77] Creating layer conv5_dw/scale
I0510 16:30:45.209672 39557 net.cpp:122] Setting up conv5_dw/scale
I0510 16:30:45.209681 39557 net.cpp:129] Top shape: 64 512 1 1 (32768)
I0510 16:30:45.209684 39557 net.cpp:137] Memory required for data: 2963473152
I0510 16:30:45.209689 39557 layer_factory.hpp:77] Creating layer fc5
I0510 16:30:45.209699 39557 net.cpp:84] Creating Layer fc5
I0510 16:30:45.209702 39557 net.cpp:406] fc5 <- conv5_dw
I0510 16:30:45.209707 39557 net.cpp:380] fc5 -> fc5
I0510 16:30:45.210114 39557 net.cpp:122] Setting up fc5
I0510 16:30:45.210124 39557 net.cpp:129] Top shape: 64 128 (8192)
I0510 16:30:45.210126 39557 net.cpp:137] Memory required for data: 2963505920
I0510 16:30:45.210142 39557 layer_factory.hpp:77] Creating layer norm1
I0510 16:30:45.210149 39557 net.cpp:84] Creating Layer norm1
I0510 16:30:45.210152 39557 net.cpp:406] norm1 <- fc5
I0510 16:30:45.210157 39557 net.cpp:380] norm1 -> norm1
I0510 16:30:45.210218 39557 net.cpp:122] Setting up norm1
I0510 16:30:45.210227 39557 net.cpp:129] Top shape: 64 128 1 1 (8192)
I0510 16:30:45.210230 39557 net.cpp:137] Memory required for data: 2963538688
I0510 16:30:45.210233 39557 layer_factory.hpp:77] Creating layer fc6_l2
I0510 16:30:45.210242 39557 net.cpp:84] Creating Layer fc6_l2
I0510 16:30:45.210244 39557 net.cpp:406] fc6_l2 <- norm1
I0510 16:30:45.210249 39557 net.cpp:380] fc6_l2 -> fc6
I0510 16:30:45.217748 39557 net.cpp:122] Setting up fc6_l2
I0510 16:30:45.217761 39557 net.cpp:129] Top shape: 64 9131 (584384)
I0510 16:30:45.217775 39557 net.cpp:137] Memory required for data: 2965876224
I0510 16:30:45.217780 39557 layer_factory.hpp:77] Creating layer label_specific_margin
I0510 16:30:45.217787 39557 net.cpp:84] Creating Layer label_specific_margin
I0510 16:30:45.217792 39557 net.cpp:406] label_specific_margin <- fc6
I0510 16:30:45.217797 39557 net.cpp:406] label_specific_margin <- label_data_1_split_0
I0510 16:30:45.217802 39557 net.cpp:380] label_specific_margin -> fc6_margin
I0510 16:30:45.217835 39557 net.cpp:122] Setting up label_specific_margin
I0510 16:30:45.217844 39557 net.cpp:129] Top shape: 64 9131 (584384)
I0510 16:30:45.217846 39557 net.cpp:137] Memory required for data: 2968213760
I0510 16:30:45.217849 39557 layer_factory.hpp:77] Creating layer fc6_margin_scale
I0510 16:30:45.217856 39557 net.cpp:84] Creating Layer fc6_margin_scale
I0510 16:30:45.217860 39557 net.cpp:406] fc6_margin_scale <- fc6_margin
I0510 16:30:45.217865 39557 net.cpp:380] fc6_margin_scale -> fc6_margin_scale
I0510 16:30:45.217975 39557 net.cpp:122] Setting up fc6_margin_scale
I0510 16:30:45.217985 39557 net.cpp:129] Top shape: 64 9131 (584384)
I0510 16:30:45.217988 39557 net.cpp:137] Memory required for data: 2970551296
I0510 16:30:45.217993 39557 layer_factory.hpp:77] Creating layer softmax_loss
I0510 16:30:45.217999 39557 net.cpp:84] Creating Layer softmax_loss
I0510 16:30:45.218003 39557 net.cpp:406] softmax_loss <- fc6_margin_scale
I0510 16:30:45.218008 39557 net.cpp:406] softmax_loss <- label_data_1_split_1
I0510 16:30:45.218014 39557 net.cpp:380] softmax_loss -> softmax_loss
I0510 16:30:45.218021 39557 layer_factory.hpp:77] Creating layer softmax_loss
I0510 16:30:45.220355 39557 net.cpp:122] Setting up softmax_loss
I0510 16:30:45.220386 39557 net.cpp:129] Top shape: (1)
I0510 16:30:45.220402 39557 net.cpp:132]     with loss weight 1
I0510 16:30:45.220430 39557 net.cpp:137] Memory required for data: 2970551300
I0510 16:30:45.220434 39557 net.cpp:198] softmax_loss needs backward computation.
I0510 16:30:45.220443 39557 net.cpp:198] fc6_margin_scale needs backward computation.
I0510 16:30:45.220445 39557 net.cpp:198] label_specific_margin needs backward computation.
I0510 16:30:45.220449 39557 net.cpp:198] fc6_l2 needs backward computation.
I0510 16:30:45.220453 39557 net.cpp:198] norm1 needs backward computation.
I0510 16:30:45.220455 39557 net.cpp:198] fc5 needs backward computation.
I0510 16:30:45.220458 39557 net.cpp:198] conv5_dw/scale needs backward computation.
I0510 16:30:45.220461 39557 net.cpp:198] conv5_dw/bn needs backward computation.
I0510 16:30:45.220463 39557 net.cpp:198] conv5_dw needs backward computation.
I0510 16:30:45.220481 39557 net.cpp:198] relu5_ex needs backward computation.
I0510 16:30:45.220484 39557 net.cpp:198] conv5_ex/scale needs backward computation.
I0510 16:30:45.220487 39557 net.cpp:198] conv5_ex/bn needs backward computation.
I0510 16:30:45.220489 39557 net.cpp:198] conv5_ex needs backward computation.
I0510 16:30:45.220494 39557 net.cpp:198] res4_2 needs backward computation.
I0510 16:30:45.220499 39557 net.cpp:198] conv4_2_em/scale needs backward computation.
I0510 16:30:45.220502 39557 net.cpp:198] conv4_2_em/bn needs backward computation.
I0510 16:30:45.220505 39557 net.cpp:198] conv4_2_em needs backward computation.
I0510 16:30:45.220520 39557 net.cpp:198] relu4_2_dw needs backward computation.
I0510 16:30:45.220522 39557 net.cpp:198] conv4_2_dw/scale needs backward computation.
I0510 16:30:45.220525 39557 net.cpp:198] conv4_2_dw/bn needs backward computation.
I0510 16:30:45.220527 39557 net.cpp:198] conv4_2_dw needs backward computation.
I0510 16:30:45.220531 39557 net.cpp:198] relu4_2_ex needs backward computation.
I0510 16:30:45.220533 39557 net.cpp:198] conv4_2_ex/scale needs backward computation.
I0510 16:30:45.220536 39557 net.cpp:198] conv4_2_ex/bn needs backward computation.
I0510 16:30:45.220540 39557 net.cpp:198] conv4_2_ex needs backward computation.
I0510 16:30:45.220543 39557 net.cpp:198] res4_1_res4_1_0_split needs backward computation.
I0510 16:30:45.220546 39557 net.cpp:198] res4_1 needs backward computation.
I0510 16:30:45.220549 39557 net.cpp:198] conv4_1_em/scale needs backward computation.
I0510 16:30:45.220552 39557 net.cpp:198] conv4_1_em/bn needs backward computation.
I0510 16:30:45.220556 39557 net.cpp:198] conv4_1_em needs backward computation.
I0510 16:30:45.220558 39557 net.cpp:198] relu4_1_dw needs backward computation.
I0510 16:30:45.220561 39557 net.cpp:198] conv4_1_dw/scale needs backward computation.
I0510 16:30:45.220563 39557 net.cpp:198] conv4_1_dw/bn needs backward computation.
I0510 16:30:45.220567 39557 net.cpp:198] conv4_1_dw needs backward computation.
I0510 16:30:45.220571 39557 net.cpp:198] relu4_1_ex needs backward computation.
I0510 16:30:45.220572 39557 net.cpp:198] conv4_1_ex/scale needs backward computation.
I0510 16:30:45.220576 39557 net.cpp:198] conv4_1_ex/bn needs backward computation.
I0510 16:30:45.220578 39557 net.cpp:198] conv4_1_ex needs backward computation.
I0510 16:30:45.220582 39557 net.cpp:198] conv4_em_conv4_em/scale_0_split needs backward computation.
I0510 16:30:45.220584 39557 net.cpp:198] conv4_em/scale needs backward computation.
I0510 16:30:45.220587 39557 net.cpp:198] conv4_em/bn needs backward computation.
I0510 16:30:45.220590 39557 net.cpp:198] conv4_em needs backward computation.
I0510 16:30:45.220594 39557 net.cpp:198] relu4_dw needs backward computation.
I0510 16:30:45.220597 39557 net.cpp:198] conv4_dw/scale needs backward computation.
I0510 16:30:45.220599 39557 net.cpp:198] conv4_dw/bn needs backward computation.
I0510 16:30:45.220602 39557 net.cpp:198] conv4_dw needs backward computation.
I0510 16:30:45.220605 39557 net.cpp:198] relu4_ex needs backward computation.
I0510 16:30:45.220607 39557 net.cpp:198] conv4_ex/scale needs backward computation.
I0510 16:30:45.220610 39557 net.cpp:198] conv4_ex/bn needs backward computation.
I0510 16:30:45.220613 39557 net.cpp:198] conv4_ex needs backward computation.
I0510 16:30:45.220616 39557 net.cpp:198] res3_6 needs backward computation.
I0510 16:30:45.220619 39557 net.cpp:198] conv3_6_em/scale needs backward computation.
I0510 16:30:45.220623 39557 net.cpp:198] conv3_6_em/bn needs backward computation.
I0510 16:30:45.220625 39557 net.cpp:198] conv3_6_em needs backward computation.
I0510 16:30:45.220628 39557 net.cpp:198] relu3_6_dw needs backward computation.
I0510 16:30:45.220631 39557 net.cpp:198] conv3_6_dw/scale needs backward computation.
I0510 16:30:45.220633 39557 net.cpp:198] conv3_6_dw/bn needs backward computation.
I0510 16:30:45.220636 39557 net.cpp:198] conv3_6_dw needs backward computation.
I0510 16:30:45.220639 39557 net.cpp:198] relu3_6_ex needs backward computation.
I0510 16:30:45.220643 39557 net.cpp:198] conv3_6_ex/scale needs backward computation.
I0510 16:30:45.220645 39557 net.cpp:198] conv3_6_ex/bn needs backward computation.
I0510 16:30:45.220647 39557 net.cpp:198] conv3_6_ex needs backward computation.
I0510 16:30:45.220651 39557 net.cpp:198] res3_5_res3_5_0_split needs backward computation.
I0510 16:30:45.220654 39557 net.cpp:198] res3_5 needs backward computation.
I0510 16:30:45.220657 39557 net.cpp:198] conv3_5_em/scale needs backward computation.
I0510 16:30:45.220660 39557 net.cpp:198] conv3_5_em/bn needs backward computation.
I0510 16:30:45.220664 39557 net.cpp:198] conv3_5_em needs backward computation.
I0510 16:30:45.220671 39557 net.cpp:198] relu3_5_dw needs backward computation.
I0510 16:30:45.220674 39557 net.cpp:198] conv3_5_dw/scale needs backward computation.
I0510 16:30:45.220677 39557 net.cpp:198] conv3_5_dw/bn needs backward computation.
I0510 16:30:45.220680 39557 net.cpp:198] conv3_5_dw needs backward computation.
I0510 16:30:45.220683 39557 net.cpp:198] relu3_5_ex needs backward computation.
I0510 16:30:45.220686 39557 net.cpp:198] conv3_5_ex/scale needs backward computation.
I0510 16:30:45.220690 39557 net.cpp:198] conv3_5_ex/bn needs backward computation.
I0510 16:30:45.220691 39557 net.cpp:198] conv3_5_ex needs backward computation.
I0510 16:30:45.220695 39557 net.cpp:198] res3_4_res3_4_0_split needs backward computation.
I0510 16:30:45.220700 39557 net.cpp:198] res3_4 needs backward computation.
I0510 16:30:45.220705 39557 net.cpp:198] conv3_4_em/scale needs backward computation.
I0510 16:30:45.220707 39557 net.cpp:198] conv3_4_em/bn needs backward computation.
I0510 16:30:45.220710 39557 net.cpp:198] conv3_4_em needs backward computation.
I0510 16:30:45.220713 39557 net.cpp:198] relu3_4_dw needs backward computation.
I0510 16:30:45.220716 39557 net.cpp:198] conv3_4_dw/scale needs backward computation.
I0510 16:30:45.220718 39557 net.cpp:198] conv3_4_dw/bn needs backward computation.
I0510 16:30:45.220721 39557 net.cpp:198] conv3_4_dw needs backward computation.
I0510 16:30:45.220724 39557 net.cpp:198] relu3_4_ex needs backward computation.
I0510 16:30:45.220727 39557 net.cpp:198] conv3_4_ex/scale needs backward computation.
I0510 16:30:45.220731 39557 net.cpp:198] conv3_4_ex/bn needs backward computation.
I0510 16:30:45.220732 39557 net.cpp:198] conv3_4_ex needs backward computation.
I0510 16:30:45.220736 39557 net.cpp:198] res3_3_res3_3_0_split needs backward computation.
I0510 16:30:45.220739 39557 net.cpp:198] res3_3 needs backward computation.
I0510 16:30:45.220742 39557 net.cpp:198] conv3_3_em/scale needs backward computation.
I0510 16:30:45.220746 39557 net.cpp:198] conv3_3_em/bn needs backward computation.
I0510 16:30:45.220748 39557 net.cpp:198] conv3_3_em needs backward computation.
I0510 16:30:45.220752 39557 net.cpp:198] relu3_3_dw needs backward computation.
I0510 16:30:45.220754 39557 net.cpp:198] conv3_3_dw/scale needs backward computation.
I0510 16:30:45.220757 39557 net.cpp:198] conv3_3_dw/bn needs backward computation.
I0510 16:30:45.220760 39557 net.cpp:198] conv3_3_dw needs backward computation.
I0510 16:30:45.220763 39557 net.cpp:198] relu3_3_ex needs backward computation.
I0510 16:30:45.220767 39557 net.cpp:198] conv3_3_ex/scale needs backward computation.
I0510 16:30:45.220769 39557 net.cpp:198] conv3_3_ex/bn needs backward computation.
I0510 16:30:45.220772 39557 net.cpp:198] conv3_3_ex needs backward computation.
I0510 16:30:45.220775 39557 net.cpp:198] res3_2_res3_2_0_split needs backward computation.
I0510 16:30:45.220778 39557 net.cpp:198] res3_2 needs backward computation.
I0510 16:30:45.220782 39557 net.cpp:198] conv3_2_em/scale needs backward computation.
I0510 16:30:45.220784 39557 net.cpp:198] conv3_2_em/bn needs backward computation.
I0510 16:30:45.220788 39557 net.cpp:198] conv3_2_em needs backward computation.
I0510 16:30:45.220790 39557 net.cpp:198] relu3_2_dw needs backward computation.
I0510 16:30:45.220793 39557 net.cpp:198] conv3_2_dw/scale needs backward computation.
I0510 16:30:45.220795 39557 net.cpp:198] conv3_2_dw/bn needs backward computation.
I0510 16:30:45.220798 39557 net.cpp:198] conv3_2_dw needs backward computation.
I0510 16:30:45.220801 39557 net.cpp:198] relu3_2_ex needs backward computation.
I0510 16:30:45.220804 39557 net.cpp:198] conv3_2_ex/scale needs backward computation.
I0510 16:30:45.220808 39557 net.cpp:198] conv3_2_ex/bn needs backward computation.
I0510 16:30:45.220809 39557 net.cpp:198] conv3_2_ex needs backward computation.
I0510 16:30:45.220813 39557 net.cpp:198] res3_1_res3_1_0_split needs backward computation.
I0510 16:30:45.220816 39557 net.cpp:198] res3_1 needs backward computation.
I0510 16:30:45.220827 39557 net.cpp:198] conv3_1_em/scale needs backward computation.
I0510 16:30:45.220830 39557 net.cpp:198] conv3_1_em/bn needs backward computation.
I0510 16:30:45.220834 39557 net.cpp:198] conv3_1_em needs backward computation.
I0510 16:30:45.220836 39557 net.cpp:198] relu3_1_dw needs backward computation.
I0510 16:30:45.220839 39557 net.cpp:198] conv3_1_dw/scale needs backward computation.
I0510 16:30:45.220842 39557 net.cpp:198] conv3_1_dw/bn needs backward computation.
I0510 16:30:45.220845 39557 net.cpp:198] conv3_1_dw needs backward computation.
I0510 16:30:45.220849 39557 net.cpp:198] relu3_1_ex needs backward computation.
I0510 16:30:45.220862 39557 net.cpp:198] conv3_1_ex/scale needs backward computation.
I0510 16:30:45.220866 39557 net.cpp:198] conv3_1_ex/bn needs backward computation.
I0510 16:30:45.220868 39557 net.cpp:198] conv3_1_ex needs backward computation.
I0510 16:30:45.220872 39557 net.cpp:198] conv3_em_conv3_em/scale_0_split needs backward computation.
I0510 16:30:45.220876 39557 net.cpp:198] conv3_em/scale needs backward computation.
I0510 16:30:45.220880 39557 net.cpp:198] conv3_em/bn needs backward computation.
I0510 16:30:45.220882 39557 net.cpp:198] conv3_em needs backward computation.
I0510 16:30:45.220885 39557 net.cpp:198] relu3_dw needs backward computation.
I0510 16:30:45.220896 39557 net.cpp:198] conv3_dw/scale needs backward computation.
I0510 16:30:45.220898 39557 net.cpp:198] conv3_dw/bn needs backward computation.
I0510 16:30:45.220901 39557 net.cpp:198] conv3_dw needs backward computation.
I0510 16:30:45.220904 39557 net.cpp:198] relu3_ex needs backward computation.
I0510 16:30:45.220907 39557 net.cpp:198] conv3_ex/scale needs backward computation.
I0510 16:30:45.220911 39557 net.cpp:198] conv3_ex/bn needs backward computation.
I0510 16:30:45.220913 39557 net.cpp:198] conv3_ex needs backward computation.
I0510 16:30:45.220916 39557 net.cpp:198] res2_4 needs backward computation.
I0510 16:30:45.220921 39557 net.cpp:198] conv2_4_em/scale needs backward computation.
I0510 16:30:45.220923 39557 net.cpp:198] conv2_4_em/bn needs backward computation.
I0510 16:30:45.220927 39557 net.cpp:198] conv2_4_em needs backward computation.
I0510 16:30:45.220930 39557 net.cpp:198] relu2_4_dw needs backward computation.
I0510 16:30:45.220933 39557 net.cpp:198] conv2_4_dw/scale needs backward computation.
I0510 16:30:45.220937 39557 net.cpp:198] conv2_4_dw/bn needs backward computation.
I0510 16:30:45.220939 39557 net.cpp:198] conv2_4_dw needs backward computation.
I0510 16:30:45.220942 39557 net.cpp:198] relu2_4_ex needs backward computation.
I0510 16:30:45.220945 39557 net.cpp:198] conv2_4_ex/scale needs backward computation.
I0510 16:30:45.220949 39557 net.cpp:198] conv2_4_ex/bn needs backward computation.
I0510 16:30:45.220952 39557 net.cpp:198] conv2_4_ex needs backward computation.
I0510 16:30:45.220955 39557 net.cpp:198] res2_3_res2_3_0_split needs backward computation.
I0510 16:30:45.220959 39557 net.cpp:198] res2_3 needs backward computation.
I0510 16:30:45.220963 39557 net.cpp:198] conv2_3_em/scale needs backward computation.
I0510 16:30:45.220966 39557 net.cpp:198] conv2_3_em/bn needs backward computation.
I0510 16:30:45.220969 39557 net.cpp:198] conv2_3_em needs backward computation.
I0510 16:30:45.220973 39557 net.cpp:198] relu2_3_dw needs backward computation.
I0510 16:30:45.220975 39557 net.cpp:198] conv2_3_dw/scale needs backward computation.
I0510 16:30:45.220978 39557 net.cpp:198] conv2_3_dw/bn needs backward computation.
I0510 16:30:45.220981 39557 net.cpp:198] conv2_3_dw needs backward computation.
I0510 16:30:45.220985 39557 net.cpp:198] relu2_3_ex needs backward computation.
I0510 16:30:45.220988 39557 net.cpp:198] conv2_3_ex/scale needs backward computation.
I0510 16:30:45.220991 39557 net.cpp:198] conv2_3_ex/bn needs backward computation.
I0510 16:30:45.220994 39557 net.cpp:198] conv2_3_ex needs backward computation.
I0510 16:30:45.220999 39557 net.cpp:198] res2_2_res2_2_0_split needs backward computation.
I0510 16:30:45.221001 39557 net.cpp:198] res2_2 needs backward computation.
I0510 16:30:45.221010 39557 net.cpp:198] conv2_2_em/scale needs backward computation.
I0510 16:30:45.221014 39557 net.cpp:198] conv2_2_em/bn needs backward computation.
I0510 16:30:45.221016 39557 net.cpp:198] conv2_2_em needs backward computation.
I0510 16:30:45.221019 39557 net.cpp:198] relu2_2_dw needs backward computation.
I0510 16:30:45.221022 39557 net.cpp:198] conv2_2_dw/scale needs backward computation.
I0510 16:30:45.221025 39557 net.cpp:198] conv2_2_dw/bn needs backward computation.
I0510 16:30:45.221029 39557 net.cpp:198] conv2_2_dw needs backward computation.
I0510 16:30:45.221032 39557 net.cpp:198] relu2_2_ex needs backward computation.
I0510 16:30:45.221035 39557 net.cpp:198] conv2_2_ex/scale needs backward computation.
I0510 16:30:45.221038 39557 net.cpp:198] conv2_2_ex/bn needs backward computation.
I0510 16:30:45.221041 39557 net.cpp:198] conv2_2_ex needs backward computation.
I0510 16:30:45.221045 39557 net.cpp:198] res2_1_res2_1_0_split needs backward computation.
I0510 16:30:45.221050 39557 net.cpp:198] res2_1 needs backward computation.
I0510 16:30:45.221055 39557 net.cpp:198] conv2_1_em/scale needs backward computation.
I0510 16:30:45.221057 39557 net.cpp:198] conv2_1_em/bn needs backward computation.
I0510 16:30:45.221060 39557 net.cpp:198] conv2_1_em needs backward computation.
I0510 16:30:45.221063 39557 net.cpp:198] relu2_1_dw needs backward computation.
I0510 16:30:45.221067 39557 net.cpp:198] conv2_1_dw/scale needs backward computation.
I0510 16:30:45.221071 39557 net.cpp:198] conv2_1_dw/bn needs backward computation.
I0510 16:30:45.221073 39557 net.cpp:198] conv2_1_dw needs backward computation.
I0510 16:30:45.221076 39557 net.cpp:198] relu2_1_ex needs backward computation.
I0510 16:30:45.221081 39557 net.cpp:198] conv2_1_ex/scale needs backward computation.
I0510 16:30:45.221083 39557 net.cpp:198] conv2_1_ex/bn needs backward computation.
I0510 16:30:45.221086 39557 net.cpp:198] conv2_1_ex needs backward computation.
I0510 16:30:45.221091 39557 net.cpp:198] conv2_em_conv2_em/scale_0_split needs backward computation.
I0510 16:30:45.221093 39557 net.cpp:198] conv2_em/scale needs backward computation.
I0510 16:30:45.221096 39557 net.cpp:198] conv2_em/bn needs backward computation.
I0510 16:30:45.221099 39557 net.cpp:198] conv2_em needs backward computation.
I0510 16:30:45.221103 39557 net.cpp:198] relu2_dw needs backward computation.
I0510 16:30:45.221107 39557 net.cpp:198] conv2_dw/scale needs backward computation.
I0510 16:30:45.221109 39557 net.cpp:198] conv2_dw/bn needs backward computation.
I0510 16:30:45.221112 39557 net.cpp:198] conv2_dw needs backward computation.
I0510 16:30:45.221115 39557 net.cpp:198] relu2_ex needs backward computation.
I0510 16:30:45.221118 39557 net.cpp:198] conv2_ex/scale needs backward computation.
I0510 16:30:45.221122 39557 net.cpp:198] conv2_ex/bn needs backward computation.
I0510 16:30:45.221124 39557 net.cpp:198] conv2_ex needs backward computation.
I0510 16:30:45.221128 39557 net.cpp:198] relu1_dw needs backward computation.
I0510 16:30:45.221132 39557 net.cpp:198] conv1_dw/scale needs backward computation.
I0510 16:30:45.221148 39557 net.cpp:198] conv1_dw/bn needs backward computation.
I0510 16:30:45.221151 39557 net.cpp:198] conv1_dw needs backward computation.
I0510 16:30:45.221154 39557 net.cpp:198] relu1 needs backward computation.
I0510 16:30:45.221158 39557 net.cpp:198] conv1/scale needs backward computation.
I0510 16:30:45.221160 39557 net.cpp:198] conv1/bn needs backward computation.
I0510 16:30:45.221163 39557 net.cpp:198] conv1 needs backward computation.
I0510 16:30:45.221168 39557 net.cpp:200] label_data_1_split does not need backward computation.
I0510 16:30:45.221171 39557 net.cpp:200] data does not need backward computation.
I0510 16:30:45.221174 39557 net.cpp:242] This network produces output softmax_loss
I0510 16:30:45.221280 39557 net.cpp:255] Network initialization done.
I0510 16:30:45.222029 39557 solver.cpp:56] Solver scaffolding done.
I0510 16:30:45.235213 39557 caffe.cpp:248] Starting Optimization
I0510 16:30:45.909693 39624 upgrade_proto.cpp:77] Attempting to upgrade batch norm layers using deprecated params: MobileFaceNet_1.prototxt
I0510 16:30:45.909775 39624 upgrade_proto.cpp:80] Successfully upgraded batch norm layers using deprecated params.
I0510 16:30:45.911074 39624 image_data_layer.cpp:38] Opening file /data/vggface2/VGGFACE_train_list-112X96-add.txt
I0510 16:30:46.814100 39624 image_data_layer.cpp:53] Shuffling data
I0510 16:30:47.051571 39624 image_data_layer.cpp:63] A total of 3306982 images.
I0510 16:30:47.057852 39624 image_data_layer.cpp:90] output data size: 64,3,112,96
I0510 16:30:48.294826 39557 solver.cpp:272] Solving Caffe_MobileFaceNet
I0510 16:30:48.294898 39557 solver.cpp:273] Learning Rate Policy: multistep
I0510 16:30:51.666826 39557 solver.cpp:218] Iteration 0 (0 iter/s, 3.3703s/100 iters), loss = 12.536
I0510 16:30:51.666883 39557 solver.cpp:237]     Train net output #0: softmax_loss = 12.536 (* 1 = 12.536 loss)
I0510 16:30:51.666975 39557 sgd_solver.cpp:105] Iteration 0, lr = 0.1
I0510 16:31:18.041234 39557 solver.cpp:218] Iteration 100 (3.79162 iter/s, 26.374s/100 iters), loss = 11.1028
I0510 16:31:18.041404 39557 solver.cpp:237]     Train net output #0: softmax_loss = 11.1028 (* 1 = 11.1028 loss)
I0510 16:31:18.041419 39557 sgd_solver.cpp:105] Iteration 100, lr = 0.1
I0510 16:31:44.345535 39557 solver.cpp:218] Iteration 200 (3.80174 iter/s, 26.3038s/100 iters), loss = 11.3258
I0510 16:31:44.345583 39557 solver.cpp:237]     Train net output #0: softmax_loss = 11.3258 (* 1 = 11.3258 loss)
I0510 16:31:44.345597 39557 sgd_solver.cpp:105] Iteration 200, lr = 0.1
I0510 16:32:10.824427 39557 solver.cpp:218] Iteration 300 (3.77665 iter/s, 26.4785s/100 iters), loss = 10.457
I0510 16:32:10.824596 39557 solver.cpp:237]     Train net output #0: softmax_loss = 10.457 (* 1 = 10.457 loss)
I0510 16:32:10.824615 39557 sgd_solver.cpp:105] Iteration 300, lr = 0.1
I0510 16:32:37.288573 39557 solver.cpp:218] Iteration 400 (3.77877 iter/s, 26.4637s/100 iters), loss = 10.2858
I0510 16:32:37.288632 39557 solver.cpp:237]     Train net output #0: softmax_loss = 10.2858 (* 1 = 10.2858 loss)
I0510 16:32:37.288646 39557 sgd_solver.cpp:105] Iteration 400, lr = 0.1
I0510 16:33:03.697141 39557 solver.cpp:218] Iteration 500 (3.7867 iter/s, 26.4082s/100 iters), loss = 10.0853
I0510 16:33:03.697351 39557 solver.cpp:237]     Train net output #0: softmax_loss = 10.0853 (* 1 = 10.0853 loss)
I0510 16:33:03.697367 39557 sgd_solver.cpp:105] Iteration 500, lr = 0.1
I0510 16:33:30.175662 39557 solver.cpp:218] Iteration 600 (3.77671 iter/s, 26.478s/100 iters), loss = 9.57291
I0510 16:33:30.175717 39557 solver.cpp:237]     Train net output #0: softmax_loss = 9.57291 (* 1 = 9.57291 loss)
I0510 16:33:30.175730 39557 sgd_solver.cpp:105] Iteration 600, lr = 0.1
I0510 16:33:56.707626 39557 solver.cpp:218] Iteration 700 (3.76908 iter/s, 26.5317s/100 iters), loss = 9.7992
I0510 16:33:56.707805 39557 solver.cpp:237]     Train net output #0: softmax_loss = 9.7992 (* 1 = 9.7992 loss)
I0510 16:33:56.707821 39557 sgd_solver.cpp:105] Iteration 700, lr = 0.1
I0510 16:34:23.292376 39557 solver.cpp:218] Iteration 800 (3.76161 iter/s, 26.5843s/100 iters), loss = 9.84886
I0510 16:34:23.292430 39557 solver.cpp:237]     Train net output #0: softmax_loss = 9.84886 (* 1 = 9.84886 loss)
I0510 16:34:23.292443 39557 sgd_solver.cpp:105] Iteration 800, lr = 0.1
I0510 16:34:49.836494 39557 solver.cpp:218] Iteration 900 (3.76735 iter/s, 26.5439s/100 iters), loss = 9.54274
I0510 16:34:49.836722 39557 solver.cpp:237]     Train net output #0: softmax_loss = 9.54274 (* 1 = 9.54274 loss)
I0510 16:34:49.836738 39557 sgd_solver.cpp:105] Iteration 900, lr = 0.1
I0510 16:35:16.403558 39557 solver.cpp:218] Iteration 1000 (3.76412 iter/s, 26.5666s/100 iters), loss = 10.0234
I0510 16:35:16.403620 39557 solver.cpp:237]     Train net output #0: softmax_loss = 10.0234 (* 1 = 10.0234 loss)
I0510 16:35:16.403632 39557 sgd_solver.cpp:105] Iteration 1000, lr = 0.1
I0510 16:35:42.923848 39557 solver.cpp:218] Iteration 1100 (3.77073 iter/s, 26.52s/100 iters), loss = 9.96777
I0510 16:35:42.924072 39557 solver.cpp:237]     Train net output #0: softmax_loss = 9.96777 (* 1 = 9.96777 loss)
I0510 16:35:42.924089 39557 sgd_solver.cpp:105] Iteration 1100, lr = 0.1
I0510 16:36:09.408329 39557 solver.cpp:218] Iteration 1200 (3.77585 iter/s, 26.4841s/100 iters), loss = 9.5465
I0510 16:36:09.408385 39557 solver.cpp:237]     Train net output #0: softmax_loss = 9.5465 (* 1 = 9.5465 loss)
I0510 16:36:09.408398 39557 sgd_solver.cpp:105] Iteration 1200, lr = 0.1
I0510 16:36:35.963322 39557 solver.cpp:218] Iteration 1300 (3.7658 iter/s, 26.5548s/100 iters), loss = 10.5895
I0510 16:36:35.963529 39557 solver.cpp:237]     Train net output #0: softmax_loss = 10.5895 (* 1 = 10.5895 loss)
I0510 16:36:35.963546 39557 sgd_solver.cpp:105] Iteration 1300, lr = 0.1
I0510 16:37:02.492852 39557 solver.cpp:218] Iteration 1400 (3.76944 iter/s, 26.5292s/100 iters), loss = 10.3743
I0510 16:37:02.492910 39557 solver.cpp:237]     Train net output #0: softmax_loss = 10.3743 (* 1 = 10.3743 loss)
I0510 16:37:02.492924 39557 sgd_solver.cpp:105] Iteration 1400, lr = 0.1
I0510 16:37:29.104971 39557 solver.cpp:218] Iteration 1500 (3.75772 iter/s, 26.6119s/100 iters), loss = 10.1805
I0510 16:37:29.105199 39557 solver.cpp:237]     Train net output #0: softmax_loss = 10.1805 (* 1 = 10.1805 loss)
I0510 16:37:29.105216 39557 sgd_solver.cpp:105] Iteration 1500, lr = 0.1
I0510 16:37:55.715894 39557 solver.cpp:218] Iteration 1600 (3.75791 iter/s, 26.6105s/100 iters), loss = 10.3606
I0510 16:37:55.715952 39557 solver.cpp:237]     Train net output #0: softmax_loss = 10.3606 (* 1 = 10.3606 loss)
I0510 16:37:55.715966 39557 sgd_solver.cpp:105] Iteration 1600, lr = 0.1
I0510 16:38:22.360563 39557 solver.cpp:218] Iteration 1700 (3.75313 iter/s, 26.6444s/100 iters), loss = 11.2666
I0510 16:38:22.360713 39557 solver.cpp:237]     Train net output #0: softmax_loss = 11.2666 (* 1 = 11.2666 loss)
I0510 16:38:22.360730 39557 sgd_solver.cpp:105] Iteration 1700, lr = 0.1
I0510 16:38:48.980559 39557 solver.cpp:218] Iteration 1800 (3.75662 iter/s, 26.6197s/100 iters), loss = 12.1168
I0510 16:38:48.980619 39557 solver.cpp:237]     Train net output #0: softmax_loss = 12.1168 (* 1 = 12.1168 loss)
I0510 16:38:48.980636 39557 sgd_solver.cpp:105] Iteration 1800, lr = 0.1
I0510 16:39:15.759759 39557 solver.cpp:218] Iteration 1900 (3.73427 iter/s, 26.779s/100 iters), loss = 11.5123
I0510 16:39:15.759928 39557 solver.cpp:237]     Train net output #0: softmax_loss = 11.5123 (* 1 = 11.5123 loss)
I0510 16:39:15.759948 39557 sgd_solver.cpp:105] Iteration 1900, lr = 0.1
I0510 16:39:17.944533 39624 blocking_queue.cpp:49] Waiting for data
I0510 16:39:46.351075 39557 solver.cpp:218] Iteration 2000 (3.26894 iter/s, 30.591s/100 iters), loss = 11.9498
I0510 16:39:46.351254 39557 solver.cpp:237]     Train net output #0: softmax_loss = 11.9498 (* 1 = 11.9498 loss)
I0510 16:39:46.351270 39557 sgd_solver.cpp:105] Iteration 2000, lr = 0.1
I0510 16:40:12.875560 39557 solver.cpp:218] Iteration 2100 (3.77015 iter/s, 26.5241s/100 iters), loss = 11.6036
I0510 16:40:12.875618 39557 solver.cpp:237]     Train net output #0: softmax_loss = 11.6036 (* 1 = 11.6036 loss)
I0510 16:40:12.875633 39557 sgd_solver.cpp:105] Iteration 2100, lr = 0.1
I0510 16:40:39.509877 39557 solver.cpp:218] Iteration 2200 (3.75459 iter/s, 26.6341s/100 iters), loss = 11.8161
I0510 16:40:39.511127 39557 solver.cpp:237]     Train net output #0: softmax_loss = 11.8161 (* 1 = 11.8161 loss)
I0510 16:40:39.511195 39557 sgd_solver.cpp:105] Iteration 2200, lr = 0.1
I0510 16:41:06.056427 39557 solver.cpp:218] Iteration 2300 (3.76717 iter/s, 26.5451s/100 iters), loss = 11.575
I0510 16:41:06.056491 39557 solver.cpp:237]     Train net output #0: softmax_loss = 11.575 (* 1 = 11.575 loss)
I0510 16:41:06.056504 39557 sgd_solver.cpp:105] Iteration 2300, lr = 0.1
I0510 16:41:32.691660 39557 solver.cpp:218] Iteration 2400 (3.75446 iter/s, 26.635s/100 iters), loss = 11.0001
I0510 16:41:32.691880 39557 solver.cpp:237]     Train net output #0: softmax_loss = 11.0001 (* 1 = 11.0001 loss)
I0510 16:41:32.691897 39557 sgd_solver.cpp:105] Iteration 2400, lr = 0.1
I0510 16:41:59.240311 39557 solver.cpp:218] Iteration 2500 (3.76672 iter/s, 26.5483s/100 iters), loss = 11.7117
I0510 16:41:59.240360 39557 solver.cpp:237]     Train net output #0: softmax_loss = 11.7117 (* 1 = 11.7117 loss)
I0510 16:41:59.240372 39557 sgd_solver.cpp:105] Iteration 2500, lr = 0.1
I0510 16:42:25.875320 39557 solver.cpp:218] Iteration 2600 (3.75449 iter/s, 26.6348s/100 iters), loss = 11.8325
I0510 16:42:25.875519 39557 solver.cpp:237]     Train net output #0: softmax_loss = 11.8325 (* 1 = 11.8325 loss)
I0510 16:42:25.875533 39557 sgd_solver.cpp:105] Iteration 2600, lr = 0.1
I0510 16:42:52.400277 39557 solver.cpp:218] Iteration 2700 (3.77009 iter/s, 26.5246s/100 iters), loss = 12.3274
I0510 16:42:52.400353 39557 solver.cpp:237]     Train net output #0: softmax_loss = 12.3274 (* 1 = 12.3274 loss)
I0510 16:42:52.400367 39557 sgd_solver.cpp:105] Iteration 2700, lr = 0.1
I0510 16:43:19.058145 39557 solver.cpp:218] Iteration 2800 (3.75127 iter/s, 26.6576s/100 iters), loss = 12.0319
I0510 16:43:19.058571 39557 solver.cpp:237]     Train net output #0: softmax_loss = 12.0319 (* 1 = 12.0319 loss)
I0510 16:43:19.058588 39557 sgd_solver.cpp:105] Iteration 2800, lr = 0.1
I0510 16:43:45.776252 39557 solver.cpp:218] Iteration 2900 (3.74286 iter/s, 26.7175s/100 iters), loss = 11.0202
I0510 16:43:45.776300 39557 solver.cpp:237]     Train net output #0: softmax_loss = 11.0202 (* 1 = 11.0202 loss)
I0510 16:43:45.776314 39557 sgd_solver.cpp:105] Iteration 2900, lr = 0.1
I0510 16:44:12.380105 39557 solver.cpp:218] Iteration 3000 (3.75889 iter/s, 26.6036s/100 iters), loss = 11.7628
I0510 16:44:12.380288 39557 solver.cpp:237]     Train net output #0: softmax_loss = 11.7628 (* 1 = 11.7628 loss)
I0510 16:44:12.380304 39557 sgd_solver.cpp:105] Iteration 3000, lr = 0.1
I0510 16:44:39.005744 39557 solver.cpp:218] Iteration 3100 (3.75583 iter/s, 26.6253s/100 iters), loss = 11.1251
I0510 16:44:39.005818 39557 solver.cpp:237]     Train net output #0: softmax_loss = 11.1251 (* 1 = 11.1251 loss)
I0510 16:44:39.005831 39557 sgd_solver.cpp:105] Iteration 3100, lr = 0.1
I0510 16:45:05.678087 39557 solver.cpp:218] Iteration 3200 (3.74924 iter/s, 26.6721s/100 iters), loss = 10.8438
I0510 16:45:05.678274 39557 solver.cpp:237]     Train net output #0: softmax_loss = 10.8438 (* 1 = 10.8438 loss)
I0510 16:45:05.678292 39557 sgd_solver.cpp:105] Iteration 3200, lr = 0.1
I0510 16:45:32.340165 39557 solver.cpp:218] Iteration 3300 (3.75064 iter/s, 26.6621s/100 iters), loss = 10.6864
I0510 16:45:32.340224 39557 solver.cpp:237]     Train net output #0: softmax_loss = 10.6864 (* 1 = 10.6864 loss)
I0510 16:45:32.340237 39557 sgd_solver.cpp:105] Iteration 3300, lr = 0.1
I0510 16:45:58.911054 39557 solver.cpp:218] Iteration 3400 (3.76349 iter/s, 26.5711s/100 iters), loss = 11.4578
I0510 16:45:58.911260 39557 solver.cpp:237]     Train net output #0: softmax_loss = 11.4578 (* 1 = 11.4578 loss)
I0510 16:45:58.911276 39557 sgd_solver.cpp:105] Iteration 3400, lr = 0.1
I0510 16:46:25.494945 39557 solver.cpp:218] Iteration 3500 (3.76167 iter/s, 26.5839s/100 iters), loss = 11.0085
I0510 16:46:25.495002 39557 solver.cpp:237]     Train net output #0: softmax_loss = 11.0085 (* 1 = 11.0085 loss)
I0510 16:46:25.495014 39557 sgd_solver.cpp:105] Iteration 3500, lr = 0.1
I0510 16:46:52.156911 39557 solver.cpp:218] Iteration 3600 (3.75064 iter/s, 26.6621s/100 iters), loss = 11.3608
I0510 16:46:52.157094 39557 solver.cpp:237]     Train net output #0: softmax_loss = 11.3608 (* 1 = 11.3608 loss)
I0510 16:46:52.157107 39557 sgd_solver.cpp:105] Iteration 3600, lr = 0.1
I0510 16:47:18.816108 39557 solver.cpp:218] Iteration 3700 (3.75105 iter/s, 26.6592s/100 iters), loss = 10.349
I0510 16:47:18.816176 39557 solver.cpp:237]     Train net output #0: softmax_loss = 10.349 (* 1 = 10.349 loss)
I0510 16:47:18.816190 39557 sgd_solver.cpp:105] Iteration 3700, lr = 0.1
I0510 16:47:45.419139 39557 solver.cpp:218] Iteration 3800 (3.75895 iter/s, 26.6032s/100 iters), loss = 10.4196
I0510 16:47:45.419381 39557 solver.cpp:237]     Train net output #0: softmax_loss = 10.4196 (* 1 = 10.4196 loss)
I0510 16:47:45.419399 39557 sgd_solver.cpp:105] Iteration 3800, lr = 0.1
I0510 16:48:12.057561 39557 solver.cpp:218] Iteration 3900 (3.75398 iter/s, 26.6384s/100 iters), loss = 11.3052
I0510 16:48:12.057607 39557 solver.cpp:237]     Train net output #0: softmax_loss = 11.3052 (* 1 = 11.3052 loss)
I0510 16:48:12.057639 39557 sgd_solver.cpp:105] Iteration 3900, lr = 0.1
I0510 16:48:38.644058 39557 solver.cpp:218] Iteration 4000 (3.76129 iter/s, 26.5866s/100 iters), loss = 11.4735
I0510 16:48:38.644264 39557 solver.cpp:237]     Train net output #0: softmax_loss = 11.4735 (* 1 = 11.4735 loss)
I0510 16:48:38.644289 39557 sgd_solver.cpp:105] Iteration 4000, lr = 0.1
I0510 16:49:05.280514 39557 solver.cpp:218] Iteration 4100 (3.75426 iter/s, 26.6364s/100 iters), loss = 10.4317
I0510 16:49:05.280572 39557 solver.cpp:237]     Train net output #0: softmax_loss = 10.4317 (* 1 = 10.4317 loss)
I0510 16:49:05.280586 39557 sgd_solver.cpp:105] Iteration 4100, lr = 0.1
I0510 16:49:31.964890 39557 solver.cpp:218] Iteration 4200 (3.7475 iter/s, 26.6844s/100 iters), loss = 10.8047
I0510 16:49:31.965080 39557 solver.cpp:237]     Train net output #0: softmax_loss = 10.8047 (* 1 = 10.8047 loss)
I0510 16:49:31.965095 39557 sgd_solver.cpp:105] Iteration 4200, lr = 0.1
I0510 16:49:58.566851 39557 solver.cpp:218] Iteration 4300 (3.75913 iter/s, 26.6019s/100 iters), loss = 11.5825
I0510 16:49:58.566913 39557 solver.cpp:237]     Train net output #0: softmax_loss = 11.5825 (* 1 = 11.5825 loss)
I0510 16:49:58.566928 39557 sgd_solver.cpp:105] Iteration 4300, lr = 0.1
I0510 16:50:25.269138 39557 solver.cpp:218] Iteration 4400 (3.74499 iter/s, 26.7023s/100 iters), loss = 10.2589
I0510 16:50:25.269315 39557 solver.cpp:237]     Train net output #0: softmax_loss = 10.2589 (* 1 = 10.2589 loss)
I0510 16:50:25.269331 39557 sgd_solver.cpp:105] Iteration 4400, lr = 0.1
I0510 16:50:51.939548 39557 solver.cpp:218] Iteration 4500 (3.74948 iter/s, 26.6703s/100 iters), loss = 10.7184
I0510 16:50:51.939604 39557 solver.cpp:237]     Train net output #0: softmax_loss = 10.7184 (* 1 = 10.7184 loss)
I0510 16:50:51.939616 39557 sgd_solver.cpp:105] Iteration 4500, lr = 0.1
I0510 16:51:18.516333 39557 solver.cpp:218] Iteration 4600 (3.76268 iter/s, 26.5768s/100 iters), loss = 10.7453
I0510 16:51:18.516542 39557 solver.cpp:237]     Train net output #0: softmax_loss = 10.7453 (* 1 = 10.7453 loss)
I0510 16:51:18.516558 39557 sgd_solver.cpp:105] Iteration 4600, lr = 0.1
I0510 16:51:45.222190 39557 solver.cpp:218] Iteration 4700 (3.74451 iter/s, 26.7057s/100 iters), loss = 10.2529
I0510 16:51:45.222251 39557 solver.cpp:237]     Train net output #0: softmax_loss = 10.2529 (* 1 = 10.2529 loss)
I0510 16:51:45.222270 39557 sgd_solver.cpp:105] Iteration 4700, lr = 0.1
I0510 16:52:11.952262 39557 solver.cpp:218] Iteration 4800 (3.7411 iter/s, 26.7301s/100 iters), loss = 9.89344
I0510 16:52:11.952448 39557 solver.cpp:237]     Train net output #0: softmax_loss = 9.89344 (* 1 = 9.89344 loss)
I0510 16:52:11.952464 39557 sgd_solver.cpp:105] Iteration 4800, lr = 0.1
I0510 16:52:38.637040 39557 solver.cpp:218] Iteration 4900 (3.74748 iter/s, 26.6846s/100 iters), loss = 10.9322
I0510 16:52:38.637100 39557 solver.cpp:237]     Train net output #0: softmax_loss = 10.9322 (* 1 = 10.9322 loss)
I0510 16:52:38.637117 39557 sgd_solver.cpp:105] Iteration 4900, lr = 0.1
I0510 16:53:05.081687 39557 solver.cpp:447] Snapshotting to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_5000.caffemodel
I0510 16:53:05.184801 39557 sgd_solver.cpp:273] Snapshotting solver state to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_5000.solverstate
I0510 16:53:05.534158 39557 solver.cpp:218] Iteration 5000 (3.71787 iter/s, 26.8971s/100 iters), loss = 9.4616
I0510 16:53:05.534219 39557 solver.cpp:237]     Train net output #0: softmax_loss = 9.4616 (* 1 = 9.4616 loss)
I0510 16:53:05.534229 39557 sgd_solver.cpp:105] Iteration 5000, lr = 0.1
I0510 16:53:32.129828 39557 solver.cpp:218] Iteration 5100 (3.76001 iter/s, 26.5956s/100 iters), loss = 10.4973
I0510 16:53:32.129889 39557 solver.cpp:237]     Train net output #0: softmax_loss = 10.4973 (* 1 = 10.4973 loss)
I0510 16:53:32.129905 39557 sgd_solver.cpp:105] Iteration 5100, lr = 0.1
I0510 16:53:58.741194 39557 solver.cpp:218] Iteration 5200 (3.7578 iter/s, 26.6113s/100 iters), loss = 9.99246
I0510 16:53:58.741441 39557 solver.cpp:237]     Train net output #0: softmax_loss = 9.99246 (* 1 = 9.99246 loss)
I0510 16:53:58.741457 39557 sgd_solver.cpp:105] Iteration 5200, lr = 0.1
I0510 16:54:25.314086 39557 solver.cpp:218] Iteration 5300 (3.76326 iter/s, 26.5727s/100 iters), loss = 10.6802
I0510 16:54:25.314147 39557 solver.cpp:237]     Train net output #0: softmax_loss = 10.6802 (* 1 = 10.6802 loss)
I0510 16:54:25.314160 39557 sgd_solver.cpp:105] Iteration 5300, lr = 0.1
I0510 16:54:51.925781 39557 solver.cpp:218] Iteration 5400 (3.75775 iter/s, 26.6117s/100 iters), loss = 9.94882
I0510 16:54:51.926023 39557 solver.cpp:237]     Train net output #0: softmax_loss = 9.94882 (* 1 = 9.94882 loss)
I0510 16:54:51.926039 39557 sgd_solver.cpp:105] Iteration 5400, lr = 0.1
I0510 16:55:18.570852 39557 solver.cpp:218] Iteration 5500 (3.75307 iter/s, 26.6448s/100 iters), loss = 9.68468
I0510 16:55:18.570915 39557 solver.cpp:237]     Train net output #0: softmax_loss = 9.68468 (* 1 = 9.68468 loss)
I0510 16:55:18.570927 39557 sgd_solver.cpp:105] Iteration 5500, lr = 0.1
I0510 16:55:45.342303 39557 solver.cpp:218] Iteration 5600 (3.73533 iter/s, 26.7714s/100 iters), loss = 9.68579
I0510 16:55:45.342499 39557 solver.cpp:237]     Train net output #0: softmax_loss = 9.68579 (* 1 = 9.68579 loss)
I0510 16:55:45.342517 39557 sgd_solver.cpp:105] Iteration 5600, lr = 0.1
I0510 16:56:12.033766 39557 solver.cpp:218] Iteration 5700 (3.74654 iter/s, 26.6913s/100 iters), loss = 9.29763
I0510 16:56:12.033824 39557 solver.cpp:237]     Train net output #0: softmax_loss = 9.29763 (* 1 = 9.29763 loss)
I0510 16:56:12.033836 39557 sgd_solver.cpp:105] Iteration 5700, lr = 0.1
I0510 16:56:38.714071 39557 solver.cpp:218] Iteration 5800 (3.74809 iter/s, 26.6802s/100 iters), loss = 9.0948
I0510 16:56:38.714279 39557 solver.cpp:237]     Train net output #0: softmax_loss = 9.0948 (* 1 = 9.0948 loss)
I0510 16:56:38.714294 39557 sgd_solver.cpp:105] Iteration 5800, lr = 0.1
I0510 16:57:05.351402 39557 solver.cpp:218] Iteration 5900 (3.75416 iter/s, 26.6371s/100 iters), loss = 8.78136
I0510 16:57:05.351464 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.78136 (* 1 = 8.78136 loss)
I0510 16:57:05.351477 39557 sgd_solver.cpp:105] Iteration 5900, lr = 0.1
I0510 16:57:32.009438 39557 solver.cpp:218] Iteration 6000 (3.75123 iter/s, 26.658s/100 iters), loss = 9.16174
I0510 16:57:32.011299 39557 solver.cpp:237]     Train net output #0: softmax_loss = 9.16174 (* 1 = 9.16174 loss)
I0510 16:57:32.011318 39557 sgd_solver.cpp:105] Iteration 6000, lr = 0.1
I0510 16:57:58.779955 39557 solver.cpp:218] Iteration 6100 (3.73571 iter/s, 26.7686s/100 iters), loss = 9.25479
I0510 16:57:58.780017 39557 solver.cpp:237]     Train net output #0: softmax_loss = 9.25479 (* 1 = 9.25479 loss)
I0510 16:57:58.780031 39557 sgd_solver.cpp:105] Iteration 6100, lr = 0.1
I0510 16:58:25.409854 39557 solver.cpp:218] Iteration 6200 (3.75519 iter/s, 26.6298s/100 iters), loss = 8.65649
I0510 16:58:25.410032 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.65649 (* 1 = 8.65649 loss)
I0510 16:58:25.410048 39557 sgd_solver.cpp:105] Iteration 6200, lr = 0.1
I0510 16:58:52.017802 39557 solver.cpp:218] Iteration 6300 (3.7583 iter/s, 26.6077s/100 iters), loss = 9.23368
I0510 16:58:52.017868 39557 solver.cpp:237]     Train net output #0: softmax_loss = 9.23368 (* 1 = 9.23368 loss)
I0510 16:58:52.017880 39557 sgd_solver.cpp:105] Iteration 6300, lr = 0.1
I0510 16:59:18.785126 39557 solver.cpp:218] Iteration 6400 (3.73591 iter/s, 26.7672s/100 iters), loss = 9.28853
I0510 16:59:18.785336 39557 solver.cpp:237]     Train net output #0: softmax_loss = 9.28853 (* 1 = 9.28853 loss)
I0510 16:59:18.785353 39557 sgd_solver.cpp:105] Iteration 6400, lr = 0.1
I0510 16:59:45.480495 39557 solver.cpp:218] Iteration 6500 (3.746 iter/s, 26.6951s/100 iters), loss = 9.82306
I0510 16:59:45.480546 39557 solver.cpp:237]     Train net output #0: softmax_loss = 9.82306 (* 1 = 9.82306 loss)
I0510 16:59:45.480559 39557 sgd_solver.cpp:105] Iteration 6500, lr = 0.1
I0510 17:00:12.153787 39557 solver.cpp:218] Iteration 6600 (3.74908 iter/s, 26.6732s/100 iters), loss = 9.21509
I0510 17:00:12.154009 39557 solver.cpp:237]     Train net output #0: softmax_loss = 9.21509 (* 1 = 9.21509 loss)
I0510 17:00:12.154026 39557 sgd_solver.cpp:105] Iteration 6600, lr = 0.1
I0510 17:00:38.834517 39557 solver.cpp:218] Iteration 6700 (3.74806 iter/s, 26.6805s/100 iters), loss = 9.12617
I0510 17:00:38.834579 39557 solver.cpp:237]     Train net output #0: softmax_loss = 9.12617 (* 1 = 9.12617 loss)
I0510 17:00:38.834594 39557 sgd_solver.cpp:105] Iteration 6700, lr = 0.1
I0510 17:01:05.426611 39557 solver.cpp:218] Iteration 6800 (3.76053 iter/s, 26.592s/100 iters), loss = 9.75062
I0510 17:01:05.426820 39557 solver.cpp:237]     Train net output #0: softmax_loss = 9.75062 (* 1 = 9.75062 loss)
I0510 17:01:05.426834 39557 sgd_solver.cpp:105] Iteration 6800, lr = 0.1
I0510 17:01:32.138738 39557 solver.cpp:218] Iteration 6900 (3.74365 iter/s, 26.7119s/100 iters), loss = 9.15429
I0510 17:01:32.138801 39557 solver.cpp:237]     Train net output #0: softmax_loss = 9.15429 (* 1 = 9.15429 loss)
I0510 17:01:32.138814 39557 sgd_solver.cpp:105] Iteration 6900, lr = 0.1
I0510 17:02:00.108433 39557 solver.cpp:218] Iteration 7000 (3.57531 iter/s, 27.9696s/100 iters), loss = 9.12599
I0510 17:02:00.108626 39557 solver.cpp:237]     Train net output #0: softmax_loss = 9.12599 (* 1 = 9.12599 loss)
I0510 17:02:00.108644 39557 sgd_solver.cpp:105] Iteration 7000, lr = 0.1
I0510 17:02:26.649701 39557 solver.cpp:218] Iteration 7100 (3.76775 iter/s, 26.541s/100 iters), loss = 9.39485
I0510 17:02:26.649760 39557 solver.cpp:237]     Train net output #0: softmax_loss = 9.39485 (* 1 = 9.39485 loss)
I0510 17:02:26.649772 39557 sgd_solver.cpp:105] Iteration 7100, lr = 0.1
I0510 17:02:53.214563 39557 solver.cpp:218] Iteration 7200 (3.76439 iter/s, 26.5647s/100 iters), loss = 9.18171
I0510 17:02:53.214758 39557 solver.cpp:237]     Train net output #0: softmax_loss = 9.18171 (* 1 = 9.18171 loss)
I0510 17:02:53.214773 39557 sgd_solver.cpp:105] Iteration 7200, lr = 0.1
I0510 17:03:19.810242 39557 solver.cpp:218] Iteration 7300 (3.76005 iter/s, 26.5954s/100 iters), loss = 10.4542
I0510 17:03:19.810298 39557 solver.cpp:237]     Train net output #0: softmax_loss = 10.4542 (* 1 = 10.4542 loss)
I0510 17:03:19.810328 39557 sgd_solver.cpp:105] Iteration 7300, lr = 0.1
I0510 17:03:46.400959 39557 solver.cpp:218] Iteration 7400 (3.76073 iter/s, 26.5906s/100 iters), loss = 9.3299
I0510 17:03:46.401141 39557 solver.cpp:237]     Train net output #0: softmax_loss = 9.3299 (* 1 = 9.3299 loss)
I0510 17:03:46.401165 39557 sgd_solver.cpp:105] Iteration 7400, lr = 0.1
I0510 17:04:13.064942 39557 solver.cpp:218] Iteration 7500 (3.75041 iter/s, 26.6637s/100 iters), loss = 8.72623
I0510 17:04:13.065001 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.72623 (* 1 = 8.72623 loss)
I0510 17:04:13.065013 39557 sgd_solver.cpp:105] Iteration 7500, lr = 0.1
I0510 17:04:39.658079 39557 solver.cpp:218] Iteration 7600 (3.76039 iter/s, 26.593s/100 iters), loss = 8.70596
I0510 17:04:39.658275 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.70596 (* 1 = 8.70596 loss)
I0510 17:04:39.658290 39557 sgd_solver.cpp:105] Iteration 7600, lr = 0.1
I0510 17:05:06.252243 39557 solver.cpp:218] Iteration 7700 (3.76026 iter/s, 26.5939s/100 iters), loss = 8.99401
I0510 17:05:06.252295 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.99401 (* 1 = 8.99401 loss)
I0510 17:05:06.252308 39557 sgd_solver.cpp:105] Iteration 7700, lr = 0.1
I0510 17:05:32.842778 39557 solver.cpp:218] Iteration 7800 (3.76075 iter/s, 26.5904s/100 iters), loss = 8.90164
I0510 17:05:32.842998 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.90164 (* 1 = 8.90164 loss)
I0510 17:05:32.843014 39557 sgd_solver.cpp:105] Iteration 7800, lr = 0.1
I0510 17:05:59.493358 39557 solver.cpp:218] Iteration 7900 (3.7523 iter/s, 26.6503s/100 iters), loss = 8.90752
I0510 17:05:59.493405 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.90752 (* 1 = 8.90752 loss)
I0510 17:05:59.493418 39557 sgd_solver.cpp:105] Iteration 7900, lr = 0.1
I0510 17:06:26.103828 39557 solver.cpp:218] Iteration 8000 (3.75794 iter/s, 26.6103s/100 iters), loss = 8.73894
I0510 17:06:26.104012 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.73894 (* 1 = 8.73894 loss)
I0510 17:06:26.104028 39557 sgd_solver.cpp:105] Iteration 8000, lr = 0.1
I0510 17:06:52.758244 39557 solver.cpp:218] Iteration 8100 (3.75176 iter/s, 26.6542s/100 iters), loss = 8.73523
I0510 17:06:52.758314 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.73523 (* 1 = 8.73523 loss)
I0510 17:06:52.758329 39557 sgd_solver.cpp:105] Iteration 8100, lr = 0.1
I0510 17:07:19.424427 39557 solver.cpp:218] Iteration 8200 (3.75009 iter/s, 26.666s/100 iters), loss = 8.95334
I0510 17:07:19.424616 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.95334 (* 1 = 8.95334 loss)
I0510 17:07:19.424633 39557 sgd_solver.cpp:105] Iteration 8200, lr = 0.1
I0510 17:07:46.102438 39557 solver.cpp:218] Iteration 8300 (3.74844 iter/s, 26.6777s/100 iters), loss = 8.70306
I0510 17:07:46.102499 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.70306 (* 1 = 8.70306 loss)
I0510 17:07:46.102516 39557 sgd_solver.cpp:105] Iteration 8300, lr = 0.1
I0510 17:08:12.655100 39557 solver.cpp:218] Iteration 8400 (3.76612 iter/s, 26.5525s/100 iters), loss = 8.80588
I0510 17:08:12.655292 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.80588 (* 1 = 8.80588 loss)
I0510 17:08:12.655308 39557 sgd_solver.cpp:105] Iteration 8400, lr = 0.1
I0510 17:08:39.675192 39557 solver.cpp:218] Iteration 8500 (3.70099 iter/s, 27.0198s/100 iters), loss = 9.00144
I0510 17:08:39.675253 39557 solver.cpp:237]     Train net output #0: softmax_loss = 9.00144 (* 1 = 9.00144 loss)
I0510 17:08:39.675266 39557 sgd_solver.cpp:105] Iteration 8500, lr = 0.1
I0510 17:09:06.299837 39557 solver.cpp:218] Iteration 8600 (3.75594 iter/s, 26.6245s/100 iters), loss = 8.41181
I0510 17:09:06.300055 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.41181 (* 1 = 8.41181 loss)
I0510 17:09:06.300071 39557 sgd_solver.cpp:105] Iteration 8600, lr = 0.1
I0510 17:09:32.920612 39557 solver.cpp:218] Iteration 8700 (3.75651 iter/s, 26.6205s/100 iters), loss = 8.76921
I0510 17:09:32.920675 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.76921 (* 1 = 8.76921 loss)
I0510 17:09:32.920691 39557 sgd_solver.cpp:105] Iteration 8700, lr = 0.1
I0510 17:09:59.696470 39557 solver.cpp:218] Iteration 8800 (3.73473 iter/s, 26.7757s/100 iters), loss = 8.59798
I0510 17:09:59.696674 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.59798 (* 1 = 8.59798 loss)
I0510 17:09:59.696688 39557 sgd_solver.cpp:105] Iteration 8800, lr = 0.1
I0510 17:10:26.406139 39557 solver.cpp:218] Iteration 8900 (3.744 iter/s, 26.7094s/100 iters), loss = 8.59151
I0510 17:10:26.406188 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.59151 (* 1 = 8.59151 loss)
I0510 17:10:26.406211 39557 sgd_solver.cpp:105] Iteration 8900, lr = 0.1
I0510 17:10:53.128281 39557 solver.cpp:218] Iteration 9000 (3.74224 iter/s, 26.722s/100 iters), loss = 8.25494
I0510 17:10:53.128463 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.25494 (* 1 = 8.25494 loss)
I0510 17:10:53.128482 39557 sgd_solver.cpp:105] Iteration 9000, lr = 0.1
I0510 17:11:19.806568 39557 solver.cpp:218] Iteration 9100 (3.7484 iter/s, 26.678s/100 iters), loss = 8.62064
I0510 17:11:19.806632 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.62064 (* 1 = 8.62064 loss)
I0510 17:11:19.806645 39557 sgd_solver.cpp:105] Iteration 9100, lr = 0.1
I0510 17:11:46.535131 39557 solver.cpp:218] Iteration 9200 (3.74134 iter/s, 26.7284s/100 iters), loss = 8.66395
I0510 17:11:46.535336 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.66395 (* 1 = 8.66395 loss)
I0510 17:11:46.535353 39557 sgd_solver.cpp:105] Iteration 9200, lr = 0.1
I0510 17:12:13.263926 39557 solver.cpp:218] Iteration 9300 (3.74132 iter/s, 26.7285s/100 iters), loss = 8.73081
I0510 17:12:13.263993 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.73081 (* 1 = 8.73081 loss)
I0510 17:12:13.264008 39557 sgd_solver.cpp:105] Iteration 9300, lr = 0.1
I0510 17:12:39.906239 39557 solver.cpp:218] Iteration 9400 (3.75345 iter/s, 26.6422s/100 iters), loss = 9.17373
I0510 17:12:39.906445 39557 solver.cpp:237]     Train net output #0: softmax_loss = 9.17373 (* 1 = 9.17373 loss)
I0510 17:12:39.906460 39557 sgd_solver.cpp:105] Iteration 9400, lr = 0.1
I0510 17:13:06.496078 39557 solver.cpp:218] Iteration 9500 (3.76088 iter/s, 26.5895s/100 iters), loss = 8.31457
I0510 17:13:06.496141 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.31457 (* 1 = 8.31457 loss)
I0510 17:13:06.496155 39557 sgd_solver.cpp:105] Iteration 9500, lr = 0.1
I0510 17:13:33.118450 39557 solver.cpp:218] Iteration 9600 (3.75626 iter/s, 26.6222s/100 iters), loss = 9.10969
I0510 17:13:33.118656 39557 solver.cpp:237]     Train net output #0: softmax_loss = 9.10969 (* 1 = 9.10969 loss)
I0510 17:13:33.118674 39557 sgd_solver.cpp:105] Iteration 9600, lr = 0.1
I0510 17:14:00.097398 39557 solver.cpp:218] Iteration 9700 (3.70663 iter/s, 26.9787s/100 iters), loss = 8.59775
I0510 17:14:00.097455 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.59775 (* 1 = 8.59775 loss)
I0510 17:14:00.097468 39557 sgd_solver.cpp:105] Iteration 9700, lr = 0.1
I0510 17:14:26.694725 39557 solver.cpp:218] Iteration 9800 (3.7598 iter/s, 26.5972s/100 iters), loss = 8.88585
I0510 17:14:26.694952 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.88585 (* 1 = 8.88585 loss)
I0510 17:14:26.694968 39557 sgd_solver.cpp:105] Iteration 9800, lr = 0.1
I0510 17:14:53.307888 39557 solver.cpp:218] Iteration 9900 (3.75758 iter/s, 26.6128s/100 iters), loss = 8.71014
I0510 17:14:53.307948 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.71014 (* 1 = 8.71014 loss)
I0510 17:14:53.307960 39557 sgd_solver.cpp:105] Iteration 9900, lr = 0.1
I0510 17:15:19.644258 39557 solver.cpp:447] Snapshotting to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_10000.caffemodel
I0510 17:15:19.730533 39557 sgd_solver.cpp:273] Snapshotting solver state to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_10000.solverstate
I0510 17:15:20.023571 39557 solver.cpp:218] Iteration 10000 (3.74314 iter/s, 26.7155s/100 iters), loss = 8.1434
I0510 17:15:20.023623 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.1434 (* 1 = 8.1434 loss)
I0510 17:15:20.023634 39557 sgd_solver.cpp:105] Iteration 10000, lr = 0.1
I0510 17:15:46.720855 39557 solver.cpp:218] Iteration 10100 (3.74572 iter/s, 26.6971s/100 iters), loss = 8.53685
I0510 17:15:46.720919 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.53685 (* 1 = 8.53685 loss)
I0510 17:15:46.720934 39557 sgd_solver.cpp:105] Iteration 10100, lr = 0.1
I0510 17:16:13.393139 39557 solver.cpp:218] Iteration 10200 (3.74923 iter/s, 26.6721s/100 iters), loss = 8.33243
I0510 17:16:13.393344 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.33243 (* 1 = 8.33243 loss)
I0510 17:16:13.393359 39557 sgd_solver.cpp:105] Iteration 10200, lr = 0.1
I0510 17:16:40.039645 39557 solver.cpp:218] Iteration 10300 (3.75288 iter/s, 26.6462s/100 iters), loss = 8.31854
I0510 17:16:40.039706 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.31854 (* 1 = 8.31854 loss)
I0510 17:16:40.039719 39557 sgd_solver.cpp:105] Iteration 10300, lr = 0.1
I0510 17:17:06.653352 39557 solver.cpp:218] Iteration 10400 (3.75749 iter/s, 26.6135s/100 iters), loss = 8.16892
I0510 17:17:06.653550 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.16892 (* 1 = 8.16892 loss)
I0510 17:17:06.653566 39557 sgd_solver.cpp:105] Iteration 10400, lr = 0.1
I0510 17:17:33.320757 39557 solver.cpp:218] Iteration 10500 (3.74994 iter/s, 26.6671s/100 iters), loss = 8.43277
I0510 17:17:33.320818 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.43277 (* 1 = 8.43277 loss)
I0510 17:17:33.320832 39557 sgd_solver.cpp:105] Iteration 10500, lr = 0.1
I0510 17:17:59.915045 39557 solver.cpp:218] Iteration 10600 (3.76023 iter/s, 26.5941s/100 iters), loss = 8.64221
I0510 17:17:59.915257 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.64221 (* 1 = 8.64221 loss)
I0510 17:17:59.915273 39557 sgd_solver.cpp:105] Iteration 10600, lr = 0.1
I0510 17:18:26.999750 39557 solver.cpp:218] Iteration 10700 (3.69216 iter/s, 27.0844s/100 iters), loss = 8.34888
I0510 17:18:26.999815 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.34888 (* 1 = 8.34888 loss)
I0510 17:18:26.999830 39557 sgd_solver.cpp:105] Iteration 10700, lr = 0.1
I0510 17:18:53.673794 39557 solver.cpp:218] Iteration 10800 (3.74899 iter/s, 26.6739s/100 iters), loss = 9.35026
I0510 17:18:53.674011 39557 solver.cpp:237]     Train net output #0: softmax_loss = 9.35026 (* 1 = 9.35026 loss)
I0510 17:18:53.674027 39557 sgd_solver.cpp:105] Iteration 10800, lr = 0.1
I0510 17:19:20.443460 39557 solver.cpp:218] Iteration 10900 (3.73562 iter/s, 26.7693s/100 iters), loss = 8.29279
I0510 17:19:20.443512 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.29279 (* 1 = 8.29279 loss)
I0510 17:19:20.443526 39557 sgd_solver.cpp:105] Iteration 10900, lr = 0.1
I0510 17:19:47.134366 39557 solver.cpp:218] Iteration 11000 (3.74666 iter/s, 26.6904s/100 iters), loss = 8.33954
I0510 17:19:47.134558 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.33954 (* 1 = 8.33954 loss)
I0510 17:19:47.134575 39557 sgd_solver.cpp:105] Iteration 11000, lr = 0.1
I0510 17:20:13.770638 39557 solver.cpp:218] Iteration 11100 (3.75437 iter/s, 26.6357s/100 iters), loss = 7.88921
I0510 17:20:13.770699 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.88921 (* 1 = 7.88921 loss)
I0510 17:20:13.770712 39557 sgd_solver.cpp:105] Iteration 11100, lr = 0.1
I0510 17:20:40.324473 39557 solver.cpp:218] Iteration 11200 (3.766 iter/s, 26.5534s/100 iters), loss = 8.35987
I0510 17:20:40.324671 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.35987 (* 1 = 8.35987 loss)
I0510 17:20:40.324688 39557 sgd_solver.cpp:105] Iteration 11200, lr = 0.1
I0510 17:21:06.960402 39557 solver.cpp:218] Iteration 11300 (3.75441 iter/s, 26.6353s/100 iters), loss = 8.60472
I0510 17:21:06.960464 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.60472 (* 1 = 8.60472 loss)
I0510 17:21:06.960479 39557 sgd_solver.cpp:105] Iteration 11300, lr = 0.1
I0510 17:21:33.582263 39557 solver.cpp:218] Iteration 11400 (3.75637 iter/s, 26.6214s/100 iters), loss = 7.38512
I0510 17:21:33.582427 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.38512 (* 1 = 7.38512 loss)
I0510 17:21:33.582443 39557 sgd_solver.cpp:105] Iteration 11400, lr = 0.1
I0510 17:22:00.168717 39557 solver.cpp:218] Iteration 11500 (3.76139 iter/s, 26.5859s/100 iters), loss = 8.46223
I0510 17:22:00.168771 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.46223 (* 1 = 8.46223 loss)
I0510 17:22:00.168799 39557 sgd_solver.cpp:105] Iteration 11500, lr = 0.1
I0510 17:22:28.202785 39557 solver.cpp:218] Iteration 11600 (3.56714 iter/s, 28.0336s/100 iters), loss = 8.36687
I0510 17:22:28.202989 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.36687 (* 1 = 8.36687 loss)
I0510 17:22:28.203007 39557 sgd_solver.cpp:105] Iteration 11600, lr = 0.1
I0510 17:22:54.813022 39557 solver.cpp:218] Iteration 11700 (3.75803 iter/s, 26.6097s/100 iters), loss = 7.70918
I0510 17:22:54.813083 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.70918 (* 1 = 7.70918 loss)
I0510 17:22:54.813097 39557 sgd_solver.cpp:105] Iteration 11700, lr = 0.1
I0510 17:23:21.444193 39557 solver.cpp:218] Iteration 11800 (3.75505 iter/s, 26.6308s/100 iters), loss = 8.46843
I0510 17:23:21.444389 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.46843 (* 1 = 8.46843 loss)
I0510 17:23:21.444404 39557 sgd_solver.cpp:105] Iteration 11800, lr = 0.1
I0510 17:23:48.048835 39557 solver.cpp:218] Iteration 11900 (3.75882 iter/s, 26.6041s/100 iters), loss = 8.1969
I0510 17:23:48.048898 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.1969 (* 1 = 8.1969 loss)
I0510 17:23:48.048913 39557 sgd_solver.cpp:105] Iteration 11900, lr = 0.1
I0510 17:24:14.822263 39557 solver.cpp:218] Iteration 12000 (3.7351 iter/s, 26.773s/100 iters), loss = 8.60358
I0510 17:24:14.822480 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.60358 (* 1 = 8.60358 loss)
I0510 17:24:14.822497 39557 sgd_solver.cpp:105] Iteration 12000, lr = 0.1
I0510 17:24:41.570382 39557 solver.cpp:218] Iteration 12100 (3.73866 iter/s, 26.7476s/100 iters), loss = 7.93841
I0510 17:24:41.570443 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.93841 (* 1 = 7.93841 loss)
I0510 17:24:41.570456 39557 sgd_solver.cpp:105] Iteration 12100, lr = 0.1
I0510 17:25:08.176225 39557 solver.cpp:218] Iteration 12200 (3.75863 iter/s, 26.6055s/100 iters), loss = 7.97983
I0510 17:25:08.176404 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.97983 (* 1 = 7.97983 loss)
I0510 17:25:08.176419 39557 sgd_solver.cpp:105] Iteration 12200, lr = 0.1
I0510 17:25:34.816655 39557 solver.cpp:218] Iteration 12300 (3.75376 iter/s, 26.64s/100 iters), loss = 8.83177
I0510 17:25:34.816715 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.83177 (* 1 = 8.83177 loss)
I0510 17:25:34.816728 39557 sgd_solver.cpp:105] Iteration 12300, lr = 0.1
I0510 17:26:01.362257 39557 solver.cpp:218] Iteration 12400 (3.76715 iter/s, 26.5453s/100 iters), loss = 8.37192
I0510 17:26:01.362464 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.37192 (* 1 = 8.37192 loss)
I0510 17:26:01.362481 39557 sgd_solver.cpp:105] Iteration 12400, lr = 0.1
I0510 17:26:28.048640 39557 solver.cpp:218] Iteration 12500 (3.7473 iter/s, 26.6859s/100 iters), loss = 8.46346
I0510 17:26:28.048701 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.46346 (* 1 = 8.46346 loss)
I0510 17:26:28.048713 39557 sgd_solver.cpp:105] Iteration 12500, lr = 0.1
I0510 17:26:54.644204 39557 solver.cpp:218] Iteration 12600 (3.76007 iter/s, 26.5952s/100 iters), loss = 7.73131
I0510 17:26:54.644395 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.73131 (* 1 = 7.73131 loss)
I0510 17:26:54.644415 39557 sgd_solver.cpp:105] Iteration 12600, lr = 0.1
I0510 17:27:21.330152 39557 solver.cpp:218] Iteration 12700 (3.74735 iter/s, 26.6855s/100 iters), loss = 7.58236
I0510 17:27:21.330206 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.58236 (* 1 = 7.58236 loss)
I0510 17:27:21.330219 39557 sgd_solver.cpp:105] Iteration 12700, lr = 0.1
I0510 17:27:47.954340 39557 solver.cpp:218] Iteration 12800 (3.75603 iter/s, 26.6239s/100 iters), loss = 8.25418
I0510 17:27:47.954540 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.25418 (* 1 = 8.25418 loss)
I0510 17:27:47.954555 39557 sgd_solver.cpp:105] Iteration 12800, lr = 0.1
I0510 17:28:14.651974 39557 solver.cpp:218] Iteration 12900 (3.74571 iter/s, 26.6972s/100 iters), loss = 7.9416
I0510 17:28:14.652027 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.9416 (* 1 = 7.9416 loss)
I0510 17:28:14.652040 39557 sgd_solver.cpp:105] Iteration 12900, lr = 0.1
I0510 17:28:41.296058 39557 solver.cpp:218] Iteration 13000 (3.75322 iter/s, 26.6438s/100 iters), loss = 7.77648
I0510 17:28:41.296264 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.77648 (* 1 = 7.77648 loss)
I0510 17:28:41.296279 39557 sgd_solver.cpp:105] Iteration 13000, lr = 0.1
I0510 17:29:07.898339 39557 solver.cpp:218] Iteration 13100 (3.75914 iter/s, 26.6018s/100 iters), loss = 8.2704
I0510 17:29:07.898389 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.2704 (* 1 = 8.2704 loss)
I0510 17:29:07.898401 39557 sgd_solver.cpp:105] Iteration 13100, lr = 0.1
I0510 17:29:34.663442 39557 solver.cpp:218] Iteration 13200 (3.73625 iter/s, 26.7648s/100 iters), loss = 7.57209
I0510 17:29:34.663689 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.57209 (* 1 = 7.57209 loss)
I0510 17:29:34.663712 39557 sgd_solver.cpp:105] Iteration 13200, lr = 0.1
I0510 17:30:01.446602 39557 solver.cpp:218] Iteration 13300 (3.73376 iter/s, 26.7827s/100 iters), loss = 7.85369
I0510 17:30:01.446653 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.85369 (* 1 = 7.85369 loss)
I0510 17:30:01.446667 39557 sgd_solver.cpp:105] Iteration 13300, lr = 0.1
I0510 17:30:28.029861 39557 solver.cpp:218] Iteration 13400 (3.76181 iter/s, 26.583s/100 iters), loss = 7.03704
I0510 17:30:28.030056 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.03704 (* 1 = 7.03704 loss)
I0510 17:30:28.030072 39557 sgd_solver.cpp:105] Iteration 13400, lr = 0.1
I0510 17:30:54.692132 39557 solver.cpp:218] Iteration 13500 (3.75068 iter/s, 26.6618s/100 iters), loss = 8.08046
