I0428 13:50:30.138387 30475 upgrade_proto.cpp:1082] Attempting to upgrade input file specified using deprecated 'solver_type' field (enum)': /mnt/bigdisk/DIGITS-AMB-2/digits/jobs/20210428-120256-6c36/solver.prototxt
I0428 13:50:30.138696 30475 upgrade_proto.cpp:1089] Successfully upgraded file specified using deprecated 'solver_type' field (enum) to 'type' field (string).
W0428 13:50:30.138708 30475 upgrade_proto.cpp:1091] Note that future Caffe releases will only support 'type' field (string) for a solver's type.
I0428 13:50:30.138850 30475 caffe.cpp:218] Using GPUs 0
I0428 13:50:30.213814 30475 caffe.cpp:223] GPU 0: GeForce RTX 2080
I0428 13:50:30.691910 30475 solver.cpp:44] Initializing solver from parameters:
test_iter: 51
test_interval: 102
base_lr: 0.01
display: 12
max_iter: 10200
lr_policy: "exp"
gamma: 0.99980193
momentum: 0.9
weight_decay: 0.0001
snapshot: 102
snapshot_prefix: "snapshot"
solver_mode: GPU
device_id: 0
net: "train_val.prototxt"
train_state {
level: 0
stage: ""
}
type: "SGD"
I0428 13:50:30.693214 30475 solver.cpp:87] Creating training net from net file: train_val.prototxt
I0428 13:50:30.693990 30475 net.cpp:294] The NetState phase (0) differed from the phase (1) specified by a rule in layer val-data
I0428 13:50:30.694005 30475 net.cpp:294] The NetState phase (0) differed from the phase (1) specified by a rule in layer accuracy
I0428 13:50:30.694137 30475 net.cpp:51] Initializing net from parameters:
state {
phase: TRAIN
level: 0
stage: ""
}
layer {
name: "train-data"
type: "Data"
top: "data"
top: "label"
include {
phase: TRAIN
}
transform_param {
mirror: true
crop_size: 227
mean_file: "/mnt/bigdisk/DIGITS-AMB-2/digits/jobs/20210419-113214-d311/mean.binaryproto"
}
data_param {
source: "/mnt/bigdisk/DIGITS-AMB-2/digits/jobs/20210419-113214-d311/train_db"
batch_size: 128
backend: LMDB
}
}
layer {
name: "conv1"
type: "Convolution"
bottom: "data"
top: "conv1"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 96
kernel_size: 11
stride: 4
weight_filler {
type: "gaussian"
std: 0.01
}
bias_filler {
type: "constant"
value: 0
}
}
}
layer {
name: "relu1"
type: "ReLU"
bottom: "conv1"
top: "conv1"
}
layer {
name: "norm1"
type: "LRN"
bottom: "conv1"
top: "norm1"
lrn_param {
local_size: 5
alpha: 0.0001
beta: 0.75
}
}
layer {
name: "pool1"
type: "Pooling"
bottom: "norm1"
top: "pool1"
pooling_param {
pool: MAX
kernel_size: 3
stride: 2
}
}
layer {
name: "conv2"
type: "Convolution"
bottom: "pool1"
top: "conv2"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 256
pad: 2
kernel_size: 5
group: 2
weight_filler {
type: "gaussian"
std: 0.01
}
bias_filler {
type: "constant"
value: 0.1
}
}
}
layer {
name: "relu2"
type: "ReLU"
bottom: "conv2"
top: "conv2"
}
layer {
name: "norm2"
type: "LRN"
bottom: "conv2"
top: "norm2"
lrn_param {
local_size: 5
alpha: 0.0001
beta: 0.75
}
}
layer {
name: "pool2"
type: "Pooling"
bottom: "norm2"
top: "pool2"
pooling_param {
pool: MAX
kernel_size: 3
stride: 2
}
}
layer {
name: "conv3"
type: "Convolution"
bottom: "pool2"
top: "conv3"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 384
pad: 1
kernel_size: 3
weight_filler {
type: "gaussian"
std: 0.01
}
bias_filler {
type: "constant"
value: 0
}
}
}
layer {
name: "relu3"
type: "ReLU"
bottom: "conv3"
top: "conv3"
}
layer {
name: "conv3.5"
type: "Convolution"
bottom: "conv3"
top: "conv3.5"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 384
pad: 1
kernel_size: 9
weight_filler {
type: "gaussian"
std: 0.01
}
bias_filler {
type: "constant"
value: 0
}
}
}
layer {
name: "relu3.5"
type: "ReLU"
bottom: "conv3.5"
top: "conv3.5"
}
layer {
name: "conv4"
type: "Convolution"
bottom: "conv3.5"
top: "conv4"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 384
pad: 1
kernel_size: 3
group: 2
weight_filler {
type: "gaussian"
std: 0.01
}
bias_filler {
type: "constant"
value: 0.1
}
}
}
layer {
name: "relu4"
type: "ReLU"
bottom: "conv4"
top: "conv4"
}
layer {
name: "conv5"
type: "Convolution"
bottom: "conv4"
top: "conv5"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 256
pad: 1
kernel_size: 3
group: 2
weight_filler {
type: "gaussian"
std: 0.01
}
bias_filler {
type: "constant"
value: 0.1
}
}
}
layer {
name: "relu5"
type: "ReLU"
bottom: "conv5"
top: "conv5"
}
layer {
name: "pool5"
type: "Pooling"
bottom: "conv5"
top: "pool5"
pooling_param {
pool: MAX
kernel_size: 3
stride: 2
}
}
layer {
name: "fc6"
type: "InnerProduct"
bottom: "pool5"
top: "fc6"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
inner_product_param {
num_output: 4096
weight_filler {
type: "gaussian"
std: 0.005
}
bias_filler {
type: "constant"
value: 0.1
}
}
}
layer {
name: "relu6"
type: "ReLU"
bottom: "fc6"
top: "fc6"
}
layer {
name: "drop6"
type: "Dropout"
bottom: "fc6"
top: "fc6"
dropout_param {
dropout_ratio: 0.5
}
}
layer {
name: "fc7"
type: "InnerProduct"
bottom: "fc6"
top: "fc7"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
inner_product_param {
num_output: 4096
weight_filler {
type: "gaussian"
std: 0.005
}
bias_filler {
type: "constant"
value: 0.1
}
}
}
layer {
name: "relu7"
type: "ReLU"
bottom: "fc7"
top: "fc7"
}
layer {
name: "drop7"
type: "Dropout"
bottom: "fc7"
top: "fc7"
dropout_param {
dropout_ratio: 0.5
}
}
layer {
name: "fc8"
type: "InnerProduct"
bottom: "fc7"
top: "fc8"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
inner_product_param {
num_output: 196
weight_filler {
type: "gaussian"
std: 0.01
}
bias_filler {
type: "constant"
value: 0
}
}
}
layer {
name: "loss"
type: "SoftmaxWithLoss"
bottom: "fc8"
bottom: "label"
top: "loss"
}
I0428 13:50:30.694226 30475 layer_factory.hpp:77] Creating layer train-data
I0428 13:50:30.696998 30475 db_lmdb.cpp:35] Opened lmdb /mnt/bigdisk/DIGITS-AMB-2/digits/jobs/20210419-113214-d311/train_db
I0428 13:50:30.698127 30475 net.cpp:84] Creating Layer train-data
I0428 13:50:30.698139 30475 net.cpp:380] train-data -> data
I0428 13:50:30.698158 30475 net.cpp:380] train-data -> label
I0428 13:50:30.698168 30475 data_transformer.cpp:25] Loading mean file from: /mnt/bigdisk/DIGITS-AMB-2/digits/jobs/20210419-113214-d311/mean.binaryproto
I0428 13:50:30.702147 30475 data_layer.cpp:45] output data size: 128,3,227,227
I0428 13:50:30.833786 30475 net.cpp:122] Setting up train-data
I0428 13:50:30.833808 30475 net.cpp:129] Top shape: 128 3 227 227 (19787136)
I0428 13:50:30.833812 30475 net.cpp:129] Top shape: 128 (128)
I0428 13:50:30.833815 30475 net.cpp:137] Memory required for data: 79149056
I0428 13:50:30.833824 30475 layer_factory.hpp:77] Creating layer conv1
I0428 13:50:30.833843 30475 net.cpp:84] Creating Layer conv1
I0428 13:50:30.833848 30475 net.cpp:406] conv1 <- data
I0428 13:50:30.833859 30475 net.cpp:380] conv1 -> conv1
I0428 13:50:31.705168 30475 net.cpp:122] Setting up conv1
I0428 13:50:31.705188 30475 net.cpp:129] Top shape: 128 96 55 55 (37171200)
I0428 13:50:31.705190 30475 net.cpp:137] Memory required for data: 227833856
I0428 13:50:31.705209 30475 layer_factory.hpp:77] Creating layer relu1
I0428 13:50:31.705219 30475 net.cpp:84] Creating Layer relu1
I0428 13:50:31.705224 30475 net.cpp:406] relu1 <- conv1
I0428 13:50:31.705250 30475 net.cpp:367] relu1 -> conv1 (in-place)
I0428 13:50:31.705567 30475 net.cpp:122] Setting up relu1
I0428 13:50:31.705577 30475 net.cpp:129] Top shape: 128 96 55 55 (37171200)
I0428 13:50:31.705580 30475 net.cpp:137] Memory required for data: 376518656
I0428 13:50:31.705583 30475 layer_factory.hpp:77] Creating layer norm1
I0428 13:50:31.705591 30475 net.cpp:84] Creating Layer norm1
I0428 13:50:31.705595 30475 net.cpp:406] norm1 <- conv1
I0428 13:50:31.705598 30475 net.cpp:380] norm1 -> norm1
I0428 13:50:31.706104 30475 net.cpp:122] Setting up norm1
I0428 13:50:31.706113 30475 net.cpp:129] Top shape: 128 96 55 55 (37171200)
I0428 13:50:31.706116 30475 net.cpp:137] Memory required for data: 525203456
I0428 13:50:31.706120 30475 layer_factory.hpp:77] Creating layer pool1
I0428 13:50:31.706126 30475 net.cpp:84] Creating Layer pool1
I0428 13:50:31.706130 30475 net.cpp:406] pool1 <- norm1
I0428 13:50:31.706133 30475 net.cpp:380] pool1 -> pool1
I0428 13:50:31.706164 30475 net.cpp:122] Setting up pool1
I0428 13:50:31.706169 30475 net.cpp:129] Top shape: 128 96 27 27 (8957952)
I0428 13:50:31.706172 30475 net.cpp:137] Memory required for data: 561035264
I0428 13:50:31.706174 30475 layer_factory.hpp:77] Creating layer conv2
I0428 13:50:31.706183 30475 net.cpp:84] Creating Layer conv2
I0428 13:50:31.706187 30475 net.cpp:406] conv2 <- pool1
I0428 13:50:31.706192 30475 net.cpp:380] conv2 -> conv2
I0428 13:50:31.714076 30475 net.cpp:122] Setting up conv2
I0428 13:50:31.714087 30475 net.cpp:129] Top shape: 128 256 27 27 (23887872)
I0428 13:50:31.714092 30475 net.cpp:137] Memory required for data: 656586752
I0428 13:50:31.714100 30475 layer_factory.hpp:77] Creating layer relu2
I0428 13:50:31.714107 30475 net.cpp:84] Creating Layer relu2
I0428 13:50:31.714109 30475 net.cpp:406] relu2 <- conv2
I0428 13:50:31.714115 30475 net.cpp:367] relu2 -> conv2 (in-place)
I0428 13:50:31.714684 30475 net.cpp:122] Setting up relu2
I0428 13:50:31.714694 30475 net.cpp:129] Top shape: 128 256 27 27 (23887872)
I0428 13:50:31.714696 30475 net.cpp:137] Memory required for data: 752138240
I0428 13:50:31.714699 30475 layer_factory.hpp:77] Creating layer norm2
I0428 13:50:31.714709 30475 net.cpp:84] Creating Layer norm2
I0428 13:50:31.714711 30475 net.cpp:406] norm2 <- conv2
I0428 13:50:31.714716 30475 net.cpp:380] norm2 -> norm2
I0428 13:50:31.715107 30475 net.cpp:122] Setting up norm2
I0428 13:50:31.715117 30475 net.cpp:129] Top shape: 128 256 27 27 (23887872)
I0428 13:50:31.715121 30475 net.cpp:137] Memory required for data: 847689728
I0428 13:50:31.715123 30475 layer_factory.hpp:77] Creating layer pool2
I0428 13:50:31.715129 30475 net.cpp:84] Creating Layer pool2
I0428 13:50:31.715133 30475 net.cpp:406] pool2 <- norm2
I0428 13:50:31.715138 30475 net.cpp:380] pool2 -> pool2
I0428 13:50:31.715164 30475 net.cpp:122] Setting up pool2
I0428 13:50:31.715168 30475 net.cpp:129] Top shape: 128 256 13 13 (5537792)
I0428 13:50:31.715171 30475 net.cpp:137] Memory required for data: 869840896
I0428 13:50:31.715174 30475 layer_factory.hpp:77] Creating layer conv3
I0428 13:50:31.715183 30475 net.cpp:84] Creating Layer conv3
I0428 13:50:31.715186 30475 net.cpp:406] conv3 <- pool2
I0428 13:50:31.715191 30475 net.cpp:380] conv3 -> conv3
I0428 13:50:31.725786 30475 net.cpp:122] Setting up conv3
I0428 13:50:31.725800 30475 net.cpp:129] Top shape: 128 384 13 13 (8306688)
I0428 13:50:31.725803 30475 net.cpp:137] Memory required for data: 903067648
I0428 13:50:31.725811 30475 layer_factory.hpp:77] Creating layer relu3
I0428 13:50:31.725819 30475 net.cpp:84] Creating Layer relu3
I0428 13:50:31.725822 30475 net.cpp:406] relu3 <- conv3
I0428 13:50:31.725827 30475 net.cpp:367] relu3 -> conv3 (in-place)
I0428 13:50:31.726419 30475 net.cpp:122] Setting up relu3
I0428 13:50:31.726428 30475 net.cpp:129] Top shape: 128 384 13 13 (8306688)
I0428 13:50:31.726433 30475 net.cpp:137] Memory required for data: 936294400
I0428 13:50:31.726435 30475 layer_factory.hpp:77] Creating layer conv3.5
I0428 13:50:31.726444 30475 net.cpp:84] Creating Layer conv3.5
I0428 13:50:31.726466 30475 net.cpp:406] conv3.5 <- conv3
I0428 13:50:31.726472 30475 net.cpp:380] conv3.5 -> conv3.5
I0428 13:50:31.843281 30475 net.cpp:122] Setting up conv3.5
I0428 13:50:31.843299 30475 net.cpp:129] Top shape: 128 384 7 7 (2408448)
I0428 13:50:31.843302 30475 net.cpp:137] Memory required for data: 945928192
I0428 13:50:31.843310 30475 layer_factory.hpp:77] Creating layer relu3.5
I0428 13:50:31.843318 30475 net.cpp:84] Creating Layer relu3.5
I0428 13:50:31.843322 30475 net.cpp:406] relu3.5 <- conv3.5
I0428 13:50:31.843328 30475 net.cpp:367] relu3.5 -> conv3.5 (in-place)
I0428 13:50:31.843883 30475 net.cpp:122] Setting up relu3.5
I0428 13:50:31.843894 30475 net.cpp:129] Top shape: 128 384 7 7 (2408448)
I0428 13:50:31.843895 30475 net.cpp:137] Memory required for data: 955561984
I0428 13:50:31.843899 30475 layer_factory.hpp:77] Creating layer conv4
I0428 13:50:31.843909 30475 net.cpp:84] Creating Layer conv4
I0428 13:50:31.843912 30475 net.cpp:406] conv4 <- conv3.5
I0428 13:50:31.843917 30475 net.cpp:380] conv4 -> conv4
I0428 13:50:31.854135 30475 net.cpp:122] Setting up conv4
I0428 13:50:31.854152 30475 net.cpp:129] Top shape: 128 384 7 7 (2408448)
I0428 13:50:31.854156 30475 net.cpp:137] Memory required for data: 965195776
I0428 13:50:31.854167 30475 layer_factory.hpp:77] Creating layer relu4
I0428 13:50:31.854176 30475 net.cpp:84] Creating Layer relu4
I0428 13:50:31.854180 30475 net.cpp:406] relu4 <- conv4
I0428 13:50:31.854185 30475 net.cpp:367] relu4 -> conv4 (in-place)
I0428 13:50:31.854564 30475 net.cpp:122] Setting up relu4
I0428 13:50:31.854573 30475 net.cpp:129] Top shape: 128 384 7 7 (2408448)
I0428 13:50:31.854576 30475 net.cpp:137] Memory required for data: 974829568
I0428 13:50:31.854579 30475 layer_factory.hpp:77] Creating layer conv5
I0428 13:50:31.854589 30475 net.cpp:84] Creating Layer conv5
I0428 13:50:31.854593 30475 net.cpp:406] conv5 <- conv4
I0428 13:50:31.854605 30475 net.cpp:380] conv5 -> conv5
I0428 13:50:31.863696 30475 net.cpp:122] Setting up conv5
I0428 13:50:31.863713 30475 net.cpp:129] Top shape: 128 256 7 7 (1605632)
I0428 13:50:31.863715 30475 net.cpp:137] Memory required for data: 981252096
I0428 13:50:31.863723 30475 layer_factory.hpp:77] Creating layer relu5
I0428 13:50:31.863734 30475 net.cpp:84] Creating Layer relu5
I0428 13:50:31.863739 30475 net.cpp:406] relu5 <- conv5
I0428 13:50:31.863744 30475 net.cpp:367] relu5 -> conv5 (in-place)
I0428 13:50:31.864301 30475 net.cpp:122] Setting up relu5
I0428 13:50:31.864310 30475 net.cpp:129] Top shape: 128 256 7 7 (1605632)
I0428 13:50:31.864313 30475 net.cpp:137] Memory required for data: 987674624
I0428 13:50:31.864316 30475 layer_factory.hpp:77] Creating layer pool5
I0428 13:50:31.864326 30475 net.cpp:84] Creating Layer pool5
I0428 13:50:31.864329 30475 net.cpp:406] pool5 <- conv5
I0428 13:50:31.864334 30475 net.cpp:380] pool5 -> pool5
I0428 13:50:31.864370 30475 net.cpp:122] Setting up pool5
I0428 13:50:31.864375 30475 net.cpp:129] Top shape: 128 256 3 3 (294912)
I0428 13:50:31.864378 30475 net.cpp:137] Memory required for data: 988854272
I0428 13:50:31.864380 30475 layer_factory.hpp:77] Creating layer fc6
I0428 13:50:31.864388 30475 net.cpp:84] Creating Layer fc6
I0428 13:50:31.864392 30475 net.cpp:406] fc6 <- pool5
I0428 13:50:31.864395 30475 net.cpp:380] fc6 -> fc6
I0428 13:50:31.968770 30475 net.cpp:122] Setting up fc6
I0428 13:50:31.968791 30475 net.cpp:129] Top shape: 128 4096 (524288)
I0428 13:50:31.968793 30475 net.cpp:137] Memory required for data: 990951424
I0428 13:50:31.968802 30475 layer_factory.hpp:77] Creating layer relu6
I0428 13:50:31.968811 30475 net.cpp:84] Creating Layer relu6
I0428 13:50:31.968814 30475 net.cpp:406] relu6 <- fc6
I0428 13:50:31.968820 30475 net.cpp:367] relu6 -> fc6 (in-place)
I0428 13:50:31.969730 30475 net.cpp:122] Setting up relu6
I0428 13:50:31.969740 30475 net.cpp:129] Top shape: 128 4096 (524288)
I0428 13:50:31.969743 30475 net.cpp:137] Memory required for data: 993048576
I0428 13:50:31.969746 30475 layer_factory.hpp:77] Creating layer drop6
I0428 13:50:31.969753 30475 net.cpp:84] Creating Layer drop6
I0428 13:50:31.969774 30475 net.cpp:406] drop6 <- fc6
I0428 13:50:31.969779 30475 net.cpp:367] drop6 -> fc6 (in-place)
I0428 13:50:31.969808 30475 net.cpp:122] Setting up drop6
I0428 13:50:31.969815 30475 net.cpp:129] Top shape: 128 4096 (524288)
I0428 13:50:31.969817 30475 net.cpp:137] Memory required for data: 995145728
I0428 13:50:31.969820 30475 layer_factory.hpp:77] Creating layer fc7
I0428 13:50:31.969827 30475 net.cpp:84] Creating Layer fc7
I0428 13:50:31.969830 30475 net.cpp:406] fc7 <- fc6
I0428 13:50:31.969835 30475 net.cpp:380] fc7 -> fc7
I0428 13:50:32.129160 30475 net.cpp:122] Setting up fc7
I0428 13:50:32.129177 30475 net.cpp:129] Top shape: 128 4096 (524288)
I0428 13:50:32.129180 30475 net.cpp:137] Memory required for data: 997242880
I0428 13:50:32.129189 30475 layer_factory.hpp:77] Creating layer relu7
I0428 13:50:32.129199 30475 net.cpp:84] Creating Layer relu7
I0428 13:50:32.129202 30475 net.cpp:406] relu7 <- fc7
I0428 13:50:32.129209 30475 net.cpp:367] relu7 -> fc7 (in-place)
I0428 13:50:32.130399 30475 net.cpp:122] Setting up relu7
I0428 13:50:32.130410 30475 net.cpp:129] Top shape: 128 4096 (524288)
I0428 13:50:32.130414 30475 net.cpp:137] Memory required for data: 999340032
I0428 13:50:32.130417 30475 layer_factory.hpp:77] Creating layer drop7
I0428 13:50:32.130424 30475 net.cpp:84] Creating Layer drop7
I0428 13:50:32.130426 30475 net.cpp:406] drop7 <- fc7
I0428 13:50:32.130432 30475 net.cpp:367] drop7 -> fc7 (in-place)
I0428 13:50:32.130457 30475 net.cpp:122] Setting up drop7
I0428 13:50:32.130462 30475 net.cpp:129] Top shape: 128 4096 (524288)
I0428 13:50:32.130465 30475 net.cpp:137] Memory required for data: 1001437184
I0428 13:50:32.130467 30475 layer_factory.hpp:77] Creating layer fc8
I0428 13:50:32.130475 30475 net.cpp:84] Creating Layer fc8
I0428 13:50:32.130478 30475 net.cpp:406] fc8 <- fc7
I0428 13:50:32.130483 30475 net.cpp:380] fc8 -> fc8
I0428 13:50:32.138383 30475 net.cpp:122] Setting up fc8
I0428 13:50:32.138393 30475 net.cpp:129] Top shape: 128 196 (25088)
I0428 13:50:32.138396 30475 net.cpp:137] Memory required for data: 1001537536
I0428 13:50:32.138409 30475 layer_factory.hpp:77] Creating layer loss
I0428 13:50:32.138415 30475 net.cpp:84] Creating Layer loss
I0428 13:50:32.138418 30475 net.cpp:406] loss <- fc8
I0428 13:50:32.138422 30475 net.cpp:406] loss <- label
I0428 13:50:32.138429 30475 net.cpp:380] loss -> loss
I0428 13:50:32.138442 30475 layer_factory.hpp:77] Creating layer loss
I0428 13:50:32.138929 30475 net.cpp:122] Setting up loss
I0428 13:50:32.138938 30475 net.cpp:129] Top shape: (1)
I0428 13:50:32.138942 30475 net.cpp:132]     with loss weight 1
I0428 13:50:32.138958 30475 net.cpp:137] Memory required for data: 1001537540
I0428 13:50:32.138962 30475 net.cpp:198] loss needs backward computation.
I0428 13:50:32.138967 30475 net.cpp:198] fc8 needs backward computation.
I0428 13:50:32.138970 30475 net.cpp:198] drop7 needs backward computation.
I0428 13:50:32.138973 30475 net.cpp:198] relu7 needs backward computation.
I0428 13:50:32.138975 30475 net.cpp:198] fc7 needs backward computation.
I0428 13:50:32.138978 30475 net.cpp:198] drop6 needs backward computation.
I0428 13:50:32.138981 30475 net.cpp:198] relu6 needs backward computation.
I0428 13:50:32.138983 30475 net.cpp:198] fc6 needs backward computation.
I0428 13:50:32.138986 30475 net.cpp:198] pool5 needs backward computation.
I0428 13:50:32.138989 30475 net.cpp:198] relu5 needs backward computation.
I0428 13:50:32.138993 30475 net.cpp:198] conv5 needs backward computation.
I0428 13:50:32.138994 30475 net.cpp:198] relu4 needs backward computation.
I0428 13:50:32.138998 30475 net.cpp:198] conv4 needs backward computation.
I0428 13:50:32.139000 30475 net.cpp:198] relu3.5 needs backward computation.
I0428 13:50:32.139004 30475 net.cpp:198] conv3.5 needs backward computation.
I0428 13:50:32.139008 30475 net.cpp:198] relu3 needs backward computation.
I0428 13:50:32.139010 30475 net.cpp:198] conv3 needs backward computation.
I0428 13:50:32.139014 30475 net.cpp:198] pool2 needs backward computation.
I0428 13:50:32.139036 30475 net.cpp:198] norm2 needs backward computation.
I0428 13:50:32.139039 30475 net.cpp:198] relu2 needs backward computation.
I0428 13:50:32.139041 30475 net.cpp:198] conv2 needs backward computation.
I0428 13:50:32.139045 30475 net.cpp:198] pool1 needs backward computation.
I0428 13:50:32.139047 30475 net.cpp:198] norm1 needs backward computation.
I0428 13:50:32.139050 30475 net.cpp:198] relu1 needs backward computation.
I0428 13:50:32.139052 30475 net.cpp:198] conv1 needs backward computation.
I0428 13:50:32.139056 30475 net.cpp:200] train-data does not need backward computation.
I0428 13:50:32.139060 30475 net.cpp:242] This network produces output loss
I0428 13:50:32.139075 30475 net.cpp:255] Network initialization done.
I0428 13:50:32.139602 30475 solver.cpp:172] Creating test net (#0) specified by net file: train_val.prototxt
I0428 13:50:32.139633 30475 net.cpp:294] The NetState phase (1) differed from the phase (0) specified by a rule in layer train-data
I0428 13:50:32.139777 30475 net.cpp:51] Initializing net from parameters:
state {
phase: TEST
}
layer {
name: "val-data"
type: "Data"
top: "data"
top: "label"
include {
phase: TEST
}
transform_param {
crop_size: 227
mean_file: "/mnt/bigdisk/DIGITS-AMB-2/digits/jobs/20210419-113214-d311/mean.binaryproto"
}
data_param {
source: "/mnt/bigdisk/DIGITS-AMB-2/digits/jobs/20210419-113214-d311/val_db"
batch_size: 32
backend: LMDB
}
}
layer {
name: "conv1"
type: "Convolution"
bottom: "data"
top: "conv1"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 96
kernel_size: 11
stride: 4
weight_filler {
type: "gaussian"
std: 0.01
}
bias_filler {
type: "constant"
value: 0
}
}
}
layer {
name: "relu1"
type: "ReLU"
bottom: "conv1"
top: "conv1"
}
layer {
name: "norm1"
type: "LRN"
bottom: "conv1"
top: "norm1"
lrn_param {
local_size: 5
alpha: 0.0001
beta: 0.75
}
}
layer {
name: "pool1"
type: "Pooling"
bottom: "norm1"
top: "pool1"
pooling_param {
pool: MAX
kernel_size: 3
stride: 2
}
}
layer {
name: "conv2"
type: "Convolution"
bottom: "pool1"
top: "conv2"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 256
pad: 2
kernel_size: 5
group: 2
weight_filler {
type: "gaussian"
std: 0.01
}
bias_filler {
type: "constant"
value: 0.1
}
}
}
layer {
name: "relu2"
type: "ReLU"
bottom: "conv2"
top: "conv2"
}
layer {
name: "norm2"
type: "LRN"
bottom: "conv2"
top: "norm2"
lrn_param {
local_size: 5
alpha: 0.0001
beta: 0.75
}
}
layer {
name: "pool2"
type: "Pooling"
bottom: "norm2"
top: "pool2"
pooling_param {
pool: MAX
kernel_size: 3
stride: 2
}
}
layer {
name: "conv3"
type: "Convolution"
bottom: "pool2"
top: "conv3"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 384
pad: 1
kernel_size: 3
weight_filler {
type: "gaussian"
std: 0.01
}
bias_filler {
type: "constant"
value: 0
}
}
}
layer {
name: "relu3"
type: "ReLU"
bottom: "conv3"
top: "conv3"
}
layer {
name: "conv3.5"
type: "Convolution"
bottom: "conv3"
top: "conv3.5"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 384
pad: 1
kernel_size: 9
weight_filler {
type: "gaussian"
std: 0.01
}
bias_filler {
type: "constant"
value: 0
}
}
}
layer {
name: "relu3.5"
type: "ReLU"
bottom: "conv3.5"
top: "conv3.5"
}
layer {
name: "conv4"
type: "Convolution"
bottom: "conv3.5"
top: "conv4"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 384
pad: 1
kernel_size: 3
group: 2
weight_filler {
type: "gaussian"
std: 0.01
}
bias_filler {
type: "constant"
value: 0.1
}
}
}
layer {
name: "relu4"
type: "ReLU"
bottom: "conv4"
top: "conv4"
}
layer {
name: "conv5"
type: "Convolution"
bottom: "conv4"
top: "conv5"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 256
pad: 1
kernel_size: 3
group: 2
weight_filler {
type: "gaussian"
std: 0.01
}
bias_filler {
type: "constant"
value: 0.1
}
}
}
layer {
name: "relu5"
type: "ReLU"
bottom: "conv5"
top: "conv5"
}
layer {
name: "pool5"
type: "Pooling"
bottom: "conv5"
top: "pool5"
pooling_param {
pool: MAX
kernel_size: 3
stride: 2
}
}
layer {
name: "fc6"
type: "InnerProduct"
bottom: "pool5"
top: "fc6"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
inner_product_param {
num_output: 4096
weight_filler {
type: "gaussian"
std: 0.005
}
bias_filler {
type: "constant"
value: 0.1
}
}
}
layer {
name: "relu6"
type: "ReLU"
bottom: "fc6"
top: "fc6"
}
layer {
name: "drop6"
type: "Dropout"
bottom: "fc6"
top: "fc6"
dropout_param {
dropout_ratio: 0.5
}
}
layer {
name: "fc7"
type: "InnerProduct"
bottom: "fc6"
top: "fc7"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
inner_product_param {
num_output: 4096
weight_filler {
type: "gaussian"
std: 0.005
}
bias_filler {
type: "constant"
value: 0.1
}
}
}
layer {
name: "relu7"
type: "ReLU"
bottom: "fc7"
top: "fc7"
}
layer {
name: "drop7"
type: "Dropout"
bottom: "fc7"
top: "fc7"
dropout_param {
dropout_ratio: 0.5
}
}
layer {
name: "fc8"
type: "InnerProduct"
bottom: "fc7"
top: "fc8"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
inner_product_param {
num_output: 196
weight_filler {
type: "gaussian"
std: 0.01
}
bias_filler {
type: "constant"
value: 0
}
}
}
layer {
name: "accuracy"
type: "Accuracy"
bottom: "fc8"
bottom: "label"
top: "accuracy"
include {
phase: TEST
}
}
layer {
name: "loss"
type: "SoftmaxWithLoss"
bottom: "fc8"
bottom: "label"
top: "loss"
}
I0428 13:50:32.139889 30475 layer_factory.hpp:77] Creating layer val-data
I0428 13:50:32.160320 30475 db_lmdb.cpp:35] Opened lmdb /mnt/bigdisk/DIGITS-AMB-2/digits/jobs/20210419-113214-d311/val_db
I0428 13:50:32.161829 30475 net.cpp:84] Creating Layer val-data
I0428 13:50:32.161852 30475 net.cpp:380] val-data -> data
I0428 13:50:32.161866 30475 net.cpp:380] val-data -> label
I0428 13:50:32.161876 30475 data_transformer.cpp:25] Loading mean file from: /mnt/bigdisk/DIGITS-AMB-2/digits/jobs/20210419-113214-d311/mean.binaryproto
I0428 13:50:32.167361 30475 data_layer.cpp:45] output data size: 32,3,227,227
I0428 13:50:32.202591 30475 net.cpp:122] Setting up val-data
I0428 13:50:32.202616 30475 net.cpp:129] Top shape: 32 3 227 227 (4946784)
I0428 13:50:32.202621 30475 net.cpp:129] Top shape: 32 (32)
I0428 13:50:32.202623 30475 net.cpp:137] Memory required for data: 19787264
I0428 13:50:32.202630 30475 layer_factory.hpp:77] Creating layer label_val-data_1_split
I0428 13:50:32.202639 30475 net.cpp:84] Creating Layer label_val-data_1_split
I0428 13:50:32.202643 30475 net.cpp:406] label_val-data_1_split <- label
I0428 13:50:32.202649 30475 net.cpp:380] label_val-data_1_split -> label_val-data_1_split_0
I0428 13:50:32.202657 30475 net.cpp:380] label_val-data_1_split -> label_val-data_1_split_1
I0428 13:50:32.202700 30475 net.cpp:122] Setting up label_val-data_1_split
I0428 13:50:32.202705 30475 net.cpp:129] Top shape: 32 (32)
I0428 13:50:32.202708 30475 net.cpp:129] Top shape: 32 (32)
I0428 13:50:32.202710 30475 net.cpp:137] Memory required for data: 19787520
I0428 13:50:32.202734 30475 layer_factory.hpp:77] Creating layer conv1
I0428 13:50:32.202745 30475 net.cpp:84] Creating Layer conv1
I0428 13:50:32.202749 30475 net.cpp:406] conv1 <- data
I0428 13:50:32.202754 30475 net.cpp:380] conv1 -> conv1
I0428 13:50:32.206975 30475 net.cpp:122] Setting up conv1
I0428 13:50:32.206988 30475 net.cpp:129] Top shape: 32 96 55 55 (9292800)
I0428 13:50:32.206991 30475 net.cpp:137] Memory required for data: 56958720
I0428 13:50:32.207001 30475 layer_factory.hpp:77] Creating layer relu1
I0428 13:50:32.207008 30475 net.cpp:84] Creating Layer relu1
I0428 13:50:32.207011 30475 net.cpp:406] relu1 <- conv1
I0428 13:50:32.207015 30475 net.cpp:367] relu1 -> conv1 (in-place)
I0428 13:50:32.207532 30475 net.cpp:122] Setting up relu1
I0428 13:50:32.207542 30475 net.cpp:129] Top shape: 32 96 55 55 (9292800)
I0428 13:50:32.207545 30475 net.cpp:137] Memory required for data: 94129920
I0428 13:50:32.207548 30475 layer_factory.hpp:77] Creating layer norm1
I0428 13:50:32.207557 30475 net.cpp:84] Creating Layer norm1
I0428 13:50:32.207561 30475 net.cpp:406] norm1 <- conv1
I0428 13:50:32.207566 30475 net.cpp:380] norm1 -> norm1
I0428 13:50:32.207891 30475 net.cpp:122] Setting up norm1
I0428 13:50:32.207901 30475 net.cpp:129] Top shape: 32 96 55 55 (9292800)
I0428 13:50:32.207903 30475 net.cpp:137] Memory required for data: 131301120
I0428 13:50:32.207906 30475 layer_factory.hpp:77] Creating layer pool1
I0428 13:50:32.207912 30475 net.cpp:84] Creating Layer pool1
I0428 13:50:32.207916 30475 net.cpp:406] pool1 <- norm1
I0428 13:50:32.207919 30475 net.cpp:380] pool1 -> pool1
I0428 13:50:32.207945 30475 net.cpp:122] Setting up pool1
I0428 13:50:32.207949 30475 net.cpp:129] Top shape: 32 96 27 27 (2239488)
I0428 13:50:32.207952 30475 net.cpp:137] Memory required for data: 140259072
I0428 13:50:32.207955 30475 layer_factory.hpp:77] Creating layer conv2
I0428 13:50:32.207962 30475 net.cpp:84] Creating Layer conv2
I0428 13:50:32.207965 30475 net.cpp:406] conv2 <- pool1
I0428 13:50:32.207970 30475 net.cpp:380] conv2 -> conv2
I0428 13:50:32.216063 30475 net.cpp:122] Setting up conv2
I0428 13:50:32.216080 30475 net.cpp:129] Top shape: 32 256 27 27 (5971968)
I0428 13:50:32.216084 30475 net.cpp:137] Memory required for data: 164146944
I0428 13:50:32.216094 30475 layer_factory.hpp:77] Creating layer relu2
I0428 13:50:32.216101 30475 net.cpp:84] Creating Layer relu2
I0428 13:50:32.216104 30475 net.cpp:406] relu2 <- conv2
I0428 13:50:32.216109 30475 net.cpp:367] relu2 -> conv2 (in-place)
I0428 13:50:32.217821 30475 net.cpp:122] Setting up relu2
I0428 13:50:32.217833 30475 net.cpp:129] Top shape: 32 256 27 27 (5971968)
I0428 13:50:32.217835 30475 net.cpp:137] Memory required for data: 188034816
I0428 13:50:32.217839 30475 layer_factory.hpp:77] Creating layer norm2
I0428 13:50:32.217851 30475 net.cpp:84] Creating Layer norm2
I0428 13:50:32.217854 30475 net.cpp:406] norm2 <- conv2
I0428 13:50:32.217860 30475 net.cpp:380] norm2 -> norm2
I0428 13:50:32.218447 30475 net.cpp:122] Setting up norm2
I0428 13:50:32.218457 30475 net.cpp:129] Top shape: 32 256 27 27 (5971968)
I0428 13:50:32.218461 30475 net.cpp:137] Memory required for data: 211922688
I0428 13:50:32.218464 30475 layer_factory.hpp:77] Creating layer pool2
I0428 13:50:32.218472 30475 net.cpp:84] Creating Layer pool2
I0428 13:50:32.218474 30475 net.cpp:406] pool2 <- norm2
I0428 13:50:32.218478 30475 net.cpp:380] pool2 -> pool2
I0428 13:50:32.218508 30475 net.cpp:122] Setting up pool2
I0428 13:50:32.218513 30475 net.cpp:129] Top shape: 32 256 13 13 (1384448)
I0428 13:50:32.218516 30475 net.cpp:137] Memory required for data: 217460480
I0428 13:50:32.218518 30475 layer_factory.hpp:77] Creating layer conv3
I0428 13:50:32.218528 30475 net.cpp:84] Creating Layer conv3
I0428 13:50:32.218531 30475 net.cpp:406] conv3 <- pool2
I0428 13:50:32.218539 30475 net.cpp:380] conv3 -> conv3
I0428 13:50:32.229190 30475 net.cpp:122] Setting up conv3
I0428 13:50:32.229207 30475 net.cpp:129] Top shape: 32 384 13 13 (2076672)
I0428 13:50:32.229209 30475 net.cpp:137] Memory required for data: 225767168
I0428 13:50:32.229239 30475 layer_factory.hpp:77] Creating layer relu3
I0428 13:50:32.229247 30475 net.cpp:84] Creating Layer relu3
I0428 13:50:32.229251 30475 net.cpp:406] relu3 <- conv3
I0428 13:50:32.229259 30475 net.cpp:367] relu3 -> conv3 (in-place)
I0428 13:50:32.229817 30475 net.cpp:122] Setting up relu3
I0428 13:50:32.229826 30475 net.cpp:129] Top shape: 32 384 13 13 (2076672)
I0428 13:50:32.229830 30475 net.cpp:137] Memory required for data: 234073856
I0428 13:50:32.229832 30475 layer_factory.hpp:77] Creating layer conv3.5
I0428 13:50:32.229843 30475 net.cpp:84] Creating Layer conv3.5
I0428 13:50:32.229846 30475 net.cpp:406] conv3.5 <- conv3
I0428 13:50:32.229853 30475 net.cpp:380] conv3.5 -> conv3.5
I0428 13:50:32.345742 30475 net.cpp:122] Setting up conv3.5
I0428 13:50:32.345762 30475 net.cpp:129] Top shape: 32 384 7 7 (602112)
I0428 13:50:32.345764 30475 net.cpp:137] Memory required for data: 236482304
I0428 13:50:32.345772 30475 layer_factory.hpp:77] Creating layer relu3.5
I0428 13:50:32.345780 30475 net.cpp:84] Creating Layer relu3.5
I0428 13:50:32.345784 30475 net.cpp:406] relu3.5 <- conv3.5
I0428 13:50:32.345793 30475 net.cpp:367] relu3.5 -> conv3.5 (in-place)
I0428 13:50:32.346525 30475 net.cpp:122] Setting up relu3.5
I0428 13:50:32.346534 30475 net.cpp:129] Top shape: 32 384 7 7 (602112)
I0428 13:50:32.346537 30475 net.cpp:137] Memory required for data: 238890752
I0428 13:50:32.346540 30475 layer_factory.hpp:77] Creating layer conv4
I0428 13:50:32.346550 30475 net.cpp:84] Creating Layer conv4
I0428 13:50:32.346554 30475 net.cpp:406] conv4 <- conv3.5
I0428 13:50:32.346561 30475 net.cpp:380] conv4 -> conv4
I0428 13:50:32.358045 30475 net.cpp:122] Setting up conv4
I0428 13:50:32.358063 30475 net.cpp:129] Top shape: 32 384 7 7 (602112)
I0428 13:50:32.358067 30475 net.cpp:137] Memory required for data: 241299200
I0428 13:50:32.358078 30475 layer_factory.hpp:77] Creating layer relu4
I0428 13:50:32.358088 30475 net.cpp:84] Creating Layer relu4
I0428 13:50:32.358093 30475 net.cpp:406] relu4 <- conv4
I0428 13:50:32.358098 30475 net.cpp:367] relu4 -> conv4 (in-place)
I0428 13:50:32.358664 30475 net.cpp:122] Setting up relu4
I0428 13:50:32.358673 30475 net.cpp:129] Top shape: 32 384 7 7 (602112)
I0428 13:50:32.358675 30475 net.cpp:137] Memory required for data: 243707648
I0428 13:50:32.358680 30475 layer_factory.hpp:77] Creating layer conv5
I0428 13:50:32.358692 30475 net.cpp:84] Creating Layer conv5
I0428 13:50:32.358695 30475 net.cpp:406] conv5 <- conv4
I0428 13:50:32.358701 30475 net.cpp:380] conv5 -> conv5
I0428 13:50:32.368257 30475 net.cpp:122] Setting up conv5
I0428 13:50:32.368274 30475 net.cpp:129] Top shape: 32 256 7 7 (401408)
I0428 13:50:32.368278 30475 net.cpp:137] Memory required for data: 245313280
I0428 13:50:32.368285 30475 layer_factory.hpp:77] Creating layer relu5
I0428 13:50:32.368292 30475 net.cpp:84] Creating Layer relu5
I0428 13:50:32.368296 30475 net.cpp:406] relu5 <- conv5
I0428 13:50:32.368301 30475 net.cpp:367] relu5 -> conv5 (in-place)
I0428 13:50:32.368690 30475 net.cpp:122] Setting up relu5
I0428 13:50:32.368700 30475 net.cpp:129] Top shape: 32 256 7 7 (401408)
I0428 13:50:32.368703 30475 net.cpp:137] Memory required for data: 246918912
I0428 13:50:32.368706 30475 layer_factory.hpp:77] Creating layer pool5
I0428 13:50:32.368712 30475 net.cpp:84] Creating Layer pool5
I0428 13:50:32.368716 30475 net.cpp:406] pool5 <- conv5
I0428 13:50:32.368721 30475 net.cpp:380] pool5 -> pool5
I0428 13:50:32.368757 30475 net.cpp:122] Setting up pool5
I0428 13:50:32.368762 30475 net.cpp:129] Top shape: 32 256 3 3 (73728)
I0428 13:50:32.368764 30475 net.cpp:137] Memory required for data: 247213824
I0428 13:50:32.368767 30475 layer_factory.hpp:77] Creating layer fc6
I0428 13:50:32.368774 30475 net.cpp:84] Creating Layer fc6
I0428 13:50:32.368777 30475 net.cpp:406] fc6 <- pool5
I0428 13:50:32.368783 30475 net.cpp:380] fc6 -> fc6
I0428 13:50:32.458639 30475 net.cpp:122] Setting up fc6
I0428 13:50:32.458660 30475 net.cpp:129] Top shape: 32 4096 (131072)
I0428 13:50:32.458664 30475 net.cpp:137] Memory required for data: 247738112
I0428 13:50:32.458693 30475 layer_factory.hpp:77] Creating layer relu6
I0428 13:50:32.458703 30475 net.cpp:84] Creating Layer relu6
I0428 13:50:32.458706 30475 net.cpp:406] relu6 <- fc6
I0428 13:50:32.458714 30475 net.cpp:367] relu6 -> fc6 (in-place)
I0428 13:50:32.459491 30475 net.cpp:122] Setting up relu6
I0428 13:50:32.459501 30475 net.cpp:129] Top shape: 32 4096 (131072)
I0428 13:50:32.459503 30475 net.cpp:137] Memory required for data: 248262400
I0428 13:50:32.459506 30475 layer_factory.hpp:77] Creating layer drop6
I0428 13:50:32.459512 30475 net.cpp:84] Creating Layer drop6
I0428 13:50:32.459515 30475 net.cpp:406] drop6 <- fc6
I0428 13:50:32.459523 30475 net.cpp:367] drop6 -> fc6 (in-place)
I0428 13:50:32.459547 30475 net.cpp:122] Setting up drop6
I0428 13:50:32.459553 30475 net.cpp:129] Top shape: 32 4096 (131072)
I0428 13:50:32.459555 30475 net.cpp:137] Memory required for data: 248786688
I0428 13:50:32.459558 30475 layer_factory.hpp:77] Creating layer fc7
I0428 13:50:32.459563 30475 net.cpp:84] Creating Layer fc7
I0428 13:50:32.459566 30475 net.cpp:406] fc7 <- fc6
I0428 13:50:32.459573 30475 net.cpp:380] fc7 -> fc7
I0428 13:50:32.618871 30475 net.cpp:122] Setting up fc7
I0428 13:50:32.618893 30475 net.cpp:129] Top shape: 32 4096 (131072)
I0428 13:50:32.618896 30475 net.cpp:137] Memory required for data: 249310976
I0428 13:50:32.618906 30475 layer_factory.hpp:77] Creating layer relu7
I0428 13:50:32.618913 30475 net.cpp:84] Creating Layer relu7
I0428 13:50:32.618917 30475 net.cpp:406] relu7 <- fc7
I0428 13:50:32.618924 30475 net.cpp:367] relu7 -> fc7 (in-place)
I0428 13:50:32.619679 30475 net.cpp:122] Setting up relu7
I0428 13:50:32.619690 30475 net.cpp:129] Top shape: 32 4096 (131072)
I0428 13:50:32.619693 30475 net.cpp:137] Memory required for data: 249835264
I0428 13:50:32.619696 30475 layer_factory.hpp:77] Creating layer drop7
I0428 13:50:32.619702 30475 net.cpp:84] Creating Layer drop7
I0428 13:50:32.619705 30475 net.cpp:406] drop7 <- fc7
I0428 13:50:32.619710 30475 net.cpp:367] drop7 -> fc7 (in-place)
I0428 13:50:32.619735 30475 net.cpp:122] Setting up drop7
I0428 13:50:32.619740 30475 net.cpp:129] Top shape: 32 4096 (131072)
I0428 13:50:32.619742 30475 net.cpp:137] Memory required for data: 250359552
I0428 13:50:32.619745 30475 layer_factory.hpp:77] Creating layer fc8
I0428 13:50:32.619752 30475 net.cpp:84] Creating Layer fc8
I0428 13:50:32.619755 30475 net.cpp:406] fc8 <- fc7
I0428 13:50:32.619760 30475 net.cpp:380] fc8 -> fc8
I0428 13:50:32.627487 30475 net.cpp:122] Setting up fc8
I0428 13:50:32.627498 30475 net.cpp:129] Top shape: 32 196 (6272)
I0428 13:50:32.627501 30475 net.cpp:137] Memory required for data: 250384640
I0428 13:50:32.627511 30475 layer_factory.hpp:77] Creating layer fc8_fc8_0_split
I0428 13:50:32.627516 30475 net.cpp:84] Creating Layer fc8_fc8_0_split
I0428 13:50:32.627519 30475 net.cpp:406] fc8_fc8_0_split <- fc8
I0428 13:50:32.627526 30475 net.cpp:380] fc8_fc8_0_split -> fc8_fc8_0_split_0
I0428 13:50:32.627532 30475 net.cpp:380] fc8_fc8_0_split -> fc8_fc8_0_split_1
I0428 13:50:32.627562 30475 net.cpp:122] Setting up fc8_fc8_0_split
I0428 13:50:32.627568 30475 net.cpp:129] Top shape: 32 196 (6272)
I0428 13:50:32.627570 30475 net.cpp:129] Top shape: 32 196 (6272)
I0428 13:50:32.627573 30475 net.cpp:137] Memory required for data: 250434816
I0428 13:50:32.627575 30475 layer_factory.hpp:77] Creating layer accuracy
I0428 13:50:32.627581 30475 net.cpp:84] Creating Layer accuracy
I0428 13:50:32.627583 30475 net.cpp:406] accuracy <- fc8_fc8_0_split_0
I0428 13:50:32.627588 30475 net.cpp:406] accuracy <- label_val-data_1_split_0
I0428 13:50:32.627591 30475 net.cpp:380] accuracy -> accuracy
I0428 13:50:32.627599 30475 net.cpp:122] Setting up accuracy
I0428 13:50:32.627601 30475 net.cpp:129] Top shape: (1)
I0428 13:50:32.627604 30475 net.cpp:137] Memory required for data: 250434820
I0428 13:50:32.627606 30475 layer_factory.hpp:77] Creating layer loss
I0428 13:50:32.627611 30475 net.cpp:84] Creating Layer loss
I0428 13:50:32.627614 30475 net.cpp:406] loss <- fc8_fc8_0_split_1
I0428 13:50:32.627635 30475 net.cpp:406] loss <- label_val-data_1_split_1
I0428 13:50:32.627640 30475 net.cpp:380] loss -> loss
I0428 13:50:32.627646 30475 layer_factory.hpp:77] Creating layer loss
I0428 13:50:32.628293 30475 net.cpp:122] Setting up loss
I0428 13:50:32.628302 30475 net.cpp:129] Top shape: (1)
I0428 13:50:32.628304 30475 net.cpp:132]     with loss weight 1
I0428 13:50:32.628314 30475 net.cpp:137] Memory required for data: 250434824
I0428 13:50:32.628317 30475 net.cpp:198] loss needs backward computation.
I0428 13:50:32.628322 30475 net.cpp:200] accuracy does not need backward computation.
I0428 13:50:32.628325 30475 net.cpp:198] fc8_fc8_0_split needs backward computation.
I0428 13:50:32.628329 30475 net.cpp:198] fc8 needs backward computation.
I0428 13:50:32.628331 30475 net.cpp:198] drop7 needs backward computation.
I0428 13:50:32.628334 30475 net.cpp:198] relu7 needs backward computation.
I0428 13:50:32.628336 30475 net.cpp:198] fc7 needs backward computation.
I0428 13:50:32.628338 30475 net.cpp:198] drop6 needs backward computation.
I0428 13:50:32.628341 30475 net.cpp:198] relu6 needs backward computation.
I0428 13:50:32.628343 30475 net.cpp:198] fc6 needs backward computation.
I0428 13:50:32.628346 30475 net.cpp:198] pool5 needs backward computation.
I0428 13:50:32.628350 30475 net.cpp:198] relu5 needs backward computation.
I0428 13:50:32.628352 30475 net.cpp:198] conv5 needs backward computation.
I0428 13:50:32.628355 30475 net.cpp:198] relu4 needs backward computation.
I0428 13:50:32.628357 30475 net.cpp:198] conv4 needs backward computation.
I0428 13:50:32.628360 30475 net.cpp:198] relu3.5 needs backward computation.
I0428 13:50:32.628363 30475 net.cpp:198] conv3.5 needs backward computation.
I0428 13:50:32.628366 30475 net.cpp:198] relu3 needs backward computation.
I0428 13:50:32.628368 30475 net.cpp:198] conv3 needs backward computation.
I0428 13:50:32.628371 30475 net.cpp:198] pool2 needs backward computation.
I0428 13:50:32.628374 30475 net.cpp:198] norm2 needs backward computation.
I0428 13:50:32.628377 30475 net.cpp:198] relu2 needs backward computation.
I0428 13:50:32.628381 30475 net.cpp:198] conv2 needs backward computation.
I0428 13:50:32.628382 30475 net.cpp:198] pool1 needs backward computation.
I0428 13:50:32.628386 30475 net.cpp:198] norm1 needs backward computation.
I0428 13:50:32.628388 30475 net.cpp:198] relu1 needs backward computation.
I0428 13:50:32.628391 30475 net.cpp:198] conv1 needs backward computation.
I0428 13:50:32.628394 30475 net.cpp:200] label_val-data_1_split does not need backward computation.
I0428 13:50:32.628398 30475 net.cpp:200] val-data does not need backward computation.
I0428 13:50:32.628401 30475 net.cpp:242] This network produces output accuracy
I0428 13:50:32.628405 30475 net.cpp:242] This network produces output loss
I0428 13:50:32.628422 30475 net.cpp:255] Network initialization done.
I0428 13:50:32.628494 30475 solver.cpp:56] Solver scaffolding done.
