I0412 12:56:05.765990 8032 upgrade_proto.cpp:1082] Attempting to upgrade input file specified using deprecated 'solver_type' field (enum)': /mnt/bigdisk/DIGITS-MAN-2/digits/jobs/20210412-125604-40a6/solver.prototxt I0412 12:56:05.766211 8032 upgrade_proto.cpp:1089] Successfully upgraded file specified using deprecated 'solver_type' field (enum) to 'type' field (string). W0412 12:56:05.766219 8032 upgrade_proto.cpp:1091] Note that future Caffe releases will only support 'type' field (string) for a solver's type. I0412 12:56:05.766310 8032 caffe.cpp:218] Using GPUs 1 I0412 12:56:05.795572 8032 caffe.cpp:223] GPU 1: GeForce GTX 1080 Ti I0412 12:56:06.098641 8032 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: 1 net: "train_val.prototxt" train_state { level: 0 stage: "" } type: "SGD" I0412 12:56:06.099412 8032 solver.cpp:87] Creating training net from net file: train_val.prototxt I0412 12:56:06.100018 8032 net.cpp:294] The NetState phase (0) differed from the phase (1) specified by a rule in layer val-data I0412 12:56:06.100033 8032 net.cpp:294] The NetState phase (0) differed from the phase (1) specified by a rule in layer accuracy I0412 12:56:06.100180 8032 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-2/digits/jobs/20210407-214532-d396/mean.binaryproto" } data_param { source: "/mnt/bigdisk/DIGITS-MAN-2/digits/jobs/20210407-214532-d396/train_db" batch_size: 128 backend: LMDB } } layer { name: "conv1" type: "Convolution" bottom: "data" top: "conv1" param { lr_mult: 1 decay_mult: 1 } param { lr_mult: 2 decay_mult: 0 } convolution_param { num_output: 96 kernel_size: 11 stride: 4 weight_filler { type: "gaussian" std: 0.01 } bias_filler { type: "constant" value: 0 } } } layer { name: "relu1" type: "ReLU" bottom: "conv1" top: "conv1" } layer { name: "norm1" type: "LRN" bottom: "conv1" top: "norm1" lrn_param { local_size: 5 alpha: 0.0001 beta: 0.75 } } layer { name: "pool1" type: "Pooling" bottom: "norm1" top: "pool1" pooling_param { pool: MAX kernel_size: 3 stride: 2 } } layer { name: "conv2" type: "Convolution" bottom: "pool1" top: "conv2" param { lr_mult: 1 decay_mult: 1 } param { lr_mult: 2 decay_mult: 0 } convolution_param { num_output: 256 pad: 2 kernel_size: 5 group: 2 weight_filler { type: "gaussian" std: 0.01 } bias_filler { type: "constant" value: 0.1 } } } layer { name: "relu2" type: "ReLU" bottom: "conv2" top: "conv2" } layer { name: "norm2" type: "LRN" bottom: "conv2" top: "norm2" lrn_param { local_size: 5 alpha: 0.0001 beta: 0.75 } } layer { name: "pool2" type: "Pooling" bottom: "norm2" top: "pool2" pooling_param { pool: MAX kernel_size: 3 stride: 2 } } layer { name: "conv3" type: "Convolution" bottom: "pool2" top: "conv3" param { lr_mult: 1 decay_mult: 1 } param { lr_mult: 2 decay_mult: 0 } convolution_param { num_output: 384 pad: 1 kernel_size: 3 weight_filler { type: "gaussian" std: 0.01 } bias_filler { type: "constant" value: 0 } } } layer { name: "relu3" type: "ReLU" bottom: "conv3" top: "conv3" } layer { name: "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: 192 pad: 1 kernel_size: 3 group: 2 weight_filler { type: "gaussian" std: 0.01 } bias_filler { type: "constant" value: 0.1 } } } layer { name: "relu4" type: "ReLU" bottom: "conv4" top: "conv4" } layer { name: "conv4.2" type: "Convolution" bottom: "conv4" top: "conv4.2" param { lr_mult: 1 decay_mult: 1 } param { lr_mult: 2 decay_mult: 0 } convolution_param { num_output: 192 pad: 1 kernel_size: 3 group: 2 weight_filler { type: "gaussian" std: 0.01 } bias_filler { type: "constant" value: 0.1 } } } layer { name: "relu4.2" type: "ReLU" bottom: "conv4.2" top: "conv4.2" } layer { name: "conv5" type: "Convolution" bottom: "conv4.2" top: "conv5" param { lr_mult: 1 decay_mult: 1 } param { lr_mult: 2 decay_mult: 0 } convolution_param { num_output: 256 pad: 1 kernel_size: 3 group: 2 weight_filler { type: "gaussian" std: 0.01 } bias_filler { type: "constant" value: 0.1 } } } layer { name: "relu5" type: "ReLU" bottom: "conv5" top: "conv5" } layer { name: "pool5" type: "Pooling" bottom: "conv5" top: "pool5" pooling_param { pool: MAX kernel_size: 3 stride: 2 } } layer { name: "fc6" type: "InnerProduct" bottom: "pool5" top: "fc6" param { lr_mult: 1 decay_mult: 1 } param { lr_mult: 2 decay_mult: 0 } inner_product_param { num_output: 4096 weight_filler { type: "gaussian" std: 0.005 } bias_filler { type: "constant" value: 0.1 } } } layer { name: "relu6" type: "ReLU" bottom: "fc6" top: "fc6" } layer { name: "drop6" type: "Dropout" bottom: "fc6" top: "fc6" dropout_param { dropout_ratio: 0.5 } } layer { name: "fc7" type: "InnerProduct" bottom: "fc6" top: "fc7" param { lr_mult: 1 decay_mult: 1 } param { lr_mult: 2 decay_mult: 0 } inner_product_param { num_output: 4096 weight_filler { type: "gaussian" std: 0.005 } bias_filler { type: "constant" value: 0.1 } } } layer { name: "relu7" type: "ReLU" bottom: "fc7" top: "fc7" } layer { name: "drop7" type: "Dropout" bottom: "fc7" top: "fc7" dropout_param { dropout_ratio: 0.5 } } layer { name: "fc8" type: "InnerProduct" bottom: "fc7" top: "fc8" param { lr_mult: 1 decay_mult: 1 } param { lr_mult: 2 decay_mult: 0 } inner_product_param { num_output: 196 weight_filler { type: "gaussian" std: 0.01 } bias_filler { type: "constant" value: 0 } } } layer { name: "loss" type: "SoftmaxWithLoss" bottom: "fc8" bottom: "label" top: "loss" } I0412 12:56:06.100272 8032 layer_factory.hpp:77] Creating layer train-data I0412 12:56:06.102432 8032 db_lmdb.cpp:35] Opened lmdb /mnt/bigdisk/DIGITS-MAN-2/digits/jobs/20210407-214532-d396/train_db I0412 12:56:06.102643 8032 net.cpp:84] Creating Layer train-data I0412 12:56:06.102655 8032 net.cpp:380] train-data -> data I0412 12:56:06.102680 8032 net.cpp:380] train-data -> label I0412 12:56:06.102695 8032 data_transformer.cpp:25] Loading mean file from: /mnt/bigdisk/DIGITS-MAN-2/digits/jobs/20210407-214532-d396/mean.binaryproto I0412 12:56:06.107628 8032 data_layer.cpp:45] output data size: 128,3,227,227 I0412 12:56:06.240501 8032 net.cpp:122] Setting up train-data I0412 12:56:06.240526 8032 net.cpp:129] Top shape: 128 3 227 227 (19787136) I0412 12:56:06.240532 8032 net.cpp:129] Top shape: 128 (128) I0412 12:56:06.240536 8032 net.cpp:137] Memory required for data: 79149056 I0412 12:56:06.240547 8032 layer_factory.hpp:77] Creating layer conv1 I0412 12:56:06.240568 8032 net.cpp:84] Creating Layer conv1 I0412 12:56:06.240574 8032 net.cpp:406] conv1 <- data I0412 12:56:06.240586 8032 net.cpp:380] conv1 -> conv1 I0412 12:56:06.821391 8032 net.cpp:122] Setting up conv1 I0412 12:56:06.821413 8032 net.cpp:129] Top shape: 128 96 55 55 (37171200) I0412 12:56:06.821417 8032 net.cpp:137] Memory required for data: 227833856 I0412 12:56:06.821436 8032 layer_factory.hpp:77] Creating layer relu1 I0412 12:56:06.821447 8032 net.cpp:84] Creating Layer relu1 I0412 12:56:06.821471 8032 net.cpp:406] relu1 <- conv1 I0412 12:56:06.821478 8032 net.cpp:367] relu1 -> conv1 (in-place) I0412 12:56:06.821768 8032 net.cpp:122] Setting up relu1 I0412 12:56:06.821776 8032 net.cpp:129] Top shape: 128 96 55 55 (37171200) I0412 12:56:06.821780 8032 net.cpp:137] Memory required for data: 376518656 I0412 12:56:06.821784 8032 layer_factory.hpp:77] Creating layer norm1 I0412 12:56:06.821794 8032 net.cpp:84] Creating Layer norm1 I0412 12:56:06.821796 8032 net.cpp:406] norm1 <- conv1 I0412 12:56:06.821802 8032 net.cpp:380] norm1 -> norm1 I0412 12:56:06.822247 8032 net.cpp:122] Setting up norm1 I0412 12:56:06.822257 8032 net.cpp:129] Top shape: 128 96 55 55 (37171200) I0412 12:56:06.822260 8032 net.cpp:137] Memory required for data: 525203456 I0412 12:56:06.822264 8032 layer_factory.hpp:77] Creating layer pool1 I0412 12:56:06.822273 8032 net.cpp:84] Creating Layer pool1 I0412 12:56:06.822276 8032 net.cpp:406] pool1 <- norm1 I0412 12:56:06.822281 8032 net.cpp:380] pool1 -> pool1 I0412 12:56:06.822317 8032 net.cpp:122] Setting up pool1 I0412 12:56:06.822324 8032 net.cpp:129] Top shape: 128 96 27 27 (8957952) I0412 12:56:06.822326 8032 net.cpp:137] Memory required for data: 561035264 I0412 12:56:06.822329 8032 layer_factory.hpp:77] Creating layer conv2 I0412 12:56:06.822340 8032 net.cpp:84] Creating Layer conv2 I0412 12:56:06.822343 8032 net.cpp:406] conv2 <- pool1 I0412 12:56:06.822347 8032 net.cpp:380] conv2 -> conv2 I0412 12:56:06.831840 8032 net.cpp:122] Setting up conv2 I0412 12:56:06.831854 8032 net.cpp:129] Top shape: 128 256 27 27 (23887872) I0412 12:56:06.831859 8032 net.cpp:137] Memory required for data: 656586752 I0412 12:56:06.831868 8032 layer_factory.hpp:77] Creating layer relu2 I0412 12:56:06.831876 8032 net.cpp:84] Creating Layer relu2 I0412 12:56:06.831879 8032 net.cpp:406] relu2 <- conv2 I0412 12:56:06.831887 8032 net.cpp:367] relu2 -> conv2 (in-place) I0412 12:56:06.832381 8032 net.cpp:122] Setting up relu2 I0412 12:56:06.832389 8032 net.cpp:129] Top shape: 128 256 27 27 (23887872) I0412 12:56:06.832393 8032 net.cpp:137] Memory required for data: 752138240 I0412 12:56:06.832397 8032 layer_factory.hpp:77] Creating layer norm2 I0412 12:56:06.832406 8032 net.cpp:84] Creating Layer norm2 I0412 12:56:06.832410 8032 net.cpp:406] norm2 <- conv2 I0412 12:56:06.832415 8032 net.cpp:380] norm2 -> norm2 I0412 12:56:06.832774 8032 net.cpp:122] Setting up norm2 I0412 12:56:06.832783 8032 net.cpp:129] Top shape: 128 256 27 27 (23887872) I0412 12:56:06.832787 8032 net.cpp:137] Memory required for data: 847689728 I0412 12:56:06.832790 8032 layer_factory.hpp:77] Creating layer pool2 I0412 12:56:06.832799 8032 net.cpp:84] Creating Layer pool2 I0412 12:56:06.832803 8032 net.cpp:406] pool2 <- norm2 I0412 12:56:06.832808 8032 net.cpp:380] pool2 -> pool2 I0412 12:56:06.832839 8032 net.cpp:122] Setting up pool2 I0412 12:56:06.832844 8032 net.cpp:129] Top shape: 128 256 13 13 (5537792) I0412 12:56:06.832846 8032 net.cpp:137] Memory required for data: 869840896 I0412 12:56:06.832850 8032 layer_factory.hpp:77] Creating layer conv3 I0412 12:56:06.832859 8032 net.cpp:84] Creating Layer conv3 I0412 12:56:06.832863 8032 net.cpp:406] conv3 <- pool2 I0412 12:56:06.832870 8032 net.cpp:380] conv3 -> conv3 I0412 12:56:06.842919 8032 net.cpp:122] Setting up conv3 I0412 12:56:06.842931 8032 net.cpp:129] Top shape: 128 384 13 13 (8306688) I0412 12:56:06.842934 8032 net.cpp:137] Memory required for data: 903067648 I0412 12:56:06.842943 8032 layer_factory.hpp:77] Creating layer relu3 I0412 12:56:06.842950 8032 net.cpp:84] Creating Layer relu3 I0412 12:56:06.842955 8032 net.cpp:406] relu3 <- conv3 I0412 12:56:06.842960 8032 net.cpp:367] relu3 -> conv3 (in-place) I0412 12:56:06.843444 8032 net.cpp:122] Setting up relu3 I0412 12:56:06.843456 8032 net.cpp:129] Top shape: 128 384 13 13 (8306688) I0412 12:56:06.843459 8032 net.cpp:137] Memory required for data: 936294400 I0412 12:56:06.843463 8032 layer_factory.hpp:77] Creating layer conv4 I0412 12:56:06.843473 8032 net.cpp:84] Creating Layer conv4 I0412 12:56:06.843493 8032 net.cpp:406] conv4 <- conv3 I0412 12:56:06.843500 8032 net.cpp:380] conv4 -> conv4 I0412 12:56:06.850950 8032 net.cpp:122] Setting up conv4 I0412 12:56:06.850962 8032 net.cpp:129] Top shape: 128 192 13 13 (4153344) I0412 12:56:06.850966 8032 net.cpp:137] Memory required for data: 952907776 I0412 12:56:06.850975 8032 layer_factory.hpp:77] Creating layer relu4 I0412 12:56:06.850982 8032 net.cpp:84] Creating Layer relu4 I0412 12:56:06.850986 8032 net.cpp:406] relu4 <- conv4 I0412 12:56:06.850993 8032 net.cpp:367] relu4 -> conv4 (in-place) I0412 12:56:06.851327 8032 net.cpp:122] Setting up relu4 I0412 12:56:06.851337 8032 net.cpp:129] Top shape: 128 192 13 13 (4153344) I0412 12:56:06.851341 8032 net.cpp:137] Memory required for data: 969521152 I0412 12:56:06.851344 8032 layer_factory.hpp:77] Creating layer conv4.2 I0412 12:56:06.851353 8032 net.cpp:84] Creating Layer conv4.2 I0412 12:56:06.851357 8032 net.cpp:406] conv4.2 <- conv4 I0412 12:56:06.851366 8032 net.cpp:380] conv4.2 -> conv4.2 I0412 12:56:06.861027 8032 net.cpp:122] Setting up conv4.2 I0412 12:56:06.861039 8032 net.cpp:129] Top shape: 128 192 13 13 (4153344) I0412 12:56:06.861043 8032 net.cpp:137] Memory required for data: 986134528 I0412 12:56:06.861053 8032 layer_factory.hpp:77] Creating layer relu4.2 I0412 12:56:06.861061 8032 net.cpp:84] Creating Layer relu4.2 I0412 12:56:06.861065 8032 net.cpp:406] relu4.2 <- conv4.2 I0412 12:56:06.861070 8032 net.cpp:367] relu4.2 -> conv4.2 (in-place) I0412 12:56:06.861552 8032 net.cpp:122] Setting up relu4.2 I0412 12:56:06.861562 8032 net.cpp:129] Top shape: 128 192 13 13 (4153344) I0412 12:56:06.861567 8032 net.cpp:137] Memory required for data: 1002747904 I0412 12:56:06.861569 8032 layer_factory.hpp:77] Creating layer conv5 I0412 12:56:06.861579 8032 net.cpp:84] Creating Layer conv5 I0412 12:56:06.861584 8032 net.cpp:406] conv5 <- conv4.2 I0412 12:56:06.861591 8032 net.cpp:380] conv5 -> conv5 I0412 12:56:06.866694 8032 net.cpp:122] Setting up conv5 I0412 12:56:06.866705 8032 net.cpp:129] Top shape: 128 256 13 13 (5537792) I0412 12:56:06.866709 8032 net.cpp:137] Memory required for data: 1024899072 I0412 12:56:06.866715 8032 layer_factory.hpp:77] Creating layer relu5 I0412 12:56:06.866724 8032 net.cpp:84] Creating Layer relu5 I0412 12:56:06.866727 8032 net.cpp:406] relu5 <- conv5 I0412 12:56:06.866734 8032 net.cpp:367] relu5 -> conv5 (in-place) I0412 12:56:06.867223 8032 net.cpp:122] Setting up relu5 I0412 12:56:06.867233 8032 net.cpp:129] Top shape: 128 256 13 13 (5537792) I0412 12:56:06.867235 8032 net.cpp:137] Memory required for data: 1047050240 I0412 12:56:06.867239 8032 layer_factory.hpp:77] Creating layer pool5 I0412 12:56:06.867246 8032 net.cpp:84] Creating Layer pool5 I0412 12:56:06.867249 8032 net.cpp:406] pool5 <- conv5 I0412 12:56:06.867256 8032 net.cpp:380] pool5 -> pool5 I0412 12:56:06.867292 8032 net.cpp:122] Setting up pool5 I0412 12:56:06.867298 8032 net.cpp:129] Top shape: 128 256 6 6 (1179648) I0412 12:56:06.867301 8032 net.cpp:137] Memory required for data: 1051768832 I0412 12:56:06.867305 8032 layer_factory.hpp:77] Creating layer fc6 I0412 12:56:06.867312 8032 net.cpp:84] Creating Layer fc6 I0412 12:56:06.867316 8032 net.cpp:406] fc6 <- pool5 I0412 12:56:06.867321 8032 net.cpp:380] fc6 -> fc6 I0412 12:56:07.226191 8032 net.cpp:122] Setting up fc6 I0412 12:56:07.226214 8032 net.cpp:129] Top shape: 128 4096 (524288) I0412 12:56:07.226218 8032 net.cpp:137] Memory required for data: 1053865984 I0412 12:56:07.226228 8032 layer_factory.hpp:77] Creating layer relu6 I0412 12:56:07.226238 8032 net.cpp:84] Creating Layer relu6 I0412 12:56:07.226241 8032 net.cpp:406] relu6 <- fc6 I0412 12:56:07.226248 8032 net.cpp:367] relu6 -> fc6 (in-place) I0412 12:56:07.226642 8032 net.cpp:122] Setting up relu6 I0412 12:56:07.226650 8032 net.cpp:129] Top shape: 128 4096 (524288) I0412 12:56:07.226655 8032 net.cpp:137] Memory required for data: 1055963136 I0412 12:56:07.226658 8032 layer_factory.hpp:77] Creating layer drop6 I0412 12:56:07.226686 8032 net.cpp:84] Creating Layer drop6 I0412 12:56:07.226689 8032 net.cpp:406] drop6 <- fc6 I0412 12:56:07.226696 8032 net.cpp:367] drop6 -> fc6 (in-place) I0412 12:56:07.226722 8032 net.cpp:122] Setting up drop6 I0412 12:56:07.226729 8032 net.cpp:129] Top shape: 128 4096 (524288) I0412 12:56:07.226733 8032 net.cpp:137] Memory required for data: 1058060288 I0412 12:56:07.226737 8032 layer_factory.hpp:77] Creating layer fc7 I0412 12:56:07.226744 8032 net.cpp:84] Creating Layer fc7 I0412 12:56:07.226748 8032 net.cpp:406] fc7 <- fc6 I0412 12:56:07.226755 8032 net.cpp:380] fc7 -> fc7 I0412 12:56:07.385761 8032 net.cpp:122] Setting up fc7 I0412 12:56:07.385784 8032 net.cpp:129] Top shape: 128 4096 (524288) I0412 12:56:07.385788 8032 net.cpp:137] Memory required for data: 1060157440 I0412 12:56:07.385797 8032 layer_factory.hpp:77] Creating layer relu7 I0412 12:56:07.385807 8032 net.cpp:84] Creating Layer relu7 I0412 12:56:07.385812 8032 net.cpp:406] relu7 <- fc7 I0412 12:56:07.385819 8032 net.cpp:367] relu7 -> fc7 (in-place) I0412 12:56:07.390307 8032 net.cpp:122] Setting up relu7 I0412 12:56:07.390321 8032 net.cpp:129] Top shape: 128 4096 (524288) I0412 12:56:07.390323 8032 net.cpp:137] Memory required for data: 1062254592 I0412 12:56:07.390327 8032 layer_factory.hpp:77] Creating layer drop7 I0412 12:56:07.390336 8032 net.cpp:84] Creating Layer drop7 I0412 12:56:07.390339 8032 net.cpp:406] drop7 <- fc7 I0412 12:56:07.390345 8032 net.cpp:367] drop7 -> fc7 (in-place) I0412 12:56:07.390372 8032 net.cpp:122] Setting up drop7 I0412 12:56:07.390377 8032 net.cpp:129] Top shape: 128 4096 (524288) I0412 12:56:07.390380 8032 net.cpp:137] Memory required for data: 1064351744 I0412 12:56:07.390383 8032 layer_factory.hpp:77] Creating layer fc8 I0412 12:56:07.390390 8032 net.cpp:84] Creating Layer fc8 I0412 12:56:07.390393 8032 net.cpp:406] fc8 <- fc7 I0412 12:56:07.390406 8032 net.cpp:380] fc8 -> fc8 I0412 12:56:07.398075 8032 net.cpp:122] Setting up fc8 I0412 12:56:07.398088 8032 net.cpp:129] Top shape: 128 196 (25088) I0412 12:56:07.398092 8032 net.cpp:137] Memory required for data: 1064452096 I0412 12:56:07.398102 8032 layer_factory.hpp:77] Creating layer loss I0412 12:56:07.398113 8032 net.cpp:84] Creating Layer loss I0412 12:56:07.398116 8032 net.cpp:406] loss <- fc8 I0412 12:56:07.398121 8032 net.cpp:406] loss <- label I0412 12:56:07.398128 8032 net.cpp:380] loss -> loss I0412 12:56:07.398139 8032 layer_factory.hpp:77] Creating layer loss I0412 12:56:07.398749 8032 net.cpp:122] Setting up loss I0412 12:56:07.398758 8032 net.cpp:129] Top shape: (1) I0412 12:56:07.398761 8032 net.cpp:132] with loss weight 1 I0412 12:56:07.398778 8032 net.cpp:137] Memory required for data: 1064452100 I0412 12:56:07.398782 8032 net.cpp:198] loss needs backward computation. I0412 12:56:07.398789 8032 net.cpp:198] fc8 needs backward computation. I0412 12:56:07.398792 8032 net.cpp:198] drop7 needs backward computation. I0412 12:56:07.398795 8032 net.cpp:198] relu7 needs backward computation. I0412 12:56:07.398798 8032 net.cpp:198] fc7 needs backward computation. I0412 12:56:07.398802 8032 net.cpp:198] drop6 needs backward computation. I0412 12:56:07.398805 8032 net.cpp:198] relu6 needs backward computation. I0412 12:56:07.398809 8032 net.cpp:198] fc6 needs backward computation. I0412 12:56:07.398813 8032 net.cpp:198] pool5 needs backward computation. I0412 12:56:07.398816 8032 net.cpp:198] relu5 needs backward computation. I0412 12:56:07.398819 8032 net.cpp:198] conv5 needs backward computation. I0412 12:56:07.398823 8032 net.cpp:198] relu4.2 needs backward computation. I0412 12:56:07.398826 8032 net.cpp:198] conv4.2 needs backward computation. I0412 12:56:07.398831 8032 net.cpp:198] relu4 needs backward computation. I0412 12:56:07.398834 8032 net.cpp:198] conv4 needs backward computation. I0412 12:56:07.398838 8032 net.cpp:198] relu3 needs backward computation. I0412 12:56:07.398841 8032 net.cpp:198] conv3 needs backward computation. I0412 12:56:07.398844 8032 net.cpp:198] pool2 needs backward computation. I0412 12:56:07.398867 8032 net.cpp:198] norm2 needs backward computation. I0412 12:56:07.398870 8032 net.cpp:198] relu2 needs backward computation. I0412 12:56:07.398874 8032 net.cpp:198] conv2 needs backward computation. I0412 12:56:07.398877 8032 net.cpp:198] pool1 needs backward computation. I0412 12:56:07.398881 8032 net.cpp:198] norm1 needs backward computation. I0412 12:56:07.398885 8032 net.cpp:198] relu1 needs backward computation. I0412 12:56:07.398887 8032 net.cpp:198] conv1 needs backward computation. I0412 12:56:07.398891 8032 net.cpp:200] train-data does not need backward computation. I0412 12:56:07.398895 8032 net.cpp:242] This network produces output loss I0412 12:56:07.398910 8032 net.cpp:255] Network initialization done. I0412 12:56:07.399443 8032 solver.cpp:172] Creating test net (#0) specified by net file: train_val.prototxt I0412 12:56:07.399474 8032 net.cpp:294] The NetState phase (1) differed from the phase (0) specified by a rule in layer train-data I0412 12:56:07.399623 8032 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-2/digits/jobs/20210407-214532-d396/mean.binaryproto" } data_param { source: "/mnt/bigdisk/DIGITS-MAN-2/digits/jobs/20210407-214532-d396/val_db" batch_size: 32 backend: LMDB } } layer { name: "conv1" type: "Convolution" bottom: "data" top: "conv1" param { lr_mult: 1 decay_mult: 1 } param { lr_mult: 2 decay_mult: 0 } convolution_param { num_output: 96 kernel_size: 11 stride: 4 weight_filler { type: "gaussian" std: 0.01 } bias_filler { type: "constant" value: 0 } } } layer { name: "relu1" type: "ReLU" bottom: "conv1" top: "conv1" } layer { name: "norm1" type: "LRN" bottom: "conv1" top: "norm1" lrn_param { local_size: 5 alpha: 0.0001 beta: 0.75 } } layer { name: "pool1" type: "Pooling" bottom: "norm1" top: "pool1" pooling_param { pool: MAX kernel_size: 3 stride: 2 } } layer { name: "conv2" type: "Convolution" bottom: "pool1" top: "conv2" param { lr_mult: 1 decay_mult: 1 } param { lr_mult: 2 decay_mult: 0 } convolution_param { num_output: 256 pad: 2 kernel_size: 5 group: 2 weight_filler { type: "gaussian" std: 0.01 } bias_filler { type: "constant" value: 0.1 } } } layer { name: "relu2" type: "ReLU" bottom: "conv2" top: "conv2" } layer { name: "norm2" type: "LRN" bottom: "conv2" top: "norm2" lrn_param { local_size: 5 alpha: 0.0001 beta: 0.75 } } layer { name: "pool2" type: "Pooling" bottom: "norm2" top: "pool2" pooling_param { pool: MAX kernel_size: 3 stride: 2 } } layer { name: "conv3" type: "Convolution" bottom: "pool2" top: "conv3" param { lr_mult: 1 decay_mult: 1 } param { lr_mult: 2 decay_mult: 0 } convolution_param { num_output: 384 pad: 1 kernel_size: 3 weight_filler { type: "gaussian" std: 0.01 } bias_filler { type: "constant" value: 0 } } } layer { name: "relu3" type: "ReLU" bottom: "conv3" top: "conv3" } layer { name: "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: 192 pad: 1 kernel_size: 3 group: 2 weight_filler { type: "gaussian" std: 0.01 } bias_filler { type: "constant" value: 0.1 } } } layer { name: "relu4" type: "ReLU" bottom: "conv4" top: "conv4" } layer { name: "conv4.2" type: "Convolution" bottom: "conv4" top: "conv4.2" param { lr_mult: 1 decay_mult: 1 } param { lr_mult: 2 decay_mult: 0 } convolution_param { num_output: 192 pad: 1 kernel_size: 3 group: 2 weight_filler { type: "gaussian" std: 0.01 } bias_filler { type: "constant" value: 0.1 } } } layer { name: "relu4.2" type: "ReLU" bottom: "conv4.2" top: "conv4.2" } layer { name: "conv5" type: "Convolution" bottom: "conv4.2" top: "conv5" param { lr_mult: 1 decay_mult: 1 } param { lr_mult: 2 decay_mult: 0 } convolution_param { num_output: 256 pad: 1 kernel_size: 3 group: 2 weight_filler { type: "gaussian" std: 0.01 } bias_filler { type: "constant" value: 0.1 } } } layer { name: "relu5" type: "ReLU" bottom: "conv5" top: "conv5" } layer { name: "pool5" type: "Pooling" bottom: "conv5" top: "pool5" pooling_param { pool: MAX kernel_size: 3 stride: 2 } } layer { name: "fc6" type: "InnerProduct" bottom: "pool5" top: "fc6" param { lr_mult: 1 decay_mult: 1 } param { lr_mult: 2 decay_mult: 0 } inner_product_param { num_output: 4096 weight_filler { type: "gaussian" std: 0.005 } bias_filler { type: "constant" value: 0.1 } } } layer { name: "relu6" type: "ReLU" bottom: "fc6" top: "fc6" } layer { name: "drop6" type: "Dropout" bottom: "fc6" top: "fc6" dropout_param { dropout_ratio: 0.5 } } layer { name: "fc7" type: "InnerProduct" bottom: "fc6" top: "fc7" param { lr_mult: 1 decay_mult: 1 } param { lr_mult: 2 decay_mult: 0 } inner_product_param { num_output: 4096 weight_filler { type: "gaussian" std: 0.005 } bias_filler { type: "constant" value: 0.1 } } } layer { name: "relu7" type: "ReLU" bottom: "fc7" top: "fc7" } layer { name: "drop7" type: "Dropout" bottom: "fc7" top: "fc7" dropout_param { dropout_ratio: 0.5 } } layer { name: "fc8" type: "InnerProduct" bottom: "fc7" top: "fc8" param { lr_mult: 1 decay_mult: 1 } param { lr_mult: 2 decay_mult: 0 } inner_product_param { num_output: 196 weight_filler { type: "gaussian" std: 0.01 } bias_filler { type: "constant" value: 0 } } } layer { name: "accuracy" type: "Accuracy" bottom: "fc8" bottom: "label" top: "accuracy" include { phase: TEST } } layer { name: "loss" type: "SoftmaxWithLoss" bottom: "fc8" bottom: "label" top: "loss" } I0412 12:56:07.399720 8032 layer_factory.hpp:77] Creating layer val-data I0412 12:56:07.401391 8032 db_lmdb.cpp:35] Opened lmdb /mnt/bigdisk/DIGITS-MAN-2/digits/jobs/20210407-214532-d396/val_db I0412 12:56:07.401594 8032 net.cpp:84] Creating Layer val-data I0412 12:56:07.401604 8032 net.cpp:380] val-data -> data I0412 12:56:07.401612 8032 net.cpp:380] val-data -> label I0412 12:56:07.401618 8032 data_transformer.cpp:25] Loading mean file from: /mnt/bigdisk/DIGITS-MAN-2/digits/jobs/20210407-214532-d396/mean.binaryproto I0412 12:56:07.405503 8032 data_layer.cpp:45] output data size: 32,3,227,227 I0412 12:56:07.447587 8032 net.cpp:122] Setting up val-data I0412 12:56:07.447609 8032 net.cpp:129] Top shape: 32 3 227 227 (4946784) I0412 12:56:07.447613 8032 net.cpp:129] Top shape: 32 (32) I0412 12:56:07.447618 8032 net.cpp:137] Memory required for data: 19787264 I0412 12:56:07.447623 8032 layer_factory.hpp:77] Creating layer label_val-data_1_split I0412 12:56:07.447635 8032 net.cpp:84] Creating Layer label_val-data_1_split I0412 12:56:07.447639 8032 net.cpp:406] label_val-data_1_split <- label I0412 12:56:07.447646 8032 net.cpp:380] label_val-data_1_split -> label_val-data_1_split_0 I0412 12:56:07.447655 8032 net.cpp:380] label_val-data_1_split -> label_val-data_1_split_1 I0412 12:56:07.447705 8032 net.cpp:122] Setting up label_val-data_1_split I0412 12:56:07.447710 8032 net.cpp:129] Top shape: 32 (32) I0412 12:56:07.447715 8032 net.cpp:129] Top shape: 32 (32) I0412 12:56:07.447736 8032 net.cpp:137] Memory required for data: 19787520 I0412 12:56:07.447739 8032 layer_factory.hpp:77] Creating layer conv1 I0412 12:56:07.447751 8032 net.cpp:84] Creating Layer conv1 I0412 12:56:07.447754 8032 net.cpp:406] conv1 <- data I0412 12:56:07.447759 8032 net.cpp:380] conv1 -> conv1 I0412 12:56:07.449777 8032 net.cpp:122] Setting up conv1 I0412 12:56:07.449788 8032 net.cpp:129] Top shape: 32 96 55 55 (9292800) I0412 12:56:07.449792 8032 net.cpp:137] Memory required for data: 56958720 I0412 12:56:07.449801 8032 layer_factory.hpp:77] Creating layer relu1 I0412 12:56:07.449808 8032 net.cpp:84] Creating Layer relu1 I0412 12:56:07.449812 8032 net.cpp:406] relu1 <- conv1 I0412 12:56:07.449817 8032 net.cpp:367] relu1 -> conv1 (in-place) I0412 12:56:07.450273 8032 net.cpp:122] Setting up relu1 I0412 12:56:07.450282 8032 net.cpp:129] Top shape: 32 96 55 55 (9292800) I0412 12:56:07.450285 8032 net.cpp:137] Memory required for data: 94129920 I0412 12:56:07.450289 8032 layer_factory.hpp:77] Creating layer norm1 I0412 12:56:07.450299 8032 net.cpp:84] Creating Layer norm1 I0412 12:56:07.450301 8032 net.cpp:406] norm1 <- conv1 I0412 12:56:07.450306 8032 net.cpp:380] norm1 -> norm1 I0412 12:56:07.450750 8032 net.cpp:122] Setting up norm1 I0412 12:56:07.450760 8032 net.cpp:129] Top shape: 32 96 55 55 (9292800) I0412 12:56:07.450763 8032 net.cpp:137] Memory required for data: 131301120 I0412 12:56:07.450767 8032 layer_factory.hpp:77] Creating layer pool1 I0412 12:56:07.450774 8032 net.cpp:84] Creating Layer pool1 I0412 12:56:07.450778 8032 net.cpp:406] pool1 <- norm1 I0412 12:56:07.450783 8032 net.cpp:380] pool1 -> pool1 I0412 12:56:07.450811 8032 net.cpp:122] Setting up pool1 I0412 12:56:07.450816 8032 net.cpp:129] Top shape: 32 96 27 27 (2239488) I0412 12:56:07.450819 8032 net.cpp:137] Memory required for data: 140259072 I0412 12:56:07.450824 8032 layer_factory.hpp:77] Creating layer conv2 I0412 12:56:07.450830 8032 net.cpp:84] Creating Layer conv2 I0412 12:56:07.450834 8032 net.cpp:406] conv2 <- pool1 I0412 12:56:07.450839 8032 net.cpp:380] conv2 -> conv2 I0412 12:56:07.461030 8032 net.cpp:122] Setting up conv2 I0412 12:56:07.461048 8032 net.cpp:129] Top shape: 32 256 27 27 (5971968) I0412 12:56:07.461051 8032 net.cpp:137] Memory required for data: 164146944 I0412 12:56:07.461061 8032 layer_factory.hpp:77] Creating layer relu2 I0412 12:56:07.461069 8032 net.cpp:84] Creating Layer relu2 I0412 12:56:07.461073 8032 net.cpp:406] relu2 <- conv2 I0412 12:56:07.461081 8032 net.cpp:367] relu2 -> conv2 (in-place) I0412 12:56:07.461439 8032 net.cpp:122] Setting up relu2 I0412 12:56:07.461447 8032 net.cpp:129] Top shape: 32 256 27 27 (5971968) I0412 12:56:07.461450 8032 net.cpp:137] Memory required for data: 188034816 I0412 12:56:07.461454 8032 layer_factory.hpp:77] Creating layer norm2 I0412 12:56:07.461464 8032 net.cpp:84] Creating Layer norm2 I0412 12:56:07.461467 8032 net.cpp:406] norm2 <- conv2 I0412 12:56:07.461472 8032 net.cpp:380] norm2 -> norm2 I0412 12:56:07.462028 8032 net.cpp:122] Setting up norm2 I0412 12:56:07.462038 8032 net.cpp:129] Top shape: 32 256 27 27 (5971968) I0412 12:56:07.462041 8032 net.cpp:137] Memory required for data: 211922688 I0412 12:56:07.462045 8032 layer_factory.hpp:77] Creating layer pool2 I0412 12:56:07.462052 8032 net.cpp:84] Creating Layer pool2 I0412 12:56:07.462055 8032 net.cpp:406] pool2 <- norm2 I0412 12:56:07.462062 8032 net.cpp:380] pool2 -> pool2 I0412 12:56:07.462093 8032 net.cpp:122] Setting up pool2 I0412 12:56:07.462103 8032 net.cpp:129] Top shape: 32 256 13 13 (1384448) I0412 12:56:07.462106 8032 net.cpp:137] Memory required for data: 217460480 I0412 12:56:07.462110 8032 layer_factory.hpp:77] Creating layer conv3 I0412 12:56:07.462119 8032 net.cpp:84] Creating Layer conv3 I0412 12:56:07.462122 8032 net.cpp:406] conv3 <- pool2 I0412 12:56:07.462128 8032 net.cpp:380] conv3 -> conv3 I0412 12:56:07.480446 8032 net.cpp:122] Setting up conv3 I0412 12:56:07.480463 8032 net.cpp:129] Top shape: 32 384 13 13 (2076672) I0412 12:56:07.480482 8032 net.cpp:137] Memory required for data: 225767168 I0412 12:56:07.480494 8032 layer_factory.hpp:77] Creating layer relu3 I0412 12:56:07.480504 8032 net.cpp:84] Creating Layer relu3 I0412 12:56:07.480509 8032 net.cpp:406] relu3 <- conv3 I0412 12:56:07.480515 8032 net.cpp:367] relu3 -> conv3 (in-place) I0412 12:56:07.480880 8032 net.cpp:122] Setting up relu3 I0412 12:56:07.480890 8032 net.cpp:129] Top shape: 32 384 13 13 (2076672) I0412 12:56:07.480892 8032 net.cpp:137] Memory required for data: 234073856 I0412 12:56:07.480895 8032 layer_factory.hpp:77] Creating layer conv4 I0412 12:56:07.480906 8032 net.cpp:84] Creating Layer conv4 I0412 12:56:07.480911 8032 net.cpp:406] conv4 <- conv3 I0412 12:56:07.480917 8032 net.cpp:380] conv4 -> conv4 I0412 12:56:07.491151 8032 net.cpp:122] Setting up conv4 I0412 12:56:07.491168 8032 net.cpp:129] Top shape: 32 192 13 13 (1038336) I0412 12:56:07.491173 8032 net.cpp:137] Memory required for data: 238227200 I0412 12:56:07.491180 8032 layer_factory.hpp:77] Creating layer relu4 I0412 12:56:07.491189 8032 net.cpp:84] Creating Layer relu4 I0412 12:56:07.491194 8032 net.cpp:406] relu4 <- conv4 I0412 12:56:07.491200 8032 net.cpp:367] relu4 -> conv4 (in-place) I0412 12:56:07.491691 8032 net.cpp:122] Setting up relu4 I0412 12:56:07.491701 8032 net.cpp:129] Top shape: 32 192 13 13 (1038336) I0412 12:56:07.491704 8032 net.cpp:137] Memory required for data: 242380544 I0412 12:56:07.491708 8032 layer_factory.hpp:77] Creating layer conv4.2 I0412 12:56:07.491719 8032 net.cpp:84] Creating Layer conv4.2 I0412 12:56:07.491722 8032 net.cpp:406] conv4.2 <- conv4 I0412 12:56:07.491730 8032 net.cpp:380] conv4.2 -> conv4.2 I0412 12:56:07.497539 8032 net.cpp:122] Setting up conv4.2 I0412 12:56:07.497555 8032 net.cpp:129] Top shape: 32 192 13 13 (1038336) I0412 12:56:07.497557 8032 net.cpp:137] Memory required for data: 246533888 I0412 12:56:07.497570 8032 layer_factory.hpp:77] Creating layer relu4.2 I0412 12:56:07.497577 8032 net.cpp:84] Creating Layer relu4.2 I0412 12:56:07.497581 8032 net.cpp:406] relu4.2 <- conv4.2 I0412 12:56:07.497588 8032 net.cpp:367] relu4.2 -> conv4.2 (in-place) I0412 12:56:07.498102 8032 net.cpp:122] Setting up relu4.2 I0412 12:56:07.498112 