I0510 17:30:54.692191 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.08046 (* 1 = 8.08046 loss)
I0510 17:30:54.692204 39557 sgd_solver.cpp:105] Iteration 13500, lr = 0.1
I0510 17:31:21.277463 39557 solver.cpp:218] Iteration 13600 (3.76151 iter/s, 26.5851s/100 iters), loss = 7.49108
I0510 17:31:21.277643 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.49108 (* 1 = 7.49108 loss)
I0510 17:31:21.277658 39557 sgd_solver.cpp:105] Iteration 13600, lr = 0.1
I0510 17:31:47.836112 39557 solver.cpp:218] Iteration 13700 (3.76531 iter/s, 26.5582s/100 iters), loss = 7.92159
I0510 17:31:47.836174 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.92159 (* 1 = 7.92159 loss)
I0510 17:31:47.836187 39557 sgd_solver.cpp:105] Iteration 13700, lr = 0.1
I0510 17:32:14.483530 39557 solver.cpp:218] Iteration 13800 (3.75275 iter/s, 26.6471s/100 iters), loss = 8.6414
I0510 17:32:14.483716 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.6414 (* 1 = 8.6414 loss)
I0510 17:32:14.483731 39557 sgd_solver.cpp:105] Iteration 13800, lr = 0.1
I0510 17:32:41.115186 39557 solver.cpp:218] Iteration 13900 (3.75499 iter/s, 26.6313s/100 iters), loss = 8.15432
I0510 17:32:41.115249 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.15432 (* 1 = 8.15432 loss)
I0510 17:32:41.115262 39557 sgd_solver.cpp:105] Iteration 13900, lr = 0.1
I0510 17:33:07.760713 39557 solver.cpp:218] Iteration 14000 (3.75301 iter/s, 26.6453s/100 iters), loss = 7.73722
I0510 17:33:07.760906 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.73722 (* 1 = 7.73722 loss)
I0510 17:33:07.760921 39557 sgd_solver.cpp:105] Iteration 14000, lr = 0.1
I0510 17:33:34.395871 39557 solver.cpp:218] Iteration 14100 (3.75449 iter/s, 26.6348s/100 iters), loss = 7.46247
I0510 17:33:34.395932 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.46247 (* 1 = 7.46247 loss)
I0510 17:33:34.395944 39557 sgd_solver.cpp:105] Iteration 14100, lr = 0.1
I0510 17:34:00.989614 39557 solver.cpp:218] Iteration 14200 (3.76032 iter/s, 26.5935s/100 iters), loss = 8.49526
I0510 17:34:00.989797 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.49526 (* 1 = 8.49526 loss)
I0510 17:34:00.989814 39557 sgd_solver.cpp:105] Iteration 14200, lr = 0.1
I0510 17:34:27.631498 39557 solver.cpp:218] Iteration 14300 (3.75354 iter/s, 26.6415s/100 iters), loss = 7.9508
I0510 17:34:27.631561 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.9508 (* 1 = 7.9508 loss)
I0510 17:34:27.631575 39557 sgd_solver.cpp:105] Iteration 14300, lr = 0.1
I0510 17:34:54.257052 39557 solver.cpp:218] Iteration 14400 (3.75583 iter/s, 26.6253s/100 iters), loss = 8.21063
I0510 17:34:54.257238 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.21063 (* 1 = 8.21063 loss)
I0510 17:34:54.257254 39557 sgd_solver.cpp:105] Iteration 14400, lr = 0.1
I0510 17:35:20.817126 39557 solver.cpp:218] Iteration 14500 (3.7651 iter/s, 26.5597s/100 iters), loss = 7.80932
I0510 17:35:20.817190 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.80932 (* 1 = 7.80932 loss)
I0510 17:35:20.817205 39557 sgd_solver.cpp:105] Iteration 14500, lr = 0.1
I0510 17:35:47.538452 39557 solver.cpp:218] Iteration 14600 (3.74236 iter/s, 26.7211s/100 iters), loss = 8.07924
I0510 17:35:47.538663 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.07924 (* 1 = 8.07924 loss)
I0510 17:35:47.538681 39557 sgd_solver.cpp:105] Iteration 14600, lr = 0.1
I0510 17:36:14.159368 39557 solver.cpp:218] Iteration 14700 (3.7565 iter/s, 26.6205s/100 iters), loss = 7.63123
I0510 17:36:14.159430 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.63123 (* 1 = 7.63123 loss)
I0510 17:36:14.159441 39557 sgd_solver.cpp:105] Iteration 14700, lr = 0.1
I0510 17:36:40.828917 39557 solver.cpp:218] Iteration 14800 (3.74963 iter/s, 26.6693s/100 iters), loss = 7.45361
I0510 17:36:40.829104 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.45361 (* 1 = 7.45361 loss)
I0510 17:36:40.829119 39557 sgd_solver.cpp:105] Iteration 14800, lr = 0.1
I0510 17:37:07.474699 39557 solver.cpp:218] Iteration 14900 (3.75299 iter/s, 26.6454s/100 iters), loss = 7.19834
I0510 17:37:07.474762 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.19834 (* 1 = 7.19834 loss)
I0510 17:37:07.474774 39557 sgd_solver.cpp:105] Iteration 14900, lr = 0.1
I0510 17:37:33.831854 39557 solver.cpp:447] Snapshotting to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_15000.caffemodel
I0510 17:37:33.873108 39557 sgd_solver.cpp:273] Snapshotting solver state to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_15000.solverstate
I0510 17:37:34.165683 39557 solver.cpp:218] Iteration 15000 (3.74662 iter/s, 26.6907s/100 iters), loss = 7.28784
I0510 17:37:34.165742 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.28784 (* 1 = 7.28784 loss)
I0510 17:37:34.165753 39557 sgd_solver.cpp:105] Iteration 15000, lr = 0.1
I0510 17:38:00.876016 39557 solver.cpp:218] Iteration 15100 (3.7439 iter/s, 26.7101s/100 iters), loss = 7.84083
I0510 17:38:00.876082 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.84083 (* 1 = 7.84083 loss)
I0510 17:38:00.876096 39557 sgd_solver.cpp:105] Iteration 15100, lr = 0.1
I0510 17:38:27.569624 39557 solver.cpp:218] Iteration 15200 (3.74625 iter/s, 26.6934s/100 iters), loss = 7.25799
I0510 17:38:27.569823 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.25799 (* 1 = 7.25799 loss)
I0510 17:38:27.569839 39557 sgd_solver.cpp:105] Iteration 15200, lr = 0.1
I0510 17:38:54.196426 39557 solver.cpp:218] Iteration 15300 (3.75567 iter/s, 26.6264s/100 iters), loss = 7.33421
I0510 17:38:54.196487 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.33421 (* 1 = 7.33421 loss)
I0510 17:38:54.196501 39557 sgd_solver.cpp:105] Iteration 15300, lr = 0.1
I0510 17:39:20.805507 39557 solver.cpp:218] Iteration 15400 (3.75815 iter/s, 26.6088s/100 iters), loss = 7.8192
I0510 17:39:20.805680 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.8192 (* 1 = 7.8192 loss)
I0510 17:39:20.805696 39557 sgd_solver.cpp:105] Iteration 15400, lr = 0.1
I0510 17:39:47.506479 39557 solver.cpp:218] Iteration 15500 (3.74523 iter/s, 26.7006s/100 iters), loss = 6.73623
I0510 17:39:47.506544 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.73623 (* 1 = 6.73623 loss)
I0510 17:39:47.506557 39557 sgd_solver.cpp:105] Iteration 15500, lr = 0.1
I0510 17:40:14.217116 39557 solver.cpp:218] Iteration 15600 (3.74386 iter/s, 26.7104s/100 iters), loss = 7.63754
I0510 17:40:14.217324 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.63754 (* 1 = 7.63754 loss)
I0510 17:40:14.217355 39557 sgd_solver.cpp:105] Iteration 15600, lr = 0.1
I0510 17:40:40.932061 39557 solver.cpp:218] Iteration 15700 (3.74328 iter/s, 26.7146s/100 iters), loss = 6.91343
I0510 17:40:40.932122 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.91343 (* 1 = 6.91343 loss)
I0510 17:40:40.932137 39557 sgd_solver.cpp:105] Iteration 15700, lr = 0.1
I0510 17:41:07.557451 39557 solver.cpp:218] Iteration 15800 (3.75585 iter/s, 26.6252s/100 iters), loss = 7.31873
I0510 17:41:07.557621 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.31873 (* 1 = 7.31873 loss)
I0510 17:41:07.557644 39557 sgd_solver.cpp:105] Iteration 15800, lr = 0.1
I0510 17:41:34.155540 39557 solver.cpp:218] Iteration 15900 (3.75972 iter/s, 26.5977s/100 iters), loss = 8.18536
I0510 17:41:34.155604 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.18536 (* 1 = 8.18536 loss)
I0510 17:41:34.155617 39557 sgd_solver.cpp:105] Iteration 15900, lr = 0.1
I0510 17:42:00.790050 39557 solver.cpp:218] Iteration 16000 (3.75456 iter/s, 26.6343s/100 iters), loss = 7.34497
I0510 17:42:00.790241 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.34497 (* 1 = 7.34497 loss)
I0510 17:42:00.790258 39557 sgd_solver.cpp:105] Iteration 16000, lr = 0.1
I0510 17:42:27.345171 39557 solver.cpp:218] Iteration 16100 (3.7658 iter/s, 26.5548s/100 iters), loss = 7.50814
I0510 17:42:27.345237 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.50814 (* 1 = 7.50814 loss)
I0510 17:42:27.345249 39557 sgd_solver.cpp:105] Iteration 16100, lr = 0.1
I0510 17:42:53.980263 39557 solver.cpp:218] Iteration 16200 (3.75448 iter/s, 26.6349s/100 iters), loss = 7.66619
I0510 17:42:53.980435 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.66619 (* 1 = 7.66619 loss)
I0510 17:42:53.980450 39557 sgd_solver.cpp:105] Iteration 16200, lr = 0.1
I0510 17:43:20.591403 39557 solver.cpp:218] Iteration 16300 (3.75787 iter/s, 26.6108s/100 iters), loss = 7.30956
I0510 17:43:20.591464 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.30956 (* 1 = 7.30956 loss)
I0510 17:43:20.591477 39557 sgd_solver.cpp:105] Iteration 16300, lr = 0.1
I0510 17:43:47.258579 39557 solver.cpp:218] Iteration 16400 (3.74996 iter/s, 26.667s/100 iters), loss = 8.07502
I0510 17:43:47.258772 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.07502 (* 1 = 8.07502 loss)
I0510 17:43:47.258788 39557 sgd_solver.cpp:105] Iteration 16400, lr = 0.1
I0510 17:44:13.907279 39557 solver.cpp:218] Iteration 16500 (3.75258 iter/s, 26.6483s/100 iters), loss = 7.96665
I0510 17:44:13.907357 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.96665 (* 1 = 7.96665 loss)
I0510 17:44:13.907371 39557 sgd_solver.cpp:105] Iteration 16500, lr = 0.1
I0510 17:44:40.555707 39557 solver.cpp:218] Iteration 16600 (3.7526 iter/s, 26.6482s/100 iters), loss = 7.86983
I0510 17:44:40.555876 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.86983 (* 1 = 7.86983 loss)
I0510 17:44:40.555892 39557 sgd_solver.cpp:105] Iteration 16600, lr = 0.1
I0510 17:45:07.213645 39557 solver.cpp:218] Iteration 16700 (3.75127 iter/s, 26.6576s/100 iters), loss = 7.50941
I0510 17:45:07.213708 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.50941 (* 1 = 7.50941 loss)
I0510 17:45:07.213722 39557 sgd_solver.cpp:105] Iteration 16700, lr = 0.1
I0510 17:45:33.790846 39557 solver.cpp:218] Iteration 16800 (3.76266 iter/s, 26.577s/100 iters), loss = 7.30301
I0510 17:45:33.791038 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.30301 (* 1 = 7.30301 loss)
I0510 17:45:33.791054 39557 sgd_solver.cpp:105] Iteration 16800, lr = 0.1
I0510 17:46:00.470816 39557 solver.cpp:218] Iteration 16900 (3.74818 iter/s, 26.6796s/100 iters), loss = 7.0984
I0510 17:46:00.470875 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.0984 (* 1 = 7.0984 loss)
I0510 17:46:00.470888 39557 sgd_solver.cpp:105] Iteration 16900, lr = 0.1
I0510 17:46:27.050189 39557 solver.cpp:218] Iteration 17000 (3.76235 iter/s, 26.5792s/100 iters), loss = 6.31261
I0510 17:46:27.050376 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.31261 (* 1 = 6.31261 loss)
I0510 17:46:27.050392 39557 sgd_solver.cpp:105] Iteration 17000, lr = 0.1
I0510 17:46:53.603973 39557 solver.cpp:218] Iteration 17100 (3.76599 iter/s, 26.5534s/100 iters), loss = 8.15016
I0510 17:46:53.604038 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.15016 (* 1 = 8.15016 loss)
I0510 17:46:53.604050 39557 sgd_solver.cpp:105] Iteration 17100, lr = 0.1
I0510 17:47:20.363270 39557 solver.cpp:218] Iteration 17200 (3.73705 iter/s, 26.7591s/100 iters), loss = 7.88503
I0510 17:47:20.363492 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.88503 (* 1 = 7.88503 loss)
I0510 17:47:20.363509 39557 sgd_solver.cpp:105] Iteration 17200, lr = 0.1
I0510 17:47:47.069800 39557 solver.cpp:218] Iteration 17300 (3.74445 iter/s, 26.7062s/100 iters), loss = 7.86275
I0510 17:47:47.069854 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.86275 (* 1 = 7.86275 loss)
I0510 17:47:47.069866 39557 sgd_solver.cpp:105] Iteration 17300, lr = 0.1
I0510 17:48:15.178782 39557 solver.cpp:218] Iteration 17400 (3.55761 iter/s, 28.1088s/100 iters), loss = 7.45749
I0510 17:48:15.178951 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.45749 (* 1 = 7.45749 loss)
I0510 17:48:15.178966 39557 sgd_solver.cpp:105] Iteration 17400, lr = 0.1
I0510 17:48:41.790529 39557 solver.cpp:218] Iteration 17500 (3.75779 iter/s, 26.6114s/100 iters), loss = 7.52282
I0510 17:48:41.790607 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.52282 (* 1 = 7.52282 loss)
I0510 17:48:41.790619 39557 sgd_solver.cpp:105] Iteration 17500, lr = 0.1
I0510 17:49:08.424264 39557 solver.cpp:218] Iteration 17600 (3.75467 iter/s, 26.6335s/100 iters), loss = 7.88404
I0510 17:49:08.424437 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.88404 (* 1 = 7.88404 loss)
I0510 17:49:08.424453 39557 sgd_solver.cpp:105] Iteration 17600, lr = 0.1
I0510 17:49:35.030450 39557 solver.cpp:218] Iteration 17700 (3.75857 iter/s, 26.6059s/100 iters), loss = 8.0086
I0510 17:49:35.030514 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.0086 (* 1 = 8.0086 loss)
I0510 17:49:35.030525 39557 sgd_solver.cpp:105] Iteration 17700, lr = 0.1
I0510 17:50:01.672840 39557 solver.cpp:218] Iteration 17800 (3.75345 iter/s, 26.6422s/100 iters), loss = 6.92362
I0510 17:50:01.673022 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.92362 (* 1 = 6.92362 loss)
I0510 17:50:01.673038 39557 sgd_solver.cpp:105] Iteration 17800, lr = 0.1
I0510 17:50:28.253970 39557 solver.cpp:218] Iteration 17900 (3.76211 iter/s, 26.5808s/100 iters), loss = 7.65731
I0510 17:50:28.254034 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.65731 (* 1 = 7.65731 loss)
I0510 17:50:28.254046 39557 sgd_solver.cpp:105] Iteration 17900, lr = 0.1
I0510 17:50:54.840131 39557 solver.cpp:218] Iteration 18000 (3.76139 iter/s, 26.5859s/100 iters), loss = 7.33394
I0510 17:50:54.840386 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.33394 (* 1 = 7.33394 loss)
I0510 17:50:54.840432 39557 sgd_solver.cpp:105] Iteration 18000, lr = 0.1
I0510 17:51:21.475714 39557 solver.cpp:218] Iteration 18100 (3.75443 iter/s, 26.6352s/100 iters), loss = 7.7734
I0510 17:51:21.475764 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.7734 (* 1 = 7.7734 loss)
I0510 17:51:21.475776 39557 sgd_solver.cpp:105] Iteration 18100, lr = 0.1
I0510 17:51:48.044507 39557 solver.cpp:218] Iteration 18200 (3.76384 iter/s, 26.5686s/100 iters), loss = 7.33134
I0510 17:51:48.044708 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.33134 (* 1 = 7.33134 loss)
I0510 17:51:48.044723 39557 sgd_solver.cpp:105] Iteration 18200, lr = 0.1
I0510 17:52:14.650784 39557 solver.cpp:218] Iteration 18300 (3.75856 iter/s, 26.6059s/100 iters), loss = 7.09186
I0510 17:52:14.650840 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.09186 (* 1 = 7.09186 loss)
I0510 17:52:14.650856 39557 sgd_solver.cpp:105] Iteration 18300, lr = 0.1
I0510 17:52:41.359383 39557 solver.cpp:218] Iteration 18400 (3.74414 iter/s, 26.7084s/100 iters), loss = 8.10586
I0510 17:52:41.359786 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.10586 (* 1 = 8.10586 loss)
I0510 17:52:41.359802 39557 sgd_solver.cpp:105] Iteration 18400, lr = 0.1
I0510 17:53:08.045454 39557 solver.cpp:218] Iteration 18500 (3.74735 iter/s, 26.6855s/100 iters), loss = 7.80698
I0510 17:53:08.045519 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.80698 (* 1 = 7.80698 loss)
I0510 17:53:08.045534 39557 sgd_solver.cpp:105] Iteration 18500, lr = 0.1
I0510 17:53:34.770432 39557 solver.cpp:218] Iteration 18600 (3.74184 iter/s, 26.7248s/100 iters), loss = 7.62438
I0510 17:53:34.770642 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.62438 (* 1 = 7.62438 loss)
I0510 17:53:34.770665 39557 sgd_solver.cpp:105] Iteration 18600, lr = 0.1
I0510 17:54:01.552460 39557 solver.cpp:218] Iteration 18700 (3.73389 iter/s, 26.7817s/100 iters), loss = 8.18468
I0510 17:54:01.552525 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.18468 (* 1 = 8.18468 loss)
I0510 17:54:01.552539 39557 sgd_solver.cpp:105] Iteration 18700, lr = 0.1
I0510 17:54:28.237284 39557 solver.cpp:218] Iteration 18800 (3.74747 iter/s, 26.6847s/100 iters), loss = 6.9019
I0510 17:54:28.237480 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.9019 (* 1 = 6.9019 loss)
I0510 17:54:28.237495 39557 sgd_solver.cpp:105] Iteration 18800, lr = 0.1
I0510 17:54:54.877030 39557 solver.cpp:218] Iteration 18900 (3.75383 iter/s, 26.6395s/100 iters), loss = 7.87617
I0510 17:54:54.877094 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.87617 (* 1 = 7.87617 loss)
I0510 17:54:54.877106 39557 sgd_solver.cpp:105] Iteration 18900, lr = 0.1
I0510 17:55:21.505213 39557 solver.cpp:218] Iteration 19000 (3.75545 iter/s, 26.628s/100 iters), loss = 7.00283
I0510 17:55:21.505409 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.00283 (* 1 = 7.00283 loss)
I0510 17:55:21.505424 39557 sgd_solver.cpp:105] Iteration 19000, lr = 0.1
I0510 17:55:48.053607 39557 solver.cpp:218] Iteration 19100 (3.76675 iter/s, 26.5481s/100 iters), loss = 6.84899
I0510 17:55:48.053669 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.84899 (* 1 = 6.84899 loss)
I0510 17:55:48.053683 39557 sgd_solver.cpp:105] Iteration 19100, lr = 0.1
I0510 17:56:14.682402 39557 solver.cpp:218] Iteration 19200 (3.75536 iter/s, 26.6286s/100 iters), loss = 7.1353
I0510 17:56:14.682570 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.1353 (* 1 = 7.1353 loss)
I0510 17:56:14.682586 39557 sgd_solver.cpp:105] Iteration 19200, lr = 0.1
I0510 17:56:41.300812 39557 solver.cpp:218] Iteration 19300 (3.75684 iter/s, 26.6181s/100 iters), loss = 6.65865
I0510 17:56:41.300873 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.65865 (* 1 = 6.65865 loss)
I0510 17:56:41.300885 39557 sgd_solver.cpp:105] Iteration 19300, lr = 0.1
I0510 17:57:07.895023 39557 solver.cpp:218] Iteration 19400 (3.76024 iter/s, 26.594s/100 iters), loss = 6.81562
I0510 17:57:07.895211 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.81562 (* 1 = 6.81562 loss)
I0510 17:57:07.895227 39557 sgd_solver.cpp:105] Iteration 19400, lr = 0.1
I0510 17:57:34.487614 39557 solver.cpp:218] Iteration 19500 (3.76049 iter/s, 26.5923s/100 iters), loss = 7.50622
I0510 17:57:34.487665 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.50622 (* 1 = 7.50622 loss)
I0510 17:57:34.487679 39557 sgd_solver.cpp:105] Iteration 19500, lr = 0.1
I0510 17:58:01.130174 39557 solver.cpp:218] Iteration 19600 (3.75342 iter/s, 26.6424s/100 iters), loss = 7.82467
I0510 17:58:01.130349 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.82467 (* 1 = 7.82467 loss)
I0510 17:58:01.130364 39557 sgd_solver.cpp:105] Iteration 19600, lr = 0.1
I0510 17:58:27.643208 39557 solver.cpp:218] Iteration 19700 (3.77177 iter/s, 26.5128s/100 iters), loss = 7.3811
I0510 17:58:27.643270 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.3811 (* 1 = 7.3811 loss)
I0510 17:58:27.643285 39557 sgd_solver.cpp:105] Iteration 19700, lr = 0.1
I0510 17:58:54.288372 39557 solver.cpp:218] Iteration 19800 (3.75305 iter/s, 26.645s/100 iters), loss = 7.33267
I0510 17:58:54.288575 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.33267 (* 1 = 7.33267 loss)
I0510 17:58:54.288591 39557 sgd_solver.cpp:105] Iteration 19800, lr = 0.1
I0510 17:59:20.841562 39557 solver.cpp:218] Iteration 19900 (3.76607 iter/s, 26.5529s/100 iters), loss = 7.57448
I0510 17:59:20.841621 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.57448 (* 1 = 7.57448 loss)
I0510 17:59:20.841634 39557 sgd_solver.cpp:105] Iteration 19900, lr = 0.1
I0510 17:59:47.114949 39557 solver.cpp:447] Snapshotting to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_20000.caffemodel
I0510 17:59:47.154994 39557 sgd_solver.cpp:273] Snapshotting solver state to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_20000.solverstate
I0510 17:59:47.443930 39557 solver.cpp:218] Iteration 20000 (3.75909 iter/s, 26.6022s/100 iters), loss = 7.5326
I0510 17:59:47.443986 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.5326 (* 1 = 7.5326 loss)
I0510 17:59:47.443997 39557 sgd_solver.cpp:105] Iteration 20000, lr = 0.1
I0510 18:00:14.056943 39557 solver.cpp:218] Iteration 20100 (3.75758 iter/s, 26.6128s/100 iters), loss = 7.00854
I0510 18:00:14.057006 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.00854 (* 1 = 7.00854 loss)
I0510 18:00:14.057021 39557 sgd_solver.cpp:105] Iteration 20100, lr = 0.1
I0510 18:00:40.692253 39557 solver.cpp:218] Iteration 20200 (3.75444 iter/s, 26.6351s/100 iters), loss = 8.16949
I0510 18:00:40.692492 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.16949 (* 1 = 8.16949 loss)
I0510 18:00:40.692507 39557 sgd_solver.cpp:105] Iteration 20200, lr = 0.1
I0510 18:01:07.365068 39557 solver.cpp:218] Iteration 20300 (3.74919 iter/s, 26.6725s/100 iters), loss = 7.74783
I0510 18:01:07.365129 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.74783 (* 1 = 7.74783 loss)
I0510 18:01:07.365142 39557 sgd_solver.cpp:105] Iteration 20300, lr = 0.1
I0510 18:01:33.961305 39557 solver.cpp:218] Iteration 20400 (3.75996 iter/s, 26.5961s/100 iters), loss = 6.89838
I0510 18:01:33.961474 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.89838 (* 1 = 6.89838 loss)
I0510 18:01:33.961489 39557 sgd_solver.cpp:105] Iteration 20400, lr = 0.1
I0510 18:02:00.542217 39557 solver.cpp:218] Iteration 20500 (3.76214 iter/s, 26.5806s/100 iters), loss = 6.82685
I0510 18:02:00.542276 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.82685 (* 1 = 6.82685 loss)
I0510 18:02:00.542289 39557 sgd_solver.cpp:105] Iteration 20500, lr = 0.1
I0510 18:02:27.205734 39557 solver.cpp:218] Iteration 20600 (3.75047 iter/s, 26.6633s/100 iters), loss = 7.5697
I0510 18:02:27.205925 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.5697 (* 1 = 7.5697 loss)
I0510 18:02:27.205945 39557 sgd_solver.cpp:105] Iteration 20600, lr = 0.1
I0510 18:02:53.860086 39557 solver.cpp:218] Iteration 20700 (3.75178 iter/s, 26.654s/100 iters), loss = 8.15413
I0510 18:02:53.860159 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.15413 (* 1 = 8.15413 loss)
I0510 18:02:53.860173 39557 sgd_solver.cpp:105] Iteration 20700, lr = 0.1
I0510 18:03:20.623761 39557 solver.cpp:218] Iteration 20800 (3.73644 iter/s, 26.7635s/100 iters), loss = 7.47673
I0510 18:03:20.623970 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.47673 (* 1 = 7.47673 loss)
I0510 18:03:20.623994 39557 sgd_solver.cpp:105] Iteration 20800, lr = 0.1
I0510 18:03:47.413617 39557 solver.cpp:218] Iteration 20900 (3.7328 iter/s, 26.7895s/100 iters), loss = 6.95681
I0510 18:03:47.413673 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.95681 (* 1 = 6.95681 loss)
I0510 18:03:47.413687 39557 sgd_solver.cpp:105] Iteration 20900, lr = 0.1
I0510 18:04:14.202431 39557 solver.cpp:218] Iteration 21000 (3.73293 iter/s, 26.7886s/100 iters), loss = 7.32313
I0510 18:04:14.202601 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.32313 (* 1 = 7.32313 loss)
I0510 18:04:14.202616 39557 sgd_solver.cpp:105] Iteration 21000, lr = 0.1
I0510 18:04:40.748853 39557 solver.cpp:218] Iteration 21100 (3.76703 iter/s, 26.5461s/100 iters), loss = 8.07178
I0510 18:04:40.748908 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.07178 (* 1 = 8.07178 loss)
I0510 18:04:40.748919 39557 sgd_solver.cpp:105] Iteration 21100, lr = 0.1
I0510 18:05:07.329373 39557 solver.cpp:218] Iteration 21200 (3.76218 iter/s, 26.5803s/100 iters), loss = 7.58878
I0510 18:05:07.329581 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.58878 (* 1 = 7.58878 loss)
I0510 18:05:07.329596 39557 sgd_solver.cpp:105] Iteration 21200, lr = 0.1
I0510 18:05:33.945528 39557 solver.cpp:218] Iteration 21300 (3.75716 iter/s, 26.6158s/100 iters), loss = 7.14925
I0510 18:05:33.945580 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.14925 (* 1 = 7.14925 loss)
I0510 18:05:33.945593 39557 sgd_solver.cpp:105] Iteration 21300, lr = 0.1
I0510 18:06:00.555016 39557 solver.cpp:218] Iteration 21400 (3.75808 iter/s, 26.6093s/100 iters), loss = 7.97776
I0510 18:06:00.555238 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.97776 (* 1 = 7.97776 loss)
I0510 18:06:00.555253 39557 sgd_solver.cpp:105] Iteration 21400, lr = 0.1
I0510 18:06:27.222843 39557 solver.cpp:218] Iteration 21500 (3.74989 iter/s, 26.6675s/100 iters), loss = 7.9095
I0510 18:06:27.222905 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.9095 (* 1 = 7.9095 loss)
I0510 18:06:27.222919 39557 sgd_solver.cpp:105] Iteration 21500, lr = 0.1
I0510 18:06:53.856611 39557 solver.cpp:218] Iteration 21600 (3.75466 iter/s, 26.6336s/100 iters), loss = 7.63173
I0510 18:06:53.856812 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.63173 (* 1 = 7.63173 loss)
I0510 18:06:53.856827 39557 sgd_solver.cpp:105] Iteration 21600, lr = 0.1
I0510 18:07:20.373492 39557 solver.cpp:218] Iteration 21700 (3.77123 iter/s, 26.5166s/100 iters), loss = 6.77752
I0510 18:07:20.373554 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.77752 (* 1 = 6.77752 loss)
I0510 18:07:20.373567 39557 sgd_solver.cpp:105] Iteration 21700, lr = 0.1
I0510 18:07:47.013948 39557 solver.cpp:218] Iteration 21800 (3.75372 iter/s, 26.6402s/100 iters), loss = 7.28933
I0510 18:07:47.014133 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.28933 (* 1 = 7.28933 loss)
I0510 18:07:47.014148 39557 sgd_solver.cpp:105] Iteration 21800, lr = 0.1
I0510 18:08:13.590740 39557 solver.cpp:218] Iteration 21900 (3.76273 iter/s, 26.5765s/100 iters), loss = 7.16876
I0510 18:08:13.590800 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.16876 (* 1 = 7.16876 loss)
I0510 18:08:13.590812 39557 sgd_solver.cpp:105] Iteration 21900, lr = 0.1
I0510 18:08:40.221771 39557 solver.cpp:218] Iteration 22000 (3.75504 iter/s, 26.6308s/100 iters), loss = 7.34515
I0510 18:08:40.221941 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.34515 (* 1 = 7.34515 loss)
I0510 18:08:40.221956 39557 sgd_solver.cpp:105] Iteration 22000, lr = 0.1
I0510 18:09:06.851765 39557 solver.cpp:218] Iteration 22100 (3.75521 iter/s, 26.6297s/100 iters), loss = 7.29772
I0510 18:09:06.851848 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.29772 (* 1 = 7.29772 loss)
I0510 18:09:06.851861 39557 sgd_solver.cpp:105] Iteration 22100, lr = 0.1
I0510 18:09:33.483994 39557 solver.cpp:218] Iteration 22200 (3.75488 iter/s, 26.632s/100 iters), loss = 7.21329
I0510 18:09:33.484166 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.21329 (* 1 = 7.21329 loss)
I0510 18:09:33.484182 39557 sgd_solver.cpp:105] Iteration 22200, lr = 0.1
I0510 18:10:00.051054 39557 solver.cpp:218] Iteration 22300 (3.7641 iter/s, 26.5668s/100 iters), loss = 7.5719
I0510 18:10:00.051115 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.5719 (* 1 = 7.5719 loss)
I0510 18:10:00.051127 39557 sgd_solver.cpp:105] Iteration 22300, lr = 0.1
I0510 18:10:26.628186 39557 solver.cpp:218] Iteration 22400 (3.76266 iter/s, 26.5769s/100 iters), loss = 7.38636
I0510 18:10:26.628371 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.38636 (* 1 = 7.38636 loss)
I0510 18:10:26.628386 39557 sgd_solver.cpp:105] Iteration 22400, lr = 0.1
I0510 18:10:53.298317 39557 solver.cpp:218] Iteration 22500 (3.74956 iter/s, 26.6698s/100 iters), loss = 7.88658
I0510 18:10:53.298382 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.88658 (* 1 = 7.88658 loss)
I0510 18:10:53.298395 39557 sgd_solver.cpp:105] Iteration 22500, lr = 0.1
I0510 18:11:19.892444 39557 solver.cpp:218] Iteration 22600 (3.76026 iter/s, 26.5939s/100 iters), loss = 7.67766
I0510 18:11:19.892678 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.67766 (* 1 = 7.67766 loss)
I0510 18:11:19.892693 39557 sgd_solver.cpp:105] Iteration 22600, lr = 0.1
I0510 18:11:46.508198 39557 solver.cpp:218] Iteration 22700 (3.75722 iter/s, 26.6154s/100 iters), loss = 7.12234
I0510 18:11:46.508262 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.12234 (* 1 = 7.12234 loss)
I0510 18:11:46.508276 39557 sgd_solver.cpp:105] Iteration 22700, lr = 0.1
I0510 18:12:13.203771 39557 solver.cpp:218] Iteration 22800 (3.74597 iter/s, 26.6954s/100 iters), loss = 7.07274
I0510 18:12:13.203961 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.07274 (* 1 = 7.07274 loss)
I0510 18:12:13.203977 39557 sgd_solver.cpp:105] Iteration 22800, lr = 0.1
I0510 18:12:39.805915 39557 solver.cpp:218] Iteration 22900 (3.75914 iter/s, 26.6018s/100 iters), loss = 7.52191
I0510 18:12:39.805979 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.52191 (* 1 = 7.52191 loss)
I0510 18:12:39.805995 39557 sgd_solver.cpp:105] Iteration 22900, lr = 0.1
I0510 18:13:06.367614 39557 solver.cpp:218] Iteration 23000 (3.76485 iter/s, 26.5615s/100 iters), loss = 7.56848
I0510 18:13:06.367796 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.56848 (* 1 = 7.56848 loss)
I0510 18:13:06.367811 39557 sgd_solver.cpp:105] Iteration 23000, lr = 0.1
I0510 18:13:32.919061 39557 solver.cpp:218] Iteration 23100 (3.76632 iter/s, 26.5511s/100 iters), loss = 6.80812
I0510 18:13:32.919122 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.80812 (* 1 = 6.80812 loss)
I0510 18:13:32.919136 39557 sgd_solver.cpp:105] Iteration 23100, lr = 0.1
I0510 18:13:59.481695 39557 solver.cpp:218] Iteration 23200 (3.76471 iter/s, 26.5624s/100 iters), loss = 7.40055
I0510 18:13:59.481871 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.40055 (* 1 = 7.40055 loss)
I0510 18:13:59.481886 39557 sgd_solver.cpp:105] Iteration 23200, lr = 0.1
I0510 18:14:26.229193 39557 solver.cpp:218] Iteration 23300 (3.73871 iter/s, 26.7472s/100 iters), loss = 6.48882
I0510 18:14:26.229256 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.48882 (* 1 = 6.48882 loss)
I0510 18:14:26.229269 39557 sgd_solver.cpp:105] Iteration 23300, lr = 0.1
I0510 18:14:52.848740 39557 solver.cpp:218] Iteration 23400 (3.75667 iter/s, 26.6193s/100 iters), loss = 7.53402
I0510 18:14:52.848945 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.53402 (* 1 = 7.53402 loss)
I0510 18:14:52.848961 39557 sgd_solver.cpp:105] Iteration 23400, lr = 0.1
I0510 18:15:19.580204 39557 solver.cpp:218] Iteration 23500 (3.74096 iter/s, 26.7311s/100 iters), loss = 7.12718
I0510 18:15:19.580269 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.12718 (* 1 = 7.12718 loss)
I0510 18:15:19.580286 39557 sgd_solver.cpp:105] Iteration 23500, lr = 0.1
I0510 18:15:46.179468 39557 solver.cpp:218] Iteration 23600 (3.75953 iter/s, 26.5991s/100 iters), loss = 7.2289
I0510 18:15:46.179657 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.2289 (* 1 = 7.2289 loss)
I0510 18:15:46.179673 39557 sgd_solver.cpp:105] Iteration 23600, lr = 0.1
I0510 18:16:12.812580 39557 solver.cpp:218] Iteration 23700 (3.75477 iter/s, 26.6328s/100 iters), loss = 7.6335
I0510 18:16:12.812641 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.6335 (* 1 = 7.6335 loss)
I0510 18:16:12.812655 39557 sgd_solver.cpp:105] Iteration 23700, lr = 0.1
I0510 18:16:39.424558 39557 solver.cpp:218] Iteration 23800 (3.75773 iter/s, 26.6118s/100 iters), loss = 6.2693
I0510 18:16:39.424754 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.2693 (* 1 = 6.2693 loss)
I0510 18:16:39.424769 39557 sgd_solver.cpp:105] Iteration 23800, lr = 0.1
I0510 18:17:06.075940 39557 solver.cpp:218] Iteration 23900 (3.7522 iter/s, 26.6511s/100 iters), loss = 7.28749
I0510 18:17:06.076004 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.28749 (* 1 = 7.28749 loss)
I0510 18:17:06.076017 39557 sgd_solver.cpp:105] Iteration 23900, lr = 0.1
I0510 18:17:32.678036 39557 solver.cpp:218] Iteration 24000 (3.75913 iter/s, 26.6019s/100 iters), loss = 7.40634
I0510 18:17:32.678252 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.40634 (* 1 = 7.40634 loss)
I0510 18:17:32.678268 39557 sgd_solver.cpp:105] Iteration 24000, lr = 0.1
I0510 18:17:59.331904 39557 solver.cpp:218] Iteration 24100 (3.75185 iter/s, 26.6535s/100 iters), loss = 7.62761
I0510 18:17:59.331965 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.62761 (* 1 = 7.62761 loss)
I0510 18:17:59.331979 39557 sgd_solver.cpp:105] Iteration 24100, lr = 0.1
I0510 18:18:26.057638 39557 solver.cpp:218] Iteration 24200 (3.74174 iter/s, 26.7255s/100 iters), loss = 7.62122
I0510 18:18:26.057806 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.62122 (* 1 = 7.62122 loss)
I0510 18:18:26.057822 39557 sgd_solver.cpp:105] Iteration 24200, lr = 0.1
I0510 18:18:52.790444 39557 solver.cpp:218] Iteration 24300 (3.74077 iter/s, 26.7325s/100 iters), loss = 6.93072
I0510 18:18:52.790506 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.93072 (* 1 = 6.93072 loss)
I0510 18:18:52.790521 39557 sgd_solver.cpp:105] Iteration 24300, lr = 0.1
I0510 18:19:20.685554 39557 solver.cpp:218] Iteration 24400 (3.58488 iter/s, 27.8949s/100 iters), loss = 6.5711
I0510 18:19:20.685732 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.5711 (* 1 = 6.5711 loss)
I0510 18:19:20.685747 39557 sgd_solver.cpp:105] Iteration 24400, lr = 0.1
I0510 18:19:47.247998 39557 solver.cpp:218] Iteration 24500 (3.76476 iter/s, 26.5621s/100 iters), loss = 7.6734
I0510 18:19:47.248061 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.6734 (* 1 = 7.6734 loss)
I0510 18:19:47.248073 39557 sgd_solver.cpp:105] Iteration 24500, lr = 0.1
I0510 18:20:13.882761 39557 solver.cpp:218] Iteration 24600 (3.75452 iter/s, 26.6346s/100 iters), loss = 6.51681
I0510 18:20:13.882941 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.51681 (* 1 = 6.51681 loss)
I0510 18:20:13.882959 39557 sgd_solver.cpp:105] Iteration 24600, lr = 0.1
I0510 18:20:40.531116 39557 solver.cpp:218] Iteration 24700 (3.75262 iter/s, 26.648s/100 iters), loss = 6.6968
I0510 18:20:40.531177 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.6968 (* 1 = 6.6968 loss)
I0510 18:20:40.531189 39557 sgd_solver.cpp:105] Iteration 24700, lr = 0.1
I0510 18:21:07.273454 39557 solver.cpp:218] Iteration 24800 (3.73942 iter/s, 26.7421s/100 iters), loss = 7.79642
I0510 18:21:07.273634 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.79642 (* 1 = 7.79642 loss)
I0510 18:21:07.273651 39557 sgd_solver.cpp:105] Iteration 24800, lr = 0.1
I0510 18:21:33.923835 39557 solver.cpp:218] Iteration 24900 (3.75234 iter/s, 26.6501s/100 iters), loss = 7.39934
I0510 18:21:33.923895 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.39934 (* 1 = 7.39934 loss)
I0510 18:21:33.923910 39557 sgd_solver.cpp:105] Iteration 24900, lr = 0.1
I0510 18:22:00.401412 39557 solver.cpp:447] Snapshotting to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_25000.caffemodel
I0510 18:22:00.439326 39557 sgd_solver.cpp:273] Snapshotting solver state to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_25000.solverstate
I0510 18:22:00.721351 39557 solver.cpp:218] Iteration 25000 (3.73172 iter/s, 26.7973s/100 iters), loss = 7.10974
I0510 18:22:00.721406 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.10974 (* 1 = 7.10974 loss)
I0510 18:22:00.721417 39557 sgd_solver.cpp:105] Iteration 25000, lr = 0.1
I0510 18:22:27.333663 39557 solver.cpp:218] Iteration 25100 (3.75769 iter/s, 26.6121s/100 iters), loss = 7.50078
I0510 18:22:27.333725 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.50078 (* 1 = 7.50078 loss)
I0510 18:22:27.333739 39557 sgd_solver.cpp:105] Iteration 25100, lr = 0.1
I0510 18:22:53.981875 39557 solver.cpp:218] Iteration 25200 (3.75263 iter/s, 26.648s/100 iters), loss = 6.76206
I0510 18:22:53.982077 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.76206 (* 1 = 6.76206 loss)
I0510 18:22:53.982094 39557 sgd_solver.cpp:105] Iteration 25200, lr = 0.1
I0510 18:23:20.566256 39557 solver.cpp:218] Iteration 25300 (3.76165 iter/s, 26.584s/100 iters), loss = 6.70895
I0510 18:23:20.566318 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.70895 (* 1 = 6.70895 loss)
I0510 18:23:20.566330 39557 sgd_solver.cpp:105] Iteration 25300, lr = 0.1
I0510 18:23:47.095479 39557 solver.cpp:218] Iteration 25400 (3.76946 iter/s, 26.529s/100 iters), loss = 7.11814
I0510 18:23:47.095647 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.11814 (* 1 = 7.11814 loss)
I0510 18:23:47.095662 39557 sgd_solver.cpp:105] Iteration 25400, lr = 0.1
I0510 18:24:13.674129 39557 solver.cpp:218] Iteration 25500 (3.76246 iter/s, 26.5783s/100 iters), loss = 7.15488
I0510 18:24:13.674192 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.15488 (* 1 = 7.15488 loss)
I0510 18:24:13.674207 39557 sgd_solver.cpp:105] Iteration 25500, lr = 0.1
I0510 18:24:40.209029 39557 solver.cpp:218] Iteration 25600 (3.76865 iter/s, 26.5347s/100 iters), loss = 7.97129
I0510 18:24:40.209226 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.97129 (* 1 = 7.97129 loss)
I0510 18:24:40.209242 39557 sgd_solver.cpp:105] Iteration 25600, lr = 0.1
I0510 18:25:06.988533 39557 solver.cpp:218] Iteration 25700 (3.73424 iter/s, 26.7792s/100 iters), loss = 6.77561
I0510 18:25:06.988595 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.77561 (* 1 = 6.77561 loss)
I0510 18:25:06.988607 39557 sgd_solver.cpp:105] Iteration 25700, lr = 0.1
I0510 18:25:33.659813 39557 solver.cpp:218] Iteration 25800 (3.74938 iter/s, 26.6711s/100 iters), loss = 8.58176
I0510 18:25:33.660043 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.58176 (* 1 = 8.58176 loss)
I0510 18:25:33.660058 39557 sgd_solver.cpp:105] Iteration 25800, lr = 0.1
I0510 18:26:00.211604 39557 solver.cpp:218] Iteration 25900 (3.76628 iter/s, 26.5514s/100 iters), loss = 7.34337
I0510 18:26:00.211665 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.34337 (* 1 = 7.34337 loss)
I0510 18:26:00.211678 39557 sgd_solver.cpp:105] Iteration 25900, lr = 0.1
I0510 18:26:26.804935 39557 solver.cpp:218] Iteration 26000 (3.76037 iter/s, 26.5931s/100 iters), loss = 7.39856
I0510 18:26:26.805115 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.39856 (* 1 = 7.39856 loss)
I0510 18:26:26.805131 39557 sgd_solver.cpp:105] Iteration 26000, lr = 0.1
I0510 18:26:53.382750 39557 solver.cpp:218] Iteration 26100 (3.76258 iter/s, 26.5775s/100 iters), loss = 7.12957
I0510 18:26:53.382809 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.12957 (* 1 = 7.12957 loss)
I0510 18:26:53.382822 39557 sgd_solver.cpp:105] Iteration 26100, lr = 0.1
I0510 18:27:20.000013 39557 solver.cpp:218] Iteration 26200 (3.75699 iter/s, 26.6171s/100 iters), loss = 6.83253
I0510 18:27:20.000210 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.83253 (* 1 = 6.83253 loss)
I0510 18:27:20.000224 39557 sgd_solver.cpp:105] Iteration 26200, lr = 0.1
I0510 18:27:46.691812 39557 solver.cpp:218] Iteration 26300 (3.74651 iter/s, 26.6915s/100 iters), loss = 6.85833
I0510 18:27:46.691875 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.85833 (* 1 = 6.85833 loss)
I0510 18:27:46.691889 39557 sgd_solver.cpp:105] Iteration 26300, lr = 0.1
I0510 18:28:13.292361 39557 solver.cpp:218] Iteration 26400 (3.75934 iter/s, 26.6004s/100 iters), loss = 7.06032
I0510 18:28:13.292544 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.06032 (* 1 = 7.06032 loss)
I0510 18:28:13.292558 39557 sgd_solver.cpp:105] Iteration 26400, lr = 0.1
I0510 18:28:39.827600 39557 solver.cpp:218] Iteration 26500 (3.76861 iter/s, 26.535s/100 iters), loss = 7.02872
I0510 18:28:39.827661 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.02872 (* 1 = 7.02872 loss)
I0510 18:28:39.827675 39557 sgd_solver.cpp:105] Iteration 26500, lr = 0.1
I0510 18:29:06.367089 39557 solver.cpp:218] Iteration 26600 (3.76799 iter/s, 26.5393s/100 iters), loss = 6.70544
I0510 18:29:06.367281 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.70544 (* 1 = 6.70544 loss)
I0510 18:29:06.367297 39557 sgd_solver.cpp:105] Iteration 26600, lr = 0.1
I0510 18:29:33.023141 39557 solver.cpp:218] Iteration 26700 (3.75153 iter/s, 26.6558s/100 iters), loss = 6.98966
I0510 18:29:33.023205 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.98966 (* 1 = 6.98966 loss)
I0510 18:29:33.023221 39557 sgd_solver.cpp:105] Iteration 26700, lr = 0.1
I0510 18:29:59.574815 39557 solver.cpp:218] Iteration 26800 (3.76626 iter/s, 26.5515s/100 iters), loss = 6.5942
I0510 18:29:59.574985 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.5942 (* 1 = 6.5942 loss)
I0510 18:29:59.575001 39557 sgd_solver.cpp:105] Iteration 26800, lr = 0.1
I0510 18:30:26.324064 39557 solver.cpp:218] Iteration 26900 (3.73846 iter/s, 26.749s/100 iters), loss = 6.85168
I0510 18:30:26.324126 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.85168 (* 1 = 6.85168 loss)
I0510 18:30:26.324141 39557 sgd_solver.cpp:105] Iteration 26900, lr = 0.1
I0510 18:30:53.019860 39557 solver.cpp:218] Iteration 27000 (3.74593 iter/s, 26.6956s/100 iters), loss = 7.3679
I0510 18:30:53.020061 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.3679 (* 1 = 7.3679 loss)
I0510 18:30:53.020076 39557 sgd_solver.cpp:105] Iteration 27000, lr = 0.1
I0510 18:31:19.647925 39557 solver.cpp:218] Iteration 27100 (3.75548 iter/s, 26.6278s/100 iters), loss = 7.14305
I0510 18:31:19.647984 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.14305 (* 1 = 7.14305 loss)
I0510 18:31:19.647996 39557 sgd_solver.cpp:105] Iteration 27100, lr = 0.1
I0510 18:31:46.359771 39557 solver.cpp:218] Iteration 27200 (3.74368 iter/s, 26.7117s/100 iters), loss = 6.79671
I0510 18:31:46.359956 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.79671 (* 1 = 6.79671 loss)
I0510 18:31:46.359972 39557 sgd_solver.cpp:105] Iteration 27200, lr = 0.1
I0510 18:32:12.991510 39557 solver.cpp:218] Iteration 27300 (3.75496 iter/s, 26.6315s/100 iters), loss = 6.24667
I0510 18:32:12.991575 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.24667 (* 1 = 6.24667 loss)
I0510 18:32:12.991590 39557 sgd_solver.cpp:105] Iteration 27300, lr = 0.1
I0510 18:32:39.687614 39557 solver.cpp:218] Iteration 27400 (3.74589 iter/s, 26.6959s/100 iters), loss = 8.06394
I0510 18:32:39.687826 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.06394 (* 1 = 8.06394 loss)
I0510 18:32:39.687841 39557 sgd_solver.cpp:105] Iteration 27400, lr = 0.1
I0510 18:33:06.285827 39557 solver.cpp:218] Iteration 27500 (3.7597 iter/s, 26.5979s/100 iters), loss = 7.38201
I0510 18:33:06.285890 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.38201 (* 1 = 7.38201 loss)
I0510 18:33:06.285903 39557 sgd_solver.cpp:105] Iteration 27500, lr = 0.1
I0510 18:33:32.923626 39557 solver.cpp:218] Iteration 27600 (3.75409 iter/s, 26.6376s/100 iters), loss = 6.52462
I0510 18:33:32.923815 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.52462 (* 1 = 6.52462 loss)
I0510 18:33:32.923828 39557 sgd_solver.cpp:105] Iteration 27600, lr = 0.1
I0510 18:33:59.520475 39557 solver.cpp:218] Iteration 27700 (3.75989 iter/s, 26.5965s/100 iters), loss = 6.43808
I0510 18:33:59.520534 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.43808 (* 1 = 6.43808 loss)
I0510 18:33:59.520547 39557 sgd_solver.cpp:105] Iteration 27700, lr = 0.1
I0510 18:34:26.265681 39557 solver.cpp:218] Iteration 27800 (3.73901 iter/s, 26.745s/100 iters), loss = 7.1297
I0510 18:34:26.265867 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.1297 (* 1 = 7.1297 loss)
I0510 18:34:26.265885 39557 sgd_solver.cpp:105] Iteration 27800, lr = 0.1
I0510 18:34:53.091255 39557 solver.cpp:218] Iteration 27900 (3.72783 iter/s, 26.8253s/100 iters), loss = 7.1657
I0510 18:34:53.091320 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.1657 (* 1 = 7.1657 loss)
I0510 18:34:53.091346 39557 sgd_solver.cpp:105] Iteration 27900, lr = 0.1
I0510 18:35:19.700618 39557 solver.cpp:218] Iteration 28000 (3.7581 iter/s, 26.6092s/100 iters), loss = 7.62944
I0510 18:35:19.700848 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.62944 (* 1 = 7.62944 loss)
I0510 18:35:19.700865 39557 sgd_solver.cpp:105] Iteration 28000, lr = 0.1
I0510 18:35:46.452587 39557 solver.cpp:218] Iteration 28100 (3.73809 iter/s, 26.7516s/100 iters), loss = 7.08231
I0510 18:35:46.452647 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.08231 (* 1 = 7.08231 loss)
I0510 18:35:46.452659 39557 sgd_solver.cpp:105] Iteration 28100, lr = 0.1
I0510 18:36:13.123750 39557 solver.cpp:218] Iteration 28200 (3.74939 iter/s, 26.671s/100 iters), loss = 7.51702
I0510 18:36:13.124102 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.51702 (* 1 = 7.51702 loss)
I0510 18:36:13.124117 39557 sgd_solver.cpp:105] Iteration 28200, lr = 0.1
I0510 18:36:39.863009 39557 solver.cpp:218] Iteration 28300 (3.73988 iter/s, 26.7388s/100 iters), loss = 6.64265
I0510 18:36:39.863061 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.64265 (* 1 = 6.64265 loss)
I0510 18:36:39.863075 39557 sgd_solver.cpp:105] Iteration 28300, lr = 0.1
I0510 18:37:06.503150 39557 solver.cpp:218] Iteration 28400 (3.75376 iter/s, 26.64s/100 iters), loss = 6.63651
I0510 18:37:06.503360 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.63651 (* 1 = 6.63651 loss)
I0510 18:37:06.503377 39557 sgd_solver.cpp:105] Iteration 28400, lr = 0.1
I0510 18:37:33.040701 39557 solver.cpp:218] Iteration 28500 (3.76829 iter/s, 26.5372s/100 iters), loss = 7.16992
I0510 18:37:33.040763 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.16992 (* 1 = 7.16992 loss)
I0510 18:37:33.040778 39557 sgd_solver.cpp:105] Iteration 28500, lr = 0.1
I0510 18:37:59.726378 39557 solver.cpp:218] Iteration 28600 (3.74735 iter/s, 26.6855s/100 iters), loss = 7.97017
I0510 18:37:59.726611 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.97017 (* 1 = 7.97017 loss)
I0510 18:37:59.726627 39557 sgd_solver.cpp:105] Iteration 28600, lr = 0.1
I0510 18:38:26.323750 39557 solver.cpp:218] Iteration 28700 (3.75982 iter/s, 26.597s/100 iters), loss = 7.88988
I0510 18:38:26.323812 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.88988 (* 1 = 7.88988 loss)
I0510 18:38:26.323825 39557 sgd_solver.cpp:105] Iteration 28700, lr = 0.1
I0510 18:38:52.927341 39557 solver.cpp:218] Iteration 28800 (3.75892 iter/s, 26.6034s/100 iters), loss = 7.4174
I0510 18:38:52.927548 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.4174 (* 1 = 7.4174 loss)
I0510 18:38:52.927564 39557 sgd_solver.cpp:105] Iteration 28800, lr = 0.1
I0510 18:39:19.540499 39557 solver.cpp:218] Iteration 28900 (3.75759 iter/s, 26.6128s/100 iters), loss = 8.04833
I0510 18:39:19.540561 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.04833 (* 1 = 8.04833 loss)
I0510 18:39:19.540575 39557 sgd_solver.cpp:105] Iteration 28900, lr = 0.1
I0510 18:39:46.108460 39557 solver.cpp:218] Iteration 29000 (3.76396 iter/s, 26.5678s/100 iters), loss = 6.9871
I0510 18:39:46.108650 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.9871 (* 1 = 6.9871 loss)
I0510 18:39:46.108666 39557 sgd_solver.cpp:105] Iteration 29000, lr = 0.1
I0510 18:40:12.757808 39557 solver.cpp:218] Iteration 29100 (3.75248 iter/s, 26.649s/100 iters), loss = 7.88581
I0510 18:40:12.757869 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.88581 (* 1 = 7.88581 loss)
I0510 18:40:12.757881 39557 sgd_solver.cpp:105] Iteration 29100, lr = 0.1
I0510 18:40:39.350497 39557 solver.cpp:218] Iteration 29200 (3.76046 iter/s, 26.5925s/100 iters), loss = 7.55036
I0510 18:40:39.350704 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.55036 (* 1 = 7.55036 loss)
I0510 18:40:39.350719 39557 sgd_solver.cpp:105] Iteration 29200, lr = 0.1
I0510 18:41:05.957942 39557 solver.cpp:218] Iteration 29300 (3.75839 iter/s, 26.6071s/100 iters), loss = 6.25441
I0510 18:41:05.958004 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.25441 (* 1 = 6.25441 loss)
I0510 18:41:05.958017 39557 sgd_solver.cpp:105] Iteration 29300, lr = 0.1
I0510 18:41:32.577422 39557 solver.cpp:218] Iteration 29400 (3.75667 iter/s, 26.6193s/100 iters), loss = 7.05555
I0510 18:41:32.577636 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.05555 (* 1 = 7.05555 loss)
I0510 18:41:32.577652 39557 sgd_solver.cpp:105] Iteration 29400, lr = 0.1
I0510 18:41:59.174650 39557 solver.cpp:218] Iteration 29500 (3.75984 iter/s, 26.5969s/100 iters), loss = 7.30777
I0510 18:41:59.174710 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.30777 (* 1 = 7.30777 loss)
I0510 18:41:59.174721 39557 sgd_solver.cpp:105] Iteration 29500, lr = 0.1
I0510 18:42:25.895051 39557 solver.cpp:218] Iteration 29600 (3.74249 iter/s, 26.7202s/100 iters), loss = 7.76378
I0510 18:42:25.895238 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.76378 (* 1 = 7.76378 loss)
I0510 18:42:25.895256 39557 sgd_solver.cpp:105] Iteration 29600, lr = 0.1
I0510 18:42:52.614089 39557 solver.cpp:218] Iteration 29700 (3.74269 iter/s, 26.7187s/100 iters), loss = 7.12365
I0510 18:42:52.614151 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.12365 (* 1 = 7.12365 loss)
I0510 18:42:52.614164 39557 sgd_solver.cpp:105] Iteration 29700, lr = 0.1
I0510 18:43:19.259848 39557 solver.cpp:218] Iteration 29800 (3.75297 iter/s, 26.6456s/100 iters), loss = 7.78356
I0510 18:43:19.260032 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.78356 (* 1 = 7.78356 loss)
I0510 18:43:19.260051 39557 sgd_solver.cpp:105] Iteration 29800, lr = 0.1
I0510 18:43:45.982992 39557 solver.cpp:218] Iteration 29900 (3.74212 iter/s, 26.7228s/100 iters), loss = 6.8636
I0510 18:43:45.983054 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.8636 (* 1 = 6.8636 loss)
I0510 18:43:45.983068 39557 sgd_solver.cpp:105] Iteration 29900, lr = 0.1
I0510 18:44:12.350177 39557 solver.cpp:447] Snapshotting to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_30000.caffemodel
I0510 18:44:12.388469 39557 sgd_solver.cpp:273] Snapshotting solver state to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_30000.solverstate
I0510 18:44:12.670953 39557 solver.cpp:218] Iteration 30000 (3.74703 iter/s, 26.6878s/100 iters), loss = 6.54331
I0510 18:44:12.671005 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.54331 (* 1 = 6.54331 loss)
I0510 18:44:12.671016 39557 sgd_solver.cpp:105] Iteration 30000, lr = 0.1
I0510 18:44:39.371136 39557 solver.cpp:218] Iteration 30100 (3.74532 iter/s, 26.7s/100 iters), loss = 6.88831
I0510 18:44:39.371199 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.88831 (* 1 = 6.88831 loss)
I0510 18:44:39.371212 39557 sgd_solver.cpp:105] Iteration 30100, lr = 0.1
I0510 18:45:06.173063 39557 solver.cpp:218] Iteration 30200 (3.7311 iter/s, 26.8017s/100 iters), loss = 7.60534
I0510 18:45:06.173815 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.60534 (* 1 = 7.60534 loss)
I0510 18:45:06.173830 39557 sgd_solver.cpp:105] Iteration 30200, lr = 0.1
I0510 18:45:32.882779 39557 solver.cpp:218] Iteration 30300 (3.74408 iter/s, 26.7088s/100 iters), loss = 6.59359
I0510 18:45:32.882840 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.59359 (* 1 = 6.59359 loss)
I0510 18:45:32.882853 39557 sgd_solver.cpp:105] Iteration 30300, lr = 0.1
I0510 18:45:59.469528 39557 solver.cpp:218] Iteration 30400 (3.7613 iter/s, 26.5866s/100 iters), loss = 7.38913
I0510 18:45:59.469703 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.38913 (* 1 = 7.38913 loss)
I0510 18:45:59.469719 39557 sgd_solver.cpp:105] Iteration 30400, lr = 0.1
I0510 18:46:25.976024 39557 solver.cpp:218] Iteration 30500 (3.7727 iter/s, 26.5062s/100 iters), loss = 6.9993
I0510 18:46:25.976083 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.9993 (* 1 = 6.9993 loss)
I0510 18:46:25.976095 39557 sgd_solver.cpp:105] Iteration 30500, lr = 0.1
I0510 18:46:52.584306 39557 solver.cpp:218] Iteration 30600 (3.75826 iter/s, 26.6081s/100 iters), loss = 6.82206
I0510 18:46:52.584513 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.82206 (* 1 = 6.82206 loss)
I0510 18:46:52.584529 39557 sgd_solver.cpp:105] Iteration 30600, lr = 0.1
I0510 18:47:19.308884 39557 solver.cpp:218] Iteration 30700 (3.74192 iter/s, 26.7242s/100 iters), loss = 7.83129
I0510 18:47:19.308944 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.83129 (* 1 = 7.83129 loss)
I0510 18:47:19.308959 39557 sgd_solver.cpp:105] Iteration 30700, lr = 0.1
I0510 18:47:45.931769 39557 solver.cpp:218] Iteration 30800 (3.75619 iter/s, 26.6227s/100 iters), loss = 7.84292
I0510 18:47:45.931946 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.84292 (* 1 = 7.84292 loss)
I0510 18:47:45.931964 39557 sgd_solver.cpp:105] Iteration 30800, lr = 0.1
I0510 18:48:12.492090 39557 solver.cpp:218] Iteration 30900 (3.76506 iter/s, 26.56s/100 iters), loss = 6.54673
I0510 18:48:12.492149 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.54673 (* 1 = 6.54673 loss)
I0510 18:48:12.492162 39557 sgd_solver.cpp:105] Iteration 30900, lr = 0.1
I0510 18:48:39.024602 39557 solver.cpp:218] Iteration 31000 (3.76899 iter/s, 26.5323s/100 iters), loss = 7.41804
I0510 18:48:39.024776 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.41804 (* 1 = 7.41804 loss)
I0510 18:48:39.024792 39557 sgd_solver.cpp:105] Iteration 31000, lr = 0.1
I0510 18:49:05.695816 39557 solver.cpp:218] Iteration 31100 (3.7494 iter/s, 26.6709s/100 iters), loss = 7.5951
I0510 18:49:05.695878 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.5951 (* 1 = 7.5951 loss)
I0510 18:49:05.695891 39557 sgd_solver.cpp:105] Iteration 31100, lr = 0.1
I0510 18:49:32.322855 39557 solver.cpp:218] Iteration 31200 (3.75561 iter/s, 26.6268s/100 iters), loss = 6.68962
I0510 18:49:32.323031 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.68962 (* 1 = 6.68962 loss)
I0510 18:49:32.323048 39557 sgd_solver.cpp:105] Iteration 31200, lr = 0.1
I0510 18:49:59.047175 39557 solver.cpp:218] Iteration 31300 (3.74195 iter/s, 26.724s/100 iters), loss = 7.1107
I0510 18:49:59.047225 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.1107 (* 1 = 7.1107 loss)
I0510 18:49:59.047238 39557 sgd_solver.cpp:105] Iteration 31300, lr = 0.1
I0510 18:50:25.699415 39557 solver.cpp:218] Iteration 31400 (3.75206 iter/s, 26.6521s/100 iters), loss = 7.53646
I0510 18:50:25.699607 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.53646 (* 1 = 7.53646 loss)
I0510 18:50:25.699622 39557 sgd_solver.cpp:105] Iteration 31400, lr = 0.1
I0510 18:50:52.316552 39557 solver.cpp:218] Iteration 31500 (3.75702 iter/s, 26.6168s/100 iters), loss = 7.39815
I0510 18:50:52.316601 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.39815 (* 1 = 7.39815 loss)
I0510 18:50:52.316614 39557 sgd_solver.cpp:105] Iteration 31500, lr = 0.1
I0510 18:51:18.993288 39557 solver.cpp:218] Iteration 31600 (3.74861 iter/s, 26.6765s/100 iters), loss = 6.69167
I0510 18:51:18.993481 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.69167 (* 1 = 6.69167 loss)
I0510 18:51:18.993497 39557 sgd_solver.cpp:105] Iteration 31600, lr = 0.1
I0510 18:51:45.674995 39557 solver.cpp:218] Iteration 31700 (3.74793 iter/s, 26.6814s/100 iters), loss = 7.31182
I0510 18:51:45.675045 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.31182 (* 1 = 7.31182 loss)
I0510 18:51:45.675057 39557 sgd_solver.cpp:105] Iteration 31700, lr = 0.1
I0510 18:52:12.390305 39557 solver.cpp:218] Iteration 31800 (3.7432 iter/s, 26.7151s/100 iters), loss = 6.96759
I0510 18:52:12.390496 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.96759 (* 1 = 6.96759 loss)
I0510 18:52:12.390512 39557 sgd_solver.cpp:105] Iteration 31800, lr = 0.1
I0510 18:52:39.072633 39557 solver.cpp:218] Iteration 31900 (3.74784 iter/s, 26.682s/100 iters), loss = 7.22152
I0510 18:52:39.072695 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.22152 (* 1 = 7.22152 loss)
I0510 18:52:39.072707 39557 sgd_solver.cpp:105] Iteration 31900, lr = 0.1
I0510 18:53:05.738339 39557 solver.cpp:218] Iteration 32000 (3.75016 iter/s, 26.6655s/100 iters), loss = 6.95697
I0510 18:53:05.738554 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.95697 (* 1 = 6.95697 loss)
I0510 18:53:05.738570 39557 sgd_solver.cpp:105] Iteration 32000, lr = 0.1
I0510 18:53:32.318472 39557 solver.cpp:218] Iteration 32100 (3.76226 iter/s, 26.5798s/100 iters), loss = 7.18567
I0510 18:53:32.318538 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.18567 (* 1 = 7.18567 loss)
I0510 18:53:32.318550 39557 sgd_solver.cpp:105] Iteration 32100, lr = 0.1
I0510 18:53:59.003455 39557 solver.cpp:218] Iteration 32200 (3.74745 iter/s, 26.6848s/100 iters), loss = 7.46232
I0510 18:53:59.003638 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.46232 (* 1 = 7.46232 loss)
I0510 18:53:59.003654 39557 sgd_solver.cpp:105] Iteration 32200, lr = 0.1
I0510 18:54:25.635800 39557 solver.cpp:218] Iteration 32300 (3.75488 iter/s, 26.632s/100 iters), loss = 6.98867
I0510 18:54:25.635859 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.98867 (* 1 = 6.98867 loss)
I0510 18:54:25.635870 39557 sgd_solver.cpp:105] Iteration 32300, lr = 0.1
I0510 18:54:52.356575 39557 solver.cpp:218] Iteration 32400 (3.74243 iter/s, 26.7206s/100 iters), loss = 7.1179
I0510 18:54:52.356737 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.1179 (* 1 = 7.1179 loss)
I0510 18:54:52.356755 39557 sgd_solver.cpp:105] Iteration 32400, lr = 0.1
I0510 18:55:18.970844 39557 solver.cpp:218] Iteration 32500 (3.75742 iter/s, 26.614s/100 iters), loss = 7.4822
I0510 18:55:18.970906 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.4822 (* 1 = 7.4822 loss)
I0510 18:55:18.970917 39557 sgd_solver.cpp:105] Iteration 32500, lr = 0.1
I0510 18:55:45.563789 39557 solver.cpp:218] Iteration 32600 (3.76042 iter/s, 26.5928s/100 iters), loss = 6.88317
I0510 18:55:45.563974 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.88317 (* 1 = 6.88317 loss)
I0510 18:55:45.563989 39557 sgd_solver.cpp:105] Iteration 32600, lr = 0.1
I0510 18:56:13.452430 39557 solver.cpp:218] Iteration 32700 (3.58573 iter/s, 27.8883s/100 iters), loss = 7.89535
I0510 18:56:13.452491 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.89535 (* 1 = 7.89535 loss)
I0510 18:56:13.452507 39557 sgd_solver.cpp:105] Iteration 32700, lr = 0.1
I0510 18:56:40.031675 39557 solver.cpp:218] Iteration 32800 (3.76236 iter/s, 26.579s/100 iters), loss = 7.32856
I0510 18:56:40.031842 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.32856 (* 1 = 7.32856 loss)
I0510 18:56:40.031857 39557 sgd_solver.cpp:105] Iteration 32800, lr = 0.1
I0510 18:57:06.741703 39557 solver.cpp:218] Iteration 32900 (3.74396 iter/s, 26.7097s/100 iters), loss = 7.11518
I0510 18:57:06.741765 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.11518 (* 1 = 7.11518 loss)
I0510 18:57:06.741780 39557 sgd_solver.cpp:105] Iteration 32900, lr = 0.1
I0510 18:57:33.421150 39557 solver.cpp:218] Iteration 33000 (3.74823 iter/s, 26.6793s/100 iters), loss = 6.2815
I0510 18:57:33.421363 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.2815 (* 1 = 6.2815 loss)
I0510 18:57:33.421380 39557 sgd_solver.cpp:105] Iteration 33000, lr = 0.1
I0510 18:58:00.047642 39557 solver.cpp:218] Iteration 33100 (3.75571 iter/s, 26.6262s/100 iters), loss = 7.3588
I0510 18:58:00.047703 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.3588 (* 1 = 7.3588 loss)
I0510 18:58:00.047715 39557 sgd_solver.cpp:105] Iteration 33100, lr = 0.1
I0510 18:58:26.658339 39557 solver.cpp:218] Iteration 33200 (3.75791 iter/s, 26.6105s/100 iters), loss = 7.00557
I0510 18:58:26.658542 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.00557 (* 1 = 7.00557 loss)
I0510 18:58:26.658557 39557 sgd_solver.cpp:105] Iteration 33200, lr = 0.1
I0510 18:58:53.240463 39557 solver.cpp:218] Iteration 33300 (3.76197 iter/s, 26.5818s/100 iters), loss = 7.49961
I0510 18:58:53.240525 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.49961 (* 1 = 7.49961 loss)
I0510 18:58:53.240538 39557 sgd_solver.cpp:105] Iteration 33300, lr = 0.1
I0510 18:59:19.785518 39557 solver.cpp:218] Iteration 33400 (3.76721 iter/s, 26.5449s/100 iters), loss = 6.9401
I0510 18:59:19.785768 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.9401 (* 1 = 6.9401 loss)
I0510 18:59:19.785784 39557 sgd_solver.cpp:105] Iteration 33400, lr = 0.1
I0510 18:59:46.411777 39557 solver.cpp:218] Iteration 33500 (3.75574 iter/s, 26.6259s/100 iters), loss = 6.11081
I0510 18:59:46.411835 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.11081 (* 1 = 6.11081 loss)
I0510 18:59:46.411849 39557 sgd_solver.cpp:105] Iteration 33500, lr = 0.1
I0510 19:00:13.103319 39557 solver.cpp:218] Iteration 33600 (3.74653 iter/s, 26.6914s/100 iters), loss = 7.48478
I0510 19:00:13.103520 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.48478 (* 1 = 7.48478 loss)
I0510 19:00:13.103539 39557 sgd_solver.cpp:105] Iteration 33600, lr = 0.1
I0510 19:00:39.818966 39557 solver.cpp:218] Iteration 33700 (3.74317 iter/s, 26.7153s/100 iters), loss = 7.60659
I0510 19:00:39.819026 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.60659 (* 1 = 7.60659 loss)
I0510 19:00:39.819038 39557 sgd_solver.cpp:105] Iteration 33700, lr = 0.1
I0510 19:01:06.418196 39557 solver.cpp:218] Iteration 33800 (3.75953 iter/s, 26.599s/100 iters), loss = 8.07172
I0510 19:01:06.418381 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.07172 (* 1 = 8.07172 loss)
I0510 19:01:06.418400 39557 sgd_solver.cpp:105] Iteration 33800, lr = 0.1
I0510 19:01:33.044474 39557 solver.cpp:218] Iteration 33900 (3.75573 iter/s, 26.626s/100 iters), loss = 6.75512
I0510 19:01:33.044538 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.75512 (* 1 = 6.75512 loss)
I0510 19:01:33.044550 39557 sgd_solver.cpp:105] Iteration 33900, lr = 0.1
I0510 19:01:59.701582 39557 solver.cpp:218] Iteration 34000 (3.75136 iter/s, 26.657s/100 iters), loss = 7.2582
I0510 19:01:59.701761 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.2582 (* 1 = 7.2582 loss)
I0510 19:01:59.701776 39557 sgd_solver.cpp:105] Iteration 34000, lr = 0.1
I0510 19:02:26.302433 39557 solver.cpp:218] Iteration 34100 (3.75931 iter/s, 26.6006s/100 iters), loss = 7.41684
I0510 19:02:26.302494 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.41684 (* 1 = 7.41684 loss)
I0510 19:02:26.302506 39557 sgd_solver.cpp:105] Iteration 34100, lr = 0.1
I0510 19:02:52.886831 39557 solver.cpp:218] Iteration 34200 (3.76162 iter/s, 26.5843s/100 iters), loss = 6.85679
I0510 19:02:52.887004 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.85679 (* 1 = 6.85679 loss)
I0510 19:02:52.887019 39557 sgd_solver.cpp:105] Iteration 34200, lr = 0.1
I0510 19:03:19.553469 39557 solver.cpp:218] Iteration 34300 (3.75004 iter/s, 26.6664s/100 iters), loss = 6.64632
I0510 19:03:19.554292 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.64632 (* 1 = 6.64632 loss)
I0510 19:03:19.554309 39557 sgd_solver.cpp:105] Iteration 34300, lr = 0.1
I0510 19:03:46.260345 39557 solver.cpp:218] Iteration 34400 (3.74448 iter/s, 26.706s/100 iters), loss = 6.85976
I0510 19:03:46.260516 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.85976 (* 1 = 6.85976 loss)
I0510 19:03:46.260530 39557 sgd_solver.cpp:105] Iteration 34400, lr = 0.1
I0510 19:04:12.851888 39557 solver.cpp:218] Iteration 34500 (3.76062 iter/s, 26.5913s/100 iters), loss = 6.82327
I0510 19:04:12.851949 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.82327 (* 1 = 6.82327 loss)
I0510 19:04:12.851963 39557 sgd_solver.cpp:105] Iteration 34500, lr = 0.1
I0510 19:04:39.426203 39557 solver.cpp:218] Iteration 34600 (3.76305 iter/s, 26.5742s/100 iters), loss = 6.55932
I0510 19:04:39.426379 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.55932 (* 1 = 6.55932 loss)