I0428 13:50:32.628891 30475 caffe.cpp:248] Starting Optimization
I0428 13:50:32.628902 30475 solver.cpp:272] Solving
I0428 13:50:32.628906 30475 solver.cpp:273] Learning Rate Policy: exp
I0428 13:50:32.630568 30475 solver.cpp:330] Iteration 0, Testing net (#0)
I0428 13:50:32.630578 30475 net.cpp:676] Ignoring source layer train-data
I0428 13:50:32.722115 30475 blocking_queue.cpp:49] Waiting for data
I0428 13:50:37.386014 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 13:50:37.437781 30475 solver.cpp:397]     Test net output #0: accuracy = 0.00367647
I0428 13:50:37.437809 30475 solver.cpp:397]     Test net output #1: loss = 5.28391 (* 1 = 5.28391 loss)
I0428 13:50:37.600502 30475 solver.cpp:218] Iteration 0 (0 iter/s, 4.97156s/12 iters), loss = 5.27916
I0428 13:50:37.602013 30475 solver.cpp:237]     Train net output #0: loss = 5.27916 (* 1 = 5.27916 loss)
I0428 13:50:37.602030 30475 sgd_solver.cpp:105] Iteration 0, lr = 0.01
I0428 13:50:42.033695 30475 solver.cpp:218] Iteration 12 (2.70779 iter/s, 4.43166s/12 iters), loss = 5.2884
I0428 13:50:42.033737 30475 solver.cpp:237]     Train net output #0: loss = 5.2884 (* 1 = 5.2884 loss)
I0428 13:50:42.033769 30475 sgd_solver.cpp:105] Iteration 12, lr = 0.00997626
I0428 13:50:47.664503 30475 solver.cpp:218] Iteration 24 (2.13115 iter/s, 5.63075s/12 iters), loss = 5.29751
I0428 13:50:47.664546 30475 solver.cpp:237]     Train net output #0: loss = 5.29751 (* 1 = 5.29751 loss)
I0428 13:50:47.664554 30475 sgd_solver.cpp:105] Iteration 24, lr = 0.00995257
I0428 13:50:53.315459 30475 solver.cpp:218] Iteration 36 (2.12355 iter/s, 5.6509s/12 iters), loss = 5.30878
I0428 13:50:53.315502 30475 solver.cpp:237]     Train net output #0: loss = 5.30878 (* 1 = 5.30878 loss)
I0428 13:50:53.315510 30475 sgd_solver.cpp:105] Iteration 36, lr = 0.00992894
I0428 13:50:58.907531 30475 solver.cpp:218] Iteration 48 (2.14592 iter/s, 5.59202s/12 iters), loss = 5.31643
I0428 13:50:58.907577 30475 solver.cpp:237]     Train net output #0: loss = 5.31643 (* 1 = 5.31643 loss)
I0428 13:50:58.907584 30475 sgd_solver.cpp:105] Iteration 48, lr = 0.00990537
I0428 13:51:04.567687 30475 solver.cpp:218] Iteration 60 (2.1201 iter/s, 5.6601s/12 iters), loss = 5.2787
I0428 13:51:04.567770 30475 solver.cpp:237]     Train net output #0: loss = 5.2787 (* 1 = 5.2787 loss)
I0428 13:51:04.567778 30475 sgd_solver.cpp:105] Iteration 60, lr = 0.00988185
I0428 13:51:10.207355 30475 solver.cpp:218] Iteration 72 (2.12782 iter/s, 5.63958s/12 iters), loss = 5.29372
I0428 13:51:10.207396 30475 solver.cpp:237]     Train net output #0: loss = 5.29372 (* 1 = 5.29372 loss)
I0428 13:51:10.207406 30475 sgd_solver.cpp:105] Iteration 72, lr = 0.00985839
I0428 13:51:15.755107 30475 solver.cpp:218] Iteration 84 (2.16306 iter/s, 5.54771s/12 iters), loss = 5.2866
I0428 13:51:15.755146 30475 solver.cpp:237]     Train net output #0: loss = 5.2866 (* 1 = 5.2866 loss)
I0428 13:51:15.755154 30475 sgd_solver.cpp:105] Iteration 84, lr = 0.00983498
I0428 13:51:21.423348 30475 solver.cpp:218] Iteration 96 (2.11707 iter/s, 5.6682s/12 iters), loss = 5.29528
I0428 13:51:21.423384 30475 solver.cpp:237]     Train net output #0: loss = 5.29528 (* 1 = 5.29528 loss)
I0428 13:51:21.423393 30475 sgd_solver.cpp:105] Iteration 96, lr = 0.00981163
I0428 13:51:23.360157 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 13:51:23.684931 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_102.caffemodel
I0428 13:51:32.727982 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_102.solverstate
I0428 13:51:39.164858 30475 solver.cpp:330] Iteration 102, Testing net (#0)
I0428 13:51:39.164947 30475 net.cpp:676] Ignoring source layer train-data
I0428 13:51:44.067075 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 13:51:44.160976 30475 solver.cpp:397]     Test net output #0: accuracy = 0.00551471
I0428 13:51:44.161012 30475 solver.cpp:397]     Test net output #1: loss = 5.29023 (* 1 = 5.29023 loss)
I0428 13:51:46.299455 30475 solver.cpp:218] Iteration 108 (0.482391 iter/s, 24.8761s/12 iters), loss = 5.3027
I0428 13:51:46.299504 30475 solver.cpp:237]     Train net output #0: loss = 5.3027 (* 1 = 5.3027 loss)
I0428 13:51:46.299511 30475 sgd_solver.cpp:105] Iteration 108, lr = 0.00978834
I0428 13:51:51.927512 30475 solver.cpp:218] Iteration 120 (2.1322 iter/s, 5.62799s/12 iters), loss = 5.28149
I0428 13:51:51.927561 30475 solver.cpp:237]     Train net output #0: loss = 5.28149 (* 1 = 5.28149 loss)
I0428 13:51:51.927569 30475 sgd_solver.cpp:105] Iteration 120, lr = 0.0097651
I0428 13:51:57.552453 30475 solver.cpp:218] Iteration 132 (2.13338 iter/s, 5.62488s/12 iters), loss = 5.27879
I0428 13:51:57.552503 30475 solver.cpp:237]     Train net output #0: loss = 5.27879 (* 1 = 5.27879 loss)
I0428 13:51:57.552513 30475 sgd_solver.cpp:105] Iteration 132, lr = 0.00974192
I0428 13:52:03.096251 30475 solver.cpp:218] Iteration 144 (2.1646 iter/s, 5.54374s/12 iters), loss = 5.29808
I0428 13:52:03.096297 30475 solver.cpp:237]     Train net output #0: loss = 5.29808 (* 1 = 5.29808 loss)
I0428 13:52:03.096305 30475 sgd_solver.cpp:105] Iteration 144, lr = 0.00971879
I0428 13:52:08.729060 30475 solver.cpp:218] Iteration 156 (2.1304 iter/s, 5.63275s/12 iters), loss = 5.29452
I0428 13:52:08.729099 30475 solver.cpp:237]     Train net output #0: loss = 5.29452 (* 1 = 5.29452 loss)
I0428 13:52:08.729108 30475 sgd_solver.cpp:105] Iteration 156, lr = 0.00969571
I0428 13:52:14.357404 30475 solver.cpp:218] Iteration 168 (2.13208 iter/s, 5.6283s/12 iters), loss = 5.29734
I0428 13:52:14.357564 30475 solver.cpp:237]     Train net output #0: loss = 5.29734 (* 1 = 5.29734 loss)
I0428 13:52:14.357573 30475 sgd_solver.cpp:105] Iteration 168, lr = 0.00967269
I0428 13:52:20.008169 30475 solver.cpp:218] Iteration 180 (2.12367 iter/s, 5.6506s/12 iters), loss = 5.29555
I0428 13:52:20.008213 30475 solver.cpp:237]     Train net output #0: loss = 5.29555 (* 1 = 5.29555 loss)
I0428 13:52:20.008220 30475 sgd_solver.cpp:105] Iteration 180, lr = 0.00964973
I0428 13:52:25.671630 30475 solver.cpp:218] Iteration 192 (2.11886 iter/s, 5.66342s/12 iters), loss = 5.26417
I0428 13:52:25.671666 30475 solver.cpp:237]     Train net output #0: loss = 5.26417 (* 1 = 5.26417 loss)
I0428 13:52:25.671674 30475 sgd_solver.cpp:105] Iteration 192, lr = 0.00962682
I0428 13:52:30.047729 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 13:52:30.792158 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_204.caffemodel
I0428 13:52:35.591545 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_204.solverstate
I0428 13:52:39.722892 30475 solver.cpp:330] Iteration 204, Testing net (#0)
I0428 13:52:39.722910 30475 net.cpp:676] Ignoring source layer train-data
I0428 13:52:44.707180 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 13:52:44.858592 30475 solver.cpp:397]     Test net output #0: accuracy = 0.00857843
I0428 13:52:44.858644 30475 solver.cpp:397]     Test net output #1: loss = 5.20838 (* 1 = 5.20838 loss)
I0428 13:52:45.020311 30475 solver.cpp:218] Iteration 204 (0.620198 iter/s, 19.3487s/12 iters), loss = 5.23895
I0428 13:52:45.020357 30475 solver.cpp:237]     Train net output #0: loss = 5.23895 (* 1 = 5.23895 loss)
I0428 13:52:45.020366 30475 sgd_solver.cpp:105] Iteration 204, lr = 0.00960396
I0428 13:52:49.725901 30475 solver.cpp:218] Iteration 216 (2.55019 iter/s, 4.70553s/12 iters), loss = 5.22144
I0428 13:52:49.725944 30475 solver.cpp:237]     Train net output #0: loss = 5.22144 (* 1 = 5.22144 loss)
I0428 13:52:49.725953 30475 sgd_solver.cpp:105] Iteration 216, lr = 0.00958116
I0428 13:52:55.292248 30475 solver.cpp:218] Iteration 228 (2.15583 iter/s, 5.56629s/12 iters), loss = 5.21448
I0428 13:52:55.292286 30475 solver.cpp:237]     Train net output #0: loss = 5.21448 (* 1 = 5.21448 loss)
I0428 13:52:55.292294 30475 sgd_solver.cpp:105] Iteration 228, lr = 0.00955841
I0428 13:53:00.922482 30475 solver.cpp:218] Iteration 240 (2.13137 iter/s, 5.63019s/12 iters), loss = 5.10576
I0428 13:53:00.922526 30475 solver.cpp:237]     Train net output #0: loss = 5.10576 (* 1 = 5.10576 loss)
I0428 13:53:00.922535 30475 sgd_solver.cpp:105] Iteration 240, lr = 0.00953572
I0428 13:53:06.582753 30475 solver.cpp:218] Iteration 252 (2.12006 iter/s, 5.66021s/12 iters), loss = 5.14787
I0428 13:53:06.582815 30475 solver.cpp:237]     Train net output #0: loss = 5.14787 (* 1 = 5.14787 loss)
I0428 13:53:06.582829 30475 sgd_solver.cpp:105] Iteration 252, lr = 0.00951308
I0428 13:53:12.233088 30475 solver.cpp:218] Iteration 264 (2.12379 iter/s, 5.65027s/12 iters), loss = 5.16817
I0428 13:53:12.233124 30475 solver.cpp:237]     Train net output #0: loss = 5.16817 (* 1 = 5.16817 loss)
I0428 13:53:12.233134 30475 sgd_solver.cpp:105] Iteration 264, lr = 0.00949049
I0428 13:53:17.878846 30475 solver.cpp:218] Iteration 276 (2.12554 iter/s, 5.64562s/12 iters), loss = 5.23431
I0428 13:53:17.878978 30475 solver.cpp:237]     Train net output #0: loss = 5.23431 (* 1 = 5.23431 loss)
I0428 13:53:17.878988 30475 sgd_solver.cpp:105] Iteration 276, lr = 0.00946796
I0428 13:53:23.425081 30475 solver.cpp:218] Iteration 288 (2.16368 iter/s, 5.54609s/12 iters), loss = 5.197
I0428 13:53:23.425125 30475 solver.cpp:237]     Train net output #0: loss = 5.197 (* 1 = 5.197 loss)
I0428 13:53:23.425134 30475 sgd_solver.cpp:105] Iteration 288, lr = 0.00944548
I0428 13:53:28.995436 30475 solver.cpp:218] Iteration 300 (2.15428 iter/s, 5.5703s/12 iters), loss = 5.15434
I0428 13:53:28.995483 30475 solver.cpp:237]     Train net output #0: loss = 5.15434 (* 1 = 5.15434 loss)
I0428 13:53:28.995492 30475 sgd_solver.cpp:105] Iteration 300, lr = 0.00942305
I0428 13:53:30.108307 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 13:53:31.313571 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_306.caffemodel
I0428 13:53:34.024454 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_306.solverstate
I0428 13:53:35.734288 30475 solver.cpp:330] Iteration 306, Testing net (#0)
I0428 13:53:35.734315 30475 net.cpp:676] Ignoring source layer train-data
I0428 13:53:40.649081 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 13:53:40.835680 30475 solver.cpp:397]     Test net output #0: accuracy = 0.00857843
I0428 13:53:40.835707 30475 solver.cpp:397]     Test net output #1: loss = 5.16696 (* 1 = 5.16696 loss)
I0428 13:53:42.987705 30475 solver.cpp:218] Iteration 312 (0.857619 iter/s, 13.9922s/12 iters), loss = 5.20381
I0428 13:53:42.987751 30475 solver.cpp:237]     Train net output #0: loss = 5.20381 (* 1 = 5.20381 loss)
I0428 13:53:42.987761 30475 sgd_solver.cpp:105] Iteration 312, lr = 0.00940068
I0428 13:53:48.630005 30475 solver.cpp:218] Iteration 324 (2.12681 iter/s, 5.64225s/12 iters), loss = 5.16282
I0428 13:53:48.630067 30475 solver.cpp:237]     Train net output #0: loss = 5.16282 (* 1 = 5.16282 loss)
I0428 13:53:48.630076 30475 sgd_solver.cpp:105] Iteration 324, lr = 0.00937836
I0428 13:53:54.318727 30475 solver.cpp:218] Iteration 336 (2.10946 iter/s, 5.68865s/12 iters), loss = 5.09569
I0428 13:53:54.318775 30475 solver.cpp:237]     Train net output #0: loss = 5.09569 (* 1 = 5.09569 loss)
I0428 13:53:54.318783 30475 sgd_solver.cpp:105] Iteration 336, lr = 0.0093561
I0428 13:54:00.068447 30475 solver.cpp:218] Iteration 348 (2.08708 iter/s, 5.74967s/12 iters), loss = 5.16975
I0428 13:54:00.068485 30475 solver.cpp:237]     Train net output #0: loss = 5.16975 (* 1 = 5.16975 loss)
I0428 13:54:00.068492 30475 sgd_solver.cpp:105] Iteration 348, lr = 0.00933388
I0428 13:54:05.710755 30475 solver.cpp:218] Iteration 360 (2.12681 iter/s, 5.64226s/12 iters), loss = 5.12954
I0428 13:54:05.710801 30475 solver.cpp:237]     Train net output #0: loss = 5.12954 (* 1 = 5.12954 loss)
I0428 13:54:05.710809 30475 sgd_solver.cpp:105] Iteration 360, lr = 0.00931172
I0428 13:54:11.358386 30475 solver.cpp:218] Iteration 372 (2.1248 iter/s, 5.64758s/12 iters), loss = 5.12653
I0428 13:54:11.358425 30475 solver.cpp:237]     Train net output #0: loss = 5.12653 (* 1 = 5.12653 loss)
I0428 13:54:11.358433 30475 sgd_solver.cpp:105] Iteration 372, lr = 0.00928961
I0428 13:54:16.959683 30475 solver.cpp:218] Iteration 384 (2.14238 iter/s, 5.60125s/12 iters), loss = 5.1659
I0428 13:54:16.959723 30475 solver.cpp:237]     Train net output #0: loss = 5.1659 (* 1 = 5.1659 loss)
I0428 13:54:16.959731 30475 sgd_solver.cpp:105] Iteration 384, lr = 0.00926756
I0428 13:54:22.450481 30475 solver.cpp:218] Iteration 396 (2.18549 iter/s, 5.49075s/12 iters), loss = 5.18915
I0428 13:54:22.450582 30475 solver.cpp:237]     Train net output #0: loss = 5.18915 (* 1 = 5.18915 loss)
I0428 13:54:22.450592 30475 sgd_solver.cpp:105] Iteration 396, lr = 0.00924556
I0428 13:54:25.985378 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 13:54:27.580430 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_408.caffemodel
I0428 13:54:31.628453 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_408.solverstate
I0428 13:54:33.379047 30475 solver.cpp:330] Iteration 408, Testing net (#0)
I0428 13:54:33.379065 30475 net.cpp:676] Ignoring source layer train-data
I0428 13:54:38.278023 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 13:54:38.524130 30475 solver.cpp:397]     Test net output #0: accuracy = 0.0110294
I0428 13:54:38.524163 30475 solver.cpp:397]     Test net output #1: loss = 5.14659 (* 1 = 5.14659 loss)
I0428 13:54:38.685390 30475 solver.cpp:218] Iteration 408 (0.739152 iter/s, 16.2348s/12 iters), loss = 5.11792
I0428 13:54:38.685446 30475 solver.cpp:237]     Train net output #0: loss = 5.11792 (* 1 = 5.11792 loss)
I0428 13:54:38.685456 30475 sgd_solver.cpp:105] Iteration 408, lr = 0.00922361
I0428 13:54:43.473865 30475 solver.cpp:218] Iteration 420 (2.50605 iter/s, 4.78841s/12 iters), loss = 5.11905
I0428 13:54:43.473910 30475 solver.cpp:237]     Train net output #0: loss = 5.11905 (* 1 = 5.11905 loss)
I0428 13:54:43.473918 30475 sgd_solver.cpp:105] Iteration 420, lr = 0.00920171
I0428 13:54:49.141443 30475 solver.cpp:218] Iteration 432 (2.11733 iter/s, 5.66752s/12 iters), loss = 5.12302
I0428 13:54:49.141497 30475 solver.cpp:237]     Train net output #0: loss = 5.12302 (* 1 = 5.12302 loss)
I0428 13:54:49.141512 30475 sgd_solver.cpp:105] Iteration 432, lr = 0.00917986
I0428 13:54:54.800514 30475 solver.cpp:218] Iteration 444 (2.12051 iter/s, 5.65902s/12 iters), loss = 5.15946
I0428 13:54:54.800662 30475 solver.cpp:237]     Train net output #0: loss = 5.15946 (* 1 = 5.15946 loss)
I0428 13:54:54.800671 30475 sgd_solver.cpp:105] Iteration 444, lr = 0.00915807
I0428 13:55:00.352258 30475 solver.cpp:218] Iteration 456 (2.16154 iter/s, 5.55159s/12 iters), loss = 5.11505
I0428 13:55:00.352303 30475 solver.cpp:237]     Train net output #0: loss = 5.11505 (* 1 = 5.11505 loss)
I0428 13:55:00.352313 30475 sgd_solver.cpp:105] Iteration 456, lr = 0.00913632
I0428 13:55:06.015220 30475 solver.cpp:218] Iteration 468 (2.11905 iter/s, 5.66291s/12 iters), loss = 5.12209
I0428 13:55:06.015260 30475 solver.cpp:237]     Train net output #0: loss = 5.12209 (* 1 = 5.12209 loss)
I0428 13:55:06.015269 30475 sgd_solver.cpp:105] Iteration 468, lr = 0.00911463
I0428 13:55:11.667688 30475 solver.cpp:218] Iteration 480 (2.12298 iter/s, 5.65242s/12 iters), loss = 5.0796
I0428 13:55:11.667735 30475 solver.cpp:237]     Train net output #0: loss = 5.0796 (* 1 = 5.0796 loss)
I0428 13:55:11.667743 30475 sgd_solver.cpp:105] Iteration 480, lr = 0.00909299
I0428 13:55:17.240926 30475 solver.cpp:218] Iteration 492 (2.15317 iter/s, 5.57318s/12 iters), loss = 5.09225
I0428 13:55:17.240964 30475 solver.cpp:237]     Train net output #0: loss = 5.09225 (* 1 = 5.09225 loss)
I0428 13:55:17.240973 30475 sgd_solver.cpp:105] Iteration 492, lr = 0.0090714
I0428 13:55:22.820681 30475 solver.cpp:218] Iteration 504 (2.15065 iter/s, 5.57971s/12 iters), loss = 5.10075
I0428 13:55:22.820724 30475 solver.cpp:237]     Train net output #0: loss = 5.10075 (* 1 = 5.10075 loss)
I0428 13:55:22.820734 30475 sgd_solver.cpp:105] Iteration 504, lr = 0.00904986
I0428 13:55:23.067625 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 13:55:25.079035 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_510.caffemodel
I0428 13:55:29.376660 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_510.solverstate
I0428 13:55:32.005380 30475 solver.cpp:330] Iteration 510, Testing net (#0)
I0428 13:55:32.005403 30475 net.cpp:676] Ignoring source layer train-data
I0428 13:55:36.899276 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 13:55:37.182487 30475 solver.cpp:397]     Test net output #0: accuracy = 0.0104167
I0428 13:55:37.182521 30475 solver.cpp:397]     Test net output #1: loss = 5.10721 (* 1 = 5.10721 loss)
I0428 13:55:39.307811 30475 solver.cpp:218] Iteration 516 (0.727842 iter/s, 16.4871s/12 iters), loss = 5.10821
I0428 13:55:39.307855 30475 solver.cpp:237]     Train net output #0: loss = 5.10821 (* 1 = 5.10821 loss)
I0428 13:55:39.307863 30475 sgd_solver.cpp:105] Iteration 516, lr = 0.00902838
I0428 13:55:44.959540 30475 solver.cpp:218] Iteration 528 (2.12326 iter/s, 5.65168s/12 iters), loss = 5.10914
I0428 13:55:44.959579 30475 solver.cpp:237]     Train net output #0: loss = 5.10914 (* 1 = 5.10914 loss)
I0428 13:55:44.959587 30475 sgd_solver.cpp:105] Iteration 528, lr = 0.00900694
I0428 13:55:50.609437 30475 solver.cpp:218] Iteration 540 (2.12395 iter/s, 5.64985s/12 iters), loss = 5.15491
I0428 13:55:50.609483 30475 solver.cpp:237]     Train net output #0: loss = 5.15491 (* 1 = 5.15491 loss)
I0428 13:55:50.609493 30475 sgd_solver.cpp:105] Iteration 540, lr = 0.00898556
I0428 13:55:56.261286 30475 solver.cpp:218] Iteration 552 (2.12322 iter/s, 5.65179s/12 iters), loss = 5.03198
I0428 13:55:56.261437 30475 solver.cpp:237]     Train net output #0: loss = 5.03198 (* 1 = 5.03198 loss)
I0428 13:55:56.261447 30475 sgd_solver.cpp:105] Iteration 552, lr = 0.00896423
I0428 13:56:01.909600 30475 solver.cpp:218] Iteration 564 (2.12459 iter/s, 5.64816s/12 iters), loss = 5.08766
I0428 13:56:01.909641 30475 solver.cpp:237]     Train net output #0: loss = 5.08766 (* 1 = 5.08766 loss)
I0428 13:56:01.909648 30475 sgd_solver.cpp:105] Iteration 564, lr = 0.00894294
I0428 13:56:07.558849 30475 solver.cpp:218] Iteration 576 (2.12419 iter/s, 5.6492s/12 iters), loss = 5.1013
I0428 13:56:07.558892 30475 solver.cpp:237]     Train net output #0: loss = 5.1013 (* 1 = 5.1013 loss)
I0428 13:56:07.558899 30475 sgd_solver.cpp:105] Iteration 576, lr = 0.00892171
I0428 13:56:13.138530 30475 solver.cpp:218] Iteration 588 (2.15068 iter/s, 5.57963s/12 iters), loss = 5.05994
I0428 13:56:13.138572 30475 solver.cpp:237]     Train net output #0: loss = 5.05994 (* 1 = 5.05994 loss)
I0428 13:56:13.138581 30475 sgd_solver.cpp:105] Iteration 588, lr = 0.00890053
I0428 13:56:18.807941 30475 solver.cpp:218] Iteration 600 (2.11664 iter/s, 5.66936s/12 iters), loss = 5.06077
I0428 13:56:18.807979 30475 solver.cpp:237]     Train net output #0: loss = 5.06077 (* 1 = 5.06077 loss)
I0428 13:56:18.807987 30475 sgd_solver.cpp:105] Iteration 600, lr = 0.0088794
I0428 13:56:21.473711 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 13:56:23.912153 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_612.caffemodel
I0428 13:56:26.091022 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_612.solverstate
I0428 13:56:27.784711 30475 solver.cpp:330] Iteration 612, Testing net (#0)
I0428 13:56:27.784826 30475 net.cpp:676] Ignoring source layer train-data
I0428 13:56:32.541131 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 13:56:32.877563 30475 solver.cpp:397]     Test net output #0: accuracy = 0.0177696
I0428 13:56:32.877609 30475 solver.cpp:397]     Test net output #1: loss = 5.05112 (* 1 = 5.05112 loss)
I0428 13:56:33.032261 30475 solver.cpp:218] Iteration 612 (0.843627 iter/s, 14.2243s/12 iters), loss = 5.08647
I0428 13:56:33.033809 30475 solver.cpp:237]     Train net output #0: loss = 5.08647 (* 1 = 5.08647 loss)
I0428 13:56:33.033823 30475 sgd_solver.cpp:105] Iteration 612, lr = 0.00885831
I0428 13:56:37.725644 30475 solver.cpp:218] Iteration 624 (2.55764 iter/s, 4.69182s/12 iters), loss = 5.032
I0428 13:56:37.725713 30475 solver.cpp:237]     Train net output #0: loss = 5.032 (* 1 = 5.032 loss)
I0428 13:56:37.725726 30475 sgd_solver.cpp:105] Iteration 624, lr = 0.00883728
I0428 13:56:43.374437 30475 solver.cpp:218] Iteration 636 (2.12438 iter/s, 5.64872s/12 iters), loss = 5.02704
I0428 13:56:43.374483 30475 solver.cpp:237]     Train net output #0: loss = 5.02704 (* 1 = 5.02704 loss)
I0428 13:56:43.374491 30475 sgd_solver.cpp:105] Iteration 636, lr = 0.0088163
I0428 13:56:49.034399 30475 solver.cpp:218] Iteration 648 (2.12018 iter/s, 5.65991s/12 iters), loss = 5.05211
I0428 13:56:49.034440 30475 solver.cpp:237]     Train net output #0: loss = 5.05211 (* 1 = 5.05211 loss)
I0428 13:56:49.034447 30475 sgd_solver.cpp:105] Iteration 648, lr = 0.00879537
I0428 13:56:54.664204 30475 solver.cpp:218] Iteration 660 (2.13153 iter/s, 5.62976s/12 iters), loss = 5.03588
I0428 13:56:54.664244 30475 solver.cpp:237]     Train net output #0: loss = 5.03588 (* 1 = 5.03588 loss)
I0428 13:56:54.664252 30475 sgd_solver.cpp:105] Iteration 660, lr = 0.00877449
I0428 13:57:00.343564 30475 solver.cpp:218] Iteration 672 (2.11293 iter/s, 5.67931s/12 iters), loss = 5.05965
I0428 13:57:00.343677 30475 solver.cpp:237]     Train net output #0: loss = 5.05965 (* 1 = 5.05965 loss)
I0428 13:57:00.343685 30475 sgd_solver.cpp:105] Iteration 672, lr = 0.00875366
I0428 13:57:06.176483 30475 solver.cpp:218] Iteration 684 (2.05733 iter/s, 5.8328s/12 iters), loss = 5.09388
I0428 13:57:06.176528 30475 solver.cpp:237]     Train net output #0: loss = 5.09388 (* 1 = 5.09388 loss)
I0428 13:57:06.176537 30475 sgd_solver.cpp:105] Iteration 684, lr = 0.00873287
I0428 13:57:07.078431 30475 blocking_queue.cpp:49] Waiting for data
I0428 13:57:12.023353 30475 solver.cpp:218] Iteration 696 (2.0524 iter/s, 5.84682s/12 iters), loss = 5.01412
I0428 13:57:12.023386 30475 solver.cpp:237]     Train net output #0: loss = 5.01412 (* 1 = 5.01412 loss)
I0428 13:57:12.023396 30475 sgd_solver.cpp:105] Iteration 696, lr = 0.00871214
I0428 13:57:17.313989 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 13:57:17.752702 30475 solver.cpp:218] Iteration 708 (2.09449 iter/s, 5.72931s/12 iters), loss = 4.95756
I0428 13:57:17.752743 30475 solver.cpp:237]     Train net output #0: loss = 4.95756 (* 1 = 4.95756 loss)
I0428 13:57:17.752749 30475 sgd_solver.cpp:105] Iteration 708, lr = 0.00869145
I0428 13:57:20.017194 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_714.caffemodel
I0428 13:57:23.471415 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_714.solverstate
I0428 13:57:26.846135 30475 solver.cpp:330] Iteration 714, Testing net (#0)
I0428 13:57:26.846164 30475 net.cpp:676] Ignoring source layer train-data
I0428 13:57:31.630331 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 13:57:32.017324 30475 solver.cpp:397]     Test net output #0: accuracy = 0.0159314
I0428 13:57:32.017366 30475 solver.cpp:397]     Test net output #1: loss = 4.99965 (* 1 = 4.99965 loss)
I0428 13:57:34.104157 30475 solver.cpp:218] Iteration 720 (0.733881 iter/s, 16.3514s/12 iters), loss = 4.94759
I0428 13:57:34.104212 30475 solver.cpp:237]     Train net output #0: loss = 4.94759 (* 1 = 4.94759 loss)
I0428 13:57:34.104225 30475 sgd_solver.cpp:105] Iteration 720, lr = 0.00867082
I0428 13:57:39.738051 30475 solver.cpp:218] Iteration 732 (2.12999 iter/s, 5.63383s/12 iters), loss = 4.98743
I0428 13:57:39.738095 30475 solver.cpp:237]     Train net output #0: loss = 4.98743 (* 1 = 4.98743 loss)
I0428 13:57:39.738104 30475 sgd_solver.cpp:105] Iteration 732, lr = 0.00865023
I0428 13:57:45.395651 30475 solver.cpp:218] Iteration 744 (2.12106 iter/s, 5.65754s/12 iters), loss = 4.88565
I0428 13:57:45.395711 30475 solver.cpp:237]     Train net output #0: loss = 4.88565 (* 1 = 4.88565 loss)
I0428 13:57:45.395725 30475 sgd_solver.cpp:105] Iteration 744, lr = 0.0086297
I0428 13:57:51.060354 30475 solver.cpp:218] Iteration 756 (2.1184 iter/s, 5.66464s/12 iters), loss = 4.97743
I0428 13:57:51.060400 30475 solver.cpp:237]     Train net output #0: loss = 4.97743 (* 1 = 4.97743 loss)
I0428 13:57:51.060410 30475 sgd_solver.cpp:105] Iteration 756, lr = 0.00860921
I0428 13:57:56.753118 30475 solver.cpp:218] Iteration 768 (2.10796 iter/s, 5.6927s/12 iters), loss = 4.81901
I0428 13:57:56.753165 30475 solver.cpp:237]     Train net output #0: loss = 4.81901 (* 1 = 4.81901 loss)
I0428 13:57:56.753172 30475 sgd_solver.cpp:105] Iteration 768, lr = 0.00858877
I0428 13:58:02.447806 30475 solver.cpp:218] Iteration 780 (2.10725 iter/s, 5.69463s/12 iters), loss = 4.97308
I0428 13:58:02.447897 30475 solver.cpp:237]     Train net output #0: loss = 4.97308 (* 1 = 4.97308 loss)
I0428 13:58:02.447907 30475 sgd_solver.cpp:105] Iteration 780, lr = 0.00856838
I0428 13:58:08.038992 30475 solver.cpp:218] Iteration 792 (2.14627 iter/s, 5.59109s/12 iters), loss = 4.97614
I0428 13:58:08.039036 30475 solver.cpp:237]     Train net output #0: loss = 4.97614 (* 1 = 4.97614 loss)
I0428 13:58:08.039045 30475 sgd_solver.cpp:105] Iteration 792, lr = 0.00854803
I0428 13:58:13.711478 30475 solver.cpp:218] Iteration 804 (2.11549 iter/s, 5.67243s/12 iters), loss = 4.9891
I0428 13:58:13.711521 30475 solver.cpp:237]     Train net output #0: loss = 4.9891 (* 1 = 4.9891 loss)
I0428 13:58:13.711530 30475 sgd_solver.cpp:105] Iteration 804, lr = 0.00852774
I0428 13:58:15.684684 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 13:58:18.822986 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_816.caffemodel
I0428 13:58:21.512404 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_816.solverstate
I0428 13:58:23.327736 30475 solver.cpp:330] Iteration 816, Testing net (#0)
I0428 13:58:23.327755 30475 net.cpp:676] Ignoring source layer train-data
I0428 13:58:28.056617 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 13:58:28.478341 30475 solver.cpp:397]     Test net output #0: accuracy = 0.0226716
I0428 13:58:28.478377 30475 solver.cpp:397]     Test net output #1: loss = 4.90149 (* 1 = 4.90149 loss)
I0428 13:58:28.640224 30475 solver.cpp:218] Iteration 816 (0.80382 iter/s, 14.9287s/12 iters), loss = 4.95406
I0428 13:58:28.640270 30475 solver.cpp:237]     Train net output #0: loss = 4.95406 (* 1 = 4.95406 loss)
I0428 13:58:28.640278 30475 sgd_solver.cpp:105] Iteration 816, lr = 0.00850749
I0428 13:58:33.147513 30475 solver.cpp:218] Iteration 828 (2.66239 iter/s, 4.50723s/12 iters), loss = 4.90035
I0428 13:58:33.147653 30475 solver.cpp:237]     Train net output #0: loss = 4.90035 (* 1 = 4.90035 loss)
I0428 13:58:33.147662 30475 sgd_solver.cpp:105] Iteration 828, lr = 0.00848729
I0428 13:58:38.712817 30475 solver.cpp:218] Iteration 840 (2.15627 iter/s, 5.56516s/12 iters), loss = 4.87204
I0428 13:58:38.712857 30475 solver.cpp:237]     Train net output #0: loss = 4.87204 (* 1 = 4.87204 loss)
I0428 13:58:38.712865 30475 sgd_solver.cpp:105] Iteration 840, lr = 0.00846714
I0428 13:58:44.416973 30475 solver.cpp:218] Iteration 852 (2.10375 iter/s, 5.70411s/12 iters), loss = 4.95871
I0428 13:58:44.417021 30475 solver.cpp:237]     Train net output #0: loss = 4.95871 (* 1 = 4.95871 loss)
I0428 13:58:44.417029 30475 sgd_solver.cpp:105] Iteration 852, lr = 0.00844704
I0428 13:58:50.058578 30475 solver.cpp:218] Iteration 864 (2.12707 iter/s, 5.64155s/12 iters), loss = 4.84873
I0428 13:58:50.058624 30475 solver.cpp:237]     Train net output #0: loss = 4.84873 (* 1 = 4.84873 loss)
I0428 13:58:50.058631 30475 sgd_solver.cpp:105] Iteration 864, lr = 0.00842698
I0428 13:58:55.798848 30475 solver.cpp:218] Iteration 876 (2.09051 iter/s, 5.74022s/12 iters), loss = 4.97362
I0428 13:58:55.798895 30475 solver.cpp:237]     Train net output #0: loss = 4.97362 (* 1 = 4.97362 loss)
I0428 13:58:55.798905 30475 sgd_solver.cpp:105] Iteration 876, lr = 0.00840698
I0428 13:59:01.655787 30475 solver.cpp:218] Iteration 888 (2.04887 iter/s, 5.85689s/12 iters), loss = 4.82349
I0428 13:59:01.655831 30475 solver.cpp:237]     Train net output #0: loss = 4.82349 (* 1 = 4.82349 loss)
I0428 13:59:01.655839 30475 sgd_solver.cpp:105] Iteration 888, lr = 0.00838702
I0428 13:59:07.373596 30475 solver.cpp:218] Iteration 900 (2.09873 iter/s, 5.71775s/12 iters), loss = 4.79362
I0428 13:59:07.373708 30475 solver.cpp:237]     Train net output #0: loss = 4.79362 (* 1 = 4.79362 loss)
I0428 13:59:07.373723 30475 sgd_solver.cpp:105] Iteration 900, lr = 0.0083671
I0428 13:59:11.793323 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 13:59:13.083735 30475 solver.cpp:218] Iteration 912 (2.10157 iter/s, 5.71003s/12 iters), loss = 4.78333
I0428 13:59:13.083778 30475 solver.cpp:237]     Train net output #0: loss = 4.78333 (* 1 = 4.78333 loss)
I0428 13:59:13.083786 30475 sgd_solver.cpp:105] Iteration 912, lr = 0.00834724
I0428 13:59:15.366273 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_918.caffemodel
I0428 13:59:19.843119 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_918.solverstate
I0428 13:59:24.091316 30475 solver.cpp:330] Iteration 918, Testing net (#0)
I0428 13:59:24.091336 30475 net.cpp:676] Ignoring source layer train-data
I0428 13:59:28.757354 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 13:59:29.232435 30475 solver.cpp:397]     Test net output #0: accuracy = 0.0257353
I0428 13:59:29.232465 30475 solver.cpp:397]     Test net output #1: loss = 4.82069 (* 1 = 4.82069 loss)
I0428 13:59:31.346675 30475 solver.cpp:218] Iteration 924 (0.657069 iter/s, 18.2629s/12 iters), loss = 4.8255
I0428 13:59:31.346721 30475 solver.cpp:237]     Train net output #0: loss = 4.8255 (* 1 = 4.8255 loss)
I0428 13:59:31.346729 30475 sgd_solver.cpp:105] Iteration 924, lr = 0.00832742
I0428 13:59:36.983520 30475 solver.cpp:218] Iteration 936 (2.12887 iter/s, 5.63679s/12 iters), loss = 4.83723
I0428 13:59:36.983564 30475 solver.cpp:237]     Train net output #0: loss = 4.83723 (* 1 = 4.83723 loss)
I0428 13:59:36.983572 30475 sgd_solver.cpp:105] Iteration 936, lr = 0.00830765
I0428 13:59:42.658809 30475 solver.cpp:218] Iteration 948 (2.11445 iter/s, 5.67524s/12 iters), loss = 4.82519
I0428 13:59:42.658962 30475 solver.cpp:237]     Train net output #0: loss = 4.82519 (* 1 = 4.82519 loss)
I0428 13:59:42.658972 30475 sgd_solver.cpp:105] Iteration 948, lr = 0.00828793
I0428 13:59:48.292244 30475 solver.cpp:218] Iteration 960 (2.1302 iter/s, 5.63328s/12 iters), loss = 4.71014
I0428 13:59:48.292289 30475 solver.cpp:237]     Train net output #0: loss = 4.71014 (* 1 = 4.71014 loss)
I0428 13:59:48.292297 30475 sgd_solver.cpp:105] Iteration 960, lr = 0.00826825
I0428 13:59:54.102466 30475 solver.cpp:218] Iteration 972 (2.06534 iter/s, 5.81017s/12 iters), loss = 4.71073
I0428 13:59:54.102509 30475 solver.cpp:237]     Train net output #0: loss = 4.71073 (* 1 = 4.71073 loss)
I0428 13:59:54.102517 30475 sgd_solver.cpp:105] Iteration 972, lr = 0.00824862
I0428 13:59:59.749714 30475 solver.cpp:218] Iteration 984 (2.12495 iter/s, 5.6472s/12 iters), loss = 4.57517
I0428 13:59:59.749755 30475 solver.cpp:237]     Train net output #0: loss = 4.57517 (* 1 = 4.57517 loss)
I0428 13:59:59.749763 30475 sgd_solver.cpp:105] Iteration 984, lr = 0.00822903
I0428 14:00:05.224483 30475 solver.cpp:218] Iteration 996 (2.19189 iter/s, 5.47472s/12 iters), loss = 4.7035
I0428 14:00:05.224521 30475 solver.cpp:237]     Train net output #0: loss = 4.7035 (* 1 = 4.7035 loss)
I0428 14:00:05.224529 30475 sgd_solver.cpp:105] Iteration 996, lr = 0.0082095
I0428 14:00:10.885596 30475 solver.cpp:218] Iteration 1008 (2.11974 iter/s, 5.66107s/12 iters), loss = 4.60675
I0428 14:00:10.885632 30475 solver.cpp:237]     Train net output #0: loss = 4.60675 (* 1 = 4.60675 loss)
I0428 14:00:10.885640 30475 sgd_solver.cpp:105] Iteration 1008, lr = 0.00819001
I0428 14:00:12.012233 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:00:15.896728 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1020.caffemodel
I0428 14:00:21.049700 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1020.solverstate
I0428 14:00:25.318666 30475 solver.cpp:330] Iteration 1020, Testing net (#0)
I0428 14:00:25.318686 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:00:29.923182 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:00:30.435190 30475 solver.cpp:397]     Test net output #0: accuracy = 0.0410539
I0428 14:00:30.435225 30475 solver.cpp:397]     Test net output #1: loss = 4.61472 (* 1 = 4.61472 loss)
I0428 14:00:30.596305 30475 solver.cpp:218] Iteration 1020 (0.608806 iter/s, 19.7107s/12 iters), loss = 4.64068
I0428 14:00:30.596346 30475 solver.cpp:237]     Train net output #0: loss = 4.64068 (* 1 = 4.64068 loss)
I0428 14:00:30.596354 30475 sgd_solver.cpp:105] Iteration 1020, lr = 0.00817056
I0428 14:00:35.308252 30475 solver.cpp:218] Iteration 1032 (2.54674 iter/s, 4.7119s/12 iters), loss = 4.53453
I0428 14:00:35.308292 30475 solver.cpp:237]     Train net output #0: loss = 4.53453 (* 1 = 4.53453 loss)
I0428 14:00:35.308300 30475 sgd_solver.cpp:105] Iteration 1032, lr = 0.00815116
I0428 14:00:40.942658 30475 solver.cpp:218] Iteration 1044 (2.12979 iter/s, 5.63436s/12 iters), loss = 4.55396
I0428 14:00:40.942705 30475 solver.cpp:237]     Train net output #0: loss = 4.55396 (* 1 = 4.55396 loss)
I0428 14:00:40.942713 30475 sgd_solver.cpp:105] Iteration 1044, lr = 0.00813181
I0428 14:00:46.601621 30475 solver.cpp:218] Iteration 1056 (2.12055 iter/s, 5.65891s/12 iters), loss = 4.56164
I0428 14:00:46.601747 30475 solver.cpp:237]     Train net output #0: loss = 4.56164 (* 1 = 4.56164 loss)
I0428 14:00:46.601758 30475 sgd_solver.cpp:105] Iteration 1056, lr = 0.0081125
I0428 14:00:52.250039 30475 solver.cpp:218] Iteration 1068 (2.12454 iter/s, 5.64829s/12 iters), loss = 4.76003
I0428 14:00:52.250079 30475 solver.cpp:237]     Train net output #0: loss = 4.76003 (* 1 = 4.76003 loss)
I0428 14:00:52.250088 30475 sgd_solver.cpp:105] Iteration 1068, lr = 0.00809324
I0428 14:00:57.972280 30475 solver.cpp:218] Iteration 1080 (2.0971 iter/s, 5.72219s/12 iters), loss = 4.51135
I0428 14:00:57.972333 30475 solver.cpp:237]     Train net output #0: loss = 4.51135 (* 1 = 4.51135 loss)
I0428 14:00:57.972342 30475 sgd_solver.cpp:105] Iteration 1080, lr = 0.00807403
I0428 14:01:03.903806 30475 solver.cpp:218] Iteration 1092 (2.02311 iter/s, 5.93147s/12 iters), loss = 4.56372
I0428 14:01:03.903846 30475 solver.cpp:237]     Train net output #0: loss = 4.56372 (* 1 = 4.56372 loss)
I0428 14:01:03.903856 30475 sgd_solver.cpp:105] Iteration 1092, lr = 0.00805486
I0428 14:01:09.683887 30475 solver.cpp:218] Iteration 1104 (2.07611 iter/s, 5.78003s/12 iters), loss = 4.4893
I0428 14:01:09.683933 30475 solver.cpp:237]     Train net output #0: loss = 4.4893 (* 1 = 4.4893 loss)
I0428 14:01:09.683941 30475 sgd_solver.cpp:105] Iteration 1104, lr = 0.00803573
I0428 14:01:13.250516 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:01:15.402856 30475 solver.cpp:218] Iteration 1116 (2.0983 iter/s, 5.71891s/12 iters), loss = 4.37089
I0428 14:01:15.402901 30475 solver.cpp:237]     Train net output #0: loss = 4.37089 (* 1 = 4.37089 loss)
I0428 14:01:15.402910 30475 sgd_solver.cpp:105] Iteration 1116, lr = 0.00801666
I0428 14:01:17.692401 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1122.caffemodel
I0428 14:01:21.408638 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1122.solverstate
I0428 14:01:26.718475 30475 solver.cpp:330] Iteration 1122, Testing net (#0)
I0428 14:01:26.718497 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:01:31.102012 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:01:31.633781 30475 solver.cpp:397]     Test net output #0: accuracy = 0.0398284
I0428 14:01:31.633823 30475 solver.cpp:397]     Test net output #1: loss = 4.62157 (* 1 = 4.62157 loss)
I0428 14:01:33.784814 30475 solver.cpp:218] Iteration 1128 (0.652815 iter/s, 18.3819s/12 iters), loss = 4.41566
I0428 14:01:33.784860 30475 solver.cpp:237]     Train net output #0: loss = 4.41566 (* 1 = 4.41566 loss)
I0428 14:01:33.784868 30475 sgd_solver.cpp:105] Iteration 1128, lr = 0.00799762
I0428 14:01:39.464668 30475 solver.cpp:218] Iteration 1140 (2.11275 iter/s, 5.6798s/12 iters), loss = 4.64666
I0428 14:01:39.464712 30475 solver.cpp:237]     Train net output #0: loss = 4.64666 (* 1 = 4.64666 loss)
I0428 14:01:39.464721 30475 sgd_solver.cpp:105] Iteration 1140, lr = 0.00797863
I0428 14:01:45.130273 30475 solver.cpp:218] Iteration 1152 (2.11806 iter/s, 5.66555s/12 iters), loss = 4.53918
I0428 14:01:45.130316 30475 solver.cpp:237]     Train net output #0: loss = 4.53918 (* 1 = 4.53918 loss)
I0428 14:01:45.130323 30475 sgd_solver.cpp:105] Iteration 1152, lr = 0.00795969
I0428 14:01:50.799769 30475 solver.cpp:218] Iteration 1164 (2.11661 iter/s, 5.66944s/12 iters), loss = 4.34379
I0428 14:01:50.799894 30475 solver.cpp:237]     Train net output #0: loss = 4.34379 (* 1 = 4.34379 loss)
I0428 14:01:50.799903 30475 sgd_solver.cpp:105] Iteration 1164, lr = 0.00794079
I0428 14:01:56.362751 30475 solver.cpp:218] Iteration 1176 (2.15717 iter/s, 5.56285s/12 iters), loss = 4.49239
I0428 14:01:56.362797 30475 solver.cpp:237]     Train net output #0: loss = 4.49239 (* 1 = 4.49239 loss)
I0428 14:01:56.362805 30475 sgd_solver.cpp:105] Iteration 1176, lr = 0.00792194
I0428 14:02:01.895709 30475 solver.cpp:218] Iteration 1188 (2.16884 iter/s, 5.53291s/12 iters), loss = 4.36766
I0428 14:02:01.895745 30475 solver.cpp:237]     Train net output #0: loss = 4.36766 (* 1 = 4.36766 loss)
I0428 14:02:01.895754 30475 sgd_solver.cpp:105] Iteration 1188, lr = 0.00790313
I0428 14:02:07.556591 30475 solver.cpp:218] Iteration 1200 (2.11983 iter/s, 5.66083s/12 iters), loss = 4.35903
I0428 14:02:07.556635 30475 solver.cpp:237]     Train net output #0: loss = 4.35903 (* 1 = 4.35903 loss)
I0428 14:02:07.556643 30475 sgd_solver.cpp:105] Iteration 1200, lr = 0.00788437
I0428 14:02:13.222657 30475 solver.cpp:218] Iteration 1212 (2.11789 iter/s, 5.66601s/12 iters), loss = 4.43072
I0428 14:02:13.222704 30475 solver.cpp:237]     Train net output #0: loss = 4.43072 (* 1 = 4.43072 loss)
I0428 14:02:13.222713 30475 sgd_solver.cpp:105] Iteration 1212, lr = 0.00786565
I0428 14:02:13.501029 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:02:18.326442 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1224.caffemodel
I0428 14:02:21.363488 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1224.solverstate
I0428 14:02:23.204196 30475 solver.cpp:330] Iteration 1224, Testing net (#0)
I0428 14:02:23.204216 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:02:27.727389 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:02:28.338812 30475 solver.cpp:397]     Test net output #0: accuracy = 0.0533088
I0428 14:02:28.338841 30475 solver.cpp:397]     Test net output #1: loss = 4.38751 (* 1 = 4.38751 loss)
I0428 14:02:28.500522 30475 solver.cpp:218] Iteration 1224 (0.785452 iter/s, 15.2778s/12 iters), loss = 4.31516
I0428 14:02:28.500573 30475 solver.cpp:237]     Train net output #0: loss = 4.31516 (* 1 = 4.31516 loss)
I0428 14:02:28.500582 30475 sgd_solver.cpp:105] Iteration 1224, lr = 0.00784697
I0428 14:02:33.211472 30475 solver.cpp:218] Iteration 1236 (2.54729 iter/s, 4.71089s/12 iters), loss = 4.419
I0428 14:02:33.211511 30475 solver.cpp:237]     Train net output #0: loss = 4.419 (* 1 = 4.419 loss)
I0428 14:02:33.211519 30475 sgd_solver.cpp:105] Iteration 1236, lr = 0.00782834
I0428 14:02:38.853314 30475 solver.cpp:218] Iteration 1248 (2.12698 iter/s, 5.6418s/12 iters), loss = 4.38652
I0428 14:02:38.853354 30475 solver.cpp:237]     Train net output #0: loss = 4.38652 (* 1 = 4.38652 loss)
I0428 14:02:38.853363 30475 sgd_solver.cpp:105] Iteration 1248, lr = 0.00780976
I0428 14:02:44.407471 30475 solver.cpp:218] Iteration 1260 (2.16056 iter/s, 5.55411s/12 iters), loss = 4.23806
I0428 14:02:44.407517 30475 solver.cpp:237]     Train net output #0: loss = 4.23806 (* 1 = 4.23806 loss)
I0428 14:02:44.407526 30475 sgd_solver.cpp:105] Iteration 1260, lr = 0.00779122
I0428 14:02:50.046763 30475 solver.cpp:218] Iteration 1272 (2.12795 iter/s, 5.63924s/12 iters), loss = 4.17541
I0428 14:02:50.046808 30475 solver.cpp:237]     Train net output #0: loss = 4.17541 (* 1 = 4.17541 loss)
I0428 14:02:50.046816 30475 sgd_solver.cpp:105] Iteration 1272, lr = 0.00777272
I0428 14:02:55.696110 30475 solver.cpp:218] Iteration 1284 (2.12416 iter/s, 5.6493s/12 iters), loss = 4.3981
I0428 14:02:55.696193 30475 solver.cpp:237]     Train net output #0: loss = 4.3981 (* 1 = 4.3981 loss)
I0428 14:02:55.696203 30475 sgd_solver.cpp:105] Iteration 1284, lr = 0.00775426
I0428 14:03:01.428086 30475 solver.cpp:218] Iteration 1296 (2.09355 iter/s, 5.73189s/12 iters), loss = 4.32865
I0428 14:03:01.428129 30475 solver.cpp:237]     Train net output #0: loss = 4.32865 (* 1 = 4.32865 loss)
I0428 14:03:01.428138 30475 sgd_solver.cpp:105] Iteration 1296, lr = 0.00773585
I0428 14:03:07.258544 30475 solver.cpp:218] Iteration 1308 (2.05818 iter/s, 5.83041s/12 iters), loss = 4.04935
I0428 14:03:07.258589 30475 solver.cpp:237]     Train net output #0: loss = 4.04935 (* 1 = 4.04935 loss)
I0428 14:03:07.258601 30475 sgd_solver.cpp:105] Iteration 1308, lr = 0.00771749
I0428 14:03:10.119096 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:03:12.939373 30475 solver.cpp:218] Iteration 1320 (2.11239 iter/s, 5.68078s/12 iters), loss = 4.00635
I0428 14:03:12.939414 30475 solver.cpp:237]     Train net output #0: loss = 4.00635 (* 1 = 4.00635 loss)
I0428 14:03:12.939424 30475 sgd_solver.cpp:105] Iteration 1320, lr = 0.00769916
I0428 14:03:15.196907 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1326.caffemodel
I0428 14:03:18.681941 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1326.solverstate
I0428 14:03:20.384253 30475 solver.cpp:330] Iteration 1326, Testing net (#0)
I0428 14:03:20.384274 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:03:24.727082 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:03:25.381687 30475 solver.cpp:397]     Test net output #0: accuracy = 0.0667892
I0428 14:03:25.381723 30475 solver.cpp:397]     Test net output #1: loss = 4.25035 (* 1 = 4.25035 loss)
I0428 14:03:27.503872 30475 solver.cpp:218] Iteration 1332 (0.823923 iter/s, 14.5645s/12 iters), loss = 4.22714
I0428 14:03:27.504040 30475 solver.cpp:237]     Train net output #0: loss = 4.22714 (* 1 = 4.22714 loss)
I0428 14:03:27.504050 30475 sgd_solver.cpp:105] Iteration 1332, lr = 0.00768088
I0428 14:03:33.153365 30475 solver.cpp:218] Iteration 1344 (2.12415 iter/s, 5.64932s/12 iters), loss = 4.15937
I0428 14:03:33.153412 30475 solver.cpp:237]     Train net output #0: loss = 4.15937 (* 1 = 4.15937 loss)
I0428 14:03:33.153420 30475 sgd_solver.cpp:105] Iteration 1344, lr = 0.00766265
I0428 14:03:38.750494 30475 solver.cpp:218] Iteration 1356 (2.14398 iter/s, 5.59708s/12 iters), loss = 4.33306
I0428 14:03:38.750536 30475 solver.cpp:237]     Train net output #0: loss = 4.33306 (* 1 = 4.33306 loss)
I0428 14:03:38.750545 30475 sgd_solver.cpp:105] Iteration 1356, lr = 0.00764446
I0428 14:03:44.346844 30475 solver.cpp:218] Iteration 1368 (2.14427 iter/s, 5.5963s/12 iters), loss = 4.18782
I0428 14:03:44.346892 30475 solver.cpp:237]     Train net output #0: loss = 4.18782 (* 1 = 4.18782 loss)
I0428 14:03:44.346900 30475 sgd_solver.cpp:105] Iteration 1368, lr = 0.00762631
I0428 14:03:45.689986 30475 blocking_queue.cpp:49] Waiting for data
I0428 14:03:50.044754 30475 solver.cpp:218] Iteration 1380 (2.10605 iter/s, 5.69786s/12 iters), loss = 4.11592
I0428 14:03:50.044795 30475 solver.cpp:237]     Train net output #0: loss = 4.11592 (* 1 = 4.11592 loss)
I0428 14:03:50.044808 30475 sgd_solver.cpp:105] Iteration 1380, lr = 0.0076082
I0428 14:03:55.492739 30475 solver.cpp:218] Iteration 1392 (2.20267 iter/s, 5.44794s/12 iters), loss = 4.11679
I0428 14:03:55.492777 30475 solver.cpp:237]     Train net output #0: loss = 4.11679 (* 1 = 4.11679 loss)
I0428 14:03:55.492785 30475 sgd_solver.cpp:105] Iteration 1392, lr = 0.00759014
I0428 14:04:01.104605 30475 solver.cpp:218] Iteration 1404 (2.13834 iter/s, 5.61182s/12 iters), loss = 3.89342
I0428 14:04:01.104718 30475 solver.cpp:237]     Train net output #0: loss = 3.89342 (* 1 = 3.89342 loss)
I0428 14:04:01.104727 30475 sgd_solver.cpp:105] Iteration 1404, lr = 0.00757212
I0428 14:04:06.358711 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:04:06.768301 30475 solver.cpp:218] Iteration 1416 (2.1188 iter/s, 5.66358s/12 iters), loss = 3.99754
I0428 14:04:06.768347 30475 solver.cpp:237]     Train net output #0: loss = 3.99754 (* 1 = 3.99754 loss)
I0428 14:04:06.768354 30475 sgd_solver.cpp:105] Iteration 1416, lr = 0.00755414
I0428 14:04:11.865207 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1428.caffemodel
I0428 14:04:14.593605 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1428.solverstate
I0428 14:04:17.425864 30475 solver.cpp:330] Iteration 1428, Testing net (#0)
I0428 14:04:17.425884 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:04:21.845726 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:04:22.557462 30475 solver.cpp:397]     Test net output #0: accuracy = 0.0833333
I0428 14:04:22.557490 30475 solver.cpp:397]     Test net output #1: loss = 4.14611 (* 1 = 4.14611 loss)
I0428 14:04:22.719259 30475 solver.cpp:218] Iteration 1428 (0.752307 iter/s, 15.9509s/12 iters), loss = 3.90725
I0428 14:04:22.719296 30475 solver.cpp:237]     Train net output #0: loss = 3.90725 (* 1 = 3.90725 loss)
I0428 14:04:22.719305 30475 sgd_solver.cpp:105] Iteration 1428, lr = 0.0075362
I0428 14:04:27.510289 30475 solver.cpp:218] Iteration 1440 (2.5047 iter/s, 4.79099s/12 iters), loss = 4.01998
I0428 14:04:27.510327 30475 solver.cpp:237]     Train net output #0: loss = 4.01998 (* 1 = 4.01998 loss)
I0428 14:04:27.510335 30475 sgd_solver.cpp:105] Iteration 1440, lr = 0.00751831
I0428 14:04:33.048151 30475 solver.cpp:218] Iteration 1452 (2.16692 iter/s, 5.53781s/12 iters), loss = 3.74343
I0428 14:04:33.048249 30475 solver.cpp:237]     Train net output #0: loss = 3.74343 (* 1 = 3.74343 loss)
I0428 14:04:33.048259 30475 sgd_solver.cpp:105] Iteration 1452, lr = 0.00750046
