DIGITS-CNN/cars/architecture-investigations/best/caffe_output.log

I0428 17:05:02.593752  8468 upgrade_proto.cpp:1082] Attempting to upgrade input file specified using deprecated 'solver_type' field (enum)': /mnt/bigdisk/DIGITS-MAN-3/digits/jobs/20210428-170501-06d1/solver.prototxt
I0428 17:05:02.593987  8468 upgrade_proto.cpp:1089] Successfully upgraded file specified using deprecated 'solver_type' field (enum) to 'type' field (string).
W0428 17:05:02.593997  8468 upgrade_proto.cpp:1091] Note that future Caffe releases will only support 'type' field (string) for a solver's type.
I0428 17:05:02.594100  8468 caffe.cpp:218] Using GPUs 3
I0428 17:05:02.623229  8468 caffe.cpp:223] GPU 3: GeForce GTX 1080 Ti
I0428 17:05:03.071666  8468 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: 3
net: "train_val.prototxt"
train_state {
level: 0
stage: ""
}
type: "SGD"
I0428 17:05:03.072530  8468 solver.cpp:87] Creating training net from net file: train_val.prototxt
I0428 17:05:03.073154  8468 net.cpp:294] The NetState phase (0) differed from the phase (1) specified by a rule in layer val-data
I0428 17:05:03.073172  8468 net.cpp:294] The NetState phase (0) differed from the phase (1) specified by a rule in layer accuracy
I0428 17:05:03.073321  8468 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-MAN-3/digits/jobs/20210421-230320-902c/mean.binaryproto"
}
data_param {
source: "/mnt/bigdisk/DIGITS-MAN-3/digits/jobs/20210421-230320-902c/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: 7
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: 7
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: 5
weight_filler {
type: "gaussian"
std: 0.01
}
bias_filler {
type: "constant"
value: 0
}
}
}
layer {
name: "relu3"
type: "ReLU"
bottom: "conv3"
top: "conv3"
}
layer {
name: "conv4"
type: "Convolution"
bottom: "conv3"
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: 7
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: 7
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: 1024
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: 1024
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 17:05:03.073421  8468 layer_factory.hpp:77] Creating layer train-data
I0428 17:05:03.078701  8468 db_lmdb.cpp:35] Opened lmdb /mnt/bigdisk/DIGITS-MAN-3/digits/jobs/20210421-230320-902c/train_db
I0428 17:05:03.078907  8468 net.cpp:84] Creating Layer train-data
I0428 17:05:03.078924  8468 net.cpp:380] train-data -> data
I0428 17:05:03.078950  8468 net.cpp:380] train-data -> label
I0428 17:05:03.078964  8468 data_transformer.cpp:25] Loading mean file from: /mnt/bigdisk/DIGITS-MAN-3/digits/jobs/20210421-230320-902c/mean.binaryproto
I0428 17:05:03.086239  8468 data_layer.cpp:45] output data size: 128,3,227,227
I0428 17:05:03.269457  8468 net.cpp:122] Setting up train-data
I0428 17:05:03.269486  8468 net.cpp:129] Top shape: 128 3 227 227 (19787136)
I0428 17:05:03.269493  8468 net.cpp:129] Top shape: 128 (128)
I0428 17:05:03.269497  8468 net.cpp:137] Memory required for data: 79149056
I0428 17:05:03.269510  8468 layer_factory.hpp:77] Creating layer conv1
I0428 17:05:03.269533  8468 net.cpp:84] Creating Layer conv1
I0428 17:05:03.269541  8468 net.cpp:406] conv1 <- data
I0428 17:05:03.269556  8468 net.cpp:380] conv1 -> conv1
I0428 17:05:04.033461  8468 net.cpp:122] Setting up conv1
I0428 17:05:04.033485  8468 net.cpp:129] Top shape: 128 96 56 56 (38535168)
I0428 17:05:04.033490  8468 net.cpp:137] Memory required for data: 233289728
I0428 17:05:04.033512  8468 layer_factory.hpp:77] Creating layer relu1
I0428 17:05:04.033524  8468 net.cpp:84] Creating Layer relu1
I0428 17:05:04.033529  8468 net.cpp:406] relu1 <- conv1
I0428 17:05:04.033535  8468 net.cpp:367] relu1 -> conv1 (in-place)
I0428 17:05:04.033846  8468 net.cpp:122] Setting up relu1
I0428 17:05:04.033856  8468 net.cpp:129] Top shape: 128 96 56 56 (38535168)
I0428 17:05:04.033860  8468 net.cpp:137] Memory required for data: 387430400
I0428 17:05:04.033864  8468 layer_factory.hpp:77] Creating layer norm1
I0428 17:05:04.033874  8468 net.cpp:84] Creating Layer norm1
I0428 17:05:04.033877  8468 net.cpp:406] norm1 <- conv1
I0428 17:05:04.033905  8468 net.cpp:380] norm1 -> norm1
I0428 17:05:04.034390  8468 net.cpp:122] Setting up norm1
I0428 17:05:04.034401  8468 net.cpp:129] Top shape: 128 96 56 56 (38535168)
I0428 17:05:04.034405  8468 net.cpp:137] Memory required for data: 541571072
I0428 17:05:04.034409  8468 layer_factory.hpp:77] Creating layer pool1
I0428 17:05:04.034417  8468 net.cpp:84] Creating Layer pool1
I0428 17:05:04.034421  8468 net.cpp:406] pool1 <- norm1
I0428 17:05:04.034427  8468 net.cpp:380] pool1 -> pool1
I0428 17:05:04.034466  8468 net.cpp:122] Setting up pool1
I0428 17:05:04.034472  8468 net.cpp:129] Top shape: 128 96 28 28 (9633792)
I0428 17:05:04.034476  8468 net.cpp:137] Memory required for data: 580106240
I0428 17:05:04.034479  8468 layer_factory.hpp:77] Creating layer conv2
I0428 17:05:04.034489  8468 net.cpp:84] Creating Layer conv2
I0428 17:05:04.034493  8468 net.cpp:406] conv2 <- pool1
I0428 17:05:04.034498  8468 net.cpp:380] conv2 -> conv2
I0428 17:05:04.044618  8468 net.cpp:122] Setting up conv2
I0428 17:05:04.044638  8468 net.cpp:129] Top shape: 128 256 26 26 (22151168)
I0428 17:05:04.044642  8468 net.cpp:137] Memory required for data: 668710912
I0428 17:05:04.044653  8468 layer_factory.hpp:77] Creating layer relu2
I0428 17:05:04.044661  8468 net.cpp:84] Creating Layer relu2
I0428 17:05:04.044667  8468 net.cpp:406] relu2 <- conv2
I0428 17:05:04.044672  8468 net.cpp:367] relu2 -> conv2 (in-place)
I0428 17:05:04.045120  8468 net.cpp:122] Setting up relu2
I0428 17:05:04.045131  8468 net.cpp:129] Top shape: 128 256 26 26 (22151168)
I0428 17:05:04.045135  8468 net.cpp:137] Memory required for data: 757315584
I0428 17:05:04.045138  8468 layer_factory.hpp:77] Creating layer norm2
I0428 17:05:04.045146  8468 net.cpp:84] Creating Layer norm2
I0428 17:05:04.045150  8468 net.cpp:406] norm2 <- conv2
I0428 17:05:04.045156  8468 net.cpp:380] norm2 -> norm2
I0428 17:05:04.045470  8468 net.cpp:122] Setting up norm2
I0428 17:05:04.045480  8468 net.cpp:129] Top shape: 128 256 26 26 (22151168)
I0428 17:05:04.045483  8468 net.cpp:137] Memory required for data: 845920256
I0428 17:05:04.045487  8468 layer_factory.hpp:77] Creating layer pool2
I0428 17:05:04.045495  8468 net.cpp:84] Creating Layer pool2
I0428 17:05:04.045500  8468 net.cpp:406] pool2 <- norm2
I0428 17:05:04.045504  8468 net.cpp:380] pool2 -> pool2
I0428 17:05:04.045534  8468 net.cpp:122] Setting up pool2
I0428 17:05:04.045539  8468 net.cpp:129] Top shape: 128 256 13 13 (5537792)
I0428 17:05:04.045543  8468 net.cpp:137] Memory required for data: 868071424
I0428 17:05:04.045547  8468 layer_factory.hpp:77] Creating layer conv3
I0428 17:05:04.045557  8468 net.cpp:84] Creating Layer conv3
I0428 17:05:04.045560  8468 net.cpp:406] conv3 <- pool2
I0428 17:05:04.045565  8468 net.cpp:380] conv3 -> conv3
I0428 17:05:04.073928  8468 net.cpp:122] Setting up conv3
I0428 17:05:04.073949  8468 net.cpp:129] Top shape: 128 384 11 11 (5947392)
I0428 17:05:04.073953  8468 net.cpp:137] Memory required for data: 891860992
I0428 17:05:04.073966  8468 layer_factory.hpp:77] Creating layer relu3
I0428 17:05:04.073973  8468 net.cpp:84] Creating Layer relu3
I0428 17:05:04.073978  8468 net.cpp:406] relu3 <- conv3
I0428 17:05:04.073984  8468 net.cpp:367] relu3 -> conv3 (in-place)
I0428 17:05:04.074456  8468 net.cpp:122] Setting up relu3
I0428 17:05:04.074466  8468 net.cpp:129] Top shape: 128 384 11 11 (5947392)
I0428 17:05:04.074470  8468 net.cpp:137] Memory required for data: 915650560
I0428 17:05:04.074473  8468 layer_factory.hpp:77] Creating layer conv4
I0428 17:05:04.074486  8468 net.cpp:84] Creating Layer conv4
I0428 17:05:04.074489  8468 net.cpp:406] conv4 <- conv3
I0428 17:05:04.074496  8468 net.cpp:380] conv4 -> conv4
I0428 17:05:04.116331  8468 net.cpp:122] Setting up conv4
I0428 17:05:04.116351  8468 net.cpp:129] Top shape: 128 384 7 7 (2408448)
I0428 17:05:04.116356  8468 net.cpp:137] Memory required for data: 925284352
I0428 17:05:04.116366  8468 layer_factory.hpp:77] Creating layer relu4
I0428 17:05:04.116376  8468 net.cpp:84] Creating Layer relu4
I0428 17:05:04.116401  8468 net.cpp:406] relu4 <- conv4
I0428 17:05:04.116408  8468 net.cpp:367] relu4 -> conv4 (in-place)
I0428 17:05:04.116804  8468 net.cpp:122] Setting up relu4
I0428 17:05:04.116816  8468 net.cpp:129] Top shape: 128 384 7 7 (2408448)
I0428 17:05:04.116820  8468 net.cpp:137] Memory required for data: 934918144
I0428 17:05:04.116825  8468 layer_factory.hpp:77] Creating layer conv5
I0428 17:05:04.116837  8468 net.cpp:84] Creating Layer conv5
I0428 17:05:04.116842  8468 net.cpp:406] conv5 <- conv4
I0428 17:05:04.116848  8468 net.cpp:380] conv5 -> conv5
I0428 17:05:04.149544  8468 net.cpp:122] Setting up conv5
I0428 17:05:04.149565  8468 net.cpp:129] Top shape: 128 256 3 3 (294912)
I0428 17:05:04.149570  8468 net.cpp:137] Memory required for data: 936097792
I0428 17:05:04.149585  8468 layer_factory.hpp:77] Creating layer relu5
I0428 17:05:04.149595  8468 net.cpp:84] Creating Layer relu5
I0428 17:05:04.149600  8468 net.cpp:406] relu5 <- conv5
I0428 17:05:04.149608  8468 net.cpp:367] relu5 -> conv5 (in-place)
I0428 17:05:04.150146  8468 net.cpp:122] Setting up relu5
I0428 17:05:04.150156  8468 net.cpp:129] Top shape: 128 256 3 3 (294912)
I0428 17:05:04.150161  8468 net.cpp:137] Memory required for data: 937277440
I0428 17:05:04.150166  8468 layer_factory.hpp:77] Creating layer pool5
I0428 17:05:04.150175  8468 net.cpp:84] Creating Layer pool5
I0428 17:05:04.150179  8468 net.cpp:406] pool5 <- conv5
I0428 17:05:04.150185  8468 net.cpp:380] pool5 -> pool5
I0428 17:05:04.150225  8468 net.cpp:122] Setting up pool5
I0428 17:05:04.150233  8468 net.cpp:129] Top shape: 128 256 1 1 (32768)
I0428 17:05:04.150236  8468 net.cpp:137] Memory required for data: 937408512
I0428 17:05:04.150240  8468 layer_factory.hpp:77] Creating layer fc6
I0428 17:05:04.150251  8468 net.cpp:84] Creating Layer fc6
I0428 17:05:04.150255  8468 net.cpp:406] fc6 <- pool5
I0428 17:05:04.150264  8468 net.cpp:380] fc6 -> fc6
I0428 17:05:04.152822  8468 net.cpp:122] Setting up fc6
I0428 17:05:04.152829  8468 net.cpp:129] Top shape: 128 1024 (131072)
I0428 17:05:04.152833  8468 net.cpp:137] Memory required for data: 937932800
I0428 17:05:04.152840  8468 layer_factory.hpp:77] Creating layer relu6
I0428 17:05:04.152848  8468 net.cpp:84] Creating Layer relu6
I0428 17:05:04.152851  8468 net.cpp:406] relu6 <- fc6
I0428 17:05:04.152858  8468 net.cpp:367] relu6 -> fc6 (in-place)
I0428 17:05:04.153388  8468 net.cpp:122] Setting up relu6
I0428 17:05:04.153398  8468 net.cpp:129] Top shape: 128 1024 (131072)
I0428 17:05:04.153401  8468 net.cpp:137] Memory required for data: 938457088
I0428 17:05:04.153405  8468 layer_factory.hpp:77] Creating layer drop6
I0428 17:05:04.153412  8468 net.cpp:84] Creating Layer drop6
I0428 17:05:04.153417  8468 net.cpp:406] drop6 <- fc6
I0428 17:05:04.153424  8468 net.cpp:367] drop6 -> fc6 (in-place)
I0428 17:05:04.153450  8468 net.cpp:122] Setting up drop6
I0428 17:05:04.153458  8468 net.cpp:129] Top shape: 128 1024 (131072)
I0428 17:05:04.153462  8468 net.cpp:137] Memory required for data: 938981376
I0428 17:05:04.153466  8468 layer_factory.hpp:77] Creating layer fc7
I0428 17:05:04.153473  8468 net.cpp:84] Creating Layer fc7
I0428 17:05:04.153477  8468 net.cpp:406] fc7 <- fc6
I0428 17:05:04.153486  8468 net.cpp:380] fc7 -> fc7
I0428 17:05:04.164450  8468 net.cpp:122] Setting up fc7
I0428 17:05:04.164466  8468 net.cpp:129] Top shape: 128 1024 (131072)
I0428 17:05:04.164470  8468 net.cpp:137] Memory required for data: 939505664
I0428 17:05:04.164479  8468 layer_factory.hpp:77] Creating layer relu7
I0428 17:05:04.164508  8468 net.cpp:84] Creating Layer relu7
I0428 17:05:04.164515  8468 net.cpp:406] relu7 <- fc7
I0428 17:05:04.164521  8468 net.cpp:367] relu7 -> fc7 (in-place)
I0428 17:05:04.191749  8468 net.cpp:122] Setting up relu7
I0428 17:05:04.191767  8468 net.cpp:129] Top shape: 128 1024 (131072)
I0428 17:05:04.191771  8468 net.cpp:137] Memory required for data: 940029952
I0428 17:05:04.191776  8468 layer_factory.hpp:77] Creating layer drop7
I0428 17:05:04.191785  8468 net.cpp:84] Creating Layer drop7
I0428 17:05:04.191790  8468 net.cpp:406] drop7 <- fc7
I0428 17:05:04.191819  8468 net.cpp:367] drop7 -> fc7 (in-place)
I0428 17:05:04.191862  8468 net.cpp:122] Setting up drop7
I0428 17:05:04.191869  8468 net.cpp:129] Top shape: 128 1024 (131072)
I0428 17:05:04.191874  8468 net.cpp:137] Memory required for data: 940554240
I0428 17:05:04.191876  8468 layer_factory.hpp:77] Creating layer fc8
I0428 17:05:04.191884  8468 net.cpp:84] Creating Layer fc8
I0428 17:05:04.191888  8468 net.cpp:406] fc8 <- fc7
I0428 17:05:04.191895  8468 net.cpp:380] fc8 -> fc8
I0428 17:05:04.194640  8468 net.cpp:122] Setting up fc8
I0428 17:05:04.194651  8468 net.cpp:129] Top shape: 128 196 (25088)
I0428 17:05:04.194655  8468 net.cpp:137] Memory required for data: 940654592
I0428 17:05:04.194664  8468 layer_factory.hpp:77] Creating layer loss
I0428 17:05:04.194670  8468 net.cpp:84] Creating Layer loss
I0428 17:05:04.194674  8468 net.cpp:406] loss <- fc8
I0428 17:05:04.194679  8468 net.cpp:406] loss <- label
I0428 17:05:04.194686  8468 net.cpp:380] loss -> loss
I0428 17:05:04.194695  8468 layer_factory.hpp:77] Creating layer loss
I0428 17:05:04.195420  8468 net.cpp:122] Setting up loss
I0428 17:05:04.195430  8468 net.cpp:129] Top shape: (1)
I0428 17:05:04.195432  8468 net.cpp:132]     with loss weight 1
I0428 17:05:04.195451  8468 net.cpp:137] Memory required for data: 940654596
I0428 17:05:04.195456  8468 net.cpp:198] loss needs backward computation.
I0428 17:05:04.195462  8468 net.cpp:198] fc8 needs backward computation.
I0428 17:05:04.195466  8468 net.cpp:198] drop7 needs backward computation.
I0428 17:05:04.195470  8468 net.cpp:198] relu7 needs backward computation.
I0428 17:05:04.195473  8468 net.cpp:198] fc7 needs backward computation.
I0428 17:05:04.195477  8468 net.cpp:198] drop6 needs backward computation.
I0428 17:05:04.195480  8468 net.cpp:198] relu6 needs backward computation.
I0428 17:05:04.195484  8468 net.cpp:198] fc6 needs backward computation.
I0428 17:05:04.195488  8468 net.cpp:198] pool5 needs backward computation.
I0428 17:05:04.195492  8468 net.cpp:198] relu5 needs backward computation.
I0428 17:05:04.195497  8468 net.cpp:198] conv5 needs backward computation.
I0428 17:05:04.195499  8468 net.cpp:198] relu4 needs backward computation.
I0428 17:05:04.195503  8468 net.cpp:198] conv4 needs backward computation.
I0428 17:05:04.195508  8468 net.cpp:198] relu3 needs backward computation.
I0428 17:05:04.195510  8468 net.cpp:198] conv3 needs backward computation.
I0428 17:05:04.195514  8468 net.cpp:198] pool2 needs backward computation.
I0428 17:05:04.195518  8468 net.cpp:198] norm2 needs backward computation.
I0428 17:05:04.195521  8468 net.cpp:198] relu2 needs backward computation.
I0428 17:05:04.195525  8468 net.cpp:198] conv2 needs backward computation.
I0428 17:05:04.195529  8468 net.cpp:198] pool1 needs backward computation.
I0428 17:05:04.195533  8468 net.cpp:198] norm1 needs backward computation.
I0428 17:05:04.195536  8468 net.cpp:198] relu1 needs backward computation.
I0428 17:05:04.195540  8468 net.cpp:198] conv1 needs backward computation.
I0428 17:05:04.195544  8468 net.cpp:200] train-data does not need backward computation.
I0428 17:05:04.195549  8468 net.cpp:242] This network produces output loss
I0428 17:05:04.195564  8468 net.cpp:255] Network initialization done.
I0428 17:05:04.196044  8468 solver.cpp:172] Creating test net (#0) specified by net file: train_val.prototxt
I0428 17:05:04.196076  8468 net.cpp:294] The NetState phase (1) differed from the phase (0) specified by a rule in layer train-data
I0428 17:05:04.196228  8468 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-MAN-3/digits/jobs/20210421-230320-902c/mean.binaryproto"
}
data_param {
source: "/mnt/bigdisk/DIGITS-MAN-3/digits/jobs/20210421-230320-902c/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: 7
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: 7
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: 5
weight_filler {
type: "gaussian"
std: 0.01
}
bias_filler {
type: "constant"
value: 0
}
}
}
layer {
name: "relu3"
type: "ReLU"
bottom: "conv3"
top: "conv3"
}
layer {
name: "conv4"
type: "Convolution"
bottom: "conv3"
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: 7
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: 7
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: 1024
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: 1024
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 17:05:04.196333  8468 layer_factory.hpp:77] Creating layer val-data
I0428 17:05:04.197921  8468 db_lmdb.cpp:35] Opened lmdb /mnt/bigdisk/DIGITS-MAN-3/digits/jobs/20210421-230320-902c/val_db
I0428 17:05:04.198091  8468 net.cpp:84] Creating Layer val-data
I0428 17:05:04.198101  8468 net.cpp:380] val-data -> data
I0428 17:05:04.198110  8468 net.cpp:380] val-data -> label
I0428 17:05:04.198117  8468 data_transformer.cpp:25] Loading mean file from: /mnt/bigdisk/DIGITS-MAN-3/digits/jobs/20210421-230320-902c/mean.binaryproto
I0428 17:05:04.201850  8468 data_layer.cpp:45] output data size: 32,3,227,227
I0428 17:05:04.247722  8468 net.cpp:122] Setting up val-data
I0428 17:05:04.247745  8468 net.cpp:129] Top shape: 32 3 227 227 (4946784)
I0428 17:05:04.247750  8468 net.cpp:129] Top shape: 32 (32)
I0428 17:05:04.247754  8468 net.cpp:137] Memory required for data: 19787264
I0428 17:05:04.247761  8468 layer_factory.hpp:77] Creating layer label_val-data_1_split
I0428 17:05:04.247774  8468 net.cpp:84] Creating Layer label_val-data_1_split
I0428 17:05:04.247778  8468 net.cpp:406] label_val-data_1_split <- label
I0428 17:05:04.247786  8468 net.cpp:380] label_val-data_1_split -> label_val-data_1_split_0
I0428 17:05:04.247797  8468 net.cpp:380] label_val-data_1_split -> label_val-data_1_split_1
I0428 17:05:04.247851  8468 net.cpp:122] Setting up label_val-data_1_split
I0428 17:05:04.247857  8468 net.cpp:129] Top shape: 32 (32)
I0428 17:05:04.247860  8468 net.cpp:129] Top shape: 32 (32)
I0428 17:05:04.247864  8468 net.cpp:137] Memory required for data: 19787520
I0428 17:05:04.247867  8468 layer_factory.hpp:77] Creating layer conv1
I0428 17:05:04.247879  8468 net.cpp:84] Creating Layer conv1
I0428 17:05:04.247884  8468 net.cpp:406] conv1 <- data
I0428 17:05:04.247889  8468 net.cpp:380] conv1 -> conv1
I0428 17:05:04.254287  8468 net.cpp:122] Setting up conv1
I0428 17:05:04.254302  8468 net.cpp:129] Top shape: 32 96 56 56 (9633792)
I0428 17:05:04.254307  8468 net.cpp:137] Memory required for data: 58322688
I0428 17:05:04.254317  8468 layer_factory.hpp:77] Creating layer relu1
I0428 17:05:04.254325  8468 net.cpp:84] Creating Layer relu1
I0428 17:05:04.254329  8468 net.cpp:406] relu1 <- conv1
I0428 17:05:04.254334  8468 net.cpp:367] relu1 -> conv1 (in-place)
I0428 17:05:04.254647  8468 net.cpp:122] Setting up relu1
I0428 17:05:04.254655  8468 net.cpp:129] Top shape: 32 96 56 56 (9633792)
I0428 17:05:04.254659  8468 net.cpp:137] Memory required for data: 96857856
I0428 17:05:04.254663  8468 layer_factory.hpp:77] Creating layer norm1
I0428 17:05:04.254673  8468 net.cpp:84] Creating Layer norm1
I0428 17:05:04.254675  8468 net.cpp:406] norm1 <- conv1
I0428 17:05:04.254681  8468 net.cpp:380] norm1 -> norm1
I0428 17:05:04.255165  8468 net.cpp:122] Setting up norm1
I0428 17:05:04.255175  8468 net.cpp:129] Top shape: 32 96 56 56 (9633792)
I0428 17:05:04.255179  8468 net.cpp:137] Memory required for data: 135393024
I0428 17:05:04.255183  8468 layer_factory.hpp:77] Creating layer pool1
I0428 17:05:04.255190  8468 net.cpp:84] Creating Layer pool1
I0428 17:05:04.255194  8468 net.cpp:406] pool1 <- norm1
I0428 17:05:04.255199  8468 net.cpp:380] pool1 -> pool1
I0428 17:05:04.255232  8468 net.cpp:122] Setting up pool1
I0428 17:05:04.255239  8468 net.cpp:129] Top shape: 32 96 28 28 (2408448)
I0428 17:05:04.255241  8468 net.cpp:137] Memory required for data: 145026816
I0428 17:05:04.255245  8468 layer_factory.hpp:77] Creating layer conv2
I0428 17:05:04.255254  8468 net.cpp:84] Creating Layer conv2
I0428 17:05:04.255257  8468 net.cpp:406] conv2 <- pool1
I0428 17:05:04.255287  8468 net.cpp:380] conv2 -> conv2
I0428 17:05:04.267192  8468 net.cpp:122] Setting up conv2
I0428 17:05:04.267213  8468 net.cpp:129] Top shape: 32 256 26 26 (5537792)
I0428 17:05:04.267217  8468 net.cpp:137] Memory required for data: 167177984
I0428 17:05:04.267230  8468 layer_factory.hpp:77] Creating layer relu2
I0428 17:05:04.267239  8468 net.cpp:84] Creating Layer relu2
I0428 17:05:04.267243  8468 net.cpp:406] relu2 <- conv2
I0428 17:05:04.267249  8468 net.cpp:367] relu2 -> conv2 (in-place)
I0428 17:05:04.267784  8468 net.cpp:122] Setting up relu2
I0428 17:05:04.267794  8468 net.cpp:129] Top shape: 32 256 26 26 (5537792)
I0428 17:05:04.267797  8468 net.cpp:137] Memory required for data: 189329152
I0428 17:05:04.267802  8468 layer_factory.hpp:77] Creating layer norm2
I0428 17:05:04.267812  8468 net.cpp:84] Creating Layer norm2
I0428 17:05:04.267817  8468 net.cpp:406] norm2 <- conv2
I0428 17:05:04.267822  8468 net.cpp:380] norm2 -> norm2
I0428 17:05:04.268384  8468 net.cpp:122] Setting up norm2
I0428 17:05:04.268394  8468 net.cpp:129] Top shape: 32 256 26 26 (5537792)
I0428 17:05:04.268399  8468 net.cpp:137] Memory required for data: 211480320
I0428 17:05:04.268402  8468 layer_factory.hpp:77] Creating layer pool2
I0428 17:05:04.268409  8468 net.cpp:84] Creating Layer pool2
I0428 17:05:04.268414  8468 net.cpp:406] pool2 <- norm2
I0428 17:05:04.268420  8468 net.cpp:380] pool2 -> pool2
I0428 17:05:04.268452  8468 net.cpp:122] Setting up pool2
I0428 17:05:04.268460  8468 net.cpp:129] Top shape: 32 256 13 13 (1384448)
I0428 17:05:04.268463  8468 net.cpp:137] Memory required for data: 217018112
I0428 17:05:04.268466  8468 layer_factory.hpp:77] Creating layer conv3
I0428 17:05:04.268478  8468 net.cpp:84] Creating Layer conv3
I0428 17:05:04.268482  8468 net.cpp:406] conv3 <- pool2
I0428 17:05:04.268515  8468 net.cpp:380] conv3 -> conv3
I0428 17:05:04.297034  8468 net.cpp:122] Setting up conv3
I0428 17:05:04.297055  8468 net.cpp:129] Top shape: 32 384 11 11 (1486848)
I0428 17:05:04.297060  8468 net.cpp:137] Memory required for data: 222965504
I0428 17:05:04.297071  8468 layer_factory.hpp:77] Creating layer relu3
I0428 17:05:04.297083  8468 net.cpp:84] Creating Layer relu3
I0428 17:05:04.297088  8468 net.cpp:406] relu3 <- conv3
I0428 17:05:04.297096  8468 net.cpp:367] relu3 -> conv3 (in-place)
I0428 17:05:04.297641  8468 net.cpp:122] Setting up relu3
I0428 17:05:04.297652  8468 net.cpp:129] Top shape: 32 384 11 11 (1486848)
I0428 17:05:04.297655  8468 net.cpp:137] Memory required for data: 228912896
I0428 17:05:04.297659  8468 layer_factory.hpp:77] Creating layer conv4
I0428 17:05:04.297672  8468 net.cpp:84] Creating Layer conv4
I0428 17:05:04.297677  8468 net.cpp:406] conv4 <- conv3
I0428 17:05:04.297684  8468 net.cpp:380] conv4 -> conv4
I0428 17:05:04.338552  8468 net.cpp:122] Setting up conv4
I0428 17:05:04.338572  8468 net.cpp:129] Top shape: 32 384 7 7 (602112)
I0428 17:05:04.338577  8468 net.cpp:137] Memory required for data: 231321344
I0428 17:05:04.338587  8468 layer_factory.hpp:77] Creating layer relu4
I0428 17:05:04.338595  8468 net.cpp:84] Creating Layer relu4
I0428 17:05:04.338601  8468 net.cpp:406] relu4 <- conv4
I0428 17:05:04.338611  8468 net.cpp:367] relu4 -> conv4 (in-place)
I0428 17:05:04.339000  8468 net.cpp:122] Setting up relu4
I0428 17:05:04.339010  8468 net.cpp:129] Top shape: 32 384 7 7 (602112)
I0428 17:05:04.339015  8468 net.cpp:137] Memory required for data: 233729792
I0428 17:05:04.339020  8468 layer_factory.hpp:77] Creating layer conv5
I0428 17:05:04.339031  8468 net.cpp:84] Creating Layer conv5
I0428 17:05:04.339035  8468 net.cpp:406] conv5 <- conv4
I0428 17:05:04.339042  8468 net.cpp:380] conv5 -> conv5
I0428 17:05:04.369298  8468 net.cpp:122] Setting up conv5
I0428 17:05:04.369318  8468 net.cpp:129] Top shape: 32 256 3 3 (73728)
I0428 17:05:04.369323  8468 net.cpp:137] Memory required for data: 234024704
I0428 17:05:04.369339  8468 layer_factory.hpp:77] Creating layer relu5
I0428 17:05:04.369350  8468 net.cpp:84] Creating Layer relu5
I0428 17:05:04.369356  8468 net.cpp:406] relu5 <- conv5
I0428 17:05:04.369385  8468 net.cpp:367] relu5 -> conv5 (in-place)
I0428 17:05:04.371474  8468 net.cpp:122] Setting up relu5
I0428 17:05:04.371484  8468 net.cpp:129] Top shape: 32 256 3 3 (73728)
I0428 17:05:04.371490  8468 net.cpp:137] Memory required for data: 234319616
I0428 17:05:04.371496  8468 layer_factory.hpp:77] Creating layer pool5
I0428 17:05:04.371507  8468 net.cpp:84] Creating Layer pool5
I0428 17:05:04.371512  8468 net.cpp:406] pool5 <- conv5
I0428 17:05:04.371521  8468 net.cpp:380] pool5 -> pool5
I0428 17:05:04.371563  8468 net.cpp:122] Setting up pool5
I0428 17:05:04.371570  8468 net.cpp:129] Top shape: 32 256 1 1 (8192)
I0428 17:05:04.371574  8468 net.cpp:137] Memory required for data: 234352384
I0428 17:05:04.371579  8468 layer_factory.hpp:77] Creating layer fc6
I0428 17:05:04.371588  8468 net.cpp:84] Creating Layer fc6
I0428 17:05:04.371592  8468 net.cpp:406] fc6 <- pool5
I0428 17:05:04.371600  8468 net.cpp:380] fc6 -> fc6
I0428 17:05:04.374823  8468 net.cpp:122] Setting up fc6
I0428 17:05:04.374835  8468 net.cpp:129] Top shape: 32 1024 (32768)
I0428 17:05:04.374840  8468 net.cpp:137] Memory required for data: 234483456
I0428 17:05:04.374850  8468 layer_factory.hpp:77] Creating layer relu6
I0428 17:05:04.374857  8468 net.cpp:84] Creating Layer relu6
I0428 17:05:04.374863  8468 net.cpp:406] relu6 <- fc6
I0428 17:05:04.374869  8468 net.cpp:367] relu6 -> fc6 (in-place)
I0428 17:05:04.376003  8468 net.cpp:122] Setting up relu6
I0428 17:05:04.376014  8468 net.cpp:129] Top shape: 32 1024 (32768)
I0428 17:05:04.376019  8468 net.cpp:137] Memory required for data: 234614528
I0428 17:05:04.376024  8468 layer_factory.hpp:77] Creating layer drop6
I0428 17:05:04.376034  8468 net.cpp:84] Creating Layer drop6
I0428 17:05:04.376039  8468 net.cpp:406] drop6 <- fc6
I0428 17:05:04.376045  8468 net.cpp:367] drop6 -> fc6 (in-place)
I0428 17:05:04.376073  8468 net.cpp:122] Setting up drop6
I0428 17:05:04.376080  8468 net.cpp:129] Top shape: 32 1024 (32768)
I0428 17:05:04.376083  8468 net.cpp:137] Memory required for data: 234745600
I0428 17:05:04.376088  8468 layer_factory.hpp:77] Creating layer fc7
I0428 17:05:04.376096  8468 net.cpp:84] Creating Layer fc7
I0428 17:05:04.376101  8468 net.cpp:406] fc7 <- fc6
I0428 17:05:04.376108  8468 net.cpp:380] fc7 -> fc7
I0428 17:05:04.388618  8468 net.cpp:122] Setting up fc7
I0428 17:05:04.388635  8468 net.cpp:129] Top shape: 32 1024 (32768)
I0428 17:05:04.388639  8468 net.cpp:137] Memory required for data: 234876672
I0428 17:05:04.388648  8468 layer_factory.hpp:77] Creating layer relu7
I0428 17:05:04.388657  8468 net.cpp:84] Creating Layer relu7
I0428 17:05:04.388662  8468 net.cpp:406] relu7 <- fc7
I0428 17:05:04.388670  8468 net.cpp:367] relu7 -> fc7 (in-place)
I0428 17:05:04.389096  8468 net.cpp:122] Setting up relu7
I0428 17:05:04.389106  8468 net.cpp:129] Top shape: 32 1024 (32768)
I0428 17:05:04.389109  8468 net.cpp:137] Memory required for data: 235007744
I0428 17:05:04.389115  8468 layer_factory.hpp:77] Creating layer drop7
I0428 17:05:04.389122  8468 net.cpp:84] Creating Layer drop7
I0428 17:05:04.389127  8468 net.cpp:406] drop7 <- fc7
I0428 17:05:04.389134  8468 net.cpp:367] drop7 -> fc7 (in-place)
I0428 17:05:04.389160  8468 net.cpp:122] Setting up drop7
I0428 17:05:04.389168  8468 net.cpp:129] Top shape: 32 1024 (32768)
I0428 17:05:04.389173  8468 net.cpp:137] Memory required for data: 235138816
I0428 17:05:04.389178  8468 layer_factory.hpp:77] Creating layer fc8
I0428 17:05:04.389184  8468 net.cpp:84] Creating Layer fc8
I0428 17:05:04.389189  8468 net.cpp:406] fc8 <- fc7
I0428 17:05:04.389195  8468 net.cpp:380] fc8 -> fc8
I0428 17:05:04.391175  8468 net.cpp:122] Setting up fc8
I0428 17:05:04.391182  8468 net.cpp:129] Top shape: 32 196 (6272)
I0428 17:05:04.391186  8468 net.cpp:137] Memory required for data: 235163904
I0428 17:05:04.391193  8468 layer_factory.hpp:77] Creating layer fc8_fc8_0_split
I0428 17:05:04.391201  8468 net.cpp:84] Creating Layer fc8_fc8_0_split
I0428 17:05:04.391206  8468 net.cpp:406] fc8_fc8_0_split <- fc8
I0428 17:05:04.391211  8468 net.cpp:380] fc8_fc8_0_split -> fc8_fc8_0_split_0
I0428 17:05:04.391243  8468 net.cpp:380] fc8_fc8_0_split -> fc8_fc8_0_split_1
I0428 17:05:04.391281  8468 net.cpp:122] Setting up fc8_fc8_0_split
I0428 17:05:04.391288  8468 net.cpp:129] Top shape: 32 196 (6272)
I0428 17:05:04.391291  8468 net.cpp:129] Top shape: 32 196 (6272)
I0428 17:05:04.391295  8468 net.cpp:137] Memory required for data: 235214080
I0428 17:05:04.391299  8468 layer_factory.hpp:77] Creating layer accuracy
I0428 17:05:04.391309  8468 net.cpp:84] Creating Layer accuracy
I0428 17:05:04.391312  8468 net.cpp:406] accuracy <- fc8_fc8_0_split_0
I0428 17:05:04.391319  8468 net.cpp:406] accuracy <- label_val-data_1_split_0
I0428 17:05:04.391324  8468 net.cpp:380] accuracy -> accuracy
I0428 17:05:04.391332  8468 net.cpp:122] Setting up accuracy
I0428 17:05:04.391336  8468 net.cpp:129] Top shape: (1)
I0428 17:05:04.391340  8468 net.cpp:137] Memory required for data: 235214084
I0428 17:05:04.391345  8468 layer_factory.hpp:77] Creating layer loss
I0428 17:05:04.391352  8468 net.cpp:84] Creating Layer loss
I0428 17:05:04.391356  8468 net.cpp:406] loss <- fc8_fc8_0_split_1
I0428 17:05:04.391361  8468 net.cpp:406] loss <- label_val-data_1_split_1
I0428 17:05:04.391367  8468 net.cpp:380] loss -> loss
I0428 17:05:04.391374  8468 layer_factory.hpp:77] Creating layer loss
I0428 17:05:04.411280  8468 net.cpp:122] Setting up loss
I0428 17:05:04.411294  8468 net.cpp:129] Top shape: (1)
I0428 17:05:04.411300  8468 net.cpp:132]     with loss weight 1
I0428 17:05:04.411311  8468 net.cpp:137] Memory required for data: 235214088
I0428 17:05:04.411316  8468 net.cpp:198] loss needs backward computation.
I0428 17:05:04.411322  8468 net.cpp:200] accuracy does not need backward computation.
I0428 17:05:04.411327  8468 net.cpp:198] fc8_fc8_0_split needs backward computation.
I0428 17:05:04.411332  8468 net.cpp:198] fc8 needs backward computation.
I0428 17:05:04.411337  8468 net.cpp:198] drop7 needs backward computation.
I0428 17:05:04.411342  8468 net.cpp:198] relu7 needs backward computation.
I0428 17:05:04.411347  8468 net.cpp:198] fc7 needs backward computation.
I0428 17:05:04.411351  8468 net.cpp:198] drop6 needs backward computation.
I0428 17:05:04.411355  8468 net.cpp:198] relu6 needs backward computation.
I0428 17:05:04.411360  8468 net.cpp:198] fc6 needs backward computation.
I0428 17:05:04.411365  8468 net.cpp:198] pool5 needs backward computation.
I0428 17:05:04.411368  8468 net.cpp:198] relu5 needs backward computation.
I0428 17:05:04.411373  8468 net.cpp:198] conv5 needs backward computation.
I0428 17:05:04.411377  8468 net.cpp:198] relu4 needs backward computation.
I0428 17:05:04.411381  8468 net.cpp:198] conv4 needs backward computation.
I0428 17:05:04.411387  8468 net.cpp:198] relu3 needs backward computation.
I0428 17:05:04.411392  8468 net.cpp:198] conv3 needs backward computation.
I0428 17:05:04.411396  8468 net.cpp:198] pool2 needs backward computation.
I0428 17:05:04.411401  8468 net.cpp:198] norm2 needs backward computation.
I0428 17:05:04.411406  8468 net.cpp:198] relu2 needs backward computation.
I0428 17:05:04.411409  8468 net.cpp:198] conv2 needs backward computation.
I0428 17:05:04.411414  8468 net.cpp:198] pool1 needs backward computation.
I0428 17:05:04.411418  8468 net.cpp:198] norm1 needs backward computation.
I0428 17:05:04.411422  8468 net.cpp:198] relu1 needs backward computation.
I0428 17:05:04.411427  8468 net.cpp:198] conv1 needs backward computation.
I0428 17:05:04.411433  8468 net.cpp:200] label_val-data_1_split does not need backward computation.
I0428 17:05:04.411438  8468 net.cpp:200] val-data does not need backward computation.
I0428 17:05:04.411443  8468 net.cpp:242] This network produces output accuracy
I0428 17:05:04.411449  8468 net.cpp:242] This network produces output loss
I0428 17:05:04.411469  8468 net.cpp:255] Network initialization done.
I0428 17:05:04.411542  8468 solver.cpp:56] Solver scaffolding done.