8032 net.cpp:129] Top shape: 32 192 13 13 (1038336) I0412 12:56:07.498116 8032 net.cpp:137] Memory required for data: 250687232 I0412 12:56:07.498119 8032 layer_factory.hpp:77] Creating layer conv5 I0412 12:56:07.498133 8032 net.cpp:84] Creating Layer conv5 I0412 12:56:07.498136 8032 net.cpp:406] conv5 <- conv4.2 I0412 12:56:07.498142 8032 net.cpp:380] conv5 -> conv5 I0412 12:56:07.503863 8032 net.cpp:122] Setting up conv5 I0412 12:56:07.503875 8032 net.cpp:129] Top shape: 32 256 13 13 (1384448) I0412 12:56:07.503880 8032 net.cpp:137] Memory required for data: 256225024 I0412 12:56:07.503886 8032 layer_factory.hpp:77] Creating layer relu5 I0412 12:56:07.503892 8032 net.cpp:84] Creating Layer relu5 I0412 12:56:07.503896 8032 net.cpp:406] relu5 <- conv5 I0412 12:56:07.503902 8032 net.cpp:367] relu5 -> conv5 (in-place) I0412 12:56:07.504398 8032 net.cpp:122] Setting up relu5 I0412 12:56:07.504408 8032 net.cpp:129] Top shape: 32 256 13 13 (1384448) I0412 12:56:07.504412 8032 net.cpp:137] Memory required for data: 261762816 I0412 12:56:07.504415 8032 layer_factory.hpp:77] Creating layer pool5 I0412 12:56:07.504422 8032 net.cpp:84] Creating Layer pool5 I0412 12:56:07.504426 8032 net.cpp:406] pool5 <- conv5 I0412 12:56:07.504431 8032 net.cpp:380] pool5 -> pool5 I0412 12:56:07.504469 8032 net.cpp:122] Setting up pool5 I0412 12:56:07.504475 8032 net.cpp:129] Top shape: 32 256 6 6 (294912) I0412 12:56:07.504478 8032 net.cpp:137] Memory required for data: 262942464 I0412 12:56:07.504482 8032 layer_factory.hpp:77] Creating layer fc6 I0412 12:56:07.504490 8032 net.cpp:84] Creating Layer fc6 I0412 12:56:07.504493 8032 net.cpp:406] fc6 <- pool5 I0412 12:56:07.504498 8032 net.cpp:380] fc6 -> fc6 I0412 12:56:07.865087 8032 net.cpp:122] Setting up fc6 I0412 12:56:07.865108 8032 net.cpp:129] Top shape: 32 4096 (131072) I0412 12:56:07.865135 8032 net.cpp:137] Memory required for data: 263466752 I0412 12:56:07.865145 8032 layer_factory.hpp:77] Creating layer relu6 I0412 12:56:07.865154 8032 net.cpp:84] Creating Layer relu6 I0412 12:56:07.865159 8032 net.cpp:406] relu6 <- fc6 I0412 12:56:07.865166 8032 net.cpp:367] relu6 -> fc6 (in-place) I0412 12:56:07.868643 8032 net.cpp:122] Setting up relu6 I0412 12:56:07.868654 8032 net.cpp:129] Top shape: 32 4096 (131072) I0412 12:56:07.868659 8032 net.cpp:137] Memory required for data: 263991040 I0412 12:56:07.868662 8032 layer_factory.hpp:77] Creating layer drop6 I0412 12:56:07.868669 8032 net.cpp:84] Creating Layer drop6 I0412 12:56:07.868672 8032 net.cpp:406] drop6 <- fc6 I0412 12:56:07.868677 8032 net.cpp:367] drop6 -> fc6 (in-place) I0412 12:56:07.868706 8032 net.cpp:122] Setting up drop6 I0412 12:56:07.868711 8032 net.cpp:129] Top shape: 32 4096 (131072) I0412 12:56:07.868714 8032 net.cpp:137] Memory required for data: 264515328 I0412 12:56:07.868718 8032 layer_factory.hpp:77] Creating layer fc7 I0412 12:56:07.868726 8032 net.cpp:84] Creating Layer fc7 I0412 12:56:07.868728 8032 net.cpp:406] fc7 <- fc6 I0412 12:56:07.868733 8032 net.cpp:380] fc7 -> fc7 I0412 12:56:08.029983 8032 net.cpp:122] Setting up fc7 I0412 12:56:08.030006 8032 net.cpp:129] Top shape: 32 4096 (131072) I0412 12:56:08.030010 8032 net.cpp:137] Memory required for data: 265039616 I0412 12:56:08.030019 8032 layer_factory.hpp:77] Creating layer relu7 I0412 12:56:08.030028 8032 net.cpp:84] Creating Layer relu7 I0412 12:56:08.030033 8032 net.cpp:406] relu7 <- fc7 I0412 12:56:08.030041 8032 net.cpp:367] relu7 -> fc7 (in-place) I0412 12:56:08.030694 8032 net.cpp:122] Setting up relu7 I0412 12:56:08.030704 8032 net.cpp:129] Top shape: 32 4096 (131072) I0412 12:56:08.030707 8032 net.cpp:137] Memory required for data: 265563904 I0412 12:56:08.030711 8032 layer_factory.hpp:77] Creating layer drop7 I0412 12:56:08.030719 8032 net.cpp:84] Creating Layer drop7 I0412 12:56:08.030722 8032 net.cpp:406] drop7 <- fc7 I0412 12:56:08.030728 8032 net.cpp:367] drop7 -> fc7 (in-place) I0412 12:56:08.030753 8032 net.cpp:122] Setting up drop7 I0412 12:56:08.030759 8032 net.cpp:129] Top shape: 32 4096 (131072) I0412 12:56:08.030762 8032 net.cpp:137] Memory required for data: 266088192 I0412 12:56:08.030766 8032 layer_factory.hpp:77] Creating layer fc8 I0412 12:56:08.030773 8032 net.cpp:84] Creating Layer fc8 I0412 12:56:08.030777 8032 net.cpp:406] fc8 <- fc7 I0412 12:56:08.030783 8032 net.cpp:380] fc8 -> fc8 I0412 12:56:08.038452 8032 net.cpp:122] Setting up fc8 I0412 12:56:08.038465 8032 net.cpp:129] Top shape: 32 196 (6272) I0412 12:56:08.038467 8032 net.cpp:137] Memory required for data: 266113280 I0412 12:56:08.038480 8032 layer_factory.hpp:77] Creating layer fc8_fc8_0_split I0412 12:56:08.038487 8032 net.cpp:84] Creating Layer fc8_fc8_0_split I0412 12:56:08.038491 8032 net.cpp:406] fc8_fc8_0_split <- fc8 I0412 12:56:08.038498 8032 net.cpp:380] fc8_fc8_0_split -> fc8_fc8_0_split_0 I0412 12:56:08.038506 8032 net.cpp:380] fc8_fc8_0_split -> fc8_fc8_0_split_1 I0412 12:56:08.038538 8032 net.cpp:122] Setting up fc8_fc8_0_split I0412 12:56:08.038544 8032 net.cpp:129] Top shape: 32 196 (6272) I0412 12:56:08.038547 8032 net.cpp:129] Top shape: 32 196 (6272) I0412 12:56:08.038550 8032 net.cpp:137] Memory required for data: 266163456 I0412 12:56:08.038554 8032 layer_factory.hpp:77] Creating layer accuracy I0412 12:56:08.038561 8032 net.cpp:84] Creating Layer accuracy I0412 12:56:08.038564 8032 net.cpp:406] accuracy <- fc8_fc8_0_split_0 I0412 12:56:08.038569 8032 net.cpp:406] accuracy <- label_val-data_1_split_0 I0412 12:56:08.038575 8032 net.cpp:380] accuracy -> accuracy I0412 12:56:08.038583 8032 net.cpp:122] Setting up accuracy I0412 12:56:08.038587 8032 net.cpp:129] Top shape: (1) I0412 12:56:08.038589 8032 net.cpp:137] Memory required for data: 266163460 I0412 12:56:08.038594 8032 layer_factory.hpp:77] Creating layer loss I0412 12:56:08.038599 8032 net.cpp:84] Creating Layer loss I0412 12:56:08.038623 8032 net.cpp:406] loss <- fc8_fc8_0_split_1 I0412 12:56:08.038628 8032 net.cpp:406] loss <- label_val-data_1_split_1 I0412 12:56:08.038633 8032 net.cpp:380] loss -> loss I0412 12:56:08.038641 8032 layer_factory.hpp:77] Creating layer loss I0412 12:56:08.039081 8032 net.cpp:122] Setting up loss I0412 12:56:08.039089 8032 net.cpp:129] Top shape: (1) I0412 12:56:08.039093 8032 net.cpp:132] with loss weight 1 I0412 12:56:08.039103 8032 net.cpp:137] Memory required for data: 266163464 I0412 12:56:08.039108 8032 net.cpp:198] loss needs backward computation. I0412 12:56:08.039113 8032 net.cpp:200] accuracy does not need backward computation. I0412 12:56:08.039117 8032 net.cpp:198] fc8_fc8_0_split needs backward computation. I0412 12:56:08.039120 8032 net.cpp:198] fc8 needs backward computation. I0412 12:56:08.039124 8032 net.cpp:198] drop7 needs backward computation. I0412 12:56:08.039129 8032 net.cpp:198] relu7 needs backward computation. I0412 12:56:08.039131 8032 net.cpp:198] fc7 needs backward computation. I0412 12:56:08.039135 8032 net.cpp:198] drop6 needs backward computation. I0412 12:56:08.039139 8032 net.cpp:198] relu6 needs backward computation. I0412 12:56:08.039142 8032 net.cpp:198] fc6 needs backward computation. I0412 12:56:08.039146 8032 net.cpp:198] pool5 needs backward computation. I0412 12:56:08.039150 8032 net.cpp:198] relu5 needs backward computation. I0412 12:56:08.039153 8032 net.cpp:198] conv5 needs backward computation. I0412 12:56:08.039157 8032 net.cpp:198] relu4.2 needs backward computation. I0412 12:56:08.039161 8032 net.cpp:198] conv4.2 needs backward computation. I0412 12:56:08.039165 8032 net.cpp:198] relu4 needs backward computation. I0412 12:56:08.039168 8032 net.cpp:198] conv4 needs backward computation. I0412 12:56:08.039171 8032 net.cpp:198] relu3 needs backward computation. I0412 12:56:08.039175 8032 net.cpp:198] conv3 needs backward computation. I0412 12:56:08.039180 8032 net.cpp:198] pool2 needs backward computation. I0412 12:56:08.039182 8032 net.cpp:198] norm2 needs backward computation. I0412 12:56:08.039186 8032 net.cpp:198] relu2 needs backward computation. I0412 12:56:08.039191 8032 net.cpp:198] conv2 needs backward computation. I0412 12:56:08.039194 8032 net.cpp:198] pool1 needs backward computation. I0412 12:56:08.039198 8032 net.cpp:198] norm1 needs backward computation. I0412 12:56:08.039201 8032 net.cpp:198] relu1 needs backward computation. I0412 12:56:08.039206 8032 net.cpp:198] conv1 needs backward computation. I0412 12:56:08.039209 8032 net.cpp:200] label_val-data_1_split does not need backward computation. I0412 12:56:08.039214 8032 net.cpp:200] val-data does not need backward computation. I0412 12:56:08.039217 8032 net.cpp:242] This network produces output accuracy I0412 12:56:08.039220 8032 net.cpp:242] This network produces output loss I0412 12:56:08.039237 8032 net.cpp:255] Network initialization done. I0412 12:56:08.039321 8032 solver.cpp:56] Solver scaffolding done. I0412 12:56:08.039829 8032 caffe.cpp:248] Starting Optimization I0412 12:56:08.039839 8032 solver.cpp:272] Solving I0412 12:56:08.039842 8032 solver.cpp:273] Learning Rate Policy: exp I0412 12:56:08.041286 8032 solver.cpp:330] Iteration 0, Testing net (#0) I0412 12:56:08.041296 8032 net.cpp:676] Ignoring source layer train-data I0412 12:56:08.127015 8032 blocking_queue.cpp:49] Waiting for data I0412 12:56:12.307719 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 12:56:12.352257 8032 solver.cpp:397] Test net output #0: accuracy = 0.00551471 I0412 12:56:12.352303 8032 solver.cpp:397] Test net output #1: loss = 5.2796 (* 1 = 5.2796 loss) I0412 12:56:12.453359 8032 solver.cpp:218] Iteration 0 (-5.04467e-44 iter/s, 4.41333s/12 iters), loss = 5.28824 I0412 12:56:12.454871 8032 solver.cpp:237] Train net output #0: loss = 5.28824 (* 1 = 5.28824 loss) I0412 12:56:12.454892 8032 sgd_solver.cpp:105] Iteration 0, lr = 0.01 I0412 12:56:16.326056 8032 solver.cpp:218] Iteration 12 (3.09995 iter/s, 3.87103s/12 iters), loss = 5.29409 I0412 12:56:16.326139 8032 solver.cpp:237] Train net output #0: loss = 5.29409 (* 1 = 5.29409 loss) I0412 12:56:16.326150 8032 sgd_solver.cpp:105] Iteration 12, lr = 0.00997626 I0412 12:56:21.211719 8032 solver.cpp:218] Iteration 24 (2.45629 iter/s, 4.88542s/12 iters), loss = 5.28236 I0412 12:56:21.211767 8032 solver.cpp:237] Train net output #0: loss = 5.28236 (* 1 = 5.28236 loss) I0412 12:56:21.211776 8032 sgd_solver.cpp:105] Iteration 24, lr = 0.00995257 I0412 12:56:26.114434 8032 solver.cpp:218] Iteration 36 (2.44773 iter/s, 4.9025s/12 iters), loss = 5.29315 I0412 12:56:26.114482 8032 solver.cpp:237] Train net output #0: loss = 5.29315 (* 1 = 5.29315 loss) I0412 12:56:26.114493 8032 sgd_solver.cpp:105] Iteration 36, lr = 0.00992894 I0412 12:56:31.013247 8032 solver.cpp:218] Iteration 48 (2.44968 iter/s, 4.8986s/12 iters), loss = 5.31025 I0412 12:56:31.013288 8032 solver.cpp:237] Train net output #0: loss = 5.31025 (* 1 = 5.31025 loss) I0412 12:56:31.013298 8032 sgd_solver.cpp:105] Iteration 48, lr = 0.00990537 I0412 12:56:36.159621 8032 solver.cpp:218] Iteration 60 (2.33184 iter/s, 5.14615s/12 iters), loss = 5.29013 I0412 12:56:36.159802 8032 solver.cpp:237] Train net output #0: loss = 5.29013 (* 1 = 5.29013 loss) I0412 12:56:36.159816 8032 sgd_solver.cpp:105] Iteration 60, lr = 0.00988185 I0412 12:56:41.149011 8032 solver.cpp:218] Iteration 72 (2.40527 iter/s, 4.98904s/12 iters), loss = 5.30697 I0412 12:56:41.149056 8032 solver.cpp:237] Train net output #0: loss = 5.30697 (* 1 = 5.30697 loss) I0412 12:56:41.149068 8032 sgd_solver.cpp:105] Iteration 72, lr = 0.00985839 I0412 12:56:46.140774 8032 solver.cpp:218] Iteration 84 (2.40407 iter/s, 4.99154s/12 iters), loss = 5.29812 I0412 12:56:46.140818 8032 solver.cpp:237] Train net output #0: loss = 5.29812 (* 1 = 5.29812 loss) I0412 12:56:46.140827 8032 sgd_solver.cpp:105] Iteration 84, lr = 0.00983498 I0412 12:56:51.177175 8032 solver.cpp:218] Iteration 96 (2.38276 iter/s, 5.03618s/12 iters), loss = 5.30993 I0412 12:56:51.177230 8032 solver.cpp:237] Train net output #0: loss = 5.30993 (* 1 = 5.30993 loss) I0412 12:56:51.177243 8032 sgd_solver.cpp:105] Iteration 96, lr = 0.00981163 I0412 12:56:52.896311 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 12:56:53.239117 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_102.caffemodel I0412 12:56:57.988775 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_102.solverstate I0412 12:57:02.018167 8032 solver.cpp:330] Iteration 102, Testing net (#0) I0412 12:57:02.018195 8032 net.cpp:676] Ignoring source layer train-data I0412 12:57:06.583135 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 12:57:06.659936 8032 solver.cpp:397] Test net output #0: accuracy = 0.00551471 I0412 12:57:06.659987 8032 solver.cpp:397] Test net output #1: loss = 5.29047 (* 1 = 5.29047 loss) I0412 12:57:08.632896 8032 solver.cpp:218] Iteration 108 (0.687478 iter/s, 17.4551s/12 iters), loss = 5.31307 I0412 12:57:08.632953 8032 solver.cpp:237] Train net output #0: loss = 5.31307 (* 1 = 5.31307 loss) I0412 12:57:08.632966 8032 sgd_solver.cpp:105] Iteration 108, lr = 0.00978834 I0412 12:57:13.857015 8032 solver.cpp:218] Iteration 120 (2.29714 iter/s, 5.22389s/12 iters), loss = 5.27878 I0412 12:57:13.857061 8032 solver.cpp:237] Train net output #0: loss = 5.27878 (* 1 = 5.27878 loss) I0412 12:57:13.857071 8032 sgd_solver.cpp:105] Iteration 120, lr = 0.0097651 I0412 12:57:18.802996 8032 solver.cpp:218] Iteration 132 (2.42632 iter/s, 4.94577s/12 iters), loss = 5.24911 I0412 12:57:18.803035 8032 solver.cpp:237] Train net output #0: loss = 5.24911 (* 1 = 5.24911 loss) I0412 12:57:18.803043 8032 sgd_solver.cpp:105] Iteration 132, lr = 0.00974192 I0412 12:57:23.927465 8032 solver.cpp:218] Iteration 144 (2.34181 iter/s, 5.12425s/12 iters), loss = 5.32002 I0412 12:57:23.927511 8032 solver.cpp:237] Train net output #0: loss = 5.32002 (* 1 = 5.32002 loss) I0412 12:57:23.927520 8032 sgd_solver.cpp:105] Iteration 144, lr = 0.00971879 I0412 12:57:28.955991 8032 solver.cpp:218] Iteration 156 (2.38649 iter/s, 5.02831s/12 iters), loss = 5.26659 I0412 12:57:28.956032 8032 solver.cpp:237] Train net output #0: loss = 5.26659 (* 1 = 5.26659 loss) I0412 12:57:28.956041 8032 sgd_solver.cpp:105] Iteration 156, lr = 0.00969571 I0412 12:57:34.000895 8032 solver.cpp:218] Iteration 168 (2.37874 iter/s, 5.04468s/12 iters), loss = 5.27264 I0412 12:57:34.000941 8032 solver.cpp:237] Train net output #0: loss = 5.27264 (* 1 = 5.27264 loss) I0412 12:57:34.000952 8032 sgd_solver.cpp:105] Iteration 168, lr = 0.00967269 I0412 12:57:39.101243 8032 solver.cpp:218] Iteration 180 (2.35288 iter/s, 5.10012s/12 iters), loss = 5.2831 I0412 12:57:39.101411 8032 solver.cpp:237] Train net output #0: loss = 5.2831 (* 1 = 5.2831 loss) I0412 12:57:39.101425 8032 sgd_solver.cpp:105] Iteration 180, lr = 0.00964973 I0412 12:57:44.216595 8032 solver.cpp:218] Iteration 192 (2.34603 iter/s, 5.11501s/12 iters), loss = 5.29391 I0412 12:57:44.216645 8032 solver.cpp:237] Train net output #0: loss = 5.29391 (* 1 = 5.29391 loss) I0412 12:57:44.216656 8032 sgd_solver.cpp:105] Iteration 192, lr = 0.00962682 I0412 12:57:47.997571 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 12:57:48.682423 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_204.caffemodel I0412 12:57:53.357458 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_204.solverstate I0412 12:58:00.935190 8032 solver.cpp:330] Iteration 204, Testing net (#0) I0412 12:58:00.935216 8032 net.cpp:676] Ignoring source layer train-data I0412 12:58:05.305629 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 12:58:05.428984 8032 solver.cpp:397] Test net output #0: accuracy = 0.00551471 I0412 12:58:05.429018 8032 solver.cpp:397] Test net output #1: loss = 5.28821 (* 1 = 5.28821 loss) I0412 12:58:05.516860 8032 solver.cpp:218] Iteration 204 (0.563393 iter/s, 21.2995s/12 iters), loss = 5.27231 I0412 12:58:05.516906 8032 solver.cpp:237] Train net output #0: loss = 5.27231 (* 1 = 5.27231 loss) I0412 12:58:05.516916 8032 sgd_solver.cpp:105] Iteration 204, lr = 0.00960396 I0412 12:58:09.919549 8032 solver.cpp:218] Iteration 216 (2.72574 iter/s, 4.40248s/12 iters), loss = 5.28874 I0412 12:58:09.925128 8032 solver.cpp:237] Train net output #0: loss = 5.28874 (* 1 = 5.28874 loss) I0412 12:58:09.925148 8032 sgd_solver.cpp:105] Iteration 216, lr = 0.00958116 I0412 12:58:15.358954 8032 solver.cpp:218] Iteration 228 (2.20846 iter/s, 5.43366s/12 iters), loss = 5.28309 I0412 12:58:15.358994 8032 solver.cpp:237] Train net output #0: loss = 5.28309 (* 1 = 5.28309 loss) I0412 12:58:15.359002 8032 sgd_solver.cpp:105] Iteration 228, lr = 0.00955841 I0412 12:58:20.184244 8032 solver.cpp:218] Iteration 240 (2.487 iter/s, 4.82508s/12 iters), loss = 5.29764 I0412 12:58:20.184286 8032 solver.cpp:237] Train net output #0: loss = 5.29764 (* 1 = 5.29764 loss) I0412 12:58:20.184294 8032 sgd_solver.cpp:105] Iteration 240, lr = 0.00953572 I0412 12:58:25.611981 8032 solver.cpp:218] Iteration 252 (2.21096 iter/s, 5.42751s/12 iters), loss = 5.28075 I0412 12:58:25.612028 8032 solver.cpp:237] Train net output #0: loss = 5.28075 (* 1 = 5.28075 loss) I0412 12:58:25.612036 8032 sgd_solver.cpp:105] Iteration 252, lr = 0.00951308 I0412 12:58:30.691871 8032 solver.cpp:218] Iteration 264 (2.36236 iter/s, 5.07966s/12 iters), loss = 5.27737 I0412 12:58:30.691929 8032 solver.cpp:237] Train net output #0: loss = 5.27737 (* 1 = 5.27737 loss) I0412 12:58:30.691941 8032 sgd_solver.cpp:105] Iteration 264, lr = 0.00949049 I0412 12:58:35.761399 8032 solver.cpp:218] Iteration 276 (2.36719 iter/s, 5.06929s/12 iters), loss = 5.29669 I0412 12:58:35.761454 8032 solver.cpp:237] Train net output #0: loss = 5.29669 (* 1 = 5.29669 loss) I0412 12:58:35.761466 8032 sgd_solver.cpp:105] Iteration 276, lr = 0.00946796 I0412 12:58:40.730834 8032 solver.cpp:218] Iteration 288 (2.41487 iter/s, 4.96921s/12 iters), loss = 5.29286 I0412 12:58:40.733122 8032 solver.cpp:237] Train net output #0: loss = 5.29286 (* 1 = 5.29286 loss) I0412 12:58:40.733132 8032 sgd_solver.cpp:105] Iteration 288, lr = 0.00944548 I0412 12:58:45.870052 8032 solver.cpp:218] Iteration 300 (2.3361 iter/s, 5.13676s/12 iters), loss = 5.28804 I0412 12:58:45.870095 8032 solver.cpp:237] Train net output #0: loss = 5.28804 (* 1 = 5.28804 loss) I0412 12:58:45.870105 8032 sgd_solver.cpp:105] Iteration 300, lr = 0.00942305 I0412 12:58:46.958283 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 12:58:48.034027 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_306.caffemodel I0412 12:58:50.988782 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_306.solverstate I0412 12:58:53.290102 8032 solver.cpp:330] Iteration 306, Testing net (#0) I0412 12:58:53.290123 8032 net.cpp:676] Ignoring source layer train-data I0412 12:58:57.629873 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 12:58:57.787333 8032 solver.cpp:397] Test net output #0: accuracy = 0.00551471 I0412 12:58:57.787370 8032 solver.cpp:397] Test net output #1: loss = 5.28416 (* 1 = 5.28416 loss) I0412 12:58:59.707466 8032 solver.cpp:218] Iteration 312 (0.867245 iter/s, 13.8369s/12 iters), loss = 5.28155 I0412 12:58:59.707520 8032 solver.cpp:237] Train net output #0: loss = 5.28155 (* 1 = 5.28155 loss) I0412 12:58:59.707532 8032 sgd_solver.cpp:105] Iteration 312, lr = 0.00940068 I0412 12:59:04.715936 8032 solver.cpp:218] Iteration 324 (2.39605 iter/s, 5.00825s/12 iters), loss = 5.2578 I0412 12:59:04.715984 8032 solver.cpp:237] Train net output #0: loss = 5.2578 (* 1 = 5.2578 loss) I0412 12:59:04.715996 8032 sgd_solver.cpp:105] Iteration 324, lr = 0.00937836 I0412 12:59:09.689811 8032 solver.cpp:218] Iteration 336 (2.41271 iter/s, 4.97366s/12 iters), loss = 5.27012 I0412 12:59:09.689867 8032 solver.cpp:237] Train net output #0: loss = 5.27012 (* 1 = 5.27012 loss) I0412 12:59:09.689879 8032 sgd_solver.cpp:105] Iteration 336, lr = 0.0093561 I0412 12:59:14.694465 8032 solver.cpp:218] Iteration 348 (2.39787 iter/s, 5.00443s/12 iters), loss = 5.25753 I0412 12:59:14.696569 8032 solver.cpp:237] Train net output #0: loss = 5.25753 (* 1 = 5.25753 loss) I0412 12:59:14.696580 8032 sgd_solver.cpp:105] Iteration 348, lr = 0.00933388 I0412 12:59:19.901705 8032 solver.cpp:218] Iteration 360 (2.30549 iter/s, 5.20496s/12 iters), loss = 5.27633 I0412 12:59:19.901758 8032 solver.cpp:237] Train net output #0: loss = 5.27633 (* 1 = 5.27633 loss) I0412 12:59:19.901770 8032 sgd_solver.cpp:105] Iteration 360, lr = 0.00931172 I0412 12:59:24.934355 8032 solver.cpp:218] Iteration 372 (2.38454 iter/s, 5.03242s/12 iters), loss = 5.24685 I0412 12:59:24.934409 8032 solver.cpp:237] Train net output #0: loss = 5.24685 (* 1 = 5.24685 loss) I0412 12:59:24.934422 8032 sgd_solver.cpp:105] Iteration 372, lr = 0.00928961 I0412 12:59:29.959259 8032 solver.cpp:218] Iteration 384 (2.38821 iter/s, 5.02468s/12 iters), loss = 5.22751 I0412 12:59:29.959311 8032 solver.cpp:237] Train net output #0: loss = 5.22751 (* 1 = 5.22751 loss) I0412 12:59:29.959321 8032 sgd_solver.cpp:105] Iteration 384, lr = 0.00926756 I0412 12:59:35.032203 8032 solver.cpp:218] Iteration 396 (2.36559 iter/s, 5.07273s/12 iters), loss = 5.12188 I0412 12:59:35.032246 8032 solver.cpp:237] Train net output #0: loss = 5.12188 (* 1 = 5.12188 loss) I0412 12:59:35.032258 8032 sgd_solver.cpp:105] Iteration 396, lr = 0.00924556 I0412 12:59:38.093830 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 12:59:39.504258 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_408.caffemodel I0412 12:59:43.371423 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_408.solverstate I0412 12:59:47.249428 8032 solver.cpp:330] Iteration 408, Testing net (#0) I0412 12:59:47.249552 8032 net.cpp:676] Ignoring source layer train-data I0412 12:59:51.628481 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 12:59:51.833102 8032 solver.cpp:397] Test net output #0: accuracy = 0.00735294 I0412 12:59:51.833135 8032 solver.cpp:397] Test net output #1: loss = 5.19702 (* 1 = 5.19702 loss) I0412 12:59:51.921051 8032 solver.cpp:218] Iteration 408 (0.710552 iter/s, 16.8883s/12 iters), loss = 5.25369 I0412 12:59:51.921097 8032 solver.cpp:237] Train net output #0: loss = 5.25369 (* 1 = 5.25369 loss) I0412 12:59:51.921106 8032 sgd_solver.cpp:105] Iteration 408, lr = 0.00922361 I0412 12:59:56.765784 8032 solver.cpp:218] Iteration 420 (2.47703 iter/s, 4.84452s/12 iters), loss = 5.27075 I0412 12:59:56.765832 8032 solver.cpp:237] Train net output #0: loss = 5.27075 (* 1 = 5.27075 loss) I0412 12:59:56.765843 8032 sgd_solver.cpp:105] Iteration 420, lr = 0.00920171 I0412 13:00:01.719800 8032 solver.cpp:218] Iteration 432 (2.42238 iter/s, 4.9538s/12 iters), loss = 5.18013 I0412 13:00:01.719853 8032 solver.cpp:237] Train net output #0: loss = 5.18013 (* 1 = 5.18013 loss) I0412 13:00:01.719866 8032 sgd_solver.cpp:105] Iteration 432, lr = 0.00917986 I0412 13:00:06.697278 8032 solver.cpp:218] Iteration 444 (2.41097 iter/s, 4.97726s/12 iters), loss = 5.1706 I0412 13:00:06.697327 8032 solver.cpp:237] Train net output #0: loss = 5.1706 (* 1 = 5.1706 loss) I0412 13:00:06.697340 8032 sgd_solver.cpp:105] Iteration 444, lr = 0.00915807 I0412 13:00:11.705672 8032 solver.cpp:218] Iteration 456 (2.39608 iter/s, 5.00817s/12 iters), loss = 5.22719 I0412 13:00:11.705729 8032 solver.cpp:237] Train net output #0: loss = 5.22719 (* 1 = 5.22719 loss) I0412 13:00:11.705742 8032 sgd_solver.cpp:105] Iteration 456, lr = 0.00913632 I0412 13:00:16.722728 8032 solver.cpp:218] Iteration 468 (2.39195 iter/s, 5.01684s/12 iters), loss = 5.17774 I0412 13:00:16.722766 8032 solver.cpp:237] Train net output #0: loss = 5.17774 (* 1 = 5.17774 loss) I0412 13:00:16.722775 8032 sgd_solver.cpp:105] Iteration 468, lr = 0.00911463 I0412 13:00:21.729199 8032 solver.cpp:218] Iteration 480 (2.397 iter/s, 5.00626s/12 iters), loss = 5.14215 I0412 13:00:21.729318 8032 solver.cpp:237] Train net output #0: loss = 5.14215 (* 1 = 5.14215 loss) I0412 13:00:21.729331 8032 sgd_solver.cpp:105] Iteration 480, lr = 0.00909299 I0412 13:00:26.685840 8032 solver.cpp:218] Iteration 492 (2.42113 iter/s, 4.95635s/12 iters), loss = 5.17009 I0412 13:00:26.685899 8032 solver.cpp:237] Train net output #0: loss = 5.17009 (* 1 = 5.17009 loss) I0412 13:00:26.685910 8032 sgd_solver.cpp:105] Iteration 492, lr = 0.0090714 I0412 13:00:31.663373 8032 solver.cpp:218] Iteration 504 (2.41094 iter/s, 4.9773s/12 iters), loss = 5.19825 I0412 13:00:31.663431 8032 solver.cpp:237] Train net output #0: loss = 5.19825 (* 1 = 5.19825 loss) I0412 13:00:31.663444 8032 sgd_solver.cpp:105] Iteration 504, lr = 0.00904986 I0412 13:00:31.933831 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:00:33.655474 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_510.caffemodel I0412 13:00:39.373415 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_510.solverstate I0412 13:00:43.557799 8032 solver.cpp:330] Iteration 510, Testing net (#0) I0412 13:00:43.557826 8032 net.cpp:676] Ignoring source layer train-data I0412 13:00:47.973470 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:00:48.210806 8032 solver.cpp:397] Test net output #0: accuracy = 0.00857843 I0412 13:00:48.210857 8032 solver.cpp:397] Test net output #1: loss = 5.15945 (* 1 = 5.15945 loss) I0412 13:00:50.194761 8032 solver.cpp:218] Iteration 516 (0.647572 iter/s, 18.5307s/12 iters), loss = 5.14721 I0412 13:00:50.194818 8032 solver.cpp:237] Train net output #0: loss = 5.14721 (* 1 = 5.14721 loss) I0412 13:00:50.194829 8032 sgd_solver.cpp:105] Iteration 516, lr = 0.00902838 I0412 13:00:55.116175 8032 solver.cpp:218] Iteration 528 (2.43843 iter/s, 4.92119s/12 iters), loss = 5.19665 I0412 13:00:55.116582 8032 solver.cpp:237] Train net output #0: loss = 5.19665 (* 1 = 5.19665 loss) I0412 13:00:55.116595 8032 sgd_solver.cpp:105] Iteration 528, lr = 0.00900694 I0412 13:01:00.119969 8032 solver.cpp:218] Iteration 540 (2.39846 iter/s, 5.00322s/12 iters), loss = 5.16401 I0412 13:01:00.120015 8032 solver.cpp:237] Train net output #0: loss = 5.16401 (* 1 = 5.16401 loss) I0412 13:01:00.120024 8032 sgd_solver.cpp:105] Iteration 540, lr = 0.00898556 I0412 13:01:05.084097 8032 solver.cpp:218] Iteration 552 (2.41745 iter/s, 4.96392s/12 iters), loss = 5.12106 I0412 13:01:05.084143 8032 solver.cpp:237] Train net output #0: loss = 5.12106 (* 1 = 5.12106 loss) I0412 13:01:05.084152 8032 sgd_solver.cpp:105] Iteration 552, lr = 0.00896423 I0412 13:01:10.141256 8032 solver.cpp:218] Iteration 564 (2.37297 iter/s, 5.05694s/12 iters), loss = 5.17253 I0412 13:01:10.141307 8032 solver.cpp:237] Train net output #0: loss = 5.17253 (* 1 = 5.17253 loss) I0412 13:01:10.141319 8032 sgd_solver.cpp:105] Iteration 564, lr = 0.00894294 I0412 13:01:15.149981 8032 solver.cpp:218] Iteration 576 (2.39593 iter/s, 5.0085s/12 iters), loss = 5.09646 I0412 13:01:15.150020 8032 solver.cpp:237] Train net output #0: loss = 5.09646 (* 1 = 5.09646 loss) I0412 13:01:15.150030 8032 sgd_solver.cpp:105] Iteration 576, lr = 0.00892171 I0412 13:01:20.262459 8032 solver.cpp:218] Iteration 588 (2.3473 iter/s, 5.11227s/12 iters), loss = 5.0929 I0412 13:01:20.262508 8032 solver.cpp:237] Train net output #0: loss = 5.0929 (* 1 = 5.0929 loss) I0412 13:01:20.262517 8032 sgd_solver.cpp:105] Iteration 588, lr = 0.00890053 I0412 13:01:25.205271 8032 solver.cpp:218] Iteration 600 (2.42787 iter/s, 4.9426s/12 iters), loss = 5.15156 I0412 13:01:25.205379 8032 solver.cpp:237] Train net output #0: loss = 5.15156 (* 1 = 5.15156 loss) I0412 13:01:25.205390 8032 sgd_solver.cpp:105] Iteration 600, lr = 0.0088794 I0412 13:01:27.579768 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:01:29.696189 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_612.caffemodel I0412 13:01:35.291451 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_612.solverstate I0412 13:01:38.564833 8032 solver.cpp:330] Iteration 612, Testing net (#0) I0412 13:01:38.564862 8032 net.cpp:676] Ignoring source layer train-data I0412 13:01:42.871809 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:01:43.156450 8032 solver.cpp:397] Test net output #0: accuracy = 0.00980392 I0412 13:01:43.156499 8032 solver.cpp:397] Test net output #1: loss = 5.1112 (* 1 = 5.1112 loss) I0412 13:01:43.244493 8032 solver.cpp:218] Iteration 612 (0.665242 iter/s, 18.0385s/12 iters), loss = 5.10494 I0412 13:01:43.244542 8032 solver.cpp:237] Train net output #0: loss = 5.10494 (* 1 = 5.10494 loss) I0412 13:01:43.244554 8032 sgd_solver.cpp:105] Iteration 612, lr = 0.00885831 I0412 13:01:47.524574 8032 solver.cpp:218] Iteration 624 (2.80381 iter/s, 4.27989s/12 iters), loss = 5.14054 I0412 13:01:47.524616 8032 solver.cpp:237] Train net output #0: loss = 5.14054 (* 1 = 5.14054 loss) I0412 13:01:47.524623 8032 sgd_solver.cpp:105] Iteration 624, lr = 0.00883728 I0412 13:01:52.483906 8032 solver.cpp:218] Iteration 636 (2.41978 iter/s, 4.95912s/12 iters), loss = 4.99235 I0412 13:01:52.483960 8032 solver.cpp:237] Train net output #0: loss = 4.99235 (* 1 = 4.99235 loss) I0412 13:01:52.483973 8032 sgd_solver.cpp:105] Iteration 636, lr = 0.0088163 I0412 13:01:57.470811 8032 solver.cpp:218] Iteration 648 (2.40641 iter/s, 4.98669s/12 iters), loss = 5.1953 I0412 13:01:57.470919 8032 solver.cpp:237] Train net output #0: loss = 5.1953 (* 1 = 5.1953 loss) I0412 13:01:57.470932 8032 sgd_solver.cpp:105] Iteration 648, lr = 0.00879537 I0412 13:02:02.467705 8032 solver.cpp:218] Iteration 660 (2.40163 iter/s, 4.99662s/12 iters), loss = 5.11587 I0412 13:02:02.467758 8032 solver.cpp:237] Train net output #0: loss = 5.11587 (* 1 = 5.11587 loss) I0412 13:02:02.467772 8032 sgd_solver.cpp:105] Iteration 660, lr = 0.00877449 I0412 13:02:07.387569 8032 solver.cpp:218] Iteration 672 (2.4392 iter/s, 4.91965s/12 iters), loss = 5.07346 I0412 13:02:07.387629 8032 solver.cpp:237] Train net output #0: loss = 5.07346 (* 1 = 5.07346 loss) I0412 13:02:07.387643 8032 sgd_solver.cpp:105] Iteration 672, lr = 0.00875366 I0412 13:02:12.310245 8032 solver.cpp:218] Iteration 684 (2.43781 iter/s, 4.92245s/12 iters), loss = 4.95721 I0412 13:02:12.310302 8032 solver.cpp:237] Train net output #0: loss = 4.95721 (* 1 = 4.95721 loss) I0412 13:02:12.310313 8032 sgd_solver.cpp:105] Iteration 684, lr = 0.00873287 I0412 13:02:13.083784 8032 blocking_queue.cpp:49] Waiting for data I0412 13:02:17.408983 8032 solver.cpp:218] Iteration 696 (2.35363 iter/s, 5.09851s/12 iters), loss = 5.03226 I0412 13:02:17.409034 8032 solver.cpp:237] Train net output #0: loss = 5.03226 (* 1 = 5.03226 loss) I0412 13:02:17.409044 8032 sgd_solver.cpp:105] Iteration 696, lr = 0.00871214 I0412 13:02:22.455507 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:02:22.857208 8032 solver.cpp:218] Iteration 708 (2.20265 iter/s, 5.44799s/12 iters), loss = 5.15022 I0412 13:02:22.857264 8032 solver.cpp:237] Train net output #0: loss = 5.15022 (* 1 = 5.15022 loss) I0412 13:02:22.857276 8032 sgd_solver.cpp:105] Iteration 708, lr = 0.00869145 I0412 13:02:24.969043 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_714.caffemodel I0412 13:02:33.207159 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_714.solverstate I0412 13:02:36.043423 8032 solver.cpp:330] Iteration 714, Testing net (#0) I0412 13:02:36.043452 8032 net.cpp:676] Ignoring source layer train-data I0412 13:02:40.426287 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:02:40.745467 8032 solver.cpp:397] Test net output #0: accuracy = 0.0128676 I0412 13:02:40.745517 8032 solver.cpp:397] Test net output #1: loss = 5.07169 (* 1 = 5.07169 loss) I0412 13:02:42.772181 8032 solver.cpp:218] Iteration 720 (0.602582 iter/s, 19.9143s/12 iters), loss = 5.15602 I0412 13:02:42.772224 8032 solver.cpp:237] Train net output #0: loss = 5.15602 (* 1 = 5.15602 loss) I0412 13:02:42.772233 8032 sgd_solver.cpp:105] Iteration 720, lr = 0.00867082 I0412 13:02:47.735023 8032 solver.cpp:218] Iteration 732 (2.41807 iter/s, 4.96263s/12 iters), loss = 4.96896 I0412 13:02:47.735074 8032 solver.cpp:237] Train net output #0: loss = 4.96896 (* 1 = 4.96896 loss) I0412 13:02:47.735086 8032 sgd_solver.cpp:105] Iteration 732, lr = 0.00865023 I0412 13:02:52.707341 8032 solver.cpp:218] Iteration 744 (2.41347 iter/s, 4.9721s/12 iters), loss = 5.01426 I0412 13:02:52.707392 8032 solver.cpp:237] Train net output #0: loss = 5.01426 (* 1 = 5.01426 loss) I0412 13:02:52.707404 8032 sgd_solver.cpp:105] Iteration 744, lr = 0.0086297 I0412 13:02:57.768354 8032 solver.cpp:218] Iteration 756 (2.37117 iter/s, 5.0608s/12 iters), loss = 5.07753 I0412 13:02:57.768393 8032 solver.cpp:237] Train net output #0: loss = 5.07753 (* 1 = 5.07753 loss) I0412 13:02:57.768402 8032 sgd_solver.cpp:105] Iteration 756, lr = 0.00860921 I0412 13:03:03.087517 8032 solver.cpp:218] Iteration 768 (2.25609 iter/s, 5.31895s/12 iters), loss = 5.14375 I0412 13:03:03.087560 8032 solver.cpp:237] Train net output #0: loss = 5.14375 (* 1 = 5.14375 loss) I0412 13:03:03.087569 8032 sgd_solver.cpp:105] Iteration 768, lr = 0.00858877 I0412 13:03:08.212044 8032 solver.cpp:218] Iteration 780 (2.34178 iter/s, 5.12431s/12 iters), loss = 5.0934 I0412 13:03:08.212193 8032 solver.cpp:237] Train net output #0: loss = 5.0934 (* 1 = 5.0934 loss) I0412 13:03:08.212208 8032 sgd_solver.cpp:105] Iteration 780, lr = 0.00856838 I0412 13:03:13.131500 8032 solver.cpp:218] Iteration 792 (2.43945 iter/s, 4.91915s/12 iters), loss = 4.95669 I0412 13:03:13.131558 8032 solver.cpp:237] Train net output #0: loss = 4.95669 (* 1 = 4.95669 loss) I0412 13:03:13.131570 8032 sgd_solver.cpp:105] Iteration 792, lr = 0.00854803 I0412 13:03:18.122802 8032 solver.cpp:218] Iteration 804 (2.40429 iter/s, 4.99108s/12 iters), loss = 5.08503 I0412 13:03:18.122843 8032 solver.cpp:237] Train net output #0: loss = 5.08503 (* 1 = 5.08503 loss) I0412 13:03:18.122850 8032 sgd_solver.cpp:105] Iteration 804, lr = 0.00852774 I0412 13:03:19.886520 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:03:22.688232 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_816.caffemodel I0412 13:03:29.395213 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_816.solverstate I0412 13:03:33.959380 8032 solver.cpp:330] Iteration 816, Testing net (#0) I0412 13:03:33.959403 8032 net.cpp:676] Ignoring source layer train-data I0412 13:03:38.211899 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:03:38.566380 8032 solver.cpp:397] Test net output #0: accuracy = 0.0165441 I0412 13:03:38.566485 8032 solver.cpp:397] Test net output #1: loss = 5.0328 (* 1 = 5.0328 loss) I0412 13:03:38.651634 8032 solver.cpp:218] Iteration 816 (0.584563 iter/s, 20.5281s/12 iters), loss = 5.06284 I0412 13:03:38.651686 8032 solver.cpp:237] Train net output #0: loss = 5.06284 (* 1 = 5.06284 loss) I0412 13:03:38.651698 8032 sgd_solver.cpp:105] Iteration 816, lr = 0.00850749 I0412 13:03:43.005252 8032 solver.cpp:218] Iteration 828 (2.75645 iter/s, 4.35342s/12 iters), loss = 5.13101 I0412 13:03:43.005300 8032 solver.cpp:237] Train net output #0: loss = 5.13101 (* 1 = 5.13101 loss) I0412 13:03:43.005312 8032 sgd_solver.cpp:105] Iteration 828, lr = 0.00848729 I0412 13:03:48.189924 8032 solver.cpp:218] Iteration 840 (2.31461 iter/s, 5.18445s/12 iters), loss = 4.98066 I0412 13:03:48.189970 8032 solver.cpp:237] Train net output #0: loss = 4.98066 (* 1 = 4.98066 loss) I0412 13:03:48.189980 8032 sgd_solver.cpp:105] Iteration 840, lr = 0.00846714 I0412 13:03:53.258026 8032 solver.cpp:218] Iteration 852 (2.36788 iter/s, 5.06783s/12 iters), loss = 4.9437 I0412 13:03:53.258116 8032 solver.cpp:237] Train net output #0: loss = 4.9437 (* 1 = 4.9437 loss) I0412 13:03:53.258143 8032 sgd_solver.cpp:105] Iteration 852, lr = 0.00844704 I0412 13:03:58.358598 8032 solver.cpp:218] Iteration 864 (2.35279 iter/s, 5.10032s/12 iters), loss = 4.99605 I0412 13:03:58.358637 8032 solver.cpp:237] Train net output #0: loss = 4.99605 (* 1 = 4.99605 loss) I0412 13:03:58.358645 8032 sgd_solver.cpp:105] Iteration 864, lr = 0.00842698 I0412 13:04:03.293066 8032 solver.cpp:218] Iteration 876 (2.43197 iter/s, 4.93426s/12 iters), loss = 4.98043 I0412 13:04:03.293112 8032 solver.cpp:237] Train net output #0: loss = 4.98043 (* 1 = 4.98043 loss) I0412 13:04:03.293120 8032 sgd_solver.cpp:105] Iteration 876, lr = 0.00840698 I0412 13:04:08.372081 8032 solver.cpp:218] Iteration 888 (2.36276 iter/s, 5.0788s/12 iters), loss = 4.88912 I0412 13:04:08.372133 8032 solver.cpp:237] Train net output #0: loss = 4.88912 (* 1 = 4.88912 loss) I0412 13:04:08.372143 8032 sgd_solver.cpp:105] Iteration 888, lr = 0.00838702 I0412 13:04:13.630178 8032 solver.cpp:218] Iteration 900 (2.28229 iter/s, 5.25787s/12 iters), loss = 5.01801 I0412 13:04:13.630260 8032 solver.cpp:237] Train net output #0: loss = 5.01801 (* 1 = 5.01801 loss) I0412 13:04:13.630280 8032 sgd_solver.cpp:105] Iteration 900, lr = 0.0083671 I0412 13:04:17.794450 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:04:18.892385 8032 solver.cpp:218] Iteration 912 (2.28052 iter/s, 5.26195s/12 iters), loss = 4.94136 I0412 13:04:18.892428 8032 solver.cpp:237] Train net output #0: loss = 4.94136 (* 1 = 4.94136 loss) I0412 13:04:18.892439 8032 sgd_solver.cpp:105] Iteration 912, lr = 0.00834724 I0412 13:04:20.912603 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_918.caffemodel I0412 13:04:28.013753 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_918.solverstate I0412 13:04:32.088620 8032 solver.cpp:330] Iteration 918, Testing net (#0) I0412 13:04:32.088647 8032 net.cpp:676] Ignoring source layer train-data I0412 13:04:36.662223 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:04:37.062839 8032 solver.cpp:397] Test net output #0: accuracy = 0.0220588 I0412 13:04:37.062876 8032 solver.cpp:397] Test net output #1: loss = 4.99736 (* 1 = 4.99736 loss) I0412 13:04:38.974347 8032 solver.cpp:218] Iteration 924 (0.597571 iter/s, 20.0813s/12 iters), loss = 5.03162 I0412 13:04:38.974393 8032 solver.cpp:237] Train net output #0: loss = 5.03162 (* 1 = 5.03162 loss) I0412 13:04:38.974401 8032 sgd_solver.cpp:105] Iteration 924, lr = 0.00832742 I0412 13:04:44.416997 8032 solver.cpp:218] Iteration 936 (2.2049 iter/s, 5.44242s/12 iters), loss = 5.01727 I0412 13:04:44.417111 8032 solver.cpp:237] Train net output #0: loss = 5.01727 (* 1 = 5.01727 loss) I0412 13:04:44.417124 8032 sgd_solver.cpp:105] Iteration 936, lr = 0.00830765 I0412 13:04:49.591997 8032 solver.cpp:218] Iteration 948 (2.31897 iter/s, 5.17472s/12 iters), loss = 4.95314 I0412 13:04:49.592047 8032 solver.cpp:237] Train net output #0: loss = 4.95314 (* 1 = 4.95314 loss) I0412 13:04:49.592058 8032 sgd_solver.cpp:105] Iteration 948, lr = 0.00828793 I0412 13:04:54.642838 8032 solver.cpp:218] Iteration 960 (2.37594 iter/s, 5.05062s/12 iters), loss = 4.89703 I0412 13:04:54.642884 8032 solver.cpp:237] Train net output #0: loss = 4.89703 (* 1 = 4.89703 loss) I0412 13:04:54.642895 8032 sgd_solver.cpp:105] Iteration 960, lr = 0.00826825 I0412 13:04:59.666731 8032 solver.cpp:218] Iteration 972 (2.38868 iter/s, 5.02369s/12 iters), loss = 5.00577 I0412 13:04:59.666774 8032 solver.cpp:237] Train net output #0: loss = 5.00577 (* 1 = 5.00577 loss) I0412 13:04:59.666785 8032 sgd_solver.cpp:105] Iteration 972, lr = 0.00824862 I0412 13:05:04.712164 8032 solver.cpp:218] Iteration 984 (2.37849 iter/s, 5.04522s/12 iters), loss = 4.91794 I0412 13:05:04.712214 8032 solver.cpp:237] Train net output #0: loss = 4.91794 (* 1 = 4.91794 loss) I0412 13:05:04.712225 8032 sgd_solver.cpp:105] Iteration 984, lr = 0.00822903 I0412 13:05:09.671134 8032 solver.cpp:218] Iteration 996 (2.41996 iter/s, 4.95875s/12 iters), loss = 4.79772 I0412 13:05:09.671181 8032 solver.cpp:237] Train net output #0: loss = 4.79772 (* 1 = 4.79772 loss) I0412 13:05:09.671190 8032 sgd_solver.cpp:105] Iteration 996, lr = 0.0082095 I0412 13:05:14.816756 8032 solver.cpp:218] Iteration 1008 (2.33218 iter/s, 5.1454s/12 iters), loss = 4.99861 I0412 13:05:14.816845 8032 solver.cpp:237] Train net output #0: loss = 4.99861 (* 1 = 4.99861 loss) I0412 13:05:14.816855 8032 sgd_solver.cpp:105] Iteration 1008, lr = 0.00819001 I0412 13:05:15.844970 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:05:19.401829 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1020.caffemodel I0412 13:05:24.022734 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1020.solverstate I0412 13:05:26.553241 8032 solver.cpp:330] Iteration 1020, Testing net (#0) I0412 13:05:26.553263 8032 net.cpp:676] Ignoring source layer train-data I0412 13:05:30.748991 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:05:31.190091 8032 solver.cpp:397] Test net output #0: accuracy = 0.0208333 I0412 13:05:31.190129 8032 solver.cpp:397] Test net output #1: loss = 4.95093 (* 1 = 4.95093 loss) I0412 13:05:31.278234 8032 solver.cpp:218] Iteration 1020 (0.729002 iter/s, 16.4609s/12 iters), loss = 4.79213 I0412 13:05:31.278295 8032 solver.cpp:237] Train net output #0: loss = 4.79213 (* 1 = 4.79213 loss) I0412 13:05:31.278307 8032 sgd_solver.cpp:105] Iteration 1020, lr = 0.00817056 I0412 13:05:35.596974 8032 solver.cpp:218] Iteration 1032 (2.77872 iter/s, 4.31853s/12 iters), loss = 4.87935 I0412 13:05:35.597023 8032 solver.cpp:237] Train net output #0: loss = 4.87935 (* 1 = 4.87935 loss) I0412 13:05:35.597035 8032 sgd_solver.cpp:105] Iteration 1032, lr = 0.00815116 I0412 13:05:40.754048 8032 solver.cpp:218] Iteration 1044 (2.327 iter/s, 5.15685s/12 iters), loss = 4.94582 I0412 13:05:40.754108 8032 solver.cpp:237] Train net output #0: loss = 4.94582 (* 1 = 4.94582 loss) I0412 13:05:40.754122 8032 sgd_solver.cpp:105] Iteration 1044, lr = 0.00813181 I0412 13:05:45.788826 8032 solver.cpp:218] Iteration 1056 (2.38353 iter/s, 5.03455s/12 iters), loss = 4.84512 I0412 13:05:45.788944 8032 solver.cpp:237] Train net output #0: loss = 4.84512 (* 1 = 4.84512 loss) I0412 13:05:45.788954 8032 sgd_solver.cpp:105] Iteration 1056, lr = 0.0081125 I0412 13:05:50.896013 8032 solver.cpp:218] Iteration 1068 (2.34976 iter/s, 5.1069s/12 iters), loss = 4.94456 I0412 13:05:50.896064 8032 solver.cpp:237] Train net output #0: loss = 4.94456 (* 1 = 4.94456 loss) I0412 13:05:50.896076 8032 sgd_solver.cpp:105] Iteration 1068, lr = 0.00809324 I0412 13:05:56.013512 8032 solver.cpp:218] Iteration 1080 (2.345 iter/s, 5.11728s/12 iters), loss = 4.89893 I0412 13:05:56.013566 8032 solver.cpp:237] Train net output #0: loss = 4.89893 (* 1 = 4.89893 loss) I0412 13:05:56.013576 8032 sgd_solver.cpp:105] Iteration 1080, lr = 0.00807403 I0412 13:06:01.071319 8032 solver.cpp:218] Iteration 1092 (2.37267 iter/s, 5.05758s/12 iters), loss = 4.84712 I0412 13:06:01.071377 8032 solver.cpp:237] Train net output #0: loss = 4.84712 (* 1 = 4.84712 loss) I0412 13:06:01.071393 8032 sgd_solver.cpp:105] Iteration 1092, lr = 0.00805486 I0412 13:06:06.156479 8032 solver.cpp:218] Iteration 1104 (2.35991 iter/s, 5.08493s/12 iters), loss = 4.85184 I0412 13:06:06.156522 8032 solver.cpp:237] Train net output #0: loss = 4.85184 (* 1 = 4.85184 loss) I0412 13:06:06.156530 8032 sgd_solver.cpp:105] Iteration 1104, lr = 0.00803573 I0412 13:06:09.259141 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:06:11.131399 8032 solver.cpp:218] Iteration 1116 (2.4122 iter/s, 4.97471s/12 iters), loss = 4.92975 I0412 13:06:11.131458 8032 solver.cpp:237] Train net output #0: loss = 4.92975 (* 1 = 4.92975 loss) I0412 13:06:11.131470 8032 sgd_solver.cpp:105] Iteration 1116, lr = 0.00801666 I0412 13:06:13.164556 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1122.caffemodel I0412 13:06:16.125387 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1122.solverstate I0412 13:06:21.347937 8032 solver.cpp:330] Iteration 1122, Testing net (#0) I0412 13:06:21.347965 8032 net.cpp:676] Ignoring source layer train-data I0412 13:06:25.380223 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:06:25.867929 8032 solver.cpp:397] Test net output #0: accuracy = 0.0300245 I0412 13:06:25.867983 8032 solver.cpp:397] Test net output #1: loss = 4.88031 (* 1 = 4.88031 loss) I0412 13:06:27.737499 8032 solver.cpp:218] Iteration 1128 (0.722651 iter/s, 16.6055s/12 iters), loss = 4.88932 I0412 13:06:27.737542 8032 solver.cpp:237] Train net output #0: loss = 4.88932 (* 1 = 4.88932 loss) I0412 13:06:27.737552 8032 sgd_solver.cpp:105] Iteration 1128, lr = 0.00799762 I0412 13:06:33.000514 8032 solver.cpp:218] Iteration 1140 (2.28016 iter/s, 5.26279s/12 iters), loss = 4.91539 I0412 13:06:33.000560 8032 solver.cpp:237] Train net output #0: loss = 4.91539 (* 1 = 4.91539 loss) I0412 13:06:33.000571 8032 sgd_solver.cpp:105] Iteration 1140, lr = 0.00797863 I0412 13:06:38.099587 8032 solver.cpp:218] Iteration 1152 (2.35347 iter/s, 5.09886s/12 iters), loss = 4.83856 I0412 13:06:38.099637 8032 solver.cpp:237] Train net output #0: loss = 4.83856 (* 1 = 4.83856 loss) I0412 13:06:38.099648 8032 sgd_solver.cpp:105] Iteration 1152, lr = 0.00795969 I0412 13:06:43.109547 8032 solver.cpp:218] Iteration 1164 (2.39533 iter/s, 5.00974s/12 iters), loss = 4.77672 I0412 13:06:43.109601 8032 solver.cpp:237] Train net output #0: loss = 4.77672 (* 1 = 4.77672 loss) I0412 13:06:43.109611 8032 sgd_solver.cpp:105] Iteration 1164, lr = 0.00794079 I0412 13:06:48.133512 8032 solver.cpp:218] Iteration 1176 (2.38866 iter/s, 5.02374s/12 iters), loss = 4.8013 I0412 13:06:48.133671 8032 solver.cpp:237] Train net output #0: loss = 4.8013 (* 1 = 4.8013 loss) I0412 13:06:48.133682 8032 sgd_solver.cpp:105] Iteration 1176, lr = 0.00792194 I0412 13:06:53.555794 8032 solver.cpp:218] Iteration 1188 (2.21323 iter/s, 5.42195s/12 iters), loss = 4.78486 I0412 13:06:53.555838 8032 solver.cpp:237] Train net output #0: loss = 4.78486 (* 1 = 4.78486 loss) I0412 13:06:53.555847 8032 sgd_solver.cpp:105] Iteration 1188, lr = 0.00790313 I0412 13:06:58.674710 8032 solver.cpp:218] Iteration 1200 (2.34435 iter/s, 5.11869s/12 iters), loss = 4.78735 I0412 13:06:58.674765 8032 solver.cpp:237] Train net output #0: loss = 4.78735 (* 1 = 4.78735 loss) I0412 13:06:58.674777 8032 sgd_solver.cpp:105] Iteration 1200, lr = 0.00788437 I0412 13:07:03.752833 8032 solver.cpp:218] Iteration 1212 (2.36318 iter/s, 5.0779s/12 iters), loss = 4.8711 I0412 13:07:03.752883 8032 solver.cpp:237] Train net output #0: loss = 4.8711 (* 1 = 4.8711 loss) I0412 13:07:03.752894 8032 sgd_solver.cpp:105] Iteration 1212, lr = 0.00786565 I0412 13:07:04.031002 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:07:08.412937 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1224.caffemodel I0412 13:07:11.446923 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1224.solverstate I0412 13:07:14.531919 8032 solver.cpp:330] Iteration 1224, Testing net (#0) I0412 13:07:14.531940 8032 net.cpp:676] Ignoring source layer train-data I0412 13:07:18.354262 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:07:18.865804 8032 solver.cpp:397] Test net output #0: accuracy = 0.0349265 I0412 13:07:18.865870 8032 solver.cpp:397] Test net output #1: loss = 4.81447 (* 1 = 4.81447 loss) I0412 13:07:18.954345 8032 solver.cpp:218] Iteration 1224 (0.789423 iter/s, 15.201s/12 iters), loss = 4.77563 I0412 13:07:18.954404 8032 solver.cpp:237] Train net output #0: loss = 4.77563 (* 1 = 4.77563 loss) I0412 13:07:18.954418 8032 sgd_solver.cpp:105] Iteration 1224, lr = 0.00784697 I0412 13:07:23.172227 8032 solver.cpp:218] Iteration 1236 (2.84516 iter/s, 4.21768s/12 iters), loss = 4.90139 I0412 13:07:23.172269 8032 solver.cpp:237] Train net output #0: loss = 4.90139 (* 1 = 4.90139 loss) I0412 13:07:23.172278 8032 sgd_solver.cpp:105] Iteration 1236, lr = 0.00782834 I0412 13:07:28.303530 8032 solver.cpp:218] Iteration 1248 (2.33868 iter/s, 5.13109s/12 iters), loss = 4.65549 I0412 13:07:28.303578 8032 solver.cpp:237] Train net output #0: loss = 4.65549 (* 1 = 4.65549 loss) I0412 13:07:28.303591 8032 sgd_solver.cpp:105] Iteration 1248, lr = 0.00780976 I0412 13:07:33.517220 8032 solver.cpp:218] Iteration 1260 (2.30173 iter/s, 5.21347s/12 iters), loss = 4.71664 I0412 13:07:33.517262 8032 solver.cpp:237] Train net output #0: loss = 4.71664 (* 1 = 4.71664 loss) I0412 13:07:33.517272 8032 sgd_solver.cpp:105] Iteration 1260, lr = 0.00779122 I0412 13:07:38.852314 8032 solver.cpp:218] Iteration 1272 (2.24935 iter/s, 5.33487s/12 iters), loss = 4.74571 I0412 13:07:38.852365 8032 solver.cpp:237] Train net output #0: loss = 4.74571 (* 1 = 4.74571 loss) I0412 13:07:38.852377 8032 sgd_solver.cpp:105] Iteration 1272, lr = 0.00777272 I0412 13:07:44.348006 8032 solver.cpp:218] Iteration 1284 (2.18362 iter/s, 5.49546s/12 iters), loss = 4.77594 I0412 13:07:44.348062 8032 solver.cpp:237] Train net output #0: loss = 4.77594 (* 1 = 4.77594 loss) I0412 13:07:44.348073 8032 sgd_solver.cpp:105] Iteration 1284, lr = 0.00775426 I0412 13:07:49.555094 8032 solver.cpp:218] Iteration 1296 (2.30465 iter/s, 5.20685s/12 iters), loss = 4.51665 I0412 13:07:49.555214 8032 solver.cpp:237] Train net output #0: loss = 4.51665 (* 1 = 4.51665 loss) I0412 13:07:49.555227 8032 sgd_solver.cpp:105] Iteration 1296, lr = 0.00773585 I0412 13:07:55.040943 8032 solver.cpp:218] Iteration 1308 (2.18756 iter/s, 5.48555s/12 iters), loss = 4.60515 I0412 13:07:55.040992 8032 solver.cpp:237] Train net output #0: loss = 4.60515 (* 1 = 4.60515 loss) I0412 13:07:55.041002 8032 sgd_solver.cpp:105] Iteration 1308, lr = 0.00771749 I0412 13:07:57.746807 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:08:00.425029 8032 solver.cpp:218] Iteration 1320 (2.22888 iter/s, 5.38386s/12 iters), loss = 4.62997 I0412 13:08:00.425083 8032 solver.cpp:237] Train net output #0: loss = 4.62997 (* 1 = 4.62997 loss) I0412 13:08:00.425093 8032 sgd_solver.cpp:105] Iteration 1320, lr = 0.00769916 I0412 13:08:02.478305 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1326.caffemodel I0412 13:08:05.532095 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1326.solverstate I0412 13:08:08.985507 8032 solver.cpp:330] Iteration 1326, Testing net (#0) I0412 13:08:08.985534 8032 net.cpp:676] Ignoring source layer train-data I0412 13:08:13.143754 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:08:13.707940 8032 solver.cpp:397] Test net output #0: accuracy = 0.0349265 I0412 13:08:13.707989 8032 solver.cpp:397] Test net output #1: loss = 4.70052 (* 1 = 4.70052 loss) I0412 13:08:15.640480 8032 solver.cpp:218] Iteration 1332 (0.7887 iter/s, 15.2149s/12 iters), loss = 4.56171 I0412 13:08:15.640534 8032 solver.cpp:237] Train net output #0: loss = 4.56171 (* 1 = 4.56171 loss) I0412 13:08:15.640547 8032 sgd_solver.cpp:105] Iteration 1332, lr = 0.00768088 I0412 13:08:20.556267 8032 solver.cpp:218] Iteration 1344 (2.44122 iter/s, 4.91557s/12 iters), loss = 4.5732 I0412 13:08:20.556387 8032 solver.cpp:237] Train net output #0: loss = 4.5732 (* 1 = 4.5732 loss) I0412 13:08:20.556401 8032 sgd_solver.cpp:105] Iteration 1344, lr = 0.00766265 I0412 13:08:25.601815 8032 solver.cpp:218] Iteration 1356 (2.37847 iter/s, 5.04527s/12 iters), loss = 4.69156 I0412 13:08:25.601857 8032 solver.cpp:237] Train net output #0: loss = 4.69156 (* 1 = 4.69156 loss) I0412 13:08:25.601866 8032 sgd_solver.cpp:105] Iteration 1356, lr = 0.00764446 I0412 13:08:30.584084 8032 solver.cpp:218] Iteration 1368 (2.40864 iter/s, 4.98206s/12 iters), loss = 4.54843 I0412 13:08:30.584132 8032 solver.cpp:237] Train net output #0: loss = 4.54843 (* 1 = 4.54843 loss) I0412 13:08:30.584143 8032 sgd_solver.cpp:105] Iteration 1368, lr = 0.00762631 I0412 13:08:31.781409 8032 blocking_queue.cpp:49] Waiting for data I0412 13:08:35.609997 8032 solver.cpp:218] Iteration 1380 (2.38774 iter/s, 5.02567s/12 iters), loss = 4.46416 I0412 13:08:35.610045 8032 solver.cpp:237] Train net output #0: loss = 4.46416 (* 1 = 4.46416 loss) I0412 13:08:35.610054 8032 sgd_solver.cpp:105] Iteration 1380, lr = 0.0076082 I0412 13:08:40.857386 8032 solver.cpp:218] Iteration 1392 (2.28695 iter/s, 5.24716s/12 iters), loss = 4.55988 I0412 13:08:40.857442 8032 solver.cpp:237] Train net output #0: loss = 4.55988 (* 1 = 4.55988 loss) I0412 13:08:40.857455 8032 sgd_solver.cpp:105] Iteration 1392, lr = 0.00759014 I0412 13:08:45.969810 8032 solver.cpp:218] Iteration 1404 (2.34733 iter/s, 5.1122s/12 iters), loss = 4.52012 I0412 13:08:45.969851 8032 solver.cpp:237] Train net output #0: loss = 4.52012 (* 1 = 4.52012 loss) I0412 13:08:45.969861 8032 sgd_solver.cpp:105] Iteration 1404, lr = 0.00757212 I0412 13:08:50.688789 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:08:51.043490 8032 solver.cpp:218] Iteration 1416 (2.36525 iter/s, 5.07347s/12 iters), loss = 4.43863 I0412 13:08:51.043534 8032 solver.cpp:237] Train net output #0: loss = 4.43863 (* 1 = 4.43863 loss) I0412 13:08:51.043543 8032 sgd_solver.cpp:105] Iteration 1416, lr = 0.00755414 I0412 13:08:55.696169 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1428.caffemodel I0412 13:08:58.714083 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1428.solverstate I0412 13:09:01.818372 8032 solver.cpp:330] Iteration 1428, Testing net (#0) I0412 13:09:01.818392 8032 net.cpp:676] Ignoring source layer train-data I0412 13:09:05.737920 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:09:06.359019 8032 solver.cpp:397] Test net output #0: accuracy = 0.0526961 I0412 13:09:06.359047 8032 solver.cpp:397] Test net output #1: loss = 4.51223 (* 1 = 4.51223 loss) I0412 13:09:06.444195 8032 solver.cpp:218] Iteration 1428 (0.779212 iter/s, 15.4002s/12 iters), loss = 4.47687 I0412 13:09:06.444234 8032 solver.cpp:237] Train net output #0: loss = 4.47687 (* 1 = 4.47687 loss) I0412 13:09:06.444244 8032 sgd_solver.cpp:105] Iteration 1428, lr = 0.0075362 I0412 13:09:10.698917 8032 solver.cpp:218] Iteration 1440 (2.82052 iter/s, 4.25454s/12 iters), loss = 4.30954 I0412 13:09:10.698966 8032 solver.cpp:237] Train net output #0: loss = 4.30954 (* 1 = 4.30954 loss) I0412 13:09:10.698977 8032 sgd_solver.cpp:105] Iteration 1440, lr = 0.00751831 I0412 13:09:15.723733 8032 solver.cpp:218] Iteration 1452 (2.38825 iter/s, 5.0246s/12 iters), loss = 4.78505 I0412 13:09:15.723784 8032 solver.cpp:237] Train net output #0: loss = 4.78505 (* 1 = 4.78505 loss) I0412 13:09:15.723796 8032 sgd_solver.cpp:105] Iteration 1452, lr = 0.00750046 I0412 13:09:20.728947 8032 solver.cpp:218] Iteration 1464 (2.3976 iter/s, 5.005s/12 iters), loss = 4.48388 I0412 13:09:20.729097 8032 solver.cpp:237] Train net output #0: loss = 4.48388 (* 1 = 4.48388 loss) I0412 13:09:20.729115 8032 sgd_solver.cpp:105] Iteration 1464, lr = 0.00748265 I0412 13:09:25.806028 8032 solver.cpp:218] Iteration 1476 (2.36371 iter/s, 5.07676s/12 iters), loss = 4.48415 I0412 13:09:25.806094 8032 solver.cpp:237] Train net output #0: loss = 4.48415 (* 1 = 4.48415 loss) I0412 13:09:25.806107 8032 sgd_solver.cpp:105] Iteration 1476, lr = 0.00746489 I0412 13:09:30.892637 8032 solver.cpp:218] Iteration 1488 (2.35924 iter/s, 5.08638s/12 iters), loss = 4.55361 I0412 13:09:30.892689 8032 solver.cpp:237] Train net output #0: loss = 4.55361 (* 1 = 4.55361 loss) I0412 13:09:30.892700 8032 sgd_solver.cpp:105] Iteration 1488, lr = 0.00744716 I0412 13:09:35.946929 8032 solver.cpp:218] Iteration 1500 (2.37432 iter/s, 5.05407s/12 iters), loss = 4.2888 I0412 13:09:35.946981 8032 solver.cpp:237] Train net output #0: loss = 4.2888 (* 1 = 4.2888 loss) I0412 13:09:35.946993 8032 sgd_solver.cpp:105] Iteration 1500, lr = 0.00742948 I0412 13:09:41.025033 8032 solver.cpp:218] Iteration 1512 (2.36319 iter/s, 5.07789s/12 iters), loss = 4.59271 I0412 13:09:41.025074 8032 solver.cpp:237] Train net output #0: loss = 4.59271 (* 1 = 4.59271 loss) I0412 13:09:41.025082 8032 sgd_solver.cpp:105] Iteration 1512, lr = 0.00741184 I0412 13:09:42.792960 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:09:46.044054 8032 solver.cpp:218] Iteration 1524 (2.39101 iter/s, 5.01881s/12 iters), loss = 4.53495 I0412 13:09:46.044098 8032 solver.cpp:237] Train net output #0: loss = 4.53495 (* 1 = 4.53495 loss) I0412 13:09:46.044106 8032 sgd_solver.cpp:105] Iteration 1524, lr = 0.00739425 I0412 13:09:48.110991 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1530.caffemodel I0412 13:09:51.131624 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1530.solverstate I0412 13:09:53.451434 8032 solver.cpp:330] Iteration 1530, Testing net (#0) I0412 13:09:53.451462 8032 net.cpp:676] Ignoring source layer train-data I0412 13:09:57.555150 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:09:58.276401 8032 solver.cpp:397] Test net output #0: accuracy = 0.0569853 I0412 13:09:58.276435 8032 solver.cpp:397] Test net output #1: loss = 4.54833 (* 1 = 4.54833 loss) I0412 13:10:00.282917 8032 solver.cpp:218] Iteration 1536 (0.842794 iter/s, 14.2384s/12 iters), loss = 4.44751 I0412 13:10:00.282968 8032 solver.cpp:237] Train net output #0: loss = 4.44751 (* 1 = 4.44751 loss) I0412 13:10:00.282979 8032 sgd_solver.cpp:105] Iteration 1536, lr = 0.00737669 I0412 13:10:05.400028 8032 solver.cpp:218] Iteration 1548 (2.34518 iter/s, 5.11688s/12 iters), loss = 4.10244 I0412 13:10:05.400081 8032 solver.cpp:237] Train net output #0: loss = 4.10244 (* 1 = 4.10244 loss) I0412 13:10:05.400094 8032 sgd_solver.cpp:105] Iteration 1548, lr = 0.00735918 I0412 13:10:10.569984 8032 solver.cpp:218] Iteration 1560 (2.32121 iter/s, 5.16971s/12 iters), loss = 4.43521 I0412 13:10:10.570050 8032 solver.cpp:237] Train net output #0: loss = 4.43521 (* 1 = 4.43521 loss) I0412 13:10:10.570067 8032 sgd_solver.cpp:105] Iteration 1560, lr = 0.00734171 I0412 13:10:15.536999 8032 solver.cpp:218] Iteration 1572 (2.41605 iter/s, 4.96679s/12 iters), loss = 4.49348 I0412 13:10:15.537043 8032 solver.cpp:237] Train net output #0: loss = 4.49348 (* 1 = 4.49348 loss) I0412 13:10:15.537053 8032 sgd_solver.cpp:105] Iteration 1572, lr = 0.00732427 I0412 13:10:20.446377 8032 solver.cpp:218] Iteration 1584 (2.44441 iter/s, 4.90917s/12 iters), loss = 4.35628 I0412 13:10:20.446429 8032 solver.cpp:237] Train net output #0: loss = 4.35628 (* 1 = 4.35628 loss) I0412 13:10:20.446440 8032 sgd_solver.cpp:105] Iteration 1584, lr = 0.00730688 I0412 13:10:25.698385 8032 solver.cpp:218] Iteration 1596 (2.28494 iter/s, 5.25178s/12 iters), loss = 4.43161 I0412 13:10:25.698529 8032 solver.cpp:237] Train net output #0: loss = 4.43161 (* 1 = 4.43161 loss) I0412 13:10:25.698541 8032 sgd_solver.cpp:105] Iteration 1596, lr = 0.00728954 I0412 13:10:30.762184 8032 solver.cpp:218] Iteration 1608 (2.36991 iter/s, 5.06349s/12 iters), loss = 4.32174 I0412 13:10:30.762238 8032 solver.cpp:237] Train net output #0: loss = 4.32174 (* 1 = 4.32174 loss) I0412 13:10:30.762249 8032 sgd_solver.cpp:105] Iteration 1608, lr = 0.00727223 I0412 13:10:34.772616 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:10:35.872375 8032 solver.cpp:218] Iteration 1620 (2.34835 iter/s, 5.10996s/12 iters), loss = 4.23378 I0412 13:10:35.872428 8032 solver.cpp:237] Train net output #0: loss = 4.23378 (* 1 = 4.23378 loss) I0412 13:10:35.872440 8032 sgd_solver.cpp:105] Iteration 1620, lr = 0.00725496 I0412 13:10:40.427394 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1632.caffemodel I0412 13:10:44.553946 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1632.solverstate I0412 13:10:48.553676 8032 solver.cpp:330] Iteration 1632, Testing net (#0) I0412 13:10:48.553704 8032 net.cpp:676] Ignoring source layer train-data I0412 13:10:52.389621 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:10:53.060293 8032 solver.cpp:397] Test net output #0: accuracy = 0.0661765 I0412 13:10:53.060346 8032 solver.cpp:397] Test net output #1: loss = 4.33669 (* 1 = 4.33669 loss) I0412 13:10:53.148450 8032 solver.cpp:218] Iteration 1632 (0.694626 iter/s, 17.2755s/12 iters), loss = 4.1539 I0412 13:10:53.148501 8032 solver.cpp:237] Train net output #0: loss = 4.1539 (* 1 = 4.1539 loss) I0412 13:10:53.148511 8032 sgd_solver.cpp:105] Iteration 1632, lr = 0.00723774 I0412 13:10:57.596278 8032 solver.cpp:218] Iteration 1644 (2.69807 iter/s, 4.44763s/12 iters), loss = 4.32828 I0412 13:10:57.596371 8032 solver.cpp:237] Train net output #0: loss = 4.32828 (* 1 = 4.32828 loss) I0412 13:10:57.596382 8032 sgd_solver.cpp:105] Iteration 1644, lr = 0.00722056 I0412 13:11:02.708993 8032 solver.cpp:218] Iteration 1656 (2.34721 iter/s, 5.11245s/12 iters), loss = 4.3165 I0412 13:11:02.709043 8032 solver.cpp:237] Train net output #0: loss = 4.3165 (* 1 = 4.3165 loss) I0412 13:11:02.709053 8032 sgd_solver.cpp:105] Iteration 1656, lr = 0.00720341 I0412 13:11:07.661096 8032 solver.cpp:218] Iteration 1668 (2.42332 iter/s, 4.95188s/12 iters), loss = 4.08486 I0412 13:11:07.661147 8032 solver.cpp:237] Train net output #0: loss = 4.08486 (* 1 = 4.08486 loss) I0412 13:11:07.661159 8032 sgd_solver.cpp:105] Iteration 1668, lr = 0.00718631 I0412 13:11:12.759428 8032 solver.cpp:218] Iteration 1680 (2.35381 iter/s, 5.09811s/12 iters), loss = 4.2563 I0412 13:11:12.759479 8032 solver.cpp:237] Train net output #0: loss = 4.2563 (* 1 = 4.2563 loss) I0412 13:11:12.759490 8032 sgd_solver.cpp:105] Iteration 1680, lr = 0.00716925 I0412 13:11:17.829852 8032 solver.cpp:218] Iteration 1692 (2.36677 iter/s, 5.0702s/12 iters), loss = 4.28217 I0412 13:11:17.829903 8032 solver.cpp:237] Train net output #0: loss = 4.28217 (* 1 = 4.28217 loss) I0412 13:11:17.829914 8032 sgd_solver.cpp:105] Iteration 1692, lr = 0.00715223 I0412 13:11:22.938131 8032 solver.cpp:218] Iteration 1704 (2.34923 iter/s, 5.10806s/12 iters), loss = 3.8998 I0412 13:11:22.938179 8032 solver.cpp:237] Train net output #0: loss = 3.8998 (* 1 = 3.8998 loss) I0412 13:11:22.938189 8032 sgd_solver.cpp:105] Iteration 1704, lr = 0.00713525 I0412 13:11:27.989006 8032 solver.cpp:218] Iteration 1716 (2.37593 iter/s, 5.05066s/12 iters), loss = 4.12206 I0412 13:11:27.989158 8032 solver.cpp:237] Train net output #0: loss = 4.12206 (* 1 = 4.12206 loss) I0412 13:11:27.989172 8032 sgd_solver.cpp:105] Iteration 1716, lr = 0.00711831 I0412 13:11:29.067955 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:11:33.094873 8032 solver.cpp:218] Iteration 1728 (2.35038 iter/s, 5.10555s/12 iters), loss = 4.12931 I0412 13:11:33.094928 8032 solver.cpp:237] Train net output #0: loss = 4.12931 (* 1 = 4.12931 loss) I0412 13:11:33.094940 8032 sgd_solver.cpp:105] Iteration 1728, lr = 0.00710141 I0412 13:11:35.185554 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1734.caffemodel I0412 13:11:40.792161 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1734.solverstate I0412 13:11:45.476166 8032 solver.cpp:330] Iteration 1734, Testing net (#0) I0412 13:11:45.476194 8032 net.cpp:676] Ignoring source layer train-data I0412 13:11:49.317521 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:11:50.025768 8032 solver.cpp:397] Test net output #0: accuracy = 0.0759804 I0412 13:11:50.025808 8032 solver.cpp:397] Test net output #1: loss = 4.33976 (* 1 = 4.33976 loss) I0412 13:11:51.790113 8032 solver.cpp:218] Iteration 1740 (0.641897 iter/s, 18.6946s/12 iters), loss = 4.01946 I0412 13:11:51.790158 8032 solver.cpp:237] Train net output #0: loss = 4.01946 (* 1 = 4.01946 loss) I0412 13:11:51.790166 8032 sgd_solver.cpp:105] Iteration 1740, lr = 0.00708455 I0412 13:11:56.839299 8032 solver.cpp:218] Iteration 1752 (2.37672 iter/s, 5.04897s/12 iters), loss = 4.07616 I0412 13:11:56.839354 8032 solver.cpp:237] Train net output #0: loss = 4.07616 (* 1 = 4.07616 loss) I0412 13:11:56.839365 8032 sgd_solver.cpp:105] Iteration 1752, lr = 0.00706773 I0412 13:12:01.917241 8032 solver.cpp:218] Iteration 1764 (2.36327 iter/s, 5.07771s/12 iters), loss = 4.21458 I0412 13:12:01.917354 8032 solver.cpp:237] Train net output #0: loss = 4.21458 (* 1 = 4.21458 loss) I0412 13:12:01.917366 8032 sgd_solver.cpp:105] Iteration 1764, lr = 0.00705094 I0412 13:12:07.151803 8032 solver.cpp:218] Iteration 1776 (2.29258 iter/s, 5.23428s/12 iters), loss = 4.30918 I0412 13:12:07.151845 8032 solver.cpp:237] Train net output #0: loss = 4.30918 (* 1 = 4.30918 loss) I0412 13:12:07.151854 8032 sgd_solver.cpp:105] Iteration 1776, lr = 0.0070342 I0412 13:12:12.197059 8032 solver.cpp:218] Iteration 1788 (2.37857 iter/s, 5.04504s/12 iters), loss = 4.21914 I0412 13:12:12.197111 8032 solver.cpp:237] Train net output #0: loss = 4.21914 (* 1 = 4.21914 loss) I0412 13:12:12.197124 8032 sgd_solver.cpp:105] Iteration 1788, lr = 0.0070175 I0412 13:12:17.140013 8032 solver.cpp:218] Iteration 1800 (2.42781 iter/s, 4.94273s/12 iters), loss = 4.05374 I0412 13:12:17.140065 8032 solver.cpp:237] Train net output #0: loss = 4.05374 (* 1 = 4.05374 loss) I0412 13:12:17.140075 8032 sgd_solver.cpp:105] Iteration 1800, lr = 0.00700084 I0412 13:12:22.246701 8032 solver.cpp:218] Iteration 1812 (2.34996 iter/s, 5.10647s/12 iters), loss = 4.12332 I0412 13:12:22.246747 8032 solver.cpp:237] Train net output #0: loss = 4.12332 (* 1 = 4.12332 loss) I0412 13:12:22.246757 8032 sgd_solver.cpp:105] Iteration 1812, lr = 0.00698422 I0412 13:12:25.481926 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:12:27.361241 8032 solver.cpp:218] Iteration 1824 (2.34635 iter/s, 5.11432s/12 iters), loss = 4.1738 I0412 13:12:27.361287 8032 solver.cpp:237] Train net output #0: loss = 4.1738 (* 1 = 4.1738 loss) I0412 13:12:27.361296 8032 sgd_solver.cpp:105] Iteration 1824, lr = 0.00696764 I0412 13:12:32.019724 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1836.caffemodel I0412 13:12:36.772804 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1836.solverstate I0412 13:12:40.059373 8032 solver.cpp:330] Iteration 1836, Testing net (#0) I0412 13:12:40.059401 8032 net.cpp:676] Ignoring source layer train-data I0412 13:12:43.949417 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:12:44.696799 8032 solver.cpp:397] Test net output #0: accuracy = 0.0778186 I0412 13:12:44.696837 8032 solver.cpp:397] Test net output #1: loss = 4.27359 (* 1 = 4.27359 loss) I0412 13:12:44.784814 8032 solver.cpp:218] Iteration 1836 (0.688746 iter/s, 17.423s/12 iters), loss = 4.2145 I0412 13:12:44.784862 8032 solver.cpp:237] Train net output #0: loss = 4.2145 (* 1 = 4.2145 loss) I0412 13:12:44.784869 8032 sgd_solver.cpp:105] Iteration 1836, lr = 0.0069511 I0412 13:12:48.996047 8032 solver.cpp:218] Iteration 1848 (2.84966 iter/s, 4.21103s/12 iters), loss = 4.21158 I0412 13:12:48.996102 8032 solver.cpp:237] Train net output #0: loss = 4.21158 (* 1 = 4.21158 loss) I0412 13:12:48.996114 8032 sgd_solver.cpp:105] Iteration 1848, lr = 0.00693459 I0412 13:12:54.066890 8032 solver.cpp:218] Iteration 1860 (2.36658 iter/s, 5.07062s/12 iters), loss = 4.13306 I0412 13:12:54.066946 8032 solver.cpp:237] Train net output #0: loss = 4.13306 (* 1 = 4.13306 loss) I0412 13:12:54.066958 8032 sgd_solver.cpp:105] Iteration 1860, lr = 0.00691813 I0412 13:12:59.283342 8032 solver.cpp:218] Iteration 1872 (2.30052 iter/s, 5.21622s/12 iters), loss = 4.13747 I0412 13:12:59.283393 8032 solver.cpp:237] Train net output #0: loss = 4.13747 (* 1 = 4.13747 loss) I0412 13:12:59.283404 8032 sgd_solver.cpp:105] Iteration 1872, lr = 0.0069017 I0412 13:13:04.360867 8032 solver.cpp:218] Iteration 1884 (2.36346 iter/s, 5.07731s/12 iters), loss = 4.11452 I0412 13:13:04.360949 8032 solver.cpp:237] Train net output #0: loss = 4.11452 (* 1 = 4.11452 loss) I0412 13:13:04.360960 8032 sgd_solver.cpp:105] Iteration 1884, lr = 0.00688532 I0412 13:13:09.478262 8032 solver.cpp:218] Iteration 1896 (2.34506 iter/s, 5.11714s/12 iters), loss = 4.08499 I0412 13:13:09.478315 8032 solver.cpp:237] Train net output #0: loss = 4.08499 (* 1 = 4.08499 loss) I0412 13:13:09.478327 8032 sgd_solver.cpp:105] Iteration 1896, lr = 0.00686897 I0412 13:13:14.457105 8032 solver.cpp:218] Iteration 1908 (2.4103 iter/s, 4.97862s/12 iters), loss = 4.13032 I0412 13:13:14.457145 8032 solver.cpp:237] Train net output #0: loss = 4.13032 (* 1 = 4.13032 loss) I0412 13:13:14.457154 8032 sgd_solver.cpp:105] Iteration 1908, lr = 0.00685266 I0412 13:13:19.676627 8032 solver.cpp:218] Iteration 1920 (2.29916 iter/s, 5.2193s/12 iters), loss = 3.86511 I0412 13:13:19.676685 8032 solver.cpp:237] Train net output #0: loss = 3.86511 (* 1 = 3.86511 loss) I0412 13:13:19.676697 8032 sgd_solver.cpp:105] Iteration 1920, lr = 0.00683639 I0412 13:13:20.014505 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:13:24.937402 8032 solver.cpp:218] Iteration 1932 (2.28113 iter/s, 5.26054s/12 iters), loss = 4.09307 I0412 13:13:24.937448 8032 solver.cpp:237] Train net output #0: loss = 4.09307 (* 1 = 4.09307 loss) I0412 13:13:24.937459 8032 sgd_solver.cpp:105] Iteration 1932, lr = 0.00682016 I0412 13:13:27.070981 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1938.caffemodel I0412 13:13:34.625907 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1938.solverstate I0412 13:13:45.391598 8032 solver.cpp:330] Iteration 1938, Testing net (#0) I0412 13:13:45.391628 8032 net.cpp:676] Ignoring source layer train-data I0412 13:13:49.172957 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:13:49.959863 8032 solver.cpp:397] Test net output #0: accuracy = 0.109069 I0412 13:13:49.959897 8032 solver.cpp:397] Test net output #1: loss = 4.0387 (* 1 = 4.0387 loss) I0412 13:13:51.871666 8032 solver.cpp:218] Iteration 1944 (0.445544 iter/s, 26.9334s/12 iters), loss = 3.87122 I0412 13:13:51.871716 8032 solver.cpp:237] Train net output #0: loss = 3.87122 (* 1 = 3.87122 loss) I0412 13:13:51.871727 8032 sgd_solver.cpp:105] Iteration 1944, lr = 0.00680397 I0412 13:13:56.959084 8032 solver.cpp:218] Iteration 1956 (2.35886 iter/s, 5.0872s/12 iters), loss = 3.76589 I0412 13:13:56.959144 8032 solver.cpp:237] Train net output #0: loss = 3.76589 (* 1 = 3.76589 loss) I0412 13:13:56.959161 8032 sgd_solver.cpp:105] Iteration 1956, lr = 0.00678782 I0412 13:14:02.168911 8032 solver.cpp:218] Iteration 1968 (2.30344 iter/s, 5.20959s/12 iters), loss = 3.8276 I0412 13:14:02.168959 8032 solver.cpp:237] Train net output #0: loss = 3.8276 (* 1 = 3.8276 loss) I0412 13:14:02.168969 8032 sgd_solver.cpp:105] Iteration 1968, lr = 0.0067717 I0412 13:14:07.315536 8032 solver.cpp:218] Iteration 1980 (2.33172 iter/s, 5.14641s/12 iters), loss = 3.96816 I0412 13:14:07.315634 8032 solver.cpp:237] Train net output #0: loss = 3.96816 (* 1 = 3.96816 loss) I0412 13:14:07.315644 8032 sgd_solver.cpp:105] Iteration 1980, lr = 0.00675562 I0412 13:14:12.419891 8032 solver.cpp:218] Iteration 1992 (2.35106 iter/s, 5.10408s/12 iters), loss = 3.91976 I0412 13:14:12.419945 8032 solver.cpp:237] Train net output #0: loss = 3.91976 (* 1 = 3.91976 loss) I0412 13:14:12.419955 8032 sgd_solver.cpp:105] Iteration 1992, lr = 0.00673958 I0412 13:14:17.494855 8032 solver.cpp:218] Iteration 2004 (2.36465 iter/s, 5.07474s/12 iters), loss = 3.7267 I0412 13:14:17.494904 8032 solver.cpp:237] Train net output #0: loss = 3.7267 (* 1 = 3.7267 loss) I0412 13:14:17.494916 8032 sgd_solver.cpp:105] Iteration 2004, lr = 0.00672358 I0412 13:14:22.595520 8032 solver.cpp:218] Iteration 2016 (2.35274 iter/s, 5.10044s/12 iters), loss = 3.92854 I0412 13:14:22.595562 8032 solver.cpp:237] Train net output #0: loss = 3.92854 (* 1 = 3.92854 loss) I0412 13:14:22.595571 8032 sgd_solver.cpp:105] Iteration 2016, lr = 0.00670762 I0412 13:14:25.216336 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:14:27.764160 8032 solver.cpp:218] Iteration 2028 (2.32179 iter/s, 5.16842s/12 iters), loss = 3.81987 I0412 13:14:27.764210 8032 solver.cpp:237] Train net output #0: loss = 3.81987 (* 1 = 3.81987 loss) I0412 13:14:27.764221 8032 sgd_solver.cpp:105] Iteration 2028, lr = 0.00669169 I0412 13:14:32.333520 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2040.caffemodel I0412 13:14:39.525389 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2040.solverstate I0412 13:14:42.854260 8032 solver.cpp:330] Iteration 2040, Testing net (#0) I0412 13:14:42.854288 8032 net.cpp:676] Ignoring source layer train-data I0412 13:14:47.294237 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:14:48.159199 8032 solver.cpp:397] Test net output #0: accuracy = 0.114583 I0412 13:14:48.159233 8032 solver.cpp:397] Test net output #1: loss = 3.94061 (* 1 = 3.94061 loss) I0412 13:14:48.247457 8032 solver.cpp:218] Iteration 2040 (0.585863 iter/s, 20.4826s/12 iters), loss = 3.50446 I0412 13:14:48.247498 8032 solver.cpp:237] Train net output #0: loss = 3.50446 (* 1 = 3.50446 loss) I0412 13:14:48.247507 8032 sgd_solver.cpp:105] Iteration 2040, lr = 0.00667581 I0412 13:14:52.325502 8032 solver.cpp:218] Iteration 2052 (2.94272 iter/s, 4.07786s/12 iters), loss = 3.90014 I0412 13:14:52.325554 8032 solver.cpp:237] Train net output #0: loss = 3.90014 (* 1 = 3.90014 loss) I0412 13:14:52.325567 8032 sgd_solver.cpp:105] Iteration 2052, lr = 0.00665996 I0412 13:14:53.927244 8032 blocking_queue.cpp:49] Waiting for data I0412 13:14:57.528520 8032 solver.cpp:218] Iteration 2064 (2.30645 iter/s, 5.20279s/12 iters), loss = 3.75183 I0412 13:14:57.528564 8032 solver.cpp:237] Train net output #0: loss = 3.75183 (* 1 = 3.75183 loss) I0412 13:14:57.528573 8032 sgd_solver.cpp:105] Iteration 2064, lr = 0.00664414 I0412 13:15:02.777305 8032 solver.cpp:218] Iteration 2076 (2.28634 iter/s, 5.24856s/12 iters), loss = 3.78889 I0412 13:15:02.777362 8032 solver.cpp:237] Train net output #0: loss = 3.78889 (* 1 = 3.78889 loss) I0412 13:15:02.777374 8032 sgd_solver.cpp:105] Iteration 2076, lr = 0.00662837 I0412 13:15:07.870055 8032 solver.cpp:218] Iteration 2088 (2.35639 iter/s, 5.09253s/12 iters), loss = 3.81699 I0412 13:15:07.870100 8032 solver.cpp:237] Train net output #0: loss = 3.81699 (* 1 = 3.81699 loss) I0412 13:15:07.870111 8032 sgd_solver.cpp:105] Iteration 2088, lr = 0.00661263 I0412 13:15:12.878357 8032 solver.cpp:218] Iteration 2100 (2.39612 iter/s, 5.00809s/12 iters), loss = 3.86132 I0412 13:15:12.878470 8032 solver.cpp:237] Train net output #0: loss = 3.86132 (* 1 = 3.86132 loss) I0412 13:15:12.878479 8032 sgd_solver.cpp:105] Iteration 2100, lr = 0.00659693 I0412 13:15:17.987522 8032 solver.cpp:218] Iteration 2112 (2.34885 iter/s, 5.10888s/12 iters), loss = 3.78934 I0412 13:15:17.987562 8032 solver.cpp:237] Train net output #0: loss = 3.78934 (* 1 = 3.78934 loss) I0412 13:15:17.987571 8032 sgd_solver.cpp:105] Iteration 2112, lr = 0.00658127 I0412 13:15:22.896342 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:15:23.242063 8032 solver.cpp:218] Iteration 2124 (2.28384 iter/s, 5.25432s/12 iters), loss = 3.66652 I0412 13:15:23.242125 8032 solver.cpp:237] Train net output #0: loss = 3.66652 (* 1 = 3.66652 loss) I0412 13:15:23.242139 8032 sgd_solver.cpp:105] Iteration 2124, lr = 0.00656564 I0412 13:15:28.649684 8032 solver.cpp:218] Iteration 2136 (2.21919 iter/s, 5.40738s/12 iters), loss = 3.80894 I0412 13:15:28.649729 8032 solver.cpp:237] Train net output #0: loss = 3.80894 (* 1 = 3.80894 loss) I0412 13:15:28.649737 8032 sgd_solver.cpp:105] Iteration 2136, lr = 0.00655006 I0412 13:15:30.660553 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2142.caffemodel I0412 13:15:39.759253 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2142.solverstate I0412 13:15:42.108490 8032 solver.cpp:330] Iteration 2142, Testing net (#0) I0412 13:15:42.108511 8032 net.cpp:676] Ignoring source layer train-data I0412 13:15:45.888012 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:15:46.788237 8032 solver.cpp:397] Test net output #0: accuracy = 0.122549 I0412 13:15:46.788267 8032 solver.cpp:397] Test net output #1: loss = 3.89458 (* 1 = 3.89458 loss) I0412 13:15:48.778371 8032 solver.cpp:218] Iteration 2148 (0.596184 iter/s, 20.128s/12 iters), loss = 3.59458 I0412 13:15:48.778426 8032 solver.cpp:237] Train net output #0: loss = 3.59458 (* 1 = 3.59458 loss) I0412 13:15:48.778436 8032 sgd_solver.cpp:105] Iteration 2148, lr = 0.00653451 I0412 13:15:53.799844 8032 solver.cpp:218] Iteration 2160 (2.38985 iter/s, 5.02124s/12 iters), loss = 3.97154 I0412 13:15:53.799897 8032 solver.cpp:237] Train net output #0: loss = 3.97154 (* 1 = 3.97154 loss) I0412 13:15:53.799909 8032 sgd_solver.cpp:105] Iteration 2160, lr = 0.00651899 I0412 13:15:59.023932 8032 solver.cpp:218] Iteration 2172 (2.29715 iter/s, 5.22386s/12 iters), loss = 3.8417 I0412 13:15:59.023969 8032 solver.cpp:237] Train net output #0: loss = 3.8417 (* 1 = 3.8417 loss) I0412 13:15:59.023978 8032 sgd_solver.cpp:105] Iteration 2172, lr = 0.00650351 I0412 13:16:04.429761 8032 solver.cpp:218] Iteration 2184 (2.21992 iter/s, 5.4056s/12 iters), loss = 3.74603 I0412 13:16:04.429818 8032 solver.cpp:237] Train net output #0: loss = 3.74603 (* 1 = 3.74603 loss) I0412 13:16:04.429829 8032 sgd_solver.cpp:105] Iteration 2184, lr = 0.00648807 I0412 13:16:09.455029 8032 solver.cpp:218] Iteration 2196 (2.38804 iter/s, 5.02504s/12 iters), loss = 3.86774 I0412 13:16:09.455071 8032 solver.cpp:237] Train net output #0: loss = 3.86774 (* 1 = 3.86774 loss) I0412 13:16:09.455080 8032 sgd_solver.cpp:105] Iteration 2196, lr = 0.00647267 I0412 13:16:14.516216 8032 solver.cpp:218] Iteration 2208 (2.37109 iter/s, 5.06097s/12 iters), loss = 3.44665 I0412 13:16:14.516260 8032 solver.cpp:237] Train net output #0: loss = 3.44665 (* 1 = 3.44665 loss) I0412 13:16:14.516269 8032 sgd_solver.cpp:105] Iteration 2208, lr = 0.0064573 I0412 13:16:19.616017 8032 solver.cpp:218] Iteration 2220 (2.35313 iter/s, 5.09959s/12 iters), loss = 3.50992 I0412 13:16:19.616124 8032 solver.cpp:237] Train net output #0: loss = 3.50992 (* 1 = 3.50992 loss) I0412 13:16:19.616133 8032 sgd_solver.cpp:105] Iteration 2220, lr = 0.00644197 I0412 13:16:21.464926 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:16:24.940572 8032 solver.cpp:218] Iteration 2232 (2.25383 iter/s, 5.32427s/12 iters), loss = 3.71471 I0412 13:16:24.940613 8032 solver.cpp:237] Train net output #0: loss = 3.71471 (* 1 = 3.71471 loss) I0412 13:16:24.940623 8032 sgd_solver.cpp:105] Iteration 2232, lr = 0.00642668 I0412 13:16:29.560240 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2244.caffemodel I0412 13:16:38.386909 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2244.solverstate I0412 13:16:41.138708 8032 solver.cpp:330] Iteration 2244, Testing net (#0) I0412 13:16:41.138734 8032 net.cpp:676] Ignoring source layer train-data I0412 13:16:44.676457 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:16:45.585651 8032 solver.cpp:397] Test net output #0: accuracy = 0.141544 I0412 13:16:45.585698 8032 solver.cpp:397] Test net output #1: loss = 3.81237 (* 1 = 3.81237 loss) I0412 13:16:45.673840 8032 solver.cpp:218] Iteration 2244 (0.578799 iter/s, 20.7326s/12 iters), loss = 3.73727 I0412 13:16:45.673897 8032 solver.cpp:237] Train net output #0: loss = 3.73727 (* 1 = 3.73727 loss) I0412 13:16:45.673907 8032 sgd_solver.cpp:105] Iteration 2244, lr = 0.00641142 I0412 13:16:50.060482 8032 solver.cpp:218] Iteration 2256 (2.73571 iter/s, 4.38644s/12 iters), loss = 3.20893 I0412 13:16:50.060590 8032 solver.cpp:237] Train net output #0: loss = 3.20893 (* 1 = 3.20893 loss) I0412 13:16:50.060601 8032 sgd_solver.cpp:105] Iteration 2256, lr = 0.0063962 I0412 13:16:55.046785 8032 solver.cpp:218] Iteration 2268 (2.40672 iter/s, 4.98603s/12 iters), loss = 3.69013 I0412 13:16:55.046825 8032 solver.cpp:237] Train net output #0: loss = 3.69013 (* 1 = 3.69013 loss) I0412 13:16:55.046834 8032 sgd_solver.cpp:105] Iteration 2268, lr = 0.00638101 I0412 13:17:00.190827 8032 solver.cpp:218] Iteration 2280 (2.33289 iter/s, 5.14383s/12 iters), loss = 3.63568 I0412 13:17:00.190882 8032 solver.cpp:237] Train net output #0: loss = 3.63568 (* 1 = 3.63568 loss) I0412 13:17:00.190898 8032 sgd_solver.cpp:105] Iteration 2280, lr = 0.00636586 I0412 13:17:05.240708 8032 solver.cpp:218] Iteration 2292 (2.3764 iter/s, 5.04966s/12 iters), loss = 3.47965 I0412 13:17:05.240754 8032 solver.cpp:237] Train net output #0: loss = 3.47965 (* 1 = 3.47965 loss) I0412 13:17:05.240764 8032 sgd_solver.cpp:105] Iteration 2292, lr = 0.00635075 I0412 13:17:10.360602 8032 solver.cpp:218] Iteration 2304 (2.3439 iter/s, 5.11967s/12 iters), loss = 3.72957 I0412 13:17:10.360666 8032 solver.cpp:237] Train net output #0: loss = 3.72957 (* 1 = 3.72957 loss) I0412 13:17:10.360680 8032 sgd_solver.cpp:105] Iteration 2304, lr = 0.00633567 I0412 13:17:15.627494 8032 solver.cpp:218] Iteration 2316 (2.27849 iter/s, 5.26666s/12 iters), loss = 3.35113 I0412 13:17:15.627539 8032 solver.cpp:237] Train net output #0: loss = 3.35113 (* 1 = 3.35113 loss) I0412 13:17:15.627548 8032 sgd_solver.cpp:105] Iteration 2316, lr = 0.00632063 I0412 13:17:19.551343 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:17:20.631309 8032 solver.cpp:218] Iteration 2328 (2.39827 iter/s, 5.0036s/12 iters), loss = 3.45853 I0412 13:17:20.631426 8032 solver.cpp:237] Train net output #0: loss = 3.45853 (* 1 = 3.45853 loss) I0412 13:17:20.631436 8032 sgd_solver.cpp:105] Iteration 2328, lr = 0.00630562 I0412 13:17:25.623149 8032 solver.cpp:218] Iteration 2340 (2.40406 iter/s, 4.99156s/12 iters), loss = 3.40605 I0412 13:17:25.623191 8032 solver.cpp:237] Train net output #0: loss = 3.40605 (* 1 = 3.40605 loss) I0412 13:17:25.623200 8032 sgd_solver.cpp:105] Iteration 2340, lr = 0.00629065 I0412 13:17:27.733690 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2346.caffemodel I0412 13:17:34.366202 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2346.solverstate I0412 13:17:36.762137 8032 solver.cpp:330] Iteration 2346, Testing net (#0) I0412 13:17:36.762162 8032 net.cpp:676] Ignoring source layer train-data I0412 13:17:40.266531 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:17:41.210364 8032 solver.cpp:397] Test net output #0: accuracy = 0.150735 I0412 13:17:41.210413 8032 solver.cpp:397] Test net output #1: loss = 3.80982 (* 1 = 3.80982 loss) I0412 13:17:43.169842 8032 solver.cpp:218] Iteration 2352 (0.683913 iter/s, 17.5461s/12 iters), loss = 3.54456 I0412 13:17:43.169910 8032 solver.cpp:237] Train net output #0: loss = 3.54456 (* 1 = 3.54456 loss) I0412 13:17:43.169926 8032 sgd_solver.cpp:105] Iteration 2352, lr = 0.00627571 I0412 13:17:48.246989 8032 solver.cpp:218] Iteration 2364 (2.36364 iter/s, 5.07691s/12 iters), loss = 3.32086 I0412 13:17:48.247036 8032 solver.cpp:237] Train net output #0: loss = 3.32086 (* 1 = 3.32086 loss) I0412 13:17:48.247045 8032 sgd_solver.cpp:105] Iteration 2364, lr = 0.00626081 I0412 13:17:53.280594 8032 solver.cpp:218] Iteration 2376 (2.38408 iter/s, 5.03339s/12 iters), loss = 3.31129 I0412 13:17:53.286059 8032 solver.cpp:237] Train net output #0: loss = 3.31129 (* 1 = 3.31129 loss) I0412 13:17:53.286079 8032 sgd_solver.cpp:105] Iteration 2376, lr = 0.00624595 I0412 13:17:58.310544 8032 solver.cpp:218] Iteration 2388 (2.38838 iter/s, 5.02433s/12 iters), loss = 3.50438 I0412 13:17:58.310591 8032 solver.cpp:237] Train net output #0: loss = 3.50438 (* 1 = 3.50438 loss) I0412 13:17:58.310602 8032 sgd_solver.cpp:105] Iteration 2388, lr = 0.00623112 I0412 13:18:03.442296 8032 solver.cpp:218] Iteration 2400 (2.33848 iter/s, 5.13154s/12 iters), loss = 3.18178 I0412 13:18:03.442338 8032 solver.cpp:237] Train net output #0: loss = 3.18178 (* 1 = 3.18178 loss) I0412 13:18:03.442348 8032 sgd_solver.cpp:105] Iteration 2400, lr = 0.00621633 I0412 13:18:08.503615 8032 solver.cpp:218] Iteration 2412 (2.37102 iter/s, 5.0611s/12 iters), loss = 3.12125 I0412 13:18:08.503659 8032 solver.cpp:237] Train net output #0: loss = 3.12125 (* 1 = 3.12125 loss) I0412 13:18:08.503669 8032 sgd_solver.cpp:105] Iteration 2412, lr = 0.00620157 I0412 13:18:13.466953 8032 solver.cpp:218] Iteration 2424 (2.41783 iter/s, 4.96313s/12 iters), loss = 3.28977 I0412 13:18:13.466996 8032 solver.cpp:237] Train net output #0: loss = 3.28977 (* 1 = 3.28977 loss) I0412 13:18:13.467005 8032 sgd_solver.cpp:105] Iteration 2424, lr = 0.00618684 I0412 13:18:14.537755 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:18:18.642717 8032 solver.cpp:218] Iteration 2436 (2.3186 iter/s, 5.17555s/12 iters), loss = 3.33779 I0412 13:18:18.642761 8032 solver.cpp:237] Train net output #0: loss = 3.33779 (* 1 = 3.33779 loss) I0412 13:18:18.642771 8032 sgd_solver.cpp:105] Iteration 2436, lr = 0.00617215 I0412 13:18:23.312038 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2448.caffemodel I0412 13:18:29.868366 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2448.solverstate I0412 13:18:35.054368 8032 solver.cpp:330] Iteration 2448, Testing net (#0) I0412 13:18:35.054391 8032 net.cpp:676] Ignoring source layer train-data I0412 13:18:38.455382 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:18:39.440073 8032 solver.cpp:397] Test net output #0: accuracy = 0.166054 I0412 13:18:39.440112 8032 solver.cpp:397] Test net output #1: loss = 3.7174 (* 1 = 3.7174 loss) I0412 13:18:39.528406 8032 solver.cpp:218] Iteration 2448 (0.574576 iter/s, 20.885s/12 iters), loss = 3.46236 I0412 13:18:39.528465 8032 solver.cpp:237] Train net output #0: loss = 3.46236 (* 1 = 3.46236 loss) I0412 13:18:39.528481 8032 sgd_solver.cpp:105] Iteration 2448, lr = 0.0061575 I0412 13:18:43.597759 8032 solver.cpp:218] Iteration 2460 (2.94902 iter/s, 4.06915s/12 iters), loss = 3.28646 I0412 13:18:43.597817 8032 solver.cpp:237] Train net output #0: loss = 3.28646 (* 1 = 3.28646 loss) I0412 13:18:43.597833 8032 sgd_solver.cpp:105] Iteration 2460, lr = 0.00614288 I0412 13:18:48.715785 8032 solver.cpp:218] Iteration 2472 (2.34476 iter/s, 5.1178s/12 iters), loss = 3.35594 I0412 13:18:48.715832 8032 solver.cpp:237] Train net output #0: loss = 3.35594 (* 1 = 3.35594 loss) I0412 13:18:48.715843 8032 sgd_solver.cpp:105] Iteration 2472, lr = 0.0061283 I0412 13:18:53.830392 8032 solver.cpp:218] Iteration 2484 (2.34632 iter/s, 5.11438s/12 iters), loss = 3.39719 I0412 13:18:53.830533 8032 solver.cpp:237] Train net output #0: loss = 3.39719 (* 1 = 3.39719 loss) I0412 13:18:53.830547 8032 sgd_solver.cpp:105] Iteration 2484, lr = 0.00611375 I0412 13:18:58.761202 8032 solver.cpp:218] Iteration 2496 (2.43383 iter/s, 4.93051s/12 iters), loss = 3.12007 I0412 13:18:58.761255 8032 solver.cpp:237] Train net output #0: loss = 3.12007 (* 1 = 3.12007 loss) I0412 13:18:58.761267 8032 sgd_solver.cpp:105] Iteration 2496, lr = 0.00609923 I0412 13:19:04.206529 8032 solver.cpp:218] Iteration 2508 (2.20382 iter/s, 5.44509s/12 iters), loss = 3.22709 I0412 13:19:04.206581 8032 solver.cpp:237] Train net output #0: loss = 3.22709 (* 1 = 3.22709 loss) I0412 13:19:04.206593 8032 sgd_solver.cpp:105] Iteration 2508, lr = 0.00608475 I0412 13:19:09.711338 8032 solver.cpp:218] Iteration 2520 (2.18001 iter/s, 5.50457s/12 iters), loss = 3.18205 I0412 13:19:09.711391 8032 solver.cpp:237] Train net output #0: loss = 3.18205 (* 1 = 3.18205 loss) I0412 13:19:09.711403 8032 sgd_solver.cpp:105] Iteration 2520, lr = 0.0060703 I0412 13:19:13.251595 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:19:15.034955 8032 solver.cpp:218] Iteration 2532 (2.2542 iter/s, 5.32338s/12 iters), loss = 3.29079 I0412 13:19:15.035001 8032 solver.cpp:237] Train net output #0: loss = 3.29079 (* 1 = 3.29079 loss) I0412 13:19:15.035012 8032 sgd_solver.cpp:105] Iteration 2532, lr = 0.00605589 I0412 13:19:20.040621 8032 solver.cpp:218] Iteration 2544 (2.39739 iter/s, 5.00545s/12 iters), loss = 3.248 I0412 13:19:20.040669 8032 solver.cpp:237] Train net output #0: loss = 3.248 (* 1 = 3.248 loss) I0412 13:19:20.040681 8032 sgd_solver.cpp:105] Iteration 2544, lr = 0.00604151 I0412 13:19:22.215288 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2550.caffemodel I0412 13:19:25.276964 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2550.solverstate I0412 13:19:32.578279 8032 solver.cpp:330] Iteration 2550, Testing net (#0) I0412 13:19:32.578310 8032 net.cpp:676] Ignoring source layer train-data I0412 13:19:36.033779 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:19:37.072841 8032 solver.cpp:397] Test net output #0: accuracy = 0.189951 I0412 13:19:37.072893 8032 solver.cpp:397] Test net output #1: loss = 3.5856 (* 1 = 3.5856 loss) I0412 13:19:38.932407 8032 solver.cpp:218] Iteration 2556 (0.635218 iter/s, 18.8911s/12 iters), loss = 3.5144 I0412 13:19:38.932457 8032 solver.cpp:237] Train net output #0: loss = 3.5144 (* 1 = 3.5144 loss) I0412 13:19:38.932468 8032 sgd_solver.cpp:105] Iteration 2556, lr = 0.00602717 I0412 13:19:44.140497 8032 solver.cpp:218] Iteration 2568 (2.30421 iter/s, 5.20787s/12 iters), loss = 3.09954 I0412 13:19:44.140537 8032 solver.cpp:237] Train net output #0: loss = 3.09954 (* 1 = 3.09954 loss) I0412 13:19:44.140545 8032 sgd_solver.cpp:105] Iteration 2568, lr = 0.00601286 I0412 13:19:49.274238 8032 solver.cpp:218] Iteration 2580 (2.33757 iter/s, 5.13353s/12 iters), loss = 3.19201 I0412 13:19:49.274286 8032 solver.cpp:237] Train net output #0: loss = 3.19201 (* 1 = 3.19201 loss) I0412 13:19:49.274296 8032 sgd_solver.cpp:105] Iteration 2580, lr = 0.00599858 I0412 13:19:54.543313 8032 solver.cpp:218] Iteration 2592 (2.27756 iter/s, 5.2688s/12 iters), loss = 3.42228 I0412 13:19:54.543360 8032 solver.cpp:237] Train net output #0: loss = 3.42228 (* 1 = 3.42228 loss) I0412 13:19:54.543370 8032 sgd_solver.cpp:105] Iteration 2592, lr = 0.00598434 I0412 13:19:59.633246 8032 solver.cpp:218] Iteration 2604 (2.3577 iter/s, 5.08971s/12 iters), loss = 3.39662 I0412 13:19:59.633399 8032 solver.cpp:237] Train net output #0: loss = 3.39662 (* 1 = 3.39662 loss) I0412 13:19:59.633414 8032 sgd_solver.cpp:105] Iteration 2604, lr = 0.00597013 I0412 13:20:04.848928 8032 solver.cpp:218] Iteration 2616 (2.3009 iter/s, 5.21535s/12 iters), loss = 3.12293 I0412 13:20:04.848989 8032 solver.cpp:237] Train net output #0: loss = 3.12293 (* 1 = 3.12293 loss) I0412 13:20:04.849002 8032 sgd_solver.cpp:105] Iteration 2616, lr = 0.00595596 I0412 13:20:09.820956 8032 solver.cpp:218] Iteration 2628 (2.41361 iter/s, 4.9718s/12 iters), loss = 2.68216 I0412 13:20:09.821013 8032 solver.cpp:237] Train net output #0: loss = 2.68216 (* 1 = 2.68216 loss) I0412 13:20:09.821025 8032 sgd_solver.cpp:105] Iteration 2628, lr = 0.00594182 I0412 13:20:10.219758 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:20:14.858312 8032 solver.cpp:218] Iteration 2640 (2.38231 iter/s, 5.03713s/12 iters), loss = 3.11465 I0412 13:20:14.858355 8032 solver.cpp:237] Train net output #0: loss = 3.11465 (* 1 = 3.11465 loss) I0412 13:20:14.858363 8032 sgd_solver.cpp:105] Iteration 2640, lr = 0.00592771 I0412 13:20:19.611515 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2652.caffemodel I0412 13:20:22.636848 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2652.solverstate I0412 13:20:27.070006 8032 solver.cpp:330] Iteration 2652, Testing net (#0) I0412 13:20:27.070035 8032 net.cpp:676] Ignoring source layer train-data I0412 13:20:30.397125 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:20:31.478415 8032 solver.cpp:397] Test net output #0: accuracy = 0.177083 I0412 13:20:31.478449 8032 solver.cpp:397] Test net output #1: loss = 3.60818 (* 1 = 3.60818 loss) I0412 13:20:31.566562 8032 solver.cpp:218] Iteration 2652 (0.718233 iter/s, 16.7077s/12 iters), loss = 3.06118 I0412 13:20:31.566607 8032 solver.cpp:237] Train net output #0: loss = 3.06118 (* 1 = 3.06118 loss) I0412 13:20:31.566617 8032 sgd_solver.cpp:105] Iteration 2652, lr = 0.00591364 I0412 13:20:35.680183 8032 solver.cpp:218] Iteration 2664 (2.91727 iter/s, 4.11343s/12 iters), loss = 2.76293 I0412 13:20:35.680240 8032 solver.cpp:237] Train net output #0: loss = 2.76293 (* 1 = 2.76293 loss) I0412 13:20:35.680253 8032 sgd_solver.cpp:105] Iteration 2664, lr = 0.0058996 I0412 13:20:40.480311 8032 solver.cpp:218] Iteration 2676 (2.50005 iter/s, 4.79991s/12 iters), loss = 3.19928 I0412 13:20:40.480351 8032 solver.cpp:237] Train net output #0: loss = 3.19928 (* 1 = 3.19928 loss) I0412 13:20:40.480360 8032 sgd_solver.cpp:105] Iteration 2676, lr = 0.00588559 I0412 13:20:45.836870 8032 solver.cpp:218] Iteration 2688 (2.24034 iter/s, 5.35634s/12 iters), loss = 3.15388 I0412 13:20:45.836925 8032 solver.cpp:237] Train net output #0: loss = 3.15388 (* 1 = 3.15388 loss) I0412 13:20:45.836935 8032 sgd_solver.cpp:105] Iteration 2688, lr = 0.00587162 I0412 13:20:51.253749 8032 solver.cpp:218] Iteration 2700 (2.2154 iter/s, 5.41664s/12 iters), loss = 3.10841 I0412 13:20:51.253803 8032 solver.cpp:237] Train net output #0: loss = 3.10841 (* 1 = 3.10841 loss) I0412 13:20:51.253814 8032 sgd_solver.cpp:105] Iteration 2700, lr = 0.00585768 I0412 13:20:56.542186 8032 solver.cpp:218] Iteration 2712 (2.2692 iter/s, 5.2882s/12 iters), loss = 2.85167 I0412 13:20:56.542230 8032 solver.cpp:237] Train net output #0: loss = 2.85167 (* 1 = 2.85167 loss) I0412 13:20:56.542240 8032 sgd_solver.cpp:105] Iteration 2712, lr = 0.00584377 I0412 13:21:01.561090 8032 solver.cpp:218] Iteration 2724 (2.39106 iter/s, 5.01869s/12 iters), loss = 3.14679 I0412 13:21:01.561260 8032 solver.cpp:237] Train net output #0: loss = 3.14679 (* 1 = 3.14679 loss) I0412 13:21:01.561273 8032 sgd_solver.cpp:105] Iteration 2724, lr = 0.0058299 I0412 13:21:04.334072 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:21:07.007086 8032 solver.cpp:218] Iteration 2736 (2.2036 iter/s, 5.44564s/12 iters), loss = 2.7821 I0412 13:21:07.007136 8032 solver.cpp:237] Train net output #0: loss = 2.7821 (* 1 = 2.7821 loss) I0412 13:21:07.007148 8032 sgd_solver.cpp:105] Iteration 2736, lr = 0.00581605 I0412 13:21:12.352103 8032 solver.cpp:218] Iteration 2748 (2.24518 iter/s, 5.34479s/12 iters), loss = 2.81386 I0412 13:21:12.352146 8032 solver.cpp:237] Train net output #0: loss = 2.81386 (* 1 = 2.81386 loss) I0412 13:21:12.352154 8032 sgd_solver.cpp:105] Iteration 2748, lr = 0.00580225 I0412 13:21:14.417625 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2754.caffemodel I0412 13:21:17.754271 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2754.solverstate I0412 13:21:20.053488 8032 solver.cpp:330] Iteration 2754, Testing net (#0) I0412 13:21:20.053512 8032 net.cpp:676] Ignoring source layer train-data I0412 13:21:23.398612 8032 blocking_queue.cpp:49] Waiting for data I0412 13:21:23.634470 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:21:24.868624 8032 solver.cpp:397] Test net output #0: accuracy = 0.20527 I0412 13:21:24.868667 8032 solver.cpp:397] Test net output #1: loss = 3.5195 (* 1 = 3.5195 loss) I0412 13:21:26.875874 8032 solver.cpp:218] Iteration 2760 (0.82626 iter/s, 14.5233s/12 iters), loss = 2.98651 I0412 13:21:26.875916 8032 solver.cpp:237] Train net output #0: loss = 2.98651 (* 1 = 2.98651 loss) I0412 13:21:26.875926 8032 sgd_solver.cpp:105] Iteration 2760, lr = 0.00578847 I0412 13:21:32.116447 8032 solver.cpp:218] Iteration 2772 (2.28992 iter/s, 5.24035s/12 iters), loss = 2.99767 I0412 13:21:32.116565 8032 solver.cpp:237] Train net output #0: loss = 2.99767 (* 1 = 2.99767 loss) I0412 13:21:32.116578 8032 sgd_solver.cpp:105] Iteration 2772, lr = 0.00577473 I0412 13:21:37.051916 8032 solver.cpp:218] Iteration 2784 (2.43152 iter/s, 4.93519s/12 iters), loss = 3.07716 I0412 13:21:37.051965 8032 solver.cpp:237] Train net output #0: loss = 3.07716 (* 1 = 3.07716 loss) I0412 13:21:37.051975 8032 sgd_solver.cpp:105] Iteration 2784, lr = 0.00576102 I0412 13:21:41.910944 8032 solver.cpp:218] Iteration 2796 (2.46974 iter/s, 4.85882s/12 iters), loss = 3.05295 I0412 13:21:41.910988 8032 solver.cpp:237] Train net output #0: loss = 3.05295 (* 1 = 3.05295 loss) I0412 13:21:41.910996 8032 sgd_solver.cpp:105] Iteration 2796, lr = 0.00574734 I0412 13:21:47.055070 8032 solver.cpp:218] Iteration 2808 (2.33286 iter/s, 5.14391s/12 iters), loss = 3.03448 I0412 13:21:47.055122 8032 solver.cpp:237] Train net output #0: loss = 3.03448 (* 1 = 3.03448 loss) I0412 13:21:47.055135 8032 sgd_solver.cpp:105] Iteration 2808, lr = 0.00573369 I0412 13:21:52.357553 8032 solver.cpp:218] Iteration 2820 (2.26319 iter/s, 5.30225s/12 iters), loss = 2.8205 I0412 13:21:52.357596 8032 solver.cpp:237] Train net output #0: loss = 2.8205 (* 1 = 2.8205 loss) I0412 13:21:52.357606 8032 sgd_solver.cpp:105] Iteration 2820, lr = 0.00572008 I0412 13:21:57.219280 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:21:57.507098 8032 solver.cpp:218] Iteration 2832 (2.3304 iter/s, 5.14933s/12 iters), loss = 2.71223 I0412 13:21:57.507149 8032 solver.cpp:237] Train net output #0: loss = 2.71223 (* 1 = 2.71223 loss) I0412 13:21:57.507160 8032 sgd_solver.cpp:105] Iteration 2832, lr = 0.0057065 I0412 13:22:02.603966 8032 solver.cpp:218] Iteration 2844 (2.35449 iter/s, 5.09664s/12 iters), loss = 2.85606 I0412 13:22:02.606423 8032 solver.cpp:237] Train net output #0: loss = 2.85606 (* 1 = 2.85606 loss) I0412 13:22:02.606437 8032 sgd_solver.cpp:105] Iteration 2844, lr = 0.00569295 I0412 13:22:07.387336 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2856.caffemodel I0412 