I0428 14:04:38.712872 30475 solver.cpp:218] Iteration 1464 (2.11841 iter/s, 5.66462s/12 iters), loss = 4.03164
I0428 14:04:38.712915 30475 solver.cpp:237]     Train net output #0: loss = 4.03164 (* 1 = 4.03164 loss)
I0428 14:04:38.712924 30475 sgd_solver.cpp:105] Iteration 1464, lr = 0.00748265
I0428 14:04:44.357645 30475 solver.cpp:218] Iteration 1476 (2.12588 iter/s, 5.64472s/12 iters), loss = 3.88725
I0428 14:04:44.357684 30475 solver.cpp:237]     Train net output #0: loss = 3.88725 (* 1 = 3.88725 loss)
I0428 14:04:44.357693 30475 sgd_solver.cpp:105] Iteration 1476, lr = 0.00746489
I0428 14:04:50.020529 30475 solver.cpp:218] Iteration 1488 (2.11908 iter/s, 5.66284s/12 iters), loss = 3.84664
I0428 14:04:50.020573 30475 solver.cpp:237]     Train net output #0: loss = 3.84664 (* 1 = 3.84664 loss)
I0428 14:04:50.020581 30475 sgd_solver.cpp:105] Iteration 1488, lr = 0.00744716
I0428 14:04:55.592670 30475 solver.cpp:218] Iteration 1500 (2.15359 iter/s, 5.5721s/12 iters), loss = 4.093
I0428 14:04:55.592706 30475 solver.cpp:237]     Train net output #0: loss = 4.093 (* 1 = 4.093 loss)
I0428 14:04:55.592715 30475 sgd_solver.cpp:105] Iteration 1500, lr = 0.00742948
I0428 14:05:01.271374 30475 solver.cpp:218] Iteration 1512 (2.11318 iter/s, 5.67866s/12 iters), loss = 3.93274
I0428 14:05:01.271414 30475 solver.cpp:237]     Train net output #0: loss = 3.93274 (* 1 = 3.93274 loss)
I0428 14:05:01.271422 30475 sgd_solver.cpp:105] Iteration 1512, lr = 0.00741184
I0428 14:05:03.279129 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:05:06.942929 30475 solver.cpp:218] Iteration 1524 (2.11584 iter/s, 5.67151s/12 iters), loss = 4.02426
I0428 14:05:06.942970 30475 solver.cpp:237]     Train net output #0: loss = 4.02426 (* 1 = 4.02426 loss)
I0428 14:05:06.942979 30475 sgd_solver.cpp:105] Iteration 1524, lr = 0.00739425
I0428 14:05:09.191506 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1530.caffemodel
I0428 14:05:11.583703 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1530.solverstate
I0428 14:05:14.031188 30475 solver.cpp:330] Iteration 1530, Testing net (#0)
I0428 14:05:14.031208 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:05:18.364125 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:05:19.117435 30475 solver.cpp:397]     Test net output #0: accuracy = 0.0882353
I0428 14:05:19.117470 30475 solver.cpp:397]     Test net output #1: loss = 4.09241 (* 1 = 4.09241 loss)
I0428 14:05:21.273454 30475 solver.cpp:218] Iteration 1536 (0.837375 iter/s, 14.3305s/12 iters), loss = 3.79058
I0428 14:05:21.273499 30475 solver.cpp:237]     Train net output #0: loss = 3.79058 (* 1 = 3.79058 loss)
I0428 14:05:21.273509 30475 sgd_solver.cpp:105] Iteration 1536, lr = 0.00737669
I0428 14:05:26.931329 30475 solver.cpp:218] Iteration 1548 (2.12096 iter/s, 5.65782s/12 iters), loss = 3.81357
I0428 14:05:26.931375 30475 solver.cpp:237]     Train net output #0: loss = 3.81357 (* 1 = 3.81357 loss)
I0428 14:05:26.931383 30475 sgd_solver.cpp:105] Iteration 1548, lr = 0.00735918
I0428 14:05:32.589748 30475 solver.cpp:218] Iteration 1560 (2.12075 iter/s, 5.65837s/12 iters), loss = 4.00047
I0428 14:05:32.589792 30475 solver.cpp:237]     Train net output #0: loss = 4.00047 (* 1 = 4.00047 loss)
I0428 14:05:32.589800 30475 sgd_solver.cpp:105] Iteration 1560, lr = 0.00734171
I0428 14:05:38.253477 30475 solver.cpp:218] Iteration 1572 (2.11876 iter/s, 5.66368s/12 iters), loss = 3.70531
I0428 14:05:38.253639 30475 solver.cpp:237]     Train net output #0: loss = 3.70531 (* 1 = 3.70531 loss)
I0428 14:05:38.253649 30475 sgd_solver.cpp:105] Iteration 1572, lr = 0.00732427
I0428 14:05:44.203264 30475 solver.cpp:218] Iteration 1584 (2.01693 iter/s, 5.94962s/12 iters), loss = 4.09849
I0428 14:05:44.203307 30475 solver.cpp:237]     Train net output #0: loss = 4.09849 (* 1 = 4.09849 loss)
I0428 14:05:44.203315 30475 sgd_solver.cpp:105] Iteration 1584, lr = 0.00730688
I0428 14:05:49.851035 30475 solver.cpp:218] Iteration 1596 (2.12475 iter/s, 5.64773s/12 iters), loss = 3.96507
I0428 14:05:49.851070 30475 solver.cpp:237]     Train net output #0: loss = 3.96507 (* 1 = 3.96507 loss)
I0428 14:05:49.851078 30475 sgd_solver.cpp:105] Iteration 1596, lr = 0.00728954
I0428 14:05:55.516789 30475 solver.cpp:218] Iteration 1608 (2.118 iter/s, 5.66571s/12 iters), loss = 3.54411
I0428 14:05:55.516830 30475 solver.cpp:237]     Train net output #0: loss = 3.54411 (* 1 = 3.54411 loss)
I0428 14:05:55.516839 30475 sgd_solver.cpp:105] Iteration 1608, lr = 0.00727223
I0428 14:05:59.935689 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:06:01.190043 30475 solver.cpp:218] Iteration 1620 (2.11521 iter/s, 5.6732s/12 iters), loss = 3.89601
I0428 14:06:01.190088 30475 solver.cpp:237]     Train net output #0: loss = 3.89601 (* 1 = 3.89601 loss)
I0428 14:06:01.190096 30475 sgd_solver.cpp:105] Iteration 1620, lr = 0.00725496
I0428 14:06:06.281924 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1632.caffemodel
I0428 14:06:08.533427 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1632.solverstate
I0428 14:06:11.272125 30475 solver.cpp:330] Iteration 1632, Testing net (#0)
I0428 14:06:11.272143 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:06:15.530105 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:06:16.311708 30475 solver.cpp:397]     Test net output #0: accuracy = 0.114583
I0428 14:06:16.311744 30475 solver.cpp:397]     Test net output #1: loss = 3.8703 (* 1 = 3.8703 loss)
I0428 14:06:16.473667 30475 solver.cpp:218] Iteration 1632 (0.785156 iter/s, 15.2836s/12 iters), loss = 3.80613
I0428 14:06:16.473711 30475 solver.cpp:237]     Train net output #0: loss = 3.80613 (* 1 = 3.80613 loss)
I0428 14:06:16.473718 30475 sgd_solver.cpp:105] Iteration 1632, lr = 0.00723774
I0428 14:06:21.162405 30475 solver.cpp:218] Iteration 1644 (2.55935 iter/s, 4.68869s/12 iters), loss = 3.5387
I0428 14:06:21.162446 30475 solver.cpp:237]     Train net output #0: loss = 3.5387 (* 1 = 3.5387 loss)
I0428 14:06:21.162454 30475 sgd_solver.cpp:105] Iteration 1644, lr = 0.00722056
I0428 14:06:26.642169 30475 solver.cpp:218] Iteration 1656 (2.18989 iter/s, 5.47971s/12 iters), loss = 3.73329
I0428 14:06:26.642210 30475 solver.cpp:237]     Train net output #0: loss = 3.73329 (* 1 = 3.73329 loss)
I0428 14:06:26.642218 30475 sgd_solver.cpp:105] Iteration 1656, lr = 0.00720341
I0428 14:06:32.310286 30475 solver.cpp:218] Iteration 1668 (2.11712 iter/s, 5.66808s/12 iters), loss = 3.78716
I0428 14:06:32.310326 30475 solver.cpp:237]     Train net output #0: loss = 3.78716 (* 1 = 3.78716 loss)
I0428 14:06:32.310334 30475 sgd_solver.cpp:105] Iteration 1668, lr = 0.00718631
I0428 14:06:37.987879 30475 solver.cpp:218] Iteration 1680 (2.11358 iter/s, 5.67756s/12 iters), loss = 3.4931
I0428 14:06:37.987926 30475 solver.cpp:237]     Train net output #0: loss = 3.4931 (* 1 = 3.4931 loss)
I0428 14:06:37.987934 30475 sgd_solver.cpp:105] Iteration 1680, lr = 0.00716925
I0428 14:06:43.643661 30475 solver.cpp:218] Iteration 1692 (2.12174 iter/s, 5.65573s/12 iters), loss = 3.6442
I0428 14:06:43.643817 30475 solver.cpp:237]     Train net output #0: loss = 3.6442 (* 1 = 3.6442 loss)
I0428 14:06:43.643831 30475 sgd_solver.cpp:105] Iteration 1692, lr = 0.00715223
I0428 14:06:49.329486 30475 solver.cpp:218] Iteration 1704 (2.11057 iter/s, 5.68568s/12 iters), loss = 3.79362
I0428 14:06:49.329531 30475 solver.cpp:237]     Train net output #0: loss = 3.79362 (* 1 = 3.79362 loss)
I0428 14:06:49.329540 30475 sgd_solver.cpp:105] Iteration 1704, lr = 0.00713525
I0428 14:06:55.001785 30475 solver.cpp:218] Iteration 1716 (2.11556 iter/s, 5.67226s/12 iters), loss = 3.59488
I0428 14:06:55.001827 30475 solver.cpp:237]     Train net output #0: loss = 3.59488 (* 1 = 3.59488 loss)
I0428 14:06:55.001834 30475 sgd_solver.cpp:105] Iteration 1716, lr = 0.00711831
I0428 14:06:56.160667 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:07:00.576884 30475 solver.cpp:218] Iteration 1728 (2.15245 iter/s, 5.57505s/12 iters), loss = 3.54049
I0428 14:07:00.576942 30475 solver.cpp:237]     Train net output #0: loss = 3.54049 (* 1 = 3.54049 loss)
I0428 14:07:00.576956 30475 sgd_solver.cpp:105] Iteration 1728, lr = 0.00710141
I0428 14:07:02.816910 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1734.caffemodel
I0428 14:07:05.043499 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1734.solverstate
I0428 14:07:06.738878 30475 solver.cpp:330] Iteration 1734, Testing net (#0)
I0428 14:07:06.738903 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:07:10.932622 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:07:11.727665 30475 solver.cpp:397]     Test net output #0: accuracy = 0.130515
I0428 14:07:11.727694 30475 solver.cpp:397]     Test net output #1: loss = 3.76485 (* 1 = 3.76485 loss)
I0428 14:07:13.860375 30475 solver.cpp:218] Iteration 1740 (0.903378 iter/s, 13.2835s/12 iters), loss = 3.42147
I0428 14:07:13.860483 30475 solver.cpp:237]     Train net output #0: loss = 3.42147 (* 1 = 3.42147 loss)
I0428 14:07:13.860494 30475 sgd_solver.cpp:105] Iteration 1740, lr = 0.00708455
I0428 14:07:19.526939 30475 solver.cpp:218] Iteration 1752 (2.11772 iter/s, 5.66647s/12 iters), loss = 3.46998
I0428 14:07:19.526981 30475 solver.cpp:237]     Train net output #0: loss = 3.46998 (* 1 = 3.46998 loss)
I0428 14:07:19.526989 30475 sgd_solver.cpp:105] Iteration 1752, lr = 0.00706773
I0428 14:07:25.175933 30475 solver.cpp:218] Iteration 1764 (2.12429 iter/s, 5.64896s/12 iters), loss = 3.39528
I0428 14:07:25.175977 30475 solver.cpp:237]     Train net output #0: loss = 3.39528 (* 1 = 3.39528 loss)
I0428 14:07:25.175987 30475 sgd_solver.cpp:105] Iteration 1764, lr = 0.00705094
I0428 14:07:30.815778 30475 solver.cpp:218] Iteration 1776 (2.12773 iter/s, 5.6398s/12 iters), loss = 3.45761
I0428 14:07:30.815827 30475 solver.cpp:237]     Train net output #0: loss = 3.45761 (* 1 = 3.45761 loss)
I0428 14:07:30.815835 30475 sgd_solver.cpp:105] Iteration 1776, lr = 0.0070342
I0428 14:07:36.450273 30475 solver.cpp:218] Iteration 1788 (2.12976 iter/s, 5.63445s/12 iters), loss = 3.21644
I0428 14:07:36.450316 30475 solver.cpp:237]     Train net output #0: loss = 3.21644 (* 1 = 3.21644 loss)
I0428 14:07:36.450325 30475 sgd_solver.cpp:105] Iteration 1788, lr = 0.0070175
I0428 14:07:42.093480 30475 solver.cpp:218] Iteration 1800 (2.12646 iter/s, 5.64317s/12 iters), loss = 3.48439
I0428 14:07:42.093519 30475 solver.cpp:237]     Train net output #0: loss = 3.48439 (* 1 = 3.48439 loss)
I0428 14:07:42.093528 30475 sgd_solver.cpp:105] Iteration 1800, lr = 0.00700084
I0428 14:07:47.744813 30475 solver.cpp:218] Iteration 1812 (2.12341 iter/s, 5.65129s/12 iters), loss = 3.70953
I0428 14:07:47.744952 30475 solver.cpp:237]     Train net output #0: loss = 3.70953 (* 1 = 3.70953 loss)
I0428 14:07:47.744962 30475 sgd_solver.cpp:105] Iteration 1812, lr = 0.00698422
I0428 14:07:51.309032 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:07:53.410281 30475 solver.cpp:218] Iteration 1824 (2.11815 iter/s, 5.66533s/12 iters), loss = 3.24268
I0428 14:07:53.410323 30475 solver.cpp:237]     Train net output #0: loss = 3.24268 (* 1 = 3.24268 loss)
I0428 14:07:53.410332 30475 sgd_solver.cpp:105] Iteration 1824, lr = 0.00696764
I0428 14:07:58.492558 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1836.caffemodel
I0428 14:08:01.305105 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1836.solverstate
I0428 14:08:03.020313 30475 solver.cpp:330] Iteration 1836, Testing net (#0)
I0428 14:08:03.020334 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:08:07.351164 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:08:08.261905 30475 solver.cpp:397]     Test net output #0: accuracy = 0.147059
I0428 14:08:08.261973 30475 solver.cpp:397]     Test net output #1: loss = 3.6818 (* 1 = 3.6818 loss)
I0428 14:08:08.423377 30475 solver.cpp:218] Iteration 1836 (0.799302 iter/s, 15.0131s/12 iters), loss = 3.497
I0428 14:08:08.424988 30475 solver.cpp:237]     Train net output #0: loss = 3.497 (* 1 = 3.497 loss)
I0428 14:08:08.425001 30475 sgd_solver.cpp:105] Iteration 1836, lr = 0.0069511
I0428 14:08:13.145610 30475 solver.cpp:218] Iteration 1848 (2.54203 iter/s, 4.72063s/12 iters), loss = 3.56657
I0428 14:08:13.145649 30475 solver.cpp:237]     Train net output #0: loss = 3.56657 (* 1 = 3.56657 loss)
I0428 14:08:13.145658 30475 sgd_solver.cpp:105] Iteration 1848, lr = 0.00693459
I0428 14:08:18.783960 30475 solver.cpp:218] Iteration 1860 (2.1283 iter/s, 5.63831s/12 iters), loss = 3.54519
I0428 14:08:18.784041 30475 solver.cpp:237]     Train net output #0: loss = 3.54519 (* 1 = 3.54519 loss)
I0428 14:08:18.784051 30475 sgd_solver.cpp:105] Iteration 1860, lr = 0.00691813
I0428 14:08:24.504393 30475 solver.cpp:218] Iteration 1872 (2.09778 iter/s, 5.72034s/12 iters), loss = 3.43386
I0428 14:08:24.504460 30475 solver.cpp:237]     Train net output #0: loss = 3.43386 (* 1 = 3.43386 loss)
I0428 14:08:24.504473 30475 sgd_solver.cpp:105] Iteration 1872, lr = 0.0069017
I0428 14:08:30.130754 30475 solver.cpp:218] Iteration 1884 (2.13284 iter/s, 5.6263s/12 iters), loss = 3.23295
I0428 14:08:30.130800 30475 solver.cpp:237]     Train net output #0: loss = 3.23295 (* 1 = 3.23295 loss)
I0428 14:08:30.130807 30475 sgd_solver.cpp:105] Iteration 1884, lr = 0.00688532
I0428 14:08:35.754127 30475 solver.cpp:218] Iteration 1896 (2.13396 iter/s, 5.62334s/12 iters), loss = 3.2059
I0428 14:08:35.754161 30475 solver.cpp:237]     Train net output #0: loss = 3.2059 (* 1 = 3.2059 loss)
I0428 14:08:35.754168 30475 sgd_solver.cpp:105] Iteration 1896, lr = 0.00686897
I0428 14:08:41.385481 30475 solver.cpp:218] Iteration 1908 (2.13094 iter/s, 5.63132s/12 iters), loss = 3.4353
I0428 14:08:41.385526 30475 solver.cpp:237]     Train net output #0: loss = 3.4353 (* 1 = 3.4353 loss)
I0428 14:08:41.385535 30475 sgd_solver.cpp:105] Iteration 1908, lr = 0.00685266
I0428 14:08:47.132766 30475 solver.cpp:218] Iteration 1920 (2.08796 iter/s, 5.74724s/12 iters), loss = 3.43101
I0428 14:08:47.132812 30475 solver.cpp:237]     Train net output #0: loss = 3.43101 (* 1 = 3.43101 loss)
I0428 14:08:47.132819 30475 sgd_solver.cpp:105] Iteration 1920, lr = 0.00683639
I0428 14:08:47.444375 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:08:52.826409 30475 solver.cpp:218] Iteration 1932 (2.10763 iter/s, 5.6936s/12 iters), loss = 3.1239
I0428 14:08:52.826539 30475 solver.cpp:237]     Train net output #0: loss = 3.1239 (* 1 = 3.1239 loss)
I0428 14:08:52.826546 30475 sgd_solver.cpp:105] Iteration 1932, lr = 0.00682016
I0428 14:08:55.070528 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1938.caffemodel
I0428 14:09:00.328902 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1938.solverstate
I0428 14:09:02.545264 30475 solver.cpp:330] Iteration 1938, Testing net (#0)
I0428 14:09:02.545286 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:09:06.776865 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:09:07.742990 30475 solver.cpp:397]     Test net output #0: accuracy = 0.156863
I0428 14:09:07.743022 30475 solver.cpp:397]     Test net output #1: loss = 3.57131 (* 1 = 3.57131 loss)
I0428 14:09:09.906857 30475 solver.cpp:218] Iteration 1944 (0.702561 iter/s, 17.0804s/12 iters), loss = 3.29557
I0428 14:09:09.906909 30475 solver.cpp:237]     Train net output #0: loss = 3.29557 (* 1 = 3.29557 loss)
I0428 14:09:09.906920 30475 sgd_solver.cpp:105] Iteration 1944, lr = 0.00680397
I0428 14:09:15.566359 30475 solver.cpp:218] Iteration 1956 (2.12035 iter/s, 5.65945s/12 iters), loss = 3.40377
I0428 14:09:15.566402 30475 solver.cpp:237]     Train net output #0: loss = 3.40377 (* 1 = 3.40377 loss)
I0428 14:09:15.566411 30475 sgd_solver.cpp:105] Iteration 1956, lr = 0.00678782
I0428 14:09:21.151697 30475 solver.cpp:218] Iteration 1968 (2.1485 iter/s, 5.58529s/12 iters), loss = 3.34903
I0428 14:09:21.151741 30475 solver.cpp:237]     Train net output #0: loss = 3.34903 (* 1 = 3.34903 loss)
I0428 14:09:21.151749 30475 sgd_solver.cpp:105] Iteration 1968, lr = 0.0067717
I0428 14:09:26.799561 30475 solver.cpp:218] Iteration 1980 (2.12471 iter/s, 5.64782s/12 iters), loss = 3.03165
I0428 14:09:26.799676 30475 solver.cpp:237]     Train net output #0: loss = 3.03165 (* 1 = 3.03165 loss)
I0428 14:09:26.799686 30475 sgd_solver.cpp:105] Iteration 1980, lr = 0.00675562
I0428 14:09:32.458454 30475 solver.cpp:218] Iteration 1992 (2.1206 iter/s, 5.65878s/12 iters), loss = 3.23579
I0428 14:09:32.458498 30475 solver.cpp:237]     Train net output #0: loss = 3.23579 (* 1 = 3.23579 loss)
I0428 14:09:32.458506 30475 sgd_solver.cpp:105] Iteration 1992, lr = 0.00673958
I0428 14:09:38.013166 30475 solver.cpp:218] Iteration 2004 (2.16035 iter/s, 5.55466s/12 iters), loss = 3.28752
I0428 14:09:38.013213 30475 solver.cpp:237]     Train net output #0: loss = 3.28752 (* 1 = 3.28752 loss)
I0428 14:09:38.013221 30475 sgd_solver.cpp:105] Iteration 2004, lr = 0.00672358
I0428 14:09:43.684675 30475 solver.cpp:218] Iteration 2016 (2.11585 iter/s, 5.67147s/12 iters), loss = 2.93763
I0428 14:09:43.684715 30475 solver.cpp:237]     Train net output #0: loss = 2.93763 (* 1 = 2.93763 loss)
I0428 14:09:43.684723 30475 sgd_solver.cpp:105] Iteration 2016, lr = 0.00670762
I0428 14:09:46.638125 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:09:49.423560 30475 solver.cpp:218] Iteration 2028 (2.09101 iter/s, 5.73885s/12 iters), loss = 3.03867
I0428 14:09:49.423599 30475 solver.cpp:237]     Train net output #0: loss = 3.03867 (* 1 = 3.03867 loss)
I0428 14:09:49.423607 30475 sgd_solver.cpp:105] Iteration 2028, lr = 0.00669169
I0428 14:09:54.384775 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2040.caffemodel
I0428 14:09:57.475366 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2040.solverstate
I0428 14:09:59.937777 30475 solver.cpp:330] Iteration 2040, Testing net (#0)
I0428 14:09:59.937805 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:10:04.079155 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:10:05.071379 30475 solver.cpp:397]     Test net output #0: accuracy = 0.155637
I0428 14:10:05.071408 30475 solver.cpp:397]     Test net output #1: loss = 3.65342 (* 1 = 3.65342 loss)
I0428 14:10:05.233088 30475 solver.cpp:218] Iteration 2040 (0.759036 iter/s, 15.8095s/12 iters), loss = 3.2603
I0428 14:10:05.233153 30475 solver.cpp:237]     Train net output #0: loss = 3.2603 (* 1 = 3.2603 loss)
I0428 14:10:05.233163 30475 sgd_solver.cpp:105] Iteration 2040, lr = 0.00667581
I0428 14:10:10.010336 30475 solver.cpp:218] Iteration 2052 (2.51194 iter/s, 4.77719s/12 iters), loss = 3.02324
I0428 14:10:10.010377 30475 solver.cpp:237]     Train net output #0: loss = 3.02324 (* 1 = 3.02324 loss)
I0428 14:10:10.010385 30475 sgd_solver.cpp:105] Iteration 2052, lr = 0.00665996
I0428 14:10:11.799494 30475 blocking_queue.cpp:49] Waiting for data
I0428 14:10:15.653349 30475 solver.cpp:218] Iteration 2064 (2.12654 iter/s, 5.64297s/12 iters), loss = 3.10772
I0428 14:10:15.653396 30475 solver.cpp:237]     Train net output #0: loss = 3.10772 (* 1 = 3.10772 loss)
I0428 14:10:15.653404 30475 sgd_solver.cpp:105] Iteration 2064, lr = 0.00664414
I0428 14:10:21.259917 30475 solver.cpp:218] Iteration 2076 (2.14037 iter/s, 5.60652s/12 iters), loss = 3.0795
I0428 14:10:21.259954 30475 solver.cpp:237]     Train net output #0: loss = 3.0795 (* 1 = 3.0795 loss)
I0428 14:10:21.259963 30475 sgd_solver.cpp:105] Iteration 2076, lr = 0.00662837
I0428 14:10:26.880005 30475 solver.cpp:218] Iteration 2088 (2.13521 iter/s, 5.62005s/12 iters), loss = 2.97495
I0428 14:10:26.880053 30475 solver.cpp:237]     Train net output #0: loss = 2.97495 (* 1 = 2.97495 loss)
I0428 14:10:26.880061 30475 sgd_solver.cpp:105] Iteration 2088, lr = 0.00661263
I0428 14:10:32.722822 30475 solver.cpp:218] Iteration 2100 (2.05382 iter/s, 5.84277s/12 iters), loss = 2.97303
I0428 14:10:32.722956 30475 solver.cpp:237]     Train net output #0: loss = 2.97303 (* 1 = 2.97303 loss)
I0428 14:10:32.722965 30475 sgd_solver.cpp:105] Iteration 2100, lr = 0.00659693
I0428 14:10:38.386837 30475 solver.cpp:218] Iteration 2112 (2.11869 iter/s, 5.66388s/12 iters), loss = 3.12171
I0428 14:10:38.386881 30475 solver.cpp:237]     Train net output #0: loss = 3.12171 (* 1 = 3.12171 loss)
I0428 14:10:38.386889 30475 sgd_solver.cpp:105] Iteration 2112, lr = 0.00658127
I0428 14:10:43.692574 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:10:44.072190 30475 solver.cpp:218] Iteration 2124 (2.1107 iter/s, 5.68531s/12 iters), loss = 2.82527
I0428 14:10:44.072237 30475 solver.cpp:237]     Train net output #0: loss = 2.82527 (* 1 = 2.82527 loss)
I0428 14:10:44.072244 30475 sgd_solver.cpp:105] Iteration 2124, lr = 0.00656564
I0428 14:10:49.730140 30475 solver.cpp:218] Iteration 2136 (2.12093 iter/s, 5.65791s/12 iters), loss = 2.92059
I0428 14:10:49.730175 30475 solver.cpp:237]     Train net output #0: loss = 2.92059 (* 1 = 2.92059 loss)
I0428 14:10:49.730182 30475 sgd_solver.cpp:105] Iteration 2136, lr = 0.00655006
I0428 14:10:51.976330 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2142.caffemodel
I0428 14:10:54.575006 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2142.solverstate
I0428 14:10:56.343727 30475 solver.cpp:330] Iteration 2142, Testing net (#0)
I0428 14:10:56.343745 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:11:00.633649 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:11:01.710153 30475 solver.cpp:397]     Test net output #0: accuracy = 0.185662
I0428 14:11:01.710196 30475 solver.cpp:397]     Test net output #1: loss = 3.40891 (* 1 = 3.40891 loss)
I0428 14:11:03.849591 30475 solver.cpp:218] Iteration 2148 (0.849892 iter/s, 14.1194s/12 iters), loss = 2.73885
I0428 14:11:03.849709 30475 solver.cpp:237]     Train net output #0: loss = 2.73885 (* 1 = 2.73885 loss)
I0428 14:11:03.849718 30475 sgd_solver.cpp:105] Iteration 2148, lr = 0.00653451
I0428 14:11:09.525023 30475 solver.cpp:218] Iteration 2160 (2.11442 iter/s, 5.67531s/12 iters), loss = 2.68446
I0428 14:11:09.525068 30475 solver.cpp:237]     Train net output #0: loss = 2.68446 (* 1 = 2.68446 loss)
I0428 14:11:09.525077 30475 sgd_solver.cpp:105] Iteration 2160, lr = 0.00651899
I0428 14:11:15.170894 30475 solver.cpp:218] Iteration 2172 (2.12546 iter/s, 5.64583s/12 iters), loss = 3.21586
I0428 14:11:15.170933 30475 solver.cpp:237]     Train net output #0: loss = 3.21586 (* 1 = 3.21586 loss)
I0428 14:11:15.170941 30475 sgd_solver.cpp:105] Iteration 2172, lr = 0.00650351
I0428 14:11:20.798143 30475 solver.cpp:218] Iteration 2184 (2.1325 iter/s, 5.62721s/12 iters), loss = 2.79373
I0428 14:11:20.798187 30475 solver.cpp:237]     Train net output #0: loss = 2.79373 (* 1 = 2.79373 loss)
I0428 14:11:20.798195 30475 sgd_solver.cpp:105] Iteration 2184, lr = 0.00648807
I0428 14:11:26.317514 30475 solver.cpp:218] Iteration 2196 (2.17418 iter/s, 5.51933s/12 iters), loss = 2.8198
I0428 14:11:26.317559 30475 solver.cpp:237]     Train net output #0: loss = 2.8198 (* 1 = 2.8198 loss)
I0428 14:11:26.317566 30475 sgd_solver.cpp:105] Iteration 2196, lr = 0.00647267
I0428 14:11:31.951294 30475 solver.cpp:218] Iteration 2208 (2.13003 iter/s, 5.63373s/12 iters), loss = 2.87583
I0428 14:11:31.951339 30475 solver.cpp:237]     Train net output #0: loss = 2.87583 (* 1 = 2.87583 loss)
I0428 14:11:31.951349 30475 sgd_solver.cpp:105] Iteration 2208, lr = 0.0064573
I0428 14:11:37.612671 30475 solver.cpp:218] Iteration 2220 (2.11964 iter/s, 5.66133s/12 iters), loss = 2.75405
I0428 14:11:37.612860 30475 solver.cpp:237]     Train net output #0: loss = 2.75405 (* 1 = 2.75405 loss)
I0428 14:11:37.612869 30475 sgd_solver.cpp:105] Iteration 2220, lr = 0.00644197
I0428 14:11:39.652756 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:11:43.219668 30475 solver.cpp:218] Iteration 2232 (2.14025 iter/s, 5.60681s/12 iters), loss = 2.87666
I0428 14:11:43.219713 30475 solver.cpp:237]     Train net output #0: loss = 2.87666 (* 1 = 2.87666 loss)
I0428 14:11:43.219722 30475 sgd_solver.cpp:105] Iteration 2232, lr = 0.00642668
I0428 14:11:48.261345 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2244.caffemodel
I0428 14:11:50.456579 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2244.solverstate
I0428 14:11:52.148361 30475 solver.cpp:330] Iteration 2244, Testing net (#0)
I0428 14:11:52.148387 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:11:56.217247 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:11:57.304700 30475 solver.cpp:397]     Test net output #0: accuracy = 0.199142
I0428 14:11:57.304728 30475 solver.cpp:397]     Test net output #1: loss = 3.44643 (* 1 = 3.44643 loss)
I0428 14:11:57.466085 30475 solver.cpp:218] Iteration 2244 (0.842318 iter/s, 14.2464s/12 iters), loss = 2.77278
I0428 14:11:57.466147 30475 solver.cpp:237]     Train net output #0: loss = 2.77278 (* 1 = 2.77278 loss)
I0428 14:11:57.466156 30475 sgd_solver.cpp:105] Iteration 2244, lr = 0.00641142
I0428 14:12:02.171808 30475 solver.cpp:218] Iteration 2256 (2.55012 iter/s, 4.70565s/12 iters), loss = 2.79983
I0428 14:12:02.171852 30475 solver.cpp:237]     Train net output #0: loss = 2.79983 (* 1 = 2.79983 loss)
I0428 14:12:02.171860 30475 sgd_solver.cpp:105] Iteration 2256, lr = 0.0063962
I0428 14:12:07.813838 30475 solver.cpp:218] Iteration 2268 (2.12691 iter/s, 5.64199s/12 iters), loss = 3.083
I0428 14:12:07.813932 30475 solver.cpp:237]     Train net output #0: loss = 3.083 (* 1 = 3.083 loss)
I0428 14:12:07.813942 30475 sgd_solver.cpp:105] Iteration 2268, lr = 0.00638101
I0428 14:12:13.492569 30475 solver.cpp:218] Iteration 2280 (2.11318 iter/s, 5.67864s/12 iters), loss = 2.75661
I0428 14:12:13.492614 30475 solver.cpp:237]     Train net output #0: loss = 2.75661 (* 1 = 2.75661 loss)
I0428 14:12:13.492621 30475 sgd_solver.cpp:105] Iteration 2280, lr = 0.00636586
I0428 14:12:19.103840 30475 solver.cpp:218] Iteration 2292 (2.13857 iter/s, 5.61123s/12 iters), loss = 3.00607
I0428 14:12:19.103883 30475 solver.cpp:237]     Train net output #0: loss = 3.00607 (* 1 = 3.00607 loss)
I0428 14:12:19.103893 30475 sgd_solver.cpp:105] Iteration 2292, lr = 0.00635075
I0428 14:12:24.755051 30475 solver.cpp:218] Iteration 2304 (2.12346 iter/s, 5.65117s/12 iters), loss = 2.8839
I0428 14:12:24.755096 30475 solver.cpp:237]     Train net output #0: loss = 2.8839 (* 1 = 2.8839 loss)
I0428 14:12:24.755105 30475 sgd_solver.cpp:105] Iteration 2304, lr = 0.00633567
I0428 14:12:30.415241 30475 solver.cpp:218] Iteration 2316 (2.12009 iter/s, 5.66014s/12 iters), loss = 2.33158
I0428 14:12:30.415285 30475 solver.cpp:237]     Train net output #0: loss = 2.33158 (* 1 = 2.33158 loss)
I0428 14:12:30.415292 30475 sgd_solver.cpp:105] Iteration 2316, lr = 0.00632063
I0428 14:12:34.861099 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:12:36.080461 30475 solver.cpp:218] Iteration 2328 (2.11821 iter/s, 5.66517s/12 iters), loss = 2.71855
I0428 14:12:36.080525 30475 solver.cpp:237]     Train net output #0: loss = 2.71855 (* 1 = 2.71855 loss)
I0428 14:12:36.080538 30475 sgd_solver.cpp:105] Iteration 2328, lr = 0.00630562
I0428 14:12:41.714787 30475 solver.cpp:218] Iteration 2340 (2.12982 iter/s, 5.63427s/12 iters), loss = 2.74261
I0428 14:12:41.714924 30475 solver.cpp:237]     Train net output #0: loss = 2.74261 (* 1 = 2.74261 loss)
I0428 14:12:41.714934 30475 sgd_solver.cpp:105] Iteration 2340, lr = 0.00629065
I0428 14:12:43.966539 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2346.caffemodel
I0428 14:12:46.180050 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2346.solverstate
I0428 14:12:48.619446 30475 solver.cpp:330] Iteration 2346, Testing net (#0)
I0428 14:12:48.619477 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:12:52.611780 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:12:53.728917 30475 solver.cpp:397]     Test net output #0: accuracy = 0.221814
I0428 14:12:53.728945 30475 solver.cpp:397]     Test net output #1: loss = 3.27653 (* 1 = 3.27653 loss)
I0428 14:12:55.885401 30475 solver.cpp:218] Iteration 2352 (0.846829 iter/s, 14.1705s/12 iters), loss = 2.55094
I0428 14:12:55.885444 30475 solver.cpp:237]     Train net output #0: loss = 2.55094 (* 1 = 2.55094 loss)
I0428 14:12:55.885452 30475 sgd_solver.cpp:105] Iteration 2352, lr = 0.00627571
I0428 14:13:01.557870 30475 solver.cpp:218] Iteration 2364 (2.1155 iter/s, 5.67243s/12 iters), loss = 2.78828
I0428 14:13:01.557910 30475 solver.cpp:237]     Train net output #0: loss = 2.78828 (* 1 = 2.78828 loss)
I0428 14:13:01.557917 30475 sgd_solver.cpp:105] Iteration 2364, lr = 0.00626081
I0428 14:13:07.197630 30475 solver.cpp:218] Iteration 2376 (2.12776 iter/s, 5.63972s/12 iters), loss = 3.0355
I0428 14:13:07.197669 30475 solver.cpp:237]     Train net output #0: loss = 3.0355 (* 1 = 3.0355 loss)
I0428 14:13:07.197679 30475 sgd_solver.cpp:105] Iteration 2376, lr = 0.00624595
I0428 14:13:12.847385 30475 solver.cpp:218] Iteration 2388 (2.124 iter/s, 5.64971s/12 iters), loss = 2.34986
I0428 14:13:12.847510 30475 solver.cpp:237]     Train net output #0: loss = 2.34986 (* 1 = 2.34986 loss)
I0428 14:13:12.847520 30475 sgd_solver.cpp:105] Iteration 2388, lr = 0.00623112
I0428 14:13:18.482775 30475 solver.cpp:218] Iteration 2400 (2.12945 iter/s, 5.63527s/12 iters), loss = 2.2902
I0428 14:13:18.482820 30475 solver.cpp:237]     Train net output #0: loss = 2.2902 (* 1 = 2.2902 loss)
I0428 14:13:18.482828 30475 sgd_solver.cpp:105] Iteration 2400, lr = 0.00621633
I0428 14:13:24.130386 30475 solver.cpp:218] Iteration 2412 (2.12481 iter/s, 5.64757s/12 iters), loss = 2.47686
I0428 14:13:24.130426 30475 solver.cpp:237]     Train net output #0: loss = 2.47686 (* 1 = 2.47686 loss)
I0428 14:13:24.130434 30475 sgd_solver.cpp:105] Iteration 2412, lr = 0.00620157
I0428 14:13:29.798710 30475 solver.cpp:218] Iteration 2424 (2.11704 iter/s, 5.66829s/12 iters), loss = 2.35123
I0428 14:13:29.798753 30475 solver.cpp:237]     Train net output #0: loss = 2.35123 (* 1 = 2.35123 loss)
I0428 14:13:29.798761 30475 sgd_solver.cpp:105] Iteration 2424, lr = 0.00618684
I0428 14:13:30.980660 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:13:35.466084 30475 solver.cpp:218] Iteration 2436 (2.1174 iter/s, 5.66733s/12 iters), loss = 2.8206
I0428 14:13:35.466128 30475 solver.cpp:237]     Train net output #0: loss = 2.8206 (* 1 = 2.8206 loss)
I0428 14:13:35.466136 30475 sgd_solver.cpp:105] Iteration 2436, lr = 0.00617215
I0428 14:13:40.525012 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2448.caffemodel
I0428 14:13:43.015787 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2448.solverstate
I0428 14:13:44.718228 30475 solver.cpp:330] Iteration 2448, Testing net (#0)
I0428 14:13:44.718253 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:13:48.593238 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:13:49.732093 30475 solver.cpp:397]     Test net output #0: accuracy = 0.224877
I0428 14:13:49.732129 30475 solver.cpp:397]     Test net output #1: loss = 3.22576 (* 1 = 3.22576 loss)
I0428 14:13:49.886715 30475 solver.cpp:218] Iteration 2448 (0.832142 iter/s, 14.4206s/12 iters), loss = 2.52829
I0428 14:13:49.886761 30475 solver.cpp:237]     Train net output #0: loss = 2.52829 (* 1 = 2.52829 loss)
I0428 14:13:49.886770 30475 sgd_solver.cpp:105] Iteration 2448, lr = 0.0061575
I0428 14:13:54.622357 30475 solver.cpp:218] Iteration 2460 (2.534 iter/s, 4.73559s/12 iters), loss = 2.65625
I0428 14:13:54.622400 30475 solver.cpp:237]     Train net output #0: loss = 2.65625 (* 1 = 2.65625 loss)
I0428 14:13:54.622408 30475 sgd_solver.cpp:105] Iteration 2460, lr = 0.00614288
I0428 14:14:00.339584 30475 solver.cpp:218] Iteration 2472 (2.09894 iter/s, 5.71718s/12 iters), loss = 2.44084
I0428 14:14:00.339630 30475 solver.cpp:237]     Train net output #0: loss = 2.44084 (* 1 = 2.44084 loss)
I0428 14:14:00.339639 30475 sgd_solver.cpp:105] Iteration 2472, lr = 0.0061283
I0428 14:14:05.906816 30475 solver.cpp:218] Iteration 2484 (2.15549 iter/s, 5.56718s/12 iters), loss = 2.40199
I0428 14:14:05.906860 30475 solver.cpp:237]     Train net output #0: loss = 2.40199 (* 1 = 2.40199 loss)
I0428 14:14:05.906868 30475 sgd_solver.cpp:105] Iteration 2484, lr = 0.00611375
I0428 14:14:11.565858 30475 solver.cpp:218] Iteration 2496 (2.12052 iter/s, 5.65899s/12 iters), loss = 2.00635
I0428 14:14:11.565907 30475 solver.cpp:237]     Train net output #0: loss = 2.00635 (* 1 = 2.00635 loss)
I0428 14:14:11.565915 30475 sgd_solver.cpp:105] Iteration 2496, lr = 0.00609923
I0428 14:14:17.251574 30475 solver.cpp:218] Iteration 2508 (2.11057 iter/s, 5.68567s/12 iters), loss = 2.60564
I0428 14:14:17.251718 30475 solver.cpp:237]     Train net output #0: loss = 2.60564 (* 1 = 2.60564 loss)
I0428 14:14:17.251727 30475 sgd_solver.cpp:105] Iteration 2508, lr = 0.00608475
I0428 14:14:22.827018 30475 solver.cpp:218] Iteration 2520 (2.15235 iter/s, 5.5753s/12 iters), loss = 2.25279
I0428 14:14:22.827054 30475 solver.cpp:237]     Train net output #0: loss = 2.25279 (* 1 = 2.25279 loss)
I0428 14:14:22.827061 30475 sgd_solver.cpp:105] Iteration 2520, lr = 0.0060703
I0428 14:14:26.329452 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:14:28.396855 30475 solver.cpp:218] Iteration 2532 (2.15448 iter/s, 5.5698s/12 iters), loss = 2.19422
I0428 14:14:28.396893 30475 solver.cpp:237]     Train net output #0: loss = 2.19422 (* 1 = 2.19422 loss)
I0428 14:14:28.396901 30475 sgd_solver.cpp:105] Iteration 2532, lr = 0.00605589
I0428 14:14:33.948745 30475 solver.cpp:218] Iteration 2544 (2.16144 iter/s, 5.55185s/12 iters), loss = 2.58591
I0428 14:14:33.948784 30475 solver.cpp:237]     Train net output #0: loss = 2.58591 (* 1 = 2.58591 loss)
I0428 14:14:33.948792 30475 sgd_solver.cpp:105] Iteration 2544, lr = 0.00604151
I0428 14:14:36.225100 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2550.caffemodel
I0428 14:14:38.425962 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2550.solverstate
I0428 14:14:41.167431 30475 solver.cpp:330] Iteration 2550, Testing net (#0)
I0428 14:14:41.167449 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:14:45.089079 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:14:46.290179 30475 solver.cpp:397]     Test net output #0: accuracy = 0.238971
I0428 14:14:46.290215 30475 solver.cpp:397]     Test net output #1: loss = 3.17772 (* 1 = 3.17772 loss)
I0428 14:14:48.425262 30475 solver.cpp:218] Iteration 2556 (0.828929 iter/s, 14.4765s/12 iters), loss = 2.50562
I0428 14:14:48.425360 30475 solver.cpp:237]     Train net output #0: loss = 2.50562 (* 1 = 2.50562 loss)
I0428 14:14:48.425369 30475 sgd_solver.cpp:105] Iteration 2556, lr = 0.00602717
I0428 14:14:54.048365 30475 solver.cpp:218] Iteration 2568 (2.13409 iter/s, 5.62301s/12 iters), loss = 2.44966
I0428 14:14:54.048410 30475 solver.cpp:237]     Train net output #0: loss = 2.44966 (* 1 = 2.44966 loss)
I0428 14:14:54.048419 30475 sgd_solver.cpp:105] Iteration 2568, lr = 0.00601286
I0428 14:14:59.866987 30475 solver.cpp:218] Iteration 2580 (2.06236 iter/s, 5.81857s/12 iters), loss = 2.75304
I0428 14:14:59.867029 30475 solver.cpp:237]     Train net output #0: loss = 2.75304 (* 1 = 2.75304 loss)
I0428 14:14:59.867038 30475 sgd_solver.cpp:105] Iteration 2580, lr = 0.00599858
I0428 14:15:05.521517 30475 solver.cpp:218] Iteration 2592 (2.12221 iter/s, 5.65448s/12 iters), loss = 2.38287
I0428 14:15:05.521564 30475 solver.cpp:237]     Train net output #0: loss = 2.38287 (* 1 = 2.38287 loss)
I0428 14:15:05.521572 30475 sgd_solver.cpp:105] Iteration 2592, lr = 0.00598434
I0428 14:15:11.158347 30475 solver.cpp:218] Iteration 2604 (2.12887 iter/s, 5.63679s/12 iters), loss = 2.01413
I0428 14:15:11.158382 30475 solver.cpp:237]     Train net output #0: loss = 2.01413 (* 1 = 2.01413 loss)
I0428 14:15:11.158388 30475 sgd_solver.cpp:105] Iteration 2604, lr = 0.00597013
I0428 14:15:16.796377 30475 solver.cpp:218] Iteration 2616 (2.12842 iter/s, 5.63799s/12 iters), loss = 2.38781
I0428 14:15:16.796424 30475 solver.cpp:237]     Train net output #0: loss = 2.38781 (* 1 = 2.38781 loss)
I0428 14:15:16.796432 30475 sgd_solver.cpp:105] Iteration 2616, lr = 0.00595596
I0428 14:15:22.474211 30475 solver.cpp:218] Iteration 2628 (2.1135 iter/s, 5.67778s/12 iters), loss = 2.38908
I0428 14:15:22.474372 30475 solver.cpp:237]     Train net output #0: loss = 2.38908 (* 1 = 2.38908 loss)
I0428 14:15:22.474381 30475 sgd_solver.cpp:105] Iteration 2628, lr = 0.00594182
I0428 14:15:22.967033 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:15:28.137831 30475 solver.cpp:218] Iteration 2640 (2.11885 iter/s, 5.66346s/12 iters), loss = 2.14555
I0428 14:15:28.137876 30475 solver.cpp:237]     Train net output #0: loss = 2.14555 (* 1 = 2.14555 loss)
I0428 14:15:28.137885 30475 sgd_solver.cpp:105] Iteration 2640, lr = 0.00592771
I0428 14:15:33.110028 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2652.caffemodel
I0428 14:15:36.435883 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2652.solverstate
I0428 14:15:39.091969 30475 solver.cpp:330] Iteration 2652, Testing net (#0)
I0428 14:15:39.091997 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:15:42.964069 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:15:44.216528 30475 solver.cpp:397]     Test net output #0: accuracy = 0.258578
I0428 14:15:44.216557 30475 solver.cpp:397]     Test net output #1: loss = 3.09392 (* 1 = 3.09392 loss)
I0428 14:15:44.378664 30475 solver.cpp:218] Iteration 2652 (0.738879 iter/s, 16.2408s/12 iters), loss = 2.38431
I0428 14:15:44.378713 30475 solver.cpp:237]     Train net output #0: loss = 2.38431 (* 1 = 2.38431 loss)
I0428 14:15:44.378722 30475 sgd_solver.cpp:105] Iteration 2652, lr = 0.00591364
I0428 14:15:49.073381 30475 solver.cpp:218] Iteration 2664 (2.55609 iter/s, 4.69467s/12 iters), loss = 2.07364
I0428 14:15:49.073426 30475 solver.cpp:237]     Train net output #0: loss = 2.07364 (* 1 = 2.07364 loss)
I0428 14:15:49.073436 30475 sgd_solver.cpp:105] Iteration 2664, lr = 0.0058996
I0428 14:15:54.606725 30475 solver.cpp:218] Iteration 2676 (2.16869 iter/s, 5.53329s/12 iters), loss = 1.91269
I0428 14:15:54.606840 30475 solver.cpp:237]     Train net output #0: loss = 1.91269 (* 1 = 1.91269 loss)
I0428 14:15:54.606850 30475 sgd_solver.cpp:105] Iteration 2676, lr = 0.00588559
I0428 14:16:00.229482 30475 solver.cpp:218] Iteration 2688 (2.13423 iter/s, 5.62264s/12 iters), loss = 2.07045
I0428 14:16:00.229522 30475 solver.cpp:237]     Train net output #0: loss = 2.07045 (* 1 = 2.07045 loss)
I0428 14:16:00.229530 30475 sgd_solver.cpp:105] Iteration 2688, lr = 0.00587162
I0428 14:16:05.857604 30475 solver.cpp:218] Iteration 2700 (2.13217 iter/s, 5.62808s/12 iters), loss = 2.17975
I0428 14:16:05.857650 30475 solver.cpp:237]     Train net output #0: loss = 2.17975 (* 1 = 2.17975 loss)
I0428 14:16:05.857657 30475 sgd_solver.cpp:105] Iteration 2700, lr = 0.00585768
I0428 14:16:11.503166 30475 solver.cpp:218] Iteration 2712 (2.12558 iter/s, 5.64552s/12 iters), loss = 1.95084
I0428 14:16:11.503209 30475 solver.cpp:237]     Train net output #0: loss = 1.95084 (* 1 = 1.95084 loss)
I0428 14:16:11.503217 30475 sgd_solver.cpp:105] Iteration 2712, lr = 0.00584377
I0428 14:16:17.157938 30475 solver.cpp:218] Iteration 2724 (2.12212 iter/s, 5.65473s/12 iters), loss = 1.91692
I0428 14:16:17.157982 30475 solver.cpp:237]     Train net output #0: loss = 1.91692 (* 1 = 1.91692 loss)
I0428 14:16:17.157991 30475 sgd_solver.cpp:105] Iteration 2724, lr = 0.0058299
I0428 14:16:19.955399 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:16:22.709650 30475 solver.cpp:218] Iteration 2736 (2.16151 iter/s, 5.55167s/12 iters), loss = 1.83597
I0428 14:16:22.709688 30475 solver.cpp:237]     Train net output #0: loss = 1.83597 (* 1 = 1.83597 loss)
I0428 14:16:22.709695 30475 sgd_solver.cpp:105] Iteration 2736, lr = 0.00581605
I0428 14:16:28.334868 30475 solver.cpp:218] Iteration 2748 (2.13327 iter/s, 5.62518s/12 iters), loss = 2.42303
I0428 14:16:28.335033 30475 solver.cpp:237]     Train net output #0: loss = 2.42303 (* 1 = 2.42303 loss)
I0428 14:16:28.335043 30475 sgd_solver.cpp:105] Iteration 2748, lr = 0.00580225
I0428 14:16:30.584703 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2754.caffemodel
I0428 14:16:34.708048 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2754.solverstate
I0428 14:16:38.747186 30475 solver.cpp:330] Iteration 2754, Testing net (#0)
I0428 14:16:38.747206 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:16:42.232904 30475 blocking_queue.cpp:49] Waiting for data
I0428 14:16:42.499567 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:16:43.771100 30475 solver.cpp:397]     Test net output #0: accuracy = 0.283701
I0428 14:16:43.771129 30475 solver.cpp:397]     Test net output #1: loss = 3.04075 (* 1 = 3.04075 loss)
I0428 14:16:45.897037 30475 solver.cpp:218] Iteration 2760 (0.683292 iter/s, 17.562s/12 iters), loss = 1.98344
I0428 14:16:45.897081 30475 solver.cpp:237]     Train net output #0: loss = 1.98344 (* 1 = 1.98344 loss)
I0428 14:16:45.897091 30475 sgd_solver.cpp:105] Iteration 2760, lr = 0.00578847
I0428 14:16:51.572456 30475 solver.cpp:218] Iteration 2772 (2.1144 iter/s, 5.67537s/12 iters), loss = 1.96795
I0428 14:16:51.572501 30475 solver.cpp:237]     Train net output #0: loss = 1.96795 (* 1 = 1.96795 loss)
I0428 14:16:51.572510 30475 sgd_solver.cpp:105] Iteration 2772, lr = 0.00577473
I0428 14:16:57.165939 30475 solver.cpp:218] Iteration 2784 (2.14537 iter/s, 5.59343s/12 iters), loss = 2.05272
I0428 14:16:57.165980 30475 solver.cpp:237]     Train net output #0: loss = 2.05272 (* 1 = 2.05272 loss)
I0428 14:16:57.165988 30475 sgd_solver.cpp:105] Iteration 2784, lr = 0.00576102
I0428 14:17:02.779253 30475 solver.cpp:218] Iteration 2796 (2.13779 iter/s, 5.61327s/12 iters), loss = 1.93346
I0428 14:17:02.779366 30475 solver.cpp:237]     Train net output #0: loss = 1.93346 (* 1 = 1.93346 loss)
I0428 14:17:02.779376 30475 sgd_solver.cpp:105] Iteration 2796, lr = 0.00574734
I0428 14:17:08.323168 30475 solver.cpp:218] Iteration 2808 (2.16458 iter/s, 5.5438s/12 iters), loss = 2.08431
I0428 14:17:08.323208 30475 solver.cpp:237]     Train net output #0: loss = 2.08431 (* 1 = 2.08431 loss)
I0428 14:17:08.323216 30475 sgd_solver.cpp:105] Iteration 2808, lr = 0.00573369
I0428 14:17:14.016957 30475 solver.cpp:218] Iteration 2820 (2.10758 iter/s, 5.69374s/12 iters), loss = 2.05613
I0428 14:17:14.016996 30475 solver.cpp:237]     Train net output #0: loss = 2.05613 (* 1 = 2.05613 loss)
I0428 14:17:14.017004 30475 sgd_solver.cpp:105] Iteration 2820, lr = 0.00572008
I0428 14:17:19.327049 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:17:19.675915 30475 solver.cpp:218] Iteration 2832 (2.12055 iter/s, 5.65892s/12 iters), loss = 1.82664
I0428 14:17:19.675954 30475 solver.cpp:237]     Train net output #0: loss = 1.82664 (* 1 = 1.82664 loss)
I0428 14:17:19.675962 30475 sgd_solver.cpp:105] Iteration 2832, lr = 0.0057065
I0428 14:17:25.329231 30475 solver.cpp:218] Iteration 2844 (2.12266 iter/s, 5.65328s/12 iters), loss = 1.77614
I0428 14:17:25.329272 30475 solver.cpp:237]     Train net output #0: loss = 1.77614 (* 1 = 1.77614 loss)
I0428 14:17:25.329280 30475 sgd_solver.cpp:105] Iteration 2844, lr = 0.00569295
I0428 14:17:30.301656 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2856.caffemodel
I0428 14:17:35.608665 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2856.solverstate
I0428 14:17:38.887089 30475 solver.cpp:330] Iteration 2856, Testing net (#0)
I0428 14:17:38.887107 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:17:42.686139 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:17:44.027586 30475 solver.cpp:397]     Test net output #0: accuracy = 0.292892
I0428 14:17:44.027612 30475 solver.cpp:397]     Test net output #1: loss = 2.91334 (* 1 = 2.91334 loss)
I0428 14:17:44.185616 30475 solver.cpp:218] Iteration 2856 (0.636389 iter/s, 18.8564s/12 iters), loss = 1.86623
I0428 14:17:44.185667 30475 solver.cpp:237]     Train net output #0: loss = 1.86623 (* 1 = 1.86623 loss)
I0428 14:17:44.185676 30475 sgd_solver.cpp:105] Iteration 2856, lr = 0.00567944
I0428 14:17:48.949188 30475 solver.cpp:218] Iteration 2868 (2.51916 iter/s, 4.76349s/12 iters), loss = 1.68139
I0428 14:17:48.949249 30475 solver.cpp:237]     Train net output #0: loss = 1.68139 (* 1 = 1.68139 loss)
I0428 14:17:48.949286 30475 sgd_solver.cpp:105] Iteration 2868, lr = 0.00566595
I0428 14:17:54.680940 30475 solver.cpp:218] Iteration 2880 (2.09362 iter/s, 5.73169s/12 iters), loss = 1.86104
I0428 14:17:54.680984 30475 solver.cpp:237]     Train net output #0: loss = 1.86104 (* 1 = 1.86104 loss)
I0428 14:17:54.680990 30475 sgd_solver.cpp:105] Iteration 2880, lr = 0.0056525
I0428 14:18:00.314446 30475 solver.cpp:218] Iteration 2892 (2.13013 iter/s, 5.63346s/12 iters), loss = 1.67834
I0428 14:18:00.314494 30475 solver.cpp:237]     Train net output #0: loss = 1.67834 (* 1 = 1.67834 loss)
I0428 14:18:00.314502 30475 sgd_solver.cpp:105] Iteration 2892, lr = 0.00563908
I0428 14:18:05.936285 30475 solver.cpp:218] Iteration 2904 (2.13455 iter/s, 5.62179s/12 iters), loss = 1.91542
I0428 14:18:05.936399 30475 solver.cpp:237]     Train net output #0: loss = 1.91542 (* 1 = 1.91542 loss)
I0428 14:18:05.936414 30475 sgd_solver.cpp:105] Iteration 2904, lr = 0.00562569
I0428 14:18:11.572134 30475 solver.cpp:218] Iteration 2916 (2.12927 iter/s, 5.63573s/12 iters), loss = 1.90911
I0428 14:18:11.572196 30475 solver.cpp:237]     Train net output #0: loss = 1.90911 (* 1 = 1.90911 loss)
I0428 14:18:11.572208 30475 sgd_solver.cpp:105] Iteration 2916, lr = 0.00561233
I0428 14:18:17.120147 30475 solver.cpp:218] Iteration 2928 (2.16296 iter/s, 5.54795s/12 iters), loss = 1.72401
I0428 14:18:17.120190 30475 solver.cpp:237]     Train net output #0: loss = 1.72401 (* 1 = 1.72401 loss)
I0428 14:18:17.120199 30475 sgd_solver.cpp:105] Iteration 2928, lr = 0.00559901
I0428 14:18:19.174989 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:18:22.782624 30475 solver.cpp:218] Iteration 2940 (2.11923 iter/s, 5.66243s/12 iters), loss = 1.7271
I0428 14:18:22.782663 30475 solver.cpp:237]     Train net output #0: loss = 1.7271 (* 1 = 1.7271 loss)