I0428 17:05:04.412016  8468 caffe.cpp:248] Starting Optimization
I0428 17:05:04.412025  8468 solver.cpp:272] Solving
I0428 17:05:04.412029  8468 solver.cpp:273] Learning Rate Policy: exp
I0428 17:05:04.416846  8468 solver.cpp:330] Iteration 0, Testing net (#0)
I0428 17:05:04.416857  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:05:04.457778  8468 blocking_queue.cpp:49] Waiting for data
I0428 17:05:09.118793  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:05:09.166906  8468 solver.cpp:397]     Test net output #0: accuracy = 0.00245098
I0428 17:05:09.166949  8468 solver.cpp:397]     Test net output #1: loss = 5.28004 (* 1 = 5.28004 loss)
I0428 17:05:09.319537  8468 solver.cpp:218] Iteration 0 (-3.72531e-39 iter/s, 4.90724s/12 iters), loss = 5.27436
I0428 17:05:09.321087  8468 solver.cpp:237]     Train net output #0: loss = 5.27436 (* 1 = 5.27436 loss)
I0428 17:05:09.321106  8468 sgd_solver.cpp:105] Iteration 0, lr = 0.01
I0428 17:05:13.500245  8468 solver.cpp:218] Iteration 12 (2.87152 iter/s, 4.17897s/12 iters), loss = 5.27126
I0428 17:05:13.500294  8468 solver.cpp:237]     Train net output #0: loss = 5.27126 (* 1 = 5.27126 loss)
I0428 17:05:13.500304  8468 sgd_solver.cpp:105] Iteration 12, lr = 0.00997626
I0428 17:05:18.910413  8468 solver.cpp:218] Iteration 24 (2.21817 iter/s, 5.40987s/12 iters), loss = 5.27332
I0428 17:05:18.910459  8468 solver.cpp:237]     Train net output #0: loss = 5.27332 (* 1 = 5.27332 loss)
I0428 17:05:18.910468  8468 sgd_solver.cpp:105] Iteration 24, lr = 0.00995257
I0428 17:05:24.436957  8468 solver.cpp:218] Iteration 36 (2.17145 iter/s, 5.52626s/12 iters), loss = 5.29045
I0428 17:05:24.436990  8468 solver.cpp:237]     Train net output #0: loss = 5.29045 (* 1 = 5.29045 loss)
I0428 17:05:24.437000  8468 sgd_solver.cpp:105] Iteration 36, lr = 0.00992894
I0428 17:05:29.841836  8468 solver.cpp:218] Iteration 48 (2.22033 iter/s, 5.4046s/12 iters), loss = 5.2729
I0428 17:05:29.841881  8468 solver.cpp:237]     Train net output #0: loss = 5.2729 (* 1 = 5.2729 loss)
I0428 17:05:29.841892  8468 sgd_solver.cpp:105] Iteration 48, lr = 0.00990537
I0428 17:05:35.193605  8468 solver.cpp:218] Iteration 60 (2.24237 iter/s, 5.35149s/12 iters), loss = 5.28498
I0428 17:05:35.196002  8468 solver.cpp:237]     Train net output #0: loss = 5.28498 (* 1 = 5.28498 loss)
I0428 17:05:35.196014  8468 sgd_solver.cpp:105] Iteration 60, lr = 0.00988185
I0428 17:05:40.541934  8468 solver.cpp:218] Iteration 72 (2.2448 iter/s, 5.3457s/12 iters), loss = 5.27915
I0428 17:05:40.541977  8468 solver.cpp:237]     Train net output #0: loss = 5.27915 (* 1 = 5.27915 loss)
I0428 17:05:40.541986  8468 sgd_solver.cpp:105] Iteration 72, lr = 0.00985839
I0428 17:05:45.918064  8468 solver.cpp:218] Iteration 84 (2.23221 iter/s, 5.37585s/12 iters), loss = 5.2813
I0428 17:05:45.918104  8468 solver.cpp:237]     Train net output #0: loss = 5.2813 (* 1 = 5.2813 loss)
I0428 17:05:45.918114  8468 sgd_solver.cpp:105] Iteration 84, lr = 0.00983498
I0428 17:05:51.374163  8468 solver.cpp:218] Iteration 96 (2.19949 iter/s, 5.45581s/12 iters), loss = 5.28716
I0428 17:05:51.374209  8468 solver.cpp:237]     Train net output #0: loss = 5.28716 (* 1 = 5.28716 loss)
I0428 17:05:51.374219  8468 sgd_solver.cpp:105] Iteration 96, lr = 0.00981163
I0428 17:05:53.268280  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:05:53.583232  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_102.caffemodel
I0428 17:05:54.193843  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_102.solverstate
I0428 17:05:54.626422  8468 solver.cpp:330] Iteration 102, Testing net (#0)
I0428 17:05:54.626447  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:05:59.079623  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:05:59.163388  8468 solver.cpp:397]     Test net output #0: accuracy = 0.00551471
I0428 17:05:59.163426  8468 solver.cpp:397]     Test net output #1: loss = 5.28244 (* 1 = 5.28244 loss)
I0428 17:06:01.046315  8468 solver.cpp:218] Iteration 108 (1.24073 iter/s, 9.6717s/12 iters), loss = 5.27672
I0428 17:06:01.046361  8468 solver.cpp:237]     Train net output #0: loss = 5.27672 (* 1 = 5.27672 loss)
I0428 17:06:01.046371  8468 sgd_solver.cpp:105] Iteration 108, lr = 0.00978834
I0428 17:06:06.434146  8468 solver.cpp:218] Iteration 120 (2.22736 iter/s, 5.38755s/12 iters), loss = 5.26979
I0428 17:06:06.434286  8468 solver.cpp:237]     Train net output #0: loss = 5.26979 (* 1 = 5.26979 loss)
I0428 17:06:06.434296  8468 sgd_solver.cpp:105] Iteration 120, lr = 0.0097651
I0428 17:06:11.829610  8468 solver.cpp:218] Iteration 132 (2.22424 iter/s, 5.39509s/12 iters), loss = 5.26569
I0428 17:06:11.829656  8468 solver.cpp:237]     Train net output #0: loss = 5.26569 (* 1 = 5.26569 loss)
I0428 17:06:11.829668  8468 sgd_solver.cpp:105] Iteration 132, lr = 0.00974192
I0428 17:06:17.292146  8468 solver.cpp:218] Iteration 144 (2.1969 iter/s, 5.46225s/12 iters), loss = 5.27889
I0428 17:06:17.292189  8468 solver.cpp:237]     Train net output #0: loss = 5.27889 (* 1 = 5.27889 loss)
I0428 17:06:17.292199  8468 sgd_solver.cpp:105] Iteration 144, lr = 0.00971879
I0428 17:06:22.701609  8468 solver.cpp:218] Iteration 156 (2.21845 iter/s, 5.40919s/12 iters), loss = 5.28487
I0428 17:06:22.701644  8468 solver.cpp:237]     Train net output #0: loss = 5.28487 (* 1 = 5.28487 loss)
I0428 17:06:22.701655  8468 sgd_solver.cpp:105] Iteration 156, lr = 0.00969571
I0428 17:06:28.330230  8468 solver.cpp:218] Iteration 168 (2.13207 iter/s, 5.62834s/12 iters), loss = 5.27293
I0428 17:06:28.330271  8468 solver.cpp:237]     Train net output #0: loss = 5.27293 (* 1 = 5.27293 loss)
I0428 17:06:28.330281  8468 sgd_solver.cpp:105] Iteration 168, lr = 0.00967269
I0428 17:06:33.862092  8468 solver.cpp:218] Iteration 180 (2.16936 iter/s, 5.53158s/12 iters), loss = 5.28362
I0428 17:06:33.862140  8468 solver.cpp:237]     Train net output #0: loss = 5.28362 (* 1 = 5.28362 loss)
I0428 17:06:33.862151  8468 sgd_solver.cpp:105] Iteration 180, lr = 0.00964973
I0428 17:06:39.582921  8468 solver.cpp:218] Iteration 192 (2.09771 iter/s, 5.72054s/12 iters), loss = 5.26917
I0428 17:06:39.583106  8468 solver.cpp:237]     Train net output #0: loss = 5.26917 (* 1 = 5.26917 loss)
I0428 17:06:39.583117  8468 sgd_solver.cpp:105] Iteration 192, lr = 0.00962682
I0428 17:06:43.732290  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:06:44.446611  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_204.caffemodel
I0428 17:06:45.670266  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_204.solverstate
I0428 17:06:46.734509  8468 solver.cpp:330] Iteration 204, Testing net (#0)
I0428 17:06:46.734534  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:06:51.315469  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:06:51.447312  8468 solver.cpp:397]     Test net output #0: accuracy = 0.00551471
I0428 17:06:51.447345  8468 solver.cpp:397]     Test net output #1: loss = 5.28285 (* 1 = 5.28285 loss)
I0428 17:06:51.573474  8468 solver.cpp:218] Iteration 204 (1.00084 iter/s, 11.9899s/12 iters), loss = 5.26147
I0428 17:06:51.573514  8468 solver.cpp:237]     Train net output #0: loss = 5.26147 (* 1 = 5.26147 loss)
I0428 17:06:51.573523  8468 sgd_solver.cpp:105] Iteration 204, lr = 0.00960396
I0428 17:06:56.282714  8468 solver.cpp:218] Iteration 216 (2.54832 iter/s, 4.70899s/12 iters), loss = 5.27448
I0428 17:06:56.282763  8468 solver.cpp:237]     Train net output #0: loss = 5.27448 (* 1 = 5.27448 loss)
I0428 17:06:56.282775  8468 sgd_solver.cpp:105] Iteration 216, lr = 0.00958116
I0428 17:07:01.728842  8468 solver.cpp:218] Iteration 228 (2.20351 iter/s, 5.44585s/12 iters), loss = 5.27
I0428 17:07:01.728879  8468 solver.cpp:237]     Train net output #0: loss = 5.27 (* 1 = 5.27 loss)
I0428 17:07:01.728888  8468 sgd_solver.cpp:105] Iteration 228, lr = 0.00955841
I0428 17:07:07.093895  8468 solver.cpp:218] Iteration 240 (2.23681 iter/s, 5.36479s/12 iters), loss = 5.25976
I0428 17:07:07.093935  8468 solver.cpp:237]     Train net output #0: loss = 5.25976 (* 1 = 5.25976 loss)
I0428 17:07:07.093943  8468 sgd_solver.cpp:105] Iteration 240, lr = 0.00953572
I0428 17:07:12.481320  8468 solver.cpp:218] Iteration 252 (2.22752 iter/s, 5.38715s/12 iters), loss = 5.28528
I0428 17:07:12.481456  8468 solver.cpp:237]     Train net output #0: loss = 5.28528 (* 1 = 5.28528 loss)
I0428 17:07:12.481467  8468 sgd_solver.cpp:105] Iteration 252, lr = 0.00951308
I0428 17:07:17.881352  8468 solver.cpp:218] Iteration 264 (2.22236 iter/s, 5.39967s/12 iters), loss = 5.25807
I0428 17:07:17.881392  8468 solver.cpp:237]     Train net output #0: loss = 5.25807 (* 1 = 5.25807 loss)
I0428 17:07:17.881402  8468 sgd_solver.cpp:105] Iteration 264, lr = 0.00949049
I0428 17:07:23.266052  8468 solver.cpp:218] Iteration 276 (2.22865 iter/s, 5.38443s/12 iters), loss = 5.26004
I0428 17:07:23.266091  8468 solver.cpp:237]     Train net output #0: loss = 5.26004 (* 1 = 5.26004 loss)
I0428 17:07:23.266101  8468 sgd_solver.cpp:105] Iteration 276, lr = 0.00946796
I0428 17:07:28.636466  8468 solver.cpp:218] Iteration 288 (2.23458 iter/s, 5.37014s/12 iters), loss = 5.28638
I0428 17:07:28.636528  8468 solver.cpp:237]     Train net output #0: loss = 5.28638 (* 1 = 5.28638 loss)
I0428 17:07:28.636538  8468 sgd_solver.cpp:105] Iteration 288, lr = 0.00944548
I0428 17:07:34.153259  8468 solver.cpp:218] Iteration 300 (2.17529 iter/s, 5.51649s/12 iters), loss = 5.27914
I0428 17:07:34.153299  8468 solver.cpp:237]     Train net output #0: loss = 5.27914 (* 1 = 5.27914 loss)
I0428 17:07:34.153311  8468 sgd_solver.cpp:105] Iteration 300, lr = 0.00942305
I0428 17:07:35.274613  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:07:36.382230  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_306.caffemodel
I0428 17:07:37.823722  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_306.solverstate
I0428 17:07:38.437075  8468 solver.cpp:330] Iteration 306, Testing net (#0)
I0428 17:07:38.437095  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:07:42.760650  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:07:42.925802  8468 solver.cpp:397]     Test net output #0: accuracy = 0.0128676
I0428 17:07:42.925832  8468 solver.cpp:397]     Test net output #1: loss = 5.20103 (* 1 = 5.20103 loss)
I0428 17:07:44.952443  8468 solver.cpp:218] Iteration 312 (1.11125 iter/s, 10.7987s/12 iters), loss = 5.2372
I0428 17:07:44.952510  8468 solver.cpp:237]     Train net output #0: loss = 5.2372 (* 1 = 5.2372 loss)
I0428 17:07:44.952520  8468 sgd_solver.cpp:105] Iteration 312, lr = 0.00940068
I0428 17:07:50.319155  8468 solver.cpp:218] Iteration 324 (2.23612 iter/s, 5.36644s/12 iters), loss = 5.19498
I0428 17:07:50.319197  8468 solver.cpp:237]     Train net output #0: loss = 5.19498 (* 1 = 5.19498 loss)
I0428 17:07:50.319207  8468 sgd_solver.cpp:105] Iteration 324, lr = 0.00937836
I0428 17:07:55.732169  8468 solver.cpp:218] Iteration 336 (2.21699 iter/s, 5.41274s/12 iters), loss = 5.22829
I0428 17:07:55.732213  8468 solver.cpp:237]     Train net output #0: loss = 5.22829 (* 1 = 5.22829 loss)
I0428 17:07:55.732223  8468 sgd_solver.cpp:105] Iteration 336, lr = 0.0093561
I0428 17:08:01.331585  8468 solver.cpp:218] Iteration 348 (2.14319 iter/s, 5.59914s/12 iters), loss = 5.16491
I0428 17:08:01.331622  8468 solver.cpp:237]     Train net output #0: loss = 5.16491 (* 1 = 5.16491 loss)
I0428 17:08:01.331632  8468 sgd_solver.cpp:105] Iteration 348, lr = 0.00933388
I0428 17:08:06.846345  8468 solver.cpp:218] Iteration 360 (2.17609 iter/s, 5.51449s/12 iters), loss = 5.18139
I0428 17:08:06.846385  8468 solver.cpp:237]     Train net output #0: loss = 5.18139 (* 1 = 5.18139 loss)
I0428 17:08:06.846393  8468 sgd_solver.cpp:105] Iteration 360, lr = 0.00931172
I0428 17:08:12.304538  8468 solver.cpp:218] Iteration 372 (2.19865 iter/s, 5.45789s/12 iters), loss = 5.22535
I0428 17:08:12.304581  8468 solver.cpp:237]     Train net output #0: loss = 5.22535 (* 1 = 5.22535 loss)
I0428 17:08:12.304590  8468 sgd_solver.cpp:105] Iteration 372, lr = 0.00928961
I0428 17:08:17.573839  8468 solver.cpp:218] Iteration 384 (2.27746 iter/s, 5.26904s/12 iters), loss = 5.13936
I0428 17:08:17.573968  8468 solver.cpp:237]     Train net output #0: loss = 5.13936 (* 1 = 5.13936 loss)
I0428 17:08:17.573980  8468 sgd_solver.cpp:105] Iteration 384, lr = 0.00926756
I0428 17:08:22.996433  8468 solver.cpp:218] Iteration 396 (2.21311 iter/s, 5.42224s/12 iters), loss = 5.12522
I0428 17:08:22.996477  8468 solver.cpp:237]     Train net output #0: loss = 5.12522 (* 1 = 5.12522 loss)
I0428 17:08:22.996505  8468 sgd_solver.cpp:105] Iteration 396, lr = 0.00924556
I0428 17:08:26.359738  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:08:27.869374  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_408.caffemodel
I0428 17:08:28.816205  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_408.solverstate
I0428 17:08:30.810431  8468 solver.cpp:330] Iteration 408, Testing net (#0)
I0428 17:08:30.810453  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:08:35.135363  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:08:35.351131  8468 solver.cpp:397]     Test net output #0: accuracy = 0.00857843
I0428 17:08:35.351163  8468 solver.cpp:397]     Test net output #1: loss = 5.13222 (* 1 = 5.13222 loss)
I0428 17:08:35.482503  8468 solver.cpp:218] Iteration 408 (0.961114 iter/s, 12.4855s/12 iters), loss = 5.18013
I0428 17:08:35.482547  8468 solver.cpp:237]     Train net output #0: loss = 5.18013 (* 1 = 5.18013 loss)
I0428 17:08:35.482554  8468 sgd_solver.cpp:105] Iteration 408, lr = 0.00922361
I0428 17:08:40.044705  8468 solver.cpp:218] Iteration 420 (2.63045 iter/s, 4.56196s/12 iters), loss = 5.01401
I0428 17:08:40.044751  8468 solver.cpp:237]     Train net output #0: loss = 5.01401 (* 1 = 5.01401 loss)
I0428 17:08:40.044762  8468 sgd_solver.cpp:105] Iteration 420, lr = 0.00920171
I0428 17:08:45.779515  8468 solver.cpp:218] Iteration 432 (2.09259 iter/s, 5.73453s/12 iters), loss = 5.01146
I0428 17:08:45.779551  8468 solver.cpp:237]     Train net output #0: loss = 5.01146 (* 1 = 5.01146 loss)
I0428 17:08:45.779561  8468 sgd_solver.cpp:105] Iteration 432, lr = 0.00917986
I0428 17:08:51.364676  8468 solver.cpp:218] Iteration 444 (2.14866 iter/s, 5.58489s/12 iters), loss = 5.14291
I0428 17:08:51.364894  8468 solver.cpp:237]     Train net output #0: loss = 5.14291 (* 1 = 5.14291 loss)
I0428 17:08:51.364905  8468 sgd_solver.cpp:105] Iteration 444, lr = 0.00915807
I0428 17:08:56.837961  8468 solver.cpp:218] Iteration 456 (2.19265 iter/s, 5.47284s/12 iters), loss = 5.10121
I0428 17:08:56.838003  8468 solver.cpp:237]     Train net output #0: loss = 5.10121 (* 1 = 5.10121 loss)
I0428 17:08:56.838013  8468 sgd_solver.cpp:105] Iteration 456, lr = 0.00913632
I0428 17:09:02.239782  8468 solver.cpp:218] Iteration 468 (2.22158 iter/s, 5.40155s/12 iters), loss = 5.11242
I0428 17:09:02.239822  8468 solver.cpp:237]     Train net output #0: loss = 5.11242 (* 1 = 5.11242 loss)
I0428 17:09:02.239832  8468 sgd_solver.cpp:105] Iteration 468, lr = 0.00911463
I0428 17:09:07.711228  8468 solver.cpp:218] Iteration 480 (2.19331 iter/s, 5.47118s/12 iters), loss = 5.05996
I0428 17:09:07.711274  8468 solver.cpp:237]     Train net output #0: loss = 5.05996 (* 1 = 5.05996 loss)
I0428 17:09:07.711284  8468 sgd_solver.cpp:105] Iteration 480, lr = 0.00909299
I0428 17:09:13.192277  8468 solver.cpp:218] Iteration 492 (2.18947 iter/s, 5.48077s/12 iters), loss = 5.15082
I0428 17:09:13.192310  8468 solver.cpp:237]     Train net output #0: loss = 5.15082 (* 1 = 5.15082 loss)
I0428 17:09:13.192318  8468 sgd_solver.cpp:105] Iteration 492, lr = 0.0090714
I0428 17:09:18.609208  8468 solver.cpp:218] Iteration 504 (2.21538 iter/s, 5.41667s/12 iters), loss = 5.09469
I0428 17:09:18.609244  8468 solver.cpp:237]     Train net output #0: loss = 5.09469 (* 1 = 5.09469 loss)
I0428 17:09:18.609254  8468 sgd_solver.cpp:105] Iteration 504, lr = 0.00904986
I0428 17:09:18.866868  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:09:20.803237  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_510.caffemodel
I0428 17:09:21.426465  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_510.solverstate
I0428 17:09:21.873039  8468 solver.cpp:330] Iteration 510, Testing net (#0)
I0428 17:09:21.873057  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:09:26.106806  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:09:26.358701  8468 solver.cpp:397]     Test net output #0: accuracy = 0.0104167
I0428 17:09:26.358748  8468 solver.cpp:397]     Test net output #1: loss = 5.08826 (* 1 = 5.08826 loss)
I0428 17:09:28.432018  8468 solver.cpp:218] Iteration 516 (1.2217 iter/s, 9.82236s/12 iters), loss = 4.99447
I0428 17:09:28.432070  8468 solver.cpp:237]     Train net output #0: loss = 4.99447 (* 1 = 4.99447 loss)
I0428 17:09:28.432080  8468 sgd_solver.cpp:105] Iteration 516, lr = 0.00902838
I0428 17:09:33.910023  8468 solver.cpp:218] Iteration 528 (2.19069 iter/s, 5.47773s/12 iters), loss = 5.0436
I0428 17:09:33.910064  8468 solver.cpp:237]     Train net output #0: loss = 5.0436 (* 1 = 5.0436 loss)
I0428 17:09:33.910074  8468 sgd_solver.cpp:105] Iteration 528, lr = 0.00900694
I0428 17:09:39.313974  8468 solver.cpp:218] Iteration 540 (2.22071 iter/s, 5.40368s/12 iters), loss = 5.06489
I0428 17:09:39.314015  8468 solver.cpp:237]     Train net output #0: loss = 5.06489 (* 1 = 5.06489 loss)
I0428 17:09:39.314025  8468 sgd_solver.cpp:105] Iteration 540, lr = 0.00898556
I0428 17:09:44.685997  8468 solver.cpp:218] Iteration 552 (2.23391 iter/s, 5.37176s/12 iters), loss = 5.17721
I0428 17:09:44.686038  8468 solver.cpp:237]     Train net output #0: loss = 5.17721 (* 1 = 5.17721 loss)
I0428 17:09:44.686048  8468 sgd_solver.cpp:105] Iteration 552, lr = 0.00896423
I0428 17:09:50.097889  8468 solver.cpp:218] Iteration 564 (2.21745 iter/s, 5.41162s/12 iters), loss = 5.15701
I0428 17:09:50.097937  8468 solver.cpp:237]     Train net output #0: loss = 5.15701 (* 1 = 5.15701 loss)
I0428 17:09:50.097946  8468 sgd_solver.cpp:105] Iteration 564, lr = 0.00894294
I0428 17:09:55.500936  8468 solver.cpp:218] Iteration 576 (2.22108 iter/s, 5.40277s/12 iters), loss = 5.02861
I0428 17:09:55.501036  8468 solver.cpp:237]     Train net output #0: loss = 5.02861 (* 1 = 5.02861 loss)
I0428 17:09:55.501047  8468 sgd_solver.cpp:105] Iteration 576, lr = 0.00892171
I0428 17:10:00.871152  8468 solver.cpp:218] Iteration 588 (2.23468 iter/s, 5.36989s/12 iters), loss = 5.04939
I0428 17:10:00.871198  8468 solver.cpp:237]     Train net output #0: loss = 5.04939 (* 1 = 5.04939 loss)
I0428 17:10:00.871210  8468 sgd_solver.cpp:105] Iteration 588, lr = 0.00890053
I0428 17:10:06.439690  8468 solver.cpp:218] Iteration 600 (2.15507 iter/s, 5.56826s/12 iters), loss = 5.10244
I0428 17:10:06.439733  8468 solver.cpp:237]     Train net output #0: loss = 5.10244 (* 1 = 5.10244 loss)
I0428 17:10:06.439743  8468 sgd_solver.cpp:105] Iteration 600, lr = 0.0088794
I0428 17:10:09.000190  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:10:11.305197  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_612.caffemodel
I0428 17:10:12.978655  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_612.solverstate
I0428 17:10:13.447384  8468 solver.cpp:330] Iteration 612, Testing net (#0)
I0428 17:10:13.447404  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:10:17.637253  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:10:17.936543  8468 solver.cpp:397]     Test net output #0: accuracy = 0.0110294
I0428 17:10:17.936589  8468 solver.cpp:397]     Test net output #1: loss = 5.06223 (* 1 = 5.06223 loss)
I0428 17:10:18.065049  8468 solver.cpp:218] Iteration 612 (1.03227 iter/s, 11.6248s/12 iters), loss = 5.0226
I0428 17:10:18.065091  8468 solver.cpp:237]     Train net output #0: loss = 5.0226 (* 1 = 5.0226 loss)
I0428 17:10:18.065102  8468 sgd_solver.cpp:105] Iteration 612, lr = 0.00885831
I0428 17:10:22.566565  8468 solver.cpp:218] Iteration 624 (2.66591 iter/s, 4.50129s/12 iters), loss = 5.14907
I0428 17:10:22.566612  8468 solver.cpp:237]     Train net output #0: loss = 5.14907 (* 1 = 5.14907 loss)
I0428 17:10:22.566623  8468 sgd_solver.cpp:105] Iteration 624, lr = 0.00883728
I0428 17:10:27.904968  8468 solver.cpp:218] Iteration 636 (2.24798 iter/s, 5.33814s/12 iters), loss = 5.154
I0428 17:10:27.905200  8468 solver.cpp:237]     Train net output #0: loss = 5.154 (* 1 = 5.154 loss)
I0428 17:10:27.905211  8468 sgd_solver.cpp:105] Iteration 636, lr = 0.0088163
I0428 17:10:33.327661  8468 solver.cpp:218] Iteration 648 (2.21311 iter/s, 5.42223s/12 iters), loss = 5.11436
I0428 17:10:33.327708  8468 solver.cpp:237]     Train net output #0: loss = 5.11436 (* 1 = 5.11436 loss)
I0428 17:10:33.327719  8468 sgd_solver.cpp:105] Iteration 648, lr = 0.00879537
I0428 17:10:38.712905  8468 solver.cpp:218] Iteration 660 (2.22842 iter/s, 5.38498s/12 iters), loss = 5.05737
I0428 17:10:38.712940  8468 solver.cpp:237]     Train net output #0: loss = 5.05737 (* 1 = 5.05737 loss)
I0428 17:10:38.712949  8468 sgd_solver.cpp:105] Iteration 660, lr = 0.00877449
I0428 17:10:44.187678  8468 solver.cpp:218] Iteration 672 (2.19198 iter/s, 5.47451s/12 iters), loss = 5.0456
I0428 17:10:44.187717  8468 solver.cpp:237]     Train net output #0: loss = 5.0456 (* 1 = 5.0456 loss)
I0428 17:10:44.187728  8468 sgd_solver.cpp:105] Iteration 672, lr = 0.00875366
I0428 17:10:49.064927  8468 blocking_queue.cpp:49] Waiting for data
I0428 17:10:49.588500  8468 solver.cpp:218] Iteration 684 (2.222 iter/s, 5.40054s/12 iters), loss = 4.97355
I0428 17:10:49.588542  8468 solver.cpp:237]     Train net output #0: loss = 4.97355 (* 1 = 4.97355 loss)
I0428 17:10:49.588552  8468 sgd_solver.cpp:105] Iteration 684, lr = 0.00873287
I0428 17:10:55.155923  8468 solver.cpp:218] Iteration 696 (2.1555 iter/s, 5.56715s/12 iters), loss = 5.05126
I0428 17:10:55.155972  8468 solver.cpp:237]     Train net output #0: loss = 5.05126 (* 1 = 5.05126 loss)
I0428 17:10:55.155982  8468 sgd_solver.cpp:105] Iteration 696, lr = 0.00871214
I0428 17:11:00.507278  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:11:00.950453  8468 solver.cpp:218] Iteration 708 (2.07102 iter/s, 5.79424s/12 iters), loss = 4.98734
I0428 17:11:00.950498  8468 solver.cpp:237]     Train net output #0: loss = 4.98734 (* 1 = 4.98734 loss)
I0428 17:11:00.950507  8468 sgd_solver.cpp:105] Iteration 708, lr = 0.00869145
I0428 17:11:03.224025  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_714.caffemodel
I0428 17:11:04.352757  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_714.solverstate
I0428 17:11:06.713780  8468 solver.cpp:330] Iteration 714, Testing net (#0)
I0428 17:11:06.713809  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:11:11.015446  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:11:11.348263  8468 solver.cpp:397]     Test net output #0: accuracy = 0.0196078
I0428 17:11:11.348300  8468 solver.cpp:397]     Test net output #1: loss = 5.01757 (* 1 = 5.01757 loss)
I0428 17:11:13.500391  8468 solver.cpp:218] Iteration 720 (0.956222 iter/s, 12.5494s/12 iters), loss = 5.09171
I0428 17:11:13.500427  8468 solver.cpp:237]     Train net output #0: loss = 5.09171 (* 1 = 5.09171 loss)
I0428 17:11:13.500437  8468 sgd_solver.cpp:105] Iteration 720, lr = 0.00867082
I0428 17:11:18.950918  8468 solver.cpp:218] Iteration 732 (2.20173 iter/s, 5.45026s/12 iters), loss = 5.07471
I0428 17:11:18.950956  8468 solver.cpp:237]     Train net output #0: loss = 5.07471 (* 1 = 5.07471 loss)
I0428 17:11:18.950965  8468 sgd_solver.cpp:105] Iteration 732, lr = 0.00865023
I0428 17:11:24.374450  8468 solver.cpp:218] Iteration 744 (2.21269 iter/s, 5.42327s/12 iters), loss = 4.91155
I0428 17:11:24.374490  8468 solver.cpp:237]     Train net output #0: loss = 4.91155 (* 1 = 4.91155 loss)
I0428 17:11:24.374500  8468 sgd_solver.cpp:105] Iteration 744, lr = 0.0086297
I0428 17:11:29.899490  8468 solver.cpp:218] Iteration 756 (2.17204 iter/s, 5.52477s/12 iters), loss = 4.98481
I0428 17:11:29.899530  8468 solver.cpp:237]     Train net output #0: loss = 4.98481 (* 1 = 4.98481 loss)
I0428 17:11:29.899539  8468 sgd_solver.cpp:105] Iteration 756, lr = 0.00860921
I0428 17:11:35.290779  8468 solver.cpp:218] Iteration 768 (2.22592 iter/s, 5.39102s/12 iters), loss = 5.00341
I0428 17:11:35.290915  8468 solver.cpp:237]     Train net output #0: loss = 5.00341 (* 1 = 5.00341 loss)
I0428 17:11:35.290925  8468 sgd_solver.cpp:105] Iteration 768, lr = 0.00858877
I0428 17:11:40.669930  8468 solver.cpp:218] Iteration 780 (2.23099 iter/s, 5.37879s/12 iters), loss = 4.8663
I0428 17:11:40.669988  8468 solver.cpp:237]     Train net output #0: loss = 4.8663 (* 1 = 4.8663 loss)
I0428 17:11:40.670004  8468 sgd_solver.cpp:105] Iteration 780, lr = 0.00856838
I0428 17:11:46.276527  8468 solver.cpp:218] Iteration 792 (2.14045 iter/s, 5.60631s/12 iters), loss = 4.86655
I0428 17:11:46.276567  8468 solver.cpp:237]     Train net output #0: loss = 4.86655 (* 1 = 4.86655 loss)
I0428 17:11:46.276577  8468 sgd_solver.cpp:105] Iteration 792, lr = 0.00854803
I0428 17:11:51.827461  8468 solver.cpp:218] Iteration 804 (2.16191 iter/s, 5.55066s/12 iters), loss = 5.03654
I0428 17:11:51.827505  8468 solver.cpp:237]     Train net output #0: loss = 5.03654 (* 1 = 5.03654 loss)
I0428 17:11:51.827513  8468 sgd_solver.cpp:105] Iteration 804, lr = 0.00852774
I0428 17:11:53.704169  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:11:56.838284  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_816.caffemodel
I0428 17:11:57.445858  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_816.solverstate
I0428 17:11:57.888985  8468 solver.cpp:330] Iteration 816, Testing net (#0)
I0428 17:11:57.889003  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:12:02.010901  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:12:02.378180  8468 solver.cpp:397]     Test net output #0: accuracy = 0.0251225
I0428 17:12:02.378226  8468 solver.cpp:397]     Test net output #1: loss = 4.95976 (* 1 = 4.95976 loss)
I0428 17:12:02.509510  8468 solver.cpp:218] Iteration 816 (1.12343 iter/s, 10.6816s/12 iters), loss = 4.90248
I0428 17:12:02.509557  8468 solver.cpp:237]     Train net output #0: loss = 4.90248 (* 1 = 4.90248 loss)
I0428 17:12:02.509569  8468 sgd_solver.cpp:105] Iteration 816, lr = 0.00850749
I0428 17:12:06.997107  8468 solver.cpp:218] Iteration 828 (2.67418 iter/s, 4.48736s/12 iters), loss = 4.894
I0428 17:12:06.997195  8468 solver.cpp:237]     Train net output #0: loss = 4.894 (* 1 = 4.894 loss)
I0428 17:12:06.997208  8468 sgd_solver.cpp:105] Iteration 828, lr = 0.00848729
I0428 17:12:12.601701  8468 solver.cpp:218] Iteration 840 (2.14122 iter/s, 5.60427s/12 iters), loss = 4.8474
I0428 17:12:12.601740  8468 solver.cpp:237]     Train net output #0: loss = 4.8474 (* 1 = 4.8474 loss)
I0428 17:12:12.601752  8468 sgd_solver.cpp:105] Iteration 840, lr = 0.00846714
I0428 17:12:18.119065  8468 solver.cpp:218] Iteration 852 (2.17506 iter/s, 5.5171s/12 iters), loss = 4.83236
I0428 17:12:18.119097  8468 solver.cpp:237]     Train net output #0: loss = 4.83236 (* 1 = 4.83236 loss)
I0428 17:12:18.119107  8468 sgd_solver.cpp:105] Iteration 852, lr = 0.00844704
I0428 17:12:23.535920  8468 solver.cpp:218] Iteration 864 (2.21542 iter/s, 5.41659s/12 iters), loss = 4.95609
I0428 17:12:23.535967  8468 solver.cpp:237]     Train net output #0: loss = 4.95609 (* 1 = 4.95609 loss)
I0428 17:12:23.535979  8468 sgd_solver.cpp:105] Iteration 864, lr = 0.00842698
I0428 17:12:29.144287  8468 solver.cpp:218] Iteration 876 (2.13977 iter/s, 5.60808s/12 iters), loss = 4.83863
I0428 17:12:29.144330  8468 solver.cpp:237]     Train net output #0: loss = 4.83863 (* 1 = 4.83863 loss)
I0428 17:12:29.144338  8468 sgd_solver.cpp:105] Iteration 876, lr = 0.00840698
I0428 17:12:34.529731  8468 solver.cpp:218] Iteration 888 (2.22834 iter/s, 5.38517s/12 iters), loss = 4.73402
I0428 17:12:34.529776  8468 solver.cpp:237]     Train net output #0: loss = 4.73402 (* 1 = 4.73402 loss)
I0428 17:12:34.529786  8468 sgd_solver.cpp:105] Iteration 888, lr = 0.00838702
I0428 17:12:39.895275  8468 solver.cpp:218] Iteration 900 (2.23661 iter/s, 5.36527s/12 iters), loss = 4.83889
I0428 17:12:39.895401  8468 solver.cpp:237]     Train net output #0: loss = 4.83889 (* 1 = 4.83889 loss)
I0428 17:12:39.895411  8468 sgd_solver.cpp:105] Iteration 900, lr = 0.0083671
I0428 17:12:44.042697  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:12:45.261747  8468 solver.cpp:218] Iteration 912 (2.23625 iter/s, 5.36612s/12 iters), loss = 4.92283
I0428 17:12:45.261795  8468 solver.cpp:237]     Train net output #0: loss = 4.92283 (* 1 = 4.92283 loss)
I0428 17:12:45.261806  8468 sgd_solver.cpp:105] Iteration 912, lr = 0.00834724
I0428 17:12:47.400522  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_918.caffemodel
I0428 17:12:48.034525  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_918.solverstate
I0428 17:12:48.490084  8468 solver.cpp:330] Iteration 918, Testing net (#0)
I0428 17:12:48.490113  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:12:52.564455  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:12:52.980118  8468 solver.cpp:397]     Test net output #0: accuracy = 0.0318627
I0428 17:12:52.980165  8468 solver.cpp:397]     Test net output #1: loss = 4.84772 (* 1 = 4.84772 loss)
I0428 17:12:55.129108  8468 solver.cpp:218] Iteration 924 (1.21619 iter/s, 9.86691s/12 iters), loss = 4.68118
I0428 17:12:55.129143  8468 solver.cpp:237]     Train net output #0: loss = 4.68118 (* 1 = 4.68118 loss)
I0428 17:12:55.129151  8468 sgd_solver.cpp:105] Iteration 924, lr = 0.00832742
I0428 17:13:00.541551  8468 solver.cpp:218] Iteration 936 (2.21722 iter/s, 5.41218s/12 iters), loss = 4.69367
I0428 17:13:00.541592  8468 solver.cpp:237]     Train net output #0: loss = 4.69367 (* 1 = 4.69367 loss)
I0428 17:13:00.541602  8468 sgd_solver.cpp:105] Iteration 936, lr = 0.00830765
I0428 17:13:06.234524  8468 solver.cpp:218] Iteration 948 (2.10797 iter/s, 5.69269s/12 iters), loss = 4.74001
I0428 17:13:06.234565  8468 solver.cpp:237]     Train net output #0: loss = 4.74001 (* 1 = 4.74001 loss)
I0428 17:13:06.234575  8468 sgd_solver.cpp:105] Iteration 948, lr = 0.00828793
I0428 17:13:12.051380  8468 solver.cpp:218] Iteration 960 (2.06307 iter/s, 5.81657s/12 iters), loss = 4.86918
I0428 17:13:12.051497  8468 solver.cpp:237]     Train net output #0: loss = 4.86918 (* 1 = 4.86918 loss)
I0428 17:13:12.051507  8468 sgd_solver.cpp:105] Iteration 960, lr = 0.00826825
I0428 17:13:17.674300  8468 solver.cpp:218] Iteration 972 (2.13425 iter/s, 5.62257s/12 iters), loss = 4.71764
I0428 17:13:17.674333  8468 solver.cpp:237]     Train net output #0: loss = 4.71764 (* 1 = 4.71764 loss)
I0428 17:13:17.674341  8468 sgd_solver.cpp:105] Iteration 972, lr = 0.00824862
I0428 17:13:23.266707  8468 solver.cpp:218] Iteration 984 (2.14587 iter/s, 5.59214s/12 iters), loss = 4.70578
I0428 17:13:23.266752  8468 solver.cpp:237]     Train net output #0: loss = 4.70578 (* 1 = 4.70578 loss)
I0428 17:13:23.266765  8468 sgd_solver.cpp:105] Iteration 984, lr = 0.00822903
I0428 17:13:28.622541  8468 solver.cpp:218] Iteration 996 (2.24066 iter/s, 5.35556s/12 iters), loss = 4.79299
I0428 17:13:28.622586  8468 solver.cpp:237]     Train net output #0: loss = 4.79299 (* 1 = 4.79299 loss)
I0428 17:13:28.622596  8468 sgd_solver.cpp:105] Iteration 996, lr = 0.0082095
I0428 17:13:33.993476  8468 solver.cpp:218] Iteration 1008 (2.23436 iter/s, 5.37066s/12 iters), loss = 4.69624
I0428 17:13:33.993520  8468 solver.cpp:237]     Train net output #0: loss = 4.69624 (* 1 = 4.69624 loss)
I0428 17:13:33.993530  8468 sgd_solver.cpp:105] Iteration 1008, lr = 0.00819001
I0428 17:13:35.067310  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:13:38.844363  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1020.caffemodel
I0428 17:13:39.414171  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1020.solverstate
I0428 17:13:39.840910  8468 solver.cpp:330] Iteration 1020, Testing net (#0)
I0428 17:13:39.840934  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:13:43.980310  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:13:44.431190  8468 solver.cpp:397]     Test net output #0: accuracy = 0.0428922
I0428 17:13:44.431228  8468 solver.cpp:397]     Test net output #1: loss = 4.68907 (* 1 = 4.68907 loss)
I0428 17:13:44.562284  8468 solver.cpp:218] Iteration 1020 (1.13547 iter/s, 10.5683s/12 iters), loss = 4.81366
I0428 17:13:44.562325  8468 solver.cpp:237]     Train net output #0: loss = 4.81366 (* 1 = 4.81366 loss)
I0428 17:13:44.562335  8468 sgd_solver.cpp:105] Iteration 1020, lr = 0.00817056
I0428 17:13:49.058552  8468 solver.cpp:218] Iteration 1032 (2.66902 iter/s, 4.49603s/12 iters), loss = 4.77885
I0428 17:13:49.058598  8468 solver.cpp:237]     Train net output #0: loss = 4.77885 (* 1 = 4.77885 loss)
I0428 17:13:49.058609  8468 sgd_solver.cpp:105] Iteration 1032, lr = 0.00815116
I0428 17:13:54.576115  8468 solver.cpp:218] Iteration 1044 (2.17498 iter/s, 5.51728s/12 iters), loss = 4.7512
I0428 17:13:54.576164  8468 solver.cpp:237]     Train net output #0: loss = 4.7512 (* 1 = 4.7512 loss)
I0428 17:13:54.576172  8468 sgd_solver.cpp:105] Iteration 1044, lr = 0.00813181
I0428 17:13:59.946899  8468 solver.cpp:218] Iteration 1056 (2.23442 iter/s, 5.37051s/12 iters), loss = 4.55923
I0428 17:13:59.946940  8468 solver.cpp:237]     Train net output #0: loss = 4.55923 (* 1 = 4.55923 loss)
I0428 17:13:59.946950  8468 sgd_solver.cpp:105] Iteration 1056, lr = 0.0081125
I0428 17:14:05.334434  8468 solver.cpp:218] Iteration 1068 (2.22748 iter/s, 5.38727s/12 iters), loss = 4.67573
I0428 17:14:05.334481  8468 solver.cpp:237]     Train net output #0: loss = 4.67573 (* 1 = 4.67573 loss)
I0428 17:14:05.334491  8468 sgd_solver.cpp:105] Iteration 1068, lr = 0.00809324
I0428 17:14:10.799676  8468 solver.cpp:218] Iteration 1080 (2.1958 iter/s, 5.46498s/12 iters), loss = 4.62943
I0428 17:14:10.799710  8468 solver.cpp:237]     Train net output #0: loss = 4.62943 (* 1 = 4.62943 loss)
I0428 17:14:10.799718  8468 sgd_solver.cpp:105] Iteration 1080, lr = 0.00807403
I0428 17:14:16.245736  8468 solver.cpp:218] Iteration 1092 (2.20354 iter/s, 5.44579s/12 iters), loss = 4.66901
I0428 17:14:16.245878  8468 solver.cpp:237]     Train net output #0: loss = 4.66901 (* 1 = 4.66901 loss)
I0428 17:14:16.245894  8468 sgd_solver.cpp:105] Iteration 1092, lr = 0.00805486
I0428 17:14:21.719558  8468 solver.cpp:218] Iteration 1104 (2.1924 iter/s, 5.47346s/12 iters), loss = 4.52904
I0428 17:14:21.719596  8468 solver.cpp:237]     Train net output #0: loss = 4.52904 (* 1 = 4.52904 loss)
I0428 17:14:21.719604  8468 sgd_solver.cpp:105] Iteration 1104, lr = 0.00803573
I0428 17:14:25.216936  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:14:27.278514  8468 solver.cpp:218] Iteration 1116 (2.15878 iter/s, 5.55869s/12 iters), loss = 4.69837
I0428 17:14:27.278550  8468 solver.cpp:237]     Train net output #0: loss = 4.69837 (* 1 = 4.69837 loss)
I0428 17:14:27.278558  8468 sgd_solver.cpp:105] Iteration 1116, lr = 0.00801666
I0428 17:14:29.517220  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1122.caffemodel
I0428 17:14:30.117799  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1122.solverstate
I0428 17:14:30.551839  8468 solver.cpp:330] Iteration 1122, Testing net (#0)
I0428 17:14:30.551859  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:14:34.758577  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:14:35.299196  8468 solver.cpp:397]     Test net output #0: accuracy = 0.0422794
I0428 17:14:35.299232  8468 solver.cpp:397]     Test net output #1: loss = 4.58472 (* 1 = 4.58472 loss)
I0428 17:14:37.286199  8468 solver.cpp:218] Iteration 1128 (1.19913 iter/s, 10.0072s/12 iters), loss = 4.35168
I0428 17:14:37.286245  8468 solver.cpp:237]     Train net output #0: loss = 4.35168 (* 1 = 4.35168 loss)
I0428 17:14:37.286255  8468 sgd_solver.cpp:105] Iteration 1128, lr = 0.00799762
I0428 17:14:42.819391  8468 solver.cpp:218] Iteration 1140 (2.16884 iter/s, 5.53292s/12 iters), loss = 4.33376
I0428 17:14:42.819427  8468 solver.cpp:237]     Train net output #0: loss = 4.33376 (* 1 = 4.33376 loss)
I0428 17:14:42.819437  8468 sgd_solver.cpp:105] Iteration 1140, lr = 0.00797863
I0428 17:14:48.309165  8468 solver.cpp:218] Iteration 1152 (2.18599 iter/s, 5.48951s/12 iters), loss = 4.58961
I0428 17:14:48.309952  8468 solver.cpp:237]     Train net output #0: loss = 4.58961 (* 1 = 4.58961 loss)
I0428 17:14:48.309962  8468 sgd_solver.cpp:105] Iteration 1152, lr = 0.00795969
I0428 17:14:53.708199  8468 solver.cpp:218] Iteration 1164 (2.22304 iter/s, 5.39802s/12 iters), loss = 4.45267
I0428 17:14:53.708243  8468 solver.cpp:237]     Train net output #0: loss = 4.45267 (* 1 = 4.45267 loss)
I0428 17:14:53.708254  8468 sgd_solver.cpp:105] Iteration 1164, lr = 0.00794079
I0428 17:14:59.050473  8468 solver.cpp:218] Iteration 1176 (2.24635 iter/s, 5.34201s/12 iters), loss = 4.50819
I0428 17:14:59.050511  8468 solver.cpp:237]     Train net output #0: loss = 4.50819 (* 1 = 4.50819 loss)
I0428 17:14:59.050521  8468 sgd_solver.cpp:105] Iteration 1176, lr = 0.00792194
I0428 17:15:04.587962  8468 solver.cpp:218] Iteration 1188 (2.16715 iter/s, 5.53722s/12 iters), loss = 4.42853
I0428 17:15:04.588007  8468 solver.cpp:237]     Train net output #0: loss = 4.42853 (* 1 = 4.42853 loss)
I0428 17:15:04.588018  8468 sgd_solver.cpp:105] Iteration 1188, lr = 0.00790313
I0428 17:15:10.061631  8468 solver.cpp:218] Iteration 1200 (2.19242 iter/s, 5.4734s/12 iters), loss = 4.55942
I0428 17:15:10.061671  8468 solver.cpp:237]     Train net output #0: loss = 4.55942 (* 1 = 4.55942 loss)
I0428 17:15:10.061681  8468 sgd_solver.cpp:105] Iteration 1200, lr = 0.00788437
I0428 17:15:15.504047  8468 solver.cpp:218] Iteration 1212 (2.20501 iter/s, 5.44215s/12 iters), loss = 4.61584
I0428 17:15:15.504086  8468 solver.cpp:237]     Train net output #0: loss = 4.61584 (* 1 = 4.61584 loss)
I0428 17:15:15.504096  8468 sgd_solver.cpp:105] Iteration 1212, lr = 0.00786565
I0428 17:15:15.825845  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:15:20.733345  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1224.caffemodel
I0428 17:15:21.392565  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1224.solverstate
I0428 17:15:21.843149  8468 solver.cpp:330] Iteration 1224, Testing net (#0)
I0428 17:15:21.843176  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:15:25.988382  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:15:26.521697  8468 solver.cpp:397]     Test net output #0: accuracy = 0.0612745
I0428 17:15:26.521744  8468 solver.cpp:397]     Test net output #1: loss = 4.37915 (* 1 = 4.37915 loss)
I0428 17:15:26.652791  8468 solver.cpp:218] Iteration 1224 (1.0764 iter/s, 11.1483s/12 iters), loss = 4.21562
I0428 17:15:26.652838  8468 solver.cpp:237]     Train net output #0: loss = 4.21562 (* 1 = 4.21562 loss)
I0428 17:15:26.652848  8468 sgd_solver.cpp:105] Iteration 1224, lr = 0.00784697
I0428 17:15:31.307992  8468 solver.cpp:218] Iteration 1236 (2.5779 iter/s, 4.65495s/12 iters), loss = 4.37777
I0428 17:15:31.308035  8468 solver.cpp:237]     Train net output #0: loss = 4.37777 (* 1 = 4.37777 loss)
I0428 17:15:31.308044  8468 sgd_solver.cpp:105] Iteration 1236, lr = 0.00782834
I0428 17:15:36.678403  8468 solver.cpp:218] Iteration 1248 (2.23458 iter/s, 5.37014s/12 iters), loss = 4.66711
I0428 17:15:36.678448  8468 solver.cpp:237]     Train net output #0: loss = 4.66711 (* 1 = 4.66711 loss)
I0428 17:15:36.678457  8468 sgd_solver.cpp:105] Iteration 1248, lr = 0.00780976
I0428 17:15:42.041462  8468 solver.cpp:218] Iteration 1260 (2.23764 iter/s, 5.36279s/12 iters), loss = 4.58
I0428 17:15:42.041505  8468 solver.cpp:237]     Train net output #0: loss = 4.58 (* 1 = 4.58 loss)
I0428 17:15:42.041515  8468 sgd_solver.cpp:105] Iteration 1260, lr = 0.00779122
I0428 17:15:47.380082  8468 solver.cpp:218] Iteration 1272 (2.24789 iter/s, 5.33835s/12 iters), loss = 4.44055
I0428 17:15:47.380127  8468 solver.cpp:237]     Train net output #0: loss = 4.44055 (* 1 = 4.44055 loss)
I0428 17:15:47.380137  8468 sgd_solver.cpp:105] Iteration 1272, lr = 0.00777272
I0428 17:15:52.815054  8468 solver.cpp:218] Iteration 1284 (2.20803 iter/s, 5.43471s/12 iters), loss = 4.36151
I0428 17:15:52.815388  8468 solver.cpp:237]     Train net output #0: loss = 4.36151 (* 1 = 4.36151 loss)
I0428 17:15:52.815400  8468 sgd_solver.cpp:105] Iteration 1284, lr = 0.00775426
I0428 17:15:58.233530  8468 solver.cpp:218] Iteration 1296 (2.21487 iter/s, 5.41792s/12 iters), loss = 4.408
I0428 17:15:58.233569  8468 solver.cpp:237]     Train net output #0: loss = 4.408 (* 1 = 4.408 loss)
I0428 17:15:58.233582  8468 sgd_solver.cpp:105] Iteration 1296, lr = 0.00773585
I0428 17:16:03.611009  8468 solver.cpp:218] Iteration 1308 (2.23164 iter/s, 5.37722s/12 iters), loss = 4.47339
I0428 17:16:03.611050  8468 solver.cpp:237]     Train net output #0: loss = 4.47339 (* 1 = 4.47339 loss)
I0428 17:16:03.611059  8468 sgd_solver.cpp:105] Iteration 1308, lr = 0.00771749
I0428 17:16:06.279927  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:16:08.947788  8468 solver.cpp:218] Iteration 1320 (2.24866 iter/s, 5.33652s/12 iters), loss = 4.42188
I0428 17:16:08.947829  8468 solver.cpp:237]     Train net output #0: loss = 4.42188 (* 1 = 4.42188 loss)
I0428 17:16:08.947839  8468 sgd_solver.cpp:105] Iteration 1320, lr = 0.00769916
I0428 17:16:11.109923  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1326.caffemodel
I0428 17:16:11.685899  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1326.solverstate
I0428 17:16:12.124800  8468 solver.cpp:330] Iteration 1326, Testing net (#0)
I0428 17:16:12.124819  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:16:16.108569  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:16:16.693035  8468 solver.cpp:397]     Test net output #0: accuracy = 0.0637255
I0428 17:16:16.693064  8468 solver.cpp:397]     Test net output #1: loss = 4.26672 (* 1 = 4.26672 loss)
I0428 17:16:18.741119  8468 solver.cpp:218] Iteration 1332 (1.22538 iter/s, 9.79289s/12 iters), loss = 4.47959
I0428 17:16:18.741163  8468 solver.cpp:237]     Train net output #0: loss = 4.47959 (* 1 = 4.47959 loss)
I0428 17:16:18.741173  8468 sgd_solver.cpp:105] Iteration 1332, lr = 0.00768088
I0428 17:16:24.159502  8468 solver.cpp:218] Iteration 1344 (2.21479 iter/s, 5.41812s/12 iters), loss = 4.30435
I0428 17:16:24.159628  8468 solver.cpp:237]     Train net output #0: loss = 4.30435 (* 1 = 4.30435 loss)
I0428 17:16:24.159641  8468 sgd_solver.cpp:105] Iteration 1344, lr = 0.00766265
I0428 17:16:29.654168  8468 solver.cpp:218] Iteration 1356 (2.18408 iter/s, 5.49431s/12 iters), loss = 4.31725
I0428 17:16:29.654214  8468 solver.cpp:237]     Train net output #0: loss = 4.31725 (* 1 = 4.31725 loss)
I0428 17:16:29.654229  8468 sgd_solver.cpp:105] Iteration 1356, lr = 0.00764446
I0428 17:16:35.051504  8468 solver.cpp:218] Iteration 1368 (2.22343 iter/s, 5.39707s/12 iters), loss = 4.30708
I0428 17:16:35.051545  8468 solver.cpp:237]     Train net output #0: loss = 4.30708 (* 1 = 4.30708 loss)
I0428 17:16:35.051556  8468 sgd_solver.cpp:105] Iteration 1368, lr = 0.00762631
I0428 17:16:35.051748  8468 blocking_queue.cpp:49] Waiting for data
I0428 17:16:40.407663  8468 solver.cpp:218] Iteration 1380 (2.24052 iter/s, 5.35589s/12 iters), loss = 4.29363
I0428 17:16:40.407706  8468 solver.cpp:237]     Train net output #0: loss = 4.29363 (* 1 = 4.29363 loss)
I0428 17:16:40.407716  8468 sgd_solver.cpp:105] Iteration 1380, lr = 0.0076082
I0428 17:16:45.793097  8468 solver.cpp:218] Iteration 1392 (2.22834 iter/s, 5.38517s/12 iters), loss = 4.21257
I0428 17:16:45.793138  8468 solver.cpp:237]     Train net output #0: loss = 4.21257 (* 1 = 4.21257 loss)
I0428 17:16:45.793148  8468 sgd_solver.cpp:105] Iteration 1392, lr = 0.00759014
I0428 17:16:51.155963  8468 solver.cpp:218] Iteration 1404 (2.23772 iter/s, 5.3626s/12 iters), loss = 4.32133
I0428 17:16:51.156006  8468 solver.cpp:237]     Train net output #0: loss = 4.32133 (* 1 = 4.32133 loss)
I0428 17:16:51.156016  8468 sgd_solver.cpp:105] Iteration 1404, lr = 0.00757212
I0428 17:16:56.132803  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:16:56.545696  8468 solver.cpp:218] Iteration 1416 (2.22657 iter/s, 5.38946s/12 iters), loss = 4.17097
I0428 17:16:56.545743  8468 solver.cpp:237]     Train net output #0: loss = 4.17097 (* 1 = 4.17097 loss)
I0428 17:16:56.545754  8468 sgd_solver.cpp:105] Iteration 1416, lr = 0.00755414
I0428 17:17:01.443385  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1428.caffemodel
I0428 17:17:02.537979  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1428.solverstate
I0428 17:17:03.948308  8468 solver.cpp:330] Iteration 1428, Testing net (#0)
I0428 17:17:03.948331  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:17:07.886624  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:17:08.503361  8468 solver.cpp:397]     Test net output #0: accuracy = 0.0790441
I0428 17:17:08.503391  8468 solver.cpp:397]     Test net output #1: loss = 4.21468 (* 1 = 4.21468 loss)
I0428 17:17:08.634099  8468 solver.cpp:218] Iteration 1428 (0.99273 iter/s, 12.0879s/12 iters), loss = 4.13891
I0428 17:17:08.634140  8468 solver.cpp:237]     Train net output #0: loss = 4.13891 (* 1 = 4.13891 loss)
I0428 17:17:08.634150  8468 sgd_solver.cpp:105] Iteration 1428, lr = 0.0075362
I0428 17:17:13.324101  8468 solver.cpp:218] Iteration 1440 (2.55877 iter/s, 4.68976s/12 iters), loss = 4.04427
I0428 17:17:13.324144  8468 solver.cpp:237]     Train net output #0: loss = 4.04427 (* 1 = 4.04427 loss)
I0428 17:17:13.324152  8468 sgd_solver.cpp:105] Iteration 1440, lr = 0.00751831
I0428 17:17:18.798471  8468 solver.cpp:218] Iteration 1452 (2.19214 iter/s, 5.4741s/12 iters), loss = 4.07235
I0428 17:17:18.798516  8468 solver.cpp:237]     Train net output #0: loss = 4.07235 (* 1 = 4.07235 loss)