13:22:10.417264 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2856.solverstate I0412 13:22:12.720969 8032 solver.cpp:330] Iteration 2856, Testing net (#0) I0412 13:22:12.720996 8032 net.cpp:676] Ignoring source layer train-data I0412 13:22:16.135504 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:22:17.275112 8032 solver.cpp:397] Test net output #0: accuracy = 0.208333 I0412 13:22:17.275154 8032 solver.cpp:397] Test net output #1: loss = 3.44801 (* 1 = 3.44801 loss) I0412 13:22:17.363310 8032 solver.cpp:218] Iteration 2856 (0.813206 iter/s, 14.7564s/12 iters), loss = 2.71009 I0412 13:22:17.363373 8032 solver.cpp:237] Train net output #0: loss = 2.71009 (* 1 = 2.71009 loss) I0412 13:22:17.363385 8032 sgd_solver.cpp:105] Iteration 2856, lr = 0.00567944 I0412 13:22:21.586905 8032 solver.cpp:218] Iteration 2868 (2.84132 iter/s, 4.22339s/12 iters), loss = 3.08069 I0412 13:22:21.586966 8032 solver.cpp:237] Train net output #0: loss = 3.08069 (* 1 = 3.08069 loss) I0412 13:22:21.586977 8032 sgd_solver.cpp:105] Iteration 2868, lr = 0.00566595 I0412 13:22:26.573609 8032 solver.cpp:218] Iteration 2880 (2.40651 iter/s, 4.98648s/12 iters), loss = 2.82909 I0412 13:22:26.573657 8032 solver.cpp:237] Train net output #0: loss = 2.82909 (* 1 = 2.82909 loss) I0412 13:22:26.573668 8032 sgd_solver.cpp:105] Iteration 2880, lr = 0.0056525 I0412 13:22:31.574929 8032 solver.cpp:218] Iteration 2892 (2.39947 iter/s, 5.0011s/12 iters), loss = 2.71815 I0412 13:22:31.574990 8032 solver.cpp:237] Train net output #0: loss = 2.71815 (* 1 = 2.71815 loss) I0412 13:22:31.575002 8032 sgd_solver.cpp:105] Iteration 2892, lr = 0.00563908 I0412 13:22:36.590140 8032 solver.cpp:218] Iteration 2904 (2.39283 iter/s, 5.01498s/12 iters), loss = 2.95686 I0412 13:22:36.590261 8032 solver.cpp:237] Train net output #0: loss = 2.95686 (* 1 = 2.95686 loss) I0412 13:22:36.590274 8032 sgd_solver.cpp:105] Iteration 2904, lr = 0.00562569 I0412 13:22:41.714926 8032 solver.cpp:218] Iteration 2916 (2.34169 iter/s, 5.1245s/12 iters), loss = 2.75387 I0412 13:22:41.714970 8032 solver.cpp:237] Train net output #0: loss = 2.75387 (* 1 = 2.75387 loss) I0412 13:22:41.714979 8032 sgd_solver.cpp:105] Iteration 2916, lr = 0.00561233 I0412 13:22:47.364241 8032 solver.cpp:218] Iteration 2928 (2.12424 iter/s, 5.64908s/12 iters), loss = 2.91014 I0412 13:22:47.364285 8032 solver.cpp:237] Train net output #0: loss = 2.91014 (* 1 = 2.91014 loss) I0412 13:22:47.364293 8032 sgd_solver.cpp:105] Iteration 2928, lr = 0.00559901 I0412 13:22:49.264147 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:22:52.367815 8032 solver.cpp:218] Iteration 2940 (2.39839 iter/s, 5.00336s/12 iters), loss = 2.98391 I0412 13:22:52.367862 8032 solver.cpp:237] Train net output #0: loss = 2.98391 (* 1 = 2.98391 loss) I0412 13:22:52.367872 8032 sgd_solver.cpp:105] Iteration 2940, lr = 0.00558572 I0412 13:22:57.497509 8032 solver.cpp:218] Iteration 2952 (2.33942 iter/s, 5.12948s/12 iters), loss = 2.78897 I0412 13:22:57.497557 8032 solver.cpp:237] Train net output #0: loss = 2.78897 (* 1 = 2.78897 loss) I0412 13:22:57.497570 8032 sgd_solver.cpp:105] Iteration 2952, lr = 0.00557245 I0412 13:22:59.535138 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2958.caffemodel I0412 13:23:03.931502 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2958.solverstate I0412 13:23:07.128304 8032 solver.cpp:330] Iteration 2958, Testing net (#0) I0412 13:23:07.128412 8032 net.cpp:676] Ignoring source layer train-data I0412 13:23:10.364400 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:23:11.554637 8032 solver.cpp:397] Test net output #0: accuracy = 0.242647 I0412 13:23:11.554672 8032 solver.cpp:397] Test net output #1: loss = 3.30152 (* 1 = 3.30152 loss) I0412 13:23:13.580771 8032 solver.cpp:218] Iteration 2964 (0.746143 iter/s, 16.0827s/12 iters), loss = 2.43774 I0412 13:23:13.580826 8032 solver.cpp:237] Train net output #0: loss = 2.43774 (* 1 = 2.43774 loss) I0412 13:23:13.580837 8032 sgd_solver.cpp:105] Iteration 2964, lr = 0.00555922 I0412 13:23:19.012853 8032 solver.cpp:218] Iteration 2976 (2.20919 iter/s, 5.43185s/12 iters), loss = 2.67973 I0412 13:23:19.012897 8032 solver.cpp:237] Train net output #0: loss = 2.67973 (* 1 = 2.67973 loss) I0412 13:23:19.012909 8032 sgd_solver.cpp:105] Iteration 2976, lr = 0.00554603 I0412 13:23:24.409229 8032 solver.cpp:218] Iteration 2988 (2.22381 iter/s, 5.39615s/12 iters), loss = 2.82124 I0412 13:23:24.409277 8032 solver.cpp:237] Train net output #0: loss = 2.82124 (* 1 = 2.82124 loss) I0412 13:23:24.409288 8032 sgd_solver.cpp:105] Iteration 2988, lr = 0.00553286 I0412 13:23:29.603292 8032 solver.cpp:218] Iteration 3000 (2.31043 iter/s, 5.19384s/12 iters), loss = 2.68606 I0412 13:23:29.603344 8032 solver.cpp:237] Train net output #0: loss = 2.68606 (* 1 = 2.68606 loss) I0412 13:23:29.603355 8032 sgd_solver.cpp:105] Iteration 3000, lr = 0.00551972 I0412 13:23:34.979177 8032 solver.cpp:218] Iteration 3012 (2.23229 iter/s, 5.37565s/12 iters), loss = 2.89523 I0412 13:23:34.979235 8032 solver.cpp:237] Train net output #0: loss = 2.89523 (* 1 = 2.89523 loss) I0412 13:23:34.979246 8032 sgd_solver.cpp:105] Iteration 3012, lr = 0.00550662 I0412 13:23:40.361685 8032 solver.cpp:218] Iteration 3024 (2.22954 iter/s, 5.38227s/12 iters), loss = 2.9071 I0412 13:23:40.361789 8032 solver.cpp:237] Train net output #0: loss = 2.9071 (* 1 = 2.9071 loss) I0412 13:23:40.361802 8032 sgd_solver.cpp:105] Iteration 3024, lr = 0.00549354 I0412 13:23:44.297246 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:23:45.336997 8032 solver.cpp:218] Iteration 3036 (2.41204 iter/s, 4.97504s/12 iters), loss = 2.92618 I0412 13:23:45.337044 8032 solver.cpp:237] Train net output #0: loss = 2.92618 (* 1 = 2.92618 loss) I0412 13:23:45.337054 8032 sgd_solver.cpp:105] Iteration 3036, lr = 0.0054805 I0412 13:23:50.398754 8032 solver.cpp:218] Iteration 3048 (2.37082 iter/s, 5.06154s/12 iters), loss = 2.87351 I0412 13:23:50.398798 8032 solver.cpp:237] Train net output #0: loss = 2.87351 (* 1 = 2.87351 loss) I0412 13:23:50.398808 8032 sgd_solver.cpp:105] Iteration 3048, lr = 0.00546749 I0412 13:23:55.007449 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3060.caffemodel I0412 13:24:01.825683 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3060.solverstate I0412 13:24:05.682060 8032 solver.cpp:330] Iteration 3060, Testing net (#0) I0412 13:24:05.682088 8032 net.cpp:676] Ignoring source layer train-data I0412 13:24:08.889943 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:24:10.115051 8032 solver.cpp:397] Test net output #0: accuracy = 0.25 I0412 13:24:10.115097 8032 solver.cpp:397] Test net output #1: loss = 3.17958 (* 1 = 3.17958 loss) I0412 13:24:10.203531 8032 solver.cpp:218] Iteration 3060 (0.605935 iter/s, 19.8041s/12 iters), loss = 2.59694 I0412 13:24:10.203568 8032 solver.cpp:237] Train net output #0: loss = 2.59694 (* 1 = 2.59694 loss) I0412 13:24:10.203578 8032 sgd_solver.cpp:105] Iteration 3060, lr = 0.00545451 I0412 13:24:14.759922 8032 solver.cpp:218] Iteration 3072 (2.63378 iter/s, 4.55619s/12 iters), loss = 2.52023 I0412 13:24:14.760140 8032 solver.cpp:237] Train net output #0: loss = 2.52023 (* 1 = 2.52023 loss) I0412 13:24:14.760164 8032 sgd_solver.cpp:105] Iteration 3072, lr = 0.00544156 I0412 13:24:20.105585 8032 solver.cpp:218] Iteration 3084 (2.24497 iter/s, 5.34528s/12 iters), loss = 2.64738 I0412 13:24:20.105636 8032 solver.cpp:237] Train net output #0: loss = 2.64738 (* 1 = 2.64738 loss) I0412 13:24:20.105648 8032 sgd_solver.cpp:105] Iteration 3084, lr = 0.00542864 I0412 13:24:25.330680 8032 solver.cpp:218] Iteration 3096 (2.29671 iter/s, 5.22487s/12 iters), loss = 2.78296 I0412 13:24:25.330721 8032 solver.cpp:237] Train net output #0: loss = 2.78296 (* 1 = 2.78296 loss) I0412 13:24:25.330730 8032 sgd_solver.cpp:105] Iteration 3096, lr = 0.00541575 I0412 13:24:30.847815 8032 solver.cpp:218] Iteration 3108 (2.17513 iter/s, 5.51691s/12 iters), loss = 2.35977 I0412 13:24:30.847854 8032 solver.cpp:237] Train net output #0: loss = 2.35977 (* 1 = 2.35977 loss) I0412 13:24:30.847863 8032 sgd_solver.cpp:105] Iteration 3108, lr = 0.00540289 I0412 13:24:36.229377 8032 solver.cpp:218] Iteration 3120 (2.22993 iter/s, 5.38134s/12 iters), loss = 2.36242 I0412 13:24:36.229434 8032 solver.cpp:237] Train net output #0: loss = 2.36242 (* 1 = 2.36242 loss) I0412 13:24:36.229446 8032 sgd_solver.cpp:105] Iteration 3120, lr = 0.00539006 I0412 13:24:41.312829 8032 solver.cpp:218] Iteration 3132 (2.36071 iter/s, 5.08323s/12 iters), loss = 2.69775 I0412 13:24:41.312877 8032 solver.cpp:237] Train net output #0: loss = 2.69775 (* 1 = 2.69775 loss) I0412 13:24:41.312889 8032 sgd_solver.cpp:105] Iteration 3132, lr = 0.00537727 I0412 13:24:42.429563 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:24:46.414422 8032 solver.cpp:218] Iteration 3144 (2.35231 iter/s, 5.10138s/12 iters), loss = 2.26129 I0412 13:24:46.414520 8032 solver.cpp:237] Train net output #0: loss = 2.26129 (* 1 = 2.26129 loss) I0412 13:24:46.414530 8032 sgd_solver.cpp:105] Iteration 3144, lr = 0.0053645 I0412 13:24:51.825270 8032 solver.cpp:218] Iteration 3156 (2.21788 iter/s, 5.41057s/12 iters), loss = 2.43423 I0412 13:24:51.825320 8032 solver.cpp:237] Train net output #0: loss = 2.43423 (* 1 = 2.43423 loss) I0412 13:24:51.825330 8032 sgd_solver.cpp:105] Iteration 3156, lr = 0.00535176 I0412 13:24:53.824945 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3162.caffemodel I0412 13:25:00.814482 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3162.solverstate I0412 13:25:04.918721 8032 solver.cpp:330] Iteration 3162, Testing net (#0) I0412 13:25:04.918751 8032 net.cpp:676] Ignoring source layer train-data I0412 13:25:08.073192 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:25:09.340181 8032 solver.cpp:397] Test net output #0: accuracy = 0.254289 I0412 13:25:09.340219 8032 solver.cpp:397] Test net output #1: loss = 3.21982 (* 1 = 3.21982 loss) I0412 13:25:11.286756 8032 solver.cpp:218] Iteration 3168 (0.616624 iter/s, 19.4608s/12 iters), loss = 2.46443 I0412 13:25:11.286809 8032 solver.cpp:237] Train net output #0: loss = 2.46443 (* 1 = 2.46443 loss) I0412 13:25:11.286821 8032 sgd_solver.cpp:105] Iteration 3168, lr = 0.00533906 I0412 13:25:16.301386 8032 solver.cpp:218] Iteration 3180 (2.3931 iter/s, 5.01441s/12 iters), loss = 2.64677 I0412 13:25:16.301425 8032 solver.cpp:237] Train net output #0: loss = 2.64677 (* 1 = 2.64677 loss) I0412 13:25:16.301434 8032 sgd_solver.cpp:105] Iteration 3180, lr = 0.00532638 I0412 13:25:21.441015 8032 solver.cpp:218] Iteration 3192 (2.3349 iter/s, 5.13941s/12 iters), loss = 2.65569 I0412 13:25:21.441138 8032 solver.cpp:237] Train net output #0: loss = 2.65569 (* 1 = 2.65569 loss) I0412 13:25:21.441150 8032 sgd_solver.cpp:105] Iteration 3192, lr = 0.00531374 I0412 13:25:26.429311 8032 solver.cpp:218] Iteration 3204 (2.40577 iter/s, 4.988s/12 iters), loss = 2.2948 I0412 13:25:26.429369 8032 solver.cpp:237] Train net output #0: loss = 2.2948 (* 1 = 2.2948 loss) I0412 13:25:26.429383 8032 sgd_solver.cpp:105] Iteration 3204, lr = 0.00530112 I0412 13:25:31.434511 8032 solver.cpp:218] Iteration 3216 (2.39762 iter/s, 5.00497s/12 iters), loss = 2.37607 I0412 13:25:31.434574 8032 solver.cpp:237] Train net output #0: loss = 2.37607 (* 1 = 2.37607 loss) I0412 13:25:31.434589 8032 sgd_solver.cpp:105] Iteration 3216, lr = 0.00528853 I0412 13:25:36.506600 8032 solver.cpp:218] Iteration 3228 (2.366 iter/s, 5.07185s/12 iters), loss = 2.46055 I0412 13:25:36.506651 8032 solver.cpp:237] Train net output #0: loss = 2.46055 (* 1 = 2.46055 loss) I0412 13:25:36.506661 8032 sgd_solver.cpp:105] Iteration 3228, lr = 0.00527598 I0412 13:25:39.785765 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:25:41.529790 8032 solver.cpp:218] Iteration 3240 (2.38903 iter/s, 5.02296s/12 iters), loss = 2.94779 I0412 13:25:41.529845 8032 solver.cpp:237] Train net output #0: loss = 2.94779 (* 1 = 2.94779 loss) I0412 13:25:41.529858 8032 sgd_solver.cpp:105] Iteration 3240, lr = 0.00526345 I0412 13:25:46.475523 8032 solver.cpp:218] Iteration 3252 (2.42644 iter/s, 4.94552s/12 iters), loss = 2.41645 I0412 13:25:46.475565 8032 solver.cpp:237] Train net output #0: loss = 2.41645 (* 1 = 2.41645 loss) I0412 13:25:46.475574 8032 sgd_solver.cpp:105] Iteration 3252, lr = 0.00525095 I0412 13:25:50.970285 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3264.caffemodel I0412 13:26:00.260248 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3264.solverstate I0412 13:26:03.354216 8032 solver.cpp:330] Iteration 3264, Testing net (#0) I0412 13:26:03.354243 8032 net.cpp:676] Ignoring source layer train-data I0412 13:26:06.450623 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:26:07.756057 8032 solver.cpp:397] Test net output #0: accuracy = 0.258578 I0412 13:26:07.756108 8032 solver.cpp:397] Test net output #1: loss = 3.20451 (* 1 = 3.20451 loss) I0412 13:26:07.844012 8032 solver.cpp:218] Iteration 3264 (0.561593 iter/s, 21.3678s/12 iters), loss = 2.49051 I0412 13:26:07.844061 8032 solver.cpp:237] Train net output #0: loss = 2.49051 (* 1 = 2.49051 loss) I0412 13:26:07.844072 8032 sgd_solver.cpp:105] Iteration 3264, lr = 0.00523849 I0412 13:26:12.270177 8032 solver.cpp:218] Iteration 3276 (2.71127 iter/s, 4.42597s/12 iters), loss = 2.62758 I0412 13:26:12.270223 8032 solver.cpp:237] Train net output #0: loss = 2.62758 (* 1 = 2.62758 loss) I0412 13:26:12.270236 8032 sgd_solver.cpp:105] Iteration 3276, lr = 0.00522605 I0412 13:26:17.447472 8032 solver.cpp:218] Iteration 3288 (2.31791 iter/s, 5.17707s/12 iters), loss = 2.48828 I0412 13:26:17.447599 8032 solver.cpp:237] Train net output #0: loss = 2.48828 (* 1 = 2.48828 loss) I0412 13:26:17.447611 8032 sgd_solver.cpp:105] Iteration 3288, lr = 0.00521364 I0412 13:26:22.724958 8032 solver.cpp:218] Iteration 3300 (2.27391 iter/s, 5.27726s/12 iters), loss = 2.71639 I0412 13:26:22.725003 8032 solver.cpp:237] Train net output #0: loss = 2.71639 (* 1 = 2.71639 loss) I0412 13:26:22.725011 8032 sgd_solver.cpp:105] Iteration 3300, lr = 0.00520126 I0412 13:26:27.769594 8032 solver.cpp:218] Iteration 3312 (2.37887 iter/s, 5.04442s/12 iters), loss = 2.31623 I0412 13:26:27.769646 8032 solver.cpp:237] Train net output #0: loss = 2.31623 (* 1 = 2.31623 loss) I0412 13:26:27.769659 8032 sgd_solver.cpp:105] Iteration 3312, lr = 0.00518892 I0412 13:26:32.833398 8032 solver.cpp:218] Iteration 3324 (2.36986 iter/s, 5.06358s/12 iters), loss = 2.05275 I0412 13:26:32.833487 8032 solver.cpp:237] Train net output #0: loss = 2.05275 (* 1 = 2.05275 loss) I0412 13:26:32.833498 8032 sgd_solver.cpp:105] Iteration 3324, lr = 0.0051766 I0412 13:26:37.936115 8032 solver.cpp:218] Iteration 3336 (2.35181 iter/s, 5.10245s/12 iters), loss = 2.06808 I0412 13:26:37.936161 8032 solver.cpp:237] Train net output #0: loss = 2.06808 (* 1 = 2.06808 loss) I0412 13:26:37.936172 8032 sgd_solver.cpp:105] Iteration 3336, lr = 0.00516431 I0412 13:26:38.397588 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:26:43.261438 8032 solver.cpp:218] Iteration 3348 (2.25348 iter/s, 5.32509s/12 iters), loss = 2.46995 I0412 13:26:43.261508 8032 solver.cpp:237] Train net output #0: loss = 2.46995 (* 1 = 2.46995 loss) I0412 13:26:43.261524 8032 sgd_solver.cpp:105] Iteration 3348, lr = 0.00515204 I0412 13:26:48.493291 8032 solver.cpp:218] Iteration 3360 (2.29375 iter/s, 5.2316s/12 iters), loss = 2.15003 I0412 13:26:48.493346 8032 solver.cpp:237] Train net output #0: loss = 2.15003 (* 1 = 2.15003 loss) I0412 13:26:48.493356 8032 sgd_solver.cpp:105] Iteration 3360, lr = 0.00513981 I0412 13:26:50.576817 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3366.caffemodel I0412 13:26:55.791153 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3366.solverstate I0412 13:26:58.108265 8032 solver.cpp:330] Iteration 3366, Testing net (#0) I0412 13:26:58.108290 8032 net.cpp:676] Ignoring source layer train-data I0412 13:27:01.184062 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:27:02.526392 8032 solver.cpp:397] Test net output #0: accuracy = 0.258578 I0412 13:27:02.526450 8032 solver.cpp:397] Test net output #1: loss = 3.18938 (* 1 = 3.18938 loss) I0412 13:27:04.315197 8032 solver.cpp:218] Iteration 3372 (0.758469 iter/s, 15.8213s/12 iters), loss = 2.18724 I0412 13:27:04.315337 8032 solver.cpp:237] Train net output #0: loss = 2.18724 (* 1 = 2.18724 loss) I0412 13:27:04.315349 8032 sgd_solver.cpp:105] Iteration 3372, lr = 0.00512761 I0412 13:27:09.406942 8032 solver.cpp:218] Iteration 3384 (2.3569 iter/s, 5.09144s/12 iters), loss = 2.484 I0412 13:27:09.406985 8032 solver.cpp:237] Train net output #0: loss = 2.484 (* 1 = 2.484 loss) I0412 13:27:09.406993 8032 sgd_solver.cpp:105] Iteration 3384, lr = 0.00511544 I0412 13:27:14.397989 8032 solver.cpp:218] Iteration 3396 (2.40442 iter/s, 4.9908s/12 iters), loss = 2.52458 I0412 13:27:14.398033 8032 solver.cpp:237] Train net output #0: loss = 2.52458 (* 1 = 2.52458 loss) I0412 13:27:14.398042 8032 sgd_solver.cpp:105] Iteration 3396, lr = 0.00510329 I0412 13:27:19.295776 8032 solver.cpp:218] Iteration 3408 (2.45019 iter/s, 4.89758s/12 iters), loss = 2.34152 I0412 13:27:19.295816 8032 solver.cpp:237] Train net output #0: loss = 2.34152 (* 1 = 2.34152 loss) I0412 13:27:19.295826 8032 sgd_solver.cpp:105] Iteration 3408, lr = 0.00509117 I0412 13:27:24.352942 8032 solver.cpp:218] Iteration 3420 (2.37297 iter/s, 5.05695s/12 iters), loss = 2.16518 I0412 13:27:24.352984 8032 solver.cpp:237] Train net output #0: loss = 2.16518 (* 1 = 2.16518 loss) I0412 13:27:24.352993 8032 sgd_solver.cpp:105] Iteration 3420, lr = 0.00507909 I0412 13:27:29.325469 8032 solver.cpp:218] Iteration 3432 (2.41336 iter/s, 4.97231s/12 iters), loss = 2.17867 I0412 13:27:29.325515 8032 solver.cpp:237] Train net output #0: loss = 2.17867 (* 1 = 2.17867 loss) I0412 13:27:29.325526 8032 sgd_solver.cpp:105] Iteration 3432, lr = 0.00506703 I0412 13:27:31.876350 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:27:34.320155 8032 solver.cpp:218] Iteration 3444 (2.40266 iter/s, 4.99447s/12 iters), loss = 2.16156 I0412 13:27:34.320240 8032 solver.cpp:237] Train net output #0: loss = 2.16156 (* 1 = 2.16156 loss) I0412 13:27:34.320250 8032 sgd_solver.cpp:105] Iteration 3444, lr = 0.005055 I0412 13:27:39.396553 8032 solver.cpp:218] Iteration 3456 (2.364 iter/s, 5.07615s/12 iters), loss = 2.21358 I0412 13:27:39.396595 8032 solver.cpp:237] Train net output #0: loss = 2.21358 (* 1 = 2.21358 loss) I0412 13:27:39.396605 8032 sgd_solver.cpp:105] Iteration 3456, lr = 0.005043 I0412 13:27:44.324942 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3468.caffemodel I0412 13:27:52.621742 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3468.solverstate I0412 13:27:54.950513 8032 solver.cpp:330] Iteration 3468, Testing net (#0) I0412 13:27:54.950541 8032 net.cpp:676] Ignoring source layer train-data I0412 13:27:55.463038 8032 blocking_queue.cpp:49] Waiting for data I0412 13:27:58.133271 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:27:59.519456 8032 solver.cpp:397] Test net output #0: accuracy = 0.253064 I0412 13:27:59.519493 8032 solver.cpp:397] Test net output #1: loss = 3.19809 (* 1 = 3.19809 loss) I0412 13:27:59.607262 8032 solver.cpp:218] Iteration 3468 (0.593765 iter/s, 20.21s/12 iters), loss = 2.36247 I0412 13:27:59.607321 8032 solver.cpp:237] Train net output #0: loss = 2.36247 (* 1 = 2.36247 loss) I0412 13:27:59.607332 8032 sgd_solver.cpp:105] Iteration 3468, lr = 0.00503102 I0412 13:28:04.328624 8032 solver.cpp:218] Iteration 3480 (2.54176 iter/s, 4.72114s/12 iters), loss = 2.27544 I0412 13:28:04.328780 8032 solver.cpp:237] Train net output #0: loss = 2.27544 (* 1 = 2.27544 loss) I0412 13:28:04.328792 8032 sgd_solver.cpp:105] Iteration 3480, lr = 0.00501908 I0412 13:28:09.351884 8032 solver.cpp:218] Iteration 3492 (2.38904 iter/s, 5.02293s/12 iters), loss = 2.56693 I0412 13:28:09.351938 8032 solver.cpp:237] Train net output #0: loss = 2.56693 (* 1 = 2.56693 loss) I0412 13:28:09.351950 8032 sgd_solver.cpp:105] Iteration 3492, lr = 0.00500716 I0412 13:28:14.573729 8032 solver.cpp:218] Iteration 3504 (2.29814 iter/s, 5.22162s/12 iters), loss = 2.13096 I0412 13:28:14.573777 8032 solver.cpp:237] Train net output #0: loss = 2.13096 (* 1 = 2.13096 loss) I0412 13:28:14.573789 8032 sgd_solver.cpp:105] Iteration 3504, lr = 0.00499527 I0412 13:28:19.629456 8032 solver.cpp:218] Iteration 3516 (2.37365 iter/s, 5.0555s/12 iters), loss = 1.95403 I0412 13:28:19.629510 8032 solver.cpp:237] Train net output #0: loss = 1.95403 (* 1 = 1.95403 loss) I0412 13:28:19.629586 8032 sgd_solver.cpp:105] Iteration 3516, lr = 0.00498341 I0412 13:28:24.857836 8032 solver.cpp:218] Iteration 3528 (2.29527 iter/s, 5.22814s/12 iters), loss = 2.16614 I0412 13:28:24.857903 8032 solver.cpp:237] Train net output #0: loss = 2.16614 (* 1 = 2.16614 loss) I0412 13:28:24.857919 8032 sgd_solver.cpp:105] Iteration 3528, lr = 0.00497158 I0412 13:28:29.824622 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:28:30.106055 8032 solver.cpp:218] Iteration 3540 (2.2866 iter/s, 5.24797s/12 iters), loss = 1.71088 I0412 13:28:30.106104 8032 solver.cpp:237] Train net output #0: loss = 1.71088 (* 1 = 1.71088 loss) I0412 13:28:30.106115 8032 sgd_solver.cpp:105] Iteration 3540, lr = 0.00495978 I0412 13:28:35.299276 8032 solver.cpp:218] Iteration 3552 (2.3108 iter/s, 5.193s/12 iters), loss = 2.12728 I0412 13:28:35.309866 8032 solver.cpp:237] Train net output #0: loss = 2.12728 (* 1 = 2.12728 loss) I0412 13:28:35.309881 8032 sgd_solver.cpp:105] Iteration 3552, lr = 0.004948 I0412 13:28:40.372151 8032 solver.cpp:218] Iteration 3564 (2.37055 iter/s, 5.06212s/12 iters), loss = 2.06077 I0412 13:28:40.372198 8032 solver.cpp:237] Train net output #0: loss = 2.06077 (* 1 = 2.06077 loss) I0412 13:28:40.372210 8032 sgd_solver.cpp:105] Iteration 3564, lr = 0.00493626 I0412 13:28:42.433709 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3570.caffemodel I0412 13:28:45.571302 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3570.solverstate I0412 13:28:47.893352 8032 solver.cpp:330] Iteration 3570, Testing net (#0) I0412 13:28:47.893379 8032 net.cpp:676] Ignoring source layer train-data I0412 13:28:50.864727 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:28:52.317562 8032 solver.cpp:397] Test net output #0: accuracy = 0.287377 I0412 13:28:52.317602 8032 solver.cpp:397] Test net output #1: loss = 3.14161 (* 1 = 3.14161 loss) I0412 13:28:54.288786 8032 solver.cpp:218] Iteration 3576 (0.862308 iter/s, 13.9161s/12 iters), loss = 2.21601 I0412 13:28:54.288831 8032 solver.cpp:237] Train net output #0: loss = 2.21601 (* 1 = 2.21601 loss) I0412 13:28:54.288841 8032 sgd_solver.cpp:105] Iteration 3576, lr = 0.00492454 I0412 13:28:59.391623 8032 solver.cpp:218] Iteration 3588 (2.35173 iter/s, 5.10262s/12 iters), loss = 2.19535 I0412 13:28:59.391670 8032 solver.cpp:237] Train net output #0: loss = 2.19535 (* 1 = 2.19535 loss) I0412 13:28:59.391682 8032 sgd_solver.cpp:105] Iteration 3588, lr = 0.00491284 I0412 13:29:04.538069 8032 solver.cpp:218] Iteration 3600 (2.33181 iter/s, 5.14622s/12 iters), loss = 2.20516 I0412 13:29:04.538125 8032 solver.cpp:237] Train net output #0: loss = 2.20516 (* 1 = 2.20516 loss) I0412 13:29:04.538137 8032 sgd_solver.cpp:105] Iteration 3600, lr = 0.00490118 I0412 13:29:09.723675 8032 solver.cpp:218] Iteration 3612 (2.3142 iter/s, 5.18538s/12 iters), loss = 2.28089 I0412 13:29:09.723820 8032 solver.cpp:237] Train net output #0: loss = 2.28089 (* 1 = 2.28089 loss) I0412 13:29:09.723834 8032 sgd_solver.cpp:105] Iteration 3612, lr = 0.00488954 I0412 13:29:14.868283 8032 solver.cpp:218] Iteration 3624 (2.33268 iter/s, 5.1443s/12 iters), loss = 2.28518 I0412 13:29:14.868325 8032 solver.cpp:237] Train net output #0: loss = 2.28518 (* 1 = 2.28518 loss) I0412 13:29:14.868335 8032 sgd_solver.cpp:105] Iteration 3624, lr = 0.00487793 I0412 13:29:20.127190 8032 solver.cpp:218] Iteration 3636 (2.28194 iter/s, 5.25867s/12 iters), loss = 1.94745 I0412 13:29:20.127239 8032 solver.cpp:237] Train net output #0: loss = 1.94745 (* 1 = 1.94745 loss) I0412 13:29:20.127250 8032 sgd_solver.cpp:105] Iteration 3636, lr = 0.00486635 I0412 13:29:22.043129 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:29:25.475184 8032 solver.cpp:218] Iteration 3648 (2.24393 iter/s, 5.34776s/12 iters), loss = 2.10033 I0412 13:29:25.475235 8032 solver.cpp:237] Train net output #0: loss = 2.10033 (* 1 = 2.10033 loss) I0412 13:29:25.475246 8032 sgd_solver.cpp:105] Iteration 3648, lr = 0.0048548 I0412 13:29:30.612711 8032 solver.cpp:218] Iteration 3660 (2.33586 iter/s, 5.1373s/12 iters), loss = 1.98879 I0412 13:29:30.612772 8032 solver.cpp:237] Train net output #0: loss = 1.98879 (* 1 = 1.98879 loss) I0412 13:29:30.612787 8032 sgd_solver.cpp:105] Iteration 3660, lr = 0.00484327 I0412 13:29:35.541438 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3672.caffemodel I0412 13:29:40.949805 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3672.solverstate I0412 13:29:44.904290 8032 solver.cpp:330] Iteration 3672, Testing net (#0) I0412 13:29:44.904311 8032 net.cpp:676] Ignoring source layer train-data I0412 13:29:47.927487 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:29:49.452090 8032 solver.cpp:397] Test net output #0: accuracy = 0.302083 I0412 13:29:49.452127 8032 solver.cpp:397] Test net output #1: loss = 3.01262 (* 1 = 3.01262 loss) I0412 13:29:49.540505 8032 solver.cpp:218] Iteration 3672 (0.63401 iter/s, 18.9271s/12 iters), loss = 2.20981 I0412 13:29:49.540552 8032 solver.cpp:237] Train net output #0: loss = 2.20981 (* 1 = 2.20981 loss) I0412 13:29:49.540561 8032 sgd_solver.cpp:105] Iteration 3672, lr = 0.00483177 I0412 13:29:53.844338 8032 solver.cpp:218] Iteration 3684 (2.78834 iter/s, 4.30363s/12 iters), loss = 1.8899 I0412 13:29:53.844393 8032 solver.cpp:237] Train net output #0: loss = 1.8899 (* 1 = 1.8899 loss) I0412 13:29:53.844404 8032 sgd_solver.cpp:105] Iteration 3684, lr = 0.0048203 I0412 13:29:58.902436 8032 solver.cpp:218] Iteration 3696 (2.37254 iter/s, 5.05787s/12 iters), loss = 1.93972 I0412 13:29:58.902493 8032 solver.cpp:237] Train net output #0: loss = 1.93972 (* 1 = 1.93972 loss) I0412 13:29:58.902504 8032 sgd_solver.cpp:105] Iteration 3696, lr = 0.00480886 I0412 13:30:04.081220 8032 solver.cpp:218] Iteration 3708 (2.31725 iter/s, 5.17855s/12 iters), loss = 1.91732 I0412 13:30:04.081274 8032 solver.cpp:237] Train net output #0: loss = 1.91732 (* 1 = 1.91732 loss) I0412 13:30:04.081285 8032 sgd_solver.cpp:105] Iteration 3708, lr = 0.00479744 I0412 13:30:09.187534 8032 solver.cpp:218] Iteration 3720 (2.35014 iter/s, 5.10608s/12 iters), loss = 2.12419 I0412 13:30:09.187609 8032 solver.cpp:237] Train net output #0: loss = 2.12419 (* 1 = 2.12419 loss) I0412 13:30:09.187630 8032 sgd_solver.cpp:105] Iteration 3720, lr = 0.00478605 I0412 13:30:14.165787 8032 solver.cpp:218] Iteration 3732 (2.4106 iter/s, 4.97802s/12 iters), loss = 1.97178 I0412 13:30:14.165936 8032 solver.cpp:237] Train net output #0: loss = 1.97178 (* 1 = 1.97178 loss) I0412 13:30:14.165948 8032 sgd_solver.cpp:105] Iteration 3732, lr = 0.00477469 I0412 13:30:18.232698 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:30:19.265375 8032 solver.cpp:218] Iteration 3744 (2.35328 iter/s, 5.09927s/12 iters), loss = 1.87305 I0412 13:30:19.265426 8032 solver.cpp:237] Train net output #0: loss = 1.87305 (* 1 = 1.87305 loss) I0412 13:30:19.265439 8032 sgd_solver.cpp:105] Iteration 3744, lr = 0.00476335 I0412 13:30:24.426611 8032 solver.cpp:218] Iteration 3756 (2.32515 iter/s, 5.16096s/12 iters), loss = 1.98858 I0412 13:30:24.426652 8032 solver.cpp:237] Train net output #0: loss = 1.98858 (* 1 = 1.98858 loss) I0412 13:30:24.426661 8032 sgd_solver.cpp:105] Iteration 3756, lr = 0.00475204 I0412 13:30:29.675001 8032 solver.cpp:218] Iteration 3768 (2.28651 iter/s, 5.24817s/12 iters), loss = 1.93981 I0412 13:30:29.675053 8032 solver.cpp:237] Train net output #0: loss = 1.93981 (* 1 = 1.93981 loss) I0412 13:30:29.675066 8032 sgd_solver.cpp:105] Iteration 3768, lr = 0.00474076 I0412 13:30:31.733167 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3774.caffemodel I0412 13:30:34.721577 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3774.solverstate I0412 13:30:37.370419 8032 solver.cpp:330] Iteration 3774, Testing net (#0) I0412 13:30:37.370450 8032 net.cpp:676] Ignoring source layer train-data I0412 13:30:40.257333 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:30:41.751648 8032 solver.cpp:397] Test net output #0: accuracy = 0.318015 I0412 13:30:41.751688 8032 solver.cpp:397] Test net output #1: loss = 3.00299 (* 1 = 3.00299 loss) I0412 13:30:43.518412 8032 solver.cpp:218] Iteration 3780 (0.86687 iter/s, 13.8429s/12 iters), loss = 2.15157 I0412 13:30:43.518466 8032 solver.cpp:237] Train net output #0: loss = 2.15157 (* 1 = 2.15157 loss) I0412 13:30:43.518477 8032 sgd_solver.cpp:105] Iteration 3780, lr = 0.00472951 I0412 13:30:48.904402 8032 solver.cpp:218] Iteration 3792 (2.2281 iter/s, 5.38575s/12 iters), loss = 1.88733 I0412 13:30:48.904516 8032 solver.cpp:237] Train net output #0: loss = 1.88733 (* 1 = 1.88733 loss) I0412 13:30:48.904531 8032 sgd_solver.cpp:105] Iteration 3792, lr = 0.00471828 I0412 13:30:53.842830 8032 solver.cpp:218] Iteration 3804 (2.43006 iter/s, 4.93815s/12 iters), loss = 1.83613 I0412 13:30:53.842880 8032 solver.cpp:237] Train net output #0: loss = 1.83613 (* 1 = 1.83613 loss) I0412 13:30:53.842891 8032 sgd_solver.cpp:105] Iteration 3804, lr = 0.00470707 I0412 13:30:58.787724 8032 solver.cpp:218] Iteration 3816 (2.42685 iter/s, 4.94468s/12 iters), loss = 2.04534 I0412 13:30:58.787775 8032 solver.cpp:237] Train net output #0: loss = 2.04534 (* 1 = 2.04534 loss) I0412 13:30:58.787786 8032 sgd_solver.cpp:105] Iteration 3816, lr = 0.0046959 I0412 13:31:03.771318 8032 solver.cpp:218] Iteration 3828 (2.40801 iter/s, 4.98337s/12 iters), loss = 1.41818 I0412 13:31:03.771376 8032 solver.cpp:237] Train net output #0: loss = 1.41818 (* 1 = 1.41818 loss) I0412 13:31:03.771389 8032 sgd_solver.cpp:105] Iteration 3828, lr = 0.00468475 I0412 13:31:08.848213 8032 solver.cpp:218] Iteration 3840 (2.36376 iter/s, 5.07667s/12 iters), loss = 1.86828 I0412 13:31:08.848260 8032 solver.cpp:237] Train net output #0: loss = 1.86828 (* 1 = 1.86828 loss) I0412 13:31:08.848269 8032 sgd_solver.cpp:105] Iteration 3840, lr = 0.00467363 I0412 13:31:09.970268 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:31:13.956677 8032 solver.cpp:218] Iteration 3852 (2.34914 iter/s, 5.10824s/12 iters), loss = 1.7255 I0412 13:31:13.956729 8032 solver.cpp:237] Train net output #0: loss = 1.7255 (* 1 = 1.7255 loss) I0412 13:31:13.956741 8032 sgd_solver.cpp:105] Iteration 3852, lr = 0.00466253 I0412 13:31:19.010761 8032 solver.cpp:218] Iteration 3864 (2.37442 iter/s, 5.05386s/12 iters), loss = 2.05781 I0412 13:31:19.010900 8032 solver.cpp:237] Train net output #0: loss = 2.05781 (* 1 = 2.05781 loss) I0412 13:31:19.010912 8032 sgd_solver.cpp:105] Iteration 3864, lr = 0.00465146 I0412 13:31:23.489605 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3876.caffemodel I0412 13:31:26.478999 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3876.solverstate I0412 13:31:28.821159 8032 solver.cpp:330] Iteration 3876, Testing net (#0) I0412 13:31:28.821188 8032 net.cpp:676] Ignoring source layer train-data I0412 13:31:31.819813 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:31:33.459597 8032 solver.cpp:397] Test net output #0: accuracy = 0.318015 I0412 13:31:33.459633 8032 solver.cpp:397] Test net output #1: loss = 2.89415 (* 1 = 2.89415 loss) I0412 13:31:33.547762 8032 solver.cpp:218] Iteration 3876 (0.825514 iter/s, 14.5364s/12 iters), loss = 1.71869 I0412 13:31:33.547816 8032 solver.cpp:237] Train net output #0: loss = 1.71869 (* 1 = 1.71869 loss) I0412 13:31:33.547827 8032 sgd_solver.cpp:105] Iteration 3876, lr = 0.00464042 I0412 13:31:37.887472 8032 solver.cpp:218] Iteration 3888 (2.76529 iter/s, 4.3395s/12 iters), loss = 1.75712 I0412 13:31:37.887526 8032 