I0510 19:04:39.426395 39557 sgd_solver.cpp:105] Iteration 34600, lr = 0.1
I0510 19:05:06.115218 39557 solver.cpp:218] Iteration 34700 (3.74689 iter/s, 26.6888s/100 iters), loss = 7.60091
I0510 19:05:06.115301 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.60091 (* 1 = 7.60091 loss)
I0510 19:05:06.115316 39557 sgd_solver.cpp:105] Iteration 34700, lr = 0.1
I0510 19:05:32.805704 39557 solver.cpp:218] Iteration 34800 (3.74667 iter/s, 26.6903s/100 iters), loss = 7.40333
I0510 19:05:32.805920 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.40333 (* 1 = 7.40333 loss)
I0510 19:05:32.805936 39557 sgd_solver.cpp:105] Iteration 34800, lr = 0.1
I0510 19:05:59.466368 39557 solver.cpp:218] Iteration 34900 (3.75088 iter/s, 26.6604s/100 iters), loss = 7.01983
I0510 19:05:59.466429 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.01983 (* 1 = 7.01983 loss)
I0510 19:05:59.466441 39557 sgd_solver.cpp:105] Iteration 34900, lr = 0.1
I0510 19:06:25.851397 39557 solver.cpp:447] Snapshotting to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_35000.caffemodel
I0510 19:06:25.891026 39557 sgd_solver.cpp:273] Snapshotting solver state to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_35000.solverstate
I0510 19:06:26.174716 39557 solver.cpp:218] Iteration 35000 (3.74416 iter/s, 26.7082s/100 iters), loss = 6.94106
I0510 19:06:26.174777 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.94106 (* 1 = 6.94106 loss)
I0510 19:06:26.174788 39557 sgd_solver.cpp:105] Iteration 35000, lr = 0.1
I0510 19:06:52.843484 39557 solver.cpp:218] Iteration 35100 (3.74972 iter/s, 26.6686s/100 iters), loss = 7.0026
I0510 19:06:52.843544 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.0026 (* 1 = 7.0026 loss)
I0510 19:06:52.843557 39557 sgd_solver.cpp:105] Iteration 35100, lr = 0.1
I0510 19:07:19.387670 39557 solver.cpp:218] Iteration 35200 (3.76732 iter/s, 26.5441s/100 iters), loss = 7.37331
I0510 19:07:19.387847 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.37331 (* 1 = 7.37331 loss)
I0510 19:07:19.387862 39557 sgd_solver.cpp:105] Iteration 35200, lr = 0.1
I0510 19:07:46.011303 39557 solver.cpp:218] Iteration 35300 (3.7561 iter/s, 26.6234s/100 iters), loss = 7.7905
I0510 19:07:46.011361 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.7905 (* 1 = 7.7905 loss)
I0510 19:07:46.011374 39557 sgd_solver.cpp:105] Iteration 35300, lr = 0.1
I0510 19:08:12.651036 39557 solver.cpp:218] Iteration 35400 (3.75381 iter/s, 26.6396s/100 iters), loss = 7.22884
I0510 19:08:12.651286 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.22884 (* 1 = 7.22884 loss)
I0510 19:08:12.651347 39557 sgd_solver.cpp:105] Iteration 35400, lr = 0.1
I0510 19:08:39.301744 39557 solver.cpp:218] Iteration 35500 (3.75229 iter/s, 26.6504s/100 iters), loss = 6.48003
I0510 19:08:39.301795 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.48003 (* 1 = 6.48003 loss)
I0510 19:08:39.301807 39557 sgd_solver.cpp:105] Iteration 35500, lr = 0.1
I0510 19:09:05.961356 39557 solver.cpp:218] Iteration 35600 (3.75101 iter/s, 26.6595s/100 iters), loss = 6.75889
I0510 19:09:05.961597 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.75889 (* 1 = 6.75889 loss)
I0510 19:09:05.961645 39557 sgd_solver.cpp:105] Iteration 35600, lr = 0.1
I0510 19:09:32.593449 39557 solver.cpp:218] Iteration 35700 (3.75491 iter/s, 26.6318s/100 iters), loss = 7.44504
I0510 19:09:32.593498 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.44504 (* 1 = 7.44504 loss)
I0510 19:09:32.593511 39557 sgd_solver.cpp:105] Iteration 35700, lr = 0.1
I0510 19:09:59.224766 39557 solver.cpp:218] Iteration 35800 (3.755 iter/s, 26.6312s/100 iters), loss = 6.81646
I0510 19:09:59.224936 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.81646 (* 1 = 6.81646 loss)
I0510 19:09:59.224952 39557 sgd_solver.cpp:105] Iteration 35800, lr = 0.1
I0510 19:10:25.912134 39557 solver.cpp:218] Iteration 35900 (3.74713 iter/s, 26.6871s/100 iters), loss = 7.34807
I0510 19:10:25.912184 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.34807 (* 1 = 7.34807 loss)
I0510 19:10:25.912196 39557 sgd_solver.cpp:105] Iteration 35900, lr = 0.1
I0510 19:10:52.517005 39557 solver.cpp:218] Iteration 36000 (3.75873 iter/s, 26.6047s/100 iters), loss = 6.71562
I0510 19:10:52.517235 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.71562 (* 1 = 6.71562 loss)
I0510 19:10:52.517254 39557 sgd_solver.cpp:105] Iteration 36000, lr = 0.1
I0510 19:11:19.160156 39557 solver.cpp:218] Iteration 36100 (3.75335 iter/s, 26.6428s/100 iters), loss = 7.03109
I0510 19:11:19.160223 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.03109 (* 1 = 7.03109 loss)
I0510 19:11:19.160235 39557 sgd_solver.cpp:105] Iteration 36100, lr = 0.1
I0510 19:11:45.805661 39557 solver.cpp:218] Iteration 36200 (3.753 iter/s, 26.6453s/100 iters), loss = 6.75685
I0510 19:11:45.805877 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.75685 (* 1 = 6.75685 loss)
I0510 19:11:45.805893 39557 sgd_solver.cpp:105] Iteration 36200, lr = 0.1
I0510 19:12:12.525456 39557 solver.cpp:218] Iteration 36300 (3.74259 iter/s, 26.7195s/100 iters), loss = 7.58906
I0510 19:12:12.525509 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.58906 (* 1 = 7.58906 loss)
I0510 19:12:12.525522 39557 sgd_solver.cpp:105] Iteration 36300, lr = 0.1
I0510 19:12:39.093080 39557 solver.cpp:218] Iteration 36400 (3.764 iter/s, 26.5675s/100 iters), loss = 6.60021
I0510 19:12:39.093277 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.60021 (* 1 = 6.60021 loss)
I0510 19:12:39.093293 39557 sgd_solver.cpp:105] Iteration 36400, lr = 0.1
I0510 19:13:05.682577 39557 solver.cpp:218] Iteration 36500 (3.76092 iter/s, 26.5892s/100 iters), loss = 6.73911
I0510 19:13:05.682628 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.73911 (* 1 = 6.73911 loss)
I0510 19:13:05.682641 39557 sgd_solver.cpp:105] Iteration 36500, lr = 0.1
I0510 19:13:32.326676 39557 solver.cpp:218] Iteration 36600 (3.7532 iter/s, 26.644s/100 iters), loss = 6.86395
I0510 19:13:32.326846 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.86395 (* 1 = 6.86395 loss)
I0510 19:13:32.326862 39557 sgd_solver.cpp:105] Iteration 36600, lr = 0.1
I0510 19:13:58.909747 39557 solver.cpp:218] Iteration 36700 (3.76183 iter/s, 26.5828s/100 iters), loss = 6.84964
I0510 19:13:58.909802 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.84964 (* 1 = 6.84964 loss)
I0510 19:13:58.909816 39557 sgd_solver.cpp:105] Iteration 36700, lr = 0.1
I0510 19:14:25.647255 39557 solver.cpp:218] Iteration 36800 (3.74009 iter/s, 26.7374s/100 iters), loss = 7.23074
I0510 19:14:25.647445 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.23074 (* 1 = 7.23074 loss)
I0510 19:14:25.647462 39557 sgd_solver.cpp:105] Iteration 36800, lr = 0.1
I0510 19:14:52.258671 39557 solver.cpp:218] Iteration 36900 (3.75783 iter/s, 26.6111s/100 iters), loss = 6.84445
I0510 19:14:52.258725 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.84445 (* 1 = 6.84445 loss)
I0510 19:14:52.258738 39557 sgd_solver.cpp:105] Iteration 36900, lr = 0.1
I0510 19:15:18.997197 39557 solver.cpp:218] Iteration 37000 (3.73994 iter/s, 26.7384s/100 iters), loss = 7.48159
I0510 19:15:18.997356 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.48159 (* 1 = 7.48159 loss)
I0510 19:15:18.997371 39557 sgd_solver.cpp:105] Iteration 37000, lr = 0.1
I0510 19:15:45.584832 39557 solver.cpp:218] Iteration 37100 (3.76118 iter/s, 26.5874s/100 iters), loss = 7.07585
I0510 19:15:45.584897 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.07585 (* 1 = 7.07585 loss)
I0510 19:15:45.584911 39557 sgd_solver.cpp:105] Iteration 37100, lr = 0.1
I0510 19:16:12.165432 39557 solver.cpp:218] Iteration 37200 (3.76217 iter/s, 26.5804s/100 iters), loss = 6.91899
I0510 19:16:12.165594 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.91899 (* 1 = 6.91899 loss)
I0510 19:16:12.165609 39557 sgd_solver.cpp:105] Iteration 37200, lr = 0.1
I0510 19:16:38.699915 39557 solver.cpp:218] Iteration 37300 (3.76872 iter/s, 26.5342s/100 iters), loss = 6.65715
I0510 19:16:38.699977 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.65715 (* 1 = 6.65715 loss)
I0510 19:16:38.699992 39557 sgd_solver.cpp:105] Iteration 37300, lr = 0.1
I0510 19:17:05.287065 39557 solver.cpp:218] Iteration 37400 (3.76124 iter/s, 26.587s/100 iters), loss = 8.21382
I0510 19:17:05.287262 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.21382 (* 1 = 8.21382 loss)
I0510 19:17:05.287276 39557 sgd_solver.cpp:105] Iteration 37400, lr = 0.1
I0510 19:17:31.844995 39557 solver.cpp:218] Iteration 37500 (3.7654 iter/s, 26.5576s/100 iters), loss = 7.33218
I0510 19:17:31.845049 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.33218 (* 1 = 7.33218 loss)
I0510 19:17:31.845062 39557 sgd_solver.cpp:105] Iteration 37500, lr = 0.1
I0510 19:17:58.412520 39557 solver.cpp:218] Iteration 37600 (3.76402 iter/s, 26.5674s/100 iters), loss = 7.36254
I0510 19:17:58.412694 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.36254 (* 1 = 7.36254 loss)
I0510 19:17:58.412710 39557 sgd_solver.cpp:105] Iteration 37600, lr = 0.1
I0510 19:18:25.039710 39557 solver.cpp:218] Iteration 37700 (3.7556 iter/s, 26.6269s/100 iters), loss = 7.2747
I0510 19:18:25.039773 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.2747 (* 1 = 7.2747 loss)
I0510 19:18:25.039786 39557 sgd_solver.cpp:105] Iteration 37700, lr = 0.1
I0510 19:18:51.684532 39557 solver.cpp:218] Iteration 37800 (3.7531 iter/s, 26.6446s/100 iters), loss = 6.34154
I0510 19:18:51.684725 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.34154 (* 1 = 6.34154 loss)
I0510 19:18:51.684739 39557 sgd_solver.cpp:105] Iteration 37800, lr = 0.1
I0510 19:19:18.302529 39557 solver.cpp:218] Iteration 37900 (3.7569 iter/s, 26.6177s/100 iters), loss = 7.56948
I0510 19:19:18.302590 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.56948 (* 1 = 7.56948 loss)
I0510 19:19:18.302605 39557 sgd_solver.cpp:105] Iteration 37900, lr = 0.1
I0510 19:19:44.958284 39557 solver.cpp:218] Iteration 38000 (3.75156 iter/s, 26.6556s/100 iters), loss = 6.95856
I0510 19:19:44.958456 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.95856 (* 1 = 6.95856 loss)
I0510 19:19:44.958472 39557 sgd_solver.cpp:105] Iteration 38000, lr = 0.1
I0510 19:20:11.555527 39557 solver.cpp:218] Iteration 38100 (3.75983 iter/s, 26.597s/100 iters), loss = 6.76376
I0510 19:20:11.555591 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.76376 (* 1 = 6.76376 loss)
I0510 19:20:11.555604 39557 sgd_solver.cpp:105] Iteration 38100, lr = 0.1
I0510 19:20:38.183600 39557 solver.cpp:218] Iteration 38200 (3.75546 iter/s, 26.6279s/100 iters), loss = 7.04928
I0510 19:20:38.183756 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.04928 (* 1 = 7.04928 loss)
I0510 19:20:38.183773 39557 sgd_solver.cpp:105] Iteration 38200, lr = 0.1
I0510 19:21:04.718698 39557 solver.cpp:218] Iteration 38300 (3.76863 iter/s, 26.5348s/100 iters), loss = 6.20286
I0510 19:21:04.718760 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.20286 (* 1 = 6.20286 loss)
I0510 19:21:04.718773 39557 sgd_solver.cpp:105] Iteration 38300, lr = 0.1
I0510 19:21:31.316303 39557 solver.cpp:218] Iteration 38400 (3.75976 iter/s, 26.5974s/100 iters), loss = 6.46999
I0510 19:21:31.316485 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.46999 (* 1 = 6.46999 loss)
I0510 19:21:31.316500 39557 sgd_solver.cpp:105] Iteration 38400, lr = 0.1
I0510 19:21:57.943452 39557 solver.cpp:218] Iteration 38500 (3.75561 iter/s, 26.6269s/100 iters), loss = 6.76715
I0510 19:21:57.943516 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.76715 (* 1 = 6.76715 loss)
I0510 19:21:57.943528 39557 sgd_solver.cpp:105] Iteration 38500, lr = 0.1
I0510 19:22:24.606457 39557 solver.cpp:218] Iteration 38600 (3.75054 iter/s, 26.6628s/100 iters), loss = 7.12134
I0510 19:22:24.606655 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.12134 (* 1 = 7.12134 loss)
I0510 19:22:24.606673 39557 sgd_solver.cpp:105] Iteration 38600, lr = 0.1
I0510 19:22:51.175912 39557 solver.cpp:218] Iteration 38700 (3.76376 iter/s, 26.5692s/100 iters), loss = 6.68701
I0510 19:22:51.175974 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.68701 (* 1 = 6.68701 loss)
I0510 19:22:51.175990 39557 sgd_solver.cpp:105] Iteration 38700, lr = 0.1
I0510 19:23:17.787194 39557 solver.cpp:218] Iteration 38800 (3.75783 iter/s, 26.6111s/100 iters), loss = 6.6477
I0510 19:23:17.787469 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.6477 (* 1 = 6.6477 loss)
I0510 19:23:17.787520 39557 sgd_solver.cpp:105] Iteration 38800, lr = 0.1
I0510 19:23:44.452683 39557 solver.cpp:218] Iteration 38900 (3.75022 iter/s, 26.6651s/100 iters), loss = 6.88692
I0510 19:23:44.452750 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.88692 (* 1 = 6.88692 loss)
I0510 19:23:44.452764 39557 sgd_solver.cpp:105] Iteration 38900, lr = 0.1
I0510 19:24:11.089076 39557 solver.cpp:218] Iteration 39000 (3.75429 iter/s, 26.6362s/100 iters), loss = 6.53296
I0510 19:24:11.089277 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.53296 (* 1 = 6.53296 loss)
I0510 19:24:11.089292 39557 sgd_solver.cpp:105] Iteration 39000, lr = 0.1
I0510 19:24:37.779268 39557 solver.cpp:218] Iteration 39100 (3.74674 iter/s, 26.6899s/100 iters), loss = 6.4098
I0510 19:24:37.779336 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.4098 (* 1 = 6.4098 loss)
I0510 19:24:37.779350 39557 sgd_solver.cpp:105] Iteration 39100, lr = 0.1
I0510 19:25:04.427057 39557 solver.cpp:218] Iteration 39200 (3.75268 iter/s, 26.6476s/100 iters), loss = 7.31553
I0510 19:25:04.427240 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.31553 (* 1 = 7.31553 loss)
I0510 19:25:04.427255 39557 sgd_solver.cpp:105] Iteration 39200, lr = 0.1
I0510 19:25:31.026844 39557 solver.cpp:218] Iteration 39300 (3.75947 iter/s, 26.5995s/100 iters), loss = 6.63297
I0510 19:25:31.026906 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.63297 (* 1 = 6.63297 loss)
I0510 19:25:31.026918 39557 sgd_solver.cpp:105] Iteration 39300, lr = 0.1
I0510 19:25:57.635427 39557 solver.cpp:218] Iteration 39400 (3.75821 iter/s, 26.6084s/100 iters), loss = 7.12225
I0510 19:25:57.635602 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.12225 (* 1 = 7.12225 loss)
I0510 19:25:57.635617 39557 sgd_solver.cpp:105] Iteration 39400, lr = 0.1
I0510 19:26:24.350970 39557 solver.cpp:218] Iteration 39500 (3.74318 iter/s, 26.7153s/100 iters), loss = 6.07152
I0510 19:26:24.351040 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.07152 (* 1 = 6.07152 loss)
I0510 19:26:24.351054 39557 sgd_solver.cpp:105] Iteration 39500, lr = 0.1
I0510 19:26:50.989675 39557 solver.cpp:218] Iteration 39600 (3.75396 iter/s, 26.6385s/100 iters), loss = 7.04376
I0510 19:26:50.989840 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.04376 (* 1 = 7.04376 loss)
I0510 19:26:50.989854 39557 sgd_solver.cpp:105] Iteration 39600, lr = 0.1
I0510 19:27:17.623466 39557 solver.cpp:218] Iteration 39700 (3.75467 iter/s, 26.6335s/100 iters), loss = 7.24451
I0510 19:27:17.623527 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.24451 (* 1 = 7.24451 loss)
I0510 19:27:17.623539 39557 sgd_solver.cpp:105] Iteration 39700, lr = 0.1
I0510 19:27:44.277218 39557 solver.cpp:218] Iteration 39800 (3.75184 iter/s, 26.6536s/100 iters), loss = 6.84965
I0510 19:27:44.277376 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.84965 (* 1 = 6.84965 loss)
I0510 19:27:44.277391 39557 sgd_solver.cpp:105] Iteration 39800, lr = 0.1
I0510 19:28:10.837713 39557 solver.cpp:218] Iteration 39900 (3.76503 iter/s, 26.5602s/100 iters), loss = 7.2579
I0510 19:28:10.837779 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.2579 (* 1 = 7.2579 loss)
I0510 19:28:10.837793 39557 sgd_solver.cpp:105] Iteration 39900, lr = 0.1
I0510 19:28:37.173225 39557 solver.cpp:447] Snapshotting to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_40000.caffemodel
I0510 19:28:37.211288 39557 sgd_solver.cpp:273] Snapshotting solver state to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_40000.solverstate
I0510 19:28:37.495190 39557 solver.cpp:218] Iteration 40000 (3.75132 iter/s, 26.6573s/100 iters), loss = 7.1158
I0510 19:28:37.495247 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.1158 (* 1 = 7.1158 loss)
I0510 19:28:37.495257 39557 sgd_solver.cpp:105] Iteration 40000, lr = 0.1
I0510 19:29:04.146322 39557 solver.cpp:218] Iteration 40100 (3.75221 iter/s, 26.651s/100 iters), loss = 7.02234
I0510 19:29:04.146383 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.02234 (* 1 = 7.02234 loss)
I0510 19:29:04.146396 39557 sgd_solver.cpp:105] Iteration 40100, lr = 0.1
I0510 19:29:30.793154 39557 solver.cpp:218] Iteration 40200 (3.75282 iter/s, 26.6467s/100 iters), loss = 7.38344
I0510 19:29:30.793393 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.38344 (* 1 = 7.38344 loss)
I0510 19:29:30.793411 39557 sgd_solver.cpp:105] Iteration 40200, lr = 0.1
I0510 19:29:57.378284 39557 solver.cpp:218] Iteration 40300 (3.76155 iter/s, 26.5848s/100 iters), loss = 6.42492
I0510 19:29:57.378351 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.42492 (* 1 = 6.42492 loss)
I0510 19:29:57.378365 39557 sgd_solver.cpp:105] Iteration 40300, lr = 0.1
I0510 19:30:24.013926 39557 solver.cpp:218] Iteration 40400 (3.75439 iter/s, 26.6355s/100 iters), loss = 7.40126
I0510 19:30:24.014122 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.40126 (* 1 = 7.40126 loss)
I0510 19:30:24.014142 39557 sgd_solver.cpp:105] Iteration 40400, lr = 0.1
I0510 19:30:50.621577 39557 solver.cpp:218] Iteration 40500 (3.75836 iter/s, 26.6073s/100 iters), loss = 7.88158
I0510 19:30:50.621646 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.88158 (* 1 = 7.88158 loss)
I0510 19:30:50.621666 39557 sgd_solver.cpp:105] Iteration 40500, lr = 0.1
I0510 19:31:17.433723 39557 solver.cpp:218] Iteration 40600 (3.72968 iter/s, 26.812s/100 iters), loss = 7.02318
I0510 19:31:17.433933 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.02318 (* 1 = 7.02318 loss)
I0510 19:31:17.433950 39557 sgd_solver.cpp:105] Iteration 40600, lr = 0.1
I0510 19:31:44.159564 39557 solver.cpp:218] Iteration 40700 (3.74174 iter/s, 26.7255s/100 iters), loss = 6.52387
I0510 19:31:44.159616 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.52387 (* 1 = 6.52387 loss)
I0510 19:31:44.159628 39557 sgd_solver.cpp:105] Iteration 40700, lr = 0.1
I0510 19:32:10.852279 39557 solver.cpp:218] Iteration 40800 (3.74636 iter/s, 26.6925s/100 iters), loss = 6.67453
I0510 19:32:10.852483 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.67453 (* 1 = 6.67453 loss)
I0510 19:32:10.852500 39557 sgd_solver.cpp:105] Iteration 40800, lr = 0.1
I0510 19:32:37.550415 39557 solver.cpp:218] Iteration 40900 (3.74562 iter/s, 26.6978s/100 iters), loss = 6.59567
I0510 19:32:37.550465 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.59567 (* 1 = 6.59567 loss)
I0510 19:32:37.550480 39557 sgd_solver.cpp:105] Iteration 40900, lr = 0.1
I0510 19:33:04.221695 39557 solver.cpp:218] Iteration 41000 (3.74938 iter/s, 26.6711s/100 iters), loss = 7.29287
I0510 19:33:04.221868 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.29287 (* 1 = 7.29287 loss)
I0510 19:33:04.221884 39557 sgd_solver.cpp:105] Iteration 41000, lr = 0.1
I0510 19:33:30.878659 39557 solver.cpp:218] Iteration 41100 (3.75141 iter/s, 26.6567s/100 iters), loss = 6.71968
I0510 19:33:30.878726 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.71968 (* 1 = 6.71968 loss)
I0510 19:33:30.878739 39557 sgd_solver.cpp:105] Iteration 41100, lr = 0.1
I0510 19:33:57.518913 39557 solver.cpp:218] Iteration 41200 (3.75375 iter/s, 26.6401s/100 iters), loss = 6.65349
I0510 19:33:57.519098 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.65349 (* 1 = 6.65349 loss)
I0510 19:33:57.519114 39557 sgd_solver.cpp:105] Iteration 41200, lr = 0.1
I0510 19:34:24.057961 39557 solver.cpp:218] Iteration 41300 (3.76808 iter/s, 26.5387s/100 iters), loss = 6.75642
I0510 19:34:24.058022 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.75642 (* 1 = 6.75642 loss)
I0510 19:34:24.058033 39557 sgd_solver.cpp:105] Iteration 41300, lr = 0.1
I0510 19:34:50.627879 39557 solver.cpp:218] Iteration 41400 (3.76368 iter/s, 26.5697s/100 iters), loss = 7.14693
I0510 19:34:50.628096 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.14693 (* 1 = 7.14693 loss)
I0510 19:34:50.628113 39557 sgd_solver.cpp:105] Iteration 41400, lr = 0.1
I0510 19:35:17.394276 39557 solver.cpp:218] Iteration 41500 (3.73607 iter/s, 26.7661s/100 iters), loss = 6.32132
I0510 19:35:17.394330 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.32132 (* 1 = 6.32132 loss)
I0510 19:35:17.394342 39557 sgd_solver.cpp:105] Iteration 41500, lr = 0.1
I0510 19:35:43.944272 39557 solver.cpp:218] Iteration 41600 (3.7665 iter/s, 26.5498s/100 iters), loss = 7.0719
I0510 19:35:43.944458 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.0719 (* 1 = 7.0719 loss)
I0510 19:35:43.944473 39557 sgd_solver.cpp:105] Iteration 41600, lr = 0.1
I0510 19:36:10.584193 39557 solver.cpp:218] Iteration 41700 (3.7538 iter/s, 26.6397s/100 iters), loss = 7.01825
I0510 19:36:10.584244 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.01825 (* 1 = 7.01825 loss)
I0510 19:36:10.584255 39557 sgd_solver.cpp:105] Iteration 41700, lr = 0.1
I0510 19:36:37.208231 39557 solver.cpp:218] Iteration 41800 (3.75602 iter/s, 26.6239s/100 iters), loss = 6.20058
I0510 19:36:37.208406 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.20058 (* 1 = 6.20058 loss)
I0510 19:36:37.208421 39557 sgd_solver.cpp:105] Iteration 41800, lr = 0.1
I0510 19:37:03.833348 39557 solver.cpp:218] Iteration 41900 (3.75589 iter/s, 26.6249s/100 iters), loss = 7.41832
I0510 19:37:03.833410 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.41832 (* 1 = 7.41832 loss)
I0510 19:37:03.833422 39557 sgd_solver.cpp:105] Iteration 41900, lr = 0.1
I0510 19:37:30.455273 39557 solver.cpp:218] Iteration 42000 (3.75632 iter/s, 26.6218s/100 iters), loss = 8.00558
I0510 19:37:30.455440 39557 solver.cpp:237]     Train net output #0: softmax_loss = 8.00558 (* 1 = 8.00558 loss)
I0510 19:37:30.455453 39557 sgd_solver.cpp:105] Iteration 42000, lr = 0.1
I0510 19:37:57.053939 39557 solver.cpp:218] Iteration 42100 (3.75962 iter/s, 26.5984s/100 iters), loss = 7.20001
I0510 19:37:57.054003 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.20001 (* 1 = 7.20001 loss)
I0510 19:37:57.054015 39557 sgd_solver.cpp:105] Iteration 42100, lr = 0.1
I0510 19:38:23.672264 39557 solver.cpp:218] Iteration 42200 (3.75683 iter/s, 26.6182s/100 iters), loss = 6.32397
I0510 19:38:23.672418 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.32397 (* 1 = 6.32397 loss)
I0510 19:38:23.672432 39557 sgd_solver.cpp:105] Iteration 42200, lr = 0.1
I0510 19:38:50.309727 39557 solver.cpp:218] Iteration 42300 (3.75414 iter/s, 26.6372s/100 iters), loss = 6.09192
I0510 19:38:50.309790 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.09192 (* 1 = 6.09192 loss)
I0510 19:38:50.309803 39557 sgd_solver.cpp:105] Iteration 42300, lr = 0.1
I0510 19:39:16.953815 39557 solver.cpp:218] Iteration 42400 (3.7532 iter/s, 26.6439s/100 iters), loss = 7.54666
I0510 19:39:16.953986 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.54666 (* 1 = 7.54666 loss)
I0510 19:39:16.954001 39557 sgd_solver.cpp:105] Iteration 42400, lr = 0.1
I0510 19:39:43.581080 39557 solver.cpp:218] Iteration 42500 (3.75559 iter/s, 26.627s/100 iters), loss = 6.96378
I0510 19:39:43.581140 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.96378 (* 1 = 6.96378 loss)
I0510 19:39:43.581153 39557 sgd_solver.cpp:105] Iteration 42500, lr = 0.1
I0510 19:40:10.231806 39557 solver.cpp:218] Iteration 42600 (3.75226 iter/s, 26.6506s/100 iters), loss = 7.09122
I0510 19:40:10.231973 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.09122 (* 1 = 7.09122 loss)
I0510 19:40:10.231987 39557 sgd_solver.cpp:105] Iteration 42600, lr = 0.1
I0510 19:40:36.820413 39557 solver.cpp:218] Iteration 42700 (3.76105 iter/s, 26.5884s/100 iters), loss = 6.94721
I0510 19:40:36.820472 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.94721 (* 1 = 6.94721 loss)
I0510 19:40:36.820482 39557 sgd_solver.cpp:105] Iteration 42700, lr = 0.1
I0510 19:41:03.421965 39557 solver.cpp:218] Iteration 42800 (3.7592 iter/s, 26.6014s/100 iters), loss = 5.98114
I0510 19:41:03.422179 39557 solver.cpp:237]     Train net output #0: softmax_loss = 5.98114 (* 1 = 5.98114 loss)
I0510 19:41:03.422194 39557 sgd_solver.cpp:105] Iteration 42800, lr = 0.1
I0510 19:41:30.062234 39557 solver.cpp:218] Iteration 42900 (3.75376 iter/s, 26.64s/100 iters), loss = 7.09993
I0510 19:41:30.062299 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.09993 (* 1 = 7.09993 loss)
I0510 19:41:30.062310 39557 sgd_solver.cpp:105] Iteration 42900, lr = 0.1
I0510 19:41:56.678376 39557 solver.cpp:218] Iteration 43000 (3.75714 iter/s, 26.616s/100 iters), loss = 6.61027
I0510 19:41:56.678519 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.61027 (* 1 = 6.61027 loss)
I0510 19:41:56.678531 39557 sgd_solver.cpp:105] Iteration 43000, lr = 0.1
I0510 19:42:23.316588 39557 solver.cpp:218] Iteration 43100 (3.75404 iter/s, 26.638s/100 iters), loss = 7.12183
I0510 19:42:23.316649 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.12183 (* 1 = 7.12183 loss)
I0510 19:42:23.316663 39557 sgd_solver.cpp:105] Iteration 43100, lr = 0.1
I0510 19:42:49.913087 39557 solver.cpp:218] Iteration 43200 (3.75991 iter/s, 26.5963s/100 iters), loss = 7.32158
I0510 19:42:49.913234 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.32158 (* 1 = 7.32158 loss)
I0510 19:42:49.913250 39557 sgd_solver.cpp:105] Iteration 43200, lr = 0.1
I0510 19:43:16.490933 39557 solver.cpp:218] Iteration 43300 (3.76257 iter/s, 26.5776s/100 iters), loss = 7.09365
I0510 19:43:16.490993 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.09365 (* 1 = 7.09365 loss)
I0510 19:43:16.491003 39557 sgd_solver.cpp:105] Iteration 43300, lr = 0.1
I0510 19:43:43.111686 39557 solver.cpp:218] Iteration 43400 (3.75649 iter/s, 26.6206s/100 iters), loss = 7.11564
I0510 19:43:43.111840 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.11564 (* 1 = 7.11564 loss)
I0510 19:43:43.111855 39557 sgd_solver.cpp:105] Iteration 43400, lr = 0.1
I0510 19:44:09.795869 39557 solver.cpp:218] Iteration 43500 (3.74757 iter/s, 26.6839s/100 iters), loss = 7.02294
I0510 19:44:09.795925 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.02294 (* 1 = 7.02294 loss)
I0510 19:44:09.795936 39557 sgd_solver.cpp:105] Iteration 43500, lr = 0.1
I0510 19:44:36.483669 39557 solver.cpp:218] Iteration 43600 (3.74705 iter/s, 26.6876s/100 iters), loss = 7.05594
I0510 19:44:36.483840 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.05594 (* 1 = 7.05594 loss)
I0510 19:44:36.483856 39557 sgd_solver.cpp:105] Iteration 43600, lr = 0.1
I0510 19:45:03.158073 39557 solver.cpp:218] Iteration 43700 (3.74895 iter/s, 26.6741s/100 iters), loss = 6.55196
I0510 19:45:03.158128 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.55196 (* 1 = 6.55196 loss)
I0510 19:45:03.158139 39557 sgd_solver.cpp:105] Iteration 43700, lr = 0.1
I0510 19:45:29.804559 39557 solver.cpp:218] Iteration 43800 (3.75286 iter/s, 26.6463s/100 iters), loss = 7.10119
I0510 19:45:29.804756 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.10119 (* 1 = 7.10119 loss)
I0510 19:45:29.804774 39557 sgd_solver.cpp:105] Iteration 43800, lr = 0.1
I0510 19:45:56.483693 39557 solver.cpp:218] Iteration 43900 (3.74829 iter/s, 26.6788s/100 iters), loss = 6.48572
I0510 19:45:56.483747 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.48572 (* 1 = 6.48572 loss)
I0510 19:45:56.483760 39557 sgd_solver.cpp:105] Iteration 43900, lr = 0.1
I0510 19:46:23.137182 39557 solver.cpp:218] Iteration 44000 (3.75188 iter/s, 26.6533s/100 iters), loss = 7.67063
I0510 19:46:23.137377 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.67063 (* 1 = 7.67063 loss)
I0510 19:46:23.137392 39557 sgd_solver.cpp:105] Iteration 44000, lr = 0.1
I0510 19:46:49.740798 39557 solver.cpp:218] Iteration 44100 (3.75893 iter/s, 26.6033s/100 iters), loss = 6.59676
I0510 19:46:49.740859 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.59676 (* 1 = 6.59676 loss)
I0510 19:46:49.740871 39557 sgd_solver.cpp:105] Iteration 44100, lr = 0.1
I0510 19:47:16.389322 39557 solver.cpp:218] Iteration 44200 (3.75258 iter/s, 26.6484s/100 iters), loss = 7.05422
I0510 19:47:16.389513 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.05422 (* 1 = 7.05422 loss)
I0510 19:47:16.389526 39557 sgd_solver.cpp:105] Iteration 44200, lr = 0.1
I0510 19:47:43.046171 39557 solver.cpp:218] Iteration 44300 (3.75142 iter/s, 26.6566s/100 iters), loss = 7.01252
I0510 19:47:43.046227 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.01252 (* 1 = 7.01252 loss)
I0510 19:47:43.046241 39557 sgd_solver.cpp:105] Iteration 44300, lr = 0.1
I0510 19:48:09.747715 39557 solver.cpp:218] Iteration 44400 (3.74512 iter/s, 26.7014s/100 iters), loss = 6.99661
I0510 19:48:09.747921 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.99661 (* 1 = 6.99661 loss)
I0510 19:48:09.747934 39557 sgd_solver.cpp:105] Iteration 44400, lr = 0.1
I0510 19:48:36.360998 39557 solver.cpp:218] Iteration 44500 (3.75757 iter/s, 26.613s/100 iters), loss = 7.76495
I0510 19:48:36.361059 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.76495 (* 1 = 7.76495 loss)
I0510 19:48:36.361073 39557 sgd_solver.cpp:105] Iteration 44500, lr = 0.1
I0510 19:49:02.971881 39557 solver.cpp:218] Iteration 44600 (3.75788 iter/s, 26.6107s/100 iters), loss = 7.36034
I0510 19:49:02.972024 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.36034 (* 1 = 7.36034 loss)
I0510 19:49:02.972038 39557 sgd_solver.cpp:105] Iteration 44600, lr = 0.1
I0510 19:49:29.559226 39557 solver.cpp:218] Iteration 44700 (3.76122 iter/s, 26.5871s/100 iters), loss = 6.898
I0510 19:49:29.559281 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.898 (* 1 = 6.898 loss)
I0510 19:49:29.559295 39557 sgd_solver.cpp:105] Iteration 44700, lr = 0.1
I0510 19:49:56.190788 39557 solver.cpp:218] Iteration 44800 (3.75497 iter/s, 26.6314s/100 iters), loss = 6.90435
I0510 19:49:56.190943 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.90435 (* 1 = 6.90435 loss)
I0510 19:49:56.190955 39557 sgd_solver.cpp:105] Iteration 44800, lr = 0.1
I0510 19:50:22.879250 39557 solver.cpp:218] Iteration 44900 (3.74698 iter/s, 26.6882s/100 iters), loss = 6.22648
I0510 19:50:22.879315 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.22648 (* 1 = 6.22648 loss)
I0510 19:50:22.879336 39557 sgd_solver.cpp:105] Iteration 44900, lr = 0.1
I0510 19:50:49.242518 39557 solver.cpp:447] Snapshotting to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_45000.caffemodel
I0510 19:50:49.279989 39557 sgd_solver.cpp:273] Snapshotting solver state to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_45000.solverstate
I0510 19:50:49.562575 39557 solver.cpp:218] Iteration 45000 (3.74768 iter/s, 26.6832s/100 iters), loss = 6.91614
I0510 19:50:49.562633 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.91614 (* 1 = 6.91614 loss)
I0510 19:50:49.562644 39557 sgd_solver.cpp:105] Iteration 45000, lr = 0.1
I0510 19:51:16.269731 39557 solver.cpp:218] Iteration 45100 (3.74434 iter/s, 26.707s/100 iters), loss = 6.7892
I0510 19:51:16.269785 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.7892 (* 1 = 6.7892 loss)
I0510 19:51:16.269798 39557 sgd_solver.cpp:105] Iteration 45100, lr = 0.1
I0510 19:51:42.981561 39557 solver.cpp:218] Iteration 45200 (3.74368 iter/s, 26.7117s/100 iters), loss = 7.47266
I0510 19:51:42.981734 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.47266 (* 1 = 7.47266 loss)
I0510 19:51:42.981747 39557 sgd_solver.cpp:105] Iteration 45200, lr = 0.1
I0510 19:52:09.645424 39557 solver.cpp:218] Iteration 45300 (3.75043 iter/s, 26.6636s/100 iters), loss = 6.92548
I0510 19:52:09.645479 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.92548 (* 1 = 6.92548 loss)
I0510 19:52:09.645491 39557 sgd_solver.cpp:105] Iteration 45300, lr = 0.1
I0510 19:52:36.360186 39557 solver.cpp:218] Iteration 45400 (3.74327 iter/s, 26.7146s/100 iters), loss = 7.16687
I0510 19:52:36.360389 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.16687 (* 1 = 7.16687 loss)
I0510 19:52:36.360406 39557 sgd_solver.cpp:105] Iteration 45400, lr = 0.1
I0510 19:53:03.051731 39557 solver.cpp:218] Iteration 45500 (3.74655 iter/s, 26.6912s/100 iters), loss = 7.26759
I0510 19:53:03.051800 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.26759 (* 1 = 7.26759 loss)
I0510 19:53:03.051815 39557 sgd_solver.cpp:105] Iteration 45500, lr = 0.1
I0510 19:53:29.788550 39557 solver.cpp:218] Iteration 45600 (3.74018 iter/s, 26.7366s/100 iters), loss = 7.6413
I0510 19:53:29.788748 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.6413 (* 1 = 7.6413 loss)
I0510 19:53:29.788765 39557 sgd_solver.cpp:105] Iteration 45600, lr = 0.1
I0510 19:53:56.563954 39557 solver.cpp:218] Iteration 45700 (3.73481 iter/s, 26.7751s/100 iters), loss = 7.6768
I0510 19:53:56.564018 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.6768 (* 1 = 7.6768 loss)
I0510 19:53:56.564041 39557 sgd_solver.cpp:105] Iteration 45700, lr = 0.1
I0510 19:54:23.280015 39557 solver.cpp:218] Iteration 45800 (3.74309 iter/s, 26.7159s/100 iters), loss = 6.74753
I0510 19:54:23.280216 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.74753 (* 1 = 6.74753 loss)
I0510 19:54:23.280233 39557 sgd_solver.cpp:105] Iteration 45800, lr = 0.1
I0510 19:54:50.064077 39557 solver.cpp:218] Iteration 45900 (3.73361 iter/s, 26.7838s/100 iters), loss = 7.79875
I0510 19:54:50.064138 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.79875 (* 1 = 7.79875 loss)
I0510 19:54:50.064153 39557 sgd_solver.cpp:105] Iteration 45900, lr = 0.1
I0510 19:55:16.777484 39557 solver.cpp:218] Iteration 46000 (3.74346 iter/s, 26.7132s/100 iters), loss = 6.88394
I0510 19:55:16.777719 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.88394 (* 1 = 6.88394 loss)
I0510 19:55:16.777735 39557 sgd_solver.cpp:105] Iteration 46000, lr = 0.1
I0510 19:55:43.433575 39557 solver.cpp:218] Iteration 46100 (3.75154 iter/s, 26.6558s/100 iters), loss = 7.63392
I0510 19:55:43.433639 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.63392 (* 1 = 7.63392 loss)
I0510 19:55:43.433650 39557 sgd_solver.cpp:105] Iteration 46100, lr = 0.1
I0510 19:56:10.125154 39557 solver.cpp:218] Iteration 46200 (3.74652 iter/s, 26.6914s/100 iters), loss = 6.77789
I0510 19:56:10.125358 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.77789 (* 1 = 6.77789 loss)
I0510 19:56:10.125375 39557 sgd_solver.cpp:105] Iteration 46200, lr = 0.1
I0510 19:56:36.749085 39557 solver.cpp:218] Iteration 46300 (3.75606 iter/s, 26.6236s/100 iters), loss = 6.71586
I0510 19:56:36.749148 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.71586 (* 1 = 6.71586 loss)
I0510 19:56:36.749161 39557 sgd_solver.cpp:105] Iteration 46300, lr = 0.1
I0510 19:57:03.422461 39557 solver.cpp:218] Iteration 46400 (3.74908 iter/s, 26.6732s/100 iters), loss = 6.93071
I0510 19:57:03.422611 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.93071 (* 1 = 6.93071 loss)
I0510 19:57:03.422624 39557 sgd_solver.cpp:105] Iteration 46400, lr = 0.1
I0510 19:57:30.075428 39557 solver.cpp:218] Iteration 46500 (3.75197 iter/s, 26.6527s/100 iters), loss = 6.92897
I0510 19:57:30.075490 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.92897 (* 1 = 6.92897 loss)
I0510 19:57:30.075505 39557 sgd_solver.cpp:105] Iteration 46500, lr = 0.1
I0510 19:57:56.732023 39557 solver.cpp:218] Iteration 46600 (3.75144 iter/s, 26.6564s/100 iters), loss = 6.96818
I0510 19:57:56.732172 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.96818 (* 1 = 6.96818 loss)
I0510 19:57:56.732187 39557 sgd_solver.cpp:105] Iteration 46600, lr = 0.1
I0510 19:58:23.381682 39557 solver.cpp:218] Iteration 46700 (3.75243 iter/s, 26.6494s/100 iters), loss = 6.66386
I0510 19:58:23.381737 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.66386 (* 1 = 6.66386 loss)
I0510 19:58:23.381750 39557 sgd_solver.cpp:105] Iteration 46700, lr = 0.1
I0510 19:58:50.037422 39557 solver.cpp:218] Iteration 46800 (3.75156 iter/s, 26.6556s/100 iters), loss = 6.81375
I0510 19:58:50.037600 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.81375 (* 1 = 6.81375 loss)
I0510 19:58:50.037616 39557 sgd_solver.cpp:105] Iteration 46800, lr = 0.1
I0510 19:59:16.670405 39557 solver.cpp:218] Iteration 46900 (3.75478 iter/s, 26.6327s/100 iters), loss = 7.38464
I0510 19:59:16.670467 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.38464 (* 1 = 7.38464 loss)
I0510 19:59:16.670483 39557 sgd_solver.cpp:105] Iteration 46900, lr = 0.1
I0510 19:59:43.333582 39557 solver.cpp:218] Iteration 47000 (3.75051 iter/s, 26.663s/100 iters), loss = 7.40565
I0510 19:59:43.333742 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.40565 (* 1 = 7.40565 loss)
I0510 19:59:43.333760 39557 sgd_solver.cpp:105] Iteration 47000, lr = 0.1
I0510 20:00:09.942353 39557 solver.cpp:218] Iteration 47100 (3.7582 iter/s, 26.6085s/100 iters), loss = 7.09314
I0510 20:00:09.942407 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.09314 (* 1 = 7.09314 loss)
I0510 20:00:09.942420 39557 sgd_solver.cpp:105] Iteration 47100, lr = 0.1
I0510 20:00:36.811062 39557 solver.cpp:218] Iteration 47200 (3.72183 iter/s, 26.8685s/100 iters), loss = 6.99015
I0510 20:00:36.811239 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.99015 (* 1 = 6.99015 loss)
I0510 20:00:36.811254 39557 sgd_solver.cpp:105] Iteration 47200, lr = 0.1
I0510 20:01:03.515215 39557 solver.cpp:218] Iteration 47300 (3.74478 iter/s, 26.7039s/100 iters), loss = 6.4556
I0510 20:01:03.515278 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.4556 (* 1 = 6.4556 loss)
I0510 20:01:03.515293 39557 sgd_solver.cpp:105] Iteration 47300, lr = 0.1
I0510 20:01:30.178428 39557 solver.cpp:218] Iteration 47400 (3.75051 iter/s, 26.663s/100 iters), loss = 6.81173
I0510 20:01:30.178580 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.81173 (* 1 = 6.81173 loss)
I0510 20:01:30.178596 39557 sgd_solver.cpp:105] Iteration 47400, lr = 0.1
I0510 20:01:56.960530 39557 solver.cpp:218] Iteration 47500 (3.73387 iter/s, 26.7818s/100 iters), loss = 7.16837
I0510 20:01:56.960593 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.16837 (* 1 = 7.16837 loss)
I0510 20:01:56.960608 39557 sgd_solver.cpp:105] Iteration 47500, lr = 0.1
I0510 20:02:23.616883 39557 solver.cpp:218] Iteration 47600 (3.75148 iter/s, 26.6562s/100 iters), loss = 6.88146
I0510 20:02:23.617075 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.88146 (* 1 = 6.88146 loss)
I0510 20:02:23.617091 39557 sgd_solver.cpp:105] Iteration 47600, lr = 0.1
I0510 20:02:50.353878 39557 solver.cpp:218] Iteration 47700 (3.74018 iter/s, 26.7367s/100 iters), loss = 6.51213
I0510 20:02:50.353943 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.51213 (* 1 = 6.51213 loss)
I0510 20:02:50.353958 39557 sgd_solver.cpp:105] Iteration 47700, lr = 0.1
I0510 20:03:17.228786 39557 solver.cpp:218] Iteration 47800 (3.72097 iter/s, 26.8747s/100 iters), loss = 7.13802
I0510 20:03:17.228976 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.13802 (* 1 = 7.13802 loss)
I0510 20:03:17.228993 39557 sgd_solver.cpp:105] Iteration 47800, lr = 0.1
I0510 20:03:43.912112 39557 solver.cpp:218] Iteration 47900 (3.7477 iter/s, 26.683s/100 iters), loss = 6.69289
I0510 20:03:43.912176 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.69289 (* 1 = 6.69289 loss)
I0510 20:03:43.912189 39557 sgd_solver.cpp:105] Iteration 47900, lr = 0.1
I0510 20:04:10.503271 39557 solver.cpp:218] Iteration 48000 (3.76067 iter/s, 26.591s/100 iters), loss = 6.58713
I0510 20:04:10.503442 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.58713 (* 1 = 6.58713 loss)
I0510 20:04:10.503455 39557 sgd_solver.cpp:105] Iteration 48000, lr = 0.1
I0510 20:04:37.121918 39557 solver.cpp:218] Iteration 48100 (3.7568 iter/s, 26.6184s/100 iters), loss = 6.79854
I0510 20:04:37.121986 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.79854 (* 1 = 6.79854 loss)
I0510 20:04:37.121999 39557 sgd_solver.cpp:105] Iteration 48100, lr = 0.1
I0510 20:05:03.779052 39557 solver.cpp:218] Iteration 48200 (3.75137 iter/s, 26.657s/100 iters), loss = 6.69827
I0510 20:05:03.779259 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.69827 (* 1 = 6.69827 loss)
I0510 20:05:03.779276 39557 sgd_solver.cpp:105] Iteration 48200, lr = 0.1
I0510 20:05:30.563763 39557 solver.cpp:218] Iteration 48300 (3.73352 iter/s, 26.7844s/100 iters), loss = 6.90369
I0510 20:05:30.563829 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.90369 (* 1 = 6.90369 loss)
I0510 20:05:30.563843 39557 sgd_solver.cpp:105] Iteration 48300, lr = 0.1
I0510 20:05:57.260073 39557 solver.cpp:218] Iteration 48400 (3.74586 iter/s, 26.6961s/100 iters), loss = 6.95159
I0510 20:05:57.260283 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.95159 (* 1 = 6.95159 loss)
I0510 20:05:57.260300 39557 sgd_solver.cpp:105] Iteration 48400, lr = 0.1
I0510 20:06:24.040352 39557 solver.cpp:218] Iteration 48500 (3.73413 iter/s, 26.78s/100 iters), loss = 7.98572
I0510 20:06:24.040417 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.98572 (* 1 = 7.98572 loss)
I0510 20:06:24.040432 39557 sgd_solver.cpp:105] Iteration 48500, lr = 0.1
I0510 20:06:50.812532 39557 solver.cpp:218] Iteration 48600 (3.73525 iter/s, 26.772s/100 iters), loss = 7.71324
I0510 20:06:50.812706 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.71324 (* 1 = 7.71324 loss)
I0510 20:06:50.812722 39557 sgd_solver.cpp:105] Iteration 48600, lr = 0.1
I0510 20:07:17.527801 39557 solver.cpp:218] Iteration 48700 (3.74322 iter/s, 26.715s/100 iters), loss = 6.73139
I0510 20:07:17.527861 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.73139 (* 1 = 6.73139 loss)
I0510 20:07:17.527876 39557 sgd_solver.cpp:105] Iteration 48700, lr = 0.1
I0510 20:07:44.297276 39557 solver.cpp:218] Iteration 48800 (3.73562 iter/s, 26.7693s/100 iters), loss = 6.79512
I0510 20:07:44.297453 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.79512 (* 1 = 6.79512 loss)
I0510 20:07:44.297467 39557 sgd_solver.cpp:105] Iteration 48800, lr = 0.1
I0510 20:08:11.000233 39557 solver.cpp:218] Iteration 48900 (3.74494 iter/s, 26.7027s/100 iters), loss = 7.41165
I0510 20:08:11.000311 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.41165 (* 1 = 7.41165 loss)
I0510 20:08:11.000324 39557 sgd_solver.cpp:105] Iteration 48900, lr = 0.1
I0510 20:08:37.646458 39557 solver.cpp:218] Iteration 49000 (3.7529 iter/s, 26.6461s/100 iters), loss = 6.83765
I0510 20:08:37.646610 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.83765 (* 1 = 6.83765 loss)
I0510 20:08:37.646622 39557 sgd_solver.cpp:105] Iteration 49000, lr = 0.1
I0510 20:09:04.257292 39557 solver.cpp:218] Iteration 49100 (3.75791 iter/s, 26.6106s/100 iters), loss = 7.13552
I0510 20:09:04.257349 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.13552 (* 1 = 7.13552 loss)
I0510 20:09:04.257362 39557 sgd_solver.cpp:105] Iteration 49100, lr = 0.1
I0510 20:09:30.872619 39557 solver.cpp:218] Iteration 49200 (3.75726 iter/s, 26.6152s/100 iters), loss = 7.16862
I0510 20:09:30.872771 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.16862 (* 1 = 7.16862 loss)
I0510 20:09:30.872783 39557 sgd_solver.cpp:105] Iteration 49200, lr = 0.1
I0510 20:09:57.521289 39557 solver.cpp:218] Iteration 49300 (3.75257 iter/s, 26.6484s/100 iters), loss = 6.51295
I0510 20:09:57.521356 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.51295 (* 1 = 6.51295 loss)
I0510 20:09:57.521370 39557 sgd_solver.cpp:105] Iteration 49300, lr = 0.1
I0510 20:10:24.303934 39557 solver.cpp:218] Iteration 49400 (3.73381 iter/s, 26.7823s/100 iters), loss = 7.70182
I0510 20:10:24.304108 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.70182 (* 1 = 7.70182 loss)
I0510 20:10:24.304124 39557 sgd_solver.cpp:105] Iteration 49400, lr = 0.1
I0510 20:10:51.041663 39557 solver.cpp:218] Iteration 49500 (3.74009 iter/s, 26.7373s/100 iters), loss = 6.82309
I0510 20:10:51.041725 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.82309 (* 1 = 6.82309 loss)
I0510 20:10:51.041738 39557 sgd_solver.cpp:105] Iteration 49500, lr = 0.1
I0510 20:11:17.825814 39557 solver.cpp:218] Iteration 49600 (3.7336 iter/s, 26.7838s/100 iters), loss = 7.01644
I0510 20:11:17.826006 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.01644 (* 1 = 7.01644 loss)
I0510 20:11:17.826022 39557 sgd_solver.cpp:105] Iteration 49600, lr = 0.1
I0510 20:11:44.614274 39557 solver.cpp:218] Iteration 49700 (3.73301 iter/s, 26.788s/100 iters), loss = 6.92883
I0510 20:11:44.614336 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.92883 (* 1 = 6.92883 loss)
I0510 20:11:44.614349 39557 sgd_solver.cpp:105] Iteration 49700, lr = 0.1
I0510 20:12:11.386778 39557 solver.cpp:218] Iteration 49800 (3.73522 iter/s, 26.7722s/100 iters), loss = 6.39521
I0510 20:12:11.386942 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.39521 (* 1 = 6.39521 loss)
I0510 20:12:11.386958 39557 sgd_solver.cpp:105] Iteration 49800, lr = 0.1
I0510 20:12:38.167255 39557 solver.cpp:218] Iteration 49900 (3.73412 iter/s, 26.7801s/100 iters), loss = 7.28062
I0510 20:12:38.167315 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.28062 (* 1 = 7.28062 loss)
I0510 20:12:38.167330 39557 sgd_solver.cpp:105] Iteration 49900, lr = 0.1
I0510 20:13:04.637893 39557 solver.cpp:447] Snapshotting to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_50000.caffemodel
I0510 20:13:04.677840 39557 sgd_solver.cpp:273] Snapshotting solver state to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_50000.solverstate
I0510 20:13:04.960445 39557 solver.cpp:218] Iteration 50000 (3.73233 iter/s, 26.7929s/100 iters), loss = 7.0145
I0510 20:13:04.960500 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.0145 (* 1 = 7.0145 loss)
I0510 20:13:04.960510 39557 sgd_solver.cpp:105] Iteration 50000, lr = 0.1
I0510 20:13:31.715472 39557 solver.cpp:218] Iteration 50100 (3.73766 iter/s, 26.7547s/100 iters), loss = 6.19918
I0510 20:13:31.715533 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.19918 (* 1 = 6.19918 loss)
I0510 20:13:31.715545 39557 sgd_solver.cpp:105] Iteration 50100, lr = 0.1
I0510 20:13:58.475517 39557 solver.cpp:218] Iteration 50200 (3.73696 iter/s, 26.7598s/100 iters), loss = 7.5721
I0510 20:13:58.475680 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.5721 (* 1 = 7.5721 loss)
I0510 20:13:58.475697 39557 sgd_solver.cpp:105] Iteration 50200, lr = 0.1
I0510 20:14:25.111575 39557 solver.cpp:218] Iteration 50300 (3.75436 iter/s, 26.6357s/100 iters), loss = 7.41678
I0510 20:14:25.111644 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.41678 (* 1 = 7.41678 loss)
I0510 20:14:25.111660 39557 sgd_solver.cpp:105] Iteration 50300, lr = 0.1
I0510 20:14:51.762370 39557 solver.cpp:218] Iteration 50400 (3.75227 iter/s, 26.6505s/100 iters), loss = 6.85493
I0510 20:14:51.762534 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.85493 (* 1 = 6.85493 loss)
I0510 20:14:51.762547 39557 sgd_solver.cpp:105] Iteration 50400, lr = 0.1
I0510 20:15:18.451370 39557 solver.cpp:218] Iteration 50500 (3.74691 iter/s, 26.6886s/100 iters), loss = 7.21767
I0510 20:15:18.451426 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.21767 (* 1 = 7.21767 loss)
I0510 20:15:18.451438 39557 sgd_solver.cpp:105] Iteration 50500, lr = 0.1
I0510 20:15:45.110649 39557 solver.cpp:218] Iteration 50600 (3.75108 iter/s, 26.659s/100 iters), loss = 6.971
I0510 20:15:45.110821 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.971 (* 1 = 6.971 loss)
I0510 20:15:45.110834 39557 sgd_solver.cpp:105] Iteration 50600, lr = 0.1
I0510 20:16:11.688503 39557 solver.cpp:218] Iteration 50700 (3.76258 iter/s, 26.5775s/100 iters), loss = 6.4963
I0510 20:16:11.688565 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.4963 (* 1 = 6.4963 loss)
I0510 20:16:11.688583 39557 sgd_solver.cpp:105] Iteration 50700, lr = 0.1
I0510 20:16:38.331599 39557 solver.cpp:218] Iteration 50800 (3.75336 iter/s, 26.6428s/100 iters), loss = 7.06672
I0510 20:16:38.331781 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.06672 (* 1 = 7.06672 loss)
I0510 20:16:38.331797 39557 sgd_solver.cpp:105] Iteration 50800, lr = 0.1
I0510 20:17:05.003021 39557 solver.cpp:218] Iteration 50900 (3.74939 iter/s, 26.671s/100 iters), loss = 6.97196
I0510 20:17:05.003085 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.97196 (* 1 = 6.97196 loss)
I0510 20:17:05.003099 39557 sgd_solver.cpp:105] Iteration 50900, lr = 0.1
I0510 20:17:31.653550 39557 solver.cpp:218] Iteration 51000 (3.75231 iter/s, 26.6503s/100 iters), loss = 6.47084
I0510 20:17:31.653714 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.47084 (* 1 = 6.47084 loss)
I0510 20:17:31.653729 39557 sgd_solver.cpp:105] Iteration 51000, lr = 0.1
I0510 20:17:58.303908 39557 solver.cpp:218] Iteration 51100 (3.75234 iter/s, 26.65s/100 iters), loss = 6.49778
I0510 20:17:58.303969 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.49778 (* 1 = 6.49778 loss)
I0510 20:17:58.303982 39557 sgd_solver.cpp:105] Iteration 51100, lr = 0.1
I0510 20:18:24.950400 39557 solver.cpp:218] Iteration 51200 (3.75288 iter/s, 26.6462s/100 iters), loss = 6.61593
I0510 20:18:24.950570 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.61593 (* 1 = 6.61593 loss)
I0510 20:18:24.950585 39557 sgd_solver.cpp:105] Iteration 51200, lr = 0.1
I0510 20:18:51.544358 39557 solver.cpp:218] Iteration 51300 (3.7603 iter/s, 26.5936s/100 iters), loss = 7.1643
I0510 20:18:51.544418 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.1643 (* 1 = 7.1643 loss)
I0510 20:18:51.544428 39557 sgd_solver.cpp:105] Iteration 51300, lr = 0.1
I0510 20:19:18.182101 39557 solver.cpp:218] Iteration 51400 (3.75411 iter/s, 26.6375s/100 iters), loss = 6.28479
I0510 20:19:18.182263 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.28479 (* 1 = 6.28479 loss)
I0510 20:19:18.182278 39557 sgd_solver.cpp:105] Iteration 51400, lr = 0.1
I0510 20:19:44.846474 39557 solver.cpp:218] Iteration 51500 (3.75037 iter/s, 26.664s/100 iters), loss = 7.1559
I0510 20:19:44.846532 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.1559 (* 1 = 7.1559 loss)
I0510 20:19:44.846545 39557 sgd_solver.cpp:105] Iteration 51500, lr = 0.1
I0510 20:20:11.445668 39557 solver.cpp:218] Iteration 51600 (3.75955 iter/s, 26.599s/100 iters), loss = 7.35963
I0510 20:20:11.445819 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.35963 (* 1 = 7.35963 loss)
I0510 20:20:11.445845 39557 sgd_solver.cpp:105] Iteration 51600, lr = 0.1
I0510 20:20:38.053624 39557 solver.cpp:218] Iteration 51700 (3.75832 iter/s, 26.6076s/100 iters), loss = 6.60433
I0510 20:20:38.053680 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.60433 (* 1 = 6.60433 loss)
I0510 20:20:38.053690 39557 sgd_solver.cpp:105] Iteration 51700, lr = 0.1
I0510 20:21:04.641336 39557 solver.cpp:218] Iteration 51800 (3.76117 iter/s, 26.5875s/100 iters), loss = 7.30245
I0510 20:21:04.641491 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.30245 (* 1 = 7.30245 loss)
I0510 20:21:04.641506 39557 sgd_solver.cpp:105] Iteration 51800, lr = 0.1
I0510 20:21:31.222436 39557 solver.cpp:218] Iteration 51900 (3.76212 iter/s, 26.5808s/100 iters), loss = 6.7098
I0510 20:21:31.222497 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.7098 (* 1 = 6.7098 loss)
I0510 20:21:31.222510 39557 sgd_solver.cpp:105] Iteration 51900, lr = 0.1
I0510 20:21:57.836983 39557 solver.cpp:218] Iteration 52000 (3.75738 iter/s, 26.6143s/100 iters), loss = 6.85458
I0510 20:21:57.837137 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.85458 (* 1 = 6.85458 loss)
I0510 20:21:57.837152 39557 sgd_solver.cpp:105] Iteration 52000, lr = 0.1
I0510 20:22:24.423512 39557 solver.cpp:218] Iteration 52100 (3.76135 iter/s, 26.5862s/100 iters), loss = 6.6443
I0510 20:22:24.423571 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.6443 (* 1 = 6.6443 loss)
I0510 20:22:24.423584 39557 sgd_solver.cpp:105] Iteration 52100, lr = 0.1
I0510 20:22:51.024233 39557 solver.cpp:218] Iteration 52200 (3.75933 iter/s, 26.6005s/100 iters), loss = 6.68918
I0510 20:22:51.024427 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.68918 (* 1 = 6.68918 loss)
I0510 20:22:51.024453 39557 sgd_solver.cpp:105] Iteration 52200, lr = 0.1
I0510 20:23:17.666496 39557 solver.cpp:218] Iteration 52300 (3.75349 iter/s, 26.6419s/100 iters), loss = 6.82501
I0510 20:23:17.666563 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.82501 (* 1 = 6.82501 loss)
I0510 20:23:17.666573 39557 sgd_solver.cpp:105] Iteration 52300, lr = 0.1
I0510 20:23:44.243880 39557 solver.cpp:218] Iteration 52400 (3.76263 iter/s, 26.5772s/100 iters), loss = 7.0357
I0510 20:23:44.244027 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.0357 (* 1 = 7.0357 loss)
I0510 20:23:44.244052 39557 sgd_solver.cpp:105] Iteration 52400, lr = 0.1
I0510 20:24:10.866506 39557 solver.cpp:218] Iteration 52500 (3.75625 iter/s, 26.6223s/100 iters), loss = 6.65671
I0510 20:24:10.866564 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.65671 (* 1 = 6.65671 loss)
I0510 20:24:10.866575 39557 sgd_solver.cpp:105] Iteration 52500, lr = 0.1
I0510 20:24:37.531903 39557 solver.cpp:218] Iteration 52600 (3.75021 iter/s, 26.6652s/100 iters), loss = 7.0574
I0510 20:24:37.532047 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.0574 (* 1 = 7.0574 loss)
I0510 20:24:37.532071 39557 sgd_solver.cpp:105] Iteration 52600, lr = 0.1
I0510 20:25:04.133332 39557 solver.cpp:218] Iteration 52700 (3.75924 iter/s, 26.6011s/100 iters), loss = 7.09083
I0510 20:25:04.133388 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.09083 (* 1 = 7.09083 loss)
I0510 20:25:04.133399 39557 sgd_solver.cpp:105] Iteration 52700, lr = 0.1
I0510 20:25:30.745086 39557 solver.cpp:218] Iteration 52800 (3.75777 iter/s, 26.6115s/100 iters), loss = 6.1964
I0510 20:25:30.745236 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.1964 (* 1 = 6.1964 loss)
I0510 20:25:30.745252 39557 sgd_solver.cpp:105] Iteration 52800, lr = 0.1
I0510 20:25:57.390978 39557 solver.cpp:218] Iteration 52900 (3.75297 iter/s, 26.6456s/100 iters), loss = 7.73997
I0510 20:25:57.391034 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.73997 (* 1 = 7.73997 loss)
I0510 20:25:57.391044 39557 sgd_solver.cpp:105] Iteration 52900, lr = 0.1
I0510 20:26:24.041138 39557 solver.cpp:218] Iteration 53000 (3.75235 iter/s, 26.6499s/100 iters), loss = 6.48276
I0510 20:26:24.041301 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.48276 (* 1 = 6.48276 loss)
I0510 20:26:24.041316 39557 sgd_solver.cpp:105] Iteration 53000, lr = 0.1
I0510 20:26:50.736737 39557 solver.cpp:218] Iteration 53100 (3.74598 iter/s, 26.6953s/100 iters), loss = 6.52019
I0510 20:26:50.736799 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.52019 (* 1 = 6.52019 loss)
I0510 20:26:50.736811 39557 sgd_solver.cpp:105] Iteration 53100, lr = 0.1
I0510 20:27:17.463040 39557 solver.cpp:218] Iteration 53200 (3.74166 iter/s, 26.7261s/100 iters), loss = 7.05126
I0510 20:27:17.463186 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.05126 (* 1 = 7.05126 loss)
I0510 20:27:17.463203 39557 sgd_solver.cpp:105] Iteration 53200, lr = 0.1
I0510 20:27:44.105520 39557 solver.cpp:218] Iteration 53300 (3.75345 iter/s, 26.6422s/100 iters), loss = 7.13839
I0510 20:27:44.105582 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.13839 (* 1 = 7.13839 loss)
I0510 20:27:44.105594 39557 sgd_solver.cpp:105] Iteration 53300, lr = 0.1
I0510 20:28:10.753141 39557 solver.cpp:218] Iteration 53400 (3.75271 iter/s, 26.6474s/100 iters), loss = 6.89165
I0510 20:28:10.753298 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.89165 (* 1 = 6.89165 loss)
I0510 20:28:10.753315 39557 sgd_solver.cpp:105] Iteration 53400, lr = 0.1
I0510 20:28:37.391948 39557 solver.cpp:218] Iteration 53500 (3.75397 iter/s, 26.6385s/100 iters), loss = 7.12678
I0510 20:28:37.392005 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.12678 (* 1 = 7.12678 loss)
I0510 20:28:37.392016 39557 sgd_solver.cpp:105] Iteration 53500, lr = 0.1
I0510 20:29:04.004175 39557 solver.cpp:218] Iteration 53600 (3.7577 iter/s, 26.612s/100 iters), loss = 6.17817
I0510 20:29:04.004379 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.17817 (* 1 = 6.17817 loss)
I0510 20:29:04.004395 39557 sgd_solver.cpp:105] Iteration 53600, lr = 0.1
I0510 20:29:30.644634 39557 solver.cpp:218] Iteration 53700 (3.75374 iter/s, 26.6401s/100 iters), loss = 7.20096
I0510 20:29:30.644690 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.20096 (* 1 = 7.20096 loss)
I0510 20:29:30.644706 39557 sgd_solver.cpp:105] Iteration 53700, lr = 0.1
I0510 20:29:57.288467 39557 solver.cpp:218] Iteration 53800 (3.75324 iter/s, 26.6436s/100 iters), loss = 7.01716
I0510 20:29:57.288635 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.01716 (* 1 = 7.01716 loss)
I0510 20:29:57.288650 39557 sgd_solver.cpp:105] Iteration 53800, lr = 0.1
I0510 20:30:23.982363 39557 solver.cpp:218] Iteration 53900 (3.74622 iter/s, 26.6936s/100 iters), loss = 7.50779
I0510 20:30:23.982424 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.50779 (* 1 = 7.50779 loss)
I0510 20:30:23.982435 39557 sgd_solver.cpp:105] Iteration 53900, lr = 0.1
I0510 20:30:50.577409 39557 solver.cpp:218] Iteration 54000 (3.76013 iter/s, 26.5948s/100 iters), loss = 6.24062
I0510 20:30:50.577576 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.24062 (* 1 = 6.24062 loss)
I0510 20:30:50.577594 39557 sgd_solver.cpp:105] Iteration 54000, lr = 0.1
I0510 20:31:17.406829 39557 solver.cpp:218] Iteration 54100 (3.7273 iter/s, 26.8291s/100 iters), loss = 6.5878
I0510 20:31:17.406898 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.5878 (* 1 = 6.5878 loss)
I0510 20:31:17.406913 39557 sgd_solver.cpp:105] Iteration 54100, lr = 0.1
I0510 20:31:44.188964 39557 solver.cpp:218] Iteration 54200 (3.73386 iter/s, 26.7819s/100 iters), loss = 6.81113
I0510 20:31:44.189123 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.81113 (* 1 = 6.81113 loss)
I0510 20:31:44.189138 39557 sgd_solver.cpp:105] Iteration 54200, lr = 0.1
I0510 20:32:10.897620 39557 solver.cpp:218] Iteration 54300 (3.74415 iter/s, 26.7083s/100 iters), loss = 7.1628
I0510 20:32:10.897686 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.1628 (* 1 = 7.1628 loss)
I0510 20:32:10.897701 39557 sgd_solver.cpp:105] Iteration 54300, lr = 0.1
I0510 20:32:37.721724 39557 solver.cpp:218] Iteration 54400 (3.72802 iter/s, 26.8239s/100 iters), loss = 6.72818
I0510 20:32:37.721874 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.72818 (* 1 = 6.72818 loss)
I0510 20:32:37.721890 39557 sgd_solver.cpp:105] Iteration 54400, lr = 0.1
I0510 20:33:04.502003 39557 solver.cpp:218] Iteration 54500 (3.73413 iter/s, 26.78s/100 iters), loss = 6.54597
I0510 20:33:04.502074 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.54597 (* 1 = 6.54597 loss)
I0510 20:33:04.502087 39557 sgd_solver.cpp:105] Iteration 54500, lr = 0.1
I0510 20:33:31.241477 39557 solver.cpp:218] Iteration 54600 (3.73982 iter/s, 26.7393s/100 iters), loss = 6.56488
I0510 20:33:31.241636 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.56488 (* 1 = 6.56488 loss)
I0510 20:33:31.241648 39557 sgd_solver.cpp:105] Iteration 54600, lr = 0.1
I0510 20:33:57.923537 39557 solver.cpp:218] Iteration 54700 (3.74788 iter/s, 26.6817s/100 iters), loss = 7.07652
I0510 20:33:57.923596 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.07652 (* 1 = 7.07652 loss)
I0510 20:33:57.923612 39557 sgd_solver.cpp:105] Iteration 54700, lr = 0.1
I0510 20:34:24.696951 39557 solver.cpp:218] Iteration 54800 (3.73508 iter/s, 26.7732s/100 iters), loss = 6.64923
I0510 20:34:24.697149 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.64923 (* 1 = 6.64923 loss)
I0510 20:34:24.697162 39557 sgd_solver.cpp:105] Iteration 54800, lr = 0.1
I0510 20:34:51.484205 39557 solver.cpp:218] Iteration 54900 (3.73317 iter/s, 26.7869s/100 iters), loss = 6.34297
I0510 20:34:51.484268 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.34297 (* 1 = 6.34297 loss)
I0510 20:34:51.484285 39557 sgd_solver.cpp:105] Iteration 54900, lr = 0.1
I0510 20:35:17.970623 39557 solver.cpp:447] Snapshotting to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_55000.caffemodel
I0510 20:35:18.010152 39557 sgd_solver.cpp:273] Snapshotting solver state to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_55000.solverstate
I0510 20:35:18.295323 39557 solver.cpp:218] Iteration 55000 (3.72982 iter/s, 26.8109s/100 iters), loss = 6.4825
I0510 20:35:18.295382 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.4825 (* 1 = 6.4825 loss)
I0510 20:35:18.295392 39557 sgd_solver.cpp:105] Iteration 55000, lr = 0.1
I0510 20:35:45.012006 39557 solver.cpp:218] Iteration 55100 (3.74301 iter/s, 26.7165s/100 iters), loss = 6.62753
I0510 20:35:45.012064 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.62753 (* 1 = 6.62753 loss)
I0510 20:35:45.012078 39557 sgd_solver.cpp:105] Iteration 55100, lr = 0.1
I0510 20:36:11.708891 39557 solver.cpp:218] Iteration 55200 (3.74578 iter/s, 26.6967s/100 iters), loss = 6.67975
I0510 20:36:11.709048 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.67975 (* 1 = 6.67975 loss)
I0510 20:36:11.709061 39557 sgd_solver.cpp:105] Iteration 55200, lr = 0.1
I0510 20:36:38.468848 39557 solver.cpp:218] Iteration 55300 (3.73697 iter/s, 26.7597s/100 iters), loss = 6.99611
I0510 20:36:38.468909 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.99611 (* 1 = 6.99611 loss)
I0510 20:36:38.468922 39557 sgd_solver.cpp:105] Iteration 55300, lr = 0.1
I0510 20:37:05.260202 39557 solver.cpp:218] Iteration 55400 (3.73258 iter/s, 26.7911s/100 iters), loss = 6.74046
I0510 20:37:05.260370 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.74046 (* 1 = 6.74046 loss)
I0510 20:37:05.260386 39557 sgd_solver.cpp:105] Iteration 55400, lr = 0.1
I0510 20:37:32.084892 39557 solver.cpp:218] Iteration 55500 (3.72795 iter/s, 26.8244s/100 iters), loss = 7.49274
I0510 20:37:32.084954 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.49274 (* 1 = 7.49274 loss)
I0510 20:37:32.084966 39557 sgd_solver.cpp:105] Iteration 55500, lr = 0.1
I0510 20:37:58.675576 39557 solver.cpp:218] Iteration 55600 (3.76074 iter/s, 26.5905s/100 iters), loss = 6.32324
I0510 20:37:58.675726 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.32324 (* 1 = 6.32324 loss)
I0510 20:37:58.675743 39557 sgd_solver.cpp:105] Iteration 55600, lr = 0.1