I0428 14:18:22.782675 30475 sgd_solver.cpp:105] Iteration 2940, lr = 0.00558572
I0428 14:18:28.515561 30475 solver.cpp:218] Iteration 2952 (2.09319 iter/s, 5.73289s/12 iters), loss = 1.62488
I0428 14:18:28.515619 30475 solver.cpp:237]     Train net output #0: loss = 1.62488 (* 1 = 1.62488 loss)
I0428 14:18:28.515632 30475 sgd_solver.cpp:105] Iteration 2952, lr = 0.00557245
I0428 14:18:30.835541 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2958.caffemodel
I0428 14:18:36.299306 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2958.solverstate
I0428 14:18:40.204174 30475 solver.cpp:330] Iteration 2958, Testing net (#0)
I0428 14:18:40.204201 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:18:43.829185 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:18:45.200896 30475 solver.cpp:397]     Test net output #0: accuracy = 0.302083
I0428 14:18:45.200929 30475 solver.cpp:397]     Test net output #1: loss = 2.85668 (* 1 = 2.85668 loss)
I0428 14:18:47.322873 30475 solver.cpp:218] Iteration 2964 (0.638051 iter/s, 18.8073s/12 iters), loss = 1.82298
I0428 14:18:47.322918 30475 solver.cpp:237]     Train net output #0: loss = 1.82298 (* 1 = 1.82298 loss)
I0428 14:18:47.322927 30475 sgd_solver.cpp:105] Iteration 2964, lr = 0.00555922
I0428 14:18:52.966436 30475 solver.cpp:218] Iteration 2976 (2.12634 iter/s, 5.64351s/12 iters), loss = 1.72657
I0428 14:18:52.966482 30475 solver.cpp:237]     Train net output #0: loss = 1.72657 (* 1 = 1.72657 loss)
I0428 14:18:52.966492 30475 sgd_solver.cpp:105] Iteration 2976, lr = 0.00554603
I0428 14:18:58.401134 30475 solver.cpp:218] Iteration 2988 (2.20806 iter/s, 5.43465s/12 iters), loss = 1.60527
I0428 14:18:58.401180 30475 solver.cpp:237]     Train net output #0: loss = 1.60527 (* 1 = 1.60527 loss)
I0428 14:18:58.401188 30475 sgd_solver.cpp:105] Iteration 2988, lr = 0.00553286
I0428 14:19:04.023911 30475 solver.cpp:218] Iteration 3000 (2.1342 iter/s, 5.62272s/12 iters), loss = 1.75224
I0428 14:19:04.023955 30475 solver.cpp:237]     Train net output #0: loss = 1.75224 (* 1 = 1.75224 loss)
I0428 14:19:04.023963 30475 sgd_solver.cpp:105] Iteration 3000, lr = 0.00551972
I0428 14:19:09.667624 30475 solver.cpp:218] Iteration 3012 (2.12628 iter/s, 5.64366s/12 iters), loss = 1.7524
I0428 14:19:09.667757 30475 solver.cpp:237]     Train net output #0: loss = 1.7524 (* 1 = 1.7524 loss)
I0428 14:19:09.667765 30475 sgd_solver.cpp:105] Iteration 3012, lr = 0.00550662
I0428 14:19:15.415143 30475 solver.cpp:218] Iteration 3024 (2.08791 iter/s, 5.74738s/12 iters), loss = 1.671
I0428 14:19:15.415184 30475 solver.cpp:237]     Train net output #0: loss = 1.671 (* 1 = 1.671 loss)
I0428 14:19:15.415190 30475 sgd_solver.cpp:105] Iteration 3024, lr = 0.00549354
I0428 14:19:19.888850 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:19:21.083775 30475 solver.cpp:218] Iteration 3036 (2.11693 iter/s, 5.66858s/12 iters), loss = 1.74515
I0428 14:19:21.083818 30475 solver.cpp:237]     Train net output #0: loss = 1.74515 (* 1 = 1.74515 loss)
I0428 14:19:21.083827 30475 sgd_solver.cpp:105] Iteration 3036, lr = 0.0054805
I0428 14:19:26.724475 30475 solver.cpp:218] Iteration 3048 (2.12741 iter/s, 5.64066s/12 iters), loss = 1.70258
I0428 14:19:26.724514 30475 solver.cpp:237]     Train net output #0: loss = 1.70258 (* 1 = 1.70258 loss)
I0428 14:19:26.724521 30475 sgd_solver.cpp:105] Iteration 3048, lr = 0.00546749
I0428 14:19:31.865406 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3060.caffemodel
I0428 14:19:36.125360 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3060.solverstate
I0428 14:19:39.444624 30475 solver.cpp:330] Iteration 3060, Testing net (#0)
I0428 14:19:39.444646 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:19:43.097196 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:19:44.547430 30475 solver.cpp:397]     Test net output #0: accuracy = 0.303922
I0428 14:19:44.547457 30475 solver.cpp:397]     Test net output #1: loss = 3.0859 (* 1 = 3.0859 loss)
I0428 14:19:44.704659 30475 solver.cpp:218] Iteration 3060 (0.667402 iter/s, 17.9802s/12 iters), loss = 1.51239
I0428 14:19:44.704700 30475 solver.cpp:237]     Train net output #0: loss = 1.51239 (* 1 = 1.51239 loss)
I0428 14:19:44.704708 30475 sgd_solver.cpp:105] Iteration 3060, lr = 0.00545451
I0428 14:19:49.403373 30475 solver.cpp:218] Iteration 3072 (2.55392 iter/s, 4.69867s/12 iters), loss = 1.86407
I0428 14:19:49.403419 30475 solver.cpp:237]     Train net output #0: loss = 1.86407 (* 1 = 1.86407 loss)
I0428 14:19:49.403426 30475 sgd_solver.cpp:105] Iteration 3072, lr = 0.00544156
I0428 14:19:54.860939 30475 solver.cpp:218] Iteration 3084 (2.1988 iter/s, 5.45751s/12 iters), loss = 1.72112
I0428 14:19:54.860980 30475 solver.cpp:237]     Train net output #0: loss = 1.72112 (* 1 = 1.72112 loss)
I0428 14:19:54.860987 30475 sgd_solver.cpp:105] Iteration 3084, lr = 0.00542864
I0428 14:20:00.510658 30475 solver.cpp:218] Iteration 3096 (2.12402 iter/s, 5.64968s/12 iters), loss = 1.56955
I0428 14:20:00.510699 30475 solver.cpp:237]     Train net output #0: loss = 1.56955 (* 1 = 1.56955 loss)
I0428 14:20:00.510707 30475 sgd_solver.cpp:105] Iteration 3096, lr = 0.00541575
I0428 14:20:06.052079 30475 solver.cpp:218] Iteration 3108 (2.16553 iter/s, 5.54138s/12 iters), loss = 1.45083
I0428 14:20:06.052119 30475 solver.cpp:237]     Train net output #0: loss = 1.45083 (* 1 = 1.45083 loss)
I0428 14:20:06.052129 30475 sgd_solver.cpp:105] Iteration 3108, lr = 0.00540289
I0428 14:20:11.580945 30475 solver.cpp:218] Iteration 3120 (2.17045 iter/s, 5.52882s/12 iters), loss = 1.73686
I0428 14:20:11.580991 30475 solver.cpp:237]     Train net output #0: loss = 1.73686 (* 1 = 1.73686 loss)
I0428 14:20:11.580999 30475 sgd_solver.cpp:105] Iteration 3120, lr = 0.00539006
I0428 14:20:17.220758 30475 solver.cpp:218] Iteration 3132 (2.12775 iter/s, 5.63976s/12 iters), loss = 1.61402
I0428 14:20:17.220975 30475 solver.cpp:237]     Train net output #0: loss = 1.61402 (* 1 = 1.61402 loss)
I0428 14:20:17.220990 30475 sgd_solver.cpp:105] Iteration 3132, lr = 0.00537727
I0428 14:20:18.431934 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:20:22.872517 30475 solver.cpp:218] Iteration 3144 (2.12331 iter/s, 5.65154s/12 iters), loss = 1.70316
I0428 14:20:22.872557 30475 solver.cpp:237]     Train net output #0: loss = 1.70316 (* 1 = 1.70316 loss)
I0428 14:20:22.872566 30475 sgd_solver.cpp:105] Iteration 3144, lr = 0.0053645
I0428 14:20:28.493212 30475 solver.cpp:218] Iteration 3156 (2.13498 iter/s, 5.62065s/12 iters), loss = 1.27695
I0428 14:20:28.493249 30475 solver.cpp:237]     Train net output #0: loss = 1.27695 (* 1 = 1.27695 loss)
I0428 14:20:28.493257 30475 sgd_solver.cpp:105] Iteration 3156, lr = 0.00535176
I0428 14:20:30.771679 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3162.caffemodel
I0428 14:20:37.000267 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3162.solverstate
I0428 14:20:40.162891 30475 solver.cpp:330] Iteration 3162, Testing net (#0)
I0428 14:20:40.162921 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:20:43.834600 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:20:45.343334 30475 solver.cpp:397]     Test net output #0: accuracy = 0.318627
I0428 14:20:45.343367 30475 solver.cpp:397]     Test net output #1: loss = 2.98523 (* 1 = 2.98523 loss)
I0428 14:20:47.497444 30475 solver.cpp:218] Iteration 3168 (0.631439 iter/s, 19.0042s/12 iters), loss = 1.71453
I0428 14:20:47.497565 30475 solver.cpp:237]     Train net output #0: loss = 1.71453 (* 1 = 1.71453 loss)
I0428 14:20:47.497575 30475 sgd_solver.cpp:105] Iteration 3168, lr = 0.00533906
I0428 14:20:53.046548 30475 solver.cpp:218] Iteration 3180 (2.16256 iter/s, 5.54897s/12 iters), loss = 1.27635
I0428 14:20:53.046592 30475 solver.cpp:237]     Train net output #0: loss = 1.27635 (* 1 = 1.27635 loss)
I0428 14:20:53.046607 30475 sgd_solver.cpp:105] Iteration 3180, lr = 0.00532638
I0428 14:20:58.665772 30475 solver.cpp:218] Iteration 3192 (2.13555 iter/s, 5.61917s/12 iters), loss = 1.6689
I0428 14:20:58.665818 30475 solver.cpp:237]     Train net output #0: loss = 1.6689 (* 1 = 1.6689 loss)
I0428 14:20:58.665827 30475 sgd_solver.cpp:105] Iteration 3192, lr = 0.00531374
I0428 14:21:04.186355 30475 solver.cpp:218] Iteration 3204 (2.17371 iter/s, 5.52053s/12 iters), loss = 1.34996
I0428 14:21:04.186403 30475 solver.cpp:237]     Train net output #0: loss = 1.34996 (* 1 = 1.34996 loss)
I0428 14:21:04.186411 30475 sgd_solver.cpp:105] Iteration 3204, lr = 0.00530112
I0428 14:21:09.719189 30475 solver.cpp:218] Iteration 3216 (2.16889 iter/s, 5.53278s/12 iters), loss = 1.57578
I0428 14:21:09.719228 30475 solver.cpp:237]     Train net output #0: loss = 1.57578 (* 1 = 1.57578 loss)
I0428 14:21:09.719235 30475 sgd_solver.cpp:105] Iteration 3216, lr = 0.00528853
I0428 14:21:15.353379 30475 solver.cpp:218] Iteration 3228 (2.12987 iter/s, 5.63414s/12 iters), loss = 1.65136
I0428 14:21:15.353423 30475 solver.cpp:237]     Train net output #0: loss = 1.65136 (* 1 = 1.65136 loss)
I0428 14:21:15.353431 30475 sgd_solver.cpp:105] Iteration 3228, lr = 0.00527598
I0428 14:21:18.967676 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:21:21.000855 30475 solver.cpp:218] Iteration 3240 (2.12486 iter/s, 5.64743s/12 iters), loss = 1.2748
I0428 14:21:21.000896 30475 solver.cpp:237]     Train net output #0: loss = 1.2748 (* 1 = 1.2748 loss)
I0428 14:21:21.000905 30475 sgd_solver.cpp:105] Iteration 3240, lr = 0.00526345
I0428 14:21:26.624411 30475 solver.cpp:218] Iteration 3252 (2.1339 iter/s, 5.6235s/12 iters), loss = 1.39613
I0428 14:21:26.624466 30475 solver.cpp:237]     Train net output #0: loss = 1.39613 (* 1 = 1.39613 loss)
I0428 14:21:26.624477 30475 sgd_solver.cpp:105] Iteration 3252, lr = 0.00525095
I0428 14:21:31.712721 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3264.caffemodel
I0428 14:21:35.591545 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3264.solverstate
I0428 14:21:41.353631 30475 solver.cpp:330] Iteration 3264, Testing net (#0)
I0428 14:21:41.353648 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:21:44.831133 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:21:46.277829 30475 solver.cpp:397]     Test net output #0: accuracy = 0.329657
I0428 14:21:46.277870 30475 solver.cpp:397]     Test net output #1: loss = 3.01945 (* 1 = 3.01945 loss)
I0428 14:21:46.436893 30475 solver.cpp:218] Iteration 3264 (0.605679 iter/s, 19.8125s/12 iters), loss = 1.56431
I0428 14:21:46.436937 30475 solver.cpp:237]     Train net output #0: loss = 1.56431 (* 1 = 1.56431 loss)
I0428 14:21:46.436946 30475 sgd_solver.cpp:105] Iteration 3264, lr = 0.00523849
I0428 14:21:51.126935 30475 solver.cpp:218] Iteration 3276 (2.55864 iter/s, 4.68999s/12 iters), loss = 1.3323
I0428 14:21:51.127050 30475 solver.cpp:237]     Train net output #0: loss = 1.3323 (* 1 = 1.3323 loss)
I0428 14:21:51.127058 30475 sgd_solver.cpp:105] Iteration 3276, lr = 0.00522605
I0428 14:21:56.760524 30475 solver.cpp:218] Iteration 3288 (2.13013 iter/s, 5.63347s/12 iters), loss = 1.4895
I0428 14:21:56.760568 30475 solver.cpp:237]     Train net output #0: loss = 1.4895 (* 1 = 1.4895 loss)
I0428 14:21:56.760576 30475 sgd_solver.cpp:105] Iteration 3288, lr = 0.00521364
I0428 14:22:02.394909 30475 solver.cpp:218] Iteration 3300 (2.1298 iter/s, 5.63434s/12 iters), loss = 1.38324
I0428 14:22:02.394954 30475 solver.cpp:237]     Train net output #0: loss = 1.38324 (* 1 = 1.38324 loss)
I0428 14:22:02.394963 30475 sgd_solver.cpp:105] Iteration 3300, lr = 0.00520126
I0428 14:22:07.933614 30475 solver.cpp:218] Iteration 3312 (2.16659 iter/s, 5.53866s/12 iters), loss = 1.28969
I0428 14:22:07.933658 30475 solver.cpp:237]     Train net output #0: loss = 1.28969 (* 1 = 1.28969 loss)
I0428 14:22:07.933667 30475 sgd_solver.cpp:105] Iteration 3312, lr = 0.00518892
I0428 14:22:13.647620 30475 solver.cpp:218] Iteration 3324 (2.10012 iter/s, 5.71395s/12 iters), loss = 1.45002
I0428 14:22:13.647680 30475 solver.cpp:237]     Train net output #0: loss = 1.45002 (* 1 = 1.45002 loss)
I0428 14:22:13.647693 30475 sgd_solver.cpp:105] Iteration 3324, lr = 0.0051766
I0428 14:22:19.282678 30475 solver.cpp:218] Iteration 3336 (2.12955 iter/s, 5.635s/12 iters), loss = 1.7144
I0428 14:22:19.282718 30475 solver.cpp:237]     Train net output #0: loss = 1.7144 (* 1 = 1.7144 loss)
I0428 14:22:19.282725 30475 sgd_solver.cpp:105] Iteration 3336, lr = 0.00516431
I0428 14:22:19.803906 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:22:24.842067 30475 solver.cpp:218] Iteration 3348 (2.15853 iter/s, 5.55934s/12 iters), loss = 1.17599
I0428 14:22:24.842217 30475 solver.cpp:237]     Train net output #0: loss = 1.17599 (* 1 = 1.17599 loss)
I0428 14:22:24.842227 30475 sgd_solver.cpp:105] Iteration 3348, lr = 0.00515204
I0428 14:22:30.490586 30475 solver.cpp:218] Iteration 3360 (2.12451 iter/s, 5.64836s/12 iters), loss = 1.40069
I0428 14:22:30.490638 30475 solver.cpp:237]     Train net output #0: loss = 1.40069 (* 1 = 1.40069 loss)
I0428 14:22:30.490646 30475 sgd_solver.cpp:105] Iteration 3360, lr = 0.00513981
I0428 14:22:32.750787 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3366.caffemodel
I0428 14:22:39.525384 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3366.solverstate
I0428 14:22:43.180836 30475 solver.cpp:330] Iteration 3366, Testing net (#0)
I0428 14:22:43.180866 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:22:46.748929 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:22:48.342278 30475 solver.cpp:397]     Test net output #0: accuracy = 0.329044
I0428 14:22:48.342322 30475 solver.cpp:397]     Test net output #1: loss = 3.07972 (* 1 = 3.07972 loss)
I0428 14:22:50.438844 30475 solver.cpp:218] Iteration 3372 (0.601557 iter/s, 19.9482s/12 iters), loss = 1.40162
I0428 14:22:50.438885 30475 solver.cpp:237]     Train net output #0: loss = 1.40162 (* 1 = 1.40162 loss)
I0428 14:22:50.438894 30475 sgd_solver.cpp:105] Iteration 3372, lr = 0.00512761
I0428 14:22:56.115597 30475 solver.cpp:218] Iteration 3384 (2.1139 iter/s, 5.67671s/12 iters), loss = 1.11876
I0428 14:22:56.115689 30475 solver.cpp:237]     Train net output #0: loss = 1.11876 (* 1 = 1.11876 loss)
I0428 14:22:56.115697 30475 sgd_solver.cpp:105] Iteration 3384, lr = 0.00511544
I0428 14:23:01.741235 30475 solver.cpp:218] Iteration 3396 (2.13313 iter/s, 5.62554s/12 iters), loss = 1.3233
I0428 14:23:01.741281 30475 solver.cpp:237]     Train net output #0: loss = 1.3233 (* 1 = 1.3233 loss)
I0428 14:23:01.741290 30475 sgd_solver.cpp:105] Iteration 3396, lr = 0.00510329
I0428 14:23:07.254914 30475 solver.cpp:218] Iteration 3408 (2.17643 iter/s, 5.51363s/12 iters), loss = 1.22107
I0428 14:23:07.254959 30475 solver.cpp:237]     Train net output #0: loss = 1.22107 (* 1 = 1.22107 loss)
I0428 14:23:07.254968 30475 sgd_solver.cpp:105] Iteration 3408, lr = 0.00509117
I0428 14:23:12.881182 30475 solver.cpp:218] Iteration 3420 (2.13287 iter/s, 5.62622s/12 iters), loss = 1.27301
I0428 14:23:12.881227 30475 solver.cpp:237]     Train net output #0: loss = 1.27301 (* 1 = 1.27301 loss)
I0428 14:23:12.881235 30475 sgd_solver.cpp:105] Iteration 3420, lr = 0.00507909
I0428 14:23:18.514252 30475 solver.cpp:218] Iteration 3432 (2.1303 iter/s, 5.63302s/12 iters), loss = 1.23906
I0428 14:23:18.514290 30475 solver.cpp:237]     Train net output #0: loss = 1.23906 (* 1 = 1.23906 loss)
I0428 14:23:18.514299 30475 sgd_solver.cpp:105] Iteration 3432, lr = 0.00506703
I0428 14:23:21.439915 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:23:24.154453 30475 solver.cpp:218] Iteration 3444 (2.1276 iter/s, 5.64016s/12 iters), loss = 1.16984
I0428 14:23:24.154501 30475 solver.cpp:237]     Train net output #0: loss = 1.16984 (* 1 = 1.16984 loss)
I0428 14:23:24.154510 30475 sgd_solver.cpp:105] Iteration 3444, lr = 0.005055
I0428 14:23:29.770361 30475 solver.cpp:218] Iteration 3456 (2.13681 iter/s, 5.61585s/12 iters), loss = 1.42141
I0428 14:23:29.770947 30475 solver.cpp:237]     Train net output #0: loss = 1.42141 (* 1 = 1.42141 loss)
I0428 14:23:29.770958 30475 sgd_solver.cpp:105] Iteration 3456, lr = 0.005043
I0428 14:23:34.868324 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3468.caffemodel
I0428 14:23:38.399533 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3468.solverstate
I0428 14:23:44.127763 30475 solver.cpp:330] Iteration 3468, Testing net (#0)
I0428 14:23:44.127786 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:23:44.588934 30475 blocking_queue.cpp:49] Waiting for data
I0428 14:23:47.615073 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:23:49.261189 30475 solver.cpp:397]     Test net output #0: accuracy = 0.33027
I0428 14:23:49.261225 30475 solver.cpp:397]     Test net output #1: loss = 3.03261 (* 1 = 3.03261 loss)
I0428 14:23:49.422099 30475 solver.cpp:218] Iteration 3468 (0.61065 iter/s, 19.6512s/12 iters), loss = 1.17466
I0428 14:23:49.422153 30475 solver.cpp:237]     Train net output #0: loss = 1.17466 (* 1 = 1.17466 loss)
I0428 14:23:49.422163 30475 sgd_solver.cpp:105] Iteration 3468, lr = 0.00503102
I0428 14:23:54.050361 30475 solver.cpp:218] Iteration 3480 (2.5928 iter/s, 4.6282s/12 iters), loss = 1.05456
I0428 14:23:54.050408 30475 solver.cpp:237]     Train net output #0: loss = 1.05456 (* 1 = 1.05456 loss)
I0428 14:23:54.050416 30475 sgd_solver.cpp:105] Iteration 3480, lr = 0.00501908
I0428 14:23:59.668889 30475 solver.cpp:218] Iteration 3492 (2.13581 iter/s, 5.61847s/12 iters), loss = 1.43321
I0428 14:23:59.668934 30475 solver.cpp:237]     Train net output #0: loss = 1.43321 (* 1 = 1.43321 loss)
I0428 14:23:59.668943 30475 sgd_solver.cpp:105] Iteration 3492, lr = 0.00500716
I0428 14:24:05.266897 30475 solver.cpp:218] Iteration 3504 (2.14364 iter/s, 5.59796s/12 iters), loss = 1.07513
I0428 14:24:05.267035 30475 solver.cpp:237]     Train net output #0: loss = 1.07513 (* 1 = 1.07513 loss)
I0428 14:24:05.267045 30475 sgd_solver.cpp:105] Iteration 3504, lr = 0.00499527
I0428 14:24:10.798852 30475 solver.cpp:218] Iteration 3516 (2.16927 iter/s, 5.53181s/12 iters), loss = 1.21359
I0428 14:24:10.798897 30475 solver.cpp:237]     Train net output #0: loss = 1.21359 (* 1 = 1.21359 loss)
I0428 14:24:10.798907 30475 sgd_solver.cpp:105] Iteration 3516, lr = 0.00498341
I0428 14:24:16.439337 30475 solver.cpp:218] Iteration 3528 (2.12749 iter/s, 5.64044s/12 iters), loss = 1.19396
I0428 14:24:16.439376 30475 solver.cpp:237]     Train net output #0: loss = 1.19396 (* 1 = 1.19396 loss)
I0428 14:24:16.439385 30475 sgd_solver.cpp:105] Iteration 3528, lr = 0.00497158
I0428 14:24:21.715214 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:24:22.036396 30475 solver.cpp:218] Iteration 3540 (2.144 iter/s, 5.59702s/12 iters), loss = 1.07549
I0428 14:24:22.036437 30475 solver.cpp:237]     Train net output #0: loss = 1.07549 (* 1 = 1.07549 loss)
I0428 14:24:22.036445 30475 sgd_solver.cpp:105] Iteration 3540, lr = 0.00495978
I0428 14:24:27.782207 30475 solver.cpp:218] Iteration 3552 (2.0885 iter/s, 5.74576s/12 iters), loss = 0.993678
I0428 14:24:27.782246 30475 solver.cpp:237]     Train net output #0: loss = 0.993678 (* 1 = 0.993678 loss)
I0428 14:24:27.782254 30475 sgd_solver.cpp:105] Iteration 3552, lr = 0.004948
I0428 14:24:33.429718 30475 solver.cpp:218] Iteration 3564 (2.12485 iter/s, 5.64747s/12 iters), loss = 1.02287
I0428 14:24:33.429760 30475 solver.cpp:237]     Train net output #0: loss = 1.02287 (* 1 = 1.02287 loss)
I0428 14:24:33.429769 30475 sgd_solver.cpp:105] Iteration 3564, lr = 0.00493626
I0428 14:24:35.688050 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3570.caffemodel
I0428 14:24:39.684064 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3570.solverstate
I0428 14:24:41.644208 30475 solver.cpp:330] Iteration 3570, Testing net (#0)
I0428 14:24:41.644230 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:24:45.090868 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:24:46.810384 30475 solver.cpp:397]     Test net output #0: accuracy = 0.351103
I0428 14:24:46.810425 30475 solver.cpp:397]     Test net output #1: loss = 3.03184 (* 1 = 3.03184 loss)
I0428 14:24:48.873322 30475 solver.cpp:218] Iteration 3576 (0.777022 iter/s, 15.4436s/12 iters), loss = 0.927345
I0428 14:24:48.873370 30475 solver.cpp:237]     Train net output #0: loss = 0.927345 (* 1 = 0.927345 loss)
I0428 14:24:48.873379 30475 sgd_solver.cpp:105] Iteration 3576, lr = 0.00492454
I0428 14:24:54.505141 30475 solver.cpp:218] Iteration 3588 (2.13077 iter/s, 5.63176s/12 iters), loss = 1.33428
I0428 14:24:54.505185 30475 solver.cpp:237]     Train net output #0: loss = 1.33428 (* 1 = 1.33428 loss)
I0428 14:24:54.505193 30475 sgd_solver.cpp:105] Iteration 3588, lr = 0.00491284
I0428 14:25:00.130635 30475 solver.cpp:218] Iteration 3600 (2.13316 iter/s, 5.62545s/12 iters), loss = 1.02188
I0428 14:25:00.130681 30475 solver.cpp:237]     Train net output #0: loss = 1.02188 (* 1 = 1.02188 loss)
I0428 14:25:00.130689 30475 sgd_solver.cpp:105] Iteration 3600, lr = 0.00490118
I0428 14:25:05.552574 30475 solver.cpp:218] Iteration 3612 (2.21325 iter/s, 5.42189s/12 iters), loss = 1.07031
I0428 14:25:05.552618 30475 solver.cpp:237]     Train net output #0: loss = 1.07031 (* 1 = 1.07031 loss)
I0428 14:25:05.552625 30475 sgd_solver.cpp:105] Iteration 3612, lr = 0.00488954
I0428 14:25:11.183001 30475 solver.cpp:218] Iteration 3624 (2.1313 iter/s, 5.63037s/12 iters), loss = 0.929249
I0428 14:25:11.183166 30475 solver.cpp:237]     Train net output #0: loss = 0.929249 (* 1 = 0.929249 loss)
I0428 14:25:11.183176 30475 sgd_solver.cpp:105] Iteration 3624, lr = 0.00487793
I0428 14:25:16.725766 30475 solver.cpp:218] Iteration 3636 (2.16505 iter/s, 5.5426s/12 iters), loss = 1.14724
I0428 14:25:16.725806 30475 solver.cpp:237]     Train net output #0: loss = 1.14724 (* 1 = 1.14724 loss)
I0428 14:25:16.725814 30475 sgd_solver.cpp:105] Iteration 3636, lr = 0.00486635
I0428 14:25:18.816377 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:25:22.389219 30475 solver.cpp:218] Iteration 3648 (2.11887 iter/s, 5.6634s/12 iters), loss = 1.00805
I0428 14:25:22.389263 30475 solver.cpp:237]     Train net output #0: loss = 1.00805 (* 1 = 1.00805 loss)
I0428 14:25:22.389271 30475 sgd_solver.cpp:105] Iteration 3648, lr = 0.0048548
I0428 14:25:28.033924 30475 solver.cpp:218] Iteration 3660 (2.12591 iter/s, 5.64465s/12 iters), loss = 1.00746
I0428 14:25:28.033965 30475 solver.cpp:237]     Train net output #0: loss = 1.00746 (* 1 = 1.00746 loss)
I0428 14:25:28.033975 30475 sgd_solver.cpp:105] Iteration 3660, lr = 0.00484327
I0428 14:25:33.153620 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3672.caffemodel
I0428 14:25:36.124631 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3672.solverstate
I0428 14:25:38.738801 30475 solver.cpp:330] Iteration 3672, Testing net (#0)
I0428 14:25:38.738817 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:25:42.137457 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:25:43.871604 30475 solver.cpp:397]     Test net output #0: accuracy = 0.340074
I0428 14:25:43.871634 30475 solver.cpp:397]     Test net output #1: loss = 3.14593 (* 1 = 3.14593 loss)
I0428 14:25:44.023959 30475 solver.cpp:218] Iteration 3672 (0.750468 iter/s, 15.99s/12 iters), loss = 0.978084
I0428 14:25:44.024009 30475 solver.cpp:237]     Train net output #0: loss = 0.978084 (* 1 = 0.978084 loss)
I0428 14:25:44.024016 30475 sgd_solver.cpp:105] Iteration 3672, lr = 0.00483177
I0428 14:25:48.792234 30475 solver.cpp:218] Iteration 3684 (2.51666 iter/s, 4.76823s/12 iters), loss = 1.31665
I0428 14:25:48.792275 30475 solver.cpp:237]     Train net output #0: loss = 1.31665 (* 1 = 1.31665 loss)
I0428 14:25:48.792284 30475 sgd_solver.cpp:105] Iteration 3684, lr = 0.0048203
I0428 14:25:54.382972 30475 solver.cpp:218] Iteration 3696 (2.14642 iter/s, 5.59069s/12 iters), loss = 0.817142
I0428 14:25:54.383013 30475 solver.cpp:237]     Train net output #0: loss = 0.817142 (* 1 = 0.817142 loss)
I0428 14:25:54.383023 30475 sgd_solver.cpp:105] Iteration 3696, lr = 0.00480886
I0428 14:25:59.987680 30475 solver.cpp:218] Iteration 3708 (2.14108 iter/s, 5.60466s/12 iters), loss = 1.13032
I0428 14:25:59.987721 30475 solver.cpp:237]     Train net output #0: loss = 1.13032 (* 1 = 1.13032 loss)
I0428 14:25:59.987730 30475 sgd_solver.cpp:105] Iteration 3708, lr = 0.00479744
I0428 14:26:05.692688 30475 solver.cpp:218] Iteration 3720 (2.10343 iter/s, 5.70496s/12 iters), loss = 1.19061
I0428 14:26:05.692728 30475 solver.cpp:237]     Train net output #0: loss = 1.19061 (* 1 = 1.19061 loss)
I0428 14:26:05.692736 30475 sgd_solver.cpp:105] Iteration 3720, lr = 0.00478605
I0428 14:26:11.335464 30475 solver.cpp:218] Iteration 3732 (2.12663 iter/s, 5.64273s/12 iters), loss = 1.14597
I0428 14:26:11.335508 30475 solver.cpp:237]     Train net output #0: loss = 1.14597 (* 1 = 1.14597 loss)
I0428 14:26:11.335517 30475 sgd_solver.cpp:105] Iteration 3732, lr = 0.00477469
I0428 14:26:15.852455 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:26:17.020193 30475 solver.cpp:218] Iteration 3744 (2.11094 iter/s, 5.68468s/12 iters), loss = 1.14218
I0428 14:26:17.020238 30475 solver.cpp:237]     Train net output #0: loss = 1.14218 (* 1 = 1.14218 loss)
I0428 14:26:17.020247 30475 sgd_solver.cpp:105] Iteration 3744, lr = 0.00476335
I0428 14:26:22.561194 30475 solver.cpp:218] Iteration 3756 (2.16569 iter/s, 5.54095s/12 iters), loss = 1.04354
I0428 14:26:22.561234 30475 solver.cpp:237]     Train net output #0: loss = 1.04354 (* 1 = 1.04354 loss)
I0428 14:26:22.561242 30475 sgd_solver.cpp:105] Iteration 3756, lr = 0.00475204
I0428 14:26:28.204639 30475 solver.cpp:218] Iteration 3768 (2.12638 iter/s, 5.6434s/12 iters), loss = 0.839063
I0428 14:26:28.204679 30475 solver.cpp:237]     Train net output #0: loss = 0.839063 (* 1 = 0.839063 loss)
I0428 14:26:28.204686 30475 sgd_solver.cpp:105] Iteration 3768, lr = 0.00474076
I0428 14:26:30.512039 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3774.caffemodel
I0428 14:26:32.716464 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3774.solverstate
I0428 14:26:35.039976 30475 solver.cpp:330] Iteration 3774, Testing net (#0)
I0428 14:26:35.039997 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:26:38.448596 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:26:40.233521 30475 solver.cpp:397]     Test net output #0: accuracy = 0.349877
I0428 14:26:40.233569 30475 solver.cpp:397]     Test net output #1: loss = 3.12671 (* 1 = 3.12671 loss)
I0428 14:26:42.268496 30475 solver.cpp:218] Iteration 3780 (0.853253 iter/s, 14.0638s/12 iters), loss = 1.28999
I0428 14:26:42.268543 30475 solver.cpp:237]     Train net output #0: loss = 1.28999 (* 1 = 1.28999 loss)
I0428 14:26:42.268550 30475 sgd_solver.cpp:105] Iteration 3780, lr = 0.00472951
I0428 14:26:47.915606 30475 solver.cpp:218] Iteration 3792 (2.125 iter/s, 5.64706s/12 iters), loss = 1.00165
I0428 14:26:47.915999 30475 solver.cpp:237]     Train net output #0: loss = 1.00165 (* 1 = 1.00165 loss)
I0428 14:26:47.916010 30475 sgd_solver.cpp:105] Iteration 3792, lr = 0.00471828
I0428 14:26:53.549742 30475 solver.cpp:218] Iteration 3804 (2.13002 iter/s, 5.63374s/12 iters), loss = 0.991054
I0428 14:26:53.549789 30475 solver.cpp:237]     Train net output #0: loss = 0.991054 (* 1 = 0.991054 loss)
I0428 14:26:53.549798 30475 sgd_solver.cpp:105] Iteration 3804, lr = 0.00470707
I0428 14:26:59.170874 30475 solver.cpp:218] Iteration 3816 (2.13482 iter/s, 5.62108s/12 iters), loss = 0.855912
I0428 14:26:59.170917 30475 solver.cpp:237]     Train net output #0: loss = 0.855912 (* 1 = 0.855912 loss)
I0428 14:26:59.170923 30475 sgd_solver.cpp:105] Iteration 3816, lr = 0.0046959
I0428 14:27:04.786991 30475 solver.cpp:218] Iteration 3828 (2.13672 iter/s, 5.61607s/12 iters), loss = 1.00258
I0428 14:27:04.787030 30475 solver.cpp:237]     Train net output #0: loss = 1.00258 (* 1 = 1.00258 loss)
I0428 14:27:04.787039 30475 sgd_solver.cpp:105] Iteration 3828, lr = 0.00468475
I0428 14:27:10.440866 30475 solver.cpp:218] Iteration 3840 (2.12246 iter/s, 5.65383s/12 iters), loss = 1.12892
I0428 14:27:10.440913 30475 solver.cpp:237]     Train net output #0: loss = 1.12892 (* 1 = 1.12892 loss)
I0428 14:27:10.440922 30475 sgd_solver.cpp:105] Iteration 3840, lr = 0.00467363
I0428 14:27:11.681337 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:27:16.205210 30475 solver.cpp:218] Iteration 3852 (2.08178 iter/s, 5.7643s/12 iters), loss = 1.22111
I0428 14:27:16.205243 30475 solver.cpp:237]     Train net output #0: loss = 1.22111 (* 1 = 1.22111 loss)
I0428 14:27:16.205250 30475 sgd_solver.cpp:105] Iteration 3852, lr = 0.00466253
I0428 14:27:21.807060 30475 solver.cpp:218] Iteration 3864 (2.14217 iter/s, 5.6018s/12 iters), loss = 1.08813
I0428 14:27:21.807250 30475 solver.cpp:237]     Train net output #0: loss = 1.08813 (* 1 = 1.08813 loss)
I0428 14:27:21.807262 30475 sgd_solver.cpp:105] Iteration 3864, lr = 0.00465146
I0428 14:27:26.769445 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3876.caffemodel
I0428 14:27:28.982136 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3876.solverstate
I0428 14:27:30.898726 30475 solver.cpp:330] Iteration 3876, Testing net (#0)
I0428 14:27:30.898744 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:27:34.192821 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:27:36.002686 30475 solver.cpp:397]     Test net output #0: accuracy = 0.352328
I0428 14:27:36.002712 30475 solver.cpp:397]     Test net output #1: loss = 3.1381 (* 1 = 3.1381 loss)
I0428 14:27:36.155350 30475 solver.cpp:218] Iteration 3876 (0.836347 iter/s, 14.3481s/12 iters), loss = 1.06413
I0428 14:27:36.155413 30475 solver.cpp:237]     Train net output #0: loss = 1.06413 (* 1 = 1.06413 loss)
I0428 14:27:36.155427 30475 sgd_solver.cpp:105] Iteration 3876, lr = 0.00464042
I0428 14:27:40.982643 30475 solver.cpp:218] Iteration 3888 (2.4859 iter/s, 4.82723s/12 iters), loss = 0.773767
I0428 14:27:40.982687 30475 solver.cpp:237]     Train net output #0: loss = 0.773767 (* 1 = 0.773767 loss)
I0428 14:27:40.982695 30475 sgd_solver.cpp:105] Iteration 3888, lr = 0.0046294
I0428 14:27:46.680608 30475 solver.cpp:218] Iteration 3900 (2.10603 iter/s, 5.69792s/12 iters), loss = 1.08364
I0428 14:27:46.680650 30475 solver.cpp:237]     Train net output #0: loss = 1.08364 (* 1 = 1.08364 loss)
I0428 14:27:46.680660 30475 sgd_solver.cpp:105] Iteration 3900, lr = 0.00461841
I0428 14:27:52.299144 30475 solver.cpp:218] Iteration 3912 (2.13581 iter/s, 5.61849s/12 iters), loss = 0.914697
I0428 14:27:52.299263 30475 solver.cpp:237]     Train net output #0: loss = 0.914697 (* 1 = 0.914697 loss)
I0428 14:27:52.299273 30475 sgd_solver.cpp:105] Iteration 3912, lr = 0.00460744
I0428 14:27:57.929477 30475 solver.cpp:218] Iteration 3924 (2.13136 iter/s, 5.63021s/12 iters), loss = 1.09872
I0428 14:27:57.929522 30475 solver.cpp:237]     Train net output #0: loss = 1.09872 (* 1 = 1.09872 loss)
I0428 14:27:57.929530 30475 sgd_solver.cpp:105] Iteration 3924, lr = 0.0045965
I0428 14:28:03.567095 30475 solver.cpp:218] Iteration 3936 (2.12858 iter/s, 5.63756s/12 iters), loss = 0.904886
I0428 14:28:03.567140 30475 solver.cpp:237]     Train net output #0: loss = 0.904886 (* 1 = 0.904886 loss)
I0428 14:28:03.567148 30475 sgd_solver.cpp:105] Iteration 3936, lr = 0.00458559
I0428 14:28:07.379012 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:28:09.222091 30475 solver.cpp:218] Iteration 3948 (2.12203 iter/s, 5.65495s/12 iters), loss = 0.936468
I0428 14:28:09.222128 30475 solver.cpp:237]     Train net output #0: loss = 0.936468 (* 1 = 0.936468 loss)
I0428 14:28:09.222136 30475 sgd_solver.cpp:105] Iteration 3948, lr = 0.0045747
I0428 14:28:14.853857 30475 solver.cpp:218] Iteration 3960 (2.13079 iter/s, 5.63173s/12 iters), loss = 0.830002
I0428 14:28:14.853899 30475 solver.cpp:237]     Train net output #0: loss = 0.830002 (* 1 = 0.830002 loss)
I0428 14:28:14.853907 30475 sgd_solver.cpp:105] Iteration 3960, lr = 0.00456384
I0428 14:28:20.456595 30475 solver.cpp:218] Iteration 3972 (2.14183 iter/s, 5.60269s/12 iters), loss = 0.801514
I0428 14:28:20.456641 30475 solver.cpp:237]     Train net output #0: loss = 0.801514 (* 1 = 0.801514 loss)
I0428 14:28:20.456650 30475 sgd_solver.cpp:105] Iteration 3972, lr = 0.00455301
I0428 14:28:22.743160 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3978.caffemodel
I0428 14:28:26.363121 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3978.solverstate
I0428 14:28:29.063899 30475 solver.cpp:330] Iteration 3978, Testing net (#0)
I0428 14:28:29.063917 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:28:32.303745 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:28:34.163798 30475 solver.cpp:397]     Test net output #0: accuracy = 0.349877
I0428 14:28:34.163823 30475 solver.cpp:397]     Test net output #1: loss = 3.22061 (* 1 = 3.22061 loss)
I0428 14:28:36.272884 30475 solver.cpp:218] Iteration 3984 (0.758713 iter/s, 15.8163s/12 iters), loss = 1.03771
I0428 14:28:36.272928 30475 solver.cpp:237]     Train net output #0: loss = 1.03771 (* 1 = 1.03771 loss)
I0428 14:28:36.272938 30475 sgd_solver.cpp:105] Iteration 3984, lr = 0.0045422
I0428 14:28:41.897127 30475 solver.cpp:218] Iteration 3996 (2.13364 iter/s, 5.62419s/12 iters), loss = 0.82995
I0428 14:28:41.897169 30475 solver.cpp:237]     Train net output #0: loss = 0.82995 (* 1 = 0.82995 loss)
I0428 14:28:41.897177 30475 sgd_solver.cpp:105] Iteration 3996, lr = 0.00453141
I0428 14:28:47.517912 30475 solver.cpp:218] Iteration 4008 (2.13495 iter/s, 5.62073s/12 iters), loss = 0.851053
I0428 14:28:47.517956 30475 solver.cpp:237]     Train net output #0: loss = 0.851053 (* 1 = 0.851053 loss)
I0428 14:28:47.517963 30475 sgd_solver.cpp:105] Iteration 4008, lr = 0.00452066
I0428 14:28:53.124873 30475 solver.cpp:218] Iteration 4020 (2.14022 iter/s, 5.60691s/12 iters), loss = 0.687371
I0428 14:28:53.125047 30475 solver.cpp:237]     Train net output #0: loss = 0.687371 (* 1 = 0.687371 loss)
I0428 14:28:53.125057 30475 sgd_solver.cpp:105] Iteration 4020, lr = 0.00450992
I0428 14:28:58.649400 30475 solver.cpp:218] Iteration 4032 (2.1722 iter/s, 5.52435s/12 iters), loss = 0.807345
I0428 14:28:58.649447 30475 solver.cpp:237]     Train net output #0: loss = 0.807345 (* 1 = 0.807345 loss)
I0428 14:28:58.649456 30475 sgd_solver.cpp:105] Iteration 4032, lr = 0.00449921
I0428 14:29:04.274828 30475 solver.cpp:218] Iteration 4044 (2.13319 iter/s, 5.62538s/12 iters), loss = 0.914033
I0428 14:29:04.274875 30475 solver.cpp:237]     Train net output #0: loss = 0.914033 (* 1 = 0.914033 loss)
I0428 14:29:04.274883 30475 sgd_solver.cpp:105] Iteration 4044, lr = 0.00448853
I0428 14:29:04.827899 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:29:09.917994 30475 solver.cpp:218] Iteration 4056 (2.12649 iter/s, 5.64311s/12 iters), loss = 0.820119
I0428 14:29:09.918033 30475 solver.cpp:237]     Train net output #0: loss = 0.820119 (* 1 = 0.820119 loss)
I0428 14:29:09.918042 30475 sgd_solver.cpp:105] Iteration 4056, lr = 0.00447788
I0428 14:29:15.528208 30475 solver.cpp:218] Iteration 4068 (2.13897 iter/s, 5.61017s/12 iters), loss = 0.793218
I0428 14:29:15.528251 30475 solver.cpp:237]     Train net output #0: loss = 0.793218 (* 1 = 0.793218 loss)
I0428 14:29:15.528260 30475 sgd_solver.cpp:105] Iteration 4068, lr = 0.00446724
I0428 14:29:20.500068 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4080.caffemodel
I0428 14:29:27.296618 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4080.solverstate
I0428 14:29:29.995263 30475 solver.cpp:330] Iteration 4080, Testing net (#0)
I0428 14:29:29.995281 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:29:33.006500 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:29:34.904147 30475 solver.cpp:397]     Test net output #0: accuracy = 0.363358
I0428 14:29:34.904175 30475 solver.cpp:397]     Test net output #1: loss = 3.33102 (* 1 = 3.33102 loss)
I0428 14:29:35.066254 30475 solver.cpp:218] Iteration 4080 (0.614187 iter/s, 19.538s/12 iters), loss = 0.915518
I0428 14:29:35.066298 30475 solver.cpp:237]     Train net output #0: loss = 0.915518 (* 1 = 0.915518 loss)
I0428 14:29:35.066305 30475 sgd_solver.cpp:105] Iteration 4080, lr = 0.00445664
I0428 14:29:39.816382 30475 solver.cpp:218] Iteration 4092 (2.52627 iter/s, 4.75008s/12 iters), loss = 0.978847
I0428 14:29:39.816424 30475 solver.cpp:237]     Train net output #0: loss = 0.978847 (* 1 = 0.978847 loss)
I0428 14:29:39.816432 30475 sgd_solver.cpp:105] Iteration 4092, lr = 0.00444606
I0428 14:29:45.495131 30475 solver.cpp:218] Iteration 4104 (2.11316 iter/s, 5.6787s/12 iters), loss = 0.636948
I0428 14:29:45.495172 30475 solver.cpp:237]     Train net output #0: loss = 0.636948 (* 1 = 0.636948 loss)
I0428 14:29:45.495182 30475 sgd_solver.cpp:105] Iteration 4104, lr = 0.0044355
I0428 14:29:51.158632 30475 solver.cpp:218] Iteration 4116 (2.11885 iter/s, 5.66345s/12 iters), loss = 0.765151
I0428 14:29:51.158670 30475 solver.cpp:237]     Train net output #0: loss = 0.765151 (* 1 = 0.765151 loss)
I0428 14:29:51.158679 30475 sgd_solver.cpp:105] Iteration 4116, lr = 0.00442497
I0428 14:29:56.909178 30475 solver.cpp:218] Iteration 4128 (2.08677 iter/s, 5.7505s/12 iters), loss = 0.684143
I0428 14:29:56.909224 30475 solver.cpp:237]     Train net output #0: loss = 0.684143 (* 1 = 0.684143 loss)
I0428 14:29:56.909233 30475 sgd_solver.cpp:105] Iteration 4128, lr = 0.00441447
I0428 14:30:02.588393 30475 solver.cpp:218] Iteration 4140 (2.11299 iter/s, 5.67916s/12 iters), loss = 0.915793
I0428 14:30:02.588517 30475 solver.cpp:237]     Train net output #0: loss = 0.915793 (* 1 = 0.915793 loss)
I0428 14:30:02.588527 30475 sgd_solver.cpp:105] Iteration 4140, lr = 0.00440398
I0428 14:30:05.571382 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:30:08.288666 30475 solver.cpp:218] Iteration 4152 (2.10521 iter/s, 5.70015s/12 iters), loss = 0.897254
I0428 14:30:08.288704 30475 solver.cpp:237]     Train net output #0: loss = 0.897254 (* 1 = 0.897254 loss)
I0428 14:30:08.288713 30475 sgd_solver.cpp:105] Iteration 4152, lr = 0.00439353
I0428 14:30:10.100445 30475 blocking_queue.cpp:49] Waiting for data
I0428 14:30:13.970791 30475 solver.cpp:218] Iteration 4164 (2.1119 iter/s, 5.68208s/12 iters), loss = 0.76361
I0428 14:30:13.970837 30475 solver.cpp:237]     Train net output #0: loss = 0.76361 (* 1 = 0.76361 loss)
I0428 14:30:13.970845 30475 sgd_solver.cpp:105] Iteration 4164, lr = 0.0043831
I0428 14:30:19.661846 30475 solver.cpp:218] Iteration 4176 (2.10859 iter/s, 5.691s/12 iters), loss = 0.631111
I0428 14:30:19.661885 30475 solver.cpp:237]     Train net output #0: loss = 0.631111 (* 1 = 0.631111 loss)
I0428 14:30:19.661892 30475 sgd_solver.cpp:105] Iteration 4176, lr = 0.00437269
I0428 14:30:21.939234 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4182.caffemodel
I0428 14:30:28.808845 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4182.solverstate
I0428 14:30:32.625883 30475 solver.cpp:330] Iteration 4182, Testing net (#0)
I0428 14:30:32.626006 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:30:35.760046 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:30:37.718482 30475 solver.cpp:397]     Test net output #0: accuracy = 0.371936
I0428 14:30:37.718516 30475 solver.cpp:397]     Test net output #1: loss = 3.21384 (* 1 = 3.21384 loss)
I0428 14:30:39.891366 30475 solver.cpp:218] Iteration 4188 (0.593193 iter/s, 20.2295s/12 iters), loss = 0.786169
I0428 14:30:39.891413 30475 solver.cpp:237]     Train net output #0: loss = 0.786169 (* 1 = 0.786169 loss)
I0428 14:30:39.891422 30475 sgd_solver.cpp:105] Iteration 4188, lr = 0.00436231
I0428 14:30:45.501562 30475 solver.cpp:218] Iteration 4200 (2.13898 iter/s, 5.61014s/12 iters), loss = 0.736024
I0428 14:30:45.501606 30475 solver.cpp:237]     Train net output #0: loss = 0.736024 (* 1 = 0.736024 loss)
I0428 14:30:45.501613 30475 sgd_solver.cpp:105] Iteration 4200, lr = 0.00435195
I0428 14:30:51.115427 30475 solver.cpp:218] Iteration 4212 (2.13758 iter/s, 5.61382s/12 iters), loss = 0.600672
I0428 14:30:51.115473 30475 solver.cpp:237]     Train net output #0: loss = 0.600672 (* 1 = 0.600672 loss)
I0428 14:30:51.115483 30475 sgd_solver.cpp:105] Iteration 4212, lr = 0.00434162
I0428 14:30:56.743144 30475 solver.cpp:218] Iteration 4224 (2.13232 iter/s, 5.62767s/12 iters), loss = 0.747402
I0428 14:30:56.743180 30475 solver.cpp:237]     Train net output #0: loss = 0.747402 (* 1 = 0.747402 loss)
I0428 14:30:56.743187 30475 sgd_solver.cpp:105] Iteration 4224, lr = 0.00433131
I0428 14:31:02.298954 30475 solver.cpp:218] Iteration 4236 (2.15992 iter/s, 5.55576s/12 iters), loss = 0.642932
I0428 14:31:02.298995 30475 solver.cpp:237]     Train net output #0: loss = 0.642932 (* 1 = 0.642932 loss)
I0428 14:31:02.299005 30475 sgd_solver.cpp:105] Iteration 4236, lr = 0.00432103
I0428 14:31:07.675585 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:31:07.964205 30475 solver.cpp:218] Iteration 4248 (2.11819 iter/s, 5.6652s/12 iters), loss = 0.830288
I0428 14:31:07.964247 30475 solver.cpp:237]     Train net output #0: loss = 0.830288 (* 1 = 0.830288 loss)
I0428 14:31:07.964255 30475 sgd_solver.cpp:105] Iteration 4248, lr = 0.00431077
I0428 14:31:13.512439 30475 solver.cpp:218] Iteration 4260 (2.16287 iter/s, 5.54819s/12 iters), loss = 0.655906
I0428 14:31:13.512476 30475 solver.cpp:237]     Train net output #0: loss = 0.655906 (* 1 = 0.655906 loss)
I0428 14:31:13.512485 30475 sgd_solver.cpp:105] Iteration 4260, lr = 0.00430053
I0428 14:31:19.288475 30475 solver.cpp:218] Iteration 4272 (2.07757 iter/s, 5.77599s/12 iters), loss = 0.542724
I0428 14:31:19.288517 30475 solver.cpp:237]     Train net output #0: loss = 0.542724 (* 1 = 0.542724 loss)
I0428 14:31:19.288525 30475 sgd_solver.cpp:105] Iteration 4272, lr = 0.00429032
I0428 14:31:24.664355 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4284.caffemodel
I0428 14:31:29.392076 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4284.solverstate
I0428 14:31:32.115381 30475 solver.cpp:330] Iteration 4284, Testing net (#0)
I0428 14:31:32.115401 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:31:35.014413 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:31:36.877393 30475 solver.cpp:397]     Test net output #0: accuracy = 0.381127
I0428 14:31:36.877439 30475 solver.cpp:397]     Test net output #1: loss = 3.1189 (* 1 = 3.1189 loss)
I0428 14:31:37.039080 30475 solver.cpp:218] Iteration 4284 (0.676034 iter/s, 17.7506s/12 iters), loss = 0.714684
I0428 14:31:37.039147 30475 solver.cpp:237]     Train net output #0: loss = 0.714684 (* 1 = 0.714684 loss)
I0428 14:31:37.039156 30475 sgd_solver.cpp:105] Iteration 4284, lr = 0.00428014
I0428 14:31:41.751569 30475 solver.cpp:218] Iteration 4296 (2.54646 iter/s, 4.71242s/12 iters), loss = 0.681586
I0428 14:31:41.751690 30475 solver.cpp:237]     Train net output #0: loss = 0.681586 (* 1 = 0.681586 loss)
I0428 14:31:41.751699 30475 sgd_solver.cpp:105] Iteration 4296, lr = 0.00426998
I0428 14:31:47.377177 30475 solver.cpp:218] Iteration 4308 (2.13315 iter/s, 5.62549s/12 iters), loss = 0.58303
I0428 14:31:47.377223 30475 solver.cpp:237]     Train net output #0: loss = 0.58303 (* 1 = 0.58303 loss)
I0428 14:31:47.377230 30475 sgd_solver.cpp:105] Iteration 4308, lr = 0.00425984
I0428 14:31:52.892545 30475 solver.cpp:218] Iteration 4320 (2.17576 iter/s, 5.51532s/12 iters), loss = 0.650718
I0428 14:31:52.892587 30475 solver.cpp:237]     Train net output #0: loss = 0.650718 (* 1 = 0.650718 loss)
I0428 14:31:52.892596 30475 sgd_solver.cpp:105] Iteration 4320, lr = 0.00424972
I0428 14:31:58.507453 30475 solver.cpp:218] Iteration 4332 (2.13718 iter/s, 5.61486s/12 iters), loss = 0.610406
I0428 14:31:58.507493 30475 solver.cpp:237]     Train net output #0: loss = 0.610406 (* 1 = 0.610406 loss)
I0428 14:31:58.507501 30475 sgd_solver.cpp:105] Iteration 4332, lr = 0.00423964
I0428 14:32:04.136669 30475 solver.cpp:218] Iteration 4344 (2.13175 iter/s, 5.62917s/12 iters), loss = 0.611655
I0428 14:32:04.136711 30475 solver.cpp:237]     Train net output #0: loss = 0.611655 (* 1 = 0.611655 loss)
I0428 14:32:04.136720 30475 sgd_solver.cpp:105] Iteration 4344, lr = 0.00422957
I0428 14:32:06.253208 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:32:09.686628 30475 solver.cpp:218] Iteration 4356 (2.1622 iter/s, 5.54991s/12 iters), loss = 0.999279
I0428 14:32:09.686671 30475 solver.cpp:237]     Train net output #0: loss = 0.999279 (* 1 = 0.999279 loss)
I0428 14:32:09.686678 30475 sgd_solver.cpp:105] Iteration 4356, lr = 0.00421953
I0428 14:32:15.301692 30475 solver.cpp:218] Iteration 4368 (2.13713 iter/s, 5.61501s/12 iters), loss = 0.707606
I0428 14:32:15.302407 30475 solver.cpp:237]     Train net output #0: loss = 0.707606 (* 1 = 0.707606 loss)
I0428 14:32:15.302415 30475 sgd_solver.cpp:105] Iteration 4368, lr = 0.00420951
I0428 14:32:20.911192 30475 solver.cpp:218] Iteration 4380 (2.1395 iter/s, 5.60879s/12 iters), loss = 0.501423
I0428 14:32:20.911237 30475 solver.cpp:237]     Train net output #0: loss = 0.501423 (* 1 = 0.501423 loss)
I0428 14:32:20.911244 30475 sgd_solver.cpp:105] Iteration 4380, lr = 0.00419952
I0428 14:32:23.103567 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4386.caffemodel
I0428 14:32:27.221451 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4386.solverstate
I0428 14:32:29.354260 30475 solver.cpp:330] Iteration 4386, Testing net (#0)
I0428 14:32:29.354291 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:32:32.396574 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:32:34.449144 30475 solver.cpp:397]     Test net output #0: accuracy = 0.390319
I0428 14:32:34.449189 30475 solver.cpp:397]     Test net output #1: loss = 3.31953 (* 1 = 3.31953 loss)
I0428 14:32:36.609391 30475 solver.cpp:218] Iteration 4392 (0.76442 iter/s, 15.6982s/12 iters), loss = 0.622214
I0428 14:32:36.609437 30475 solver.cpp:237]     Train net output #0: loss = 0.622214 (* 1 = 0.622214 loss)
I0428 14:32:36.609447 30475 sgd_solver.cpp:105] Iteration 4392, lr = 0.00418954