solver.cpp:237] Train net output #0: loss = 1.75712 (* 1 = 1.75712 loss) I0412 13:31:37.887537 8032 sgd_solver.cpp:105] Iteration 3888, lr = 0.0046294 I0412 13:31:43.227696 8032 solver.cpp:218] Iteration 3900 (2.24719 iter/s, 5.33999s/12 iters), loss = 1.80088 I0412 13:31:43.227738 8032 solver.cpp:237] Train net output #0: loss = 1.80088 (* 1 = 1.80088 loss) I0412 13:31:43.227747 8032 sgd_solver.cpp:105] Iteration 3900, lr = 0.00461841 I0412 13:31:48.214398 8032 solver.cpp:218] Iteration 3912 (2.4065 iter/s, 4.98649s/12 iters), loss = 1.51045 I0412 13:31:48.214449 8032 solver.cpp:237] Train net output #0: loss = 1.51045 (* 1 = 1.51045 loss) I0412 13:31:48.214463 8032 sgd_solver.cpp:105] Iteration 3912, lr = 0.00460744 I0412 13:31:53.325690 8032 solver.cpp:218] Iteration 3924 (2.34784 iter/s, 5.11107s/12 iters), loss = 1.59292 I0412 13:31:53.325783 8032 solver.cpp:237] Train net output #0: loss = 1.59292 (* 1 = 1.59292 loss) I0412 13:31:53.325791 8032 sgd_solver.cpp:105] Iteration 3924, lr = 0.0045965 I0412 13:31:58.515640 8032 solver.cpp:218] Iteration 3936 (2.31228 iter/s, 5.18969s/12 iters), loss = 1.75948 I0412 13:31:58.515681 8032 solver.cpp:237] Train net output #0: loss = 1.75948 (* 1 = 1.75948 loss) I0412 13:31:58.515689 8032 sgd_solver.cpp:105] Iteration 3936, lr = 0.00458559 I0412 13:32:01.933020 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:32:03.582790 8032 solver.cpp:218] Iteration 3948 (2.36829 iter/s, 5.06694s/12 iters), loss = 1.61979 I0412 13:32:03.582844 8032 solver.cpp:237] Train net output #0: loss = 1.61979 (* 1 = 1.61979 loss) I0412 13:32:03.582855 8032 sgd_solver.cpp:105] Iteration 3948, lr = 0.0045747 I0412 13:32:08.667146 8032 solver.cpp:218] Iteration 3960 (2.36029 iter/s, 5.08413s/12 iters), loss = 1.76276 I0412 13:32:08.667199 8032 solver.cpp:237] Train net output #0: loss = 1.76276 (* 1 = 1.76276 loss) I0412 13:32:08.667210 8032 sgd_solver.cpp:105] Iteration 3960, lr = 0.00456384 I0412 13:32:14.146044 8032 solver.cpp:218] Iteration 3972 (2.19032 iter/s, 5.47866s/12 iters), loss = 1.95209 I0412 13:32:14.146095 8032 solver.cpp:237] Train net output #0: loss = 1.95209 (* 1 = 1.95209 loss) I0412 13:32:14.146106 8032 sgd_solver.cpp:105] Iteration 3972, lr = 0.00455301 I0412 13:32:16.204303 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3978.caffemodel I0412 13:32:22.259281 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3978.solverstate I0412 13:32:25.678236 8032 solver.cpp:330] Iteration 3978, Testing net (#0) I0412 13:32:25.678347 8032 net.cpp:676] Ignoring source layer train-data I0412 13:32:28.578388 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:32:30.302994 8032 solver.cpp:397] Test net output #0: accuracy = 0.327206 I0412 13:32:30.303036 8032 solver.cpp:397] Test net output #1: loss = 2.8424 (* 1 = 2.8424 loss) I0412 13:32:32.284250 8032 solver.cpp:218] Iteration 3984 (0.66161 iter/s, 18.1376s/12 iters), loss = 1.78591 I0412 13:32:32.284302 8032 solver.cpp:237] Train net output #0: loss = 1.78591 (* 1 = 1.78591 loss) I0412 13:32:32.284313 8032 sgd_solver.cpp:105] Iteration 3984, lr = 0.0045422 I0412 13:32:37.244128 8032 solver.cpp:218] Iteration 3996 (2.41952 iter/s, 4.95966s/12 iters), loss = 2.23738 I0412 13:32:37.244171 8032 solver.cpp:237] Train net output #0: loss = 2.23738 (* 1 = 2.23738 loss) I0412 13:32:37.244181 8032 sgd_solver.cpp:105] Iteration 3996, lr = 0.00453141 I0412 13:32:42.282070 8032 solver.cpp:218] Iteration 4008 (2.38203 iter/s, 5.03773s/12 iters), loss = 1.68127 I0412 13:32:42.282116 8032 solver.cpp:237] Train net output #0: loss = 1.68127 (* 1 = 1.68127 loss) I0412 13:32:42.282124 8032 sgd_solver.cpp:105] Iteration 4008, lr = 0.00452066 I0412 13:32:47.321081 8032 solver.cpp:218] Iteration 4020 (2.38152 iter/s, 5.03879s/12 iters), loss = 1.74091 I0412 13:32:47.321130 8032 solver.cpp:237] Train net output #0: loss = 1.74091 (* 1 = 1.74091 loss) I0412 13:32:47.321141 8032 sgd_solver.cpp:105] Iteration 4020, lr = 0.00450992 I0412 13:32:52.389012 8032 solver.cpp:218] Iteration 4032 (2.36793 iter/s, 5.06771s/12 iters), loss = 1.57737 I0412 13:32:52.389060 8032 solver.cpp:237] Train net output #0: loss = 1.57737 (* 1 = 1.57737 loss) I0412 13:32:52.389070 8032 sgd_solver.cpp:105] Iteration 4032, lr = 0.00449921 I0412 13:32:57.503433 8032 solver.cpp:218] Iteration 4044 (2.34641 iter/s, 5.1142s/12 iters), loss = 1.36361 I0412 13:32:57.503543 8032 solver.cpp:237] Train net output #0: loss = 1.36361 (* 1 = 1.36361 loss) I0412 13:32:57.503553 8032 sgd_solver.cpp:105] Iteration 4044, lr = 0.00448853 I0412 13:32:58.014132 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:33:02.518082 8032 solver.cpp:218] Iteration 4056 (2.39312 iter/s, 5.01437s/12 iters), loss = 1.66356 I0412 13:33:02.518139 8032 solver.cpp:237] Train net output #0: loss = 1.66356 (* 1 = 1.66356 loss) I0412 13:33:02.518152 8032 sgd_solver.cpp:105] Iteration 4056, lr = 0.00447788 I0412 13:33:07.640844 8032 solver.cpp:218] Iteration 4068 (2.34259 iter/s, 5.12253s/12 iters), loss = 1.70373 I0412 13:33:07.640905 8032 solver.cpp:237] Train net output #0: loss = 1.70373 (* 1 = 1.70373 loss) I0412 13:33:07.640918 8032 sgd_solver.cpp:105] Iteration 4068, lr = 0.00446724 I0412 13:33:12.229765 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4080.caffemodel I0412 13:33:23.172292 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4080.solverstate I0412 13:33:29.027669 8032 solver.cpp:330] Iteration 4080, Testing net (#0) I0412 13:33:29.027752 8032 net.cpp:676] Ignoring source layer train-data I0412 13:33:31.887013 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:33:33.610028 8032 solver.cpp:397] Test net output #0: accuracy = 0.348652 I0412 13:33:33.610064 8032 solver.cpp:397] Test net output #1: loss = 2.81936 (* 1 = 2.81936 loss) I0412 13:33:33.696910 8032 solver.cpp:218] Iteration 4080 (0.460561 iter/s, 26.0552s/12 iters), loss = 1.5362 I0412 13:33:33.696954 8032 solver.cpp:237] Train net output #0: loss = 1.5362 (* 1 = 1.5362 loss) I0412 13:33:33.696962 8032 sgd_solver.cpp:105] Iteration 4080, lr = 0.00445664 I0412 13:33:38.131103 8032 solver.cpp:218] Iteration 4092 (2.70637 iter/s, 4.43399s/12 iters), loss = 1.6698 I0412 13:33:38.131150 8032 solver.cpp:237] Train net output #0: loss = 1.6698 (* 1 = 1.6698 loss) I0412 13:33:38.131160 8032 sgd_solver.cpp:105] Iteration 4092, lr = 0.00444606 I0412 13:33:43.378279 8032 solver.cpp:218] Iteration 4104 (2.28704 iter/s, 5.24695s/12 iters), loss = 1.78751 I0412 13:33:43.378324 8032 solver.cpp:237] Train net output #0: loss = 1.78751 (* 1 = 1.78751 loss) I0412 13:33:43.378332 8032 sgd_solver.cpp:105] Iteration 4104, lr = 0.0044355 I0412 13:33:48.927528 8032 solver.cpp:218] Iteration 4116 (2.16254 iter/s, 5.54902s/12 iters), loss = 1.4328 I0412 13:33:48.927572 8032 solver.cpp:237] Train net output #0: loss = 1.4328 (* 1 = 1.4328 loss) I0412 13:33:48.927580 8032 sgd_solver.cpp:105] Iteration 4116, lr = 0.00442497 I0412 13:33:54.097182 8032 solver.cpp:218] Iteration 4128 (2.32134 iter/s, 5.16944s/12 iters), loss = 1.43512 I0412 13:33:54.097225 8032 solver.cpp:237] Train net output #0: loss = 1.43512 (* 1 = 1.43512 loss) I0412 13:33:54.097236 8032 sgd_solver.cpp:105] Iteration 4128, lr = 0.00441447 I0412 13:33:59.141227 8032 solver.cpp:218] Iteration 4140 (2.37914 iter/s, 5.04383s/12 iters), loss = 1.74731 I0412 13:33:59.141378 8032 solver.cpp:237] Train net output #0: loss = 1.74731 (* 1 = 1.74731 loss) I0412 13:33:59.141388 8032 sgd_solver.cpp:105] Iteration 4140, lr = 0.00440398 I0412 13:34:01.762553 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:34:04.183231 8032 solver.cpp:218] Iteration 4152 (2.38016 iter/s, 5.04169s/12 iters), loss = 1.33551 I0412 13:34:04.183281 8032 solver.cpp:237] Train net output #0: loss = 1.33551 (* 1 = 1.33551 loss) I0412 13:34:04.183293 8032 sgd_solver.cpp:105] Iteration 4152, lr = 0.00439353 I0412 13:34:05.776886 8032 blocking_queue.cpp:49] Waiting for data I0412 13:34:09.176009 8032 solver.cpp:218] Iteration 4164 (2.40358 iter/s, 4.99256s/12 iters), loss = 1.69151 I0412 13:34:09.176051 8032 solver.cpp:237] Train net output #0: loss = 1.69151 (* 1 = 1.69151 loss) I0412 13:34:09.176062 8032 sgd_solver.cpp:105] Iteration 4164, lr = 0.0043831 I0412 13:34:14.256603 8032 solver.cpp:218] Iteration 4176 (2.36203 iter/s, 5.08038s/12 iters), loss = 1.4044 I0412 13:34:14.256654 8032 solver.cpp:237] Train net output #0: loss = 1.4044 (* 1 = 1.4044 loss) I0412 13:34:14.256665 8032 sgd_solver.cpp:105] Iteration 4176, lr = 0.00437269 I0412 13:34:16.301429 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4182.caffemodel I0412 13:34:21.514183 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4182.solverstate I0412 13:34:26.559130 8032 solver.cpp:330] Iteration 4182, Testing net (#0) I0412 13:34:26.559156 8032 net.cpp:676] Ignoring source layer train-data I0412 13:34:29.406685 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:34:31.130715 8032 solver.cpp:397] Test net output #0: accuracy = 0.359069 I0412 13:34:31.130764 8032 solver.cpp:397] Test net output #1: loss = 2.80121 (* 1 = 2.80121 loss) I0412 13:34:33.129639 8032 solver.cpp:218] Iteration 4188 (0.63585 iter/s, 18.8724s/12 iters), loss = 1.28144 I0412 13:34:33.129685 8032 solver.cpp:237] Train net output #0: loss = 1.28144 (* 1 = 1.28144 loss) I0412 13:34:33.129694 8032 sgd_solver.cpp:105] Iteration 4188, lr = 0.00436231 I0412 13:34:38.451426 8032 solver.cpp:218] Iteration 4200 (2.25498 iter/s, 5.32156s/12 iters), loss = 1.56353 I0412 13:34:38.451476 8032 solver.cpp:237] Train net output #0: loss = 1.56353 (* 1 = 1.56353 loss) I0412 13:34:38.451489 8032 sgd_solver.cpp:105] Iteration 4200, lr = 0.00435195 I0412 13:34:43.826905 8032 solver.cpp:218] Iteration 4212 (2.23246 iter/s, 5.37525s/12 iters), loss = 1.47748 I0412 13:34:43.826949 8032 solver.cpp:237] Train net output #0: loss = 1.47748 (* 1 = 1.47748 loss) I0412 13:34:43.826958 8032 sgd_solver.cpp:105] Iteration 4212, lr = 0.00434162 I0412 13:34:49.281567 8032 solver.cpp:218] Iteration 4224 (2.20005 iter/s, 5.45443s/12 iters), loss = 1.39147 I0412 13:34:49.281620 8032 solver.cpp:237] Train net output #0: loss = 1.39147 (* 1 = 1.39147 loss) I0412 13:34:49.281631 8032 sgd_solver.cpp:105] Iteration 4224, lr = 0.00433131 I0412 13:34:54.505108 8032 solver.cpp:218] Iteration 4236 (2.29739 iter/s, 5.22331s/12 iters), loss = 1.40363 I0412 13:34:54.505151 8032 solver.cpp:237] Train net output #0: loss = 1.40363 (* 1 = 1.40363 loss) I0412 13:34:54.505160 8032 sgd_solver.cpp:105] Iteration 4236, lr = 0.00432103 I0412 13:34:59.274435 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:34:59.497658 8032 solver.cpp:218] Iteration 4248 (2.40368 iter/s, 4.99234s/12 iters), loss = 1.19281 I0412 13:34:59.497750 8032 solver.cpp:237] Train net output #0: loss = 1.19281 (* 1 = 1.19281 loss) I0412 13:34:59.497759 8032 sgd_solver.cpp:105] Iteration 4248, lr = 0.00431077 I0412 13:35:04.760799 8032 solver.cpp:218] Iteration 4260 (2.28012 iter/s, 5.26287s/12 iters), loss = 1.39465 I0412 13:35:04.760840 8032 solver.cpp:237] Train net output #0: loss = 1.39465 (* 1 = 1.39465 loss) I0412 13:35:04.760848 8032 sgd_solver.cpp:105] Iteration 4260, lr = 0.00430053 I0412 13:35:09.772249 8032 solver.cpp:218] Iteration 4272 (2.39462 iter/s, 5.01124s/12 iters), loss = 1.31776 I0412 13:35:09.772291 8032 solver.cpp:237] Train net output #0: loss = 1.31776 (* 1 = 1.31776 loss) I0412 13:35:09.772301 8032 sgd_solver.cpp:105] Iteration 4272, lr = 0.00429032 I0412 13:35:14.519520 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4284.caffemodel I0412 13:35:17.585983 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4284.solverstate I0412 13:35:21.767613 8032 solver.cpp:330] Iteration 4284, Testing net (#0) I0412 13:35:21.767638 8032 net.cpp:676] Ignoring source layer train-data I0412 13:35:24.671838 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:35:26.364713 8032 solver.cpp:397] Test net output #0: accuracy = 0.354167 I0412 13:35:26.364763 8032 solver.cpp:397] Test net output #1: loss = 2.8855 (* 1 = 2.8855 loss) I0412 13:35:26.453001 8032 solver.cpp:218] Iteration 4284 (0.719417 iter/s, 16.6802s/12 iters), loss = 1.40769 I0412 13:35:26.453052 8032 solver.cpp:237] Train net output #0: loss = 1.40769 (* 1 = 1.40769 loss) I0412 13:35:26.453064 8032 sgd_solver.cpp:105] Iteration 4284, lr = 0.00428014 I0412 13:35:30.754139 8032 solver.cpp:218] Iteration 4296 (2.79009 iter/s, 4.30094s/12 iters), loss = 1.2725 I0412 13:35:30.754233 8032 solver.cpp:237] Train net output #0: loss = 1.2725 (* 1 = 1.2725 loss) I0412 13:35:30.754243 8032 sgd_solver.cpp:105] Iteration 4296, lr = 0.00426998 I0412 13:35:35.927744 8032 solver.cpp:218] Iteration 4308 (2.31959 iter/s, 5.17334s/12 iters), loss = 1.52061 I0412 13:35:35.927791 8032 solver.cpp:237] Train net output #0: loss = 1.52061 (* 1 = 1.52061 loss) I0412 13:35:35.927803 8032 sgd_solver.cpp:105] Iteration 4308, lr = 0.00425984 I0412 13:35:41.102775 8032 solver.cpp:218] Iteration 4320 (2.31893 iter/s, 5.17481s/12 iters), loss = 1.58977 I0412 13:35:41.102818 8032 solver.cpp:237] Train net output #0: loss = 1.58977 (* 1 = 1.58977 loss) I0412 13:35:41.102826 8032 sgd_solver.cpp:105] Iteration 4320, lr = 0.00424972 I0412 13:35:46.318461 8032 solver.cpp:218] Iteration 4332 (2.30085 iter/s, 5.21547s/12 iters), loss = 1.43684 I0412 13:35:46.318514 8032 solver.cpp:237] Train net output #0: loss = 1.43684 (* 1 = 1.43684 loss) I0412 13:35:46.318526 8032 sgd_solver.cpp:105] Iteration 4332, lr = 0.00423964 I0412 13:35:51.291862 8032 solver.cpp:218] Iteration 4344 (2.41294 iter/s, 4.97318s/12 iters), loss = 1.33633 I0412 13:35:51.291903 8032 solver.cpp:237] Train net output #0: loss = 1.33633 (* 1 = 1.33633 loss) I0412 13:35:51.291913 8032 sgd_solver.cpp:105] Iteration 4344, lr = 0.00422957 I0412 13:35:53.188058 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:35:56.328006 8032 solver.cpp:218] Iteration 4356 (2.38287 iter/s, 5.03593s/12 iters), loss = 1.28031 I0412 13:35:56.328052 8032 solver.cpp:237] Train net output #0: loss = 1.28031 (* 1 = 1.28031 loss) I0412 13:35:56.328063 8032 sgd_solver.cpp:105] Iteration 4356, lr = 0.00421953 I0412 13:36:01.452500 8032 solver.cpp:218] Iteration 4368 (2.3418 iter/s, 5.12427s/12 iters), loss = 1.47255 I0412 13:36:01.453646 8032 solver.cpp:237] Train net output #0: loss = 1.47255 (* 1 = 1.47255 loss) I0412 13:36:01.453657 8032 sgd_solver.cpp:105] Iteration 4368, lr = 0.00420951 I0412 13:36:06.701036 8032 solver.cpp:218] Iteration 4380 (2.28693 iter/s, 5.24721s/12 iters), loss = 1.20982 I0412 13:36:06.701092 8032 solver.cpp:237] Train net output #0: loss = 1.20982 (* 1 = 1.20982 loss) I0412 13:36:06.701104 8032 sgd_solver.cpp:105] Iteration 4380, lr = 0.00419952 I0412 13:36:08.700624 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4386.caffemodel I0412 13:36:11.844907 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4386.solverstate I0412 13:36:14.172137 8032 solver.cpp:330] Iteration 4386, Testing net (#0) I0412 13:36:14.172164 8032 net.cpp:676] Ignoring source layer train-data I0412 13:36:16.846309 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:36:18.788730 8032 solver.cpp:397] Test net output #0: accuracy = 0.368873 I0412 13:36:18.788772 8032 solver.cpp:397] Test net output #1: loss = 2.71814 (* 1 = 2.71814 loss) I0412 13:36:20.938467 8032 solver.cpp:218] Iteration 4392 (0.842879 iter/s, 14.2369s/12 iters), loss = 1.39213 I0412 13:36:20.938514 8032 solver.cpp:237] Train net output #0: loss = 1.39213 (* 1 = 1.39213 loss) I0412 13:36:20.938524 8032 sgd_solver.cpp:105] Iteration 4392, lr = 0.00418954 I0412 13:36:26.060643 8032 solver.cpp:218] Iteration 4404 (2.34285 iter/s, 5.12196s/12 iters), loss = 1.21385 I0412 13:36:26.060688 8032 solver.cpp:237] Train net output #0: loss = 1.21385 (* 1 = 1.21385 loss) I0412 13:36:26.060698 8032 sgd_solver.cpp:105] Iteration 4404, lr = 0.0041796 I0412 13:36:31.154979 8032 solver.cpp:218] Iteration 4416 (2.35566 iter/s, 5.09412s/12 iters), loss = 1.21802 I0412 13:36:31.155030 8032 solver.cpp:237] Train net output #0: loss = 1.21802 (* 1 = 1.21802 loss) I0412 13:36:31.155040 8032 sgd_solver.cpp:105] Iteration 4416, lr = 0.00416967 I0412 13:36:36.549075 8032 solver.cpp:218] Iteration 4428 (2.22475 iter/s, 5.39386s/12 iters), loss = 1.11614 I0412 13:36:36.549165 8032 solver.cpp:237] Train net output #0: loss = 1.11614 (* 1 = 1.11614 loss) I0412 13:36:36.549175 8032 sgd_solver.cpp:105] Iteration 4428, lr = 0.00415977 I0412 13:36:41.468039 8032 solver.cpp:218] Iteration 4440 (2.43967 iter/s, 4.91871s/12 iters), loss = 1.03993 I0412 13:36:41.468098 8032 solver.cpp:237] Train net output #0: loss = 1.03993 (* 1 = 1.03993 loss) I0412 13:36:41.468111 8032 sgd_solver.cpp:105] Iteration 4440, lr = 0.0041499 I0412 13:36:45.682466 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:36:46.607625 8032 solver.cpp:218] Iteration 4452 (2.33493 iter/s, 5.13935s/12 iters), loss = 1.23011 I0412 13:36:46.607686 8032 solver.cpp:237] Train net output #0: loss = 1.23011 (* 1 = 1.23011 loss) I0412 13:36:46.607700 8032 sgd_solver.cpp:105] Iteration 4452, lr = 0.00414005 I0412 13:36:51.575727 8032 solver.cpp:218] Iteration 4464 (2.41552 iter/s, 4.96788s/12 iters), loss = 1.27818 I0412 13:36:51.575773 8032 solver.cpp:237] Train net output #0: loss = 1.27818 (* 1 = 1.27818 loss) I0412 13:36:51.575783 8032 sgd_solver.cpp:105] Iteration 4464, lr = 0.00413022 I0412 13:36:56.632668 8032 solver.cpp:218] Iteration 4476 (2.37308 iter/s, 5.05672s/12 iters), loss = 1.22317 I0412 13:36:56.632721 8032 solver.cpp:237] Train net output #0: loss = 1.22317 (* 1 = 1.22317 loss) I0412 13:36:56.632733 8032 sgd_solver.cpp:105] Iteration 4476, lr = 0.00412041 I0412 13:37:01.224519 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4488.caffemodel I0412 13:37:05.846987 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4488.solverstate I0412 13:37:08.161340 8032 solver.cpp:330] Iteration 4488, Testing net (#0) I0412 13:37:08.161427 8032 net.cpp:676] Ignoring source layer train-data I0412 13:37:10.835341 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:37:12.694166 8032 solver.cpp:397] Test net output #0: accuracy = 0.381127 I0412 13:37:12.694197 8032 solver.cpp:397] Test net output #1: loss = 2.73325 (* 1 = 2.73325 loss) I0412 13:37:12.782550 8032 solver.cpp:218] Iteration 4488 (0.743066 iter/s, 16.1493s/12 iters), loss = 1.18425 I0412 13:37:12.782609 8032 solver.cpp:237] Train net output #0: loss = 1.18425 (* 1 = 1.18425 loss) I0412 13:37:12.782621 8032 sgd_solver.cpp:105] Iteration 4488, lr = 0.00411063 I0412 13:37:17.150144 8032 solver.cpp:218] Iteration 4500 (2.74764 iter/s, 4.36738s/12 iters), loss = 1.20375 I0412 13:37:17.150209 8032 solver.cpp:237] Train net output #0: loss = 1.20375 (* 1 = 1.20375 loss) I0412 13:37:17.150220 8032 sgd_solver.cpp:105] Iteration 4500, lr = 0.00410087 I0412 13:37:22.232952 8032 solver.cpp:218] Iteration 4512 (2.36101 iter/s, 5.08258s/12 iters), loss = 1.45414 I0412 13:37:22.232995 8032 solver.cpp:237] Train net output #0: loss = 1.45414 (* 1 = 1.45414 loss) I0412 13:37:22.233002 8032 sgd_solver.cpp:105] Iteration 4512, lr = 0.00409113 I0412 13:37:27.353688 8032 solver.cpp:218] Iteration 4524 (2.34351 iter/s, 5.12052s/12 iters), loss = 1.30552 I0412 13:37:27.353732 8032 solver.cpp:237] Train net output #0: loss = 1.30552 (* 1 = 1.30552 loss) I0412 13:37:27.353744 8032 sgd_solver.cpp:105] Iteration 4524, lr = 0.00408142 I0412 13:37:32.566196 8032 solver.cpp:218] Iteration 4536 (2.30225 iter/s, 5.21229s/12 iters), loss = 1.07833 I0412 13:37:32.566247 8032 solver.cpp:237] Train net output #0: loss = 1.07833 (* 1 = 1.07833 loss) I0412 13:37:32.566259 8032 sgd_solver.cpp:105] Iteration 4536, lr = 0.00407173 I0412 13:37:37.643452 8032 solver.cpp:218] Iteration 4548 (2.36358 iter/s, 5.07704s/12 iters), loss = 1.18095 I0412 13:37:37.643502 8032 solver.cpp:237] Train net output #0: loss = 1.18095 (* 1 = 1.18095 loss) I0412 13:37:37.643513 8032 sgd_solver.cpp:105] Iteration 4548, lr = 0.00406206 I0412 13:37:38.944834 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:37:42.919097 8032 solver.cpp:218] Iteration 4560 (2.2747 iter/s, 5.27542s/12 iters), loss = 1.11327 I0412 13:37:42.919142 8032 solver.cpp:237] Train net output #0: loss = 1.11327 (* 1 = 1.11327 loss) I0412 13:37:42.919152 8032 sgd_solver.cpp:105] Iteration 4560, lr = 0.00405242 I0412 13:37:47.940748 8032 solver.cpp:218] Iteration 4572 (2.38976 iter/s, 5.02143s/12 iters), loss = 1.14086 I0412 13:37:47.940801 8032 solver.cpp:237] Train net output #0: loss = 1.14086 (* 1 = 1.14086 loss) I0412 13:37:47.940812 8032 sgd_solver.cpp:105] Iteration 4572, lr = 0.0040428 I0412 13:37:53.042290 8032 solver.cpp:218] Iteration 4584 (2.35233 iter/s, 5.10132s/12 iters), loss = 1.20336 I0412 13:37:53.042338 8032 solver.cpp:237] Train net output #0: loss = 1.20336 (* 1 = 1.20336 loss) I0412 13:37:53.042348 8032 sgd_solver.cpp:105] Iteration 4584, lr = 0.0040332 I0412 13:37:55.253726 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4590.caffemodel I0412 13:38:03.049695 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4590.solverstate I0412 13:38:07.154008 8032 solver.cpp:330] Iteration 4590, Testing net (#0) I0412 13:38:07.154034 8032 net.cpp:676] Ignoring source layer train-data I0412 13:38:09.804822 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:38:11.622495 8032 solver.cpp:397] Test net output #0: accuracy = 0.382966 I0412 13:38:11.622527 8032 solver.cpp:397] Test net output #1: loss = 2.81489 (* 1 = 2.81489 loss) I0412 13:38:13.717034 8032 solver.cpp:218] Iteration 4596 (0.580438 iter/s, 20.674s/12 iters), loss = 1.37189 I0412 13:38:13.717089 8032 solver.cpp:237] Train net output #0: loss = 1.37189 (* 1 = 1.37189 loss) I0412 13:38:13.717100 8032 sgd_solver.cpp:105] Iteration 4596, lr = 0.00402362 I0412 13:38:19.053887 8032 solver.cpp:218] Iteration 4608 (2.24861 iter/s, 5.33663s/12 iters), loss = 1.12304 I0412 13:38:19.053939 8032 solver.cpp:237] Train net output #0: loss = 1.12304 (* 1 = 1.12304 loss) I0412 13:38:19.053951 8032 sgd_solver.cpp:105] Iteration 4608, lr = 0.00401407 I0412 13:38:24.259908 8032 solver.cpp:218] Iteration 4620 (2.30512 iter/s, 5.2058s/12 iters), loss = 1.03584 I0412 13:38:24.259959 8032 solver.cpp:237] Train net output #0: loss = 1.03584 (* 1 = 1.03584 loss) I0412 13:38:24.259971 8032 sgd_solver.cpp:105] Iteration 4620, lr = 0.00400454 I0412 13:38:29.331020 8032 solver.cpp:218] Iteration 4632 (2.36644 iter/s, 5.0709s/12 iters), loss = 1.23911 I0412 13:38:29.331064 8032 solver.cpp:237] Train net output #0: loss = 1.23911 (* 1 = 1.23911 loss) I0412 13:38:29.331073 8032 sgd_solver.cpp:105] Iteration 4632, lr = 0.00399503 I0412 13:38:34.694839 8032 solver.cpp:218] Iteration 4644 (2.2373 iter/s, 5.36361s/12 iters), loss = 1.08047 I0412 13:38:34.694895 8032 solver.cpp:237] Train net output #0: loss = 1.08047 (* 1 = 1.08047 loss) I0412 13:38:34.694907 8032 sgd_solver.cpp:105] Iteration 4644, lr = 0.00398555 I0412 13:38:38.068982 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:38:39.634310 8032 solver.cpp:218] Iteration 4656 (2.42951 iter/s, 4.93926s/12 iters), loss = 0.947513 I0412 13:38:39.634361 8032 solver.cpp:237] Train net output #0: loss = 0.947513 (* 1 = 0.947513 loss) I0412 13:38:39.634371 8032 sgd_solver.cpp:105] Iteration 4656, lr = 0.00397608 I0412 13:38:44.756783 8032 solver.cpp:218] Iteration 4668 (2.34272 iter/s, 5.12226s/12 iters), loss = 1.1006 I0412 13:38:44.756880 8032 solver.cpp:237] Train net output #0: loss = 1.1006 (* 1 = 1.1006 loss) I0412 13:38:44.756889 8032 sgd_solver.cpp:105] Iteration 4668, lr = 0.00396664 I0412 13:38:49.896183 8032 solver.cpp:218] Iteration 4680 (2.33502 iter/s, 5.13914s/12 iters), loss = 1.18656 I0412 13:38:49.896227 8032 solver.cpp:237] Train net output #0: loss = 1.18656 (* 1 = 1.18656 loss) I0412 13:38:49.896237 8032 sgd_solver.cpp:105] Iteration 4680, lr = 0.00395723 I0412 13:38:54.414683 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4692.caffemodel I0412 13:39:02.659288 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4692.solverstate I0412 13:39:08.310109 8032 solver.cpp:330] Iteration 4692, Testing net (#0) I0412 13:39:08.310138 8032 net.cpp:676] Ignoring source layer train-data I0412 13:39:10.915787 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:39:12.855901 8032 solver.cpp:397] Test net output #0: accuracy = 0.370711 I0412 13:39:12.855932 8032 solver.cpp:397] Test net output #1: loss = 2.83234 (* 1 = 2.83234 loss) I0412 13:39:12.945858 8032 solver.cpp:218] Iteration 4692 (0.520631 iter/s, 23.0489s/12 iters), loss = 1.0495 I0412 13:39:12.945925 8032 solver.cpp:237] Train net output #0: loss = 1.0495 (* 1 = 1.0495 loss) I0412 13:39:12.945935 8032 sgd_solver.cpp:105] Iteration 4692, lr = 0.00394783 I0412 13:39:17.138664 8032 solver.cpp:218] Iteration 4704 (2.86219 iter/s, 4.1926s/12 iters), loss = 1.18947 I0412 13:39:17.138731 8032 solver.cpp:237] Train net output #0: loss = 1.18947 (* 1 = 1.18947 loss) I0412 13:39:17.138741 8032 sgd_solver.cpp:105] Iteration 4704, lr = 0.00393846 I0412 13:39:22.226809 8032 solver.cpp:218] Iteration 4716 (2.35853 iter/s, 5.08791s/12 iters), loss = 1.2322 I0412 13:39:22.226859 8032 solver.cpp:237] Train net output #0: loss = 1.2322 (* 1 = 1.2322 loss) I0412 13:39:22.226871 8032 sgd_solver.cpp:105] Iteration 4716, lr = 0.00392911 I0412 13:39:27.334385 8032 solver.cpp:218] Iteration 4728 (2.34955 iter/s, 5.10736s/12 iters), loss = 1.16409 I0412 13:39:27.334441 8032 solver.cpp:237] Train net output #0: loss = 1.16409 (* 1 = 1.16409 loss) I0412 13:39:27.334452 8032 sgd_solver.cpp:105] Iteration 4728, lr = 0.00391978 I0412 13:39:32.638367 8032 solver.cpp:218] Iteration 4740 (2.26255 iter/s, 5.30375s/12 iters), loss = 1.02082 I0412 13:39:32.638412 8032 solver.cpp:237] Train net output #0: loss = 1.02082 (* 1 = 1.02082 loss) I0412 13:39:32.638422 8032 sgd_solver.cpp:105] Iteration 4740, lr = 0.00391047 I0412 13:39:37.849133 8032 solver.cpp:218] Iteration 4752 (2.30302 iter/s, 5.21055s/12 iters), loss = 1.13832 I0412 13:39:37.849177 8032 solver.cpp:237] Train net output #0: loss = 1.13832 (* 1 = 1.13832 loss) I0412 13:39:37.849187 8032 sgd_solver.cpp:105] Iteration 4752, lr = 0.00390119 I0412 13:39:38.426307 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:39:43.160854 8032 solver.cpp:218] Iteration 4764 (2.25925 iter/s, 5.31151s/12 iters), loss = 1.11062 I0412 13:39:43.160904 8032 solver.cpp:237] Train net output #0: loss = 1.11062 (* 1 = 1.11062 loss) I0412 13:39:43.160917 8032 sgd_solver.cpp:105] Iteration 4764, lr = 0.00389193 I0412 13:39:48.298523 8032 solver.cpp:218] Iteration 4776 (2.33579 iter/s, 5.13745s/12 iters), loss = 1.05929 I0412 13:39:48.298658 8032 solver.cpp:237] Train net output #0: loss = 1.05929 (* 1 = 1.05929 loss) I0412 13:39:48.298673 8032 sgd_solver.cpp:105] Iteration 4776, lr = 0.00388269 I0412 13:39:53.392828 8032 solver.cpp:218] Iteration 4788 (2.35571 iter/s, 5.09401s/12 iters), loss = 1.24923 I0412 13:39:53.392885 8032 solver.cpp:237] Train net output #0: loss = 1.24923 (* 1 = 1.24923 loss) I0412 13:39:53.392900 8032 sgd_solver.cpp:105] Iteration 4788, lr = 0.00387347 I0412 13:39:55.377089 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4794.caffemodel I0412 13:40:14.115272 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4794.solverstate I0412 13:40:21.679862 8032 solver.cpp:330] Iteration 4794, Testing net (#0) I0412 13:40:21.679919 8032 net.cpp:676] Ignoring source layer train-data I0412 13:40:24.239573 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:40:26.173588 8032 solver.cpp:397] Test net output #0: accuracy = 0.373774 I0412 13:40:26.173648 8032 solver.cpp:397] Test net output #1: loss = 2.88007 (* 1 = 2.88007 loss) I0412 13:40:28.093173 8032 solver.cpp:218] Iteration 4800 (0.345829 iter/s, 34.6992s/12 iters), loss = 1.02722 I0412 13:40:28.093225 8032 solver.cpp:237] Train net output #0: loss = 1.02722 (* 1 = 1.02722 loss) I0412 13:40:28.093236 8032 sgd_solver.cpp:105] Iteration 4800, lr = 0.00386427 I0412 13:40:33.082669 8032 solver.cpp:218] Iteration 4812 (2.40516 iter/s, 4.98928s/12 iters), loss = 1.15693 I0412 13:40:33.082713 8032 solver.cpp:237] Train net output #0: loss = 1.15693 (* 1 = 1.15693 loss) I0412 13:40:33.082723 8032 sgd_solver.cpp:105] Iteration 4812, lr = 0.0038551 I0412 13:40:38.009384 8032 solver.cpp:218] Iteration 4824 (2.4358 iter/s, 4.92651s/12 iters), loss = 1.08398 I0412 13:40:38.009426 8032 solver.cpp:237] Train net output #0: loss = 1.08398 (* 1 = 1.08398 loss) I0412 13:40:38.009435 8032 sgd_solver.cpp:105] Iteration 4824, lr = 0.00384594 I0412 13:40:43.085341 8032 solver.cpp:218] Iteration 4836 (2.36419 iter/s, 5.07574s/12 iters), loss = 1.01156 I0412 13:40:43.085415 8032 solver.cpp:237] Train net output #0: loss = 1.01156 (* 1 = 1.01156 loss) I0412 13:40:43.085435 8032 sgd_solver.cpp:105] Iteration 4836, lr = 0.00383681 I0412 13:40:45.123968 8032 blocking_queue.cpp:49] Waiting for data I0412 13:40:48.067261 8032 solver.cpp:218] Iteration 4848 (2.40882 iter/s, 4.98169s/12 iters), loss = 1.40005 I0412 13:40:48.067310 8032 solver.cpp:237] Train net output #0: loss = 1.40005 (* 1 = 1.40005 loss) I0412 13:40:48.067322 8032 sgd_solver.cpp:105] Iteration 4848, lr = 0.0038277 I0412 13:40:50.947402 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:40:53.321666 8032 solver.cpp:218] Iteration 4860 (2.28389 iter/s, 5.25418s/12 iters), loss = 1.05346 I0412 13:40:53.321774 8032 solver.cpp:237] Train net output #0: loss = 1.05346 (* 1 = 1.05346 loss) I0412 13:40:53.321785 8032 sgd_solver.cpp:105] Iteration 4860, lr = 0.00381862 I0412 13:40:58.860049 8032 solver.cpp:218] Iteration 4872 (2.16681 iter/s, 5.5381s/12 iters), loss = 0.939489 I0412 13:40:58.860090 8032 solver.cpp:237] Train net output #0: loss = 0.939489 (* 1 = 0.939489 loss) I0412 13:40:58.860098 8032 sgd_solver.cpp:105] Iteration 4872, lr = 0.00380955 I0412 13:41:03.963353 8032 solver.cpp:218] Iteration 4884 (2.35151 iter/s, 5.10309s/12 iters), loss = 0.883472 I0412 13:41:03.963405 8032 solver.cpp:237] Train net output #0: loss = 0.883472 (* 1 = 0.883472 loss) I0412 13:41:03.963416 8032 sgd_solver.cpp:105] Iteration 4884, lr = 0.0038005 I0412 13:41:08.937948 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4896.caffemodel I0412 13:41:12.966732 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4896.solverstate I0412 13:41:16.781152 8032 solver.cpp:330] Iteration 4896, Testing net (#0) I0412 13:41:16.781177 8032 net.cpp:676] Ignoring source layer train-data I0412 13:41:19.290397 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:41:21.273255 8032 solver.cpp:397] Test net output #0: accuracy = 0.389093 I0412 13:41:21.273304 8032 solver.cpp:397] Test net output #1: loss = 2.75266 (* 1 = 2.75266 loss) I0412 13:41:21.361459 8032 solver.cpp:218] Iteration 4896 (0.689753 iter/s, 17.3975s/12 iters), loss = 1.18432 I0412 13:41:21.361510 8032 solver.cpp:237] Train net output #0: loss = 1.18432 (* 1 = 1.18432 loss) I0412 13:41:21.361521 8032 sgd_solver.cpp:105] Iteration 4896, lr = 0.00379148 I0412 13:41:25.709904 8032 solver.cpp:218] Iteration 4908 (2.75973 iter/s, 4.34825s/12 iters), loss = 1.34366 I0412 13:41:25.710063 8032 solver.cpp:237] Train net output #0: loss = 1.34366 (* 1 = 1.34366 loss) I0412 13:41:25.710075 8032 sgd_solver.cpp:105] Iteration 4908, lr = 0.00378248 I0412 13:41:30.744689 8032 solver.cpp:218] Iteration 4920 (2.38357 iter/s, 5.03446s/12 iters), loss = 0.981235 I0412 13:41:30.744740 8032 solver.cpp:237] Train net output #0: loss = 0.981235 (* 1 = 0.981235 loss) I0412 13:41:30.744750 8032 sgd_solver.cpp:105] Iteration 4920, lr = 0.0037735 I0412 13:41:35.741752 8032 solver.cpp:218] Iteration 4932 (2.40151 iter/s, 4.99685s/12 iters), loss = 0.876219 I0412 13:41:35.741816 8032 solver.cpp:237] Train net output #0: loss = 0.876219 (* 1 = 0.876219 loss) I0412 13:41:35.741828 8032 sgd_solver.cpp:105] Iteration 4932, lr = 0.00376454 I0412 13:41:40.670212 8032 solver.cpp:218] Iteration 4944 (2.43495 iter/s, 4.92824s/12 iters), loss = 