I0428 17:17:18.798527  8468 sgd_solver.cpp:105] Iteration 1452, lr = 0.00750046
I0428 17:17:24.217286  8468 solver.cpp:218] Iteration 1464 (2.21462 iter/s, 5.41855s/12 iters), loss = 4.21797
I0428 17:17:24.217325  8468 solver.cpp:237]     Train net output #0: loss = 4.21797 (* 1 = 4.21797 loss)
I0428 17:17:24.217335  8468 sgd_solver.cpp:105] Iteration 1464, lr = 0.00748265
I0428 17:17:29.627152  8468 solver.cpp:218] Iteration 1476 (2.21828 iter/s, 5.4096s/12 iters), loss = 4.15995
I0428 17:17:29.627254  8468 solver.cpp:237]     Train net output #0: loss = 4.15995 (* 1 = 4.15995 loss)
I0428 17:17:29.627264  8468 sgd_solver.cpp:105] Iteration 1476, lr = 0.00746489
I0428 17:17:35.168551  8468 solver.cpp:218] Iteration 1488 (2.16565 iter/s, 5.54107s/12 iters), loss = 3.98703
I0428 17:17:35.168601  8468 solver.cpp:237]     Train net output #0: loss = 3.98703 (* 1 = 3.98703 loss)
I0428 17:17:35.168612  8468 sgd_solver.cpp:105] Iteration 1488, lr = 0.00744716
I0428 17:17:40.565022  8468 solver.cpp:218] Iteration 1500 (2.22379 iter/s, 5.3962s/12 iters), loss = 4.09219
I0428 17:17:40.565069  8468 solver.cpp:237]     Train net output #0: loss = 4.09219 (* 1 = 4.09219 loss)
I0428 17:17:40.565079  8468 sgd_solver.cpp:105] Iteration 1500, lr = 0.00742948
I0428 17:17:45.978698  8468 solver.cpp:218] Iteration 1512 (2.21672 iter/s, 5.41341s/12 iters), loss = 3.9573
I0428 17:17:45.978737  8468 solver.cpp:237]     Train net output #0: loss = 3.9573 (* 1 = 3.9573 loss)
I0428 17:17:45.978747  8468 sgd_solver.cpp:105] Iteration 1512, lr = 0.00741184
I0428 17:17:47.937486  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:17:51.506393  8468 solver.cpp:218] Iteration 1524 (2.17099 iter/s, 5.52743s/12 iters), loss = 3.65154
I0428 17:17:51.506438  8468 solver.cpp:237]     Train net output #0: loss = 3.65154 (* 1 = 3.65154 loss)
I0428 17:17:51.506449  8468 sgd_solver.cpp:105] Iteration 1524, lr = 0.00739425
I0428 17:17:53.710327  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1530.caffemodel
I0428 17:17:55.958719  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1530.solverstate
I0428 17:17:58.236944  8468 solver.cpp:330] Iteration 1530, Testing net (#0)
I0428 17:17:58.236974  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:18:02.175673  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:18:02.854951  8468 solver.cpp:397]     Test net output #0: accuracy = 0.0839461
I0428 17:18:02.854993  8468 solver.cpp:397]     Test net output #1: loss = 4.03887 (* 1 = 4.03887 loss)
I0428 17:18:04.799152  8468 solver.cpp:218] Iteration 1536 (0.902786 iter/s, 13.2922s/12 iters), loss = 3.9096
I0428 17:18:04.799204  8468 solver.cpp:237]     Train net output #0: loss = 3.9096 (* 1 = 3.9096 loss)
I0428 17:18:04.799216  8468 sgd_solver.cpp:105] Iteration 1536, lr = 0.00737669
I0428 17:18:10.198578  8468 solver.cpp:218] Iteration 1548 (2.22257 iter/s, 5.39915s/12 iters), loss = 3.98094
I0428 17:18:10.198621  8468 solver.cpp:237]     Train net output #0: loss = 3.98094 (* 1 = 3.98094 loss)
I0428 17:18:10.198629  8468 sgd_solver.cpp:105] Iteration 1548, lr = 0.00735918
I0428 17:18:15.564859  8468 solver.cpp:218] Iteration 1560 (2.2363 iter/s, 5.36602s/12 iters), loss = 3.86983
I0428 17:18:15.564908  8468 solver.cpp:237]     Train net output #0: loss = 3.86983 (* 1 = 3.86983 loss)
I0428 17:18:15.564920  8468 sgd_solver.cpp:105] Iteration 1560, lr = 0.00734171
I0428 17:18:21.026067  8468 solver.cpp:218] Iteration 1572 (2.19742 iter/s, 5.46094s/12 iters), loss = 3.95333
I0428 17:18:21.026110  8468 solver.cpp:237]     Train net output #0: loss = 3.95333 (* 1 = 3.95333 loss)
I0428 17:18:21.026120  8468 sgd_solver.cpp:105] Iteration 1572, lr = 0.00732427
I0428 17:18:26.395061  8468 solver.cpp:218] Iteration 1584 (2.23517 iter/s, 5.36873s/12 iters), loss = 3.69217
I0428 17:18:26.395105  8468 solver.cpp:237]     Train net output #0: loss = 3.69217 (* 1 = 3.69217 loss)
I0428 17:18:26.395114  8468 sgd_solver.cpp:105] Iteration 1584, lr = 0.00730688
I0428 17:18:31.786880  8468 solver.cpp:218] Iteration 1596 (2.2257 iter/s, 5.39155s/12 iters), loss = 3.69305
I0428 17:18:31.786924  8468 solver.cpp:237]     Train net output #0: loss = 3.69305 (* 1 = 3.69305 loss)
I0428 17:18:31.786934  8468 sgd_solver.cpp:105] Iteration 1596, lr = 0.00728954
I0428 17:18:37.259068  8468 solver.cpp:218] Iteration 1608 (2.19302 iter/s, 5.47192s/12 iters), loss = 3.73013
I0428 17:18:37.259172  8468 solver.cpp:237]     Train net output #0: loss = 3.73013 (* 1 = 3.73013 loss)
I0428 17:18:37.259187  8468 sgd_solver.cpp:105] Iteration 1608, lr = 0.00727223
I0428 17:18:41.491746  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:18:42.685894  8468 solver.cpp:218] Iteration 1620 (2.21137 iter/s, 5.4265s/12 iters), loss = 3.82118
I0428 17:18:42.685940  8468 solver.cpp:237]     Train net output #0: loss = 3.82118 (* 1 = 3.82118 loss)
I0428 17:18:42.685951  8468 sgd_solver.cpp:105] Iteration 1620, lr = 0.00725496
I0428 17:18:47.558706  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1632.caffemodel
I0428 17:18:48.156687  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1632.solverstate
I0428 17:18:48.783658  8468 solver.cpp:330] Iteration 1632, Testing net (#0)
I0428 17:18:48.783689  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:18:52.657752  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:18:53.379070  8468 solver.cpp:397]     Test net output #0: accuracy = 0.110907
I0428 17:18:53.379114  8468 solver.cpp:397]     Test net output #1: loss = 3.90807 (* 1 = 3.90807 loss)
I0428 17:18:53.500370  8468 solver.cpp:218] Iteration 1632 (1.10967 iter/s, 10.814s/12 iters), loss = 3.86968
I0428 17:18:53.500433  8468 solver.cpp:237]     Train net output #0: loss = 3.86968 (* 1 = 3.86968 loss)
I0428 17:18:53.500447  8468 sgd_solver.cpp:105] Iteration 1632, lr = 0.00723774
I0428 17:18:58.000432  8468 solver.cpp:218] Iteration 1644 (2.66678 iter/s, 4.49981s/12 iters), loss = 3.72193
I0428 17:18:58.000473  8468 solver.cpp:237]     Train net output #0: loss = 3.72193 (* 1 = 3.72193 loss)
I0428 17:18:58.000483  8468 sgd_solver.cpp:105] Iteration 1644, lr = 0.00722056
I0428 17:19:03.363554  8468 solver.cpp:218] Iteration 1656 (2.23761 iter/s, 5.36286s/12 iters), loss = 3.85367
I0428 17:19:03.363600  8468 solver.cpp:237]     Train net output #0: loss = 3.85367 (* 1 = 3.85367 loss)
I0428 17:19:03.363612  8468 sgd_solver.cpp:105] Iteration 1656, lr = 0.00720341
I0428 17:19:08.801438  8468 solver.cpp:218] Iteration 1668 (2.20685 iter/s, 5.43762s/12 iters), loss = 3.93195
I0428 17:19:08.801565  8468 solver.cpp:237]     Train net output #0: loss = 3.93195 (* 1 = 3.93195 loss)
I0428 17:19:08.801576  8468 sgd_solver.cpp:105] Iteration 1668, lr = 0.00718631
I0428 17:19:14.229370  8468 solver.cpp:218] Iteration 1680 (2.21093 iter/s, 5.42758s/12 iters), loss = 3.67725
I0428 17:19:14.229411  8468 solver.cpp:237]     Train net output #0: loss = 3.67725 (* 1 = 3.67725 loss)
I0428 17:19:14.229421  8468 sgd_solver.cpp:105] Iteration 1680, lr = 0.00716925
I0428 17:19:19.616279  8468 solver.cpp:218] Iteration 1692 (2.22773 iter/s, 5.38664s/12 iters), loss = 3.70248
I0428 17:19:19.616331  8468 solver.cpp:237]     Train net output #0: loss = 3.70248 (* 1 = 3.70248 loss)
I0428 17:19:19.616343  8468 sgd_solver.cpp:105] Iteration 1692, lr = 0.00715223
I0428 17:19:26.235999  8468 solver.cpp:218] Iteration 1704 (1.81286 iter/s, 6.61939s/12 iters), loss = 3.93202
I0428 17:19:26.236065  8468 solver.cpp:237]     Train net output #0: loss = 3.93202 (* 1 = 3.93202 loss)
I0428 17:19:26.236079  8468 sgd_solver.cpp:105] Iteration 1704, lr = 0.00713525
I0428 17:19:33.034912  8468 solver.cpp:218] Iteration 1716 (1.76508 iter/s, 6.79857s/12 iters), loss = 3.70874
I0428 17:19:33.034966  8468 solver.cpp:237]     Train net output #0: loss = 3.70874 (* 1 = 3.70874 loss)
I0428 17:19:33.034978  8468 sgd_solver.cpp:105] Iteration 1716, lr = 0.00711831
I0428 17:19:34.375176  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:19:38.728271  8468 solver.cpp:218] Iteration 1728 (2.10782 iter/s, 5.69308s/12 iters), loss = 3.84139
I0428 17:19:38.728313  8468 solver.cpp:237]     Train net output #0: loss = 3.84139 (* 1 = 3.84139 loss)
I0428 17:19:38.728323  8468 sgd_solver.cpp:105] Iteration 1728, lr = 0.00710141
I0428 17:19:40.970368  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1734.caffemodel
I0428 17:19:41.930577  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1734.solverstate
I0428 17:19:43.594262  8468 solver.cpp:330] Iteration 1734, Testing net (#0)
I0428 17:19:43.594285  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:19:47.390987  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:19:48.126010  8468 solver.cpp:397]     Test net output #0: accuracy = 0.139093
I0428 17:19:48.126039  8468 solver.cpp:397]     Test net output #1: loss = 3.74527 (* 1 = 3.74527 loss)
I0428 17:19:50.167234  8468 solver.cpp:218] Iteration 1740 (1.04909 iter/s, 11.4385s/12 iters), loss = 3.7069
I0428 17:19:50.167277  8468 solver.cpp:237]     Train net output #0: loss = 3.7069 (* 1 = 3.7069 loss)
I0428 17:19:50.167287  8468 sgd_solver.cpp:105] Iteration 1740, lr = 0.00708455
I0428 17:19:55.617883  8468 solver.cpp:218] Iteration 1752 (2.20168 iter/s, 5.45038s/12 iters), loss = 3.70593
I0428 17:19:55.617931  8468 solver.cpp:237]     Train net output #0: loss = 3.70593 (* 1 = 3.70593 loss)
I0428 17:19:55.617941  8468 sgd_solver.cpp:105] Iteration 1752, lr = 0.00706773
I0428 17:20:01.197592  8468 solver.cpp:218] Iteration 1764 (2.15076 iter/s, 5.57943s/12 iters), loss = 3.78703
I0428 17:20:01.197643  8468 solver.cpp:237]     Train net output #0: loss = 3.78703 (* 1 = 3.78703 loss)
I0428 17:20:01.197651  8468 sgd_solver.cpp:105] Iteration 1764, lr = 0.00705094
I0428 17:20:06.643080  8468 solver.cpp:218] Iteration 1776 (2.20377 iter/s, 5.44522s/12 iters), loss = 3.56465
I0428 17:20:06.643126  8468 solver.cpp:237]     Train net output #0: loss = 3.56465 (* 1 = 3.56465 loss)
I0428 17:20:06.643134  8468 sgd_solver.cpp:105] Iteration 1776, lr = 0.0070342
I0428 17:20:12.212182  8468 solver.cpp:218] Iteration 1788 (2.15485 iter/s, 5.56883s/12 iters), loss = 3.6503
I0428 17:20:12.212323  8468 solver.cpp:237]     Train net output #0: loss = 3.6503 (* 1 = 3.6503 loss)
I0428 17:20:12.212333  8468 sgd_solver.cpp:105] Iteration 1788, lr = 0.0070175
I0428 17:20:17.751186  8468 solver.cpp:218] Iteration 1800 (2.1666 iter/s, 5.53864s/12 iters), loss = 3.62357
I0428 17:20:17.751233  8468 solver.cpp:237]     Train net output #0: loss = 3.62357 (* 1 = 3.62357 loss)
I0428 17:20:17.751245  8468 sgd_solver.cpp:105] Iteration 1800, lr = 0.00700084
I0428 17:20:23.348393  8468 solver.cpp:218] Iteration 1812 (2.14403 iter/s, 5.59693s/12 iters), loss = 3.55268
I0428 17:20:23.348436  8468 solver.cpp:237]     Train net output #0: loss = 3.55268 (* 1 = 3.55268 loss)
I0428 17:20:23.348445  8468 sgd_solver.cpp:105] Iteration 1812, lr = 0.00698422
I0428 17:20:26.792870  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:20:28.766366  8468 solver.cpp:218] Iteration 1824 (2.21496 iter/s, 5.41771s/12 iters), loss = 3.73003
I0428 17:20:28.766403  8468 solver.cpp:237]     Train net output #0: loss = 3.73003 (* 1 = 3.73003 loss)
I0428 17:20:28.766413  8468 sgd_solver.cpp:105] Iteration 1824, lr = 0.00696764
I0428 17:20:35.315822  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1836.caffemodel
I0428 17:20:36.930574  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1836.solverstate
I0428 17:20:38.397437  8468 solver.cpp:330] Iteration 1836, Testing net (#0)
I0428 17:20:38.397460  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:20:43.436605  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:20:44.376669  8468 solver.cpp:397]     Test net output #0: accuracy = 0.133578
I0428 17:20:44.376698  8468 solver.cpp:397]     Test net output #1: loss = 3.68746 (* 1 = 3.68746 loss)
I0428 17:20:44.503706  8468 solver.cpp:218] Iteration 1836 (0.76255 iter/s, 15.7367s/12 iters), loss = 3.42551
I0428 17:20:44.503760  8468 solver.cpp:237]     Train net output #0: loss = 3.42551 (* 1 = 3.42551 loss)
I0428 17:20:44.503775  8468 sgd_solver.cpp:105] Iteration 1836, lr = 0.0069511
I0428 17:20:49.095917  8468 solver.cpp:218] Iteration 1848 (2.61326 iter/s, 4.59197s/12 iters), loss = 3.49271
I0428 17:20:49.095959  8468 solver.cpp:237]     Train net output #0: loss = 3.49271 (* 1 = 3.49271 loss)
I0428 17:20:49.095969  8468 sgd_solver.cpp:105] Iteration 1848, lr = 0.00693459
I0428 17:20:54.604562  8468 solver.cpp:218] Iteration 1860 (2.1785 iter/s, 5.50838s/12 iters), loss = 3.66506
I0428 17:20:54.604602  8468 solver.cpp:237]     Train net output #0: loss = 3.66506 (* 1 = 3.66506 loss)
I0428 17:20:54.604611  8468 sgd_solver.cpp:105] Iteration 1860, lr = 0.00691813
I0428 17:21:00.111090  8468 solver.cpp:218] Iteration 1872 (2.17934 iter/s, 5.50626s/12 iters), loss = 3.76481
I0428 17:21:00.111135  8468 solver.cpp:237]     Train net output #0: loss = 3.76481 (* 1 = 3.76481 loss)
I0428 17:21:00.111147  8468 sgd_solver.cpp:105] Iteration 1872, lr = 0.0069017
I0428 17:21:05.815102  8468 solver.cpp:218] Iteration 1884 (2.10388 iter/s, 5.70374s/12 iters), loss = 3.88833
I0428 17:21:05.815143  8468 solver.cpp:237]     Train net output #0: loss = 3.88833 (* 1 = 3.88833 loss)
I0428 17:21:05.815153  8468 sgd_solver.cpp:105] Iteration 1884, lr = 0.00688532
I0428 17:21:11.243142  8468 solver.cpp:218] Iteration 1896 (2.21085 iter/s, 5.42778s/12 iters), loss = 3.56089
I0428 17:21:11.243188  8468 solver.cpp:237]     Train net output #0: loss = 3.56089 (* 1 = 3.56089 loss)
I0428 17:21:11.243198  8468 sgd_solver.cpp:105] Iteration 1896, lr = 0.00686897
I0428 17:21:16.645087  8468 solver.cpp:218] Iteration 1908 (2.22153 iter/s, 5.40168s/12 iters), loss = 3.61566
I0428 17:21:16.645205  8468 solver.cpp:237]     Train net output #0: loss = 3.61566 (* 1 = 3.61566 loss)
I0428 17:21:16.645215  8468 sgd_solver.cpp:105] Iteration 1908, lr = 0.00685266
I0428 17:21:22.009874  8468 solver.cpp:218] Iteration 1920 (2.23695 iter/s, 5.36445s/12 iters), loss = 3.37863
I0428 17:21:22.009918  8468 solver.cpp:237]     Train net output #0: loss = 3.37863 (* 1 = 3.37863 loss)
I0428 17:21:22.009927  8468 sgd_solver.cpp:105] Iteration 1920, lr = 0.00683639
I0428 17:21:22.337713  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:21:27.476420  8468 solver.cpp:218] Iteration 1932 (2.19528 iter/s, 5.46628s/12 iters), loss = 3.24738
I0428 17:21:27.476459  8468 solver.cpp:237]     Train net output #0: loss = 3.24738 (* 1 = 3.24738 loss)
I0428 17:21:27.476469  8468 sgd_solver.cpp:105] Iteration 1932, lr = 0.00682016
I0428 17:21:29.622498  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1938.caffemodel
I0428 17:21:30.820369  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1938.solverstate
I0428 17:21:34.899111  8468 solver.cpp:330] Iteration 1938, Testing net (#0)
I0428 17:21:34.899138  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:21:38.684957  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:21:39.502570  8468 solver.cpp:397]     Test net output #0: accuracy = 0.158088
I0428 17:21:39.502609  8468 solver.cpp:397]     Test net output #1: loss = 3.62667 (* 1 = 3.62667 loss)
I0428 17:21:41.560818  8468 solver.cpp:218] Iteration 1944 (0.852042 iter/s, 14.0838s/12 iters), loss = 3.59444
I0428 17:21:41.560856  8468 solver.cpp:237]     Train net output #0: loss = 3.59444 (* 1 = 3.59444 loss)
I0428 17:21:41.560868  8468 sgd_solver.cpp:105] Iteration 1944, lr = 0.00680397
I0428 17:21:47.033947  8468 solver.cpp:218] Iteration 1956 (2.19264 iter/s, 5.47286s/12 iters), loss = 3.6179
I0428 17:21:47.034158  8468 solver.cpp:237]     Train net output #0: loss = 3.6179 (* 1 = 3.6179 loss)
I0428 17:21:47.034198  8468 sgd_solver.cpp:105] Iteration 1956, lr = 0.00678782
I0428 17:21:52.490379  8468 solver.cpp:218] Iteration 1968 (2.19941 iter/s, 5.45601s/12 iters), loss = 3.60205
I0428 17:21:52.490419  8468 solver.cpp:237]     Train net output #0: loss = 3.60205 (* 1 = 3.60205 loss)
I0428 17:21:52.490428  8468 sgd_solver.cpp:105] Iteration 1968, lr = 0.0067717
I0428 17:21:58.047560  8468 solver.cpp:218] Iteration 1980 (2.15947 iter/s, 5.55691s/12 iters), loss = 3.38457
I0428 17:21:58.047605  8468 solver.cpp:237]     Train net output #0: loss = 3.38457 (* 1 = 3.38457 loss)
I0428 17:21:58.047614  8468 sgd_solver.cpp:105] Iteration 1980, lr = 0.00675562
I0428 17:22:05.468274  8468 solver.cpp:218] Iteration 1992 (1.61717 iter/s, 7.42036s/12 iters), loss = 3.34913
I0428 17:22:05.474328  8468 solver.cpp:237]     Train net output #0: loss = 3.34913 (* 1 = 3.34913 loss)
I0428 17:22:05.474356  8468 sgd_solver.cpp:105] Iteration 1992, lr = 0.00673958
I0428 17:22:12.122825  8468 solver.cpp:218] Iteration 2004 (1.80499 iter/s, 6.64825s/12 iters), loss = 3.43513
I0428 17:22:12.122867  8468 solver.cpp:237]     Train net output #0: loss = 3.43513 (* 1 = 3.43513 loss)
I0428 17:22:12.122876  8468 sgd_solver.cpp:105] Iteration 2004, lr = 0.00672358
I0428 17:22:17.485971  8468 solver.cpp:218] Iteration 2016 (2.2376 iter/s, 5.36288s/12 iters), loss = 3.26304
I0428 17:22:17.486358  8468 solver.cpp:237]     Train net output #0: loss = 3.26304 (* 1 = 3.26304 loss)
I0428 17:22:17.486368  8468 sgd_solver.cpp:105] Iteration 2016, lr = 0.00670762
I0428 17:22:20.258517  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:22:22.939996  8468 solver.cpp:218] Iteration 2028 (2.20046 iter/s, 5.45342s/12 iters), loss = 3.25339
I0428 17:22:22.940040  8468 solver.cpp:237]     Train net output #0: loss = 3.25339 (* 1 = 3.25339 loss)
I0428 17:22:22.940049  8468 sgd_solver.cpp:105] Iteration 2028, lr = 0.00669169
I0428 17:22:27.796748  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2040.caffemodel
I0428 17:22:28.403828  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2040.solverstate
I0428 17:22:28.849102  8468 solver.cpp:330] Iteration 2040, Testing net (#0)
I0428 17:22:28.849123  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:22:32.491737  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:22:33.354507  8468 solver.cpp:397]     Test net output #0: accuracy = 0.158701
I0428 17:22:33.354555  8468 solver.cpp:397]     Test net output #1: loss = 3.54367 (* 1 = 3.54367 loss)
I0428 17:22:33.484006  8468 solver.cpp:218] Iteration 2040 (1.13814 iter/s, 10.5435s/12 iters), loss = 3.42819
I0428 17:22:33.484062  8468 solver.cpp:237]     Train net output #0: loss = 3.42819 (* 1 = 3.42819 loss)
I0428 17:22:33.484073  8468 sgd_solver.cpp:105] Iteration 2040, lr = 0.00667581
I0428 17:22:38.008842  8468 solver.cpp:218] Iteration 2052 (2.65217 iter/s, 4.5246s/12 iters), loss = 3.31466
I0428 17:22:38.008877  8468 solver.cpp:237]     Train net output #0: loss = 3.31466 (* 1 = 3.31466 loss)
I0428 17:22:38.008886  8468 sgd_solver.cpp:105] Iteration 2052, lr = 0.00665996
I0428 17:22:38.375455  8468 blocking_queue.cpp:49] Waiting for data
I0428 17:22:43.405364  8468 solver.cpp:218] Iteration 2064 (2.22376 iter/s, 5.39626s/12 iters), loss = 3.29423
I0428 17:22:43.405409  8468 solver.cpp:237]     Train net output #0: loss = 3.29423 (* 1 = 3.29423 loss)
I0428 17:22:43.405418  8468 sgd_solver.cpp:105] Iteration 2064, lr = 0.00664414
I0428 17:22:48.797094  8468 solver.cpp:218] Iteration 2076 (2.22574 iter/s, 5.39146s/12 iters), loss = 3.26617
I0428 17:22:48.797247  8468 solver.cpp:237]     Train net output #0: loss = 3.26617 (* 1 = 3.26617 loss)
I0428 17:22:48.797259  8468 sgd_solver.cpp:105] Iteration 2076, lr = 0.00662837
I0428 17:22:54.267096  8468 solver.cpp:218] Iteration 2088 (2.19393 iter/s, 5.46962s/12 iters), loss = 3.14062
I0428 17:22:54.267138  8468 solver.cpp:237]     Train net output #0: loss = 3.14062 (* 1 = 3.14062 loss)
I0428 17:22:54.267148  8468 sgd_solver.cpp:105] Iteration 2088, lr = 0.00661263
I0428 17:23:02.115865  8468 solver.cpp:218] Iteration 2100 (1.52897 iter/s, 7.8484s/12 iters), loss = 3.29223
I0428 17:23:02.128559  8468 solver.cpp:237]     Train net output #0: loss = 3.29223 (* 1 = 3.29223 loss)
I0428 17:23:02.128585  8468 sgd_solver.cpp:105] Iteration 2100, lr = 0.00659693
I0428 17:23:10.109184  8468 solver.cpp:218] Iteration 2112 (1.5037 iter/s, 7.98032s/12 iters), loss = 3.22899
I0428 17:23:10.109241  8468 solver.cpp:237]     Train net output #0: loss = 3.22899 (* 1 = 3.22899 loss)
I0428 17:23:10.109252  8468 sgd_solver.cpp:105] Iteration 2112, lr = 0.00658127
I0428 17:23:16.518862  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:23:16.987742  8468 solver.cpp:218] Iteration 2124 (1.74464 iter/s, 6.87822s/12 iters), loss = 3.31118
I0428 17:23:16.987807  8468 solver.cpp:237]     Train net output #0: loss = 3.31118 (* 1 = 3.31118 loss)
I0428 17:23:16.987818  8468 sgd_solver.cpp:105] Iteration 2124, lr = 0.00656564
I0428 17:23:23.642092  8468 solver.cpp:218] Iteration 2136 (1.80342 iter/s, 6.65401s/12 iters), loss = 3.45711
I0428 17:23:23.642226  8468 solver.cpp:237]     Train net output #0: loss = 3.45711 (* 1 = 3.45711 loss)
I0428 17:23:23.642239  8468 sgd_solver.cpp:105] Iteration 2136, lr = 0.00655006
I0428 17:23:26.469543  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2142.caffemodel
I0428 17:23:27.427289  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2142.solverstate
I0428 17:23:30.150591  8468 solver.cpp:330] Iteration 2142, Testing net (#0)
I0428 17:23:30.150611  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:23:34.997936  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:23:36.237815  8468 solver.cpp:397]     Test net output #0: accuracy = 0.18076
I0428 17:23:36.237855  8468 solver.cpp:397]     Test net output #1: loss = 3.40808 (* 1 = 3.40808 loss)
I0428 17:23:38.633059  8468 solver.cpp:218] Iteration 2148 (0.800521 iter/s, 14.9902s/12 iters), loss = 3.20639
I0428 17:23:38.633114  8468 solver.cpp:237]     Train net output #0: loss = 3.20639 (* 1 = 3.20639 loss)
I0428 17:23:38.633126  8468 sgd_solver.cpp:105] Iteration 2148, lr = 0.00653451
I0428 17:23:45.302579  8468 solver.cpp:218] Iteration 2160 (1.79932 iter/s, 6.66919s/12 iters), loss = 3.18811
I0428 17:23:45.302635  8468 solver.cpp:237]     Train net output #0: loss = 3.18811 (* 1 = 3.18811 loss)
I0428 17:23:45.302647  8468 sgd_solver.cpp:105] Iteration 2160, lr = 0.00651899
I0428 17:23:52.206758  8468 solver.cpp:218] Iteration 2172 (1.73816 iter/s, 6.90384s/12 iters), loss = 3.33188
I0428 17:23:52.206812  8468 solver.cpp:237]     Train net output #0: loss = 3.33188 (* 1 = 3.33188 loss)
I0428 17:23:52.206825  8468 sgd_solver.cpp:105] Iteration 2172, lr = 0.00650351
I0428 17:23:59.002820  8468 solver.cpp:218] Iteration 2184 (1.76582 iter/s, 6.79572s/12 iters), loss = 3.23883
I0428 17:23:59.003002  8468 solver.cpp:237]     Train net output #0: loss = 3.23883 (* 1 = 3.23883 loss)
I0428 17:23:59.003015  8468 sgd_solver.cpp:105] Iteration 2184, lr = 0.00648807
I0428 17:24:05.567088  8468 solver.cpp:218] Iteration 2196 (1.8282 iter/s, 6.56382s/12 iters), loss = 3.18041
I0428 17:24:05.567135  8468 solver.cpp:237]     Train net output #0: loss = 3.18041 (* 1 = 3.18041 loss)
I0428 17:24:05.567144  8468 sgd_solver.cpp:105] Iteration 2196, lr = 0.00647267
I0428 17:24:12.237692  8468 solver.cpp:218] Iteration 2208 (1.79902 iter/s, 6.67028s/12 iters), loss = 3.19836
I0428 17:24:12.237744  8468 solver.cpp:237]     Train net output #0: loss = 3.19836 (* 1 = 3.19836 loss)
I0428 17:24:12.237754  8468 sgd_solver.cpp:105] Iteration 2208, lr = 0.0064573
I0428 17:24:19.242553  8468 solver.cpp:218] Iteration 2220 (1.71318 iter/s, 7.00452s/12 iters), loss = 3.12636
I0428 17:24:19.242609  8468 solver.cpp:237]     Train net output #0: loss = 3.12636 (* 1 = 3.12636 loss)
I0428 17:24:19.242620  8468 sgd_solver.cpp:105] Iteration 2220, lr = 0.00644197
I0428 17:24:21.713809  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:24:26.098119  8468 solver.cpp:218] Iteration 2232 (1.75049 iter/s, 6.85523s/12 iters), loss = 2.76746
I0428 17:24:26.098176  8468 solver.cpp:237]     Train net output #0: loss = 2.76746 (* 1 = 2.76746 loss)
I0428 17:24:26.098188  8468 sgd_solver.cpp:105] Iteration 2232, lr = 0.00642668
I0428 17:24:31.050856  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2244.caffemodel
I0428 17:24:31.697371  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2244.solverstate
I0428 17:24:32.735823  8468 solver.cpp:330] Iteration 2244, Testing net (#0)
I0428 17:24:32.735842  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:24:36.486995  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:24:37.568284  8468 solver.cpp:397]     Test net output #0: accuracy = 0.178309
I0428 17:24:37.568313  8468 solver.cpp:397]     Test net output #1: loss = 3.36866 (* 1 = 3.36866 loss)
I0428 17:24:37.699061  8468 solver.cpp:218] Iteration 2244 (1.03445 iter/s, 11.6004s/12 iters), loss = 2.8291
I0428 17:24:37.699115  8468 solver.cpp:237]     Train net output #0: loss = 2.8291 (* 1 = 2.8291 loss)
I0428 17:24:37.699126  8468 sgd_solver.cpp:105] Iteration 2244, lr = 0.00641142
I0428 17:24:42.226796  8468 solver.cpp:218] Iteration 2256 (2.65047 iter/s, 4.52749s/12 iters), loss = 3.10117
I0428 17:24:42.226840  8468 solver.cpp:237]     Train net output #0: loss = 3.10117 (* 1 = 3.10117 loss)
I0428 17:24:42.226850  8468 sgd_solver.cpp:105] Iteration 2256, lr = 0.0063962
I0428 17:24:47.584262  8468 solver.cpp:218] Iteration 2268 (2.23998 iter/s, 5.3572s/12 iters), loss = 3.02105
I0428 17:24:47.584306  8468 solver.cpp:237]     Train net output #0: loss = 3.02105 (* 1 = 3.02105 loss)
I0428 17:24:47.584317  8468 sgd_solver.cpp:105] Iteration 2268, lr = 0.00638101
I0428 17:24:53.014900  8468 solver.cpp:218] Iteration 2280 (2.2098 iter/s, 5.43037s/12 iters), loss = 3.10771
I0428 17:24:53.014950  8468 solver.cpp:237]     Train net output #0: loss = 3.10771 (* 1 = 3.10771 loss)
I0428 17:24:53.014959  8468 sgd_solver.cpp:105] Iteration 2280, lr = 0.00636586
I0428 17:25:00.532564  8468 solver.cpp:218] Iteration 2292 (1.59967 iter/s, 7.50156s/12 iters), loss = 3.07415
I0428 17:25:00.532620  8468 solver.cpp:237]     Train net output #0: loss = 3.07415 (* 1 = 3.07415 loss)
I0428 17:25:00.532632  8468 sgd_solver.cpp:105] Iteration 2292, lr = 0.00635075
I0428 17:25:10.352058  8468 solver.cpp:218] Iteration 2304 (1.22212 iter/s, 9.81904s/12 iters), loss = 2.90217
I0428 17:25:10.352232  8468 solver.cpp:237]     Train net output #0: loss = 2.90217 (* 1 = 2.90217 loss)
I0428 17:25:10.352246  8468 sgd_solver.cpp:105] Iteration 2304, lr = 0.00633567
I0428 17:25:18.560339  8468 solver.cpp:218] Iteration 2316 (1.46203 iter/s, 8.20777s/12 iters), loss = 3.05273
I0428 17:25:18.560397  8468 solver.cpp:237]     Train net output #0: loss = 3.05273 (* 1 = 3.05273 loss)
I0428 17:25:18.560410  8468 sgd_solver.cpp:105] Iteration 2316, lr = 0.00632063
I0428 17:25:23.585989  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:25:25.035759  8468 solver.cpp:218] Iteration 2328 (1.85326 iter/s, 6.47509s/12 iters), loss = 2.94265
I0428 17:25:25.035815  8468 solver.cpp:237]     Train net output #0: loss = 2.94265 (* 1 = 2.94265 loss)
I0428 17:25:25.035826  8468 sgd_solver.cpp:105] Iteration 2328, lr = 0.00630562
I0428 17:25:31.614753  8468 solver.cpp:218] Iteration 2340 (1.82408 iter/s, 6.57867s/12 iters), loss = 2.7241
I0428 17:25:31.620810  8468 solver.cpp:237]     Train net output #0: loss = 2.7241 (* 1 = 2.7241 loss)
I0428 17:25:31.620836  8468 sgd_solver.cpp:105] Iteration 2340, lr = 0.00629065
I0428 17:25:34.471979  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2346.caffemodel
I0428 17:25:36.997349  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2346.solverstate
I0428 17:25:39.109495  8468 solver.cpp:330] Iteration 2346, Testing net (#0)
I0428 17:25:39.109520  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:25:43.829397  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:25:45.167310  8468 solver.cpp:397]     Test net output #0: accuracy = 0.205882
I0428 17:25:45.167351  8468 solver.cpp:397]     Test net output #1: loss = 3.23485 (* 1 = 3.23485 loss)
I0428 17:25:47.423992  8468 solver.cpp:218] Iteration 2352 (0.75937 iter/s, 15.8026s/12 iters), loss = 2.77624
I0428 17:25:47.424058  8468 solver.cpp:237]     Train net output #0: loss = 2.77624 (* 1 = 2.77624 loss)
I0428 17:25:47.424072  8468 sgd_solver.cpp:105] Iteration 2352, lr = 0.00627571
I0428 17:25:54.164466  8468 solver.cpp:218] Iteration 2364 (1.78038 iter/s, 6.74013s/12 iters), loss = 2.92037
I0428 17:25:54.164561  8468 solver.cpp:237]     Train net output #0: loss = 2.92037 (* 1 = 2.92037 loss)
I0428 17:25:54.164574  8468 sgd_solver.cpp:105] Iteration 2364, lr = 0.00626081
I0428 17:26:00.751896  8468 solver.cpp:218] Iteration 2376 (1.82175 iter/s, 6.58706s/12 iters), loss = 3.1406
I0428 17:26:00.751941  8468 solver.cpp:237]     Train net output #0: loss = 3.1406 (* 1 = 3.1406 loss)
I0428 17:26:00.751950  8468 sgd_solver.cpp:105] Iteration 2376, lr = 0.00624595
I0428 17:26:07.708673  8468 solver.cpp:218] Iteration 2388 (1.72502 iter/s, 6.95644s/12 iters), loss = 2.77269
I0428 17:26:07.708731  8468 solver.cpp:237]     Train net output #0: loss = 2.77269 (* 1 = 2.77269 loss)
I0428 17:26:07.708745  8468 sgd_solver.cpp:105] Iteration 2388, lr = 0.00623112
I0428 17:26:14.592926  8468 solver.cpp:218] Iteration 2400 (1.7432 iter/s, 6.88391s/12 iters), loss = 2.94094
I0428 17:26:14.593068  8468 solver.cpp:237]     Train net output #0: loss = 2.94094 (* 1 = 2.94094 loss)
I0428 17:26:14.593080  8468 sgd_solver.cpp:105] Iteration 2400, lr = 0.00621633
I0428 17:26:21.294103  8468 solver.cpp:218] Iteration 2412 (1.79084 iter/s, 6.70076s/12 iters), loss = 3.05969
I0428 17:26:21.294153  8468 solver.cpp:237]     Train net output #0: loss = 3.05969 (* 1 = 3.05969 loss)
I0428 17:26:21.294165  8468 sgd_solver.cpp:105] Iteration 2412, lr = 0.00620157
I0428 17:26:27.080397  8468 solver.cpp:218] Iteration 2424 (2.07397 iter/s, 5.78601s/12 iters), loss = 3.05228
I0428 17:26:27.080436  8468 solver.cpp:237]     Train net output #0: loss = 3.05228 (* 1 = 3.05228 loss)
I0428 17:26:27.080446  8468 sgd_solver.cpp:105] Iteration 2424, lr = 0.00618684
I0428 17:26:28.201804  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:26:32.464663  8468 solver.cpp:218] Iteration 2436 (2.22883 iter/s, 5.384s/12 iters), loss = 3.15231
I0428 17:26:32.464709  8468 solver.cpp:237]     Train net output #0: loss = 3.15231 (* 1 = 3.15231 loss)
I0428 17:26:32.464718  8468 sgd_solver.cpp:105] Iteration 2436, lr = 0.00617215
I0428 17:26:37.406975  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2448.caffemodel
I0428 17:26:38.712987  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2448.solverstate
I0428 17:26:39.156949  8468 solver.cpp:330] Iteration 2448, Testing net (#0)
I0428 17:26:39.156972  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:26:42.627274  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:26:43.651152  8468 solver.cpp:397]     Test net output #0: accuracy = 0.208946
I0428 17:26:43.651185  8468 solver.cpp:397]     Test net output #1: loss = 3.23616 (* 1 = 3.23616 loss)
I0428 17:26:43.781888  8468 solver.cpp:218] Iteration 2448 (1.06038 iter/s, 11.3167s/12 iters), loss = 2.88839
I0428 17:26:43.781932  8468 solver.cpp:237]     Train net output #0: loss = 2.88839 (* 1 = 2.88839 loss)
I0428 17:26:43.781942  8468 sgd_solver.cpp:105] Iteration 2448, lr = 0.0061575
I0428 17:26:50.217931  8468 solver.cpp:218] Iteration 2460 (1.86681 iter/s, 6.42809s/12 iters), loss = 2.68653
I0428 17:26:50.220896  8468 solver.cpp:237]     Train net output #0: loss = 2.68653 (* 1 = 2.68653 loss)
I0428 17:26:50.220916  8468 sgd_solver.cpp:105] Iteration 2460, lr = 0.00614288
I0428 17:27:00.145754  8468 solver.cpp:218] Iteration 2472 (1.20963 iter/s, 9.9204s/12 iters), loss = 2.91339
I0428 17:27:00.153380  8468 solver.cpp:237]     Train net output #0: loss = 2.91339 (* 1 = 2.91339 loss)
I0428 17:27:00.153407  8468 sgd_solver.cpp:105] Iteration 2472, lr = 0.0061283
I0428 17:27:10.186146  8468 solver.cpp:218] Iteration 2484 (1.19613 iter/s, 10.0324s/12 iters), loss = 2.71023
I0428 17:27:10.186205  8468 solver.cpp:237]     Train net output #0: loss = 2.71023 (* 1 = 2.71023 loss)
I0428 17:27:10.186218  8468 sgd_solver.cpp:105] Iteration 2484, lr = 0.00611375
I0428 17:27:17.342424  8468 solver.cpp:218] Iteration 2496 (1.67694 iter/s, 7.15591s/12 iters), loss = 2.90706
I0428 17:27:17.348462  8468 solver.cpp:237]     Train net output #0: loss = 2.90706 (* 1 = 2.90706 loss)
I0428 17:27:17.348529  8468 sgd_solver.cpp:105] Iteration 2496, lr = 0.00609923
I0428 17:27:24.278096  8468 solver.cpp:218] Iteration 2508 (1.73176 iter/s, 6.92937s/12 iters), loss = 2.83404
I0428 17:27:24.278198  8468 solver.cpp:237]     Train net output #0: loss = 2.83404 (* 1 = 2.83404 loss)
I0428 17:27:24.278210  8468 sgd_solver.cpp:105] Iteration 2508, lr = 0.00608475
I0428 17:27:30.417320  8468 solver.cpp:218] Iteration 2520 (1.95476 iter/s, 6.13887s/12 iters), loss = 2.7811
I0428 17:27:30.417376  8468 solver.cpp:237]     Train net output #0: loss = 2.7811 (* 1 = 2.7811 loss)
I0428 17:27:30.417387  8468 sgd_solver.cpp:105] Iteration 2520, lr = 0.0060703
I0428 17:27:34.332417  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:27:36.810390  8468 solver.cpp:218] Iteration 2532 (1.87713 iter/s, 6.39275s/12 iters), loss = 2.72326
I0428 17:27:36.810439  8468 solver.cpp:237]     Train net output #0: loss = 2.72326 (* 1 = 2.72326 loss)
I0428 17:27:36.810451  8468 sgd_solver.cpp:105] Iteration 2532, lr = 0.00605589
I0428 17:27:43.638952  8468 solver.cpp:218] Iteration 2544 (1.75741 iter/s, 6.82822s/12 iters), loss = 2.58473
I0428 17:27:43.639011  8468 solver.cpp:237]     Train net output #0: loss = 2.58473 (* 1 = 2.58473 loss)
I0428 17:27:43.639024  8468 sgd_solver.cpp:105] Iteration 2544, lr = 0.00604151
I0428 17:27:46.333825  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2550.caffemodel
I0428 17:27:47.037818  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2550.solverstate
I0428 17:27:47.591413  8468 solver.cpp:330] Iteration 2550, Testing net (#0)
I0428 17:27:47.591437  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:27:52.429067  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:27:54.022825  8468 solver.cpp:397]     Test net output #0: accuracy = 0.20527
I0428 17:27:54.022864  8468 solver.cpp:397]     Test net output #1: loss = 3.23785 (* 1 = 3.23785 loss)
I0428 17:27:56.434743  8468 solver.cpp:218] Iteration 2556 (0.93785 iter/s, 12.7952s/12 iters), loss = 2.61431
I0428 17:27:56.437618  8468 solver.cpp:237]     Train net output #0: loss = 2.61431 (* 1 = 2.61431 loss)
I0428 17:27:56.437636  8468 sgd_solver.cpp:105] Iteration 2556, lr = 0.00602717
I0428 17:28:03.152562  8468 solver.cpp:218] Iteration 2568 (1.78752 iter/s, 6.7132s/12 iters), loss = 2.69825
I0428 17:28:03.152614  8468 solver.cpp:237]     Train net output #0: loss = 2.69825 (* 1 = 2.69825 loss)
I0428 17:28:03.152628  8468 sgd_solver.cpp:105] Iteration 2568, lr = 0.00601286
I0428 17:28:10.056713  8468 solver.cpp:218] Iteration 2580 (1.73817 iter/s, 6.90381s/12 iters), loss = 2.7156
I0428 17:28:10.056767  8468 solver.cpp:237]     Train net output #0: loss = 2.7156 (* 1 = 2.7156 loss)
I0428 17:28:10.056780  8468 sgd_solver.cpp:105] Iteration 2580, lr = 0.00599858
I0428 17:28:16.627660  8468 solver.cpp:218] Iteration 2592 (1.82631 iter/s, 6.57062s/12 iters), loss = 2.91684
I0428 17:28:16.627712  8468 solver.cpp:237]     Train net output #0: loss = 2.91684 (* 1 = 2.91684 loss)
I0428 17:28:16.627722  8468 sgd_solver.cpp:105] Iteration 2592, lr = 0.00598434
I0428 17:28:23.010004  8468 solver.cpp:218] Iteration 2604 (1.88028 iter/s, 6.38203s/12 iters), loss = 2.64107
I0428 17:28:23.010048  8468 solver.cpp:237]     Train net output #0: loss = 2.64107 (* 1 = 2.64107 loss)
I0428 17:28:23.010058  8468 sgd_solver.cpp:105] Iteration 2604, lr = 0.00597013
I0428 17:28:28.383689  8468 solver.cpp:218] Iteration 2616 (2.23322 iter/s, 5.37341s/12 iters), loss = 2.69201
I0428 17:28:28.383903  8468 solver.cpp:237]     Train net output #0: loss = 2.69201 (* 1 = 2.69201 loss)
I0428 17:28:28.383913  8468 sgd_solver.cpp:105] Iteration 2616, lr = 0.00595596
I0428 17:28:33.761811  8468 solver.cpp:218] Iteration 2628 (2.23144 iter/s, 5.37768s/12 iters), loss = 2.61098
I0428 17:28:33.761852  8468 solver.cpp:237]     Train net output #0: loss = 2.61098 (* 1 = 2.61098 loss)
I0428 17:28:33.761860  8468 sgd_solver.cpp:105] Iteration 2628, lr = 0.00594182
I0428 17:28:34.240191  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:28:39.109138  8468 solver.cpp:218] Iteration 2640 (2.24423 iter/s, 5.34706s/12 iters), loss = 2.47148
I0428 17:28:39.109184  8468 solver.cpp:237]     Train net output #0: loss = 2.47148 (* 1 = 2.47148 loss)
I0428 17:28:39.109194  8468 sgd_solver.cpp:105] Iteration 2640, lr = 0.00592771
I0428 17:28:43.946854  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2652.caffemodel
I0428 17:28:44.542898  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2652.solverstate
I0428 17:28:44.978935  8468 solver.cpp:330] Iteration 2652, Testing net (#0)
I0428 17:28:44.978952  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:28:48.490907  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:28:49.590055  8468 solver.cpp:397]     Test net output #0: accuracy = 0.235907
I0428 17:28:49.590083  8468 solver.cpp:397]     Test net output #1: loss = 3.02536 (* 1 = 3.02536 loss)
I0428 17:28:49.720165  8468 solver.cpp:218] Iteration 2652 (1.13095 iter/s, 10.6105s/12 iters), loss = 2.61757
I0428 17:28:49.720222  8468 solver.cpp:237]     Train net output #0: loss = 2.61757 (* 1 = 2.61757 loss)
I0428 17:28:49.720240  8468 sgd_solver.cpp:105] Iteration 2652, lr = 0.00591364
I0428 17:28:54.211241  8468 solver.cpp:218] Iteration 2664 (2.67211 iter/s, 4.49083s/12 iters), loss = 2.54936
I0428 17:28:54.211282  8468 solver.cpp:237]     Train net output #0: loss = 2.54936 (* 1 = 2.54936 loss)
I0428 17:28:54.211292  8468 sgd_solver.cpp:105] Iteration 2664, lr = 0.0058996
I0428 17:28:59.713800  8468 solver.cpp:218] Iteration 2676 (2.18091 iter/s, 5.50229s/12 iters), loss = 2.64168
I0428 17:28:59.713932  8468 solver.cpp:237]     Train net output #0: loss = 2.64168 (* 1 = 2.64168 loss)
I0428 17:28:59.713943  8468 sgd_solver.cpp:105] Iteration 2676, lr = 0.00588559
I0428 17:29:05.071377  8468 solver.cpp:218] Iteration 2688 (2.23997 iter/s, 5.35721s/12 iters), loss = 2.73432
I0428 17:29:05.071434  8468 solver.cpp:237]     Train net output #0: loss = 2.73432 (* 1 = 2.73432 loss)
I0428 17:29:05.071446  8468 sgd_solver.cpp:105] Iteration 2688, lr = 0.00587162
I0428 17:29:10.480386  8468 solver.cpp:218] Iteration 2700 (2.21864 iter/s, 5.40873s/12 iters), loss = 2.32579
I0428 17:29:10.480438  8468 solver.cpp:237]     Train net output #0: loss = 2.32579 (* 1 = 2.32579 loss)
I0428 17:29:10.480449  8468 sgd_solver.cpp:105] Iteration 2700, lr = 0.00585768
I0428 17:29:15.919517  8468 solver.cpp:218] Iteration 2712 (2.20635 iter/s, 5.43885s/12 iters), loss = 2.66187
I0428 17:29:15.919557  8468 solver.cpp:237]     Train net output #0: loss = 2.66187 (* 1 = 2.66187 loss)
I0428 17:29:15.919566  8468 sgd_solver.cpp:105] Iteration 2712, lr = 0.00584377
I0428 17:29:21.354594  8468 solver.cpp:218] Iteration 2724 (2.20799 iter/s, 5.43481s/12 iters), loss = 2.68519
I0428 17:29:21.354643  8468 solver.cpp:237]     Train net output #0: loss = 2.68519 (* 1 = 2.68519 loss)
I0428 17:29:21.354656  8468 sgd_solver.cpp:105] Iteration 2724, lr = 0.0058299
I0428 17:29:24.229476  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:29:26.982707  8468 solver.cpp:218] Iteration 2736 (2.13226 iter/s, 5.62783s/12 iters), loss = 2.68977
I0428 17:29:26.982748  8468 solver.cpp:237]     Train net output #0: loss = 2.68977 (* 1 = 2.68977 loss)
I0428 17:29:26.982756  8468 sgd_solver.cpp:105] Iteration 2736, lr = 0.00581605
I0428 17:29:32.455777  8468 solver.cpp:218] Iteration 2748 (2.19266 iter/s, 5.4728s/12 iters), loss = 2.75542
I0428 17:29:32.455871  8468 solver.cpp:237]     Train net output #0: loss = 2.75542 (* 1 = 2.75542 loss)
I0428 17:29:32.455880  8468 sgd_solver.cpp:105] Iteration 2748, lr = 0.00580225
I0428 17:29:34.649425  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2754.caffemodel
I0428 17:29:35.737736  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2754.solverstate
I0428 17:29:37.232888  8468 solver.cpp:330] Iteration 2754, Testing net (#0)
I0428 17:29:37.232906  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:29:40.197444  8468 blocking_queue.cpp:49] Waiting for data
I0428 17:29:40.721248  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:29:41.867516  8468 solver.cpp:397]     Test net output #0: accuracy = 0.253064
I0428 17:29:41.867565  8468 solver.cpp:397]     Test net output #1: loss = 3.01773 (* 1 = 3.01773 loss)
I0428 17:29:43.957546  8468 solver.cpp:218] Iteration 2760 (1.04337 iter/s, 11.5012s/12 iters), loss = 2.28683
I0428 17:29:43.957583  8468 solver.cpp:237]     Train net output #0: loss = 2.28683 (* 1 = 2.28683 loss)
I0428 17:29:43.957594  8468 sgd_solver.cpp:105] Iteration 2760, lr = 0.00578847
I0428 17:29:49.358738  8468 solver.cpp:218] Iteration 2772 (2.22184 iter/s, 5.40092s/12 iters), loss = 2.43644
I0428 17:29:49.358783  8468 solver.cpp:237]     Train net output #0: loss = 2.43644 (* 1 = 2.43644 loss)
I0428 17:29:49.358791  8468 sgd_solver.cpp:105] Iteration 2772, lr = 0.00577473
I0428 17:29:54.810688  8468 solver.cpp:218] Iteration 2784 (2.20116 iter/s, 5.45168s/12 iters), loss = 2.77656
I0428 17:29:54.810724  8468 solver.cpp:237]     Train net output #0: loss = 2.77656 (* 1 = 2.77656 loss)
I0428 17:29:54.810732  8468 sgd_solver.cpp:105] Iteration 2784, lr = 0.00576102
I0428 17:30:00.214377  8468 solver.cpp:218] Iteration 2796 (2.22082 iter/s, 5.40342s/12 iters), loss = 2.31733
I0428 17:30:00.214424  8468 solver.cpp:237]     Train net output #0: loss = 2.31733 (* 1 = 2.31733 loss)
I0428 17:30:00.214434  8468 sgd_solver.cpp:105] Iteration 2796, lr = 0.00574734
I0428 17:30:05.667865  8468 solver.cpp:218] Iteration 2808 (2.20054 iter/s, 5.45321s/12 iters), loss = 2.39262
I0428 17:30:05.668041  8468 solver.cpp:237]     Train net output #0: loss = 2.39262 (* 1 = 2.39262 loss)
I0428 17:30:05.668056  8468 sgd_solver.cpp:105] Iteration 2808, lr = 0.00573369
I0428 17:30:11.076508  8468 solver.cpp:218] Iteration 2820 (2.21884 iter/s, 5.40824s/12 iters), loss = 2.153
I0428 17:30:11.076545  8468 solver.cpp:237]     Train net output #0: loss = 2.153 (* 1 = 2.153 loss)
I0428 17:30:11.076555  8468 sgd_solver.cpp:105] Iteration 2820, lr = 0.00572008
I0428 17:30:16.225950  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:30:16.554339  8468 solver.cpp:218] Iteration 2832 (2.19075 iter/s, 5.47756s/12 iters), loss = 2.3914
I0428 17:30:16.554375  8468 solver.cpp:237]     Train net output #0: loss = 2.3914 (* 1 = 2.3914 loss)
I0428 17:30:16.554384  8468 sgd_solver.cpp:105] Iteration 2832, lr = 0.0057065
I0428 17:30:21.946550  8468 solver.cpp:218] Iteration 2844 (2.22554 iter/s, 5.39195s/12 iters), loss = 2.44721
I0428 17:30:21.946590  8468 solver.cpp:237]     Train net output #0: loss = 2.44721 (* 1 = 2.44721 loss)
I0428 17:30:21.946601  8468 sgd_solver.cpp:105] Iteration 2844, lr = 0.00569295
I0428 17:30:26.892856  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2856.caffemodel
I0428 17:30:27.854924  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2856.solverstate
I0428 17:30:29.249331  8468 solver.cpp:330] Iteration 2856, Testing net (#0)
I0428 17:30:29.249362  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:30:32.746642  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:30:33.927736  8468 solver.cpp:397]     Test net output #0: accuracy = 0.270833
I0428 17:30:33.927770  8468 solver.cpp:397]     Test net output #1: loss = 3.00052 (* 1 = 3.00052 loss)
I0428 17:30:34.056562  8468 solver.cpp:218] Iteration 2856 (0.990959 iter/s, 12.1095s/12 iters), loss = 2.41214
I0428 17:30:34.056618  8468 solver.cpp:237]     Train net output #0: loss = 2.41214 (* 1 = 2.41214 loss)
I0428 17:30:34.056635  8468 sgd_solver.cpp:105] Iteration 2856, lr = 0.00567944
I0428 17:30:38.563678  8468 solver.cpp:218] Iteration 2868 (2.6626 iter/s, 4.50688s/12 iters), loss = 2.33902
I0428 17:30:38.563793  8468 solver.cpp:237]     Train net output #0: loss = 2.33902 (* 1 = 2.33902 loss)
I0428 17:30:38.563805  8468 sgd_solver.cpp:105] Iteration 2868, lr = 0.00566595
I0428 17:30:43.932629  8468 solver.cpp:218] Iteration 2880 (2.23521 iter/s, 5.36861s/12 iters), loss = 2.64347
I0428 17:30:43.932672  8468 solver.cpp:237]     Train net output #0: loss = 2.64347 (* 1 = 2.64347 loss)
I0428 17:30:43.932682  8468 sgd_solver.cpp:105] Iteration 2880, lr = 0.0056525
I0428 17:30:49.326314  8468 solver.cpp:218] Iteration 2892 (2.22493 iter/s, 5.39342s/12 iters), loss = 2.51994
I0428 17:30:49.326354  8468 solver.cpp:237]     Train net output #0: loss = 2.51994 (* 1 = 2.51994 loss)
I0428 17:30:49.326364  8468 sgd_solver.cpp:105] Iteration 2892, lr = 0.00563908
I0428 17:30:54.808970  8468 solver.cpp:218] Iteration 2904 (2.18883 iter/s, 5.48239s/12 iters), loss = 2.11426
I0428 17:30:54.809020  8468 solver.cpp:237]     Train net output #0: loss = 2.11426 (* 1 = 2.11426 loss)
I0428 17:30:54.809028  8468 sgd_solver.cpp:105] Iteration 2904, lr = 0.00562569
I0428 17:31:00.345371  8468 solver.cpp:218] Iteration 2916 (2.16758 iter/s, 5.53612s/12 iters), loss = 2.52598
I0428 17:31:00.345413  8468 solver.cpp:237]     Train net output #0: loss = 2.52598 (* 1 = 2.52598 loss)
I0428 17:31:00.345423  8468 sgd_solver.cpp:105] Iteration 2916, lr = 0.00561233
I0428 17:31:05.957438  8468 solver.cpp:218] Iteration 2928 (2.13836 iter/s, 5.61179s/12 iters), loss = 2.24003
I0428 17:31:05.957484  8468 solver.cpp:237]     Train net output #0: loss = 2.24003 (* 1 = 2.24003 loss)
I0428 17:31:05.957495  8468 sgd_solver.cpp:105] Iteration 2928, lr = 0.00559901