I0428 14:32:42.248385 30475 solver.cpp:218] Iteration 4404 (2.12806 iter/s, 5.63894s/12 iters), loss = 0.468406
I0428 14:32:42.248431 30475 solver.cpp:237]     Train net output #0: loss = 0.468406 (* 1 = 0.468406 loss)
I0428 14:32:42.248440 30475 sgd_solver.cpp:105] Iteration 4404, lr = 0.0041796
I0428 14:32:47.941220 30475 solver.cpp:218] Iteration 4416 (2.10793 iter/s, 5.69278s/12 iters), loss = 0.69868
I0428 14:32:47.941334 30475 solver.cpp:237]     Train net output #0: loss = 0.69868 (* 1 = 0.69868 loss)
I0428 14:32:47.941344 30475 sgd_solver.cpp:105] Iteration 4416, lr = 0.00416967
I0428 14:32:53.515352 30475 solver.cpp:218] Iteration 4428 (2.15285 iter/s, 5.57401s/12 iters), loss = 0.645782
I0428 14:32:53.515398 30475 solver.cpp:237]     Train net output #0: loss = 0.645782 (* 1 = 0.645782 loss)
I0428 14:32:53.515406 30475 sgd_solver.cpp:105] Iteration 4428, lr = 0.00415977
I0428 14:32:59.233724 30475 solver.cpp:218] Iteration 4440 (2.09852 iter/s, 5.71832s/12 iters), loss = 0.61683
I0428 14:32:59.233770 30475 solver.cpp:237]     Train net output #0: loss = 0.61683 (* 1 = 0.61683 loss)
I0428 14:32:59.233779 30475 sgd_solver.cpp:105] Iteration 4440, lr = 0.0041499
I0428 14:33:03.797550 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:33:04.948972 30475 solver.cpp:218] Iteration 4452 (2.09967 iter/s, 5.71519s/12 iters), loss = 0.739787
I0428 14:33:04.949021 30475 solver.cpp:237]     Train net output #0: loss = 0.739787 (* 1 = 0.739787 loss)
I0428 14:33:04.949030 30475 sgd_solver.cpp:105] Iteration 4452, lr = 0.00414005
I0428 14:33:10.779165 30475 solver.cpp:218] Iteration 4464 (2.05827 iter/s, 5.83015s/12 iters), loss = 0.723309
I0428 14:33:10.779191 30475 solver.cpp:237]     Train net output #0: loss = 0.723309 (* 1 = 0.723309 loss)
I0428 14:33:10.779198 30475 sgd_solver.cpp:105] Iteration 4464, lr = 0.00413022
I0428 14:33:16.393414 30475 solver.cpp:218] Iteration 4476 (2.13743 iter/s, 5.61421s/12 iters), loss = 0.529277
I0428 14:33:16.393455 30475 solver.cpp:237]     Train net output #0: loss = 0.529277 (* 1 = 0.529277 loss)
I0428 14:33:16.393465 30475 sgd_solver.cpp:105] Iteration 4476, lr = 0.00412041
I0428 14:33:21.484773 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4488.caffemodel
I0428 14:33:25.856323 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4488.solverstate
I0428 14:33:28.179085 30475 solver.cpp:330] Iteration 4488, Testing net (#0)
I0428 14:33:28.179102 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:33:31.260546 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:33:33.625962 30475 solver.cpp:397]     Test net output #0: accuracy = 0.384191
I0428 14:33:33.625991 30475 solver.cpp:397]     Test net output #1: loss = 3.34736 (* 1 = 3.34736 loss)
I0428 14:33:33.782258 30475 solver.cpp:218] Iteration 4488 (0.690099 iter/s, 17.3888s/12 iters), loss = 0.621208
I0428 14:33:33.782305 30475 solver.cpp:237]     Train net output #0: loss = 0.621208 (* 1 = 0.621208 loss)
I0428 14:33:33.782315 30475 sgd_solver.cpp:105] Iteration 4488, lr = 0.00411063
I0428 14:33:38.549602 30475 solver.cpp:218] Iteration 4500 (2.51715 iter/s, 4.7673s/12 iters), loss = 0.631857
I0428 14:33:38.549643 30475 solver.cpp:237]     Train net output #0: loss = 0.631857 (* 1 = 0.631857 loss)
I0428 14:33:38.549650 30475 sgd_solver.cpp:105] Iteration 4500, lr = 0.00410087
I0428 14:33:44.082638 30475 solver.cpp:218] Iteration 4512 (2.16881 iter/s, 5.53298s/12 iters), loss = 0.588642
I0428 14:33:44.082680 30475 solver.cpp:237]     Train net output #0: loss = 0.588642 (* 1 = 0.588642 loss)
I0428 14:33:44.082690 30475 sgd_solver.cpp:105] Iteration 4512, lr = 0.00409113
I0428 14:33:49.705940 30475 solver.cpp:218] Iteration 4524 (2.134 iter/s, 5.62325s/12 iters), loss = 0.52659
I0428 14:33:49.705979 30475 solver.cpp:237]     Train net output #0: loss = 0.52659 (* 1 = 0.52659 loss)
I0428 14:33:49.705987 30475 sgd_solver.cpp:105] Iteration 4524, lr = 0.00408142
I0428 14:33:55.321343 30475 solver.cpp:218] Iteration 4536 (2.137 iter/s, 5.61536s/12 iters), loss = 0.3812
I0428 14:33:55.321509 30475 solver.cpp:237]     Train net output #0: loss = 0.3812 (* 1 = 0.3812 loss)
I0428 14:33:55.321519 30475 sgd_solver.cpp:105] Iteration 4536, lr = 0.00407173
I0428 14:34:00.760325 30475 solver.cpp:218] Iteration 4548 (2.20636 iter/s, 5.43882s/12 iters), loss = 0.933716
I0428 14:34:00.760362 30475 solver.cpp:237]     Train net output #0: loss = 0.933716 (* 1 = 0.933716 loss)
I0428 14:34:00.760370 30475 sgd_solver.cpp:105] Iteration 4548, lr = 0.00406206
I0428 14:34:02.183794 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:34:06.408244 30475 solver.cpp:218] Iteration 4560 (2.12469 iter/s, 5.64787s/12 iters), loss = 0.685736
I0428 14:34:06.408293 30475 solver.cpp:237]     Train net output #0: loss = 0.685736 (* 1 = 0.685736 loss)
I0428 14:34:06.408301 30475 sgd_solver.cpp:105] Iteration 4560, lr = 0.00405242
I0428 14:34:12.036142 30475 solver.cpp:218] Iteration 4572 (2.13226 iter/s, 5.62784s/12 iters), loss = 0.62442
I0428 14:34:12.036197 30475 solver.cpp:237]     Train net output #0: loss = 0.62442 (* 1 = 0.62442 loss)
I0428 14:34:12.036209 30475 sgd_solver.cpp:105] Iteration 4572, lr = 0.0040428
I0428 14:34:17.658128 30475 solver.cpp:218] Iteration 4584 (2.1345 iter/s, 5.62193s/12 iters), loss = 0.670401
I0428 14:34:17.658169 30475 solver.cpp:237]     Train net output #0: loss = 0.670401 (* 1 = 0.670401 loss)
I0428 14:34:17.658176 30475 sgd_solver.cpp:105] Iteration 4584, lr = 0.0040332
I0428 14:34:19.926350 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4590.caffemodel
I0428 14:34:22.544873 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4590.solverstate
I0428 14:34:24.257453 30475 solver.cpp:330] Iteration 4590, Testing net (#0)
I0428 14:34:24.257473 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:34:27.145226 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:34:29.172646 30475 solver.cpp:397]     Test net output #0: accuracy = 0.389706
I0428 14:34:29.172674 30475 solver.cpp:397]     Test net output #1: loss = 3.27343 (* 1 = 3.27343 loss)
I0428 14:34:31.195993 30475 solver.cpp:218] Iteration 4596 (0.886405 iter/s, 13.5378s/12 iters), loss = 0.518852
I0428 14:34:31.196038 30475 solver.cpp:237]     Train net output #0: loss = 0.518852 (* 1 = 0.518852 loss)
I0428 14:34:31.196046 30475 sgd_solver.cpp:105] Iteration 4596, lr = 0.00402362
I0428 14:34:36.811347 30475 solver.cpp:218] Iteration 4608 (2.13702 iter/s, 5.61531s/12 iters), loss = 0.73867
I0428 14:34:36.811390 30475 solver.cpp:237]     Train net output #0: loss = 0.73867 (* 1 = 0.73867 loss)
I0428 14:34:36.811399 30475 sgd_solver.cpp:105] Iteration 4608, lr = 0.00401407
I0428 14:34:42.426064 30475 solver.cpp:218] Iteration 4620 (2.13726 iter/s, 5.61468s/12 iters), loss = 0.512568
I0428 14:34:42.426096 30475 solver.cpp:237]     Train net output #0: loss = 0.512568 (* 1 = 0.512568 loss)
I0428 14:34:42.426103 30475 sgd_solver.cpp:105] Iteration 4620, lr = 0.00400454
I0428 14:34:47.852468 30475 solver.cpp:218] Iteration 4632 (2.21142 iter/s, 5.42637s/12 iters), loss = 0.816389
I0428 14:34:47.852500 30475 solver.cpp:237]     Train net output #0: loss = 0.816389 (* 1 = 0.816389 loss)
I0428 14:34:47.852509 30475 sgd_solver.cpp:105] Iteration 4632, lr = 0.00399503
I0428 14:34:53.378804 30475 solver.cpp:218] Iteration 4644 (2.17144 iter/s, 5.52629s/12 iters), loss = 0.498332
I0428 14:34:53.378854 30475 solver.cpp:237]     Train net output #0: loss = 0.498332 (* 1 = 0.498332 loss)
I0428 14:34:53.378862 30475 sgd_solver.cpp:105] Iteration 4644, lr = 0.00398555
I0428 14:34:57.200817 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:34:59.017958 30475 solver.cpp:218] Iteration 4656 (2.128 iter/s, 5.6391s/12 iters), loss = 0.561577
I0428 14:34:59.018002 30475 solver.cpp:237]     Train net output #0: loss = 0.561577 (* 1 = 0.561577 loss)
I0428 14:34:59.018010 30475 sgd_solver.cpp:105] Iteration 4656, lr = 0.00397608
I0428 14:35:04.629966 30475 solver.cpp:218] Iteration 4668 (2.13829 iter/s, 5.61196s/12 iters), loss = 0.537365
I0428 14:35:04.630007 30475 solver.cpp:237]     Train net output #0: loss = 0.537365 (* 1 = 0.537365 loss)
I0428 14:35:04.630014 30475 sgd_solver.cpp:105] Iteration 4668, lr = 0.00396664
I0428 14:35:10.331140 30475 solver.cpp:218] Iteration 4680 (2.10485 iter/s, 5.70113s/12 iters), loss = 0.618632
I0428 14:35:10.331182 30475 solver.cpp:237]     Train net output #0: loss = 0.618632 (* 1 = 0.618632 loss)
I0428 14:35:10.331190 30475 sgd_solver.cpp:105] Iteration 4680, lr = 0.00395723
I0428 14:35:15.420199 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4692.caffemodel
I0428 14:35:17.626632 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4692.solverstate
I0428 14:35:19.933462 30475 solver.cpp:330] Iteration 4692, Testing net (#0)
I0428 14:35:19.933483 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:35:22.861829 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:35:25.061131 30475 solver.cpp:397]     Test net output #0: accuracy = 0.393995
I0428 14:35:25.061159 30475 solver.cpp:397]     Test net output #1: loss = 3.18507 (* 1 = 3.18507 loss)
I0428 14:35:25.213502 30475 solver.cpp:218] Iteration 4692 (0.806325 iter/s, 14.8823s/12 iters), loss = 0.589518
I0428 14:35:25.213546 30475 solver.cpp:237]     Train net output #0: loss = 0.589518 (* 1 = 0.589518 loss)
I0428 14:35:25.213553 30475 sgd_solver.cpp:105] Iteration 4692, lr = 0.00394783
I0428 14:35:29.781383 30475 solver.cpp:218] Iteration 4704 (2.62707 iter/s, 4.56782s/12 iters), loss = 0.53004
I0428 14:35:29.781495 30475 solver.cpp:237]     Train net output #0: loss = 0.53004 (* 1 = 0.53004 loss)
I0428 14:35:29.781505 30475 sgd_solver.cpp:105] Iteration 4704, lr = 0.00393846
I0428 14:35:35.480047 30475 solver.cpp:218] Iteration 4716 (2.1058 iter/s, 5.69855s/12 iters), loss = 0.501277
I0428 14:35:35.480090 30475 solver.cpp:237]     Train net output #0: loss = 0.501277 (* 1 = 0.501277 loss)
I0428 14:35:35.480098 30475 sgd_solver.cpp:105] Iteration 4716, lr = 0.00392911
I0428 14:35:41.060020 30475 solver.cpp:218] Iteration 4728 (2.15056 iter/s, 5.57993s/12 iters), loss = 0.418965
I0428 14:35:41.060060 30475 solver.cpp:237]     Train net output #0: loss = 0.418965 (* 1 = 0.418965 loss)
I0428 14:35:41.060067 30475 sgd_solver.cpp:105] Iteration 4728, lr = 0.00391978
I0428 14:35:46.710642 30475 solver.cpp:218] Iteration 4740 (2.12368 iter/s, 5.65056s/12 iters), loss = 0.559189
I0428 14:35:46.710713 30475 solver.cpp:237]     Train net output #0: loss = 0.559189 (* 1 = 0.559189 loss)
I0428 14:35:46.710726 30475 sgd_solver.cpp:105] Iteration 4740, lr = 0.00391047
I0428 14:35:52.345492 30475 solver.cpp:218] Iteration 4752 (2.12963 iter/s, 5.63478s/12 iters), loss = 0.6332
I0428 14:35:52.345533 30475 solver.cpp:237]     Train net output #0: loss = 0.6332 (* 1 = 0.6332 loss)
I0428 14:35:52.345541 30475 sgd_solver.cpp:105] Iteration 4752, lr = 0.00390119
I0428 14:35:52.927136 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:35:57.986258 30475 solver.cpp:218] Iteration 4764 (2.12739 iter/s, 5.64072s/12 iters), loss = 0.298634
I0428 14:35:57.986296 30475 solver.cpp:237]     Train net output #0: loss = 0.298634 (* 1 = 0.298634 loss)
I0428 14:35:57.986304 30475 sgd_solver.cpp:105] Iteration 4764, lr = 0.00389193
I0428 14:36:03.510174 30475 solver.cpp:218] Iteration 4776 (2.17239 iter/s, 5.52387s/12 iters), loss = 0.427342
I0428 14:36:03.510350 30475 solver.cpp:237]     Train net output #0: loss = 0.427342 (* 1 = 0.427342 loss)
I0428 14:36:03.510360 30475 sgd_solver.cpp:105] Iteration 4776, lr = 0.00388269
I0428 14:36:09.141950 30475 solver.cpp:218] Iteration 4788 (2.13083 iter/s, 5.6316s/12 iters), loss = 0.372048
I0428 14:36:09.141989 30475 solver.cpp:237]     Train net output #0: loss = 0.372048 (* 1 = 0.372048 loss)
I0428 14:36:09.141996 30475 sgd_solver.cpp:105] Iteration 4788, lr = 0.00387347
I0428 14:36:11.409461 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4794.caffemodel
I0428 14:36:13.610178 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4794.solverstate
I0428 14:36:15.301395 30475 solver.cpp:330] Iteration 4794, Testing net (#0)
I0428 14:36:15.301415 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:36:18.161581 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:36:20.414062 30475 solver.cpp:397]     Test net output #0: accuracy = 0.410539
I0428 14:36:20.414098 30475 solver.cpp:397]     Test net output #1: loss = 3.35624 (* 1 = 3.35624 loss)
I0428 14:36:22.422284 30475 solver.cpp:218] Iteration 4800 (0.903593 iter/s, 13.2803s/12 iters), loss = 0.399556
I0428 14:36:22.422324 30475 solver.cpp:237]     Train net output #0: loss = 0.399556 (* 1 = 0.399556 loss)
I0428 14:36:22.422331 30475 sgd_solver.cpp:105] Iteration 4800, lr = 0.00386427
I0428 14:36:28.033460 30475 solver.cpp:218] Iteration 4812 (2.13861 iter/s, 5.61113s/12 iters), loss = 0.413272
I0428 14:36:28.033506 30475 solver.cpp:237]     Train net output #0: loss = 0.413272 (* 1 = 0.413272 loss)
I0428 14:36:28.033514 30475 sgd_solver.cpp:105] Iteration 4812, lr = 0.0038551
I0428 14:36:33.639434 30475 solver.cpp:218] Iteration 4824 (2.14059 iter/s, 5.60593s/12 iters), loss = 0.1614
I0428 14:36:33.639514 30475 solver.cpp:237]     Train net output #0: loss = 0.1614 (* 1 = 0.1614 loss)
I0428 14:36:33.639523 30475 sgd_solver.cpp:105] Iteration 4824, lr = 0.00384594
I0428 14:36:39.262418 30475 solver.cpp:218] Iteration 4836 (2.13413 iter/s, 5.6229s/12 iters), loss = 0.367303
I0428 14:36:39.262461 30475 solver.cpp:237]     Train net output #0: loss = 0.367303 (* 1 = 0.367303 loss)
I0428 14:36:39.262470 30475 sgd_solver.cpp:105] Iteration 4836, lr = 0.00383681
I0428 14:36:41.560920 30475 blocking_queue.cpp:49] Waiting for data
I0428 14:36:44.997773 30475 solver.cpp:218] Iteration 4848 (2.0923 iter/s, 5.73531s/12 iters), loss = 0.439351
I0428 14:36:44.997813 30475 solver.cpp:237]     Train net output #0: loss = 0.439351 (* 1 = 0.439351 loss)
I0428 14:36:44.997821 30475 sgd_solver.cpp:105] Iteration 4848, lr = 0.0038277
I0428 14:36:47.997813 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:36:50.820735 30475 solver.cpp:218] Iteration 4860 (2.06082 iter/s, 5.82292s/12 iters), loss = 0.446032
I0428 14:36:50.820775 30475 solver.cpp:237]     Train net output #0: loss = 0.446032 (* 1 = 0.446032 loss)
I0428 14:36:50.820782 30475 sgd_solver.cpp:105] Iteration 4860, lr = 0.00381862
I0428 14:36:56.432919 30475 solver.cpp:218] Iteration 4872 (2.13822 iter/s, 5.61213s/12 iters), loss = 0.434522
I0428 14:36:56.432973 30475 solver.cpp:237]     Train net output #0: loss = 0.434522 (* 1 = 0.434522 loss)
I0428 14:36:56.432984 30475 sgd_solver.cpp:105] Iteration 4872, lr = 0.00380955
I0428 14:37:02.061569 30475 solver.cpp:218] Iteration 4884 (2.13197 iter/s, 5.62859s/12 iters), loss = 0.430597
I0428 14:37:02.061610 30475 solver.cpp:237]     Train net output #0: loss = 0.430597 (* 1 = 0.430597 loss)
I0428 14:37:02.061619 30475 sgd_solver.cpp:105] Iteration 4884, lr = 0.0038005
I0428 14:37:07.023216 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4896.caffemodel
I0428 14:37:09.225018 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4896.solverstate
I0428 14:37:10.935400 30475 solver.cpp:330] Iteration 4896, Testing net (#0)
I0428 14:37:10.935420 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:37:13.756296 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:37:16.016404 30475 solver.cpp:397]     Test net output #0: accuracy = 0.416667
I0428 14:37:16.016433 30475 solver.cpp:397]     Test net output #1: loss = 3.37249 (* 1 = 3.37249 loss)
I0428 14:37:16.178503 30475 solver.cpp:218] Iteration 4896 (0.850045 iter/s, 14.1169s/12 iters), loss = 0.303714
I0428 14:37:16.178552 30475 solver.cpp:237]     Train net output #0: loss = 0.303714 (* 1 = 0.303714 loss)
I0428 14:37:16.178561 30475 sgd_solver.cpp:105] Iteration 4896, lr = 0.00379148
I0428 14:37:20.876281 30475 solver.cpp:218] Iteration 4908 (2.55443 iter/s, 4.69772s/12 iters), loss = 0.353016
I0428 14:37:20.876327 30475 solver.cpp:237]     Train net output #0: loss = 0.353016 (* 1 = 0.353016 loss)
I0428 14:37:20.876335 30475 sgd_solver.cpp:105] Iteration 4908, lr = 0.00378248
I0428 14:37:26.563907 30475 solver.cpp:218] Iteration 4920 (2.10986 iter/s, 5.68757s/12 iters), loss = 0.450371
I0428 14:37:26.563957 30475 solver.cpp:237]     Train net output #0: loss = 0.450371 (* 1 = 0.450371 loss)
I0428 14:37:26.563966 30475 sgd_solver.cpp:105] Iteration 4920, lr = 0.0037735
I0428 14:37:32.204670 30475 solver.cpp:218] Iteration 4932 (2.12739 iter/s, 5.6407s/12 iters), loss = 0.287245
I0428 14:37:32.204713 30475 solver.cpp:237]     Train net output #0: loss = 0.287245 (* 1 = 0.287245 loss)
I0428 14:37:32.204721 30475 sgd_solver.cpp:105] Iteration 4932, lr = 0.00376454
I0428 14:37:37.836620 30475 solver.cpp:218] Iteration 4944 (2.13072 iter/s, 5.63189s/12 iters), loss = 0.399454
I0428 14:37:37.836761 30475 solver.cpp:237]     Train net output #0: loss = 0.399454 (* 1 = 0.399454 loss)
I0428 14:37:37.836772 30475 sgd_solver.cpp:105] Iteration 4944, lr = 0.0037556
I0428 14:37:43.302198 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:37:43.561607 30475 solver.cpp:218] Iteration 4956 (2.09613 iter/s, 5.72484s/12 iters), loss = 0.229551
I0428 14:37:43.561652 30475 solver.cpp:237]     Train net output #0: loss = 0.229551 (* 1 = 0.229551 loss)
I0428 14:37:43.561661 30475 sgd_solver.cpp:105] Iteration 4956, lr = 0.00374669
I0428 14:37:49.101203 30475 solver.cpp:218] Iteration 4968 (2.16624 iter/s, 5.53954s/12 iters), loss = 0.410413
I0428 14:37:49.101244 30475 solver.cpp:237]     Train net output #0: loss = 0.410413 (* 1 = 0.410413 loss)
I0428 14:37:49.101253 30475 sgd_solver.cpp:105] Iteration 4968, lr = 0.00373779
I0428 14:37:54.711710 30475 solver.cpp:218] Iteration 4980 (2.13886 iter/s, 5.61046s/12 iters), loss = 0.447708
I0428 14:37:54.711751 30475 solver.cpp:237]     Train net output #0: loss = 0.447708 (* 1 = 0.447708 loss)
I0428 14:37:54.711758 30475 sgd_solver.cpp:105] Iteration 4980, lr = 0.00372892
I0428 14:38:00.447491 30475 solver.cpp:218] Iteration 4992 (2.09215 iter/s, 5.73573s/12 iters), loss = 0.302347
I0428 14:38:00.447538 30475 solver.cpp:237]     Train net output #0: loss = 0.302347 (* 1 = 0.302347 loss)
I0428 14:38:00.447546 30475 sgd_solver.cpp:105] Iteration 4992, lr = 0.00372006
I0428 14:38:02.715600 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4998.caffemodel
I0428 14:38:05.861480 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4998.solverstate
I0428 14:38:07.999689 30475 solver.cpp:330] Iteration 4998, Testing net (#0)
I0428 14:38:07.999862 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:38:10.781728 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:38:13.097894 30475 solver.cpp:397]     Test net output #0: accuracy = 0.406863
I0428 14:38:13.097931 30475 solver.cpp:397]     Test net output #1: loss = 3.43328 (* 1 = 3.43328 loss)
I0428 14:38:15.238994 30475 solver.cpp:218] Iteration 5004 (0.811279 iter/s, 14.7915s/12 iters), loss = 0.375882
I0428 14:38:15.239043 30475 solver.cpp:237]     Train net output #0: loss = 0.375882 (* 1 = 0.375882 loss)
I0428 14:38:15.239051 30475 sgd_solver.cpp:105] Iteration 5004, lr = 0.00371123
I0428 14:38:20.847702 30475 solver.cpp:218] Iteration 5016 (2.13955 iter/s, 5.60865s/12 iters), loss = 0.438418
I0428 14:38:20.847740 30475 solver.cpp:237]     Train net output #0: loss = 0.438418 (* 1 = 0.438418 loss)
I0428 14:38:20.847749 30475 sgd_solver.cpp:105] Iteration 5016, lr = 0.00370242
I0428 14:38:26.501215 30475 solver.cpp:218] Iteration 5028 (2.12259 iter/s, 5.65347s/12 iters), loss = 0.300076
I0428 14:38:26.501260 30475 solver.cpp:237]     Train net output #0: loss = 0.300076 (* 1 = 0.300076 loss)
I0428 14:38:26.501267 30475 sgd_solver.cpp:105] Iteration 5028, lr = 0.00369363
I0428 14:38:32.218257 30475 solver.cpp:218] Iteration 5040 (2.09901 iter/s, 5.71699s/12 iters), loss = 0.363777
I0428 14:38:32.218295 30475 solver.cpp:237]     Train net output #0: loss = 0.363777 (* 1 = 0.363777 loss)
I0428 14:38:32.218302 30475 sgd_solver.cpp:105] Iteration 5040, lr = 0.00368486
I0428 14:38:37.903923 30475 solver.cpp:218] Iteration 5052 (2.11059 iter/s, 5.68562s/12 iters), loss = 0.553489
I0428 14:38:37.903962 30475 solver.cpp:237]     Train net output #0: loss = 0.553489 (* 1 = 0.553489 loss)
I0428 14:38:37.903970 30475 sgd_solver.cpp:105] Iteration 5052, lr = 0.00367611
I0428 14:38:40.066087 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:38:43.595674 30475 solver.cpp:218] Iteration 5064 (2.10833 iter/s, 5.69171s/12 iters), loss = 0.42088
I0428 14:38:43.595721 30475 solver.cpp:237]     Train net output #0: loss = 0.42088 (* 1 = 0.42088 loss)
I0428 14:38:43.595728 30475 sgd_solver.cpp:105] Iteration 5064, lr = 0.00366738
I0428 14:38:49.263844 30475 solver.cpp:218] Iteration 5076 (2.1171 iter/s, 5.66812s/12 iters), loss = 0.349174
I0428 14:38:49.263881 30475 solver.cpp:237]     Train net output #0: loss = 0.349174 (* 1 = 0.349174 loss)
I0428 14:38:49.263890 30475 sgd_solver.cpp:105] Iteration 5076, lr = 0.00365868
I0428 14:38:55.007198 30475 solver.cpp:218] Iteration 5088 (2.08939 iter/s, 5.74331s/12 iters), loss = 0.385922
I0428 14:38:55.007238 30475 solver.cpp:237]     Train net output #0: loss = 0.385922 (* 1 = 0.385922 loss)
I0428 14:38:55.007247 30475 sgd_solver.cpp:105] Iteration 5088, lr = 0.00364999
I0428 14:39:00.110620 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_5100.caffemodel
I0428 14:39:04.032717 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_5100.solverstate
I0428 14:39:09.187060 30475 solver.cpp:330] Iteration 5100, Testing net (#0)
I0428 14:39:09.187078 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:39:11.934566 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:39:14.311403 30475 solver.cpp:397]     Test net output #0: accuracy = 0.392157
I0428 14:39:14.311430 30475 solver.cpp:397]     Test net output #1: loss = 3.65437 (* 1 = 3.65437 loss)
I0428 14:39:14.473088 30475 solver.cpp:218] Iteration 5100 (0.616463 iter/s, 19.4659s/12 iters), loss = 0.412323
I0428 14:39:14.473155 30475 solver.cpp:237]     Train net output #0: loss = 0.412323 (* 1 = 0.412323 loss)
I0428 14:39:14.473165 30475 sgd_solver.cpp:105] Iteration 5100, lr = 0.00364132
I0428 14:39:19.168282 30475 solver.cpp:218] Iteration 5112 (2.55584 iter/s, 4.69512s/12 iters), loss = 0.471152
I0428 14:39:19.168327 30475 solver.cpp:237]     Train net output #0: loss = 0.471152 (* 1 = 0.471152 loss)
I0428 14:39:19.168335 30475 sgd_solver.cpp:105] Iteration 5112, lr = 0.00363268
I0428 14:39:24.679653 30475 solver.cpp:218] Iteration 5124 (2.17734 iter/s, 5.51132s/12 iters), loss = 0.300501
I0428 14:39:24.679693 30475 solver.cpp:237]     Train net output #0: loss = 0.300501 (* 1 = 0.300501 loss)
I0428 14:39:24.679702 30475 sgd_solver.cpp:105] Iteration 5124, lr = 0.00362405
I0428 14:39:30.304860 30475 solver.cpp:218] Iteration 5136 (2.13327 iter/s, 5.62516s/12 iters), loss = 0.332063
I0428 14:39:30.304903 30475 solver.cpp:237]     Train net output #0: loss = 0.332063 (* 1 = 0.332063 loss)
I0428 14:39:30.304913 30475 sgd_solver.cpp:105] Iteration 5136, lr = 0.00361545
I0428 14:39:35.732903 30475 solver.cpp:218] Iteration 5148 (2.21076 iter/s, 5.42799s/12 iters), loss = 0.351556
I0428 14:39:35.732950 30475 solver.cpp:237]     Train net output #0: loss = 0.351556 (* 1 = 0.351556 loss)
I0428 14:39:35.732959 30475 sgd_solver.cpp:105] Iteration 5148, lr = 0.00360687
I0428 14:39:40.180986 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:39:41.287801 30475 solver.cpp:218] Iteration 5160 (2.16028 iter/s, 5.55485s/12 iters), loss = 0.261883
I0428 14:39:41.287847 30475 solver.cpp:237]     Train net output #0: loss = 0.261883 (* 1 = 0.261883 loss)
I0428 14:39:41.287855 30475 sgd_solver.cpp:105] Iteration 5160, lr = 0.0035983
I0428 14:39:46.930248 30475 solver.cpp:218] Iteration 5172 (2.12676 iter/s, 5.6424s/12 iters), loss = 0.334737
I0428 14:39:46.930406 30475 solver.cpp:237]     Train net output #0: loss = 0.334737 (* 1 = 0.334737 loss)
I0428 14:39:46.930416 30475 sgd_solver.cpp:105] Iteration 5172, lr = 0.00358976
I0428 14:39:52.541849 30475 solver.cpp:218] Iteration 5184 (2.13849 iter/s, 5.61144s/12 iters), loss = 0.105574
I0428 14:39:52.541890 30475 solver.cpp:237]     Train net output #0: loss = 0.105574 (* 1 = 0.105574 loss)
I0428 14:39:52.541899 30475 sgd_solver.cpp:105] Iteration 5184, lr = 0.00358124
I0428 14:39:58.063540 30475 solver.cpp:218] Iteration 5196 (2.17327 iter/s, 5.52164s/12 iters), loss = 0.468392
I0428 14:39:58.063577 30475 solver.cpp:237]     Train net output #0: loss = 0.468392 (* 1 = 0.468392 loss)
I0428 14:39:58.063586 30475 sgd_solver.cpp:105] Iteration 5196, lr = 0.00357273
I0428 14:40:00.328493 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_5202.caffemodel
I0428 14:40:03.356953 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_5202.solverstate
I0428 14:40:05.988853 30475 solver.cpp:330] Iteration 5202, Testing net (#0)
I0428 14:40:05.988883 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:40:08.684132 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:40:11.030848 30475 solver.cpp:397]     Test net output #0: accuracy = 0.395221
I0428 14:40:11.030874 30475 solver.cpp:397]     Test net output #1: loss = 3.7151 (* 1 = 3.7151 loss)
I0428 14:40:13.143848 30475 solver.cpp:218] Iteration 5208 (0.795741 iter/s, 15.0803s/12 iters), loss = 0.379225
I0428 14:40:13.143893 30475 solver.cpp:237]     Train net output #0: loss = 0.379225 (* 1 = 0.379225 loss)
I0428 14:40:13.143900 30475 sgd_solver.cpp:105] Iteration 5208, lr = 0.00356425
I0428 14:40:18.673260 30475 solver.cpp:218] Iteration 5220 (2.17023 iter/s, 5.52937s/12 iters), loss = 0.269088
I0428 14:40:18.673373 30475 solver.cpp:237]     Train net output #0: loss = 0.269088 (* 1 = 0.269088 loss)
I0428 14:40:18.673382 30475 sgd_solver.cpp:105] Iteration 5220, lr = 0.00355579
I0428 14:40:24.264914 30475 solver.cpp:218] Iteration 5232 (2.1461 iter/s, 5.59154s/12 iters), loss = 0.306543
I0428 14:40:24.264952 30475 solver.cpp:237]     Train net output #0: loss = 0.306543 (* 1 = 0.306543 loss)
I0428 14:40:24.264961 30475 sgd_solver.cpp:105] Iteration 5232, lr = 0.00354735
I0428 14:40:29.879923 30475 solver.cpp:218] Iteration 5244 (2.13714 iter/s, 5.61497s/12 iters), loss = 0.274406
I0428 14:40:29.879961 30475 solver.cpp:237]     Train net output #0: loss = 0.274406 (* 1 = 0.274406 loss)
I0428 14:40:29.879968 30475 sgd_solver.cpp:105] Iteration 5244, lr = 0.00353892
I0428 14:40:35.511016 30475 solver.cpp:218] Iteration 5256 (2.13104 iter/s, 5.63105s/12 iters), loss = 0.295315
I0428 14:40:35.511059 30475 solver.cpp:237]     Train net output #0: loss = 0.295315 (* 1 = 0.295315 loss)
I0428 14:40:35.511068 30475 sgd_solver.cpp:105] Iteration 5256, lr = 0.00353052
I0428 14:40:36.957547 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:40:41.149049 30475 solver.cpp:218] Iteration 5268 (2.12842 iter/s, 5.63798s/12 iters), loss = 0.300086
I0428 14:40:41.149093 30475 solver.cpp:237]     Train net output #0: loss = 0.300086 (* 1 = 0.300086 loss)
I0428 14:40:41.149102 30475 sgd_solver.cpp:105] Iteration 5268, lr = 0.00352214
I0428 14:40:46.666090 30475 solver.cpp:218] Iteration 5280 (2.1751 iter/s, 5.51698s/12 iters), loss = 0.264141
I0428 14:40:46.666131 30475 solver.cpp:237]     Train net output #0: loss = 0.264141 (* 1 = 0.264141 loss)
I0428 14:40:46.666139 30475 sgd_solver.cpp:105] Iteration 5280, lr = 0.00351378
I0428 14:40:52.282766 30475 solver.cpp:218] Iteration 5292 (2.13651 iter/s, 5.61662s/12 iters), loss = 0.261104
I0428 14:40:52.282938 30475 solver.cpp:237]     Train net output #0: loss = 0.261104 (* 1 = 0.261104 loss)
I0428 14:40:52.282948 30475 sgd_solver.cpp:105] Iteration 5292, lr = 0.00350544
I0428 14:40:57.344902 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_5304.caffemodel
I0428 14:40:59.567145 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_5304.solverstate
I0428 14:41:02.419915 30475 solver.cpp:330] Iteration 5304, Testing net (#0)
I0428 14:41:02.419934 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:41:05.009083 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:41:07.455809 30475 solver.cpp:397]     Test net output #0: accuracy = 0.425245
I0428 14:41:07.455839 30475 solver.cpp:397]     Test net output #1: loss = 3.52759 (* 1 = 3.52759 loss)
I0428 14:41:07.616570 30475 solver.cpp:218] Iteration 5304 (0.782593 iter/s, 15.3336s/12 iters), loss = 0.153986
I0428 14:41:07.616621 30475 solver.cpp:237]     Train net output #0: loss = 0.153986 (* 1 = 0.153986 loss)
I0428 14:41:07.616631 30475 sgd_solver.cpp:105] Iteration 5304, lr = 0.00349711
I0428 14:41:12.329591 30475 solver.cpp:218] Iteration 5316 (2.54617 iter/s, 4.71297s/12 iters), loss = 0.258119
I0428 14:41:12.329623 30475 solver.cpp:237]     Train net output #0: loss = 0.258119 (* 1 = 0.258119 loss)
I0428 14:41:12.329630 30475 sgd_solver.cpp:105] Iteration 5316, lr = 0.00348881
I0428 14:41:17.870385 30475 solver.cpp:218] Iteration 5328 (2.16577 iter/s, 5.54075s/12 iters), loss = 0.234884
I0428 14:41:17.870425 30475 solver.cpp:237]     Train net output #0: loss = 0.234884 (* 1 = 0.234884 loss)
I0428 14:41:17.870434 30475 sgd_solver.cpp:105] Iteration 5328, lr = 0.00348053
I0428 14:41:23.487818 30475 solver.cpp:218] Iteration 5340 (2.13623 iter/s, 5.61738s/12 iters), loss = 0.384423
I0428 14:41:23.487931 30475 solver.cpp:237]     Train net output #0: loss = 0.384423 (* 1 = 0.384423 loss)
I0428 14:41:23.487941 30475 sgd_solver.cpp:105] Iteration 5340, lr = 0.00347226
I0428 14:41:29.206010 30475 solver.cpp:218] Iteration 5352 (2.09861 iter/s, 5.71807s/12 iters), loss = 0.31803
I0428 14:41:29.206055 30475 solver.cpp:237]     Train net output #0: loss = 0.31803 (* 1 = 0.31803 loss)
I0428 14:41:29.206064 30475 sgd_solver.cpp:105] Iteration 5352, lr = 0.00346402
I0428 14:41:33.063925 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:41:34.859881 30475 solver.cpp:218] Iteration 5364 (2.12246 iter/s, 5.65381s/12 iters), loss = 0.287589
I0428 14:41:34.859926 30475 solver.cpp:237]     Train net output #0: loss = 0.287589 (* 1 = 0.287589 loss)
I0428 14:41:34.859935 30475 sgd_solver.cpp:105] Iteration 5364, lr = 0.0034558
I0428 14:41:40.529112 30475 solver.cpp:218] Iteration 5376 (2.11671 iter/s, 5.66917s/12 iters), loss = 0.236762
I0428 14:41:40.529155 30475 solver.cpp:237]     Train net output #0: loss = 0.236762 (* 1 = 0.236762 loss)
I0428 14:41:40.529162 30475 sgd_solver.cpp:105] Iteration 5376, lr = 0.00344759
I0428 14:41:46.201917 30475 solver.cpp:218] Iteration 5388 (2.11552 iter/s, 5.67238s/12 iters), loss = 0.258114
I0428 14:41:46.202008 30475 solver.cpp:237]     Train net output #0: loss = 0.258114 (* 1 = 0.258114 loss)
I0428 14:41:46.202019 30475 sgd_solver.cpp:105] Iteration 5388, lr = 0.00343941
I0428 14:41:51.812613 30475 solver.cpp:218] Iteration 5400 (2.13881 iter/s, 5.6106s/12 iters), loss = 0.246072
I0428 14:41:51.812652 30475 solver.cpp:237]     Train net output #0: loss = 0.246072 (* 1 = 0.246072 loss)
I0428 14:41:51.812661 30475 sgd_solver.cpp:105] Iteration 5400, lr = 0.00343124
I0428 14:41:54.077704 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_5406.caffemodel
I0428 14:41:56.305891 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_5406.solverstate
I0428 14:41:58.072759 30475 solver.cpp:330] Iteration 5406, Testing net (#0)
I0428 14:41:58.072777 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:42:00.721583 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:42:03.251133 30475 solver.cpp:397]     Test net output #0: accuracy = 0.422794
I0428 14:42:03.251166 30475 solver.cpp:397]     Test net output #1: loss = 3.73409 (* 1 = 3.73409 loss)
I0428 14:42:05.358283 30475 solver.cpp:218] Iteration 5412 (0.885895 iter/s, 13.5456s/12 iters), loss = 0.294713
I0428 14:42:05.358325 30475 solver.cpp:237]     Train net output #0: loss = 0.294713 (* 1 = 0.294713 loss)
I0428 14:42:05.358333 30475 sgd_solver.cpp:105] Iteration 5412, lr = 0.00342309
I0428 14:42:10.992038 30475 solver.cpp:218] Iteration 5424 (2.13003 iter/s, 5.63371s/12 iters), loss = 0.297376
I0428 14:42:10.992072 30475 solver.cpp:237]     Train net output #0: loss = 0.297376 (* 1 = 0.297376 loss)
I0428 14:42:10.992080 30475 sgd_solver.cpp:105] Iteration 5424, lr = 0.00341497
I0428 14:42:16.505242 30475 solver.cpp:218] Iteration 5436 (2.17661 iter/s, 5.51317s/12 iters), loss = 0.328091
I0428 14:42:16.505275 30475 solver.cpp:237]     Train net output #0: loss = 0.328091 (* 1 = 0.328091 loss)
I0428 14:42:16.505283 30475 sgd_solver.cpp:105] Iteration 5436, lr = 0.00340686
I0428 14:42:22.190925 30475 solver.cpp:218] Iteration 5448 (2.11058 iter/s, 5.68564s/12 iters), loss = 0.212514
I0428 14:42:22.190968 30475 solver.cpp:237]     Train net output #0: loss = 0.212514 (* 1 = 0.212514 loss)
I0428 14:42:22.190975 30475 sgd_solver.cpp:105] Iteration 5448, lr = 0.00339877
I0428 14:42:27.874220 30475 solver.cpp:218] Iteration 5460 (2.11147 iter/s, 5.68324s/12 iters), loss = 0.263454
I0428 14:42:27.874358 30475 solver.cpp:237]     Train net output #0: loss = 0.263454 (* 1 = 0.263454 loss)
I0428 14:42:27.874368 30475 sgd_solver.cpp:105] Iteration 5460, lr = 0.0033907
I0428 14:42:28.495882 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:42:33.483387 30475 solver.cpp:218] Iteration 5472 (2.13941 iter/s, 5.60902s/12 iters), loss = 0.196515
I0428 14:42:33.483429 30475 solver.cpp:237]     Train net output #0: loss = 0.196515 (* 1 = 0.196515 loss)
I0428 14:42:33.483438 30475 sgd_solver.cpp:105] Iteration 5472, lr = 0.00338265
I0428 14:42:39.159363 30475 solver.cpp:218] Iteration 5484 (2.11419 iter/s, 5.67592s/12 iters), loss = 0.328168
I0428 14:42:39.159408 30475 solver.cpp:237]     Train net output #0: loss = 0.328168 (* 1 = 0.328168 loss)
I0428 14:42:39.159417 30475 sgd_solver.cpp:105] Iteration 5484, lr = 0.00337462
I0428 14:42:44.850627 30475 solver.cpp:218] Iteration 5496 (2.10852 iter/s, 5.69121s/12 iters), loss = 0.232133
I0428 14:42:44.850672 30475 solver.cpp:237]     Train net output #0: loss = 0.232133 (* 1 = 0.232133 loss)
I0428 14:42:44.850682 30475 sgd_solver.cpp:105] Iteration 5496, lr = 0.00336661
I0428 14:42:49.950508 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_5508.caffemodel
I0428 14:42:52.144273 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_5508.solverstate
I0428 14:42:53.836258 30475 solver.cpp:330] Iteration 5508, Testing net (#0)
I0428 14:42:53.836282 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:42:56.328521 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:42:58.844254 30475 solver.cpp:397]     Test net output #0: accuracy = 0.433211
I0428 14:42:58.844429 30475 solver.cpp:397]     Test net output #1: loss = 3.77839 (* 1 = 3.77839 loss)
I0428 14:42:58.996896 30475 solver.cpp:218] Iteration 5508 (0.848283 iter/s, 14.1462s/12 iters), loss = 0.17256
I0428 14:42:58.996942 30475 solver.cpp:237]     Train net output #0: loss = 0.17256 (* 1 = 0.17256 loss)
I0428 14:42:58.996949 30475 sgd_solver.cpp:105] Iteration 5508, lr = 0.00335861
I0428 14:43:03.693610 30475 solver.cpp:218] Iteration 5520 (2.55501 iter/s, 4.69665s/12 iters), loss = 0.27077
I0428 14:43:03.693658 30475 solver.cpp:237]     Train net output #0: loss = 0.27077 (* 1 = 0.27077 loss)
I0428 14:43:03.693667 30475 sgd_solver.cpp:105] Iteration 5520, lr = 0.00335064
I0428 14:43:06.401993 30475 blocking_queue.cpp:49] Waiting for data
I0428 14:43:09.323055 30475 solver.cpp:218] Iteration 5532 (2.13167 iter/s, 5.62939s/12 iters), loss = 0.323298
I0428 14:43:09.323099 30475 solver.cpp:237]     Train net output #0: loss = 0.323298 (* 1 = 0.323298 loss)
I0428 14:43:09.323107 30475 sgd_solver.cpp:105] Iteration 5532, lr = 0.00334268
I0428 14:43:15.008525 30475 solver.cpp:218] Iteration 5544 (2.11066 iter/s, 5.68542s/12 iters), loss = 0.297825
I0428 14:43:15.008563 30475 solver.cpp:237]     Train net output #0: loss = 0.297825 (* 1 = 0.297825 loss)
I0428 14:43:15.008571 30475 sgd_solver.cpp:105] Iteration 5544, lr = 0.00333475
I0428 14:43:20.725977 30475 solver.cpp:218] Iteration 5556 (2.09886 iter/s, 5.7174s/12 iters), loss = 0.258622
I0428 14:43:20.726019 30475 solver.cpp:237]     Train net output #0: loss = 0.258622 (* 1 = 0.258622 loss)
I0428 14:43:20.726028 30475 sgd_solver.cpp:105] Iteration 5556, lr = 0.00332683
I0428 14:43:23.752085 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:43:26.381105 30475 solver.cpp:218] Iteration 5568 (2.12199 iter/s, 5.65508s/12 iters), loss = 0.19686
I0428 14:43:26.381150 30475 solver.cpp:237]     Train net output #0: loss = 0.19686 (* 1 = 0.19686 loss)
I0428 14:43:26.381160 30475 sgd_solver.cpp:105] Iteration 5568, lr = 0.00331893
I0428 14:43:31.990846 30475 solver.cpp:218] Iteration 5580 (2.13916 iter/s, 5.60968s/12 iters), loss = 0.247538
I0428 14:43:31.990983 30475 solver.cpp:237]     Train net output #0: loss = 0.247538 (* 1 = 0.247538 loss)
I0428 14:43:31.990993 30475 sgd_solver.cpp:105] Iteration 5580, lr = 0.00331105
I0428 14:43:37.620066 30475 solver.cpp:218] Iteration 5592 (2.13179 iter/s, 5.62908s/12 iters), loss = 0.338731
I0428 14:43:37.620108 30475 solver.cpp:237]     Train net output #0: loss = 0.338731 (* 1 = 0.338731 loss)
I0428 14:43:37.620117 30475 sgd_solver.cpp:105] Iteration 5592, lr = 0.00330319
I0428 14:43:43.297452 30475 solver.cpp:218] Iteration 5604 (2.11367 iter/s, 5.67733s/12 iters), loss = 0.341057
I0428 14:43:43.297497 30475 solver.cpp:237]     Train net output #0: loss = 0.341057 (* 1 = 0.341057 loss)
I0428 14:43:43.297505 30475 sgd_solver.cpp:105] Iteration 5604, lr = 0.00329535
I0428 14:43:45.548733 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_5610.caffemodel
I0428 14:43:47.732338 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_5610.solverstate
I0428 14:43:49.432451 30475 solver.cpp:330] Iteration 5610, Testing net (#0)
I0428 14:43:49.432469 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:43:51.936534 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:43:54.547029 30475 solver.cpp:397]     Test net output #0: accuracy = 0.43076
I0428 14:43:54.547055 30475 solver.cpp:397]     Test net output #1: loss = 3.69328 (* 1 = 3.69328 loss)
I0428 14:43:56.769070 30475 solver.cpp:218] Iteration 5616 (0.890765 iter/s, 13.4716s/12 iters), loss = 0.239984
I0428 14:43:56.769122 30475 solver.cpp:237]     Train net output #0: loss = 0.239984 (* 1 = 0.239984 loss)
I0428 14:43:56.769131 30475 sgd_solver.cpp:105] Iteration 5616, lr = 0.00328752
I0428 14:44:02.517247 30475 solver.cpp:218] Iteration 5628 (2.08764 iter/s, 5.74812s/12 iters), loss = 0.264277
I0428 14:44:02.517369 30475 solver.cpp:237]     Train net output #0: loss = 0.264277 (* 1 = 0.264277 loss)
I0428 14:44:02.517379 30475 sgd_solver.cpp:105] Iteration 5628, lr = 0.00327972
I0428 14:44:08.134665 30475 solver.cpp:218] Iteration 5640 (2.13626 iter/s, 5.61729s/12 iters), loss = 0.1368
I0428 14:44:08.134699 30475 solver.cpp:237]     Train net output #0: loss = 0.1368 (* 1 = 0.1368 loss)
I0428 14:44:08.134708 30475 sgd_solver.cpp:105] Iteration 5640, lr = 0.00327193
I0428 14:44:13.952162 30475 solver.cpp:218] Iteration 5652 (2.06276 iter/s, 5.81745s/12 iters), loss = 0.294965
I0428 14:44:13.952206 30475 solver.cpp:237]     Train net output #0: loss = 0.294965 (* 1 = 0.294965 loss)
I0428 14:44:13.952215 30475 sgd_solver.cpp:105] Iteration 5652, lr = 0.00326416
I0428 14:44:19.373909 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:44:19.603317 30475 solver.cpp:218] Iteration 5664 (2.12348 iter/s, 5.6511s/12 iters), loss = 0.116082
I0428 14:44:19.603359 30475 solver.cpp:237]     Train net output #0: loss = 0.116082 (* 1 = 0.116082 loss)
I0428 14:44:19.603366 30475 sgd_solver.cpp:105] Iteration 5664, lr = 0.00325641
I0428 14:44:25.240072 30475 solver.cpp:218] Iteration 5676 (2.12891 iter/s, 5.6367s/12 iters), loss = 0.17258
I0428 14:44:25.240118 30475 solver.cpp:237]     Train net output #0: loss = 0.17258 (* 1 = 0.17258 loss)
I0428 14:44:25.240125 30475 sgd_solver.cpp:105] Iteration 5676, lr = 0.00324868
I0428 14:44:30.896121 30475 solver.cpp:218] Iteration 5688 (2.12164 iter/s, 5.65599s/12 iters), loss = 0.24021
I0428 14:44:30.896163 30475 solver.cpp:237]     Train net output #0: loss = 0.24021 (* 1 = 0.24021 loss)
I0428 14:44:30.896173 30475 sgd_solver.cpp:105] Iteration 5688, lr = 0.00324097
I0428 14:44:36.518908 30475 solver.cpp:218] Iteration 5700 (2.13419 iter/s, 5.62273s/12 iters), loss = 0.261218
I0428 14:44:36.519026 30475 solver.cpp:237]     Train net output #0: loss = 0.261218 (* 1 = 0.261218 loss)
I0428 14:44:36.519034 30475 sgd_solver.cpp:105] Iteration 5700, lr = 0.00323328
I0428 14:44:41.592927 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_5712.caffemodel
I0428 14:44:43.781718 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_5712.solverstate
I0428 14:44:45.476909 30475 solver.cpp:330] Iteration 5712, Testing net (#0)
I0428 14:44:45.476927 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:44:47.938500 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:44:50.593905 30475 solver.cpp:397]     Test net output #0: accuracy = 0.431985
I0428 14:44:50.593955 30475 solver.cpp:397]     Test net output #1: loss = 3.82186 (* 1 = 3.82186 loss)
I0428 14:44:50.756049 30475 solver.cpp:218] Iteration 5712 (0.842873 iter/s, 14.237s/12 iters), loss = 0.137598
I0428 14:44:50.756114 30475 solver.cpp:237]     Train net output #0: loss = 0.137598 (* 1 = 0.137598 loss)
I0428 14:44:50.756124 30475 sgd_solver.cpp:105] Iteration 5712, lr = 0.0032256
I0428 14:44:55.512404 30475 solver.cpp:218] Iteration 5724 (2.52298 iter/s, 4.75628s/12 iters), loss = 0.178905
I0428 14:44:55.512455 30475 solver.cpp:237]     Train net output #0: loss = 0.178905 (* 1 = 0.178905 loss)
I0428 14:44:55.512465 30475 sgd_solver.cpp:105] Iteration 5724, lr = 0.00321794
I0428 14:45:01.152411 30475 solver.cpp:218] Iteration 5736 (2.12768 iter/s, 5.63995s/12 iters), loss = 0.244328
I0428 14:45:01.152441 30475 solver.cpp:237]     Train net output #0: loss = 0.244328 (* 1 = 0.244328 loss)
I0428 14:45:01.152449 30475 sgd_solver.cpp:105] Iteration 5736, lr = 0.0032103
I0428 14:45:06.782097 30475 solver.cpp:218] Iteration 5748 (2.13157 iter/s, 5.62965s/12 iters), loss = 0.257732
I0428 14:45:06.782212 30475 solver.cpp:237]     Train net output #0: loss = 0.257732 (* 1 = 0.257732 loss)
I0428 14:45:06.782222 30475 sgd_solver.cpp:105] Iteration 5748, lr = 0.00320268
I0428 14:45:12.315994 30475 solver.cpp:218] Iteration 5760 (2.1685 iter/s, 5.53378s/12 iters), loss = 0.241809
I0428 14:45:12.316033 30475 solver.cpp:237]     Train net output #0: loss = 0.241809 (* 1 = 0.241809 loss)
I0428 14:45:12.316041 30475 sgd_solver.cpp:105] Iteration 5760, lr = 0.00319508
I0428 14:45:14.497359 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:45:17.974838 30475 solver.cpp:218] Iteration 5772 (2.12059 iter/s, 5.65879s/12 iters), loss = 0.118994
I0428 14:45:17.974881 30475 solver.cpp:237]     Train net output #0: loss = 0.118994 (* 1 = 0.118994 loss)
I0428 14:45:17.974889 30475 sgd_solver.cpp:105] Iteration 5772, lr = 0.00318749
I0428 14:45:23.592329 30475 solver.cpp:218] Iteration 5784 (2.1362 iter/s, 5.61744s/12 iters), loss = 0.215914
I0428 14:45:23.592366 30475 solver.cpp:237]     Train net output #0: loss = 0.215914 (* 1 = 0.215914 loss)
I0428 14:45:23.592375 30475 sgd_solver.cpp:105] Iteration 5784, lr = 0.00317992
I0428 14:45:29.217491 30475 solver.cpp:218] Iteration 5796 (2.13329 iter/s, 5.62511s/12 iters), loss = 0.257776
I0428 14:45:29.217538 30475 solver.cpp:237]     Train net output #0: loss = 0.257776 (* 1 = 0.257776 loss)
I0428 14:45:29.217547 30475 sgd_solver.cpp:105] Iteration 5796, lr = 0.00317237
I0428 14:45:34.862879 30475 solver.cpp:218] Iteration 5808 (2.12565 iter/s, 5.64534s/12 iters), loss = 0.278923
I0428 14:45:34.862918 30475 solver.cpp:237]     Train net output #0: loss = 0.278923 (* 1 = 0.278923 loss)
I0428 14:45:34.862926 30475 sgd_solver.cpp:105] Iteration 5808, lr = 0.00316484
I0428 14:45:37.106472 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_5814.caffemodel
I0428 14:45:40.208721 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_5814.solverstate
I0428 14:45:42.746309 30475 solver.cpp:330] Iteration 5814, Testing net (#0)
I0428 14:45:42.746336 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:45:45.111276 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:45:47.639472 30475 solver.cpp:397]     Test net output #0: accuracy = 0.421569
I0428 14:45:47.639513 30475 solver.cpp:397]     Test net output #1: loss = 3.76278 (* 1 = 3.76278 loss)
I0428 14:45:49.745244 30475 solver.cpp:218] Iteration 5820 (0.806325 iter/s, 14.8823s/12 iters), loss = 0.186828
I0428 14:45:49.745286 30475 solver.cpp:237]     Train net output #0: loss = 0.186828 (* 1 = 0.186828 loss)
I0428 14:45:49.745294 30475 sgd_solver.cpp:105] Iteration 5820, lr = 0.00315733
I0428 14:45:55.466073 30475 solver.cpp:218] Iteration 5832 (2.09762 iter/s, 5.72078s/12 iters), loss = 0.156113
I0428 14:45:55.466118 30475 solver.cpp:237]     Train net output #0: loss = 0.156113 (* 1 = 0.156113 loss)
I0428 14:45:55.466126 30475 sgd_solver.cpp:105] Iteration 5832, lr = 0.00314983
I0428 14:46:01.129539 30475 solver.cpp:218] Iteration 5844 (2.11886 iter/s, 5.66341s/12 iters), loss = 0.102424
I0428 14:46:01.129586 30475 solver.cpp:237]     Train net output #0: loss = 0.102424 (* 1 = 0.102424 loss)
I0428 14:46:01.129595 30475 sgd_solver.cpp:105] Iteration 5844, lr = 0.00314235
I0428 14:46:06.779503 30475 solver.cpp:218] Iteration 5856 (2.12393 iter/s, 5.64991s/12 iters), loss = 0.224164
I0428 14:46:06.779543 30475 solver.cpp:237]     Train net output #0: loss = 0.224164 (* 1 = 0.224164 loss)
I0428 14:46:06.779551 30475 sgd_solver.cpp:105] Iteration 5856, lr = 0.00313489
I0428 14:46:11.525940 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:46:12.456857 30475 solver.cpp:218] Iteration 5868 (2.11368 iter/s, 5.6773s/12 iters), loss = 0.0994823