0.885811 I0412 13:41:40.670261 8032 solver.cpp:237] Train net output #0: loss = 0.885811 (* 1 = 0.885811 loss) I0412 13:41:40.670270 8032 sgd_solver.cpp:105] Iteration 4944, lr = 0.0037556 I0412 13:41:45.835319 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:41:46.048576 8032 solver.cpp:218] Iteration 4956 (2.23125 iter/s, 5.37814s/12 iters), loss = 0.719517 I0412 13:41:46.048619 8032 solver.cpp:237] Train net output #0: loss = 0.719517 (* 1 = 0.719517 loss) I0412 13:41:46.048629 8032 sgd_solver.cpp:105] Iteration 4956, lr = 0.00374669 I0412 13:41:51.467344 8032 solver.cpp:218] Iteration 4968 (2.21462 iter/s, 5.41855s/12 iters), loss = 0.933844 I0412 13:41:51.467396 8032 solver.cpp:237] Train net output #0: loss = 0.933844 (* 1 = 0.933844 loss) I0412 13:41:51.467409 8032 sgd_solver.cpp:105] Iteration 4968, lr = 0.00373779 I0412 13:41:56.433430 8032 solver.cpp:218] Iteration 4980 (2.4165 iter/s, 4.96587s/12 iters), loss = 0.780992 I0412 13:41:56.433514 8032 solver.cpp:237] Train net output #0: loss = 0.780992 (* 1 = 0.780992 loss) I0412 13:41:56.433527 8032 sgd_solver.cpp:105] Iteration 4980, lr = 0.00372892 I0412 13:42:01.603855 8032 solver.cpp:218] Iteration 4992 (2.321 iter/s, 5.17018s/12 iters), loss = 0.997907 I0412 13:42:01.603904 8032 solver.cpp:237] Train net output #0: loss = 0.997907 (* 1 = 0.997907 loss) I0412 13:42:01.603915 8032 sgd_solver.cpp:105] Iteration 4992, lr = 0.00372006 I0412 13:42:03.672506 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4998.caffemodel I0412 13:42:07.697068 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4998.solverstate I0412 13:42:10.017866 8032 solver.cpp:330] Iteration 4998, Testing net (#0) I0412 13:42:10.017894 8032 net.cpp:676] Ignoring source layer train-data I0412 13:42:12.691977 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:42:14.734071 8032 solver.cpp:397] Test net output #0: accuracy = 0.403799 I0412 13:42:14.734122 8032 solver.cpp:397] Test net output #1: loss = 2.67105 (* 1 = 2.67105 loss) I0412 13:42:16.688803 8032 solver.cpp:218] Iteration 5004 (0.795522 iter/s, 15.0844s/12 iters), loss = 0.865084 I0412 13:42:16.688853 8032 solver.cpp:237] Train net output #0: loss = 0.865084 (* 1 = 0.865084 loss) I0412 13:42:16.688864 8032 sgd_solver.cpp:105] Iteration 5004, lr = 0.00371123 I0412 13:42:21.702282 8032 solver.cpp:218] Iteration 5016 (2.39365 iter/s, 5.01327s/12 iters), loss = 0.98601 I0412 13:42:21.702330 8032 solver.cpp:237] Train net output #0: loss = 0.98601 (* 1 = 0.98601 loss) I0412 13:42:21.702340 8032 sgd_solver.cpp:105] Iteration 5016, lr = 0.00370242 I0412 13:42:26.854319 8032 solver.cpp:218] Iteration 5028 (2.32927 iter/s, 5.15182s/12 iters), loss = 0.890042 I0412 13:42:26.854462 8032 solver.cpp:237] Train net output #0: loss = 0.890042 (* 1 = 0.890042 loss) I0412 13:42:26.854480 8032 sgd_solver.cpp:105] Iteration 5028, lr = 0.00369363 I0412 13:42:32.208729 8032 solver.cpp:218] Iteration 5040 (2.24128 iter/s, 5.35409s/12 iters), loss = 0.915572 I0412 13:42:32.208777 8032 solver.cpp:237] Train net output #0: loss = 0.915572 (* 1 = 0.915572 loss) I0412 13:42:32.208786 8032 sgd_solver.cpp:105] Iteration 5040, lr = 0.00368486 I0412 13:42:37.366104 8032 solver.cpp:218] Iteration 5052 (2.32687 iter/s, 5.15715s/12 iters), loss = 0.810976 I0412 13:42:37.366166 8032 solver.cpp:237] Train net output #0: loss = 0.810976 (* 1 = 0.810976 loss) I0412 13:42:37.366179 8032 sgd_solver.cpp:105] Iteration 5052, lr = 0.00367611 I0412 13:42:39.232434 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:42:42.307766 8032 solver.cpp:218] Iteration 5064 (2.42844 iter/s, 4.94144s/12 iters), loss = 1.15219 I0412 13:42:42.307816 8032 solver.cpp:237] Train net output #0: loss = 1.15219 (* 1 = 1.15219 loss) I0412 13:42:42.307826 8032 sgd_solver.cpp:105] Iteration 5064, lr = 0.00366738 I0412 13:42:47.253183 8032 solver.cpp:218] Iteration 5076 (2.4266 iter/s, 4.9452s/12 iters), loss = 0.895168 I0412 13:42:47.253243 8032 solver.cpp:237] Train net output #0: loss = 0.895168 (* 1 = 0.895168 loss) I0412 13:42:47.253257 8032 sgd_solver.cpp:105] Iteration 5076, lr = 0.00365868 I0412 13:42:52.280231 8032 solver.cpp:218] Iteration 5088 (2.38719 iter/s, 5.02683s/12 iters), loss = 0.913304 I0412 13:42:52.280277 8032 solver.cpp:237] Train net output #0: loss = 0.913304 (* 1 = 0.913304 loss) I0412 13:42:52.280284 8032 sgd_solver.cpp:105] Iteration 5088, lr = 0.00364999 I0412 13:42:56.836668 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_5100.caffemodel I0412 13:43:00.780557 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_5100.solverstate I0412 13:43:03.862136 8032 solver.cpp:330] Iteration 5100, Testing net (#0) I0412 13:43:03.862162 8032 net.cpp:676] Ignoring source layer train-data I0412 13:43:06.393110 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:43:08.426542 8032 solver.cpp:397] Test net output #0: accuracy = 0.421569 I0412 13:43:08.426604 8032 solver.cpp:397] Test net output #1: loss = 2.6262 (* 1 = 2.6262 loss) I0412 13:43:08.515038 8032 solver.cpp:218] Iteration 5100 (0.739177 iter/s, 16.2343s/12 iters), loss = 0.825493 I0412 13:43:08.515089 8032 solver.cpp:237] Train net output #0: loss = 0.825493 (* 1 = 0.825493 loss) I0412 13:43:08.515100 8032 sgd_solver.cpp:105] Iteration 5100, lr = 0.00364132 I0412 13:43:12.833489 8032 solver.cpp:218] Iteration 5112 (2.7789 iter/s, 4.31825s/12 iters), loss = 0.922046 I0412 13:43:12.833539 8032 solver.cpp:237] Train net output #0: loss = 0.922046 (* 1 = 0.922046 loss) I0412 13:43:12.833549 8032 sgd_solver.cpp:105] Iteration 5112, lr = 0.00363268 I0412 13:43:18.036237 8032 solver.cpp:218] Iteration 5124 (2.30657 iter/s, 5.20252s/12 iters), loss = 0.726298 I0412 13:43:18.036283 8032 solver.cpp:237] Train net output #0: loss = 0.726298 (* 1 = 0.726298 loss) I0412 13:43:18.036295 8032 sgd_solver.cpp:105] Iteration 5124, lr = 0.00362405 I0412 13:43:23.514303 8032 solver.cpp:218] Iteration 5136 (2.19065 iter/s, 5.47784s/12 iters), loss = 0.833272 I0412 13:43:23.514361 8032 solver.cpp:237] Train net output #0: loss = 0.833272 (* 1 = 0.833272 loss) I0412 13:43:23.514374 8032 sgd_solver.cpp:105] Iteration 5136, lr = 0.00361545 I0412 13:43:28.820200 8032 solver.cpp:218] Iteration 5148 (2.26173 iter/s, 5.30568s/12 iters), loss = 0.781012 I0412 13:43:28.820240 8032 solver.cpp:237] Train net output #0: loss = 0.781012 (* 1 = 0.781012 loss) I0412 13:43:28.820247 8032 sgd_solver.cpp:105] Iteration 5148, lr = 0.00360687 I0412 13:43:32.851608 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:43:33.777279 8032 solver.cpp:218] Iteration 5160 (2.42088 iter/s, 4.95687s/12 iters), loss = 0.860746 I0412 13:43:33.777333 8032 solver.cpp:237] Train net output #0: loss = 0.860746 (* 1 = 0.860746 loss) I0412 13:43:33.777345 8032 sgd_solver.cpp:105] Iteration 5160, lr = 0.0035983 I0412 13:43:38.927654 8032 solver.cpp:218] Iteration 5172 (2.33003 iter/s, 5.15015s/12 iters), loss = 0.858186 I0412 13:43:38.927700 8032 solver.cpp:237] Train net output #0: loss = 0.858186 (* 1 = 0.858186 loss) I0412 13:43:38.927709 8032 sgd_solver.cpp:105] Iteration 5172, lr = 0.00358976 I0412 13:43:44.094550 8032 solver.cpp:218] Iteration 5184 (2.32257 iter/s, 5.16668s/12 iters), loss = 0.656308 I0412 13:43:44.094592 8032 solver.cpp:237] Train net output #0: loss = 0.656308 (* 1 = 0.656308 loss) I0412 13:43:44.094601 8032 sgd_solver.cpp:105] Iteration 5184, lr = 0.00358124 I0412 13:43:49.252041 8032 solver.cpp:218] Iteration 5196 (2.32681 iter/s, 5.15728s/12 iters), loss = 0.857863 I0412 13:43:49.252096 8032 solver.cpp:237] Train net output #0: loss = 0.857863 (* 1 = 0.857863 loss) I0412 13:43:49.252108 8032 sgd_solver.cpp:105] Iteration 5196, lr = 0.00357273 I0412 13:43:51.491837 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_5202.caffemodel I0412 13:43:55.967739 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_5202.solverstate I0412 13:43:58.598088 8032 solver.cpp:330] Iteration 5202, Testing net (#0) I0412 13:43:58.598119 8032 net.cpp:676] Ignoring source layer train-data I0412 13:44:00.993778 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:44:03.085145 8032 solver.cpp:397] Test net output #0: accuracy = 0.404412 I0412 13:44:03.085230 8032 solver.cpp:397] Test net output #1: loss = 2.74904 (* 1 = 2.74904 loss) I0412 13:44:04.773372 8032 solver.cpp:218] Iteration 5208 (0.773156 iter/s, 15.5208s/12 iters), loss = 0.916829 I0412 13:44:04.773427 8032 solver.cpp:237] Train net output #0: loss = 0.916829 (* 1 = 0.916829 loss) I0412 13:44:04.773439 8032 sgd_solver.cpp:105] Iteration 5208, lr = 0.00356425 I0412 13:44:09.808184 8032 solver.cpp:218] Iteration 5220 (2.38351 iter/s, 5.03459s/12 iters), loss = 0.77636 I0412 13:44:09.808241 8032 solver.cpp:237] Train net output #0: loss = 0.77636 (* 1 = 0.77636 loss) I0412 13:44:09.808254 8032 sgd_solver.cpp:105] Iteration 5220, lr = 0.00355579 I0412 13:44:14.884555 8032 solver.cpp:218] Iteration 5232 (2.36399 iter/s, 5.07615s/12 iters), loss = 0.63869 I0412 13:44:14.884594 8032 solver.cpp:237] Train net output #0: loss = 0.63869 (* 1 = 0.63869 loss) I0412 13:44:14.884603 8032 sgd_solver.cpp:105] Iteration 5232, lr = 0.00354735 I0412 13:44:20.239980 8032 solver.cpp:218] Iteration 5244 (2.24081 iter/s, 5.3552s/12 iters), loss = 0.911289 I0412 13:44:20.240044 8032 solver.cpp:237] Train net output #0: loss = 0.911289 (* 1 = 0.911289 loss) I0412 13:44:20.240061 8032 sgd_solver.cpp:105] Iteration 5244, lr = 0.00353892 I0412 13:44:25.299463 8032 solver.cpp:218] Iteration 5256 (2.37189 iter/s, 5.05925s/12 iters), loss = 0.801762 I0412 13:44:25.299512 8032 solver.cpp:237] Train net output #0: loss = 0.801762 (* 1 = 0.801762 loss) I0412 13:44:25.299523 8032 sgd_solver.cpp:105] Iteration 5256, lr = 0.00353052 I0412 13:44:26.498776 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:44:30.315153 8032 solver.cpp:218] Iteration 5268 (2.3926 iter/s, 5.01547s/12 iters), loss = 0.672972 I0412 13:44:30.315204 8032 solver.cpp:237] Train net output #0: loss = 0.672972 (* 1 = 0.672972 loss) I0412 13:44:30.315215 8032 sgd_solver.cpp:105] Iteration 5268, lr = 0.00352214 I0412 13:44:35.552431 8032 solver.cpp:218] Iteration 5280 (2.29136 iter/s, 5.23705s/12 iters), loss = 0.875995 I0412 13:44:35.552603 8032 solver.cpp:237] Train net output #0: loss = 0.875995 (* 1 = 0.875995 loss) I0412 13:44:35.552618 8032 sgd_solver.cpp:105] Iteration 5280, lr = 0.00351378 I0412 13:44:40.562786 8032 solver.cpp:218] Iteration 5292 (2.3952 iter/s, 5.01003s/12 iters), loss = 0.575709 I0412 13:44:40.562834 8032 solver.cpp:237] Train net output #0: loss = 0.575709 (* 1 = 0.575709 loss) I0412 13:44:40.562844 8032 sgd_solver.cpp:105] Iteration 5292, lr = 0.00350544 I0412 13:44:45.158404 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_5304.caffemodel I0412 13:44:50.055243 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_5304.solverstate I0412 13:44:55.090278 8032 solver.cpp:330] Iteration 5304, Testing net (#0) I0412 13:44:55.090301 8032 net.cpp:676] Ignoring source layer train-data I0412 13:44:57.358150 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:44:59.476338 8032 solver.cpp:397] Test net output #0: accuracy = 0.414216 I0412 13:44:59.476387 8032 solver.cpp:397] Test net output #1: loss = 2.67444 (* 1 = 2.67444 loss) I0412 13:44:59.564391 8032 solver.cpp:218] Iteration 5304 (0.631546 iter/s, 19.001s/12 iters), loss = 0.743374 I0412 13:44:59.564445 8032 solver.cpp:237] Train net output #0: loss = 0.743374 (* 1 = 0.743374 loss) I0412 13:44:59.564455 8032 sgd_solver.cpp:105] Iteration 5304, lr = 0.00349711 I0412 13:45:03.770983 8032 solver.cpp:218] Iteration 5316 (2.8528 iter/s, 4.2064s/12 iters), loss = 0.76467 I0412 13:45:03.771028 8032 solver.cpp:237] Train net output #0: loss = 0.76467 (* 1 = 0.76467 loss) I0412 13:45:03.771036 8032 sgd_solver.cpp:105] Iteration 5316, lr = 0.00348881 I0412 13:45:08.742118 8032 solver.cpp:218] Iteration 5328 (2.41404 iter/s, 4.97093s/12 iters), loss = 1.03331 I0412 13:45:08.742240 8032 solver.cpp:237] Train net output #0: loss = 1.03331 (* 1 = 1.03331 loss) I0412 13:45:08.742250 8032 sgd_solver.cpp:105] Iteration 5328, lr = 0.00348053 I0412 13:45:14.041716 8032 solver.cpp:218] Iteration 5340 (2.26445 iter/s, 5.29931s/12 iters), loss = 0.931166 I0412 13:45:14.041760 8032 solver.cpp:237] Train net output #0: loss = 0.931166 (* 1 = 0.931166 loss) I0412 13:45:14.041772 8032 sgd_solver.cpp:105] Iteration 5340, lr = 0.00347226 I0412 13:45:18.995669 8032 solver.cpp:218] Iteration 5352 (2.42241 iter/s, 4.95374s/12 iters), loss = 0.785755 I0412 13:45:18.995721 8032 solver.cpp:237] Train net output #0: loss = 0.785755 (* 1 = 0.785755 loss) I0412 13:45:18.995733 8032 sgd_solver.cpp:105] Iteration 5352, lr = 0.00346402 I0412 13:45:22.554662 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:45:24.127854 8032 solver.cpp:218] Iteration 5364 (2.33829 iter/s, 5.13196s/12 iters), loss = 0.728403 I0412 13:45:24.127903 8032 solver.cpp:237] Train net output #0: loss = 0.728403 (* 1 = 0.728403 loss) I0412 13:45:24.127914 8032 sgd_solver.cpp:105] Iteration 5364, lr = 0.0034558 I0412 13:45:29.089787 8032 solver.cpp:218] Iteration 5376 (2.41851 iter/s, 4.96172s/12 iters), loss = 0.66427 I0412 13:45:29.089834 8032 solver.cpp:237] Train net output #0: loss = 0.66427 (* 1 = 0.66427 loss) I0412 13:45:29.089843 8032 sgd_solver.cpp:105] Iteration 5376, lr = 0.00344759 I0412 13:45:34.106849 8032 solver.cpp:218] Iteration 5388 (2.39194 iter/s, 5.01685s/12 iters), loss = 0.593727 I0412 13:45:34.106897 8032 solver.cpp:237] Train net output #0: loss = 0.593727 (* 1 = 0.593727 loss) I0412 13:45:34.106907 8032 sgd_solver.cpp:105] Iteration 5388, lr = 0.00343941 I0412 13:45:39.108140 8032 solver.cpp:218] Iteration 5400 (2.39948 iter/s, 5.00108s/12 iters), loss = 0.682391 I0412 13:45:39.108278 8032 solver.cpp:237] Train net output #0: loss = 0.682391 (* 1 = 0.682391 loss) I0412 13:45:39.108289 8032 sgd_solver.cpp:105] Iteration 5400, lr = 0.00343124 I0412 13:45:41.268620 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_5406.caffemodel I0412 13:45:49.314070 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_5406.solverstate I0412 13:45:53.843159 8032 solver.cpp:330] Iteration 5406, Testing net (#0) I0412 13:45:53.843186 8032 net.cpp:676] Ignoring source layer train-data I0412 13:45:56.210219 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:45:58.388993 8032 solver.cpp:397] Test net output #0: accuracy = 0.416054 I0412 13:45:58.389042 8032 solver.cpp:397] Test net output #1: loss = 2.62302 (* 1 = 2.62302 loss) I0412 13:46:00.162226 8032 solver.cpp:218] Iteration 5412 (0.569982 iter/s, 21.0533s/12 iters), loss = 1.3395 I0412 13:46:00.162271 8032 solver.cpp:237] Train net output #0: loss = 1.3395 (* 1 = 1.3395 loss) I0412 13:46:00.162279 8032 sgd_solver.cpp:105] Iteration 5412, lr = 0.00342309 I0412 13:46:05.129937 8032 solver.cpp:218] Iteration 5424 (2.4157 iter/s, 4.9675s/12 iters), loss = 0.557344 I0412 13:46:05.129999 8032 solver.cpp:237] Train net output #0: loss = 0.557344 (* 1 = 0.557344 loss) I0412 13:46:05.130008 8032 sgd_solver.cpp:105] Iteration 5424, lr = 0.00341497 I0412 13:46:10.141052 8032 solver.cpp:218] Iteration 5436 (2.39478 iter/s, 5.01089s/12 iters), loss = 0.692448 I0412 13:46:10.141139 8032 solver.cpp:237] Train net output #0: loss = 0.692448 (* 1 = 0.692448 loss) I0412 13:46:10.141147 8032 sgd_solver.cpp:105] Iteration 5436, lr = 0.00340686 I0412 13:46:15.136883 8032 solver.cpp:218] Iteration 5448 (2.40213 iter/s, 4.99557s/12 iters), loss = 0.904668 I0412 13:46:15.136940 8032 solver.cpp:237] Train net output #0: loss = 0.904668 (* 1 = 0.904668 loss) I0412 13:46:15.136952 8032 sgd_solver.cpp:105] Iteration 5448, lr = 0.00339877 I0412 13:46:20.273288 8032 solver.cpp:218] Iteration 5460 (2.33637 iter/s, 5.13618s/12 iters), loss = 0.869146 I0412 13:46:20.273341 8032 solver.cpp:237] Train net output #0: loss = 0.869146 (* 1 = 0.869146 loss) I0412 13:46:20.273352 8032 sgd_solver.cpp:105] Iteration 5460, lr = 0.0033907 I0412 13:46:20.840313 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:46:25.372514 8032 solver.cpp:218] Iteration 5472 (2.3534 iter/s, 5.099s/12 iters), loss = 0.718199 I0412 13:46:25.372570 8032 solver.cpp:237] Train net output #0: loss = 0.718199 (* 1 = 0.718199 loss) I0412 13:46:25.372584 8032 sgd_solver.cpp:105] Iteration 5472, lr = 0.00338265 I0412 13:46:30.416038 8032 solver.cpp:218] Iteration 5484 (2.37939 iter/s, 5.04331s/12 iters), loss = 0.48533 I0412 13:46:30.416080 8032 solver.cpp:237] Train net output #0: loss = 0.48533 (* 1 = 0.48533 loss) I0412 13:46:30.416090 8032 sgd_solver.cpp:105] Iteration 5484, lr = 0.00337462 I0412 13:46:35.427918 8032 solver.cpp:218] Iteration 5496 (2.39441 iter/s, 5.01167s/12 iters), loss = 0.645921 I0412 13:46:35.427969 8032 solver.cpp:237] Train net output #0: loss = 0.645921 (* 1 = 0.645921 loss) I0412 13:46:35.427980 8032 sgd_solver.cpp:105] Iteration 5496, lr = 0.00336661 I0412 13:46:40.065325 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_5508.caffemodel I0412 13:46:43.141031 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_5508.solverstate I0412 13:46:47.478978 8032 solver.cpp:330] Iteration 5508, Testing net (#0) I0412 13:46:47.478999 8032 net.cpp:676] Ignoring source layer train-data I0412 13:46:49.779211 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:46:52.146391 8032 solver.cpp:397] Test net output #0: accuracy = 0.440564 I0412 13:46:52.146437 8032 solver.cpp:397] Test net output #1: loss = 2.62419 (* 1 = 2.62419 loss) I0412 13:46:52.233650 8032 solver.cpp:218] Iteration 5508 (0.714066 iter/s, 16.8052s/12 iters), loss = 0.522163 I0412 13:46:52.233697 8032 solver.cpp:237] Train net output #0: loss = 0.522163 (* 1 = 0.522163 loss) I0412 13:46:52.233708 8032 sgd_solver.cpp:105] Iteration 5508, lr = 0.00335861 I0412 13:46:56.843268 8032 solver.cpp:218] Iteration 5520 (2.60337 iter/s, 4.60942s/12 iters), loss = 0.782961 I0412 13:46:56.843314 8032 solver.cpp:237] Train net output #0: loss = 0.782961 (* 1 = 0.782961 loss) I0412 13:46:56.843322 8032 sgd_solver.cpp:105] Iteration 5520, lr = 0.00335064 I0412 13:46:59.531950 8032 blocking_queue.cpp:49] Waiting for data I0412 13:47:02.326272 8032 solver.cpp:218] Iteration 5532 (2.18867 iter/s, 5.48277s/12 iters), loss = 0.809523 I0412 13:47:02.326314 8032 solver.cpp:237] Train net output #0: loss = 0.809523 (* 1 = 0.809523 loss) I0412 13:47:02.326324 8032 sgd_solver.cpp:105] Iteration 5532, lr = 0.00334268 I0412 13:47:07.820181 8032 solver.cpp:218] Iteration 5544 (2.18433 iter/s, 5.49369s/12 iters), loss = 0.594185 I0412 13:47:07.820230 8032 solver.cpp:237] Train net output #0: loss = 0.594185 (* 1 = 0.594185 loss) I0412 13:47:07.820241 8032 sgd_solver.cpp:105] Iteration 5544, lr = 0.00333475 I0412 13:47:12.850100 8032 solver.cpp:218] Iteration 5556 (2.38583 iter/s, 5.0297s/12 iters), loss = 0.5927 I0412 13:47:12.850147 8032 solver.cpp:237] Train net output #0: loss = 0.5927 (* 1 = 0.5927 loss) I0412 13:47:12.850158 8032 sgd_solver.cpp:105] Iteration 5556, lr = 0.00332683 I0412 13:47:15.556497 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:47:17.850507 8032 solver.cpp:218] Iteration 5568 (2.39991 iter/s, 5.0002s/12 iters), loss = 0.532312 I0412 13:47:17.850553 8032 solver.cpp:237] Train net output #0: loss = 0.532312 (* 1 = 0.532312 loss) I0412 13:47:17.850562 8032 sgd_solver.cpp:105] Iteration 5568, lr = 0.00331893 I0412 13:47:22.901201 8032 solver.cpp:218] Iteration 5580 (2.37601 iter/s, 5.05048s/12 iters), loss = 0.81186 I0412 13:47:22.901257 8032 solver.cpp:237] Train net output #0: loss = 0.81186 (* 1 = 0.81186 loss) I0412 13:47:22.901269 8032 sgd_solver.cpp:105] Iteration 5580, lr = 0.00331105 I0412 13:47:27.865573 8032 solver.cpp:218] Iteration 5592 (2.41733 iter/s, 4.96416s/12 iters), loss = 0.68974 I0412 13:47:27.865620 8032 solver.cpp:237] Train net output #0: loss = 0.68974 (* 1 = 0.68974 loss) I0412 13:47:27.865631 8032 sgd_solver.cpp:105] Iteration 5592, lr = 0.00330319 I0412 13:47:32.994771 8032 solver.cpp:218] Iteration 5604 (2.33965 iter/s, 5.12898s/12 iters), loss = 0.827058 I0412 13:47:32.994823 8032 solver.cpp:237] Train net output #0: loss = 0.827058 (* 1 = 0.827058 loss) I0412 13:47:32.994835 8032 sgd_solver.cpp:105] Iteration 5604, lr = 0.00329535 I0412 13:47:35.097239 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_5610.caffemodel I0412 13:47:38.129706 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_5610.solverstate I0412 13:47:45.028861 8032 solver.cpp:330] Iteration 5610, Testing net (#0) I0412 13:47:45.028888 8032 net.cpp:676] Ignoring source layer train-data I0412 13:47:47.398684 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:47:49.821681 8032 solver.cpp:397] Test net output #0: accuracy = 0.431985 I0412 13:47:49.821727 8032 solver.cpp:397] Test net output #1: loss = 2.73961 (* 1 = 2.73961 loss) I0412 13:47:51.765667 8032 solver.cpp:218] Iteration 5616 (0.639309 iter/s, 18.7703s/12 iters), loss = 0.573297 I0412 13:47:51.765725 8032 solver.cpp:237] Train net output #0: loss = 0.573297 (* 1 = 0.573297 loss) I0412 13:47:51.765738 8032 sgd_solver.cpp:105] Iteration 5616, lr = 0.00328752 I0412 13:47:57.052523 8032 solver.cpp:218] Iteration 5628 (2.26988 iter/s, 5.28662s/12 iters), loss = 0.654063 I0412 13:47:57.052577 8032 solver.cpp:237] Train net output #0: loss = 0.654063 (* 1 = 0.654063 loss) I0412 13:47:57.052589 8032 sgd_solver.cpp:105] Iteration 5628, lr = 0.00327972 I0412 13:48:02.105141 8032 solver.cpp:218] Iteration 5640 (2.37511 iter/s, 5.0524s/12 iters), loss = 0.681815 I0412 13:48:02.105185 8032 solver.cpp:237] Train net output #0: loss = 0.681815 (* 1 = 0.681815 loss) I0412 13:48:02.105193 8032 sgd_solver.cpp:105] Iteration 5640, lr = 0.00327193 I0412 13:48:07.123132 8032 solver.cpp:218] Iteration 5652 (2.39149 iter/s, 5.01779s/12 iters), loss = 0.511187 I0412 13:48:07.123170 8032 solver.cpp:237] Train net output #0: loss = 0.511187 (* 1 = 0.511187 loss) I0412 13:48:07.123178 8032 sgd_solver.cpp:105] Iteration 5652, lr = 0.00326416 I0412 13:48:12.092872 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:48:12.272274 8032 solver.cpp:218] Iteration 5664 (2.33058 iter/s, 5.14893s/12 iters), loss = 0.710445 I0412 13:48:12.272326 8032 solver.cpp:237] Train net output #0: loss = 0.710445 (* 1 = 0.710445 loss) I0412 13:48:12.272337 8032 sgd_solver.cpp:105] Iteration 5664, lr = 0.00325641 I0412 13:48:17.363587 8032 solver.cpp:218] Iteration 5676 (2.35706 iter/s, 5.09109s/12 iters), loss = 0.647534 I0412 13:48:17.363641 8032 solver.cpp:237] Train net output #0: loss = 0.647534 (* 1 = 0.647534 loss) I0412 13:48:17.363652 8032 sgd_solver.cpp:105] Iteration 5676, lr = 0.00324868 I0412 13:48:22.648279 8032 solver.cpp:218] Iteration 5688 (2.27081 iter/s, 5.28446s/12 iters), loss = 0.61363 I0412 13:48:22.648442 8032 solver.cpp:237] Train net output #0: loss = 0.61363 (* 1 = 0.61363 loss) I0412 13:48:22.648453 8032 sgd_solver.cpp:105] Iteration 5688, lr = 0.00324097 I0412 13:48:27.771713 8032 solver.cpp:218] Iteration 5700 (2.34233 iter/s, 5.1231s/12 iters), loss = 0.60316 I0412 13:48:27.771770 8032 solver.cpp:237] Train net output #0: loss = 0.60316 (* 1 = 0.60316 loss) I0412 13:48:27.771782 8032 sgd_solver.cpp:105] Iteration 5700, lr = 0.00323328 I0412 13:48:32.372822 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_5712.caffemodel I0412 13:48:36.993499 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_5712.solverstate I0412 13:48:40.780544 8032 solver.cpp:330] Iteration 5712, Testing net (#0) I0412 13:48:40.780568 8032 net.cpp:676] Ignoring source layer train-data I0412 13:48:43.002480 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:48:45.377094 8032 solver.cpp:397] Test net output #0: accuracy = 0.433211 I0412 13:48:45.377135 8032 solver.cpp:397] Test net output #1: loss = 2.64936 (* 1 = 2.64936 loss) I0412 13:48:45.465062 8032 solver.cpp:218] Iteration 5712 (0.678244 iter/s, 17.6927s/12 iters), loss = 0.55574 I0412 13:48:45.465107 8032 solver.cpp:237] Train net output #0: loss = 0.55574 (* 1 = 0.55574 loss) I0412 13:48:45.465117 8032 sgd_solver.cpp:105] Iteration 5712, lr = 0.0032256 I0412 13:48:49.860154 8032 solver.cpp:218] Iteration 5724 (2.73044 iter/s, 4.39489s/12 iters), loss = 0.656344 I0412 13:48:49.860208 8032 solver.cpp:237] Train net output #0: loss = 0.656344 (* 1 = 0.656344 loss) I0412 13:48:49.860219 8032 sgd_solver.cpp:105] Iteration 5724, lr = 0.00321794 I0412 13:48:55.354441 8032 solver.cpp:218] Iteration 5736 (2.18418 iter/s, 5.49405s/12 iters), loss = 0.421302 I0412 13:48:55.354938 8032 solver.cpp:237] Train net output #0: loss = 0.421302 (* 1 = 0.421302 loss) I0412 13:48:55.354954 8032 sgd_solver.cpp:105] Iteration 5736, lr = 0.0032103 I0412 13:49:00.431428 8032 solver.cpp:218] Iteration 5748 (2.36391 iter/s, 5.07633s/12 iters), loss = 0.529667 I0412 13:49:00.431478 8032 solver.cpp:237] Train net output #0: loss = 0.529667 (* 1 = 0.529667 loss) I0412 13:49:00.431489 8032 sgd_solver.cpp:105] Iteration 5748, lr = 0.00320268 I0412 13:49:05.511390 8032 solver.cpp:218] Iteration 5760 (2.36232 iter/s, 5.07974s/12 iters), loss = 0.543959 I0412 13:49:05.511445 8032 solver.cpp:237] Train net output #0: loss = 0.543959 (* 1 = 0.543959 loss) I0412 13:49:05.511457 8032 sgd_solver.cpp:105] Iteration 5760, lr = 0.00319508 I0412 13:49:07.471925 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:49:10.538933 8032 solver.cpp:218] Iteration 5772 (2.38696 iter/s, 5.02732s/12 iters), loss = 0.531806 I0412 13:49:10.538991 8032 solver.cpp:237] Train net output #0: loss = 0.531806 (* 1 = 0.531806 loss) I0412 13:49:10.539002 8032 sgd_solver.cpp:105] Iteration 5772, lr = 0.00318749 I0412 13:49:15.618010 8032 solver.cpp:218] Iteration 5784 (2.36274 iter/s, 5.07885s/12 iters), loss = 0.448747 I0412 13:49:15.618059 8032 solver.cpp:237] Train net output #0: loss = 0.448747 (* 1 = 0.448747 loss) I0412 13:49:15.618069 8032 sgd_solver.cpp:105] Iteration 5784, lr = 0.00317992 I0412 13:49:20.794085 8032 solver.cpp:218] Iteration 5796 (2.31846 iter/s, 5.17585s/12 iters), loss = 0.583601 I0412 13:49:20.794135 8032 solver.cpp:237] Train net output #0: loss = 0.583601 (* 1 = 0.583601 loss) I0412 13:49:20.794147 8032 sgd_solver.cpp:105] Iteration 5796, lr = 0.00317237 I0412 13:49:26.026564 8032 solver.cpp:218] Iteration 5808 (2.29347 iter/s, 5.23224s/12 iters), loss = 0.610117 I0412 13:49:26.026723 8032 solver.cpp:237] Train net output #0: loss = 0.610117 (* 1 = 0.610117 loss) I0412 13:49:26.026736 8032 sgd_solver.cpp:105] Iteration 5808, lr = 0.00316484 I0412 13:49:28.094571 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_5814.caffemodel I0412 13:49:33.748107 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_5814.solverstate I0412 13:49:36.078294 8032 solver.cpp:330] Iteration 5814, Testing net (#0) I0412 13:49:36.078320 8032 net.cpp:676] Ignoring source layer train-data I0412 13:49:38.121317 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:49:40.537670 8032 solver.cpp:397] Test net output #0: accuracy = 0.428922 I0412 13:49:40.537714 8032 solver.cpp:397] Test net output #1: loss = 2.60024 (* 1 = 2.60024 loss) I0412 13:49:42.537175 8032 solver.cpp:218] Iteration 5820 (0.726835 iter/s, 16.5099s/12 iters), loss = 0.44928 I0412 13:49:42.537228 8032 solver.cpp:237] Train net output #0: loss = 0.44928 (* 1 = 0.44928 loss) I0412 13:49:42.537240 8032 sgd_solver.cpp:105] Iteration 5820, lr = 0.00315733 I0412 13:49:47.828418 8032 solver.cpp:218] Iteration 5832 (2.26799 iter/s, 5.29102s/12 iters), loss = 0.378925 I0412 13:49:47.828464 8032 solver.cpp:237] Train net output #0: loss = 0.378925 (* 1 = 0.378925 loss) I0412 13:49:47.828474 8032 sgd_solver.cpp:105] Iteration 5832, lr = 0.00314983 I0412 13:49:52.956218 8032 solver.cpp:218] Iteration 5844 (2.34028 iter/s, 5.12758s/12 iters), loss = 0.540987 I0412 13:49:52.956257 8032 solver.cpp:237] Train net output #0: loss = 0.540987 (* 1 = 0.540987 loss) I0412 13:49:52.956265 8032 sgd_solver.cpp:105] Iteration 5844, lr = 0.00314235 I0412 13:49:58.314491 8032 solver.cpp:218] Iteration 5856 (2.23962 iter/s, 5.35806s/12 iters), loss = 0.405285 I0412 13:49:58.314560 8032 solver.cpp:237] Train net output #0: loss = 0.405285 (* 1 = 0.405285 loss) I0412 13:49:58.314570 8032 sgd_solver.cpp:105] Iteration 5856, lr = 0.00313489 I0412 13:50:02.556493 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:50:03.378208 8032 solver.cpp:218] Iteration 5868 (2.36991 iter/s, 5.06349s/12 iters), loss = 0.562451 I0412 13:50:03.378250 8032 solver.cpp:237] Train net output #0: loss = 0.562451 (* 1 = 0.562451 loss) I0412 13:50:03.378259 8032 sgd_solver.cpp:105] Iteration 5868, lr = 0.00312745 I0412 13:50:08.493690 8032 solver.cpp:218] Iteration 5880 (2.34592 iter/s, 5.11527s/12 iters), loss = 0.463132 I0412 13:50:08.493743 8032 solver.cpp:237] Train net output #0: loss = 0.463132 (* 1 = 0.463132 loss) I0412 13:50:08.493755 8032 sgd_solver.cpp:105] Iteration 5880, lr = 0.00312002 I0412 13:50:13.861925 8032 solver.cpp:218] Iteration 5892 (2.23547 iter/s, 5.368s/12 iters), loss = 0.63882 I0412 13:50:13.862000 8032 solver.cpp:237] Train net output #0: loss = 0.63882 (* 1 = 0.63882 loss) I0412 13:50:13.862012 8032 sgd_solver.cpp:105] Iteration 5892, lr = 0.00311262 I0412 13:50:18.825266 8032 solver.cpp:218] Iteration 5904 (2.41784 iter/s, 4.9631s/12 iters), loss = 0.614667 I0412 13:50:18.825317 8032 solver.cpp:237] Train net output #0: loss = 0.614667 (* 1 = 0.614667 loss) I0412 13:50:18.825330 8032 sgd_solver.cpp:105] Iteration 5904, lr = 0.00310523 I0412 13:50:23.399298 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_5916.caffemodel I0412 13:50:26.517170 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_5916.solverstate I0412 13:50:28.887768 8032 solver.cpp:330] Iteration 5916, Testing net (#0) I0412 13:50:28.887872 8032 net.cpp:676] Ignoring source layer train-data I0412 13:50:31.006510 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:50:33.392163 8032 solver.cpp:397] Test net output #0: accuracy = 0.422181 I0412 13:50:33.392192 8032 solver.cpp:397] Test net output #1: loss = 2.84669 (* 1 = 2.84669 loss) I0412 13:50:33.480181 8032 solver.cpp:218] Iteration 5916 (0.818866 iter/s, 14.6544s/12 iters), loss = 0.512181 I0412 13:50:33.480229 8032 solver.cpp:237] Train net output #0: loss = 0.512181 (* 1 = 0.512181 loss) I0412 13:50:33.480240 8032 sgd_solver.cpp:105] Iteration 5916, lr = 0.00309785 I0412 13:50:37.809981 8032 solver.cpp:218] Iteration 5928 (2.77162 iter/s, 4.32959s/12 iters), loss = 0.520547 I0412 13:50:37.810026 8032 solver.cpp:237] Train net output #0: loss = 0.520547 (* 1 = 0.520547 loss) I0412 13:50:37.810035 8032 sgd_solver.cpp:105] Iteration 5928, lr = 0.0030905 I0412 13:50:42.913367 8032 solver.cpp:218] Iteration 5940 (2.35148 iter/s, 5.10317s/12 iters), loss = 0.417839 I0412 13:50:42.913414 8032 solver.cpp:237] Train net output #0: loss = 0.417839 (* 1 = 0.417839 loss) I0412 13:50:42.913422 8032 sgd_solver.cpp:105] Iteration 5940, lr = 0.00308316 I0412 13:50:48.050673 8032 solver.cpp:218] Iteration 5952 (2.33595 iter/s, 5.13709s/12 iters), loss = 0.512849 I0412 13:50:48.050720 8032 solver.cpp:237] Train net output #0: loss = 0.512849 (* 1 = 0.512849 loss) I0412 13:50:48.050730 8032 sgd_solver.cpp:105] Iteration 5952, lr = 0.00307584 I0412 13:50:53.101094 8032 solver.cpp:218] Iteration 5964 (2.37614 iter/s, 5.05021s/12 iters), loss = 0.498528 I0412 13:50:53.101145 8032 solver.cpp:237] Train net output #0: loss = 0.498528 (* 1 = 0.498528 loss) I0412 13:50:53.101157 8032 sgd_solver.cpp:105] Iteration 5964, lr = 0.00306854 I0412 13:50:54.458338 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:50:58.152333 8032 solver.cpp:218] Iteration 5976 (2.37576 iter/s, 5.05103s/12 iters), loss = 0.713401 I0412 13:50:58.152375 8032 solver.cpp:237] Train net output #0: loss = 0.713401 (* 1 = 0.713401 loss) I0412 13:50:58.152384 8032 sgd_solver.cpp:105] Iteration 5976, lr = 0.00306125 I0412 13:51:03.402113 8032 solver.cpp:218] Iteration 5988 (2.28591 iter/s, 5.24956s/12 iters), loss = 0.497363 I0412 