I0428 17:31:08.204672  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:31:11.713191  8468 solver.cpp:218] Iteration 2940 (2.08498 iter/s, 5.75546s/12 iters), loss = 1.93968
I0428 17:31:11.713368  8468 solver.cpp:237]     Train net output #0: loss = 1.93968 (* 1 = 1.93968 loss)
I0428 17:31:11.713380  8468 sgd_solver.cpp:105] Iteration 2940, lr = 0.00558572
I0428 17:31:17.200999  8468 solver.cpp:218] Iteration 2952 (2.18683 iter/s, 5.4874s/12 iters), loss = 1.96517
I0428 17:31:17.201050  8468 solver.cpp:237]     Train net output #0: loss = 1.96517 (* 1 = 1.96517 loss)
I0428 17:31:17.201061  8468 sgd_solver.cpp:105] Iteration 2952, lr = 0.00557245
I0428 17:31:19.383787  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2958.caffemodel
I0428 17:31:19.990890  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2958.solverstate
I0428 17:31:20.419834  8468 solver.cpp:330] Iteration 2958, Testing net (#0)
I0428 17:31:20.419852  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:31:23.953208  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:31:25.309284  8468 solver.cpp:397]     Test net output #0: accuracy = 0.276961
I0428 17:31:25.309322  8468 solver.cpp:397]     Test net output #1: loss = 2.94561 (* 1 = 2.94561 loss)
I0428 17:31:27.503227  8468 solver.cpp:218] Iteration 2964 (1.16485 iter/s, 10.3018s/12 iters), loss = 2.19411
I0428 17:31:27.503273  8468 solver.cpp:237]     Train net output #0: loss = 2.19411 (* 1 = 2.19411 loss)
I0428 17:31:27.503281  8468 sgd_solver.cpp:105] Iteration 2964, lr = 0.00555922
I0428 17:31:33.138201  8468 solver.cpp:218] Iteration 2976 (2.12967 iter/s, 5.63469s/12 iters), loss = 2.02982
I0428 17:31:33.138247  8468 solver.cpp:237]     Train net output #0: loss = 2.02982 (* 1 = 2.02982 loss)
I0428 17:31:33.138257  8468 sgd_solver.cpp:105] Iteration 2976, lr = 0.00554603
I0428 17:31:38.681000  8468 solver.cpp:218] Iteration 2988 (2.16508 iter/s, 5.54252s/12 iters), loss = 2.13046
I0428 17:31:38.681057  8468 solver.cpp:237]     Train net output #0: loss = 2.13046 (* 1 = 2.13046 loss)
I0428 17:31:38.681069  8468 sgd_solver.cpp:105] Iteration 2988, lr = 0.00553286
I0428 17:31:44.308290  8468 solver.cpp:218] Iteration 3000 (2.13258 iter/s, 5.62699s/12 iters), loss = 2.41968
I0428 17:31:44.308416  8468 solver.cpp:237]     Train net output #0: loss = 2.41968 (* 1 = 2.41968 loss)
I0428 17:31:44.308429  8468 sgd_solver.cpp:105] Iteration 3000, lr = 0.00551972
I0428 17:31:50.518499  8468 solver.cpp:218] Iteration 3012 (1.93242 iter/s, 6.20983s/12 iters), loss = 2.45325
I0428 17:31:50.518549  8468 solver.cpp:237]     Train net output #0: loss = 2.45325 (* 1 = 2.45325 loss)
I0428 17:31:50.518558  8468 sgd_solver.cpp:105] Iteration 3012, lr = 0.00550662
I0428 17:31:56.245975  8468 solver.cpp:218] Iteration 3024 (2.09527 iter/s, 5.72718s/12 iters), loss = 2.28745
I0428 17:31:56.246027  8468 solver.cpp:237]     Train net output #0: loss = 2.28745 (* 1 = 2.28745 loss)
I0428 17:31:56.246039  8468 sgd_solver.cpp:105] Iteration 3024, lr = 0.00549354
I0428 17:32:00.606885  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:32:01.794616  8468 solver.cpp:218] Iteration 3036 (2.1628 iter/s, 5.54836s/12 iters), loss = 2.04048
I0428 17:32:01.794659  8468 solver.cpp:237]     Train net output #0: loss = 2.04048 (* 1 = 2.04048 loss)
I0428 17:32:01.794669  8468 sgd_solver.cpp:105] Iteration 3036, lr = 0.0054805
I0428 17:32:07.432718  8468 solver.cpp:218] Iteration 3048 (2.12848 iter/s, 5.63782s/12 iters), loss = 2.33274
I0428 17:32:07.432766  8468 solver.cpp:237]     Train net output #0: loss = 2.33274 (* 1 = 2.33274 loss)
I0428 17:32:07.432777  8468 sgd_solver.cpp:105] Iteration 3048, lr = 0.00546749
I0428 17:32:12.334995  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3060.caffemodel
I0428 17:32:14.055398  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3060.solverstate
I0428 17:32:14.837311  8468 solver.cpp:330] Iteration 3060, Testing net (#0)
I0428 17:32:14.837393  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:32:18.391427  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:32:19.729436  8468 solver.cpp:397]     Test net output #0: accuracy = 0.262868
I0428 17:32:19.729483  8468 solver.cpp:397]     Test net output #1: loss = 2.91784 (* 1 = 2.91784 loss)
I0428 17:32:19.852921  8468 solver.cpp:218] Iteration 3060 (0.96621 iter/s, 12.4197s/12 iters), loss = 2.11616
I0428 17:32:19.852968  8468 solver.cpp:237]     Train net output #0: loss = 2.11616 (* 1 = 2.11616 loss)
I0428 17:32:19.852977  8468 sgd_solver.cpp:105] Iteration 3060, lr = 0.00545451
I0428 17:32:24.510123  8468 solver.cpp:218] Iteration 3072 (2.57679 iter/s, 4.65695s/12 iters), loss = 2.38044
I0428 17:32:24.510179  8468 solver.cpp:237]     Train net output #0: loss = 2.38044 (* 1 = 2.38044 loss)
I0428 17:32:24.510190  8468 sgd_solver.cpp:105] Iteration 3072, lr = 0.00544156
I0428 17:32:30.233350  8468 solver.cpp:218] Iteration 3084 (2.09683 iter/s, 5.72293s/12 iters), loss = 2.17204
I0428 17:32:30.233395  8468 solver.cpp:237]     Train net output #0: loss = 2.17204 (* 1 = 2.17204 loss)
I0428 17:32:30.233405  8468 sgd_solver.cpp:105] Iteration 3084, lr = 0.00542864
I0428 17:32:35.769065  8468 solver.cpp:218] Iteration 3096 (2.16785 iter/s, 5.53544s/12 iters), loss = 2.11892
I0428 17:32:35.769111  8468 solver.cpp:237]     Train net output #0: loss = 2.11892 (* 1 = 2.11892 loss)
I0428 17:32:35.769120  8468 sgd_solver.cpp:105] Iteration 3096, lr = 0.00541575
I0428 17:32:41.319368  8468 solver.cpp:218] Iteration 3108 (2.16215 iter/s, 5.55003s/12 iters), loss = 1.95684
I0428 17:32:41.319409  8468 solver.cpp:237]     Train net output #0: loss = 1.95684 (* 1 = 1.95684 loss)
I0428 17:32:41.319420  8468 sgd_solver.cpp:105] Iteration 3108, lr = 0.00540289
I0428 17:32:46.970244  8468 solver.cpp:218] Iteration 3120 (2.12367 iter/s, 5.6506s/12 iters), loss = 2.33969
I0428 17:32:46.970372  8468 solver.cpp:237]     Train net output #0: loss = 2.33969 (* 1 = 2.33969 loss)
I0428 17:32:46.970383  8468 sgd_solver.cpp:105] Iteration 3120, lr = 0.00539006
I0428 17:32:52.569721  8468 solver.cpp:218] Iteration 3132 (2.1432 iter/s, 5.59911s/12 iters), loss = 2.18817
I0428 17:32:52.569766  8468 solver.cpp:237]     Train net output #0: loss = 2.18817 (* 1 = 2.18817 loss)
I0428 17:32:52.569773  8468 sgd_solver.cpp:105] Iteration 3132, lr = 0.00537727
I0428 17:32:53.772365  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:32:58.123661  8468 solver.cpp:218] Iteration 3144 (2.16074 iter/s, 5.55366s/12 iters), loss = 2.0634
I0428 17:32:58.123713  8468 solver.cpp:237]     Train net output #0: loss = 2.0634 (* 1 = 2.0634 loss)
I0428 17:32:58.123724  8468 sgd_solver.cpp:105] Iteration 3144, lr = 0.0053645
I0428 17:33:03.628545  8468 solver.cpp:218] Iteration 3156 (2.18 iter/s, 5.50459s/12 iters), loss = 2.07187
I0428 17:33:03.628602  8468 solver.cpp:237]     Train net output #0: loss = 2.07187 (* 1 = 2.07187 loss)
I0428 17:33:03.628618  8468 sgd_solver.cpp:105] Iteration 3156, lr = 0.00535176
I0428 17:33:05.899125  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3162.caffemodel
I0428 17:33:06.927294  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3162.solverstate
I0428 17:33:07.887991  8468 solver.cpp:330] Iteration 3162, Testing net (#0)
I0428 17:33:07.888017  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:33:11.400365  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:33:12.924158  8468 solver.cpp:397]     Test net output #0: accuracy = 0.294118
I0428 17:33:12.924187  8468 solver.cpp:397]     Test net output #1: loss = 2.89958 (* 1 = 2.89958 loss)
I0428 17:33:15.171692  8468 solver.cpp:218] Iteration 3168 (1.03963 iter/s, 11.5426s/12 iters), loss = 2.18548
I0428 17:33:15.171753  8468 solver.cpp:237]     Train net output #0: loss = 2.18548 (* 1 = 2.18548 loss)
I0428 17:33:15.171766  8468 sgd_solver.cpp:105] Iteration 3168, lr = 0.00533906
I0428 17:33:20.673813  8468 solver.cpp:218] Iteration 3180 (2.18109 iter/s, 5.50182s/12 iters), loss = 2.08307
I0428 17:33:20.673969  8468 solver.cpp:237]     Train net output #0: loss = 2.08307 (* 1 = 2.08307 loss)
I0428 17:33:20.673982  8468 sgd_solver.cpp:105] Iteration 3180, lr = 0.00532638
I0428 17:33:26.252780  8468 solver.cpp:218] Iteration 3192 (2.15108 iter/s, 5.57858s/12 iters), loss = 2.16515
I0428 17:33:26.252825  8468 solver.cpp:237]     Train net output #0: loss = 2.16515 (* 1 = 2.16515 loss)
I0428 17:33:26.252832  8468 sgd_solver.cpp:105] Iteration 3192, lr = 0.00531374
I0428 17:33:31.699918  8468 solver.cpp:218] Iteration 3204 (2.2031 iter/s, 5.44686s/12 iters), loss = 1.97876
I0428 17:33:31.699973  8468 solver.cpp:237]     Train net output #0: loss = 1.97876 (* 1 = 1.97876 loss)
I0428 17:33:31.699985  8468 sgd_solver.cpp:105] Iteration 3204, lr = 0.00530112
I0428 17:33:37.030750  8468 solver.cpp:218] Iteration 3216 (2.25117 iter/s, 5.33055s/12 iters), loss = 1.95539
I0428 17:33:37.030802  8468 solver.cpp:237]     Train net output #0: loss = 1.95539 (* 1 = 1.95539 loss)
I0428 17:33:37.030815  8468 sgd_solver.cpp:105] Iteration 3216, lr = 0.00528853
I0428 17:33:42.405979  8468 solver.cpp:218] Iteration 3228 (2.23258 iter/s, 5.37495s/12 iters), loss = 1.96201
I0428 17:33:42.406033  8468 solver.cpp:237]     Train net output #0: loss = 1.96201 (* 1 = 1.96201 loss)
I0428 17:33:42.406044  8468 sgd_solver.cpp:105] Iteration 3228, lr = 0.00527598
I0428 17:33:45.920053  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:33:47.761703  8468 solver.cpp:218] Iteration 3240 (2.24071 iter/s, 5.35544s/12 iters), loss = 2.15625
I0428 17:33:47.761747  8468 solver.cpp:237]     Train net output #0: loss = 2.15625 (* 1 = 2.15625 loss)
I0428 17:33:47.761755  8468 sgd_solver.cpp:105] Iteration 3240, lr = 0.00526345
I0428 17:33:53.386463  8468 solver.cpp:218] Iteration 3252 (2.13353 iter/s, 5.62447s/12 iters), loss = 1.79171
I0428 17:33:53.386602  8468 solver.cpp:237]     Train net output #0: loss = 1.79171 (* 1 = 1.79171 loss)
I0428 17:33:53.386617  8468 sgd_solver.cpp:105] Iteration 3252, lr = 0.00525095
I0428 17:33:58.542819  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3264.caffemodel
I0428 17:33:59.129010  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3264.solverstate
I0428 17:33:59.656906  8468 solver.cpp:330] Iteration 3264, Testing net (#0)
I0428 17:33:59.656927  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:34:03.503140  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:34:04.880131  8468 solver.cpp:397]     Test net output #0: accuracy = 0.306373
I0428 17:34:04.880174  8468 solver.cpp:397]     Test net output #1: loss = 2.89938 (* 1 = 2.89938 loss)
I0428 17:34:05.009680  8468 solver.cpp:218] Iteration 3264 (1.03247 iter/s, 11.6226s/12 iters), loss = 2.03465
I0428 17:34:05.009734  8468 solver.cpp:237]     Train net output #0: loss = 2.03465 (* 1 = 2.03465 loss)
I0428 17:34:05.009748  8468 sgd_solver.cpp:105] Iteration 3264, lr = 0.00523849
I0428 17:34:09.722921  8468 solver.cpp:218] Iteration 3276 (2.54616 iter/s, 4.71298s/12 iters), loss = 1.87754
I0428 17:34:09.722971  8468 solver.cpp:237]     Train net output #0: loss = 1.87754 (* 1 = 1.87754 loss)
I0428 17:34:09.722982  8468 sgd_solver.cpp:105] Iteration 3276, lr = 0.00522605
I0428 17:34:15.516767  8468 solver.cpp:218] Iteration 3288 (2.07127 iter/s, 5.79355s/12 iters), loss = 1.75211
I0428 17:34:15.516806  8468 solver.cpp:237]     Train net output #0: loss = 1.75211 (* 1 = 1.75211 loss)
I0428 17:34:15.516815  8468 sgd_solver.cpp:105] Iteration 3288, lr = 0.00521364
I0428 17:34:21.262063  8468 solver.cpp:218] Iteration 3300 (2.08877 iter/s, 5.74501s/12 iters), loss = 2.06456
I0428 17:34:21.262113  8468 solver.cpp:237]     Train net output #0: loss = 2.06456 (* 1 = 2.06456 loss)
I0428 17:34:21.262126  8468 sgd_solver.cpp:105] Iteration 3300, lr = 0.00520126
I0428 17:34:26.882848  8468 solver.cpp:218] Iteration 3312 (2.13504 iter/s, 5.6205s/12 iters), loss = 1.9154
I0428 17:34:26.883006  8468 solver.cpp:237]     Train net output #0: loss = 1.9154 (* 1 = 1.9154 loss)
I0428 17:34:26.883016  8468 sgd_solver.cpp:105] Iteration 3312, lr = 0.00518892
I0428 17:34:32.436583  8468 solver.cpp:218] Iteration 3324 (2.16086 iter/s, 5.55334s/12 iters), loss = 2.14143
I0428 17:34:32.436621  8468 solver.cpp:237]     Train net output #0: loss = 2.14143 (* 1 = 2.14143 loss)
I0428 17:34:32.436630  8468 sgd_solver.cpp:105] Iteration 3324, lr = 0.0051766
I0428 17:34:38.080372  8468 solver.cpp:218] Iteration 3336 (2.12634 iter/s, 5.64351s/12 iters), loss = 1.98913
I0428 17:34:38.080421  8468 solver.cpp:237]     Train net output #0: loss = 1.98913 (* 1 = 1.98913 loss)
I0428 17:34:38.080431  8468 sgd_solver.cpp:105] Iteration 3336, lr = 0.00516431
I0428 17:34:38.583112  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:34:43.749799  8468 solver.cpp:218] Iteration 3348 (2.11673 iter/s, 5.66913s/12 iters), loss = 1.78493
I0428 17:34:43.749843  8468 solver.cpp:237]     Train net output #0: loss = 1.78493 (* 1 = 1.78493 loss)
I0428 17:34:43.749853  8468 sgd_solver.cpp:105] Iteration 3348, lr = 0.00515204
I0428 17:34:49.553491  8468 solver.cpp:218] Iteration 3360 (2.06775 iter/s, 5.8034s/12 iters), loss = 1.76932
I0428 17:34:49.553534  8468 solver.cpp:237]     Train net output #0: loss = 1.76932 (* 1 = 1.76932 loss)
I0428 17:34:49.553542  8468 sgd_solver.cpp:105] Iteration 3360, lr = 0.00513981
I0428 17:34:51.883648  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3366.caffemodel
I0428 17:34:53.293792  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3366.solverstate
I0428 17:34:54.299662  8468 solver.cpp:330] Iteration 3366, Testing net (#0)
I0428 17:34:54.299692  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:34:57.991981  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:34:59.580803  8468 solver.cpp:397]     Test net output #0: accuracy = 0.294118
I0428 17:34:59.580847  8468 solver.cpp:397]     Test net output #1: loss = 2.94822 (* 1 = 2.94822 loss)
I0428 17:35:01.624847  8468 solver.cpp:218] Iteration 3372 (0.994133 iter/s, 12.0708s/12 iters), loss = 1.80641
I0428 17:35:01.624889  8468 solver.cpp:237]     Train net output #0: loss = 1.80641 (* 1 = 1.80641 loss)
I0428 17:35:01.624899  8468 sgd_solver.cpp:105] Iteration 3372, lr = 0.00512761
I0428 17:35:07.407696  8468 solver.cpp:218] Iteration 3384 (2.07521 iter/s, 5.78256s/12 iters), loss = 2.02962
I0428 17:35:07.407749  8468 solver.cpp:237]     Train net output #0: loss = 2.02962 (* 1 = 2.02962 loss)
I0428 17:35:07.407763  8468 sgd_solver.cpp:105] Iteration 3384, lr = 0.00511544
I0428 17:35:13.218371  8468 solver.cpp:218] Iteration 3396 (2.06527 iter/s, 5.81038s/12 iters), loss = 1.76526
I0428 17:35:13.218413  8468 solver.cpp:237]     Train net output #0: loss = 1.76526 (* 1 = 1.76526 loss)
I0428 17:35:13.218422  8468 sgd_solver.cpp:105] Iteration 3396, lr = 0.00510329
I0428 17:35:19.078003  8468 solver.cpp:218] Iteration 3408 (2.04801 iter/s, 5.85934s/12 iters), loss = 1.66164
I0428 17:35:19.078058  8468 solver.cpp:237]     Train net output #0: loss = 1.66164 (* 1 = 1.66164 loss)
I0428 17:35:19.078070  8468 sgd_solver.cpp:105] Iteration 3408, lr = 0.00509117
I0428 17:35:24.814005  8468 solver.cpp:218] Iteration 3420 (2.09216 iter/s, 5.7357s/12 iters), loss = 1.67038
I0428 17:35:24.814052  8468 solver.cpp:237]     Train net output #0: loss = 1.67038 (* 1 = 1.67038 loss)
I0428 17:35:24.814060  8468 sgd_solver.cpp:105] Iteration 3420, lr = 0.00507909
I0428 17:35:30.469831  8468 solver.cpp:218] Iteration 3432 (2.12182 iter/s, 5.65554s/12 iters), loss = 1.9025
I0428 17:35:30.470835  8468 solver.cpp:237]     Train net output #0: loss = 1.9025 (* 1 = 1.9025 loss)
I0428 17:35:30.470866  8468 sgd_solver.cpp:105] Iteration 3432, lr = 0.00506703
I0428 17:35:33.427343  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:35:36.218475  8468 solver.cpp:218] Iteration 3444 (2.0879 iter/s, 5.7474s/12 iters), loss = 1.86351
I0428 17:35:36.218519  8468 solver.cpp:237]     Train net output #0: loss = 1.86351 (* 1 = 1.86351 loss)
I0428 17:35:36.218528  8468 sgd_solver.cpp:105] Iteration 3444, lr = 0.005055
I0428 17:35:41.982260  8468 solver.cpp:218] Iteration 3456 (2.08207 iter/s, 5.76349s/12 iters), loss = 2.04378
I0428 17:35:41.982308  8468 solver.cpp:237]     Train net output #0: loss = 2.04378 (* 1 = 2.04378 loss)
I0428 17:35:41.982319  8468 sgd_solver.cpp:105] Iteration 3456, lr = 0.005043
I0428 17:35:47.245661  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3468.caffemodel
I0428 17:35:47.878902  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3468.solverstate
I0428 17:35:48.320688  8468 solver.cpp:330] Iteration 3468, Testing net (#0)
I0428 17:35:48.320709  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:35:48.438446  8468 blocking_queue.cpp:49] Waiting for data
I0428 17:35:51.736654  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:35:53.394634  8468 solver.cpp:397]     Test net output #0: accuracy = 0.295956
I0428 17:35:53.394665  8468 solver.cpp:397]     Test net output #1: loss = 2.93937 (* 1 = 2.93937 loss)
I0428 17:35:53.519793  8468 solver.cpp:218] Iteration 3468 (1.04013 iter/s, 11.537s/12 iters), loss = 1.5291
I0428 17:35:53.519834  8468 solver.cpp:237]     Train net output #0: loss = 1.5291 (* 1 = 1.5291 loss)
I0428 17:35:53.519842  8468 sgd_solver.cpp:105] Iteration 3468, lr = 0.00503102
I0428 17:35:58.277577  8468 solver.cpp:218] Iteration 3480 (2.52231 iter/s, 4.75753s/12 iters), loss = 1.68211
I0428 17:35:58.277634  8468 solver.cpp:237]     Train net output #0: loss = 1.68211 (* 1 = 1.68211 loss)
I0428 17:35:58.277647  8468 sgd_solver.cpp:105] Iteration 3480, lr = 0.00501908
I0428 17:36:04.056543  8468 solver.cpp:218] Iteration 3492 (2.07661 iter/s, 5.77865s/12 iters), loss = 1.70495
I0428 17:36:04.057237  8468 solver.cpp:237]     Train net output #0: loss = 1.70495 (* 1 = 1.70495 loss)
I0428 17:36:04.057248  8468 sgd_solver.cpp:105] Iteration 3492, lr = 0.00500716
I0428 17:36:09.941968  8468 solver.cpp:218] Iteration 3504 (2.03926 iter/s, 5.88448s/12 iters), loss = 1.57269
I0428 17:36:09.942029  8468 solver.cpp:237]     Train net output #0: loss = 1.57269 (* 1 = 1.57269 loss)
I0428 17:36:09.942045  8468 sgd_solver.cpp:105] Iteration 3504, lr = 0.00499527
I0428 17:36:15.801949  8468 solver.cpp:218] Iteration 3516 (2.04789 iter/s, 5.85968s/12 iters), loss = 1.85599
I0428 17:36:15.801986  8468 solver.cpp:237]     Train net output #0: loss = 1.85599 (* 1 = 1.85599 loss)
I0428 17:36:15.801995  8468 sgd_solver.cpp:105] Iteration 3516, lr = 0.00498341
I0428 17:36:21.413518  8468 solver.cpp:218] Iteration 3528 (2.13855 iter/s, 5.61129s/12 iters), loss = 1.74412
I0428 17:36:21.413566  8468 solver.cpp:237]     Train net output #0: loss = 1.74412 (* 1 = 1.74412 loss)
I0428 17:36:21.413578  8468 sgd_solver.cpp:105] Iteration 3528, lr = 0.00497158
I0428 17:36:26.808645  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:36:27.119614  8468 solver.cpp:218] Iteration 3540 (2.10312 iter/s, 5.7058s/12 iters), loss = 1.67242
I0428 17:36:27.119663  8468 solver.cpp:237]     Train net output #0: loss = 1.67242 (* 1 = 1.67242 loss)
I0428 17:36:27.119673  8468 sgd_solver.cpp:105] Iteration 3540, lr = 0.00495978
I0428 17:36:32.661026  8468 solver.cpp:218] Iteration 3552 (2.16562 iter/s, 5.54113s/12 iters), loss = 1.64977
I0428 17:36:32.661072  8468 solver.cpp:237]     Train net output #0: loss = 1.64977 (* 1 = 1.64977 loss)
I0428 17:36:32.661079  8468 sgd_solver.cpp:105] Iteration 3552, lr = 0.004948
I0428 17:36:38.231357  8468 solver.cpp:218] Iteration 3564 (2.15438 iter/s, 5.57005s/12 iters), loss = 1.57766
I0428 17:36:38.232023  8468 solver.cpp:237]     Train net output #0: loss = 1.57766 (* 1 = 1.57766 loss)
I0428 17:36:38.232033  8468 sgd_solver.cpp:105] Iteration 3564, lr = 0.00493626
I0428 17:36:40.432755  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3570.caffemodel
I0428 17:36:44.095284  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3570.solverstate
I0428 17:36:44.743263  8468 solver.cpp:330] Iteration 3570, Testing net (#0)
I0428 17:36:44.743289  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:36:48.081667  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:36:49.769727  8468 solver.cpp:397]     Test net output #0: accuracy = 0.325368
I0428 17:36:49.769762  8468 solver.cpp:397]     Test net output #1: loss = 2.9777 (* 1 = 2.9777 loss)
I0428 17:36:51.813643  8468 solver.cpp:218] Iteration 3576 (0.883583 iter/s, 13.5811s/12 iters), loss = 1.50351
I0428 17:36:51.813694  8468 solver.cpp:237]     Train net output #0: loss = 1.50351 (* 1 = 1.50351 loss)
I0428 17:36:51.813705  8468 sgd_solver.cpp:105] Iteration 3576, lr = 0.00492454
I0428 17:36:57.587621  8468 solver.cpp:218] Iteration 3588 (2.07839 iter/s, 5.77369s/12 iters), loss = 1.76215
I0428 17:36:57.587661  8468 solver.cpp:237]     Train net output #0: loss = 1.76215 (* 1 = 1.76215 loss)
I0428 17:36:57.587669  8468 sgd_solver.cpp:105] Iteration 3588, lr = 0.00491284
I0428 17:37:03.093869  8468 solver.cpp:218] Iteration 3600 (2.17945 iter/s, 5.50597s/12 iters), loss = 1.82702
I0428 17:37:03.093926  8468 solver.cpp:237]     Train net output #0: loss = 1.82702 (* 1 = 1.82702 loss)
I0428 17:37:03.093938  8468 sgd_solver.cpp:105] Iteration 3600, lr = 0.00490118
I0428 17:37:08.667374  8468 solver.cpp:218] Iteration 3612 (2.15316 iter/s, 5.57321s/12 iters), loss = 1.34343
I0428 17:37:08.667513  8468 solver.cpp:237]     Train net output #0: loss = 1.34343 (* 1 = 1.34343 loss)
I0428 17:37:08.667522  8468 sgd_solver.cpp:105] Iteration 3612, lr = 0.00488954
I0428 17:37:14.237067  8468 solver.cpp:218] Iteration 3624 (2.15466 iter/s, 5.56932s/12 iters), loss = 1.79099
I0428 17:37:14.237107  8468 solver.cpp:237]     Train net output #0: loss = 1.79099 (* 1 = 1.79099 loss)
I0428 17:37:14.237116  8468 sgd_solver.cpp:105] Iteration 3624, lr = 0.00487793
I0428 17:37:19.882380  8468 solver.cpp:218] Iteration 3636 (2.12576 iter/s, 5.64503s/12 iters), loss = 1.55094
I0428 17:37:19.882423  8468 solver.cpp:237]     Train net output #0: loss = 1.55094 (* 1 = 1.55094 loss)
I0428 17:37:19.882436  8468 sgd_solver.cpp:105] Iteration 3636, lr = 0.00486635
I0428 17:37:21.921191  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:37:25.324911  8468 solver.cpp:218] Iteration 3648 (2.20497 iter/s, 5.44225s/12 iters), loss = 1.48127
I0428 17:37:25.324954  8468 solver.cpp:237]     Train net output #0: loss = 1.48127 (* 1 = 1.48127 loss)
I0428 17:37:25.324963  8468 sgd_solver.cpp:105] Iteration 3648, lr = 0.0048548
I0428 17:37:30.767611  8468 solver.cpp:218] Iteration 3660 (2.2049 iter/s, 5.44242s/12 iters), loss = 1.60449
I0428 17:37:30.767652  8468 solver.cpp:237]     Train net output #0: loss = 1.60449 (* 1 = 1.60449 loss)
I0428 17:37:30.767661  8468 sgd_solver.cpp:105] Iteration 3660, lr = 0.00484327
I0428 17:37:35.674810  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3672.caffemodel
I0428 17:37:41.788524  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3672.solverstate
I0428 17:37:47.469812  8468 solver.cpp:330] Iteration 3672, Testing net (#0)
I0428 17:37:47.469849  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:37:50.664829  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:37:52.271486  8468 solver.cpp:397]     Test net output #0: accuracy = 0.323529
I0428 17:37:52.271523  8468 solver.cpp:397]     Test net output #1: loss = 2.98268 (* 1 = 2.98268 loss)
I0428 17:37:52.403750  8468 solver.cpp:218] Iteration 3672 (0.554651 iter/s, 21.6352s/12 iters), loss = 1.61623
I0428 17:37:52.403811  8468 solver.cpp:237]     Train net output #0: loss = 1.61623 (* 1 = 1.61623 loss)
I0428 17:37:52.403820  8468 sgd_solver.cpp:105] Iteration 3672, lr = 0.00483177
I0428 17:37:56.913278  8468 solver.cpp:218] Iteration 3684 (2.66119 iter/s, 4.50927s/12 iters), loss = 1.53781
I0428 17:37:56.913316  8468 solver.cpp:237]     Train net output #0: loss = 1.53781 (* 1 = 1.53781 loss)
I0428 17:37:56.913326  8468 sgd_solver.cpp:105] Iteration 3684, lr = 0.0048203
I0428 17:38:02.324209  8468 solver.cpp:218] Iteration 3696 (2.21784 iter/s, 5.41066s/12 iters), loss = 1.5746
I0428 17:38:02.324249  8468 solver.cpp:237]     Train net output #0: loss = 1.5746 (* 1 = 1.5746 loss)
I0428 17:38:02.324261  8468 sgd_solver.cpp:105] Iteration 3696, lr = 0.00480886
I0428 17:38:07.712051  8468 solver.cpp:218] Iteration 3708 (2.22735 iter/s, 5.38757s/12 iters), loss = 1.71392
I0428 17:38:07.712090  8468 solver.cpp:237]     Train net output #0: loss = 1.71392 (* 1 = 1.71392 loss)
I0428 17:38:07.712098  8468 sgd_solver.cpp:105] Iteration 3708, lr = 0.00479744
I0428 17:38:13.413609  8468 solver.cpp:218] Iteration 3720 (2.10479 iter/s, 5.70128s/12 iters), loss = 1.86677
I0428 17:38:13.413851  8468 solver.cpp:237]     Train net output #0: loss = 1.86677 (* 1 = 1.86677 loss)
I0428 17:38:13.413862  8468 sgd_solver.cpp:105] Iteration 3720, lr = 0.00478605
I0428 17:38:18.983410  8468 solver.cpp:218] Iteration 3732 (2.15466 iter/s, 5.56932s/12 iters), loss = 1.77104
I0428 17:38:18.983453  8468 solver.cpp:237]     Train net output #0: loss = 1.77104 (* 1 = 1.77104 loss)
I0428 17:38:18.983464  8468 sgd_solver.cpp:105] Iteration 3732, lr = 0.00477469
I0428 17:38:23.389160  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:38:24.485776  8468 solver.cpp:218] Iteration 3744 (2.18099 iter/s, 5.50209s/12 iters), loss = 1.64387
I0428 17:38:24.485812  8468 solver.cpp:237]     Train net output #0: loss = 1.64387 (* 1 = 1.64387 loss)
I0428 17:38:24.485822  8468 sgd_solver.cpp:105] Iteration 3744, lr = 0.00476335
I0428 17:38:30.122089  8468 solver.cpp:218] Iteration 3756 (2.12916 iter/s, 5.63603s/12 iters), loss = 1.50058
I0428 17:38:30.122138  8468 solver.cpp:237]     Train net output #0: loss = 1.50058 (* 1 = 1.50058 loss)
I0428 17:38:30.122150  8468 sgd_solver.cpp:105] Iteration 3756, lr = 0.00475204
I0428 17:38:35.683288  8468 solver.cpp:218] Iteration 3768 (2.15792 iter/s, 5.56091s/12 iters), loss = 1.48356
I0428 17:38:35.683327  8468 solver.cpp:237]     Train net output #0: loss = 1.48356 (* 1 = 1.48356 loss)
I0428 17:38:35.683337  8468 sgd_solver.cpp:105] Iteration 3768, lr = 0.00474076
I0428 17:38:37.871193  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3774.caffemodel
I0428 17:38:39.468806  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3774.solverstate
I0428 17:38:40.240329  8468 solver.cpp:330] Iteration 3774, Testing net (#0)
I0428 17:38:40.240351  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:38:43.458250  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:38:45.120586  8468 solver.cpp:397]     Test net output #0: accuracy = 0.32598
I0428 17:38:45.120613  8468 solver.cpp:397]     Test net output #1: loss = 3.03422 (* 1 = 3.03422 loss)
I0428 17:38:47.196817  8468 solver.cpp:218] Iteration 3780 (1.0423 iter/s, 11.513s/12 iters), loss = 1.73625
I0428 17:38:47.196858  8468 solver.cpp:237]     Train net output #0: loss = 1.73625 (* 1 = 1.73625 loss)
I0428 17:38:47.196867  8468 sgd_solver.cpp:105] Iteration 3780, lr = 0.00472951
I0428 17:38:52.723517  8468 solver.cpp:218] Iteration 3792 (2.17139 iter/s, 5.52643s/12 iters), loss = 1.623
I0428 17:38:52.723557  8468 solver.cpp:237]     Train net output #0: loss = 1.623 (* 1 = 1.623 loss)
I0428 17:38:52.723565  8468 sgd_solver.cpp:105] Iteration 3792, lr = 0.00471828
I0428 17:38:58.174346  8468 solver.cpp:218] Iteration 3804 (2.20161 iter/s, 5.45055s/12 iters), loss = 1.51203
I0428 17:38:58.174397  8468 solver.cpp:237]     Train net output #0: loss = 1.51203 (* 1 = 1.51203 loss)
I0428 17:38:58.174414  8468 sgd_solver.cpp:105] Iteration 3804, lr = 0.00470707
I0428 17:39:03.677918  8468 solver.cpp:218] Iteration 3816 (2.18051 iter/s, 5.50329s/12 iters), loss = 1.31874
I0428 17:39:03.677968  8468 solver.cpp:237]     Train net output #0: loss = 1.31874 (* 1 = 1.31874 loss)
I0428 17:39:03.677978  8468 sgd_solver.cpp:105] Iteration 3816, lr = 0.0046959
I0428 17:39:09.096042  8468 solver.cpp:218] Iteration 3828 (2.2149 iter/s, 5.41784s/12 iters), loss = 1.41622
I0428 17:39:09.096099  8468 solver.cpp:237]     Train net output #0: loss = 1.41622 (* 1 = 1.41622 loss)
I0428 17:39:09.096112  8468 sgd_solver.cpp:105] Iteration 3828, lr = 0.00468475
I0428 17:39:14.419648  8468 solver.cpp:218] Iteration 3840 (2.25423 iter/s, 5.32332s/12 iters), loss = 1.51121
I0428 17:39:14.420861  8468 solver.cpp:237]     Train net output #0: loss = 1.51121 (* 1 = 1.51121 loss)
I0428 17:39:14.420876  8468 sgd_solver.cpp:105] Iteration 3840, lr = 0.00467363
I0428 17:39:15.625785  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:39:19.773205  8468 solver.cpp:218] Iteration 3852 (2.2421 iter/s, 5.35212s/12 iters), loss = 1.2848
I0428 17:39:19.773254  8468 solver.cpp:237]     Train net output #0: loss = 1.2848 (* 1 = 1.2848 loss)
I0428 17:39:19.773265  8468 sgd_solver.cpp:105] Iteration 3852, lr = 0.00466253
I0428 17:39:25.198482  8468 solver.cpp:218] Iteration 3864 (2.21199 iter/s, 5.42499s/12 iters), loss = 1.59569
I0428 17:39:25.198530  8468 solver.cpp:237]     Train net output #0: loss = 1.59569 (* 1 = 1.59569 loss)
I0428 17:39:25.198545  8468 sgd_solver.cpp:105] Iteration 3864, lr = 0.00465146
I0428 17:39:30.266368  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3876.caffemodel
I0428 17:39:31.661631  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3876.solverstate
I0428 17:39:33.343683  8468 solver.cpp:330] Iteration 3876, Testing net (#0)
I0428 17:39:33.343713  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:39:36.400632  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:39:38.183374  8468 solver.cpp:397]     Test net output #0: accuracy = 0.33027
I0428 17:39:38.183404  8468 solver.cpp:397]     Test net output #1: loss = 2.9755 (* 1 = 2.9755 loss)
I0428 17:39:38.311877  8468 solver.cpp:218] Iteration 3876 (0.915136 iter/s, 13.1128s/12 iters), loss = 1.43392
I0428 17:39:38.311916  8468 solver.cpp:237]     Train net output #0: loss = 1.43392 (* 1 = 1.43392 loss)
I0428 17:39:38.311925  8468 sgd_solver.cpp:105] Iteration 3876, lr = 0.00464042
I0428 17:39:42.819839  8468 solver.cpp:218] Iteration 3888 (2.6621 iter/s, 4.50772s/12 iters), loss = 1.30713
I0428 17:39:42.819887  8468 solver.cpp:237]     Train net output #0: loss = 1.30713 (* 1 = 1.30713 loss)
I0428 17:39:42.819898  8468 sgd_solver.cpp:105] Iteration 3888, lr = 0.0046294
I0428 17:39:48.221621  8468 solver.cpp:218] Iteration 3900 (2.2216 iter/s, 5.4015s/12 iters), loss = 1.45476
I0428 17:39:48.221733  8468 solver.cpp:237]     Train net output #0: loss = 1.45476 (* 1 = 1.45476 loss)
I0428 17:39:48.221743  8468 sgd_solver.cpp:105] Iteration 3900, lr = 0.00461841
I0428 17:39:53.577301  8468 solver.cpp:218] Iteration 3912 (2.24076 iter/s, 5.35534s/12 iters), loss = 1.35995
I0428 17:39:53.577417  8468 solver.cpp:237]     Train net output #0: loss = 1.35995 (* 1 = 1.35995 loss)
I0428 17:39:53.577433  8468 sgd_solver.cpp:105] Iteration 3912, lr = 0.00460744
I0428 17:39:59.087698  8468 solver.cpp:218] Iteration 3924 (2.17784 iter/s, 5.51005s/12 iters), loss = 1.40345
I0428 17:39:59.087738  8468 solver.cpp:237]     Train net output #0: loss = 1.40345 (* 1 = 1.40345 loss)
I0428 17:39:59.087747  8468 sgd_solver.cpp:105] Iteration 3924, lr = 0.0045965
I0428 17:40:04.655050  8468 solver.cpp:218] Iteration 3936 (2.15553 iter/s, 5.56708s/12 iters), loss = 1.35754
I0428 17:40:04.655090  8468 solver.cpp:237]     Train net output #0: loss = 1.35754 (* 1 = 1.35754 loss)
I0428 17:40:04.655099  8468 sgd_solver.cpp:105] Iteration 3936, lr = 0.00458559
I0428 17:40:08.301116  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:40:10.081784  8468 solver.cpp:218] Iteration 3948 (2.21139 iter/s, 5.42646s/12 iters), loss = 1.59247
I0428 17:40:10.081822  8468 solver.cpp:237]     Train net output #0: loss = 1.59247 (* 1 = 1.59247 loss)
I0428 17:40:10.081831  8468 sgd_solver.cpp:105] Iteration 3948, lr = 0.0045747
I0428 17:40:15.531280  8468 solver.cpp:218] Iteration 3960 (2.20215 iter/s, 5.44922s/12 iters), loss = 1.32107
I0428 17:40:15.531327  8468 solver.cpp:237]     Train net output #0: loss = 1.32107 (* 1 = 1.32107 loss)
I0428 17:40:15.531338  8468 sgd_solver.cpp:105] Iteration 3960, lr = 0.00456384
I0428 17:40:20.961148  8468 solver.cpp:218] Iteration 3972 (2.21011 iter/s, 5.42958s/12 iters), loss = 1.70977
I0428 17:40:20.961319  8468 solver.cpp:237]     Train net output #0: loss = 1.70977 (* 1 = 1.70977 loss)
I0428 17:40:20.961328  8468 sgd_solver.cpp:105] Iteration 3972, lr = 0.00455301
I0428 17:40:23.191725  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3978.caffemodel
I0428 17:40:23.790320  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3978.solverstate
I0428 17:40:24.785653  8468 solver.cpp:330] Iteration 3978, Testing net (#0)
I0428 17:40:24.785672  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:40:27.701373  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:40:29.338317  8468 solver.cpp:397]     Test net output #0: accuracy = 0.346814
I0428 17:40:29.338349  8468 solver.cpp:397]     Test net output #1: loss = 2.85169 (* 1 = 2.85169 loss)
I0428 17:40:31.389266  8468 solver.cpp:218] Iteration 3984 (1.1508 iter/s, 10.4275s/12 iters), loss = 1.41111
I0428 17:40:31.389309  8468 solver.cpp:237]     Train net output #0: loss = 1.41111 (* 1 = 1.41111 loss)
I0428 17:40:31.389318  8468 sgd_solver.cpp:105] Iteration 3984, lr = 0.0045422
I0428 17:40:37.107579  8468 solver.cpp:218] Iteration 3996 (2.09863 iter/s, 5.71802s/12 iters), loss = 1.23288
I0428 17:40:37.107620  8468 solver.cpp:237]     Train net output #0: loss = 1.23288 (* 1 = 1.23288 loss)
I0428 17:40:37.107631  8468 sgd_solver.cpp:105] Iteration 3996, lr = 0.00453141
I0428 17:40:42.674317  8468 solver.cpp:218] Iteration 4008 (2.15577 iter/s, 5.56646s/12 iters), loss = 1.47905
I0428 17:40:42.674358  8468 solver.cpp:237]     Train net output #0: loss = 1.47905 (* 1 = 1.47905 loss)
I0428 17:40:42.674368  8468 sgd_solver.cpp:105] Iteration 4008, lr = 0.00452066
I0428 17:40:48.300979  8468 solver.cpp:218] Iteration 4020 (2.13281 iter/s, 5.62638s/12 iters), loss = 1.46131
I0428 17:40:48.301021  8468 solver.cpp:237]     Train net output #0: loss = 1.46131 (* 1 = 1.46131 loss)
I0428 17:40:48.301031  8468 sgd_solver.cpp:105] Iteration 4020, lr = 0.00450992
I0428 17:40:53.919828  8468 solver.cpp:218] Iteration 4032 (2.13578 iter/s, 5.61856s/12 iters), loss = 1.34191
I0428 17:40:53.919948  8468 solver.cpp:237]     Train net output #0: loss = 1.34191 (* 1 = 1.34191 loss)
I0428 17:40:53.919970  8468 sgd_solver.cpp:105] Iteration 4032, lr = 0.00449921
I0428 17:40:59.343145  8468 solver.cpp:218] Iteration 4044 (2.21281 iter/s, 5.42298s/12 iters), loss = 1.47853
I0428 17:40:59.343183  8468 solver.cpp:237]     Train net output #0: loss = 1.47853 (* 1 = 1.47853 loss)
I0428 17:40:59.343192  8468 sgd_solver.cpp:105] Iteration 4044, lr = 0.00448853
I0428 17:40:59.875064  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:41:04.767264  8468 solver.cpp:218] Iteration 4056 (2.21245 iter/s, 5.42385s/12 iters), loss = 1.05514
I0428 17:41:04.767304  8468 solver.cpp:237]     Train net output #0: loss = 1.05514 (* 1 = 1.05514 loss)
I0428 17:41:04.767314  8468 sgd_solver.cpp:105] Iteration 4056, lr = 0.00447788
I0428 17:41:10.224119  8468 solver.cpp:218] Iteration 4068 (2.19918 iter/s, 5.45658s/12 iters), loss = 1.1496
I0428 17:41:10.224155  8468 solver.cpp:237]     Train net output #0: loss = 1.1496 (* 1 = 1.1496 loss)
I0428 17:41:10.224166  8468 sgd_solver.cpp:105] Iteration 4068, lr = 0.00446724
I0428 17:41:15.114300  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4080.caffemodel
I0428 17:41:15.736812  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4080.solverstate
I0428 17:41:16.166232  8468 solver.cpp:330] Iteration 4080, Testing net (#0)
I0428 17:41:16.166254  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:41:19.170894  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:41:20.883407  8468 solver.cpp:397]     Test net output #0: accuracy = 0.356618
I0428 17:41:20.883440  8468 solver.cpp:397]     Test net output #1: loss = 2.96314 (* 1 = 2.96314 loss)
I0428 17:41:21.008256  8468 solver.cpp:218] Iteration 4080 (1.11279 iter/s, 10.7837s/12 iters), loss = 1.12266
I0428 17:41:21.008309  8468 solver.cpp:237]     Train net output #0: loss = 1.12266 (* 1 = 1.12266 loss)
I0428 17:41:21.008322  8468 sgd_solver.cpp:105] Iteration 4080, lr = 0.00445664
I0428 17:41:25.514258  8468 solver.cpp:218] Iteration 4092 (2.66326 iter/s, 4.50575s/12 iters), loss = 1.43533
I0428 17:41:25.514391  8468 solver.cpp:237]     Train net output #0: loss = 1.43533 (* 1 = 1.43533 loss)
I0428 17:41:25.514401  8468 sgd_solver.cpp:105] Iteration 4092, lr = 0.00444606
I0428 17:41:30.903822  8468 solver.cpp:218] Iteration 4104 (2.22668 iter/s, 5.3892s/12 iters), loss = 1.13799
I0428 17:41:30.903862  8468 solver.cpp:237]     Train net output #0: loss = 1.13799 (* 1 = 1.13799 loss)
I0428 17:41:30.903874  8468 sgd_solver.cpp:105] Iteration 4104, lr = 0.0044355
I0428 17:41:36.292449  8468 solver.cpp:218] Iteration 4116 (2.22702 iter/s, 5.38836s/12 iters), loss = 1.48287
I0428 17:41:36.292523  8468 solver.cpp:237]     Train net output #0: loss = 1.48287 (* 1 = 1.48287 loss)
I0428 17:41:36.292533  8468 sgd_solver.cpp:105] Iteration 4116, lr = 0.00442497
I0428 17:41:41.703379  8468 solver.cpp:218] Iteration 4128 (2.21785 iter/s, 5.41066s/12 iters), loss = 1.28033
I0428 17:41:41.703420  8468 solver.cpp:237]     Train net output #0: loss = 1.28033 (* 1 = 1.28033 loss)
I0428 17:41:41.703431  8468 sgd_solver.cpp:105] Iteration 4128, lr = 0.00441447
I0428 17:41:47.171339  8468 solver.cpp:218] Iteration 4140 (2.19471 iter/s, 5.46769s/12 iters), loss = 1.34781
I0428 17:41:47.171373  8468 solver.cpp:237]     Train net output #0: loss = 1.34781 (* 1 = 1.34781 loss)
I0428 17:41:47.171383  8468 sgd_solver.cpp:105] Iteration 4140, lr = 0.00440398
I0428 17:41:49.974761  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:41:52.768117  8468 solver.cpp:218] Iteration 4152 (2.14419 iter/s, 5.59651s/12 iters), loss = 1.26655
I0428 17:41:52.768160  8468 solver.cpp:237]     Train net output #0: loss = 1.26655 (* 1 = 1.26655 loss)
I0428 17:41:52.768169  8468 sgd_solver.cpp:105] Iteration 4152, lr = 0.00439353
I0428 17:41:53.128172  8468 blocking_queue.cpp:49] Waiting for data
I0428 17:41:58.098143  8468 solver.cpp:218] Iteration 4164 (2.25151 iter/s, 5.32975s/12 iters), loss = 1.20197
I0428 17:41:58.098278  8468 solver.cpp:237]     Train net output #0: loss = 1.20197 (* 1 = 1.20197 loss)
I0428 17:41:58.098291  8468 sgd_solver.cpp:105] Iteration 4164, lr = 0.0043831
I0428 17:42:03.444983  8468 solver.cpp:218] Iteration 4176 (2.24447 iter/s, 5.34647s/12 iters), loss = 1.14274
I0428 17:42:03.445034  8468 solver.cpp:237]     Train net output #0: loss = 1.14274 (* 1 = 1.14274 loss)
I0428 17:42:03.445048  8468 sgd_solver.cpp:105] Iteration 4176, lr = 0.00437269
I0428 17:42:05.580094  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4182.caffemodel
I0428 17:42:07.314091  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4182.solverstate
I0428 17:42:07.986655  8468 solver.cpp:330] Iteration 4182, Testing net (#0)
I0428 17:42:07.986675  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:42:10.820832  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:42:12.563719  8468 solver.cpp:397]     Test net output #0: accuracy = 0.334559
I0428 17:42:12.563752  8468 solver.cpp:397]     Test net output #1: loss = 3.03912 (* 1 = 3.03912 loss)
I0428 17:42:14.558014  8468 solver.cpp:218] Iteration 4188 (1.07986 iter/s, 11.1125s/12 iters), loss = 1.18118
I0428 17:42:14.558063  8468 solver.cpp:237]     Train net output #0: loss = 1.18118 (* 1 = 1.18118 loss)
I0428 17:42:14.558073  8468 sgd_solver.cpp:105] Iteration 4188, lr = 0.00436231
I0428 17:42:19.949329  8468 solver.cpp:218] Iteration 4200 (2.22592 iter/s, 5.39104s/12 iters), loss = 1.19428
I0428 17:42:19.949368  8468 solver.cpp:237]     Train net output #0: loss = 1.19428 (* 1 = 1.19428 loss)
I0428 17:42:19.949376  8468 sgd_solver.cpp:105] Iteration 4200, lr = 0.00435195
I0428 17:42:25.426604  8468 solver.cpp:218] Iteration 4212 (2.19098 iter/s, 5.477s/12 iters), loss = 1.15763
I0428 17:42:25.426645  8468 solver.cpp:237]     Train net output #0: loss = 1.15763 (* 1 = 1.15763 loss)
I0428 17:42:25.426653  8468 sgd_solver.cpp:105] Iteration 4212, lr = 0.00434162
I0428 17:42:30.774989  8468 solver.cpp:218] Iteration 4224 (2.24378 iter/s, 5.34811s/12 iters), loss = 1.38209
I0428 17:42:30.775116  8468 solver.cpp:237]     Train net output #0: loss = 1.38209 (* 1 = 1.38209 loss)
I0428 17:42:30.775125  8468 sgd_solver.cpp:105] Iteration 4224, lr = 0.00433131
I0428 17:42:36.247613  8468 solver.cpp:218] Iteration 4236 (2.19288 iter/s, 5.47226s/12 iters), loss = 1.30894
I0428 17:42:36.247661  8468 solver.cpp:237]     Train net output #0: loss = 1.30894 (* 1 = 1.30894 loss)
I0428 17:42:36.247673  8468 sgd_solver.cpp:105] Iteration 4236, lr = 0.00432103
I0428 17:42:41.384080  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:42:41.650053  8468 solver.cpp:218] Iteration 4248 (2.22134 iter/s, 5.40216s/12 iters), loss = 1.09823
I0428 17:42:41.650111  8468 solver.cpp:237]     Train net output #0: loss = 1.09823 (* 1 = 1.09823 loss)
I0428 17:42:41.650127  8468 sgd_solver.cpp:105] Iteration 4248, lr = 0.00431077
I0428 17:42:47.069272  8468 solver.cpp:218] Iteration 4260 (2.21446 iter/s, 5.41893s/12 iters), loss = 1.08128
I0428 17:42:47.069317  8468 solver.cpp:237]     Train net output #0: loss = 1.08128 (* 1 = 1.08128 loss)
I0428 17:42:47.069329  8468 sgd_solver.cpp:105] Iteration 4260, lr = 0.00430053
I0428 17:42:52.496062  8468 solver.cpp:218] Iteration 4272 (2.21136 iter/s, 5.42652s/12 iters), loss = 1.25113
I0428 17:42:52.496100  8468 solver.cpp:237]     Train net output #0: loss = 1.25113 (* 1 = 1.25113 loss)
I0428 17:42:52.496109  8468 sgd_solver.cpp:105] Iteration 4272, lr = 0.00429032
I0428 17:42:57.329811  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4284.caffemodel
I0428 17:42:59.192422  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4284.solverstate
I0428 17:43:00.632599  8468 solver.cpp:330] Iteration 4284, Testing net (#0)
I0428 17:43:00.632623  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:43:03.657964  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:43:05.458074  8468 solver.cpp:397]     Test net output #0: accuracy = 0.353554
I0428 17:43:05.458110  8468 solver.cpp:397]     Test net output #1: loss = 3.03418 (* 1 = 3.03418 loss)
I0428 17:43:05.584740  8468 solver.cpp:218] Iteration 4284 (0.916863 iter/s, 13.0881s/12 iters), loss = 1.34198
I0428 17:43:05.584791  8468 solver.cpp:237]     Train net output #0: loss = 1.34198 (* 1 = 1.34198 loss)
I0428 17:43:05.584802  8468 sgd_solver.cpp:105] Iteration 4284, lr = 0.00428014
I0428 17:43:10.077638  8468 solver.cpp:218] Iteration 4296 (2.67103 iter/s, 4.49265s/12 iters), loss = 1.0241
I0428 17:43:10.077680  8468 solver.cpp:237]     Train net output #0: loss = 1.0241 (* 1 = 1.0241 loss)
I0428 17:43:10.077690  8468 sgd_solver.cpp:105] Iteration 4296, lr = 0.00426998
I0428 17:43:16.007647  8468 solver.cpp:218] Iteration 4308 (2.02371 iter/s, 5.92971s/12 iters), loss = 1.50578
I0428 17:43:16.007683  8468 solver.cpp:237]     Train net output #0: loss = 1.50578 (* 1 = 1.50578 loss)
I0428 17:43:16.007692  8468 sgd_solver.cpp:105] Iteration 4308, lr = 0.00425984
I0428 17:43:21.662235  8468 solver.cpp:218] Iteration 4320 (2.12228 iter/s, 5.65431s/12 iters), loss = 1.21202
I0428 17:43:21.662282  8468 solver.cpp:237]     Train net output #0: loss = 1.21202 (* 1 = 1.21202 loss)
I0428 17:43:21.662294  8468 sgd_solver.cpp:105] Iteration 4320, lr = 0.00424972
I0428 17:43:27.286556  8468 solver.cpp:218] Iteration 4332 (2.1337 iter/s, 5.62403s/12 iters), loss = 1.16968
I0428 17:43:27.286597  8468 solver.cpp:237]     Train net output #0: loss = 1.16968 (* 1 = 1.16968 loss)
I0428 17:43:27.286607  8468 sgd_solver.cpp:105] Iteration 4332, lr = 0.00423964
I0428 17:43:32.676374  8468 solver.cpp:218] Iteration 4344 (2.22653 iter/s, 5.38955s/12 iters), loss = 1.28333
I0428 17:43:32.676414  8468 solver.cpp:237]     Train net output #0: loss = 1.28333 (* 1 = 1.28333 loss)
I0428 17:43:32.676422  8468 sgd_solver.cpp:105] Iteration 4344, lr = 0.00422957
I0428 17:43:34.704982  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:43:38.046250  8468 solver.cpp:218] Iteration 4356 (2.2348 iter/s, 5.36961s/12 iters), loss = 0.945464
I0428 17:43:38.046291  8468 solver.cpp:237]     Train net output #0: loss = 0.945464 (* 1 = 0.945464 loss)
I0428 17:43:38.046301  8468 sgd_solver.cpp:105] Iteration 4356, lr = 0.00421953
I0428 17:43:43.424031  8468 solver.cpp:218] Iteration 4368 (2.23152 iter/s, 5.37751s/12 iters), loss = 1.09019
I0428 17:43:43.424069  8468 solver.cpp:237]     Train net output #0: loss = 1.09019 (* 1 = 1.09019 loss)
I0428 17:43:43.424078  8468 sgd_solver.cpp:105] Iteration 4368, lr = 0.00420951
I0428 17:43:48.882064  8468 solver.cpp:218] Iteration 4380 (2.1987 iter/s, 5.45777s/12 iters), loss = 0.977747
I0428 17:43:48.882104  8468 solver.cpp:237]     Train net output #0: loss = 0.977747 (* 1 = 0.977747 loss)
I0428 17:43:48.882113  8468 sgd_solver.cpp:105] Iteration 4380, lr = 0.00419952
I0428 17:43:51.093924  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4386.caffemodel
I0428 17:43:53.122642  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4386.solverstate
I0428 17:43:54.132355  8468 solver.cpp:330] Iteration 4386, Testing net (#0)
I0428 17:43:54.132380  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:43:56.943043  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:43:58.778780  8468 solver.cpp:397]     Test net output #0: accuracy = 0.356005
I0428 17:43:58.778815  8468 solver.cpp:397]     Test net output #1: loss = 2.95882 (* 1 = 2.95882 loss)
I0428 17:44:00.810194  8468 solver.cpp:218] Iteration 4392 (1.00607 iter/s, 11.9276s/12 iters), loss = 1.06655