I0428 14:46:12.456905 30475 solver.cpp:237]     Train net output #0: loss = 0.0994823 (* 1 = 0.0994823 loss)
I0428 14:46:12.456913 30475 sgd_solver.cpp:105] Iteration 5868, lr = 0.00312745
I0428 14:46:18.132972 30475 solver.cpp:218] Iteration 5880 (2.11414 iter/s, 5.67606s/12 iters), loss = 0.156155
I0428 14:46:18.133013 30475 solver.cpp:237]     Train net output #0: loss = 0.156155 (* 1 = 0.156155 loss)
I0428 14:46:18.133020 30475 sgd_solver.cpp:105] Iteration 5880, lr = 0.00312002
I0428 14:46:23.793306 30475 solver.cpp:218] Iteration 5892 (2.12003 iter/s, 5.66029s/12 iters), loss = 0.21734
I0428 14:46:23.793339 30475 solver.cpp:237]     Train net output #0: loss = 0.21734 (* 1 = 0.21734 loss)
I0428 14:46:23.793347 30475 sgd_solver.cpp:105] Iteration 5892, lr = 0.00311262
I0428 14:46:29.414624 30475 solver.cpp:218] Iteration 5904 (2.13475 iter/s, 5.62127s/12 iters), loss = 0.210142
I0428 14:46:29.414676 30475 solver.cpp:237]     Train net output #0: loss = 0.210142 (* 1 = 0.210142 loss)
I0428 14:46:29.414685 30475 sgd_solver.cpp:105] Iteration 5904, lr = 0.00310523
I0428 14:46:34.297156 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_5916.caffemodel
I0428 14:46:36.843768 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_5916.solverstate
I0428 14:46:40.250211 30475 solver.cpp:330] Iteration 5916, Testing net (#0)
I0428 14:46:40.250241 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:46:42.629920 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:46:45.385974 30475 solver.cpp:397]     Test net output #0: accuracy = 0.428309
I0428 14:46:45.386009 30475 solver.cpp:397]     Test net output #1: loss = 3.82418 (* 1 = 3.82418 loss)
I0428 14:46:45.546869 30475 solver.cpp:218] Iteration 5916 (0.743854 iter/s, 16.1322s/12 iters), loss = 0.123688
I0428 14:46:45.546917 30475 solver.cpp:237]     Train net output #0: loss = 0.123688 (* 1 = 0.123688 loss)
I0428 14:46:45.546926 30475 sgd_solver.cpp:105] Iteration 5916, lr = 0.00309785
I0428 14:46:50.265480 30475 solver.cpp:218] Iteration 5928 (2.54315 iter/s, 4.71855s/12 iters), loss = 0.16178
I0428 14:46:50.265525 30475 solver.cpp:237]     Train net output #0: loss = 0.16178 (* 1 = 0.16178 loss)
I0428 14:46:50.265533 30475 sgd_solver.cpp:105] Iteration 5928, lr = 0.0030905
I0428 14:46:55.910615 30475 solver.cpp:218] Iteration 5940 (2.12574 iter/s, 5.64508s/12 iters), loss = 0.159647
I0428 14:46:55.910653 30475 solver.cpp:237]     Train net output #0: loss = 0.159647 (* 1 = 0.159647 loss)
I0428 14:46:55.910660 30475 sgd_solver.cpp:105] Iteration 5940, lr = 0.00308316
I0428 14:47:01.565886 30475 solver.cpp:218] Iteration 5952 (2.12193 iter/s, 5.65522s/12 iters), loss = 0.23238
I0428 14:47:01.565932 30475 solver.cpp:237]     Train net output #0: loss = 0.23238 (* 1 = 0.23238 loss)
I0428 14:47:01.565940 30475 sgd_solver.cpp:105] Iteration 5952, lr = 0.00307584
I0428 14:47:07.106331 30475 solver.cpp:218] Iteration 5964 (2.16591 iter/s, 5.54039s/12 iters), loss = 0.115912
I0428 14:47:07.106376 30475 solver.cpp:237]     Train net output #0: loss = 0.115912 (* 1 = 0.115912 loss)
I0428 14:47:07.106385 30475 sgd_solver.cpp:105] Iteration 5964, lr = 0.00306854
I0428 14:47:08.584295 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:47:12.750445 30475 solver.cpp:218] Iteration 5976 (2.12613 iter/s, 5.64405s/12 iters), loss = 0.176857
I0428 14:47:12.750613 30475 solver.cpp:237]     Train net output #0: loss = 0.176857 (* 1 = 0.176857 loss)
I0428 14:47:12.750627 30475 sgd_solver.cpp:105] Iteration 5976, lr = 0.00306125
I0428 14:47:18.374094 30475 solver.cpp:218] Iteration 5988 (2.13391 iter/s, 5.62349s/12 iters), loss = 0.153721
I0428 14:47:18.374143 30475 solver.cpp:237]     Train net output #0: loss = 0.153721 (* 1 = 0.153721 loss)
I0428 14:47:18.374151 30475 sgd_solver.cpp:105] Iteration 5988, lr = 0.00305398
I0428 14:47:24.015655 30475 solver.cpp:218] Iteration 6000 (2.12709 iter/s, 5.6415s/12 iters), loss = 0.175496
I0428 14:47:24.015699 30475 solver.cpp:237]     Train net output #0: loss = 0.175496 (* 1 = 0.175496 loss)
I0428 14:47:24.015708 30475 sgd_solver.cpp:105] Iteration 6000, lr = 0.00304673
I0428 14:47:29.644933 30475 solver.cpp:218] Iteration 6012 (2.13173 iter/s, 5.62923s/12 iters), loss = 0.265391
I0428 14:47:29.644979 30475 solver.cpp:237]     Train net output #0: loss = 0.265392 (* 1 = 0.265392 loss)
I0428 14:47:29.644987 30475 sgd_solver.cpp:105] Iteration 6012, lr = 0.0030395
I0428 14:47:31.883786 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6018.caffemodel
I0428 14:47:35.919229 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6018.solverstate
I0428 14:47:38.664686 30475 solver.cpp:330] Iteration 6018, Testing net (#0)
I0428 14:47:38.664712 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:47:40.996203 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:47:43.819985 30475 solver.cpp:397]     Test net output #0: accuracy = 0.450368
I0428 14:47:43.820197 30475 solver.cpp:397]     Test net output #1: loss = 3.64271 (* 1 = 3.64271 loss)
I0428 14:47:46.047677 30475 solver.cpp:218] Iteration 6024 (0.731587 iter/s, 16.4027s/12 iters), loss = 0.149928
I0428 14:47:46.047724 30475 solver.cpp:237]     Train net output #0: loss = 0.149928 (* 1 = 0.149928 loss)
I0428 14:47:46.047734 30475 sgd_solver.cpp:105] Iteration 6024, lr = 0.00303228
I0428 14:47:51.883181 30475 solver.cpp:218] Iteration 6036 (2.0564 iter/s, 5.83545s/12 iters), loss = 0.274803
I0428 14:47:51.883222 30475 solver.cpp:237]     Train net output #0: loss = 0.274804 (* 1 = 0.274804 loss)
I0428 14:47:51.883230 30475 sgd_solver.cpp:105] Iteration 6036, lr = 0.00302508
I0428 14:47:57.537919 30475 solver.cpp:218] Iteration 6048 (2.12213 iter/s, 5.65469s/12 iters), loss = 0.219472
I0428 14:47:57.537966 30475 solver.cpp:237]     Train net output #0: loss = 0.219472 (* 1 = 0.219472 loss)
I0428 14:47:57.537974 30475 sgd_solver.cpp:105] Iteration 6048, lr = 0.0030179
I0428 14:48:03.182637 30475 solver.cpp:218] Iteration 6060 (2.1259 iter/s, 5.64466s/12 iters), loss = 0.196293
I0428 14:48:03.182685 30475 solver.cpp:237]     Train net output #0: loss = 0.196293 (* 1 = 0.196293 loss)
I0428 14:48:03.182694 30475 sgd_solver.cpp:105] Iteration 6060, lr = 0.00301074
I0428 14:48:07.074322 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:48:08.829404 30475 solver.cpp:218] Iteration 6072 (2.12513 iter/s, 5.64671s/12 iters), loss = 0.125658
I0428 14:48:08.829447 30475 solver.cpp:237]     Train net output #0: loss = 0.125658 (* 1 = 0.125658 loss)
I0428 14:48:08.829455 30475 sgd_solver.cpp:105] Iteration 6072, lr = 0.00300359
I0428 14:48:14.371340 30475 solver.cpp:218] Iteration 6084 (2.16533 iter/s, 5.54188s/12 iters), loss = 0.19078
I0428 14:48:14.371469 30475 solver.cpp:237]     Train net output #0: loss = 0.19078 (* 1 = 0.19078 loss)
I0428 14:48:14.371479 30475 sgd_solver.cpp:105] Iteration 6084, lr = 0.00299646
I0428 14:48:20.020499 30475 solver.cpp:218] Iteration 6096 (2.12426 iter/s, 5.64902s/12 iters), loss = 0.130642
I0428 14:48:20.020547 30475 solver.cpp:237]     Train net output #0: loss = 0.130642 (* 1 = 0.130642 loss)
I0428 14:48:20.020556 30475 sgd_solver.cpp:105] Iteration 6096, lr = 0.00298934
I0428 14:48:25.543498 30475 solver.cpp:218] Iteration 6108 (2.17276 iter/s, 5.52294s/12 iters), loss = 0.0966013
I0428 14:48:25.543545 30475 solver.cpp:237]     Train net output #0: loss = 0.0966014 (* 1 = 0.0966014 loss)
I0428 14:48:25.543555 30475 sgd_solver.cpp:105] Iteration 6108, lr = 0.00298225
I0428 14:48:30.613624 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6120.caffemodel
I0428 14:48:35.080333 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6120.solverstate
I0428 14:48:37.741756 30475 solver.cpp:330] Iteration 6120, Testing net (#0)
I0428 14:48:37.741784 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:48:40.074009 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:48:43.027139 30475 solver.cpp:397]     Test net output #0: accuracy = 0.449142
I0428 14:48:43.027170 30475 solver.cpp:397]     Test net output #1: loss = 3.7787 (* 1 = 3.7787 loss)
I0428 14:48:43.188712 30475 solver.cpp:218] Iteration 6120 (0.680072 iter/s, 17.6452s/12 iters), loss = 0.176103
I0428 14:48:43.188753 30475 solver.cpp:237]     Train net output #0: loss = 0.176103 (* 1 = 0.176103 loss)
I0428 14:48:43.188763 30475 sgd_solver.cpp:105] Iteration 6120, lr = 0.00297517
I0428 14:48:47.895119 30475 solver.cpp:218] Iteration 6132 (2.54974 iter/s, 4.70636s/12 iters), loss = 0.0685038
I0428 14:48:47.895251 30475 solver.cpp:237]     Train net output #0: loss = 0.0685039 (* 1 = 0.0685039 loss)
I0428 14:48:47.895260 30475 sgd_solver.cpp:105] Iteration 6132, lr = 0.0029681
I0428 14:48:53.415689 30475 solver.cpp:218] Iteration 6144 (2.17374 iter/s, 5.52043s/12 iters), loss = 0.109014
I0428 14:48:53.415729 30475 solver.cpp:237]     Train net output #0: loss = 0.109014 (* 1 = 0.109014 loss)
I0428 14:48:53.415736 30475 sgd_solver.cpp:105] Iteration 6144, lr = 0.00296105
I0428 14:48:59.052336 30475 solver.cpp:218] Iteration 6156 (2.12894 iter/s, 5.6366s/12 iters), loss = 0.148062
I0428 14:48:59.052381 30475 solver.cpp:237]     Train net output #0: loss = 0.148062 (* 1 = 0.148062 loss)
I0428 14:48:59.052389 30475 sgd_solver.cpp:105] Iteration 6156, lr = 0.00295402
I0428 14:49:04.580617 30475 solver.cpp:218] Iteration 6168 (2.17068 iter/s, 5.52823s/12 iters), loss = 0.134345
I0428 14:49:04.580662 30475 solver.cpp:237]     Train net output #0: loss = 0.134345 (* 1 = 0.134345 loss)
I0428 14:49:04.580669 30475 sgd_solver.cpp:105] Iteration 6168, lr = 0.00294701
I0428 14:49:05.228655 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:49:10.230454 30475 solver.cpp:218] Iteration 6180 (2.12397 iter/s, 5.64979s/12 iters), loss = 0.115619
I0428 14:49:10.230499 30475 solver.cpp:237]     Train net output #0: loss = 0.115619 (* 1 = 0.115619 loss)
I0428 14:49:10.230509 30475 sgd_solver.cpp:105] Iteration 6180, lr = 0.00294001
I0428 14:49:15.842142 30475 solver.cpp:218] Iteration 6192 (2.13841 iter/s, 5.61163s/12 iters), loss = 0.224859
I0428 14:49:15.842188 30475 solver.cpp:237]     Train net output #0: loss = 0.22486 (* 1 = 0.22486 loss)
I0428 14:49:15.842197 30475 sgd_solver.cpp:105] Iteration 6192, lr = 0.00293303
I0428 14:49:21.397670 30475 solver.cpp:218] Iteration 6204 (2.16003 iter/s, 5.55547s/12 iters), loss = 0.0918527
I0428 14:49:21.397868 30475 solver.cpp:237]     Train net output #0: loss = 0.0918528 (* 1 = 0.0918528 loss)
I0428 14:49:21.397877 30475 sgd_solver.cpp:105] Iteration 6204, lr = 0.00292607
I0428 14:49:26.946841 30475 solver.cpp:218] Iteration 6216 (2.16256 iter/s, 5.54897s/12 iters), loss = 0.0819754
I0428 14:49:26.946888 30475 solver.cpp:237]     Train net output #0: loss = 0.0819755 (* 1 = 0.0819755 loss)
I0428 14:49:26.946897 30475 sgd_solver.cpp:105] Iteration 6216, lr = 0.00291912
I0428 14:49:29.187748 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6222.caffemodel
I0428 14:49:31.368942 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6222.solverstate
I0428 14:49:33.064249 30475 solver.cpp:330] Iteration 6222, Testing net (#0)
I0428 14:49:33.064270 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:49:35.319375 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:49:36.834410 30475 blocking_queue.cpp:49] Waiting for data
I0428 14:49:38.226531 30475 solver.cpp:397]     Test net output #0: accuracy = 0.453431
I0428 14:49:38.226569 30475 solver.cpp:397]     Test net output #1: loss = 3.66545 (* 1 = 3.66545 loss)
I0428 14:49:40.361160 30475 solver.cpp:218] Iteration 6228 (0.894569 iter/s, 13.4143s/12 iters), loss = 0.0832416
I0428 14:49:40.361205 30475 solver.cpp:237]     Train net output #0: loss = 0.0832417 (* 1 = 0.0832417 loss)
I0428 14:49:40.361214 30475 sgd_solver.cpp:105] Iteration 6228, lr = 0.00291219
I0428 14:49:46.058406 30475 solver.cpp:218] Iteration 6240 (2.1063 iter/s, 5.6972s/12 iters), loss = 0.118989
I0428 14:49:46.058447 30475 solver.cpp:237]     Train net output #0: loss = 0.118989 (* 1 = 0.118989 loss)
I0428 14:49:46.058455 30475 sgd_solver.cpp:105] Iteration 6240, lr = 0.00290528
I0428 14:49:51.843495 30475 solver.cpp:218] Iteration 6252 (2.07432 iter/s, 5.78504s/12 iters), loss = 0.175629
I0428 14:49:51.843667 30475 solver.cpp:237]     Train net output #0: loss = 0.175629 (* 1 = 0.175629 loss)
I0428 14:49:51.843677 30475 sgd_solver.cpp:105] Iteration 6252, lr = 0.00289838
I0428 14:49:57.487985 30475 solver.cpp:218] Iteration 6264 (2.12603 iter/s, 5.64431s/12 iters), loss = 0.0773531
I0428 14:49:57.488026 30475 solver.cpp:237]     Train net output #0: loss = 0.0773531 (* 1 = 0.0773531 loss)
I0428 14:49:57.488034 30475 sgd_solver.cpp:105] Iteration 6264, lr = 0.0028915
I0428 14:50:00.536550 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:50:03.185247 30475 solver.cpp:218] Iteration 6276 (2.10629 iter/s, 5.69721s/12 iters), loss = 0.154919
I0428 14:50:03.185290 30475 solver.cpp:237]     Train net output #0: loss = 0.154919 (* 1 = 0.154919 loss)
I0428 14:50:03.185299 30475 sgd_solver.cpp:105] Iteration 6276, lr = 0.00288463
I0428 14:50:08.815444 30475 solver.cpp:218] Iteration 6288 (2.13138 iter/s, 5.63014s/12 iters), loss = 0.111654
I0428 14:50:08.815487 30475 solver.cpp:237]     Train net output #0: loss = 0.111654 (* 1 = 0.111654 loss)
I0428 14:50:08.815495 30475 sgd_solver.cpp:105] Iteration 6288, lr = 0.00287779
I0428 14:50:14.467679 30475 solver.cpp:218] Iteration 6300 (2.12308 iter/s, 5.65218s/12 iters), loss = 0.0637102
I0428 14:50:14.467725 30475 solver.cpp:237]     Train net output #0: loss = 0.0637103 (* 1 = 0.0637103 loss)
I0428 14:50:14.467733 30475 sgd_solver.cpp:105] Iteration 6300, lr = 0.00287095
I0428 14:50:20.020803 30475 solver.cpp:218] Iteration 6312 (2.16097 iter/s, 5.55307s/12 iters), loss = 0.11073
I0428 14:50:20.020843 30475 solver.cpp:237]     Train net output #0: loss = 0.11073 (* 1 = 0.11073 loss)
I0428 14:50:20.020853 30475 sgd_solver.cpp:105] Iteration 6312, lr = 0.00286414
I0428 14:50:24.890606 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6324.caffemodel
I0428 14:50:27.090003 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6324.solverstate
I0428 14:50:28.818485 30475 solver.cpp:330] Iteration 6324, Testing net (#0)
I0428 14:50:28.818504 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:50:31.034584 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:50:33.984741 30475 solver.cpp:397]     Test net output #0: accuracy = 0.452206
I0428 14:50:33.984786 30475 solver.cpp:397]     Test net output #1: loss = 3.68755 (* 1 = 3.68755 loss)
I0428 14:50:34.146147 30475 solver.cpp:218] Iteration 6324 (0.849539 iter/s, 14.1253s/12 iters), loss = 0.152719
I0428 14:50:34.146205 30475 solver.cpp:237]     Train net output #0: loss = 0.152719 (* 1 = 0.152719 loss)
I0428 14:50:34.146214 30475 sgd_solver.cpp:105] Iteration 6324, lr = 0.00285734
I0428 14:50:38.886283 30475 solver.cpp:218] Iteration 6336 (2.53161 iter/s, 4.74007s/12 iters), loss = 0.140261
I0428 14:50:38.886329 30475 solver.cpp:237]     Train net output #0: loss = 0.140261 (* 1 = 0.140261 loss)
I0428 14:50:38.886338 30475 sgd_solver.cpp:105] Iteration 6336, lr = 0.00285055
I0428 14:50:44.429440 30475 solver.cpp:218] Iteration 6348 (2.16485 iter/s, 5.5431s/12 iters), loss = 0.0711775
I0428 14:50:44.429482 30475 solver.cpp:237]     Train net output #0: loss = 0.0711776 (* 1 = 0.0711776 loss)
I0428 14:50:44.429489 30475 sgd_solver.cpp:105] Iteration 6348, lr = 0.00284379
I0428 14:50:50.186512 30475 solver.cpp:218] Iteration 6360 (2.08441 iter/s, 5.75702s/12 iters), loss = 0.0814629
I0428 14:50:50.186555 30475 solver.cpp:237]     Train net output #0: loss = 0.081463 (* 1 = 0.081463 loss)
I0428 14:50:50.186564 30475 sgd_solver.cpp:105] Iteration 6360, lr = 0.00283703
I0428 14:50:55.742630 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:50:55.943275 30475 solver.cpp:218] Iteration 6372 (2.08452 iter/s, 5.75671s/12 iters), loss = 0.172923
I0428 14:50:55.943315 30475 solver.cpp:237]     Train net output #0: loss = 0.172923 (* 1 = 0.172923 loss)
I0428 14:50:55.943325 30475 sgd_solver.cpp:105] Iteration 6372, lr = 0.0028303
I0428 14:51:01.691679 30475 solver.cpp:218] Iteration 6384 (2.08755 iter/s, 5.74836s/12 iters), loss = 0.163987
I0428 14:51:01.691720 30475 solver.cpp:237]     Train net output #0: loss = 0.163988 (* 1 = 0.163988 loss)
I0428 14:51:01.691728 30475 sgd_solver.cpp:105] Iteration 6384, lr = 0.00282358
I0428 14:51:07.324532 30475 solver.cpp:218] Iteration 6396 (2.13038 iter/s, 5.6328s/12 iters), loss = 0.0708687
I0428 14:51:07.324580 30475 solver.cpp:237]     Train net output #0: loss = 0.0708688 (* 1 = 0.0708688 loss)
I0428 14:51:07.324589 30475 sgd_solver.cpp:105] Iteration 6396, lr = 0.00281687
I0428 14:51:12.967178 30475 solver.cpp:218] Iteration 6408 (2.12668 iter/s, 5.64259s/12 iters), loss = 0.070461
I0428 14:51:12.967226 30475 solver.cpp:237]     Train net output #0: loss = 0.070461 (* 1 = 0.070461 loss)
I0428 14:51:12.967236 30475 sgd_solver.cpp:105] Iteration 6408, lr = 0.00281019
I0428 14:51:18.526371 30475 solver.cpp:218] Iteration 6420 (2.15861 iter/s, 5.55913s/12 iters), loss = 0.0875261
I0428 14:51:18.526433 30475 solver.cpp:237]     Train net output #0: loss = 0.0875262 (* 1 = 0.0875262 loss)
I0428 14:51:18.526448 30475 sgd_solver.cpp:105] Iteration 6420, lr = 0.00280351
I0428 14:51:20.770066 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6426.caffemodel
I0428 14:51:22.973333 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6426.solverstate
I0428 14:51:24.672343 30475 solver.cpp:330] Iteration 6426, Testing net (#0)
I0428 14:51:24.672369 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:51:26.979050 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:51:30.154780 30475 solver.cpp:397]     Test net output #0: accuracy = 0.447304
I0428 14:51:30.154812 30475 solver.cpp:397]     Test net output #1: loss = 3.71233 (* 1 = 3.71233 loss)
I0428 14:51:32.263070 30475 solver.cpp:218] Iteration 6432 (0.873575 iter/s, 13.7367s/12 iters), loss = 0.173978
I0428 14:51:32.263110 30475 solver.cpp:237]     Train net output #0: loss = 0.173978 (* 1 = 0.173978 loss)
I0428 14:51:32.263118 30475 sgd_solver.cpp:105] Iteration 6432, lr = 0.00279686
I0428 14:51:37.850615 30475 solver.cpp:218] Iteration 6444 (2.14766 iter/s, 5.58749s/12 iters), loss = 0.105734
I0428 14:51:37.850656 30475 solver.cpp:237]     Train net output #0: loss = 0.105734 (* 1 = 0.105734 loss)
I0428 14:51:37.850664 30475 sgd_solver.cpp:105] Iteration 6444, lr = 0.00279022
I0428 14:51:43.414064 30475 solver.cpp:218] Iteration 6456 (2.15696 iter/s, 5.5634s/12 iters), loss = 0.0635445
I0428 14:51:43.414111 30475 solver.cpp:237]     Train net output #0: loss = 0.0635446 (* 1 = 0.0635446 loss)
I0428 14:51:43.414120 30475 sgd_solver.cpp:105] Iteration 6456, lr = 0.00278359
I0428 14:51:49.075711 30475 solver.cpp:218] Iteration 6468 (2.11955 iter/s, 5.66159s/12 iters), loss = 0.0995724
I0428 14:51:49.075753 30475 solver.cpp:237]     Train net output #0: loss = 0.0995725 (* 1 = 0.0995725 loss)
I0428 14:51:49.075762 30475 sgd_solver.cpp:105] Iteration 6468, lr = 0.00277698
I0428 14:51:51.283224 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:51:54.779119 30475 solver.cpp:218] Iteration 6480 (2.10402 iter/s, 5.70336s/12 iters), loss = 0.105355
I0428 14:51:54.779163 30475 solver.cpp:237]     Train net output #0: loss = 0.105355 (* 1 = 0.105355 loss)
I0428 14:51:54.779172 30475 sgd_solver.cpp:105] Iteration 6480, lr = 0.00277039
I0428 14:52:00.406806 30475 solver.cpp:218] Iteration 6492 (2.13234 iter/s, 5.62763s/12 iters), loss = 0.102052
I0428 14:52:00.406908 30475 solver.cpp:237]     Train net output #0: loss = 0.102052 (* 1 = 0.102052 loss)
I0428 14:52:00.406917 30475 sgd_solver.cpp:105] Iteration 6492, lr = 0.00276381
I0428 14:52:06.042304 30475 solver.cpp:218] Iteration 6504 (2.1294 iter/s, 5.63539s/12 iters), loss = 0.0481655
I0428 14:52:06.042346 30475 solver.cpp:237]     Train net output #0: loss = 0.0481656 (* 1 = 0.0481656 loss)
I0428 14:52:06.042356 30475 sgd_solver.cpp:105] Iteration 6504, lr = 0.00275725
I0428 14:52:11.747220 30475 solver.cpp:218] Iteration 6516 (2.10347 iter/s, 5.70487s/12 iters), loss = 0.202781
I0428 14:52:11.747267 30475 solver.cpp:237]     Train net output #0: loss = 0.202781 (* 1 = 0.202781 loss)
I0428 14:52:11.747274 30475 sgd_solver.cpp:105] Iteration 6516, lr = 0.00275071
I0428 14:52:16.846966 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6528.caffemodel
I0428 14:52:19.597896 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6528.solverstate
I0428 14:52:24.373018 30475 solver.cpp:330] Iteration 6528, Testing net (#0)
I0428 14:52:24.373045 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:52:26.453771 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:52:29.450219 30475 solver.cpp:397]     Test net output #0: accuracy = 0.433211
I0428 14:52:29.450258 30475 solver.cpp:397]     Test net output #1: loss = 3.76614 (* 1 = 3.76614 loss)
I0428 14:52:29.611346 30475 solver.cpp:218] Iteration 6528 (0.671738 iter/s, 17.8641s/12 iters), loss = 0.0881352
I0428 14:52:29.611389 30475 solver.cpp:237]     Train net output #0: loss = 0.0881353 (* 1 = 0.0881353 loss)
I0428 14:52:29.611398 30475 sgd_solver.cpp:105] Iteration 6528, lr = 0.00274418
I0428 14:52:34.525372 30475 solver.cpp:218] Iteration 6540 (2.44202 iter/s, 4.91397s/12 iters), loss = 0.180651
I0428 14:52:34.525517 30475 solver.cpp:237]     Train net output #0: loss = 0.180651 (* 1 = 0.180651 loss)
I0428 14:52:34.525527 30475 sgd_solver.cpp:105] Iteration 6540, lr = 0.00273766
I0428 14:52:40.251945 30475 solver.cpp:218] Iteration 6552 (2.09555 iter/s, 5.72642s/12 iters), loss = 0.0487486
I0428 14:52:40.251996 30475 solver.cpp:237]     Train net output #0: loss = 0.0487486 (* 1 = 0.0487486 loss)
I0428 14:52:40.252003 30475 sgd_solver.cpp:105] Iteration 6552, lr = 0.00273116
I0428 14:52:45.765272 30475 solver.cpp:218] Iteration 6564 (2.17657 iter/s, 5.51327s/12 iters), loss = 0.0265631
I0428 14:52:45.765317 30475 solver.cpp:237]     Train net output #0: loss = 0.0265632 (* 1 = 0.0265632 loss)
I0428 14:52:45.765324 30475 sgd_solver.cpp:105] Iteration 6564, lr = 0.00272468
I0428 14:52:50.602916 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:52:51.497426 30475 solver.cpp:218] Iteration 6576 (2.09347 iter/s, 5.7321s/12 iters), loss = 0.119994
I0428 14:52:51.497467 30475 solver.cpp:237]     Train net output #0: loss = 0.119994 (* 1 = 0.119994 loss)
I0428 14:52:51.497475 30475 sgd_solver.cpp:105] Iteration 6576, lr = 0.00271821
I0428 14:52:57.161936 30475 solver.cpp:218] Iteration 6588 (2.11847 iter/s, 5.66446s/12 iters), loss = 0.0907687
I0428 14:52:57.161980 30475 solver.cpp:237]     Train net output #0: loss = 0.0907688 (* 1 = 0.0907688 loss)
I0428 14:52:57.161996 30475 sgd_solver.cpp:105] Iteration 6588, lr = 0.00271175
I0428 14:53:02.920476 30475 solver.cpp:218] Iteration 6600 (2.08388 iter/s, 5.75848s/12 iters), loss = 0.0418944
I0428 14:53:02.920536 30475 solver.cpp:237]     Train net output #0: loss = 0.0418944 (* 1 = 0.0418944 loss)
I0428 14:53:02.920547 30475 sgd_solver.cpp:105] Iteration 6600, lr = 0.00270532
I0428 14:53:08.632521 30475 solver.cpp:218] Iteration 6612 (2.10085 iter/s, 5.71198s/12 iters), loss = 0.0874379
I0428 14:53:08.632628 30475 solver.cpp:237]     Train net output #0: loss = 0.0874379 (* 1 = 0.0874379 loss)
I0428 14:53:08.632638 30475 sgd_solver.cpp:105] Iteration 6612, lr = 0.00269889
I0428 14:53:14.324105 30475 solver.cpp:218] Iteration 6624 (2.10842 iter/s, 5.69146s/12 iters), loss = 0.0587536
I0428 14:53:14.324154 30475 solver.cpp:237]     Train net output #0: loss = 0.0587536 (* 1 = 0.0587536 loss)
I0428 14:53:14.324163 30475 sgd_solver.cpp:105] Iteration 6624, lr = 0.00269248
I0428 14:53:16.619256 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6630.caffemodel
I0428 14:53:19.578485 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6630.solverstate
I0428 14:53:22.127678 30475 solver.cpp:330] Iteration 6630, Testing net (#0)
I0428 14:53:22.127696 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:53:24.340811 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:53:27.492744 30475 solver.cpp:397]     Test net output #0: accuracy = 0.447304
I0428 14:53:27.492774 30475 solver.cpp:397]     Test net output #1: loss = 3.83136 (* 1 = 3.83136 loss)
I0428 14:53:29.639400 30475 solver.cpp:218] Iteration 6636 (0.783533 iter/s, 15.3153s/12 iters), loss = 0.0542856
I0428 14:53:29.639448 30475 solver.cpp:237]     Train net output #0: loss = 0.0542856 (* 1 = 0.0542856 loss)
I0428 14:53:29.639457 30475 sgd_solver.cpp:105] Iteration 6636, lr = 0.00268609
I0428 14:53:35.395210 30475 solver.cpp:218] Iteration 6648 (2.08487 iter/s, 5.75575s/12 iters), loss = 0.1463
I0428 14:53:35.395253 30475 solver.cpp:237]     Train net output #0: loss = 0.1463 (* 1 = 0.1463 loss)
I0428 14:53:35.395262 30475 sgd_solver.cpp:105] Iteration 6648, lr = 0.00267971
I0428 14:53:41.031299 30475 solver.cpp:218] Iteration 6660 (2.12916 iter/s, 5.63603s/12 iters), loss = 0.117866
I0428 14:53:41.031476 30475 solver.cpp:237]     Train net output #0: loss = 0.117866 (* 1 = 0.117866 loss)
I0428 14:53:41.031488 30475 sgd_solver.cpp:105] Iteration 6660, lr = 0.00267335
I0428 14:53:46.700428 30475 solver.cpp:218] Iteration 6672 (2.1168 iter/s, 5.66894s/12 iters), loss = 0.0707195
I0428 14:53:46.700498 30475 solver.cpp:237]     Train net output #0: loss = 0.0707195 (* 1 = 0.0707195 loss)
I0428 14:53:46.700511 30475 sgd_solver.cpp:105] Iteration 6672, lr = 0.00266701
I0428 14:53:48.278692 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:53:52.550952 30475 solver.cpp:218] Iteration 6684 (2.05112 iter/s, 5.85045s/12 iters), loss = 0.122856
I0428 14:53:52.550998 30475 solver.cpp:237]     Train net output #0: loss = 0.122856 (* 1 = 0.122856 loss)
I0428 14:53:52.551005 30475 sgd_solver.cpp:105] Iteration 6684, lr = 0.00266067
I0428 14:53:58.251468 30475 solver.cpp:218] Iteration 6696 (2.1051 iter/s, 5.70046s/12 iters), loss = 0.221696
I0428 14:53:58.251533 30475 solver.cpp:237]     Train net output #0: loss = 0.221696 (* 1 = 0.221696 loss)
I0428 14:53:58.251546 30475 sgd_solver.cpp:105] Iteration 6696, lr = 0.00265436
I0428 14:54:04.019718 30475 solver.cpp:218] Iteration 6708 (2.08038 iter/s, 5.76818s/12 iters), loss = 0.0694519
I0428 14:54:04.019769 30475 solver.cpp:237]     Train net output #0: loss = 0.0694519 (* 1 = 0.0694519 loss)
I0428 14:54:04.019781 30475 sgd_solver.cpp:105] Iteration 6708, lr = 0.00264805
I0428 14:54:09.890506 30475 solver.cpp:218] Iteration 6720 (2.04404 iter/s, 5.87073s/12 iters), loss = 0.144012
I0428 14:54:09.890553 30475 solver.cpp:237]     Train net output #0: loss = 0.144012 (* 1 = 0.144012 loss)
I0428 14:54:09.890561 30475 sgd_solver.cpp:105] Iteration 6720, lr = 0.00264177
I0428 14:54:14.986718 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6732.caffemodel
I0428 14:54:19.339699 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6732.solverstate
I0428 14:54:21.422513 30475 solver.cpp:330] Iteration 6732, Testing net (#0)
I0428 14:54:21.422535 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:54:23.407395 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:54:26.544112 30475 solver.cpp:397]     Test net output #0: accuracy = 0.446078
I0428 14:54:26.544150 30475 solver.cpp:397]     Test net output #1: loss = 3.88143 (* 1 = 3.88143 loss)
I0428 14:54:26.696426 30475 solver.cpp:218] Iteration 6732 (0.714036 iter/s, 16.8059s/12 iters), loss = 0.155282
I0428 14:54:26.696496 30475 solver.cpp:237]     Train net output #0: loss = 0.155282 (* 1 = 0.155282 loss)
I0428 14:54:26.696507 30475 sgd_solver.cpp:105] Iteration 6732, lr = 0.0026355
I0428 14:54:31.529979 30475 solver.cpp:218] Iteration 6744 (2.48269 iter/s, 4.83347s/12 iters), loss = 0.0897973
I0428 14:54:31.530028 30475 solver.cpp:237]     Train net output #0: loss = 0.0897974 (* 1 = 0.0897974 loss)
I0428 14:54:31.530037 30475 sgd_solver.cpp:105] Iteration 6744, lr = 0.00262924
I0428 14:54:37.238926 30475 solver.cpp:218] Iteration 6756 (2.10198 iter/s, 5.70889s/12 iters), loss = 0.0551401
I0428 14:54:37.238970 30475 solver.cpp:237]     Train net output #0: loss = 0.0551402 (* 1 = 0.0551402 loss)
I0428 14:54:37.238978 30475 sgd_solver.cpp:105] Iteration 6756, lr = 0.002623
I0428 14:54:43.309465 30475 solver.cpp:218] Iteration 6768 (1.97678 iter/s, 6.07049s/12 iters), loss = 0.137288
I0428 14:54:43.309509 30475 solver.cpp:237]     Train net output #0: loss = 0.137288 (* 1 = 0.137288 loss)
I0428 14:54:43.309518 30475 sgd_solver.cpp:105] Iteration 6768, lr = 0.00261677
I0428 14:54:47.290398 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:54:49.019024 30475 solver.cpp:218] Iteration 6780 (2.10176 iter/s, 5.7095s/12 iters), loss = 0.18832
I0428 14:54:49.019073 30475 solver.cpp:237]     Train net output #0: loss = 0.18832 (* 1 = 0.18832 loss)
I0428 14:54:49.019081 30475 sgd_solver.cpp:105] Iteration 6780, lr = 0.00261056
I0428 14:54:54.841953 30475 solver.cpp:218] Iteration 6792 (2.06084 iter/s, 5.82287s/12 iters), loss = 0.115354
I0428 14:54:54.842002 30475 solver.cpp:237]     Train net output #0: loss = 0.115354 (* 1 = 0.115354 loss)
I0428 14:54:54.842010 30475 sgd_solver.cpp:105] Iteration 6792, lr = 0.00260436
I0428 14:55:00.574231 30475 solver.cpp:218] Iteration 6804 (2.09343 iter/s, 5.73222s/12 iters), loss = 0.0869737
I0428 14:55:00.574275 30475 solver.cpp:237]     Train net output #0: loss = 0.0869737 (* 1 = 0.0869737 loss)
I0428 14:55:00.574281 30475 sgd_solver.cpp:105] Iteration 6804, lr = 0.00259817
I0428 14:55:06.422384 30475 solver.cpp:218] Iteration 6816 (2.05195 iter/s, 5.8481s/12 iters), loss = 0.0548976
I0428 14:55:06.422448 30475 solver.cpp:237]     Train net output #0: loss = 0.0548976 (* 1 = 0.0548976 loss)
I0428 14:55:06.422461 30475 sgd_solver.cpp:105] Iteration 6816, lr = 0.00259201
I0428 14:55:12.039849 30475 solver.cpp:218] Iteration 6828 (2.13622 iter/s, 5.6174s/12 iters), loss = 0.100641
I0428 14:55:12.039892 30475 solver.cpp:237]     Train net output #0: loss = 0.100641 (* 1 = 0.100641 loss)
I0428 14:55:12.039901 30475 sgd_solver.cpp:105] Iteration 6828, lr = 0.00258585
I0428 14:55:14.316099 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6834.caffemodel
I0428 14:55:17.519917 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6834.solverstate
I0428 14:55:20.945310 30475 solver.cpp:330] Iteration 6834, Testing net (#0)
I0428 14:55:20.945331 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:55:22.940701 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:55:26.046046 30475 solver.cpp:397]     Test net output #0: accuracy = 0.446078
I0428 14:55:26.046082 30475 solver.cpp:397]     Test net output #1: loss = 3.89985 (* 1 = 3.89985 loss)
I0428 14:55:28.186060 30475 solver.cpp:218] Iteration 6840 (0.74321 iter/s, 16.1462s/12 iters), loss = 0.032619
I0428 14:55:28.186103 30475 solver.cpp:237]     Train net output #0: loss = 0.032619 (* 1 = 0.032619 loss)
I0428 14:55:28.186110 30475 sgd_solver.cpp:105] Iteration 6840, lr = 0.00257971
I0428 14:55:33.704545 30475 solver.cpp:218] Iteration 6852 (2.17453 iter/s, 5.51843s/12 iters), loss = 0.0186361
I0428 14:55:33.704586 30475 solver.cpp:237]     Train net output #0: loss = 0.0186361 (* 1 = 0.0186361 loss)
I0428 14:55:33.704593 30475 sgd_solver.cpp:105] Iteration 6852, lr = 0.00257359
I0428 14:55:39.414075 30475 solver.cpp:218] Iteration 6864 (2.10177 iter/s, 5.70948s/12 iters), loss = 0.0986078
I0428 14:55:39.414139 30475 solver.cpp:237]     Train net output #0: loss = 0.0986078 (* 1 = 0.0986078 loss)
I0428 14:55:39.414149 30475 sgd_solver.cpp:105] Iteration 6864, lr = 0.00256748
I0428 14:55:45.120503 30475 solver.cpp:218] Iteration 6876 (2.10292 iter/s, 5.70636s/12 iters), loss = 0.123657
I0428 14:55:45.120553 30475 solver.cpp:237]     Train net output #0: loss = 0.123657 (* 1 = 0.123657 loss)
I0428 14:55:45.120561 30475 sgd_solver.cpp:105] Iteration 6876, lr = 0.00256138
I0428 14:55:45.794467 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:55:50.805474 30475 solver.cpp:218] Iteration 6888 (2.11085 iter/s, 5.68491s/12 iters), loss = 0.132012
I0428 14:55:50.805616 30475 solver.cpp:237]     Train net output #0: loss = 0.132012 (* 1 = 0.132012 loss)
I0428 14:55:50.805626 30475 sgd_solver.cpp:105] Iteration 6888, lr = 0.0025553
I0428 14:55:56.523613 30475 solver.cpp:218] Iteration 6900 (2.09864 iter/s, 5.71799s/12 iters), loss = 0.08801
I0428 14:55:56.523663 30475 solver.cpp:237]     Train net output #0: loss = 0.0880101 (* 1 = 0.0880101 loss)
I0428 14:55:56.523670 30475 sgd_solver.cpp:105] Iteration 6900, lr = 0.00254923
I0428 14:56:02.232761 30475 solver.cpp:218] Iteration 6912 (2.10191 iter/s, 5.70909s/12 iters), loss = 0.0603386
I0428 14:56:02.232802 30475 solver.cpp:237]     Train net output #0: loss = 0.0603387 (* 1 = 0.0603387 loss)
I0428 14:56:02.232808 30475 sgd_solver.cpp:105] Iteration 6912, lr = 0.00254318
I0428 14:56:07.815977 30475 solver.cpp:218] Iteration 6924 (2.14932 iter/s, 5.58317s/12 iters), loss = 0.137808
I0428 14:56:07.816025 30475 solver.cpp:237]     Train net output #0: loss = 0.137808 (* 1 = 0.137808 loss)
I0428 14:56:07.816032 30475 sgd_solver.cpp:105] Iteration 6924, lr = 0.00253714
I0428 14:56:12.999933 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6936.caffemodel
I0428 14:56:16.431352 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6936.solverstate
I0428 14:56:22.429005 30475 solver.cpp:330] Iteration 6936, Testing net (#0)
I0428 14:56:22.429111 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:56:23.113818 30475 blocking_queue.cpp:49] Waiting for data
I0428 14:56:24.361943 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:56:27.576056 30475 solver.cpp:397]     Test net output #0: accuracy = 0.444853
I0428 14:56:27.576082 30475 solver.cpp:397]     Test net output #1: loss = 3.78362 (* 1 = 3.78362 loss)
I0428 14:56:27.737643 30475 solver.cpp:218] Iteration 6936 (0.60236 iter/s, 19.9216s/12 iters), loss = 0.0834175
I0428 14:56:27.737699 30475 solver.cpp:237]     Train net output #0: loss = 0.0834176 (* 1 = 0.0834176 loss)
I0428 14:56:27.737707 30475 sgd_solver.cpp:105] Iteration 6936, lr = 0.00253112
I0428 14:56:32.386618 30475 solver.cpp:218] Iteration 6948 (2.58125 iter/s, 4.64891s/12 iters), loss = 0.0879464
I0428 14:56:32.386664 30475 solver.cpp:237]     Train net output #0: loss = 0.0879464 (* 1 = 0.0879464 loss)
I0428 14:56:32.386672 30475 sgd_solver.cpp:105] Iteration 6948, lr = 0.00252511
I0428 14:56:38.092507 30475 solver.cpp:218] Iteration 6960 (2.10311 iter/s, 5.70583s/12 iters), loss = 0.0209004
I0428 14:56:38.092555 30475 solver.cpp:237]     Train net output #0: loss = 0.0209004 (* 1 = 0.0209004 loss)
I0428 14:56:38.092562 30475 sgd_solver.cpp:105] Iteration 6960, lr = 0.00251911
I0428 14:56:43.822037 30475 solver.cpp:218] Iteration 6972 (2.09443 iter/s, 5.72947s/12 iters), loss = 0.0463635
I0428 14:56:43.822089 30475 solver.cpp:237]     Train net output #0: loss = 0.0463635 (* 1 = 0.0463635 loss)
I0428 14:56:43.822098 30475 sgd_solver.cpp:105] Iteration 6972, lr = 0.00251313
I0428 14:56:46.809062 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:56:49.393419 30475 solver.cpp:218] Iteration 6984 (2.15389 iter/s, 5.57132s/12 iters), loss = 0.0497285
I0428 14:56:49.393469 30475 solver.cpp:237]     Train net output #0: loss = 0.0497285 (* 1 = 0.0497285 loss)
I0428 14:56:49.393477 30475 sgd_solver.cpp:105] Iteration 6984, lr = 0.00250717
I0428 14:56:55.074276 30475 solver.cpp:218] Iteration 6996 (2.11238 iter/s, 5.6808s/12 iters), loss = 0.0380989
I0428 14:56:55.074359 30475 solver.cpp:237]     Train net output #0: loss = 0.0380989 (* 1 = 0.0380989 loss)
I0428 14:56:55.074368 30475 sgd_solver.cpp:105] Iteration 6996, lr = 0.00250121
I0428 14:57:00.761395 30475 solver.cpp:218] Iteration 7008 (2.11007 iter/s, 5.68703s/12 iters), loss = 0.0564651
I0428 14:57:00.761438 30475 solver.cpp:237]     Train net output #0: loss = 0.0564651 (* 1 = 0.0564651 loss)
I0428 14:57:00.761446 30475 sgd_solver.cpp:105] Iteration 7008, lr = 0.00249528
I0428 14:57:06.423825 30475 solver.cpp:218] Iteration 7020 (2.11925 iter/s, 5.66237s/12 iters), loss = 0.0541627
I0428 14:57:06.423894 30475 solver.cpp:237]     Train net output #0: loss = 0.0541628 (* 1 = 0.0541628 loss)
I0428 14:57:06.423905 30475 sgd_solver.cpp:105] Iteration 7020, lr = 0.00248935
I0428 14:57:12.116261 30475 solver.cpp:218] Iteration 7032 (2.10809 iter/s, 5.69237s/12 iters), loss = 0.0405545
I0428 14:57:12.116309 30475 solver.cpp:237]     Train net output #0: loss = 0.0405546 (* 1 = 0.0405546 loss)
I0428 14:57:12.116318 30475 sgd_solver.cpp:105] Iteration 7032, lr = 0.00248344
I0428 14:57:14.397974 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7038.caffemodel
I0428 14:57:17.218495 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7038.solverstate
I0428 14:57:20.323659 30475 solver.cpp:330] Iteration 7038, Testing net (#0)
I0428 14:57:20.323681 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:57:22.137998 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:57:25.314452 30475 solver.cpp:397]     Test net output #0: accuracy = 0.454044
I0428 14:57:25.314618 30475 solver.cpp:397]     Test net output #1: loss = 3.98183 (* 1 = 3.98183 loss)
I0428 14:57:27.386391 30475 solver.cpp:218] Iteration 7044 (0.78585 iter/s, 15.2701s/12 iters), loss = 0.220667
I0428 14:57:27.386445 30475 solver.cpp:237]     Train net output #0: loss = 0.220667 (* 1 = 0.220667 loss)
I0428 14:57:27.386457 30475 sgd_solver.cpp:105] Iteration 7044, lr = 0.00247755
I0428 14:57:32.957087 30475 solver.cpp:218] Iteration 7056 (2.15415 iter/s, 5.57064s/12 iters), loss = 0.0552927
I0428 14:57:32.957130 30475 solver.cpp:237]     Train net output #0: loss = 0.0552927 (* 1 = 0.0552927 loss)
I0428 14:57:32.957139 30475 sgd_solver.cpp:105] Iteration 7056, lr = 0.00247166
I0428 14:57:38.825176 30475 solver.cpp:218] Iteration 7068 (2.04498 iter/s, 5.86804s/12 iters), loss = 0.135491
I0428 14:57:38.825224 30475 solver.cpp:237]     Train net output #0: loss = 0.135491 (* 1 = 0.135491 loss)
I0428 14:57:38.825232 30475 sgd_solver.cpp:105] Iteration 7068, lr = 0.0024658
I0428 14:57:44.319550 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:57:44.491927 30475 solver.cpp:218] Iteration 7080 (2.11764 iter/s, 5.6667s/12 iters), loss = 0.162879
I0428 14:57:44.491972 30475 solver.cpp:237]     Train net output #0: loss = 0.162879 (* 1 = 0.162879 loss)
I0428 14:57:44.491981 30475 sgd_solver.cpp:105] Iteration 7080, lr = 0.00245994
I0428 14:57:50.278343 30475 solver.cpp:218] Iteration 7092 (2.07384 iter/s, 5.78636s/12 iters), loss = 0.0576326
I0428 14:57:50.278389 30475 solver.cpp:237]     Train net output #0: loss = 0.0576326 (* 1 = 0.0576326 loss)
I0428 14:57:50.278398 30475 sgd_solver.cpp:105] Iteration 7092, lr = 0.0024541
I0428 14:57:55.898942 30475 solver.cpp:218] Iteration 7104 (2.13503 iter/s, 5.62054s/12 iters), loss = 0.045072
I0428 14:57:55.899078 30475 solver.cpp:237]     Train net output #0: loss = 0.045072 (* 1 = 0.045072 loss)
I0428 14:57:55.899086 30475 sgd_solver.cpp:105] Iteration 7104, lr = 0.00244827
I0428 14:58:01.543574 30475 solver.cpp:218] Iteration 7116 (2.12597 iter/s, 5.64448s/12 iters), loss = 0.134954
I0428 14:58:01.543649 30475 solver.cpp:237]     Train net output #0: loss = 0.134954 (* 1 = 0.134954 loss)
I0428 14:58:01.543663 30475 sgd_solver.cpp:105] Iteration 7116, lr = 0.00244246
I0428 14:58:07.236512 30475 solver.cpp:218] Iteration 7128 (2.1079 iter/s, 5.69286s/12 iters), loss = 0.0968751
I0428 14:58:07.236558 30475 solver.cpp:237]     Train net output #0: loss = 0.0968751 (* 1 = 0.0968751 loss)
I0428 14:58:07.236567 30475 sgd_solver.cpp:105] Iteration 7128, lr = 0.00243666
I0428 14:58:12.336130 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7140.caffemodel
I0428 14:58:15.854060 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7140.solverstate
I0428 14:58:19.138108 30475 solver.cpp:330] Iteration 7140, Testing net (#0)
I0428 14:58:19.138126 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:58:21.035202 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:58:24.547492 30475 solver.cpp:397]     Test net output #0: accuracy = 0.446691
I0428 14:58:24.547534 30475 solver.cpp:397]     Test net output #1: loss = 3.80477 (* 1 = 3.80477 loss)
I0428 14:58:24.708925 30475 solver.cpp:218] Iteration 7140 (0.686798 iter/s, 17.4724s/12 iters), loss = 0.127346
I0428 14:58:24.708983 30475 solver.cpp:237]     Train net output #0: loss = 0.127346 (* 1 = 0.127346 loss)
I0428 14:58:24.708997 30475 sgd_solver.cpp:105] Iteration 7140, lr = 0.00243088
I0428 14:58:29.353886 30475 solver.cpp:218] Iteration 7152 (2.58348 iter/s, 4.6449s/12 iters), loss = 0.032676
I0428 14:58:29.354027 30475 solver.cpp:237]     Train net output #0: loss = 0.032676 (* 1 = 0.032676 loss)
I0428 14:58:29.354036 30475 sgd_solver.cpp:105] Iteration 7152, lr = 0.00242511
I0428 14:58:35.092552 30475 solver.cpp:218] Iteration 7164 (2.09113 iter/s, 5.73852s/12 iters), loss = 0.0598092
I0428 14:58:35.092593 30475 solver.cpp:237]     Train net output #0: loss = 0.0598092 (* 1 = 0.0598092 loss)
I0428 14:58:35.092602 30475 sgd_solver.cpp:105] Iteration 7164, lr = 0.00241935
I0428 14:58:40.833057 30475 solver.cpp:218] Iteration 7176 (2.09043 iter/s, 5.74046s/12 iters), loss = 0.0283509
I0428 14:58:40.833103 30475 solver.cpp:237]     Train net output #0: loss = 0.0283509 (* 1 = 0.0283509 loss)
I0428 14:58:40.833110 30475 sgd_solver.cpp:105] Iteration 7176, lr = 0.0024136
I0428 14:58:43.224426 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:58:46.555864 30475 solver.cpp:218] Iteration 7188 (2.09689 iter/s, 5.72275s/12 iters), loss = 0.0687454
I0428 14:58:46.555907 30475 solver.cpp:237]     Train net output #0: loss = 0.0687454 (* 1 = 0.0687454 loss)
I0428 14:58:46.555915 30475 sgd_solver.cpp:105] Iteration 7188, lr = 0.00240787
I0428 14:58:52.294822 30475 solver.cpp:218] Iteration 7200 (2.09099 iter/s, 5.7389s/12 iters), loss = 0.0355843
I0428 14:58:52.294893 30475 solver.cpp:237]     Train net output #0: loss = 0.0355843 (* 1 = 0.0355843 loss)
I0428 14:58:52.294905 30475 sgd_solver.cpp:105] Iteration 7200, lr = 0.00240216
I0428 14:58:57.967871 30475 solver.cpp:218] Iteration 7212 (2.11529 iter/s, 5.67298s/12 iters), loss = 0.044256
I0428 14:58:57.967912 30475 solver.cpp:237]     Train net output #0: loss = 0.044256 (* 1 = 0.044256 loss)
I0428 14:58:57.967921 30475 sgd_solver.cpp:105] Iteration 7212, lr = 0.00239645
I0428 14:59:03.744372 30475 solver.cpp:218] Iteration 7224 (2.0774 iter/s, 5.77645s/12 iters), loss = 0.031315
I0428 14:59:03.744489 30475 solver.cpp:237]     Train net output #0: loss = 0.031315 (* 1 = 0.031315 loss)
I0428 14:59:03.744499 30475 sgd_solver.cpp:105] Iteration 7224, lr = 0.00239076
I0428 14:59:09.390148 30475 solver.cpp:218] Iteration 7236 (2.12553 iter/s, 5.64565s/12 iters), loss = 0.144195
I0428 14:59:09.390200 30475 solver.cpp:237]     Train net output #0: loss = 0.144195 (* 1 = 0.144195 loss)
I0428 14:59:09.390209 30475 sgd_solver.cpp:105] Iteration 7236, lr = 0.00238509
I0428 14:59:11.744590 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7242.caffemodel
I0428 14:59:18.402904 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7242.solverstate
I0428 14:59:21.973021 30475 solver.cpp:330] Iteration 7242, Testing net (#0)
I0428 14:59:21.973040 30475 net.cpp:676] Ignoring source layer train-data
I0428 14:59:23.628724 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:59:26.995769 30475 solver.cpp:397]     Test net output #0: accuracy = 0.446691
I0428 14:59:26.995800 30475 solver.cpp:397]     Test net output #1: loss = 3.95859 (* 1 = 3.95859 loss)
I0428 14:59:29.161177 30475 solver.cpp:218] Iteration 7248 (0.606949 iter/s, 19.771s/12 iters), loss = 0.0247922
I0428 14:59:29.161221 30475 solver.cpp:237]     Train net output #0: loss = 0.0247922 (* 1 = 0.0247922 loss)
I0428 14:59:29.161229 30475 sgd_solver.cpp:105] Iteration 7248, lr = 0.00237942
I0428 14:59:34.879859 30475 solver.cpp:218] Iteration 7260 (2.09841 iter/s, 5.71863s/12 iters), loss = 0.071195
I0428 14:59:34.880028 30475 solver.cpp:237]     Train net output #0: loss = 0.071195 (* 1 = 0.071195 loss)
I0428 14:59:34.880038 30475 sgd_solver.cpp:105] Iteration 7260, lr = 0.00237378
I0428 14:59:40.516794 30475 solver.cpp:218] Iteration 7272 (2.12888 iter/s, 5.63676s/12 iters), loss = 0.0436387
I0428 14:59:40.516842 30475 solver.cpp:237]     Train net output #0: loss = 0.0436387 (* 1 = 0.0436387 loss)
I0428 14:59:40.516851 30475 sgd_solver.cpp:105] Iteration 7272, lr = 0.00236814
I0428 14:59:45.219236 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 14:59:46.103978 30475 solver.cpp:218] Iteration 7284 (2.14779 iter/s, 5.58713s/12 iters), loss = 0.0958591
I0428 14:59:46.104024 30475 solver.cpp:237]     Train net output #0: loss = 0.0958591 (* 1 = 0.0958591 loss)
I0428 14:59:46.104032 30475 sgd_solver.cpp:105] Iteration 7284, lr = 0.00236252
I0428 14:59:51.650045 30475 solver.cpp:218] Iteration 7296 (2.16372 iter/s, 5.54601s/12 iters), loss = 0.0613992
I0428 14:59:51.650094 30475 solver.cpp:237]     Train net output #0: loss = 0.0613992 (* 1 = 0.0613992 loss)
I0428 14:59:51.650101 30475 sgd_solver.cpp:105] Iteration 7296, lr = 0.00235691
I0428 14:59:57.183180 30475 solver.cpp:218] Iteration 7308 (2.16877 iter/s, 5.53308s/12 iters), loss = 0.102167
I0428 14:59:57.183225 30475 solver.cpp:237]     Train net output #0: loss = 0.102167 (* 1 = 0.102167 loss)
I0428 14:59:57.183233 30475 sgd_solver.cpp:105] Iteration 7308, lr = 0.00235131
I0428 15:00:02.864886 30475 solver.cpp:218] Iteration 7320 (2.11206 iter/s, 5.68165s/12 iters), loss = 0.038474
I0428 15:00:02.864934 30475 solver.cpp:237]     Train net output #0: loss = 0.038474 (* 1 = 0.038474 loss)
I0428 15:00:02.864943 30475 sgd_solver.cpp:105] Iteration 7320, lr = 0.00234573
I0428 15:00:08.531102 30475 solver.cpp:218] Iteration 7332 (2.11784 iter/s, 5.66615s/12 iters), loss = 0.0659452