I0510 20:38:25.409932 39557 solver.cpp:218] Iteration 55700 (3.74055 iter/s, 26.7341s/100 iters), loss = 6.89047
I0510 20:38:25.409996 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.89047 (* 1 = 6.89047 loss)
I0510 20:38:25.410009 39557 sgd_solver.cpp:105] Iteration 55700, lr = 0.1
I0510 20:38:52.163094 39557 solver.cpp:218] Iteration 55800 (3.73791 iter/s, 26.753s/100 iters), loss = 6.27308
I0510 20:38:52.163242 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.27308 (* 1 = 6.27308 loss)
I0510 20:38:52.163257 39557 sgd_solver.cpp:105] Iteration 55800, lr = 0.1
I0510 20:39:18.838510 39557 solver.cpp:218] Iteration 55900 (3.74881 iter/s, 26.6751s/100 iters), loss = 6.75769
I0510 20:39:18.838569 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.75769 (* 1 = 6.75769 loss)
I0510 20:39:18.838579 39557 sgd_solver.cpp:105] Iteration 55900, lr = 0.1
I0510 20:39:45.452939 39557 solver.cpp:218] Iteration 56000 (3.75739 iter/s, 26.6142s/100 iters), loss = 6.96713
I0510 20:39:45.453099 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.96713 (* 1 = 6.96713 loss)
I0510 20:39:45.453114 39557 sgd_solver.cpp:105] Iteration 56000, lr = 0.1
I0510 20:40:12.069505 39557 solver.cpp:218] Iteration 56100 (3.7571 iter/s, 26.6163s/100 iters), loss = 6.50879
I0510 20:40:12.069564 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.50879 (* 1 = 6.50879 loss)
I0510 20:40:12.069576 39557 sgd_solver.cpp:105] Iteration 56100, lr = 0.1
I0510 20:40:38.721586 39557 solver.cpp:218] Iteration 56200 (3.75208 iter/s, 26.6519s/100 iters), loss = 6.68187
I0510 20:40:38.721798 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.68187 (* 1 = 6.68187 loss)
I0510 20:40:38.721814 39557 sgd_solver.cpp:105] Iteration 56200, lr = 0.1
I0510 20:41:05.377854 39557 solver.cpp:218] Iteration 56300 (3.75151 iter/s, 26.6559s/100 iters), loss = 7.23777
I0510 20:41:05.377912 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.23777 (* 1 = 7.23777 loss)
I0510 20:41:05.377923 39557 sgd_solver.cpp:105] Iteration 56300, lr = 0.1
I0510 20:41:32.172521 39557 solver.cpp:218] Iteration 56400 (3.73212 iter/s, 26.7945s/100 iters), loss = 7.17041
I0510 20:41:32.172739 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.17041 (* 1 = 7.17041 loss)
I0510 20:41:32.172756 39557 sgd_solver.cpp:105] Iteration 56400, lr = 0.1
I0510 20:41:58.967078 39557 solver.cpp:218] Iteration 56500 (3.73215 iter/s, 26.7942s/100 iters), loss = 6.44929
I0510 20:41:58.967149 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.44929 (* 1 = 6.44929 loss)
I0510 20:41:58.967164 39557 sgd_solver.cpp:105] Iteration 56500, lr = 0.1
I0510 20:42:25.748116 39557 solver.cpp:218] Iteration 56600 (3.73401 iter/s, 26.7808s/100 iters), loss = 7.04846
I0510 20:42:25.748292 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.04846 (* 1 = 7.04846 loss)
I0510 20:42:25.748311 39557 sgd_solver.cpp:105] Iteration 56600, lr = 0.1
I0510 20:42:52.489161 39557 solver.cpp:218] Iteration 56700 (3.73961 iter/s, 26.7407s/100 iters), loss = 7.09203
I0510 20:42:52.489228 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.09203 (* 1 = 7.09203 loss)
I0510 20:42:52.489241 39557 sgd_solver.cpp:105] Iteration 56700, lr = 0.1
I0510 20:43:19.229043 39557 solver.cpp:218] Iteration 56800 (3.73976 iter/s, 26.7397s/100 iters), loss = 6.79877
I0510 20:43:19.229223 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.79877 (* 1 = 6.79877 loss)
I0510 20:43:19.229244 39557 sgd_solver.cpp:105] Iteration 56800, lr = 0.1
I0510 20:43:45.922413 39557 solver.cpp:218] Iteration 56900 (3.74629 iter/s, 26.6931s/100 iters), loss = 6.30432
I0510 20:43:45.922474 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.30432 (* 1 = 6.30432 loss)
I0510 20:43:45.922489 39557 sgd_solver.cpp:105] Iteration 56900, lr = 0.1
I0510 20:44:12.628774 39557 solver.cpp:218] Iteration 57000 (3.74445 iter/s, 26.7062s/100 iters), loss = 6.29963
I0510 20:44:12.628947 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.29963 (* 1 = 6.29963 loss)
I0510 20:44:12.628968 39557 sgd_solver.cpp:105] Iteration 57000, lr = 0.1
I0510 20:44:39.351770 39557 solver.cpp:218] Iteration 57100 (3.74214 iter/s, 26.7227s/100 iters), loss = 7.13145
I0510 20:44:39.351832 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.13145 (* 1 = 7.13145 loss)
I0510 20:44:39.351847 39557 sgd_solver.cpp:105] Iteration 57100, lr = 0.1
I0510 20:45:06.030370 39557 solver.cpp:218] Iteration 57200 (3.74835 iter/s, 26.6784s/100 iters), loss = 7.49593
I0510 20:45:06.030542 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.49593 (* 1 = 7.49593 loss)
I0510 20:45:06.030557 39557 sgd_solver.cpp:105] Iteration 57200, lr = 0.1
I0510 20:45:32.627707 39557 solver.cpp:218] Iteration 57300 (3.75982 iter/s, 26.597s/100 iters), loss = 6.86618
I0510 20:45:32.627764 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.86618 (* 1 = 6.86618 loss)
I0510 20:45:32.627774 39557 sgd_solver.cpp:105] Iteration 57300, lr = 0.1
I0510 20:45:59.220655 39557 solver.cpp:218] Iteration 57400 (3.76042 iter/s, 26.5927s/100 iters), loss = 6.59293
I0510 20:45:59.220799 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.59293 (* 1 = 6.59293 loss)
I0510 20:45:59.220824 39557 sgd_solver.cpp:105] Iteration 57400, lr = 0.1
I0510 20:46:25.868445 39557 solver.cpp:218] Iteration 57500 (3.7527 iter/s, 26.6475s/100 iters), loss = 6.72762
I0510 20:46:25.868525 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.72762 (* 1 = 6.72762 loss)
I0510 20:46:25.868535 39557 sgd_solver.cpp:105] Iteration 57500, lr = 0.1
I0510 20:46:52.475317 39557 solver.cpp:218] Iteration 57600 (3.75846 iter/s, 26.6066s/100 iters), loss = 7.45934
I0510 20:46:52.475502 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.45934 (* 1 = 7.45934 loss)
I0510 20:46:52.475528 39557 sgd_solver.cpp:105] Iteration 57600, lr = 0.1
I0510 20:47:19.130594 39557 solver.cpp:218] Iteration 57700 (3.75165 iter/s, 26.6549s/100 iters), loss = 6.55056
I0510 20:47:19.130653 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.55056 (* 1 = 6.55056 loss)
I0510 20:47:19.130663 39557 sgd_solver.cpp:105] Iteration 57700, lr = 0.1
I0510 20:47:45.727438 39557 solver.cpp:218] Iteration 57800 (3.75987 iter/s, 26.5966s/100 iters), loss = 6.4498
I0510 20:47:45.727586 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.4498 (* 1 = 6.4498 loss)
I0510 20:47:45.727612 39557 sgd_solver.cpp:105] Iteration 57800, lr = 0.1
I0510 20:48:12.296969 39557 solver.cpp:218] Iteration 57900 (3.76375 iter/s, 26.5692s/100 iters), loss = 6.36816
I0510 20:48:12.297027 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.36816 (* 1 = 6.36816 loss)
I0510 20:48:12.297039 39557 sgd_solver.cpp:105] Iteration 57900, lr = 0.1
I0510 20:48:38.868646 39557 solver.cpp:218] Iteration 58000 (3.76343 iter/s, 26.5715s/100 iters), loss = 6.73146
I0510 20:48:38.868791 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.73146 (* 1 = 6.73146 loss)
I0510 20:48:38.868816 39557 sgd_solver.cpp:105] Iteration 58000, lr = 0.1
I0510 20:49:05.467128 39557 solver.cpp:218] Iteration 58100 (3.75966 iter/s, 26.5982s/100 iters), loss = 6.7041
I0510 20:49:05.467195 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.7041 (* 1 = 6.7041 loss)
I0510 20:49:05.467205 39557 sgd_solver.cpp:105] Iteration 58100, lr = 0.1
I0510 20:49:32.041393 39557 solver.cpp:218] Iteration 58200 (3.76307 iter/s, 26.5741s/100 iters), loss = 7.18503
I0510 20:49:32.041537 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.18503 (* 1 = 7.18503 loss)
I0510 20:49:32.041551 39557 sgd_solver.cpp:105] Iteration 58200, lr = 0.1
I0510 20:49:58.729993 39557 solver.cpp:218] Iteration 58300 (3.74696 iter/s, 26.6883s/100 iters), loss = 7.2585
I0510 20:49:58.730063 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.2585 (* 1 = 7.2585 loss)
I0510 20:49:58.730080 39557 sgd_solver.cpp:105] Iteration 58300, lr = 0.1
I0510 20:50:25.547075 39557 solver.cpp:218] Iteration 58400 (3.729 iter/s, 26.8169s/100 iters), loss = 6.81547
I0510 20:50:25.547255 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.81547 (* 1 = 6.81547 loss)
I0510 20:50:25.547271 39557 sgd_solver.cpp:105] Iteration 58400, lr = 0.1
I0510 20:50:52.295672 39557 solver.cpp:218] Iteration 58500 (3.73856 iter/s, 26.7483s/100 iters), loss = 6.49714
I0510 20:50:52.295735 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.49714 (* 1 = 6.49714 loss)
I0510 20:50:52.295749 39557 sgd_solver.cpp:105] Iteration 58500, lr = 0.1
I0510 20:51:18.994240 39557 solver.cpp:218] Iteration 58600 (3.74555 iter/s, 26.6984s/100 iters), loss = 6.29523
I0510 20:51:18.994405 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.29523 (* 1 = 6.29523 loss)
I0510 20:51:18.994419 39557 sgd_solver.cpp:105] Iteration 58600, lr = 0.1
I0510 20:51:45.645833 39557 solver.cpp:218] Iteration 58700 (3.75216 iter/s, 26.6513s/100 iters), loss = 6.62898
I0510 20:51:45.645898 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.62898 (* 1 = 6.62898 loss)
I0510 20:51:45.645912 39557 sgd_solver.cpp:105] Iteration 58700, lr = 0.1
I0510 20:52:12.245750 39557 solver.cpp:218] Iteration 58800 (3.75944 iter/s, 26.5997s/100 iters), loss = 7.20182
I0510 20:52:12.245952 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.20182 (* 1 = 7.20182 loss)
I0510 20:52:12.245980 39557 sgd_solver.cpp:105] Iteration 58800, lr = 0.1
I0510 20:52:38.835803 39557 solver.cpp:218] Iteration 58900 (3.76085 iter/s, 26.5897s/100 iters), loss = 7.35867
I0510 20:52:38.835868 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.35867 (* 1 = 7.35867 loss)
I0510 20:52:38.835881 39557 sgd_solver.cpp:105] Iteration 58900, lr = 0.1
I0510 20:53:05.446441 39557 solver.cpp:218] Iteration 59000 (3.75792 iter/s, 26.6104s/100 iters), loss = 7.20871
I0510 20:53:05.446600 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.20871 (* 1 = 7.20871 loss)
I0510 20:53:05.446616 39557 sgd_solver.cpp:105] Iteration 59000, lr = 0.1
I0510 20:53:32.087054 39557 solver.cpp:218] Iteration 59100 (3.75371 iter/s, 26.6403s/100 iters), loss = 6.77249
I0510 20:53:32.087108 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.77249 (* 1 = 6.77249 loss)
I0510 20:53:32.087121 39557 sgd_solver.cpp:105] Iteration 59100, lr = 0.1
I0510 20:53:58.680094 39557 solver.cpp:218] Iteration 59200 (3.76041 iter/s, 26.5928s/100 iters), loss = 6.61016
I0510 20:53:58.680248 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.61016 (* 1 = 6.61016 loss)
I0510 20:53:58.680259 39557 sgd_solver.cpp:105] Iteration 59200, lr = 0.1
I0510 20:54:25.574893 39557 solver.cpp:218] Iteration 59300 (3.71823 iter/s, 26.8945s/100 iters), loss = 5.85472
I0510 20:54:25.574972 39557 solver.cpp:237]     Train net output #0: softmax_loss = 5.85472 (* 1 = 5.85472 loss)
I0510 20:54:25.574985 39557 sgd_solver.cpp:105] Iteration 59300, lr = 0.1
I0510 20:54:52.250267 39557 solver.cpp:218] Iteration 59400 (3.74881 iter/s, 26.6751s/100 iters), loss = 7.00621
I0510 20:54:52.250434 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.00621 (* 1 = 7.00621 loss)
I0510 20:54:52.250450 39557 sgd_solver.cpp:105] Iteration 59400, lr = 0.1
I0510 20:55:18.931928 39557 solver.cpp:218] Iteration 59500 (3.74794 iter/s, 26.6814s/100 iters), loss = 6.43959
I0510 20:55:18.931995 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.43959 (* 1 = 6.43959 loss)
I0510 20:55:18.932010 39557 sgd_solver.cpp:105] Iteration 59500, lr = 0.1
I0510 20:55:45.560062 39557 solver.cpp:218] Iteration 59600 (3.75546 iter/s, 26.6279s/100 iters), loss = 6.25223
I0510 20:55:45.560221 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.25223 (* 1 = 6.25223 loss)
I0510 20:55:45.560238 39557 sgd_solver.cpp:105] Iteration 59600, lr = 0.1
I0510 20:56:12.248538 39557 solver.cpp:218] Iteration 59700 (3.74698 iter/s, 26.6882s/100 iters), loss = 6.10597
I0510 20:56:12.248602 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.10597 (* 1 = 6.10597 loss)
I0510 20:56:12.248615 39557 sgd_solver.cpp:105] Iteration 59700, lr = 0.1
I0510 20:56:38.913539 39557 solver.cpp:218] Iteration 59800 (3.75026 iter/s, 26.6648s/100 iters), loss = 6.95244
I0510 20:56:38.913705 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.95244 (* 1 = 6.95244 loss)
I0510 20:56:38.913722 39557 sgd_solver.cpp:105] Iteration 59800, lr = 0.1
I0510 20:57:05.569344 39557 solver.cpp:218] Iteration 59900 (3.75157 iter/s, 26.6555s/100 iters), loss = 6.80898
I0510 20:57:05.569403 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.80898 (* 1 = 6.80898 loss)
I0510 20:57:05.569416 39557 sgd_solver.cpp:105] Iteration 59900, lr = 0.1
I0510 20:57:31.989018 39557 solver.cpp:447] Snapshotting to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_60000.caffemodel
I0510 20:57:32.028648 39557 sgd_solver.cpp:273] Snapshotting solver state to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_60000.solverstate
I0510 20:57:32.311432 39557 solver.cpp:218] Iteration 60000 (3.73945 iter/s, 26.7419s/100 iters), loss = 6.45281
I0510 20:57:32.311491 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.45281 (* 1 = 6.45281 loss)
I0510 20:57:32.311501 39557 sgd_solver.cpp:105] Iteration 60000, lr = 0.1
I0510 20:57:59.065294 39557 solver.cpp:218] Iteration 60100 (3.73781 iter/s, 26.7536s/100 iters), loss = 6.26006
I0510 20:57:59.065356 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.26006 (* 1 = 6.26006 loss)
I0510 20:57:59.065367 39557 sgd_solver.cpp:105] Iteration 60100, lr = 0.1
I0510 20:58:25.724622 39557 solver.cpp:218] Iteration 60200 (3.75106 iter/s, 26.6591s/100 iters), loss = 7.32064
I0510 20:58:25.724836 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.32064 (* 1 = 7.32064 loss)
I0510 20:58:25.724853 39557 sgd_solver.cpp:105] Iteration 60200, lr = 0.1
I0510 20:58:52.408195 39557 solver.cpp:218] Iteration 60300 (3.74767 iter/s, 26.6832s/100 iters), loss = 7.25348
I0510 20:58:52.408255 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.25348 (* 1 = 7.25348 loss)
I0510 20:58:52.408267 39557 sgd_solver.cpp:105] Iteration 60300, lr = 0.1
I0510 20:59:19.085980 39557 solver.cpp:218] Iteration 60400 (3.74847 iter/s, 26.6776s/100 iters), loss = 6.47922
I0510 20:59:19.086139 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.47922 (* 1 = 6.47922 loss)
I0510 20:59:19.086164 39557 sgd_solver.cpp:105] Iteration 60400, lr = 0.1
I0510 20:59:45.768580 39557 solver.cpp:218] Iteration 60500 (3.7478 iter/s, 26.6823s/100 iters), loss = 6.70422
I0510 20:59:45.768642 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.70422 (* 1 = 6.70422 loss)
I0510 20:59:45.768652 39557 sgd_solver.cpp:105] Iteration 60500, lr = 0.1
I0510 21:00:12.460181 39557 solver.cpp:218] Iteration 60600 (3.74653 iter/s, 26.6914s/100 iters), loss = 7.08959
I0510 21:00:12.460336 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.08959 (* 1 = 7.08959 loss)
I0510 21:00:12.460362 39557 sgd_solver.cpp:105] Iteration 60600, lr = 0.1
I0510 21:00:39.124554 39557 solver.cpp:218] Iteration 60700 (3.75036 iter/s, 26.6641s/100 iters), loss = 7.48745
I0510 21:00:39.124632 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.48745 (* 1 = 7.48745 loss)
I0510 21:00:39.124642 39557 sgd_solver.cpp:105] Iteration 60700, lr = 0.1
I0510 21:01:06.074684 39557 solver.cpp:218] Iteration 60800 (3.71059 iter/s, 26.9499s/100 iters), loss = 7.29116
I0510 21:01:06.074898 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.29116 (* 1 = 7.29116 loss)
I0510 21:01:06.074915 39557 sgd_solver.cpp:105] Iteration 60800, lr = 0.1
I0510 21:01:33.056064 39557 solver.cpp:218] Iteration 60900 (3.70631 iter/s, 26.981s/100 iters), loss = 7.56589
I0510 21:01:33.056123 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.56589 (* 1 = 7.56589 loss)
I0510 21:01:33.056135 39557 sgd_solver.cpp:105] Iteration 60900, lr = 0.1
I0510 21:01:59.638264 39557 solver.cpp:218] Iteration 61000 (3.76194 iter/s, 26.582s/100 iters), loss = 6.63649
I0510 21:01:59.638494 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.63649 (* 1 = 6.63649 loss)
I0510 21:01:59.638506 39557 sgd_solver.cpp:105] Iteration 61000, lr = 0.1
I0510 21:02:26.204479 39557 solver.cpp:218] Iteration 61100 (3.76423 iter/s, 26.5658s/100 iters), loss = 6.47576
I0510 21:02:26.204530 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.47576 (* 1 = 6.47576 loss)
I0510 21:02:26.204540 39557 sgd_solver.cpp:105] Iteration 61100, lr = 0.1
I0510 21:02:53.394990 39557 solver.cpp:218] Iteration 61200 (3.67778 iter/s, 27.1903s/100 iters), loss = 7.01902
I0510 21:02:53.395231 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.01902 (* 1 = 7.01902 loss)
I0510 21:02:53.395246 39557 sgd_solver.cpp:105] Iteration 61200, lr = 0.1
I0510 21:03:20.021193 39557 solver.cpp:218] Iteration 61300 (3.75575 iter/s, 26.6258s/100 iters), loss = 6.44025
I0510 21:03:20.021260 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.44025 (* 1 = 6.44025 loss)
I0510 21:03:20.021270 39557 sgd_solver.cpp:105] Iteration 61300, lr = 0.1
I0510 21:03:46.643527 39557 solver.cpp:218] Iteration 61400 (3.75627 iter/s, 26.6221s/100 iters), loss = 6.67098
I0510 21:03:46.643697 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.67098 (* 1 = 6.67098 loss)
I0510 21:03:46.643712 39557 sgd_solver.cpp:105] Iteration 61400, lr = 0.1
I0510 21:04:13.319538 39557 solver.cpp:218] Iteration 61500 (3.74873 iter/s, 26.6757s/100 iters), loss = 6.67842
I0510 21:04:13.319598 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.67842 (* 1 = 6.67842 loss)
I0510 21:04:13.319612 39557 sgd_solver.cpp:105] Iteration 61500, lr = 0.1
I0510 21:04:40.050814 39557 solver.cpp:218] Iteration 61600 (3.74097 iter/s, 26.7311s/100 iters), loss = 6.68107
I0510 21:04:40.051030 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.68107 (* 1 = 6.68107 loss)
I0510 21:04:40.051048 39557 sgd_solver.cpp:105] Iteration 61600, lr = 0.1
I0510 21:05:06.785157 39557 solver.cpp:218] Iteration 61700 (3.74056 iter/s, 26.734s/100 iters), loss = 7.59583
I0510 21:05:06.785224 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.59583 (* 1 = 7.59583 loss)
I0510 21:05:06.785238 39557 sgd_solver.cpp:105] Iteration 61700, lr = 0.1
I0510 21:05:33.482525 39557 solver.cpp:218] Iteration 61800 (3.74572 iter/s, 26.6972s/100 iters), loss = 7.25828
I0510 21:05:33.482703 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.25828 (* 1 = 7.25828 loss)
I0510 21:05:33.482722 39557 sgd_solver.cpp:105] Iteration 61800, lr = 0.1
I0510 21:06:00.221304 39557 solver.cpp:218] Iteration 61900 (3.73993 iter/s, 26.7385s/100 iters), loss = 6.71234
I0510 21:06:00.221370 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.71234 (* 1 = 6.71234 loss)
I0510 21:06:00.221385 39557 sgd_solver.cpp:105] Iteration 61900, lr = 0.1
I0510 21:06:26.980401 39557 solver.cpp:218] Iteration 62000 (3.73708 iter/s, 26.7589s/100 iters), loss = 7.01288
I0510 21:06:26.980617 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.01288 (* 1 = 7.01288 loss)
I0510 21:06:26.980634 39557 sgd_solver.cpp:105] Iteration 62000, lr = 0.1
I0510 21:06:53.708094 39557 solver.cpp:218] Iteration 62100 (3.74149 iter/s, 26.7273s/100 iters), loss = 6.76501
I0510 21:06:53.708154 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.76501 (* 1 = 6.76501 loss)
I0510 21:06:53.708169 39557 sgd_solver.cpp:105] Iteration 62100, lr = 0.1
I0510 21:07:20.456837 39557 solver.cpp:218] Iteration 62200 (3.73852 iter/s, 26.7485s/100 iters), loss = 6.70587
I0510 21:07:20.457010 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.70587 (* 1 = 6.70587 loss)
I0510 21:07:20.457026 39557 sgd_solver.cpp:105] Iteration 62200, lr = 0.1
I0510 21:07:47.126174 39557 solver.cpp:218] Iteration 62300 (3.74967 iter/s, 26.669s/100 iters), loss = 6.47592
I0510 21:07:47.126236 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.47592 (* 1 = 6.47592 loss)
I0510 21:07:47.126253 39557 sgd_solver.cpp:105] Iteration 62300, lr = 0.1
I0510 21:08:13.743762 39557 solver.cpp:218] Iteration 62400 (3.75694 iter/s, 26.6174s/100 iters), loss = 6.19061
I0510 21:08:13.743932 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.19061 (* 1 = 6.19061 loss)
I0510 21:08:13.743949 39557 sgd_solver.cpp:105] Iteration 62400, lr = 0.1
I0510 21:08:40.391481 39557 solver.cpp:218] Iteration 62500 (3.75271 iter/s, 26.6474s/100 iters), loss = 6.51329
I0510 21:08:40.391541 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.51329 (* 1 = 6.51329 loss)
I0510 21:08:40.391554 39557 sgd_solver.cpp:105] Iteration 62500, lr = 0.1
I0510 21:09:07.001283 39557 solver.cpp:218] Iteration 62600 (3.75804 iter/s, 26.6096s/100 iters), loss = 7.16359
I0510 21:09:07.001447 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.16359 (* 1 = 7.16359 loss)
I0510 21:09:07.001461 39557 sgd_solver.cpp:105] Iteration 62600, lr = 0.1
I0510 21:09:33.597661 39557 solver.cpp:218] Iteration 62700 (3.75995 iter/s, 26.5961s/100 iters), loss = 7.02761
I0510 21:09:33.597720 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.02761 (* 1 = 7.02761 loss)
I0510 21:09:33.597733 39557 sgd_solver.cpp:105] Iteration 62700, lr = 0.1
I0510 21:10:00.192505 39557 solver.cpp:218] Iteration 62800 (3.76016 iter/s, 26.5946s/100 iters), loss = 7.37198
I0510 21:10:00.192674 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.37198 (* 1 = 7.37198 loss)
I0510 21:10:00.192687 39557 sgd_solver.cpp:105] Iteration 62800, lr = 0.1
I0510 21:10:26.916298 39557 solver.cpp:218] Iteration 62900 (3.74203 iter/s, 26.7235s/100 iters), loss = 6.93264
I0510 21:10:26.916364 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.93264 (* 1 = 6.93264 loss)
I0510 21:10:26.916379 39557 sgd_solver.cpp:105] Iteration 62900, lr = 0.1
I0510 21:10:53.583961 39557 solver.cpp:218] Iteration 63000 (3.74989 iter/s, 26.6675s/100 iters), loss = 6.50131
I0510 21:10:53.584154 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.50131 (* 1 = 6.50131 loss)
I0510 21:10:53.584170 39557 sgd_solver.cpp:105] Iteration 63000, lr = 0.1
I0510 21:11:20.253594 39557 solver.cpp:218] Iteration 63100 (3.74963 iter/s, 26.6693s/100 iters), loss = 6.44418
I0510 21:11:20.253655 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.44418 (* 1 = 6.44418 loss)
I0510 21:11:20.253667 39557 sgd_solver.cpp:105] Iteration 63100, lr = 0.1
I0510 21:11:46.833874 39557 solver.cpp:218] Iteration 63200 (3.76222 iter/s, 26.5801s/100 iters), loss = 7.27083
I0510 21:11:46.834033 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.27083 (* 1 = 7.27083 loss)
I0510 21:11:46.834048 39557 sgd_solver.cpp:105] Iteration 63200, lr = 0.1
I0510 21:12:13.383469 39557 solver.cpp:218] Iteration 63300 (3.76658 iter/s, 26.5493s/100 iters), loss = 7.44667
I0510 21:12:13.383525 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.44667 (* 1 = 7.44667 loss)
I0510 21:12:13.383539 39557 sgd_solver.cpp:105] Iteration 63300, lr = 0.1
I0510 21:12:39.965628 39557 solver.cpp:218] Iteration 63400 (3.76195 iter/s, 26.582s/100 iters), loss = 6.24751
I0510 21:12:39.965806 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.24751 (* 1 = 6.24751 loss)
I0510 21:12:39.965817 39557 sgd_solver.cpp:105] Iteration 63400, lr = 0.1
I0510 21:13:06.515908 39557 solver.cpp:218] Iteration 63500 (3.76648 iter/s, 26.55s/100 iters), loss = 7.09334
I0510 21:13:06.515964 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.09334 (* 1 = 7.09334 loss)
I0510 21:13:06.515975 39557 sgd_solver.cpp:105] Iteration 63500, lr = 0.1
I0510 21:13:33.092104 39557 solver.cpp:218] Iteration 63600 (3.76279 iter/s, 26.576s/100 iters), loss = 6.64068
I0510 21:13:33.092259 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.64068 (* 1 = 6.64068 loss)
I0510 21:13:33.092272 39557 sgd_solver.cpp:105] Iteration 63600, lr = 0.1
I0510 21:13:59.640215 39557 solver.cpp:218] Iteration 63700 (3.76679 iter/s, 26.5478s/100 iters), loss = 7.58895
I0510 21:13:59.640272 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.58895 (* 1 = 7.58895 loss)
I0510 21:13:59.640287 39557 sgd_solver.cpp:105] Iteration 63700, lr = 0.1
I0510 21:14:26.261013 39557 solver.cpp:218] Iteration 63800 (3.75649 iter/s, 26.6206s/100 iters), loss = 7.30552
I0510 21:14:26.261176 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.30552 (* 1 = 7.30552 loss)
I0510 21:14:26.261188 39557 sgd_solver.cpp:105] Iteration 63800, lr = 0.1
I0510 21:14:52.856943 39557 solver.cpp:218] Iteration 63900 (3.76002 iter/s, 26.5956s/100 iters), loss = 7.07115
I0510 21:14:52.856997 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.07115 (* 1 = 7.07115 loss)
I0510 21:14:52.857012 39557 sgd_solver.cpp:105] Iteration 63900, lr = 0.1
I0510 21:15:19.426066 39557 solver.cpp:218] Iteration 64000 (3.7638 iter/s, 26.5689s/100 iters), loss = 6.84622
I0510 21:15:19.426235 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.84622 (* 1 = 6.84622 loss)
I0510 21:15:19.426247 39557 sgd_solver.cpp:105] Iteration 64000, lr = 0.1
I0510 21:15:46.054241 39557 solver.cpp:218] Iteration 64100 (3.75546 iter/s, 26.6279s/100 iters), loss = 6.34419
I0510 21:15:46.054308 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.34419 (* 1 = 6.34419 loss)
I0510 21:15:46.054322 39557 sgd_solver.cpp:105] Iteration 64100, lr = 0.1
I0510 21:16:12.800017 39557 solver.cpp:218] Iteration 64200 (3.73894 iter/s, 26.7456s/100 iters), loss = 7.00687
I0510 21:16:12.800246 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.00687 (* 1 = 7.00687 loss)
I0510 21:16:12.800264 39557 sgd_solver.cpp:105] Iteration 64200, lr = 0.1
I0510 21:16:39.415778 39557 solver.cpp:218] Iteration 64300 (3.75722 iter/s, 26.6154s/100 iters), loss = 6.98862
I0510 21:16:39.415843 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.98862 (* 1 = 6.98862 loss)
I0510 21:16:39.415855 39557 sgd_solver.cpp:105] Iteration 64300, lr = 0.1
I0510 21:17:06.041802 39557 solver.cpp:218] Iteration 64400 (3.75575 iter/s, 26.6258s/100 iters), loss = 6.74965
I0510 21:17:06.041973 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.74965 (* 1 = 6.74965 loss)
I0510 21:17:06.041990 39557 sgd_solver.cpp:105] Iteration 64400, lr = 0.1
I0510 21:17:32.664952 39557 solver.cpp:218] Iteration 64500 (3.75617 iter/s, 26.6228s/100 iters), loss = 6.42353
I0510 21:17:32.665012 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.42353 (* 1 = 6.42353 loss)
I0510 21:17:32.665024 39557 sgd_solver.cpp:105] Iteration 64500, lr = 0.1
I0510 21:17:59.290830 39557 solver.cpp:218] Iteration 64600 (3.75577 iter/s, 26.6257s/100 iters), loss = 5.99017
I0510 21:17:59.291015 39557 solver.cpp:237]     Train net output #0: softmax_loss = 5.99017 (* 1 = 5.99017 loss)
I0510 21:17:59.291033 39557 sgd_solver.cpp:105] Iteration 64600, lr = 0.1
I0510 21:18:25.911085 39557 solver.cpp:218] Iteration 64700 (3.75658 iter/s, 26.62s/100 iters), loss = 6.8267
I0510 21:18:25.911139 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.8267 (* 1 = 6.8267 loss)
I0510 21:18:25.911150 39557 sgd_solver.cpp:105] Iteration 64700, lr = 0.1
I0510 21:18:52.513021 39557 solver.cpp:218] Iteration 64800 (3.75914 iter/s, 26.6018s/100 iters), loss = 6.7594
I0510 21:18:52.513208 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.7594 (* 1 = 6.7594 loss)
I0510 21:18:52.513226 39557 sgd_solver.cpp:105] Iteration 64800, lr = 0.1
I0510 21:19:19.192579 39557 solver.cpp:218] Iteration 64900 (3.74822 iter/s, 26.6793s/100 iters), loss = 6.82861
I0510 21:19:19.192646 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.82861 (* 1 = 6.82861 loss)
I0510 21:19:19.192662 39557 sgd_solver.cpp:105] Iteration 64900, lr = 0.1
I0510 21:19:45.627475 39557 solver.cpp:447] Snapshotting to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_65000.caffemodel
I0510 21:19:45.666451 39557 sgd_solver.cpp:273] Snapshotting solver state to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_65000.solverstate
I0510 21:19:45.947407 39557 solver.cpp:218] Iteration 65000 (3.73766 iter/s, 26.7547s/100 iters), loss = 6.27968
I0510 21:19:45.947465 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.27968 (* 1 = 6.27968 loss)
I0510 21:19:45.947479 39557 sgd_solver.cpp:105] Iteration 65000, lr = 0.1
I0510 21:20:12.598172 39557 solver.cpp:218] Iteration 65100 (3.75226 iter/s, 26.6506s/100 iters), loss = 6.1852
I0510 21:20:12.598238 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.1852 (* 1 = 6.1852 loss)
I0510 21:20:12.598253 39557 sgd_solver.cpp:105] Iteration 65100, lr = 0.1
I0510 21:20:39.296914 39557 solver.cpp:218] Iteration 65200 (3.74551 iter/s, 26.6986s/100 iters), loss = 7.10906
I0510 21:20:39.297083 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.10906 (* 1 = 7.10906 loss)
I0510 21:20:39.297099 39557 sgd_solver.cpp:105] Iteration 65200, lr = 0.1
I0510 21:21:05.935091 39557 solver.cpp:218] Iteration 65300 (3.75405 iter/s, 26.6379s/100 iters), loss = 6.90007
I0510 21:21:05.935154 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.90007 (* 1 = 6.90007 loss)
I0510 21:21:05.935169 39557 sgd_solver.cpp:105] Iteration 65300, lr = 0.1
I0510 21:21:32.650648 39557 solver.cpp:218] Iteration 65400 (3.74316 iter/s, 26.7154s/100 iters), loss = 6.39546
I0510 21:21:32.650830 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.39546 (* 1 = 6.39546 loss)
I0510 21:21:32.650847 39557 sgd_solver.cpp:105] Iteration 65400, lr = 0.1
I0510 21:21:59.308605 39557 solver.cpp:218] Iteration 65500 (3.75126 iter/s, 26.6577s/100 iters), loss = 6.48059
I0510 21:21:59.308667 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.48059 (* 1 = 6.48059 loss)
I0510 21:21:59.308681 39557 sgd_solver.cpp:105] Iteration 65500, lr = 0.1
I0510 21:22:26.037982 39557 solver.cpp:218] Iteration 65600 (3.74122 iter/s, 26.7292s/100 iters), loss = 6.16173
I0510 21:22:26.038185 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.16173 (* 1 = 6.16173 loss)
I0510 21:22:26.038203 39557 sgd_solver.cpp:105] Iteration 65600, lr = 0.1
I0510 21:22:52.731998 39557 solver.cpp:218] Iteration 65700 (3.7462 iter/s, 26.6937s/100 iters), loss = 6.14151
I0510 21:22:52.732059 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.14151 (* 1 = 6.14151 loss)
I0510 21:22:52.732071 39557 sgd_solver.cpp:105] Iteration 65700, lr = 0.1
I0510 21:23:19.419601 39557 solver.cpp:218] Iteration 65800 (3.74708 iter/s, 26.6875s/100 iters), loss = 7.28386
I0510 21:23:19.419777 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.28386 (* 1 = 7.28386 loss)
I0510 21:23:19.419793 39557 sgd_solver.cpp:105] Iteration 65800, lr = 0.1
I0510 21:23:46.066260 39557 solver.cpp:218] Iteration 65900 (3.75285 iter/s, 26.6464s/100 iters), loss = 6.78244
I0510 21:23:46.066318 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.78244 (* 1 = 6.78244 loss)
I0510 21:23:46.066331 39557 sgd_solver.cpp:105] Iteration 65900, lr = 0.1
I0510 21:24:12.884753 39557 solver.cpp:218] Iteration 66000 (3.72879 iter/s, 26.8183s/100 iters), loss = 7.28353
I0510 21:24:12.884961 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.28353 (* 1 = 7.28353 loss)
I0510 21:24:12.884976 39557 sgd_solver.cpp:105] Iteration 66000, lr = 0.1
I0510 21:24:39.589396 39557 solver.cpp:218] Iteration 66100 (3.74471 iter/s, 26.7043s/100 iters), loss = 6.29452
I0510 21:24:39.589462 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.29452 (* 1 = 6.29452 loss)
I0510 21:24:39.589476 39557 sgd_solver.cpp:105] Iteration 66100, lr = 0.1
I0510 21:25:06.260177 39557 solver.cpp:218] Iteration 66200 (3.74945 iter/s, 26.6706s/100 iters), loss = 6.64746
I0510 21:25:06.260361 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.64746 (* 1 = 6.64746 loss)
I0510 21:25:06.260390 39557 sgd_solver.cpp:105] Iteration 66200, lr = 0.1
I0510 21:25:32.959002 39557 solver.cpp:218] Iteration 66300 (3.74552 iter/s, 26.6985s/100 iters), loss = 6.52711
I0510 21:25:32.959069 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.52711 (* 1 = 6.52711 loss)
I0510 21:25:32.959084 39557 sgd_solver.cpp:105] Iteration 66300, lr = 0.1
I0510 21:25:59.609766 39557 solver.cpp:218] Iteration 66400 (3.75226 iter/s, 26.6506s/100 iters), loss = 6.88943
I0510 21:25:59.609949 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.88943 (* 1 = 6.88943 loss)
I0510 21:25:59.609978 39557 sgd_solver.cpp:105] Iteration 66400, lr = 0.1
I0510 21:26:26.277500 39557 solver.cpp:218] Iteration 66500 (3.74989 iter/s, 26.6674s/100 iters), loss = 7.20212
I0510 21:26:26.277565 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.20212 (* 1 = 7.20212 loss)
I0510 21:26:26.277580 39557 sgd_solver.cpp:105] Iteration 66500, lr = 0.1
I0510 21:26:52.973194 39557 solver.cpp:218] Iteration 66600 (3.74595 iter/s, 26.6955s/100 iters), loss = 7.05975
I0510 21:26:52.973371 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.05975 (* 1 = 7.05975 loss)
I0510 21:26:52.973387 39557 sgd_solver.cpp:105] Iteration 66600, lr = 0.1
I0510 21:27:19.658664 39557 solver.cpp:218] Iteration 66700 (3.7474 iter/s, 26.6852s/100 iters), loss = 7.18031
I0510 21:27:19.658728 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.18031 (* 1 = 7.18031 loss)
I0510 21:27:19.658744 39557 sgd_solver.cpp:105] Iteration 66700, lr = 0.1
I0510 21:27:46.335381 39557 solver.cpp:218] Iteration 66800 (3.74861 iter/s, 26.6766s/100 iters), loss = 6.2208
I0510 21:27:46.335597 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.2208 (* 1 = 6.2208 loss)
I0510 21:27:46.335611 39557 sgd_solver.cpp:105] Iteration 66800, lr = 0.1
I0510 21:28:13.011356 39557 solver.cpp:218] Iteration 66900 (3.74874 iter/s, 26.6756s/100 iters), loss = 6.69366
I0510 21:28:13.011418 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.69366 (* 1 = 6.69366 loss)
I0510 21:28:13.011433 39557 sgd_solver.cpp:105] Iteration 66900, lr = 0.1
I0510 21:28:39.609804 39557 solver.cpp:218] Iteration 67000 (3.75964 iter/s, 26.5983s/100 iters), loss = 7.56729
I0510 21:28:39.609983 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.56729 (* 1 = 7.56729 loss)
I0510 21:28:39.609997 39557 sgd_solver.cpp:105] Iteration 67000, lr = 0.1
I0510 21:29:06.158277 39557 solver.cpp:218] Iteration 67100 (3.76673 iter/s, 26.5482s/100 iters), loss = 6.22548
I0510 21:29:06.158334 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.22548 (* 1 = 6.22548 loss)
I0510 21:29:06.158347 39557 sgd_solver.cpp:105] Iteration 67100, lr = 0.1
I0510 21:29:32.856950 39557 solver.cpp:218] Iteration 67200 (3.74553 iter/s, 26.6985s/100 iters), loss = 6.18775
I0510 21:29:32.857095 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.18775 (* 1 = 6.18775 loss)
I0510 21:29:32.857107 39557 sgd_solver.cpp:105] Iteration 67200, lr = 0.1
I0510 21:29:59.503154 39557 solver.cpp:218] Iteration 67300 (3.75292 iter/s, 26.646s/100 iters), loss = 7.30387
I0510 21:29:59.503213 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.30387 (* 1 = 7.30387 loss)
I0510 21:29:59.503227 39557 sgd_solver.cpp:105] Iteration 67300, lr = 0.1
I0510 21:30:26.169672 39557 solver.cpp:218] Iteration 67400 (3.75004 iter/s, 26.6664s/100 iters), loss = 6.73982
I0510 21:30:26.169817 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.73982 (* 1 = 6.73982 loss)
I0510 21:30:26.169831 39557 sgd_solver.cpp:105] Iteration 67400, lr = 0.1
I0510 21:30:52.819571 39557 solver.cpp:218] Iteration 67500 (3.75239 iter/s, 26.6496s/100 iters), loss = 7.28883
I0510 21:30:52.819627 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.28883 (* 1 = 7.28883 loss)
I0510 21:30:52.819640 39557 sgd_solver.cpp:105] Iteration 67500, lr = 0.1
I0510 21:31:19.496351 39557 solver.cpp:218] Iteration 67600 (3.7486 iter/s, 26.6766s/100 iters), loss = 7.01489
I0510 21:31:19.496489 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.01489 (* 1 = 7.01489 loss)
I0510 21:31:19.496501 39557 sgd_solver.cpp:105] Iteration 67600, lr = 0.1
I0510 21:31:46.181565 39557 solver.cpp:218] Iteration 67700 (3.74743 iter/s, 26.685s/100 iters), loss = 6.87673
I0510 21:31:46.181627 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.87673 (* 1 = 6.87673 loss)
I0510 21:31:46.181643 39557 sgd_solver.cpp:105] Iteration 67700, lr = 0.1
I0510 21:32:12.942944 39557 solver.cpp:218] Iteration 67800 (3.73675 iter/s, 26.7612s/100 iters), loss = 6.62454
I0510 21:32:12.943122 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.62454 (* 1 = 6.62454 loss)
I0510 21:32:12.943141 39557 sgd_solver.cpp:105] Iteration 67800, lr = 0.1
I0510 21:32:39.516193 39557 solver.cpp:218] Iteration 67900 (3.76322 iter/s, 26.573s/100 iters), loss = 7.50727
I0510 21:32:39.516259 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.50727 (* 1 = 7.50727 loss)
I0510 21:32:39.516273 39557 sgd_solver.cpp:105] Iteration 67900, lr = 0.1
I0510 21:33:06.221948 39557 solver.cpp:218] Iteration 68000 (3.74454 iter/s, 26.7056s/100 iters), loss = 7.24022
I0510 21:33:06.222131 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.24022 (* 1 = 7.24022 loss)
I0510 21:33:06.222147 39557 sgd_solver.cpp:105] Iteration 68000, lr = 0.1
I0510 21:33:32.867244 39557 solver.cpp:218] Iteration 68100 (3.75305 iter/s, 26.645s/100 iters), loss = 6.67242
I0510 21:33:32.867305 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.67242 (* 1 = 6.67242 loss)
I0510 21:33:32.867316 39557 sgd_solver.cpp:105] Iteration 68100, lr = 0.1
I0510 21:33:59.485520 39557 solver.cpp:218] Iteration 68200 (3.75684 iter/s, 26.6181s/100 iters), loss = 7.06036
I0510 21:33:59.485679 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.06036 (* 1 = 7.06036 loss)
I0510 21:33:59.485703 39557 sgd_solver.cpp:105] Iteration 68200, lr = 0.1
I0510 21:34:26.107298 39557 solver.cpp:218] Iteration 68300 (3.75636 iter/s, 26.6215s/100 iters), loss = 6.99283
I0510 21:34:26.107362 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.99283 (* 1 = 6.99283 loss)
I0510 21:34:26.107375 39557 sgd_solver.cpp:105] Iteration 68300, lr = 0.1
I0510 21:34:52.699193 39557 solver.cpp:218] Iteration 68400 (3.76057 iter/s, 26.5917s/100 iters), loss = 6.45604
I0510 21:34:52.699348 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.45604 (* 1 = 6.45604 loss)
I0510 21:34:52.699363 39557 sgd_solver.cpp:105] Iteration 68400, lr = 0.1
I0510 21:35:19.302541 39557 solver.cpp:218] Iteration 68500 (3.75896 iter/s, 26.6031s/100 iters), loss = 6.80282
I0510 21:35:19.302595 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.80282 (* 1 = 6.80282 loss)
I0510 21:35:19.302605 39557 sgd_solver.cpp:105] Iteration 68500, lr = 0.1
I0510 21:35:45.914433 39557 solver.cpp:218] Iteration 68600 (3.75774 iter/s, 26.6117s/100 iters), loss = 6.50036
I0510 21:35:45.914563 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.50036 (* 1 = 6.50036 loss)
I0510 21:35:45.914589 39557 sgd_solver.cpp:105] Iteration 68600, lr = 0.1
I0510 21:36:12.494632 39557 solver.cpp:218] Iteration 68700 (3.76223 iter/s, 26.58s/100 iters), loss = 7.23027
I0510 21:36:12.494688 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.23027 (* 1 = 7.23027 loss)
I0510 21:36:12.494698 39557 sgd_solver.cpp:105] Iteration 68700, lr = 0.1
I0510 21:36:39.151753 39557 solver.cpp:218] Iteration 68800 (3.75137 iter/s, 26.6569s/100 iters), loss = 7.01707
I0510 21:36:39.151881 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.01707 (* 1 = 7.01707 loss)
I0510 21:36:39.151907 39557 sgd_solver.cpp:105] Iteration 68800, lr = 0.1
I0510 21:37:05.831821 39557 solver.cpp:218] Iteration 68900 (3.74815 iter/s, 26.6798s/100 iters), loss = 6.25464
I0510 21:37:05.831884 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.25464 (* 1 = 6.25464 loss)
I0510 21:37:05.831897 39557 sgd_solver.cpp:105] Iteration 68900, lr = 0.1
I0510 21:37:32.504349 39557 solver.cpp:218] Iteration 69000 (3.7492 iter/s, 26.6723s/100 iters), loss = 6.1797
I0510 21:37:32.504489 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.1797 (* 1 = 6.1797 loss)
I0510 21:37:32.504504 39557 sgd_solver.cpp:105] Iteration 69000, lr = 0.1
I0510 21:37:59.217165 39557 solver.cpp:218] Iteration 69100 (3.74356 iter/s, 26.7126s/100 iters), loss = 7.65098
I0510 21:37:59.217236 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.65098 (* 1 = 7.65098 loss)
I0510 21:37:59.217252 39557 sgd_solver.cpp:105] Iteration 69100, lr = 0.1
I0510 21:38:25.957126 39557 solver.cpp:218] Iteration 69200 (3.73975 iter/s, 26.7398s/100 iters), loss = 6.45991
I0510 21:38:25.957300 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.45991 (* 1 = 6.45991 loss)
I0510 21:38:25.957314 39557 sgd_solver.cpp:105] Iteration 69200, lr = 0.1
I0510 21:38:52.578723 39557 solver.cpp:218] Iteration 69300 (3.75639 iter/s, 26.6213s/100 iters), loss = 6.45758
I0510 21:38:52.578785 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.45758 (* 1 = 6.45758 loss)
I0510 21:38:52.578799 39557 sgd_solver.cpp:105] Iteration 69300, lr = 0.1
I0510 21:39:19.264266 39557 solver.cpp:218] Iteration 69400 (3.74737 iter/s, 26.6854s/100 iters), loss = 6.87277
I0510 21:39:19.264438 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.87277 (* 1 = 6.87277 loss)
I0510 21:39:19.264453 39557 sgd_solver.cpp:105] Iteration 69400, lr = 0.1
I0510 21:39:45.938530 39557 solver.cpp:218] Iteration 69500 (3.74897 iter/s, 26.674s/100 iters), loss = 6.68349
I0510 21:39:45.938594 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.68349 (* 1 = 6.68349 loss)
I0510 21:39:45.938607 39557 sgd_solver.cpp:105] Iteration 69500, lr = 0.1
I0510 21:40:12.683702 39557 solver.cpp:218] Iteration 69600 (3.73902 iter/s, 26.745s/100 iters), loss = 6.50063
I0510 21:40:12.683902 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.50063 (* 1 = 6.50063 loss)
I0510 21:40:12.683919 39557 sgd_solver.cpp:105] Iteration 69600, lr = 0.1
I0510 21:40:39.449571 39557 solver.cpp:218] Iteration 69700 (3.73615 iter/s, 26.7656s/100 iters), loss = 7.22093
I0510 21:40:39.449637 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.22093 (* 1 = 7.22093 loss)
I0510 21:40:39.449651 39557 sgd_solver.cpp:105] Iteration 69700, lr = 0.1
I0510 21:41:06.176728 39557 solver.cpp:218] Iteration 69800 (3.74154 iter/s, 26.727s/100 iters), loss = 7.33295
I0510 21:41:06.176892 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.33295 (* 1 = 7.33295 loss)
I0510 21:41:06.176908 39557 sgd_solver.cpp:105] Iteration 69800, lr = 0.1
I0510 21:41:32.850193 39557 solver.cpp:218] Iteration 69900 (3.74908 iter/s, 26.6732s/100 iters), loss = 6.3018
I0510 21:41:32.850255 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.3018 (* 1 = 6.3018 loss)
I0510 21:41:32.850270 39557 sgd_solver.cpp:105] Iteration 69900, lr = 0.1
I0510 21:41:59.351801 39557 solver.cpp:447] Snapshotting to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_70000.caffemodel
I0510 21:41:59.390888 39557 sgd_solver.cpp:273] Snapshotting solver state to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_70000.solverstate
I0510 21:41:59.680002 39557 solver.cpp:218] Iteration 70000 (3.72722 iter/s, 26.8296s/100 iters), loss = 7.49806
I0510 21:41:59.680063 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.49806 (* 1 = 7.49806 loss)
I0510 21:41:59.680076 39557 sgd_solver.cpp:105] Iteration 70000, lr = 0.1
I0510 21:42:26.417775 39557 solver.cpp:218] Iteration 70100 (3.74005 iter/s, 26.7376s/100 iters), loss = 6.46345
I0510 21:42:26.417835 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.46345 (* 1 = 6.46345 loss)
I0510 21:42:26.417845 39557 sgd_solver.cpp:105] Iteration 70100, lr = 0.1
I0510 21:42:53.087839 39557 solver.cpp:218] Iteration 70200 (3.74955 iter/s, 26.6699s/100 iters), loss = 7.20333
I0510 21:42:53.087990 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.20333 (* 1 = 7.20333 loss)
I0510 21:42:53.088014 39557 sgd_solver.cpp:105] Iteration 70200, lr = 0.1
I0510 21:43:19.734608 39557 solver.cpp:218] Iteration 70300 (3.75284 iter/s, 26.6465s/100 iters), loss = 6.81228
I0510 21:43:19.734673 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.81228 (* 1 = 6.81228 loss)
I0510 21:43:19.734686 39557 sgd_solver.cpp:105] Iteration 70300, lr = 0.1
I0510 21:43:46.401896 39557 solver.cpp:218] Iteration 70400 (3.74994 iter/s, 26.6671s/100 iters), loss = 6.93153
I0510 21:43:46.402042 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.93153 (* 1 = 6.93153 loss)
I0510 21:43:46.402068 39557 sgd_solver.cpp:105] Iteration 70400, lr = 0.1
I0510 21:44:13.079461 39557 solver.cpp:218] Iteration 70500 (3.74851 iter/s, 26.6773s/100 iters), loss = 6.94246
I0510 21:44:13.079529 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.94246 (* 1 = 6.94246 loss)
I0510 21:44:13.079543 39557 sgd_solver.cpp:105] Iteration 70500, lr = 0.1
I0510 21:44:39.836141 39557 solver.cpp:218] Iteration 70600 (3.73741 iter/s, 26.7565s/100 iters), loss = 7.04976
I0510 21:44:39.836305 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.04976 (* 1 = 7.04976 loss)
I0510 21:44:39.836323 39557 sgd_solver.cpp:105] Iteration 70600, lr = 0.1
I0510 21:45:06.603997 39557 solver.cpp:218] Iteration 70700 (3.73586 iter/s, 26.7676s/100 iters), loss = 6.35386
I0510 21:45:06.604068 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.35386 (* 1 = 6.35386 loss)
I0510 21:45:06.604082 39557 sgd_solver.cpp:105] Iteration 70700, lr = 0.1
I0510 21:45:33.317762 39557 solver.cpp:218] Iteration 70800 (3.74341 iter/s, 26.7136s/100 iters), loss = 6.54488
I0510 21:45:33.317944 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.54488 (* 1 = 6.54488 loss)
I0510 21:45:33.317961 39557 sgd_solver.cpp:105] Iteration 70800, lr = 0.1
I0510 21:45:59.977432 39557 solver.cpp:218] Iteration 70900 (3.75103 iter/s, 26.6594s/100 iters), loss = 7.19041
I0510 21:45:59.977493 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.19041 (* 1 = 7.19041 loss)
I0510 21:45:59.977505 39557 sgd_solver.cpp:105] Iteration 70900, lr = 0.1
I0510 21:46:26.725543 39557 solver.cpp:218] Iteration 71000 (3.73861 iter/s, 26.7479s/100 iters), loss = 6.38225
I0510 21:46:26.725705 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.38225 (* 1 = 6.38225 loss)
I0510 21:46:26.725723 39557 sgd_solver.cpp:105] Iteration 71000, lr = 0.1
I0510 21:46:53.416158 39557 solver.cpp:218] Iteration 71100 (3.74667 iter/s, 26.6903s/100 iters), loss = 7.11174
I0510 21:46:53.416224 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.11174 (* 1 = 7.11174 loss)
I0510 21:46:53.416235 39557 sgd_solver.cpp:105] Iteration 71100, lr = 0.1
I0510 21:47:20.108023 39557 solver.cpp:218] Iteration 71200 (3.74649 iter/s, 26.6917s/100 iters), loss = 7.00893
I0510 21:47:20.108180 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.00893 (* 1 = 7.00893 loss)
I0510 21:47:20.108197 39557 sgd_solver.cpp:105] Iteration 71200, lr = 0.1
I0510 21:47:46.824137 39557 solver.cpp:218] Iteration 71300 (3.7431 iter/s, 26.7158s/100 iters), loss = 6.15198
I0510 21:47:46.824198 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.15198 (* 1 = 6.15198 loss)
I0510 21:47:46.824211 39557 sgd_solver.cpp:105] Iteration 71300, lr = 0.1
I0510 21:48:13.548326 39557 solver.cpp:218] Iteration 71400 (3.74195 iter/s, 26.724s/100 iters), loss = 6.72574
I0510 21:48:13.548532 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.72574 (* 1 = 6.72574 loss)
I0510 21:48:13.548583 39557 sgd_solver.cpp:105] Iteration 71400, lr = 0.1
I0510 21:48:40.279014 39557 solver.cpp:218] Iteration 71500 (3.74106 iter/s, 26.7304s/100 iters), loss = 7.08466
I0510 21:48:40.279067 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.08466 (* 1 = 7.08466 loss)
I0510 21:48:40.279081 39557 sgd_solver.cpp:105] Iteration 71500, lr = 0.1
I0510 21:49:06.923188 39557 solver.cpp:218] Iteration 71600 (3.75319 iter/s, 26.644s/100 iters), loss = 6.70726
I0510 21:49:06.923358 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.70726 (* 1 = 6.70726 loss)
I0510 21:49:06.923374 39557 sgd_solver.cpp:105] Iteration 71600, lr = 0.1
I0510 21:49:33.503911 39557 solver.cpp:218] Iteration 71700 (3.76217 iter/s, 26.5804s/100 iters), loss = 7.42048
I0510 21:49:33.503964 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.42048 (* 1 = 7.42048 loss)
I0510 21:49:33.503976 39557 sgd_solver.cpp:105] Iteration 71700, lr = 0.1
I0510 21:50:00.221501 39557 solver.cpp:218] Iteration 71800 (3.74288 iter/s, 26.7174s/100 iters), loss = 6.57309
I0510 21:50:00.221691 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.57309 (* 1 = 6.57309 loss)
I0510 21:50:00.221709 39557 sgd_solver.cpp:105] Iteration 71800, lr = 0.1
I0510 21:50:26.875342 39557 solver.cpp:218] Iteration 71900 (3.75185 iter/s, 26.6535s/100 iters), loss = 6.77967
I0510 21:50:26.875391 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.77967 (* 1 = 6.77967 loss)
I0510 21:50:26.875401 39557 sgd_solver.cpp:105] Iteration 71900, lr = 0.1
I0510 21:50:53.446447 39557 solver.cpp:218] Iteration 72000 (3.76351 iter/s, 26.5709s/100 iters), loss = 6.36679
I0510 21:50:53.446583 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.36679 (* 1 = 6.36679 loss)
I0510 21:50:53.446597 39557 sgd_solver.cpp:105] Iteration 72000, lr = 0.1
I0510 21:51:20.078671 39557 solver.cpp:218] Iteration 72100 (3.75489 iter/s, 26.632s/100 iters), loss = 6.86963
I0510 21:51:20.078717 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.86963 (* 1 = 6.86963 loss)
I0510 21:51:20.078727 39557 sgd_solver.cpp:105] Iteration 72100, lr = 0.1
I0510 21:51:46.702981 39557 solver.cpp:218] Iteration 72200 (3.75599 iter/s, 26.6241s/100 iters), loss = 6.87353
I0510 21:51:46.703158 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.87353 (* 1 = 6.87353 loss)
I0510 21:51:46.703174 39557 sgd_solver.cpp:105] Iteration 72200, lr = 0.1
I0510 21:52:13.306545 39557 solver.cpp:218] Iteration 72300 (3.75894 iter/s, 26.6033s/100 iters), loss = 6.20436
I0510 21:52:13.306596 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.20436 (* 1 = 6.20436 loss)
I0510 21:52:13.306607 39557 sgd_solver.cpp:105] Iteration 72300, lr = 0.1
I0510 21:52:40.008929 39557 solver.cpp:218] Iteration 72400 (3.74501 iter/s, 26.7022s/100 iters), loss = 6.79608
I0510 21:52:40.009119 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.79608 (* 1 = 6.79608 loss)
I0510 21:52:40.009135 39557 sgd_solver.cpp:105] Iteration 72400, lr = 0.1
I0510 21:53:06.656363 39557 solver.cpp:218] Iteration 72500 (3.75275 iter/s, 26.6472s/100 iters), loss = 7.25016
I0510 21:53:06.656427 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.25016 (* 1 = 7.25016 loss)
I0510 21:53:06.656441 39557 sgd_solver.cpp:105] Iteration 72500, lr = 0.1
I0510 21:53:33.289819 39557 solver.cpp:218] Iteration 72600 (3.7547 iter/s, 26.6333s/100 iters), loss = 6.27632
I0510 21:53:33.289997 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.27632 (* 1 = 6.27632 loss)
I0510 21:53:33.290011 39557 sgd_solver.cpp:105] Iteration 72600, lr = 0.1
I0510 21:53:59.974975 39557 solver.cpp:218] Iteration 72700 (3.74744 iter/s, 26.6849s/100 iters), loss = 6.48643
I0510 21:53:59.975030 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.48643 (* 1 = 6.48643 loss)
I0510 21:53:59.975044 39557 sgd_solver.cpp:105] Iteration 72700, lr = 0.1
I0510 21:54:26.707281 39557 solver.cpp:218] Iteration 72800 (3.74081 iter/s, 26.7322s/100 iters), loss = 6.19106
I0510 21:54:26.707489 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.19106 (* 1 = 6.19106 loss)
I0510 21:54:26.707505 39557 sgd_solver.cpp:105] Iteration 72800, lr = 0.1
I0510 21:54:53.554041 39557 solver.cpp:218] Iteration 72900 (3.72489 iter/s, 26.8465s/100 iters), loss = 6.7853
I0510 21:54:53.554107 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.7853 (* 1 = 6.7853 loss)
I0510 21:54:53.554122 39557 sgd_solver.cpp:105] Iteration 72900, lr = 0.1
I0510 21:55:20.262347 39557 solver.cpp:218] Iteration 73000 (3.74418 iter/s, 26.7081s/100 iters), loss = 6.92081
I0510 21:55:20.262511 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.92081 (* 1 = 6.92081 loss)
I0510 21:55:20.262528 39557 sgd_solver.cpp:105] Iteration 73000, lr = 0.1
I0510 21:55:46.940731 39557 solver.cpp:218] Iteration 73100 (3.74839 iter/s, 26.6781s/100 iters), loss = 6.43707
I0510 21:55:46.940796 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.43707 (* 1 = 6.43707 loss)
I0510 21:55:46.940811 39557 sgd_solver.cpp:105] Iteration 73100, lr = 0.1
I0510 21:56:13.621291 39557 solver.cpp:218] Iteration 73200 (3.74807 iter/s, 26.6804s/100 iters), loss = 6.28413
I0510 21:56:13.621449 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.28413 (* 1 = 6.28413 loss)
I0510 21:56:13.621466 39557 sgd_solver.cpp:105] Iteration 73200, lr = 0.1
I0510 21:56:40.323851 39557 solver.cpp:218] Iteration 73300 (3.745 iter/s, 26.7023s/100 iters), loss = 6.6128
I0510 21:56:40.323913 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.6128 (* 1 = 6.6128 loss)
I0510 21:56:40.323930 39557 sgd_solver.cpp:105] Iteration 73300, lr = 0.1
I0510 21:57:07.004597 39557 solver.cpp:218] Iteration 73400 (3.74804 iter/s, 26.6806s/100 iters), loss = 6.74713
I0510 21:57:07.004747 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.74713 (* 1 = 6.74713 loss)
I0510 21:57:07.004766 39557 sgd_solver.cpp:105] Iteration 73400, lr = 0.1
I0510 21:57:33.857576 39557 solver.cpp:218] Iteration 73500 (3.72402 iter/s, 26.8527s/100 iters), loss = 6.14046
I0510 21:57:33.857637 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.14046 (* 1 = 6.14046 loss)
I0510 21:57:33.857650 39557 sgd_solver.cpp:105] Iteration 73500, lr = 0.1
I0510 21:58:00.574489 39557 solver.cpp:218] Iteration 73600 (3.74297 iter/s, 26.7167s/100 iters), loss = 6.17908
I0510 21:58:00.574689 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.17908 (* 1 = 6.17908 loss)
I0510 21:58:00.574707 39557 sgd_solver.cpp:105] Iteration 73600, lr = 0.1
I0510 21:58:27.310245 39557 solver.cpp:218] Iteration 73700 (3.74035 iter/s, 26.7354s/100 iters), loss = 6.71512
I0510 21:58:27.310305 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.71512 (* 1 = 6.71512 loss)
I0510 21:58:27.310319 39557 sgd_solver.cpp:105] Iteration 73700, lr = 0.1
I0510 21:58:54.043639 39557 solver.cpp:218] Iteration 73800 (3.74066 iter/s, 26.7332s/100 iters), loss = 6.75872
I0510 21:58:54.043812 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.75872 (* 1 = 6.75872 loss)
I0510 21:58:54.043828 39557 sgd_solver.cpp:105] Iteration 73800, lr = 0.1
I0510 21:59:20.702992 39557 solver.cpp:218] Iteration 73900 (3.75107 iter/s, 26.6591s/100 iters), loss = 6.72736
I0510 21:59:20.703058 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.72736 (* 1 = 6.72736 loss)
I0510 21:59:20.703071 39557 sgd_solver.cpp:105] Iteration 73900, lr = 0.1
I0510 21:59:47.338852 39557 solver.cpp:218] Iteration 74000 (3.75436 iter/s, 26.6357s/100 iters), loss = 6.3046
I0510 21:59:47.339007 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.3046 (* 1 = 6.3046 loss)
I0510 21:59:47.339021 39557 sgd_solver.cpp:105] Iteration 74000, lr = 0.1
I0510 22:00:13.987005 39557 solver.cpp:218] Iteration 74100 (3.75264 iter/s, 26.6479s/100 iters), loss = 7.42889
I0510 22:00:13.987062 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.42889 (* 1 = 7.42889 loss)
I0510 22:00:13.987076 39557 sgd_solver.cpp:105] Iteration 74100, lr = 0.1
I0510 22:00:40.676894 39557 solver.cpp:218] Iteration 74200 (3.74676 iter/s, 26.6897s/100 iters), loss = 6.3686
I0510 22:00:40.677052 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.3686 (* 1 = 6.3686 loss)
I0510 22:00:40.677064 39557 sgd_solver.cpp:105] Iteration 74200, lr = 0.1
I0510 22:01:07.302920 39557 solver.cpp:218] Iteration 74300 (3.75576 iter/s, 26.6258s/100 iters), loss = 7.153
I0510 22:01:07.302979 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.153 (* 1 = 7.153 loss)
I0510 22:01:07.302989 39557 sgd_solver.cpp:105] Iteration 74300, lr = 0.1
I0510 22:01:33.874214 39557 solver.cpp:218] Iteration 74400 (3.76348 iter/s, 26.5711s/100 iters), loss = 6.74659
I0510 22:01:33.874404 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.74659 (* 1 = 6.74659 loss)
I0510 22:01:33.874419 39557 sgd_solver.cpp:105] Iteration 74400, lr = 0.1
I0510 22:02:00.459741 39557 solver.cpp:218] Iteration 74500 (3.76149 iter/s, 26.5852s/100 iters), loss = 7.32467
I0510 22:02:00.459803 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.32467 (* 1 = 7.32467 loss)
I0510 22:02:00.459816 39557 sgd_solver.cpp:105] Iteration 74500, lr = 0.1
I0510 22:02:27.100214 39557 solver.cpp:218] Iteration 74600 (3.75371 iter/s, 26.6403s/100 iters), loss = 7.15007
I0510 22:02:27.100389 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.15007 (* 1 = 7.15007 loss)
I0510 22:02:27.100401 39557 sgd_solver.cpp:105] Iteration 74600, lr = 0.1
I0510 22:02:53.743974 39557 solver.cpp:218] Iteration 74700 (3.75326 iter/s, 26.6435s/100 iters), loss = 7.16418
I0510 22:02:53.744042 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.16418 (* 1 = 7.16418 loss)
I0510 22:02:53.744052 39557 sgd_solver.cpp:105] Iteration 74700, lr = 0.1
I0510 22:03:20.309917 39557 solver.cpp:218] Iteration 74800 (3.76424 iter/s, 26.5658s/100 iters), loss = 6.69077
I0510 22:03:20.310091 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.69077 (* 1 = 6.69077 loss)
I0510 22:03:20.310106 39557 sgd_solver.cpp:105] Iteration 74800, lr = 0.1
I0510 22:03:46.949893 39557 solver.cpp:218] Iteration 74900 (3.7538 iter/s, 26.6397s/100 iters), loss = 6.73984
I0510 22:03:46.949960 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.73984 (* 1 = 6.73984 loss)
I0510 22:03:46.949975 39557 sgd_solver.cpp:105] Iteration 74900, lr = 0.1
I0510 22:04:13.295599 39557 solver.cpp:447] Snapshotting to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_75000.caffemodel
I0510 22:04:13.334748 39557 sgd_solver.cpp:273] Snapshotting solver state to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_75000.solverstate
I0510 22:04:13.615149 39557 solver.cpp:218] Iteration 75000 (3.75022 iter/s, 26.6651s/100 iters), loss = 6.13265
I0510 22:04:13.615206 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.13265 (* 1 = 6.13265 loss)
I0510 22:04:13.615216 39557 sgd_solver.cpp:105] Iteration 75000, lr = 0.1
I0510 22:04:40.194839 39557 solver.cpp:218] Iteration 75100 (3.7623 iter/s, 26.5795s/100 iters), loss = 6.05542
I0510 22:04:40.194895 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.05542 (* 1 = 6.05542 loss)
I0510 22:04:40.194907 39557 sgd_solver.cpp:105] Iteration 75100, lr = 0.1
I0510 22:05:06.787499 39557 solver.cpp:218] Iteration 75200 (3.76046 iter/s, 26.5925s/100 iters), loss = 6.00657
I0510 22:05:06.787685 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.00657 (* 1 = 6.00657 loss)
I0510 22:05:06.787700 39557 sgd_solver.cpp:105] Iteration 75200, lr = 0.1
I0510 22:05:33.515137 39557 solver.cpp:218] Iteration 75300 (3.74149 iter/s, 26.7273s/100 iters), loss = 7.58852
I0510 22:05:33.515198 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.58852 (* 1 = 7.58852 loss)
I0510 22:05:33.515210 39557 sgd_solver.cpp:105] Iteration 75300, lr = 0.1
I0510 22:06:00.300050 39557 solver.cpp:218] Iteration 75400 (3.73347 iter/s, 26.7847s/100 iters), loss = 5.81591
I0510 22:06:00.300210 39557 solver.cpp:237]     Train net output #0: softmax_loss = 5.81591 (* 1 = 5.81591 loss)
I0510 22:06:00.300225 39557 sgd_solver.cpp:105] Iteration 75400, lr = 0.1
I0510 22:06:27.000715 39557 solver.cpp:218] Iteration 75500 (3.74526 iter/s, 26.7004s/100 iters), loss = 6.28128
I0510 22:06:27.000797 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.28128 (* 1 = 6.28128 loss)
I0510 22:06:27.000809 39557 sgd_solver.cpp:105] Iteration 75500, lr = 0.1
I0510 22:06:53.709892 39557 solver.cpp:218] Iteration 75600 (3.74406 iter/s, 26.709s/100 iters), loss = 6.25691
I0510 22:06:53.710053 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.25691 (* 1 = 6.25691 loss)
I0510 22:06:53.710078 39557 sgd_solver.cpp:105] Iteration 75600, lr = 0.1
I0510 22:07:20.357720 39557 solver.cpp:218] Iteration 75700 (3.75269 iter/s, 26.6475s/100 iters), loss = 6.71192
I0510 22:07:20.357784 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.71192 (* 1 = 6.71192 loss)
I0510 22:07:20.357796 39557 sgd_solver.cpp:105] Iteration 75700, lr = 0.1
I0510 22:07:47.034159 39557 solver.cpp:218] Iteration 75800 (3.74865 iter/s, 26.6763s/100 iters), loss = 6.37124
I0510 22:07:47.034319 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.37124 (* 1 = 6.37124 loss)
I0510 22:07:47.034345 39557 sgd_solver.cpp:105] Iteration 75800, lr = 0.1
I0510 22:08:13.748103 39557 solver.cpp:218] Iteration 75900 (3.7434 iter/s, 26.7137s/100 iters), loss = 6.87735
I0510 22:08:13.748164 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.87735 (* 1 = 6.87735 loss)
I0510 22:08:13.748175 39557 sgd_solver.cpp:105] Iteration 75900, lr = 0.1
I0510 22:08:40.490737 39557 solver.cpp:218] Iteration 76000 (3.73937 iter/s, 26.7425s/100 iters), loss = 6.1921
I0510 22:08:40.490916 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.1921 (* 1 = 6.1921 loss)
I0510 22:08:40.490932 39557 sgd_solver.cpp:105] Iteration 76000, lr = 0.1
I0510 22:09:07.324899 39557 solver.cpp:218] Iteration 76100 (3.72663 iter/s, 26.8339s/100 iters), loss = 6.9859
I0510 22:09:07.324962 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.9859 (* 1 = 6.9859 loss)
I0510 22:09:07.324975 39557 sgd_solver.cpp:105] Iteration 76100, lr = 0.1
I0510 22:09:34.053887 39557 solver.cpp:218] Iteration 76200 (3.74128 iter/s, 26.7288s/100 iters), loss = 7.36896
I0510 22:09:34.054111 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.36896 (* 1 = 7.36896 loss)
I0510 22:09:34.054127 39557 sgd_solver.cpp:105] Iteration 76200, lr = 0.1
I0510 22:10:00.677263 39557 solver.cpp:218] Iteration 76300 (3.75615 iter/s, 26.623s/100 iters), loss = 7.65591
I0510 22:10:00.677325 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.65591 (* 1 = 7.65591 loss)
I0510 22:10:00.677337 39557 sgd_solver.cpp:105] Iteration 76300, lr = 0.1
I0510 22:10:27.311269 39557 solver.cpp:218] Iteration 76400 (3.75462 iter/s, 26.6338s/100 iters), loss = 6.7492
I0510 22:10:27.311430 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.7492 (* 1 = 6.7492 loss)
I0510 22:10:27.311445 39557 sgd_solver.cpp:105] Iteration 76400, lr = 0.1
I0510 22:10:53.907475 39557 solver.cpp:218] Iteration 76500 (3.75997 iter/s, 26.5959s/100 iters), loss = 7.32729
I0510 22:10:53.907536 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.32729 (* 1 = 7.32729 loss)
I0510 22:10:53.907546 39557 sgd_solver.cpp:105] Iteration 76500, lr = 0.1
I0510 22:11:20.534073 39557 solver.cpp:218] Iteration 76600 (3.75567 iter/s, 26.6264s/100 iters), loss = 6.3179
I0510 22:11:20.534219 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.3179 (* 1 = 6.3179 loss)
I0510 22:11:20.534234 39557 sgd_solver.cpp:105] Iteration 76600, lr = 0.1
I0510 22:11:47.235985 39557 solver.cpp:218] Iteration 76700 (3.74509 iter/s, 26.7016s/100 iters), loss = 6.44731
I0510 22:11:47.236050 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.44731 (* 1 = 6.44731 loss)
I0510 22:11:47.236065 39557 sgd_solver.cpp:105] Iteration 76700, lr = 0.1
I0510 22:12:13.932754 39557 solver.cpp:218] Iteration 76800 (3.7458 iter/s, 26.6966s/100 iters), loss = 6.85216