I0428 17:44:00.810236  8468 solver.cpp:237]     Train net output #0: loss = 1.06655 (* 1 = 1.06655 loss)
I0428 17:44:00.810245  8468 sgd_solver.cpp:105] Iteration 4392, lr = 0.00418954
I0428 17:44:06.292796  8468 solver.cpp:218] Iteration 4404 (2.18885 iter/s, 5.48232s/12 iters), loss = 1.3072
I0428 17:44:06.292909  8468 solver.cpp:237]     Train net output #0: loss = 1.3072 (* 1 = 1.3072 loss)
I0428 17:44:06.292919  8468 sgd_solver.cpp:105] Iteration 4404, lr = 0.0041796
I0428 17:44:11.696715  8468 solver.cpp:218] Iteration 4416 (2.22075 iter/s, 5.40357s/12 iters), loss = 1.26519
I0428 17:44:11.696768  8468 solver.cpp:237]     Train net output #0: loss = 1.26519 (* 1 = 1.26519 loss)
I0428 17:44:11.696779  8468 sgd_solver.cpp:105] Iteration 4416, lr = 0.00416967
I0428 17:44:17.034554  8468 solver.cpp:218] Iteration 4428 (2.24822 iter/s, 5.33756s/12 iters), loss = 1.21158
I0428 17:44:17.034590  8468 solver.cpp:237]     Train net output #0: loss = 1.21158 (* 1 = 1.21158 loss)
I0428 17:44:17.034598  8468 sgd_solver.cpp:105] Iteration 4428, lr = 0.00415977
I0428 17:44:22.667269  8468 solver.cpp:218] Iteration 4440 (2.13052 iter/s, 5.63244s/12 iters), loss = 1.24009
I0428 17:44:22.667311  8468 solver.cpp:237]     Train net output #0: loss = 1.24009 (* 1 = 1.24009 loss)
I0428 17:44:22.667320  8468 sgd_solver.cpp:105] Iteration 4440, lr = 0.0041499
I0428 17:44:26.965427  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:44:28.015843  8468 solver.cpp:218] Iteration 4452 (2.2437 iter/s, 5.3483s/12 iters), loss = 1.18084
I0428 17:44:28.015889  8468 solver.cpp:237]     Train net output #0: loss = 1.18084 (* 1 = 1.18084 loss)
I0428 17:44:28.015899  8468 sgd_solver.cpp:105] Iteration 4452, lr = 0.00414005
I0428 17:44:33.569205  8468 solver.cpp:218] Iteration 4464 (2.16096 iter/s, 5.55308s/12 iters), loss = 1.11352
I0428 17:44:33.569245  8468 solver.cpp:237]     Train net output #0: loss = 1.11352 (* 1 = 1.11352 loss)
I0428 17:44:33.569254  8468 sgd_solver.cpp:105] Iteration 4464, lr = 0.00413022
I0428 17:44:38.997537  8468 solver.cpp:218] Iteration 4476 (2.21074 iter/s, 5.42806s/12 iters), loss = 1.19322
I0428 17:44:38.997700  8468 solver.cpp:237]     Train net output #0: loss = 1.19322 (* 1 = 1.19322 loss)
I0428 17:44:38.997714  8468 sgd_solver.cpp:105] Iteration 4476, lr = 0.00412041
I0428 17:44:43.895462  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4488.caffemodel
I0428 17:44:44.999861  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4488.solverstate
I0428 17:44:45.919193  8468 solver.cpp:330] Iteration 4488, Testing net (#0)
I0428 17:44:45.919212  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:44:48.784343  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:44:50.637547  8468 solver.cpp:397]     Test net output #0: accuracy = 0.338235
I0428 17:44:50.637583  8468 solver.cpp:397]     Test net output #1: loss = 3.08541 (* 1 = 3.08541 loss)
I0428 17:44:50.769520  8468 solver.cpp:218] Iteration 4488 (1.01943 iter/s, 11.7713s/12 iters), loss = 1.05536
I0428 17:44:50.769562  8468 solver.cpp:237]     Train net output #0: loss = 1.05536 (* 1 = 1.05536 loss)
I0428 17:44:50.769570  8468 sgd_solver.cpp:105] Iteration 4488, lr = 0.00411063
I0428 17:44:55.424571  8468 solver.cpp:218] Iteration 4500 (2.57798 iter/s, 4.65481s/12 iters), loss = 0.914786
I0428 17:44:55.424618  8468 solver.cpp:237]     Train net output #0: loss = 0.914786 (* 1 = 0.914786 loss)
I0428 17:44:55.424628  8468 sgd_solver.cpp:105] Iteration 4500, lr = 0.00410087
I0428 17:45:01.082696  8468 solver.cpp:218] Iteration 4512 (2.12095 iter/s, 5.65784s/12 iters), loss = 1.13059
I0428 17:45:01.082739  8468 solver.cpp:237]     Train net output #0: loss = 1.13059 (* 1 = 1.13059 loss)
I0428 17:45:01.082751  8468 sgd_solver.cpp:105] Iteration 4512, lr = 0.00409113
I0428 17:45:06.534590  8468 solver.cpp:218] Iteration 4524 (2.20118 iter/s, 5.45162s/12 iters), loss = 0.957494
I0428 17:45:06.534627  8468 solver.cpp:237]     Train net output #0: loss = 0.957494 (* 1 = 0.957494 loss)
I0428 17:45:06.534638  8468 sgd_solver.cpp:105] Iteration 4524, lr = 0.00408142
I0428 17:45:12.138401  8468 solver.cpp:218] Iteration 4536 (2.14151 iter/s, 5.60353s/12 iters), loss = 1.08411
I0428 17:45:12.138527  8468 solver.cpp:237]     Train net output #0: loss = 1.08411 (* 1 = 1.08411 loss)
I0428 17:45:12.138536  8468 sgd_solver.cpp:105] Iteration 4536, lr = 0.00407173
I0428 17:45:17.556891  8468 solver.cpp:218] Iteration 4548 (2.21478 iter/s, 5.41814s/12 iters), loss = 0.907159
I0428 17:45:17.556929  8468 solver.cpp:237]     Train net output #0: loss = 0.907159 (* 1 = 0.907159 loss)
I0428 17:45:17.556941  8468 sgd_solver.cpp:105] Iteration 4548, lr = 0.00406206
I0428 17:45:18.931892  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:45:22.968148  8468 solver.cpp:218] Iteration 4560 (2.21771 iter/s, 5.41098s/12 iters), loss = 1.24683
I0428 17:45:22.968201  8468 solver.cpp:237]     Train net output #0: loss = 1.24683 (* 1 = 1.24683 loss)
I0428 17:45:22.968214  8468 sgd_solver.cpp:105] Iteration 4560, lr = 0.00405242
I0428 17:45:28.418013  8468 solver.cpp:218] Iteration 4572 (2.202 iter/s, 5.44958s/12 iters), loss = 1.11551
I0428 17:45:28.418051  8468 solver.cpp:237]     Train net output #0: loss = 1.11551 (* 1 = 1.11551 loss)
I0428 17:45:28.418062  8468 sgd_solver.cpp:105] Iteration 4572, lr = 0.0040428
I0428 17:45:33.880229  8468 solver.cpp:218] Iteration 4584 (2.19702 iter/s, 5.46194s/12 iters), loss = 0.873448
I0428 17:45:33.880266  8468 solver.cpp:237]     Train net output #0: loss = 0.873448 (* 1 = 0.873448 loss)
I0428 17:45:33.880275  8468 sgd_solver.cpp:105] Iteration 4584, lr = 0.0040332
I0428 17:45:36.094099  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4590.caffemodel
I0428 17:45:38.328698  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4590.solverstate
I0428 17:45:40.972591  8468 solver.cpp:330] Iteration 4590, Testing net (#0)
I0428 17:45:40.972613  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:45:43.677330  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:45:45.901691  8468 solver.cpp:397]     Test net output #0: accuracy = 0.371324
I0428 17:45:45.901743  8468 solver.cpp:397]     Test net output #1: loss = 3.11403 (* 1 = 3.11403 loss)
I0428 17:45:47.962476  8468 solver.cpp:218] Iteration 4596 (0.852174 iter/s, 14.0816s/12 iters), loss = 1.14536
I0428 17:45:47.962517  8468 solver.cpp:237]     Train net output #0: loss = 1.14536 (* 1 = 1.14536 loss)
I0428 17:45:47.962528  8468 sgd_solver.cpp:105] Iteration 4596, lr = 0.00402362
I0428 17:45:53.447772  8468 solver.cpp:218] Iteration 4608 (2.18777 iter/s, 5.48503s/12 iters), loss = 1.03592
I0428 17:45:53.447804  8468 solver.cpp:237]     Train net output #0: loss = 1.03592 (* 1 = 1.03592 loss)
I0428 17:45:53.447813  8468 sgd_solver.cpp:105] Iteration 4608, lr = 0.00401407
I0428 17:45:58.907783  8468 solver.cpp:218] Iteration 4620 (2.1979 iter/s, 5.45975s/12 iters), loss = 1.20053
I0428 17:45:58.907819  8468 solver.cpp:237]     Train net output #0: loss = 1.20053 (* 1 = 1.20053 loss)
I0428 17:45:58.907827  8468 sgd_solver.cpp:105] Iteration 4620, lr = 0.00400454
I0428 17:46:04.459268  8468 solver.cpp:218] Iteration 4632 (2.16169 iter/s, 5.55121s/12 iters), loss = 1.00673
I0428 17:46:04.459306  8468 solver.cpp:237]     Train net output #0: loss = 1.00673 (* 1 = 1.00673 loss)
I0428 17:46:04.459316  8468 sgd_solver.cpp:105] Iteration 4632, lr = 0.00399503
I0428 17:46:10.043920  8468 solver.cpp:218] Iteration 4644 (2.14885 iter/s, 5.58438s/12 iters), loss = 0.95195
I0428 17:46:10.043962  8468 solver.cpp:237]     Train net output #0: loss = 0.95195 (* 1 = 0.95195 loss)
I0428 17:46:10.043972  8468 sgd_solver.cpp:105] Iteration 4644, lr = 0.00398555
I0428 17:46:13.816215  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:46:15.771601  8468 solver.cpp:218] Iteration 4656 (2.09519 iter/s, 5.7274s/12 iters), loss = 1.01703
I0428 17:46:15.771638  8468 solver.cpp:237]     Train net output #0: loss = 1.01703 (* 1 = 1.01703 loss)
I0428 17:46:15.771646  8468 sgd_solver.cpp:105] Iteration 4656, lr = 0.00397608
I0428 17:46:21.143431  8468 solver.cpp:218] Iteration 4668 (2.23399 iter/s, 5.37156s/12 iters), loss = 1.0241
I0428 17:46:21.143473  8468 solver.cpp:237]     Train net output #0: loss = 1.0241 (* 1 = 1.0241 loss)
I0428 17:46:21.143481  8468 sgd_solver.cpp:105] Iteration 4668, lr = 0.00396664
I0428 17:46:26.661335  8468 solver.cpp:218] Iteration 4680 (2.17485 iter/s, 5.51763s/12 iters), loss = 0.960551
I0428 17:46:26.661386  8468 solver.cpp:237]     Train net output #0: loss = 0.960551 (* 1 = 0.960551 loss)
I0428 17:46:26.661394  8468 sgd_solver.cpp:105] Iteration 4680, lr = 0.00395723
I0428 17:46:31.537106  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4692.caffemodel
I0428 17:46:32.128628  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4692.solverstate
I0428 17:46:32.570268  8468 solver.cpp:330] Iteration 4692, Testing net (#0)
I0428 17:46:32.570286  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:46:35.239161  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:46:37.167634  8468 solver.cpp:397]     Test net output #0: accuracy = 0.367647
I0428 17:46:37.167670  8468 solver.cpp:397]     Test net output #1: loss = 3.1164 (* 1 = 3.1164 loss)
I0428 17:46:37.299407  8468 solver.cpp:218] Iteration 4692 (1.12808 iter/s, 10.6376s/12 iters), loss = 1.11412
I0428 17:46:37.299458  8468 solver.cpp:237]     Train net output #0: loss = 1.11412 (* 1 = 1.11412 loss)
I0428 17:46:37.299469  8468 sgd_solver.cpp:105] Iteration 4692, lr = 0.00394783
I0428 17:46:41.791623  8468 solver.cpp:218] Iteration 4704 (2.67143 iter/s, 4.49197s/12 iters), loss = 0.890307
I0428 17:46:41.791666  8468 solver.cpp:237]     Train net output #0: loss = 0.890307 (* 1 = 0.890307 loss)
I0428 17:46:41.791676  8468 sgd_solver.cpp:105] Iteration 4704, lr = 0.00393846
I0428 17:46:47.302503  8468 solver.cpp:218] Iteration 4716 (2.17762 iter/s, 5.5106s/12 iters), loss = 1.2413
I0428 17:46:47.302645  8468 solver.cpp:237]     Train net output #0: loss = 1.2413 (* 1 = 1.2413 loss)
I0428 17:46:47.302657  8468 sgd_solver.cpp:105] Iteration 4716, lr = 0.00392911
I0428 17:46:52.831156  8468 solver.cpp:218] Iteration 4728 (2.17066 iter/s, 5.52828s/12 iters), loss = 1.10757
I0428 17:46:52.831197  8468 solver.cpp:237]     Train net output #0: loss = 1.10757 (* 1 = 1.10757 loss)
I0428 17:46:52.831205  8468 sgd_solver.cpp:105] Iteration 4728, lr = 0.00391978
I0428 17:46:58.225349  8468 solver.cpp:218] Iteration 4740 (2.22473 iter/s, 5.39392s/12 iters), loss = 1.21627
I0428 17:46:58.225405  8468 solver.cpp:237]     Train net output #0: loss = 1.21627 (* 1 = 1.21627 loss)
I0428 17:46:58.225420  8468 sgd_solver.cpp:105] Iteration 4740, lr = 0.00391047
I0428 17:47:03.542412  8468 solver.cpp:218] Iteration 4752 (2.25701 iter/s, 5.31678s/12 iters), loss = 0.860933
I0428 17:47:03.542466  8468 solver.cpp:237]     Train net output #0: loss = 0.860933 (* 1 = 0.860933 loss)
I0428 17:47:03.542479  8468 sgd_solver.cpp:105] Iteration 4752, lr = 0.00390119
I0428 17:47:04.106925  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:47:08.921983  8468 solver.cpp:218] Iteration 4764 (2.23078 iter/s, 5.37929s/12 iters), loss = 0.802402
I0428 17:47:08.922019  8468 solver.cpp:237]     Train net output #0: loss = 0.802402 (* 1 = 0.802402 loss)
I0428 17:47:08.922029  8468 sgd_solver.cpp:105] Iteration 4764, lr = 0.00389193
I0428 17:47:14.427367  8468 solver.cpp:218] Iteration 4776 (2.17979 iter/s, 5.50511s/12 iters), loss = 0.743593
I0428 17:47:14.427405  8468 solver.cpp:237]     Train net output #0: loss = 0.743593 (* 1 = 0.743593 loss)
I0428 17:47:14.427415  8468 sgd_solver.cpp:105] Iteration 4776, lr = 0.00388269
I0428 17:47:19.884403  8468 solver.cpp:218] Iteration 4788 (2.1991 iter/s, 5.45677s/12 iters), loss = 0.975918
I0428 17:47:19.884546  8468 solver.cpp:237]     Train net output #0: loss = 0.975918 (* 1 = 0.975918 loss)
I0428 17:47:19.884554  8468 sgd_solver.cpp:105] Iteration 4788, lr = 0.00387347
I0428 17:47:22.088258  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4794.caffemodel
I0428 17:47:23.305272  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4794.solverstate
I0428 17:47:25.367741  8468 solver.cpp:330] Iteration 4794, Testing net (#0)
I0428 17:47:25.367763  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:47:28.296888  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:47:30.355350  8468 solver.cpp:397]     Test net output #0: accuracy = 0.36826
I0428 17:47:30.355384  8468 solver.cpp:397]     Test net output #1: loss = 3.16746 (* 1 = 3.16746 loss)
I0428 17:47:32.295805  8468 solver.cpp:218] Iteration 4800 (0.966904 iter/s, 12.4108s/12 iters), loss = 0.93234
I0428 17:47:32.295846  8468 solver.cpp:237]     Train net output #0: loss = 0.93234 (* 1 = 0.93234 loss)
I0428 17:47:32.295856  8468 sgd_solver.cpp:105] Iteration 4800, lr = 0.00386427
I0428 17:47:37.695345  8468 solver.cpp:218] Iteration 4812 (2.22252 iter/s, 5.39927s/12 iters), loss = 0.914465
I0428 17:47:37.695387  8468 solver.cpp:237]     Train net output #0: loss = 0.914465 (* 1 = 0.914465 loss)
I0428 17:47:37.695397  8468 sgd_solver.cpp:105] Iteration 4812, lr = 0.0038551
I0428 17:47:43.218232  8468 solver.cpp:218] Iteration 4824 (2.17289 iter/s, 5.5226s/12 iters), loss = 1.05685
I0428 17:47:43.218281  8468 solver.cpp:237]     Train net output #0: loss = 1.05685 (* 1 = 1.05685 loss)
I0428 17:47:43.218293  8468 sgd_solver.cpp:105] Iteration 4824, lr = 0.00384594
I0428 17:47:48.654718  8468 solver.cpp:218] Iteration 4836 (2.20742 iter/s, 5.43621s/12 iters), loss = 1.00205
I0428 17:47:48.654762  8468 solver.cpp:237]     Train net output #0: loss = 1.00205 (* 1 = 1.00205 loss)
I0428 17:47:48.654772  8468 sgd_solver.cpp:105] Iteration 4836, lr = 0.00383681
I0428 17:47:49.461032  8468 blocking_queue.cpp:49] Waiting for data
I0428 17:47:54.020440  8468 solver.cpp:218] Iteration 4848 (2.23653 iter/s, 5.36545s/12 iters), loss = 0.740278
I0428 17:47:54.020606  8468 solver.cpp:237]     Train net output #0: loss = 0.740278 (* 1 = 0.740278 loss)
I0428 17:47:54.020615  8468 sgd_solver.cpp:105] Iteration 4848, lr = 0.0038277
I0428 17:47:56.955641  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:47:59.535576  8468 solver.cpp:218] Iteration 4860 (2.17599 iter/s, 5.51474s/12 iters), loss = 0.77842
I0428 17:47:59.535616  8468 solver.cpp:237]     Train net output #0: loss = 0.77842 (* 1 = 0.77842 loss)
I0428 17:47:59.535624  8468 sgd_solver.cpp:105] Iteration 4860, lr = 0.00381862
I0428 17:48:05.018786  8468 solver.cpp:218] Iteration 4872 (2.18861 iter/s, 5.48292s/12 iters), loss = 0.984421
I0428 17:48:05.018842  8468 solver.cpp:237]     Train net output #0: loss = 0.984421 (* 1 = 0.984421 loss)
I0428 17:48:05.018855  8468 sgd_solver.cpp:105] Iteration 4872, lr = 0.00380955
I0428 17:48:10.375396  8468 solver.cpp:218] Iteration 4884 (2.24034 iter/s, 5.35632s/12 iters), loss = 0.758592
I0428 17:48:10.375439  8468 solver.cpp:237]     Train net output #0: loss = 0.758592 (* 1 = 0.758592 loss)
I0428 17:48:10.375449  8468 sgd_solver.cpp:105] Iteration 4884, lr = 0.0038005
I0428 17:48:15.432865  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4896.caffemodel
I0428 17:48:16.008826  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4896.solverstate
I0428 17:48:16.458868  8468 solver.cpp:330] Iteration 4896, Testing net (#0)
I0428 17:48:16.458889  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:48:19.079165  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:48:21.189038  8468 solver.cpp:397]     Test net output #0: accuracy = 0.371324
I0428 17:48:21.189069  8468 solver.cpp:397]     Test net output #1: loss = 3.11304 (* 1 = 3.11304 loss)
I0428 17:48:21.313225  8468 solver.cpp:218] Iteration 4896 (1.09716 iter/s, 10.9373s/12 iters), loss = 0.874247
I0428 17:48:21.313263  8468 solver.cpp:237]     Train net output #0: loss = 0.874247 (* 1 = 0.874247 loss)
I0428 17:48:21.313271  8468 sgd_solver.cpp:105] Iteration 4896, lr = 0.00379148
I0428 17:48:25.793380  8468 solver.cpp:218] Iteration 4908 (2.67862 iter/s, 4.47992s/12 iters), loss = 0.863705
I0428 17:48:25.793504  8468 solver.cpp:237]     Train net output #0: loss = 0.863705 (* 1 = 0.863705 loss)
I0428 17:48:25.793515  8468 sgd_solver.cpp:105] Iteration 4908, lr = 0.00378248
I0428 17:48:31.227005  8468 solver.cpp:218] Iteration 4920 (2.20862 iter/s, 5.43327s/12 iters), loss = 0.910319
I0428 17:48:31.227044  8468 solver.cpp:237]     Train net output #0: loss = 0.910319 (* 1 = 0.910319 loss)
I0428 17:48:31.227053  8468 sgd_solver.cpp:105] Iteration 4920, lr = 0.0037735
I0428 17:48:36.659436  8468 solver.cpp:218] Iteration 4932 (2.20907 iter/s, 5.43216s/12 iters), loss = 0.999637
I0428 17:48:36.659473  8468 solver.cpp:237]     Train net output #0: loss = 0.999637 (* 1 = 0.999637 loss)
I0428 17:48:36.659485  8468 sgd_solver.cpp:105] Iteration 4932, lr = 0.00376454
I0428 17:48:42.208092  8468 solver.cpp:218] Iteration 4944 (2.1628 iter/s, 5.54837s/12 iters), loss = 0.832636
I0428 17:48:42.208143  8468 solver.cpp:237]     Train net output #0: loss = 0.832636 (* 1 = 0.832636 loss)
I0428 17:48:42.208154  8468 sgd_solver.cpp:105] Iteration 4944, lr = 0.0037556
I0428 17:48:47.278393  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:48:47.513454  8468 solver.cpp:218] Iteration 4956 (2.26198 iter/s, 5.30508s/12 iters), loss = 0.856759
I0428 17:48:47.513494  8468 solver.cpp:237]     Train net output #0: loss = 0.856759 (* 1 = 0.856759 loss)
I0428 17:48:47.513504  8468 sgd_solver.cpp:105] Iteration 4956, lr = 0.00374669
I0428 17:48:52.938333  8468 solver.cpp:218] Iteration 4968 (2.21214 iter/s, 5.4246s/12 iters), loss = 0.666333
I0428 17:48:52.938383  8468 solver.cpp:237]     Train net output #0: loss = 0.666333 (* 1 = 0.666333 loss)
I0428 17:48:52.938395  8468 sgd_solver.cpp:105] Iteration 4968, lr = 0.00373779
I0428 17:48:58.404229  8468 solver.cpp:218] Iteration 4980 (2.19555 iter/s, 5.46561s/12 iters), loss = 0.65488
I0428 17:48:58.404414  8468 solver.cpp:237]     Train net output #0: loss = 0.65488 (* 1 = 0.65488 loss)
I0428 17:48:58.404428  8468 sgd_solver.cpp:105] Iteration 4980, lr = 0.00372892
I0428 17:49:03.967875  8468 solver.cpp:218] Iteration 4992 (2.15702 iter/s, 5.56323s/12 iters), loss = 0.786931
I0428 17:49:03.967911  8468 solver.cpp:237]     Train net output #0: loss = 0.786931 (* 1 = 0.786931 loss)
I0428 17:49:03.967919  8468 sgd_solver.cpp:105] Iteration 4992, lr = 0.00372006
I0428 17:49:06.186271  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4998.caffemodel
I0428 17:49:07.745800  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4998.solverstate
I0428 17:49:09.160414  8468 solver.cpp:330] Iteration 4998, Testing net (#0)
I0428 17:49:09.160432  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:49:11.815682  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:49:13.961311  8468 solver.cpp:397]     Test net output #0: accuracy = 0.390931
I0428 17:49:13.961351  8468 solver.cpp:397]     Test net output #1: loss = 3.12209 (* 1 = 3.12209 loss)
I0428 17:49:15.979162  8468 solver.cpp:218] Iteration 5004 (0.999105 iter/s, 12.0107s/12 iters), loss = 0.995208
I0428 17:49:15.979208  8468 solver.cpp:237]     Train net output #0: loss = 0.995208 (* 1 = 0.995208 loss)
I0428 17:49:15.979219  8468 sgd_solver.cpp:105] Iteration 5004, lr = 0.00371123
I0428 17:49:21.374142  8468 solver.cpp:218] Iteration 5016 (2.22441 iter/s, 5.3947s/12 iters), loss = 0.785567
I0428 17:49:21.374181  8468 solver.cpp:237]     Train net output #0: loss = 0.785567 (* 1 = 0.785567 loss)
I0428 17:49:21.374191  8468 sgd_solver.cpp:105] Iteration 5016, lr = 0.00370242
I0428 17:49:26.786681  8468 solver.cpp:218] Iteration 5028 (2.21719 iter/s, 5.41226s/12 iters), loss = 0.72998
I0428 17:49:26.786725  8468 solver.cpp:237]     Train net output #0: loss = 0.72998 (* 1 = 0.72998 loss)
I0428 17:49:26.786733  8468 sgd_solver.cpp:105] Iteration 5028, lr = 0.00369363
I0428 17:49:32.206027  8468 solver.cpp:218] Iteration 5040 (2.2144 iter/s, 5.41906s/12 iters), loss = 0.800788
I0428 17:49:32.206157  8468 solver.cpp:237]     Train net output #0: loss = 0.800788 (* 1 = 0.800788 loss)
I0428 17:49:32.206169  8468 sgd_solver.cpp:105] Iteration 5040, lr = 0.00368486
I0428 17:49:37.776350  8468 solver.cpp:218] Iteration 5052 (2.15442 iter/s, 5.56995s/12 iters), loss = 0.84946
I0428 17:49:37.776403  8468 solver.cpp:237]     Train net output #0: loss = 0.84946 (* 1 = 0.84946 loss)
I0428 17:49:37.776417  8468 sgd_solver.cpp:105] Iteration 5052, lr = 0.00367611
I0428 17:49:39.839232  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:49:43.230552  8468 solver.cpp:218] Iteration 5064 (2.20025 iter/s, 5.45392s/12 iters), loss = 0.762099
I0428 17:49:43.230604  8468 solver.cpp:237]     Train net output #0: loss = 0.762099 (* 1 = 0.762099 loss)
I0428 17:49:43.230618  8468 sgd_solver.cpp:105] Iteration 5064, lr = 0.00366738
I0428 17:49:48.774330  8468 solver.cpp:218] Iteration 5076 (2.1647 iter/s, 5.54349s/12 iters), loss = 0.745801
I0428 17:49:48.774369  8468 solver.cpp:237]     Train net output #0: loss = 0.745801 (* 1 = 0.745801 loss)
I0428 17:49:48.774379  8468 sgd_solver.cpp:105] Iteration 5076, lr = 0.00365868
I0428 17:49:54.178025  8468 solver.cpp:218] Iteration 5088 (2.22082 iter/s, 5.40342s/12 iters), loss = 0.814838
I0428 17:49:54.178076  8468 solver.cpp:237]     Train net output #0: loss = 0.814838 (* 1 = 0.814838 loss)
I0428 17:49:54.178087  8468 sgd_solver.cpp:105] Iteration 5088, lr = 0.00364999
I0428 17:49:59.217453  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_5100.caffemodel
I0428 17:49:59.802321  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_5100.solverstate
I0428 17:50:00.251600  8468 solver.cpp:330] Iteration 5100, Testing net (#0)
I0428 17:50:00.251623  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:50:02.774519  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:50:05.054327  8468 solver.cpp:397]     Test net output #0: accuracy = 0.378676
I0428 17:50:05.054358  8468 solver.cpp:397]     Test net output #1: loss = 3.13944 (* 1 = 3.13944 loss)
I0428 17:50:05.185420  8468 solver.cpp:218] Iteration 5100 (1.09023 iter/s, 11.0069s/12 iters), loss = 0.730811
I0428 17:50:05.185467  8468 solver.cpp:237]     Train net output #0: loss = 0.730811 (* 1 = 0.730811 loss)
I0428 17:50:05.185477  8468 sgd_solver.cpp:105] Iteration 5100, lr = 0.00364132
I0428 17:50:09.673179  8468 solver.cpp:218] Iteration 5112 (2.67409 iter/s, 4.48751s/12 iters), loss = 0.693284
I0428 17:50:09.673236  8468 solver.cpp:237]     Train net output #0: loss = 0.693284 (* 1 = 0.693284 loss)
I0428 17:50:09.673249  8468 sgd_solver.cpp:105] Iteration 5112, lr = 0.00363268
I0428 17:50:15.048292  8468 solver.cpp:218] Iteration 5124 (2.23263 iter/s, 5.37482s/12 iters), loss = 0.900779
I0428 17:50:15.048346  8468 solver.cpp:237]     Train net output #0: loss = 0.900779 (* 1 = 0.900779 loss)
I0428 17:50:15.048358  8468 sgd_solver.cpp:105] Iteration 5124, lr = 0.00362405
I0428 17:50:20.472878  8468 solver.cpp:218] Iteration 5136 (2.21227 iter/s, 5.4243s/12 iters), loss = 0.645711
I0428 17:50:20.472935  8468 solver.cpp:237]     Train net output #0: loss = 0.645711 (* 1 = 0.645711 loss)
I0428 17:50:20.472949  8468 sgd_solver.cpp:105] Iteration 5136, lr = 0.00361545
I0428 17:50:26.068372  8468 solver.cpp:218] Iteration 5148 (2.1447 iter/s, 5.5952s/12 iters), loss = 0.900007
I0428 17:50:26.068408  8468 solver.cpp:237]     Train net output #0: loss = 0.900007 (* 1 = 0.900007 loss)
I0428 17:50:26.068418  8468 sgd_solver.cpp:105] Iteration 5148, lr = 0.00360687
I0428 17:50:30.481637  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:50:31.508385  8468 solver.cpp:218] Iteration 5160 (2.20599 iter/s, 5.43974s/12 iters), loss = 0.736091
I0428 17:50:31.508433  8468 solver.cpp:237]     Train net output #0: loss = 0.736091 (* 1 = 0.736091 loss)
I0428 17:50:31.508445  8468 sgd_solver.cpp:105] Iteration 5160, lr = 0.0035983
I0428 17:50:36.874819  8468 solver.cpp:218] Iteration 5172 (2.23624 iter/s, 5.36615s/12 iters), loss = 0.965784
I0428 17:50:36.874950  8468 solver.cpp:237]     Train net output #0: loss = 0.965784 (* 1 = 0.965784 loss)
I0428 17:50:36.874961  8468 sgd_solver.cpp:105] Iteration 5172, lr = 0.00358976
I0428 17:50:42.365674  8468 solver.cpp:218] Iteration 5184 (2.1856 iter/s, 5.49049s/12 iters), loss = 0.781846
I0428 17:50:42.365731  8468 solver.cpp:237]     Train net output #0: loss = 0.781846 (* 1 = 0.781846 loss)
I0428 17:50:42.365744  8468 sgd_solver.cpp:105] Iteration 5184, lr = 0.00358124
I0428 17:50:47.683837  8468 solver.cpp:218] Iteration 5196 (2.25654 iter/s, 5.31788s/12 iters), loss = 0.890157
I0428 17:50:47.683890  8468 solver.cpp:237]     Train net output #0: loss = 0.890157 (* 1 = 0.890157 loss)
I0428 17:50:47.683904  8468 sgd_solver.cpp:105] Iteration 5196, lr = 0.00357273
I0428 17:50:49.940605  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_5202.caffemodel
I0428 17:50:51.137336  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_5202.solverstate
I0428 17:50:52.426414  8468 solver.cpp:330] Iteration 5202, Testing net (#0)
I0428 17:50:52.426440  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:50:54.923692  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:50:57.054370  8468 solver.cpp:397]     Test net output #0: accuracy = 0.365809
I0428 17:50:57.054399  8468 solver.cpp:397]     Test net output #1: loss = 3.25544 (* 1 = 3.25544 loss)
I0428 17:50:59.106398  8468 solver.cpp:218] Iteration 5208 (1.0506 iter/s, 11.422s/12 iters), loss = 0.77476
I0428 17:50:59.106444  8468 solver.cpp:237]     Train net output #0: loss = 0.77476 (* 1 = 0.77476 loss)
I0428 17:50:59.106456  8468 sgd_solver.cpp:105] Iteration 5208, lr = 0.00356425
I0428 17:51:04.479846  8468 solver.cpp:218] Iteration 5220 (2.23332 iter/s, 5.37317s/12 iters), loss = 0.846049
I0428 17:51:04.479887  8468 solver.cpp:237]     Train net output #0: loss = 0.846049 (* 1 = 0.846049 loss)
I0428 17:51:04.479897  8468 sgd_solver.cpp:105] Iteration 5220, lr = 0.00355579
I0428 17:51:09.997431  8468 solver.cpp:218] Iteration 5232 (2.17497 iter/s, 5.51731s/12 iters), loss = 0.657488
I0428 17:51:09.997558  8468 solver.cpp:237]     Train net output #0: loss = 0.657488 (* 1 = 0.657488 loss)
I0428 17:51:09.997568  8468 sgd_solver.cpp:105] Iteration 5232, lr = 0.00354735
I0428 17:51:15.349637  8468 solver.cpp:218] Iteration 5244 (2.24222 iter/s, 5.35185s/12 iters), loss = 0.794706
I0428 17:51:15.349696  8468 solver.cpp:237]     Train net output #0: loss = 0.794706 (* 1 = 0.794706 loss)
I0428 17:51:15.349712  8468 sgd_solver.cpp:105] Iteration 5244, lr = 0.00353892
I0428 17:51:20.697541  8468 solver.cpp:218] Iteration 5256 (2.24399 iter/s, 5.34762s/12 iters), loss = 0.697988
I0428 17:51:20.697594  8468 solver.cpp:237]     Train net output #0: loss = 0.697988 (* 1 = 0.697988 loss)
I0428 17:51:20.697607  8468 sgd_solver.cpp:105] Iteration 5256, lr = 0.00353052
I0428 17:51:22.070025  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:51:26.007920  8468 solver.cpp:218] Iteration 5268 (2.25985 iter/s, 5.3101s/12 iters), loss = 0.653765
I0428 17:51:26.007973  8468 solver.cpp:237]     Train net output #0: loss = 0.653765 (* 1 = 0.653765 loss)
I0428 17:51:26.007985  8468 sgd_solver.cpp:105] Iteration 5268, lr = 0.00352214
I0428 17:51:31.335263  8468 solver.cpp:218] Iteration 5280 (2.25265 iter/s, 5.32706s/12 iters), loss = 0.629717
I0428 17:51:31.335319  8468 solver.cpp:237]     Train net output #0: loss = 0.629717 (* 1 = 0.629717 loss)
I0428 17:51:31.335330  8468 sgd_solver.cpp:105] Iteration 5280, lr = 0.00351378
I0428 17:51:36.870496  8468 solver.cpp:218] Iteration 5292 (2.16804 iter/s, 5.53494s/12 iters), loss = 0.529399
I0428 17:51:36.870535  8468 solver.cpp:237]     Train net output #0: loss = 0.529399 (* 1 = 0.529399 loss)
I0428 17:51:36.870546  8468 sgd_solver.cpp:105] Iteration 5292, lr = 0.00350544
I0428 17:51:41.983815  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_5304.caffemodel
I0428 17:51:42.592370  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_5304.solverstate
I0428 17:51:43.040052  8468 solver.cpp:330] Iteration 5304, Testing net (#0)
I0428 17:51:43.040076  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:51:45.400205  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:51:47.677848  8468 solver.cpp:397]     Test net output #0: accuracy = 0.384804
I0428 17:51:47.677889  8468 solver.cpp:397]     Test net output #1: loss = 3.18093 (* 1 = 3.18093 loss)
I0428 17:51:47.808743  8468 solver.cpp:218] Iteration 5304 (1.09712 iter/s, 10.9377s/12 iters), loss = 0.708282
I0428 17:51:47.808807  8468 solver.cpp:237]     Train net output #0: loss = 0.708282 (* 1 = 0.708282 loss)
I0428 17:51:47.808820  8468 sgd_solver.cpp:105] Iteration 5304, lr = 0.00349711
I0428 17:51:52.386581  8468 solver.cpp:218] Iteration 5316 (2.62148 iter/s, 4.57757s/12 iters), loss = 0.571965
I0428 17:51:52.386637  8468 solver.cpp:237]     Train net output #0: loss = 0.571965 (* 1 = 0.571965 loss)
I0428 17:51:52.386651  8468 sgd_solver.cpp:105] Iteration 5316, lr = 0.00348881
I0428 17:51:57.855113  8468 solver.cpp:218] Iteration 5328 (2.19449 iter/s, 5.46824s/12 iters), loss = 0.645203
I0428 17:51:57.855173  8468 solver.cpp:237]     Train net output #0: loss = 0.645203 (* 1 = 0.645203 loss)
I0428 17:51:57.855187  8468 sgd_solver.cpp:105] Iteration 5328, lr = 0.00348053
I0428 17:52:03.339264  8468 solver.cpp:218] Iteration 5340 (2.18824 iter/s, 5.48386s/12 iters), loss = 0.471589
I0428 17:52:03.339320  8468 solver.cpp:237]     Train net output #0: loss = 0.471589 (* 1 = 0.471589 loss)
I0428 17:52:03.339332  8468 sgd_solver.cpp:105] Iteration 5340, lr = 0.00347226
I0428 17:52:08.675074  8468 solver.cpp:218] Iteration 5352 (2.24908 iter/s, 5.33553s/12 iters), loss = 0.830403
I0428 17:52:08.675130  8468 solver.cpp:237]     Train net output #0: loss = 0.830403 (* 1 = 0.830403 loss)
I0428 17:52:08.675143  8468 sgd_solver.cpp:105] Iteration 5352, lr = 0.00346402
I0428 17:52:12.309276  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:52:13.994529  8468 solver.cpp:218] Iteration 5364 (2.25599 iter/s, 5.31917s/12 iters), loss = 0.570808
I0428 17:52:13.994586  8468 solver.cpp:237]     Train net output #0: loss = 0.570808 (* 1 = 0.570808 loss)
I0428 17:52:13.994601  8468 sgd_solver.cpp:105] Iteration 5364, lr = 0.0034558
I0428 17:52:19.309195  8468 solver.cpp:218] Iteration 5376 (2.25802 iter/s, 5.31438s/12 iters), loss = 0.681939
I0428 17:52:19.309240  8468 solver.cpp:237]     Train net output #0: loss = 0.681939 (* 1 = 0.681939 loss)
I0428 17:52:19.309255  8468 sgd_solver.cpp:105] Iteration 5376, lr = 0.00344759
I0428 17:52:24.826156  8468 solver.cpp:218] Iteration 5388 (2.17522 iter/s, 5.51668s/12 iters), loss = 0.62887
I0428 17:52:24.826197  8468 solver.cpp:237]     Train net output #0: loss = 0.62887 (* 1 = 0.62887 loss)
I0428 17:52:24.826206  8468 sgd_solver.cpp:105] Iteration 5388, lr = 0.00343941
I0428 17:52:30.311172  8468 solver.cpp:218] Iteration 5400 (2.18789 iter/s, 5.48474s/12 iters), loss = 0.616765
I0428 17:52:30.311214  8468 solver.cpp:237]     Train net output #0: loss = 0.616765 (* 1 = 0.616765 loss)
I0428 17:52:30.311223  8468 sgd_solver.cpp:105] Iteration 5400, lr = 0.00343124
I0428 17:52:32.511884  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_5406.caffemodel
I0428 17:52:33.112673  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_5406.solverstate
I0428 17:52:33.539885  8468 solver.cpp:330] Iteration 5406, Testing net (#0)
I0428 17:52:33.539903  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:52:35.964006  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:52:38.327014  8468 solver.cpp:397]     Test net output #0: accuracy = 0.392157
I0428 17:52:38.327055  8468 solver.cpp:397]     Test net output #1: loss = 3.34842 (* 1 = 3.34842 loss)
I0428 17:52:40.256183  8468 solver.cpp:218] Iteration 5412 (1.20669 iter/s, 9.94456s/12 iters), loss = 0.886899
I0428 17:52:40.256222  8468 solver.cpp:237]     Train net output #0: loss = 0.886899 (* 1 = 0.886899 loss)
I0428 17:52:40.256233  8468 sgd_solver.cpp:105] Iteration 5412, lr = 0.00342309
I0428 17:52:45.822803  8468 solver.cpp:218] Iteration 5424 (2.15581 iter/s, 5.56635s/12 iters), loss = 0.620767
I0428 17:52:45.822909  8468 solver.cpp:237]     Train net output #0: loss = 0.620767 (* 1 = 0.620767 loss)
I0428 17:52:45.822919  8468 sgd_solver.cpp:105] Iteration 5424, lr = 0.00341497
I0428 17:52:51.203070  8468 solver.cpp:218] Iteration 5436 (2.23051 iter/s, 5.37993s/12 iters), loss = 0.692051
I0428 17:52:51.203112  8468 solver.cpp:237]     Train net output #0: loss = 0.692051 (* 1 = 0.692051 loss)
I0428 17:52:51.203121  8468 sgd_solver.cpp:105] Iteration 5436, lr = 0.00340686
I0428 17:52:56.702859  8468 solver.cpp:218] Iteration 5448 (2.18201 iter/s, 5.49951s/12 iters), loss = 0.789411
I0428 17:52:56.702903  8468 solver.cpp:237]     Train net output #0: loss = 0.789411 (* 1 = 0.789411 loss)
I0428 17:52:56.702914  8468 sgd_solver.cpp:105] Iteration 5448, lr = 0.00339877
I0428 17:53:02.180953  8468 solver.cpp:218] Iteration 5460 (2.19066 iter/s, 5.47781s/12 iters), loss = 0.576937
I0428 17:53:02.180994  8468 solver.cpp:237]     Train net output #0: loss = 0.576937 (* 1 = 0.576937 loss)
I0428 17:53:02.181002  8468 sgd_solver.cpp:105] Iteration 5460, lr = 0.0033907
I0428 17:53:02.831326  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:53:07.653295  8468 solver.cpp:218] Iteration 5472 (2.19295 iter/s, 5.47207s/12 iters), loss = 0.642968
I0428 17:53:07.653331  8468 solver.cpp:237]     Train net output #0: loss = 0.642968 (* 1 = 0.642968 loss)
I0428 17:53:07.653340  8468 sgd_solver.cpp:105] Iteration 5472, lr = 0.00338265
I0428 17:53:13.420943  8468 solver.cpp:218] Iteration 5484 (2.08068 iter/s, 5.76736s/12 iters), loss = 0.625096
I0428 17:53:13.421002  8468 solver.cpp:237]     Train net output #0: loss = 0.625096 (* 1 = 0.625096 loss)
I0428 17:53:13.421022  8468 sgd_solver.cpp:105] Iteration 5484, lr = 0.00337462
I0428 17:53:19.112021  8468 solver.cpp:218] Iteration 5496 (2.10867 iter/s, 5.69078s/12 iters), loss = 0.716169
I0428 17:53:19.112155  8468 solver.cpp:237]     Train net output #0: loss = 0.716169 (* 1 = 0.716169 loss)
I0428 17:53:19.112171  8468 sgd_solver.cpp:105] Iteration 5496, lr = 0.00336661
I0428 17:53:23.999975  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_5508.caffemodel
I0428 17:53:24.665741  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_5508.solverstate
I0428 17:53:25.110208  8468 solver.cpp:330] Iteration 5508, Testing net (#0)
I0428 17:53:25.110227  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:53:27.405925  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:53:29.747085  8468 solver.cpp:397]     Test net output #0: accuracy = 0.417892
I0428 17:53:29.747123  8468 solver.cpp:397]     Test net output #1: loss = 3.05641 (* 1 = 3.05641 loss)
I0428 17:53:29.878520  8468 solver.cpp:218] Iteration 5508 (1.11463 iter/s, 10.7659s/12 iters), loss = 0.52864
I0428 17:53:29.878571  8468 solver.cpp:237]     Train net output #0: loss = 0.52864 (* 1 = 0.52864 loss)
I0428 17:53:29.878585  8468 sgd_solver.cpp:105] Iteration 5508, lr = 0.00335861
I0428 17:53:34.492188  8468 solver.cpp:218] Iteration 5520 (2.60111 iter/s, 4.61341s/12 iters), loss = 0.644999
I0428 17:53:34.492235  8468 solver.cpp:237]     Train net output #0: loss = 0.644999 (* 1 = 0.644999 loss)
I0428 17:53:34.492247  8468 sgd_solver.cpp:105] Iteration 5520, lr = 0.00335064
I0428 17:53:35.746039  8468 blocking_queue.cpp:49] Waiting for data
I0428 17:53:40.032575  8468 solver.cpp:218] Iteration 5532 (2.16602 iter/s, 5.54011s/12 iters), loss = 0.53841
I0428 17:53:40.032618  8468 solver.cpp:237]     Train net output #0: loss = 0.53841 (* 1 = 0.53841 loss)
I0428 17:53:40.032627  8468 sgd_solver.cpp:105] Iteration 5532, lr = 0.00334268
I0428 17:53:45.386929  8468 solver.cpp:218] Iteration 5544 (2.24128 iter/s, 5.35407s/12 iters), loss = 0.609924
I0428 17:53:45.386971  8468 solver.cpp:237]     Train net output #0: loss = 0.609924 (* 1 = 0.609924 loss)
I0428 17:53:45.386981  8468 sgd_solver.cpp:105] Iteration 5544, lr = 0.00333475
I0428 17:53:50.926035  8468 solver.cpp:218] Iteration 5556 (2.16652 iter/s, 5.53882s/12 iters), loss = 0.444861
I0428 17:53:50.926170  8468 solver.cpp:237]     Train net output #0: loss = 0.444861 (* 1 = 0.444861 loss)
I0428 17:53:50.926183  8468 sgd_solver.cpp:105] Iteration 5556, lr = 0.00332683
I0428 17:53:53.780762  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:53:56.406232  8468 solver.cpp:218] Iteration 5568 (2.18985 iter/s, 5.47983s/12 iters), loss = 0.55411
I0428 17:53:56.406276  8468 solver.cpp:237]     Train net output #0: loss = 0.55411 (* 1 = 0.55411 loss)
I0428 17:53:56.406286  8468 sgd_solver.cpp:105] Iteration 5568, lr = 0.00331893
I0428 17:54:01.850782  8468 solver.cpp:218] Iteration 5580 (2.20415 iter/s, 5.44427s/12 iters), loss = 0.525508
I0428 17:54:01.850836  8468 solver.cpp:237]     Train net output #0: loss = 0.525508 (* 1 = 0.525508 loss)
I0428 17:54:01.850852  8468 sgd_solver.cpp:105] Iteration 5580, lr = 0.00331105
I0428 17:54:07.374255  8468 solver.cpp:218] Iteration 5592 (2.17266 iter/s, 5.52319s/12 iters), loss = 0.673033
I0428 17:54:07.374294  8468 solver.cpp:237]     Train net output #0: loss = 0.673033 (* 1 = 0.673033 loss)
I0428 17:54:07.374305  8468 sgd_solver.cpp:105] Iteration 5592, lr = 0.00330319
I0428 17:54:12.731369  8468 solver.cpp:218] Iteration 5604 (2.24013 iter/s, 5.35684s/12 iters), loss = 0.503431
I0428 17:54:12.731410  8468 solver.cpp:237]     Train net output #0: loss = 0.503431 (* 1 = 0.503431 loss)
I0428 17:54:12.731420  8468 sgd_solver.cpp:105] Iteration 5604, lr = 0.00329535
I0428 17:54:14.973413  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_5610.caffemodel
I0428 17:54:17.792018  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_5610.solverstate
I0428 17:54:18.676008  8468 solver.cpp:330] Iteration 5610, Testing net (#0)
I0428 17:54:18.676038  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:54:20.931241  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:54:23.347576  8468 solver.cpp:397]     Test net output #0: accuracy = 0.387255
I0428 17:54:23.347611  8468 solver.cpp:397]     Test net output #1: loss = 3.17157 (* 1 = 3.17157 loss)
I0428 17:54:25.335374  8468 solver.cpp:218] Iteration 5616 (0.952121 iter/s, 12.6034s/12 iters), loss = 0.613823
I0428 17:54:25.335417  8468 solver.cpp:237]     Train net output #0: loss = 0.613823 (* 1 = 0.613823 loss)
I0428 17:54:25.335428  8468 sgd_solver.cpp:105] Iteration 5616, lr = 0.00328752
I0428 17:54:30.677120  8468 solver.cpp:218] Iteration 5628 (2.24657 iter/s, 5.34148s/12 iters), loss = 0.461659
I0428 17:54:30.677157  8468 solver.cpp:237]     Train net output #0: loss = 0.461659 (* 1 = 0.461659 loss)
I0428 17:54:30.677167  8468 sgd_solver.cpp:105] Iteration 5628, lr = 0.00327972
I0428 17:54:36.057014  8468 solver.cpp:218] Iteration 5640 (2.23064 iter/s, 5.37963s/12 iters), loss = 0.644911
I0428 17:54:36.057052  8468 solver.cpp:237]     Train net output #0: loss = 0.644911 (* 1 = 0.644911 loss)
I0428 17:54:36.057061  8468 sgd_solver.cpp:105] Iteration 5640, lr = 0.00327193
I0428 17:54:41.542610  8468 solver.cpp:218] Iteration 5652 (2.18766 iter/s, 5.48532s/12 iters), loss = 0.386837
I0428 17:54:41.542651  8468 solver.cpp:237]     Train net output #0: loss = 0.386837 (* 1 = 0.386837 loss)
I0428 17:54:41.542661  8468 sgd_solver.cpp:105] Iteration 5652, lr = 0.00326416
I0428 17:54:46.844475  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:54:47.050290  8468 solver.cpp:218] Iteration 5664 (2.17888 iter/s, 5.50741s/12 iters), loss = 0.578293
I0428 17:54:47.050330  8468 solver.cpp:237]     Train net output #0: loss = 0.578293 (* 1 = 0.578293 loss)
I0428 17:54:47.050339  8468 sgd_solver.cpp:105] Iteration 5664, lr = 0.00325641
I0428 17:54:52.496289  8468 solver.cpp:218] Iteration 5676 (2.20357 iter/s, 5.44572s/12 iters), loss = 0.520539
I0428 17:54:52.496423  8468 solver.cpp:237]     Train net output #0: loss = 0.520539 (* 1 = 0.520539 loss)
I0428 17:54:52.496434  8468 sgd_solver.cpp:105] Iteration 5676, lr = 0.00324868
I0428 17:54:57.923890  8468 solver.cpp:218] Iteration 5688 (2.21107 iter/s, 5.42724s/12 iters), loss = 0.590221
I0428 17:54:57.923930  8468 solver.cpp:237]     Train net output #0: loss = 0.590221 (* 1 = 0.590221 loss)
I0428 17:54:57.923939  8468 sgd_solver.cpp:105] Iteration 5688, lr = 0.00324097
I0428 17:55:03.319367  8468 solver.cpp:218] Iteration 5700 (2.2242 iter/s, 5.3952s/12 iters), loss = 0.691121
I0428 17:55:03.319419  8468 solver.cpp:237]     Train net output #0: loss = 0.691121 (* 1 = 0.691121 loss)
I0428 17:55:03.319430  8468 sgd_solver.cpp:105] Iteration 5700, lr = 0.00323328
I0428 17:55:08.158268  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_5712.caffemodel
I0428 17:55:08.783246  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_5712.solverstate
I0428 17:55:09.212707  8468 solver.cpp:330] Iteration 5712, Testing net (#0)
I0428 17:55:09.212724  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:55:11.452787  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:55:13.951041  8468 solver.cpp:397]     Test net output #0: accuracy = 0.389706
I0428 17:55:13.951081  8468 solver.cpp:397]     Test net output #1: loss = 3.25837 (* 1 = 3.25837 loss)
I0428 17:55:14.080633  8468 solver.cpp:218] Iteration 5712 (1.11516 iter/s, 10.7608s/12 iters), loss = 0.513439
I0428 17:55:14.080682  8468 solver.cpp:237]     Train net output #0: loss = 0.513439 (* 1 = 0.513439 loss)
I0428 17:55:14.080695  8468 sgd_solver.cpp:105] Iteration 5712, lr = 0.0032256
I0428 17:55:18.716676  8468 solver.cpp:218] Iteration 5724 (2.58856 iter/s, 4.63579s/12 iters), loss = 0.655048
I0428 17:55:18.716732  8468 solver.cpp:237]     Train net output #0: loss = 0.655048 (* 1 = 0.655048 loss)
I0428 17:55:18.716744  8468 sgd_solver.cpp:105] Iteration 5724, lr = 0.00321794
I0428 17:55:24.078817  8468 solver.cpp:218] Iteration 5736 (2.23803 iter/s, 5.36185s/12 iters), loss = 0.573068
I0428 17:55:24.078976  8468 solver.cpp:237]     Train net output #0: loss = 0.573068 (* 1 = 0.573068 loss)
I0428 17:55:24.078989  8468 sgd_solver.cpp:105] Iteration 5736, lr = 0.0032103
I0428 17:55:29.585952  8468 solver.cpp:218] Iteration 5748 (2.17915 iter/s, 5.50674s/12 iters), loss = 0.567456
I0428 17:55:29.586004  8468 solver.cpp:237]     Train net output #0: loss = 0.567456 (* 1 = 0.567456 loss)
I0428 17:55:29.586015  8468 sgd_solver.cpp:105] Iteration 5748, lr = 0.00320268
I0428 17:55:35.144524  8468 solver.cpp:218] Iteration 5760 (2.15895 iter/s, 5.55826s/12 iters), loss = 0.50568
I0428 17:55:35.144564  8468 solver.cpp:237]     Train net output #0: loss = 0.50568 (* 1 = 0.50568 loss)
I0428 17:55:35.144573  8468 sgd_solver.cpp:105] Iteration 5760, lr = 0.00319508
I0428 17:55:37.281088  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:55:40.582506  8468 solver.cpp:218] Iteration 5772 (2.20681 iter/s, 5.43771s/12 iters), loss = 0.49446
I0428 17:55:40.582558  8468 solver.cpp:237]     Train net output #0: loss = 0.49446 (* 1 = 0.49446 loss)
I0428 17:55:40.582572  8468 sgd_solver.cpp:105] Iteration 5772, lr = 0.00318749
I0428 17:55:46.045024  8468 solver.cpp:218] Iteration 5784 (2.1969 iter/s, 5.46224s/12 iters), loss = 0.588894
I0428 17:55:46.045060  8468 solver.cpp:237]     Train net output #0: loss = 0.588894 (* 1 = 0.588894 loss)
I0428 17:55:46.045070  8468 sgd_solver.cpp:105] Iteration 5784, lr = 0.00317992
I0428 17:55:51.519230  8468 solver.cpp:218] Iteration 5796 (2.19221 iter/s, 5.47393s/12 iters), loss = 0.394999
I0428 17:55:51.519271  8468 solver.cpp:237]     Train net output #0: loss = 0.394999 (* 1 = 0.394999 loss)
I0428 17:55:51.519282  8468 sgd_solver.cpp:105] Iteration 5796, lr = 0.00317237
I0428 17:55:56.970468  8468 solver.cpp:218] Iteration 5808 (2.20144 iter/s, 5.45097s/12 iters), loss = 0.506208
I0428 17:55:56.970566  8468 solver.cpp:237]     Train net output #0: loss = 0.506208 (* 1 = 0.506208 loss)
I0428 17:55:56.970575  8468 sgd_solver.cpp:105] Iteration 5808, lr = 0.00316484
I0428 17:55:59.129446  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_5814.caffemodel
I0428 17:55:59.862511  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_5814.solverstate
I0428 17:56:00.308890  8468 solver.cpp:330] Iteration 5814, Testing net (#0)
I0428 17:56:00.308909  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:56:02.656723  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:56:05.226509  8468 solver.cpp:397]     Test net output #0: accuracy = 0.39951
I0428 17:56:05.226537  8468 solver.cpp:397]     Test net output #1: loss = 3.30419 (* 1 = 3.30419 loss)
I0428 17:56:07.257341  8468 solver.cpp:218] Iteration 5820 (1.16659 iter/s, 10.2864s/12 iters), loss = 0.515416
I0428 17:56:07.257385  8468 solver.cpp:237]     Train net output #0: loss = 0.515416 (* 1 = 0.515416 loss)
I0428 17:56:07.257395  8468 sgd_solver.cpp:105] Iteration 5820, lr = 0.00315733
I0428 17:56:12.678334  8468 solver.cpp:218] Iteration 5832 (2.21373 iter/s, 5.42071s/12 iters), loss = 0.42016
I0428 17:56:12.678375  8468 solver.cpp:237]     Train net output #0: loss = 0.42016 (* 1 = 0.42016 loss)
I0428 17:56:12.678385  8468 sgd_solver.cpp:105] Iteration 5832, lr = 0.00314983
I0428 17:56:18.022958  8468 solver.cpp:218] Iteration 5844 (2.24536 iter/s, 5.34434s/12 iters), loss = 0.623039
I0428 17:56:18.023010  8468 solver.cpp:237]     Train net output #0: loss = 0.623039 (* 1 = 0.623039 loss)
I0428 17:56:18.023025  8468 sgd_solver.cpp:105] Iteration 5844, lr = 0.00314235
I0428 17:56:23.401531  8468 solver.cpp:218] Iteration 5856 (2.23119 iter/s, 5.37829s/12 iters), loss = 0.67557
I0428 17:56:23.401573  8468 solver.cpp:237]     Train net output #0: loss = 0.67557 (* 1 = 0.67557 loss)