I0428 15:00:08.531299 30475 solver.cpp:237]     Train net output #0: loss = 0.0659452 (* 1 = 0.0659452 loss)
I0428 15:00:08.531316 30475 sgd_solver.cpp:105] Iteration 7332, lr = 0.00234016
I0428 15:00:13.668293 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7344.caffemodel
I0428 15:00:18.344206 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7344.solverstate
I0428 15:00:29.863211 30475 solver.cpp:330] Iteration 7344, Testing net (#0)
I0428 15:00:29.863229 30475 net.cpp:676] Ignoring source layer train-data
I0428 15:00:31.640594 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:00:35.066282 30475 solver.cpp:397]     Test net output #0: accuracy = 0.45098
I0428 15:00:35.066318 30475 solver.cpp:397]     Test net output #1: loss = 4.15751 (* 1 = 4.15751 loss)
I0428 15:00:35.222654 30475 solver.cpp:218] Iteration 7344 (0.449583 iter/s, 26.6914s/12 iters), loss = 0.0753084
I0428 15:00:35.222723 30475 solver.cpp:237]     Train net output #0: loss = 0.0753084 (* 1 = 0.0753084 loss)
I0428 15:00:35.222733 30475 sgd_solver.cpp:105] Iteration 7344, lr = 0.0023346
I0428 15:00:40.062959 30475 solver.cpp:218] Iteration 7356 (2.47922 iter/s, 4.84023s/12 iters), loss = 0.0215389
I0428 15:00:40.063067 30475 solver.cpp:237]     Train net output #0: loss = 0.0215389 (* 1 = 0.0215389 loss)
I0428 15:00:40.063079 30475 sgd_solver.cpp:105] Iteration 7356, lr = 0.00232906
I0428 15:00:45.725710 30475 solver.cpp:218] Iteration 7368 (2.11916 iter/s, 5.66263s/12 iters), loss = 0.0520238
I0428 15:00:45.725755 30475 solver.cpp:237]     Train net output #0: loss = 0.0520238 (* 1 = 0.0520238 loss)
I0428 15:00:45.725764 30475 sgd_solver.cpp:105] Iteration 7368, lr = 0.00232353
I0428 15:00:51.418962 30475 solver.cpp:218] Iteration 7380 (2.10778 iter/s, 5.6932s/12 iters), loss = 0.074725
I0428 15:00:51.419006 30475 solver.cpp:237]     Train net output #0: loss = 0.074725 (* 1 = 0.074725 loss)
I0428 15:00:51.419013 30475 sgd_solver.cpp:105] Iteration 7380, lr = 0.00231802
I0428 15:00:52.964205 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:00:57.075995 30475 solver.cpp:218] Iteration 7392 (2.12127 iter/s, 5.65698s/12 iters), loss = 0.0490398
I0428 15:00:57.076037 30475 solver.cpp:237]     Train net output #0: loss = 0.0490398 (* 1 = 0.0490398 loss)
I0428 15:00:57.076045 30475 sgd_solver.cpp:105] Iteration 7392, lr = 0.00231251
I0428 15:01:02.750907 30475 solver.cpp:218] Iteration 7404 (2.11459 iter/s, 5.67486s/12 iters), loss = 0.0710756
I0428 15:01:02.750957 30475 solver.cpp:237]     Train net output #0: loss = 0.0710756 (* 1 = 0.0710756 loss)
I0428 15:01:02.750964 30475 sgd_solver.cpp:105] Iteration 7404, lr = 0.00230702
I0428 15:01:08.447837 30475 solver.cpp:218] Iteration 7416 (2.10642 iter/s, 5.69687s/12 iters), loss = 0.0827278
I0428 15:01:08.447887 30475 solver.cpp:237]     Train net output #0: loss = 0.0827278 (* 1 = 0.0827278 loss)
I0428 15:01:08.447896 30475 sgd_solver.cpp:105] Iteration 7416, lr = 0.00230154
I0428 15:01:14.137576 30475 solver.cpp:218] Iteration 7428 (2.10908 iter/s, 5.68968s/12 iters), loss = 0.032996
I0428 15:01:14.137713 30475 solver.cpp:237]     Train net output #0: loss = 0.032996 (* 1 = 0.032996 loss)
I0428 15:01:14.137722 30475 sgd_solver.cpp:105] Iteration 7428, lr = 0.00229608
I0428 15:01:19.841527 30475 solver.cpp:218] Iteration 7440 (2.10386 iter/s, 5.70381s/12 iters), loss = 0.0377374
I0428 15:01:19.841570 30475 solver.cpp:237]     Train net output #0: loss = 0.0377374 (* 1 = 0.0377374 loss)
I0428 15:01:19.841578 30475 sgd_solver.cpp:105] Iteration 7440, lr = 0.00229063
I0428 15:01:22.125864 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7446.caffemodel
I0428 15:01:29.153128 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7446.solverstate
I0428 15:01:34.269523 30475 solver.cpp:330] Iteration 7446, Testing net (#0)
I0428 15:01:34.269554 30475 net.cpp:676] Ignoring source layer train-data
I0428 15:01:36.036409 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:01:39.556692 30475 solver.cpp:397]     Test net output #0: accuracy = 0.446078
I0428 15:01:39.556721 30475 solver.cpp:397]     Test net output #1: loss = 3.96717 (* 1 = 3.96717 loss)
I0428 15:01:41.719467 30475 solver.cpp:218] Iteration 7452 (0.548498 iter/s, 21.8779s/12 iters), loss = 0.0270963
I0428 15:01:41.719509 30475 solver.cpp:237]     Train net output #0: loss = 0.0270963 (* 1 = 0.0270963 loss)
I0428 15:01:41.719516 30475 sgd_solver.cpp:105] Iteration 7452, lr = 0.00228519
I0428 15:01:47.276487 30475 solver.cpp:218] Iteration 7464 (2.15945 iter/s, 5.55697s/12 iters), loss = 0.0883231
I0428 15:01:47.276607 30475 solver.cpp:237]     Train net output #0: loss = 0.0883231 (* 1 = 0.0883231 loss)
I0428 15:01:47.276618 30475 sgd_solver.cpp:105] Iteration 7464, lr = 0.00227976
I0428 15:01:52.930776 30475 solver.cpp:218] Iteration 7476 (2.12233 iter/s, 5.65416s/12 iters), loss = 0.11119
I0428 15:01:52.930819 30475 solver.cpp:237]     Train net output #0: loss = 0.11119 (* 1 = 0.11119 loss)
I0428 15:01:52.930827 30475 sgd_solver.cpp:105] Iteration 7476, lr = 0.00227435
I0428 15:01:56.880328 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:01:58.578490 30475 solver.cpp:218] Iteration 7488 (2.12477 iter/s, 5.64766s/12 iters), loss = 0.0231382
I0428 15:01:58.578536 30475 solver.cpp:237]     Train net output #0: loss = 0.0231382 (* 1 = 0.0231382 loss)
I0428 15:01:58.578545 30475 sgd_solver.cpp:105] Iteration 7488, lr = 0.00226895
I0428 15:02:04.079170 30475 solver.cpp:218] Iteration 7500 (2.18157 iter/s, 5.50063s/12 iters), loss = 0.0151761
I0428 15:02:04.079208 30475 solver.cpp:237]     Train net output #0: loss = 0.0151761 (* 1 = 0.0151761 loss)
I0428 15:02:04.079216 30475 sgd_solver.cpp:105] Iteration 7500, lr = 0.00226357
I0428 15:02:09.583636 30475 solver.cpp:218] Iteration 7512 (2.18007 iter/s, 5.50442s/12 iters), loss = 0.0533026
I0428 15:02:09.583683 30475 solver.cpp:237]     Train net output #0: loss = 0.0533026 (* 1 = 0.0533026 loss)
I0428 15:02:09.583693 30475 sgd_solver.cpp:105] Iteration 7512, lr = 0.00225819
I0428 15:02:15.222349 30475 solver.cpp:218] Iteration 7524 (2.12816 iter/s, 5.63866s/12 iters), loss = 0.0208883
I0428 15:02:15.222384 30475 solver.cpp:237]     Train net output #0: loss = 0.0208883 (* 1 = 0.0208883 loss)
I0428 15:02:15.222393 30475 sgd_solver.cpp:105] Iteration 7524, lr = 0.00225283
I0428 15:02:20.980365 30475 solver.cpp:218] Iteration 7536 (2.08407 iter/s, 5.75798s/12 iters), loss = 0.0892992
I0428 15:02:20.980507 30475 solver.cpp:237]     Train net output #0: loss = 0.0892993 (* 1 = 0.0892993 loss)
I0428 15:02:20.980516 30475 sgd_solver.cpp:105] Iteration 7536, lr = 0.00224748
I0428 15:02:26.212939 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7548.caffemodel
I0428 15:02:30.835508 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7548.solverstate
I0428 15:02:35.226547 30475 solver.cpp:330] Iteration 7548, Testing net (#0)
I0428 15:02:35.226569 30475 net.cpp:676] Ignoring source layer train-data
I0428 15:02:36.862846 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:02:40.210259 30475 solver.cpp:397]     Test net output #0: accuracy = 0.448529
I0428 15:02:40.210304 30475 solver.cpp:397]     Test net output #1: loss = 3.95218 (* 1 = 3.95218 loss)
I0428 15:02:40.371477 30475 solver.cpp:218] Iteration 7548 (0.618844 iter/s, 19.391s/12 iters), loss = 0.0370153
I0428 15:02:40.371536 30475 solver.cpp:237]     Train net output #0: loss = 0.0370154 (* 1 = 0.0370154 loss)
I0428 15:02:40.371546 30475 sgd_solver.cpp:105] Iteration 7548, lr = 0.00224215
I0428 15:02:45.135861 30475 solver.cpp:218] Iteration 7560 (2.51872 iter/s, 4.76432s/12 iters), loss = 0.0534832
I0428 15:02:45.135906 30475 solver.cpp:237]     Train net output #0: loss = 0.0534832 (* 1 = 0.0534832 loss)
I0428 15:02:45.135915 30475 sgd_solver.cpp:105] Iteration 7560, lr = 0.00223682
I0428 15:02:50.837631 30475 solver.cpp:218] Iteration 7572 (2.10463 iter/s, 5.70172s/12 iters), loss = 0.024938
I0428 15:02:50.837677 30475 solver.cpp:237]     Train net output #0: loss = 0.0249381 (* 1 = 0.0249381 loss)
I0428 15:02:50.837684 30475 sgd_solver.cpp:105] Iteration 7572, lr = 0.00223151
I0428 15:02:56.388731 30475 solver.cpp:218] Iteration 7584 (2.16176 iter/s, 5.55104s/12 iters), loss = 0.0193546
I0428 15:02:56.388845 30475 solver.cpp:237]     Train net output #0: loss = 0.0193547 (* 1 = 0.0193547 loss)
I0428 15:02:56.388855 30475 sgd_solver.cpp:105] Iteration 7584, lr = 0.00222621
I0428 15:02:57.111685 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:03:01.975665 30475 solver.cpp:218] Iteration 7596 (2.14792 iter/s, 5.58681s/12 iters), loss = 0.0481596
I0428 15:03:01.975709 30475 solver.cpp:237]     Train net output #0: loss = 0.0481596 (* 1 = 0.0481596 loss)
I0428 15:03:01.975718 30475 sgd_solver.cpp:105] Iteration 7596, lr = 0.00222093
I0428 15:03:07.587709 30475 solver.cpp:218] Iteration 7608 (2.13828 iter/s, 5.61198s/12 iters), loss = 0.083599
I0428 15:03:07.587774 30475 solver.cpp:237]     Train net output #0: loss = 0.083599 (* 1 = 0.083599 loss)
I0428 15:03:07.587786 30475 sgd_solver.cpp:105] Iteration 7608, lr = 0.00221565
I0428 15:03:13.346318 30475 solver.cpp:218] Iteration 7620 (2.08386 iter/s, 5.75854s/12 iters), loss = 0.0349331
I0428 15:03:13.346362 30475 solver.cpp:237]     Train net output #0: loss = 0.0349331 (* 1 = 0.0349331 loss)
I0428 15:03:13.346370 30475 sgd_solver.cpp:105] Iteration 7620, lr = 0.00221039
I0428 15:03:16.146667 30475 blocking_queue.cpp:49] Waiting for data
I0428 15:03:19.162959 30475 solver.cpp:218] Iteration 7632 (2.06306 iter/s, 5.81659s/12 iters), loss = 0.0273792
I0428 15:03:19.163000 30475 solver.cpp:237]     Train net output #0: loss = 0.0273793 (* 1 = 0.0273793 loss)
I0428 15:03:19.163008 30475 sgd_solver.cpp:105] Iteration 7632, lr = 0.00220515
I0428 15:03:24.853513 30475 solver.cpp:218] Iteration 7644 (2.10878 iter/s, 5.6905s/12 iters), loss = 0.0120613
I0428 15:03:24.853561 30475 solver.cpp:237]     Train net output #0: loss = 0.0120613 (* 1 = 0.0120613 loss)
I0428 15:03:24.853569 30475 sgd_solver.cpp:105] Iteration 7644, lr = 0.00219991
I0428 15:03:27.198102 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7650.caffemodel
I0428 15:03:32.605446 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7650.solverstate
I0428 15:03:35.441787 30475 solver.cpp:330] Iteration 7650, Testing net (#0)
I0428 15:03:35.441808 30475 net.cpp:676] Ignoring source layer train-data
I0428 15:03:37.015913 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:03:40.553493 30475 solver.cpp:397]     Test net output #0: accuracy = 0.454657
I0428 15:03:40.553536 30475 solver.cpp:397]     Test net output #1: loss = 3.90247 (* 1 = 3.90247 loss)
I0428 15:03:42.688032 30475 solver.cpp:218] Iteration 7656 (0.672854 iter/s, 17.8345s/12 iters), loss = 0.0328797
I0428 15:03:42.688079 30475 solver.cpp:237]     Train net output #0: loss = 0.0328797 (* 1 = 0.0328797 loss)
I0428 15:03:42.688088 30475 sgd_solver.cpp:105] Iteration 7656, lr = 0.00219469
I0428 15:03:48.305150 30475 solver.cpp:218] Iteration 7668 (2.13635 iter/s, 5.61706s/12 iters), loss = 0.0292696
I0428 15:03:48.305199 30475 solver.cpp:237]     Train net output #0: loss = 0.0292696 (* 1 = 0.0292696 loss)
I0428 15:03:48.305208 30475 sgd_solver.cpp:105] Iteration 7668, lr = 0.00218948
I0428 15:03:53.748178 30475 solver.cpp:218] Iteration 7680 (2.20468 iter/s, 5.44297s/12 iters), loss = 0.0768137
I0428 15:03:53.748220 30475 solver.cpp:237]     Train net output #0: loss = 0.0768137 (* 1 = 0.0768137 loss)
I0428 15:03:53.748229 30475 sgd_solver.cpp:105] Iteration 7680, lr = 0.00218428
I0428 15:03:56.887663 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:03:59.444702 30475 solver.cpp:218] Iteration 7692 (2.10657 iter/s, 5.69647s/12 iters), loss = 0.0257133
I0428 15:03:59.444839 30475 solver.cpp:237]     Train net output #0: loss = 0.0257133 (* 1 = 0.0257133 loss)
I0428 15:03:59.444849 30475 sgd_solver.cpp:105] Iteration 7692, lr = 0.00217909
I0428 15:04:04.975162 30475 solver.cpp:218] Iteration 7704 (2.16986 iter/s, 5.53031s/12 iters), loss = 0.0458577
I0428 15:04:04.975221 30475 solver.cpp:237]     Train net output #0: loss = 0.0458577 (* 1 = 0.0458577 loss)
I0428 15:04:04.975234 30475 sgd_solver.cpp:105] Iteration 7704, lr = 0.00217392
I0428 15:04:10.612880 30475 solver.cpp:218] Iteration 7716 (2.12855 iter/s, 5.63765s/12 iters), loss = 0.0496735
I0428 15:04:10.612937 30475 solver.cpp:237]     Train net output #0: loss = 0.0496735 (* 1 = 0.0496735 loss)
I0428 15:04:10.612951 30475 sgd_solver.cpp:105] Iteration 7716, lr = 0.00216876
I0428 15:04:16.252198 30475 solver.cpp:218] Iteration 7728 (2.12794 iter/s, 5.63926s/12 iters), loss = 0.042497
I0428 15:04:16.252243 30475 solver.cpp:237]     Train net output #0: loss = 0.042497 (* 1 = 0.042497 loss)
I0428 15:04:16.252251 30475 sgd_solver.cpp:105] Iteration 7728, lr = 0.00216361
I0428 15:04:21.892702 30475 solver.cpp:218] Iteration 7740 (2.12749 iter/s, 5.64045s/12 iters), loss = 0.0280217
I0428 15:04:21.892748 30475 solver.cpp:237]     Train net output #0: loss = 0.0280217 (* 1 = 0.0280217 loss)
I0428 15:04:21.892756 30475 sgd_solver.cpp:105] Iteration 7740, lr = 0.00215847
I0428 15:04:26.976563 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7752.caffemodel
I0428 15:04:31.035543 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7752.solverstate
I0428 15:04:34.919796 30475 solver.cpp:330] Iteration 7752, Testing net (#0)
I0428 15:04:34.919826 30475 net.cpp:676] Ignoring source layer train-data
I0428 15:04:36.474983 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:04:40.122783 30475 solver.cpp:397]     Test net output #0: accuracy = 0.458946
I0428 15:04:40.122812 30475 solver.cpp:397]     Test net output #1: loss = 4.02367 (* 1 = 4.02367 loss)
I0428 15:04:40.274940 30475 solver.cpp:218] Iteration 7752 (0.652805 iter/s, 18.3822s/12 iters), loss = 0.0371606
I0428 15:04:40.274988 30475 solver.cpp:237]     Train net output #0: loss = 0.0371606 (* 1 = 0.0371606 loss)
I0428 15:04:40.274996 30475 sgd_solver.cpp:105] Iteration 7752, lr = 0.00215335
I0428 15:04:45.075955 30475 solver.cpp:218] Iteration 7764 (2.4995 iter/s, 4.80096s/12 iters), loss = 0.0526995
I0428 15:04:45.076000 30475 solver.cpp:237]     Train net output #0: loss = 0.0526995 (* 1 = 0.0526995 loss)
I0428 15:04:45.076009 30475 sgd_solver.cpp:105] Iteration 7764, lr = 0.00214823
I0428 15:04:50.792582 30475 solver.cpp:218] Iteration 7776 (2.09916 iter/s, 5.71657s/12 iters), loss = 0.0179384
I0428 15:04:50.792626 30475 solver.cpp:237]     Train net output #0: loss = 0.0179384 (* 1 = 0.0179384 loss)
I0428 15:04:50.792635 30475 sgd_solver.cpp:105] Iteration 7776, lr = 0.00214313
I0428 15:04:56.350322 30475 solver.cpp:218] Iteration 7788 (2.15917 iter/s, 5.55769s/12 iters), loss = 0.079674
I0428 15:04:56.350366 30475 solver.cpp:237]     Train net output #0: loss = 0.079674 (* 1 = 0.079674 loss)
I0428 15:04:56.350375 30475 sgd_solver.cpp:105] Iteration 7788, lr = 0.00213805
I0428 15:04:56.357986 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:05:01.994663 30475 solver.cpp:218] Iteration 7800 (2.12604 iter/s, 5.64429s/12 iters), loss = 0.0734149
I0428 15:05:01.994814 30475 solver.cpp:237]     Train net output #0: loss = 0.0734149 (* 1 = 0.0734149 loss)
I0428 15:05:01.994824 30475 sgd_solver.cpp:105] Iteration 7800, lr = 0.00213297
I0428 15:05:07.646440 30475 solver.cpp:218] Iteration 7812 (2.12328 iter/s, 5.65163s/12 iters), loss = 0.0158482
I0428 15:05:07.646481 30475 solver.cpp:237]     Train net output #0: loss = 0.0158482 (* 1 = 0.0158482 loss)
I0428 15:05:07.646488 30475 sgd_solver.cpp:105] Iteration 7812, lr = 0.00212791
I0428 15:05:13.311450 30475 solver.cpp:218] Iteration 7824 (2.11829 iter/s, 5.66496s/12 iters), loss = 0.0691332
I0428 15:05:13.311496 30475 solver.cpp:237]     Train net output #0: loss = 0.0691332 (* 1 = 0.0691332 loss)
I0428 15:05:13.311504 30475 sgd_solver.cpp:105] Iteration 7824, lr = 0.00212285
I0428 15:05:18.955868 30475 solver.cpp:218] Iteration 7836 (2.12601 iter/s, 5.64436s/12 iters), loss = 0.0408925
I0428 15:05:18.955914 30475 solver.cpp:237]     Train net output #0: loss = 0.0408925 (* 1 = 0.0408925 loss)
I0428 15:05:18.955922 30475 sgd_solver.cpp:105] Iteration 7836, lr = 0.00211781
I0428 15:05:24.496552 30475 solver.cpp:218] Iteration 7848 (2.16582 iter/s, 5.54063s/12 iters), loss = 0.0487207
I0428 15:05:24.496599 30475 solver.cpp:237]     Train net output #0: loss = 0.0487207 (* 1 = 0.0487207 loss)
I0428 15:05:24.496608 30475 sgd_solver.cpp:105] Iteration 7848, lr = 0.00211279
I0428 15:05:26.738267 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7854.caffemodel
I0428 15:05:31.218168 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7854.solverstate
I0428 15:05:34.277674 30475 solver.cpp:330] Iteration 7854, Testing net (#0)
I0428 15:05:34.277794 30475 net.cpp:676] Ignoring source layer train-data
I0428 15:05:35.830549 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:05:39.464689 30475 solver.cpp:397]     Test net output #0: accuracy = 0.463848
I0428 15:05:39.464712 30475 solver.cpp:397]     Test net output #1: loss = 3.857 (* 1 = 3.857 loss)
I0428 15:05:41.485635 30475 solver.cpp:218] Iteration 7860 (0.706337 iter/s, 16.9891s/12 iters), loss = 0.0173527
I0428 15:05:41.485684 30475 solver.cpp:237]     Train net output #0: loss = 0.0173527 (* 1 = 0.0173527 loss)
I0428 15:05:41.485693 30475 sgd_solver.cpp:105] Iteration 7860, lr = 0.00210777
I0428 15:05:47.065840 30475 solver.cpp:218] Iteration 7872 (2.15048 iter/s, 5.58015s/12 iters), loss = 0.0547729
I0428 15:05:47.065883 30475 solver.cpp:237]     Train net output #0: loss = 0.0547729 (* 1 = 0.0547729 loss)
I0428 15:05:47.065891 30475 sgd_solver.cpp:105] Iteration 7872, lr = 0.00210277
I0428 15:05:52.703619 30475 solver.cpp:218] Iteration 7884 (2.12852 iter/s, 5.63772s/12 iters), loss = 0.0233936
I0428 15:05:52.703672 30475 solver.cpp:237]     Train net output #0: loss = 0.0233936 (* 1 = 0.0233936 loss)
I0428 15:05:52.703681 30475 sgd_solver.cpp:105] Iteration 7884, lr = 0.00209777
I0428 15:05:55.034837 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:05:58.275854 30475 solver.cpp:218] Iteration 7896 (2.15356 iter/s, 5.57218s/12 iters), loss = 0.0319481
I0428 15:05:58.275902 30475 solver.cpp:237]     Train net output #0: loss = 0.0319482 (* 1 = 0.0319482 loss)
I0428 15:05:58.275909 30475 sgd_solver.cpp:105] Iteration 7896, lr = 0.00209279
I0428 15:06:03.897195 30475 solver.cpp:218] Iteration 7908 (2.13474 iter/s, 5.62129s/12 iters), loss = 0.0474058
I0428 15:06:03.897236 30475 solver.cpp:237]     Train net output #0: loss = 0.0474058 (* 1 = 0.0474058 loss)
I0428 15:06:03.897245 30475 sgd_solver.cpp:105] Iteration 7908, lr = 0.00208782
I0428 15:06:09.432209 30475 solver.cpp:218] Iteration 7920 (2.16803 iter/s, 5.53497s/12 iters), loss = 0.0180098
I0428 15:06:09.432354 30475 solver.cpp:237]     Train net output #0: loss = 0.0180099 (* 1 = 0.0180099 loss)
I0428 15:06:09.432363 30475 sgd_solver.cpp:105] Iteration 7920, lr = 0.00208287
I0428 15:06:15.158454 30475 solver.cpp:218] Iteration 7932 (2.09567 iter/s, 5.7261s/12 iters), loss = 0.035596
I0428 15:06:15.158497 30475 solver.cpp:237]     Train net output #0: loss = 0.035596 (* 1 = 0.035596 loss)
I0428 15:06:15.158506 30475 sgd_solver.cpp:105] Iteration 7932, lr = 0.00207792
I0428 15:06:20.793524 30475 solver.cpp:218] Iteration 7944 (2.12954 iter/s, 5.63502s/12 iters), loss = 0.0231777
I0428 15:06:20.793570 30475 solver.cpp:237]     Train net output #0: loss = 0.0231777 (* 1 = 0.0231777 loss)
I0428 15:06:20.793578 30475 sgd_solver.cpp:105] Iteration 7944, lr = 0.00207299
I0428 15:06:25.866370 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7956.caffemodel
I0428 15:06:30.664608 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7956.solverstate
I0428 15:06:34.366703 30475 solver.cpp:330] Iteration 7956, Testing net (#0)
I0428 15:06:34.366722 30475 net.cpp:676] Ignoring source layer train-data
I0428 15:06:35.808730 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:06:39.238031 30475 solver.cpp:397]     Test net output #0: accuracy = 0.476103
I0428 15:06:39.238076 30475 solver.cpp:397]     Test net output #1: loss = 4.0689 (* 1 = 4.0689 loss)
I0428 15:06:39.399488 30475 solver.cpp:218] Iteration 7956 (0.644955 iter/s, 18.6059s/12 iters), loss = 0.0133597
I0428 15:06:39.399541 30475 solver.cpp:237]     Train net output #0: loss = 0.0133598 (* 1 = 0.0133598 loss)
I0428 15:06:39.399551 30475 sgd_solver.cpp:105] Iteration 7956, lr = 0.00206807
I0428 15:06:44.202167 30475 solver.cpp:218] Iteration 7968 (2.49864 iter/s, 4.80262s/12 iters), loss = 0.0216546
I0428 15:06:44.202283 30475 solver.cpp:237]     Train net output #0: loss = 0.0216546 (* 1 = 0.0216546 loss)
I0428 15:06:44.202294 30475 sgd_solver.cpp:105] Iteration 7968, lr = 0.00206316
I0428 15:06:49.750370 30475 solver.cpp:218] Iteration 7980 (2.16291 iter/s, 5.54808s/12 iters), loss = 0.0106728
I0428 15:06:49.750416 30475 solver.cpp:237]     Train net output #0: loss = 0.0106728 (* 1 = 0.0106728 loss)
I0428 15:06:49.750424 30475 sgd_solver.cpp:105] Iteration 7980, lr = 0.00205826
I0428 15:06:54.647222 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:06:55.477851 30475 solver.cpp:218] Iteration 7992 (2.09518 iter/s, 5.72743s/12 iters), loss = 0.024637
I0428 15:06:55.477895 30475 solver.cpp:237]     Train net output #0: loss = 0.0246371 (* 1 = 0.0246371 loss)
I0428 15:06:55.477902 30475 sgd_solver.cpp:105] Iteration 7992, lr = 0.00205337
I0428 15:07:01.124646 30475 solver.cpp:218] Iteration 8004 (2.12512 iter/s, 5.64674s/12 iters), loss = 0.0228985
I0428 15:07:01.124689 30475 solver.cpp:237]     Train net output #0: loss = 0.0228985 (* 1 = 0.0228985 loss)
I0428 15:07:01.124696 30475 sgd_solver.cpp:105] Iteration 8004, lr = 0.0020485
I0428 15:07:06.696728 30475 solver.cpp:218] Iteration 8016 (2.15361 iter/s, 5.57203s/12 iters), loss = 0.022651
I0428 15:07:06.696771 30475 solver.cpp:237]     Train net output #0: loss = 0.022651 (* 1 = 0.022651 loss)
I0428 15:07:06.696780 30475 sgd_solver.cpp:105] Iteration 8016, lr = 0.00204363
I0428 15:07:12.329212 30475 solver.cpp:218] Iteration 8028 (2.13052 iter/s, 5.63244s/12 iters), loss = 0.0684708
I0428 15:07:12.329254 30475 solver.cpp:237]     Train net output #0: loss = 0.0684708 (* 1 = 0.0684708 loss)
I0428 15:07:12.329262 30475 sgd_solver.cpp:105] Iteration 8028, lr = 0.00203878
I0428 15:07:17.962869 30475 solver.cpp:218] Iteration 8040 (2.13007 iter/s, 5.63361s/12 iters), loss = 0.0267009
I0428 15:07:17.963033 30475 solver.cpp:237]     Train net output #0: loss = 0.026701 (* 1 = 0.026701 loss)
I0428 15:07:17.963043 30475 sgd_solver.cpp:105] Iteration 8040, lr = 0.00203394
I0428 15:07:23.896235 30475 solver.cpp:218] Iteration 8052 (2.02252 iter/s, 5.9332s/12 iters), loss = 0.0403374
I0428 15:07:23.896276 30475 solver.cpp:237]     Train net output #0: loss = 0.0403375 (* 1 = 0.0403375 loss)
I0428 15:07:23.896286 30475 sgd_solver.cpp:105] Iteration 8052, lr = 0.00202911
I0428 15:07:26.215191 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8058.caffemodel
I0428 15:07:30.724488 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8058.solverstate
I0428 15:07:36.499722 30475 solver.cpp:330] Iteration 8058, Testing net (#0)
I0428 15:07:36.499738 30475 net.cpp:676] Ignoring source layer train-data
I0428 15:07:37.859025 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:07:41.490931 30475 solver.cpp:397]     Test net output #0: accuracy = 0.460172
I0428 15:07:41.490960 30475 solver.cpp:397]     Test net output #1: loss = 4.08812 (* 1 = 4.08812 loss)
I0428 15:07:43.720002 30475 solver.cpp:218] Iteration 8064 (0.605334 iter/s, 19.8238s/12 iters), loss = 0.0288256
I0428 15:07:43.720046 30475 solver.cpp:237]     Train net output #0: loss = 0.0288256 (* 1 = 0.0288256 loss)
I0428 15:07:43.720054 30475 sgd_solver.cpp:105] Iteration 8064, lr = 0.00202429
I0428 15:07:49.425439 30475 solver.cpp:218] Iteration 8076 (2.10328 iter/s, 5.70539s/12 iters), loss = 0.0859959
I0428 15:07:49.425549 30475 solver.cpp:237]     Train net output #0: loss = 0.0859959 (* 1 = 0.0859959 loss)
I0428 15:07:49.425559 30475 sgd_solver.cpp:105] Iteration 8076, lr = 0.00201949
I0428 15:07:55.072259 30475 solver.cpp:218] Iteration 8088 (2.12514 iter/s, 5.6467s/12 iters), loss = 0.053328
I0428 15:07:55.072304 30475 solver.cpp:237]     Train net output #0: loss = 0.053328 (* 1 = 0.053328 loss)
I0428 15:07:55.072312 30475 sgd_solver.cpp:105] Iteration 8088, lr = 0.00201469
I0428 15:07:56.642346 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:08:00.715301 30475 solver.cpp:218] Iteration 8100 (2.12653 iter/s, 5.64299s/12 iters), loss = 0.0215123
I0428 15:08:00.715340 30475 solver.cpp:237]     Train net output #0: loss = 0.0215123 (* 1 = 0.0215123 loss)
I0428 15:08:00.715349 30475 sgd_solver.cpp:105] Iteration 8100, lr = 0.00200991
I0428 15:08:06.328377 30475 solver.cpp:218] Iteration 8112 (2.13788 iter/s, 5.61303s/12 iters), loss = 0.0550813
I0428 15:08:06.328423 30475 solver.cpp:237]     Train net output #0: loss = 0.0550813 (* 1 = 0.0550813 loss)
I0428 15:08:06.328431 30475 sgd_solver.cpp:105] Iteration 8112, lr = 0.00200514
I0428 15:08:11.856761 30475 solver.cpp:218] Iteration 8124 (2.17064 iter/s, 5.52833s/12 iters), loss = 0.0332082
I0428 15:08:11.856806 30475 solver.cpp:237]     Train net output #0: loss = 0.0332082 (* 1 = 0.0332082 loss)
I0428 15:08:11.856812 30475 sgd_solver.cpp:105] Iteration 8124, lr = 0.00200038
I0428 15:08:17.491829 30475 solver.cpp:218] Iteration 8136 (2.12954 iter/s, 5.63502s/12 iters), loss = 0.00684697
I0428 15:08:17.491875 30475 solver.cpp:237]     Train net output #0: loss = 0.00684702 (* 1 = 0.00684702 loss)
I0428 15:08:17.491884 30475 sgd_solver.cpp:105] Iteration 8136, lr = 0.00199563
I0428 15:08:23.111699 30475 solver.cpp:218] Iteration 8148 (2.1353 iter/s, 5.61982s/12 iters), loss = 0.0274722
I0428 15:08:23.111806 30475 solver.cpp:237]     Train net output #0: loss = 0.0274722 (* 1 = 0.0274722 loss)
I0428 15:08:23.111815 30475 sgd_solver.cpp:105] Iteration 8148, lr = 0.00199089
I0428 15:08:28.170367 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8160.caffemodel
I0428 15:08:30.398761 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8160.solverstate
I0428 15:08:34.784312 30475 solver.cpp:330] Iteration 8160, Testing net (#0)
I0428 15:08:34.784332 30475 net.cpp:676] Ignoring source layer train-data
I0428 15:08:36.134088 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:08:39.896844 30475 solver.cpp:397]     Test net output #0: accuracy = 0.467524
I0428 15:08:39.896888 30475 solver.cpp:397]     Test net output #1: loss = 4.07288 (* 1 = 4.07288 loss)
I0428 15:08:40.051259 30475 solver.cpp:218] Iteration 8160 (0.708405 iter/s, 16.9395s/12 iters), loss = 0.036864
I0428 15:08:40.052780 30475 solver.cpp:237]     Train net output #0: loss = 0.036864 (* 1 = 0.036864 loss)
I0428 15:08:40.052793 30475 sgd_solver.cpp:105] Iteration 8160, lr = 0.00198616
I0428 15:08:44.716574 30475 solver.cpp:218] Iteration 8172 (2.57301 iter/s, 4.66379s/12 iters), loss = 0.0215298
I0428 15:08:44.716614 30475 solver.cpp:237]     Train net output #0: loss = 0.0215299 (* 1 = 0.0215299 loss)
I0428 15:08:44.716622 30475 sgd_solver.cpp:105] Iteration 8172, lr = 0.00198145
I0428 15:08:50.433624 30475 solver.cpp:218] Iteration 8184 (2.099 iter/s, 5.717s/12 iters), loss = 0.0206996
I0428 15:08:50.433667 30475 solver.cpp:237]     Train net output #0: loss = 0.0206996 (* 1 = 0.0206996 loss)
I0428 15:08:50.433676 30475 sgd_solver.cpp:105] Iteration 8184, lr = 0.00197674
I0428 15:08:54.416182 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:08:56.089720 30475 solver.cpp:218] Iteration 8196 (2.12162 iter/s, 5.65604s/12 iters), loss = 0.0252607
I0428 15:08:56.089766 30475 solver.cpp:237]     Train net output #0: loss = 0.0252608 (* 1 = 0.0252608 loss)
I0428 15:08:56.089776 30475 sgd_solver.cpp:105] Iteration 8196, lr = 0.00197205
I0428 15:09:01.620849 30475 solver.cpp:218] Iteration 8208 (2.16956 iter/s, 5.53108s/12 iters), loss = 0.0360405
I0428 15:09:01.620889 30475 solver.cpp:237]     Train net output #0: loss = 0.0360405 (* 1 = 0.0360405 loss)
I0428 15:09:01.620898 30475 sgd_solver.cpp:105] Iteration 8208, lr = 0.00196737
I0428 15:09:07.262136 30475 solver.cpp:218] Iteration 8220 (2.12719 iter/s, 5.64124s/12 iters), loss = 0.0145698
I0428 15:09:07.262181 30475 solver.cpp:237]     Train net output #0: loss = 0.0145699 (* 1 = 0.0145699 loss)
I0428 15:09:07.262188 30475 sgd_solver.cpp:105] Iteration 8220, lr = 0.0019627
I0428 15:09:12.882838 30475 solver.cpp:218] Iteration 8232 (2.13499 iter/s, 5.62065s/12 iters), loss = 0.0185226
I0428 15:09:12.882886 30475 solver.cpp:237]     Train net output #0: loss = 0.0185227 (* 1 = 0.0185227 loss)
I0428 15:09:12.882895 30475 sgd_solver.cpp:105] Iteration 8232, lr = 0.00195804
I0428 15:09:18.519491 30475 solver.cpp:218] Iteration 8244 (2.12894 iter/s, 5.6366s/12 iters), loss = 0.0322007
I0428 15:09:18.519541 30475 solver.cpp:237]     Train net output #0: loss = 0.0322007 (* 1 = 0.0322007 loss)
I0428 15:09:18.519549 30475 sgd_solver.cpp:105] Iteration 8244, lr = 0.00195339
I0428 15:09:24.182493 30475 solver.cpp:218] Iteration 8256 (2.11904 iter/s, 5.66294s/12 iters), loss = 0.0712548
I0428 15:09:24.182548 30475 solver.cpp:237]     Train net output #0: loss = 0.0712548 (* 1 = 0.0712548 loss)
I0428 15:09:24.182559 30475 sgd_solver.cpp:105] Iteration 8256, lr = 0.00194875
I0428 15:09:26.420646 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8262.caffemodel
I0428 15:09:28.626933 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8262.solverstate
I0428 15:09:30.330770 30475 solver.cpp:330] Iteration 8262, Testing net (#0)
I0428 15:09:30.330792 30475 net.cpp:676] Ignoring source layer train-data
I0428 15:09:31.650286 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:09:35.459656 30475 solver.cpp:397]     Test net output #0: accuracy = 0.457108
I0428 15:09:35.459694 30475 solver.cpp:397]     Test net output #1: loss = 3.96595 (* 1 = 3.96595 loss)
I0428 15:09:37.576787 30475 solver.cpp:218] Iteration 8268 (0.895907 iter/s, 13.3943s/12 iters), loss = 0.0465776
I0428 15:09:37.576826 30475 solver.cpp:237]     Train net output #0: loss = 0.0465776 (* 1 = 0.0465776 loss)
I0428 15:09:37.576834 30475 sgd_solver.cpp:105] Iteration 8268, lr = 0.00194412
I0428 15:09:43.224602 30475 solver.cpp:218] Iteration 8280 (2.12473 iter/s, 5.64777s/12 iters), loss = 0.0262709
I0428 15:09:43.224651 30475 solver.cpp:237]     Train net output #0: loss = 0.0262709 (* 1 = 0.0262709 loss)
I0428 15:09:43.224659 30475 sgd_solver.cpp:105] Iteration 8280, lr = 0.00193951
I0428 15:09:48.779371 30475 solver.cpp:218] Iteration 8292 (2.16033 iter/s, 5.55471s/12 iters), loss = 0.0246077
I0428 15:09:48.779412 30475 solver.cpp:237]     Train net output #0: loss = 0.0246077 (* 1 = 0.0246077 loss)
I0428 15:09:48.779420 30475 sgd_solver.cpp:105] Iteration 8292, lr = 0.0019349
I0428 15:09:49.512024 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:09:54.339984 30475 solver.cpp:218] Iteration 8304 (2.15806 iter/s, 5.56056s/12 iters), loss = 0.00995857
I0428 15:09:54.340030 30475 solver.cpp:237]     Train net output #0: loss = 0.00995861 (* 1 = 0.00995861 loss)
I0428 15:09:54.340039 30475 sgd_solver.cpp:105] Iteration 8304, lr = 0.00193031
I0428 15:09:57.514652 30475 blocking_queue.cpp:49] Waiting for data
I0428 15:09:59.958081 30475 solver.cpp:218] Iteration 8316 (2.13597 iter/s, 5.61805s/12 iters), loss = 0.052139
I0428 15:09:59.958124 30475 solver.cpp:237]     Train net output #0: loss = 0.0521391 (* 1 = 0.0521391 loss)
I0428 15:09:59.958132 30475 sgd_solver.cpp:105] Iteration 8316, lr = 0.00192573
I0428 15:10:05.559810 30475 solver.cpp:218] Iteration 8328 (2.14222 iter/s, 5.60168s/12 iters), loss = 0.013769
I0428 15:10:05.559856 30475 solver.cpp:237]     Train net output #0: loss = 0.013769 (* 1 = 0.013769 loss)
I0428 15:10:05.559864 30475 sgd_solver.cpp:105] Iteration 8328, lr = 0.00192115
I0428 15:10:11.112164 30475 solver.cpp:218] Iteration 8340 (2.16127 iter/s, 5.5523s/12 iters), loss = 0.00538887
I0428 15:10:11.112207 30475 solver.cpp:237]     Train net output #0: loss = 0.0053889 (* 1 = 0.0053889 loss)
I0428 15:10:11.112217 30475 sgd_solver.cpp:105] Iteration 8340, lr = 0.00191659
I0428 15:10:16.775892 30475 solver.cpp:218] Iteration 8352 (2.11877 iter/s, 5.66367s/12 iters), loss = 0.0302173
I0428 15:10:16.775940 30475 solver.cpp:237]     Train net output #0: loss = 0.0302173 (* 1 = 0.0302173 loss)
I0428 15:10:16.775949 30475 sgd_solver.cpp:105] Iteration 8352, lr = 0.00191204
I0428 15:10:21.849539 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8364.caffemodel
I0428 15:10:24.040910 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8364.solverstate
I0428 15:10:25.739262 30475 solver.cpp:330] Iteration 8364, Testing net (#0)
I0428 15:10:25.739280 30475 net.cpp:676] Ignoring source layer train-data
I0428 15:10:27.018844 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:10:30.934000 30475 solver.cpp:397]     Test net output #0: accuracy = 0.480392
I0428 15:10:30.934110 30475 solver.cpp:397]     Test net output #1: loss = 3.92598 (* 1 = 3.92598 loss)
I0428 15:10:31.087275 30475 solver.cpp:218] Iteration 8364 (0.838495 iter/s, 14.3114s/12 iters), loss = 0.0257921
I0428 15:10:31.087327 30475 solver.cpp:237]     Train net output #0: loss = 0.0257922 (* 1 = 0.0257922 loss)
I0428 15:10:31.087337 30475 sgd_solver.cpp:105] Iteration 8364, lr = 0.0019075
I0428 15:10:35.698804 30475 solver.cpp:218] Iteration 8376 (2.60221 iter/s, 4.61147s/12 iters), loss = 0.0303633
I0428 15:10:35.698846 30475 solver.cpp:237]     Train net output #0: loss = 0.0303634 (* 1 = 0.0303634 loss)
I0428 15:10:35.698855 30475 sgd_solver.cpp:105] Iteration 8376, lr = 0.00190297
I0428 15:10:41.347586 30475 solver.cpp:218] Iteration 8388 (2.12437 iter/s, 5.64873s/12 iters), loss = 0.00828446
I0428 15:10:41.347633 30475 solver.cpp:237]     Train net output #0: loss = 0.00828448 (* 1 = 0.00828448 loss)
I0428 15:10:41.347642 30475 sgd_solver.cpp:105] Iteration 8388, lr = 0.00189846
I0428 15:10:44.394438 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:10:46.957840 30475 solver.cpp:218] Iteration 8400 (2.13896 iter/s, 5.6102s/12 iters), loss = 0.0193697
I0428 15:10:46.957881 30475 solver.cpp:237]     Train net output #0: loss = 0.0193697 (* 1 = 0.0193697 loss)
I0428 15:10:46.957890 30475 sgd_solver.cpp:105] Iteration 8400, lr = 0.00189395
I0428 15:10:52.651206 30475 solver.cpp:218] Iteration 8412 (2.10773 iter/s, 5.69332s/12 iters), loss = 0.0127044
I0428 15:10:52.651247 30475 solver.cpp:237]     Train net output #0: loss = 0.0127045 (* 1 = 0.0127045 loss)
I0428 15:10:52.651254 30475 sgd_solver.cpp:105] Iteration 8412, lr = 0.00188945
I0428 15:10:58.349473 30475 solver.cpp:218] Iteration 8424 (2.10592 iter/s, 5.69822s/12 iters), loss = 0.0159782
I0428 15:10:58.349517 30475 solver.cpp:237]     Train net output #0: loss = 0.0159783 (* 1 = 0.0159783 loss)
I0428 15:10:58.349526 30475 sgd_solver.cpp:105] Iteration 8424, lr = 0.00188497
I0428 15:11:03.936362 30475 solver.cpp:218] Iteration 8436 (2.14791 iter/s, 5.58683s/12 iters), loss = 0.0372708
I0428 15:11:03.936529 30475 solver.cpp:237]     Train net output #0: loss = 0.0372708 (* 1 = 0.0372708 loss)
I0428 15:11:03.936539 30475 sgd_solver.cpp:105] Iteration 8436, lr = 0.00188049
I0428 15:11:09.640504 30475 solver.cpp:218] Iteration 8448 (2.1038 iter/s, 5.70397s/12 iters), loss = 0.0245848
I0428 15:11:09.640548 30475 solver.cpp:237]     Train net output #0: loss = 0.0245848 (* 1 = 0.0245848 loss)
I0428 15:11:09.640558 30475 sgd_solver.cpp:105] Iteration 8448, lr = 0.00187603
I0428 15:11:15.268430 30475 solver.cpp:218] Iteration 8460 (2.13224 iter/s, 5.62787s/12 iters), loss = 0.0166277
I0428 15:11:15.268474 30475 solver.cpp:237]     Train net output #0: loss = 0.0166278 (* 1 = 0.0166278 loss)
I0428 15:11:15.268483 30475 sgd_solver.cpp:105] Iteration 8460, lr = 0.00187157
I0428 15:11:17.526298 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8466.caffemodel
I0428 15:11:20.660748 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8466.solverstate
I0428 15:11:23.186290 30475 solver.cpp:330] Iteration 8466, Testing net (#0)
I0428 15:11:23.186319 30475 net.cpp:676] Ignoring source layer train-data
I0428 15:11:24.413569 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:11:28.298307 30475 solver.cpp:397]     Test net output #0: accuracy = 0.471201
I0428 15:11:28.298334 30475 solver.cpp:397]     Test net output #1: loss = 3.98377 (* 1 = 3.98377 loss)
I0428 15:11:30.403424 30475 solver.cpp:218] Iteration 8472 (0.792866 iter/s, 15.135s/12 iters), loss = 0.0115699
I0428 15:11:30.403476 30475 solver.cpp:237]     Train net output #0: loss = 0.0115699 (* 1 = 0.0115699 loss)
I0428 15:11:30.403486 30475 sgd_solver.cpp:105] Iteration 8472, lr = 0.00186713
I0428 15:11:36.032125 30475 solver.cpp:218] Iteration 8484 (2.13195 iter/s, 5.62864s/12 iters), loss = 0.107683
I0428 15:11:36.032220 30475 solver.cpp:237]     Train net output #0: loss = 0.107683 (* 1 = 0.107683 loss)
I0428 15:11:36.032229 30475 sgd_solver.cpp:105] Iteration 8484, lr = 0.0018627
I0428 15:11:41.892648 30475 solver.cpp:218] Iteration 8496 (2.04763 iter/s, 5.86042s/12 iters), loss = 0.0156041
I0428 15:11:41.892686 30475 solver.cpp:237]     Train net output #0: loss = 0.0156041 (* 1 = 0.0156041 loss)
I0428 15:11:41.892693 30475 sgd_solver.cpp:105] Iteration 8496, lr = 0.00185827
I0428 15:11:41.929026 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:11:47.435021 30475 solver.cpp:218] Iteration 8508 (2.16515 iter/s, 5.54233s/12 iters), loss = 0.0313765
I0428 15:11:47.435063 30475 solver.cpp:237]     Train net output #0: loss = 0.0313765 (* 1 = 0.0313765 loss)
I0428 15:11:47.435072 30475 sgd_solver.cpp:105] Iteration 8508, lr = 0.00185386
I0428 15:11:53.053246 30475 solver.cpp:218] Iteration 8520 (2.13593 iter/s, 5.61817s/12 iters), loss = 0.0209625
I0428 15:11:53.053290 30475 solver.cpp:237]     Train net output #0: loss = 0.0209625 (* 1 = 0.0209625 loss)
I0428 15:11:53.053298 30475 sgd_solver.cpp:105] Iteration 8520, lr = 0.00184946
I0428 15:11:58.684845 30475 solver.cpp:218] Iteration 8532 (2.13085 iter/s, 5.63154s/12 iters), loss = 0.0631172
I0428 15:11:58.684892 30475 solver.cpp:237]     Train net output #0: loss = 0.0631172 (* 1 = 0.0631172 loss)
I0428 15:11:58.684904 30475 sgd_solver.cpp:105] Iteration 8532, lr = 0.00184507
I0428 15:12:04.340154 30475 solver.cpp:218] Iteration 8544 (2.12192 iter/s, 5.65526s/12 iters), loss = 0.00719106
I0428 15:12:04.340195 30475 solver.cpp:237]     Train net output #0: loss = 0.00719109 (* 1 = 0.00719109 loss)
I0428 15:12:04.340204 30475 sgd_solver.cpp:105] Iteration 8544, lr = 0.00184069
I0428 15:12:09.969447 30475 solver.cpp:218] Iteration 8556 (2.13173 iter/s, 5.62924s/12 iters), loss = 0.0448844
I0428 15:12:09.969580 30475 solver.cpp:237]     Train net output #0: loss = 0.0448845 (* 1 = 0.0448845 loss)
I0428 15:12:09.969590 30475 sgd_solver.cpp:105] Iteration 8556, lr = 0.00183632
I0428 15:12:14.833251 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8568.caffemodel
I0428 15:12:20.575528 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8568.solverstate
I0428 15:12:24.459089 30475 solver.cpp:330] Iteration 8568, Testing net (#0)
I0428 15:12:24.459112 30475 net.cpp:676] Ignoring source layer train-data
I0428 15:12:25.642410 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:12:29.620995 30475 solver.cpp:397]     Test net output #0: accuracy = 0.460172
I0428 15:12:29.621023 30475 solver.cpp:397]     Test net output #1: loss = 4.01609 (* 1 = 4.01609 loss)
I0428 15:12:29.782747 30475 solver.cpp:218] Iteration 8568 (0.605657 iter/s, 19.8132s/12 iters), loss = 0.0128492
I0428 15:12:29.782799 30475 solver.cpp:237]     Train net output #0: loss = 0.0128493 (* 1 = 0.0128493 loss)
I0428 15:12:29.782809 30475 sgd_solver.cpp:105] Iteration 8568, lr = 0.00183196
I0428 15:12:34.480473 30475 solver.cpp:218] Iteration 8580 (2.55446 iter/s, 4.69767s/12 iters), loss = 0.014463
I0428 15:12:34.480520 30475 solver.cpp:237]     Train net output #0: loss = 0.014463 (* 1 = 0.014463 loss)
I0428 15:12:34.480528 30475 sgd_solver.cpp:105] Iteration 8580, lr = 0.00182761
I0428 15:12:40.166544 30475 solver.cpp:218] Iteration 8592 (2.11044 iter/s, 5.68602s/12 iters), loss = 0.0423651
I0428 15:12:40.166687 30475 solver.cpp:237]     Train net output #0: loss = 0.0423651 (* 1 = 0.0423651 loss)
I0428 15:12:40.166695 30475 sgd_solver.cpp:105] Iteration 8592, lr = 0.00182327
I0428 15:12:42.620431 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:12:45.855967 30475 solver.cpp:218] Iteration 8604 (2.10923 iter/s, 5.68928s/12 iters), loss = 0.0727721
I0428 15:12:45.856016 30475 solver.cpp:237]     Train net output #0: loss = 0.0727722 (* 1 = 0.0727722 loss)
I0428 15:12:45.856024 30475 sgd_solver.cpp:105] Iteration 8604, lr = 0.00181894
I0428 15:12:51.615422 30475 solver.cpp:218] Iteration 8616 (2.08355 iter/s, 5.7594s/12 iters), loss = 0.0467446
I0428 15:12:51.615469 30475 solver.cpp:237]     Train net output #0: loss = 0.0467446 (* 1 = 0.0467446 loss)
I0428 15:12:51.615478 30475 sgd_solver.cpp:105] Iteration 8616, lr = 0.00181462
I0428 15:12:57.287817 30475 solver.cpp:218] Iteration 8628 (2.11553 iter/s, 5.67234s/12 iters), loss = 0.00925837
I0428 15:12:57.287861 30475 solver.cpp:237]     Train net output #0: loss = 0.00925839 (* 1 = 0.00925839 loss)
I0428 15:12:57.287870 30475 sgd_solver.cpp:105] Iteration 8628, lr = 0.00181031
I0428 15:13:02.940645 30475 solver.cpp:218] Iteration 8640 (2.12285 iter/s, 5.65277s/12 iters), loss = 0.0354258
I0428 15:13:02.940693 30475 solver.cpp:237]     Train net output #0: loss = 0.0354259 (* 1 = 0.0354259 loss)
I0428 15:13:02.940701 30475 sgd_solver.cpp:105] Iteration 8640, lr = 0.00180602
I0428 15:13:08.568267 30475 solver.cpp:218] Iteration 8652 (2.13236 iter/s, 5.62756s/12 iters), loss = 0.0427126
I0428 15:13:08.568311 30475 solver.cpp:237]     Train net output #0: loss = 0.0427126 (* 1 = 0.0427126 loss)
I0428 15:13:08.568320 30475 sgd_solver.cpp:105] Iteration 8652, lr = 0.00180173
I0428 15:13:14.014714 30475 solver.cpp:218] Iteration 8664 (2.20329 iter/s, 5.4464s/12 iters), loss = 0.0158378
I0428 15:13:14.014843 30475 solver.cpp:237]     Train net output #0: loss = 0.0158378 (* 1 = 0.0158378 loss)
I0428 15:13:14.014853 30475 sgd_solver.cpp:105] Iteration 8664, lr = 0.00179745
I0428 15:13:16.272491 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8670.caffemodel
I0428 15:13:18.449604 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8670.solverstate
I0428 15:13:20.143926 30475 solver.cpp:330] Iteration 8670, Testing net (#0)
I0428 15:13:20.143950 30475 net.cpp:676] Ignoring source layer train-data
I0428 15:13:21.279920 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:13:25.284935 30475 solver.cpp:397]     Test net output #0: accuracy = 0.469976
I0428 15:13:25.284965 30475 solver.cpp:397]     Test net output #1: loss = 3.95616 (* 1 = 3.95616 loss)
I0428 15:13:27.386613 30475 solver.cpp:218] Iteration 8676 (0.897412 iter/s, 13.3718s/12 iters), loss = 0.0249012
I0428 15:13:27.386660 30475 solver.cpp:237]     Train net output #0: loss = 0.0249012 (* 1 = 0.0249012 loss)
I0428 15:13:27.386668 30475 sgd_solver.cpp:105] Iteration 8676, lr = 0.00179318
I0428 15:13:33.072515 30475 solver.cpp:218] Iteration 8688 (2.1105 iter/s, 5.68585s/12 iters), loss = 0.0205128
I0428 15:13:33.072557 30475 solver.cpp:237]     Train net output #0: loss = 0.0205128 (* 1 = 0.0205128 loss)
I0428 15:13:33.072566 30475 sgd_solver.cpp:105] Iteration 8688, lr = 0.00178893
I0428 15:13:37.932152 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:13:38.736032 30475 solver.cpp:218] Iteration 8700 (2.11884 iter/s, 5.66346s/12 iters), loss = 0.0451227
I0428 15:13:38.736076 30475 solver.cpp:237]     Train net output #0: loss = 0.0451227 (* 1 = 0.0451227 loss)
I0428 15:13:38.736085 30475 sgd_solver.cpp:105] Iteration 8700, lr = 0.00178468
I0428 15:13:44.397718 30475 solver.cpp:218] Iteration 8712 (2.11953 iter/s, 5.66163s/12 iters), loss = 0.00837603
I0428 15:13:44.397850 30475 solver.cpp:237]     Train net output #0: loss = 0.00837606 (* 1 = 0.00837606 loss)
I0428 15:13:44.397859 30475 sgd_solver.cpp:105] Iteration 8712, lr = 0.00178044
I0428 15:13:50.023459 30475 solver.cpp:218] Iteration 8724 (2.13311 iter/s, 5.6256s/12 iters), loss = 0.0108612
I0428 15:13:50.023512 30475 solver.cpp:237]     Train net output #0: loss = 0.0108613 (* 1 = 0.0108613 loss)
I0428 15:13:50.023520 30475 sgd_solver.cpp:105] Iteration 8724, lr = 0.00177621
I0428 15:13:55.659076 30475 solver.cpp:218] Iteration 8736 (2.12934 iter/s, 5.63556s/12 iters), loss = 0.0105029
I0428 15:13:55.659119 30475 solver.cpp:237]     Train net output #0: loss = 0.010503 (* 1 = 0.010503 loss)
I0428 15:13:55.659127 30475 sgd_solver.cpp:105] Iteration 8736, lr = 0.001772
I0428 15:14:01.308475 30475 solver.cpp:218] Iteration 8748 (2.12414 iter/s, 5.64935s/12 iters), loss = 0.0232757
I0428 15:14:01.308522 30475 solver.cpp:237]     Train net output #0: loss = 0.0232758 (* 1 = 0.0232758 loss)
I0428 15:14:01.308531 30475 sgd_solver.cpp:105] Iteration 8748, lr = 0.00176779
I0428 15:14:06.958277 30475 solver.cpp:218] Iteration 8760 (2.12399 iter/s, 5.64975s/12 iters), loss = 0.00655295
I0428 15:14:06.958326 30475 solver.cpp:237]     Train net output #0: loss = 0.00655299 (* 1 = 0.00655299 loss)
I0428 15:14:06.958336 30475 sgd_solver.cpp:105] Iteration 8760, lr = 0.00176359
I0428 15:14:12.058688 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8772.caffemodel
I0428 15:14:14.260151 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8772.solverstate
I0428 15:14:15.959663 30475 solver.cpp:330] Iteration 8772, Testing net (#0)
I0428 15:14:15.959743 30475 net.cpp:676] Ignoring source layer train-data
I0428 15:14:16.985458 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:14:21.041947 30475 solver.cpp:397]     Test net output #0: accuracy = 0.46875