I0510 22:12:13.932916 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.85216 (* 1 = 6.85216 loss)
I0510 22:12:13.932930 39557 sgd_solver.cpp:105] Iteration 76800, lr = 0.1
I0510 22:12:40.627176 39557 solver.cpp:218] Iteration 76900 (3.74614 iter/s, 26.6941s/100 iters), loss = 6.58914
I0510 22:12:40.627238 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.58914 (* 1 = 6.58914 loss)
I0510 22:12:40.627251 39557 sgd_solver.cpp:105] Iteration 76900, lr = 0.1
I0510 22:13:07.361943 39557 solver.cpp:218] Iteration 77000 (3.74047 iter/s, 26.7346s/100 iters), loss = 6.74594
I0510 22:13:07.362104 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.74594 (* 1 = 6.74594 loss)
I0510 22:13:07.362123 39557 sgd_solver.cpp:105] Iteration 77000, lr = 0.1
I0510 22:13:34.139235 39557 solver.cpp:218] Iteration 77100 (3.73455 iter/s, 26.777s/100 iters), loss = 6.5598
I0510 22:13:34.139305 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.5598 (* 1 = 6.5598 loss)
I0510 22:13:34.139323 39557 sgd_solver.cpp:105] Iteration 77100, lr = 0.1
I0510 22:14:00.939770 39557 solver.cpp:218] Iteration 77200 (3.73129 iter/s, 26.8004s/100 iters), loss = 6.83905
I0510 22:14:00.939977 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.83905 (* 1 = 6.83905 loss)
I0510 22:14:00.939996 39557 sgd_solver.cpp:105] Iteration 77200, lr = 0.1
I0510 22:14:27.730166 39557 solver.cpp:218] Iteration 77300 (3.73273 iter/s, 26.7901s/100 iters), loss = 7.52678
I0510 22:14:27.730250 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.52678 (* 1 = 7.52678 loss)
I0510 22:14:27.730264 39557 sgd_solver.cpp:105] Iteration 77300, lr = 0.1
I0510 22:14:54.652226 39557 solver.cpp:218] Iteration 77400 (3.71445 iter/s, 26.9219s/100 iters), loss = 6.42036
I0510 22:14:54.652412 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.42036 (* 1 = 6.42036 loss)
I0510 22:14:54.652429 39557 sgd_solver.cpp:105] Iteration 77400, lr = 0.1
I0510 22:15:21.448371 39557 solver.cpp:218] Iteration 77500 (3.73192 iter/s, 26.7958s/100 iters), loss = 6.67577
I0510 22:15:21.448436 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.67577 (* 1 = 6.67577 loss)
I0510 22:15:21.448451 39557 sgd_solver.cpp:105] Iteration 77500, lr = 0.1
I0510 22:15:48.124500 39557 solver.cpp:218] Iteration 77600 (3.7487 iter/s, 26.6759s/100 iters), loss = 6.83615
I0510 22:15:48.124696 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.83615 (* 1 = 6.83615 loss)
I0510 22:15:48.124712 39557 sgd_solver.cpp:105] Iteration 77600, lr = 0.1
I0510 22:16:14.877684 39557 solver.cpp:218] Iteration 77700 (3.73792 iter/s, 26.7529s/100 iters), loss = 6.81037
I0510 22:16:14.877750 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.81037 (* 1 = 6.81037 loss)
I0510 22:16:14.877764 39557 sgd_solver.cpp:105] Iteration 77700, lr = 0.1
I0510 22:16:41.677806 39557 solver.cpp:218] Iteration 77800 (3.73135 iter/s, 26.7999s/100 iters), loss = 7.19636
I0510 22:16:41.677963 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.19636 (* 1 = 7.19636 loss)
I0510 22:16:41.677980 39557 sgd_solver.cpp:105] Iteration 77800, lr = 0.1
I0510 22:17:08.491971 39557 solver.cpp:218] Iteration 77900 (3.72941 iter/s, 26.8139s/100 iters), loss = 6.51017
I0510 22:17:08.492035 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.51017 (* 1 = 6.51017 loss)
I0510 22:17:08.492048 39557 sgd_solver.cpp:105] Iteration 77900, lr = 0.1
I0510 22:17:35.210077 39557 solver.cpp:218] Iteration 78000 (3.74281 iter/s, 26.7179s/100 iters), loss = 6.32735
I0510 22:17:35.210228 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.32735 (* 1 = 6.32735 loss)
I0510 22:17:35.210245 39557 sgd_solver.cpp:105] Iteration 78000, lr = 0.1
I0510 22:18:01.966352 39557 solver.cpp:218] Iteration 78100 (3.73748 iter/s, 26.756s/100 iters), loss = 7.02766
I0510 22:18:01.966426 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.02766 (* 1 = 7.02766 loss)
I0510 22:18:01.966439 39557 sgd_solver.cpp:105] Iteration 78100, lr = 0.1
I0510 22:18:28.582814 39557 solver.cpp:218] Iteration 78200 (3.7571 iter/s, 26.6163s/100 iters), loss = 7.21366
I0510 22:18:28.582993 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.21366 (* 1 = 7.21366 loss)
I0510 22:18:28.583009 39557 sgd_solver.cpp:105] Iteration 78200, lr = 0.1
I0510 22:18:55.249299 39557 solver.cpp:218] Iteration 78300 (3.75007 iter/s, 26.6662s/100 iters), loss = 7.19575
I0510 22:18:55.249366 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.19575 (* 1 = 7.19575 loss)
I0510 22:18:55.249379 39557 sgd_solver.cpp:105] Iteration 78300, lr = 0.1
I0510 22:19:21.883190 39557 solver.cpp:218] Iteration 78400 (3.75464 iter/s, 26.6337s/100 iters), loss = 7.18215
I0510 22:19:21.883357 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.18215 (* 1 = 7.18215 loss)
I0510 22:19:21.883371 39557 sgd_solver.cpp:105] Iteration 78400, lr = 0.1
I0510 22:19:48.502629 39557 solver.cpp:218] Iteration 78500 (3.75669 iter/s, 26.6192s/100 iters), loss = 6.93461
I0510 22:19:48.502688 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.93461 (* 1 = 6.93461 loss)
I0510 22:19:48.502699 39557 sgd_solver.cpp:105] Iteration 78500, lr = 0.1
I0510 22:20:15.196547 39557 solver.cpp:218] Iteration 78600 (3.7462 iter/s, 26.6937s/100 iters), loss = 6.22776
I0510 22:20:15.196691 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.22776 (* 1 = 6.22776 loss)
I0510 22:20:15.196717 39557 sgd_solver.cpp:105] Iteration 78600, lr = 0.1
I0510 22:20:41.863725 39557 solver.cpp:218] Iteration 78700 (3.74996 iter/s, 26.6669s/100 iters), loss = 6.31413
I0510 22:20:41.863783 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.31413 (* 1 = 6.31413 loss)
I0510 22:20:41.863795 39557 sgd_solver.cpp:105] Iteration 78700, lr = 0.1
I0510 22:21:08.525967 39557 solver.cpp:218] Iteration 78800 (3.75065 iter/s, 26.6621s/100 iters), loss = 7.02506
I0510 22:21:08.526111 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.02506 (* 1 = 7.02506 loss)
I0510 22:21:08.526125 39557 sgd_solver.cpp:105] Iteration 78800, lr = 0.1
I0510 22:21:35.140666 39557 solver.cpp:218] Iteration 78900 (3.75736 iter/s, 26.6144s/100 iters), loss = 6.3945
I0510 22:21:35.140724 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.3945 (* 1 = 6.3945 loss)
I0510 22:21:35.140735 39557 sgd_solver.cpp:105] Iteration 78900, lr = 0.1
I0510 22:22:01.773051 39557 solver.cpp:218] Iteration 79000 (3.75485 iter/s, 26.6322s/100 iters), loss = 7.03118
I0510 22:22:01.773232 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.03118 (* 1 = 7.03118 loss)
I0510 22:22:01.773257 39557 sgd_solver.cpp:105] Iteration 79000, lr = 0.1
I0510 22:22:28.490744 39557 solver.cpp:218] Iteration 79100 (3.74288 iter/s, 26.7174s/100 iters), loss = 6.28915
I0510 22:22:28.490808 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.28915 (* 1 = 6.28915 loss)
I0510 22:22:28.490819 39557 sgd_solver.cpp:105] Iteration 79100, lr = 0.1
I0510 22:22:55.247611 39557 solver.cpp:218] Iteration 79200 (3.73738 iter/s, 26.7567s/100 iters), loss = 6.88477
I0510 22:22:55.247763 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.88477 (* 1 = 6.88477 loss)
I0510 22:22:55.247777 39557 sgd_solver.cpp:105] Iteration 79200, lr = 0.1
I0510 22:23:21.911239 39557 solver.cpp:218] Iteration 79300 (3.75047 iter/s, 26.6634s/100 iters), loss = 6.72476
I0510 22:23:21.911303 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.72476 (* 1 = 6.72476 loss)
I0510 22:23:21.911314 39557 sgd_solver.cpp:105] Iteration 79300, lr = 0.1
I0510 22:23:48.639436 39557 solver.cpp:218] Iteration 79400 (3.74139 iter/s, 26.728s/100 iters), loss = 6.66124
I0510 22:23:48.639580 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.66124 (* 1 = 6.66124 loss)
I0510 22:23:48.639595 39557 sgd_solver.cpp:105] Iteration 79400, lr = 0.1
I0510 22:24:15.348423 39557 solver.cpp:218] Iteration 79500 (3.7441 iter/s, 26.7087s/100 iters), loss = 6.64025
I0510 22:24:15.348486 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.64025 (* 1 = 6.64025 loss)
I0510 22:24:15.348500 39557 sgd_solver.cpp:105] Iteration 79500, lr = 0.1
I0510 22:24:42.039819 39557 solver.cpp:218] Iteration 79600 (3.74655 iter/s, 26.6912s/100 iters), loss = 6.65218
I0510 22:24:42.039981 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.65218 (* 1 = 6.65218 loss)
I0510 22:24:42.039995 39557 sgd_solver.cpp:105] Iteration 79600, lr = 0.1
I0510 22:25:08.774466 39557 solver.cpp:218] Iteration 79700 (3.7405 iter/s, 26.7344s/100 iters), loss = 6.66031
I0510 22:25:08.774529 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.66031 (* 1 = 6.66031 loss)
I0510 22:25:08.774543 39557 sgd_solver.cpp:105] Iteration 79700, lr = 0.1
I0510 22:25:35.547807 39557 solver.cpp:218] Iteration 79800 (3.73509 iter/s, 26.7732s/100 iters), loss = 7.12731
I0510 22:25:35.547952 39557 solver.cpp:237]     Train net output #0: softmax_loss = 7.12731 (* 1 = 7.12731 loss)
I0510 22:25:35.547967 39557 sgd_solver.cpp:105] Iteration 79800, lr = 0.1
I0510 22:26:02.233260 39557 solver.cpp:218] Iteration 79900 (3.7474 iter/s, 26.6852s/100 iters), loss = 6.79159
I0510 22:26:02.233321 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.79159 (* 1 = 6.79159 loss)
I0510 22:26:02.233335 39557 sgd_solver.cpp:105] Iteration 79900, lr = 0.1
I0510 22:26:28.646785 39557 solver.cpp:447] Snapshotting to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_80000.caffemodel
I0510 22:26:28.686518 39557 sgd_solver.cpp:273] Snapshotting solver state to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_80000.solverstate
I0510 22:26:28.971118 39557 solver.cpp:218] Iteration 80000 (3.74004 iter/s, 26.7377s/100 iters), loss = 6.65869
I0510 22:26:28.971148 39624 sgd_solver.cpp:46] MultiStep Status: Iteration 80000, step = 1
I0510 22:26:28.971170 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.65869 (* 1 = 6.65869 loss)
I0510 22:26:28.971179 39557 sgd_solver.cpp:46] MultiStep Status: Iteration 80000, step = 1
I0510 22:26:28.971184 39557 sgd_solver.cpp:105] Iteration 80000, lr = 0.01
I0510 22:26:55.614521 39557 solver.cpp:218] Iteration 80100 (3.75331 iter/s, 26.6431s/100 iters), loss = 6.56848
I0510 22:26:55.614580 39557 solver.cpp:237]     Train net output #0: softmax_loss = 6.56848 (* 1 = 6.56848 loss)
I0510 22:26:55.614593 39557 sgd_solver.cpp:105] Iteration 80100, lr = 0.01
I0510 22:27:22.277081 39557 solver.cpp:218] Iteration 80200 (3.75062 iter/s, 26.6623s/100 iters), loss = 5.31302
I0510 22:27:22.277259 39557 solver.cpp:237]     Train net output #0: softmax_loss = 5.31302 (* 1 = 5.31302 loss)
I0510 22:27:22.277272 39557 sgd_solver.cpp:105] Iteration 80200, lr = 0.01
I0510 22:27:48.934429 39557 solver.cpp:218] Iteration 80300 (3.75137 iter/s, 26.6569s/100 iters), loss = 5.4221
I0510 22:27:48.934489 39557 solver.cpp:237]     Train net output #0: softmax_loss = 5.4221 (* 1 = 5.4221 loss)
I0510 22:27:48.934501 39557 sgd_solver.cpp:105] Iteration 80300, lr = 0.01
I0510 22:28:15.599301 39557 solver.cpp:218] Iteration 80400 (3.75029 iter/s, 26.6646s/100 iters), loss = 5.46662
I0510 22:28:15.599453 39557 solver.cpp:237]     Train net output #0: softmax_loss = 5.46662 (* 1 = 5.46662 loss)
I0510 22:28:15.599467 39557 sgd_solver.cpp:105] Iteration 80400, lr = 0.01
I0510 22:28:42.342674 39557 solver.cpp:218] Iteration 80500 (3.7393 iter/s, 26.743s/100 iters), loss = 5.85237
I0510 22:28:42.342725 39557 solver.cpp:237]     Train net output #0: softmax_loss = 5.85237 (* 1 = 5.85237 loss)
I0510 22:28:42.342737 39557 sgd_solver.cpp:105] Iteration 80500, lr = 0.01
I0510 22:29:09.125926 39557 solver.cpp:218] Iteration 80600 (3.73371 iter/s, 26.783s/100 iters), loss = 5.21151
I0510 22:29:09.126085 39557 solver.cpp:237]     Train net output #0: softmax_loss = 5.21151 (* 1 = 5.21151 loss)
I0510 22:29:09.126097 39557 sgd_solver.cpp:105] Iteration 80600, lr = 0.01
I0510 22:29:35.876888 39557 solver.cpp:218] Iteration 80700 (3.73824 iter/s, 26.7506s/100 iters), loss = 4.74013
I0510 22:29:35.876962 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.74013 (* 1 = 4.74013 loss)
I0510 22:29:35.876974 39557 sgd_solver.cpp:105] Iteration 80700, lr = 0.01
I0510 22:30:02.653594 39557 solver.cpp:218] Iteration 80800 (3.73463 iter/s, 26.7764s/100 iters), loss = 4.46667
I0510 22:30:02.653745 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.46667 (* 1 = 4.46667 loss)
I0510 22:30:02.653761 39557 sgd_solver.cpp:105] Iteration 80800, lr = 0.01
I0510 22:30:29.414598 39557 solver.cpp:218] Iteration 80900 (3.73683 iter/s, 26.7607s/100 iters), loss = 4.56529
I0510 22:30:29.414662 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.56529 (* 1 = 4.56529 loss)
I0510 22:30:29.414677 39557 sgd_solver.cpp:105] Iteration 80900, lr = 0.01
I0510 22:30:56.218883 39557 solver.cpp:218] Iteration 81000 (3.73078 iter/s, 26.804s/100 iters), loss = 4.61787
I0510 22:30:56.219048 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.61787 (* 1 = 4.61787 loss)
I0510 22:30:56.219063 39557 sgd_solver.cpp:105] Iteration 81000, lr = 0.01
I0510 22:31:22.999954 39557 solver.cpp:218] Iteration 81100 (3.73403 iter/s, 26.7807s/100 iters), loss = 5.62593
I0510 22:31:23.000015 39557 solver.cpp:237]     Train net output #0: softmax_loss = 5.62593 (* 1 = 5.62593 loss)
I0510 22:31:23.000030 39557 sgd_solver.cpp:105] Iteration 81100, lr = 0.01
I0510 22:31:49.616278 39557 solver.cpp:218] Iteration 81200 (3.75713 iter/s, 26.6161s/100 iters), loss = 5.82431
I0510 22:31:49.616447 39557 solver.cpp:237]     Train net output #0: softmax_loss = 5.82431 (* 1 = 5.82431 loss)
I0510 22:31:49.616462 39557 sgd_solver.cpp:105] Iteration 81200, lr = 0.01
I0510 22:32:16.250643 39557 solver.cpp:218] Iteration 81300 (3.7546 iter/s, 26.634s/100 iters), loss = 4.96749
I0510 22:32:16.250705 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.96749 (* 1 = 4.96749 loss)
I0510 22:32:16.250716 39557 sgd_solver.cpp:105] Iteration 81300, lr = 0.01
I0510 22:32:42.878857 39557 solver.cpp:218] Iteration 81400 (3.75545 iter/s, 26.628s/100 iters), loss = 4.75524
I0510 22:32:42.879007 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.75524 (* 1 = 4.75524 loss)
I0510 22:32:42.879022 39557 sgd_solver.cpp:105] Iteration 81400, lr = 0.01
I0510 22:33:09.525434 39557 solver.cpp:218] Iteration 81500 (3.75287 iter/s, 26.6462s/100 iters), loss = 4.96684
I0510 22:33:09.525497 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.96684 (* 1 = 4.96684 loss)
I0510 22:33:09.525512 39557 sgd_solver.cpp:105] Iteration 81500, lr = 0.01
I0510 22:33:36.275585 39557 solver.cpp:218] Iteration 81600 (3.73833 iter/s, 26.7499s/100 iters), loss = 4.66607
I0510 22:33:36.275765 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.66607 (* 1 = 4.66607 loss)
I0510 22:33:36.275782 39557 sgd_solver.cpp:105] Iteration 81600, lr = 0.01
I0510 22:34:02.968559 39557 solver.cpp:218] Iteration 81700 (3.74635 iter/s, 26.6926s/100 iters), loss = 4.72986
I0510 22:34:02.968624 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.72986 (* 1 = 4.72986 loss)
I0510 22:34:02.968638 39557 sgd_solver.cpp:105] Iteration 81700, lr = 0.01
I0510 22:34:29.705413 39557 solver.cpp:218] Iteration 81800 (3.74019 iter/s, 26.7366s/100 iters), loss = 4.85636
I0510 22:34:29.705590 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.85636 (* 1 = 4.85636 loss)
I0510 22:34:29.705602 39557 sgd_solver.cpp:105] Iteration 81800, lr = 0.01
I0510 22:34:56.541152 39557 solver.cpp:218] Iteration 81900 (3.72642 iter/s, 26.8354s/100 iters), loss = 4.51485
I0510 22:34:56.541218 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.51485 (* 1 = 4.51485 loss)
I0510 22:34:56.541231 39557 sgd_solver.cpp:105] Iteration 81900, lr = 0.01
I0510 22:35:23.382508 39557 solver.cpp:218] Iteration 82000 (3.72563 iter/s, 26.8411s/100 iters), loss = 5.19728
I0510 22:35:23.382671 39557 solver.cpp:237]     Train net output #0: softmax_loss = 5.19728 (* 1 = 5.19728 loss)
I0510 22:35:23.382688 39557 sgd_solver.cpp:105] Iteration 82000, lr = 0.01
I0510 22:35:50.142735 39557 solver.cpp:218] Iteration 82100 (3.73693 iter/s, 26.7599s/100 iters), loss = 5.21221
I0510 22:35:50.142802 39557 solver.cpp:237]     Train net output #0: softmax_loss = 5.21221 (* 1 = 5.21221 loss)
I0510 22:35:50.142817 39557 sgd_solver.cpp:105] Iteration 82100, lr = 0.01
I0510 22:36:16.943825 39557 solver.cpp:218] Iteration 82200 (3.73122 iter/s, 26.8009s/100 iters), loss = 4.52043
I0510 22:36:16.943994 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.52043 (* 1 = 4.52043 loss)
I0510 22:36:16.944012 39557 sgd_solver.cpp:105] Iteration 82200, lr = 0.01
I0510 22:36:43.779168 39557 solver.cpp:218] Iteration 82300 (3.72648 iter/s, 26.835s/100 iters), loss = 5.28864
I0510 22:36:43.779230 39557 solver.cpp:237]     Train net output #0: softmax_loss = 5.28864 (* 1 = 5.28864 loss)
I0510 22:36:43.779248 39557 sgd_solver.cpp:105] Iteration 82300, lr = 0.01
I0510 22:37:10.579010 39557 solver.cpp:218] Iteration 82400 (3.7314 iter/s, 26.7996s/100 iters), loss = 4.56799
I0510 22:37:10.579172 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.56799 (* 1 = 4.56799 loss)
I0510 22:37:10.579190 39557 sgd_solver.cpp:105] Iteration 82400, lr = 0.01
I0510 22:37:37.388883 39557 solver.cpp:218] Iteration 82500 (3.73002 iter/s, 26.8095s/100 iters), loss = 4.68675
I0510 22:37:37.388974 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.68675 (* 1 = 4.68675 loss)
I0510 22:37:37.388989 39557 sgd_solver.cpp:105] Iteration 82500, lr = 0.01
I0510 22:38:04.114934 39557 solver.cpp:218] Iteration 82600 (3.74171 iter/s, 26.7258s/100 iters), loss = 4.23374
I0510 22:38:04.115106 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.23374 (* 1 = 4.23374 loss)
I0510 22:38:04.115121 39557 sgd_solver.cpp:105] Iteration 82600, lr = 0.01
I0510 22:38:30.908363 39557 solver.cpp:218] Iteration 82700 (3.7323 iter/s, 26.7931s/100 iters), loss = 4.88081
I0510 22:38:30.908442 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.88081 (* 1 = 4.88081 loss)
I0510 22:38:30.908455 39557 sgd_solver.cpp:105] Iteration 82700, lr = 0.01
I0510 22:38:57.743083 39557 solver.cpp:218] Iteration 82800 (3.72655 iter/s, 26.8345s/100 iters), loss = 4.23613
I0510 22:38:57.743252 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.23613 (* 1 = 4.23613 loss)
I0510 22:38:57.743268 39557 sgd_solver.cpp:105] Iteration 82800, lr = 0.01
I0510 22:39:24.591199 39557 solver.cpp:218] Iteration 82900 (3.7247 iter/s, 26.8478s/100 iters), loss = 4.49936
I0510 22:39:24.591267 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.49936 (* 1 = 4.49936 loss)
I0510 22:39:24.591282 39557 sgd_solver.cpp:105] Iteration 82900, lr = 0.01
I0510 22:39:51.341938 39557 solver.cpp:218] Iteration 83000 (3.73825 iter/s, 26.7505s/100 iters), loss = 4.23563
I0510 22:39:51.342131 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.23563 (* 1 = 4.23563 loss)
I0510 22:39:51.342147 39557 sgd_solver.cpp:105] Iteration 83000, lr = 0.01
I0510 22:40:17.980147 39557 solver.cpp:218] Iteration 83100 (3.75405 iter/s, 26.6379s/100 iters), loss = 3.93998
I0510 22:40:17.980209 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.93998 (* 1 = 3.93998 loss)
I0510 22:40:17.980221 39557 sgd_solver.cpp:105] Iteration 83100, lr = 0.01
I0510 22:40:44.599014 39557 solver.cpp:218] Iteration 83200 (3.75676 iter/s, 26.6186s/100 iters), loss = 3.92228
I0510 22:40:44.599185 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.92228 (* 1 = 3.92228 loss)
I0510 22:40:44.599200 39557 sgd_solver.cpp:105] Iteration 83200, lr = 0.01
I0510 22:41:11.244356 39557 solver.cpp:218] Iteration 83300 (3.75305 iter/s, 26.645s/100 iters), loss = 4.24733
I0510 22:41:11.244415 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.24733 (* 1 = 4.24733 loss)
I0510 22:41:11.244426 39557 sgd_solver.cpp:105] Iteration 83300, lr = 0.01
I0510 22:41:37.816453 39557 solver.cpp:218] Iteration 83400 (3.76338 iter/s, 26.5719s/100 iters), loss = 3.81923
I0510 22:41:37.816601 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.81923 (* 1 = 3.81923 loss)
I0510 22:41:37.816615 39557 sgd_solver.cpp:105] Iteration 83400, lr = 0.01
I0510 22:42:04.445693 39557 solver.cpp:218] Iteration 83500 (3.75531 iter/s, 26.6289s/100 iters), loss = 4.19054
I0510 22:42:04.445751 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.19054 (* 1 = 4.19054 loss)
I0510 22:42:04.445761 39557 sgd_solver.cpp:105] Iteration 83500, lr = 0.01
I0510 22:42:31.051699 39557 solver.cpp:218] Iteration 83600 (3.75858 iter/s, 26.6058s/100 iters), loss = 3.96177
I0510 22:42:31.051851 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.96177 (* 1 = 3.96177 loss)
I0510 22:42:31.051865 39557 sgd_solver.cpp:105] Iteration 83600, lr = 0.01
I0510 22:42:57.619644 39557 solver.cpp:218] Iteration 83700 (3.76398 iter/s, 26.5676s/100 iters), loss = 5.31814
I0510 22:42:57.619701 39557 solver.cpp:237]     Train net output #0: softmax_loss = 5.31814 (* 1 = 5.31814 loss)
I0510 22:42:57.619711 39557 sgd_solver.cpp:105] Iteration 83700, lr = 0.01
I0510 22:43:24.265527 39557 solver.cpp:218] Iteration 83800 (3.75295 iter/s, 26.6457s/100 iters), loss = 3.72753
I0510 22:43:24.265674 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.72753 (* 1 = 3.72753 loss)
I0510 22:43:24.265691 39557 sgd_solver.cpp:105] Iteration 83800, lr = 0.01
I0510 22:43:50.867977 39557 solver.cpp:218] Iteration 83900 (3.75909 iter/s, 26.6022s/100 iters), loss = 3.60556
I0510 22:43:50.868033 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.60556 (* 1 = 3.60556 loss)
I0510 22:43:50.868044 39557 sgd_solver.cpp:105] Iteration 83900, lr = 0.01
I0510 22:44:17.458484 39557 solver.cpp:218] Iteration 84000 (3.76077 iter/s, 26.5903s/100 iters), loss = 3.91964
I0510 22:44:17.458634 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.91964 (* 1 = 3.91964 loss)
I0510 22:44:17.458649 39557 sgd_solver.cpp:105] Iteration 84000, lr = 0.01
I0510 22:44:44.051277 39557 solver.cpp:218] Iteration 84100 (3.76046 iter/s, 26.5925s/100 iters), loss = 3.66382
I0510 22:44:44.051338 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.66382 (* 1 = 3.66382 loss)
I0510 22:44:44.051350 39557 sgd_solver.cpp:105] Iteration 84100, lr = 0.01
I0510 22:45:10.662263 39557 solver.cpp:218] Iteration 84200 (3.75788 iter/s, 26.6108s/100 iters), loss = 4.47785
I0510 22:45:10.662451 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.47785 (* 1 = 4.47785 loss)
I0510 22:45:10.662468 39557 sgd_solver.cpp:105] Iteration 84200, lr = 0.01
I0510 22:45:37.270311 39557 solver.cpp:218] Iteration 84300 (3.75831 iter/s, 26.6077s/100 iters), loss = 3.54742
I0510 22:45:37.270375 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.54742 (* 1 = 3.54742 loss)
I0510 22:45:37.270388 39557 sgd_solver.cpp:105] Iteration 84300, lr = 0.01
I0510 22:46:04.026027 39557 solver.cpp:218] Iteration 84400 (3.73755 iter/s, 26.7555s/100 iters), loss = 4.91008
I0510 22:46:04.026197 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.91008 (* 1 = 4.91008 loss)
I0510 22:46:04.026214 39557 sgd_solver.cpp:105] Iteration 84400, lr = 0.01
I0510 22:46:30.751106 39557 solver.cpp:218] Iteration 84500 (3.74185 iter/s, 26.7248s/100 iters), loss = 4.94408
I0510 22:46:30.751173 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.94408 (* 1 = 4.94408 loss)
I0510 22:46:30.751186 39557 sgd_solver.cpp:105] Iteration 84500, lr = 0.01
I0510 22:46:57.424932 39557 solver.cpp:218] Iteration 84600 (3.74902 iter/s, 26.6736s/100 iters), loss = 3.78765
I0510 22:46:57.425117 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.78765 (* 1 = 3.78765 loss)
I0510 22:46:57.425132 39557 sgd_solver.cpp:105] Iteration 84600, lr = 0.01
I0510 22:47:24.178269 39557 solver.cpp:218] Iteration 84700 (3.7379 iter/s, 26.753s/100 iters), loss = 3.97045
I0510 22:47:24.178333 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.97045 (* 1 = 3.97045 loss)
I0510 22:47:24.178347 39557 sgd_solver.cpp:105] Iteration 84700, lr = 0.01
I0510 22:47:50.831822 39557 solver.cpp:218] Iteration 84800 (3.75188 iter/s, 26.6533s/100 iters), loss = 5.14284
I0510 22:47:50.831985 39557 solver.cpp:237]     Train net output #0: softmax_loss = 5.14284 (* 1 = 5.14284 loss)
I0510 22:47:50.832000 39557 sgd_solver.cpp:105] Iteration 84800, lr = 0.01
I0510 22:48:17.546383 39557 solver.cpp:218] Iteration 84900 (3.74332 iter/s, 26.7142s/100 iters), loss = 4.05012
I0510 22:48:17.546456 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.05012 (* 1 = 4.05012 loss)
I0510 22:48:17.546469 39557 sgd_solver.cpp:105] Iteration 84900, lr = 0.01
I0510 22:48:43.997946 39557 solver.cpp:447] Snapshotting to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_85000.caffemodel
I0510 22:48:44.037817 39557 sgd_solver.cpp:273] Snapshotting solver state to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_85000.solverstate
I0510 22:48:44.322211 39557 solver.cpp:218] Iteration 85000 (3.73474 iter/s, 26.7756s/100 iters), loss = 4.35079
I0510 22:48:44.322268 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.35079 (* 1 = 4.35079 loss)
I0510 22:48:44.322278 39557 sgd_solver.cpp:105] Iteration 85000, lr = 0.01
I0510 22:49:11.066403 39557 solver.cpp:218] Iteration 85100 (3.73916 iter/s, 26.744s/100 iters), loss = 3.308
I0510 22:49:11.066465 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.308 (* 1 = 3.308 loss)
I0510 22:49:11.066479 39557 sgd_solver.cpp:105] Iteration 85100, lr = 0.01
I0510 22:49:37.801118 39557 solver.cpp:218] Iteration 85200 (3.74048 iter/s, 26.7345s/100 iters), loss = 3.70932
I0510 22:49:37.801280 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.70932 (* 1 = 3.70932 loss)
I0510 22:49:37.801291 39557 sgd_solver.cpp:105] Iteration 85200, lr = 0.01
I0510 22:50:04.475175 39557 solver.cpp:218] Iteration 85300 (3.749 iter/s, 26.6737s/100 iters), loss = 3.77158
I0510 22:50:04.475235 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.77158 (* 1 = 3.77158 loss)
I0510 22:50:04.475250 39557 sgd_solver.cpp:105] Iteration 85300, lr = 0.01
I0510 22:50:31.160548 39557 solver.cpp:218] Iteration 85400 (3.7474 iter/s, 26.6852s/100 iters), loss = 4.24059
I0510 22:50:31.160706 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.24059 (* 1 = 4.24059 loss)
I0510 22:50:31.160717 39557 sgd_solver.cpp:105] Iteration 85400, lr = 0.01
I0510 22:50:57.923501 39557 solver.cpp:218] Iteration 85500 (3.73655 iter/s, 26.7626s/100 iters), loss = 4.2593
I0510 22:50:57.923558 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.2593 (* 1 = 4.2593 loss)
I0510 22:50:57.923574 39557 sgd_solver.cpp:105] Iteration 85500, lr = 0.01
I0510 22:51:24.623663 39557 solver.cpp:218] Iteration 85600 (3.74532 iter/s, 26.7s/100 iters), loss = 3.61758
I0510 22:51:24.623845 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.61758 (* 1 = 3.61758 loss)
I0510 22:51:24.623859 39557 sgd_solver.cpp:105] Iteration 85600, lr = 0.01
I0510 22:51:51.248592 39557 solver.cpp:218] Iteration 85700 (3.75592 iter/s, 26.6246s/100 iters), loss = 4.15491
I0510 22:51:51.248653 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.15491 (* 1 = 4.15491 loss)
I0510 22:51:51.248668 39557 sgd_solver.cpp:105] Iteration 85700, lr = 0.01
I0510 22:52:17.891258 39557 solver.cpp:218] Iteration 85800 (3.75341 iter/s, 26.6425s/100 iters), loss = 3.989
I0510 22:52:17.891412 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.989 (* 1 = 3.989 loss)
I0510 22:52:17.891425 39557 sgd_solver.cpp:105] Iteration 85800, lr = 0.01
I0510 22:52:44.455811 39557 solver.cpp:218] Iteration 85900 (3.76446 iter/s, 26.5643s/100 iters), loss = 4.20513
I0510 22:52:44.455868 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.20513 (* 1 = 4.20513 loss)
I0510 22:52:44.455881 39557 sgd_solver.cpp:105] Iteration 85900, lr = 0.01
I0510 22:53:11.030123 39557 solver.cpp:218] Iteration 86000 (3.76306 iter/s, 26.5741s/100 iters), loss = 4.25422
I0510 22:53:11.030269 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.25422 (* 1 = 4.25422 loss)
I0510 22:53:11.030282 39557 sgd_solver.cpp:105] Iteration 86000, lr = 0.01
I0510 22:53:37.698161 39557 solver.cpp:218] Iteration 86100 (3.74985 iter/s, 26.6678s/100 iters), loss = 3.88451
I0510 22:53:37.698222 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.88451 (* 1 = 3.88451 loss)
I0510 22:53:37.698235 39557 sgd_solver.cpp:105] Iteration 86100, lr = 0.01
I0510 22:54:04.416771 39557 solver.cpp:218] Iteration 86200 (3.74274 iter/s, 26.7184s/100 iters), loss = 3.5794
I0510 22:54:04.416944 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.5794 (* 1 = 3.5794 loss)
I0510 22:54:04.416962 39557 sgd_solver.cpp:105] Iteration 86200, lr = 0.01
I0510 22:54:31.100198 39557 solver.cpp:218] Iteration 86300 (3.74769 iter/s, 26.6831s/100 iters), loss = 3.14586
I0510 22:54:31.100258 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.14586 (* 1 = 3.14586 loss)
I0510 22:54:31.100270 39557 sgd_solver.cpp:105] Iteration 86300, lr = 0.01
I0510 22:54:57.688614 39557 solver.cpp:218] Iteration 86400 (3.76106 iter/s, 26.5882s/100 iters), loss = 3.87646
I0510 22:54:57.688762 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.87646 (* 1 = 3.87646 loss)
I0510 22:54:57.688776 39557 sgd_solver.cpp:105] Iteration 86400, lr = 0.01
I0510 22:55:24.373476 39557 solver.cpp:218] Iteration 86500 (3.74748 iter/s, 26.6846s/100 iters), loss = 3.86687
I0510 22:55:24.373540 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.86687 (* 1 = 3.86687 loss)
I0510 22:55:24.373553 39557 sgd_solver.cpp:105] Iteration 86500, lr = 0.01
I0510 22:55:50.972610 39557 solver.cpp:218] Iteration 86600 (3.75955 iter/s, 26.5989s/100 iters), loss = 3.44183
I0510 22:55:50.972779 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.44183 (* 1 = 3.44183 loss)
I0510 22:55:50.972793 39557 sgd_solver.cpp:105] Iteration 86600, lr = 0.01
I0510 22:56:17.628049 39557 solver.cpp:218] Iteration 86700 (3.75162 iter/s, 26.6551s/100 iters), loss = 4.10009
I0510 22:56:17.628109 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.10009 (* 1 = 4.10009 loss)
I0510 22:56:17.628123 39557 sgd_solver.cpp:105] Iteration 86700, lr = 0.01
I0510 22:56:44.396474 39557 solver.cpp:218] Iteration 86800 (3.73577 iter/s, 26.7682s/100 iters), loss = 3.80685
I0510 22:56:44.396627 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.80685 (* 1 = 3.80685 loss)
I0510 22:56:44.396641 39557 sgd_solver.cpp:105] Iteration 86800, lr = 0.01
I0510 22:57:11.064039 39557 solver.cpp:218] Iteration 86900 (3.74992 iter/s, 26.6673s/100 iters), loss = 4.73146
I0510 22:57:11.064100 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.73146 (* 1 = 4.73146 loss)
I0510 22:57:11.064112 39557 sgd_solver.cpp:105] Iteration 86900, lr = 0.01
I0510 22:57:37.810117 39557 solver.cpp:218] Iteration 87000 (3.73889 iter/s, 26.7459s/100 iters), loss = 4.30432
I0510 22:57:37.810293 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.30432 (* 1 = 4.30432 loss)
I0510 22:57:37.810307 39557 sgd_solver.cpp:105] Iteration 87000, lr = 0.01
I0510 22:58:04.621054 39557 solver.cpp:218] Iteration 87100 (3.72986 iter/s, 26.8106s/100 iters), loss = 3.42116
I0510 22:58:04.621116 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.42116 (* 1 = 3.42116 loss)
I0510 22:58:04.621129 39557 sgd_solver.cpp:105] Iteration 87100, lr = 0.01
I0510 22:58:31.425498 39557 solver.cpp:218] Iteration 87200 (3.73075 iter/s, 26.8042s/100 iters), loss = 3.71942
I0510 22:58:31.425690 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.71942 (* 1 = 3.71942 loss)
I0510 22:58:31.425709 39557 sgd_solver.cpp:105] Iteration 87200, lr = 0.01
I0510 22:58:58.180673 39557 solver.cpp:218] Iteration 87300 (3.73764 iter/s, 26.7548s/100 iters), loss = 4.41576
I0510 22:58:58.180747 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.41576 (* 1 = 4.41576 loss)
I0510 22:58:58.180769 39557 sgd_solver.cpp:105] Iteration 87300, lr = 0.01
I0510 22:59:24.972725 39557 solver.cpp:218] Iteration 87400 (3.73248 iter/s, 26.7918s/100 iters), loss = 3.7027
I0510 22:59:24.972908 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.7027 (* 1 = 3.7027 loss)
I0510 22:59:24.972926 39557 sgd_solver.cpp:105] Iteration 87400, lr = 0.01
I0510 22:59:51.736184 39557 solver.cpp:218] Iteration 87500 (3.73648 iter/s, 26.7631s/100 iters), loss = 3.41662
I0510 22:59:51.736248 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.41662 (* 1 = 3.41662 loss)
I0510 22:59:51.736263 39557 sgd_solver.cpp:105] Iteration 87500, lr = 0.01
I0510 23:00:18.498208 39557 solver.cpp:218] Iteration 87600 (3.73667 iter/s, 26.7618s/100 iters), loss = 3.57781
I0510 23:00:18.498355 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.57781 (* 1 = 3.57781 loss)
I0510 23:00:18.498370 39557 sgd_solver.cpp:105] Iteration 87600, lr = 0.01
I0510 23:00:45.361338 39557 solver.cpp:218] Iteration 87700 (3.72261 iter/s, 26.8628s/100 iters), loss = 3.80683
I0510 23:00:45.361392 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.80683 (* 1 = 3.80683 loss)
I0510 23:00:45.361408 39557 sgd_solver.cpp:105] Iteration 87700, lr = 0.01
I0510 23:01:12.107659 39557 solver.cpp:218] Iteration 87800 (3.73886 iter/s, 26.7461s/100 iters), loss = 3.70467
I0510 23:01:12.107830 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.70467 (* 1 = 3.70467 loss)
I0510 23:01:12.107844 39557 sgd_solver.cpp:105] Iteration 87800, lr = 0.01
I0510 23:01:38.807416 39557 solver.cpp:218] Iteration 87900 (3.7454 iter/s, 26.6995s/100 iters), loss = 4.60556
I0510 23:01:38.807480 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.60556 (* 1 = 4.60556 loss)
I0510 23:01:38.807492 39557 sgd_solver.cpp:105] Iteration 87900, lr = 0.01
I0510 23:02:05.466599 39557 solver.cpp:218] Iteration 88000 (3.75108 iter/s, 26.659s/100 iters), loss = 4.57301
I0510 23:02:05.466776 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.57301 (* 1 = 4.57301 loss)
I0510 23:02:05.466791 39557 sgd_solver.cpp:105] Iteration 88000, lr = 0.01
I0510 23:02:32.126646 39557 solver.cpp:218] Iteration 88100 (3.75098 iter/s, 26.6597s/100 iters), loss = 3.0838
I0510 23:02:32.126706 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.0838 (* 1 = 3.0838 loss)
I0510 23:02:32.126718 39557 sgd_solver.cpp:105] Iteration 88100, lr = 0.01
I0510 23:02:58.814261 39557 solver.cpp:218] Iteration 88200 (3.74709 iter/s, 26.6874s/100 iters), loss = 3.90531
I0510 23:02:58.814483 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.90531 (* 1 = 3.90531 loss)
I0510 23:02:58.814501 39557 sgd_solver.cpp:105] Iteration 88200, lr = 0.01
I0510 23:03:25.612241 39557 solver.cpp:218] Iteration 88300 (3.73168 iter/s, 26.7976s/100 iters), loss = 3.63503
I0510 23:03:25.612304 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.63503 (* 1 = 3.63503 loss)
I0510 23:03:25.612318 39557 sgd_solver.cpp:105] Iteration 88300, lr = 0.01
I0510 23:03:52.319576 39557 solver.cpp:218] Iteration 88400 (3.74432 iter/s, 26.7071s/100 iters), loss = 3.83223
I0510 23:03:52.319751 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.83223 (* 1 = 3.83223 loss)
I0510 23:03:52.319764 39557 sgd_solver.cpp:105] Iteration 88400, lr = 0.01
I0510 23:04:19.044533 39557 solver.cpp:218] Iteration 88500 (3.74187 iter/s, 26.7246s/100 iters), loss = 3.58054
I0510 23:04:19.044600 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.58054 (* 1 = 3.58054 loss)
I0510 23:04:19.044616 39557 sgd_solver.cpp:105] Iteration 88500, lr = 0.01
I0510 23:04:45.778923 39557 solver.cpp:218] Iteration 88600 (3.74053 iter/s, 26.7342s/100 iters), loss = 4.39137
I0510 23:04:45.779094 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.39137 (* 1 = 4.39137 loss)
I0510 23:04:45.779106 39557 sgd_solver.cpp:105] Iteration 88600, lr = 0.01
I0510 23:05:12.539862 39557 solver.cpp:218] Iteration 88700 (3.73683 iter/s, 26.7606s/100 iters), loss = 4.03402
I0510 23:05:12.539921 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.03402 (* 1 = 4.03402 loss)
I0510 23:05:12.539937 39557 sgd_solver.cpp:105] Iteration 88700, lr = 0.01
I0510 23:05:39.248311 39557 solver.cpp:218] Iteration 88800 (3.74416 iter/s, 26.7083s/100 iters), loss = 3.84916
I0510 23:05:39.248481 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.84916 (* 1 = 3.84916 loss)
I0510 23:05:39.248494 39557 sgd_solver.cpp:105] Iteration 88800, lr = 0.01
I0510 23:06:05.844835 39557 solver.cpp:218] Iteration 88900 (3.75993 iter/s, 26.5962s/100 iters), loss = 3.88355
I0510 23:06:05.844892 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.88355 (* 1 = 3.88355 loss)
I0510 23:06:05.844905 39557 sgd_solver.cpp:105] Iteration 88900, lr = 0.01
I0510 23:06:32.503582 39557 solver.cpp:218] Iteration 89000 (3.75114 iter/s, 26.6586s/100 iters), loss = 4.22213
I0510 23:06:32.503738 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.22213 (* 1 = 4.22213 loss)
I0510 23:06:32.503751 39557 sgd_solver.cpp:105] Iteration 89000, lr = 0.01
I0510 23:06:59.188484 39557 solver.cpp:218] Iteration 89100 (3.74748 iter/s, 26.6846s/100 iters), loss = 3.62611
I0510 23:06:59.188544 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.62611 (* 1 = 3.62611 loss)
I0510 23:06:59.188555 39557 sgd_solver.cpp:105] Iteration 89100, lr = 0.01
I0510 23:07:25.790755 39557 solver.cpp:218] Iteration 89200 (3.75911 iter/s, 26.6021s/100 iters), loss = 3.91763
I0510 23:07:25.790907 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.91763 (* 1 = 3.91763 loss)
I0510 23:07:25.790921 39557 sgd_solver.cpp:105] Iteration 89200, lr = 0.01
I0510 23:07:52.409454 39557 solver.cpp:218] Iteration 89300 (3.7568 iter/s, 26.6184s/100 iters), loss = 3.84617
I0510 23:07:52.409512 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.84617 (* 1 = 3.84617 loss)
I0510 23:07:52.409525 39557 sgd_solver.cpp:105] Iteration 89300, lr = 0.01
I0510 23:08:19.033743 39557 solver.cpp:218] Iteration 89400 (3.756 iter/s, 26.6241s/100 iters), loss = 3.46081
I0510 23:08:19.033900 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.46081 (* 1 = 3.46081 loss)
I0510 23:08:19.033913 39557 sgd_solver.cpp:105] Iteration 89400, lr = 0.01
I0510 23:08:45.605569 39557 solver.cpp:218] Iteration 89500 (3.76343 iter/s, 26.5715s/100 iters), loss = 4.28996
I0510 23:08:45.605626 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.28996 (* 1 = 4.28996 loss)
I0510 23:08:45.605639 39557 sgd_solver.cpp:105] Iteration 89500, lr = 0.01
I0510 23:09:12.252183 39557 solver.cpp:218] Iteration 89600 (3.75285 iter/s, 26.6464s/100 iters), loss = 3.25942
I0510 23:09:12.252394 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.25942 (* 1 = 3.25942 loss)
I0510 23:09:12.252411 39557 sgd_solver.cpp:105] Iteration 89600, lr = 0.01
I0510 23:09:38.871039 39557 solver.cpp:218] Iteration 89700 (3.75678 iter/s, 26.6185s/100 iters), loss = 3.96866
I0510 23:09:38.871098 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.96866 (* 1 = 3.96866 loss)
I0510 23:09:38.871110 39557 sgd_solver.cpp:105] Iteration 89700, lr = 0.01
I0510 23:10:05.537842 39557 solver.cpp:218] Iteration 89800 (3.75001 iter/s, 26.6666s/100 iters), loss = 4.63126
I0510 23:10:05.538007 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.63126 (* 1 = 4.63126 loss)
I0510 23:10:05.538020 39557 sgd_solver.cpp:105] Iteration 89800, lr = 0.01
I0510 23:10:32.257568 39557 solver.cpp:218] Iteration 89900 (3.7426 iter/s, 26.7194s/100 iters), loss = 5.39392
I0510 23:10:32.257627 39557 solver.cpp:237]     Train net output #0: softmax_loss = 5.39392 (* 1 = 5.39392 loss)
I0510 23:10:32.257638 39557 sgd_solver.cpp:105] Iteration 89900, lr = 0.01
I0510 23:10:58.764111 39557 solver.cpp:447] Snapshotting to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_90000.caffemodel
I0510 23:10:58.803295 39557 sgd_solver.cpp:273] Snapshotting solver state to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_90000.solverstate
I0510 23:10:59.086987 39557 solver.cpp:218] Iteration 90000 (3.72728 iter/s, 26.8292s/100 iters), loss = 4.19942
I0510 23:10:59.087047 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.19942 (* 1 = 4.19942 loss)
I0510 23:10:59.087059 39557 sgd_solver.cpp:105] Iteration 90000, lr = 0.01
I0510 23:11:25.834067 39557 solver.cpp:218] Iteration 90100 (3.73875 iter/s, 26.7469s/100 iters), loss = 3.4482
I0510 23:11:25.834131 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.4482 (* 1 = 3.4482 loss)
I0510 23:11:25.834146 39557 sgd_solver.cpp:105] Iteration 90100, lr = 0.01
I0510 23:11:52.587802 39557 solver.cpp:218] Iteration 90200 (3.73782 iter/s, 26.7535s/100 iters), loss = 4.29732
I0510 23:11:52.587954 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.29732 (* 1 = 4.29732 loss)
I0510 23:11:52.587970 39557 sgd_solver.cpp:105] Iteration 90200, lr = 0.01
I0510 23:12:19.289808 39557 solver.cpp:218] Iteration 90300 (3.74508 iter/s, 26.7017s/100 iters), loss = 3.82825
I0510 23:12:19.289867 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.82825 (* 1 = 3.82825 loss)
I0510 23:12:19.289880 39557 sgd_solver.cpp:105] Iteration 90300, lr = 0.01
I0510 23:12:45.897366 39557 solver.cpp:218] Iteration 90400 (3.75836 iter/s, 26.6074s/100 iters), loss = 3.98685
I0510 23:12:45.897538 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.98685 (* 1 = 3.98685 loss)
I0510 23:12:45.897555 39557 sgd_solver.cpp:105] Iteration 90400, lr = 0.01
I0510 23:13:12.716397 39557 solver.cpp:218] Iteration 90500 (3.72874 iter/s, 26.8187s/100 iters), loss = 4.54626
I0510 23:13:12.716462 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.54626 (* 1 = 4.54626 loss)
I0510 23:13:12.716475 39557 sgd_solver.cpp:105] Iteration 90500, lr = 0.01
I0510 23:13:39.466779 39557 solver.cpp:218] Iteration 90600 (3.73829 iter/s, 26.7502s/100 iters), loss = 3.36558
I0510 23:13:39.466944 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.36558 (* 1 = 3.36558 loss)
I0510 23:13:39.466960 39557 sgd_solver.cpp:105] Iteration 90600, lr = 0.01
I0510 23:14:06.147933 39557 solver.cpp:218] Iteration 90700 (3.74801 iter/s, 26.6809s/100 iters), loss = 3.57277
I0510 23:14:06.147992 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.57277 (* 1 = 3.57277 loss)
I0510 23:14:06.148005 39557 sgd_solver.cpp:105] Iteration 90700, lr = 0.01
I0510 23:14:32.892166 39557 solver.cpp:218] Iteration 90800 (3.73915 iter/s, 26.744s/100 iters), loss = 4.19464
I0510 23:14:32.892328 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.19464 (* 1 = 4.19464 loss)
I0510 23:14:32.892344 39557 sgd_solver.cpp:105] Iteration 90800, lr = 0.01
I0510 23:14:59.572463 39557 solver.cpp:218] Iteration 90900 (3.74813 iter/s, 26.68s/100 iters), loss = 3.62806
I0510 23:14:59.572520 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.62806 (* 1 = 3.62806 loss)
I0510 23:14:59.572530 39557 sgd_solver.cpp:105] Iteration 90900, lr = 0.01
I0510 23:15:26.392027 39557 solver.cpp:218] Iteration 91000 (3.72865 iter/s, 26.8194s/100 iters), loss = 3.68756
I0510 23:15:26.392247 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.68756 (* 1 = 3.68756 loss)
I0510 23:15:26.392266 39557 sgd_solver.cpp:105] Iteration 91000, lr = 0.01
I0510 23:15:53.062136 39557 solver.cpp:218] Iteration 91100 (3.74957 iter/s, 26.6697s/100 iters), loss = 3.25321
I0510 23:15:53.062201 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.25321 (* 1 = 3.25321 loss)
I0510 23:15:53.062211 39557 sgd_solver.cpp:105] Iteration 91100, lr = 0.01
I0510 23:16:19.715636 39557 solver.cpp:218] Iteration 91200 (3.75188 iter/s, 26.6533s/100 iters), loss = 3.4214
I0510 23:16:19.715795 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.4214 (* 1 = 3.4214 loss)
I0510 23:16:19.715816 39557 sgd_solver.cpp:105] Iteration 91200, lr = 0.01
I0510 23:16:46.503221 39557 solver.cpp:218] Iteration 91300 (3.73311 iter/s, 26.7873s/100 iters), loss = 3.32067
I0510 23:16:46.503290 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.32067 (* 1 = 3.32067 loss)
I0510 23:16:46.503307 39557 sgd_solver.cpp:105] Iteration 91300, lr = 0.01
I0510 23:17:13.350184 39557 solver.cpp:218] Iteration 91400 (3.72485 iter/s, 26.8467s/100 iters), loss = 4.08514
I0510 23:17:13.350356 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.08514 (* 1 = 4.08514 loss)
I0510 23:17:13.350373 39557 sgd_solver.cpp:105] Iteration 91400, lr = 0.01
I0510 23:17:40.178398 39557 solver.cpp:218] Iteration 91500 (3.72746 iter/s, 26.8279s/100 iters), loss = 3.84518
I0510 23:17:40.178467 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.84518 (* 1 = 3.84518 loss)
I0510 23:17:40.178481 39557 sgd_solver.cpp:105] Iteration 91500, lr = 0.01
I0510 23:18:06.965536 39557 solver.cpp:218] Iteration 91600 (3.73316 iter/s, 26.7869s/100 iters), loss = 4.20818
I0510 23:18:06.965692 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.20818 (* 1 = 4.20818 loss)
I0510 23:18:06.965710 39557 sgd_solver.cpp:105] Iteration 91600, lr = 0.01
I0510 23:18:33.749449 39557 solver.cpp:218] Iteration 91700 (3.73363 iter/s, 26.7836s/100 iters), loss = 3.83931
I0510 23:18:33.749508 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.83931 (* 1 = 3.83931 loss)
I0510 23:18:33.749522 39557 sgd_solver.cpp:105] Iteration 91700, lr = 0.01
I0510 23:19:00.433042 39557 solver.cpp:218] Iteration 91800 (3.74765 iter/s, 26.6834s/100 iters), loss = 4.42953
I0510 23:19:00.433183 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.42953 (* 1 = 4.42953 loss)
I0510 23:19:00.433197 39557 sgd_solver.cpp:105] Iteration 91800, lr = 0.01
I0510 23:19:27.093386 39557 solver.cpp:218] Iteration 91900 (3.75093 iter/s, 26.6601s/100 iters), loss = 3.62611
I0510 23:19:27.093448 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.62611 (* 1 = 3.62611 loss)
I0510 23:19:27.093462 39557 sgd_solver.cpp:105] Iteration 91900, lr = 0.01
I0510 23:19:53.709830 39557 solver.cpp:218] Iteration 92000 (3.7571 iter/s, 26.6162s/100 iters), loss = 4.52618
I0510 23:19:53.710003 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.52618 (* 1 = 4.52618 loss)
I0510 23:19:53.710021 39557 sgd_solver.cpp:105] Iteration 92000, lr = 0.01
I0510 23:20:20.340384 39557 solver.cpp:218] Iteration 92100 (3.75513 iter/s, 26.6302s/100 iters), loss = 5.0488
I0510 23:20:20.340447 39557 solver.cpp:237]     Train net output #0: softmax_loss = 5.0488 (* 1 = 5.0488 loss)
I0510 23:20:20.340461 39557 sgd_solver.cpp:105] Iteration 92100, lr = 0.01
I0510 23:20:46.973294 39557 solver.cpp:218] Iteration 92200 (3.75478 iter/s, 26.6327s/100 iters), loss = 4.0541
I0510 23:20:46.973479 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.0541 (* 1 = 4.0541 loss)
I0510 23:20:46.973495 39557 sgd_solver.cpp:105] Iteration 92200, lr = 0.01
I0510 23:21:13.600455 39557 solver.cpp:218] Iteration 92300 (3.75561 iter/s, 26.6268s/100 iters), loss = 3.43062
I0510 23:21:13.600534 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.43062 (* 1 = 3.43062 loss)
I0510 23:21:13.600546 39557 sgd_solver.cpp:105] Iteration 92300, lr = 0.01
I0510 23:21:40.225090 39557 solver.cpp:218] Iteration 92400 (3.75595 iter/s, 26.6244s/100 iters), loss = 4.12451
I0510 23:21:40.225250 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.12451 (* 1 = 4.12451 loss)
I0510 23:21:40.225262 39557 sgd_solver.cpp:105] Iteration 92400, lr = 0.01
I0510 23:22:06.861729 39557 solver.cpp:218] Iteration 92500 (3.75427 iter/s, 26.6363s/100 iters), loss = 3.80139
I0510 23:22:06.861788 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.80139 (* 1 = 3.80139 loss)
I0510 23:22:06.861800 39557 sgd_solver.cpp:105] Iteration 92500, lr = 0.01
I0510 23:22:33.514714 39557 solver.cpp:218] Iteration 92600 (3.75195 iter/s, 26.6528s/100 iters), loss = 4.32529
I0510 23:22:33.514868 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.32529 (* 1 = 4.32529 loss)
I0510 23:22:33.514880 39557 sgd_solver.cpp:105] Iteration 92600, lr = 0.01
I0510 23:23:00.290555 39557 solver.cpp:218] Iteration 92700 (3.73475 iter/s, 26.7755s/100 iters), loss = 4.02645
I0510 23:23:00.290613 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.02645 (* 1 = 4.02645 loss)
I0510 23:23:00.290627 39557 sgd_solver.cpp:105] Iteration 92700, lr = 0.01
I0510 23:23:26.994638 39557 solver.cpp:218] Iteration 92800 (3.74477 iter/s, 26.7039s/100 iters), loss = 4.43601
I0510 23:23:26.994784 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.43601 (* 1 = 4.43601 loss)
I0510 23:23:26.994798 39557 sgd_solver.cpp:105] Iteration 92800, lr = 0.01
I0510 23:23:53.573309 39557 solver.cpp:218] Iteration 92900 (3.76246 iter/s, 26.5784s/100 iters), loss = 3.21416
I0510 23:23:53.573381 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.21416 (* 1 = 3.21416 loss)
I0510 23:23:53.573396 39557 sgd_solver.cpp:105] Iteration 92900, lr = 0.01
I0510 23:24:20.368788 39557 solver.cpp:218] Iteration 93000 (3.732 iter/s, 26.7953s/100 iters), loss = 3.97804
I0510 23:24:20.368949 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.97804 (* 1 = 3.97804 loss)
I0510 23:24:20.368965 39557 sgd_solver.cpp:105] Iteration 93000, lr = 0.01
I0510 23:24:46.972641 39557 solver.cpp:218] Iteration 93100 (3.75889 iter/s, 26.6036s/100 iters), loss = 3.08788
I0510 23:24:46.972718 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.08788 (* 1 = 3.08788 loss)
I0510 23:24:46.972730 39557 sgd_solver.cpp:105] Iteration 93100, lr = 0.01
I0510 23:25:13.672341 39557 solver.cpp:218] Iteration 93200 (3.74539 iter/s, 26.6995s/100 iters), loss = 3.1569
I0510 23:25:13.672500 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.1569 (* 1 = 3.1569 loss)
I0510 23:25:13.672513 39557 sgd_solver.cpp:105] Iteration 93200, lr = 0.01
I0510 23:25:40.301771 39557 solver.cpp:218] Iteration 93300 (3.75529 iter/s, 26.6291s/100 iters), loss = 4.02224
I0510 23:25:40.301839 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.02224 (* 1 = 4.02224 loss)
I0510 23:25:40.301856 39557 sgd_solver.cpp:105] Iteration 93300, lr = 0.01
I0510 23:26:06.986146 39557 solver.cpp:218] Iteration 93400 (3.74754 iter/s, 26.6842s/100 iters), loss = 4.16799
I0510 23:26:06.986292 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.16799 (* 1 = 4.16799 loss)
I0510 23:26:06.986306 39557 sgd_solver.cpp:105] Iteration 93400, lr = 0.01
I0510 23:26:33.661993 39557 solver.cpp:218] Iteration 93500 (3.74875 iter/s, 26.6756s/100 iters), loss = 4.16094
I0510 23:26:33.662050 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.16094 (* 1 = 4.16094 loss)
I0510 23:26:33.662060 39557 sgd_solver.cpp:105] Iteration 93500, lr = 0.01
I0510 23:27:00.336685 39557 solver.cpp:218] Iteration 93600 (3.7489 iter/s, 26.6745s/100 iters), loss = 4.39202
I0510 23:27:00.336879 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.39202 (* 1 = 4.39202 loss)
I0510 23:27:00.336894 39557 sgd_solver.cpp:105] Iteration 93600, lr = 0.01
I0510 23:27:26.996320 39557 solver.cpp:218] Iteration 93700 (3.75103 iter/s, 26.6593s/100 iters), loss = 3.36287
I0510 23:27:26.996378 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.36287 (* 1 = 3.36287 loss)
I0510 23:27:26.996388 39557 sgd_solver.cpp:105] Iteration 93700, lr = 0.01
I0510 23:27:53.587505 39557 solver.cpp:218] Iteration 93800 (3.76067 iter/s, 26.591s/100 iters), loss = 3.54622
I0510 23:27:53.587658 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.54622 (* 1 = 3.54622 loss)
I0510 23:27:53.587673 39557 sgd_solver.cpp:105] Iteration 93800, lr = 0.01
I0510 23:28:20.320744 39557 solver.cpp:218] Iteration 93900 (3.7407 iter/s, 26.733s/100 iters), loss = 4.85861
I0510 23:28:20.320802 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.85861 (* 1 = 4.85861 loss)
I0510 23:28:20.320813 39557 sgd_solver.cpp:105] Iteration 93900, lr = 0.01
I0510 23:28:47.002734 39557 solver.cpp:218] Iteration 94000 (3.74787 iter/s, 26.6818s/100 iters), loss = 4.18284
I0510 23:28:47.002889 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.18284 (* 1 = 4.18284 loss)
I0510 23:28:47.002904 39557 sgd_solver.cpp:105] Iteration 94000, lr = 0.01
I0510 23:29:13.648144 39557 solver.cpp:218] Iteration 94100 (3.75303 iter/s, 26.6451s/100 iters), loss = 3.30747
I0510 23:29:13.648211 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.30747 (* 1 = 3.30747 loss)
I0510 23:29:13.648222 39557 sgd_solver.cpp:105] Iteration 94100, lr = 0.01
I0510 23:29:40.310153 39557 solver.cpp:218] Iteration 94200 (3.75068 iter/s, 26.6618s/100 iters), loss = 3.07442
I0510 23:29:40.310317 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.07442 (* 1 = 3.07442 loss)
I0510 23:29:40.310333 39557 sgd_solver.cpp:105] Iteration 94200, lr = 0.01
I0510 23:30:06.980448 39557 solver.cpp:218] Iteration 94300 (3.74953 iter/s, 26.67s/100 iters), loss = 4.00114
I0510 23:30:06.980506 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.00114 (* 1 = 4.00114 loss)
I0510 23:30:06.980516 39557 sgd_solver.cpp:105] Iteration 94300, lr = 0.01
I0510 23:30:33.591163 39557 solver.cpp:218] Iteration 94400 (3.75791 iter/s, 26.6105s/100 iters), loss = 3.75357
I0510 23:30:33.591343 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.75357 (* 1 = 3.75357 loss)
I0510 23:30:33.591358 39557 sgd_solver.cpp:105] Iteration 94400, lr = 0.01
I0510 23:31:00.241971 39557 solver.cpp:218] Iteration 94500 (3.75228 iter/s, 26.6505s/100 iters), loss = 4.52094
I0510 23:31:00.242029 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.52094 (* 1 = 4.52094 loss)
I0510 23:31:00.242039 39557 sgd_solver.cpp:105] Iteration 94500, lr = 0.01
I0510 23:31:26.845571 39557 solver.cpp:218] Iteration 94600 (3.75892 iter/s, 26.6034s/100 iters), loss = 3.31442
I0510 23:31:26.845729 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.31442 (* 1 = 3.31442 loss)
I0510 23:31:26.845743 39557 sgd_solver.cpp:105] Iteration 94600, lr = 0.01
I0510 23:31:53.395987 39557 solver.cpp:218] Iteration 94700 (3.76646 iter/s, 26.5501s/100 iters), loss = 3.95925
I0510 23:31:53.396046 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.95925 (* 1 = 3.95925 loss)
I0510 23:31:53.396056 39557 sgd_solver.cpp:105] Iteration 94700, lr = 0.01
I0510 23:32:20.004313 39557 solver.cpp:218] Iteration 94800 (3.75825 iter/s, 26.6081s/100 iters), loss = 4.22897
I0510 23:32:20.004469 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.22897 (* 1 = 4.22897 loss)
I0510 23:32:20.004483 39557 sgd_solver.cpp:105] Iteration 94800, lr = 0.01
I0510 23:32:46.559801 39557 solver.cpp:218] Iteration 94900 (3.76574 iter/s, 26.5552s/100 iters), loss = 4.62404
I0510 23:32:46.559861 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.62404 (* 1 = 4.62404 loss)
I0510 23:32:46.559871 39557 sgd_solver.cpp:105] Iteration 94900, lr = 0.01
I0510 23:33:12.921869 39557 solver.cpp:447] Snapshotting to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_95000.caffemodel
I0510 23:33:12.961709 39557 sgd_solver.cpp:273] Snapshotting solver state to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_95000.solverstate
I0510 23:33:13.243365 39557 solver.cpp:218] Iteration 95000 (3.74765 iter/s, 26.6834s/100 iters), loss = 3.92879
I0510 23:33:13.243422 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.92879 (* 1 = 3.92879 loss)
I0510 23:33:13.243432 39557 sgd_solver.cpp:105] Iteration 95000, lr = 0.01
I0510 23:33:40.033726 39557 solver.cpp:218] Iteration 95100 (3.73271 iter/s, 26.7902s/100 iters), loss = 4.12075
I0510 23:33:40.033795 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.12075 (* 1 = 4.12075 loss)
I0510 23:33:40.033807 39557 sgd_solver.cpp:105] Iteration 95100, lr = 0.01
I0510 23:34:06.788594 39557 solver.cpp:218] Iteration 95200 (3.73767 iter/s, 26.7547s/100 iters), loss = 3.96637
I0510 23:34:06.788771 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.96637 (* 1 = 3.96637 loss)
I0510 23:34:06.788787 39557 sgd_solver.cpp:105] Iteration 95200, lr = 0.01
I0510 23:34:33.488060 39557 solver.cpp:218] Iteration 95300 (3.74544 iter/s, 26.6991s/100 iters), loss = 3.72185
I0510 23:34:33.488123 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.72185 (* 1 = 3.72185 loss)
I0510 23:34:33.488138 39557 sgd_solver.cpp:105] Iteration 95300, lr = 0.01
I0510 23:35:00.260704 39557 solver.cpp:218] Iteration 95400 (3.73517 iter/s, 26.7725s/100 iters), loss = 4.20868
I0510 23:35:00.260859 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.20868 (* 1 = 4.20868 loss)
I0510 23:35:00.260874 39557 sgd_solver.cpp:105] Iteration 95400, lr = 0.01
I0510 23:35:26.926792 39557 solver.cpp:218] Iteration 95500 (3.75009 iter/s, 26.666s/100 iters), loss = 3.8812
I0510 23:35:26.926856 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.8812 (* 1 = 3.8812 loss)
I0510 23:35:26.926867 39557 sgd_solver.cpp:105] Iteration 95500, lr = 0.01
I0510 23:35:53.555194 39557 solver.cpp:218] Iteration 95600 (3.75539 iter/s, 26.6284s/100 iters), loss = 4.18993
I0510 23:35:53.555369 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.18993 (* 1 = 4.18993 loss)
I0510 23:35:53.555384 39557 sgd_solver.cpp:105] Iteration 95600, lr = 0.01
I0510 23:36:20.213258 39557 solver.cpp:218] Iteration 95700 (3.75123 iter/s, 26.658s/100 iters), loss = 3.34411
I0510 23:36:20.213322 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.34411 (* 1 = 3.34411 loss)
I0510 23:36:20.213335 39557 sgd_solver.cpp:105] Iteration 95700, lr = 0.01
I0510 23:36:46.842126 39557 solver.cpp:218] Iteration 95800 (3.75532 iter/s, 26.6289s/100 iters), loss = 4.35388
I0510 23:36:46.842296 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.35388 (* 1 = 4.35388 loss)
I0510 23:36:46.842310 39557 sgd_solver.cpp:105] Iteration 95800, lr = 0.01
I0510 23:37:13.442852 39557 solver.cpp:218] Iteration 95900 (3.75931 iter/s, 26.6006s/100 iters), loss = 3.48265
I0510 23:37:13.442908 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.48265 (* 1 = 3.48265 loss)
I0510 23:37:13.442917 39557 sgd_solver.cpp:105] Iteration 95900, lr = 0.01
I0510 23:37:40.048601 39557 solver.cpp:218] Iteration 96000 (3.75859 iter/s, 26.6057s/100 iters), loss = 4.33333
I0510 23:37:40.048768 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.33333 (* 1 = 4.33333 loss)
I0510 23:37:40.048784 39557 sgd_solver.cpp:105] Iteration 96000, lr = 0.01
I0510 23:38:06.726104 39557 solver.cpp:218] Iteration 96100 (3.7485 iter/s, 26.6774s/100 iters), loss = 3.46908
I0510 23:38:06.726162 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.46908 (* 1 = 3.46908 loss)
I0510 23:38:06.726176 39557 sgd_solver.cpp:105] Iteration 96100, lr = 0.01
I0510 23:38:33.346485 39557 solver.cpp:218] Iteration 96200 (3.75652 iter/s, 26.6203s/100 iters), loss = 3.6427
I0510 23:38:33.346693 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.6427 (* 1 = 3.6427 loss)
I0510 23:38:33.346709 39557 sgd_solver.cpp:105] Iteration 96200, lr = 0.01
I0510 23:38:59.984980 39557 solver.cpp:218] Iteration 96300 (3.75399 iter/s, 26.6383s/100 iters), loss = 3.60751
I0510 23:38:59.985047 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.60751 (* 1 = 3.60751 loss)
I0510 23:38:59.985060 39557 sgd_solver.cpp:105] Iteration 96300, lr = 0.01
I0510 23:39:26.615978 39557 solver.cpp:218] Iteration 96400 (3.75503 iter/s, 26.631s/100 iters), loss = 3.93655
I0510 23:39:26.616192 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.93655 (* 1 = 3.93655 loss)
I0510 23:39:26.616215 39557 sgd_solver.cpp:105] Iteration 96400, lr = 0.01
I0510 23:39:53.218082 39557 solver.cpp:218] Iteration 96500 (3.75913 iter/s, 26.6019s/100 iters), loss = 4.56457
I0510 23:39:53.218140 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.56457 (* 1 = 4.56457 loss)
I0510 23:39:53.218170 39557 sgd_solver.cpp:105] Iteration 96500, lr = 0.01
I0510 23:40:19.918313 39557 solver.cpp:218] Iteration 96600 (3.74529 iter/s, 26.7002s/100 iters), loss = 3.96372
I0510 23:40:19.918488 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.96372 (* 1 = 3.96372 loss)
I0510 23:40:19.918506 39557 sgd_solver.cpp:105] Iteration 96600, lr = 0.01
I0510 23:40:46.685494 39557 solver.cpp:218] Iteration 96700 (3.73594 iter/s, 26.767s/100 iters), loss = 3.91169
I0510 23:40:46.685554 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.91169 (* 1 = 3.91169 loss)
I0510 23:40:46.685569 39557 sgd_solver.cpp:105] Iteration 96700, lr = 0.01
I0510 23:41:13.395748 39557 solver.cpp:218] Iteration 96800 (3.74389 iter/s, 26.7102s/100 iters), loss = 4.11392
I0510 23:41:13.395921 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.11392 (* 1 = 4.11392 loss)
I0510 23:41:13.395936 39557 sgd_solver.cpp:105] Iteration 96800, lr = 0.01
I0510 23:41:40.013190 39557 solver.cpp:218] Iteration 96900 (3.75696 iter/s, 26.6173s/100 iters), loss = 3.50706
I0510 23:41:40.013255 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.50706 (* 1 = 3.50706 loss)
I0510 23:41:40.013267 39557 sgd_solver.cpp:105] Iteration 96900, lr = 0.01
I0510 23:42:06.728565 39557 solver.cpp:218] Iteration 97000 (3.74317 iter/s, 26.7153s/100 iters), loss = 3.34162
I0510 23:42:06.728726 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.34162 (* 1 = 3.34162 loss)
I0510 23:42:06.728744 39557 sgd_solver.cpp:105] Iteration 97000, lr = 0.01
I0510 23:42:33.390674 39557 solver.cpp:218] Iteration 97100 (3.75067 iter/s, 26.6619s/100 iters), loss = 4.3123
I0510 23:42:33.390733 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.3123 (* 1 = 4.3123 loss)
I0510 23:42:33.390744 39557 sgd_solver.cpp:105] Iteration 97100, lr = 0.01
I0510 23:43:00.037757 39557 solver.cpp:218] Iteration 97200 (3.75277 iter/s, 26.647s/100 iters), loss = 4.0599
I0510 23:43:00.037912 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.0599 (* 1 = 4.0599 loss)
I0510 23:43:00.037927 39557 sgd_solver.cpp:105] Iteration 97200, lr = 0.01
I0510 23:43:26.626488 39557 solver.cpp:218] Iteration 97300 (3.76102 iter/s, 26.5886s/100 iters), loss = 3.95306
I0510 23:43:26.626549 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.95306 (* 1 = 3.95306 loss)
I0510 23:43:26.626562 39557 sgd_solver.cpp:105] Iteration 97300, lr = 0.01
I0510 23:43:53.220377 39557 solver.cpp:218] Iteration 97400 (3.76028 iter/s, 26.5938s/100 iters), loss = 3.98521
I0510 23:43:53.220546 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.98521 (* 1 = 3.98521 loss)
I0510 23:43:53.220558 39557 sgd_solver.cpp:105] Iteration 97400, lr = 0.01
I0510 23:44:19.817795 39557 solver.cpp:218] Iteration 97500 (3.75979 iter/s, 26.5972s/100 iters), loss = 3.07075
I0510 23:44:19.817862 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.07075 (* 1 = 3.07075 loss)
I0510 23:44:19.817874 39557 sgd_solver.cpp:105] Iteration 97500, lr = 0.01
I0510 23:44:46.515774 39557 solver.cpp:218] Iteration 97600 (3.74562 iter/s, 26.6979s/100 iters), loss = 3.89075
I0510 23:44:46.515944 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.89075 (* 1 = 3.89075 loss)
I0510 23:44:46.515957 39557 sgd_solver.cpp:105] Iteration 97600, lr = 0.01
I0510 23:45:13.202481 39557 solver.cpp:218] Iteration 97700 (3.74721 iter/s, 26.6865s/100 iters), loss = 3.50402
I0510 23:45:13.202543 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.50402 (* 1 = 3.50402 loss)