I0428 17:56:23.401583  8468 sgd_solver.cpp:105] Iteration 5856, lr = 0.00313489
I0428 17:56:27.920604  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:56:28.801383  8468 solver.cpp:218] Iteration 5868 (2.2224 iter/s, 5.39958s/12 iters), loss = 0.590324
I0428 17:56:28.801421  8468 solver.cpp:237]     Train net output #0: loss = 0.590324 (* 1 = 0.590324 loss)
I0428 17:56:28.801430  8468 sgd_solver.cpp:105] Iteration 5868, lr = 0.00312745
I0428 17:56:34.313647  8468 solver.cpp:218] Iteration 5880 (2.17707 iter/s, 5.51199s/12 iters), loss = 0.681813
I0428 17:56:34.313704  8468 solver.cpp:237]     Train net output #0: loss = 0.681813 (* 1 = 0.681813 loss)
I0428 17:56:34.313716  8468 sgd_solver.cpp:105] Iteration 5880, lr = 0.00312002
I0428 17:56:39.643983  8468 solver.cpp:218] Iteration 5892 (2.25139 iter/s, 5.33005s/12 iters), loss = 0.442712
I0428 17:56:39.644038  8468 solver.cpp:237]     Train net output #0: loss = 0.442712 (* 1 = 0.442712 loss)
I0428 17:56:39.644052  8468 sgd_solver.cpp:105] Iteration 5892, lr = 0.00311262
I0428 17:56:44.940188  8468 solver.cpp:218] Iteration 5904 (2.26589 iter/s, 5.29592s/12 iters), loss = 0.665911
I0428 17:56:44.940229  8468 solver.cpp:237]     Train net output #0: loss = 0.665911 (* 1 = 0.665911 loss)
I0428 17:56:44.940241  8468 sgd_solver.cpp:105] Iteration 5904, lr = 0.00310523
I0428 17:56:49.900820  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_5916.caffemodel
I0428 17:56:51.057286  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_5916.solverstate
I0428 17:56:51.777820  8468 solver.cpp:330] Iteration 5916, Testing net (#0)
I0428 17:56:51.777839  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:56:53.913018  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:56:56.401685  8468 solver.cpp:397]     Test net output #0: accuracy = 0.409314
I0428 17:56:56.401715  8468 solver.cpp:397]     Test net output #1: loss = 3.49379 (* 1 = 3.49379 loss)
I0428 17:56:56.530709  8468 solver.cpp:218] Iteration 5916 (1.03538 iter/s, 11.59s/12 iters), loss = 0.421307
I0428 17:56:56.530755  8468 solver.cpp:237]     Train net output #0: loss = 0.421307 (* 1 = 0.421307 loss)
I0428 17:56:56.530766  8468 sgd_solver.cpp:105] Iteration 5916, lr = 0.00309785
I0428 17:57:01.137166  8468 solver.cpp:218] Iteration 5928 (2.60518 iter/s, 4.6062s/12 iters), loss = 0.56133
I0428 17:57:01.137295  8468 solver.cpp:237]     Train net output #0: loss = 0.56133 (* 1 = 0.56133 loss)
I0428 17:57:01.137308  8468 sgd_solver.cpp:105] Iteration 5928, lr = 0.0030905
I0428 17:57:06.470039  8468 solver.cpp:218] Iteration 5940 (2.25034 iter/s, 5.33252s/12 iters), loss = 0.61535
I0428 17:57:06.470095  8468 solver.cpp:237]     Train net output #0: loss = 0.61535 (* 1 = 0.61535 loss)
I0428 17:57:06.470111  8468 sgd_solver.cpp:105] Iteration 5940, lr = 0.00308316
I0428 17:57:11.959597  8468 solver.cpp:218] Iteration 5952 (2.18608 iter/s, 5.48927s/12 iters), loss = 0.579
I0428 17:57:11.959653  8468 solver.cpp:237]     Train net output #0: loss = 0.579 (* 1 = 0.579 loss)
I0428 17:57:11.959667  8468 sgd_solver.cpp:105] Iteration 5952, lr = 0.00307584
I0428 17:57:17.569310  8468 solver.cpp:218] Iteration 5964 (2.13926 iter/s, 5.60942s/12 iters), loss = 0.402545
I0428 17:57:17.569360  8468 solver.cpp:237]     Train net output #0: loss = 0.402545 (* 1 = 0.402545 loss)
I0428 17:57:17.569373  8468 sgd_solver.cpp:105] Iteration 5964, lr = 0.00306854
I0428 17:57:18.999756  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:57:22.914113  8468 solver.cpp:218] Iteration 5976 (2.24529 iter/s, 5.34452s/12 iters), loss = 0.478818
I0428 17:57:22.914153  8468 solver.cpp:237]     Train net output #0: loss = 0.478818 (* 1 = 0.478818 loss)
I0428 17:57:22.914165  8468 sgd_solver.cpp:105] Iteration 5976, lr = 0.00306125
I0428 17:57:28.422359  8468 solver.cpp:218] Iteration 5988 (2.17866 iter/s, 5.50797s/12 iters), loss = 0.554464
I0428 17:57:28.422415  8468 solver.cpp:237]     Train net output #0: loss = 0.554464 (* 1 = 0.554464 loss)
I0428 17:57:28.422427  8468 sgd_solver.cpp:105] Iteration 5988, lr = 0.00305398
I0428 17:57:33.778576  8468 solver.cpp:218] Iteration 6000 (2.2405 iter/s, 5.35594s/12 iters), loss = 0.462563
I0428 17:57:33.778714  8468 solver.cpp:237]     Train net output #0: loss = 0.462563 (* 1 = 0.462563 loss)
I0428 17:57:33.778726  8468 sgd_solver.cpp:105] Iteration 6000, lr = 0.00304673
I0428 17:57:39.145411  8468 solver.cpp:218] Iteration 6012 (2.23611 iter/s, 5.36647s/12 iters), loss = 0.498612
I0428 17:57:39.145452  8468 solver.cpp:237]     Train net output #0: loss = 0.498612 (* 1 = 0.498612 loss)
I0428 17:57:39.145464  8468 sgd_solver.cpp:105] Iteration 6012, lr = 0.0030395
I0428 17:57:41.346701  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6018.caffemodel
I0428 17:57:41.987757  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6018.solverstate
I0428 17:57:42.440229  8468 solver.cpp:330] Iteration 6018, Testing net (#0)
I0428 17:57:42.440251  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:57:44.474833  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:57:47.191788  8468 solver.cpp:397]     Test net output #0: accuracy = 0.416667
I0428 17:57:47.191833  8468 solver.cpp:397]     Test net output #1: loss = 3.26315 (* 1 = 3.26315 loss)
I0428 17:57:49.151262  8468 solver.cpp:218] Iteration 6024 (1.19935 iter/s, 10.0054s/12 iters), loss = 0.534377
I0428 17:57:49.151301  8468 solver.cpp:237]     Train net output #0: loss = 0.534377 (* 1 = 0.534377 loss)
I0428 17:57:49.151310  8468 sgd_solver.cpp:105] Iteration 6024, lr = 0.00303228
I0428 17:57:54.597756  8468 solver.cpp:218] Iteration 6036 (2.20336 iter/s, 5.44622s/12 iters), loss = 0.319217
I0428 17:57:54.597797  8468 solver.cpp:237]     Train net output #0: loss = 0.319217 (* 1 = 0.319217 loss)
I0428 17:57:54.597805  8468 sgd_solver.cpp:105] Iteration 6036, lr = 0.00302508
I0428 17:58:00.114821  8468 solver.cpp:218] Iteration 6048 (2.17518 iter/s, 5.51678s/12 iters), loss = 0.441278
I0428 17:58:00.114873  8468 solver.cpp:237]     Train net output #0: loss = 0.441278 (* 1 = 0.441278 loss)
I0428 17:58:00.114886  8468 sgd_solver.cpp:105] Iteration 6048, lr = 0.0030179
I0428 17:58:05.703646  8468 solver.cpp:218] Iteration 6060 (2.14725 iter/s, 5.58853s/12 iters), loss = 0.423046
I0428 17:58:05.703792  8468 solver.cpp:237]     Train net output #0: loss = 0.423046 (* 1 = 0.423046 loss)
I0428 17:58:05.703805  8468 sgd_solver.cpp:105] Iteration 6060, lr = 0.00301074
I0428 17:58:09.379133  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:58:11.094300  8468 solver.cpp:218] Iteration 6072 (2.22623 iter/s, 5.39028s/12 iters), loss = 0.274081
I0428 17:58:11.094341  8468 solver.cpp:237]     Train net output #0: loss = 0.274081 (* 1 = 0.274081 loss)
I0428 17:58:11.094350  8468 sgd_solver.cpp:105] Iteration 6072, lr = 0.00300359
I0428 17:58:16.612030  8468 solver.cpp:218] Iteration 6084 (2.17492 iter/s, 5.51745s/12 iters), loss = 0.371653
I0428 17:58:16.612069  8468 solver.cpp:237]     Train net output #0: loss = 0.371653 (* 1 = 0.371653 loss)
I0428 17:58:16.612078  8468 sgd_solver.cpp:105] Iteration 6084, lr = 0.00299646
I0428 17:58:22.241174  8468 solver.cpp:218] Iteration 6096 (2.13187 iter/s, 5.62887s/12 iters), loss = 0.729148
I0428 17:58:22.241210  8468 solver.cpp:237]     Train net output #0: loss = 0.729148 (* 1 = 0.729148 loss)
I0428 17:58:22.241221  8468 sgd_solver.cpp:105] Iteration 6096, lr = 0.00298934
I0428 17:58:27.673553  8468 solver.cpp:218] Iteration 6108 (2.20909 iter/s, 5.43211s/12 iters), loss = 0.365232
I0428 17:58:27.673593  8468 solver.cpp:237]     Train net output #0: loss = 0.365232 (* 1 = 0.365232 loss)
I0428 17:58:27.673604  8468 sgd_solver.cpp:105] Iteration 6108, lr = 0.00298225
I0428 17:58:32.689214  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6120.caffemodel
I0428 17:58:34.016342  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6120.solverstate
I0428 17:58:34.862221  8468 solver.cpp:330] Iteration 6120, Testing net (#0)
I0428 17:58:34.862246  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:58:36.897593  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:58:39.426085  8468 solver.cpp:397]     Test net output #0: accuracy = 0.416667
I0428 17:58:39.426123  8468 solver.cpp:397]     Test net output #1: loss = 3.37304 (* 1 = 3.37304 loss)
I0428 17:58:39.557122  8468 solver.cpp:218] Iteration 6120 (1.00984 iter/s, 11.883s/12 iters), loss = 0.471288
I0428 17:58:39.557163  8468 solver.cpp:237]     Train net output #0: loss = 0.471288 (* 1 = 0.471288 loss)
I0428 17:58:39.557176  8468 sgd_solver.cpp:105] Iteration 6120, lr = 0.00297517
I0428 17:58:44.051232  8468 solver.cpp:218] Iteration 6132 (2.6703 iter/s, 4.49387s/12 iters), loss = 0.56074
I0428 17:58:44.051270  8468 solver.cpp:237]     Train net output #0: loss = 0.56074 (* 1 = 0.56074 loss)
I0428 17:58:44.051280  8468 sgd_solver.cpp:105] Iteration 6132, lr = 0.0029681
I0428 17:58:49.609869  8468 solver.cpp:218] Iteration 6144 (2.15891 iter/s, 5.55836s/12 iters), loss = 0.409664
I0428 17:58:49.609912  8468 solver.cpp:237]     Train net output #0: loss = 0.409664 (* 1 = 0.409664 loss)
I0428 17:58:49.609922  8468 sgd_solver.cpp:105] Iteration 6144, lr = 0.00296105
I0428 17:58:55.140929  8468 solver.cpp:218] Iteration 6156 (2.16968 iter/s, 5.53078s/12 iters), loss = 0.621763
I0428 17:58:55.140980  8468 solver.cpp:237]     Train net output #0: loss = 0.621763 (* 1 = 0.621763 loss)
I0428 17:58:55.140991  8468 sgd_solver.cpp:105] Iteration 6156, lr = 0.00295402
I0428 17:59:00.714342  8468 solver.cpp:218] Iteration 6168 (2.15319 iter/s, 5.57312s/12 iters), loss = 0.430858
I0428 17:59:00.714382  8468 solver.cpp:237]     Train net output #0: loss = 0.430858 (* 1 = 0.430858 loss)
I0428 17:59:00.714394  8468 sgd_solver.cpp:105] Iteration 6168, lr = 0.00294701
I0428 17:59:01.330616  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:59:06.055285  8468 solver.cpp:218] Iteration 6180 (2.24691 iter/s, 5.34067s/12 iters), loss = 0.348172
I0428 17:59:06.055326  8468 solver.cpp:237]     Train net output #0: loss = 0.348172 (* 1 = 0.348172 loss)
I0428 17:59:06.055335  8468 sgd_solver.cpp:105] Iteration 6180, lr = 0.00294001
I0428 17:59:11.546700  8468 solver.cpp:218] Iteration 6192 (2.18534 iter/s, 5.49114s/12 iters), loss = 0.388067
I0428 17:59:11.546800  8468 solver.cpp:237]     Train net output #0: loss = 0.388067 (* 1 = 0.388067 loss)
I0428 17:59:11.546809  8468 sgd_solver.cpp:105] Iteration 6192, lr = 0.00293303
I0428 17:59:16.886446  8468 solver.cpp:218] Iteration 6204 (2.24744 iter/s, 5.33942s/12 iters), loss = 0.365513
I0428 17:59:16.886487  8468 solver.cpp:237]     Train net output #0: loss = 0.365513 (* 1 = 0.365513 loss)
I0428 17:59:16.886497  8468 sgd_solver.cpp:105] Iteration 6204, lr = 0.00292607
I0428 17:59:22.555817  8468 solver.cpp:218] Iteration 6216 (2.11674 iter/s, 5.66909s/12 iters), loss = 0.530317
I0428 17:59:22.555860  8468 solver.cpp:237]     Train net output #0: loss = 0.530317 (* 1 = 0.530317 loss)
I0428 17:59:22.555869  8468 sgd_solver.cpp:105] Iteration 6216, lr = 0.00291912
I0428 17:59:24.699527  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6222.caffemodel
I0428 17:59:25.283262  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6222.solverstate
I0428 17:59:26.527720  8468 solver.cpp:330] Iteration 6222, Testing net (#0)
I0428 17:59:26.527746  8468 net.cpp:676] Ignoring source layer train-data
I0428 17:59:28.496395  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:59:29.547951  8468 blocking_queue.cpp:49] Waiting for data
I0428 17:59:31.053025  8468 solver.cpp:397]     Test net output #0: accuracy = 0.395221
I0428 17:59:31.053064  8468 solver.cpp:397]     Test net output #1: loss = 3.31952 (* 1 = 3.31952 loss)
I0428 17:59:33.070401  8468 solver.cpp:218] Iteration 6228 (1.14132 iter/s, 10.5141s/12 iters), loss = 0.473763
I0428 17:59:33.070441  8468 solver.cpp:237]     Train net output #0: loss = 0.473763 (* 1 = 0.473763 loss)
I0428 17:59:33.070449  8468 sgd_solver.cpp:105] Iteration 6228, lr = 0.00291219
I0428 17:59:38.430588  8468 solver.cpp:218] Iteration 6240 (2.23884 iter/s, 5.35992s/12 iters), loss = 0.370887
I0428 17:59:38.430627  8468 solver.cpp:237]     Train net output #0: loss = 0.370887 (* 1 = 0.370887 loss)
I0428 17:59:38.430639  8468 sgd_solver.cpp:105] Iteration 6240, lr = 0.00290528
I0428 17:59:43.757556  8468 solver.cpp:218] Iteration 6252 (2.2528 iter/s, 5.3267s/12 iters), loss = 0.354213
I0428 17:59:43.757797  8468 solver.cpp:237]     Train net output #0: loss = 0.354213 (* 1 = 0.354213 loss)
I0428 17:59:43.757805  8468 sgd_solver.cpp:105] Iteration 6252, lr = 0.00289838
I0428 17:59:49.123086  8468 solver.cpp:218] Iteration 6264 (2.2367 iter/s, 5.36506s/12 iters), loss = 0.322013
I0428 17:59:49.123138  8468 solver.cpp:237]     Train net output #0: loss = 0.322013 (* 1 = 0.322013 loss)
I0428 17:59:49.123149  8468 sgd_solver.cpp:105] Iteration 6264, lr = 0.0028915
I0428 17:59:52.021359  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 17:59:54.537379  8468 solver.cpp:218] Iteration 6276 (2.21647 iter/s, 5.41401s/12 iters), loss = 0.417642
I0428 17:59:54.537434  8468 solver.cpp:237]     Train net output #0: loss = 0.417642 (* 1 = 0.417642 loss)
I0428 17:59:54.537448  8468 sgd_solver.cpp:105] Iteration 6276, lr = 0.00288463
I0428 18:00:00.005970  8468 solver.cpp:218] Iteration 6288 (2.19446 iter/s, 5.4683s/12 iters), loss = 0.686677
I0428 18:00:00.006012  8468 solver.cpp:237]     Train net output #0: loss = 0.686677 (* 1 = 0.686677 loss)
I0428 18:00:00.006022  8468 sgd_solver.cpp:105] Iteration 6288, lr = 0.00287779
I0428 18:00:05.651352  8468 solver.cpp:218] Iteration 6300 (2.12574 iter/s, 5.6451s/12 iters), loss = 0.370543
I0428 18:00:05.651398  8468 solver.cpp:237]     Train net output #0: loss = 0.370543 (* 1 = 0.370543 loss)
I0428 18:00:05.651410  8468 sgd_solver.cpp:105] Iteration 6300, lr = 0.00287095
I0428 18:00:11.085391  8468 solver.cpp:218] Iteration 6312 (2.20841 iter/s, 5.43377s/12 iters), loss = 0.373402
I0428 18:00:11.085427  8468 solver.cpp:237]     Train net output #0: loss = 0.373402 (* 1 = 0.373402 loss)
I0428 18:00:11.085438  8468 sgd_solver.cpp:105] Iteration 6312, lr = 0.00286414
I0428 18:00:15.898301  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6324.caffemodel
I0428 18:00:17.129936  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6324.solverstate
I0428 18:00:17.802256  8468 solver.cpp:330] Iteration 6324, Testing net (#0)
I0428 18:00:17.802275  8468 net.cpp:676] Ignoring source layer train-data
I0428 18:00:19.773798  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:00:22.674116  8468 solver.cpp:397]     Test net output #0: accuracy = 0.40625
I0428 18:00:22.674157  8468 solver.cpp:397]     Test net output #1: loss = 3.38831 (* 1 = 3.38831 loss)
I0428 18:00:22.805194  8468 solver.cpp:218] Iteration 6324 (1.02395 iter/s, 11.7193s/12 iters), loss = 0.476372
I0428 18:00:22.805235  8468 solver.cpp:237]     Train net output #0: loss = 0.476372 (* 1 = 0.476372 loss)
I0428 18:00:22.805244  8468 sgd_solver.cpp:105] Iteration 6324, lr = 0.00285734
I0428 18:00:27.245326  8468 solver.cpp:218] Iteration 6336 (2.70276 iter/s, 4.4399s/12 iters), loss = 0.448321
I0428 18:00:27.245366  8468 solver.cpp:237]     Train net output #0: loss = 0.448321 (* 1 = 0.448321 loss)
I0428 18:00:27.245378  8468 sgd_solver.cpp:105] Iteration 6336, lr = 0.00285055
I0428 18:00:32.819197  8468 solver.cpp:218] Iteration 6348 (2.15301 iter/s, 5.57359s/12 iters), loss = 0.350176
I0428 18:00:32.819237  8468 solver.cpp:237]     Train net output #0: loss = 0.350176 (* 1 = 0.350176 loss)
I0428 18:00:32.819248  8468 sgd_solver.cpp:105] Iteration 6348, lr = 0.00284379
I0428 18:00:38.265972  8468 solver.cpp:218] Iteration 6360 (2.20325 iter/s, 5.4465s/12 iters), loss = 0.435244
I0428 18:00:38.266014  8468 solver.cpp:237]     Train net output #0: loss = 0.435244 (* 1 = 0.435244 loss)
I0428 18:00:38.266024  8468 sgd_solver.cpp:105] Iteration 6360, lr = 0.00283703
I0428 18:00:43.451735  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:00:43.629346  8468 solver.cpp:218] Iteration 6372 (2.23751 iter/s, 5.3631s/12 iters), loss = 0.334895
I0428 18:00:43.629388  8468 solver.cpp:237]     Train net output #0: loss = 0.334895 (* 1 = 0.334895 loss)
I0428 18:00:43.629397  8468 sgd_solver.cpp:105] Iteration 6372, lr = 0.0028303
I0428 18:00:49.072057  8468 solver.cpp:218] Iteration 6384 (2.2049 iter/s, 5.44244s/12 iters), loss = 0.231009
I0428 18:00:49.072193  8468 solver.cpp:237]     Train net output #0: loss = 0.231009 (* 1 = 0.231009 loss)
I0428 18:00:49.072204  8468 sgd_solver.cpp:105] Iteration 6384, lr = 0.00282358
I0428 18:00:54.715829  8468 solver.cpp:218] Iteration 6396 (2.12638 iter/s, 5.6434s/12 iters), loss = 0.523784
I0428 18:00:54.715870  8468 solver.cpp:237]     Train net output #0: loss = 0.523784 (* 1 = 0.523784 loss)
I0428 18:00:54.715878  8468 sgd_solver.cpp:105] Iteration 6396, lr = 0.00281687
I0428 18:01:00.153643  8468 solver.cpp:218] Iteration 6408 (2.20688 iter/s, 5.43753s/12 iters), loss = 0.353541
I0428 18:01:00.153698  8468 solver.cpp:237]     Train net output #0: loss = 0.353541 (* 1 = 0.353541 loss)
I0428 18:01:00.153714  8468 sgd_solver.cpp:105] Iteration 6408, lr = 0.00281019
I0428 18:01:05.704162  8468 solver.cpp:218] Iteration 6420 (2.16207 iter/s, 5.55023s/12 iters), loss = 0.339242
I0428 18:01:05.704216  8468 solver.cpp:237]     Train net output #0: loss = 0.339242 (* 1 = 0.339242 loss)
I0428 18:01:05.704226  8468 sgd_solver.cpp:105] Iteration 6420, lr = 0.00280351
I0428 18:01:07.859328  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6426.caffemodel
I0428 18:01:09.073916  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6426.solverstate
I0428 18:01:09.912626  8468 solver.cpp:330] Iteration 6426, Testing net (#0)
I0428 18:01:09.912649  8468 net.cpp:676] Ignoring source layer train-data
I0428 18:01:11.808708  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:01:14.577039  8468 solver.cpp:397]     Test net output #0: accuracy = 0.405024
I0428 18:01:14.577069  8468 solver.cpp:397]     Test net output #1: loss = 3.54951 (* 1 = 3.54951 loss)
I0428 18:01:16.603688  8468 solver.cpp:218] Iteration 6432 (1.10102 iter/s, 10.899s/12 iters), loss = 0.343749
I0428 18:01:16.603729  8468 solver.cpp:237]     Train net output #0: loss = 0.343749 (* 1 = 0.343749 loss)
I0428 18:01:16.603737  8468 sgd_solver.cpp:105] Iteration 6432, lr = 0.00279686
I0428 18:01:22.182564  8468 solver.cpp:218] Iteration 6444 (2.15108 iter/s, 5.5786s/12 iters), loss = 0.441936
I0428 18:01:22.182662  8468 solver.cpp:237]     Train net output #0: loss = 0.441936 (* 1 = 0.441936 loss)
I0428 18:01:22.182673  8468 sgd_solver.cpp:105] Iteration 6444, lr = 0.00279022
I0428 18:01:27.538611  8468 solver.cpp:218] Iteration 6456 (2.2406 iter/s, 5.35572s/12 iters), loss = 0.333443
I0428 18:01:27.538650  8468 solver.cpp:237]     Train net output #0: loss = 0.333443 (* 1 = 0.333443 loss)
I0428 18:01:27.538661  8468 sgd_solver.cpp:105] Iteration 6456, lr = 0.00278359
I0428 18:01:32.960403  8468 solver.cpp:218] Iteration 6468 (2.2134 iter/s, 5.42152s/12 iters), loss = 0.355073
I0428 18:01:32.960469  8468 solver.cpp:237]     Train net output #0: loss = 0.355073 (* 1 = 0.355073 loss)
I0428 18:01:32.960482  8468 sgd_solver.cpp:105] Iteration 6468, lr = 0.00277698
I0428 18:01:35.043114  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:01:38.338176  8468 solver.cpp:218] Iteration 6480 (2.23153 iter/s, 5.37748s/12 iters), loss = 0.31271
I0428 18:01:38.338229  8468 solver.cpp:237]     Train net output #0: loss = 0.31271 (* 1 = 0.31271 loss)
I0428 18:01:38.338244  8468 sgd_solver.cpp:105] Iteration 6480, lr = 0.00277039
I0428 18:01:43.827257  8468 solver.cpp:218] Iteration 6492 (2.18627 iter/s, 5.48879s/12 iters), loss = 0.401695
I0428 18:01:43.827298  8468 solver.cpp:237]     Train net output #0: loss = 0.401695 (* 1 = 0.401695 loss)
I0428 18:01:43.827309  8468 sgd_solver.cpp:105] Iteration 6492, lr = 0.00276381
I0428 18:01:49.301824  8468 solver.cpp:218] Iteration 6504 (2.19206 iter/s, 5.47429s/12 iters), loss = 0.510296
I0428 18:01:49.301863  8468 solver.cpp:237]     Train net output #0: loss = 0.510296 (* 1 = 0.510296 loss)
I0428 18:01:49.301872  8468 sgd_solver.cpp:105] Iteration 6504, lr = 0.00275725
I0428 18:01:54.889086  8468 solver.cpp:218] Iteration 6516 (2.14785 iter/s, 5.58698s/12 iters), loss = 0.590555
I0428 18:01:54.889259  8468 solver.cpp:237]     Train net output #0: loss = 0.590555 (* 1 = 0.590555 loss)
I0428 18:01:54.889272  8468 sgd_solver.cpp:105] Iteration 6516, lr = 0.00275071
I0428 18:01:59.766480  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6528.caffemodel
I0428 18:02:01.600009  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6528.solverstate
I0428 18:02:02.023118  8468 solver.cpp:330] Iteration 6528, Testing net (#0)
I0428 18:02:02.023142  8468 net.cpp:676] Ignoring source layer train-data
I0428 18:02:04.074364  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:02:06.924988  8468 solver.cpp:397]     Test net output #0: accuracy = 0.400123
I0428 18:02:06.925030  8468 solver.cpp:397]     Test net output #1: loss = 3.42828 (* 1 = 3.42828 loss)
I0428 18:02:07.055680  8468 solver.cpp:218] Iteration 6528 (0.986362 iter/s, 12.1659s/12 iters), loss = 0.40532
I0428 18:02:07.055732  8468 solver.cpp:237]     Train net output #0: loss = 0.40532 (* 1 = 0.40532 loss)
I0428 18:02:07.055744  8468 sgd_solver.cpp:105] Iteration 6528, lr = 0.00274418
I0428 18:02:11.587195  8468 solver.cpp:218] Iteration 6540 (2.64826 iter/s, 4.53127s/12 iters), loss = 0.369416
I0428 18:02:11.587236  8468 solver.cpp:237]     Train net output #0: loss = 0.369416 (* 1 = 0.369416 loss)
I0428 18:02:11.587245  8468 sgd_solver.cpp:105] Iteration 6540, lr = 0.00273766
I0428 18:02:17.044008  8468 solver.cpp:218] Iteration 6552 (2.1992 iter/s, 5.45653s/12 iters), loss = 0.370828
I0428 18:02:17.044059  8468 solver.cpp:237]     Train net output #0: loss = 0.370828 (* 1 = 0.370828 loss)
I0428 18:02:17.044070  8468 sgd_solver.cpp:105] Iteration 6552, lr = 0.00273116
I0428 18:02:22.510870  8468 solver.cpp:218] Iteration 6564 (2.19516 iter/s, 5.46658s/12 iters), loss = 0.322033
I0428 18:02:22.510915  8468 solver.cpp:237]     Train net output #0: loss = 0.322033 (* 1 = 0.322033 loss)
I0428 18:02:22.510923  8468 sgd_solver.cpp:105] Iteration 6564, lr = 0.00272468
I0428 18:02:27.056689  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:02:27.945881  8468 solver.cpp:218] Iteration 6576 (2.20802 iter/s, 5.43474s/12 iters), loss = 0.29083
I0428 18:02:27.945919  8468 solver.cpp:237]     Train net output #0: loss = 0.29083 (* 1 = 0.29083 loss)
I0428 18:02:27.945930  8468 sgd_solver.cpp:105] Iteration 6576, lr = 0.00271821
I0428 18:02:33.433441  8468 solver.cpp:218] Iteration 6588 (2.18687 iter/s, 5.48728s/12 iters), loss = 0.544462
I0428 18:02:33.433481  8468 solver.cpp:237]     Train net output #0: loss = 0.544462 (* 1 = 0.544462 loss)
I0428 18:02:33.433491  8468 sgd_solver.cpp:105] Iteration 6588, lr = 0.00271175
I0428 18:02:38.934876  8468 solver.cpp:218] Iteration 6600 (2.18136 iter/s, 5.50115s/12 iters), loss = 0.372928
I0428 18:02:38.934928  8468 solver.cpp:237]     Train net output #0: loss = 0.372928 (* 1 = 0.372928 loss)
I0428 18:02:38.934942  8468 sgd_solver.cpp:105] Iteration 6600, lr = 0.00270532
I0428 18:02:44.586405  8468 solver.cpp:218] Iteration 6612 (2.12343 iter/s, 5.65124s/12 iters), loss = 0.387212
I0428 18:02:44.586448  8468 solver.cpp:237]     Train net output #0: loss = 0.387212 (* 1 = 0.387212 loss)
I0428 18:02:44.586459  8468 sgd_solver.cpp:105] Iteration 6612, lr = 0.00269889
I0428 18:02:50.008431  8468 solver.cpp:218] Iteration 6624 (2.21331 iter/s, 5.42174s/12 iters), loss = 0.430999
I0428 18:02:50.008522  8468 solver.cpp:237]     Train net output #0: loss = 0.430999 (* 1 = 0.430999 loss)
I0428 18:02:50.008536  8468 sgd_solver.cpp:105] Iteration 6624, lr = 0.00269248
I0428 18:02:52.193243  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6630.caffemodel
I0428 18:02:54.605149  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6630.solverstate
I0428 18:02:56.830296  8468 solver.cpp:330] Iteration 6630, Testing net (#0)
I0428 18:02:56.830314  8468 net.cpp:676] Ignoring source layer train-data
I0428 18:02:58.750672  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:03:01.530194  8468 solver.cpp:397]     Test net output #0: accuracy = 0.42402
I0428 18:03:01.530231  8468 solver.cpp:397]     Test net output #1: loss = 3.48907 (* 1 = 3.48907 loss)
I0428 18:03:03.500458  8468 solver.cpp:218] Iteration 6636 (0.889455 iter/s, 13.4914s/12 iters), loss = 0.308763
I0428 18:03:03.500537  8468 solver.cpp:237]     Train net output #0: loss = 0.308763 (* 1 = 0.308763 loss)
I0428 18:03:03.500550  8468 sgd_solver.cpp:105] Iteration 6636, lr = 0.00268609
I0428 18:03:08.991241  8468 solver.cpp:218] Iteration 6648 (2.18561 iter/s, 5.49047s/12 iters), loss = 0.252751
I0428 18:03:08.991283  8468 solver.cpp:237]     Train net output #0: loss = 0.252751 (* 1 = 0.252751 loss)
I0428 18:03:08.991297  8468 sgd_solver.cpp:105] Iteration 6648, lr = 0.00267971
I0428 18:03:14.405606  8468 solver.cpp:218] Iteration 6660 (2.21644 iter/s, 5.41409s/12 iters), loss = 0.297332
I0428 18:03:14.405647  8468 solver.cpp:237]     Train net output #0: loss = 0.297332 (* 1 = 0.297332 loss)
I0428 18:03:14.405655  8468 sgd_solver.cpp:105] Iteration 6660, lr = 0.00267335
I0428 18:03:19.951153  8468 solver.cpp:218] Iteration 6672 (2.16401 iter/s, 5.54526s/12 iters), loss = 0.289358
I0428 18:03:19.951197  8468 solver.cpp:237]     Train net output #0: loss = 0.289358 (* 1 = 0.289358 loss)
I0428 18:03:19.951206  8468 sgd_solver.cpp:105] Iteration 6672, lr = 0.00266701
I0428 18:03:21.407281  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:03:25.326848  8468 solver.cpp:218] Iteration 6684 (2.23238 iter/s, 5.37542s/12 iters), loss = 0.322618
I0428 18:03:25.326884  8468 solver.cpp:237]     Train net output #0: loss = 0.322618 (* 1 = 0.322618 loss)
I0428 18:03:25.326894  8468 sgd_solver.cpp:105] Iteration 6684, lr = 0.00266067
I0428 18:03:30.807056  8468 solver.cpp:218] Iteration 6696 (2.18981 iter/s, 5.47994s/12 iters), loss = 0.309223
I0428 18:03:30.807163  8468 solver.cpp:237]     Train net output #0: loss = 0.309223 (* 1 = 0.309223 loss)
I0428 18:03:30.807173  8468 sgd_solver.cpp:105] Iteration 6696, lr = 0.00265436
I0428 18:03:36.239104  8468 solver.cpp:218] Iteration 6708 (2.20925 iter/s, 5.43171s/12 iters), loss = 0.36083
I0428 18:03:36.239145  8468 solver.cpp:237]     Train net output #0: loss = 0.36083 (* 1 = 0.36083 loss)
I0428 18:03:36.239153  8468 sgd_solver.cpp:105] Iteration 6708, lr = 0.00264805
I0428 18:03:41.726828  8468 solver.cpp:218] Iteration 6720 (2.18681 iter/s, 5.48744s/12 iters), loss = 0.362614
I0428 18:03:41.726878  8468 solver.cpp:237]     Train net output #0: loss = 0.362614 (* 1 = 0.362614 loss)
I0428 18:03:41.726889  8468 sgd_solver.cpp:105] Iteration 6720, lr = 0.00264177
I0428 18:03:46.644403  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6732.caffemodel
I0428 18:03:47.784255  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6732.solverstate
I0428 18:03:49.431664  8468 solver.cpp:330] Iteration 6732, Testing net (#0)
I0428 18:03:49.431689  8468 net.cpp:676] Ignoring source layer train-data
I0428 18:03:51.353683  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:03:54.269656  8468 solver.cpp:397]     Test net output #0: accuracy = 0.41973
I0428 18:03:54.269686  8468 solver.cpp:397]     Test net output #1: loss = 3.48637 (* 1 = 3.48637 loss)
I0428 18:03:54.398677  8468 solver.cpp:218] Iteration 6732 (0.947024 iter/s, 12.6713s/12 iters), loss = 0.322549
I0428 18:03:54.398718  8468 solver.cpp:237]     Train net output #0: loss = 0.322549 (* 1 = 0.322549 loss)
I0428 18:03:54.398730  8468 sgd_solver.cpp:105] Iteration 6732, lr = 0.0026355
I0428 18:03:59.003156  8468 solver.cpp:218] Iteration 6744 (2.6063 iter/s, 4.60423s/12 iters), loss = 0.289209
I0428 18:03:59.003209  8468 solver.cpp:237]     Train net output #0: loss = 0.289209 (* 1 = 0.289209 loss)
I0428 18:03:59.003221  8468 sgd_solver.cpp:105] Iteration 6744, lr = 0.00262924
I0428 18:04:04.434573  8468 solver.cpp:218] Iteration 6756 (2.20948 iter/s, 5.43113s/12 iters), loss = 0.323745
I0428 18:04:04.434727  8468 solver.cpp:237]     Train net output #0: loss = 0.323745 (* 1 = 0.323745 loss)
I0428 18:04:04.434739  8468 sgd_solver.cpp:105] Iteration 6756, lr = 0.002623
I0428 18:04:10.342371  8468 solver.cpp:218] Iteration 6768 (2.03135 iter/s, 5.9074s/12 iters), loss = 0.236324
I0428 18:04:10.342412  8468 solver.cpp:237]     Train net output #0: loss = 0.236324 (* 1 = 0.236324 loss)
I0428 18:04:10.342423  8468 sgd_solver.cpp:105] Iteration 6768, lr = 0.00261677
I0428 18:04:14.146066  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:04:15.783866  8468 solver.cpp:218] Iteration 6780 (2.20539 iter/s, 5.44122s/12 iters), loss = 0.21903
I0428 18:04:15.783907  8468 solver.cpp:237]     Train net output #0: loss = 0.21903 (* 1 = 0.21903 loss)
I0428 18:04:15.783916  8468 sgd_solver.cpp:105] Iteration 6780, lr = 0.00261056
I0428 18:04:21.258460  8468 solver.cpp:218] Iteration 6792 (2.19205 iter/s, 5.47432s/12 iters), loss = 0.391115
I0428 18:04:21.258512  8468 solver.cpp:237]     Train net output #0: loss = 0.391115 (* 1 = 0.391115 loss)
I0428 18:04:21.258525  8468 sgd_solver.cpp:105] Iteration 6792, lr = 0.00260436
I0428 18:04:26.695336  8468 solver.cpp:218] Iteration 6804 (2.20726 iter/s, 5.4366s/12 iters), loss = 0.28464
I0428 18:04:26.695389  8468 solver.cpp:237]     Train net output #0: loss = 0.28464 (* 1 = 0.28464 loss)
I0428 18:04:26.695403  8468 sgd_solver.cpp:105] Iteration 6804, lr = 0.00259817
I0428 18:04:32.013823  8468 solver.cpp:218] Iteration 6816 (2.2564 iter/s, 5.3182s/12 iters), loss = 0.333655
I0428 18:04:32.013880  8468 solver.cpp:237]     Train net output #0: loss = 0.333655 (* 1 = 0.333655 loss)
I0428 18:04:32.013891  8468 sgd_solver.cpp:105] Iteration 6816, lr = 0.00259201
I0428 18:04:37.513069  8468 solver.cpp:218] Iteration 6828 (2.18223 iter/s, 5.49896s/12 iters), loss = 0.270133
I0428 18:04:37.513450  8468 solver.cpp:237]     Train net output #0: loss = 0.270133 (* 1 = 0.270133 loss)
I0428 18:04:37.513459  8468 sgd_solver.cpp:105] Iteration 6828, lr = 0.00258585
I0428 18:04:39.669755  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6834.caffemodel
I0428 18:04:40.267813  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6834.solverstate
I0428 18:04:40.695343  8468 solver.cpp:330] Iteration 6834, Testing net (#0)
I0428 18:04:40.695365  8468 net.cpp:676] Ignoring source layer train-data
I0428 18:04:42.562165  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:04:45.519721  8468 solver.cpp:397]     Test net output #0: accuracy = 0.409926
I0428 18:04:45.519759  8468 solver.cpp:397]     Test net output #1: loss = 3.40849 (* 1 = 3.40849 loss)
I0428 18:04:47.601761  8468 solver.cpp:218] Iteration 6840 (1.18954 iter/s, 10.0879s/12 iters), loss = 0.37796
I0428 18:04:47.601805  8468 solver.cpp:237]     Train net output #0: loss = 0.37796 (* 1 = 0.37796 loss)
I0428 18:04:47.601814  8468 sgd_solver.cpp:105] Iteration 6840, lr = 0.00257971
I0428 18:04:53.047928  8468 solver.cpp:218] Iteration 6852 (2.20349 iter/s, 5.4459s/12 iters), loss = 0.356983
I0428 18:04:53.047966  8468 solver.cpp:237]     Train net output #0: loss = 0.356983 (* 1 = 0.356983 loss)
I0428 18:04:53.047974  8468 sgd_solver.cpp:105] Iteration 6852, lr = 0.00257359
I0428 18:04:58.581944  8468 solver.cpp:218] Iteration 6864 (2.16851 iter/s, 5.53374s/12 iters), loss = 0.264756
I0428 18:04:58.581990  8468 solver.cpp:237]     Train net output #0: loss = 0.264756 (* 1 = 0.264756 loss)
I0428 18:04:58.582001  8468 sgd_solver.cpp:105] Iteration 6864, lr = 0.00256748
I0428 18:05:04.251026  8468 solver.cpp:218] Iteration 6876 (2.11685 iter/s, 5.6688s/12 iters), loss = 0.212193
I0428 18:05:04.251068  8468 solver.cpp:237]     Train net output #0: loss = 0.212193 (* 1 = 0.212193 loss)
I0428 18:05:04.251078  8468 sgd_solver.cpp:105] Iteration 6876, lr = 0.00256138
I0428 18:05:04.902978  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:05:09.683028  8468 solver.cpp:218] Iteration 6888 (2.20924 iter/s, 5.43173s/12 iters), loss = 0.339998
I0428 18:05:09.719424  8468 solver.cpp:237]     Train net output #0: loss = 0.339998 (* 1 = 0.339998 loss)
I0428 18:05:09.719436  8468 sgd_solver.cpp:105] Iteration 6888, lr = 0.0025553
I0428 18:05:15.085489  8468 solver.cpp:218] Iteration 6900 (2.23637 iter/s, 5.36584s/12 iters), loss = 0.467551
I0428 18:05:15.085526  8468 solver.cpp:237]     Train net output #0: loss = 0.467551 (* 1 = 0.467551 loss)
I0428 18:05:15.085536  8468 sgd_solver.cpp:105] Iteration 6900, lr = 0.00254923
I0428 18:05:20.540688  8468 solver.cpp:218] Iteration 6912 (2.19985 iter/s, 5.45492s/12 iters), loss = 0.272096
I0428 18:05:20.540729  8468 solver.cpp:237]     Train net output #0: loss = 0.272096 (* 1 = 0.272096 loss)
I0428 18:05:20.540738  8468 sgd_solver.cpp:105] Iteration 6912, lr = 0.00254318
I0428 18:05:25.893174  8468 solver.cpp:218] Iteration 6924 (2.24206 iter/s, 5.35222s/12 iters), loss = 0.21273
I0428 18:05:25.893214  8468 solver.cpp:237]     Train net output #0: loss = 0.21273 (* 1 = 0.21273 loss)
I0428 18:05:25.893226  8468 sgd_solver.cpp:105] Iteration 6924, lr = 0.00253714
I0428 18:05:30.741570  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6936.caffemodel
I0428 18:05:31.362273  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6936.solverstate
I0428 18:05:31.790155  8468 solver.cpp:330] Iteration 6936, Testing net (#0)
I0428 18:05:31.790174  8468 net.cpp:676] Ignoring source layer train-data
I0428 18:05:32.079803  8468 blocking_queue.cpp:49] Waiting for data
I0428 18:05:33.437716  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:05:36.280689  8468 solver.cpp:397]     Test net output #0: accuracy = 0.423407
I0428 18:05:36.280725  8468 solver.cpp:397]     Test net output #1: loss = 3.49238 (* 1 = 3.49238 loss)
I0428 18:05:36.407732  8468 solver.cpp:218] Iteration 6936 (1.14133 iter/s, 10.5141s/12 iters), loss = 0.226995
I0428 18:05:36.407770  8468 solver.cpp:237]     Train net output #0: loss = 0.226995 (* 1 = 0.226995 loss)
I0428 18:05:36.407779  8468 sgd_solver.cpp:105] Iteration 6936, lr = 0.00253112
I0428 18:05:40.896755  8468 solver.cpp:218] Iteration 6948 (2.67333 iter/s, 4.48879s/12 iters), loss = 0.24341
I0428 18:05:40.897428  8468 solver.cpp:237]     Train net output #0: loss = 0.24341 (* 1 = 0.24341 loss)
I0428 18:05:40.897442  8468 sgd_solver.cpp:105] Iteration 6948, lr = 0.00252511
I0428 18:05:46.226310  8468 solver.cpp:218] Iteration 6960 (2.25197 iter/s, 5.32866s/12 iters), loss = 0.322609
I0428 18:05:46.226349  8468 solver.cpp:237]     Train net output #0: loss = 0.322609 (* 1 = 0.322609 loss)
I0428 18:05:46.226359  8468 sgd_solver.cpp:105] Iteration 6960, lr = 0.00251911
I0428 18:05:51.678304  8468 solver.cpp:218] Iteration 6972 (2.20114 iter/s, 5.45172s/12 iters), loss = 0.27452
I0428 18:05:51.678355  8468 solver.cpp:237]     Train net output #0: loss = 0.27452 (* 1 = 0.27452 loss)
I0428 18:05:51.678369  8468 sgd_solver.cpp:105] Iteration 6972, lr = 0.00251313
I0428 18:05:54.673506  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:05:57.187011  8468 solver.cpp:218] Iteration 6984 (2.17848 iter/s, 5.50842s/12 iters), loss = 0.256837
I0428 18:05:57.187052  8468 solver.cpp:237]     Train net output #0: loss = 0.256837 (* 1 = 0.256837 loss)
I0428 18:05:57.187060  8468 sgd_solver.cpp:105] Iteration 6984, lr = 0.00250717
I0428 18:06:02.588347  8468 solver.cpp:218] Iteration 6996 (2.22178 iter/s, 5.40106s/12 iters), loss = 0.314635
I0428 18:06:02.588387  8468 solver.cpp:237]     Train net output #0: loss = 0.314635 (* 1 = 0.314635 loss)
I0428 18:06:02.588397  8468 sgd_solver.cpp:105] Iteration 6996, lr = 0.00250121
I0428 18:06:08.116542  8468 solver.cpp:218] Iteration 7008 (2.17081 iter/s, 5.52789s/12 iters), loss = 0.338179
I0428 18:06:08.116583  8468 solver.cpp:237]     Train net output #0: loss = 0.338179 (* 1 = 0.338179 loss)
I0428 18:06:08.116593  8468 sgd_solver.cpp:105] Iteration 7008, lr = 0.00249528
I0428 18:06:13.446604  8468 solver.cpp:218] Iteration 7020 (2.2515 iter/s, 5.32979s/12 iters), loss = 0.280241
I0428 18:06:13.446768  8468 solver.cpp:237]     Train net output #0: loss = 0.280241 (* 1 = 0.280241 loss)
I0428 18:06:13.446780  8468 sgd_solver.cpp:105] Iteration 7020, lr = 0.00248935
I0428 18:06:18.949383  8468 solver.cpp:218] Iteration 7032 (2.18087 iter/s, 5.50238s/12 iters), loss = 0.224451
I0428 18:06:18.949421  8468 solver.cpp:237]     Train net output #0: loss = 0.224451 (* 1 = 0.224451 loss)
I0428 18:06:18.949430  8468 sgd_solver.cpp:105] Iteration 7032, lr = 0.00248344
I0428 18:06:21.103364  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7038.caffemodel
I0428 18:06:21.721436  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7038.solverstate
I0428 18:06:22.152287  8468 solver.cpp:330] Iteration 7038, Testing net (#0)
I0428 18:06:22.152312  8468 net.cpp:676] Ignoring source layer train-data
I0428 18:06:23.801133  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:06:26.849586  8468 solver.cpp:397]     Test net output #0: accuracy = 0.425245
I0428 18:06:26.849622  8468 solver.cpp:397]     Test net output #1: loss = 3.46346 (* 1 = 3.46346 loss)
I0428 18:06:28.924566  8468 solver.cpp:218] Iteration 7044 (1.20304 iter/s, 9.97473s/12 iters), loss = 0.231223
I0428 18:06:28.924604  8468 solver.cpp:237]     Train net output #0: loss = 0.231223 (* 1 = 0.231223 loss)
I0428 18:06:28.924615  8468 sgd_solver.cpp:105] Iteration 7044, lr = 0.00247755
I0428 18:06:34.391755  8468 solver.cpp:218] Iteration 7056 (2.19502 iter/s, 5.46692s/12 iters), loss = 0.221365
I0428 18:06:34.391803  8468 solver.cpp:237]     Train net output #0: loss = 0.221365 (* 1 = 0.221365 loss)
I0428 18:06:34.391815  8468 sgd_solver.cpp:105] Iteration 7056, lr = 0.00247166
I0428 18:06:39.784921  8468 solver.cpp:218] Iteration 7068 (2.22515 iter/s, 5.39288s/12 iters), loss = 0.266817
I0428 18:06:39.784976  8468 solver.cpp:237]     Train net output #0: loss = 0.266817 (* 1 = 0.266817 loss)
I0428 18:06:39.784992  8468 sgd_solver.cpp:105] Iteration 7068, lr = 0.0024658
I0428 18:06:44.995469  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:06:45.139081  8468 solver.cpp:218] Iteration 7080 (2.24137 iter/s, 5.35388s/12 iters), loss = 0.268101
I0428 18:06:45.139122  8468 solver.cpp:237]     Train net output #0: loss = 0.268101 (* 1 = 0.268101 loss)
I0428 18:06:45.139132  8468 sgd_solver.cpp:105] Iteration 7080, lr = 0.00245994
I0428 18:06:50.586539  8468 solver.cpp:218] Iteration 7092 (2.20297 iter/s, 5.44718s/12 iters), loss = 0.211526
I0428 18:06:50.586591  8468 solver.cpp:237]     Train net output #0: loss = 0.211526 (* 1 = 0.211526 loss)
I0428 18:06:50.586603  8468 sgd_solver.cpp:105] Iteration 7092, lr = 0.0024541
I0428 18:06:56.069726  8468 solver.cpp:218] Iteration 7104 (2.18862 iter/s, 5.48291s/12 iters), loss = 0.267729
I0428 18:06:56.069766  8468 solver.cpp:237]     Train net output #0: loss = 0.267729 (* 1 = 0.267729 loss)
I0428 18:06:56.069777  8468 sgd_solver.cpp:105] Iteration 7104, lr = 0.00244827
I0428 18:07:01.414304  8468 solver.cpp:218] Iteration 7116 (2.24538 iter/s, 5.34431s/12 iters), loss = 0.241553
I0428 18:07:01.414343  8468 solver.cpp:237]     Train net output #0: loss = 0.241553 (* 1 = 0.241553 loss)
I0428 18:07:01.414353  8468 sgd_solver.cpp:105] Iteration 7116, lr = 0.00244246
I0428 18:07:06.901561  8468 solver.cpp:218] Iteration 7128 (2.187 iter/s, 5.48698s/12 iters), loss = 0.335122
I0428 18:07:06.901602  8468 solver.cpp:237]     Train net output #0: loss = 0.335122 (* 1 = 0.335122 loss)
I0428 18:07:06.901610  8468 sgd_solver.cpp:105] Iteration 7128, lr = 0.00243666
I0428 18:07:11.805938  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7140.caffemodel
I0428 18:07:12.584631  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7140.solverstate
I0428 18:07:13.289530  8468 solver.cpp:330] Iteration 7140, Testing net (#0)
I0428 18:07:13.289554  8468 net.cpp:676] Ignoring source layer train-data
I0428 18:07:14.866348  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:07:17.948874  8468 solver.cpp:397]     Test net output #0: accuracy = 0.438113
I0428 18:07:17.949012  8468 solver.cpp:397]     Test net output #1: loss = 3.48926 (* 1 = 3.48926 loss)
I0428 18:07:18.080155  8468 solver.cpp:218] Iteration 7140 (1.07353 iter/s, 11.1781s/12 iters), loss = 0.217267
I0428 18:07:18.080197  8468 solver.cpp:237]     Train net output #0: loss = 0.217267 (* 1 = 0.217267 loss)
I0428 18:07:18.080209  8468 sgd_solver.cpp:105] Iteration 7140, lr = 0.00243088
I0428 18:07:22.602385  8468 solver.cpp:218] Iteration 7152 (2.6537 iter/s, 4.52199s/12 iters), loss = 0.157984
I0428 18:07:22.602425  8468 solver.cpp:237]     Train net output #0: loss = 0.157984 (* 1 = 0.157984 loss)
I0428 18:07:22.602433  8468 sgd_solver.cpp:105] Iteration 7152, lr = 0.00242511
I0428 18:07:28.258168  8468 solver.cpp:218] Iteration 7164 (2.12183 iter/s, 5.6555s/12 iters), loss = 0.242612
I0428 18:07:28.258222  8468 solver.cpp:237]     Train net output #0: loss = 0.242612 (* 1 = 0.242612 loss)
I0428 18:07:28.258235  8468 sgd_solver.cpp:105] Iteration 7164, lr = 0.00241935
I0428 18:07:33.751466  8468 solver.cpp:218] Iteration 7176 (2.18459 iter/s, 5.49301s/12 iters), loss = 0.211294
I0428 18:07:33.751504  8468 solver.cpp:237]     Train net output #0: loss = 0.211294 (* 1 = 0.211294 loss)
I0428 18:07:33.751514  8468 sgd_solver.cpp:105] Iteration 7176, lr = 0.0024136
I0428 18:07:36.092033  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:07:39.191826  8468 solver.cpp:218] Iteration 7188 (2.20585 iter/s, 5.44009s/12 iters), loss = 0.284513
I0428 18:07:39.191876  8468 solver.cpp:237]     Train net output #0: loss = 0.284513 (* 1 = 0.284513 loss)
I0428 18:07:39.191890  8468 sgd_solver.cpp:105] Iteration 7188, lr = 0.00240787
I0428 18:07:44.768633  8468 solver.cpp:218] Iteration 7200 (2.15188 iter/s, 5.57652s/12 iters), loss = 0.196213
I0428 18:07:44.768687  8468 solver.cpp:237]     Train net output #0: loss = 0.196213 (* 1 = 0.196213 loss)
I0428 18:07:44.768705  8468 sgd_solver.cpp:105] Iteration 7200, lr = 0.00240216
I0428 18:07:50.197012  8468 solver.cpp:218] Iteration 7212 (2.21072 iter/s, 5.4281s/12 iters), loss = 0.309912
I0428 18:07:50.197134  8468 solver.cpp:237]     Train net output #0: loss = 0.309912 (* 1 = 0.309912 loss)
I0428 18:07:50.197150  8468 sgd_solver.cpp:105] Iteration 7212, lr = 0.00239645
I0428 18:07:55.635927  8468 solver.cpp:218] Iteration 7224 (2.20646 iter/s, 5.43857s/12 iters), loss = 0.255081
I0428 18:07:55.635968  8468 solver.cpp:237]     Train net output #0: loss = 0.255081 (* 1 = 0.255081 loss)
I0428 18:07:55.635978  8468 sgd_solver.cpp:105] Iteration 7224, lr = 0.00239076
I0428 18:08:01.230947  8468 solver.cpp:218] Iteration 7236 (2.14487 iter/s, 5.59474s/12 iters), loss = 0.266401
I0428 18:08:01.230988  8468 solver.cpp:237]     Train net output #0: loss = 0.266401 (* 1 = 0.266401 loss)
I0428 18:08:01.230998  8468 sgd_solver.cpp:105] Iteration 7236, lr = 0.00238509
I0428 18:08:03.437204  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7242.caffemodel
I0428 18:08:05.084442  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7242.solverstate
I0428 18:08:05.925941  8468 solver.cpp:330] Iteration 7242, Testing net (#0)
I0428 18:08:05.925966  8468 net.cpp:676] Ignoring source layer train-data
I0428 18:08:07.479583  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:08:10.559551  8468 solver.cpp:397]     Test net output #0: accuracy = 0.427083
I0428 18:08:10.559582  8468 solver.cpp:397]     Test net output #1: loss = 3.51863 (* 1 = 3.51863 loss)
I0428 18:08:12.544287  8468 solver.cpp:218] Iteration 7248 (1.06074 iter/s, 11.3128s/12 iters), loss = 0.344783
I0428 18:08:12.544328  8468 solver.cpp:237]     Train net output #0: loss = 0.344783 (* 1 = 0.344783 loss)
I0428 18:08:12.544337  8468 sgd_solver.cpp:105] Iteration 7248, lr = 0.00237942
I0428 18:08:18.067800  8468 solver.cpp:218] Iteration 7260 (2.17264 iter/s, 5.52323s/12 iters), loss = 0.232457
I0428 18:08:18.067849  8468 solver.cpp:237]     Train net output #0: loss = 0.232457 (* 1 = 0.232457 loss)
I0428 18:08:18.067860  8468 sgd_solver.cpp:105] Iteration 7260, lr = 0.00237378
I0428 18:08:23.575650  8468 solver.cpp:218] Iteration 7272 (2.17882 iter/s, 5.50756s/12 iters), loss = 0.177145
I0428 18:08:23.575831  8468 solver.cpp:237]     Train net output #0: loss = 0.177145 (* 1 = 0.177145 loss)
I0428 18:08:23.575845  8468 sgd_solver.cpp:105] Iteration 7272, lr = 0.00236814
I0428 18:08:28.149188  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:08:28.960183  8468 solver.cpp:218] Iteration 7284 (2.22878 iter/s, 5.38412s/12 iters), loss = 0.172017
I0428 18:08:28.960237  8468 solver.cpp:237]     Train net output #0: loss = 0.172017 (* 1 = 0.172017 loss)
I0428 18:08:28.960251  8468 sgd_solver.cpp:105] Iteration 7284, lr = 0.00236252
I0428 18:08:34.357853  8468 solver.cpp:218] Iteration 7296 (2.2233 iter/s, 5.39739s/12 iters), loss = 0.225122
I0428 18:08:34.357895  8468 solver.cpp:237]     Train net output #0: loss = 0.225122 (* 1 = 0.225122 loss)
I0428 18:08:34.357903  8468 sgd_solver.cpp:105] Iteration 7296, lr = 0.00235691
I0428 18:08:39.989686  8468 solver.cpp:218] Iteration 7308 (2.13085 iter/s, 5.63155s/12 iters), loss = 0.14413
I0428 18:08:39.989720  8468 solver.cpp:237]     Train net output #0: loss = 0.14413 (* 1 = 0.14413 loss)
I0428 18:08:39.989729  8468 sgd_solver.cpp:105] Iteration 7308, lr = 0.00235131
I0428 18:08:45.449594  8468 solver.cpp:218] Iteration 7320 (2.19795 iter/s, 5.45964s/12 iters), loss = 0.287017