I0428 15:14:21.041990 30475 solver.cpp:397]     Test net output #1: loss = 3.89889 (* 1 = 3.89889 loss)
I0428 15:14:21.200250 30475 solver.cpp:218] Iteration 8772 (0.842582 iter/s, 14.2419s/12 iters), loss = 0.0101373
I0428 15:14:21.200300 30475 solver.cpp:237]     Train net output #0: loss = 0.0101374 (* 1 = 0.0101374 loss)
I0428 15:14:21.200310 30475 sgd_solver.cpp:105] Iteration 8772, lr = 0.00175941
I0428 15:14:25.902138 30475 solver.cpp:218] Iteration 8784 (2.5522 iter/s, 4.70183s/12 iters), loss = 0.0343455
I0428 15:14:25.902181 30475 solver.cpp:237]     Train net output #0: loss = 0.0343456 (* 1 = 0.0343456 loss)
I0428 15:14:25.902189 30475 sgd_solver.cpp:105] Iteration 8784, lr = 0.00175523
I0428 15:14:31.554109 30475 solver.cpp:218] Iteration 8796 (2.12317 iter/s, 5.65192s/12 iters), loss = 0.0266273
I0428 15:14:31.554157 30475 solver.cpp:237]     Train net output #0: loss = 0.0266273 (* 1 = 0.0266273 loss)
I0428 15:14:31.554165 30475 sgd_solver.cpp:105] Iteration 8796, lr = 0.00175106
I0428 15:14:33.164880 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:14:37.211493 30475 solver.cpp:218] Iteration 8808 (2.12114 iter/s, 5.65733s/12 iters), loss = 0.00869286
I0428 15:14:37.211539 30475 solver.cpp:237]     Train net output #0: loss = 0.00869289 (* 1 = 0.00869289 loss)
I0428 15:14:37.211547 30475 sgd_solver.cpp:105] Iteration 8808, lr = 0.0017469
I0428 15:14:42.837091 30475 solver.cpp:218] Iteration 8820 (2.13313 iter/s, 5.62555s/12 iters), loss = 0.00680286
I0428 15:14:42.837134 30475 solver.cpp:237]     Train net output #0: loss = 0.00680289 (* 1 = 0.00680289 loss)
I0428 15:14:42.837142 30475 sgd_solver.cpp:105] Iteration 8820, lr = 0.00174276
I0428 15:14:48.487231 30475 solver.cpp:218] Iteration 8832 (2.12386 iter/s, 5.65009s/12 iters), loss = 0.0178454
I0428 15:14:48.487360 30475 solver.cpp:237]     Train net output #0: loss = 0.0178455 (* 1 = 0.0178455 loss)
I0428 15:14:48.487370 30475 sgd_solver.cpp:105] Iteration 8832, lr = 0.00173862
I0428 15:14:54.129060 30475 solver.cpp:218] Iteration 8844 (2.12702 iter/s, 5.64169s/12 iters), loss = 0.00364047
I0428 15:14:54.129107 30475 solver.cpp:237]     Train net output #0: loss = 0.0036405 (* 1 = 0.0036405 loss)
I0428 15:14:54.129117 30475 sgd_solver.cpp:105] Iteration 8844, lr = 0.00173449
I0428 15:14:59.805737 30475 solver.cpp:218] Iteration 8856 (2.11393 iter/s, 5.67663s/12 iters), loss = 0.0145422
I0428 15:14:59.805775 30475 solver.cpp:237]     Train net output #0: loss = 0.0145422 (* 1 = 0.0145422 loss)
I0428 15:14:59.805783 30475 sgd_solver.cpp:105] Iteration 8856, lr = 0.00173037
I0428 15:15:05.525506 30475 solver.cpp:218] Iteration 8868 (2.098 iter/s, 5.71973s/12 iters), loss = 0.00573173
I0428 15:15:05.525545 30475 solver.cpp:237]     Train net output #0: loss = 0.00573177 (* 1 = 0.00573177 loss)
I0428 15:15:05.525554 30475 sgd_solver.cpp:105] Iteration 8868, lr = 0.00172626
I0428 15:15:07.704933 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8874.caffemodel
I0428 15:15:09.892808 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8874.solverstate
I0428 15:15:11.586580 30475 solver.cpp:330] Iteration 8874, Testing net (#0)
I0428 15:15:11.586627 30475 net.cpp:676] Ignoring source layer train-data
I0428 15:15:12.637390 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:15:16.746717 30475 solver.cpp:397]     Test net output #0: accuracy = 0.479779
I0428 15:15:16.746742 30475 solver.cpp:397]     Test net output #1: loss = 4.01822 (* 1 = 4.01822 loss)
I0428 15:15:18.859077 30475 solver.cpp:218] Iteration 8880 (0.899986 iter/s, 13.3335s/12 iters), loss = 0.0150608
I0428 15:15:18.859210 30475 solver.cpp:237]     Train net output #0: loss = 0.0150609 (* 1 = 0.0150609 loss)
I0428 15:15:18.859220 30475 sgd_solver.cpp:105] Iteration 8880, lr = 0.00172217
I0428 15:15:24.497098 30475 solver.cpp:218] Iteration 8892 (2.12846 iter/s, 5.63788s/12 iters), loss = 0.030776
I0428 15:15:24.497141 30475 solver.cpp:237]     Train net output #0: loss = 0.0307761 (* 1 = 0.0307761 loss)
I0428 15:15:24.497150 30475 sgd_solver.cpp:105] Iteration 8892, lr = 0.00171808
I0428 15:15:28.517910 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:15:30.167030 30475 solver.cpp:218] Iteration 8904 (2.11645 iter/s, 5.66988s/12 iters), loss = 0.00323499
I0428 15:15:30.167076 30475 solver.cpp:237]     Train net output #0: loss = 0.00323503 (* 1 = 0.00323503 loss)
I0428 15:15:30.167084 30475 sgd_solver.cpp:105] Iteration 8904, lr = 0.001714
I0428 15:15:35.854184 30475 solver.cpp:218] Iteration 8916 (2.11004 iter/s, 5.68709s/12 iters), loss = 0.0082031
I0428 15:15:35.854228 30475 solver.cpp:237]     Train net output #0: loss = 0.00820314 (* 1 = 0.00820314 loss)
I0428 15:15:35.854236 30475 sgd_solver.cpp:105] Iteration 8916, lr = 0.00170993
I0428 15:15:41.507318 30475 solver.cpp:218] Iteration 8928 (2.12274 iter/s, 5.65308s/12 iters), loss = 0.00379644
I0428 15:15:41.507359 30475 solver.cpp:237]     Train net output #0: loss = 0.00379648 (* 1 = 0.00379648 loss)
I0428 15:15:41.507367 30475 sgd_solver.cpp:105] Iteration 8928, lr = 0.00170587
I0428 15:15:47.171244 30475 solver.cpp:218] Iteration 8940 (2.11869 iter/s, 5.66387s/12 iters), loss = 0.00342892
I0428 15:15:47.171288 30475 solver.cpp:237]     Train net output #0: loss = 0.00342897 (* 1 = 0.00342897 loss)
I0428 15:15:47.171298 30475 sgd_solver.cpp:105] Iteration 8940, lr = 0.00170182
I0428 15:15:52.835775 30475 solver.cpp:218] Iteration 8952 (2.11847 iter/s, 5.66447s/12 iters), loss = 0.0160739
I0428 15:15:52.835935 30475 solver.cpp:237]     Train net output #0: loss = 0.016074 (* 1 = 0.016074 loss)
I0428 15:15:52.835944 30475 sgd_solver.cpp:105] Iteration 8952, lr = 0.00169778
I0428 15:15:58.480538 30475 solver.cpp:218] Iteration 8964 (2.12593 iter/s, 5.6446s/12 iters), loss = 0.0189016
I0428 15:15:58.480579 30475 solver.cpp:237]     Train net output #0: loss = 0.0189016 (* 1 = 0.0189016 loss)
I0428 15:15:58.480587 30475 sgd_solver.cpp:105] Iteration 8964, lr = 0.00169375
I0428 15:16:03.553663 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8976.caffemodel
I0428 15:16:05.757525 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8976.solverstate
I0428 15:16:07.484558 30475 solver.cpp:330] Iteration 8976, Testing net (#0)
I0428 15:16:07.484577 30475 net.cpp:676] Ignoring source layer train-data
I0428 15:16:08.496107 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:16:12.611297 30475 solver.cpp:397]     Test net output #0: accuracy = 0.484069
I0428 15:16:12.611325 30475 solver.cpp:397]     Test net output #1: loss = 3.92921 (* 1 = 3.92921 loss)
I0428 15:16:12.774866 30475 solver.cpp:218] Iteration 8976 (0.839496 iter/s, 14.2943s/12 iters), loss = 0.0570285
I0428 15:16:12.774937 30475 solver.cpp:237]     Train net output #0: loss = 0.0570285 (* 1 = 0.0570285 loss)
I0428 15:16:12.774950 30475 sgd_solver.cpp:105] Iteration 8976, lr = 0.00168973
I0428 15:16:17.482668 30475 solver.cpp:218] Iteration 8988 (2.549 iter/s, 4.70772s/12 iters), loss = 0.00296076
I0428 15:16:17.482712 30475 solver.cpp:237]     Train net output #0: loss = 0.00296083 (* 1 = 0.00296083 loss)
I0428 15:16:17.482720 30475 sgd_solver.cpp:105] Iteration 8988, lr = 0.00168571
I0428 15:16:21.172571 30475 blocking_queue.cpp:49] Waiting for data
I0428 15:16:23.161938 30475 solver.cpp:218] Iteration 9000 (2.11297 iter/s, 5.67922s/12 iters), loss = 0.0152218
I0428 15:16:23.162071 30475 solver.cpp:237]     Train net output #0: loss = 0.0152219 (* 1 = 0.0152219 loss)
I0428 15:16:23.162081 30475 sgd_solver.cpp:105] Iteration 9000, lr = 0.00168171
I0428 15:16:23.930873 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:16:28.831898 30475 solver.cpp:218] Iteration 9012 (2.11647 iter/s, 5.66982s/12 iters), loss = 0.0123358
I0428 15:16:28.831948 30475 solver.cpp:237]     Train net output #0: loss = 0.0123358 (* 1 = 0.0123358 loss)
I0428 15:16:28.831955 30475 sgd_solver.cpp:105] Iteration 9012, lr = 0.00167772
I0428 15:16:34.510607 30475 solver.cpp:218] Iteration 9024 (2.11318 iter/s, 5.67865s/12 iters), loss = 0.0159438
I0428 15:16:34.510650 30475 solver.cpp:237]     Train net output #0: loss = 0.0159439 (* 1 = 0.0159439 loss)
I0428 15:16:34.510659 30475 sgd_solver.cpp:105] Iteration 9024, lr = 0.00167374
I0428 15:16:40.212812 30475 solver.cpp:218] Iteration 9036 (2.10447 iter/s, 5.70215s/12 iters), loss = 0.022087
I0428 15:16:40.212854 30475 solver.cpp:237]     Train net output #0: loss = 0.0220871 (* 1 = 0.0220871 loss)
I0428 15:16:40.212862 30475 sgd_solver.cpp:105] Iteration 9036, lr = 0.00166976
I0428 15:16:45.924330 30475 solver.cpp:218] Iteration 9048 (2.10104 iter/s, 5.71147s/12 iters), loss = 0.00684053
I0428 15:16:45.924365 30475 solver.cpp:237]     Train net output #0: loss = 0.00684059 (* 1 = 0.00684059 loss)
I0428 15:16:45.924373 30475 sgd_solver.cpp:105] Iteration 9048, lr = 0.0016658
I0428 15:16:51.564241 30475 solver.cpp:218] Iteration 9060 (2.12771 iter/s, 5.63986s/12 iters), loss = 0.0427679
I0428 15:16:51.564282 30475 solver.cpp:237]     Train net output #0: loss = 0.0427679 (* 1 = 0.0427679 loss)
I0428 15:16:51.564291 30475 sgd_solver.cpp:105] Iteration 9060, lr = 0.00166184
I0428 15:16:57.198168 30475 solver.cpp:218] Iteration 9072 (2.12997 iter/s, 5.63387s/12 iters), loss = 0.0110084
I0428 15:16:57.198335 30475 solver.cpp:237]     Train net output #0: loss = 0.0110084 (* 1 = 0.0110084 loss)
I0428 15:16:57.198345 30475 sgd_solver.cpp:105] Iteration 9072, lr = 0.0016579
I0428 15:16:59.428052 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9078.caffemodel
I0428 15:17:01.633440 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9078.solverstate
I0428 15:17:03.327894 30475 solver.cpp:330] Iteration 9078, Testing net (#0)
I0428 15:17:03.327926 30475 net.cpp:676] Ignoring source layer train-data
I0428 15:17:04.272212 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:17:08.241238 30475 solver.cpp:397]     Test net output #0: accuracy = 0.465074
I0428 15:17:08.241284 30475 solver.cpp:397]     Test net output #1: loss = 3.9468 (* 1 = 3.9468 loss)
I0428 15:17:10.366300 30475 solver.cpp:218] Iteration 9084 (0.911302 iter/s, 13.168s/12 iters), loss = 0.017992
I0428 15:17:10.366341 30475 solver.cpp:237]     Train net output #0: loss = 0.017992 (* 1 = 0.017992 loss)
I0428 15:17:10.366351 30475 sgd_solver.cpp:105] Iteration 9084, lr = 0.00165396
I0428 15:17:15.997167 30475 solver.cpp:218] Iteration 9096 (2.13113 iter/s, 5.63081s/12 iters), loss = 0.0107991
I0428 15:17:15.997207 30475 solver.cpp:237]     Train net output #0: loss = 0.0107992 (* 1 = 0.0107992 loss)
I0428 15:17:15.997215 30475 sgd_solver.cpp:105] Iteration 9096, lr = 0.00165003
I0428 15:17:19.323966 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:17:21.653213 30475 solver.cpp:218] Iteration 9108 (2.12164 iter/s, 5.656s/12 iters), loss = 0.0151249
I0428 15:17:21.653255 30475 solver.cpp:237]     Train net output #0: loss = 0.015125 (* 1 = 0.015125 loss)
I0428 15:17:21.653264 30475 sgd_solver.cpp:105] Iteration 9108, lr = 0.00164612
I0428 15:17:27.183662 30475 solver.cpp:218] Iteration 9120 (2.16983 iter/s, 5.53039s/12 iters), loss = 0.00443154
I0428 15:17:27.183718 30475 solver.cpp:237]     Train net output #0: loss = 0.00443161 (* 1 = 0.00443161 loss)
I0428 15:17:27.183732 30475 sgd_solver.cpp:105] Iteration 9120, lr = 0.00164221
I0428 15:17:32.828352 30475 solver.cpp:218] Iteration 9132 (2.12592 iter/s, 5.64463s/12 iters), loss = 0.0272477
I0428 15:17:32.828475 30475 solver.cpp:237]     Train net output #0: loss = 0.0272478 (* 1 = 0.0272478 loss)
I0428 15:17:32.828485 30475 sgd_solver.cpp:105] Iteration 9132, lr = 0.00163831
I0428 15:17:38.273938 30475 solver.cpp:218] Iteration 9144 (2.20367 iter/s, 5.44545s/12 iters), loss = 0.00896501
I0428 15:17:38.273983 30475 solver.cpp:237]     Train net output #0: loss = 0.00896507 (* 1 = 0.00896507 loss)
I0428 15:17:38.273993 30475 sgd_solver.cpp:105] Iteration 9144, lr = 0.00163442
I0428 15:17:43.800798 30475 solver.cpp:218] Iteration 9156 (2.17124 iter/s, 5.5268s/12 iters), loss = 0.0129928
I0428 15:17:43.800856 30475 solver.cpp:237]     Train net output #0: loss = 0.0129929 (* 1 = 0.0129929 loss)
I0428 15:17:43.800868 30475 sgd_solver.cpp:105] Iteration 9156, lr = 0.00163054
I0428 15:17:49.434401 30475 solver.cpp:218] Iteration 9168 (2.1301 iter/s, 5.63354s/12 iters), loss = 0.0162314
I0428 15:17:49.434443 30475 solver.cpp:237]     Train net output #0: loss = 0.0162314 (* 1 = 0.0162314 loss)
I0428 15:17:49.434453 30475 sgd_solver.cpp:105] Iteration 9168, lr = 0.00162667
I0428 15:17:54.505937 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9180.caffemodel
I0428 15:17:56.970777 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9180.solverstate
I0428 15:17:58.866622 30475 solver.cpp:330] Iteration 9180, Testing net (#0)
I0428 15:17:58.866645 30475 net.cpp:676] Ignoring source layer train-data
I0428 15:17:59.779407 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:18:04.094678 30475 solver.cpp:397]     Test net output #0: accuracy = 0.473039
I0428 15:18:04.094864 30475 solver.cpp:397]     Test net output #1: loss = 4.07678 (* 1 = 4.07678 loss)
I0428 15:18:04.246915 30475 solver.cpp:218] Iteration 9180 (0.810128 iter/s, 14.8125s/12 iters), loss = 0.0160394
I0428 15:18:04.246965 30475 solver.cpp:237]     Train net output #0: loss = 0.0160395 (* 1 = 0.0160395 loss)
I0428 15:18:04.246978 30475 sgd_solver.cpp:105] Iteration 9180, lr = 0.00162281
I0428 15:18:08.979848 30475 solver.cpp:218] Iteration 9192 (2.53545 iter/s, 4.73288s/12 iters), loss = 0.060436
I0428 15:18:08.979890 30475 solver.cpp:237]     Train net output #0: loss = 0.0604361 (* 1 = 0.0604361 loss)
I0428 15:18:08.979898 30475 sgd_solver.cpp:105] Iteration 9192, lr = 0.00161895
I0428 15:18:14.533967 30475 solver.cpp:218] Iteration 9204 (2.16058 iter/s, 5.55406s/12 iters), loss = 0.0227618
I0428 15:18:14.534013 30475 solver.cpp:237]     Train net output #0: loss = 0.0227619 (* 1 = 0.0227619 loss)
I0428 15:18:14.534021 30475 sgd_solver.cpp:105] Iteration 9204, lr = 0.00161511
I0428 15:18:14.600838 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:18:20.192965 30475 solver.cpp:218] Iteration 9216 (2.12054 iter/s, 5.65894s/12 iters), loss = 0.0178884
I0428 15:18:20.193006 30475 solver.cpp:237]     Train net output #0: loss = 0.0178885 (* 1 = 0.0178885 loss)
I0428 15:18:20.193014 30475 sgd_solver.cpp:105] Iteration 9216, lr = 0.00161128
I0428 15:18:25.821192 30475 solver.cpp:218] Iteration 9228 (2.13213 iter/s, 5.62818s/12 iters), loss = 0.00216544
I0428 15:18:25.821234 30475 solver.cpp:237]     Train net output #0: loss = 0.0021655 (* 1 = 0.0021655 loss)
I0428 15:18:25.821244 30475 sgd_solver.cpp:105] Iteration 9228, lr = 0.00160745
I0428 15:18:31.450840 30475 solver.cpp:218] Iteration 9240 (2.13159 iter/s, 5.6296s/12 iters), loss = 0.00807671
I0428 15:18:31.450882 30475 solver.cpp:237]     Train net output #0: loss = 0.00807676 (* 1 = 0.00807676 loss)
I0428 15:18:31.450891 30475 sgd_solver.cpp:105] Iteration 9240, lr = 0.00160363
I0428 15:18:37.161108 30475 solver.cpp:218] Iteration 9252 (2.1015 iter/s, 5.71022s/12 iters), loss = 0.00177039
I0428 15:18:37.161237 30475 solver.cpp:237]     Train net output #0: loss = 0.00177045 (* 1 = 0.00177045 loss)
I0428 15:18:37.161247 30475 sgd_solver.cpp:105] Iteration 9252, lr = 0.00159983
I0428 15:18:42.805195 30475 solver.cpp:218] Iteration 9264 (2.12617 iter/s, 5.64395s/12 iters), loss = 0.0126956
I0428 15:18:42.805243 30475 solver.cpp:237]     Train net output #0: loss = 0.0126957 (* 1 = 0.0126957 loss)
I0428 15:18:42.805251 30475 sgd_solver.cpp:105] Iteration 9264, lr = 0.00159603
I0428 15:18:48.347592 30475 solver.cpp:218] Iteration 9276 (2.16515 iter/s, 5.54234s/12 iters), loss = 0.00566716
I0428 15:18:48.347635 30475 solver.cpp:237]     Train net output #0: loss = 0.00566721 (* 1 = 0.00566721 loss)
I0428 15:18:48.347642 30475 sgd_solver.cpp:105] Iteration 9276, lr = 0.00159224
I0428 15:18:50.590019 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9282.caffemodel
I0428 15:18:53.417527 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9282.solverstate
I0428 15:18:55.132362 30475 solver.cpp:330] Iteration 9282, Testing net (#0)
I0428 15:18:55.132386 30475 net.cpp:676] Ignoring source layer train-data
I0428 15:18:55.917026 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:18:59.896989 30475 solver.cpp:397]     Test net output #0: accuracy = 0.473039
I0428 15:18:59.897022 30475 solver.cpp:397]     Test net output #1: loss = 4.11722 (* 1 = 4.11722 loss)
I0428 15:19:01.997648 30475 solver.cpp:218] Iteration 9288 (0.87912 iter/s, 13.65s/12 iters), loss = 0.00722243
I0428 15:19:01.997696 30475 solver.cpp:237]     Train net output #0: loss = 0.00722248 (* 1 = 0.00722248 loss)
I0428 15:19:01.997704 30475 sgd_solver.cpp:105] Iteration 9288, lr = 0.00158846
I0428 15:19:07.515403 30475 solver.cpp:218] Iteration 9300 (2.17482 iter/s, 5.5177s/12 iters), loss = 0.0257192
I0428 15:19:07.515547 30475 solver.cpp:237]     Train net output #0: loss = 0.0257193 (* 1 = 0.0257193 loss)
I0428 15:19:07.515555 30475 sgd_solver.cpp:105] Iteration 9300, lr = 0.00158469
I0428 15:19:09.995564 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:19:13.159858 30475 solver.cpp:218] Iteration 9312 (2.12604 iter/s, 5.6443s/12 iters), loss = 0.00650267
I0428 15:19:13.159900 30475 solver.cpp:237]     Train net output #0: loss = 0.00650273 (* 1 = 0.00650273 loss)
I0428 15:19:13.159909 30475 sgd_solver.cpp:105] Iteration 9312, lr = 0.00158092
I0428 15:19:18.711395 30475 solver.cpp:218] Iteration 9324 (2.16159 iter/s, 5.55148s/12 iters), loss = 0.00500832
I0428 15:19:18.711443 30475 solver.cpp:237]     Train net output #0: loss = 0.00500837 (* 1 = 0.00500837 loss)
I0428 15:19:18.711452 30475 sgd_solver.cpp:105] Iteration 9324, lr = 0.00157717
I0428 15:19:24.344031 30475 solver.cpp:218] Iteration 9336 (2.13046 iter/s, 5.63258s/12 iters), loss = 0.00815384
I0428 15:19:24.344076 30475 solver.cpp:237]     Train net output #0: loss = 0.00815389 (* 1 = 0.00815389 loss)
I0428 15:19:24.344084 30475 sgd_solver.cpp:105] Iteration 9336, lr = 0.00157343
I0428 15:19:29.890195 30475 solver.cpp:218] Iteration 9348 (2.16368 iter/s, 5.5461s/12 iters), loss = 0.0316843
I0428 15:19:29.890244 30475 solver.cpp:237]     Train net output #0: loss = 0.0316844 (* 1 = 0.0316844 loss)
I0428 15:19:29.890256 30475 sgd_solver.cpp:105] Iteration 9348, lr = 0.00156969
I0428 15:19:35.577780 30475 solver.cpp:218] Iteration 9360 (2.10988 iter/s, 5.68753s/12 iters), loss = 0.0150155
I0428 15:19:35.577827 30475 solver.cpp:237]     Train net output #0: loss = 0.0150155 (* 1 = 0.0150155 loss)
I0428 15:19:35.577837 30475 sgd_solver.cpp:105] Iteration 9360, lr = 0.00156596
I0428 15:19:41.201815 30475 solver.cpp:218] Iteration 9372 (2.13372 iter/s, 5.62398s/12 iters), loss = 0.0357112
I0428 15:19:41.201967 30475 solver.cpp:237]     Train net output #0: loss = 0.0357113 (* 1 = 0.0357113 loss)
I0428 15:19:41.201978 30475 sgd_solver.cpp:105] Iteration 9372, lr = 0.00156225
I0428 15:19:46.263936 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9384.caffemodel
I0428 15:19:48.478716 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9384.solverstate
I0428 15:19:50.173682 30475 solver.cpp:330] Iteration 9384, Testing net (#0)
I0428 15:19:50.173710 30475 net.cpp:676] Ignoring source layer train-data
I0428 15:19:50.986826 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:19:55.313261 30475 solver.cpp:397]     Test net output #0: accuracy = 0.481005
I0428 15:19:55.313306 30475 solver.cpp:397]     Test net output #1: loss = 4.03294 (* 1 = 4.03294 loss)
I0428 15:19:55.469938 30475 solver.cpp:218] Iteration 9384 (0.841044 iter/s, 14.268s/12 iters), loss = 0.0207252
I0428 15:19:55.469985 30475 solver.cpp:237]     Train net output #0: loss = 0.0207253 (* 1 = 0.0207253 loss)
I0428 15:19:55.469996 30475 sgd_solver.cpp:105] Iteration 9384, lr = 0.00155854
I0428 15:20:00.269126 30475 solver.cpp:218] Iteration 9396 (2.50045 iter/s, 4.79913s/12 iters), loss = 0.00907914
I0428 15:20:00.269174 30475 solver.cpp:237]     Train net output #0: loss = 0.00907919 (* 1 = 0.00907919 loss)
I0428 15:20:00.269183 30475 sgd_solver.cpp:105] Iteration 9396, lr = 0.00155484
I0428 15:20:05.235296 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:20:06.012425 30475 solver.cpp:218] Iteration 9408 (2.08941 iter/s, 5.74324s/12 iters), loss = 0.0346105
I0428 15:20:06.012470 30475 solver.cpp:237]     Train net output #0: loss = 0.0346106 (* 1 = 0.0346106 loss)
I0428 15:20:06.012480 30475 sgd_solver.cpp:105] Iteration 9408, lr = 0.00155114
I0428 15:20:11.660305 30475 solver.cpp:218] Iteration 9420 (2.12471 iter/s, 5.64782s/12 iters), loss = 0.0122003
I0428 15:20:11.661898 30475 solver.cpp:237]     Train net output #0: loss = 0.0122003 (* 1 = 0.0122003 loss)
I0428 15:20:11.661907 30475 sgd_solver.cpp:105] Iteration 9420, lr = 0.00154746
I0428 15:20:17.292769 30475 solver.cpp:218] Iteration 9432 (2.13111 iter/s, 5.63087s/12 iters), loss = 0.0157943
I0428 15:20:17.292810 30475 solver.cpp:237]     Train net output #0: loss = 0.0157944 (* 1 = 0.0157944 loss)
I0428 15:20:17.292819 30475 sgd_solver.cpp:105] Iteration 9432, lr = 0.00154379
I0428 15:20:22.840580 30475 solver.cpp:218] Iteration 9444 (2.16303 iter/s, 5.54777s/12 iters), loss = 0.00487112
I0428 15:20:22.840615 30475 solver.cpp:237]     Train net output #0: loss = 0.00487117 (* 1 = 0.00487117 loss)
I0428 15:20:22.840623 30475 sgd_solver.cpp:105] Iteration 9444, lr = 0.00154012
I0428 15:20:28.459008 30475 solver.cpp:218] Iteration 9456 (2.13585 iter/s, 5.61838s/12 iters), loss = 0.00418937
I0428 15:20:28.459051 30475 solver.cpp:237]     Train net output #0: loss = 0.00418943 (* 1 = 0.00418943 loss)
I0428 15:20:28.459059 30475 sgd_solver.cpp:105] Iteration 9456, lr = 0.00153647
I0428 15:20:34.040766 30475 solver.cpp:218] Iteration 9468 (2.14988 iter/s, 5.5817s/12 iters), loss = 0.0213113
I0428 15:20:34.040807 30475 solver.cpp:237]     Train net output #0: loss = 0.0213113 (* 1 = 0.0213113 loss)
I0428 15:20:34.040814 30475 sgd_solver.cpp:105] Iteration 9468, lr = 0.00153282
I0428 15:20:39.665300 30475 solver.cpp:218] Iteration 9480 (2.13353 iter/s, 5.62448s/12 iters), loss = 0.0492479
I0428 15:20:39.665346 30475 solver.cpp:237]     Train net output #0: loss = 0.049248 (* 1 = 0.049248 loss)
I0428 15:20:39.665355 30475 sgd_solver.cpp:105] Iteration 9480, lr = 0.00152918
I0428 15:20:41.901942 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9486.caffemodel
I0428 15:20:44.089067 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9486.solverstate
I0428 15:20:45.829625 30475 solver.cpp:330] Iteration 9486, Testing net (#0)
I0428 15:20:45.829653 30475 net.cpp:676] Ignoring source layer train-data
I0428 15:20:46.575870 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:20:51.002807 30475 solver.cpp:397]     Test net output #0: accuracy = 0.471201
I0428 15:20:51.002836 30475 solver.cpp:397]     Test net output #1: loss = 4.0469 (* 1 = 4.0469 loss)
I0428 15:20:53.150985 30475 solver.cpp:218] Iteration 9492 (0.889835 iter/s, 13.4856s/12 iters), loss = 0.0361478
I0428 15:20:53.151031 30475 solver.cpp:237]     Train net output #0: loss = 0.0361478 (* 1 = 0.0361478 loss)
I0428 15:20:53.151039 30475 sgd_solver.cpp:105] Iteration 9492, lr = 0.00152555
I0428 15:20:58.823107 30475 solver.cpp:218] Iteration 9504 (2.11563 iter/s, 5.67206s/12 iters), loss = 0.0534341
I0428 15:20:58.823153 30475 solver.cpp:237]     Train net output #0: loss = 0.0534341 (* 1 = 0.0534341 loss)
I0428 15:20:58.823161 30475 sgd_solver.cpp:105] Iteration 9504, lr = 0.00152193
I0428 15:21:00.464239 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:21:04.491991 30475 solver.cpp:218] Iteration 9516 (2.11684 iter/s, 5.66883s/12 iters), loss = 0.00646015
I0428 15:21:04.492041 30475 solver.cpp:237]     Train net output #0: loss = 0.0064602 (* 1 = 0.0064602 loss)
I0428 15:21:04.492050 30475 sgd_solver.cpp:105] Iteration 9516, lr = 0.00151831
I0428 15:21:10.202828 30475 solver.cpp:218] Iteration 9528 (2.10129 iter/s, 5.71078s/12 iters), loss = 0.0067635
I0428 15:21:10.202875 30475 solver.cpp:237]     Train net output #0: loss = 0.00676355 (* 1 = 0.00676355 loss)
I0428 15:21:10.202884 30475 sgd_solver.cpp:105] Iteration 9528, lr = 0.00151471
I0428 15:21:15.838584 30475 solver.cpp:218] Iteration 9540 (2.12928 iter/s, 5.6357s/12 iters), loss = 0.00660599
I0428 15:21:15.838793 30475 solver.cpp:237]     Train net output #0: loss = 0.00660604 (* 1 = 0.00660604 loss)
I0428 15:21:15.838809 30475 sgd_solver.cpp:105] Iteration 9540, lr = 0.00151111
I0428 15:21:21.370564 30475 solver.cpp:218] Iteration 9552 (2.16929 iter/s, 5.53177s/12 iters), loss = 0.00780705
I0428 15:21:21.370615 30475 solver.cpp:237]     Train net output #0: loss = 0.0078071 (* 1 = 0.0078071 loss)
I0428 15:21:21.370623 30475 sgd_solver.cpp:105] Iteration 9552, lr = 0.00150752
I0428 15:21:26.920572 30475 solver.cpp:218] Iteration 9564 (2.16218 iter/s, 5.54995s/12 iters), loss = 0.00716911
I0428 15:21:26.920615 30475 solver.cpp:237]     Train net output #0: loss = 0.00716916 (* 1 = 0.00716916 loss)
I0428 15:21:26.920624 30475 sgd_solver.cpp:105] Iteration 9564, lr = 0.00150395
I0428 15:21:32.559386 30475 solver.cpp:218] Iteration 9576 (2.12813 iter/s, 5.63876s/12 iters), loss = 0.0325881
I0428 15:21:32.559425 30475 solver.cpp:237]     Train net output #0: loss = 0.0325882 (* 1 = 0.0325882 loss)
I0428 15:21:32.559434 30475 sgd_solver.cpp:105] Iteration 9576, lr = 0.00150037
I0428 15:21:37.647619 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9588.caffemodel
I0428 15:21:39.906970 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9588.solverstate
I0428 15:21:43.038264 30475 solver.cpp:330] Iteration 9588, Testing net (#0)
I0428 15:21:43.038287 30475 net.cpp:676] Ignoring source layer train-data
I0428 15:21:43.770135 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:21:48.258106 30475 solver.cpp:397]     Test net output #0: accuracy = 0.489583
I0428 15:21:48.258245 30475 solver.cpp:397]     Test net output #1: loss = 4.05551 (* 1 = 4.05551 loss)
I0428 15:21:48.411468 30475 solver.cpp:218] Iteration 9588 (0.757 iter/s, 15.8521s/12 iters), loss = 0.00349602
I0428 15:21:48.411520 30475 solver.cpp:237]     Train net output #0: loss = 0.00349608 (* 1 = 0.00349608 loss)
I0428 15:21:48.411530 30475 sgd_solver.cpp:105] Iteration 9588, lr = 0.00149681
I0428 15:21:53.118453 30475 solver.cpp:218] Iteration 9600 (2.54944 iter/s, 4.70692s/12 iters), loss = 0.0169578
I0428 15:21:53.118522 30475 solver.cpp:237]     Train net output #0: loss = 0.0169579 (* 1 = 0.0169579 loss)
I0428 15:21:53.118536 30475 sgd_solver.cpp:105] Iteration 9600, lr = 0.00149326
I0428 15:21:57.164484 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:21:58.816890 30475 solver.cpp:218] Iteration 9612 (2.10587 iter/s, 5.69837s/12 iters), loss = 0.00197745
I0428 15:21:58.816932 30475 solver.cpp:237]     Train net output #0: loss = 0.00197751 (* 1 = 0.00197751 loss)
I0428 15:21:58.816941 30475 sgd_solver.cpp:105] Iteration 9612, lr = 0.00148971
I0428 15:22:04.543615 30475 solver.cpp:218] Iteration 9624 (2.09546 iter/s, 5.72668s/12 iters), loss = 0.00382242
I0428 15:22:04.543656 30475 solver.cpp:237]     Train net output #0: loss = 0.00382248 (* 1 = 0.00382248 loss)
I0428 15:22:04.543665 30475 sgd_solver.cpp:105] Iteration 9624, lr = 0.00148618
I0428 15:22:10.211611 30475 solver.cpp:218] Iteration 9636 (2.11717 iter/s, 5.66795s/12 iters), loss = 0.00448429
I0428 15:22:10.211653 30475 solver.cpp:237]     Train net output #0: loss = 0.00448434 (* 1 = 0.00448434 loss)
I0428 15:22:10.211663 30475 sgd_solver.cpp:105] Iteration 9636, lr = 0.00148265
I0428 15:22:15.734422 30475 solver.cpp:218] Iteration 9648 (2.17283 iter/s, 5.52276s/12 iters), loss = 0.00405756
I0428 15:22:15.734467 30475 solver.cpp:237]     Train net output #0: loss = 0.00405761 (* 1 = 0.00405761 loss)
I0428 15:22:15.734477 30475 sgd_solver.cpp:105] Iteration 9648, lr = 0.00147913
I0428 15:22:21.372503 30475 solver.cpp:218] Iteration 9660 (2.1284 iter/s, 5.63803s/12 iters), loss = 0.0261594
I0428 15:22:21.372663 30475 solver.cpp:237]     Train net output #0: loss = 0.0261595 (* 1 = 0.0261595 loss)
I0428 15:22:21.372673 30475 sgd_solver.cpp:105] Iteration 9660, lr = 0.00147562
I0428 15:22:27.039469 30475 solver.cpp:218] Iteration 9672 (2.1176 iter/s, 5.6668s/12 iters), loss = 0.00237812
I0428 15:22:27.039511 30475 solver.cpp:237]     Train net output #0: loss = 0.00237818 (* 1 = 0.00237818 loss)
I0428 15:22:27.039520 30475 sgd_solver.cpp:105] Iteration 9672, lr = 0.00147211
I0428 15:22:32.574553 30475 solver.cpp:218] Iteration 9684 (2.16801 iter/s, 5.53504s/12 iters), loss = 0.00933572
I0428 15:22:32.574599 30475 solver.cpp:237]     Train net output #0: loss = 0.00933577 (* 1 = 0.00933577 loss)
I0428 15:22:32.574607 30475 sgd_solver.cpp:105] Iteration 9684, lr = 0.00146862
I0428 15:22:34.725754 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9690.caffemodel
I0428 15:22:36.948287 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9690.solverstate
I0428 15:22:38.648094 30475 solver.cpp:330] Iteration 9690, Testing net (#0)
I0428 15:22:38.648113 30475 net.cpp:676] Ignoring source layer train-data
I0428 15:22:39.295480 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:22:42.626706 30475 blocking_queue.cpp:49] Waiting for data
I0428 15:22:43.784693 30475 solver.cpp:397]     Test net output #0: accuracy = 0.479167
I0428 15:22:43.784724 30475 solver.cpp:397]     Test net output #1: loss = 4.06425 (* 1 = 4.06425 loss)
I0428 15:22:45.958209 30475 solver.cpp:218] Iteration 9696 (0.896619 iter/s, 13.3836s/12 iters), loss = 0.00341575
I0428 15:22:45.958253 30475 solver.cpp:237]     Train net output #0: loss = 0.00341581 (* 1 = 0.00341581 loss)
I0428 15:22:45.958261 30475 sgd_solver.cpp:105] Iteration 9696, lr = 0.00146513
I0428 15:22:51.610420 30475 solver.cpp:218] Iteration 9708 (2.12308 iter/s, 5.65215s/12 iters), loss = 0.00559162
I0428 15:22:51.610538 30475 solver.cpp:237]     Train net output #0: loss = 0.00559168 (* 1 = 0.00559168 loss)
I0428 15:22:51.610546 30475 sgd_solver.cpp:105] Iteration 9708, lr = 0.00146165
I0428 15:22:52.399221 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:22:57.265666 30475 solver.cpp:218] Iteration 9720 (2.12197 iter/s, 5.65512s/12 iters), loss = 0.00255611
I0428 15:22:57.265707 30475 solver.cpp:237]     Train net output #0: loss = 0.00255617 (* 1 = 0.00255617 loss)
I0428 15:22:57.265715 30475 sgd_solver.cpp:105] Iteration 9720, lr = 0.00145818
I0428 15:23:02.782064 30475 solver.cpp:218] Iteration 9732 (2.17535 iter/s, 5.51634s/12 iters), loss = 0.0193016
I0428 15:23:02.782114 30475 solver.cpp:237]     Train net output #0: loss = 0.0193016 (* 1 = 0.0193016 loss)
I0428 15:23:02.782124 30475 sgd_solver.cpp:105] Iteration 9732, lr = 0.00145472
I0428 15:23:08.418277 30475 solver.cpp:218] Iteration 9744 (2.12911 iter/s, 5.63616s/12 iters), loss = 0.0568895
I0428 15:23:08.418316 30475 solver.cpp:237]     Train net output #0: loss = 0.0568896 (* 1 = 0.0568896 loss)
I0428 15:23:08.418324 30475 sgd_solver.cpp:105] Iteration 9744, lr = 0.00145127
I0428 15:23:14.074795 30475 solver.cpp:218] Iteration 9756 (2.12146 iter/s, 5.65647s/12 iters), loss = 0.00356434
I0428 15:23:14.074834 30475 solver.cpp:237]     Train net output #0: loss = 0.0035644 (* 1 = 0.0035644 loss)
I0428 15:23:14.074842 30475 sgd_solver.cpp:105] Iteration 9756, lr = 0.00144782
I0428 15:23:19.766924 30475 solver.cpp:218] Iteration 9768 (2.10819 iter/s, 5.69208s/12 iters), loss = 0.0154609
I0428 15:23:19.766966 30475 solver.cpp:237]     Train net output #0: loss = 0.015461 (* 1 = 0.015461 loss)
I0428 15:23:19.766975 30475 sgd_solver.cpp:105] Iteration 9768, lr = 0.00144438
I0428 15:23:25.364631 30475 solver.cpp:218] Iteration 9780 (2.14375 iter/s, 5.59766s/12 iters), loss = 0.00156651
I0428 15:23:25.364744 30475 solver.cpp:237]     Train net output #0: loss = 0.00156656 (* 1 = 0.00156656 loss)
I0428 15:23:25.364754 30475 sgd_solver.cpp:105] Iteration 9780, lr = 0.00144095
I0428 15:23:30.392657 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9792.caffemodel
I0428 15:23:32.600481 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9792.solverstate
I0428 15:23:34.311877 30475 solver.cpp:330] Iteration 9792, Testing net (#0)
I0428 15:23:34.311895 30475 net.cpp:676] Ignoring source layer train-data
I0428 15:23:34.922569 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:23:39.451376 30475 solver.cpp:397]     Test net output #0: accuracy = 0.481618
I0428 15:23:39.451413 30475 solver.cpp:397]     Test net output #1: loss = 4.03411 (* 1 = 4.03411 loss)
I0428 15:23:39.609863 30475 solver.cpp:218] Iteration 9792 (0.842393 iter/s, 14.2451s/12 iters), loss = 0.0414802
I0428 15:23:39.609913 30475 solver.cpp:237]     Train net output #0: loss = 0.0414803 (* 1 = 0.0414803 loss)
I0428 15:23:39.609923 30475 sgd_solver.cpp:105] Iteration 9792, lr = 0.00143753
I0428 15:23:44.515518 30475 solver.cpp:218] Iteration 9804 (2.44619 iter/s, 4.9056s/12 iters), loss = 0.00257147
I0428 15:23:44.515563 30475 solver.cpp:237]     Train net output #0: loss = 0.00257152 (* 1 = 0.00257152 loss)
I0428 15:23:44.515570 30475 sgd_solver.cpp:105] Iteration 9804, lr = 0.00143412
I0428 15:23:47.899730 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:23:50.205775 30475 solver.cpp:218] Iteration 9816 (2.10889 iter/s, 5.6902s/12 iters), loss = 0.00366202
I0428 15:23:50.205817 30475 solver.cpp:237]     Train net output #0: loss = 0.00366207 (* 1 = 0.00366207 loss)
I0428 15:23:50.205826 30475 sgd_solver.cpp:105] Iteration 9816, lr = 0.00143072
I0428 15:23:55.840167 30475 solver.cpp:218] Iteration 9828 (2.1298 iter/s, 5.63434s/12 iters), loss = 0.0254088
I0428 15:23:55.840328 30475 solver.cpp:237]     Train net output #0: loss = 0.0254088 (* 1 = 0.0254088 loss)
I0428 15:23:55.840338 30475 sgd_solver.cpp:105] Iteration 9828, lr = 0.00142732
I0428 15:24:01.479274 30475 solver.cpp:218] Iteration 9840 (2.12806 iter/s, 5.63894s/12 iters), loss = 0.00899379
I0428 15:24:01.479316 30475 solver.cpp:237]     Train net output #0: loss = 0.00899385 (* 1 = 0.00899385 loss)
I0428 15:24:01.479326 30475 sgd_solver.cpp:105] Iteration 9840, lr = 0.00142393
I0428 15:24:07.116644 30475 solver.cpp:218] Iteration 9852 (2.12867 iter/s, 5.63732s/12 iters), loss = 0.011451
I0428 15:24:07.116703 30475 solver.cpp:237]     Train net output #0: loss = 0.0114511 (* 1 = 0.0114511 loss)
I0428 15:24:07.116717 30475 sgd_solver.cpp:105] Iteration 9852, lr = 0.00142055
I0428 15:24:12.746881 30475 solver.cpp:218] Iteration 9864 (2.13137 iter/s, 5.63018s/12 iters), loss = 0.019648
I0428 15:24:12.746928 30475 solver.cpp:237]     Train net output #0: loss = 0.0196481 (* 1 = 0.0196481 loss)
I0428 15:24:12.746937 30475 sgd_solver.cpp:105] Iteration 9864, lr = 0.00141718
I0428 15:24:18.378334 30475 solver.cpp:218] Iteration 9876 (2.13091 iter/s, 5.63139s/12 iters), loss = 0.00914046
I0428 15:24:18.378379 30475 solver.cpp:237]     Train net output #0: loss = 0.00914052 (* 1 = 0.00914052 loss)
I0428 15:24:18.378387 30475 sgd_solver.cpp:105] Iteration 9876, lr = 0.00141381
I0428 15:24:24.021854 30475 solver.cpp:218] Iteration 9888 (2.12635 iter/s, 5.64347s/12 iters), loss = 0.0236744
I0428 15:24:24.021901 30475 solver.cpp:237]     Train net output #0: loss = 0.0236744 (* 1 = 0.0236744 loss)
I0428 15:24:24.021909 30475 sgd_solver.cpp:105] Iteration 9888, lr = 0.00141045
I0428 15:24:26.302888 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9894.caffemodel
I0428 15:24:29.141842 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9894.solverstate
I0428 15:24:31.253898 30475 solver.cpp:330] Iteration 9894, Testing net (#0)
I0428 15:24:31.253922 30475 net.cpp:676] Ignoring source layer train-data
I0428 15:24:31.850929 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:24:36.469770 30475 solver.cpp:397]     Test net output #0: accuracy = 0.478554
I0428 15:24:36.469811 30475 solver.cpp:397]     Test net output #1: loss = 4.07754 (* 1 = 4.07754 loss)
I0428 15:24:38.654004 30475 solver.cpp:218] Iteration 9900 (0.820114 iter/s, 14.6321s/12 iters), loss = 0.00967839
I0428 15:24:38.654054 30475 solver.cpp:237]     Train net output #0: loss = 0.00967845 (* 1 = 0.00967845 loss)
I0428 15:24:38.654062 30475 sgd_solver.cpp:105] Iteration 9900, lr = 0.00140711
I0428 15:24:44.447170 30475 solver.cpp:218] Iteration 9912 (2.07143 iter/s, 5.79311s/12 iters), loss = 0.0189881
I0428 15:24:44.447209 30475 solver.cpp:237]     Train net output #0: loss = 0.0189882 (* 1 = 0.0189882 loss)
I0428 15:24:44.447217 30475 sgd_solver.cpp:105] Iteration 9912, lr = 0.00140377
I0428 15:24:44.544042 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:24:50.276861 30475 solver.cpp:218] Iteration 9924 (2.05844 iter/s, 5.82965s/12 iters), loss = 0.0507253
I0428 15:24:50.276902 30475 solver.cpp:237]     Train net output #0: loss = 0.0507254 (* 1 = 0.0507254 loss)
I0428 15:24:50.276911 30475 sgd_solver.cpp:105] Iteration 9924, lr = 0.00140043
I0428 15:24:55.768323 30475 solver.cpp:218] Iteration 9936 (2.18523 iter/s, 5.49141s/12 iters), loss = 0.00864137
I0428 15:24:55.768363 30475 solver.cpp:237]     Train net output #0: loss = 0.00864143 (* 1 = 0.00864143 loss)
I0428 15:24:55.768371 30475 sgd_solver.cpp:105] Iteration 9936, lr = 0.00139711
I0428 15:25:01.518745 30475 solver.cpp:218] Iteration 9948 (2.08682 iter/s, 5.75037s/12 iters), loss = 0.00166045
I0428 15:25:01.518901 30475 solver.cpp:237]     Train net output #0: loss = 0.00166051 (* 1 = 0.00166051 loss)
I0428 15:25:01.518911 30475 sgd_solver.cpp:105] Iteration 9948, lr = 0.00139379
I0428 15:25:07.136099 30475 solver.cpp:218] Iteration 9960 (2.1363 iter/s, 5.61718s/12 iters), loss = 0.00295413
I0428 15:25:07.136162 30475 solver.cpp:237]     Train net output #0: loss = 0.00295419 (* 1 = 0.00295419 loss)
I0428 15:25:07.136176 30475 sgd_solver.cpp:105] Iteration 9960, lr = 0.00139048
I0428 15:25:12.672212 30475 solver.cpp:218] Iteration 9972 (2.16761 iter/s, 5.53605s/12 iters), loss = 0.0167499
I0428 15:25:12.672261 30475 solver.cpp:237]     Train net output #0: loss = 0.01675 (* 1 = 0.01675 loss)
I0428 15:25:12.672269 30475 sgd_solver.cpp:105] Iteration 9972, lr = 0.00138718
I0428 15:25:18.292870 30475 solver.cpp:218] Iteration 9984 (2.135 iter/s, 5.6206s/12 iters), loss = 0.0123389
I0428 15:25:18.292917 30475 solver.cpp:237]     Train net output #0: loss = 0.0123389 (* 1 = 0.0123389 loss)
I0428 15:25:18.292927 30475 sgd_solver.cpp:105] Iteration 9984, lr = 0.00138389
I0428 15:25:23.254664 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9996.caffemodel
I0428 15:25:25.450330 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9996.solverstate
I0428 15:25:27.141841 30475 solver.cpp:330] Iteration 9996, Testing net (#0)
I0428 15:25:27.141873 30475 net.cpp:676] Ignoring source layer train-data
I0428 15:25:27.662282 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:25:32.297230 30475 solver.cpp:397]     Test net output #0: accuracy = 0.478554
I0428 15:25:32.297646 30475 solver.cpp:397]     Test net output #1: loss = 4.10074 (* 1 = 4.10074 loss)
I0428 15:25:32.456769 30475 solver.cpp:218] Iteration 9996 (0.847227 iter/s, 14.1639s/12 iters), loss = 0.0031919
I0428 15:25:32.456835 30475 solver.cpp:237]     Train net output #0: loss = 0.00319196 (* 1 = 0.00319196 loss)
I0428 15:25:32.456849 30475 sgd_solver.cpp:105] Iteration 9996, lr = 0.0013806
I0428 15:25:37.189617 30475 solver.cpp:218] Iteration 10008 (2.53551 iter/s, 4.73277s/12 iters), loss = 0.0264088
I0428 15:25:37.189658 30475 solver.cpp:237]     Train net output #0: loss = 0.0264089 (* 1 = 0.0264089 loss)
I0428 15:25:37.189667 30475 sgd_solver.cpp:105] Iteration 10008, lr = 0.00137732
I0428 15:25:39.626207 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:25:42.810014 30475 solver.cpp:218] Iteration 10020 (2.1351 iter/s, 5.62035s/12 iters), loss = 0.047623
I0428 15:25:42.810060 30475 solver.cpp:237]     Train net output #0: loss = 0.047623 (* 1 = 0.047623 loss)
I0428 15:25:42.810070 30475 sgd_solver.cpp:105] Iteration 10020, lr = 0.00137405
I0428 15:25:48.497074 30475 solver.cpp:218] Iteration 10032 (2.11007 iter/s, 5.68701s/12 iters), loss = 0.0165678
I0428 15:25:48.497114 30475 solver.cpp:237]     Train net output #0: loss = 0.0165679 (* 1 = 0.0165679 loss)
I0428 15:25:48.497123 30475 sgd_solver.cpp:105] Iteration 10032, lr = 0.00137079
I0428 15:25:54.175491 30475 solver.cpp:218] Iteration 10044 (2.11328 iter/s, 5.67837s/12 iters), loss = 0.00608733
I0428 15:25:54.175537 30475 solver.cpp:237]     Train net output #0: loss = 0.00608738 (* 1 = 0.00608738 loss)
I0428 15:25:54.175547 30475 sgd_solver.cpp:105] Iteration 10044, lr = 0.00136754
I0428 15:25:59.886561 30475 solver.cpp:218] Iteration 10056 (2.1012 iter/s, 5.71101s/12 iters), loss = 0.00732922
I0428 15:25:59.886608 30475 solver.cpp:237]     Train net output #0: loss = 0.00732928 (* 1 = 0.00732928 loss)
I0428 15:25:59.886617 30475 sgd_solver.cpp:105] Iteration 10056, lr = 0.00136429
I0428 15:26:05.562865 30475 solver.cpp:218] Iteration 10068 (2.11407 iter/s, 5.67624s/12 iters), loss = 0.00306652
I0428 15:26:05.562995 30475 solver.cpp:237]     Train net output #0: loss = 0.00306658 (* 1 = 0.00306658 loss)
I0428 15:26:05.563005 30475 sgd_solver.cpp:105] Iteration 10068, lr = 0.00136105
I0428 15:26:11.241606 30475 solver.cpp:218] Iteration 10080 (2.11319 iter/s, 5.67861s/12 iters), loss = 0.00177178
I0428 15:26:11.241647 30475 solver.cpp:237]     Train net output #0: loss = 0.00177184 (* 1 = 0.00177184 loss)
I0428 15:26:11.241654 30475 sgd_solver.cpp:105] Iteration 10080, lr = 0.00135782
I0428 15:26:16.916898 30475 solver.cpp:218] Iteration 10092 (2.11445 iter/s, 5.67523s/12 iters), loss = 0.00956909
I0428 15:26:16.916958 30475 solver.cpp:237]     Train net output #0: loss = 0.00956915 (* 1 = 0.00956915 loss)
I0428 15:26:16.916970 30475 sgd_solver.cpp:105] Iteration 10092, lr = 0.0013546
I0428 15:26:19.191356 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_10098.caffemodel
I0428 15:26:21.408470 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_10098.solverstate
I0428 15:26:23.100658 30475 solver.cpp:330] Iteration 10098, Testing net (#0)
I0428 15:26:23.100677 30475 net.cpp:676] Ignoring source layer train-data
I0428 15:26:23.588001 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:26:28.238670 30475 solver.cpp:397]     Test net output #0: accuracy = 0.484681
I0428 15:26:28.238711 30475 solver.cpp:397]     Test net output #1: loss = 4.06854 (* 1 = 4.06854 loss)
I0428 15:26:30.387792 30475 solver.cpp:218] Iteration 10104 (0.890813 iter/s, 13.4708s/12 iters), loss = 0.00644229
I0428 15:26:30.387832 30475 solver.cpp:237]     Train net output #0: loss = 0.00644235 (* 1 = 0.00644235 loss)
I0428 15:26:30.387840 30475 sgd_solver.cpp:105] Iteration 10104, lr = 0.00135138
I0428 15:26:35.209100 30513 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:26:35.950158 30475 solver.cpp:218] Iteration 10116 (2.15738 iter/s, 5.56231s/12 iters), loss = 0.0061107
I0428 15:26:35.950292 30475 solver.cpp:237]     Train net output #0: loss = 0.00611076 (* 1 = 0.00611076 loss)
I0428 15:26:35.950306 30475 sgd_solver.cpp:105] Iteration 10116, lr = 0.00134817
I0428 15:26:41.608731 30475 solver.cpp:218] Iteration 10128 (2.12073 iter/s, 5.65843s/12 iters), loss = 0.00231531
I0428 15:26:41.608772 30475 solver.cpp:237]     Train net output #0: loss = 0.00231537 (* 1 = 0.00231537 loss)
I0428 15:26:41.608780 30475 sgd_solver.cpp:105] Iteration 10128, lr = 0.00134497
I0428 15:26:47.228000 30475 solver.cpp:218] Iteration 10140 (2.13553 iter/s, 5.61922s/12 iters), loss = 0.00267285
I0428 15:26:47.228046 30475 solver.cpp:237]     Train net output #0: loss = 0.00267291 (* 1 = 0.00267291 loss)
I0428 15:26:47.228055 30475 sgd_solver.cpp:105] Iteration 10140, lr = 0.00134178
I0428 15:26:52.869436 30475 solver.cpp:218] Iteration 10152 (2.12714 iter/s, 5.64138s/12 iters), loss = 0.0240031
I0428 15:26:52.869480 30475 solver.cpp:237]     Train net output #0: loss = 0.0240031 (* 1 = 0.0240031 loss)
I0428 15:26:52.869488 30475 sgd_solver.cpp:105] Iteration 10152, lr = 0.00133859
I0428 15:26:58.509790 30475 solver.cpp:218] Iteration 10164 (2.12755 iter/s, 5.6403s/12 iters), loss = 0.00298071
I0428 15:26:58.509848 30475 solver.cpp:237]     Train net output #0: loss = 0.00298077 (* 1 = 0.00298077 loss)
I0428 15:26:58.509861 30475 sgd_solver.cpp:105] Iteration 10164, lr = 0.00133541
I0428 15:27:04.069577 30475 solver.cpp:218] Iteration 10176 (2.15838 iter/s, 5.55973s/12 iters), loss = 0.0330842
I0428 15:27:04.069617 30475 solver.cpp:237]     Train net output #0: loss = 0.0330842 (* 1 = 0.0330842 loss)
I0428 15:27:04.069626 30475 sgd_solver.cpp:105] Iteration 10176, lr = 0.00133224
I0428 15:27:09.691992 30475 solver.cpp:218] Iteration 10188 (2.13433 iter/s, 5.62237s/12 iters), loss = 0.00165228
I0428 15:27:09.692139 30475 solver.cpp:237]     Train net output #0: loss = 0.00165234 (* 1 = 0.00165234 loss)
I0428 15:27:09.692148 30475 sgd_solver.cpp:105] Iteration 10188, lr = 0.00132908
I0428 15:27:14.654533 30475 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_10200.caffemodel
I0428 15:27:16.913686 30475 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_10200.solverstate
I0428 15:27:19.084185 30475 solver.cpp:310] Iteration 10200, loss = 0.0172762
I0428 15:27:19.084219 30475 solver.cpp:330] Iteration 10200, Testing net (#0)
I0428 15:27:19.084225 30475 net.cpp:676] Ignoring source layer train-data
I0428 15:27:19.525601 30524 data_layer.cpp:73] Restarting data prefetching from start.
I0428 15:27:24.474020 30475 solver.cpp:397]     Test net output #0: accuracy = 0.483456
I0428 15:27:24.474059 30475 solver.cpp:397]     Test net output #1: loss = 4.0652 (* 1 = 4.0652 loss)
I0428 15:27:24.474066 30475 solver.cpp:315] Optimization Done.
I0428 15:27:24.474071 30475 caffe.cpp:259] Optimization Done.