I0510 23:45:13.202558 39557 sgd_solver.cpp:105] Iteration 97700, lr = 0.01
I0510 23:45:39.815238 39557 solver.cpp:218] Iteration 97800 (3.75761 iter/s, 26.6126s/100 iters), loss = 4.41283
I0510 23:45:39.815388 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.41283 (* 1 = 4.41283 loss)
I0510 23:45:39.815402 39557 sgd_solver.cpp:105] Iteration 97800, lr = 0.01
I0510 23:46:06.365895 39557 solver.cpp:218] Iteration 97900 (3.76641 iter/s, 26.5505s/100 iters), loss = 2.94314
I0510 23:46:06.365952 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.94314 (* 1 = 2.94314 loss)
I0510 23:46:06.365964 39557 sgd_solver.cpp:105] Iteration 97900, lr = 0.01
I0510 23:46:32.929446 39557 solver.cpp:218] Iteration 98000 (3.76457 iter/s, 26.5634s/100 iters), loss = 3.1955
I0510 23:46:32.929615 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.1955 (* 1 = 3.1955 loss)
I0510 23:46:32.929627 39557 sgd_solver.cpp:105] Iteration 98000, lr = 0.01
I0510 23:46:59.537477 39557 solver.cpp:218] Iteration 98100 (3.75829 iter/s, 26.6078s/100 iters), loss = 3.74875
I0510 23:46:59.537533 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.74875 (* 1 = 3.74875 loss)
I0510 23:46:59.537549 39557 sgd_solver.cpp:105] Iteration 98100, lr = 0.01
I0510 23:47:26.162261 39557 solver.cpp:218] Iteration 98200 (3.75591 iter/s, 26.6247s/100 iters), loss = 3.204
I0510 23:47:26.162415 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.204 (* 1 = 3.204 loss)
I0510 23:47:26.162427 39557 sgd_solver.cpp:105] Iteration 98200, lr = 0.01
I0510 23:47:52.720108 39557 solver.cpp:218] Iteration 98300 (3.7654 iter/s, 26.5576s/100 iters), loss = 3.25215
I0510 23:47:52.720165 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.25215 (* 1 = 3.25215 loss)
I0510 23:47:52.720177 39557 sgd_solver.cpp:105] Iteration 98300, lr = 0.01
I0510 23:48:19.383541 39557 solver.cpp:218] Iteration 98400 (3.75047 iter/s, 26.6633s/100 iters), loss = 3.29274
I0510 23:48:19.383709 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.29274 (* 1 = 3.29274 loss)
I0510 23:48:19.383724 39557 sgd_solver.cpp:105] Iteration 98400, lr = 0.01
I0510 23:48:45.950913 39557 solver.cpp:218] Iteration 98500 (3.76405 iter/s, 26.5671s/100 iters), loss = 3.57991
I0510 23:48:45.950971 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.57991 (* 1 = 3.57991 loss)
I0510 23:48:45.950986 39557 sgd_solver.cpp:105] Iteration 98500, lr = 0.01
I0510 23:49:12.609050 39557 solver.cpp:218] Iteration 98600 (3.75122 iter/s, 26.658s/100 iters), loss = 4.16559
I0510 23:49:12.609230 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.16559 (* 1 = 4.16559 loss)
I0510 23:49:12.609246 39557 sgd_solver.cpp:105] Iteration 98600, lr = 0.01
I0510 23:49:39.233404 39557 solver.cpp:218] Iteration 98700 (3.75599 iter/s, 26.6241s/100 iters), loss = 3.67726
I0510 23:49:39.233460 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.67726 (* 1 = 3.67726 loss)
I0510 23:49:39.233472 39557 sgd_solver.cpp:105] Iteration 98700, lr = 0.01
I0510 23:50:05.873267 39557 solver.cpp:218] Iteration 98800 (3.75379 iter/s, 26.6397s/100 iters), loss = 3.15351
I0510 23:50:05.873433 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.15351 (* 1 = 3.15351 loss)
I0510 23:50:05.873448 39557 sgd_solver.cpp:105] Iteration 98800, lr = 0.01
I0510 23:50:32.545266 39557 solver.cpp:218] Iteration 98900 (3.74928 iter/s, 26.6718s/100 iters), loss = 3.03699
I0510 23:50:32.545326 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.03699 (* 1 = 3.03699 loss)
I0510 23:50:32.545341 39557 sgd_solver.cpp:105] Iteration 98900, lr = 0.01
I0510 23:50:59.215057 39557 solver.cpp:218] Iteration 99000 (3.74958 iter/s, 26.6697s/100 iters), loss = 3.54949
I0510 23:50:59.215225 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.54949 (* 1 = 3.54949 loss)
I0510 23:50:59.215240 39557 sgd_solver.cpp:105] Iteration 99000, lr = 0.01
I0510 23:51:25.929134 39557 solver.cpp:218] Iteration 99100 (3.74338 iter/s, 26.7138s/100 iters), loss = 4.35028
I0510 23:51:25.929196 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.35028 (* 1 = 4.35028 loss)
I0510 23:51:25.929209 39557 sgd_solver.cpp:105] Iteration 99100, lr = 0.01
I0510 23:51:52.685506 39557 solver.cpp:218] Iteration 99200 (3.73745 iter/s, 26.7562s/100 iters), loss = 4.75828
I0510 23:51:52.685667 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.75828 (* 1 = 4.75828 loss)
I0510 23:51:52.685680 39557 sgd_solver.cpp:105] Iteration 99200, lr = 0.01
I0510 23:52:19.298756 39557 solver.cpp:218] Iteration 99300 (3.75756 iter/s, 26.613s/100 iters), loss = 3.63005
I0510 23:52:19.298818 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.63005 (* 1 = 3.63005 loss)
I0510 23:52:19.298831 39557 sgd_solver.cpp:105] Iteration 99300, lr = 0.01
I0510 23:52:45.973076 39557 solver.cpp:218] Iteration 99400 (3.74894 iter/s, 26.6742s/100 iters), loss = 3.91312
I0510 23:52:45.973227 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.91312 (* 1 = 3.91312 loss)
I0510 23:52:45.973253 39557 sgd_solver.cpp:105] Iteration 99400, lr = 0.01
I0510 23:53:12.642565 39557 solver.cpp:218] Iteration 99500 (3.74963 iter/s, 26.6693s/100 iters), loss = 4.21162
I0510 23:53:12.642629 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.21162 (* 1 = 4.21162 loss)
I0510 23:53:12.642640 39557 sgd_solver.cpp:105] Iteration 99500, lr = 0.01
I0510 23:53:39.326654 39557 solver.cpp:218] Iteration 99600 (3.74757 iter/s, 26.6839s/100 iters), loss = 3.6782
I0510 23:53:39.326807 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.6782 (* 1 = 3.6782 loss)
I0510 23:53:39.326822 39557 sgd_solver.cpp:105] Iteration 99600, lr = 0.01
I0510 23:54:05.975666 39557 solver.cpp:218] Iteration 99700 (3.75252 iter/s, 26.6488s/100 iters), loss = 3.79974
I0510 23:54:05.975723 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.79974 (* 1 = 3.79974 loss)
I0510 23:54:05.975733 39557 sgd_solver.cpp:105] Iteration 99700, lr = 0.01
I0510 23:54:32.596295 39557 solver.cpp:218] Iteration 99800 (3.75651 iter/s, 26.6205s/100 iters), loss = 3.68735
I0510 23:54:32.596449 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.68735 (* 1 = 3.68735 loss)
I0510 23:54:32.596464 39557 sgd_solver.cpp:105] Iteration 99800, lr = 0.01
I0510 23:54:59.241861 39557 solver.cpp:218] Iteration 99900 (3.753 iter/s, 26.6453s/100 iters), loss = 4.5192
I0510 23:54:59.241924 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.5192 (* 1 = 4.5192 loss)
I0510 23:54:59.241936 39557 sgd_solver.cpp:105] Iteration 99900, lr = 0.01
I0510 23:55:25.588973 39557 solver.cpp:447] Snapshotting to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_100000.caffemodel
I0510 23:55:25.628687 39557 sgd_solver.cpp:273] Snapshotting solver state to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_100000.solverstate
I0510 23:55:25.912591 39557 solver.cpp:218] Iteration 100000 (3.74945 iter/s, 26.6706s/100 iters), loss = 3.27654
I0510 23:55:25.912659 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.27654 (* 1 = 3.27654 loss)
I0510 23:55:25.912669 39557 sgd_solver.cpp:105] Iteration 100000, lr = 0.01
I0510 23:55:52.676120 39557 solver.cpp:218] Iteration 100100 (3.73645 iter/s, 26.7634s/100 iters), loss = 4.0266
I0510 23:55:52.676175 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.0266 (* 1 = 4.0266 loss)
I0510 23:55:52.676187 39557 sgd_solver.cpp:105] Iteration 100100, lr = 0.01
I0510 23:56:19.310847 39557 solver.cpp:218] Iteration 100200 (3.75452 iter/s, 26.6346s/100 iters), loss = 4.40221
I0510 23:56:19.311067 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.40221 (* 1 = 4.40221 loss)
I0510 23:56:19.311084 39557 sgd_solver.cpp:105] Iteration 100200, lr = 0.01
I0510 23:56:45.981770 39557 solver.cpp:218] Iteration 100300 (3.74944 iter/s, 26.6706s/100 iters), loss = 3.83932
I0510 23:56:45.981837 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.83932 (* 1 = 3.83932 loss)
I0510 23:56:45.981851 39557 sgd_solver.cpp:105] Iteration 100300, lr = 0.01
I0510 23:57:12.568547 39557 solver.cpp:218] Iteration 100400 (3.76129 iter/s, 26.5866s/100 iters), loss = 3.51547
I0510 23:57:12.568706 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.51547 (* 1 = 3.51547 loss)
I0510 23:57:12.568719 39557 sgd_solver.cpp:105] Iteration 100400, lr = 0.01
I0510 23:57:39.218242 39557 solver.cpp:218] Iteration 100500 (3.75242 iter/s, 26.6494s/100 iters), loss = 4.09083
I0510 23:57:39.218304 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.09083 (* 1 = 4.09083 loss)
I0510 23:57:39.218314 39557 sgd_solver.cpp:105] Iteration 100500, lr = 0.01
I0510 23:58:05.874660 39557 solver.cpp:218] Iteration 100600 (3.75146 iter/s, 26.6563s/100 iters), loss = 3.8387
I0510 23:58:05.874814 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.8387 (* 1 = 3.8387 loss)
I0510 23:58:05.874828 39557 sgd_solver.cpp:105] Iteration 100600, lr = 0.01
I0510 23:58:32.541036 39557 solver.cpp:218] Iteration 100700 (3.75008 iter/s, 26.6661s/100 iters), loss = 3.5818
I0510 23:58:32.541102 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.5818 (* 1 = 3.5818 loss)
I0510 23:58:32.541115 39557 sgd_solver.cpp:105] Iteration 100700, lr = 0.01
I0510 23:58:59.301192 39557 solver.cpp:218] Iteration 100800 (3.73692 iter/s, 26.76s/100 iters), loss = 4.16178
I0510 23:58:59.301383 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.16178 (* 1 = 4.16178 loss)
I0510 23:58:59.301398 39557 sgd_solver.cpp:105] Iteration 100800, lr = 0.01
I0510 23:59:26.035713 39557 solver.cpp:218] Iteration 100900 (3.74052 iter/s, 26.7342s/100 iters), loss = 3.73406
I0510 23:59:26.035780 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.73406 (* 1 = 3.73406 loss)
I0510 23:59:26.035795 39557 sgd_solver.cpp:105] Iteration 100900, lr = 0.01
I0510 23:59:52.789415 39557 solver.cpp:218] Iteration 101000 (3.73782 iter/s, 26.7535s/100 iters), loss = 3.82192
I0510 23:59:52.789592 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.82192 (* 1 = 3.82192 loss)
I0510 23:59:52.789609 39557 sgd_solver.cpp:105] Iteration 101000, lr = 0.01
I0511 00:00:19.474475 39557 solver.cpp:218] Iteration 101100 (3.74745 iter/s, 26.6848s/100 iters), loss = 3.87648
I0511 00:00:19.474540 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.87648 (* 1 = 3.87648 loss)
I0511 00:00:19.474555 39557 sgd_solver.cpp:105] Iteration 101100, lr = 0.01
I0511 00:00:46.090353 39557 solver.cpp:218] Iteration 101200 (3.75718 iter/s, 26.6157s/100 iters), loss = 3.58738
I0511 00:00:46.090529 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.58738 (* 1 = 3.58738 loss)
I0511 00:00:46.090544 39557 sgd_solver.cpp:105] Iteration 101200, lr = 0.01
I0511 00:01:12.827024 39557 solver.cpp:218] Iteration 101300 (3.74022 iter/s, 26.7364s/100 iters), loss = 4.29458
I0511 00:01:12.827087 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.29458 (* 1 = 4.29458 loss)
I0511 00:01:12.827100 39557 sgd_solver.cpp:105] Iteration 101300, lr = 0.01
I0511 00:01:39.445500 39557 solver.cpp:218] Iteration 101400 (3.75681 iter/s, 26.6183s/100 iters), loss = 4.19248
I0511 00:01:39.445673 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.19248 (* 1 = 4.19248 loss)
I0511 00:01:39.445689 39557 sgd_solver.cpp:105] Iteration 101400, lr = 0.01
I0511 00:02:06.047358 39557 solver.cpp:218] Iteration 101500 (3.75918 iter/s, 26.6016s/100 iters), loss = 3.78258
I0511 00:02:06.047418 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.78258 (* 1 = 3.78258 loss)
I0511 00:02:06.047430 39557 sgd_solver.cpp:105] Iteration 101500, lr = 0.01
I0511 00:02:32.703938 39557 solver.cpp:218] Iteration 101600 (3.75144 iter/s, 26.6564s/100 iters), loss = 3.69154
I0511 00:02:32.704130 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.69154 (* 1 = 3.69154 loss)
I0511 00:02:32.704146 39557 sgd_solver.cpp:105] Iteration 101600, lr = 0.01
I0511 00:02:59.429158 39557 solver.cpp:218] Iteration 101700 (3.74182 iter/s, 26.7249s/100 iters), loss = 3.05797
I0511 00:02:59.429221 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.05797 (* 1 = 3.05797 loss)
I0511 00:02:59.429234 39557 sgd_solver.cpp:105] Iteration 101700, lr = 0.01
I0511 00:03:26.221225 39557 solver.cpp:218] Iteration 101800 (3.73247 iter/s, 26.7919s/100 iters), loss = 4.19627
I0511 00:03:26.221395 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.19627 (* 1 = 4.19627 loss)
I0511 00:03:26.221410 39557 sgd_solver.cpp:105] Iteration 101800, lr = 0.01
I0511 00:03:53.038435 39557 solver.cpp:218] Iteration 101900 (3.72899 iter/s, 26.8169s/100 iters), loss = 3.75093
I0511 00:03:53.038499 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.75093 (* 1 = 3.75093 loss)
I0511 00:03:53.038511 39557 sgd_solver.cpp:105] Iteration 101900, lr = 0.01
I0511 00:04:19.772441 39557 solver.cpp:218] Iteration 102000 (3.74058 iter/s, 26.7338s/100 iters), loss = 3.38807
I0511 00:04:19.772735 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.38807 (* 1 = 3.38807 loss)
I0511 00:04:19.772796 39557 sgd_solver.cpp:105] Iteration 102000, lr = 0.01
I0511 00:04:46.532811 39557 solver.cpp:218] Iteration 102100 (3.73692 iter/s, 26.76s/100 iters), loss = 3.24584
I0511 00:04:46.532866 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.24584 (* 1 = 3.24584 loss)
I0511 00:04:46.532881 39557 sgd_solver.cpp:105] Iteration 102100, lr = 0.01
I0511 00:05:13.255939 39557 solver.cpp:218] Iteration 102200 (3.7421 iter/s, 26.723s/100 iters), loss = 3.5621
I0511 00:05:13.256098 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.5621 (* 1 = 3.5621 loss)
I0511 00:05:13.256114 39557 sgd_solver.cpp:105] Iteration 102200, lr = 0.01
I0511 00:05:40.082607 39557 solver.cpp:218] Iteration 102300 (3.72767 iter/s, 26.8264s/100 iters), loss = 3.3144
I0511 00:05:40.082662 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.3144 (* 1 = 3.3144 loss)
I0511 00:05:40.082676 39557 sgd_solver.cpp:105] Iteration 102300, lr = 0.01
I0511 00:06:06.818495 39557 solver.cpp:218] Iteration 102400 (3.74031 iter/s, 26.7357s/100 iters), loss = 4.08277
I0511 00:06:06.818666 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.08277 (* 1 = 4.08277 loss)
I0511 00:06:06.818683 39557 sgd_solver.cpp:105] Iteration 102400, lr = 0.01
I0511 00:06:33.489857 39557 solver.cpp:218] Iteration 102500 (3.74938 iter/s, 26.6711s/100 iters), loss = 4.08064
I0511 00:06:33.489913 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.08064 (* 1 = 4.08064 loss)
I0511 00:06:33.489928 39557 sgd_solver.cpp:105] Iteration 102500, lr = 0.01
I0511 00:07:00.289077 39557 solver.cpp:218] Iteration 102600 (3.73147 iter/s, 26.7991s/100 iters), loss = 3.55577
I0511 00:07:00.289232 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.55577 (* 1 = 3.55577 loss)
I0511 00:07:00.289259 39557 sgd_solver.cpp:105] Iteration 102600, lr = 0.01
I0511 00:07:27.019614 39557 solver.cpp:218] Iteration 102700 (3.74107 iter/s, 26.7303s/100 iters), loss = 4.45529
I0511 00:07:27.019706 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.45529 (* 1 = 4.45529 loss)
I0511 00:07:27.019721 39557 sgd_solver.cpp:105] Iteration 102700, lr = 0.01
I0511 00:07:53.825295 39557 solver.cpp:218] Iteration 102800 (3.73058 iter/s, 26.8055s/100 iters), loss = 3.25159
I0511 00:07:53.825462 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.25159 (* 1 = 3.25159 loss)
I0511 00:07:53.825476 39557 sgd_solver.cpp:105] Iteration 102800, lr = 0.01
I0511 00:08:20.511479 39557 solver.cpp:218] Iteration 102900 (3.74729 iter/s, 26.6859s/100 iters), loss = 4.62682
I0511 00:08:20.511538 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.62682 (* 1 = 4.62682 loss)
I0511 00:08:20.511553 39557 sgd_solver.cpp:105] Iteration 102900, lr = 0.01
I0511 00:08:47.311853 39557 solver.cpp:218] Iteration 103000 (3.73131 iter/s, 26.8002s/100 iters), loss = 3.6147
I0511 00:08:47.312041 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.6147 (* 1 = 3.6147 loss)
I0511 00:08:47.312055 39557 sgd_solver.cpp:105] Iteration 103000, lr = 0.01
I0511 00:09:14.183336 39557 solver.cpp:218] Iteration 103100 (3.72146 iter/s, 26.8712s/100 iters), loss = 4.13836
I0511 00:09:14.183394 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.13836 (* 1 = 4.13836 loss)
I0511 00:09:14.183408 39557 sgd_solver.cpp:105] Iteration 103100, lr = 0.01
I0511 00:09:40.975754 39557 solver.cpp:218] Iteration 103200 (3.73243 iter/s, 26.7922s/100 iters), loss = 3.40069
I0511 00:09:40.975917 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.40069 (* 1 = 3.40069 loss)
I0511 00:09:40.975931 39557 sgd_solver.cpp:105] Iteration 103200, lr = 0.01
I0511 00:10:07.810593 39557 solver.cpp:218] Iteration 103300 (3.72654 iter/s, 26.8345s/100 iters), loss = 3.64898
I0511 00:10:07.810652 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.64898 (* 1 = 3.64898 loss)
I0511 00:10:07.810665 39557 sgd_solver.cpp:105] Iteration 103300, lr = 0.01
I0511 00:10:34.639902 39557 solver.cpp:218] Iteration 103400 (3.7273 iter/s, 26.8291s/100 iters), loss = 3.11966
I0511 00:10:34.640080 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.11966 (* 1 = 3.11966 loss)
I0511 00:10:34.640097 39557 sgd_solver.cpp:105] Iteration 103400, lr = 0.01
I0511 00:11:01.440774 39557 solver.cpp:218] Iteration 103500 (3.73127 iter/s, 26.8005s/100 iters), loss = 4.07627
I0511 00:11:01.440836 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.07627 (* 1 = 4.07627 loss)
I0511 00:11:01.440851 39557 sgd_solver.cpp:105] Iteration 103500, lr = 0.01
I0511 00:11:28.198405 39557 solver.cpp:218] Iteration 103600 (3.73728 iter/s, 26.7574s/100 iters), loss = 3.92923
I0511 00:11:28.198570 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.92923 (* 1 = 3.92923 loss)
I0511 00:11:28.198585 39557 sgd_solver.cpp:105] Iteration 103600, lr = 0.01
I0511 00:11:54.979579 39557 solver.cpp:218] Iteration 103700 (3.73401 iter/s, 26.7809s/100 iters), loss = 3.63108
I0511 00:11:54.979641 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.63108 (* 1 = 3.63108 loss)
I0511 00:11:54.979657 39557 sgd_solver.cpp:105] Iteration 103700, lr = 0.01
I0511 00:12:21.734499 39557 solver.cpp:218] Iteration 103800 (3.73766 iter/s, 26.7547s/100 iters), loss = 3.96585
I0511 00:12:21.734663 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.96585 (* 1 = 3.96585 loss)
I0511 00:12:21.734679 39557 sgd_solver.cpp:105] Iteration 103800, lr = 0.01
I0511 00:12:48.576611 39557 solver.cpp:218] Iteration 103900 (3.72553 iter/s, 26.8418s/100 iters), loss = 3.55739
I0511 00:12:48.576680 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.55739 (* 1 = 3.55739 loss)
I0511 00:12:48.576696 39557 sgd_solver.cpp:105] Iteration 103900, lr = 0.01
I0511 00:13:15.418473 39557 solver.cpp:218] Iteration 104000 (3.72555 iter/s, 26.8417s/100 iters), loss = 3.95682
I0511 00:13:15.418630 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.95682 (* 1 = 3.95682 loss)
I0511 00:13:15.418648 39557 sgd_solver.cpp:105] Iteration 104000, lr = 0.01
I0511 00:13:42.172781 39557 solver.cpp:218] Iteration 104100 (3.73776 iter/s, 26.754s/100 iters), loss = 4.42524
I0511 00:13:42.172840 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.42524 (* 1 = 4.42524 loss)
I0511 00:13:42.172853 39557 sgd_solver.cpp:105] Iteration 104100, lr = 0.01
I0511 00:14:08.864081 39557 solver.cpp:218] Iteration 104200 (3.74657 iter/s, 26.6911s/100 iters), loss = 4.0727
I0511 00:14:08.864241 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.0727 (* 1 = 4.0727 loss)
I0511 00:14:08.864259 39557 sgd_solver.cpp:105] Iteration 104200, lr = 0.01
I0511 00:14:35.637315 39557 solver.cpp:218] Iteration 104300 (3.73512 iter/s, 26.7729s/100 iters), loss = 4.23206
I0511 00:14:35.637374 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.23206 (* 1 = 4.23206 loss)
I0511 00:14:35.637388 39557 sgd_solver.cpp:105] Iteration 104300, lr = 0.01
I0511 00:15:02.455271 39557 solver.cpp:218] Iteration 104400 (3.72887 iter/s, 26.8178s/100 iters), loss = 3.53507
I0511 00:15:02.455458 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.53507 (* 1 = 3.53507 loss)
I0511 00:15:02.455477 39557 sgd_solver.cpp:105] Iteration 104400, lr = 0.01
I0511 00:15:29.227749 39557 solver.cpp:218] Iteration 104500 (3.73522 iter/s, 26.7722s/100 iters), loss = 3.66149
I0511 00:15:29.227809 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.66149 (* 1 = 3.66149 loss)
I0511 00:15:29.227823 39557 sgd_solver.cpp:105] Iteration 104500, lr = 0.01
I0511 00:15:55.994277 39557 solver.cpp:218] Iteration 104600 (3.73604 iter/s, 26.7663s/100 iters), loss = 3.5698
I0511 00:15:55.994434 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.5698 (* 1 = 3.5698 loss)
I0511 00:15:55.994463 39557 sgd_solver.cpp:105] Iteration 104600, lr = 0.01
I0511 00:16:22.658772 39557 solver.cpp:218] Iteration 104700 (3.75035 iter/s, 26.6642s/100 iters), loss = 3.98389
I0511 00:16:22.658833 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.98389 (* 1 = 3.98389 loss)
I0511 00:16:22.658846 39557 sgd_solver.cpp:105] Iteration 104700, lr = 0.01
I0511 00:16:49.429999 39557 solver.cpp:218] Iteration 104800 (3.73538 iter/s, 26.771s/100 iters), loss = 4.2955
I0511 00:16:49.430187 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.2955 (* 1 = 4.2955 loss)
I0511 00:16:49.430207 39557 sgd_solver.cpp:105] Iteration 104800, lr = 0.01
I0511 00:17:16.157598 39557 solver.cpp:218] Iteration 104900 (3.7415 iter/s, 26.7273s/100 iters), loss = 3.63235
I0511 00:17:16.157662 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.63235 (* 1 = 3.63235 loss)
I0511 00:17:16.157675 39557 sgd_solver.cpp:105] Iteration 104900, lr = 0.01
I0511 00:17:42.558787 39557 solver.cpp:447] Snapshotting to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_105000.caffemodel
I0511 00:17:42.599288 39557 sgd_solver.cpp:273] Snapshotting solver state to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_105000.solverstate
I0511 00:17:42.881824 39557 solver.cpp:218] Iteration 105000 (3.74195 iter/s, 26.724s/100 iters), loss = 3.10678
I0511 00:17:42.881875 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.10678 (* 1 = 3.10678 loss)
I0511 00:17:42.881884 39557 sgd_solver.cpp:105] Iteration 105000, lr = 0.01
I0511 00:18:09.511559 39557 solver.cpp:218] Iteration 105100 (3.75523 iter/s, 26.6295s/100 iters), loss = 3.59446
I0511 00:18:09.511636 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.59446 (* 1 = 3.59446 loss)
I0511 00:18:09.511651 39557 sgd_solver.cpp:105] Iteration 105100, lr = 0.01
I0511 00:18:36.179186 39557 solver.cpp:218] Iteration 105200 (3.74989 iter/s, 26.6674s/100 iters), loss = 3.57085
I0511 00:18:36.179353 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.57085 (* 1 = 3.57085 loss)
I0511 00:18:36.179368 39557 sgd_solver.cpp:105] Iteration 105200, lr = 0.01
I0511 00:19:02.843179 39557 solver.cpp:218] Iteration 105300 (3.75042 iter/s, 26.6637s/100 iters), loss = 3.59156
I0511 00:19:02.843237 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.59156 (* 1 = 3.59156 loss)
I0511 00:19:02.843248 39557 sgd_solver.cpp:105] Iteration 105300, lr = 0.01
I0511 00:19:29.541836 39557 solver.cpp:218] Iteration 105400 (3.74553 iter/s, 26.6985s/100 iters), loss = 3.53011
I0511 00:19:29.541985 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.53011 (* 1 = 3.53011 loss)
I0511 00:19:29.542001 39557 sgd_solver.cpp:105] Iteration 105400, lr = 0.01
I0511 00:19:56.275769 39557 solver.cpp:218] Iteration 105500 (3.7406 iter/s, 26.7337s/100 iters), loss = 3.45931
I0511 00:19:56.275828 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.45931 (* 1 = 3.45931 loss)
I0511 00:19:56.275840 39557 sgd_solver.cpp:105] Iteration 105500, lr = 0.01
I0511 00:20:22.953055 39557 solver.cpp:218] Iteration 105600 (3.74853 iter/s, 26.6771s/100 iters), loss = 3.94166
I0511 00:20:22.953269 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.94166 (* 1 = 3.94166 loss)
I0511 00:20:22.953285 39557 sgd_solver.cpp:105] Iteration 105600, lr = 0.01
I0511 00:20:49.683575 39557 solver.cpp:218] Iteration 105700 (3.74109 iter/s, 26.7302s/100 iters), loss = 4.07914
I0511 00:20:49.683640 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.07914 (* 1 = 4.07914 loss)
I0511 00:20:49.683655 39557 sgd_solver.cpp:105] Iteration 105700, lr = 0.01
I0511 00:21:16.473285 39557 solver.cpp:218] Iteration 105800 (3.7328 iter/s, 26.7895s/100 iters), loss = 3.80068
I0511 00:21:16.473464 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.80068 (* 1 = 3.80068 loss)
I0511 00:21:16.473481 39557 sgd_solver.cpp:105] Iteration 105800, lr = 0.01
I0511 00:21:43.174528 39557 solver.cpp:218] Iteration 105900 (3.74519 iter/s, 26.7009s/100 iters), loss = 3.71338
I0511 00:21:43.174590 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.71338 (* 1 = 3.71338 loss)
I0511 00:21:43.174605 39557 sgd_solver.cpp:105] Iteration 105900, lr = 0.01
I0511 00:22:09.950649 39557 solver.cpp:218] Iteration 106000 (3.7347 iter/s, 26.7759s/100 iters), loss = 3.57642
I0511 00:22:09.950806 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.57642 (* 1 = 3.57642 loss)
I0511 00:22:09.950824 39557 sgd_solver.cpp:105] Iteration 106000, lr = 0.01
I0511 00:22:36.753545 39557 solver.cpp:218] Iteration 106100 (3.73098 iter/s, 26.8026s/100 iters), loss = 4.47376
I0511 00:22:36.753605 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.47376 (* 1 = 4.47376 loss)
I0511 00:22:36.753618 39557 sgd_solver.cpp:105] Iteration 106100, lr = 0.01
I0511 00:23:03.546703 39557 solver.cpp:218] Iteration 106200 (3.73232 iter/s, 26.793s/100 iters), loss = 3.27615
I0511 00:23:03.546864 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.27615 (* 1 = 3.27615 loss)
I0511 00:23:03.546880 39557 sgd_solver.cpp:105] Iteration 106200, lr = 0.01
I0511 00:23:30.440217 39557 solver.cpp:218] Iteration 106300 (3.71841 iter/s, 26.8932s/100 iters), loss = 4.07206
I0511 00:23:30.440279 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.07206 (* 1 = 4.07206 loss)
I0511 00:23:30.440290 39557 sgd_solver.cpp:105] Iteration 106300, lr = 0.01
I0511 00:23:57.063117 39557 solver.cpp:218] Iteration 106400 (3.75619 iter/s, 26.6227s/100 iters), loss = 3.91749
I0511 00:23:57.063280 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.91749 (* 1 = 3.91749 loss)
I0511 00:23:57.063295 39557 sgd_solver.cpp:105] Iteration 106400, lr = 0.01
I0511 00:24:23.671864 39557 solver.cpp:218] Iteration 106500 (3.7582 iter/s, 26.6085s/100 iters), loss = 3.15979
I0511 00:24:23.671927 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.15979 (* 1 = 3.15979 loss)
I0511 00:24:23.671941 39557 sgd_solver.cpp:105] Iteration 106500, lr = 0.01
I0511 00:24:50.268936 39557 solver.cpp:218] Iteration 106600 (3.75984 iter/s, 26.5969s/100 iters), loss = 3.52528
I0511 00:24:50.269093 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.52528 (* 1 = 3.52528 loss)
I0511 00:24:50.269109 39557 sgd_solver.cpp:105] Iteration 106600, lr = 0.01
I0511 00:25:16.930681 39557 solver.cpp:218] Iteration 106700 (3.75073 iter/s, 26.6615s/100 iters), loss = 4.15673
I0511 00:25:16.930747 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.15673 (* 1 = 4.15673 loss)
I0511 00:25:16.930760 39557 sgd_solver.cpp:105] Iteration 106700, lr = 0.01
I0511 00:25:43.628536 39557 solver.cpp:218] Iteration 106800 (3.74565 iter/s, 26.6977s/100 iters), loss = 4.787
I0511 00:25:43.628702 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.787 (* 1 = 4.787 loss)
I0511 00:25:43.628717 39557 sgd_solver.cpp:105] Iteration 106800, lr = 0.01
I0511 00:26:10.386795 39557 solver.cpp:218] Iteration 106900 (3.7372 iter/s, 26.758s/100 iters), loss = 4.1409
I0511 00:26:10.386859 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.1409 (* 1 = 4.1409 loss)
I0511 00:26:10.386874 39557 sgd_solver.cpp:105] Iteration 106900, lr = 0.01
I0511 00:26:37.125756 39557 solver.cpp:218] Iteration 107000 (3.73989 iter/s, 26.7388s/100 iters), loss = 4.64408
I0511 00:26:37.125974 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.64408 (* 1 = 4.64408 loss)
I0511 00:26:37.125990 39557 sgd_solver.cpp:105] Iteration 107000, lr = 0.01
I0511 00:27:03.882303 39557 solver.cpp:218] Iteration 107100 (3.73745 iter/s, 26.7562s/100 iters), loss = 3.5491
I0511 00:27:03.882369 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.5491 (* 1 = 3.5491 loss)
I0511 00:27:03.882382 39557 sgd_solver.cpp:105] Iteration 107100, lr = 0.01
I0511 00:27:30.626888 39557 solver.cpp:218] Iteration 107200 (3.7391 iter/s, 26.7444s/100 iters), loss = 3.98047
I0511 00:27:30.627050 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.98047 (* 1 = 3.98047 loss)
I0511 00:27:30.627063 39557 sgd_solver.cpp:105] Iteration 107200, lr = 0.01
I0511 00:27:57.271348 39557 solver.cpp:218] Iteration 107300 (3.75317 iter/s, 26.6442s/100 iters), loss = 4.15047
I0511 00:27:57.271406 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.15047 (* 1 = 4.15047 loss)
I0511 00:27:57.271431 39557 sgd_solver.cpp:105] Iteration 107300, lr = 0.01
I0511 00:28:24.083631 39557 solver.cpp:218] Iteration 107400 (3.72966 iter/s, 26.8121s/100 iters), loss = 3.76414
I0511 00:28:24.083819 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.76414 (* 1 = 3.76414 loss)
I0511 00:28:24.083837 39557 sgd_solver.cpp:105] Iteration 107400, lr = 0.01
I0511 00:28:50.882839 39557 solver.cpp:218] Iteration 107500 (3.7315 iter/s, 26.7989s/100 iters), loss = 3.97136
I0511 00:28:50.882912 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.97136 (* 1 = 3.97136 loss)
I0511 00:28:50.882925 39557 sgd_solver.cpp:105] Iteration 107500, lr = 0.01
I0511 00:29:17.615756 39557 solver.cpp:218] Iteration 107600 (3.74073 iter/s, 26.7327s/100 iters), loss = 3.33098
I0511 00:29:17.615911 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.33098 (* 1 = 3.33098 loss)
I0511 00:29:17.615926 39557 sgd_solver.cpp:105] Iteration 107600, lr = 0.01
I0511 00:29:44.334595 39557 solver.cpp:218] Iteration 107700 (3.74272 iter/s, 26.7186s/100 iters), loss = 3.44181
I0511 00:29:44.334655 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.44181 (* 1 = 3.44181 loss)
I0511 00:29:44.334667 39557 sgd_solver.cpp:105] Iteration 107700, lr = 0.01
I0511 00:30:11.035914 39557 solver.cpp:218] Iteration 107800 (3.74516 iter/s, 26.7011s/100 iters), loss = 3.12506
I0511 00:30:11.036079 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.12506 (* 1 = 3.12506 loss)
I0511 00:30:11.036094 39557 sgd_solver.cpp:105] Iteration 107800, lr = 0.01
I0511 00:30:37.784801 39557 solver.cpp:218] Iteration 107900 (3.73851 iter/s, 26.7486s/100 iters), loss = 3.32804
I0511 00:30:37.784863 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.32804 (* 1 = 3.32804 loss)
I0511 00:30:37.784878 39557 sgd_solver.cpp:105] Iteration 107900, lr = 0.01
I0511 00:31:04.578219 39557 solver.cpp:218] Iteration 108000 (3.73229 iter/s, 26.7932s/100 iters), loss = 3.89882
I0511 00:31:04.578394 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.89882 (* 1 = 3.89882 loss)
I0511 00:31:04.578409 39557 sgd_solver.cpp:105] Iteration 108000, lr = 0.01
I0511 00:31:31.343096 39557 solver.cpp:218] Iteration 108100 (3.73628 iter/s, 26.7646s/100 iters), loss = 3.5751
I0511 00:31:31.343161 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.5751 (* 1 = 3.5751 loss)
I0511 00:31:31.343175 39557 sgd_solver.cpp:105] Iteration 108100, lr = 0.01
I0511 00:31:58.116259 39557 solver.cpp:218] Iteration 108200 (3.73511 iter/s, 26.773s/100 iters), loss = 3.94664
I0511 00:31:58.116472 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.94664 (* 1 = 3.94664 loss)
I0511 00:31:58.116487 39557 sgd_solver.cpp:105] Iteration 108200, lr = 0.01
I0511 00:32:24.842547 39557 solver.cpp:218] Iteration 108300 (3.74168 iter/s, 26.726s/100 iters), loss = 4.39789
I0511 00:32:24.842615 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.39789 (* 1 = 4.39789 loss)
I0511 00:32:24.842629 39557 sgd_solver.cpp:105] Iteration 108300, lr = 0.01
I0511 00:32:51.627697 39557 solver.cpp:218] Iteration 108400 (3.73344 iter/s, 26.785s/100 iters), loss = 3.149
I0511 00:32:51.627856 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.149 (* 1 = 3.149 loss)
I0511 00:32:51.627876 39557 sgd_solver.cpp:105] Iteration 108400, lr = 0.01
I0511 00:33:18.435885 39557 solver.cpp:218] Iteration 108500 (3.73024 iter/s, 26.8079s/100 iters), loss = 3.65903
I0511 00:33:18.435972 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.65903 (* 1 = 3.65903 loss)
I0511 00:33:18.435988 39557 sgd_solver.cpp:105] Iteration 108500, lr = 0.01
I0511 00:33:45.195658 39557 solver.cpp:218] Iteration 108600 (3.73698 iter/s, 26.7596s/100 iters), loss = 3.6853
I0511 00:33:45.195824 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.6853 (* 1 = 3.6853 loss)
I0511 00:33:45.195842 39557 sgd_solver.cpp:105] Iteration 108600, lr = 0.01
I0511 00:34:12.006497 39557 solver.cpp:218] Iteration 108700 (3.72987 iter/s, 26.8106s/100 iters), loss = 4.15137
I0511 00:34:12.006564 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.15137 (* 1 = 4.15137 loss)
I0511 00:34:12.006579 39557 sgd_solver.cpp:105] Iteration 108700, lr = 0.01
I0511 00:34:38.698690 39557 solver.cpp:218] Iteration 108800 (3.74644 iter/s, 26.692s/100 iters), loss = 4.0342
I0511 00:34:38.698858 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.0342 (* 1 = 4.0342 loss)
I0511 00:34:38.698874 39557 sgd_solver.cpp:105] Iteration 108800, lr = 0.01
I0511 00:35:05.525710 39557 solver.cpp:218] Iteration 108900 (3.72762 iter/s, 26.8267s/100 iters), loss = 4.03791
I0511 00:35:05.525774 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.03791 (* 1 = 4.03791 loss)
I0511 00:35:05.525785 39557 sgd_solver.cpp:105] Iteration 108900, lr = 0.01
I0511 00:35:32.299474 39557 solver.cpp:218] Iteration 109000 (3.73502 iter/s, 26.7736s/100 iters), loss = 3.41685
I0511 00:35:32.299620 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.41685 (* 1 = 3.41685 loss)
I0511 00:35:32.299635 39557 sgd_solver.cpp:105] Iteration 109000, lr = 0.01
I0511 00:35:59.108589 39557 solver.cpp:218] Iteration 109100 (3.73011 iter/s, 26.8089s/100 iters), loss = 4.64184
I0511 00:35:59.108651 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.64184 (* 1 = 4.64184 loss)
I0511 00:35:59.108664 39557 sgd_solver.cpp:105] Iteration 109100, lr = 0.01
I0511 00:36:25.941031 39557 solver.cpp:218] Iteration 109200 (3.72686 iter/s, 26.8323s/100 iters), loss = 3.67908
I0511 00:36:25.941195 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.67908 (* 1 = 3.67908 loss)
I0511 00:36:25.941211 39557 sgd_solver.cpp:105] Iteration 109200, lr = 0.01
I0511 00:36:52.650444 39557 solver.cpp:218] Iteration 109300 (3.74404 iter/s, 26.7091s/100 iters), loss = 3.25861
I0511 00:36:52.650506 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.25861 (* 1 = 3.25861 loss)
I0511 00:36:52.650519 39557 sgd_solver.cpp:105] Iteration 109300, lr = 0.01
I0511 00:37:19.441349 39557 solver.cpp:218] Iteration 109400 (3.73263 iter/s, 26.7907s/100 iters), loss = 4.02397
I0511 00:37:19.441499 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.02397 (* 1 = 4.02397 loss)
I0511 00:37:19.441517 39557 sgd_solver.cpp:105] Iteration 109400, lr = 0.01
I0511 00:37:46.127985 39557 solver.cpp:218] Iteration 109500 (3.74723 iter/s, 26.6864s/100 iters), loss = 3.8015
I0511 00:37:46.128046 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.8015 (* 1 = 3.8015 loss)
I0511 00:37:46.128057 39557 sgd_solver.cpp:105] Iteration 109500, lr = 0.01
I0511 00:38:13.067986 39557 solver.cpp:218] Iteration 109600 (3.71198 iter/s, 26.9398s/100 iters), loss = 3.56867
I0511 00:38:13.068190 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.56867 (* 1 = 3.56867 loss)
I0511 00:38:13.068208 39557 sgd_solver.cpp:105] Iteration 109600, lr = 0.01
I0511 00:38:39.848191 39557 solver.cpp:218] Iteration 109700 (3.73415 iter/s, 26.7799s/100 iters), loss = 3.83339
I0511 00:38:39.848246 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.83339 (* 1 = 3.83339 loss)
I0511 00:38:39.848258 39557 sgd_solver.cpp:105] Iteration 109700, lr = 0.01
I0511 00:39:06.666008 39557 solver.cpp:218] Iteration 109800 (3.72889 iter/s, 26.8176s/100 iters), loss = 4.10043
I0511 00:39:06.666218 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.10043 (* 1 = 4.10043 loss)
I0511 00:39:06.666234 39557 sgd_solver.cpp:105] Iteration 109800, lr = 0.01
I0511 00:39:33.437572 39557 solver.cpp:218] Iteration 109900 (3.73535 iter/s, 26.7712s/100 iters), loss = 3.05452
I0511 00:39:33.437638 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.05452 (* 1 = 3.05452 loss)
I0511 00:39:33.437651 39557 sgd_solver.cpp:105] Iteration 109900, lr = 0.01
I0511 00:39:59.838382 39557 solver.cpp:447] Snapshotting to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_110000.caffemodel
I0511 00:39:59.887044 39557 sgd_solver.cpp:273] Snapshotting solver state to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_110000.solverstate
I0511 00:40:00.180963 39557 solver.cpp:218] Iteration 110000 (3.73927 iter/s, 26.7432s/100 iters), loss = 3.92327
I0511 00:40:00.181030 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.92327 (* 1 = 3.92327 loss)
I0511 00:40:00.181041 39557 sgd_solver.cpp:105] Iteration 110000, lr = 0.01
I0511 00:40:26.939630 39557 solver.cpp:218] Iteration 110100 (3.73713 iter/s, 26.7585s/100 iters), loss = 3.84124
I0511 00:40:26.939694 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.84124 (* 1 = 3.84124 loss)
I0511 00:40:26.939708 39557 sgd_solver.cpp:105] Iteration 110100, lr = 0.01
I0511 00:40:53.698278 39557 solver.cpp:218] Iteration 110200 (3.73713 iter/s, 26.7585s/100 iters), loss = 3.64013
I0511 00:40:53.698449 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.64013 (* 1 = 3.64013 loss)
I0511 00:40:53.698470 39557 sgd_solver.cpp:105] Iteration 110200, lr = 0.01
I0511 00:41:20.492643 39557 solver.cpp:218] Iteration 110300 (3.73217 iter/s, 26.7941s/100 iters), loss = 3.77978
I0511 00:41:20.492710 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.77978 (* 1 = 3.77978 loss)
I0511 00:41:20.492725 39557 sgd_solver.cpp:105] Iteration 110300, lr = 0.01
I0511 00:41:47.311408 39557 solver.cpp:218] Iteration 110400 (3.72876 iter/s, 26.8186s/100 iters), loss = 3.75275
I0511 00:41:47.311573 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.75275 (* 1 = 3.75275 loss)
I0511 00:41:47.311590 39557 sgd_solver.cpp:105] Iteration 110400, lr = 0.01
I0511 00:42:13.999259 39557 solver.cpp:218] Iteration 110500 (3.74706 iter/s, 26.6876s/100 iters), loss = 4.062
I0511 00:42:13.999323 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.062 (* 1 = 4.062 loss)
I0511 00:42:13.999344 39557 sgd_solver.cpp:105] Iteration 110500, lr = 0.01
I0511 00:42:40.805124 39557 solver.cpp:218] Iteration 110600 (3.73055 iter/s, 26.8057s/100 iters), loss = 2.645
I0511 00:42:40.805274 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.645 (* 1 = 2.645 loss)
I0511 00:42:40.805294 39557 sgd_solver.cpp:105] Iteration 110600, lr = 0.01
I0511 00:43:07.557710 39557 solver.cpp:218] Iteration 110700 (3.73799 iter/s, 26.7524s/100 iters), loss = 3.67265
I0511 00:43:07.557760 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.67265 (* 1 = 3.67265 loss)
I0511 00:43:07.557771 39557 sgd_solver.cpp:105] Iteration 110700, lr = 0.01
I0511 00:43:34.259527 39557 solver.cpp:218] Iteration 110800 (3.74498 iter/s, 26.7024s/100 iters), loss = 4.00352
I0511 00:43:34.259721 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.00352 (* 1 = 4.00352 loss)
I0511 00:43:34.259738 39557 sgd_solver.cpp:105] Iteration 110800, lr = 0.01
I0511 00:44:00.864074 39557 solver.cpp:218] Iteration 110900 (3.7587 iter/s, 26.6049s/100 iters), loss = 4.65305
I0511 00:44:00.864136 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.65305 (* 1 = 4.65305 loss)
I0511 00:44:00.864148 39557 sgd_solver.cpp:105] Iteration 110900, lr = 0.01
I0511 00:44:27.519554 39557 solver.cpp:218] Iteration 111000 (3.7515 iter/s, 26.656s/100 iters), loss = 3.72766
I0511 00:44:27.519739 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.72766 (* 1 = 3.72766 loss)
I0511 00:44:27.519754 39557 sgd_solver.cpp:105] Iteration 111000, lr = 0.01
I0511 00:44:54.167457 39557 solver.cpp:218] Iteration 111100 (3.75259 iter/s, 26.6483s/100 iters), loss = 3.9759
I0511 00:44:54.167522 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.9759 (* 1 = 3.9759 loss)
I0511 00:44:54.167538 39557 sgd_solver.cpp:105] Iteration 111100, lr = 0.01
I0511 00:45:20.855979 39557 solver.cpp:218] Iteration 111200 (3.74687 iter/s, 26.689s/100 iters), loss = 3.05633
I0511 00:45:20.856148 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.05633 (* 1 = 3.05633 loss)
I0511 00:45:20.856163 39557 sgd_solver.cpp:105] Iteration 111200, lr = 0.01
I0511 00:45:47.476997 39557 solver.cpp:218] Iteration 111300 (3.75639 iter/s, 26.6213s/100 iters), loss = 4.1137
I0511 00:45:47.477061 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.1137 (* 1 = 4.1137 loss)
I0511 00:45:47.477074 39557 sgd_solver.cpp:105] Iteration 111300, lr = 0.01
I0511 00:46:14.151530 39557 solver.cpp:218] Iteration 111400 (3.74884 iter/s, 26.6749s/100 iters), loss = 3.18025
I0511 00:46:14.151691 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.18025 (* 1 = 3.18025 loss)
I0511 00:46:14.151706 39557 sgd_solver.cpp:105] Iteration 111400, lr = 0.01
I0511 00:46:40.800659 39557 solver.cpp:218] Iteration 111500 (3.75243 iter/s, 26.6494s/100 iters), loss = 3.85514
I0511 00:46:40.800716 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.85514 (* 1 = 3.85514 loss)
I0511 00:46:40.800729 39557 sgd_solver.cpp:105] Iteration 111500, lr = 0.01
I0511 00:47:07.441027 39557 solver.cpp:218] Iteration 111600 (3.75365 iter/s, 26.6407s/100 iters), loss = 4.57999
I0511 00:47:07.441191 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.57999 (* 1 = 4.57999 loss)
I0511 00:47:07.441206 39557 sgd_solver.cpp:105] Iteration 111600, lr = 0.01
I0511 00:47:34.060858 39557 solver.cpp:218] Iteration 111700 (3.75657 iter/s, 26.6201s/100 iters), loss = 3.52529
I0511 00:47:34.060923 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.52529 (* 1 = 3.52529 loss)
I0511 00:47:34.060935 39557 sgd_solver.cpp:105] Iteration 111700, lr = 0.01
I0511 00:48:00.727110 39557 solver.cpp:218] Iteration 111800 (3.75002 iter/s, 26.6666s/100 iters), loss = 4.20963
I0511 00:48:00.727262 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.20963 (* 1 = 4.20963 loss)
I0511 00:48:00.727291 39557 sgd_solver.cpp:105] Iteration 111800, lr = 0.01
I0511 00:48:27.531502 39557 solver.cpp:218] Iteration 111900 (3.7307 iter/s, 26.8046s/100 iters), loss = 4.11039
I0511 00:48:27.531559 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.11039 (* 1 = 4.11039 loss)
I0511 00:48:27.531572 39557 sgd_solver.cpp:105] Iteration 111900, lr = 0.01
I0511 00:48:54.165782 39557 solver.cpp:218] Iteration 112000 (3.75452 iter/s, 26.6346s/100 iters), loss = 3.72924
I0511 00:48:54.165940 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.72924 (* 1 = 3.72924 loss)
I0511 00:48:54.165956 39557 sgd_solver.cpp:105] Iteration 112000, lr = 0.01
I0511 00:49:20.780088 39557 solver.cpp:218] Iteration 112100 (3.75735 iter/s, 26.6145s/100 iters), loss = 3.10393
I0511 00:49:20.780144 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.10393 (* 1 = 3.10393 loss)
I0511 00:49:20.780155 39557 sgd_solver.cpp:105] Iteration 112100, lr = 0.01
I0511 00:49:47.390408 39557 solver.cpp:218] Iteration 112200 (3.75791 iter/s, 26.6106s/100 iters), loss = 3.75924
I0511 00:49:47.390604 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.75924 (* 1 = 3.75924 loss)
I0511 00:49:47.390620 39557 sgd_solver.cpp:105] Iteration 112200, lr = 0.01
I0511 00:50:14.037497 39557 solver.cpp:218] Iteration 112300 (3.75274 iter/s, 26.6472s/100 iters), loss = 3.95037
I0511 00:50:14.037554 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.95037 (* 1 = 3.95037 loss)
I0511 00:50:14.037565 39557 sgd_solver.cpp:105] Iteration 112300, lr = 0.01
I0511 00:50:40.671777 39557 solver.cpp:218] Iteration 112400 (3.75453 iter/s, 26.6345s/100 iters), loss = 3.45346
I0511 00:50:40.671939 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.45346 (* 1 = 3.45346 loss)
I0511 00:50:40.671952 39557 sgd_solver.cpp:105] Iteration 112400, lr = 0.01
I0511 00:51:07.253898 39557 solver.cpp:218] Iteration 112500 (3.76191 iter/s, 26.5822s/100 iters), loss = 3.43334
I0511 00:51:07.253944 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.43334 (* 1 = 3.43334 loss)
I0511 00:51:07.253955 39557 sgd_solver.cpp:105] Iteration 112500, lr = 0.01
I0511 00:51:33.904259 39557 solver.cpp:218] Iteration 112600 (3.75227 iter/s, 26.6506s/100 iters), loss = 3.5327
I0511 00:51:33.904412 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.5327 (* 1 = 3.5327 loss)
I0511 00:51:33.904424 39557 sgd_solver.cpp:105] Iteration 112600, lr = 0.01
I0511 00:52:00.568171 39557 solver.cpp:218] Iteration 112700 (3.75038 iter/s, 26.664s/100 iters), loss = 3.74436
I0511 00:52:00.568228 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.74436 (* 1 = 3.74436 loss)
I0511 00:52:00.568240 39557 sgd_solver.cpp:105] Iteration 112700, lr = 0.01
I0511 00:52:27.259145 39557 solver.cpp:218] Iteration 112800 (3.74656 iter/s, 26.6911s/100 iters), loss = 3.26298
I0511 00:52:27.259290 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.26298 (* 1 = 3.26298 loss)
I0511 00:52:27.259302 39557 sgd_solver.cpp:105] Iteration 112800, lr = 0.01
I0511 00:52:53.845023 39557 solver.cpp:218] Iteration 112900 (3.76139 iter/s, 26.5859s/100 iters), loss = 3.41151
I0511 00:52:53.845080 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.41151 (* 1 = 3.41151 loss)
I0511 00:52:53.845095 39557 sgd_solver.cpp:105] Iteration 112900, lr = 0.01
I0511 00:53:20.419143 39557 solver.cpp:218] Iteration 113000 (3.76304 iter/s, 26.5743s/100 iters), loss = 3.8673
I0511 00:53:20.419297 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.8673 (* 1 = 3.8673 loss)
I0511 00:53:20.419309 39557 sgd_solver.cpp:105] Iteration 113000, lr = 0.01
I0511 00:53:47.278213 39557 solver.cpp:218] Iteration 113100 (3.72313 iter/s, 26.8591s/100 iters), loss = 4.03828
I0511 00:53:47.278280 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.03828 (* 1 = 4.03828 loss)
I0511 00:53:47.278295 39557 sgd_solver.cpp:105] Iteration 113100, lr = 0.01
I0511 00:54:13.978855 39557 solver.cpp:218] Iteration 113200 (3.74521 iter/s, 26.7008s/100 iters), loss = 3.38984
I0511 00:54:13.979027 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.38984 (* 1 = 3.38984 loss)
I0511 00:54:13.979045 39557 sgd_solver.cpp:105] Iteration 113200, lr = 0.01
I0511 00:54:40.711953 39557 solver.cpp:218] Iteration 113300 (3.74068 iter/s, 26.7331s/100 iters), loss = 4.49772
I0511 00:54:40.712020 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.49772 (* 1 = 4.49772 loss)
I0511 00:54:40.712035 39557 sgd_solver.cpp:105] Iteration 113300, lr = 0.01
I0511 00:55:07.410696 39557 solver.cpp:218] Iteration 113400 (3.74548 iter/s, 26.6988s/100 iters), loss = 3.90579
I0511 00:55:07.410879 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.90579 (* 1 = 3.90579 loss)
I0511 00:55:07.410897 39557 sgd_solver.cpp:105] Iteration 113400, lr = 0.01
I0511 00:55:34.151443 39557 solver.cpp:218] Iteration 113500 (3.73962 iter/s, 26.7407s/100 iters), loss = 3.5017
I0511 00:55:34.151506 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.5017 (* 1 = 3.5017 loss)
I0511 00:55:34.151520 39557 sgd_solver.cpp:105] Iteration 113500, lr = 0.01
I0511 00:56:00.801179 39557 solver.cpp:218] Iteration 113600 (3.75237 iter/s, 26.6498s/100 iters), loss = 3.72173
I0511 00:56:00.801365 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.72173 (* 1 = 3.72173 loss)
I0511 00:56:00.801381 39557 sgd_solver.cpp:105] Iteration 113600, lr = 0.01
I0511 00:56:27.425704 39557 solver.cpp:218] Iteration 113700 (3.75594 iter/s, 26.6245s/100 iters), loss = 3.74407
I0511 00:56:27.425765 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.74407 (* 1 = 3.74407 loss)
I0511 00:56:27.425777 39557 sgd_solver.cpp:105] Iteration 113700, lr = 0.01
I0511 00:56:54.049037 39557 solver.cpp:218] Iteration 113800 (3.75609 iter/s, 26.6234s/100 iters), loss = 3.1105
I0511 00:56:54.049212 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.1105 (* 1 = 3.1105 loss)
I0511 00:56:54.049227 39557 sgd_solver.cpp:105] Iteration 113800, lr = 0.01
I0511 00:57:20.693207 39557 solver.cpp:218] Iteration 113900 (3.75317 iter/s, 26.6441s/100 iters), loss = 3.55695
I0511 00:57:20.693276 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.55695 (* 1 = 3.55695 loss)
I0511 00:57:20.693302 39557 sgd_solver.cpp:105] Iteration 113900, lr = 0.01
I0511 00:57:47.333753 39557 solver.cpp:218] Iteration 114000 (3.75367 iter/s, 26.6406s/100 iters), loss = 3.98346
I0511 00:57:47.333910 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.98346 (* 1 = 3.98346 loss)
I0511 00:57:47.333922 39557 sgd_solver.cpp:105] Iteration 114000, lr = 0.01
I0511 00:58:14.019346 39557 solver.cpp:218] Iteration 114100 (3.74735 iter/s, 26.6855s/100 iters), loss = 3.50633
I0511 00:58:14.019405 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.50633 (* 1 = 3.50633 loss)
I0511 00:58:14.019418 39557 sgd_solver.cpp:105] Iteration 114100, lr = 0.01
I0511 00:58:40.710604 39557 solver.cpp:218] Iteration 114200 (3.74654 iter/s, 26.6913s/100 iters), loss = 4.07963
I0511 00:58:40.710764 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.07963 (* 1 = 4.07963 loss)
I0511 00:58:40.710777 39557 sgd_solver.cpp:105] Iteration 114200, lr = 0.01
I0511 00:59:07.397642 39557 solver.cpp:218] Iteration 114300 (3.74715 iter/s, 26.687s/100 iters), loss = 3.51535
I0511 00:59:07.397701 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.51535 (* 1 = 3.51535 loss)
I0511 00:59:07.397713 39557 sgd_solver.cpp:105] Iteration 114300, lr = 0.01
I0511 00:59:34.019698 39557 solver.cpp:218] Iteration 114400 (3.75628 iter/s, 26.6221s/100 iters), loss = 3.15559
I0511 00:59:34.019848 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.15559 (* 1 = 3.15559 loss)
I0511 00:59:34.019861 39557 sgd_solver.cpp:105] Iteration 114400, lr = 0.01
I0511 01:00:00.657459 39557 solver.cpp:218] Iteration 114500 (3.75408 iter/s, 26.6377s/100 iters), loss = 4.35231
I0511 01:00:00.657518 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.35231 (* 1 = 4.35231 loss)
I0511 01:00:00.657531 39557 sgd_solver.cpp:105] Iteration 114500, lr = 0.01
I0511 01:00:27.238574 39557 solver.cpp:218] Iteration 114600 (3.76206 iter/s, 26.5811s/100 iters), loss = 3.28819
I0511 01:00:27.238739 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.28819 (* 1 = 3.28819 loss)
I0511 01:00:27.238752 39557 sgd_solver.cpp:105] Iteration 114600, lr = 0.01
I0511 01:00:53.871974 39557 solver.cpp:218] Iteration 114700 (3.75469 iter/s, 26.6333s/100 iters), loss = 4.12655
I0511 01:00:53.872032 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.12655 (* 1 = 4.12655 loss)
I0511 01:00:53.872046 39557 sgd_solver.cpp:105] Iteration 114700, lr = 0.01
I0511 01:01:20.533300 39557 solver.cpp:218] Iteration 114800 (3.75075 iter/s, 26.6614s/100 iters), loss = 3.98884
I0511 01:01:20.533457 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.98884 (* 1 = 3.98884 loss)
I0511 01:01:20.533471 39557 sgd_solver.cpp:105] Iteration 114800, lr = 0.01
I0511 01:01:47.155207 39557 solver.cpp:218] Iteration 114900 (3.75632 iter/s, 26.6218s/100 iters), loss = 4.09753
I0511 01:01:47.155267 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.09753 (* 1 = 4.09753 loss)
I0511 01:01:47.155280 39557 sgd_solver.cpp:105] Iteration 114900, lr = 0.01
I0511 01:02:13.557479 39557 solver.cpp:447] Snapshotting to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_115000.caffemodel
I0511 01:02:13.596567 39557 sgd_solver.cpp:273] Snapshotting solver state to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_115000.solverstate
I0511 01:02:13.879314 39557 solver.cpp:218] Iteration 115000 (3.74194 iter/s, 26.7241s/100 iters), loss = 3.37793
I0511 01:02:13.879382 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.37793 (* 1 = 3.37793 loss)
I0511 01:02:13.879393 39557 sgd_solver.cpp:105] Iteration 115000, lr = 0.01
I0511 01:02:40.522167 39557 solver.cpp:218] Iteration 115100 (3.75335 iter/s, 26.6429s/100 iters), loss = 4.48123
I0511 01:02:40.522227 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.48123 (* 1 = 4.48123 loss)
I0511 01:02:40.522239 39557 sgd_solver.cpp:105] Iteration 115100, lr = 0.01
I0511 01:03:07.160892 39557 solver.cpp:218] Iteration 115200 (3.75393 iter/s, 26.6387s/100 iters), loss = 4.14294
I0511 01:03:07.161048 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.14294 (* 1 = 4.14294 loss)
I0511 01:03:07.161062 39557 sgd_solver.cpp:105] Iteration 115200, lr = 0.01
I0511 01:03:33.752544 39557 solver.cpp:218] Iteration 115300 (3.76059 iter/s, 26.5916s/100 iters), loss = 3.06334
I0511 01:03:33.752605 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.06334 (* 1 = 3.06334 loss)
I0511 01:03:33.752619 39557 sgd_solver.cpp:105] Iteration 115300, lr = 0.01
I0511 01:04:00.358853 39557 solver.cpp:218] Iteration 115400 (3.75851 iter/s, 26.6063s/100 iters), loss = 4.43914
I0511 01:04:00.359012 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.43914 (* 1 = 4.43914 loss)
I0511 01:04:00.359026 39557 sgd_solver.cpp:105] Iteration 115400, lr = 0.01
I0511 01:04:26.890516 39557 solver.cpp:218] Iteration 115500 (3.76909 iter/s, 26.5316s/100 iters), loss = 3.66059
I0511 01:04:26.890570 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.66059 (* 1 = 3.66059 loss)
I0511 01:04:26.890580 39557 sgd_solver.cpp:105] Iteration 115500, lr = 0.01
I0511 01:04:53.505625 39557 solver.cpp:218] Iteration 115600 (3.75727 iter/s, 26.6151s/100 iters), loss = 3.51639
I0511 01:04:53.505785 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.51639 (* 1 = 3.51639 loss)
I0511 01:04:53.505800 39557 sgd_solver.cpp:105] Iteration 115600, lr = 0.01
I0511 01:05:20.142174 39557 solver.cpp:218] Iteration 115700 (3.75425 iter/s, 26.6364s/100 iters), loss = 3.39651
I0511 01:05:20.142233 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.39651 (* 1 = 3.39651 loss)
I0511 01:05:20.142246 39557 sgd_solver.cpp:105] Iteration 115700, lr = 0.01
I0511 01:05:46.706149 39557 solver.cpp:218] Iteration 115800 (3.7645 iter/s, 26.564s/100 iters), loss = 4.23165
I0511 01:05:46.706313 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.23165 (* 1 = 4.23165 loss)
I0511 01:05:46.706326 39557 sgd_solver.cpp:105] Iteration 115800, lr = 0.01
I0511 01:06:13.352195 39557 solver.cpp:218] Iteration 115900 (3.75292 iter/s, 26.6459s/100 iters), loss = 3.97526
I0511 01:06:13.352253 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.97526 (* 1 = 3.97526 loss)
I0511 01:06:13.352267 39557 sgd_solver.cpp:105] Iteration 115900, lr = 0.01
I0511 01:06:39.953022 39557 solver.cpp:218] Iteration 116000 (3.75928 iter/s, 26.6008s/100 iters), loss = 3.44311
I0511 01:06:39.953189 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.44311 (* 1 = 3.44311 loss)
I0511 01:06:39.953207 39557 sgd_solver.cpp:105] Iteration 116000, lr = 0.01
I0511 01:07:06.684928 39557 solver.cpp:218] Iteration 116100 (3.74086 iter/s, 26.7318s/100 iters), loss = 4.17715
I0511 01:07:06.685003 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.17715 (* 1 = 4.17715 loss)
I0511 01:07:06.685016 39557 sgd_solver.cpp:105] Iteration 116100, lr = 0.01
I0511 01:07:33.395579 39557 solver.cpp:218] Iteration 116200 (3.74383 iter/s, 26.7106s/100 iters), loss = 4.35748
I0511 01:07:33.395793 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.35748 (* 1 = 4.35748 loss)
I0511 01:07:33.395812 39557 sgd_solver.cpp:105] Iteration 116200, lr = 0.01
I0511 01:08:00.085111 39557 solver.cpp:218] Iteration 116300 (3.74681 iter/s, 26.6894s/100 iters), loss = 3.68258
I0511 01:08:00.085176 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.68258 (* 1 = 3.68258 loss)
I0511 01:08:00.085191 39557 sgd_solver.cpp:105] Iteration 116300, lr = 0.01
I0511 01:08:26.782121 39557 solver.cpp:218] Iteration 116400 (3.74575 iter/s, 26.6969s/100 iters), loss = 3.85057
I0511 01:08:26.782301 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.85057 (* 1 = 3.85057 loss)
I0511 01:08:26.782316 39557 sgd_solver.cpp:105] Iteration 116400, lr = 0.01
I0511 01:08:53.494887 39557 solver.cpp:218] Iteration 116500 (3.74355 iter/s, 26.7126s/100 iters), loss = 3.89759
I0511 01:08:53.494948 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.89759 (* 1 = 3.89759 loss)
I0511 01:08:53.494961 39557 sgd_solver.cpp:105] Iteration 116500, lr = 0.01
I0511 01:09:20.216379 39557 solver.cpp:218] Iteration 116600 (3.74231 iter/s, 26.7215s/100 iters), loss = 4.44166
I0511 01:09:20.216542 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.44166 (* 1 = 4.44166 loss)
I0511 01:09:20.216558 39557 sgd_solver.cpp:105] Iteration 116600, lr = 0.01
I0511 01:09:46.917779 39557 solver.cpp:218] Iteration 116700 (3.74514 iter/s, 26.7013s/100 iters), loss = 2.99361
I0511 01:09:46.917847 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.99361 (* 1 = 2.99361 loss)
I0511 01:09:46.917862 39557 sgd_solver.cpp:105] Iteration 116700, lr = 0.01
I0511 01:10:13.606678 39557 solver.cpp:218] Iteration 116800 (3.74688 iter/s, 26.6889s/100 iters), loss = 3.71054
I0511 01:10:13.606830 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.71054 (* 1 = 3.71054 loss)
I0511 01:10:13.606858 39557 sgd_solver.cpp:105] Iteration 116800, lr = 0.01
I0511 01:10:40.424995 39557 solver.cpp:218] Iteration 116900 (3.72881 iter/s, 26.8182s/100 iters), loss = 3.7074
I0511 01:10:40.425060 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.7074 (* 1 = 3.7074 loss)
I0511 01:10:40.425076 39557 sgd_solver.cpp:105] Iteration 116900, lr = 0.01
I0511 01:11:07.190615 39557 solver.cpp:218] Iteration 117000 (3.73614 iter/s, 26.7656s/100 iters), loss = 4.00852
I0511 01:11:07.190776 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.00852 (* 1 = 4.00852 loss)
I0511 01:11:07.190793 39557 sgd_solver.cpp:105] Iteration 117000, lr = 0.01
I0511 01:11:33.912542 39557 solver.cpp:218] Iteration 117100 (3.74226 iter/s, 26.7218s/100 iters), loss = 3.69319
I0511 01:11:33.912611 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.69319 (* 1 = 3.69319 loss)
I0511 01:11:33.912624 39557 sgd_solver.cpp:105] Iteration 117100, lr = 0.01
I0511 01:12:00.652257 39557 solver.cpp:218] Iteration 117200 (3.73976 iter/s, 26.7397s/100 iters), loss = 3.97203
I0511 01:12:00.652432 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.97203 (* 1 = 3.97203 loss)
I0511 01:12:00.652454 39557 sgd_solver.cpp:105] Iteration 117200, lr = 0.01
I0511 01:12:27.512307 39557 solver.cpp:218] Iteration 117300 (3.72302 iter/s, 26.8599s/100 iters), loss = 3.91912
I0511 01:12:27.512374 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.91912 (* 1 = 3.91912 loss)
I0511 01:12:27.512389 39557 sgd_solver.cpp:105] Iteration 117300, lr = 0.01
I0511 01:12:54.216523 39557 solver.cpp:218] Iteration 117400 (3.74473 iter/s, 26.7042s/100 iters), loss = 2.94168
I0511 01:12:54.216684 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.94168 (* 1 = 2.94168 loss)
I0511 01:12:54.216701 39557 sgd_solver.cpp:105] Iteration 117400, lr = 0.01
I0511 01:13:20.944535 39557 solver.cpp:218] Iteration 117500 (3.74141 iter/s, 26.7279s/100 iters), loss = 3.92075
I0511 01:13:20.944594 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.92075 (* 1 = 3.92075 loss)
I0511 01:13:20.944607 39557 sgd_solver.cpp:105] Iteration 117500, lr = 0.01
I0511 01:13:47.553113 39557 solver.cpp:218] Iteration 117600 (3.75819 iter/s, 26.6085s/100 iters), loss = 4.28761
I0511 01:13:47.553336 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.28761 (* 1 = 4.28761 loss)
I0511 01:13:47.553354 39557 sgd_solver.cpp:105] Iteration 117600, lr = 0.01
I0511 01:14:14.200820 39557 solver.cpp:218] Iteration 117700 (3.7527 iter/s, 26.6475s/100 iters), loss = 3.8868
I0511 01:14:14.200871 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.8868 (* 1 = 3.8868 loss)
I0511 01:14:14.200884 39557 sgd_solver.cpp:105] Iteration 117700, lr = 0.01
I0511 01:14:40.821923 39557 solver.cpp:218] Iteration 117800 (3.75642 iter/s, 26.6211s/100 iters), loss = 4.0305
I0511 01:14:40.822093 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.0305 (* 1 = 4.0305 loss)
I0511 01:14:40.822108 39557 sgd_solver.cpp:105] Iteration 117800, lr = 0.01
I0511 01:15:07.416898 39557 solver.cpp:218] Iteration 117900 (3.76013 iter/s, 26.5948s/100 iters), loss = 4.11785
I0511 01:15:07.416957 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.11785 (* 1 = 4.11785 loss)
I0511 01:15:07.416970 39557 sgd_solver.cpp:105] Iteration 117900, lr = 0.01
I0511 01:15:34.030704 39557 solver.cpp:218] Iteration 118000 (3.75745 iter/s, 26.6138s/100 iters), loss = 3.31007
I0511 01:15:34.030884 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.31007 (* 1 = 3.31007 loss)
I0511 01:15:34.030896 39557 sgd_solver.cpp:105] Iteration 118000, lr = 0.01
I0511 01:16:00.637352 39557 solver.cpp:218] Iteration 118100 (3.75848 iter/s, 26.6065s/100 iters), loss = 3.17674
I0511 01:16:00.637413 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.17674 (* 1 = 3.17674 loss)
I0511 01:16:00.637426 39557 sgd_solver.cpp:105] Iteration 118100, lr = 0.01
I0511 01:16:27.213718 39557 solver.cpp:218] Iteration 118200 (3.76275 iter/s, 26.5763s/100 iters), loss = 4.10402
I0511 01:16:27.213871 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.10402 (* 1 = 4.10402 loss)
I0511 01:16:27.213886 39557 sgd_solver.cpp:105] Iteration 118200, lr = 0.01
I0511 01:16:53.875828 39557 solver.cpp:218] Iteration 118300 (3.75066 iter/s, 26.662s/100 iters), loss = 3.8338
I0511 01:16:53.875890 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.8338 (* 1 = 3.8338 loss)
I0511 01:16:53.875905 39557 sgd_solver.cpp:105] Iteration 118300, lr = 0.01
I0511 01:17:20.531666 39557 solver.cpp:218] Iteration 118400 (3.75153 iter/s, 26.6558s/100 iters), loss = 4.03509
I0511 01:17:20.531810 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.03509 (* 1 = 4.03509 loss)
I0511 01:17:20.531823 39557 sgd_solver.cpp:105] Iteration 118400, lr = 0.01
I0511 01:17:47.217175 39557 solver.cpp:218] Iteration 118500 (3.74737 iter/s, 26.6854s/100 iters), loss = 3.7222
I0511 01:17:47.217234 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.7222 (* 1 = 3.7222 loss)
I0511 01:17:47.217244 39557 sgd_solver.cpp:105] Iteration 118500, lr = 0.01
I0511 01:18:13.839790 39557 solver.cpp:218] Iteration 118600 (3.75621 iter/s, 26.6226s/100 iters), loss = 4.20021
I0511 01:18:13.839941 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.20021 (* 1 = 4.20021 loss)
I0511 01:18:13.839956 39557 sgd_solver.cpp:105] Iteration 118600, lr = 0.01