I0428 18:08:45.449633  8468 solver.cpp:237]     Train net output #0: loss = 0.287017 (* 1 = 0.287017 loss)
I0428 18:08:45.449642  8468 sgd_solver.cpp:105] Iteration 7320, lr = 0.00234573
I0428 18:08:51.018020  8468 solver.cpp:218] Iteration 7332 (2.15511 iter/s, 5.56815s/12 iters), loss = 0.286699
I0428 18:08:51.018061  8468 solver.cpp:237]     Train net output #0: loss = 0.286699 (* 1 = 0.286699 loss)
I0428 18:08:51.018071  8468 sgd_solver.cpp:105] Iteration 7332, lr = 0.00234016
I0428 18:08:55.919543  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7344.caffemodel
I0428 18:08:56.526391  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7344.solverstate
I0428 18:08:56.969305  8468 solver.cpp:330] Iteration 7344, Testing net (#0)
I0428 18:08:56.969327  8468 net.cpp:676] Ignoring source layer train-data
I0428 18:08:58.608105  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:09:01.720592  8468 solver.cpp:397]     Test net output #0: accuracy = 0.4375
I0428 18:09:01.720626  8468 solver.cpp:397]     Test net output #1: loss = 3.56824 (* 1 = 3.56824 loss)
I0428 18:09:01.846284  8468 solver.cpp:218] Iteration 7344 (1.10826 iter/s, 10.8278s/12 iters), loss = 0.188547
I0428 18:09:01.846343  8468 solver.cpp:237]     Train net output #0: loss = 0.188547 (* 1 = 0.188547 loss)
I0428 18:09:01.846356  8468 sgd_solver.cpp:105] Iteration 7344, lr = 0.0023346
I0428 18:09:06.338286  8468 solver.cpp:218] Iteration 7356 (2.67156 iter/s, 4.49175s/12 iters), loss = 0.312952
I0428 18:09:06.338327  8468 solver.cpp:237]     Train net output #0: loss = 0.312952 (* 1 = 0.312952 loss)
I0428 18:09:06.338337  8468 sgd_solver.cpp:105] Iteration 7356, lr = 0.00232906
I0428 18:09:11.700342  8468 solver.cpp:218] Iteration 7368 (2.23806 iter/s, 5.36178s/12 iters), loss = 0.152558
I0428 18:09:11.700395  8468 solver.cpp:237]     Train net output #0: loss = 0.152558 (* 1 = 0.152558 loss)
I0428 18:09:11.700408  8468 sgd_solver.cpp:105] Iteration 7368, lr = 0.00232353
I0428 18:09:17.035991  8468 solver.cpp:218] Iteration 7380 (2.24914 iter/s, 5.33537s/12 iters), loss = 0.268929
I0428 18:09:17.036046  8468 solver.cpp:237]     Train net output #0: loss = 0.268929 (* 1 = 0.268929 loss)
I0428 18:09:17.036059  8468 sgd_solver.cpp:105] Iteration 7380, lr = 0.00231802
I0428 18:09:18.492779  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:09:22.343633  8468 solver.cpp:218] Iteration 7392 (2.26101 iter/s, 5.30736s/12 iters), loss = 0.207175
I0428 18:09:22.343691  8468 solver.cpp:237]     Train net output #0: loss = 0.207175 (* 1 = 0.207175 loss)
I0428 18:09:22.343704  8468 sgd_solver.cpp:105] Iteration 7392, lr = 0.00231251
I0428 18:09:27.679854  8468 solver.cpp:218] Iteration 7404 (2.2489 iter/s, 5.33594s/12 iters), loss = 0.175469
I0428 18:09:27.680030  8468 solver.cpp:237]     Train net output #0: loss = 0.175469 (* 1 = 0.175469 loss)
I0428 18:09:27.680043  8468 sgd_solver.cpp:105] Iteration 7404, lr = 0.00230702
I0428 18:09:32.994541  8468 solver.cpp:218] Iteration 7416 (2.25807 iter/s, 5.31428s/12 iters), loss = 0.189292
I0428 18:09:32.994596  8468 solver.cpp:237]     Train net output #0: loss = 0.189292 (* 1 = 0.189292 loss)
I0428 18:09:32.994611  8468 sgd_solver.cpp:105] Iteration 7416, lr = 0.00230154
I0428 18:09:38.328317  8468 solver.cpp:218] Iteration 7428 (2.24993 iter/s, 5.3335s/12 iters), loss = 0.209375
I0428 18:09:38.328372  8468 solver.cpp:237]     Train net output #0: loss = 0.209375 (* 1 = 0.209375 loss)
I0428 18:09:38.328383  8468 sgd_solver.cpp:105] Iteration 7428, lr = 0.00229608
I0428 18:09:43.775434  8468 solver.cpp:218] Iteration 7440 (2.20312 iter/s, 5.44683s/12 iters), loss = 0.224886
I0428 18:09:43.775477  8468 solver.cpp:237]     Train net output #0: loss = 0.224886 (* 1 = 0.224886 loss)
I0428 18:09:43.775488  8468 sgd_solver.cpp:105] Iteration 7440, lr = 0.00229063
I0428 18:09:46.078490  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7446.caffemodel
I0428 18:09:47.200209  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7446.solverstate
I0428 18:09:49.724898  8468 solver.cpp:330] Iteration 7446, Testing net (#0)
I0428 18:09:49.724916  8468 net.cpp:676] Ignoring source layer train-data
I0428 18:09:51.367174  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:09:54.600729  8468 solver.cpp:397]     Test net output #0: accuracy = 0.441176
I0428 18:09:54.600760  8468 solver.cpp:397]     Test net output #1: loss = 3.44962 (* 1 = 3.44962 loss)
I0428 18:09:56.566316  8468 solver.cpp:218] Iteration 7452 (0.93821 iter/s, 12.7903s/12 iters), loss = 0.276515
I0428 18:09:56.566356  8468 solver.cpp:237]     Train net output #0: loss = 0.276515 (* 1 = 0.276515 loss)
I0428 18:09:56.566365  8468 sgd_solver.cpp:105] Iteration 7452, lr = 0.00228519
I0428 18:10:02.041518  8468 solver.cpp:218] Iteration 7464 (2.19181 iter/s, 5.47493s/12 iters), loss = 0.126806
I0428 18:10:02.041651  8468 solver.cpp:237]     Train net output #0: loss = 0.126806 (* 1 = 0.126806 loss)
I0428 18:10:02.041662  8468 sgd_solver.cpp:105] Iteration 7464, lr = 0.00227976
I0428 18:10:07.377041  8468 solver.cpp:218] Iteration 7476 (2.24923 iter/s, 5.33516s/12 iters), loss = 0.259034
I0428 18:10:07.377094  8468 solver.cpp:237]     Train net output #0: loss = 0.259034 (* 1 = 0.259034 loss)
I0428 18:10:07.377105  8468 sgd_solver.cpp:105] Iteration 7476, lr = 0.00227435
I0428 18:10:11.104653  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:10:12.701375  8468 solver.cpp:218] Iteration 7488 (2.25392 iter/s, 5.32405s/12 iters), loss = 0.140633
I0428 18:10:12.701429  8468 solver.cpp:237]     Train net output #0: loss = 0.140633 (* 1 = 0.140633 loss)
I0428 18:10:12.701443  8468 sgd_solver.cpp:105] Iteration 7488, lr = 0.00226895
I0428 18:10:18.019381  8468 solver.cpp:218] Iteration 7500 (2.2566 iter/s, 5.31773s/12 iters), loss = 0.088797
I0428 18:10:18.019438  8468 solver.cpp:237]     Train net output #0: loss = 0.088797 (* 1 = 0.088797 loss)
I0428 18:10:18.019450  8468 sgd_solver.cpp:105] Iteration 7500, lr = 0.00226357
I0428 18:10:23.354579  8468 solver.cpp:218] Iteration 7512 (2.24933 iter/s, 5.33491s/12 iters), loss = 0.331284
I0428 18:10:23.354624  8468 solver.cpp:237]     Train net output #0: loss = 0.331284 (* 1 = 0.331284 loss)
I0428 18:10:23.354636  8468 sgd_solver.cpp:105] Iteration 7512, lr = 0.00225819
I0428 18:10:28.693856  8468 solver.cpp:218] Iteration 7524 (2.24761 iter/s, 5.339s/12 iters), loss = 0.213642
I0428 18:10:28.693895  8468 solver.cpp:237]     Train net output #0: loss = 0.213642 (* 1 = 0.213642 loss)
I0428 18:10:28.693905  8468 sgd_solver.cpp:105] Iteration 7524, lr = 0.00225283
I0428 18:10:34.043642  8468 solver.cpp:218] Iteration 7536 (2.24319 iter/s, 5.34951s/12 iters), loss = 0.0567069
I0428 18:10:34.043831  8468 solver.cpp:237]     Train net output #0: loss = 0.0567069 (* 1 = 0.0567069 loss)
I0428 18:10:34.043845  8468 sgd_solver.cpp:105] Iteration 7536, lr = 0.00224748
I0428 18:10:38.830756  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7548.caffemodel
I0428 18:10:39.439297  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7548.solverstate
I0428 18:10:39.869743  8468 solver.cpp:330] Iteration 7548, Testing net (#0)
I0428 18:10:39.869768  8468 net.cpp:676] Ignoring source layer train-data
I0428 18:10:41.280648  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:10:44.358454  8468 solver.cpp:397]     Test net output #0: accuracy = 0.428309
I0428 18:10:44.358494  8468 solver.cpp:397]     Test net output #1: loss = 3.67156 (* 1 = 3.67156 loss)
I0428 18:10:44.489868  8468 solver.cpp:218] Iteration 7548 (1.14881 iter/s, 10.4456s/12 iters), loss = 0.224147
I0428 18:10:44.489920  8468 solver.cpp:237]     Train net output #0: loss = 0.224147 (* 1 = 0.224147 loss)
I0428 18:10:44.489935  8468 sgd_solver.cpp:105] Iteration 7548, lr = 0.00224215
I0428 18:10:49.016124  8468 solver.cpp:218] Iteration 7560 (2.65135 iter/s, 4.526s/12 iters), loss = 0.124692
I0428 18:10:49.016175  8468 solver.cpp:237]     Train net output #0: loss = 0.124692 (* 1 = 0.124692 loss)
I0428 18:10:49.016188  8468 sgd_solver.cpp:105] Iteration 7560, lr = 0.00223682
I0428 18:10:54.457159  8468 solver.cpp:218] Iteration 7572 (2.20558 iter/s, 5.44076s/12 iters), loss = 0.273672
I0428 18:10:54.457198  8468 solver.cpp:237]     Train net output #0: loss = 0.273672 (* 1 = 0.273672 loss)
I0428 18:10:54.457207  8468 sgd_solver.cpp:105] Iteration 7572, lr = 0.00223151
I0428 18:10:59.742995  8468 solver.cpp:218] Iteration 7584 (2.27033 iter/s, 5.28558s/12 iters), loss = 0.187962
I0428 18:10:59.743032  8468 solver.cpp:237]     Train net output #0: loss = 0.187962 (* 1 = 0.187962 loss)
I0428 18:10:59.743042  8468 sgd_solver.cpp:105] Iteration 7584, lr = 0.00222621
I0428 18:11:00.474061  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:11:05.206089  8468 solver.cpp:218] Iteration 7596 (2.19667 iter/s, 5.46282s/12 iters), loss = 0.317059
I0428 18:11:05.206233  8468 solver.cpp:237]     Train net output #0: loss = 0.317059 (* 1 = 0.317059 loss)
I0428 18:11:05.206243  8468 sgd_solver.cpp:105] Iteration 7596, lr = 0.00222093
I0428 18:11:10.610764  8468 solver.cpp:218] Iteration 7608 (2.22045 iter/s, 5.4043s/12 iters), loss = 0.224805
I0428 18:11:10.610814  8468 solver.cpp:237]     Train net output #0: loss = 0.224805 (* 1 = 0.224805 loss)
I0428 18:11:10.610826  8468 sgd_solver.cpp:105] Iteration 7608, lr = 0.00221565
I0428 18:11:16.073866  8468 solver.cpp:218] Iteration 7620 (2.19667 iter/s, 5.46282s/12 iters), loss = 0.124447
I0428 18:11:16.073909  8468 solver.cpp:237]     Train net output #0: loss = 0.124447 (* 1 = 0.124447 loss)
I0428 18:11:16.073917  8468 sgd_solver.cpp:105] Iteration 7620, lr = 0.00221039
I0428 18:11:17.341874  8468 blocking_queue.cpp:49] Waiting for data
I0428 18:11:21.492580  8468 solver.cpp:218] Iteration 7632 (2.21466 iter/s, 5.41843s/12 iters), loss = 0.14238
I0428 18:11:21.492636  8468 solver.cpp:237]     Train net output #0: loss = 0.14238 (* 1 = 0.14238 loss)
I0428 18:11:21.492650  8468 sgd_solver.cpp:105] Iteration 7632, lr = 0.00220515
I0428 18:11:26.876781  8468 solver.cpp:218] Iteration 7644 (2.22886 iter/s, 5.38391s/12 iters), loss = 0.137286
I0428 18:11:26.876821  8468 solver.cpp:237]     Train net output #0: loss = 0.137286 (* 1 = 0.137286 loss)
I0428 18:11:26.876832  8468 sgd_solver.cpp:105] Iteration 7644, lr = 0.00219991
I0428 18:11:29.061465  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7650.caffemodel
I0428 18:11:30.078969  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7650.solverstate
I0428 18:11:32.120236  8468 solver.cpp:330] Iteration 7650, Testing net (#0)
I0428 18:11:32.120256  8468 net.cpp:676] Ignoring source layer train-data
I0428 18:11:33.600946  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:11:36.951268  8468 solver.cpp:397]     Test net output #0: accuracy = 0.433211
I0428 18:11:36.952530  8468 solver.cpp:397]     Test net output #1: loss = 3.60329 (* 1 = 3.60329 loss)
I0428 18:11:38.979676  8468 solver.cpp:218] Iteration 7656 (0.991542 iter/s, 12.1024s/12 iters), loss = 0.162917
I0428 18:11:38.979714  8468 solver.cpp:237]     Train net output #0: loss = 0.162917 (* 1 = 0.162917 loss)
I0428 18:11:38.979724  8468 sgd_solver.cpp:105] Iteration 7656, lr = 0.00219469
I0428 18:11:44.419478  8468 solver.cpp:218] Iteration 7668 (2.20607 iter/s, 5.43954s/12 iters), loss = 0.164272
I0428 18:11:44.419519  8468 solver.cpp:237]     Train net output #0: loss = 0.164272 (* 1 = 0.164272 loss)
I0428 18:11:44.419529  8468 sgd_solver.cpp:105] Iteration 7668, lr = 0.00218948
I0428 18:11:49.858776  8468 solver.cpp:218] Iteration 7680 (2.20628 iter/s, 5.43902s/12 iters), loss = 0.228421
I0428 18:11:49.858826  8468 solver.cpp:237]     Train net output #0: loss = 0.228421 (* 1 = 0.228421 loss)
I0428 18:11:49.858837  8468 sgd_solver.cpp:105] Iteration 7680, lr = 0.00218428
I0428 18:11:52.831378  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:11:55.230648  8468 solver.cpp:218] Iteration 7692 (2.23398 iter/s, 5.37159s/12 iters), loss = 0.182786
I0428 18:11:55.230706  8468 solver.cpp:237]     Train net output #0: loss = 0.182786 (* 1 = 0.182786 loss)
I0428 18:11:55.230718  8468 sgd_solver.cpp:105] Iteration 7692, lr = 0.00217909
I0428 18:12:00.605233  8468 solver.cpp:218] Iteration 7704 (2.23285 iter/s, 5.3743s/12 iters), loss = 0.2058
I0428 18:12:00.605273  8468 solver.cpp:237]     Train net output #0: loss = 0.2058 (* 1 = 0.2058 loss)
I0428 18:12:00.605283  8468 sgd_solver.cpp:105] Iteration 7704, lr = 0.00217392
I0428 18:12:06.161785  8468 solver.cpp:218] Iteration 7716 (2.15972 iter/s, 5.55628s/12 iters), loss = 0.0865956
I0428 18:12:06.161826  8468 solver.cpp:237]     Train net output #0: loss = 0.0865956 (* 1 = 0.0865956 loss)
I0428 18:12:06.161835  8468 sgd_solver.cpp:105] Iteration 7716, lr = 0.00216876
I0428 18:12:11.567170  8468 solver.cpp:218] Iteration 7728 (2.22012 iter/s, 5.40511s/12 iters), loss = 0.190469
I0428 18:12:11.567293  8468 solver.cpp:237]     Train net output #0: loss = 0.190469 (* 1 = 0.190469 loss)
I0428 18:12:11.567303  8468 sgd_solver.cpp:105] Iteration 7728, lr = 0.00216361
I0428 18:12:17.277549  8468 solver.cpp:218] Iteration 7740 (2.10157 iter/s, 5.71001s/12 iters), loss = 0.117419
I0428 18:12:17.277590  8468 solver.cpp:237]     Train net output #0: loss = 0.117419 (* 1 = 0.117419 loss)
I0428 18:12:17.277599  8468 sgd_solver.cpp:105] Iteration 7740, lr = 0.00215847
I0428 18:12:22.114804  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7752.caffemodel
I0428 18:12:22.694725  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7752.solverstate
I0428 18:12:23.120085  8468 solver.cpp:330] Iteration 7752, Testing net (#0)
I0428 18:12:23.120112  8468 net.cpp:676] Ignoring source layer train-data
I0428 18:12:24.586930  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:12:28.030323  8468 solver.cpp:397]     Test net output #0: accuracy = 0.434436
I0428 18:12:28.030359  8468 solver.cpp:397]     Test net output #1: loss = 3.6524 (* 1 = 3.6524 loss)
I0428 18:12:28.160256  8468 solver.cpp:218] Iteration 7752 (1.10272 iter/s, 10.8822s/12 iters), loss = 0.137458
I0428 18:12:28.160307  8468 solver.cpp:237]     Train net output #0: loss = 0.137458 (* 1 = 0.137458 loss)
I0428 18:12:28.160320  8468 sgd_solver.cpp:105] Iteration 7752, lr = 0.00215335
I0428 18:12:32.822734  8468 solver.cpp:218] Iteration 7764 (2.57388 iter/s, 4.66222s/12 iters), loss = 0.253931
I0428 18:12:32.822778  8468 solver.cpp:237]     Train net output #0: loss = 0.253931 (* 1 = 0.253931 loss)
I0428 18:12:32.822789  8468 sgd_solver.cpp:105] Iteration 7764, lr = 0.00214823
I0428 18:12:38.300952  8468 solver.cpp:218] Iteration 7776 (2.1906 iter/s, 5.47794s/12 iters), loss = 0.178485
I0428 18:12:38.300992  8468 solver.cpp:237]     Train net output #0: loss = 0.178485 (* 1 = 0.178485 loss)
I0428 18:12:38.301002  8468 sgd_solver.cpp:105] Iteration 7776, lr = 0.00214313
I0428 18:12:43.735524  8468 solver.cpp:218] Iteration 7788 (2.2082 iter/s, 5.4343s/12 iters), loss = 0.24323
I0428 18:12:43.735663  8468 solver.cpp:237]     Train net output #0: loss = 0.24323 (* 1 = 0.24323 loss)
I0428 18:12:43.735682  8468 sgd_solver.cpp:105] Iteration 7788, lr = 0.00213805
I0428 18:12:43.742202  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:12:49.266714  8468 solver.cpp:218] Iteration 7800 (2.16966 iter/s, 5.53082s/12 iters), loss = 0.120956
I0428 18:12:49.266757  8468 solver.cpp:237]     Train net output #0: loss = 0.120956 (* 1 = 0.120956 loss)
I0428 18:12:49.266767  8468 sgd_solver.cpp:105] Iteration 7800, lr = 0.00213297
I0428 18:12:54.739207  8468 solver.cpp:218] Iteration 7812 (2.1929 iter/s, 5.47222s/12 iters), loss = 0.107132
I0428 18:12:54.739248  8468 solver.cpp:237]     Train net output #0: loss = 0.107132 (* 1 = 0.107132 loss)
I0428 18:12:54.739257  8468 sgd_solver.cpp:105] Iteration 7812, lr = 0.00212791
I0428 18:13:00.062855  8468 solver.cpp:218] Iteration 7824 (2.25421 iter/s, 5.32337s/12 iters), loss = 0.196546
I0428 18:13:00.062911  8468 solver.cpp:237]     Train net output #0: loss = 0.196546 (* 1 = 0.196546 loss)
I0428 18:13:00.062925  8468 sgd_solver.cpp:105] Iteration 7824, lr = 0.00212285
I0428 18:13:05.423000  8468 solver.cpp:218] Iteration 7836 (2.23886 iter/s, 5.35986s/12 iters), loss = 0.217592
I0428 18:13:05.423045  8468 solver.cpp:237]     Train net output #0: loss = 0.217592 (* 1 = 0.217592 loss)
I0428 18:13:05.423054  8468 sgd_solver.cpp:105] Iteration 7836, lr = 0.00211781
I0428 18:13:10.873312  8468 solver.cpp:218] Iteration 7848 (2.20182 iter/s, 5.45004s/12 iters), loss = 0.151784
I0428 18:13:10.873353  8468 solver.cpp:237]     Train net output #0: loss = 0.151784 (* 1 = 0.151784 loss)
I0428 18:13:10.873363  8468 sgd_solver.cpp:105] Iteration 7848, lr = 0.00211279
I0428 18:13:13.026502  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7854.caffemodel
I0428 18:13:14.689566  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7854.solverstate
I0428 18:13:16.244551  8468 solver.cpp:330] Iteration 7854, Testing net (#0)
I0428 18:13:16.244570  8468 net.cpp:676] Ignoring source layer train-data
I0428 18:13:17.632016  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:13:21.196959  8468 solver.cpp:397]     Test net output #0: accuracy = 0.4375
I0428 18:13:21.196997  8468 solver.cpp:397]     Test net output #1: loss = 3.59462 (* 1 = 3.59462 loss)
I0428 18:13:23.190734  8468 solver.cpp:218] Iteration 7860 (0.974273 iter/s, 12.3169s/12 iters), loss = 0.13613
I0428 18:13:23.190775  8468 solver.cpp:237]     Train net output #0: loss = 0.13613 (* 1 = 0.13613 loss)
I0428 18:13:23.190785  8468 sgd_solver.cpp:105] Iteration 7860, lr = 0.00210777
I0428 18:13:28.568832  8468 solver.cpp:218] Iteration 7872 (2.23139 iter/s, 5.37782s/12 iters), loss = 0.120384
I0428 18:13:28.568873  8468 solver.cpp:237]     Train net output #0: loss = 0.120384 (* 1 = 0.120384 loss)
I0428 18:13:28.568882  8468 sgd_solver.cpp:105] Iteration 7872, lr = 0.00210277
I0428 18:13:34.017750  8468 solver.cpp:218] Iteration 7884 (2.20238 iter/s, 5.44864s/12 iters), loss = 0.150708
I0428 18:13:34.017800  8468 solver.cpp:237]     Train net output #0: loss = 0.150708 (* 1 = 0.150708 loss)
I0428 18:13:34.017812  8468 sgd_solver.cpp:105] Iteration 7884, lr = 0.00209777
I0428 18:13:36.485612  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:13:39.608021  8468 solver.cpp:218] Iteration 7896 (2.1467 iter/s, 5.58998s/12 iters), loss = 0.158864
I0428 18:13:39.608081  8468 solver.cpp:237]     Train net output #0: loss = 0.158864 (* 1 = 0.158864 loss)
I0428 18:13:39.608094  8468 sgd_solver.cpp:105] Iteration 7896, lr = 0.00209279
I0428 18:13:45.285928  8468 solver.cpp:218] Iteration 7908 (2.11357 iter/s, 5.67761s/12 iters), loss = 0.157448
I0428 18:13:45.287047  8468 solver.cpp:237]     Train net output #0: loss = 0.157448 (* 1 = 0.157448 loss)
I0428 18:13:45.287056  8468 sgd_solver.cpp:105] Iteration 7908, lr = 0.00208782
I0428 18:13:50.646668  8468 solver.cpp:218] Iteration 7920 (2.23906 iter/s, 5.3594s/12 iters), loss = 0.14125
I0428 18:13:50.646713  8468 solver.cpp:237]     Train net output #0: loss = 0.14125 (* 1 = 0.14125 loss)
I0428 18:13:50.646721  8468 sgd_solver.cpp:105] Iteration 7920, lr = 0.00208287
I0428 18:13:56.109462  8468 solver.cpp:218] Iteration 7932 (2.1968 iter/s, 5.4625s/12 iters), loss = 0.23788
I0428 18:13:56.109501  8468 solver.cpp:237]     Train net output #0: loss = 0.23788 (* 1 = 0.23788 loss)
I0428 18:13:56.109510  8468 sgd_solver.cpp:105] Iteration 7932, lr = 0.00207792
I0428 18:14:01.598891  8468 solver.cpp:218] Iteration 7944 (2.18613 iter/s, 5.48916s/12 iters), loss = 0.174911
I0428 18:14:01.598930  8468 solver.cpp:237]     Train net output #0: loss = 0.174911 (* 1 = 0.174911 loss)
I0428 18:14:01.598939  8468 sgd_solver.cpp:105] Iteration 7944, lr = 0.00207299
I0428 18:14:06.500922  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7956.caffemodel
I0428 18:14:07.103825  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7956.solverstate
I0428 18:14:07.533442  8468 solver.cpp:330] Iteration 7956, Testing net (#0)
I0428 18:14:07.533463  8468 net.cpp:676] Ignoring source layer train-data
I0428 18:14:08.830888  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:14:12.225193  8468 solver.cpp:397]     Test net output #0: accuracy = 0.44424
I0428 18:14:12.225224  8468 solver.cpp:397]     Test net output #1: loss = 3.66552 (* 1 = 3.66552 loss)
I0428 18:14:12.354332  8468 solver.cpp:218] Iteration 7956 (1.11576 iter/s, 10.755s/12 iters), loss = 0.11085
I0428 18:14:12.354373  8468 solver.cpp:237]     Train net output #0: loss = 0.11085 (* 1 = 0.11085 loss)
I0428 18:14:12.354382  8468 sgd_solver.cpp:105] Iteration 7956, lr = 0.00206807
I0428 18:14:16.909530  8468 solver.cpp:218] Iteration 7968 (2.63449 iter/s, 4.55496s/12 iters), loss = 0.15136
I0428 18:14:16.909768  8468 solver.cpp:237]     Train net output #0: loss = 0.15136 (* 1 = 0.15136 loss)
I0428 18:14:16.909780  8468 sgd_solver.cpp:105] Iteration 7968, lr = 0.00206316
I0428 18:14:22.421262  8468 solver.cpp:218] Iteration 7980 (2.17736 iter/s, 5.51126s/12 iters), loss = 0.155861
I0428 18:14:22.421314  8468 solver.cpp:237]     Train net output #0: loss = 0.155861 (* 1 = 0.155861 loss)
I0428 18:14:22.421325  8468 sgd_solver.cpp:105] Iteration 7980, lr = 0.00205826
I0428 18:14:27.076161  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:14:27.853905  8468 solver.cpp:218] Iteration 7992 (2.20898 iter/s, 5.43236s/12 iters), loss = 0.106269
I0428 18:14:27.853947  8468 solver.cpp:237]     Train net output #0: loss = 0.106269 (* 1 = 0.106269 loss)
I0428 18:14:27.853955  8468 sgd_solver.cpp:105] Iteration 7992, lr = 0.00205337
I0428 18:14:33.552950  8468 solver.cpp:218] Iteration 8004 (2.10572 iter/s, 5.69876s/12 iters), loss = 0.0957698
I0428 18:14:33.553000  8468 solver.cpp:237]     Train net output #0: loss = 0.0957698 (* 1 = 0.0957698 loss)
I0428 18:14:33.553014  8468 sgd_solver.cpp:105] Iteration 8004, lr = 0.0020485
I0428 18:14:38.893476  8468 solver.cpp:218] Iteration 8016 (2.24709 iter/s, 5.34025s/12 iters), loss = 0.175921
I0428 18:14:38.893523  8468 solver.cpp:237]     Train net output #0: loss = 0.175921 (* 1 = 0.175921 loss)
I0428 18:14:38.893537  8468 sgd_solver.cpp:105] Iteration 8016, lr = 0.00204363
I0428 18:14:44.274850  8468 solver.cpp:218] Iteration 8028 (2.23003 iter/s, 5.3811s/12 iters), loss = 0.106018
I0428 18:14:44.274888  8468 solver.cpp:237]     Train net output #0: loss = 0.106018 (* 1 = 0.106018 loss)
I0428 18:14:44.274896  8468 sgd_solver.cpp:105] Iteration 8028, lr = 0.00203878
I0428 18:14:49.740633  8468 solver.cpp:218] Iteration 8040 (2.19559 iter/s, 5.46551s/12 iters), loss = 0.200159
I0428 18:14:49.740780  8468 solver.cpp:237]     Train net output #0: loss = 0.200159 (* 1 = 0.200159 loss)
I0428 18:14:49.740794  8468 sgd_solver.cpp:105] Iteration 8040, lr = 0.00203394
I0428 18:14:55.236843  8468 solver.cpp:218] Iteration 8052 (2.18347 iter/s, 5.49583s/12 iters), loss = 0.0606009
I0428 18:14:55.236886  8468 solver.cpp:237]     Train net output #0: loss = 0.0606009 (* 1 = 0.0606009 loss)
I0428 18:14:55.236894  8468 sgd_solver.cpp:105] Iteration 8052, lr = 0.00202911
I0428 18:14:57.401146  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8058.caffemodel
I0428 18:14:59.616083  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8058.solverstate
I0428 18:15:00.089463  8468 solver.cpp:330] Iteration 8058, Testing net (#0)
I0428 18:15:00.089483  8468 net.cpp:676] Ignoring source layer train-data
I0428 18:15:01.355072  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:15:04.840754  8468 solver.cpp:397]     Test net output #0: accuracy = 0.449755
I0428 18:15:04.840786  8468 solver.cpp:397]     Test net output #1: loss = 3.7133 (* 1 = 3.7133 loss)
I0428 18:15:06.807904  8468 solver.cpp:218] Iteration 8064 (1.03712 iter/s, 11.5705s/12 iters), loss = 0.213291
I0428 18:15:06.807955  8468 solver.cpp:237]     Train net output #0: loss = 0.213291 (* 1 = 0.213291 loss)
I0428 18:15:06.807991  8468 sgd_solver.cpp:105] Iteration 8064, lr = 0.00202429
I0428 18:15:12.218634  8468 solver.cpp:218] Iteration 8076 (2.21793 iter/s, 5.41045s/12 iters), loss = 0.140587
I0428 18:15:12.218683  8468 solver.cpp:237]     Train net output #0: loss = 0.140587 (* 1 = 0.140587 loss)
I0428 18:15:12.218695  8468 sgd_solver.cpp:105] Iteration 8076, lr = 0.00201949
I0428 18:15:17.660218  8468 solver.cpp:218] Iteration 8088 (2.20535 iter/s, 5.4413s/12 iters), loss = 0.158648
I0428 18:15:17.660255  8468 solver.cpp:237]     Train net output #0: loss = 0.158648 (* 1 = 0.158648 loss)
I0428 18:15:17.660264  8468 sgd_solver.cpp:105] Iteration 8088, lr = 0.00201469
I0428 18:15:19.154857  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:15:23.100922  8468 solver.cpp:218] Iteration 8100 (2.20571 iter/s, 5.44044s/12 iters), loss = 0.16467
I0428 18:15:23.101233  8468 solver.cpp:237]     Train net output #0: loss = 0.16467 (* 1 = 0.16467 loss)
I0428 18:15:23.101243  8468 sgd_solver.cpp:105] Iteration 8100, lr = 0.00200991
I0428 18:15:28.488370  8468 solver.cpp:218] Iteration 8112 (2.22762 iter/s, 5.38691s/12 iters), loss = 0.168017
I0428 18:15:28.488407  8468 solver.cpp:237]     Train net output #0: loss = 0.168017 (* 1 = 0.168017 loss)
I0428 18:15:28.488418  8468 sgd_solver.cpp:105] Iteration 8112, lr = 0.00200514
I0428 18:15:34.004758  8468 solver.cpp:218] Iteration 8124 (2.17545 iter/s, 5.51611s/12 iters), loss = 0.0879743
I0428 18:15:34.004798  8468 solver.cpp:237]     Train net output #0: loss = 0.0879743 (* 1 = 0.0879743 loss)
I0428 18:15:34.004807  8468 sgd_solver.cpp:105] Iteration 8124, lr = 0.00200038
I0428 18:15:39.754794  8468 solver.cpp:218] Iteration 8136 (2.08705 iter/s, 5.74975s/12 iters), loss = 0.104225
I0428 18:15:39.754844  8468 solver.cpp:237]     Train net output #0: loss = 0.104225 (* 1 = 0.104225 loss)
I0428 18:15:39.754860  8468 sgd_solver.cpp:105] Iteration 8136, lr = 0.00199563
I0428 18:15:45.144354  8468 solver.cpp:218] Iteration 8148 (2.22664 iter/s, 5.38928s/12 iters), loss = 0.216089
I0428 18:15:45.144393  8468 solver.cpp:237]     Train net output #0: loss = 0.21609 (* 1 = 0.21609 loss)
I0428 18:15:45.144404  8468 sgd_solver.cpp:105] Iteration 8148, lr = 0.00199089
I0428 18:15:50.135839  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8160.caffemodel
I0428 18:15:51.039075  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8160.solverstate
I0428 18:15:51.550385  8468 solver.cpp:330] Iteration 8160, Testing net (#0)
I0428 18:15:51.550405  8468 net.cpp:676] Ignoring source layer train-data
I0428 18:15:52.724870  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:15:56.062176  8468 solver.cpp:397]     Test net output #0: accuracy = 0.459559
I0428 18:15:56.062328  8468 solver.cpp:397]     Test net output #1: loss = 3.59275 (* 1 = 3.59275 loss)
I0428 18:15:56.194576  8468 solver.cpp:218] Iteration 8160 (1.086 iter/s, 11.0497s/12 iters), loss = 0.104472
I0428 18:15:56.194618  8468 solver.cpp:237]     Train net output #0: loss = 0.104472 (* 1 = 0.104472 loss)
I0428 18:15:56.194628  8468 sgd_solver.cpp:105] Iteration 8160, lr = 0.00198616
I0428 18:16:00.774523  8468 solver.cpp:218] Iteration 8172 (2.62025 iter/s, 4.57971s/12 iters), loss = 0.0949376
I0428 18:16:00.774564  8468 solver.cpp:237]     Train net output #0: loss = 0.0949376 (* 1 = 0.0949376 loss)
I0428 18:16:00.774574  8468 sgd_solver.cpp:105] Iteration 8172, lr = 0.00198145
I0428 18:16:06.336133  8468 solver.cpp:218] Iteration 8184 (2.15776 iter/s, 5.56133s/12 iters), loss = 0.103707
I0428 18:16:06.336174  8468 solver.cpp:237]     Train net output #0: loss = 0.103707 (* 1 = 0.103707 loss)
I0428 18:16:06.336184  8468 sgd_solver.cpp:105] Iteration 8184, lr = 0.00197674
I0428 18:16:10.195322  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:16:11.761174  8468 solver.cpp:218] Iteration 8196 (2.21208 iter/s, 5.42477s/12 iters), loss = 0.213378
I0428 18:16:11.761224  8468 solver.cpp:237]     Train net output #0: loss = 0.213378 (* 1 = 0.213378 loss)
I0428 18:16:11.761236  8468 sgd_solver.cpp:105] Iteration 8196, lr = 0.00197205
I0428 18:16:17.199298  8468 solver.cpp:218] Iteration 8208 (2.20676 iter/s, 5.43785s/12 iters), loss = 0.137655
I0428 18:16:17.199339  8468 solver.cpp:237]     Train net output #0: loss = 0.137655 (* 1 = 0.137655 loss)
I0428 18:16:17.199348  8468 sgd_solver.cpp:105] Iteration 8208, lr = 0.00196737
I0428 18:16:22.640398  8468 solver.cpp:218] Iteration 8220 (2.20556 iter/s, 5.4408s/12 iters), loss = 0.0951256
I0428 18:16:22.640446  8468 solver.cpp:237]     Train net output #0: loss = 0.0951257 (* 1 = 0.0951257 loss)
I0428 18:16:22.640460  8468 sgd_solver.cpp:105] Iteration 8220, lr = 0.0019627
I0428 18:16:28.066934  8468 solver.cpp:218] Iteration 8232 (2.21147 iter/s, 5.42626s/12 iters), loss = 0.163846
I0428 18:16:28.067235  8468 solver.cpp:237]     Train net output #0: loss = 0.163846 (* 1 = 0.163846 loss)
I0428 18:16:28.067246  8468 sgd_solver.cpp:105] Iteration 8232, lr = 0.00195804
I0428 18:16:33.780146  8468 solver.cpp:218] Iteration 8244 (2.10059 iter/s, 5.71267s/12 iters), loss = 0.311337
I0428 18:16:33.780190  8468 solver.cpp:237]     Train net output #0: loss = 0.311337 (* 1 = 0.311337 loss)
I0428 18:16:33.780200  8468 sgd_solver.cpp:105] Iteration 8244, lr = 0.00195339
I0428 18:16:39.242928  8468 solver.cpp:218] Iteration 8256 (2.19679 iter/s, 5.4625s/12 iters), loss = 0.0533028
I0428 18:16:39.242972  8468 solver.cpp:237]     Train net output #0: loss = 0.0533028 (* 1 = 0.0533028 loss)
I0428 18:16:39.242981  8468 sgd_solver.cpp:105] Iteration 8256, lr = 0.00194875
I0428 18:16:41.425487  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8262.caffemodel
I0428 18:16:42.015281  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8262.solverstate
I0428 18:16:42.450250  8468 solver.cpp:330] Iteration 8262, Testing net (#0)
I0428 18:16:42.450271  8468 net.cpp:676] Ignoring source layer train-data
I0428 18:16:43.737465  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:16:47.205375  8468 solver.cpp:397]     Test net output #0: accuracy = 0.4375
I0428 18:16:47.205413  8468 solver.cpp:397]     Test net output #1: loss = 3.77688 (* 1 = 3.77688 loss)
I0428 18:16:49.107982  8468 solver.cpp:218] Iteration 8268 (1.21647 iter/s, 9.8646s/12 iters), loss = 0.0630831
I0428 18:16:49.108026  8468 solver.cpp:237]     Train net output #0: loss = 0.0630831 (* 1 = 0.0630831 loss)
I0428 18:16:49.108036  8468 sgd_solver.cpp:105] Iteration 8268, lr = 0.00194412
I0428 18:16:54.553380  8468 solver.cpp:218] Iteration 8280 (2.20381 iter/s, 5.44512s/12 iters), loss = 0.17053
I0428 18:16:54.553437  8468 solver.cpp:237]     Train net output #0: loss = 0.17053 (* 1 = 0.17053 loss)
I0428 18:16:54.553452  8468 sgd_solver.cpp:105] Iteration 8280, lr = 0.00193951
I0428 18:16:59.880242  8468 solver.cpp:218] Iteration 8292 (2.25285 iter/s, 5.32658s/12 iters), loss = 0.114944
I0428 18:16:59.880406  8468 solver.cpp:237]     Train net output #0: loss = 0.114944 (* 1 = 0.114944 loss)
I0428 18:16:59.880417  8468 sgd_solver.cpp:105] Iteration 8292, lr = 0.0019349
I0428 18:17:00.679932  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:17:05.585566  8468 solver.cpp:218] Iteration 8304 (2.10345 iter/s, 5.70492s/12 iters), loss = 0.170216
I0428 18:17:05.585618  8468 solver.cpp:237]     Train net output #0: loss = 0.170216 (* 1 = 0.170216 loss)
I0428 18:17:05.585629  8468 sgd_solver.cpp:105] Iteration 8304, lr = 0.00193031
I0428 18:17:07.447225  8468 blocking_queue.cpp:49] Waiting for data
I0428 18:17:11.110246  8468 solver.cpp:218] Iteration 8316 (2.17218 iter/s, 5.52439s/12 iters), loss = 0.149733
I0428 18:17:11.110286  8468 solver.cpp:237]     Train net output #0: loss = 0.149733 (* 1 = 0.149733 loss)
I0428 18:17:11.110296  8468 sgd_solver.cpp:105] Iteration 8316, lr = 0.00192573
I0428 18:17:16.436734  8468 solver.cpp:218] Iteration 8328 (2.25301 iter/s, 5.32622s/12 iters), loss = 0.126862
I0428 18:17:16.436791  8468 solver.cpp:237]     Train net output #0: loss = 0.126862 (* 1 = 0.126862 loss)
I0428 18:17:16.436803  8468 sgd_solver.cpp:105] Iteration 8328, lr = 0.00192115
I0428 18:17:21.754343  8468 solver.cpp:218] Iteration 8340 (2.25677 iter/s, 5.31733s/12 iters), loss = 0.0928029
I0428 18:17:21.754392  8468 solver.cpp:237]     Train net output #0: loss = 0.0928029 (* 1 = 0.0928029 loss)
I0428 18:17:21.754405  8468 sgd_solver.cpp:105] Iteration 8340, lr = 0.00191659
I0428 18:17:27.072163  8468 solver.cpp:218] Iteration 8352 (2.25668 iter/s, 5.31755s/12 iters), loss = 0.204227
I0428 18:17:27.072221  8468 solver.cpp:237]     Train net output #0: loss = 0.204227 (* 1 = 0.204227 loss)
I0428 18:17:27.072234  8468 sgd_solver.cpp:105] Iteration 8352, lr = 0.00191204
I0428 18:17:31.864423  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8364.caffemodel
I0428 18:17:34.842550  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8364.solverstate
I0428 18:17:36.830654  8468 solver.cpp:330] Iteration 8364, Testing net (#0)
I0428 18:17:36.830673  8468 net.cpp:676] Ignoring source layer train-data
I0428 18:17:37.994112  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:17:41.421569  8468 solver.cpp:397]     Test net output #0: accuracy = 0.444853
I0428 18:17:41.421610  8468 solver.cpp:397]     Test net output #1: loss = 3.68074 (* 1 = 3.68074 loss)
I0428 18:17:41.546959  8468 solver.cpp:218] Iteration 8364 (0.829064 iter/s, 14.4741s/12 iters), loss = 0.10777
I0428 18:17:41.547009  8468 solver.cpp:237]     Train net output #0: loss = 0.10777 (* 1 = 0.10777 loss)
I0428 18:17:41.547022  8468 sgd_solver.cpp:105] Iteration 8364, lr = 0.0019075
I0428 18:17:46.065860  8468 solver.cpp:218] Iteration 8376 (2.65566 iter/s, 4.51866s/12 iters), loss = 0.0877378
I0428 18:17:46.065902  8468 solver.cpp:237]     Train net output #0: loss = 0.0877378 (* 1 = 0.0877378 loss)
I0428 18:17:46.065912  8468 sgd_solver.cpp:105] Iteration 8376, lr = 0.00190297
I0428 18:17:51.780046  8468 solver.cpp:218] Iteration 8388 (2.10014 iter/s, 5.7139s/12 iters), loss = 0.202565
I0428 18:17:51.780084  8468 solver.cpp:237]     Train net output #0: loss = 0.202565 (* 1 = 0.202565 loss)
I0428 18:17:51.780093  8468 sgd_solver.cpp:105] Iteration 8388, lr = 0.00189846
I0428 18:17:54.801250  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:17:57.160145  8468 solver.cpp:218] Iteration 8400 (2.23055 iter/s, 5.37983s/12 iters), loss = 0.200196
I0428 18:17:57.160185  8468 solver.cpp:237]     Train net output #0: loss = 0.200196 (* 1 = 0.200196 loss)
I0428 18:17:57.160194  8468 sgd_solver.cpp:105] Iteration 8400, lr = 0.00189395
I0428 18:18:02.740175  8468 solver.cpp:218] Iteration 8412 (2.15063 iter/s, 5.57975s/12 iters), loss = 0.126678
I0428 18:18:02.742506  8468 solver.cpp:237]     Train net output #0: loss = 0.126678 (* 1 = 0.126678 loss)
I0428 18:18:02.742519  8468 sgd_solver.cpp:105] Iteration 8412, lr = 0.00188945
I0428 18:18:08.208248  8468 solver.cpp:218] Iteration 8424 (2.19559 iter/s, 5.46551s/12 iters), loss = 0.24558
I0428 18:18:08.208302  8468 solver.cpp:237]     Train net output #0: loss = 0.24558 (* 1 = 0.24558 loss)
I0428 18:18:08.208314  8468 sgd_solver.cpp:105] Iteration 8424, lr = 0.00188497
I0428 18:18:13.754648  8468 solver.cpp:218] Iteration 8436 (2.16368 iter/s, 5.54611s/12 iters), loss = 0.206044
I0428 18:18:13.754699  8468 solver.cpp:237]     Train net output #0: loss = 0.206044 (* 1 = 0.206044 loss)
I0428 18:18:13.754712  8468 sgd_solver.cpp:105] Iteration 8436, lr = 0.00188049
I0428 18:18:19.410465  8468 solver.cpp:218] Iteration 8448 (2.12182 iter/s, 5.65552s/12 iters), loss = 0.214765
I0428 18:18:19.410503  8468 solver.cpp:237]     Train net output #0: loss = 0.214765 (* 1 = 0.214765 loss)
I0428 18:18:19.410512  8468 sgd_solver.cpp:105] Iteration 8448, lr = 0.00187603
I0428 18:18:24.869962  8468 solver.cpp:218] Iteration 8460 (2.19812 iter/s, 5.45922s/12 iters), loss = 0.113494
I0428 18:18:24.870012  8468 solver.cpp:237]     Train net output #0: loss = 0.113494 (* 1 = 0.113494 loss)
I0428 18:18:24.870024  8468 sgd_solver.cpp:105] Iteration 8460, lr = 0.00187157
I0428 18:18:27.174854  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8466.caffemodel
I0428 18:18:27.788139  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8466.solverstate
I0428 18:18:28.218590  8468 solver.cpp:330] Iteration 8466, Testing net (#0)
I0428 18:18:28.218611  8468 net.cpp:676] Ignoring source layer train-data
I0428 18:18:29.385613  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:18:32.932407  8468 solver.cpp:397]     Test net output #0: accuracy = 0.448529
I0428 18:18:32.932550  8468 solver.cpp:397]     Test net output #1: loss = 3.56376 (* 1 = 3.56376 loss)
I0428 18:18:35.036587  8468 solver.cpp:218] Iteration 8472 (1.18039 iter/s, 10.1662s/12 iters), loss = 0.123284
I0428 18:18:35.036628  8468 solver.cpp:237]     Train net output #0: loss = 0.123284 (* 1 = 0.123284 loss)
I0428 18:18:35.036638  8468 sgd_solver.cpp:105] Iteration 8472, lr = 0.00186713
I0428 18:18:40.522289  8468 solver.cpp:218] Iteration 8484 (2.18761 iter/s, 5.48543s/12 iters), loss = 0.126472
I0428 18:18:40.522327  8468 solver.cpp:237]     Train net output #0: loss = 0.126472 (* 1 = 0.126472 loss)
I0428 18:18:40.522336  8468 sgd_solver.cpp:105] Iteration 8484, lr = 0.0018627
I0428 18:18:46.003150  8468 solver.cpp:218] Iteration 8496 (2.18955 iter/s, 5.48059s/12 iters), loss = 0.11628
I0428 18:18:46.003190  8468 solver.cpp:237]     Train net output #0: loss = 0.11628 (* 1 = 0.11628 loss)
I0428 18:18:46.003199  8468 sgd_solver.cpp:105] Iteration 8496, lr = 0.00185827
I0428 18:18:46.044162  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:18:51.543071  8468 solver.cpp:218] Iteration 8508 (2.16621 iter/s, 5.53964s/12 iters), loss = 0.122272
I0428 18:18:51.543119  8468 solver.cpp:237]     Train net output #0: loss = 0.122272 (* 1 = 0.122272 loss)
I0428 18:18:51.543133  8468 sgd_solver.cpp:105] Iteration 8508, lr = 0.00185386
I0428 18:18:56.987577  8468 solver.cpp:218] Iteration 8520 (2.20417 iter/s, 5.44423s/12 iters), loss = 0.146132
I0428 18:18:56.987628  8468 solver.cpp:237]     Train net output #0: loss = 0.146132 (* 1 = 0.146132 loss)
I0428 18:18:56.987640  8468 sgd_solver.cpp:105] Iteration 8520, lr = 0.00184946
I0428 18:19:02.445546  8468 solver.cpp:218] Iteration 8532 (2.19874 iter/s, 5.45768s/12 iters), loss = 0.127459
I0428 18:19:02.445602  8468 solver.cpp:237]     Train net output #0: loss = 0.127459 (* 1 = 0.127459 loss)
I0428 18:19:02.445616  8468 sgd_solver.cpp:105] Iteration 8532, lr = 0.00184507
I0428 18:19:07.850167  8468 solver.cpp:218] Iteration 8544 (2.22044 iter/s, 5.40434s/12 iters), loss = 0.094348
I0428 18:19:07.851887  8468 solver.cpp:237]     Train net output #0: loss = 0.094348 (* 1 = 0.094348 loss)
I0428 18:19:07.851902  8468 sgd_solver.cpp:105] Iteration 8544, lr = 0.00184069
I0428 18:19:13.378906  8468 solver.cpp:218] Iteration 8556 (2.17125 iter/s, 5.52678s/12 iters), loss = 0.130748
I0428 18:19:13.378947  8468 solver.cpp:237]     Train net output #0: loss = 0.130748 (* 1 = 0.130748 loss)
I0428 18:19:13.378958  8468 sgd_solver.cpp:105] Iteration 8556, lr = 0.00183632
I0428 18:19:18.235560  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8568.caffemodel
I0428 18:19:18.831163  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8568.solverstate
I0428 18:19:19.287426  8468 solver.cpp:330] Iteration 8568, Testing net (#0)
I0428 18:19:19.287451  8468 net.cpp:676] Ignoring source layer train-data
I0428 18:19:20.463259  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:19:24.090878  8468 solver.cpp:397]     Test net output #0: accuracy = 0.456495
I0428 18:19:24.090917  8468 solver.cpp:397]     Test net output #1: loss = 3.68266 (* 1 = 3.68266 loss)
I0428 18:19:24.222934  8468 solver.cpp:218] Iteration 8568 (1.10665 iter/s, 10.8435s/12 iters), loss = 0.0788408
I0428 18:19:24.222977  8468 solver.cpp:237]     Train net output #0: loss = 0.0788408 (* 1 = 0.0788408 loss)
I0428 18:19:24.222986  8468 sgd_solver.cpp:105] Iteration 8568, lr = 0.00183196
I0428 18:19:29.010996  8468 solver.cpp:218] Iteration 8580 (2.50637 iter/s, 4.78781s/12 iters), loss = 0.204331
I0428 18:19:29.011045  8468 solver.cpp:237]     Train net output #0: loss = 0.204331 (* 1 = 0.204331 loss)
I0428 18:19:29.011057  8468 sgd_solver.cpp:105] Iteration 8580, lr = 0.00182761
I0428 18:19:34.611786  8468 solver.cpp:218] Iteration 8592 (2.14267 iter/s, 5.6005s/12 iters), loss = 0.160912
I0428 18:19:34.611825  8468 solver.cpp:237]     Train net output #0: loss = 0.160912 (* 1 = 0.160912 loss)
I0428 18:19:34.611835  8468 sgd_solver.cpp:105] Iteration 8592, lr = 0.00182327
I0428 18:19:36.929754  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:19:40.016367  8468 solver.cpp:218] Iteration 8604 (2.22045 iter/s, 5.40431s/12 iters), loss = 0.0456946
I0428 18:19:40.016520  8468 solver.cpp:237]     Train net output #0: loss = 0.0456946 (* 1 = 0.0456946 loss)
I0428 18:19:40.016532  8468 sgd_solver.cpp:105] Iteration 8604, lr = 0.00181894
I0428 18:19:45.385658  8468 solver.cpp:218] Iteration 8616 (2.23508 iter/s, 5.36893s/12 iters), loss = 0.151901
I0428 18:19:45.385718  8468 solver.cpp:237]     Train net output #0: loss = 0.151901 (* 1 = 0.151901 loss)
I0428 18:19:45.385731  8468 sgd_solver.cpp:105] Iteration 8616, lr = 0.00181462
I0428 18:19:50.792248  8468 solver.cpp:218] Iteration 8628 (2.21963 iter/s, 5.4063s/12 iters), loss = 0.174435
I0428 18:19:50.792287  8468 solver.cpp:237]     Train net output #0: loss = 0.174436 (* 1 = 0.174436 loss)
I0428 18:19:50.792297  8468 sgd_solver.cpp:105] Iteration 8628, lr = 0.00181031
I0428 18:19:56.268230  8468 solver.cpp:218] Iteration 8640 (2.1915 iter/s, 5.4757s/12 iters), loss = 0.0903176
I0428 18:19:56.268286  8468 solver.cpp:237]     Train net output #0: loss = 0.0903176 (* 1 = 0.0903176 loss)
I0428 18:19:56.268301  8468 sgd_solver.cpp:105] Iteration 8640, lr = 0.00180602
I0428 18:20:01.905751  8468 solver.cpp:218] Iteration 8652 (2.12871 iter/s, 5.63722s/12 iters), loss = 0.127434
I0428 18:20:01.905791  8468 solver.cpp:237]     Train net output #0: loss = 0.127434 (* 1 = 0.127434 loss)
I0428 18:20:01.905800  8468 sgd_solver.cpp:105] Iteration 8652, lr = 0.00180173
I0428 18:20:07.524119  8468 solver.cpp:218] Iteration 8664 (2.13596 iter/s, 5.61809s/12 iters), loss = 0.194039
I0428 18:20:07.524160  8468 solver.cpp:237]     Train net output #0: loss = 0.194039 (* 1 = 0.194039 loss)
I0428 18:20:07.524169  8468 sgd_solver.cpp:105] Iteration 8664, lr = 0.00179745
I0428 18:20:09.763949  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8670.caffemodel
I0428 18:20:11.390866  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8670.solverstate
I0428 18:20:13.343274  8468 solver.cpp:330] Iteration 8670, Testing net (#0)
I0428 18:20:13.343297  8468 net.cpp:676] Ignoring source layer train-data
I0428 18:20:14.365799  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:20:18.061638  8468 solver.cpp:397]     Test net output #0: accuracy = 0.450368
I0428 18:20:18.061677  8468 solver.cpp:397]     Test net output #1: loss = 3.77044 (* 1 = 3.77044 loss)
I0428 18:20:20.074259  8468 solver.cpp:218] Iteration 8676 (0.956207 iter/s, 12.5496s/12 iters), loss = 0.119689
I0428 18:20:20.074331  8468 solver.cpp:237]     Train net output #0: loss = 0.119689 (* 1 = 0.119689 loss)
I0428 18:20:20.074345  8468 sgd_solver.cpp:105] Iteration 8676, lr = 0.00179318
I0428 18:20:25.580116  8468 solver.cpp:218] Iteration 8688 (2.17962 iter/s, 5.50555s/12 iters), loss = 0.0874789
I0428 18:20:25.580166  8468 solver.cpp:237]     Train net output #0: loss = 0.087479 (* 1 = 0.087479 loss)
I0428 18:20:25.580179  8468 sgd_solver.cpp:105] Iteration 8688, lr = 0.00178893
I0428 18:20:30.322980  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:20:31.083916  8468 solver.cpp:218] Iteration 8700 (2.18042 iter/s, 5.50352s/12 iters), loss = 0.165308
I0428 18:20:31.083966  8468 solver.cpp:237]     Train net output #0: loss = 0.165309 (* 1 = 0.165309 loss)
I0428 18:20:31.083976  8468 sgd_solver.cpp:105] Iteration 8700, lr = 0.00178468
I0428 18:20:36.605342  8468 solver.cpp:218] Iteration 8712 (2.17346 iter/s, 5.52114s/12 iters), loss = 0.14242
I0428 18:20:36.605396  8468 solver.cpp:237]     Train net output #0: loss = 0.14242 (* 1 = 0.14242 loss)
I0428 18:20:36.605408  8468 sgd_solver.cpp:105] Iteration 8712, lr = 0.00178044
I0428 18:20:42.297425  8468 solver.cpp:218] Iteration 8724 (2.1083 iter/s, 5.69179s/12 iters), loss = 0.113357
I0428 18:20:42.297623  8468 solver.cpp:237]     Train net output #0: loss = 0.113357 (* 1 = 0.113357 loss)
I0428 18:20:42.297632  8468 sgd_solver.cpp:105] Iteration 8724, lr = 0.00177621
I0428 18:20:47.578248  8468 solver.cpp:218] Iteration 8736 (2.27256 iter/s, 5.2804s/12 iters), loss = 0.0752215
I0428 18:20:47.578302  8468 solver.cpp:237]     Train net output #0: loss = 0.0752216 (* 1 = 0.0752216 loss)
I0428 18:20:47.578316  8468 sgd_solver.cpp:105] Iteration 8736, lr = 0.001772
I0428 18:20:53.016769  8468 solver.cpp:218] Iteration 8748 (2.2066 iter/s, 5.43824s/12 iters), loss = 0.0635986
I0428 18:20:53.016808  8468 solver.cpp:237]     Train net output #0: loss = 0.0635987 (* 1 = 0.0635987 loss)
I0428 18:20:53.016817  8468 sgd_solver.cpp:105] Iteration 8748, lr = 0.00176779
I0428 18:20:58.424741  8468 solver.cpp:218] Iteration 8760 (2.21906 iter/s, 5.4077s/12 iters), loss = 0.116993
I0428 18:20:58.424782  8468 solver.cpp:237]     Train net output #0: loss = 0.116993 (* 1 = 0.116993 loss)
I0428 18:20:58.424793  8468 sgd_solver.cpp:105] Iteration 8760, lr = 0.00176359
I0428 18:21:03.308761  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8772.caffemodel
I0428 18:21:03.900913  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8772.solverstate
I0428 18:21:04.357877  8468 solver.cpp:330] Iteration 8772, Testing net (#0)
I0428 18:21:04.357898  8468 net.cpp:676] Ignoring source layer train-data