I0511 01:18:40.476405 39557 solver.cpp:218] Iteration 118700 (3.75425 iter/s, 26.6365s/100 iters), loss = 3.44207
I0511 01:18:40.476462 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.44207 (* 1 = 3.44207 loss)
I0511 01:18:40.476472 39557 sgd_solver.cpp:105] Iteration 118700, lr = 0.01
I0511 01:19:07.157220 39557 solver.cpp:218] Iteration 118800 (3.74802 iter/s, 26.6808s/100 iters), loss = 4.0328
I0511 01:19:07.157416 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.0328 (* 1 = 4.0328 loss)
I0511 01:19:07.157431 39557 sgd_solver.cpp:105] Iteration 118800, lr = 0.01
I0511 01:19:33.813673 39557 solver.cpp:218] Iteration 118900 (3.75146 iter/s, 26.6563s/100 iters), loss = 3.53779
I0511 01:19:33.813727 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.53779 (* 1 = 3.53779 loss)
I0511 01:19:33.813740 39557 sgd_solver.cpp:105] Iteration 118900, lr = 0.01
I0511 01:20:00.478493 39557 solver.cpp:218] Iteration 119000 (3.75027 iter/s, 26.6648s/100 iters), loss = 3.39197
I0511 01:20:00.478696 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.39197 (* 1 = 3.39197 loss)
I0511 01:20:00.478711 39557 sgd_solver.cpp:105] Iteration 119000, lr = 0.01
I0511 01:20:27.291220 39557 solver.cpp:218] Iteration 119100 (3.7296 iter/s, 26.8125s/100 iters), loss = 3.74161
I0511 01:20:27.291282 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.74161 (* 1 = 3.74161 loss)
I0511 01:20:27.291296 39557 sgd_solver.cpp:105] Iteration 119100, lr = 0.01
I0511 01:20:53.950652 39557 solver.cpp:218] Iteration 119200 (3.75103 iter/s, 26.6594s/100 iters), loss = 4.0302
I0511 01:20:53.950814 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.0302 (* 1 = 4.0302 loss)
I0511 01:20:53.950829 39557 sgd_solver.cpp:105] Iteration 119200, lr = 0.01
I0511 01:21:20.610942 39557 solver.cpp:218] Iteration 119300 (3.75092 iter/s, 26.6601s/100 iters), loss = 4.31355
I0511 01:21:20.611004 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.31355 (* 1 = 4.31355 loss)
I0511 01:21:20.611019 39557 sgd_solver.cpp:105] Iteration 119300, lr = 0.01
I0511 01:21:47.289505 39557 solver.cpp:218] Iteration 119400 (3.74833 iter/s, 26.6785s/100 iters), loss = 4.3143
I0511 01:21:47.289656 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.3143 (* 1 = 4.3143 loss)
I0511 01:21:47.289671 39557 sgd_solver.cpp:105] Iteration 119400, lr = 0.01
I0511 01:22:13.896986 39557 solver.cpp:218] Iteration 119500 (3.75836 iter/s, 26.6073s/100 iters), loss = 3.56234
I0511 01:22:13.897049 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.56234 (* 1 = 3.56234 loss)
I0511 01:22:13.897068 39557 sgd_solver.cpp:105] Iteration 119500, lr = 0.01
I0511 01:22:40.507575 39557 solver.cpp:218] Iteration 119600 (3.75791 iter/s, 26.6105s/100 iters), loss = 3.91919
I0511 01:22:40.507753 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.91919 (* 1 = 3.91919 loss)
I0511 01:22:40.507768 39557 sgd_solver.cpp:105] Iteration 119600, lr = 0.01
I0511 01:23:07.123459 39557 solver.cpp:218] Iteration 119700 (3.75718 iter/s, 26.6157s/100 iters), loss = 3.67977
I0511 01:23:07.123517 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.67977 (* 1 = 3.67977 loss)
I0511 01:23:07.123530 39557 sgd_solver.cpp:105] Iteration 119700, lr = 0.01
I0511 01:23:33.859215 39557 solver.cpp:218] Iteration 119800 (3.74032 iter/s, 26.7357s/100 iters), loss = 3.86194
I0511 01:23:33.859376 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.86194 (* 1 = 3.86194 loss)
I0511 01:23:33.859388 39557 sgd_solver.cpp:105] Iteration 119800, lr = 0.01
I0511 01:24:00.550422 39557 solver.cpp:218] Iteration 119900 (3.74657 iter/s, 26.6911s/100 iters), loss = 3.19515
I0511 01:24:00.550478 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.19515 (* 1 = 3.19515 loss)
I0511 01:24:00.550493 39557 sgd_solver.cpp:105] Iteration 119900, lr = 0.01
I0511 01:24:26.967391 39557 solver.cpp:447] Snapshotting to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_120000.caffemodel
I0511 01:24:27.005954 39557 sgd_solver.cpp:273] Snapshotting solver state to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_120000.solverstate
I0511 01:24:27.291963 39557 solver.cpp:218] Iteration 120000 (3.73951 iter/s, 26.7415s/100 iters), loss = 3.75881
I0511 01:24:27.292003 39624 sgd_solver.cpp:46] MultiStep Status: Iteration 120000, step = 2
I0511 01:24:27.292021 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.75881 (* 1 = 3.75881 loss)
I0511 01:24:27.292033 39557 sgd_solver.cpp:46] MultiStep Status: Iteration 120000, step = 2
I0511 01:24:27.292037 39557 sgd_solver.cpp:105] Iteration 120000, lr = 0.001
I0511 01:24:53.994415 39557 solver.cpp:218] Iteration 120100 (3.74498 iter/s, 26.7024s/100 iters), loss = 2.98578
I0511 01:24:53.994463 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.98578 (* 1 = 2.98578 loss)
I0511 01:24:53.994477 39557 sgd_solver.cpp:105] Iteration 120100, lr = 0.001
I0511 01:25:20.791016 39557 solver.cpp:218] Iteration 120200 (3.73182 iter/s, 26.7966s/100 iters), loss = 3.67808
I0511 01:25:20.791220 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.67808 (* 1 = 3.67808 loss)
I0511 01:25:20.791236 39557 sgd_solver.cpp:105] Iteration 120200, lr = 0.001
I0511 01:25:47.456249 39557 solver.cpp:218] Iteration 120300 (3.75023 iter/s, 26.665s/100 iters), loss = 2.80051
I0511 01:25:47.456310 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.80051 (* 1 = 2.80051 loss)
I0511 01:25:47.456324 39557 sgd_solver.cpp:105] Iteration 120300, lr = 0.001
I0511 01:26:14.084794 39557 solver.cpp:218] Iteration 120400 (3.75538 iter/s, 26.6285s/100 iters), loss = 3.04603
I0511 01:26:14.084951 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.04603 (* 1 = 3.04603 loss)
I0511 01:26:14.084966 39557 sgd_solver.cpp:105] Iteration 120400, lr = 0.001
I0511 01:26:40.694449 39557 solver.cpp:218] Iteration 120500 (3.75806 iter/s, 26.6095s/100 iters), loss = 3.57198
I0511 01:26:40.694514 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.57198 (* 1 = 3.57198 loss)
I0511 01:26:40.694527 39557 sgd_solver.cpp:105] Iteration 120500, lr = 0.001
I0511 01:27:07.346890 39557 solver.cpp:218] Iteration 120600 (3.75201 iter/s, 26.6524s/100 iters), loss = 2.80355
I0511 01:27:07.347033 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.80355 (* 1 = 2.80355 loss)
I0511 01:27:07.347048 39557 sgd_solver.cpp:105] Iteration 120600, lr = 0.001
I0511 01:27:34.059770 39557 solver.cpp:218] Iteration 120700 (3.74353 iter/s, 26.7127s/100 iters), loss = 3.85316
I0511 01:27:34.059833 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.85316 (* 1 = 3.85316 loss)
I0511 01:27:34.059844 39557 sgd_solver.cpp:105] Iteration 120700, lr = 0.001
I0511 01:28:00.712798 39557 solver.cpp:218] Iteration 120800 (3.75193 iter/s, 26.653s/100 iters), loss = 3.13011
I0511 01:28:00.712970 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.13011 (* 1 = 3.13011 loss)
I0511 01:28:00.712996 39557 sgd_solver.cpp:105] Iteration 120800, lr = 0.001
I0511 01:28:27.437157 39557 solver.cpp:218] Iteration 120900 (3.74193 iter/s, 26.7242s/100 iters), loss = 2.65791
I0511 01:28:27.437220 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.65791 (* 1 = 2.65791 loss)
I0511 01:28:27.437232 39557 sgd_solver.cpp:105] Iteration 120900, lr = 0.001
I0511 01:28:54.147452 39557 solver.cpp:218] Iteration 121000 (3.74388 iter/s, 26.7102s/100 iters), loss = 3.75906
I0511 01:28:54.147609 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.75906 (* 1 = 3.75906 loss)
I0511 01:28:54.147624 39557 sgd_solver.cpp:105] Iteration 121000, lr = 0.001
I0511 01:29:20.883661 39557 solver.cpp:218] Iteration 121100 (3.74027 iter/s, 26.7361s/100 iters), loss = 2.83294
I0511 01:29:20.883719 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.83294 (* 1 = 2.83294 loss)
I0511 01:29:20.883733 39557 sgd_solver.cpp:105] Iteration 121100, lr = 0.001
I0511 01:29:47.590924 39557 solver.cpp:218] Iteration 121200 (3.74431 iter/s, 26.7072s/100 iters), loss = 3.28442
I0511 01:29:47.591085 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.28442 (* 1 = 3.28442 loss)
I0511 01:29:47.591099 39557 sgd_solver.cpp:105] Iteration 121200, lr = 0.001
I0511 01:30:14.243723 39557 solver.cpp:218] Iteration 121300 (3.75197 iter/s, 26.6526s/100 iters), loss = 2.85279
I0511 01:30:14.243786 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.85279 (* 1 = 2.85279 loss)
I0511 01:30:14.243799 39557 sgd_solver.cpp:105] Iteration 121300, lr = 0.001
I0511 01:30:40.994568 39557 solver.cpp:218] Iteration 121400 (3.73821 iter/s, 26.7508s/100 iters), loss = 3.18968
I0511 01:30:40.994781 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.18968 (* 1 = 3.18968 loss)
I0511 01:30:40.994796 39557 sgd_solver.cpp:105] Iteration 121400, lr = 0.001
I0511 01:31:07.654841 39557 solver.cpp:218] Iteration 121500 (3.75093 iter/s, 26.6601s/100 iters), loss = 3.42924
I0511 01:31:07.654901 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.42924 (* 1 = 3.42924 loss)
I0511 01:31:07.654914 39557 sgd_solver.cpp:105] Iteration 121500, lr = 0.001
I0511 01:31:34.365507 39557 solver.cpp:218] Iteration 121600 (3.74383 iter/s, 26.7106s/100 iters), loss = 3.20536
I0511 01:31:34.365674 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.20536 (* 1 = 3.20536 loss)
I0511 01:31:34.365689 39557 sgd_solver.cpp:105] Iteration 121600, lr = 0.001
I0511 01:32:01.046365 39557 solver.cpp:218] Iteration 121700 (3.74803 iter/s, 26.6807s/100 iters), loss = 3.70173
I0511 01:32:01.046427 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.70173 (* 1 = 3.70173 loss)
I0511 01:32:01.046440 39557 sgd_solver.cpp:105] Iteration 121700, lr = 0.001
I0511 01:32:27.651538 39557 solver.cpp:218] Iteration 121800 (3.75867 iter/s, 26.6051s/100 iters), loss = 2.40894
I0511 01:32:27.651695 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.40894 (* 1 = 2.40894 loss)
I0511 01:32:27.651710 39557 sgd_solver.cpp:105] Iteration 121800, lr = 0.001
I0511 01:32:54.484668 39557 solver.cpp:218] Iteration 121900 (3.72676 iter/s, 26.833s/100 iters), loss = 2.85121
I0511 01:32:54.484728 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.85121 (* 1 = 2.85121 loss)
I0511 01:32:54.484743 39557 sgd_solver.cpp:105] Iteration 121900, lr = 0.001
I0511 01:33:21.334944 39557 solver.cpp:218] Iteration 122000 (3.72436 iter/s, 26.8502s/100 iters), loss = 2.91465
I0511 01:33:21.335124 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.91465 (* 1 = 2.91465 loss)
I0511 01:33:21.335139 39557 sgd_solver.cpp:105] Iteration 122000, lr = 0.001
I0511 01:33:48.127439 39557 solver.cpp:218] Iteration 122100 (3.73241 iter/s, 26.7923s/100 iters), loss = 2.8965
I0511 01:33:48.127503 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.8965 (* 1 = 2.8965 loss)
I0511 01:33:48.127518 39557 sgd_solver.cpp:105] Iteration 122100, lr = 0.001
I0511 01:34:14.874889 39557 solver.cpp:218] Iteration 122200 (3.73868 iter/s, 26.7474s/100 iters), loss = 2.46643
I0511 01:34:14.875047 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.46643 (* 1 = 2.46643 loss)
I0511 01:34:14.875064 39557 sgd_solver.cpp:105] Iteration 122200, lr = 0.001
I0511 01:34:41.698181 39557 solver.cpp:218] Iteration 122300 (3.72812 iter/s, 26.8231s/100 iters), loss = 3.36921
I0511 01:34:41.698245 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.36921 (* 1 = 3.36921 loss)
I0511 01:34:41.698257 39557 sgd_solver.cpp:105] Iteration 122300, lr = 0.001
I0511 01:35:08.400779 39557 solver.cpp:218] Iteration 122400 (3.74496 iter/s, 26.7025s/100 iters), loss = 3.3979
I0511 01:35:08.400933 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.3979 (* 1 = 3.3979 loss)
I0511 01:35:08.400951 39557 sgd_solver.cpp:105] Iteration 122400, lr = 0.001
I0511 01:35:35.107869 39557 solver.cpp:218] Iteration 122500 (3.74434 iter/s, 26.7069s/100 iters), loss = 3.06522
I0511 01:35:35.107933 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.06522 (* 1 = 3.06522 loss)
I0511 01:35:35.107944 39557 sgd_solver.cpp:105] Iteration 122500, lr = 0.001
I0511 01:36:01.813767 39557 solver.cpp:218] Iteration 122600 (3.7445 iter/s, 26.7058s/100 iters), loss = 2.92166
I0511 01:36:01.813930 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.92166 (* 1 = 2.92166 loss)
I0511 01:36:01.813947 39557 sgd_solver.cpp:105] Iteration 122600, lr = 0.001
I0511 01:36:28.516257 39557 solver.cpp:218] Iteration 122700 (3.74499 iter/s, 26.7023s/100 iters), loss = 3.33454
I0511 01:36:28.516320 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.33454 (* 1 = 3.33454 loss)
I0511 01:36:28.516333 39557 sgd_solver.cpp:105] Iteration 122700, lr = 0.001
I0511 01:36:55.224647 39557 solver.cpp:218] Iteration 122800 (3.74415 iter/s, 26.7083s/100 iters), loss = 3.53065
I0511 01:36:55.224838 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.53065 (* 1 = 3.53065 loss)
I0511 01:36:55.224855 39557 sgd_solver.cpp:105] Iteration 122800, lr = 0.001
I0511 01:37:21.927824 39557 solver.cpp:218] Iteration 122900 (3.7449 iter/s, 26.703s/100 iters), loss = 3.27089
I0511 01:37:21.927887 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.27089 (* 1 = 3.27089 loss)
I0511 01:37:21.927901 39557 sgd_solver.cpp:105] Iteration 122900, lr = 0.001
I0511 01:37:48.716861 39557 solver.cpp:218] Iteration 123000 (3.73288 iter/s, 26.789s/100 iters), loss = 3.25046
I0511 01:37:48.717034 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.25046 (* 1 = 3.25046 loss)
I0511 01:37:48.717051 39557 sgd_solver.cpp:105] Iteration 123000, lr = 0.001
I0511 01:38:15.350728 39557 solver.cpp:218] Iteration 123100 (3.75464 iter/s, 26.6337s/100 iters), loss = 3.19958
I0511 01:38:15.350795 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.19958 (* 1 = 3.19958 loss)
I0511 01:38:15.350808 39557 sgd_solver.cpp:105] Iteration 123100, lr = 0.001
I0511 01:38:42.037899 39557 solver.cpp:218] Iteration 123200 (3.74713 iter/s, 26.6871s/100 iters), loss = 3.54281
I0511 01:38:42.038059 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.54281 (* 1 = 3.54281 loss)
I0511 01:38:42.038074 39557 sgd_solver.cpp:105] Iteration 123200, lr = 0.001
I0511 01:39:08.704569 39557 solver.cpp:218] Iteration 123300 (3.75002 iter/s, 26.6665s/100 iters), loss = 2.69815
I0511 01:39:08.704634 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.69815 (* 1 = 2.69815 loss)
I0511 01:39:08.704646 39557 sgd_solver.cpp:105] Iteration 123300, lr = 0.001
I0511 01:39:35.426131 39557 solver.cpp:218] Iteration 123400 (3.7423 iter/s, 26.7215s/100 iters), loss = 3.15567
I0511 01:39:35.426281 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.15567 (* 1 = 3.15567 loss)
I0511 01:39:35.426297 39557 sgd_solver.cpp:105] Iteration 123400, lr = 0.001
I0511 01:40:02.160231 39557 solver.cpp:218] Iteration 123500 (3.74056 iter/s, 26.734s/100 iters), loss = 2.94043
I0511 01:40:02.160297 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.94043 (* 1 = 2.94043 loss)
I0511 01:40:02.160311 39557 sgd_solver.cpp:105] Iteration 123500, lr = 0.001
I0511 01:40:28.922021 39557 solver.cpp:218] Iteration 123600 (3.73668 iter/s, 26.7617s/100 iters), loss = 3.06995
I0511 01:40:28.922188 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.06995 (* 1 = 3.06995 loss)
I0511 01:40:28.922204 39557 sgd_solver.cpp:105] Iteration 123600, lr = 0.001
I0511 01:40:55.736057 39557 solver.cpp:218] Iteration 123700 (3.72941 iter/s, 26.8139s/100 iters), loss = 3.09344
I0511 01:40:55.736120 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.09344 (* 1 = 3.09344 loss)
I0511 01:40:55.736138 39557 sgd_solver.cpp:105] Iteration 123700, lr = 0.001
I0511 01:41:22.458937 39557 solver.cpp:218] Iteration 123800 (3.74212 iter/s, 26.7228s/100 iters), loss = 3.28384
I0511 01:41:22.459095 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.28384 (* 1 = 3.28384 loss)
I0511 01:41:22.459111 39557 sgd_solver.cpp:105] Iteration 123800, lr = 0.001
I0511 01:41:49.215000 39557 solver.cpp:218] Iteration 123900 (3.73749 iter/s, 26.7559s/100 iters), loss = 3.27581
I0511 01:41:49.215065 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.27581 (* 1 = 3.27581 loss)
I0511 01:41:49.215080 39557 sgd_solver.cpp:105] Iteration 123900, lr = 0.001
I0511 01:42:15.851800 39557 solver.cpp:218] Iteration 124000 (3.75421 iter/s, 26.6367s/100 iters), loss = 3.08133
I0511 01:42:15.851984 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.08133 (* 1 = 3.08133 loss)
I0511 01:42:15.852003 39557 sgd_solver.cpp:105] Iteration 124000, lr = 0.001
I0511 01:42:42.588213 39557 solver.cpp:218] Iteration 124100 (3.74025 iter/s, 26.7362s/100 iters), loss = 2.51092
I0511 01:42:42.588315 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.51092 (* 1 = 2.51092 loss)
I0511 01:42:42.588351 39557 sgd_solver.cpp:105] Iteration 124100, lr = 0.001
I0511 01:43:09.317221 39557 solver.cpp:218] Iteration 124200 (3.74127 iter/s, 26.7289s/100 iters), loss = 3.5379
I0511 01:43:09.317440 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.5379 (* 1 = 3.5379 loss)
I0511 01:43:09.317459 39557 sgd_solver.cpp:105] Iteration 124200, lr = 0.001
I0511 01:43:35.973690 39557 solver.cpp:218] Iteration 124300 (3.75147 iter/s, 26.6563s/100 iters), loss = 2.77069
I0511 01:43:35.973758 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.77069 (* 1 = 2.77069 loss)
I0511 01:43:35.973773 39557 sgd_solver.cpp:105] Iteration 124300, lr = 0.001
I0511 01:44:02.654330 39557 solver.cpp:218] Iteration 124400 (3.74804 iter/s, 26.6806s/100 iters), loss = 3.76116
I0511 01:44:02.654530 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.76116 (* 1 = 3.76116 loss)
I0511 01:44:02.654544 39557 sgd_solver.cpp:105] Iteration 124400, lr = 0.001
I0511 01:44:29.300485 39557 solver.cpp:218] Iteration 124500 (3.75291 iter/s, 26.646s/100 iters), loss = 3.33297
I0511 01:44:29.300549 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.33297 (* 1 = 3.33297 loss)
I0511 01:44:29.300563 39557 sgd_solver.cpp:105] Iteration 124500, lr = 0.001
I0511 01:44:55.935976 39557 solver.cpp:218] Iteration 124600 (3.7544 iter/s, 26.6354s/100 iters), loss = 3.55011
I0511 01:44:55.936132 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.55011 (* 1 = 3.55011 loss)
I0511 01:44:55.936146 39557 sgd_solver.cpp:105] Iteration 124600, lr = 0.001
I0511 01:45:22.608741 39557 solver.cpp:218] Iteration 124700 (3.74916 iter/s, 26.6726s/100 iters), loss = 3.15515
I0511 01:45:22.608806 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.15515 (* 1 = 3.15515 loss)
I0511 01:45:22.608816 39557 sgd_solver.cpp:105] Iteration 124700, lr = 0.001
I0511 01:45:49.338304 39557 solver.cpp:218] Iteration 124800 (3.74118 iter/s, 26.7295s/100 iters), loss = 2.9094
I0511 01:45:49.338459 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.9094 (* 1 = 2.9094 loss)
I0511 01:45:49.338475 39557 sgd_solver.cpp:105] Iteration 124800, lr = 0.001
I0511 01:46:15.956451 39557 solver.cpp:218] Iteration 124900 (3.75686 iter/s, 26.618s/100 iters), loss = 2.56976
I0511 01:46:15.956514 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.56976 (* 1 = 2.56976 loss)
I0511 01:46:15.956526 39557 sgd_solver.cpp:105] Iteration 124900, lr = 0.001
I0511 01:46:42.370820 39557 solver.cpp:447] Snapshotting to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_125000.caffemodel
I0511 01:46:42.410785 39557 sgd_solver.cpp:273] Snapshotting solver state to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_125000.solverstate
I0511 01:46:42.694316 39557 solver.cpp:218] Iteration 125000 (3.74002 iter/s, 26.7378s/100 iters), loss = 3.36778
I0511 01:46:42.694376 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.36778 (* 1 = 3.36778 loss)
I0511 01:46:42.694386 39557 sgd_solver.cpp:105] Iteration 125000, lr = 0.001
I0511 01:47:09.412494 39557 solver.cpp:218] Iteration 125100 (3.74278 iter/s, 26.7181s/100 iters), loss = 3.12643
I0511 01:47:09.412569 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.12643 (* 1 = 3.12643 loss)
I0511 01:47:09.412581 39557 sgd_solver.cpp:105] Iteration 125100, lr = 0.001
I0511 01:47:36.019785 39557 solver.cpp:218] Iteration 125200 (3.75838 iter/s, 26.6072s/100 iters), loss = 2.97398
I0511 01:47:36.019955 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.97398 (* 1 = 2.97398 loss)
I0511 01:47:36.019970 39557 sgd_solver.cpp:105] Iteration 125200, lr = 0.001
I0511 01:48:02.648622 39557 solver.cpp:218] Iteration 125300 (3.75535 iter/s, 26.6287s/100 iters), loss = 2.67601
I0511 01:48:02.648677 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.67601 (* 1 = 2.67601 loss)
I0511 01:48:02.648689 39557 sgd_solver.cpp:105] Iteration 125300, lr = 0.001
I0511 01:48:29.286339 39557 solver.cpp:218] Iteration 125400 (3.75408 iter/s, 26.6377s/100 iters), loss = 3.50866
I0511 01:48:29.286533 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.50866 (* 1 = 3.50866 loss)
I0511 01:48:29.286547 39557 sgd_solver.cpp:105] Iteration 125400, lr = 0.001
I0511 01:48:55.883720 39557 solver.cpp:218] Iteration 125500 (3.7598 iter/s, 26.5972s/100 iters), loss = 2.74513
I0511 01:48:55.883780 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.74513 (* 1 = 2.74513 loss)
I0511 01:48:55.883792 39557 sgd_solver.cpp:105] Iteration 125500, lr = 0.001
I0511 01:49:22.600114 39557 solver.cpp:218] Iteration 125600 (3.74303 iter/s, 26.7163s/100 iters), loss = 3.05931
I0511 01:49:22.600288 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.05931 (* 1 = 3.05931 loss)
I0511 01:49:22.600303 39557 sgd_solver.cpp:105] Iteration 125600, lr = 0.001
I0511 01:49:49.236825 39557 solver.cpp:218] Iteration 125700 (3.75424 iter/s, 26.6365s/100 iters), loss = 3.14006
I0511 01:49:49.236887 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.14006 (* 1 = 3.14006 loss)
I0511 01:49:49.236901 39557 sgd_solver.cpp:105] Iteration 125700, lr = 0.001
I0511 01:50:15.857203 39557 solver.cpp:218] Iteration 125800 (3.75653 iter/s, 26.6203s/100 iters), loss = 3.51923
I0511 01:50:15.857360 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.51923 (* 1 = 3.51923 loss)
I0511 01:50:15.857375 39557 sgd_solver.cpp:105] Iteration 125800, lr = 0.001
I0511 01:50:42.508015 39557 solver.cpp:218] Iteration 125900 (3.75225 iter/s, 26.6507s/100 iters), loss = 2.82482
I0511 01:50:42.508088 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.82482 (* 1 = 2.82482 loss)
I0511 01:50:42.508100 39557 sgd_solver.cpp:105] Iteration 125900, lr = 0.001
I0511 01:51:09.156452 39557 solver.cpp:218] Iteration 126000 (3.75257 iter/s, 26.6484s/100 iters), loss = 3.42639
I0511 01:51:09.156606 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.42639 (* 1 = 3.42639 loss)
I0511 01:51:09.156621 39557 sgd_solver.cpp:105] Iteration 126000, lr = 0.001
I0511 01:51:35.822878 39557 solver.cpp:218] Iteration 126100 (3.75018 iter/s, 26.6654s/100 iters), loss = 2.46043
I0511 01:51:35.822937 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.46043 (* 1 = 2.46043 loss)
I0511 01:51:35.822947 39557 sgd_solver.cpp:105] Iteration 126100, lr = 0.001
I0511 01:52:02.419615 39557 solver.cpp:218] Iteration 126200 (3.76014 iter/s, 26.5948s/100 iters), loss = 3.25593
I0511 01:52:02.419766 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.25593 (* 1 = 3.25593 loss)
I0511 01:52:02.419781 39557 sgd_solver.cpp:105] Iteration 126200, lr = 0.001
I0511 01:52:29.003197 39557 solver.cpp:218] Iteration 126300 (3.762 iter/s, 26.5816s/100 iters), loss = 3.12441
I0511 01:52:29.003255 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.12441 (* 1 = 3.12441 loss)
I0511 01:52:29.003267 39557 sgd_solver.cpp:105] Iteration 126300, lr = 0.001
I0511 01:52:55.620605 39557 solver.cpp:218] Iteration 126400 (3.75719 iter/s, 26.6156s/100 iters), loss = 2.72614
I0511 01:52:55.620756 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.72614 (* 1 = 2.72614 loss)
I0511 01:52:55.620772 39557 sgd_solver.cpp:105] Iteration 126400, lr = 0.001
I0511 01:53:22.269734 39557 solver.cpp:218] Iteration 126500 (3.75272 iter/s, 26.6473s/100 iters), loss = 2.56228
I0511 01:53:22.269793 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.56228 (* 1 = 2.56228 loss)
I0511 01:53:22.269803 39557 sgd_solver.cpp:105] Iteration 126500, lr = 0.001
I0511 01:53:48.863621 39557 solver.cpp:218] Iteration 126600 (3.7605 iter/s, 26.5922s/100 iters), loss = 2.89912
I0511 01:53:48.863770 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.89912 (* 1 = 2.89912 loss)
I0511 01:53:48.863785 39557 sgd_solver.cpp:105] Iteration 126600, lr = 0.001
I0511 01:54:15.454318 39557 solver.cpp:218] Iteration 126700 (3.76095 iter/s, 26.589s/100 iters), loss = 3.24413
I0511 01:54:15.454376 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.24413 (* 1 = 3.24413 loss)
I0511 01:54:15.454386 39557 sgd_solver.cpp:105] Iteration 126700, lr = 0.001
I0511 01:54:42.076309 39557 solver.cpp:218] Iteration 126800 (3.75651 iter/s, 26.6205s/100 iters), loss = 3.69994
I0511 01:54:42.076529 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.69994 (* 1 = 3.69994 loss)
I0511 01:54:42.076545 39557 sgd_solver.cpp:105] Iteration 126800, lr = 0.001
I0511 01:55:08.762363 39557 solver.cpp:218] Iteration 126900 (3.7475 iter/s, 26.6844s/100 iters), loss = 3.17013
I0511 01:55:08.762421 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.17013 (* 1 = 3.17013 loss)
I0511 01:55:08.762432 39557 sgd_solver.cpp:105] Iteration 126900, lr = 0.001
I0511 01:55:35.347468 39557 solver.cpp:218] Iteration 127000 (3.7617 iter/s, 26.5837s/100 iters), loss = 3.34826
I0511 01:55:35.347632 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.34826 (* 1 = 3.34826 loss)
I0511 01:55:35.347648 39557 sgd_solver.cpp:105] Iteration 127000, lr = 0.001
I0511 01:56:01.990141 39557 solver.cpp:218] Iteration 127100 (3.75358 iter/s, 26.6412s/100 iters), loss = 3.06572
I0511 01:56:01.990201 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.06572 (* 1 = 3.06572 loss)
I0511 01:56:01.990211 39557 sgd_solver.cpp:105] Iteration 127100, lr = 0.001
I0511 01:56:28.676977 39557 solver.cpp:218] Iteration 127200 (3.74735 iter/s, 26.6855s/100 iters), loss = 2.66532
I0511 01:56:28.677134 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.66532 (* 1 = 2.66532 loss)
I0511 01:56:28.677148 39557 sgd_solver.cpp:105] Iteration 127200, lr = 0.001
I0511 01:56:55.321125 39557 solver.cpp:218] Iteration 127300 (3.75336 iter/s, 26.6428s/100 iters), loss = 3.30223
I0511 01:56:55.321187 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.30223 (* 1 = 3.30223 loss)
I0511 01:56:55.321208 39557 sgd_solver.cpp:105] Iteration 127300, lr = 0.001
I0511 01:57:21.995344 39557 solver.cpp:218] Iteration 127400 (3.74911 iter/s, 26.673s/100 iters), loss = 3.78942
I0511 01:57:21.995514 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.78942 (* 1 = 3.78942 loss)
I0511 01:57:21.995529 39557 sgd_solver.cpp:105] Iteration 127400, lr = 0.001
I0511 01:57:48.650843 39557 solver.cpp:218] Iteration 127500 (3.75175 iter/s, 26.6542s/100 iters), loss = 2.53821
I0511 01:57:48.650912 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.53821 (* 1 = 2.53821 loss)
I0511 01:57:48.650921 39557 sgd_solver.cpp:105] Iteration 127500, lr = 0.001
I0511 01:58:15.299592 39557 solver.cpp:218] Iteration 127600 (3.75268 iter/s, 26.6476s/100 iters), loss = 3.55134
I0511 01:58:15.299762 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.55134 (* 1 = 3.55134 loss)
I0511 01:58:15.299778 39557 sgd_solver.cpp:105] Iteration 127600, lr = 0.001
I0511 01:58:41.898636 39557 solver.cpp:218] Iteration 127700 (3.7597 iter/s, 26.5979s/100 iters), loss = 3.07791
I0511 01:58:41.898694 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.07791 (* 1 = 3.07791 loss)
I0511 01:58:41.898705 39557 sgd_solver.cpp:105] Iteration 127700, lr = 0.001
I0511 01:59:08.500017 39557 solver.cpp:218] Iteration 127800 (3.75935 iter/s, 26.6003s/100 iters), loss = 2.59492
I0511 01:59:08.500206 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.59492 (* 1 = 2.59492 loss)
I0511 01:59:08.500221 39557 sgd_solver.cpp:105] Iteration 127800, lr = 0.001
I0511 01:59:35.097429 39557 solver.cpp:218] Iteration 127900 (3.75992 iter/s, 26.5963s/100 iters), loss = 2.87915
I0511 01:59:35.097479 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.87915 (* 1 = 2.87915 loss)
I0511 01:59:35.097491 39557 sgd_solver.cpp:105] Iteration 127900, lr = 0.001
I0511 02:00:01.687674 39557 solver.cpp:218] Iteration 128000 (3.76091 iter/s, 26.5893s/100 iters), loss = 2.69441
I0511 02:00:01.687832 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.69441 (* 1 = 2.69441 loss)
I0511 02:00:01.687860 39557 sgd_solver.cpp:105] Iteration 128000, lr = 0.001
I0511 02:00:28.336786 39557 solver.cpp:218] Iteration 128100 (3.75262 iter/s, 26.6481s/100 iters), loss = 2.94897
I0511 02:00:28.336848 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.94897 (* 1 = 2.94897 loss)
I0511 02:00:28.336863 39557 sgd_solver.cpp:105] Iteration 128100, lr = 0.001
I0511 02:00:54.946384 39557 solver.cpp:218] Iteration 128200 (3.75817 iter/s, 26.6087s/100 iters), loss = 2.36216
I0511 02:00:54.946614 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.36216 (* 1 = 2.36216 loss)
I0511 02:00:54.946631 39557 sgd_solver.cpp:105] Iteration 128200, lr = 0.001
I0511 02:01:21.781144 39557 solver.cpp:218] Iteration 128300 (3.72666 iter/s, 26.8337s/100 iters), loss = 3.06156
I0511 02:01:21.781213 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.06156 (* 1 = 3.06156 loss)
I0511 02:01:21.781227 39557 sgd_solver.cpp:105] Iteration 128300, lr = 0.001
I0511 02:01:48.511512 39557 solver.cpp:218] Iteration 128400 (3.74118 iter/s, 26.7295s/100 iters), loss = 2.58983
I0511 02:01:48.511672 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.58983 (* 1 = 2.58983 loss)
I0511 02:01:48.511708 39557 sgd_solver.cpp:105] Iteration 128400, lr = 0.001
I0511 02:02:15.285596 39557 solver.cpp:218] Iteration 128500 (3.73508 iter/s, 26.7732s/100 iters), loss = 2.76971
I0511 02:02:15.285656 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.76971 (* 1 = 2.76971 loss)
I0511 02:02:15.285684 39557 sgd_solver.cpp:105] Iteration 128500, lr = 0.001
I0511 02:02:42.052971 39557 solver.cpp:218] Iteration 128600 (3.736 iter/s, 26.7666s/100 iters), loss = 2.67366
I0511 02:02:42.053159 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.67366 (* 1 = 2.67366 loss)
I0511 02:02:42.053180 39557 sgd_solver.cpp:105] Iteration 128600, lr = 0.001
I0511 02:03:08.731429 39557 solver.cpp:218] Iteration 128700 (3.74847 iter/s, 26.6776s/100 iters), loss = 4.04143
I0511 02:03:08.731487 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.04143 (* 1 = 4.04143 loss)
I0511 02:03:08.731516 39557 sgd_solver.cpp:105] Iteration 128700, lr = 0.001
I0511 02:03:35.374153 39557 solver.cpp:218] Iteration 128800 (3.75348 iter/s, 26.642s/100 iters), loss = 2.13358
I0511 02:03:35.374318 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.13358 (* 1 = 2.13358 loss)
I0511 02:03:35.374341 39557 sgd_solver.cpp:105] Iteration 128800, lr = 0.001
I0511 02:04:02.048182 39557 solver.cpp:218] Iteration 128900 (3.74908 iter/s, 26.6732s/100 iters), loss = 3.02084
I0511 02:04:02.048249 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.02084 (* 1 = 3.02084 loss)
I0511 02:04:02.048288 39557 sgd_solver.cpp:105] Iteration 128900, lr = 0.001
I0511 02:04:28.689123 39557 solver.cpp:218] Iteration 129000 (3.75372 iter/s, 26.6402s/100 iters), loss = 3.05757
I0511 02:04:28.689280 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.05757 (* 1 = 3.05757 loss)
I0511 02:04:28.689311 39557 sgd_solver.cpp:105] Iteration 129000, lr = 0.001
I0511 02:04:55.355756 39557 solver.cpp:218] Iteration 129100 (3.75011 iter/s, 26.6658s/100 iters), loss = 3.90874
I0511 02:04:55.355809 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.90874 (* 1 = 3.90874 loss)
I0511 02:04:55.355834 39557 sgd_solver.cpp:105] Iteration 129100, lr = 0.001
I0511 02:05:22.041390 39557 solver.cpp:218] Iteration 129200 (3.74743 iter/s, 26.685s/100 iters), loss = 3.2626
I0511 02:05:22.041570 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.2626 (* 1 = 3.2626 loss)
I0511 02:05:22.041585 39557 sgd_solver.cpp:105] Iteration 129200, lr = 0.001
I0511 02:05:48.660737 39557 solver.cpp:218] Iteration 129300 (3.75678 iter/s, 26.6186s/100 iters), loss = 2.99214
I0511 02:05:48.660799 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.99214 (* 1 = 2.99214 loss)
I0511 02:05:48.660809 39557 sgd_solver.cpp:105] Iteration 129300, lr = 0.001
I0511 02:06:15.285024 39557 solver.cpp:218] Iteration 129400 (3.75606 iter/s, 26.6236s/100 iters), loss = 3.71031
I0511 02:06:15.285253 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.71031 (* 1 = 3.71031 loss)
I0511 02:06:15.285269 39557 sgd_solver.cpp:105] Iteration 129400, lr = 0.001
I0511 02:06:41.850486 39557 solver.cpp:218] Iteration 129500 (3.7644 iter/s, 26.5647s/100 iters), loss = 3.30266
I0511 02:06:41.850549 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.30266 (* 1 = 3.30266 loss)
I0511 02:06:41.850563 39557 sgd_solver.cpp:105] Iteration 129500, lr = 0.001
I0511 02:07:08.438709 39557 solver.cpp:218] Iteration 129600 (3.76115 iter/s, 26.5876s/100 iters), loss = 2.81866
I0511 02:07:08.438884 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.81866 (* 1 = 2.81866 loss)
I0511 02:07:08.438899 39557 sgd_solver.cpp:105] Iteration 129600, lr = 0.001
I0511 02:07:35.120391 39557 solver.cpp:218] Iteration 129700 (3.74799 iter/s, 26.681s/100 iters), loss = 3.07274
I0511 02:07:35.120455 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.07274 (* 1 = 3.07274 loss)
I0511 02:07:35.120467 39557 sgd_solver.cpp:105] Iteration 129700, lr = 0.001
I0511 02:08:01.736681 39557 solver.cpp:218] Iteration 129800 (3.75718 iter/s, 26.6157s/100 iters), loss = 2.92396
I0511 02:08:01.736837 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.92396 (* 1 = 2.92396 loss)
I0511 02:08:01.736852 39557 sgd_solver.cpp:105] Iteration 129800, lr = 0.001
I0511 02:08:28.318258 39557 solver.cpp:218] Iteration 129900 (3.7621 iter/s, 26.5809s/100 iters), loss = 2.99211
I0511 02:08:28.318315 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.99211 (* 1 = 2.99211 loss)
I0511 02:08:28.318325 39557 sgd_solver.cpp:105] Iteration 129900, lr = 0.001
I0511 02:08:54.668345 39557 solver.cpp:447] Snapshotting to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_130000.caffemodel
I0511 02:08:54.708467 39557 sgd_solver.cpp:273] Snapshotting solver state to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_130000.solverstate
I0511 02:08:54.998314 39557 solver.cpp:218] Iteration 130000 (3.74819 iter/s, 26.6795s/100 iters), loss = 3.29098
I0511 02:08:54.998368 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.29098 (* 1 = 3.29098 loss)
I0511 02:08:54.998379 39557 sgd_solver.cpp:105] Iteration 130000, lr = 0.001
I0511 02:09:21.646757 39557 solver.cpp:218] Iteration 130100 (3.75264 iter/s, 26.6479s/100 iters), loss = 2.92546
I0511 02:09:21.646816 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.92546 (* 1 = 2.92546 loss)
I0511 02:09:21.646827 39557 sgd_solver.cpp:105] Iteration 130100, lr = 0.001
I0511 02:09:48.237123 39557 solver.cpp:218] Iteration 130200 (3.76083 iter/s, 26.5898s/100 iters), loss = 2.97822
I0511 02:09:48.237270 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.97822 (* 1 = 2.97822 loss)
I0511 02:09:48.237287 39557 sgd_solver.cpp:105] Iteration 130200, lr = 0.001
I0511 02:10:14.838152 39557 solver.cpp:218] Iteration 130300 (3.75934 iter/s, 26.6004s/100 iters), loss = 3.51036
I0511 02:10:14.838208 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.51036 (* 1 = 3.51036 loss)
I0511 02:10:14.838218 39557 sgd_solver.cpp:105] Iteration 130300, lr = 0.001
I0511 02:10:41.430927 39557 solver.cpp:218] Iteration 130400 (3.76049 iter/s, 26.5923s/100 iters), loss = 3.50761
I0511 02:10:41.431098 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.50761 (* 1 = 3.50761 loss)
I0511 02:10:41.431114 39557 sgd_solver.cpp:105] Iteration 130400, lr = 0.001
I0511 02:11:08.043910 39557 solver.cpp:218] Iteration 130500 (3.75765 iter/s, 26.6124s/100 iters), loss = 2.62762
I0511 02:11:08.043964 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.62762 (* 1 = 2.62762 loss)
I0511 02:11:08.043975 39557 sgd_solver.cpp:105] Iteration 130500, lr = 0.001
I0511 02:11:34.643858 39557 solver.cpp:218] Iteration 130600 (3.75947 iter/s, 26.5995s/100 iters), loss = 3.62962
I0511 02:11:34.644011 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.62962 (* 1 = 3.62962 loss)
I0511 02:11:34.644027 39557 sgd_solver.cpp:105] Iteration 130600, lr = 0.001
I0511 02:12:01.303212 39557 solver.cpp:218] Iteration 130700 (3.75111 iter/s, 26.6588s/100 iters), loss = 3.13512
I0511 02:12:01.303269 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.13512 (* 1 = 3.13512 loss)
I0511 02:12:01.303280 39557 sgd_solver.cpp:105] Iteration 130700, lr = 0.001
I0511 02:12:27.860347 39557 solver.cpp:218] Iteration 130800 (3.76553 iter/s, 26.5567s/100 iters), loss = 3.68184
I0511 02:12:27.860549 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.68184 (* 1 = 3.68184 loss)
I0511 02:12:27.860565 39557 sgd_solver.cpp:105] Iteration 130800, lr = 0.001
I0511 02:12:54.460616 39557 solver.cpp:218] Iteration 130900 (3.75945 iter/s, 26.5997s/100 iters), loss = 3.43055
I0511 02:12:54.460674 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.43055 (* 1 = 3.43055 loss)
I0511 02:12:54.460683 39557 sgd_solver.cpp:105] Iteration 130900, lr = 0.001
I0511 02:13:21.026453 39557 solver.cpp:218] Iteration 131000 (3.7643 iter/s, 26.5654s/100 iters), loss = 2.42183
I0511 02:13:21.026618 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.42183 (* 1 = 2.42183 loss)
I0511 02:13:21.026634 39557 sgd_solver.cpp:105] Iteration 131000, lr = 0.001
I0511 02:13:47.634529 39557 solver.cpp:218] Iteration 131100 (3.75834 iter/s, 26.6075s/100 iters), loss = 2.98529
I0511 02:13:47.634588 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.98529 (* 1 = 2.98529 loss)
I0511 02:13:47.634599 39557 sgd_solver.cpp:105] Iteration 131100, lr = 0.001
I0511 02:14:14.335824 39557 solver.cpp:218] Iteration 131200 (3.7452 iter/s, 26.7009s/100 iters), loss = 2.71925
I0511 02:14:14.335989 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.71925 (* 1 = 2.71925 loss)
I0511 02:14:14.336007 39557 sgd_solver.cpp:105] Iteration 131200, lr = 0.001
I0511 02:14:41.118791 39557 solver.cpp:218] Iteration 131300 (3.73379 iter/s, 26.7824s/100 iters), loss = 3.00663
I0511 02:14:41.118857 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.00663 (* 1 = 3.00663 loss)
I0511 02:14:41.118872 39557 sgd_solver.cpp:105] Iteration 131300, lr = 0.001
I0511 02:15:07.826689 39557 solver.cpp:218] Iteration 131400 (3.74427 iter/s, 26.7075s/100 iters), loss = 3.12786
I0511 02:15:07.826851 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.12786 (* 1 = 3.12786 loss)
I0511 02:15:07.826879 39557 sgd_solver.cpp:105] Iteration 131400, lr = 0.001
I0511 02:15:34.568217 39557 solver.cpp:218] Iteration 131500 (3.73958 iter/s, 26.741s/100 iters), loss = 2.73787
I0511 02:15:34.568279 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.73787 (* 1 = 2.73787 loss)
I0511 02:15:34.568292 39557 sgd_solver.cpp:105] Iteration 131500, lr = 0.001
I0511 02:16:01.251812 39557 solver.cpp:218] Iteration 131600 (3.74768 iter/s, 26.6832s/100 iters), loss = 2.925
I0511 02:16:01.251991 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.925 (* 1 = 2.925 loss)
I0511 02:16:01.252005 39557 sgd_solver.cpp:105] Iteration 131600, lr = 0.001
I0511 02:16:27.910773 39557 solver.cpp:218] Iteration 131700 (3.75116 iter/s, 26.6584s/100 iters), loss = 2.62948
I0511 02:16:27.910832 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.62948 (* 1 = 2.62948 loss)
I0511 02:16:27.910845 39557 sgd_solver.cpp:105] Iteration 131700, lr = 0.001
I0511 02:16:54.549784 39557 solver.cpp:218] Iteration 131800 (3.75395 iter/s, 26.6386s/100 iters), loss = 3.69786
I0511 02:16:54.549929 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.69786 (* 1 = 3.69786 loss)
I0511 02:16:54.549944 39557 sgd_solver.cpp:105] Iteration 131800, lr = 0.001
I0511 02:17:21.196137 39557 solver.cpp:218] Iteration 131900 (3.75293 iter/s, 26.6459s/100 iters), loss = 3.47244
I0511 02:17:21.196199 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.47244 (* 1 = 3.47244 loss)
I0511 02:17:21.196213 39557 sgd_solver.cpp:105] Iteration 131900, lr = 0.001
I0511 02:17:47.821630 39557 solver.cpp:218] Iteration 132000 (3.75586 iter/s, 26.6251s/100 iters), loss = 3.52189
I0511 02:17:47.821830 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.52189 (* 1 = 3.52189 loss)
I0511 02:17:47.821847 39557 sgd_solver.cpp:105] Iteration 132000, lr = 0.001
I0511 02:18:14.470779 39557 solver.cpp:218] Iteration 132100 (3.75254 iter/s, 26.6486s/100 iters), loss = 2.49624
I0511 02:18:14.470844 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.49624 (* 1 = 2.49624 loss)
I0511 02:18:14.470857 39557 sgd_solver.cpp:105] Iteration 132100, lr = 0.001
I0511 02:18:41.055321 39557 solver.cpp:218] Iteration 132200 (3.76164 iter/s, 26.5841s/100 iters), loss = 3.52527
I0511 02:18:41.055497 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.52527 (* 1 = 3.52527 loss)
I0511 02:18:41.055511 39557 sgd_solver.cpp:105] Iteration 132200, lr = 0.001
I0511 02:19:07.644726 39557 solver.cpp:218] Iteration 132300 (3.76097 iter/s, 26.5889s/100 iters), loss = 2.87019
I0511 02:19:07.644786 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.87019 (* 1 = 2.87019 loss)
I0511 02:19:07.644798 39557 sgd_solver.cpp:105] Iteration 132300, lr = 0.001
I0511 02:19:34.254374 39557 solver.cpp:218] Iteration 132400 (3.75809 iter/s, 26.6093s/100 iters), loss = 3.38132
I0511 02:19:34.254532 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.38132 (* 1 = 3.38132 loss)
I0511 02:19:34.254547 39557 sgd_solver.cpp:105] Iteration 132400, lr = 0.001
I0511 02:20:00.845541 39557 solver.cpp:218] Iteration 132500 (3.76072 iter/s, 26.5907s/100 iters), loss = 4.09325
I0511 02:20:00.845602 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.09325 (* 1 = 4.09325 loss)
I0511 02:20:00.845612 39557 sgd_solver.cpp:105] Iteration 132500, lr = 0.001
I0511 02:20:27.428545 39557 solver.cpp:218] Iteration 132600 (3.76186 iter/s, 26.5826s/100 iters), loss = 3.19539
I0511 02:20:27.428709 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.19539 (* 1 = 3.19539 loss)
I0511 02:20:27.428725 39557 sgd_solver.cpp:105] Iteration 132600, lr = 0.001
I0511 02:20:54.011668 39557 solver.cpp:218] Iteration 132700 (3.76185 iter/s, 26.5826s/100 iters), loss = 2.77775
I0511 02:20:54.011728 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.77775 (* 1 = 2.77775 loss)
I0511 02:20:54.011739 39557 sgd_solver.cpp:105] Iteration 132700, lr = 0.001
I0511 02:21:20.665335 39557 solver.cpp:218] Iteration 132800 (3.75188 iter/s, 26.6533s/100 iters), loss = 3.16078
I0511 02:21:20.665518 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.16078 (* 1 = 3.16078 loss)
I0511 02:21:20.665534 39557 sgd_solver.cpp:105] Iteration 132800, lr = 0.001
I0511 02:21:47.306279 39557 solver.cpp:218] Iteration 132900 (3.75369 iter/s, 26.6405s/100 iters), loss = 3.42699
I0511 02:21:47.306344 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.42699 (* 1 = 3.42699 loss)
I0511 02:21:47.306357 39557 sgd_solver.cpp:105] Iteration 132900, lr = 0.001
I0511 02:22:13.878772 39557 solver.cpp:218] Iteration 133000 (3.76334 iter/s, 26.5721s/100 iters), loss = 2.66312
I0511 02:22:13.878937 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.66312 (* 1 = 2.66312 loss)
I0511 02:22:13.878953 39557 sgd_solver.cpp:105] Iteration 133000, lr = 0.001
I0511 02:22:40.562942 39557 solver.cpp:218] Iteration 133100 (3.74761 iter/s, 26.6837s/100 iters), loss = 2.55969
I0511 02:22:40.563009 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.55969 (* 1 = 2.55969 loss)
I0511 02:22:40.563021 39557 sgd_solver.cpp:105] Iteration 133100, lr = 0.001
I0511 02:23:07.232548 39557 solver.cpp:218] Iteration 133200 (3.74964 iter/s, 26.6692s/100 iters), loss = 3.28791
I0511 02:23:07.232724 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.28791 (* 1 = 3.28791 loss)
I0511 02:23:07.232740 39557 sgd_solver.cpp:105] Iteration 133200, lr = 0.001
I0511 02:23:33.869508 39557 solver.cpp:218] Iteration 133300 (3.75425 iter/s, 26.6365s/100 iters), loss = 2.67156
I0511 02:23:33.869560 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.67156 (* 1 = 2.67156 loss)
I0511 02:23:33.869572 39557 sgd_solver.cpp:105] Iteration 133300, lr = 0.001
I0511 02:24:00.515300 39557 solver.cpp:218] Iteration 133400 (3.75299 iter/s, 26.6454s/100 iters), loss = 3.17136
I0511 02:24:00.515512 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.17136 (* 1 = 3.17136 loss)
I0511 02:24:00.515539 39557 sgd_solver.cpp:105] Iteration 133400, lr = 0.001
I0511 02:24:27.123067 39557 solver.cpp:218] Iteration 133500 (3.75837 iter/s, 26.6073s/100 iters), loss = 3.69015
I0511 02:24:27.123116 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.69015 (* 1 = 3.69015 loss)
I0511 02:24:27.123128 39557 sgd_solver.cpp:105] Iteration 133500, lr = 0.001
I0511 02:24:53.802402 39557 solver.cpp:218] Iteration 133600 (3.74827 iter/s, 26.679s/100 iters), loss = 3.66181
I0511 02:24:53.802551 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.66181 (* 1 = 3.66181 loss)
I0511 02:24:53.802578 39557 sgd_solver.cpp:105] Iteration 133600, lr = 0.001
I0511 02:25:20.421948 39557 solver.cpp:218] Iteration 133700 (3.7567 iter/s, 26.6191s/100 iters), loss = 3.7666
I0511 02:25:20.421993 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.7666 (* 1 = 3.7666 loss)
I0511 02:25:20.422003 39557 sgd_solver.cpp:105] Iteration 133700, lr = 0.001
I0511 02:25:47.013924 39557 solver.cpp:218] Iteration 133800 (3.76053 iter/s, 26.592s/100 iters), loss = 3.08643
I0511 02:25:47.014072 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.08643 (* 1 = 3.08643 loss)
I0511 02:25:47.014098 39557 sgd_solver.cpp:105] Iteration 133800, lr = 0.001
I0511 02:26:13.613378 39557 solver.cpp:218] Iteration 133900 (3.75945 iter/s, 26.5996s/100 iters), loss = 3.24034
I0511 02:26:13.613422 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.24034 (* 1 = 3.24034 loss)
I0511 02:26:13.613432 39557 sgd_solver.cpp:105] Iteration 133900, lr = 0.001
I0511 02:26:40.336701 39557 solver.cpp:218] Iteration 134000 (3.74202 iter/s, 26.7236s/100 iters), loss = 2.8013
I0511 02:26:40.336937 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.8013 (* 1 = 2.8013 loss)
I0511 02:26:40.336994 39557 sgd_solver.cpp:105] Iteration 134000, lr = 0.001
I0511 02:27:07.095429 39557 solver.cpp:218] Iteration 134100 (3.73709 iter/s, 26.7588s/100 iters), loss = 3.19211
I0511 02:27:07.095485 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.19211 (* 1 = 3.19211 loss)
I0511 02:27:07.095499 39557 sgd_solver.cpp:105] Iteration 134100, lr = 0.001
I0511 02:27:33.860129 39557 solver.cpp:218] Iteration 134200 (3.73624 iter/s, 26.7649s/100 iters), loss = 2.44748
I0511 02:27:33.860291 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.44748 (* 1 = 2.44748 loss)
I0511 02:27:33.860308 39557 sgd_solver.cpp:105] Iteration 134200, lr = 0.001
I0511 02:28:00.609145 39557 solver.cpp:218] Iteration 134300 (3.73845 iter/s, 26.7491s/100 iters), loss = 4.05133
I0511 02:28:00.609201 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.05133 (* 1 = 4.05133 loss)
I0511 02:28:00.609216 39557 sgd_solver.cpp:105] Iteration 134300, lr = 0.001
I0511 02:28:27.302265 39557 solver.cpp:218] Iteration 134400 (3.74626 iter/s, 26.6933s/100 iters), loss = 2.26916
I0511 02:28:27.302394 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.26916 (* 1 = 2.26916 loss)
I0511 02:28:27.302410 39557 sgd_solver.cpp:105] Iteration 134400, lr = 0.001
I0511 02:28:53.994519 39557 solver.cpp:218] Iteration 134500 (3.7464 iter/s, 26.6923s/100 iters), loss = 3.13062
I0511 02:28:53.994570 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.13062 (* 1 = 3.13062 loss)
I0511 02:28:53.994582 39557 sgd_solver.cpp:105] Iteration 134500, lr = 0.001
I0511 02:29:20.698032 39557 solver.cpp:218] Iteration 134600 (3.74481 iter/s, 26.7036s/100 iters), loss = 2.79648
I0511 02:29:20.698184 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.79648 (* 1 = 2.79648 loss)
I0511 02:29:20.698200 39557 sgd_solver.cpp:105] Iteration 134600, lr = 0.001
I0511 02:29:47.374433 39557 solver.cpp:218] Iteration 134700 (3.74863 iter/s, 26.6764s/100 iters), loss = 3.02822
I0511 02:29:47.374483 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.02822 (* 1 = 3.02822 loss)
I0511 02:29:47.374495 39557 sgd_solver.cpp:105] Iteration 134700, lr = 0.001
I0511 02:30:14.098918 39557 solver.cpp:218] Iteration 134800 (3.74188 iter/s, 26.7246s/100 iters), loss = 2.70602
I0511 02:30:14.099126 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.70602 (* 1 = 2.70602 loss)
I0511 02:30:14.099144 39557 sgd_solver.cpp:105] Iteration 134800, lr = 0.001
I0511 02:30:40.788779 39557 solver.cpp:218] Iteration 134900 (3.74675 iter/s, 26.6898s/100 iters), loss = 2.28306
I0511 02:30:40.788842 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.28306 (* 1 = 2.28306 loss)
I0511 02:30:40.788854 39557 sgd_solver.cpp:105] Iteration 134900, lr = 0.001
I0511 02:31:07.233307 39557 solver.cpp:447] Snapshotting to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_135000.caffemodel
I0511 02:31:07.273579 39557 sgd_solver.cpp:273] Snapshotting solver state to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_135000.solverstate
I0511 02:31:07.555382 39557 solver.cpp:218] Iteration 135000 (3.73599 iter/s, 26.7666s/100 iters), loss = 3.29227
I0511 02:31:07.555435 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.29227 (* 1 = 3.29227 loss)
I0511 02:31:07.555447 39557 sgd_solver.cpp:105] Iteration 135000, lr = 0.001
I0511 02:31:34.281823 39557 solver.cpp:218] Iteration 135100 (3.74161 iter/s, 26.7265s/100 iters), loss = 3.51034
I0511 02:31:34.281885 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.51034 (* 1 = 3.51034 loss)
I0511 02:31:34.281899 39557 sgd_solver.cpp:105] Iteration 135100, lr = 0.001
I0511 02:32:01.061893 39557 solver.cpp:218] Iteration 135200 (3.73412 iter/s, 26.7801s/100 iters), loss = 3.21385
I0511 02:32:01.062043 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.21385 (* 1 = 3.21385 loss)
I0511 02:32:01.062060 39557 sgd_solver.cpp:105] Iteration 135200, lr = 0.001
I0511 02:32:27.757074 39557 solver.cpp:218] Iteration 135300 (3.74601 iter/s, 26.6951s/100 iters), loss = 2.68907
I0511 02:32:27.757136 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.68907 (* 1 = 2.68907 loss)
I0511 02:32:27.757158 39557 sgd_solver.cpp:105] Iteration 135300, lr = 0.001
I0511 02:32:54.495651 39557 solver.cpp:218] Iteration 135400 (3.73992 iter/s, 26.7386s/100 iters), loss = 2.98885
I0511 02:32:54.495810 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.98885 (* 1 = 2.98885 loss)
I0511 02:32:54.495838 39557 sgd_solver.cpp:105] Iteration 135400, lr = 0.001
I0511 02:33:21.217526 39557 solver.cpp:218] Iteration 135500 (3.74227 iter/s, 26.7218s/100 iters), loss = 2.63075
I0511 02:33:21.217592 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.63075 (* 1 = 2.63075 loss)
I0511 02:33:21.217607 39557 sgd_solver.cpp:105] Iteration 135500, lr = 0.001
I0511 02:33:47.933197 39557 solver.cpp:218] Iteration 135600 (3.74313 iter/s, 26.7156s/100 iters), loss = 2.55852
I0511 02:33:47.933367 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.55852 (* 1 = 2.55852 loss)
I0511 02:33:47.933382 39557 sgd_solver.cpp:105] Iteration 135600, lr = 0.001
I0511 02:34:14.634739 39557 solver.cpp:218] Iteration 135700 (3.74512 iter/s, 26.7014s/100 iters), loss = 3.9424
I0511 02:34:14.634805 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.9424 (* 1 = 3.9424 loss)
I0511 02:34:14.634819 39557 sgd_solver.cpp:105] Iteration 135700, lr = 0.001
I0511 02:34:41.307634 39557 solver.cpp:218] Iteration 135800 (3.74913 iter/s, 26.6728s/100 iters), loss = 2.69813
I0511 02:34:41.307792 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.69813 (* 1 = 2.69813 loss)
I0511 02:34:41.307808 39557 sgd_solver.cpp:105] Iteration 135800, lr = 0.001
I0511 02:35:07.915761 39557 solver.cpp:218] Iteration 135900 (3.75827 iter/s, 26.608s/100 iters), loss = 2.48154
I0511 02:35:07.915819 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.48154 (* 1 = 2.48154 loss)
I0511 02:35:07.915834 39557 sgd_solver.cpp:105] Iteration 135900, lr = 0.001
I0511 02:35:34.543932 39557 solver.cpp:218] Iteration 136000 (3.75543 iter/s, 26.6281s/100 iters), loss = 3.99616
I0511 02:35:34.544129 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.99616 (* 1 = 3.99616 loss)
I0511 02:35:34.544153 39557 sgd_solver.cpp:105] Iteration 136000, lr = 0.001
I0511 02:36:01.137519 39557 solver.cpp:218] Iteration 136100 (3.76034 iter/s, 26.5934s/100 iters), loss = 2.66148
I0511 02:36:01.137575 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.66148 (* 1 = 2.66148 loss)
I0511 02:36:01.137590 39557 sgd_solver.cpp:105] Iteration 136100, lr = 0.001
I0511 02:36:27.835821 39557 solver.cpp:218] Iteration 136200 (3.74557 iter/s, 26.6982s/100 iters), loss = 2.42991
I0511 02:36:27.835990 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.42991 (* 1 = 2.42991 loss)
I0511 02:36:27.836006 39557 sgd_solver.cpp:105] Iteration 136200, lr = 0.001
I0511 02:36:54.591018 39557 solver.cpp:218] Iteration 136300 (3.73762 iter/s, 26.755s/100 iters), loss = 3.6989
I0511 02:36:54.591080 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.6989 (* 1 = 3.6989 loss)
I0511 02:36:54.591096 39557 sgd_solver.cpp:105] Iteration 136300, lr = 0.001
I0511 02:37:21.294874 39557 solver.cpp:218] Iteration 136400 (3.7448 iter/s, 26.7037s/100 iters), loss = 2.50373
I0511 02:37:21.295039 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.50373 (* 1 = 2.50373 loss)
I0511 02:37:21.295056 39557 sgd_solver.cpp:105] Iteration 136400, lr = 0.001
I0511 02:37:48.049378 39557 solver.cpp:218] Iteration 136500 (3.73772 iter/s, 26.7543s/100 iters), loss = 2.9137
I0511 02:37:48.049443 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.9137 (* 1 = 2.9137 loss)
I0511 02:37:48.049458 39557 sgd_solver.cpp:105] Iteration 136500, lr = 0.001
I0511 02:38:14.755342 39557 solver.cpp:218] Iteration 136600 (3.7445 iter/s, 26.7058s/100 iters), loss = 2.34576
I0511 02:38:14.755503 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.34576 (* 1 = 2.34576 loss)
I0511 02:38:14.755518 39557 sgd_solver.cpp:105] Iteration 136600, lr = 0.001
I0511 02:38:41.525327 39557 solver.cpp:218] Iteration 136700 (3.73556 iter/s, 26.7697s/100 iters), loss = 2.87807
I0511 02:38:41.525390 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.87807 (* 1 = 2.87807 loss)
I0511 02:38:41.525403 39557 sgd_solver.cpp:105] Iteration 136700, lr = 0.001
I0511 02:39:08.296349 39557 solver.cpp:218] Iteration 136800 (3.7354 iter/s, 26.7709s/100 iters), loss = 2.57751
I0511 02:39:08.296515 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.57751 (* 1 = 2.57751 loss)
I0511 02:39:08.296535 39557 sgd_solver.cpp:105] Iteration 136800, lr = 0.001
I0511 02:39:35.060284 39557 solver.cpp:218] Iteration 136900 (3.73641 iter/s, 26.7637s/100 iters), loss = 2.5109
I0511 02:39:35.060351 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.5109 (* 1 = 2.5109 loss)
I0511 02:39:35.060367 39557 sgd_solver.cpp:105] Iteration 136900, lr = 0.001
I0511 02:40:01.926434 39557 solver.cpp:218] Iteration 137000 (3.72218 iter/s, 26.866s/100 iters), loss = 3.05601
I0511 02:40:01.926599 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.05601 (* 1 = 3.05601 loss)
I0511 02:40:01.926626 39557 sgd_solver.cpp:105] Iteration 137000, lr = 0.001
I0511 02:40:28.635299 39557 solver.cpp:218] Iteration 137100 (3.74411 iter/s, 26.7086s/100 iters), loss = 2.96731
I0511 02:40:28.635370 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.96731 (* 1 = 2.96731 loss)
I0511 02:40:28.635385 39557 sgd_solver.cpp:105] Iteration 137100, lr = 0.001
I0511 02:40:55.499246 39557 solver.cpp:218] Iteration 137200 (3.72248 iter/s, 26.8638s/100 iters), loss = 2.57982
I0511 02:40:55.499418 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.57982 (* 1 = 2.57982 loss)
I0511 02:40:55.499446 39557 sgd_solver.cpp:105] Iteration 137200, lr = 0.001
I0511 02:41:22.311575 39557 solver.cpp:218] Iteration 137300 (3.72967 iter/s, 26.8121s/100 iters), loss = 3.11447
I0511 02:41:22.311637 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.11447 (* 1 = 3.11447 loss)
I0511 02:41:22.311652 39557 sgd_solver.cpp:105] Iteration 137300, lr = 0.001
I0511 02:41:49.029053 39557 solver.cpp:218] Iteration 137400 (3.74289 iter/s, 26.7173s/100 iters), loss = 2.36633
I0511 02:41:49.029249 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.36633 (* 1 = 2.36633 loss)
I0511 02:41:49.029275 39557 sgd_solver.cpp:105] Iteration 137400, lr = 0.001
I0511 02:42:15.827397 39557 solver.cpp:218] Iteration 137500 (3.73162 iter/s, 26.798s/100 iters), loss = 3.21723
I0511 02:42:15.827459 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.21723 (* 1 = 3.21723 loss)
I0511 02:42:15.827473 39557 sgd_solver.cpp:105] Iteration 137500, lr = 0.001
I0511 02:42:42.653488 39557 solver.cpp:218] Iteration 137600 (3.72774 iter/s, 26.8259s/100 iters), loss = 3.08803
I0511 02:42:42.653667 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.08803 (* 1 = 3.08803 loss)
I0511 02:42:42.653693 39557 sgd_solver.cpp:105] Iteration 137600, lr = 0.001
I0511 02:43:09.489318 39557 solver.cpp:218] Iteration 137700 (3.7264 iter/s, 26.8355s/100 iters), loss = 3.48293
I0511 02:43:09.489398 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.48293 (* 1 = 3.48293 loss)
I0511 02:43:09.489413 39557 sgd_solver.cpp:105] Iteration 137700, lr = 0.001
I0511 02:43:36.271939 39557 solver.cpp:218] Iteration 137800 (3.73379 iter/s, 26.7824s/100 iters), loss = 3.67129
I0511 02:43:36.272114 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.67129 (* 1 = 3.67129 loss)
I0511 02:43:36.272132 39557 sgd_solver.cpp:105] Iteration 137800, lr = 0.001
I0511 02:44:03.031494 39557 solver.cpp:218] Iteration 137900 (3.73702 iter/s, 26.7593s/100 iters), loss = 3.48466
I0511 02:44:03.031555 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.48466 (* 1 = 3.48466 loss)
I0511 02:44:03.031572 39557 sgd_solver.cpp:105] Iteration 137900, lr = 0.001
I0511 02:44:29.814813 39557 solver.cpp:218] Iteration 138000 (3.73369 iter/s, 26.7831s/100 iters), loss = 3.20057
I0511 02:44:29.814965 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.20057 (* 1 = 3.20057 loss)
I0511 02:44:29.814978 39557 sgd_solver.cpp:105] Iteration 138000, lr = 0.001
I0511 02:44:56.659868 39557 solver.cpp:218] Iteration 138100 (3.72512 iter/s, 26.8448s/100 iters), loss = 2.59612
I0511 02:44:56.659929 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.59612 (* 1 = 2.59612 loss)
I0511 02:44:56.659946 39557 sgd_solver.cpp:105] Iteration 138100, lr = 0.001
I0511 02:45:23.249047 39557 solver.cpp:218] Iteration 138200 (3.76096 iter/s, 26.589s/100 iters), loss = 3.3497
I0511 02:45:23.249189 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.3497 (* 1 = 3.3497 loss)
I0511 02:45:23.249204 39557 sgd_solver.cpp:105] Iteration 138200, lr = 0.001
I0511 02:45:49.906694 39557 solver.cpp:218] Iteration 138300 (3.75131 iter/s, 26.6574s/100 iters), loss = 2.5812
I0511 02:45:49.906751 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.5812 (* 1 = 2.5812 loss)
I0511 02:45:49.906764 39557 sgd_solver.cpp:105] Iteration 138300, lr = 0.001
I0511 02:46:16.483496 39557 solver.cpp:218] Iteration 138400 (3.76271 iter/s, 26.5766s/100 iters), loss = 3.27239
I0511 02:46:16.483652 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.27239 (* 1 = 3.27239 loss)
I0511 02:46:16.483665 39557 sgd_solver.cpp:105] Iteration 138400, lr = 0.001
I0511 02:46:43.097132 39557 solver.cpp:218] Iteration 138500 (3.75751 iter/s, 26.6133s/100 iters), loss = 2.60185
I0511 02:46:43.097188 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.60185 (* 1 = 2.60185 loss)
I0511 02:46:43.097201 39557 sgd_solver.cpp:105] Iteration 138500, lr = 0.001
I0511 02:47:09.735888 39557 solver.cpp:218] Iteration 138600 (3.75396 iter/s, 26.6386s/100 iters), loss = 2.64931
I0511 02:47:09.736027 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.64931 (* 1 = 2.64931 loss)
I0511 02:47:09.736040 39557 sgd_solver.cpp:105] Iteration 138600, lr = 0.001
I0511 02:47:36.585840 39557 solver.cpp:218] Iteration 138700 (3.72444 iter/s, 26.8497s/100 iters), loss = 3.45443
I0511 02:47:36.585912 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.45443 (* 1 = 3.45443 loss)
I0511 02:47:36.585927 39557 sgd_solver.cpp:105] Iteration 138700, lr = 0.001
I0511 02:48:03.337770 39557 solver.cpp:218] Iteration 138800 (3.73808 iter/s, 26.7517s/100 iters), loss = 2.17283
I0511 02:48:03.337958 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.17283 (* 1 = 2.17283 loss)
I0511 02:48:03.337973 39557 sgd_solver.cpp:105] Iteration 138800, lr = 0.001
I0511 02:48:29.940968 39557 solver.cpp:218] Iteration 138900 (3.759 iter/s, 26.6028s/100 iters), loss = 3.2264
I0511 02:48:29.941031 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.2264 (* 1 = 3.2264 loss)
I0511 02:48:29.941045 39557 sgd_solver.cpp:105] Iteration 138900, lr = 0.001
I0511 02:48:56.671610 39557 solver.cpp:218] Iteration 139000 (3.74106 iter/s, 26.7304s/100 iters), loss = 3.22389
I0511 02:48:56.671789 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.22389 (* 1 = 3.22389 loss)
I0511 02:48:56.671804 39557 sgd_solver.cpp:105] Iteration 139000, lr = 0.001
I0511 02:49:23.258529 39557 solver.cpp:218] Iteration 139100 (3.7613 iter/s, 26.5866s/100 iters), loss = 3.12711
I0511 02:49:23.258592 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.12711 (* 1 = 3.12711 loss)
I0511 02:49:23.258607 39557 sgd_solver.cpp:105] Iteration 139100, lr = 0.001
I0511 02:49:49.835577 39557 solver.cpp:218] Iteration 139200 (3.76268 iter/s, 26.5768s/100 iters), loss = 2.86624
I0511 02:49:49.835739 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.86624 (* 1 = 2.86624 loss)
I0511 02:49:49.835753 39557 sgd_solver.cpp:105] Iteration 139200, lr = 0.001
I0511 02:50:16.441884 39557 solver.cpp:218] Iteration 139300 (3.75855 iter/s, 26.606s/100 iters), loss = 2.45696
I0511 02:50:16.441941 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.45696 (* 1 = 2.45696 loss)
I0511 02:50:16.441953 39557 sgd_solver.cpp:105] Iteration 139300, lr = 0.001
I0511 02:50:43.061980 39557 solver.cpp:218] Iteration 139400 (3.75659 iter/s, 26.6199s/100 iters), loss = 3.11028
I0511 02:50:43.062136 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.11028 (* 1 = 3.11028 loss)