I0428 18:21:05.388607  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:21:09.230118  8468 solver.cpp:397]     Test net output #0: accuracy = 0.46201
I0428 18:21:09.230160  8468 solver.cpp:397]     Test net output #1: loss = 3.64208 (* 1 = 3.64208 loss)
I0428 18:21:09.361946  8468 solver.cpp:218] Iteration 8772 (1.09722 iter/s, 10.9367s/12 iters), loss = 0.134019
I0428 18:21:09.361987  8468 solver.cpp:237]     Train net output #0: loss = 0.134019 (* 1 = 0.134019 loss)
I0428 18:21:09.361999  8468 sgd_solver.cpp:105] Iteration 8772, lr = 0.00175941
I0428 18:21:13.794185  8468 solver.cpp:218] Iteration 8784 (2.70758 iter/s, 4.432s/12 iters), loss = 0.0797514
I0428 18:21:13.794363  8468 solver.cpp:237]     Train net output #0: loss = 0.0797514 (* 1 = 0.0797514 loss)
I0428 18:21:13.794378  8468 sgd_solver.cpp:105] Iteration 8784, lr = 0.00175523
I0428 18:21:19.381525  8468 solver.cpp:218] Iteration 8796 (2.14787 iter/s, 5.58693s/12 iters), loss = 0.0622259
I0428 18:21:19.381570  8468 solver.cpp:237]     Train net output #0: loss = 0.062226 (* 1 = 0.062226 loss)
I0428 18:21:19.381580  8468 sgd_solver.cpp:105] Iteration 8796, lr = 0.00175106
I0428 18:21:20.957549  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:21:24.808706  8468 solver.cpp:218] Iteration 8808 (2.21121 iter/s, 5.4269s/12 iters), loss = 0.0841762
I0428 18:21:24.808753  8468 solver.cpp:237]     Train net output #0: loss = 0.0841763 (* 1 = 0.0841763 loss)
I0428 18:21:24.808763  8468 sgd_solver.cpp:105] Iteration 8808, lr = 0.0017469
I0428 18:21:30.208125  8468 solver.cpp:218] Iteration 8820 (2.22257 iter/s, 5.39914s/12 iters), loss = 0.226731
I0428 18:21:30.208168  8468 solver.cpp:237]     Train net output #0: loss = 0.226731 (* 1 = 0.226731 loss)
I0428 18:21:30.208176  8468 sgd_solver.cpp:105] Iteration 8820, lr = 0.00174276
I0428 18:21:35.725581  8468 solver.cpp:218] Iteration 8832 (2.17503 iter/s, 5.51718s/12 iters), loss = 0.0588538
I0428 18:21:35.725632  8468 solver.cpp:237]     Train net output #0: loss = 0.0588539 (* 1 = 0.0588539 loss)
I0428 18:21:35.725646  8468 sgd_solver.cpp:105] Iteration 8832, lr = 0.00173862
I0428 18:21:41.299705  8468 solver.cpp:218] Iteration 8844 (2.15292 iter/s, 5.57383s/12 iters), loss = 0.0537387
I0428 18:21:41.299748  8468 solver.cpp:237]     Train net output #0: loss = 0.0537388 (* 1 = 0.0537388 loss)
I0428 18:21:41.299759  8468 sgd_solver.cpp:105] Iteration 8844, lr = 0.00173449
I0428 18:21:46.995190  8468 solver.cpp:218] Iteration 8856 (2.10704 iter/s, 5.69519s/12 iters), loss = 0.0624841
I0428 18:21:47.005407  8468 solver.cpp:237]     Train net output #0: loss = 0.0624842 (* 1 = 0.0624842 loss)
I0428 18:21:47.005422  8468 sgd_solver.cpp:105] Iteration 8856, lr = 0.00173037
I0428 18:21:52.531540  8468 solver.cpp:218] Iteration 8868 (2.17159 iter/s, 5.52591s/12 iters), loss = 0.131487
I0428 18:21:52.531589  8468 solver.cpp:237]     Train net output #0: loss = 0.131487 (* 1 = 0.131487 loss)
I0428 18:21:52.531600  8468 sgd_solver.cpp:105] Iteration 8868, lr = 0.00172626
I0428 18:21:54.859486  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8874.caffemodel
I0428 18:21:56.354972  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8874.solverstate
I0428 18:21:56.889910  8468 solver.cpp:330] Iteration 8874, Testing net (#0)
I0428 18:21:56.889930  8468 net.cpp:676] Ignoring source layer train-data
I0428 18:21:57.996836  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:22:02.258539  8468 solver.cpp:397]     Test net output #0: accuracy = 0.461397
I0428 18:22:02.258579  8468 solver.cpp:397]     Test net output #1: loss = 3.66665 (* 1 = 3.66665 loss)
I0428 18:22:04.360944  8468 solver.cpp:218] Iteration 8880 (1.01447 iter/s, 11.8289s/12 iters), loss = 0.0371151
I0428 18:22:04.360998  8468 solver.cpp:237]     Train net output #0: loss = 0.0371152 (* 1 = 0.0371152 loss)
I0428 18:22:04.361011  8468 sgd_solver.cpp:105] Iteration 8880, lr = 0.00172217
I0428 18:22:10.130014  8468 solver.cpp:218] Iteration 8892 (2.08017 iter/s, 5.76877s/12 iters), loss = 0.0241191
I0428 18:22:10.130065  8468 solver.cpp:237]     Train net output #0: loss = 0.0241192 (* 1 = 0.0241192 loss)
I0428 18:22:10.130079  8468 sgd_solver.cpp:105] Iteration 8892, lr = 0.00171808
I0428 18:22:14.097849  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:22:15.794857  8468 solver.cpp:218] Iteration 8904 (2.11843 iter/s, 5.66457s/12 iters), loss = 0.172889
I0428 18:22:15.794898  8468 solver.cpp:237]     Train net output #0: loss = 0.172889 (* 1 = 0.172889 loss)
I0428 18:22:15.794906  8468 sgd_solver.cpp:105] Iteration 8904, lr = 0.001714
I0428 18:22:21.539752  8468 solver.cpp:218] Iteration 8916 (2.08891 iter/s, 5.74463s/12 iters), loss = 0.0785297
I0428 18:22:21.539929  8468 solver.cpp:237]     Train net output #0: loss = 0.0785298 (* 1 = 0.0785298 loss)
I0428 18:22:21.539942  8468 sgd_solver.cpp:105] Iteration 8916, lr = 0.00170993
I0428 18:22:27.623243  8468 solver.cpp:218] Iteration 8928 (1.97269 iter/s, 6.08308s/12 iters), loss = 0.0619127
I0428 18:22:27.623299  8468 solver.cpp:237]     Train net output #0: loss = 0.0619128 (* 1 = 0.0619128 loss)
I0428 18:22:27.623314  8468 sgd_solver.cpp:105] Iteration 8928, lr = 0.00170587
I0428 18:22:33.378705  8468 solver.cpp:218] Iteration 8940 (2.08508 iter/s, 5.75518s/12 iters), loss = 0.0987291
I0428 18:22:33.378746  8468 solver.cpp:237]     Train net output #0: loss = 0.0987292 (* 1 = 0.0987292 loss)
I0428 18:22:33.378754  8468 sgd_solver.cpp:105] Iteration 8940, lr = 0.00170182
I0428 18:22:39.386538  8468 solver.cpp:218] Iteration 8952 (1.99749 iter/s, 6.00755s/12 iters), loss = 0.101934
I0428 18:22:39.386597  8468 solver.cpp:237]     Train net output #0: loss = 0.101934 (* 1 = 0.101934 loss)
I0428 18:22:39.386611  8468 sgd_solver.cpp:105] Iteration 8952, lr = 0.00169778
I0428 18:22:45.319981  8468 solver.cpp:218] Iteration 8964 (2.02254 iter/s, 5.93315s/12 iters), loss = 0.0283864
I0428 18:22:45.320039  8468 solver.cpp:237]     Train net output #0: loss = 0.0283865 (* 1 = 0.0283865 loss)
I0428 18:22:45.320051  8468 sgd_solver.cpp:105] Iteration 8964, lr = 0.00169375
I0428 18:22:50.723948  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8976.caffemodel
I0428 18:22:51.362571  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8976.solverstate
I0428 18:22:51.805100  8468 solver.cpp:330] Iteration 8976, Testing net (#0)
I0428 18:22:51.805182  8468 net.cpp:676] Ignoring source layer train-data
I0428 18:22:52.843451  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:22:57.178902  8468 solver.cpp:397]     Test net output #0: accuracy = 0.447304
I0428 18:22:57.178941  8468 solver.cpp:397]     Test net output #1: loss = 3.79255 (* 1 = 3.79255 loss)
I0428 18:22:57.310250  8468 solver.cpp:218] Iteration 8976 (1.00085 iter/s, 11.9898s/12 iters), loss = 0.149799
I0428 18:22:57.310297  8468 solver.cpp:237]     Train net output #0: loss = 0.149799 (* 1 = 0.149799 loss)
I0428 18:22:57.310309  8468 sgd_solver.cpp:105] Iteration 8976, lr = 0.00168973
I0428 18:23:02.460325  8468 solver.cpp:218] Iteration 8988 (2.33018 iter/s, 5.14982s/12 iters), loss = 0.118435
I0428 18:23:02.460381  8468 solver.cpp:237]     Train net output #0: loss = 0.118435 (* 1 = 0.118435 loss)
I0428 18:23:02.460393  8468 sgd_solver.cpp:105] Iteration 8988, lr = 0.00168571
I0428 18:23:04.846048  8468 blocking_queue.cpp:49] Waiting for data
I0428 18:23:08.411001  8468 solver.cpp:218] Iteration 9000 (2.01668 iter/s, 5.95039s/12 iters), loss = 0.0334523
I0428 18:23:08.417047  8468 solver.cpp:237]     Train net output #0: loss = 0.0334524 (* 1 = 0.0334524 loss)
I0428 18:23:08.417062  8468 sgd_solver.cpp:105] Iteration 9000, lr = 0.00168171
I0428 18:23:09.335213  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:23:14.488185  8468 solver.cpp:218] Iteration 9012 (1.97664 iter/s, 6.0709s/12 iters), loss = 0.119307
I0428 18:23:14.488238  8468 solver.cpp:237]     Train net output #0: loss = 0.119307 (* 1 = 0.119307 loss)
I0428 18:23:14.488250  8468 sgd_solver.cpp:105] Iteration 9012, lr = 0.00167772
I0428 18:23:20.656283  8468 solver.cpp:218] Iteration 9024 (1.94559 iter/s, 6.1678s/12 iters), loss = 0.0733395
I0428 18:23:20.656340  8468 solver.cpp:237]     Train net output #0: loss = 0.0733396 (* 1 = 0.0733396 loss)
I0428 18:23:20.656352  8468 sgd_solver.cpp:105] Iteration 9024, lr = 0.00167374
I0428 18:23:26.729873  8468 solver.cpp:218] Iteration 9036 (1.97586 iter/s, 6.0733s/12 iters), loss = 0.103789
I0428 18:23:26.731585  8468 solver.cpp:237]     Train net output #0: loss = 0.103789 (* 1 = 0.103789 loss)
I0428 18:23:26.731596  8468 sgd_solver.cpp:105] Iteration 9036, lr = 0.00166976
I0428 18:23:32.561287  8468 solver.cpp:218] Iteration 9048 (2.05851 iter/s, 5.82947s/12 iters), loss = 0.0759551
I0428 18:23:32.561338  8468 solver.cpp:237]     Train net output #0: loss = 0.0759552 (* 1 = 0.0759552 loss)
I0428 18:23:32.561349  8468 sgd_solver.cpp:105] Iteration 9048, lr = 0.0016658
I0428 18:23:38.675026  8468 solver.cpp:218] Iteration 9060 (1.96289 iter/s, 6.11345s/12 iters), loss = 0.0228899
I0428 18:23:38.675071  8468 solver.cpp:237]     Train net output #0: loss = 0.02289 (* 1 = 0.02289 loss)
I0428 18:23:38.675081  8468 sgd_solver.cpp:105] Iteration 9060, lr = 0.00166184
I0428 18:23:44.379904  8468 solver.cpp:218] Iteration 9072 (2.10356 iter/s, 5.7046s/12 iters), loss = 0.109525
I0428 18:23:44.379956  8468 solver.cpp:237]     Train net output #0: loss = 0.109525 (* 1 = 0.109525 loss)
I0428 18:23:44.379968  8468 sgd_solver.cpp:105] Iteration 9072, lr = 0.0016579
I0428 18:23:47.019018  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9078.caffemodel
I0428 18:23:47.651906  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9078.solverstate
I0428 18:23:48.102721  8468 solver.cpp:330] Iteration 9078, Testing net (#0)
I0428 18:23:48.102746  8468 net.cpp:676] Ignoring source layer train-data
I0428 18:23:49.146541  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:23:53.506635  8468 solver.cpp:397]     Test net output #0: accuracy = 0.466299
I0428 18:23:53.506675  8468 solver.cpp:397]     Test net output #1: loss = 3.68386 (* 1 = 3.68386 loss)
I0428 18:23:55.704375  8468 solver.cpp:218] Iteration 9084 (1.0597 iter/s, 11.324s/12 iters), loss = 0.11829
I0428 18:23:55.704430  8468 solver.cpp:237]     Train net output #0: loss = 0.118291 (* 1 = 0.118291 loss)
I0428 18:23:55.704442  8468 sgd_solver.cpp:105] Iteration 9084, lr = 0.00165396
I0428 18:24:01.938244  8468 solver.cpp:218] Iteration 9096 (1.92506 iter/s, 6.23357s/12 iters), loss = 0.0533628
I0428 18:24:01.938354  8468 solver.cpp:237]     Train net output #0: loss = 0.0533629 (* 1 = 0.0533629 loss)
I0428 18:24:01.938364  8468 sgd_solver.cpp:105] Iteration 9096, lr = 0.00165003
I0428 18:24:05.484892  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:24:07.960232  8468 solver.cpp:218] Iteration 9108 (1.99281 iter/s, 6.02164s/12 iters), loss = 0.0209832
I0428 18:24:07.960281  8468 solver.cpp:237]     Train net output #0: loss = 0.0209833 (* 1 = 0.0209833 loss)
I0428 18:24:07.960292  8468 sgd_solver.cpp:105] Iteration 9108, lr = 0.00164612
I0428 18:24:13.859011  8468 solver.cpp:218] Iteration 9120 (2.03442 iter/s, 5.89849s/12 iters), loss = 0.0994926
I0428 18:24:13.859061  8468 solver.cpp:237]     Train net output #0: loss = 0.0994927 (* 1 = 0.0994927 loss)
I0428 18:24:13.859074  8468 sgd_solver.cpp:105] Iteration 9120, lr = 0.00164221
I0428 18:24:19.992087  8468 solver.cpp:218] Iteration 9132 (1.9567 iter/s, 6.13278s/12 iters), loss = 0.144034
I0428 18:24:19.992141  8468 solver.cpp:237]     Train net output #0: loss = 0.144034 (* 1 = 0.144034 loss)
I0428 18:24:19.992152  8468 sgd_solver.cpp:105] Iteration 9132, lr = 0.00163831
I0428 18:24:26.180371  8468 solver.cpp:218] Iteration 9144 (1.93924 iter/s, 6.18798s/12 iters), loss = 0.0779942
I0428 18:24:26.180418  8468 solver.cpp:237]     Train net output #0: loss = 0.0779943 (* 1 = 0.0779943 loss)
I0428 18:24:26.180428  8468 sgd_solver.cpp:105] Iteration 9144, lr = 0.00163442
I0428 18:24:32.059079  8468 solver.cpp:218] Iteration 9156 (2.04136 iter/s, 5.87842s/12 iters), loss = 0.103713
I0428 18:24:32.059219  8468 solver.cpp:237]     Train net output #0: loss = 0.103713 (* 1 = 0.103713 loss)
I0428 18:24:32.059231  8468 sgd_solver.cpp:105] Iteration 9156, lr = 0.00163054
I0428 18:24:38.113371  8468 solver.cpp:218] Iteration 9168 (1.98219 iter/s, 6.05391s/12 iters), loss = 0.137435
I0428 18:24:38.113413  8468 solver.cpp:237]     Train net output #0: loss = 0.137435 (* 1 = 0.137435 loss)
I0428 18:24:38.113421  8468 sgd_solver.cpp:105] Iteration 9168, lr = 0.00162667
I0428 18:24:43.565750  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9180.caffemodel
I0428 18:24:47.097218  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9180.solverstate
I0428 18:24:48.055088  8468 solver.cpp:330] Iteration 9180, Testing net (#0)
I0428 18:24:48.055114  8468 net.cpp:676] Ignoring source layer train-data
I0428 18:24:49.055020  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:24:53.531551  8468 solver.cpp:397]     Test net output #0: accuracy = 0.46201
I0428 18:24:53.531586  8468 solver.cpp:397]     Test net output #1: loss = 3.72533 (* 1 = 3.72533 loss)
I0428 18:24:53.661054  8468 solver.cpp:218] Iteration 9180 (0.771851 iter/s, 15.547s/12 iters), loss = 0.0877608
I0428 18:24:53.661105  8468 solver.cpp:237]     Train net output #0: loss = 0.087761 (* 1 = 0.087761 loss)
I0428 18:24:53.661118  8468 sgd_solver.cpp:105] Iteration 9180, lr = 0.00162281
I0428 18:24:58.513936  8468 solver.cpp:218] Iteration 9192 (2.47289 iter/s, 4.85263s/12 iters), loss = 0.0372935
I0428 18:24:58.513988  8468 solver.cpp:237]     Train net output #0: loss = 0.0372936 (* 1 = 0.0372936 loss)
I0428 18:24:58.513998  8468 sgd_solver.cpp:105] Iteration 9192, lr = 0.00161895
I0428 18:25:04.687774  8468 solver.cpp:218] Iteration 9204 (1.94378 iter/s, 6.17354s/12 iters), loss = 0.108151
I0428 18:25:04.688736  8468 solver.cpp:237]     Train net output #0: loss = 0.108151 (* 1 = 0.108151 loss)
I0428 18:25:04.688748  8468 sgd_solver.cpp:105] Iteration 9204, lr = 0.00161511
I0428 18:25:04.773365  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:25:10.509151  8468 solver.cpp:218] Iteration 9216 (2.06179 iter/s, 5.82018s/12 iters), loss = 0.0928542
I0428 18:25:10.509193  8468 solver.cpp:237]     Train net output #0: loss = 0.0928543 (* 1 = 0.0928543 loss)
I0428 18:25:10.509203  8468 sgd_solver.cpp:105] Iteration 9216, lr = 0.00161128
I0428 18:25:16.715400  8468 solver.cpp:218] Iteration 9228 (1.93363 iter/s, 6.20595s/12 iters), loss = 0.113143
I0428 18:25:16.715454  8468 solver.cpp:237]     Train net output #0: loss = 0.113144 (* 1 = 0.113144 loss)
I0428 18:25:16.715464  8468 sgd_solver.cpp:105] Iteration 9228, lr = 0.00160745
I0428 18:25:22.798393  8468 solver.cpp:218] Iteration 9240 (1.97281 iter/s, 6.0827s/12 iters), loss = 0.0523536
I0428 18:25:22.804556  8468 solver.cpp:237]     Train net output #0: loss = 0.0523538 (* 1 = 0.0523538 loss)
I0428 18:25:22.804569  8468 sgd_solver.cpp:105] Iteration 9240, lr = 0.00160363
I0428 18:25:28.607152  8468 solver.cpp:218] Iteration 9252 (2.06812 iter/s, 5.80236s/12 iters), loss = 0.077052
I0428 18:25:28.607208  8468 solver.cpp:237]     Train net output #0: loss = 0.0770521 (* 1 = 0.0770521 loss)
I0428 18:25:28.607221  8468 sgd_solver.cpp:105] Iteration 9252, lr = 0.00159983
I0428 18:25:34.546466  8468 solver.cpp:218] Iteration 9264 (2.02054 iter/s, 5.93902s/12 iters), loss = 0.0854699
I0428 18:25:34.546515  8468 solver.cpp:237]     Train net output #0: loss = 0.08547 (* 1 = 0.08547 loss)
I0428 18:25:34.546526  8468 sgd_solver.cpp:105] Iteration 9264, lr = 0.00159603
I0428 18:25:40.574338  8468 solver.cpp:218] Iteration 9276 (1.99085 iter/s, 6.02758s/12 iters), loss = 0.0331302
I0428 18:25:40.578680  8468 solver.cpp:237]     Train net output #0: loss = 0.0331303 (* 1 = 0.0331303 loss)
I0428 18:25:40.578694  8468 sgd_solver.cpp:105] Iteration 9276, lr = 0.00159224
I0428 18:25:42.857929  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9282.caffemodel
I0428 18:25:45.061125  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9282.solverstate
I0428 18:25:46.082229  8468 solver.cpp:330] Iteration 9282, Testing net (#0)
I0428 18:25:46.082250  8468 net.cpp:676] Ignoring source layer train-data
I0428 18:25:46.961756  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:25:51.574683  8468 solver.cpp:397]     Test net output #0: accuracy = 0.465074
I0428 18:25:51.574714  8468 solver.cpp:397]     Test net output #1: loss = 3.76599 (* 1 = 3.76599 loss)
I0428 18:25:53.641914  8468 solver.cpp:218] Iteration 9288 (0.918644 iter/s, 13.0627s/12 iters), loss = 0.0485559
I0428 18:25:53.641969  8468 solver.cpp:237]     Train net output #0: loss = 0.048556 (* 1 = 0.048556 loss)
I0428 18:25:53.641981  8468 sgd_solver.cpp:105] Iteration 9288, lr = 0.00158846
I0428 18:25:59.519171  8468 solver.cpp:218] Iteration 9300 (2.04187 iter/s, 5.87696s/12 iters), loss = 0.116344
I0428 18:25:59.519215  8468 solver.cpp:237]     Train net output #0: loss = 0.116344 (* 1 = 0.116344 loss)
I0428 18:25:59.519225  8468 sgd_solver.cpp:105] Iteration 9300, lr = 0.00158469
I0428 18:26:02.046128  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:26:05.393575  8468 solver.cpp:218] Iteration 9312 (2.04286 iter/s, 5.87412s/12 iters), loss = 0.0929097
I0428 18:26:05.401423  8468 solver.cpp:237]     Train net output #0: loss = 0.0929099 (* 1 = 0.0929099 loss)
I0428 18:26:05.401448  8468 sgd_solver.cpp:105] Iteration 9312, lr = 0.00158092
I0428 18:26:11.278354  8468 solver.cpp:218] Iteration 9324 (2.04196 iter/s, 5.87671s/12 iters), loss = 0.0391375
I0428 18:26:11.281658  8468 solver.cpp:237]     Train net output #0: loss = 0.0391376 (* 1 = 0.0391376 loss)
I0428 18:26:11.281674  8468 sgd_solver.cpp:105] Iteration 9324, lr = 0.00157717
I0428 18:26:17.233115  8468 solver.cpp:218] Iteration 9336 (2.01639 iter/s, 5.95122s/12 iters), loss = 0.0980633
I0428 18:26:17.233170  8468 solver.cpp:237]     Train net output #0: loss = 0.0980635 (* 1 = 0.0980635 loss)
I0428 18:26:17.233180  8468 sgd_solver.cpp:105] Iteration 9336, lr = 0.00157343
I0428 18:26:23.359448  8468 solver.cpp:218] Iteration 9348 (1.95886 iter/s, 6.12603s/12 iters), loss = 0.0444608
I0428 18:26:23.359500  8468 solver.cpp:237]     Train net output #0: loss = 0.0444609 (* 1 = 0.0444609 loss)
I0428 18:26:23.359513  8468 sgd_solver.cpp:105] Iteration 9348, lr = 0.00156969
I0428 18:26:29.333783  8468 solver.cpp:218] Iteration 9360 (2.00869 iter/s, 5.97404s/12 iters), loss = 0.0857369
I0428 18:26:29.333838  8468 solver.cpp:237]     Train net output #0: loss = 0.085737 (* 1 = 0.085737 loss)
I0428 18:26:29.333849  8468 sgd_solver.cpp:105] Iteration 9360, lr = 0.00156596
I0428 18:26:35.291177  8468 solver.cpp:218] Iteration 9372 (2.0144 iter/s, 5.95709s/12 iters), loss = 0.0822514
I0428 18:26:35.291220  8468 solver.cpp:237]     Train net output #0: loss = 0.0822516 (* 1 = 0.0822516 loss)
I0428 18:26:35.291229  8468 sgd_solver.cpp:105] Iteration 9372, lr = 0.00156225
I0428 18:26:40.427073  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9384.caffemodel
I0428 18:26:41.044888  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9384.solverstate
I0428 18:26:41.468546  8468 solver.cpp:330] Iteration 9384, Testing net (#0)
I0428 18:26:41.468617  8468 net.cpp:676] Ignoring source layer train-data
I0428 18:26:42.362437  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:26:46.721096  8468 solver.cpp:397]     Test net output #0: accuracy = 0.461397
I0428 18:26:46.721124  8468 solver.cpp:397]     Test net output #1: loss = 3.68429 (* 1 = 3.68429 loss)
I0428 18:26:46.847702  8468 solver.cpp:218] Iteration 9384 (1.03842 iter/s, 11.556s/12 iters), loss = 0.0923721
I0428 18:26:46.847764  8468 solver.cpp:237]     Train net output #0: loss = 0.0923722 (* 1 = 0.0923722 loss)
I0428 18:26:46.847779  8468 sgd_solver.cpp:105] Iteration 9384, lr = 0.00155854
I0428 18:26:51.902545  8468 solver.cpp:218] Iteration 9396 (2.37409 iter/s, 5.05458s/12 iters), loss = 0.08998
I0428 18:26:51.902585  8468 solver.cpp:237]     Train net output #0: loss = 0.0899801 (* 1 = 0.0899801 loss)
I0428 18:26:51.902593  8468 sgd_solver.cpp:105] Iteration 9396, lr = 0.00155484
I0428 18:26:57.039300  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:26:57.848752  8468 solver.cpp:218] Iteration 9408 (2.01819 iter/s, 5.94592s/12 iters), loss = 0.117764
I0428 18:26:57.848798  8468 solver.cpp:237]     Train net output #0: loss = 0.117764 (* 1 = 0.117764 loss)
I0428 18:26:57.848805  8468 sgd_solver.cpp:105] Iteration 9408, lr = 0.00155114
I0428 18:27:03.832065  8468 solver.cpp:218] Iteration 9420 (2.00567 iter/s, 5.98302s/12 iters), loss = 0.0203846
I0428 18:27:03.832106  8468 solver.cpp:237]     Train net output #0: loss = 0.0203848 (* 1 = 0.0203848 loss)
I0428 18:27:03.832116  8468 sgd_solver.cpp:105] Iteration 9420, lr = 0.00154746
I0428 18:27:09.934024  8468 solver.cpp:218] Iteration 9432 (1.96668 iter/s, 6.10167s/12 iters), loss = 0.102909
I0428 18:27:09.934067  8468 solver.cpp:237]     Train net output #0: loss = 0.102909 (* 1 = 0.102909 loss)
I0428 18:27:09.934075  8468 sgd_solver.cpp:105] Iteration 9432, lr = 0.00154379
I0428 18:27:15.799728  8468 solver.cpp:218] Iteration 9444 (2.04589 iter/s, 5.86542s/12 iters), loss = 0.0455787
I0428 18:27:15.799913  8468 solver.cpp:237]     Train net output #0: loss = 0.0455788 (* 1 = 0.0455788 loss)
I0428 18:27:15.799927  8468 sgd_solver.cpp:105] Iteration 9444, lr = 0.00154012
I0428 18:27:21.855486  8468 solver.cpp:218] Iteration 9456 (1.98172 iter/s, 6.05533s/12 iters), loss = 0.0882428
I0428 18:27:21.855538  8468 solver.cpp:237]     Train net output #0: loss = 0.0882429 (* 1 = 0.0882429 loss)
I0428 18:27:21.855551  8468 sgd_solver.cpp:105] Iteration 9456, lr = 0.00153647
I0428 18:27:27.329180  8468 solver.cpp:218] Iteration 9468 (2.19241 iter/s, 5.47342s/12 iters), loss = 0.0339181
I0428 18:27:27.329231  8468 solver.cpp:237]     Train net output #0: loss = 0.0339182 (* 1 = 0.0339182 loss)
I0428 18:27:27.329241  8468 sgd_solver.cpp:105] Iteration 9468, lr = 0.00153282
I0428 18:27:32.899821  8468 solver.cpp:218] Iteration 9480 (2.15426 iter/s, 5.57036s/12 iters), loss = 0.106641
I0428 18:27:32.899863  8468 solver.cpp:237]     Train net output #0: loss = 0.106641 (* 1 = 0.106641 loss)
I0428 18:27:32.899879  8468 sgd_solver.cpp:105] Iteration 9480, lr = 0.00152918
I0428 18:27:35.188558  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9486.caffemodel
I0428 18:27:36.480132  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9486.solverstate
I0428 18:27:38.172428  8468 solver.cpp:330] Iteration 9486, Testing net (#0)
I0428 18:27:38.172449  8468 net.cpp:676] Ignoring source layer train-data
I0428 18:27:38.904312  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:27:43.295074  8468 solver.cpp:397]     Test net output #0: accuracy = 0.466299
I0428 18:27:43.295106  8468 solver.cpp:397]     Test net output #1: loss = 3.5948 (* 1 = 3.5948 loss)
I0428 18:27:45.348790  8468 solver.cpp:218] Iteration 9492 (0.963976 iter/s, 12.4484s/12 iters), loss = 0.0633654
I0428 18:27:45.348845  8468 solver.cpp:237]     Train net output #0: loss = 0.0633655 (* 1 = 0.0633655 loss)
I0428 18:27:45.348860  8468 sgd_solver.cpp:105] Iteration 9492, lr = 0.00152555
I0428 18:27:51.085269  8468 solver.cpp:218] Iteration 9504 (2.09198 iter/s, 5.73619s/12 iters), loss = 0.0543767
I0428 18:27:51.100231  8468 solver.cpp:237]     Train net output #0: loss = 0.0543768 (* 1 = 0.0543768 loss)
I0428 18:27:51.100244  8468 sgd_solver.cpp:105] Iteration 9504, lr = 0.00152193
I0428 18:27:52.890158  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:27:57.016309  8468 solver.cpp:218] Iteration 9516 (2.02845 iter/s, 5.91584s/12 iters), loss = 0.101924
I0428 18:27:57.016369  8468 solver.cpp:237]     Train net output #0: loss = 0.101924 (* 1 = 0.101924 loss)
I0428 18:27:57.016381  8468 sgd_solver.cpp:105] Iteration 9516, lr = 0.00151831
I0428 18:28:02.735494  8468 solver.cpp:218] Iteration 9528 (2.09831 iter/s, 5.71889s/12 iters), loss = 0.0670236
I0428 18:28:02.735555  8468 solver.cpp:237]     Train net output #0: loss = 0.0670238 (* 1 = 0.0670238 loss)
I0428 18:28:02.735566  8468 sgd_solver.cpp:105] Iteration 9528, lr = 0.00151471
I0428 18:28:08.539294  8468 solver.cpp:218] Iteration 9540 (2.06772 iter/s, 5.8035s/12 iters), loss = 0.0660224
I0428 18:28:08.539336  8468 solver.cpp:237]     Train net output #0: loss = 0.0660226 (* 1 = 0.0660226 loss)
I0428 18:28:08.539346  8468 sgd_solver.cpp:105] Iteration 9540, lr = 0.00151111
I0428 18:28:14.528988  8468 solver.cpp:218] Iteration 9552 (2.00354 iter/s, 5.9894s/12 iters), loss = 0.0925418
I0428 18:28:14.529055  8468 solver.cpp:237]     Train net output #0: loss = 0.092542 (* 1 = 0.092542 loss)
I0428 18:28:14.529071  8468 sgd_solver.cpp:105] Iteration 9552, lr = 0.00150752
I0428 18:28:20.310144  8468 solver.cpp:218] Iteration 9564 (2.07582 iter/s, 5.78085s/12 iters), loss = 0.0853169
I0428 18:28:20.310201  8468 solver.cpp:237]     Train net output #0: loss = 0.0853171 (* 1 = 0.0853171 loss)
I0428 18:28:20.310214  8468 sgd_solver.cpp:105] Iteration 9564, lr = 0.00150395
I0428 18:28:26.288060  8468 solver.cpp:218] Iteration 9576 (2.00749 iter/s, 5.97762s/12 iters), loss = 0.0465749
I0428 18:28:26.288201  8468 solver.cpp:237]     Train net output #0: loss = 0.0465751 (* 1 = 0.0465751 loss)
I0428 18:28:26.288213  8468 sgd_solver.cpp:105] Iteration 9576, lr = 0.00150037
I0428 18:28:31.209264  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9588.caffemodel
I0428 18:28:31.847236  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9588.solverstate
I0428 18:28:32.292920  8468 solver.cpp:330] Iteration 9588, Testing net (#0)
I0428 18:28:32.292940  8468 net.cpp:676] Ignoring source layer train-data
I0428 18:28:33.004135  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:28:37.506085  8468 solver.cpp:397]     Test net output #0: accuracy = 0.469363
I0428 18:28:37.506125  8468 solver.cpp:397]     Test net output #1: loss = 3.71066 (* 1 = 3.71066 loss)
I0428 18:28:37.633163  8468 solver.cpp:218] Iteration 9588 (1.05778 iter/s, 11.3445s/12 iters), loss = 0.0226783
I0428 18:28:37.633203  8468 solver.cpp:237]     Train net output #0: loss = 0.0226785 (* 1 = 0.0226785 loss)
I0428 18:28:37.633214  8468 sgd_solver.cpp:105] Iteration 9588, lr = 0.00149681
I0428 18:28:42.253960  8468 solver.cpp:218] Iteration 9600 (2.59709 iter/s, 4.62056s/12 iters), loss = 0.0405999
I0428 18:28:42.254022  8468 solver.cpp:237]     Train net output #0: loss = 0.0406001 (* 1 = 0.0406001 loss)
I0428 18:28:42.254034  8468 sgd_solver.cpp:105] Iteration 9600, lr = 0.00149326
I0428 18:28:46.178720  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:28:47.853448  8468 solver.cpp:218] Iteration 9612 (2.14316 iter/s, 5.5992s/12 iters), loss = 0.072067
I0428 18:28:47.853497  8468 solver.cpp:237]     Train net output #0: loss = 0.0720672 (* 1 = 0.0720672 loss)
I0428 18:28:47.853508  8468 sgd_solver.cpp:105] Iteration 9612, lr = 0.00148971
I0428 18:28:53.583539  8468 solver.cpp:218] Iteration 9624 (2.09431 iter/s, 5.7298s/12 iters), loss = 0.0773489
I0428 18:28:53.583580  8468 solver.cpp:237]     Train net output #0: loss = 0.077349 (* 1 = 0.077349 loss)
I0428 18:28:53.583588  8468 sgd_solver.cpp:105] Iteration 9624, lr = 0.00148618
I0428 18:28:59.183306  8468 solver.cpp:218] Iteration 9636 (2.14305 iter/s, 5.5995s/12 iters), loss = 0.117475
I0428 18:28:59.183408  8468 solver.cpp:237]     Train net output #0: loss = 0.117475 (* 1 = 0.117475 loss)
I0428 18:28:59.183418  8468 sgd_solver.cpp:105] Iteration 9636, lr = 0.00148265
I0428 18:29:04.842689  8468 solver.cpp:218] Iteration 9648 (2.1205 iter/s, 5.65905s/12 iters), loss = 0.124064
I0428 18:29:04.842758  8468 solver.cpp:237]     Train net output #0: loss = 0.124064 (* 1 = 0.124064 loss)
I0428 18:29:04.842770  8468 sgd_solver.cpp:105] Iteration 9648, lr = 0.00147913
I0428 18:29:10.539288  8468 solver.cpp:218] Iteration 9660 (2.10663 iter/s, 5.6963s/12 iters), loss = 0.129108
I0428 18:29:10.539330  8468 solver.cpp:237]     Train net output #0: loss = 0.129109 (* 1 = 0.129109 loss)
I0428 18:29:10.539341  8468 sgd_solver.cpp:105] Iteration 9660, lr = 0.00147562
I0428 18:29:16.170012  8468 solver.cpp:218] Iteration 9672 (2.13127 iter/s, 5.63045s/12 iters), loss = 0.0510295
I0428 18:29:16.170053  8468 solver.cpp:237]     Train net output #0: loss = 0.0510296 (* 1 = 0.0510296 loss)
I0428 18:29:16.170061  8468 sgd_solver.cpp:105] Iteration 9672, lr = 0.00147211
I0428 18:29:21.788218  8468 solver.cpp:218] Iteration 9684 (2.13602 iter/s, 5.61794s/12 iters), loss = 0.0375033
I0428 18:29:21.788259  8468 solver.cpp:237]     Train net output #0: loss = 0.0375034 (* 1 = 0.0375034 loss)
I0428 18:29:21.788267  8468 sgd_solver.cpp:105] Iteration 9684, lr = 0.00146862
I0428 18:29:24.047195  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9690.caffemodel
I0428 18:29:25.491814  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9690.solverstate
I0428 18:29:26.486366  8468 solver.cpp:330] Iteration 9690, Testing net (#0)
I0428 18:29:26.486384  8468 net.cpp:676] Ignoring source layer train-data
I0428 18:29:27.053489  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:29:29.742089  8468 blocking_queue.cpp:49] Waiting for data
I0428 18:29:31.063481  8468 solver.cpp:397]     Test net output #0: accuracy = 0.449142
I0428 18:29:31.063516  8468 solver.cpp:397]     Test net output #1: loss = 3.82165 (* 1 = 3.82165 loss)
I0428 18:29:32.943084  8468 solver.cpp:218] Iteration 9696 (1.07581 iter/s, 11.1544s/12 iters), loss = 0.0556669
I0428 18:29:32.943125  8468 solver.cpp:237]     Train net output #0: loss = 0.055667 (* 1 = 0.055667 loss)
I0428 18:29:32.943135  8468 sgd_solver.cpp:105] Iteration 9696, lr = 0.00146513
I0428 18:29:38.514267  8468 solver.cpp:218] Iteration 9708 (2.15405 iter/s, 5.57091s/12 iters), loss = 0.0218287
I0428 18:29:38.514309  8468 solver.cpp:237]     Train net output #0: loss = 0.0218289 (* 1 = 0.0218289 loss)
I0428 18:29:38.514319  8468 sgd_solver.cpp:105] Iteration 9708, lr = 0.00146165
I0428 18:29:39.394882  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:29:44.114703  8468 solver.cpp:218] Iteration 9720 (2.14279 iter/s, 5.60016s/12 iters), loss = 0.0111182
I0428 18:29:44.114763  8468 solver.cpp:237]     Train net output #0: loss = 0.0111184 (* 1 = 0.0111184 loss)
I0428 18:29:44.114773  8468 sgd_solver.cpp:105] Iteration 9720, lr = 0.00145818
I0428 18:29:49.552269  8468 solver.cpp:218] Iteration 9732 (2.20699 iter/s, 5.43728s/12 iters), loss = 0.0321503
I0428 18:29:49.552320  8468 solver.cpp:237]     Train net output #0: loss = 0.0321505 (* 1 = 0.0321505 loss)
I0428 18:29:49.552331  8468 sgd_solver.cpp:105] Iteration 9732, lr = 0.00145472
I0428 18:29:55.152994  8468 solver.cpp:218] Iteration 9744 (2.14269 iter/s, 5.60045s/12 iters), loss = 0.0187347
I0428 18:29:55.153031  8468 solver.cpp:237]     Train net output #0: loss = 0.0187348 (* 1 = 0.0187348 loss)
I0428 18:29:55.153041  8468 sgd_solver.cpp:105] Iteration 9744, lr = 0.00145127
I0428 18:30:00.650941  8468 solver.cpp:218] Iteration 9756 (2.18274 iter/s, 5.49768s/12 iters), loss = 0.0452748
I0428 18:30:00.651062  8468 solver.cpp:237]     Train net output #0: loss = 0.0452749 (* 1 = 0.0452749 loss)
I0428 18:30:00.651075  8468 sgd_solver.cpp:105] Iteration 9756, lr = 0.00144782
I0428 18:30:06.403512  8468 solver.cpp:218] Iteration 9768 (2.08615 iter/s, 5.75221s/12 iters), loss = 0.129603
I0428 18:30:06.403568  8468 solver.cpp:237]     Train net output #0: loss = 0.129603 (* 1 = 0.129603 loss)
I0428 18:30:06.403581  8468 sgd_solver.cpp:105] Iteration 9768, lr = 0.00144438
I0428 18:30:11.878780  8468 solver.cpp:218] Iteration 9780 (2.19179 iter/s, 5.47498s/12 iters), loss = 0.0362101
I0428 18:30:11.885059  8468 solver.cpp:237]     Train net output #0: loss = 0.0362103 (* 1 = 0.0362103 loss)
I0428 18:30:11.885083  8468 sgd_solver.cpp:105] Iteration 9780, lr = 0.00144095
I0428 18:30:16.832168  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9792.caffemodel
I0428 18:30:18.012881  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9792.solverstate
I0428 18:30:21.461824  8468 solver.cpp:330] Iteration 9792, Testing net (#0)
I0428 18:30:21.461845  8468 net.cpp:676] Ignoring source layer train-data
I0428 18:30:22.102672  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:30:26.278259  8468 solver.cpp:397]     Test net output #0: accuracy = 0.463848
I0428 18:30:26.278292  8468 solver.cpp:397]     Test net output #1: loss = 3.87764 (* 1 = 3.87764 loss)
I0428 18:30:26.407627  8468 solver.cpp:218] Iteration 9792 (0.826332 iter/s, 14.522s/12 iters), loss = 0.0428353
I0428 18:30:26.407672  8468 solver.cpp:237]     Train net output #0: loss = 0.0428355 (* 1 = 0.0428355 loss)
I0428 18:30:26.407682  8468 sgd_solver.cpp:105] Iteration 9792, lr = 0.00143753
I0428 18:30:30.878592  8468 solver.cpp:218] Iteration 9804 (2.68413 iter/s, 4.47072s/12 iters), loss = 0.104679
I0428 18:30:30.878756  8468 solver.cpp:237]     Train net output #0: loss = 0.10468 (* 1 = 0.10468 loss)
I0428 18:30:30.878768  8468 sgd_solver.cpp:105] Iteration 9804, lr = 0.00143412
I0428 18:30:34.028630  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:30:36.236600  8468 solver.cpp:218] Iteration 9816 (2.2398 iter/s, 5.35762s/12 iters), loss = 0.0520696
I0428 18:30:36.236660  8468 solver.cpp:237]     Train net output #0: loss = 0.0520697 (* 1 = 0.0520697 loss)
I0428 18:30:36.236673  8468 sgd_solver.cpp:105] Iteration 9816, lr = 0.00143072
I0428 18:30:41.691365  8468 solver.cpp:218] Iteration 9828 (2.20002 iter/s, 5.45449s/12 iters), loss = 0.0489511
I0428 18:30:41.691404  8468 solver.cpp:237]     Train net output #0: loss = 0.0489513 (* 1 = 0.0489513 loss)
I0428 18:30:41.691412  8468 sgd_solver.cpp:105] Iteration 9828, lr = 0.00142732
I0428 18:30:47.104395  8468 solver.cpp:218] Iteration 9840 (2.21698 iter/s, 5.41277s/12 iters), loss = 0.0663962
I0428 18:30:47.104449  8468 solver.cpp:237]     Train net output #0: loss = 0.0663963 (* 1 = 0.0663963 loss)
I0428 18:30:47.104462  8468 sgd_solver.cpp:105] Iteration 9840, lr = 0.00142393
I0428 18:30:52.611848  8468 solver.cpp:218] Iteration 9852 (2.17897 iter/s, 5.50718s/12 iters), loss = 0.0309977
I0428 18:30:52.611888  8468 solver.cpp:237]     Train net output #0: loss = 0.0309978 (* 1 = 0.0309978 loss)
I0428 18:30:52.611897  8468 sgd_solver.cpp:105] Iteration 9852, lr = 0.00142055
I0428 18:30:58.106540  8468 solver.cpp:218] Iteration 9864 (2.18403 iter/s, 5.49442s/12 iters), loss = 0.0820104
I0428 18:30:58.106598  8468 solver.cpp:237]     Train net output #0: loss = 0.0820105 (* 1 = 0.0820105 loss)
I0428 18:30:58.106611  8468 sgd_solver.cpp:105] Iteration 9864, lr = 0.00141718
I0428 18:31:03.416539  8468 solver.cpp:218] Iteration 9876 (2.26001 iter/s, 5.3097s/12 iters), loss = 0.0625292
I0428 18:31:03.416685  8468 solver.cpp:237]     Train net output #0: loss = 0.0625293 (* 1 = 0.0625293 loss)
I0428 18:31:03.416698  8468 sgd_solver.cpp:105] Iteration 9876, lr = 0.00141381
I0428 18:31:08.832726  8468 solver.cpp:218] Iteration 9888 (2.21573 iter/s, 5.41582s/12 iters), loss = 0.0810568
I0428 18:31:08.832767  8468 solver.cpp:237]     Train net output #0: loss = 0.081057 (* 1 = 0.081057 loss)
I0428 18:31:08.832777  8468 sgd_solver.cpp:105] Iteration 9888, lr = 0.00141045
I0428 18:31:11.016033  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9894.caffemodel
I0428 18:31:12.576824  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9894.solverstate
I0428 18:31:13.474345  8468 solver.cpp:330] Iteration 9894, Testing net (#0)
I0428 18:31:13.474370  8468 net.cpp:676] Ignoring source layer train-data
I0428 18:31:14.142401  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:31:18.292364  8468 solver.cpp:397]     Test net output #0: accuracy = 0.460784
I0428 18:31:18.292402  8468 solver.cpp:397]     Test net output #1: loss = 3.80192 (* 1 = 3.80192 loss)
I0428 18:31:20.345515  8468 solver.cpp:218] Iteration 9900 (1.04236 iter/s, 11.5123s/12 iters), loss = 0.121101
I0428 18:31:20.345556  8468 solver.cpp:237]     Train net output #0: loss = 0.121101 (* 1 = 0.121101 loss)
I0428 18:31:20.345566  8468 sgd_solver.cpp:105] Iteration 9900, lr = 0.00140711
I0428 18:31:25.789103  8468 solver.cpp:218] Iteration 9912 (2.20454 iter/s, 5.44332s/12 iters), loss = 0.0633116
I0428 18:31:25.789147  8468 solver.cpp:237]     Train net output #0: loss = 0.0633117 (* 1 = 0.0633117 loss)
I0428 18:31:25.789155  8468 sgd_solver.cpp:105] Iteration 9912, lr = 0.00140377
I0428 18:31:25.890193  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:31:31.231470  8468 solver.cpp:218] Iteration 9924 (2.20503 iter/s, 5.44209s/12 iters), loss = 0.0390234
I0428 18:31:31.231523  8468 solver.cpp:237]     Train net output #0: loss = 0.0390235 (* 1 = 0.0390235 loss)
I0428 18:31:31.231536  8468 sgd_solver.cpp:105] Iteration 9924, lr = 0.00140043
I0428 18:31:36.637235  8468 solver.cpp:218] Iteration 9936 (2.21997 iter/s, 5.40548s/12 iters), loss = 0.0799755
I0428 18:31:36.637394  8468 solver.cpp:237]     Train net output #0: loss = 0.0799756 (* 1 = 0.0799756 loss)
I0428 18:31:36.637408  8468 sgd_solver.cpp:105] Iteration 9936, lr = 0.00139711
I0428 18:31:42.133646  8468 solver.cpp:218] Iteration 9948 (2.18339 iter/s, 5.49603s/12 iters), loss = 0.213487
I0428 18:31:42.133682  8468 solver.cpp:237]     Train net output #0: loss = 0.213487 (* 1 = 0.213487 loss)
I0428 18:31:42.133693  8468 sgd_solver.cpp:105] Iteration 9948, lr = 0.00139379
I0428 18:31:47.655464  8468 solver.cpp:218] Iteration 9960 (2.1733 iter/s, 5.52156s/12 iters), loss = 0.0804947
I0428 18:31:47.655501  8468 solver.cpp:237]     Train net output #0: loss = 0.0804949 (* 1 = 0.0804949 loss)
I0428 18:31:47.655510  8468 sgd_solver.cpp:105] Iteration 9960, lr = 0.00139048
I0428 18:31:53.130744  8468 solver.cpp:218] Iteration 9972 (2.19177 iter/s, 5.47502s/12 iters), loss = 0.0882462
I0428 18:31:53.130784  8468 solver.cpp:237]     Train net output #0: loss = 0.0882464 (* 1 = 0.0882464 loss)
I0428 18:31:53.130792  8468 sgd_solver.cpp:105] Iteration 9972, lr = 0.00138718
I0428 18:31:58.712998  8468 solver.cpp:218] Iteration 9984 (2.14978 iter/s, 5.58198s/12 iters), loss = 0.0381347
I0428 18:31:58.713057  8468 solver.cpp:237]     Train net output #0: loss = 0.0381349 (* 1 = 0.0381349 loss)
I0428 18:31:58.713070  8468 sgd_solver.cpp:105] Iteration 9984, lr = 0.00138389
I0428 18:32:03.622781  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9996.caffemodel
I0428 18:32:04.206394  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9996.solverstate
I0428 18:32:04.647881  8468 solver.cpp:330] Iteration 9996, Testing net (#0)
I0428 18:32:04.647899  8468 net.cpp:676] Ignoring source layer train-data
I0428 18:32:05.145151  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:32:09.697656  8468 solver.cpp:397]     Test net output #0: accuracy = 0.470588
I0428 18:32:09.697803  8468 solver.cpp:397]     Test net output #1: loss = 3.76158 (* 1 = 3.76158 loss)
I0428 18:32:09.829000  8468 solver.cpp:218] Iteration 9996 (1.07957 iter/s, 11.1155s/12 iters), loss = 0.0717229
I0428 18:32:09.829051  8468 solver.cpp:237]     Train net output #0: loss = 0.0717231 (* 1 = 0.0717231 loss)
I0428 18:32:09.829061  8468 sgd_solver.cpp:105] Iteration 9996, lr = 0.0013806
I0428 18:32:14.588714  8468 solver.cpp:218] Iteration 10008 (2.5213 iter/s, 4.75946s/12 iters), loss = 0.0903562
I0428 18:32:14.588771  8468 solver.cpp:237]     Train net output #0: loss = 0.0903563 (* 1 = 0.0903563 loss)
I0428 18:32:14.588784  8468 sgd_solver.cpp:105] Iteration 10008, lr = 0.00137732
I0428 18:32:17.049795  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:32:20.093134  8468 solver.cpp:218] Iteration 10020 (2.18018 iter/s, 5.50414s/12 iters), loss = 0.0920812
I0428 18:32:20.093174  8468 solver.cpp:237]     Train net output #0: loss = 0.0920813 (* 1 = 0.0920813 loss)
I0428 18:32:20.093183  8468 sgd_solver.cpp:105] Iteration 10020, lr = 0.00137405
I0428 18:32:25.404541  8468 solver.cpp:218] Iteration 10032 (2.2594 iter/s, 5.31114s/12 iters), loss = 0.0323462
I0428 18:32:25.404598  8468 solver.cpp:237]     Train net output #0: loss = 0.0323463 (* 1 = 0.0323463 loss)
I0428 18:32:25.404613  8468 sgd_solver.cpp:105] Iteration 10032, lr = 0.00137079
I0428 18:32:31.214355  8468 solver.cpp:218] Iteration 10044 (2.06557 iter/s, 5.80952s/12 iters), loss = 0.0546715
I0428 18:32:31.214397  8468 solver.cpp:237]     Train net output #0: loss = 0.0546716 (* 1 = 0.0546716 loss)
I0428 18:32:31.214406  8468 sgd_solver.cpp:105] Iteration 10044, lr = 0.00136754
I0428 18:32:36.645622  8468 solver.cpp:218] Iteration 10056 (2.20954 iter/s, 5.43099s/12 iters), loss = 0.0425496
I0428 18:32:36.645663  8468 solver.cpp:237]     Train net output #0: loss = 0.0425497 (* 1 = 0.0425497 loss)
I0428 18:32:36.645673  8468 sgd_solver.cpp:105] Iteration 10056, lr = 0.00136429
I0428 18:32:42.191550  8468 solver.cpp:218] Iteration 10068 (2.16386 iter/s, 5.54566s/12 iters), loss = 0.0655215
I0428 18:32:42.191677  8468 solver.cpp:237]     Train net output #0: loss = 0.0655216 (* 1 = 0.0655216 loss)
I0428 18:32:42.191687  8468 sgd_solver.cpp:105] Iteration 10068, lr = 0.00136105
I0428 18:32:47.681375  8468 solver.cpp:218] Iteration 10080 (2.186 iter/s, 5.48947s/12 iters), loss = 0.032362
I0428 18:32:47.681430  8468 solver.cpp:237]     Train net output #0: loss = 0.0323621 (* 1 = 0.0323621 loss)
I0428 18:32:47.681444  8468 sgd_solver.cpp:105] Iteration 10080, lr = 0.00135782
I0428 18:32:53.041868  8468 solver.cpp:218] Iteration 10092 (2.23872 iter/s, 5.36022s/12 iters), loss = 0.0189454
I0428 18:32:53.041909  8468 solver.cpp:237]     Train net output #0: loss = 0.0189455 (* 1 = 0.0189455 loss)
I0428 18:32:53.041918  8468 sgd_solver.cpp:105] Iteration 10092, lr = 0.0013546
I0428 18:32:55.250614  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_10098.caffemodel
I0428 18:32:55.868134  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_10098.solverstate
I0428 18:32:57.884989  8468 solver.cpp:330] Iteration 10098, Testing net (#0)
I0428 18:32:57.885007  8468 net.cpp:676] Ignoring source layer train-data
I0428 18:32:58.365599  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:33:02.582368  8468 solver.cpp:397]     Test net output #0: accuracy = 0.465686
I0428 18:33:02.582396  8468 solver.cpp:397]     Test net output #1: loss = 3.80638 (* 1 = 3.80638 loss)
I0428 18:33:04.535938  8468 solver.cpp:218] Iteration 10104 (1.04406 iter/s, 11.4936s/12 iters), loss = 0.0133017
I0428 18:33:04.535977  8468 solver.cpp:237]     Train net output #0: loss = 0.0133018 (* 1 = 0.0133018 loss)
I0428 18:33:04.535985  8468 sgd_solver.cpp:105] Iteration 10104, lr = 0.00135138
I0428 18:33:09.369113  8481 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:33:10.061985  8468 solver.cpp:218] Iteration 10116 (2.17164 iter/s, 5.52577s/12 iters), loss = 0.105568
I0428 18:33:10.062041  8468 solver.cpp:237]     Train net output #0: loss = 0.105568 (* 1 = 0.105568 loss)
I0428 18:33:10.062055  8468 sgd_solver.cpp:105] Iteration 10116, lr = 0.00134817
I0428 18:33:15.418115  8468 solver.cpp:218] Iteration 10128 (2.24054 iter/s, 5.35585s/12 iters), loss = 0.0647885
I0428 18:33:15.418248  8468 solver.cpp:237]     Train net output #0: loss = 0.0647886 (* 1 = 0.0647886 loss)
I0428 18:33:15.418262  8468 sgd_solver.cpp:105] Iteration 10128, lr = 0.00134497
I0428 18:33:20.906349  8468 solver.cpp:218] Iteration 10140 (2.18664 iter/s, 5.48787s/12 iters), loss = 0.0230519
I0428 18:33:20.906399  8468 solver.cpp:237]     Train net output #0: loss = 0.023052 (* 1 = 0.023052 loss)
I0428 18:33:20.906410  8468 sgd_solver.cpp:105] Iteration 10140, lr = 0.00134178
I0428 18:33:26.328599  8468 solver.cpp:218] Iteration 10152 (2.21322 iter/s, 5.42198s/12 iters), loss = 0.0217821
I0428 18:33:26.328639  8468 solver.cpp:237]     Train net output #0: loss = 0.0217822 (* 1 = 0.0217822 loss)
I0428 18:33:26.328649  8468 sgd_solver.cpp:105] Iteration 10152, lr = 0.00133859
I0428 18:33:31.834997  8468 solver.cpp:218] Iteration 10164 (2.17939 iter/s, 5.50612s/12 iters), loss = 0.0989524
I0428 18:33:31.835050  8468 solver.cpp:237]     Train net output #0: loss = 0.0989525 (* 1 = 0.0989525 loss)
I0428 18:33:31.835062  8468 sgd_solver.cpp:105] Iteration 10164, lr = 0.00133541
I0428 18:33:37.408255  8468 solver.cpp:218] Iteration 10176 (2.15325 iter/s, 5.57298s/12 iters), loss = 0.031199
I0428 18:33:37.408298  8468 solver.cpp:237]     Train net output #0: loss = 0.031199 (* 1 = 0.031199 loss)
I0428 18:33:37.408306  8468 sgd_solver.cpp:105] Iteration 10176, lr = 0.00133224
I0428 18:33:42.791249  8468 solver.cpp:218] Iteration 10188 (2.22935 iter/s, 5.38273s/12 iters), loss = 0.077054
I0428 18:33:42.791287  8468 solver.cpp:237]     Train net output #0: loss = 0.0770541 (* 1 = 0.0770541 loss)
I0428 18:33:42.791296  8468 sgd_solver.cpp:105] Iteration 10188, lr = 0.00132908
I0428 18:33:47.722312  8468 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_10200.caffemodel
I0428 18:33:48.673113  8468 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_10200.solverstate
I0428 18:33:49.144469  8468 solver.cpp:310] Iteration 10200, loss = 0.0430824
I0428 18:33:49.144529  8468 solver.cpp:330] Iteration 10200, Testing net (#0)
I0428 18:33:49.144536  8468 net.cpp:676] Ignoring source layer train-data
I0428 18:33:49.547199  8497 data_layer.cpp:73] Restarting data prefetching from start.
I0428 18:33:53.852267  8468 solver.cpp:397]     Test net output #0: accuracy = 0.470588
I0428 18:33:53.852303  8468 solver.cpp:397]     Test net output #1: loss = 3.79919 (* 1 = 3.79919 loss)
I0428 18:33:53.852309  8468 solver.cpp:315] Optimization Done.
I0428 18:33:53.852313  8468 caffe.cpp:259] Optimization Done.