I0511 02:50:43.062152 39557 sgd_solver.cpp:105] Iteration 139400, lr = 0.001
I0511 02:51:09.665676 39557 solver.cpp:218] Iteration 139500 (3.75892 iter/s, 26.6034s/100 iters), loss = 3.52198
I0511 02:51:09.665747 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.52198 (* 1 = 3.52198 loss)
I0511 02:51:09.665758 39557 sgd_solver.cpp:105] Iteration 139500, lr = 0.001
I0511 02:51:36.275959 39557 solver.cpp:218] Iteration 139600 (3.75798 iter/s, 26.61s/100 iters), loss = 2.71512
I0511 02:51:36.276103 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.71512 (* 1 = 2.71512 loss)
I0511 02:51:36.276120 39557 sgd_solver.cpp:105] Iteration 139600, lr = 0.001
I0511 02:52:03.035967 39557 solver.cpp:218] Iteration 139700 (3.73696 iter/s, 26.7597s/100 iters), loss = 2.51485
I0511 02:52:03.036034 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.51485 (* 1 = 2.51485 loss)
I0511 02:52:03.036047 39557 sgd_solver.cpp:105] Iteration 139700, lr = 0.001
I0511 02:52:29.745483 39557 solver.cpp:218] Iteration 139800 (3.74402 iter/s, 26.7093s/100 iters), loss = 2.45863
I0511 02:52:29.745638 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.45863 (* 1 = 2.45863 loss)
I0511 02:52:29.745654 39557 sgd_solver.cpp:105] Iteration 139800, lr = 0.001
I0511 02:52:56.413235 39557 solver.cpp:218] Iteration 139900 (3.74989 iter/s, 26.6674s/100 iters), loss = 2.69119
I0511 02:52:56.413298 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.69119 (* 1 = 2.69119 loss)
I0511 02:52:56.413311 39557 sgd_solver.cpp:105] Iteration 139900, lr = 0.001
I0511 02:53:22.839257 39557 solver.cpp:447] Snapshotting to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_140000.caffemodel
I0511 02:53:22.880434 39557 sgd_solver.cpp:273] Snapshotting solver state to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_140000.solverstate
I0511 02:53:23.167120 39557 solver.cpp:218] Iteration 140000 (3.73781 iter/s, 26.7537s/100 iters), loss = 3.22419
I0511 02:53:23.167157 39624 sgd_solver.cpp:46] MultiStep Status: Iteration 140000, step = 3
I0511 02:53:23.167181 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.22419 (* 1 = 3.22419 loss)
I0511 02:53:23.167191 39557 sgd_solver.cpp:46] MultiStep Status: Iteration 140000, step = 3
I0511 02:53:23.167196 39557 sgd_solver.cpp:105] Iteration 140000, lr = 0.0001
I0511 02:53:49.914180 39557 solver.cpp:218] Iteration 140100 (3.73876 iter/s, 26.7468s/100 iters), loss = 2.7579
I0511 02:53:49.914248 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.7579 (* 1 = 2.7579 loss)
I0511 02:53:49.914261 39557 sgd_solver.cpp:105] Iteration 140100, lr = 0.0001
I0511 02:54:16.599716 39557 solver.cpp:218] Iteration 140200 (3.74738 iter/s, 26.6853s/100 iters), loss = 2.80978
I0511 02:54:16.599895 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.80978 (* 1 = 2.80978 loss)
I0511 02:54:16.599907 39557 sgd_solver.cpp:105] Iteration 140200, lr = 0.0001
I0511 02:54:43.375211 39557 solver.cpp:218] Iteration 140300 (3.73481 iter/s, 26.7751s/100 iters), loss = 3.75927
I0511 02:54:43.375273 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.75927 (* 1 = 3.75927 loss)
I0511 02:54:43.375288 39557 sgd_solver.cpp:105] Iteration 140300, lr = 0.0001
I0511 02:55:10.044095 39557 solver.cpp:218] Iteration 140400 (3.74972 iter/s, 26.6686s/100 iters), loss = 2.75114
I0511 02:55:10.044265 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.75114 (* 1 = 2.75114 loss)
I0511 02:55:10.044281 39557 sgd_solver.cpp:105] Iteration 140400, lr = 0.0001
I0511 02:55:36.714921 39557 solver.cpp:218] Iteration 140500 (3.74946 iter/s, 26.6705s/100 iters), loss = 2.83904
I0511 02:55:36.714988 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.83904 (* 1 = 2.83904 loss)
I0511 02:55:36.715001 39557 sgd_solver.cpp:105] Iteration 140500, lr = 0.0001
I0511 02:56:03.327186 39557 solver.cpp:218] Iteration 140600 (3.7577 iter/s, 26.612s/100 iters), loss = 2.89107
I0511 02:56:03.327373 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.89107 (* 1 = 2.89107 loss)
I0511 02:56:03.327394 39557 sgd_solver.cpp:105] Iteration 140600, lr = 0.0001
I0511 02:56:29.989423 39557 solver.cpp:218] Iteration 140700 (3.75067 iter/s, 26.6619s/100 iters), loss = 2.97728
I0511 02:56:29.989478 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.97728 (* 1 = 2.97728 loss)
I0511 02:56:29.989495 39557 sgd_solver.cpp:105] Iteration 140700, lr = 0.0001
I0511 02:56:56.646463 39557 solver.cpp:218] Iteration 140800 (3.75139 iter/s, 26.6568s/100 iters), loss = 3.42457
I0511 02:56:56.646620 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.42457 (* 1 = 3.42457 loss)
I0511 02:56:56.646632 39557 sgd_solver.cpp:105] Iteration 140800, lr = 0.0001
I0511 02:57:23.294101 39557 solver.cpp:218] Iteration 140900 (3.75272 iter/s, 26.6473s/100 iters), loss = 3.17773
I0511 02:57:23.294158 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.17773 (* 1 = 3.17773 loss)
I0511 02:57:23.294173 39557 sgd_solver.cpp:105] Iteration 140900, lr = 0.0001
I0511 02:57:49.934626 39557 solver.cpp:218] Iteration 141000 (3.75371 iter/s, 26.6403s/100 iters), loss = 2.98457
I0511 02:57:49.934794 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.98457 (* 1 = 2.98457 loss)
I0511 02:57:49.934808 39557 sgd_solver.cpp:105] Iteration 141000, lr = 0.0001
I0511 02:58:16.604889 39557 solver.cpp:218] Iteration 141100 (3.74954 iter/s, 26.6699s/100 iters), loss = 2.22721
I0511 02:58:16.604965 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.22721 (* 1 = 2.22721 loss)
I0511 02:58:16.604979 39557 sgd_solver.cpp:105] Iteration 141100, lr = 0.0001
I0511 02:58:43.402451 39557 solver.cpp:218] Iteration 141200 (3.73172 iter/s, 26.7973s/100 iters), loss = 2.41527
I0511 02:58:43.402652 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.41527 (* 1 = 2.41527 loss)
I0511 02:58:43.402665 39557 sgd_solver.cpp:105] Iteration 141200, lr = 0.0001
I0511 02:59:10.045845 39557 solver.cpp:218] Iteration 141300 (3.75333 iter/s, 26.643s/100 iters), loss = 2.53
I0511 02:59:10.045900 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.53 (* 1 = 2.53 loss)
I0511 02:59:10.045913 39557 sgd_solver.cpp:105] Iteration 141300, lr = 0.0001
I0511 02:59:36.732220 39557 solver.cpp:218] Iteration 141400 (3.74726 iter/s, 26.6861s/100 iters), loss = 2.5654
I0511 02:59:36.732388 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.5654 (* 1 = 2.5654 loss)
I0511 02:59:36.732401 39557 sgd_solver.cpp:105] Iteration 141400, lr = 0.0001
I0511 03:00:03.356892 39557 solver.cpp:218] Iteration 141500 (3.75592 iter/s, 26.6246s/100 iters), loss = 2.71239
I0511 03:00:03.356947 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.71239 (* 1 = 2.71239 loss)
I0511 03:00:03.356959 39557 sgd_solver.cpp:105] Iteration 141500, lr = 0.0001
I0511 03:00:29.968140 39557 solver.cpp:218] Iteration 141600 (3.75779 iter/s, 26.6114s/100 iters), loss = 3.77917
I0511 03:00:29.968294 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.77917 (* 1 = 3.77917 loss)
I0511 03:00:29.968307 39557 sgd_solver.cpp:105] Iteration 141600, lr = 0.0001
I0511 03:00:56.589191 39557 solver.cpp:218] Iteration 141700 (3.75642 iter/s, 26.6211s/100 iters), loss = 2.69976
I0511 03:00:56.589247 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.69976 (* 1 = 2.69976 loss)
I0511 03:00:56.589259 39557 sgd_solver.cpp:105] Iteration 141700, lr = 0.0001
I0511 03:01:23.282532 39557 solver.cpp:218] Iteration 141800 (3.74624 iter/s, 26.6934s/100 iters), loss = 3.54703
I0511 03:01:23.282686 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.54703 (* 1 = 3.54703 loss)
I0511 03:01:23.282699 39557 sgd_solver.cpp:105] Iteration 141800, lr = 0.0001
I0511 03:01:49.879266 39557 solver.cpp:218] Iteration 141900 (3.75986 iter/s, 26.5967s/100 iters), loss = 2.68202
I0511 03:01:49.879321 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.68202 (* 1 = 2.68202 loss)
I0511 03:01:49.879351 39557 sgd_solver.cpp:105] Iteration 141900, lr = 0.0001
I0511 03:02:16.567695 39557 solver.cpp:218] Iteration 142000 (3.74693 iter/s, 26.6885s/100 iters), loss = 2.49347
I0511 03:02:16.567862 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.49347 (* 1 = 2.49347 loss)
I0511 03:02:16.567874 39557 sgd_solver.cpp:105] Iteration 142000, lr = 0.0001
I0511 03:02:43.255043 39557 solver.cpp:218] Iteration 142100 (3.7471 iter/s, 26.6873s/100 iters), loss = 2.67474
I0511 03:02:43.255100 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.67474 (* 1 = 2.67474 loss)
I0511 03:02:43.255115 39557 sgd_solver.cpp:105] Iteration 142100, lr = 0.0001
I0511 03:03:10.011626 39557 solver.cpp:218] Iteration 142200 (3.73739 iter/s, 26.7566s/100 iters), loss = 3.46798
I0511 03:03:10.011785 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.46798 (* 1 = 3.46798 loss)
I0511 03:03:10.011801 39557 sgd_solver.cpp:105] Iteration 142200, lr = 0.0001
I0511 03:03:36.721544 39557 solver.cpp:218] Iteration 142300 (3.74394 iter/s, 26.7099s/100 iters), loss = 3.13473
I0511 03:03:36.721621 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.13473 (* 1 = 3.13473 loss)
I0511 03:03:36.721633 39557 sgd_solver.cpp:105] Iteration 142300, lr = 0.0001
I0511 03:04:03.402206 39557 solver.cpp:218] Iteration 142400 (3.74803 iter/s, 26.6807s/100 iters), loss = 2.9995
I0511 03:04:03.402376 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.9995 (* 1 = 2.9995 loss)
I0511 03:04:03.402391 39557 sgd_solver.cpp:105] Iteration 142400, lr = 0.0001
I0511 03:04:30.053216 39557 solver.cpp:218] Iteration 142500 (3.75222 iter/s, 26.6509s/100 iters), loss = 2.5166
I0511 03:04:30.053282 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.5166 (* 1 = 2.5166 loss)
I0511 03:04:30.053295 39557 sgd_solver.cpp:105] Iteration 142500, lr = 0.0001
I0511 03:04:56.729791 39557 solver.cpp:218] Iteration 142600 (3.74861 iter/s, 26.6766s/100 iters), loss = 2.96776
I0511 03:04:56.730002 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.96776 (* 1 = 2.96776 loss)
I0511 03:04:56.730020 39557 sgd_solver.cpp:105] Iteration 142600, lr = 0.0001
I0511 03:05:23.388934 39557 solver.cpp:218] Iteration 142700 (3.75108 iter/s, 26.659s/100 iters), loss = 2.15251
I0511 03:05:23.388998 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.15251 (* 1 = 2.15251 loss)
I0511 03:05:23.389014 39557 sgd_solver.cpp:105] Iteration 142700, lr = 0.0001
I0511 03:05:50.007942 39557 solver.cpp:218] Iteration 142800 (3.75672 iter/s, 26.619s/100 iters), loss = 2.43049
I0511 03:05:50.008106 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.43049 (* 1 = 2.43049 loss)
I0511 03:05:50.008119 39557 sgd_solver.cpp:105] Iteration 142800, lr = 0.0001
I0511 03:06:16.633280 39557 solver.cpp:218] Iteration 142900 (3.75584 iter/s, 26.6252s/100 iters), loss = 1.99392
I0511 03:06:16.633340 39557 solver.cpp:237]     Train net output #0: softmax_loss = 1.99392 (* 1 = 1.99392 loss)
I0511 03:06:16.633354 39557 sgd_solver.cpp:105] Iteration 142900, lr = 0.0001
I0511 03:06:43.344949 39557 solver.cpp:218] Iteration 143000 (3.74369 iter/s, 26.7116s/100 iters), loss = 2.54144
I0511 03:06:43.345108 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.54144 (* 1 = 2.54144 loss)
I0511 03:06:43.345124 39557 sgd_solver.cpp:105] Iteration 143000, lr = 0.0001
I0511 03:07:10.024339 39557 solver.cpp:218] Iteration 143100 (3.74823 iter/s, 26.6792s/100 iters), loss = 2.90375
I0511 03:07:10.024399 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.90375 (* 1 = 2.90375 loss)
I0511 03:07:10.024410 39557 sgd_solver.cpp:105] Iteration 143100, lr = 0.0001
I0511 03:07:36.697207 39557 solver.cpp:218] Iteration 143200 (3.74914 iter/s, 26.6728s/100 iters), loss = 2.77938
I0511 03:07:36.697362 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.77938 (* 1 = 2.77938 loss)
I0511 03:07:36.697377 39557 sgd_solver.cpp:105] Iteration 143200, lr = 0.0001
I0511 03:08:03.360353 39557 solver.cpp:218] Iteration 143300 (3.75052 iter/s, 26.663s/100 iters), loss = 3.4732
I0511 03:08:03.360414 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.4732 (* 1 = 3.4732 loss)
I0511 03:08:03.360425 39557 sgd_solver.cpp:105] Iteration 143300, lr = 0.0001
I0511 03:08:30.065407 39557 solver.cpp:218] Iteration 143400 (3.74462 iter/s, 26.705s/100 iters), loss = 3.03244
I0511 03:08:30.065553 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.03244 (* 1 = 3.03244 loss)
I0511 03:08:30.065569 39557 sgd_solver.cpp:105] Iteration 143400, lr = 0.0001
I0511 03:08:56.735204 39557 solver.cpp:218] Iteration 143500 (3.74958 iter/s, 26.6696s/100 iters), loss = 4.38624
I0511 03:08:56.735262 39557 solver.cpp:237]     Train net output #0: softmax_loss = 4.38624 (* 1 = 4.38624 loss)
I0511 03:08:56.735273 39557 sgd_solver.cpp:105] Iteration 143500, lr = 0.0001
I0511 03:09:23.341552 39557 solver.cpp:218] Iteration 143600 (3.75851 iter/s, 26.6063s/100 iters), loss = 2.99696
I0511 03:09:23.341717 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.99696 (* 1 = 2.99696 loss)
I0511 03:09:23.341734 39557 sgd_solver.cpp:105] Iteration 143600, lr = 0.0001
I0511 03:09:49.967955 39557 solver.cpp:218] Iteration 143700 (3.7557 iter/s, 26.6262s/100 iters), loss = 2.71757
I0511 03:09:49.968014 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.71757 (* 1 = 2.71757 loss)
I0511 03:09:49.968025 39557 sgd_solver.cpp:105] Iteration 143700, lr = 0.0001
I0511 03:10:16.654852 39557 solver.cpp:218] Iteration 143800 (3.74717 iter/s, 26.6868s/100 iters), loss = 2.85733
I0511 03:10:16.654995 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.85733 (* 1 = 2.85733 loss)
I0511 03:10:16.655010 39557 sgd_solver.cpp:105] Iteration 143800, lr = 0.0001
I0511 03:10:43.369110 39557 solver.cpp:218] Iteration 143900 (3.74334 iter/s, 26.7141s/100 iters), loss = 2.76917
I0511 03:10:43.369171 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.76917 (* 1 = 2.76917 loss)
I0511 03:10:43.369182 39557 sgd_solver.cpp:105] Iteration 143900, lr = 0.0001
I0511 03:11:10.024188 39557 solver.cpp:218] Iteration 144000 (3.75164 iter/s, 26.655s/100 iters), loss = 2.72968
I0511 03:11:10.024377 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.72968 (* 1 = 2.72968 loss)
I0511 03:11:10.024392 39557 sgd_solver.cpp:105] Iteration 144000, lr = 0.0001
I0511 03:11:36.696344 39557 solver.cpp:218] Iteration 144100 (3.74926 iter/s, 26.6719s/100 iters), loss = 2.77788
I0511 03:11:36.696413 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.77788 (* 1 = 2.77788 loss)
I0511 03:11:36.696422 39557 sgd_solver.cpp:105] Iteration 144100, lr = 0.0001
I0511 03:12:03.455485 39557 solver.cpp:218] Iteration 144200 (3.73706 iter/s, 26.759s/100 iters), loss = 3.26858
I0511 03:12:03.455646 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.26858 (* 1 = 3.26858 loss)
I0511 03:12:03.455662 39557 sgd_solver.cpp:105] Iteration 144200, lr = 0.0001
I0511 03:12:30.195202 39557 solver.cpp:218] Iteration 144300 (3.73978 iter/s, 26.7395s/100 iters), loss = 3.21904
I0511 03:12:30.195269 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.21904 (* 1 = 3.21904 loss)
I0511 03:12:30.195283 39557 sgd_solver.cpp:105] Iteration 144300, lr = 0.0001
I0511 03:12:56.981173 39557 solver.cpp:218] Iteration 144400 (3.73332 iter/s, 26.7858s/100 iters), loss = 2.83178
I0511 03:12:56.981346 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.83178 (* 1 = 2.83178 loss)
I0511 03:12:56.981362 39557 sgd_solver.cpp:105] Iteration 144400, lr = 0.0001
I0511 03:13:23.654016 39557 solver.cpp:218] Iteration 144500 (3.74917 iter/s, 26.6726s/100 iters), loss = 2.34402
I0511 03:13:23.654080 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.34402 (* 1 = 2.34402 loss)
I0511 03:13:23.654093 39557 sgd_solver.cpp:105] Iteration 144500, lr = 0.0001
I0511 03:13:50.351189 39557 solver.cpp:218] Iteration 144600 (3.74573 iter/s, 26.697s/100 iters), loss = 2.9917
I0511 03:13:50.351361 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.9917 (* 1 = 2.9917 loss)
I0511 03:13:50.351375 39557 sgd_solver.cpp:105] Iteration 144600, lr = 0.0001
I0511 03:14:17.003497 39557 solver.cpp:218] Iteration 144700 (3.75205 iter/s, 26.6521s/100 iters), loss = 3.69086
I0511 03:14:17.003561 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.69086 (* 1 = 3.69086 loss)
I0511 03:14:17.003574 39557 sgd_solver.cpp:105] Iteration 144700, lr = 0.0001
I0511 03:14:43.713379 39557 solver.cpp:218] Iteration 144800 (3.74395 iter/s, 26.7097s/100 iters), loss = 3.12231
I0511 03:14:43.713541 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.12231 (* 1 = 3.12231 loss)
I0511 03:14:43.713557 39557 sgd_solver.cpp:105] Iteration 144800, lr = 0.0001
I0511 03:15:10.368865 39557 solver.cpp:218] Iteration 144900 (3.75161 iter/s, 26.6552s/100 iters), loss = 2.57962
I0511 03:15:10.368927 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.57962 (* 1 = 2.57962 loss)
I0511 03:15:10.368942 39557 sgd_solver.cpp:105] Iteration 144900, lr = 0.0001
I0511 03:15:36.791576 39557 solver.cpp:447] Snapshotting to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_145000.caffemodel
I0511 03:15:36.832098 39557 sgd_solver.cpp:273] Snapshotting solver state to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_145000.solverstate
I0511 03:15:37.120544 39557 solver.cpp:218] Iteration 145000 (3.7381 iter/s, 26.7515s/100 iters), loss = 3.17245
I0511 03:15:37.120599 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.17245 (* 1 = 3.17245 loss)
I0511 03:15:37.120609 39557 sgd_solver.cpp:105] Iteration 145000, lr = 0.0001
I0511 03:16:03.818267 39557 solver.cpp:218] Iteration 145100 (3.74566 iter/s, 26.6976s/100 iters), loss = 2.83683
I0511 03:16:03.818326 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.83683 (* 1 = 2.83683 loss)
I0511 03:16:03.818339 39557 sgd_solver.cpp:105] Iteration 145100, lr = 0.0001
I0511 03:16:30.460458 39557 solver.cpp:218] Iteration 145200 (3.75346 iter/s, 26.6421s/100 iters), loss = 2.56296
I0511 03:16:30.460660 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.56296 (* 1 = 2.56296 loss)
I0511 03:16:30.460674 39557 sgd_solver.cpp:105] Iteration 145200, lr = 0.0001
I0511 03:16:57.130628 39557 solver.cpp:218] Iteration 145300 (3.74955 iter/s, 26.6699s/100 iters), loss = 2.9497
I0511 03:16:57.130683 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.9497 (* 1 = 2.9497 loss)
I0511 03:16:57.130697 39557 sgd_solver.cpp:105] Iteration 145300, lr = 0.0001
I0511 03:17:23.738526 39557 solver.cpp:218] Iteration 145400 (3.7583 iter/s, 26.6077s/100 iters), loss = 3.25605
I0511 03:17:23.738698 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.25605 (* 1 = 3.25605 loss)
I0511 03:17:23.738713 39557 sgd_solver.cpp:105] Iteration 145400, lr = 0.0001
I0511 03:17:50.381371 39557 solver.cpp:218] Iteration 145500 (3.75339 iter/s, 26.6426s/100 iters), loss = 2.33971
I0511 03:17:50.381429 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.33971 (* 1 = 2.33971 loss)
I0511 03:17:50.381446 39557 sgd_solver.cpp:105] Iteration 145500, lr = 0.0001
I0511 03:18:17.061609 39557 solver.cpp:218] Iteration 145600 (3.74811 iter/s, 26.6801s/100 iters), loss = 2.95696
I0511 03:18:17.061770 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.95696 (* 1 = 2.95696 loss)
I0511 03:18:17.061784 39557 sgd_solver.cpp:105] Iteration 145600, lr = 0.0001
I0511 03:18:43.714766 39557 solver.cpp:218] Iteration 145700 (3.75194 iter/s, 26.6529s/100 iters), loss = 2.75417
I0511 03:18:43.714824 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.75417 (* 1 = 2.75417 loss)
I0511 03:18:43.714838 39557 sgd_solver.cpp:105] Iteration 145700, lr = 0.0001
I0511 03:19:10.413556 39557 solver.cpp:218] Iteration 145800 (3.74551 iter/s, 26.6986s/100 iters), loss = 2.75982
I0511 03:19:10.413702 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.75982 (* 1 = 2.75982 loss)
I0511 03:19:10.413714 39557 sgd_solver.cpp:105] Iteration 145800, lr = 0.0001
I0511 03:19:37.019845 39557 solver.cpp:218] Iteration 145900 (3.75854 iter/s, 26.606s/100 iters), loss = 3.28289
I0511 03:19:37.019907 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.28289 (* 1 = 3.28289 loss)
I0511 03:19:37.019919 39557 sgd_solver.cpp:105] Iteration 145900, lr = 0.0001
I0511 03:20:03.889675 39557 solver.cpp:218] Iteration 146000 (3.72167 iter/s, 26.8697s/100 iters), loss = 2.31927
I0511 03:20:03.889878 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.31927 (* 1 = 2.31927 loss)
I0511 03:20:03.889897 39557 sgd_solver.cpp:105] Iteration 146000, lr = 0.0001
I0511 03:20:30.525339 39557 solver.cpp:218] Iteration 146100 (3.75441 iter/s, 26.6354s/100 iters), loss = 2.85913
I0511 03:20:30.525403 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.85913 (* 1 = 2.85913 loss)
I0511 03:20:30.525415 39557 sgd_solver.cpp:105] Iteration 146100, lr = 0.0001
I0511 03:20:57.242645 39557 solver.cpp:218] Iteration 146200 (3.74292 iter/s, 26.7171s/100 iters), loss = 2.91194
I0511 03:20:57.242803 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.91194 (* 1 = 2.91194 loss)
I0511 03:20:57.242818 39557 sgd_solver.cpp:105] Iteration 146200, lr = 0.0001
I0511 03:21:24.003728 39557 solver.cpp:218] Iteration 146300 (3.73681 iter/s, 26.7608s/100 iters), loss = 2.78522
I0511 03:21:24.003790 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.78522 (* 1 = 2.78522 loss)
I0511 03:21:24.003805 39557 sgd_solver.cpp:105] Iteration 146300, lr = 0.0001
I0511 03:21:50.729351 39557 solver.cpp:218] Iteration 146400 (3.74175 iter/s, 26.7255s/100 iters), loss = 3.05365
I0511 03:21:50.729516 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.05365 (* 1 = 3.05365 loss)
I0511 03:21:50.729529 39557 sgd_solver.cpp:105] Iteration 146400, lr = 0.0001
I0511 03:22:17.474098 39557 solver.cpp:218] Iteration 146500 (3.73909 iter/s, 26.7445s/100 iters), loss = 2.59635
I0511 03:22:17.474162 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.59635 (* 1 = 2.59635 loss)
I0511 03:22:17.474177 39557 sgd_solver.cpp:105] Iteration 146500, lr = 0.0001
I0511 03:22:44.249681 39557 solver.cpp:218] Iteration 146600 (3.73477 iter/s, 26.7754s/100 iters), loss = 3.34438
I0511 03:22:44.249874 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.34438 (* 1 = 3.34438 loss)
I0511 03:22:44.249891 39557 sgd_solver.cpp:105] Iteration 146600, lr = 0.0001
I0511 03:23:10.954355 39557 solver.cpp:218] Iteration 146700 (3.74471 iter/s, 26.7044s/100 iters), loss = 3.27336
I0511 03:23:10.954412 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.27336 (* 1 = 3.27336 loss)
I0511 03:23:10.954423 39557 sgd_solver.cpp:105] Iteration 146700, lr = 0.0001
I0511 03:23:37.801105 39557 solver.cpp:218] Iteration 146800 (3.72487 iter/s, 26.8466s/100 iters), loss = 2.68134
I0511 03:23:37.801288 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.68134 (* 1 = 2.68134 loss)
I0511 03:23:37.801306 39557 sgd_solver.cpp:105] Iteration 146800, lr = 0.0001
I0511 03:24:04.492600 39557 solver.cpp:218] Iteration 146900 (3.74655 iter/s, 26.6912s/100 iters), loss = 3.2644
I0511 03:24:04.492661 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.2644 (* 1 = 3.2644 loss)
I0511 03:24:04.492676 39557 sgd_solver.cpp:105] Iteration 146900, lr = 0.0001
I0511 03:24:31.133544 39557 solver.cpp:218] Iteration 147000 (3.75364 iter/s, 26.6408s/100 iters), loss = 2.68883
I0511 03:24:31.133704 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.68883 (* 1 = 2.68883 loss)
I0511 03:24:31.133724 39557 sgd_solver.cpp:105] Iteration 147000, lr = 0.0001
I0511 03:24:57.731648 39557 solver.cpp:218] Iteration 147100 (3.75971 iter/s, 26.5978s/100 iters), loss = 2.85418
I0511 03:24:57.731709 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.85418 (* 1 = 2.85418 loss)
I0511 03:24:57.731719 39557 sgd_solver.cpp:105] Iteration 147100, lr = 0.0001
I0511 03:25:24.365710 39557 solver.cpp:218] Iteration 147200 (3.75462 iter/s, 26.6339s/100 iters), loss = 3.28274
I0511 03:25:24.365861 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.28274 (* 1 = 3.28274 loss)
I0511 03:25:24.365878 39557 sgd_solver.cpp:105] Iteration 147200, lr = 0.0001
I0511 03:25:50.943742 39557 solver.cpp:218] Iteration 147300 (3.76254 iter/s, 26.5778s/100 iters), loss = 3.24318
I0511 03:25:50.943796 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.24318 (* 1 = 3.24318 loss)
I0511 03:25:50.943809 39557 sgd_solver.cpp:105] Iteration 147300, lr = 0.0001
I0511 03:26:17.549957 39557 solver.cpp:218] Iteration 147400 (3.75854 iter/s, 26.6061s/100 iters), loss = 2.80064
I0511 03:26:17.550106 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.80064 (* 1 = 2.80064 loss)
I0511 03:26:17.550119 39557 sgd_solver.cpp:105] Iteration 147400, lr = 0.0001
I0511 03:26:44.158278 39557 solver.cpp:218] Iteration 147500 (3.75826 iter/s, 26.6081s/100 iters), loss = 3.26996
I0511 03:26:44.158335 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.26996 (* 1 = 3.26996 loss)
I0511 03:26:44.158350 39557 sgd_solver.cpp:105] Iteration 147500, lr = 0.0001
I0511 03:27:10.789371 39557 solver.cpp:218] Iteration 147600 (3.75503 iter/s, 26.6309s/100 iters), loss = 3.04365
I0511 03:27:10.789520 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.04365 (* 1 = 3.04365 loss)
I0511 03:27:10.789533 39557 sgd_solver.cpp:105] Iteration 147600, lr = 0.0001
I0511 03:27:37.437199 39557 solver.cpp:218] Iteration 147700 (3.75269 iter/s, 26.6476s/100 iters), loss = 2.45501
I0511 03:27:37.437258 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.45501 (* 1 = 2.45501 loss)
I0511 03:27:37.437271 39557 sgd_solver.cpp:105] Iteration 147700, lr = 0.0001
I0511 03:28:04.198083 39557 solver.cpp:218] Iteration 147800 (3.73682 iter/s, 26.7607s/100 iters), loss = 3.46127
I0511 03:28:04.198256 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.46127 (* 1 = 3.46127 loss)
I0511 03:28:04.198276 39557 sgd_solver.cpp:105] Iteration 147800, lr = 0.0001
I0511 03:28:30.823164 39557 solver.cpp:218] Iteration 147900 (3.7559 iter/s, 26.6248s/100 iters), loss = 2.70453
I0511 03:28:30.823228 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.70453 (* 1 = 2.70453 loss)
I0511 03:28:30.823240 39557 sgd_solver.cpp:105] Iteration 147900, lr = 0.0001
I0511 03:28:57.588579 39557 solver.cpp:218] Iteration 148000 (3.73619 iter/s, 26.7652s/100 iters), loss = 2.58874
I0511 03:28:57.588754 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.58874 (* 1 = 2.58874 loss)
I0511 03:28:57.588773 39557 sgd_solver.cpp:105] Iteration 148000, lr = 0.0001
I0511 03:29:24.256188 39557 solver.cpp:218] Iteration 148100 (3.74991 iter/s, 26.6673s/100 iters), loss = 2.3938
I0511 03:29:24.256253 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.3938 (* 1 = 2.3938 loss)
I0511 03:29:24.256264 39557 sgd_solver.cpp:105] Iteration 148100, lr = 0.0001
I0511 03:29:50.880848 39557 solver.cpp:218] Iteration 148200 (3.75594 iter/s, 26.6245s/100 iters), loss = 2.87841
I0511 03:29:50.881000 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.87841 (* 1 = 2.87841 loss)
I0511 03:29:50.881012 39557 sgd_solver.cpp:105] Iteration 148200, lr = 0.0001
I0511 03:30:17.548828 39557 solver.cpp:218] Iteration 148300 (3.74985 iter/s, 26.6677s/100 iters), loss = 2.98893
I0511 03:30:17.548887 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.98893 (* 1 = 2.98893 loss)
I0511 03:30:17.548899 39557 sgd_solver.cpp:105] Iteration 148300, lr = 0.0001
I0511 03:30:44.276695 39557 solver.cpp:218] Iteration 148400 (3.74144 iter/s, 26.7277s/100 iters), loss = 2.74149
I0511 03:30:44.276841 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.74149 (* 1 = 2.74149 loss)
I0511 03:30:44.276854 39557 sgd_solver.cpp:105] Iteration 148400, lr = 0.0001
I0511 03:31:11.031442 39557 solver.cpp:218] Iteration 148500 (3.73769 iter/s, 26.7545s/100 iters), loss = 2.26478
I0511 03:31:11.031504 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.26478 (* 1 = 2.26478 loss)
I0511 03:31:11.031517 39557 sgd_solver.cpp:105] Iteration 148500, lr = 0.0001
I0511 03:31:37.891882 39557 solver.cpp:218] Iteration 148600 (3.72297 iter/s, 26.8602s/100 iters), loss = 2.71979
I0511 03:31:37.892045 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.71979 (* 1 = 2.71979 loss)
I0511 03:31:37.892057 39557 sgd_solver.cpp:105] Iteration 148600, lr = 0.0001
I0511 03:32:04.606614 39557 solver.cpp:218] Iteration 148700 (3.74329 iter/s, 26.7144s/100 iters), loss = 2.42924
I0511 03:32:04.606678 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.42924 (* 1 = 2.42924 loss)
I0511 03:32:04.606693 39557 sgd_solver.cpp:105] Iteration 148700, lr = 0.0001
I0511 03:32:31.394284 39557 solver.cpp:218] Iteration 148800 (3.73309 iter/s, 26.7875s/100 iters), loss = 2.58677
I0511 03:32:31.394444 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.58677 (* 1 = 2.58677 loss)
I0511 03:32:31.394459 39557 sgd_solver.cpp:105] Iteration 148800, lr = 0.0001
I0511 03:32:58.165613 39557 solver.cpp:218] Iteration 148900 (3.73538 iter/s, 26.771s/100 iters), loss = 3.72719
I0511 03:32:58.165675 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.72719 (* 1 = 3.72719 loss)
I0511 03:32:58.165688 39557 sgd_solver.cpp:105] Iteration 148900, lr = 0.0001
I0511 03:33:24.883872 39557 solver.cpp:218] Iteration 149000 (3.74279 iter/s, 26.7181s/100 iters), loss = 2.60215
I0511 03:33:24.884042 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.60215 (* 1 = 2.60215 loss)
I0511 03:33:24.884057 39557 sgd_solver.cpp:105] Iteration 149000, lr = 0.0001
I0511 03:33:51.597175 39557 solver.cpp:218] Iteration 149100 (3.74349 iter/s, 26.713s/100 iters), loss = 2.14184
I0511 03:33:51.597225 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.14184 (* 1 = 2.14184 loss)
I0511 03:33:51.597239 39557 sgd_solver.cpp:105] Iteration 149100, lr = 0.0001
I0511 03:34:18.268143 39557 solver.cpp:218] Iteration 149200 (3.74942 iter/s, 26.6708s/100 iters), loss = 3.30768
I0511 03:34:18.268326 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.30768 (* 1 = 3.30768 loss)
I0511 03:34:18.268340 39557 sgd_solver.cpp:105] Iteration 149200, lr = 0.0001
I0511 03:34:44.944841 39557 solver.cpp:218] Iteration 149300 (3.74863 iter/s, 26.6764s/100 iters), loss = 2.99673
I0511 03:34:44.944901 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.99673 (* 1 = 2.99673 loss)
I0511 03:34:44.944912 39557 sgd_solver.cpp:105] Iteration 149300, lr = 0.0001
I0511 03:35:11.660171 39557 solver.cpp:218] Iteration 149400 (3.74319 iter/s, 26.7152s/100 iters), loss = 3.65116
I0511 03:35:11.660336 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.65116 (* 1 = 3.65116 loss)
I0511 03:35:11.660362 39557 sgd_solver.cpp:105] Iteration 149400, lr = 0.0001
I0511 03:35:38.418211 39557 solver.cpp:218] Iteration 149500 (3.73723 iter/s, 26.7578s/100 iters), loss = 3.6539
I0511 03:35:38.418278 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.6539 (* 1 = 3.6539 loss)
I0511 03:35:38.418293 39557 sgd_solver.cpp:105] Iteration 149500, lr = 0.0001
I0511 03:36:05.153228 39557 solver.cpp:218] Iteration 149600 (3.74044 iter/s, 26.7348s/100 iters), loss = 2.86538
I0511 03:36:05.153388 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.86538 (* 1 = 2.86538 loss)
I0511 03:36:05.153415 39557 sgd_solver.cpp:105] Iteration 149600, lr = 0.0001
I0511 03:36:31.853569 39557 solver.cpp:218] Iteration 149700 (3.74531 iter/s, 26.7001s/100 iters), loss = 3.19728
I0511 03:36:31.853659 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.19728 (* 1 = 3.19728 loss)
I0511 03:36:31.853674 39557 sgd_solver.cpp:105] Iteration 149700, lr = 0.0001
I0511 03:36:58.641624 39557 solver.cpp:218] Iteration 149800 (3.73303 iter/s, 26.7879s/100 iters), loss = 3.45162
I0511 03:36:58.641777 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.45162 (* 1 = 3.45162 loss)
I0511 03:36:58.641801 39557 sgd_solver.cpp:105] Iteration 149800, lr = 0.0001
I0511 03:37:25.354221 39557 solver.cpp:218] Iteration 149900 (3.74359 iter/s, 26.7123s/100 iters), loss = 3.63464
I0511 03:37:25.354280 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.63464 (* 1 = 3.63464 loss)
I0511 03:37:25.354296 39557 sgd_solver.cpp:105] Iteration 149900, lr = 0.0001
I0511 03:37:51.856777 39557 solver.cpp:447] Snapshotting to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_150000.caffemodel
I0511 03:37:51.897159 39557 sgd_solver.cpp:273] Snapshotting solver state to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_150000.solverstate
I0511 03:37:52.233479 39557 solver.cpp:218] Iteration 150000 (3.72036 iter/s, 26.8791s/100 iters), loss = 2.61561
I0511 03:37:52.233534 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.61561 (* 1 = 2.61561 loss)
I0511 03:37:52.233547 39557 sgd_solver.cpp:105] Iteration 150000, lr = 0.0001
I0511 03:38:18.899319 39557 solver.cpp:218] Iteration 150100 (3.75014 iter/s, 26.6657s/100 iters), loss = 2.64158
I0511 03:38:18.899382 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.64158 (* 1 = 2.64158 loss)
I0511 03:38:18.899397 39557 sgd_solver.cpp:105] Iteration 150100, lr = 0.0001
I0511 03:38:45.705662 39557 solver.cpp:218] Iteration 150200 (3.73048 iter/s, 26.8062s/100 iters), loss = 3.52319
I0511 03:38:45.705811 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.52319 (* 1 = 3.52319 loss)
I0511 03:38:45.705826 39557 sgd_solver.cpp:105] Iteration 150200, lr = 0.0001
I0511 03:39:12.528182 39557 solver.cpp:218] Iteration 150300 (3.72825 iter/s, 26.8223s/100 iters), loss = 3.09071
I0511 03:39:12.528247 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.09071 (* 1 = 3.09071 loss)
I0511 03:39:12.528264 39557 sgd_solver.cpp:105] Iteration 150300, lr = 0.0001
I0511 03:39:39.265525 39557 solver.cpp:218] Iteration 150400 (3.74011 iter/s, 26.7372s/100 iters), loss = 2.36144
I0511 03:39:39.265728 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.36144 (* 1 = 2.36144 loss)
I0511 03:39:39.265744 39557 sgd_solver.cpp:105] Iteration 150400, lr = 0.0001
I0511 03:40:05.997210 39557 solver.cpp:218] Iteration 150500 (3.74092 iter/s, 26.7314s/100 iters), loss = 2.79916
I0511 03:40:05.997274 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.79916 (* 1 = 2.79916 loss)
I0511 03:40:05.997287 39557 sgd_solver.cpp:105] Iteration 150500, lr = 0.0001
I0511 03:40:32.680343 39557 solver.cpp:218] Iteration 150600 (3.74771 iter/s, 26.683s/100 iters), loss = 3.03592
I0511 03:40:32.680505 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.03592 (* 1 = 3.03592 loss)
I0511 03:40:32.680521 39557 sgd_solver.cpp:105] Iteration 150600, lr = 0.0001
I0511 03:40:59.437885 39557 solver.cpp:218] Iteration 150700 (3.7373 iter/s, 26.7573s/100 iters), loss = 3.23309
I0511 03:40:59.437950 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.23309 (* 1 = 3.23309 loss)
I0511 03:40:59.437964 39557 sgd_solver.cpp:105] Iteration 150700, lr = 0.0001
I0511 03:41:26.258630 39557 solver.cpp:218] Iteration 150800 (3.72848 iter/s, 26.8206s/100 iters), loss = 2.71016
I0511 03:41:26.258797 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.71016 (* 1 = 2.71016 loss)
I0511 03:41:26.258815 39557 sgd_solver.cpp:105] Iteration 150800, lr = 0.0001
I0511 03:41:53.088021 39557 solver.cpp:218] Iteration 150900 (3.72729 iter/s, 26.8291s/100 iters), loss = 3.29107
I0511 03:41:53.088085 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.29107 (* 1 = 3.29107 loss)
I0511 03:41:53.088100 39557 sgd_solver.cpp:105] Iteration 150900, lr = 0.0001
I0511 03:42:19.753378 39557 solver.cpp:218] Iteration 151000 (3.75021 iter/s, 26.6652s/100 iters), loss = 3.14108
I0511 03:42:19.753540 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.14108 (* 1 = 3.14108 loss)
I0511 03:42:19.753554 39557 sgd_solver.cpp:105] Iteration 151000, lr = 0.0001
I0511 03:42:46.625494 39557 solver.cpp:218] Iteration 151100 (3.72137 iter/s, 26.8718s/100 iters), loss = 2.81094
I0511 03:42:46.625558 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.81094 (* 1 = 2.81094 loss)
I0511 03:42:46.625572 39557 sgd_solver.cpp:105] Iteration 151100, lr = 0.0001
I0511 03:43:13.325419 39557 solver.cpp:218] Iteration 151200 (3.74536 iter/s, 26.6997s/100 iters), loss = 2.59921
I0511 03:43:13.325593 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.59921 (* 1 = 2.59921 loss)
I0511 03:43:13.325610 39557 sgd_solver.cpp:105] Iteration 151200, lr = 0.0001
I0511 03:43:39.944177 39557 solver.cpp:218] Iteration 151300 (3.75679 iter/s, 26.6185s/100 iters), loss = 3.2418
I0511 03:43:39.944241 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.2418 (* 1 = 3.2418 loss)
I0511 03:43:39.944252 39557 sgd_solver.cpp:105] Iteration 151300, lr = 0.0001
I0511 03:44:06.606710 39557 solver.cpp:218] Iteration 151400 (3.75061 iter/s, 26.6624s/100 iters), loss = 2.85954
I0511 03:44:06.606873 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.85954 (* 1 = 2.85954 loss)
I0511 03:44:06.606887 39557 sgd_solver.cpp:105] Iteration 151400, lr = 0.0001
I0511 03:44:33.241756 39557 solver.cpp:218] Iteration 151500 (3.75449 iter/s, 26.6348s/100 iters), loss = 2.03318
I0511 03:44:33.241812 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.03318 (* 1 = 2.03318 loss)
I0511 03:44:33.241823 39557 sgd_solver.cpp:105] Iteration 151500, lr = 0.0001
I0511 03:44:59.886451 39557 solver.cpp:218] Iteration 151600 (3.75312 iter/s, 26.6445s/100 iters), loss = 3.10094
I0511 03:44:59.886610 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.10094 (* 1 = 3.10094 loss)
I0511 03:44:59.886623 39557 sgd_solver.cpp:105] Iteration 151600, lr = 0.0001
I0511 03:45:26.544592 39557 solver.cpp:218] Iteration 151700 (3.75124 iter/s, 26.6579s/100 iters), loss = 3.55476
I0511 03:45:26.544648 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.55476 (* 1 = 3.55476 loss)
I0511 03:45:26.544662 39557 sgd_solver.cpp:105] Iteration 151700, lr = 0.0001
I0511 03:45:53.363649 39557 solver.cpp:218] Iteration 151800 (3.72872 iter/s, 26.8189s/100 iters), loss = 3.15631
I0511 03:45:53.363852 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.15631 (* 1 = 3.15631 loss)
I0511 03:45:53.363869 39557 sgd_solver.cpp:105] Iteration 151800, lr = 0.0001
I0511 03:46:20.086782 39557 solver.cpp:218] Iteration 151900 (3.74212 iter/s, 26.7228s/100 iters), loss = 3.28581
I0511 03:46:20.086848 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.28581 (* 1 = 3.28581 loss)
I0511 03:46:20.086863 39557 sgd_solver.cpp:105] Iteration 151900, lr = 0.0001
I0511 03:46:46.900894 39557 solver.cpp:218] Iteration 152000 (3.7294 iter/s, 26.8139s/100 iters), loss = 2.48922
I0511 03:46:46.901065 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.48922 (* 1 = 2.48922 loss)
I0511 03:46:46.901085 39557 sgd_solver.cpp:105] Iteration 152000, lr = 0.0001
I0511 03:47:13.803542 39557 solver.cpp:218] Iteration 152100 (3.71715 iter/s, 26.9024s/100 iters), loss = 2.70529
I0511 03:47:13.803606 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.70529 (* 1 = 2.70529 loss)
I0511 03:47:13.803622 39557 sgd_solver.cpp:105] Iteration 152100, lr = 0.0001
I0511 03:47:40.516824 39557 solver.cpp:218] Iteration 152200 (3.74348 iter/s, 26.7131s/100 iters), loss = 2.47962
I0511 03:47:40.516978 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.47962 (* 1 = 2.47962 loss)
I0511 03:47:40.516995 39557 sgd_solver.cpp:105] Iteration 152200, lr = 0.0001
I0511 03:48:07.302465 39557 solver.cpp:218] Iteration 152300 (3.73338 iter/s, 26.7854s/100 iters), loss = 3.16257
I0511 03:48:07.302529 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.16257 (* 1 = 3.16257 loss)
I0511 03:48:07.302544 39557 sgd_solver.cpp:105] Iteration 152300, lr = 0.0001
I0511 03:48:33.992750 39557 solver.cpp:218] Iteration 152400 (3.74671 iter/s, 26.6901s/100 iters), loss = 2.99417
I0511 03:48:33.992904 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.99417 (* 1 = 2.99417 loss)
I0511 03:48:33.992921 39557 sgd_solver.cpp:105] Iteration 152400, lr = 0.0001
I0511 03:49:00.825040 39557 solver.cpp:218] Iteration 152500 (3.72689 iter/s, 26.832s/100 iters), loss = 2.80972
I0511 03:49:00.825103 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.80972 (* 1 = 2.80972 loss)
I0511 03:49:00.825115 39557 sgd_solver.cpp:105] Iteration 152500, lr = 0.0001
I0511 03:49:27.540341 39557 solver.cpp:218] Iteration 152600 (3.7432 iter/s, 26.7151s/100 iters), loss = 2.68194
I0511 03:49:27.540486 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.68194 (* 1 = 2.68194 loss)
I0511 03:49:27.540503 39557 sgd_solver.cpp:105] Iteration 152600, lr = 0.0001
I0511 03:49:54.247560 39557 solver.cpp:218] Iteration 152700 (3.74434 iter/s, 26.707s/100 iters), loss = 3.05702
I0511 03:49:54.247619 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.05702 (* 1 = 3.05702 loss)
I0511 03:49:54.247632 39557 sgd_solver.cpp:105] Iteration 152700, lr = 0.0001
I0511 03:50:21.030192 39557 solver.cpp:218] Iteration 152800 (3.73379 iter/s, 26.7825s/100 iters), loss = 3.3463
I0511 03:50:21.030341 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.3463 (* 1 = 3.3463 loss)
I0511 03:50:21.030359 39557 sgd_solver.cpp:105] Iteration 152800, lr = 0.0001
I0511 03:50:47.843551 39557 solver.cpp:218] Iteration 152900 (3.72952 iter/s, 26.8131s/100 iters), loss = 2.75394
I0511 03:50:47.843616 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.75394 (* 1 = 2.75394 loss)
I0511 03:50:47.843628 39557 sgd_solver.cpp:105] Iteration 152900, lr = 0.0001
I0511 03:51:14.676393 39557 solver.cpp:218] Iteration 153000 (3.7268 iter/s, 26.8327s/100 iters), loss = 3.08608
I0511 03:51:14.676551 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.08608 (* 1 = 3.08608 loss)
I0511 03:51:14.676568 39557 sgd_solver.cpp:105] Iteration 153000, lr = 0.0001
I0511 03:51:41.400298 39557 solver.cpp:218] Iteration 153100 (3.74201 iter/s, 26.7236s/100 iters), loss = 3.21889
I0511 03:51:41.400364 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.21889 (* 1 = 3.21889 loss)
I0511 03:51:41.400379 39557 sgd_solver.cpp:105] Iteration 153100, lr = 0.0001
I0511 03:52:08.049209 39557 solver.cpp:218] Iteration 153200 (3.75253 iter/s, 26.6487s/100 iters), loss = 3.33143
I0511 03:52:08.049422 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.33143 (* 1 = 3.33143 loss)
I0511 03:52:08.049438 39557 sgd_solver.cpp:105] Iteration 153200, lr = 0.0001
I0511 03:52:34.628825 39557 solver.cpp:218] Iteration 153300 (3.76233 iter/s, 26.5793s/100 iters), loss = 2.43822
I0511 03:52:34.628886 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.43822 (* 1 = 2.43822 loss)
I0511 03:52:34.628899 39557 sgd_solver.cpp:105] Iteration 153300, lr = 0.0001
I0511 03:53:01.279323 39557 solver.cpp:218] Iteration 153400 (3.7523 iter/s, 26.6503s/100 iters), loss = 2.60352
I0511 03:53:01.279510 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.60352 (* 1 = 2.60352 loss)
I0511 03:53:01.279527 39557 sgd_solver.cpp:105] Iteration 153400, lr = 0.0001
I0511 03:53:27.986220 39557 solver.cpp:218] Iteration 153500 (3.74439 iter/s, 26.7066s/100 iters), loss = 2.55612
I0511 03:53:27.986285 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.55612 (* 1 = 2.55612 loss)
I0511 03:53:27.986299 39557 sgd_solver.cpp:105] Iteration 153500, lr = 0.0001
I0511 03:53:54.773963 39557 solver.cpp:218] Iteration 153600 (3.73308 iter/s, 26.7875s/100 iters), loss = 2.23082
I0511 03:53:54.774121 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.23082 (* 1 = 2.23082 loss)
I0511 03:53:54.774137 39557 sgd_solver.cpp:105] Iteration 153600, lr = 0.0001
I0511 03:54:21.420534 39557 solver.cpp:218] Iteration 153700 (3.75287 iter/s, 26.6463s/100 iters), loss = 3.06061
I0511 03:54:21.420598 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.06061 (* 1 = 3.06061 loss)
I0511 03:54:21.420611 39557 sgd_solver.cpp:105] Iteration 153700, lr = 0.0001
I0511 03:54:48.199831 39557 solver.cpp:218] Iteration 153800 (3.73426 iter/s, 26.7791s/100 iters), loss = 2.26035
I0511 03:54:48.200018 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.26035 (* 1 = 2.26035 loss)
I0511 03:54:48.200038 39557 sgd_solver.cpp:105] Iteration 153800, lr = 0.0001
I0511 03:55:14.987295 39557 solver.cpp:218] Iteration 153900 (3.73313 iter/s, 26.7872s/100 iters), loss = 2.81445
I0511 03:55:14.987365 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.81445 (* 1 = 2.81445 loss)
I0511 03:55:14.987380 39557 sgd_solver.cpp:105] Iteration 153900, lr = 0.0001
I0511 03:55:41.678196 39557 solver.cpp:218] Iteration 154000 (3.74662 iter/s, 26.6907s/100 iters), loss = 2.51323
I0511 03:55:41.678366 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.51323 (* 1 = 2.51323 loss)
I0511 03:55:41.678381 39557 sgd_solver.cpp:105] Iteration 154000, lr = 0.0001
I0511 03:56:08.343863 39557 solver.cpp:218] Iteration 154100 (3.75018 iter/s, 26.6654s/100 iters), loss = 2.76635
I0511 03:56:08.343924 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.76635 (* 1 = 2.76635 loss)
I0511 03:56:08.343937 39557 sgd_solver.cpp:105] Iteration 154100, lr = 0.0001
I0511 03:56:34.983114 39557 solver.cpp:218] Iteration 154200 (3.75389 iter/s, 26.6391s/100 iters), loss = 3.29898
I0511 03:56:34.983278 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.29898 (* 1 = 3.29898 loss)
I0511 03:56:34.983291 39557 sgd_solver.cpp:105] Iteration 154200, lr = 0.0001
I0511 03:57:01.655527 39557 solver.cpp:218] Iteration 154300 (3.74923 iter/s, 26.6721s/100 iters), loss = 3.1108
I0511 03:57:01.655591 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.1108 (* 1 = 3.1108 loss)
I0511 03:57:01.655606 39557 sgd_solver.cpp:105] Iteration 154300, lr = 0.0001
I0511 03:57:28.476430 39557 solver.cpp:218] Iteration 154400 (3.72846 iter/s, 26.8207s/100 iters), loss = 2.04356
I0511 03:57:28.476635 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.04356 (* 1 = 2.04356 loss)
I0511 03:57:28.476655 39557 sgd_solver.cpp:105] Iteration 154400, lr = 0.0001
I0511 03:57:55.310638 39557 solver.cpp:218] Iteration 154500 (3.72663 iter/s, 26.8339s/100 iters), loss = 3.43488
I0511 03:57:55.310699 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.43488 (* 1 = 3.43488 loss)
I0511 03:57:55.310714 39557 sgd_solver.cpp:105] Iteration 154500, lr = 0.0001
I0511 03:58:22.159088 39557 solver.cpp:218] Iteration 154600 (3.72464 iter/s, 26.8483s/100 iters), loss = 2.76878
I0511 03:58:22.159256 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.76878 (* 1 = 2.76878 loss)
I0511 03:58:22.159272 39557 sgd_solver.cpp:105] Iteration 154600, lr = 0.0001
I0511 03:58:48.752265 39557 solver.cpp:218] Iteration 154700 (3.7604 iter/s, 26.5929s/100 iters), loss = 3.01749
I0511 03:58:48.752331 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.01749 (* 1 = 3.01749 loss)
I0511 03:58:48.752344 39557 sgd_solver.cpp:105] Iteration 154700, lr = 0.0001
I0511 03:59:15.493515 39557 solver.cpp:218] Iteration 154800 (3.73957 iter/s, 26.7411s/100 iters), loss = 3.82145
I0511 03:59:15.493702 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.82145 (* 1 = 3.82145 loss)
I0511 03:59:15.493721 39557 sgd_solver.cpp:105] Iteration 154800, lr = 0.0001
I0511 03:59:42.111934 39557 solver.cpp:218] Iteration 154900 (3.75684 iter/s, 26.6181s/100 iters), loss = 2.2702
I0511 03:59:42.111990 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.2702 (* 1 = 2.2702 loss)
I0511 03:59:42.112004 39557 sgd_solver.cpp:105] Iteration 154900, lr = 0.0001
I0511 04:00:08.514495 39557 solver.cpp:447] Snapshotting to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_155000.caffemodel
I0511 04:00:08.554117 39557 sgd_solver.cpp:273] Snapshotting solver state to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_155000.solverstate
I0511 04:00:08.840754 39557 solver.cpp:218] Iteration 155000 (3.7413 iter/s, 26.7287s/100 iters), loss = 2.3093
I0511 04:00:08.840806 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.3093 (* 1 = 2.3093 loss)
I0511 04:00:08.840817 39557 sgd_solver.cpp:105] Iteration 155000, lr = 0.0001
I0511 04:00:35.518944 39557 solver.cpp:218] Iteration 155100 (3.74841 iter/s, 26.678s/100 iters), loss = 2.63249
I0511 04:00:35.519003 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.63249 (* 1 = 2.63249 loss)
I0511 04:00:35.519016 39557 sgd_solver.cpp:105] Iteration 155100, lr = 0.0001
I0511 04:01:02.063669 39557 solver.cpp:218] Iteration 155200 (3.76725 iter/s, 26.5445s/100 iters), loss = 3.07577
I0511 04:01:02.063829 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.07577 (* 1 = 3.07577 loss)
I0511 04:01:02.063844 39557 sgd_solver.cpp:105] Iteration 155200, lr = 0.0001
I0511 04:01:28.716205 39557 solver.cpp:218] Iteration 155300 (3.75203 iter/s, 26.6522s/100 iters), loss = 2.22045
I0511 04:01:28.716269 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.22045 (* 1 = 2.22045 loss)
I0511 04:01:28.716281 39557 sgd_solver.cpp:105] Iteration 155300, lr = 0.0001
I0511 04:01:55.440337 39557 solver.cpp:218] Iteration 155400 (3.74196 iter/s, 26.7239s/100 iters), loss = 2.70011
I0511 04:01:55.440496 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.70011 (* 1 = 2.70011 loss)
I0511 04:01:55.440512 39557 sgd_solver.cpp:105] Iteration 155400, lr = 0.0001
I0511 04:02:22.143574 39557 solver.cpp:218] Iteration 155500 (3.7449 iter/s, 26.7029s/100 iters), loss = 2.51819
I0511 04:02:22.143640 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.51819 (* 1 = 2.51819 loss)
I0511 04:02:22.143652 39557 sgd_solver.cpp:105] Iteration 155500, lr = 0.0001
I0511 04:02:48.923085 39557 solver.cpp:218] Iteration 155600 (3.73423 iter/s, 26.7793s/100 iters), loss = 3.44369
I0511 04:02:48.923266 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.44369 (* 1 = 3.44369 loss)
I0511 04:02:48.923285 39557 sgd_solver.cpp:105] Iteration 155600, lr = 0.0001
I0511 04:03:15.712323 39557 solver.cpp:218] Iteration 155700 (3.73288 iter/s, 26.7889s/100 iters), loss = 2.33978
I0511 04:03:15.712404 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.33978 (* 1 = 2.33978 loss)
I0511 04:03:15.712415 39557 sgd_solver.cpp:105] Iteration 155700, lr = 0.0001
I0511 04:03:42.422258 39557 solver.cpp:218] Iteration 155800 (3.74396 iter/s, 26.7097s/100 iters), loss = 3.58871
I0511 04:03:42.422446 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.58871 (* 1 = 3.58871 loss)
I0511 04:03:42.422459 39557 sgd_solver.cpp:105] Iteration 155800, lr = 0.0001
I0511 04:04:09.151813 39557 solver.cpp:218] Iteration 155900 (3.74122 iter/s, 26.7292s/100 iters), loss = 3.15681
I0511 04:04:09.151875 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.15681 (* 1 = 3.15681 loss)
I0511 04:04:09.151885 39557 sgd_solver.cpp:105] Iteration 155900, lr = 0.0001
I0511 04:04:35.784924 39557 solver.cpp:218] Iteration 156000 (3.75475 iter/s, 26.6329s/100 iters), loss = 2.42639
I0511 04:04:35.785085 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.42639 (* 1 = 2.42639 loss)
I0511 04:04:35.785099 39557 sgd_solver.cpp:105] Iteration 156000, lr = 0.0001
I0511 04:05:02.451375 39557 solver.cpp:218] Iteration 156100 (3.75007 iter/s, 26.6662s/100 iters), loss = 2.56682
I0511 04:05:02.451434 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.56682 (* 1 = 2.56682 loss)
I0511 04:05:02.451447 39557 sgd_solver.cpp:105] Iteration 156100, lr = 0.0001
I0511 04:05:29.191696 39557 solver.cpp:218] Iteration 156200 (3.7397 iter/s, 26.7401s/100 iters), loss = 1.97198
I0511 04:05:29.191838 39557 solver.cpp:237]     Train net output #0: softmax_loss = 1.97198 (* 1 = 1.97198 loss)
I0511 04:05:29.191850 39557 sgd_solver.cpp:105] Iteration 156200, lr = 0.0001
I0511 04:05:55.819924 39557 solver.cpp:218] Iteration 156300 (3.75545 iter/s, 26.628s/100 iters), loss = 3.67386
I0511 04:05:55.819968 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.67386 (* 1 = 3.67386 loss)
I0511 04:05:55.819980 39557 sgd_solver.cpp:105] Iteration 156300, lr = 0.0001
I0511 04:06:22.468257 39557 solver.cpp:218] Iteration 156400 (3.7526 iter/s, 26.6482s/100 iters), loss = 2.12165
I0511 04:06:22.468410 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.12165 (* 1 = 2.12165 loss)
I0511 04:06:22.468421 39557 sgd_solver.cpp:105] Iteration 156400, lr = 0.0001
I0511 04:06:49.227897 39557 solver.cpp:218] Iteration 156500 (3.73701 iter/s, 26.7594s/100 iters), loss = 3.82558
I0511 04:06:49.227970 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.82558 (* 1 = 3.82558 loss)
I0511 04:06:49.227984 39557 sgd_solver.cpp:105] Iteration 156500, lr = 0.0001
I0511 04:07:15.936306 39557 solver.cpp:218] Iteration 156600 (3.74417 iter/s, 26.7082s/100 iters), loss = 2.57019
I0511 04:07:15.936466 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.57019 (* 1 = 2.57019 loss)
I0511 04:07:15.936481 39557 sgd_solver.cpp:105] Iteration 156600, lr = 0.0001
I0511 04:07:42.722353 39557 solver.cpp:218] Iteration 156700 (3.73333 iter/s, 26.7858s/100 iters), loss = 2.74346
I0511 04:07:42.722415 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.74346 (* 1 = 2.74346 loss)
I0511 04:07:42.722429 39557 sgd_solver.cpp:105] Iteration 156700, lr = 0.0001
I0511 04:08:09.470127 39557 solver.cpp:218] Iteration 156800 (3.73865 iter/s, 26.7476s/100 iters), loss = 2.31041
I0511 04:08:09.470314 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.31041 (* 1 = 2.31041 loss)
I0511 04:08:09.470333 39557 sgd_solver.cpp:105] Iteration 156800, lr = 0.0001
I0511 04:08:36.193328 39557 solver.cpp:218] Iteration 156900 (3.7421 iter/s, 26.7229s/100 iters), loss = 1.83996
I0511 04:08:36.193397 39557 solver.cpp:237]     Train net output #0: softmax_loss = 1.83996 (* 1 = 1.83996 loss)
I0511 04:08:36.193411 39557 sgd_solver.cpp:105] Iteration 156900, lr = 0.0001
I0511 04:09:02.932750 39557 solver.cpp:218] Iteration 157000 (3.73982 iter/s, 26.7393s/100 iters), loss = 3.39237
I0511 04:09:02.932957 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.39237 (* 1 = 3.39237 loss)
I0511 04:09:02.932976 39557 sgd_solver.cpp:105] Iteration 157000, lr = 0.0001
I0511 04:09:29.636348 39557 solver.cpp:218] Iteration 157100 (3.74485 iter/s, 26.7033s/100 iters), loss = 2.93387
I0511 04:09:29.636409 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.93387 (* 1 = 2.93387 loss)
I0511 04:09:29.636426 39557 sgd_solver.cpp:105] Iteration 157100, lr = 0.0001
I0511 04:09:56.472965 39557 solver.cpp:218] Iteration 157200 (3.72627 iter/s, 26.8365s/100 iters), loss = 2.24371
I0511 04:09:56.473126 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.24371 (* 1 = 2.24371 loss)
I0511 04:09:56.473142 39557 sgd_solver.cpp:105] Iteration 157200, lr = 0.0001
I0511 04:10:23.231186 39557 solver.cpp:218] Iteration 157300 (3.7372 iter/s, 26.758s/100 iters), loss = 2.69966
I0511 04:10:23.231257 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.69966 (* 1 = 2.69966 loss)
I0511 04:10:23.231274 39557 sgd_solver.cpp:105] Iteration 157300, lr = 0.0001
I0511 04:10:50.022236 39557 solver.cpp:218] Iteration 157400 (3.73261 iter/s, 26.7909s/100 iters), loss = 2.46027
I0511 04:10:50.022397 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.46027 (* 1 = 2.46027 loss)
I0511 04:10:50.022414 39557 sgd_solver.cpp:105] Iteration 157400, lr = 0.0001
I0511 04:11:16.723276 39557 solver.cpp:218] Iteration 157500 (3.74521 iter/s, 26.7008s/100 iters), loss = 3.3593
I0511 04:11:16.723341 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.3593 (* 1 = 3.3593 loss)
I0511 04:11:16.723356 39557 sgd_solver.cpp:105] Iteration 157500, lr = 0.0001
I0511 04:11:43.393270 39557 solver.cpp:218] Iteration 157600 (3.74955 iter/s, 26.6698s/100 iters), loss = 2.75687
I0511 04:11:43.393425 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.75687 (* 1 = 2.75687 loss)
I0511 04:11:43.393438 39557 sgd_solver.cpp:105] Iteration 157600, lr = 0.0001
I0511 04:12:10.071763 39557 solver.cpp:218] Iteration 157700 (3.74837 iter/s, 26.6782s/100 iters), loss = 3.13548
I0511 04:12:10.071820 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.13548 (* 1 = 3.13548 loss)
I0511 04:12:10.071831 39557 sgd_solver.cpp:105] Iteration 157700, lr = 0.0001
I0511 04:12:36.680725 39557 solver.cpp:218] Iteration 157800 (3.75815 iter/s, 26.6088s/100 iters), loss = 2.67416
I0511 04:12:36.680881 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.67416 (* 1 = 2.67416 loss)
I0511 04:12:36.680893 39557 sgd_solver.cpp:105] Iteration 157800, lr = 0.0001
I0511 04:13:03.360036 39557 solver.cpp:218] Iteration 157900 (3.74826 iter/s, 26.6791s/100 iters), loss = 2.73425
I0511 04:13:03.360100 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.73425 (* 1 = 2.73425 loss)
I0511 04:13:03.360112 39557 sgd_solver.cpp:105] Iteration 157900, lr = 0.0001
I0511 04:13:30.064877 39557 solver.cpp:218] Iteration 158000 (3.74466 iter/s, 26.7047s/100 iters), loss = 3.21031
I0511 04:13:30.065042 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.21031 (* 1 = 3.21031 loss)
I0511 04:13:30.065057 39557 sgd_solver.cpp:105] Iteration 158000, lr = 0.0001
I0511 04:13:56.836606 39557 solver.cpp:218] Iteration 158100 (3.73532 iter/s, 26.7715s/100 iters), loss = 3.04412
I0511 04:13:56.836669 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.04412 (* 1 = 3.04412 loss)
I0511 04:13:56.836683 39557 sgd_solver.cpp:105] Iteration 158100, lr = 0.0001
I0511 04:14:23.461220 39557 solver.cpp:218] Iteration 158200 (3.75595 iter/s, 26.6244s/100 iters), loss = 2.79039
I0511 04:14:23.461371 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.79039 (* 1 = 2.79039 loss)
I0511 04:14:23.461387 39557 sgd_solver.cpp:105] Iteration 158200, lr = 0.0001
I0511 04:14:50.238081 39557 solver.cpp:218] Iteration 158300 (3.7346 iter/s, 26.7766s/100 iters), loss = 3.55428
I0511 04:14:50.238145 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.55428 (* 1 = 3.55428 loss)
I0511 04:14:50.238159 39557 sgd_solver.cpp:105] Iteration 158300, lr = 0.0001
I0511 04:15:16.964646 39557 solver.cpp:218] Iteration 158400 (3.74162 iter/s, 26.7264s/100 iters), loss = 2.45958
I0511 04:15:16.964850 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.45958 (* 1 = 2.45958 loss)
I0511 04:15:16.964869 39557 sgd_solver.cpp:105] Iteration 158400, lr = 0.0001
I0511 04:15:43.659272 39557 solver.cpp:218] Iteration 158500 (3.74611 iter/s, 26.6943s/100 iters), loss = 2.28045
I0511 04:15:43.659340 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.28045 (* 1 = 2.28045 loss)
I0511 04:15:43.659354 39557 sgd_solver.cpp:105] Iteration 158500, lr = 0.0001
I0511 04:16:10.272032 39557 solver.cpp:218] Iteration 158600 (3.75762 iter/s, 26.6126s/100 iters), loss = 2.66113
I0511 04:16:10.272192 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.66113 (* 1 = 2.66113 loss)
I0511 04:16:10.272207 39557 sgd_solver.cpp:105] Iteration 158600, lr = 0.0001
I0511 04:16:36.894317 39557 solver.cpp:218] Iteration 158700 (3.75629 iter/s, 26.622s/100 iters), loss = 2.65168
I0511 04:16:36.894383 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.65168 (* 1 = 2.65168 loss)
I0511 04:16:36.894397 39557 sgd_solver.cpp:105] Iteration 158700, lr = 0.0001
I0511 04:17:03.673365 39557 solver.cpp:218] Iteration 158800 (3.73429 iter/s, 26.7789s/100 iters), loss = 2.87057
I0511 04:17:03.673516 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.87057 (* 1 = 2.87057 loss)
I0511 04:17:03.673533 39557 sgd_solver.cpp:105] Iteration 158800, lr = 0.0001
I0511 04:17:30.478662 39557 solver.cpp:218] Iteration 158900 (3.73064 iter/s, 26.805s/100 iters), loss = 2.73069
I0511 04:17:30.478749 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.73069 (* 1 = 2.73069 loss)
I0511 04:17:30.478761 39557 sgd_solver.cpp:105] Iteration 158900, lr = 0.0001
I0511 04:17:57.235208 39557 solver.cpp:218] Iteration 159000 (3.73743 iter/s, 26.7564s/100 iters), loss = 2.9338
I0511 04:17:57.235379 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.9338 (* 1 = 2.9338 loss)
I0511 04:17:57.235397 39557 sgd_solver.cpp:105] Iteration 159000, lr = 0.0001
I0511 04:18:23.904737 39557 solver.cpp:218] Iteration 159100 (3.74964 iter/s, 26.6693s/100 iters), loss = 2.44647
I0511 04:18:23.904810 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.44647 (* 1 = 2.44647 loss)
I0511 04:18:23.904825 39557 sgd_solver.cpp:105] Iteration 159100, lr = 0.0001
I0511 04:18:50.590837 39557 solver.cpp:218] Iteration 159200 (3.74729 iter/s, 26.6859s/100 iters), loss = 2.56578
I0511 04:18:50.591003 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.56578 (* 1 = 2.56578 loss)
I0511 04:18:50.591019 39557 sgd_solver.cpp:105] Iteration 159200, lr = 0.0001
I0511 04:19:17.232353 39557 solver.cpp:218] Iteration 159300 (3.75358 iter/s, 26.6412s/100 iters), loss = 2.84496
I0511 04:19:17.232409 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.84496 (* 1 = 2.84496 loss)
I0511 04:19:17.232420 39557 sgd_solver.cpp:105] Iteration 159300, lr = 0.0001
I0511 04:19:43.937515 39557 solver.cpp:218] Iteration 159400 (3.74462 iter/s, 26.705s/100 iters), loss = 2.10802
I0511 04:19:43.937667 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.10802 (* 1 = 2.10802 loss)
I0511 04:19:43.937682 39557 sgd_solver.cpp:105] Iteration 159400, lr = 0.0001
I0511 04:20:10.564013 39557 solver.cpp:218] Iteration 159500 (3.75569 iter/s, 26.6262s/100 iters), loss = 3.7261
I0511 04:20:10.564066 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.7261 (* 1 = 3.7261 loss)
I0511 04:20:10.564077 39557 sgd_solver.cpp:105] Iteration 159500, lr = 0.0001
I0511 04:20:37.345696 39557 solver.cpp:218] Iteration 159600 (3.73392 iter/s, 26.7815s/100 iters), loss = 3.23542
I0511 04:20:37.345849 39557 solver.cpp:237]     Train net output #0: softmax_loss = 3.23542 (* 1 = 3.23542 loss)
I0511 04:20:37.345866 39557 sgd_solver.cpp:105] Iteration 159600, lr = 0.0001
I0511 04:21:04.155004 39557 solver.cpp:218] Iteration 159700 (3.73009 iter/s, 26.809s/100 iters), loss = 2.89296
I0511 04:21:04.155069 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.89296 (* 1 = 2.89296 loss)
I0511 04:21:04.155083 39557 sgd_solver.cpp:105] Iteration 159700, lr = 0.0001
I0511 04:21:31.004858 39557 solver.cpp:218] Iteration 159800 (3.72444 iter/s, 26.8497s/100 iters), loss = 2.70495
I0511 04:21:31.005059 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.70495 (* 1 = 2.70495 loss)
I0511 04:21:31.005089 39557 sgd_solver.cpp:105] Iteration 159800, lr = 0.0001
I0511 04:21:57.710686 39557 solver.cpp:218] Iteration 159900 (3.74454 iter/s, 26.7055s/100 iters), loss = 2.91284
I0511 04:21:57.710749 39557 solver.cpp:237]     Train net output #0: softmax_loss = 2.91284 (* 1 = 2.91284 loss)
I0511 04:21:57.710763 39557 sgd_solver.cpp:105] Iteration 159900, lr = 0.0001
I0511 04:22:24.192131 39557 solver.cpp:447] Snapshotting to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_160000.caffemodel
I0511 04:22:24.228380 39557 sgd_solver.cpp:273] Snapshotting solver state to binary proto file face_snapshot/MobileFaceNet_16w_0pt3_model_iter_160000.solverstate
I0511 04:22:24.310762 39557 solver.cpp:310] Iteration 160000, loss = 3.07742
I0511 04:22:24.310803 39557 solver.cpp:315] Optimization Done.
I0511 04:22:26.604254 39557 caffe.cpp:259] Optimization Done.