13:51:03.402228 8032 solver.cpp:237] Train net output #0: loss = 0.497363 (* 1 = 0.497363 loss) I0412 13:51:03.402240 8032 sgd_solver.cpp:105] Iteration 5988, lr = 0.00305398 I0412 13:51:08.383919 8032 solver.cpp:218] Iteration 6000 (2.4089 iter/s, 4.98153s/12 iters), loss = 0.550455 I0412 13:51:08.383962 8032 solver.cpp:237] Train net output #0: loss = 0.550455 (* 1 = 0.550455 loss) I0412 13:51:08.383970 8032 sgd_solver.cpp:105] Iteration 6000, lr = 0.00304673 I0412 13:51:13.449205 8032 solver.cpp:218] Iteration 6012 (2.36916 iter/s, 5.06508s/12 iters), loss = 0.647086 I0412 13:51:13.449256 8032 solver.cpp:237] Train net output #0: loss = 0.647086 (* 1 = 0.647086 loss) I0412 13:51:13.449267 8032 sgd_solver.cpp:105] Iteration 6012, lr = 0.0030395 I0412 13:51:15.628448 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6018.caffemodel I0412 13:51:21.882604 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6018.solverstate I0412 13:51:24.193233 8032 solver.cpp:330] Iteration 6018, Testing net (#0) I0412 13:51:24.193256 8032 net.cpp:676] Ignoring source layer train-data I0412 13:51:26.499579 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:51:28.870918 8032 solver.cpp:397] Test net output #0: accuracy = 0.440564 I0412 13:51:28.870968 8032 solver.cpp:397] Test net output #1: loss = 2.88678 (* 1 = 2.88678 loss) I0412 13:51:30.877542 8032 solver.cpp:218] Iteration 6024 (0.688557 iter/s, 17.4277s/12 iters), loss = 0.622547 I0412 13:51:30.877595 8032 solver.cpp:237] Train net output #0: loss = 0.622547 (* 1 = 0.622547 loss) I0412 13:51:30.877606 8032 sgd_solver.cpp:105] Iteration 6024, lr = 0.00303228 I0412 13:51:36.123050 8032 solver.cpp:218] Iteration 6036 (2.28777 iter/s, 5.24528s/12 iters), loss = 0.500513 I0412 13:51:36.123198 8032 solver.cpp:237] Train net output #0: loss = 0.500513 (* 1 = 0.500513 loss) I0412 13:51:36.123210 8032 sgd_solver.cpp:105] Iteration 6036, lr = 0.00302508 I0412 13:51:41.294409 8032 solver.cpp:218] Iteration 6048 (2.32061 iter/s, 5.17105s/12 iters), loss = 0.397394 I0412 13:51:41.294456 8032 solver.cpp:237] Train net output #0: loss = 0.397394 (* 1 = 0.397394 loss) I0412 13:51:41.294467 8032 sgd_solver.cpp:105] Iteration 6048, lr = 0.0030179 I0412 13:51:46.598621 8032 solver.cpp:218] Iteration 6060 (2.26245 iter/s, 5.30399s/12 iters), loss = 0.660294 I0412 13:51:46.598673 8032 solver.cpp:237] Train net output #0: loss = 0.660294 (* 1 = 0.660294 loss) I0412 13:51:46.598685 8032 sgd_solver.cpp:105] Iteration 6060, lr = 0.00301074 I0412 13:51:50.160364 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:51:51.725103 8032 solver.cpp:218] Iteration 6072 (2.34089 iter/s, 5.12626s/12 iters), loss = 0.439985 I0412 13:51:51.725157 8032 solver.cpp:237] Train net output #0: loss = 0.439985 (* 1 = 0.439985 loss) I0412 13:51:51.725167 8032 sgd_solver.cpp:105] Iteration 6072, lr = 0.00300359 I0412 13:51:56.825331 8032 solver.cpp:218] Iteration 6084 (2.35294 iter/s, 5.1s/12 iters), loss = 0.509449 I0412 13:51:56.825387 8032 solver.cpp:237] Train net output #0: loss = 0.509449 (* 1 = 0.509449 loss) I0412 13:51:56.825399 8032 sgd_solver.cpp:105] Iteration 6084, lr = 0.00299646 I0412 13:52:01.843922 8032 solver.cpp:218] Iteration 6096 (2.39122 iter/s, 5.01836s/12 iters), loss = 0.376423 I0412 13:52:01.843976 8032 solver.cpp:237] Train net output #0: loss = 0.376423 (* 1 = 0.376423 loss) I0412 13:52:01.843989 8032 sgd_solver.cpp:105] Iteration 6096, lr = 0.00298934 I0412 13:52:06.769502 8032 solver.cpp:218] Iteration 6108 (2.43637 iter/s, 4.92536s/12 iters), loss = 0.657602 I0412 13:52:06.769616 8032 solver.cpp:237] Train net output #0: loss = 0.657602 (* 1 = 0.657602 loss) I0412 13:52:06.769629 8032 sgd_solver.cpp:105] Iteration 6108, lr = 0.00298225 I0412 13:52:11.397292 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6120.caffemodel I0412 13:52:16.735256 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6120.solverstate I0412 13:52:19.685899 8032 solver.cpp:330] Iteration 6120, Testing net (#0) I0412 13:52:19.685920 8032 net.cpp:676] Ignoring source layer train-data I0412 13:52:21.732754 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:52:24.166296 8032 solver.cpp:397] Test net output #0: accuracy = 0.43076 I0412 13:52:24.166344 8032 solver.cpp:397] Test net output #1: loss = 2.80628 (* 1 = 2.80628 loss) I0412 13:52:24.254560 8032 solver.cpp:218] Iteration 6120 (0.686326 iter/s, 17.4844s/12 iters), loss = 0.376542 I0412 13:52:24.254611 8032 solver.cpp:237] Train net output #0: loss = 0.376542 (* 1 = 0.376542 loss) I0412 13:52:24.254621 8032 sgd_solver.cpp:105] Iteration 6120, lr = 0.00297517 I0412 13:52:28.490804 8032 solver.cpp:218] Iteration 6132 (2.83283 iter/s, 4.23605s/12 iters), loss = 0.476296 I0412 13:52:28.490856 8032 solver.cpp:237] Train net output #0: loss = 0.476296 (* 1 = 0.476296 loss) I0412 13:52:28.490869 8032 sgd_solver.cpp:105] Iteration 6132, lr = 0.0029681 I0412 13:52:33.534430 8032 solver.cpp:218] Iteration 6144 (2.37934 iter/s, 5.04341s/12 iters), loss = 0.541534 I0412 13:52:33.534485 8032 solver.cpp:237] Train net output #0: loss = 0.541534 (* 1 = 0.541534 loss) I0412 13:52:33.534497 8032 sgd_solver.cpp:105] Iteration 6144, lr = 0.00296105 I0412 13:52:38.533335 8032 solver.cpp:218] Iteration 6156 (2.40063 iter/s, 4.99868s/12 iters), loss = 0.294288 I0412 13:52:38.533476 8032 solver.cpp:237] Train net output #0: loss = 0.294288 (* 1 = 0.294288 loss) I0412 13:52:38.533488 8032 sgd_solver.cpp:105] Iteration 6156, lr = 0.00295402 I0412 13:52:43.507555 8032 solver.cpp:218] Iteration 6168 (2.41259 iter/s, 4.97391s/12 iters), loss = 0.47401 I0412 13:52:43.507606 8032 solver.cpp:237] Train net output #0: loss = 0.47401 (* 1 = 0.47401 loss) I0412 13:52:43.507617 8032 sgd_solver.cpp:105] Iteration 6168, lr = 0.00294701 I0412 13:52:44.097697 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:52:48.657369 8032 solver.cpp:218] Iteration 6180 (2.33028 iter/s, 5.14959s/12 iters), loss = 0.514648 I0412 13:52:48.657424 8032 solver.cpp:237] Train net output #0: loss = 0.514648 (* 1 = 0.514648 loss) I0412 13:52:48.657438 8032 sgd_solver.cpp:105] Iteration 6180, lr = 0.00294001 I0412 13:52:53.741407 8032 solver.cpp:218] Iteration 6192 (2.36043 iter/s, 5.08382s/12 iters), loss = 0.642243 I0412 13:52:53.741459 8032 solver.cpp:237] Train net output #0: loss = 0.642243 (* 1 = 0.642243 loss) I0412 13:52:53.741470 8032 sgd_solver.cpp:105] Iteration 6192, lr = 0.00293303 I0412 13:52:58.822930 8032 solver.cpp:218] Iteration 6204 (2.3616 iter/s, 5.0813s/12 iters), loss = 0.399468 I0412 13:52:58.822986 8032 solver.cpp:237] Train net output #0: loss = 0.399468 (* 1 = 0.399468 loss) I0412 13:52:58.822999 8032 sgd_solver.cpp:105] Iteration 6204, lr = 0.00292607 I0412 13:53:03.930892 8032 solver.cpp:218] Iteration 6216 (2.34938 iter/s, 5.10774s/12 iters), loss = 0.500883 I0412 13:53:03.930936 8032 solver.cpp:237] Train net output #0: loss = 0.500883 (* 1 = 0.500883 loss) I0412 13:53:03.930945 8032 sgd_solver.cpp:105] Iteration 6216, lr = 0.00291912 I0412 13:53:05.967878 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6222.caffemodel I0412 13:53:12.552314 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6222.solverstate I0412 13:53:15.009722 8032 solver.cpp:330] Iteration 6222, Testing net (#0) I0412 13:53:15.009748 8032 net.cpp:676] Ignoring source layer train-data I0412 13:53:17.202740 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:53:18.614686 8032 blocking_queue.cpp:49] Waiting for data I0412 13:53:19.932268 8032 solver.cpp:397] Test net output #0: accuracy = 0.433211 I0412 13:53:19.932313 8032 solver.cpp:397] Test net output #1: loss = 2.86917 (* 1 = 2.86917 loss) I0412 13:53:21.831997 8032 solver.cpp:218] Iteration 6228 (0.670372 iter/s, 17.9005s/12 iters), loss = 0.407367 I0412 13:53:21.832054 8032 solver.cpp:237] Train net output #0: loss = 0.407367 (* 1 = 0.407367 loss) I0412 13:53:21.832065 8032 sgd_solver.cpp:105] Iteration 6228, lr = 0.00291219 I0412 13:53:26.868752 8032 solver.cpp:218] Iteration 6240 (2.38259 iter/s, 5.03653s/12 iters), loss = 0.529889 I0412 13:53:26.868811 8032 solver.cpp:237] Train net output #0: loss = 0.529889 (* 1 = 0.529889 loss) I0412 13:53:26.868824 8032 sgd_solver.cpp:105] Iteration 6240, lr = 0.00290528 I0412 13:53:31.849694 8032 solver.cpp:218] Iteration 6252 (2.40929 iter/s, 4.98071s/12 iters), loss = 0.254053 I0412 13:53:31.849761 8032 solver.cpp:237] Train net output #0: loss = 0.254053 (* 1 = 0.254053 loss) I0412 13:53:31.849773 8032 sgd_solver.cpp:105] Iteration 6252, lr = 0.00289838 I0412 13:53:36.888202 8032 solver.cpp:218] Iteration 6264 (2.38177 iter/s, 5.03827s/12 iters), loss = 0.351192 I0412 13:53:36.888252 8032 solver.cpp:237] Train net output #0: loss = 0.351192 (* 1 = 0.351192 loss) I0412 13:53:36.888263 8032 sgd_solver.cpp:105] Iteration 6264, lr = 0.0028915 I0412 13:53:39.647927 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:53:41.972805 8032 solver.cpp:218] Iteration 6276 (2.36017 iter/s, 5.08439s/12 iters), loss = 0.456425 I0412 13:53:41.972856 8032 solver.cpp:237] Train net output #0: loss = 0.456425 (* 1 = 0.456425 loss) I0412 13:53:41.972867 8032 sgd_solver.cpp:105] Iteration 6276, lr = 0.00288463 I0412 13:53:47.489675 8032 solver.cpp:218] Iteration 6288 (2.17524 iter/s, 5.51664s/12 iters), loss = 0.526326 I0412 13:53:47.489815 8032 solver.cpp:237] Train net output #0: loss = 0.526326 (* 1 = 0.526326 loss) I0412 13:53:47.489827 8032 sgd_solver.cpp:105] Iteration 6288, lr = 0.00287779 I0412 13:53:52.887301 8032 solver.cpp:218] Iteration 6300 (2.22333 iter/s, 5.39731s/12 iters), loss = 0.378523 I0412 13:53:52.887342 8032 solver.cpp:237] Train net output #0: loss = 0.378523 (* 1 = 0.378523 loss) I0412 13:53:52.887353 8032 sgd_solver.cpp:105] Iteration 6300, lr = 0.00287095 I0412 13:53:58.057461 8032 solver.cpp:218] Iteration 6312 (2.32111 iter/s, 5.16994s/12 iters), loss = 0.478188 I0412 13:53:58.057503 8032 solver.cpp:237] Train net output #0: loss = 0.478188 (* 1 = 0.478188 loss) I0412 13:53:58.057512 8032 sgd_solver.cpp:105] Iteration 6312, lr = 0.00286414 I0412 13:54:02.611198 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6324.caffemodel I0412 13:54:07.848058 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6324.solverstate I0412 13:54:10.830294 8032 solver.cpp:330] Iteration 6324, Testing net (#0) I0412 13:54:10.830322 8032 net.cpp:676] Ignoring source layer train-data I0412 13:54:12.817545 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:54:15.602856 8032 solver.cpp:397] Test net output #0: accuracy = 0.417279 I0412 13:54:15.602890 8032 solver.cpp:397] Test net output #1: loss = 3.01747 (* 1 = 3.01747 loss) I0412 13:54:15.690865 8032 solver.cpp:218] Iteration 6324 (0.680549 iter/s, 17.6328s/12 iters), loss = 0.336868 I0412 13:54:15.690910 8032 solver.cpp:237] Train net output #0: loss = 0.336868 (* 1 = 0.336868 loss) I0412 13:54:15.690918 8032 sgd_solver.cpp:105] Iteration 6324, lr = 0.00285734 I0412 13:54:20.306118 8032 solver.cpp:218] Iteration 6336 (2.60019 iter/s, 4.61505s/12 iters), loss = 0.308992 I0412 13:54:20.313988 8032 solver.cpp:237] Train net output #0: loss = 0.308992 (* 1 = 0.308992 loss) I0412 13:54:20.314002 8032 sgd_solver.cpp:105] Iteration 6336, lr = 0.00285055 I0412 13:54:25.524377 8032 solver.cpp:218] Iteration 6348 (2.30315 iter/s, 5.21024s/12 iters), loss = 0.218251 I0412 13:54:25.524435 8032 solver.cpp:237] Train net output #0: loss = 0.218251 (* 1 = 0.218251 loss) I0412 13:54:25.524446 8032 sgd_solver.cpp:105] Iteration 6348, lr = 0.00284379 I0412 13:54:30.742241 8032 solver.cpp:218] Iteration 6360 (2.29989 iter/s, 5.21763s/12 iters), loss = 0.500516 I0412 13:54:30.742292 8032 solver.cpp:237] Train net output #0: loss = 0.500516 (* 1 = 0.500516 loss) I0412 13:54:30.742303 8032 sgd_solver.cpp:105] Iteration 6360, lr = 0.00283703 I0412 13:54:35.790961 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:54:35.929992 8032 solver.cpp:218] Iteration 6372 (2.31324 iter/s, 5.18753s/12 iters), loss = 0.443318 I0412 13:54:35.930040 8032 solver.cpp:237] Train net output #0: loss = 0.443318 (* 1 = 0.443318 loss) I0412 13:54:35.930052 8032 sgd_solver.cpp:105] Iteration 6372, lr = 0.0028303 I0412 13:54:40.930819 8032 solver.cpp:218] Iteration 6384 (2.3997 iter/s, 5.00062s/12 iters), loss = 0.349584 I0412 13:54:40.930866 8032 solver.cpp:237] Train net output #0: loss = 0.349584 (* 1 = 0.349584 loss) I0412 13:54:40.930876 8032 sgd_solver.cpp:105] Iteration 6384, lr = 0.00282358 I0412 13:54:45.802330 8032 solver.cpp:218] Iteration 6396 (2.46341 iter/s, 4.8713s/12 iters), loss = 0.416847 I0412 13:54:45.802387 8032 solver.cpp:237] Train net output #0: loss = 0.416847 (* 1 = 0.416847 loss) I0412 13:54:45.802399 8032 sgd_solver.cpp:105] Iteration 6396, lr = 0.00281687 I0412 13:54:50.787135 8032 solver.cpp:218] Iteration 6408 (2.40742 iter/s, 4.98458s/12 iters), loss = 0.453566 I0412 13:54:50.787235 8032 solver.cpp:237] Train net output #0: loss = 0.453566 (* 1 = 0.453566 loss) I0412 13:54:50.787245 8032 sgd_solver.cpp:105] Iteration 6408, lr = 0.00281019 I0412 13:54:56.014158 8032 solver.cpp:218] Iteration 6420 (2.29588 iter/s, 5.22675s/12 iters), loss = 0.463255 I0412 13:54:56.014214 8032 solver.cpp:237] Train net output #0: loss = 0.463255 (* 1 = 0.463255 loss) I0412 13:54:56.014226 8032 sgd_solver.cpp:105] Iteration 6420, lr = 0.00280351 I0412 13:54:58.044258 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6426.caffemodel I0412 13:55:03.153703 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6426.solverstate I0412 13:55:07.456389 8032 solver.cpp:330] Iteration 6426, Testing net (#0) I0412 13:55:07.456414 8032 net.cpp:676] Ignoring source layer train-data I0412 13:55:09.536643 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:55:12.069182 8032 solver.cpp:397] Test net output #0: accuracy = 0.439951 I0412 13:55:12.069218 8032 solver.cpp:397] Test net output #1: loss = 2.80703 (* 1 = 2.80703 loss) I0412 13:55:13.883836 8032 solver.cpp:218] Iteration 6432 (0.671552 iter/s, 17.8691s/12 iters), loss = 0.415725 I0412 13:55:13.883893 8032 solver.cpp:237] Train net output #0: loss = 0.415725 (* 1 = 0.415725 loss) I0412 13:55:13.883904 8032 sgd_solver.cpp:105] Iteration 6432, lr = 0.00279686 I0412 13:55:18.931015 8032 solver.cpp:218] Iteration 6444 (2.37767 iter/s, 5.04696s/12 iters), loss = 0.461105 I0412 13:55:18.931068 8032 solver.cpp:237] Train net output #0: loss = 0.461105 (* 1 = 0.461105 loss) I0412 13:55:18.931082 8032 sgd_solver.cpp:105] Iteration 6444, lr = 0.00279022 I0412 13:55:23.852954 8032 solver.cpp:218] Iteration 6456 (2.43817 iter/s, 4.92172s/12 iters), loss = 0.331421 I0412 13:55:23.853089 8032 solver.cpp:237] Train net output #0: loss = 0.331421 (* 1 = 0.331421 loss) I0412 13:55:23.853101 8032 sgd_solver.cpp:105] Iteration 6456, lr = 0.00278359 I0412 13:55:29.162762 8032 solver.cpp:218] Iteration 6468 (2.2601 iter/s, 5.3095s/12 iters), loss = 0.373152 I0412 13:55:29.162813 8032 solver.cpp:237] Train net output #0: loss = 0.373152 (* 1 = 0.373152 loss) I0412 13:55:29.162824 8032 sgd_solver.cpp:105] Iteration 6468, lr = 0.00277698 I0412 13:55:31.262974 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:55:34.409433 8032 solver.cpp:218] Iteration 6480 (2.28726 iter/s, 5.24645s/12 iters), loss = 0.504205 I0412 13:55:34.409487 8032 solver.cpp:237] Train net output #0: loss = 0.504205 (* 1 = 0.504205 loss) I0412 13:55:34.409499 8032 sgd_solver.cpp:105] Iteration 6480, lr = 0.00277039 I0412 13:55:39.649250 8032 solver.cpp:218] Iteration 6492 (2.29025 iter/s, 5.23959s/12 iters), loss = 0.210692 I0412 13:55:39.649298 8032 solver.cpp:237] Train net output #0: loss = 0.210692 (* 1 = 0.210692 loss) I0412 13:55:39.649308 8032 sgd_solver.cpp:105] Iteration 6492, lr = 0.00276381 I0412 13:55:44.712345 8032 solver.cpp:218] Iteration 6504 (2.37019 iter/s, 5.06287s/12 iters), loss = 0.363648 I0412 13:55:44.712401 8032 solver.cpp:237] Train net output #0: loss = 0.363648 (* 1 = 0.363648 loss) I0412 13:55:44.712414 8032 sgd_solver.cpp:105] Iteration 6504, lr = 0.00275725 I0412 13:55:49.705776 8032 solver.cpp:218] Iteration 6516 (2.40326 iter/s, 4.99321s/12 iters), loss = 0.304427 I0412 13:55:49.705821 8032 solver.cpp:237] Train net output #0: loss = 0.304427 (* 1 = 0.304427 loss) I0412 13:55:49.705830 8032 sgd_solver.cpp:105] Iteration 6516, lr = 0.00275071 I0412 13:55:54.262748 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6528.caffemodel I0412 13:55:57.298190 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6528.solverstate I0412 13:56:04.947592 8032 solver.cpp:330] Iteration 6528, Testing net (#0) I0412 13:56:04.947616 8032 net.cpp:676] Ignoring source layer train-data I0412 13:56:06.774150 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:56:09.559496 8032 solver.cpp:397] Test net output #0: accuracy = 0.434436 I0412 13:56:09.559536 8032 solver.cpp:397] Test net output #1: loss = 2.91791 (* 1 = 2.91791 loss) I0412 13:56:09.647703 8032 solver.cpp:218] Iteration 6528 (0.601767 iter/s, 19.9413s/12 iters), loss = 0.341807 I0412 13:56:09.647748 8032 solver.cpp:237] Train net output #0: loss = 0.341807 (* 1 = 0.341807 loss) I0412 13:56:09.647758 8032 sgd_solver.cpp:105] Iteration 6528, lr = 0.00274418 I0412 13:56:14.120914 8032 solver.cpp:218] Iteration 6540 (2.68275 iter/s, 4.47302s/12 iters), loss = 0.631898 I0412 13:56:14.120957 8032 solver.cpp:237] Train net output #0: loss = 0.631898 (* 1 = 0.631898 loss) I0412 13:56:14.120965 8032 sgd_solver.cpp:105] Iteration 6540, lr = 0.00273766 I0412 13:56:19.402801 8032 solver.cpp:218] Iteration 6552 (2.27201 iter/s, 5.28167s/12 iters), loss = 0.300827 I0412 13:56:19.402842 8032 solver.cpp:237] Train net output #0: loss = 0.300827 (* 1 = 0.300827 loss) I0412 13:56:19.402853 8032 sgd_solver.cpp:105] Iteration 6552, lr = 0.00273116 I0412 13:56:24.464674 8032 solver.cpp:218] Iteration 6564 (2.37076 iter/s, 5.06166s/12 iters), loss = 0.418869 I0412 13:56:24.464762 8032 solver.cpp:237] Train net output #0: loss = 0.418869 (* 1 = 0.418869 loss) I0412 13:56:24.464772 8032 sgd_solver.cpp:105] Iteration 6564, lr = 0.00272468 I0412 13:56:28.820263 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:56:29.563828 8032 solver.cpp:218] Iteration 6576 (2.35345 iter/s, 5.09889s/12 iters), loss = 0.467589 I0412 13:56:29.563884 8032 solver.cpp:237] Train net output #0: loss = 0.467589 (* 1 = 0.467589 loss) I0412 13:56:29.563894 8032 sgd_solver.cpp:105] Iteration 6576, lr = 0.00271821 I0412 13:56:34.551934 8032 solver.cpp:218] Iteration 6588 (2.40583 iter/s, 4.98788s/12 iters), loss = 0.294446 I0412 13:56:34.551983 8032 solver.cpp:237] Train net output #0: loss = 0.294446 (* 1 = 0.294446 loss) I0412 13:56:34.551995 8032 sgd_solver.cpp:105] Iteration 6588, lr = 0.00271175 I0412 13:56:39.529819 8032 solver.cpp:218] Iteration 6600 (2.41077 iter/s, 4.97767s/12 iters), loss = 0.324539 I0412 13:56:39.529872 8032 solver.cpp:237] Train net output #0: loss = 0.324539 (* 1 = 0.324539 loss) I0412 13:56:39.529886 8032 sgd_solver.cpp:105] Iteration 6600, lr = 0.00270532 I0412 13:56:44.698285 8032 solver.cpp:218] Iteration 6612 (2.32187 iter/s, 5.16824s/12 iters), loss = 0.276562 I0412 13:56:44.698345 8032 solver.cpp:237] Train net output #0: loss = 0.276562 (* 1 = 0.276562 loss) I0412 13:56:44.698359 8032 sgd_solver.cpp:105] Iteration 6612, lr = 0.00269889 I0412 13:56:49.849684 8032 solver.cpp:218] Iteration 6624 (2.32957 iter/s, 5.15117s/12 iters), loss = 0.439997 I0412 13:56:49.849735 8032 solver.cpp:237] Train net output #0: loss = 0.439997 (* 1 = 0.439997 loss) I0412 13:56:49.849746 8032 sgd_solver.cpp:105] Iteration 6624, lr = 0.00269248 I0412 13:56:51.856118 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6630.caffemodel I0412 13:56:55.060600 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6630.solverstate I0412 13:56:58.043846 8032 solver.cpp:330] Iteration 6630, Testing net (#0) I0412 13:56:58.043874 8032 net.cpp:676] Ignoring source layer train-data I0412 13:56:59.917044 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:57:02.516090 8032 solver.cpp:397] Test net output #0: accuracy = 0.435049 I0412 13:57:02.516139 8032 solver.cpp:397] Test net output #1: loss = 2.9312 (* 1 = 2.9312 loss) I0412 13:57:04.271543 8032 solver.cpp:218] Iteration 6636 (0.832099 iter/s, 14.4214s/12 iters), loss = 0.455764 I0412 13:57:04.271595 8032 solver.cpp:237] Train net output #0: loss = 0.455764 (* 1 = 0.455764 loss) I0412 13:57:04.271606 8032 sgd_solver.cpp:105] Iteration 6636, lr = 0.00268609 I0412 13:57:09.291935 8032 solver.cpp:218] Iteration 6648 (2.39035 iter/s, 5.02018s/12 iters), loss = 0.399027 I0412 13:57:09.291988 8032 solver.cpp:237] Train net output #0: loss = 0.399027 (* 1 = 0.399027 loss) I0412 13:57:09.292001 8032 sgd_solver.cpp:105] Iteration 6648, lr = 0.00267971 I0412 13:57:14.510215 8032 solver.cpp:218] Iteration 6660 (2.29971 iter/s, 5.21805s/12 iters), loss = 0.438967 I0412 13:57:14.510257 8032 solver.cpp:237] Train net output #0: loss = 0.438967 (* 1 = 0.438967 loss) I0412 13:57:14.510267 8032 sgd_solver.cpp:105] Iteration 6660, lr = 0.00267335 I0412 13:57:19.722311 8032 solver.cpp:218] Iteration 6672 (2.30243 iter/s, 5.21188s/12 iters), loss = 0.299424 I0412 13:57:19.722353 8032 solver.cpp:237] Train net output #0: loss = 0.299424 (* 1 = 0.299424 loss) I0412 13:57:19.722362 8032 sgd_solver.cpp:105] Iteration 6672, lr = 0.00266701 I0412 13:57:21.099048 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:57:24.773298 8032 solver.cpp:218] Iteration 6684 (2.37588 iter/s, 5.05077s/12 iters), loss = 0.256409 I0412 13:57:24.773352 8032 solver.cpp:237] Train net output #0: loss = 0.256409 (* 1 = 0.256409 loss) I0412 13:57:24.773365 8032 sgd_solver.cpp:105] Iteration 6684, lr = 0.00266067 I0412 13:57:29.777770 8032 solver.cpp:218] Iteration 6696 (2.39796 iter/s, 5.00425s/12 iters), loss = 0.238205 I0412 13:57:29.777896 8032 solver.cpp:237] Train net output #0: loss = 0.238205 (* 1 = 0.238205 loss) I0412 13:57:29.777907 8032 sgd_solver.cpp:105] Iteration 6696, lr = 0.00265436 I0412 13:57:34.779984 8032 solver.cpp:218] Iteration 6708 (2.39908 iter/s, 5.00192s/12 iters), loss = 0.251261 I0412 13:57:34.780045 8032 solver.cpp:237] Train net output #0: loss = 0.251261 (* 1 = 0.251261 loss) I0412 13:57:34.780056 8032 sgd_solver.cpp:105] Iteration 6708, lr = 0.00264805 I0412 13:57:39.738256 8032 solver.cpp:218] Iteration 6720 (2.42031 iter/s, 4.95805s/12 iters), loss = 0.338754 I0412 13:57:39.738310 8032 solver.cpp:237] Train net output #0: loss = 0.338754 (* 1 = 0.338754 loss) I0412 13:57:39.738323 8032 sgd_solver.cpp:105] Iteration 6720, lr = 0.00264177 I0412 13:57:44.239369 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6732.caffemodel I0412 13:57:47.332506 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6732.solverstate I0412 13:57:49.749706 8032 solver.cpp:330] Iteration 6732, Testing net (#0) I0412 13:57:49.749734 8032 net.cpp:676] Ignoring source layer train-data I0412 13:57:51.570267 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:57:54.212991 8032 solver.cpp:397] Test net output #0: accuracy = 0.441789 I0412 13:57:54.213035 8032 solver.cpp:397] Test net output #1: loss = 2.81463 (* 1 = 2.81463 loss) I0412 13:57:54.301416 8032 solver.cpp:218] Iteration 6732 (0.824026 iter/s, 14.5627s/12 iters), loss = 0.284281 I0412 13:57:54.301463 8032 solver.cpp:237] Train net output #0: loss = 0.284281 (* 1 = 0.284281 loss) I0412 13:57:54.301474 8032 sgd_solver.cpp:105] Iteration 6732, lr = 0.0026355 I0412 13:57:58.570649 8032 solver.cpp:218] Iteration 6744 (2.81094 iter/s, 4.26904s/12 iters), loss = 0.29632 I0412 13:57:58.570703 8032 solver.cpp:237] Train net output #0: loss = 0.29632 (* 1 = 0.29632 loss) I0412 13:57:58.570719 8032 sgd_solver.cpp:105] Iteration 6744, lr = 0.00262924 I0412 13:58:03.890914 8032 solver.cpp:218] Iteration 6756 (2.25563 iter/s, 5.32003s/12 iters), loss = 0.431126 I0412 13:58:03.891038 8032 solver.cpp:237] Train net output #0: loss = 0.431126 (* 1 = 0.431126 loss) I0412 13:58:03.891052 8032 sgd_solver.cpp:105] Iteration 6756, lr = 0.002623 I0412 13:58:08.868131 8032 solver.cpp:218] Iteration 6768 (2.41113 iter/s, 4.97693s/12 iters), loss = 0.313282 I0412 13:58:08.868186 8032 solver.cpp:237] Train net output #0: loss = 0.313282 (* 1 = 0.313282 loss) I0412 13:58:08.868198 8032 sgd_solver.cpp:105] Iteration 6768, lr = 0.00261677 I0412 13:58:12.343590 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:58:13.804024 8032 solver.cpp:218] Iteration 6780 (2.43128 iter/s, 4.93567s/12 iters), loss = 0.251515 I0412 13:58:13.804083 8032 solver.cpp:237] Train net output #0: loss = 0.251515 (* 1 = 0.251515 loss) I0412 13:58:13.804095 8032 sgd_solver.cpp:105] Iteration 6780, lr = 0.00261056 I0412 13:58:18.780373 8032 solver.cpp:218] Iteration 6792 (2.41152 iter/s, 4.97612s/12 iters), loss = 0.186644 I0412 13:58:18.780427 8032 solver.cpp:237] Train net output #0: loss = 0.186644 (* 1 = 0.186644 loss) I0412 13:58:18.780438 8032 sgd_solver.cpp:105] Iteration 6792, lr = 0.00260436 I0412 13:58:23.789939 8032 solver.cpp:218] Iteration 6804 (2.39552 iter/s, 5.00934s/12 iters), loss = 0.337346 I0412 13:58:23.790014 8032 solver.cpp:237] Train net output #0: loss = 0.337346 (* 1 = 0.337346 loss) I0412 13:58:23.790030 8032 sgd_solver.cpp:105] Iteration 6804, lr = 0.00259817 I0412 13:58:28.760258 8032 solver.cpp:218] Iteration 6816 (2.41445 iter/s, 4.97008s/12 iters), loss = 0.273544 I0412 13:58:28.760305 8032 solver.cpp:237] Train net output #0: loss = 0.273544 (* 1 = 0.273544 loss) I0412 13:58:28.760314 8032 sgd_solver.cpp:105] Iteration 6816, lr = 0.00259201 I0412 13:58:33.900004 8032 solver.cpp:218] Iteration 6828 (2.33485 iter/s, 5.13953s/12 iters), loss = 0.403876 I0412 13:58:33.900156 8032 solver.cpp:237] Train net output #0: loss = 0.403876 (* 1 = 0.403876 loss) I0412 13:58:33.900171 8032 sgd_solver.cpp:105] Iteration 6828, lr = 0.00258585 I0412 13:58:36.006599 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6834.caffemodel I0412 13:58:39.018476 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6834.solverstate I0412 13:58:41.808609 8032 solver.cpp:330] Iteration 6834, Testing net (#0) I0412 13:58:41.808630 8032 net.cpp:676] Ignoring source layer train-data I0412 13:58:43.584983 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:58:46.386941 8032 solver.cpp:397] Test net output #0: accuracy = 0.443015 I0412 13:58:46.386977 8032 solver.cpp:397] Test net output #1: loss = 2.97814 (* 1 = 2.97814 loss) I0412 13:58:48.154498 8032 solver.cpp:218] Iteration 6840 (0.841875 iter/s, 14.2539s/12 iters), loss = 0.292757 I0412 13:58:48.154553 8032 solver.cpp:237] Train net output #0: loss = 0.292757 (* 1 = 0.292757 loss) I0412 13:58:48.154567 8032 sgd_solver.cpp:105] Iteration 6840, lr = 0.00257971 I0412 13:58:53.519086 8032 solver.cpp:218] Iteration 6852 (2.23699 iter/s, 5.36436s/12 iters), loss = 0.557449 I0412 13:58:53.519126 8032 solver.cpp:237] Train net output #0: loss = 0.557449 (* 1 = 0.557449 loss) I0412 13:58:53.519137 8032 sgd_solver.cpp:105] Iteration 6852, lr = 0.00257359 I0412 13:58:58.532845 8032 solver.cpp:218] Iteration 6864 (2.39351 iter/s, 5.01355s/12 iters), loss = 0.44598 I0412 13:58:58.532903 8032 solver.cpp:237] Train net output #0: loss = 0.44598 (* 1 = 0.44598 loss) I0412 13:58:58.532915 8032 sgd_solver.cpp:105] Iteration 6864, lr = 0.00256748 I0412 13:59:03.542910 8032 solver.cpp:218] Iteration 6876 (2.39529 iter/s, 5.00984s/12 iters), loss = 0.228476 I0412 13:59:03.542969 8032 solver.cpp:237] Train net output #0: loss = 0.228476 (* 1 = 0.228476 loss) I0412 13:59:03.542981 8032 sgd_solver.cpp:105] Iteration 6876, lr = 0.00256138 I0412 13:59:04.153592 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:59:08.513852 8032 solver.cpp:218] Iteration 6888 (2.41414 iter/s, 4.97072s/12 iters), loss = 0.371772 I0412 13:59:08.513911 8032 solver.cpp:237] Train net output #0: loss = 0.371772 (* 1 = 0.371772 loss) I0412 13:59:08.513922 8032 sgd_solver.cpp:105] Iteration 6888, lr = 0.0025553 I0412 13:59:13.606772 8032 solver.cpp:218] Iteration 6900 (2.35632 iter/s, 5.09269s/12 iters), loss = 0.317392 I0412 13:59:13.606825 8032 solver.cpp:237] Train net output #0: loss = 0.317392 (* 1 = 0.317392 loss) I0412 13:59:13.606838 8032 sgd_solver.cpp:105] Iteration 6900, lr = 0.00254923 I0412 13:59:18.828462 8032 solver.cpp:218] Iteration 6912 (2.29821 iter/s, 5.22146s/12 iters), loss = 0.220369 I0412 13:59:18.828517 8032 solver.cpp:237] Train net output #0: loss = 0.220369 (* 1 = 0.220369 loss) I0412 13:59:18.828529 8032 sgd_solver.cpp:105] Iteration 6912, lr = 0.00254318 I0412 13:59:23.877607 8032 solver.cpp:218] Iteration 6924 (2.37674 iter/s, 5.04892s/12 iters), loss = 0.406273 I0412 13:59:23.877647 8032 solver.cpp:237] Train net output #0: loss = 0.406273 (* 1 = 0.406273 loss) I0412 13:59:23.877656 8032 sgd_solver.cpp:105] Iteration 6924, lr = 0.00253714 I0412 13:59:28.731458 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6936.caffemodel I0412 13:59:31.781582 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6936.solverstate I0412 13:59:34.490080 8032 solver.cpp:330] Iteration 6936, Testing net (#0) I0412 13:59:34.490155 8032 net.cpp:676] Ignoring source layer train-data I0412 13:59:35.153352 8032 blocking_queue.cpp:49] Waiting for data I0412 13:59:36.370586 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:59:39.236524 8032 solver.cpp:397] Test net output #0: accuracy = 0.445466 I0412 13:59:39.236575 8032 solver.cpp:397] Test net output #1: loss = 2.81872 (* 1 = 2.81872 loss) I0412 13:59:39.324672 8032 solver.cpp:218] Iteration 6936 (0.776873 iter/s, 15.4465s/12 iters), loss = 0.411323 I0412 13:59:39.324720 8032 solver.cpp:237] Train net output #0: loss = 0.411323 (* 1 = 0.411323 loss) I0412 13:59:39.324731 8032 sgd_solver.cpp:105] Iteration 6936, lr = 0.00253112 I0412 13:59:43.605046 8032 solver.cpp:218] Iteration 6948 (2.80362 iter/s, 4.28018s/12 iters), loss = 0.228508 I0412 13:59:43.605096 8032 solver.cpp:237] Train net output #0: loss = 0.228508 (* 1 = 0.228508 loss) I0412 13:59:43.605109 8032 sgd_solver.cpp:105] Iteration 6948, lr = 0.00252511 I0412 13:59:48.634836 8032 solver.cpp:218] Iteration 6960 (2.38589 iter/s, 5.02957s/12 iters), loss = 0.390699 I0412 13:59:48.634891 8032 solver.cpp:237] Train net output #0: loss = 0.390699 (* 1 = 0.390699 loss) I0412 13:59:48.634904 8032 sgd_solver.cpp:105] Iteration 6960, lr = 0.00251911 I0412 13:59:53.805622 8032 solver.cpp:218] Iteration 6972 (2.32083 iter/s, 5.17056s/12 iters), loss = 0.413845 I0412 13:59:53.805675 8032 solver.cpp:237] Train net output #0: loss = 0.413845 (* 1 = 0.413845 loss) I0412 13:59:53.805685 8032 sgd_solver.cpp:105] Iteration 6972, lr = 0.00251313 I0412 13:59:56.608757 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 13:59:58.920202 8032 solver.cpp:218] Iteration 6984 (2.34634 iter/s, 5.11435s/12 iters), loss = 0.273179 I0412 13:59:58.920255 8032 solver.cpp:237] Train net output #0: loss = 0.273179 (* 1 = 0.273179 loss) I0412 13:59:58.920267 8032 sgd_solver.cpp:105] Iteration 6984, lr = 0.00250717 I0412 14:00:04.152949 8032 solver.cpp:218] Iteration 6996 (2.29335 iter/s, 5.23252s/12 iters), loss = 0.371186 I0412 14:00:04.152990 8032 solver.cpp:237] Train net output #0: loss = 0.371186 (* 1 = 0.371186 loss) I0412 14:00:04.153000 8032 sgd_solver.cpp:105] Iteration 6996, lr = 0.00250121 I0412 14:00:09.196990 8032 solver.cpp:218] Iteration 7008 (2.37914 iter/s, 5.04383s/12 iters), loss = 0.303791 I0412 14:00:09.197088 8032 solver.cpp:237] Train net output #0: loss = 0.303791 (* 1 = 0.303791 loss) I0412 14:00:09.197098 8032 sgd_solver.cpp:105] Iteration 7008, lr = 0.00249528 I0412 14:00:14.456082 8032 solver.cpp:218] Iteration 7020 (2.28188 iter/s, 5.25882s/12 iters), loss = 0.331984 I0412 14:00:14.456130 8032 solver.cpp:237] Train net output #0: loss = 0.331984 (* 1 = 0.331984 loss) I0412 14:00:14.456141 8032 sgd_solver.cpp:105] Iteration 7020, lr = 0.00248935 I0412 14:00:19.936193 8032 solver.cpp:218] Iteration 7032 (2.18983 iter/s, 5.47988s/12 iters), loss = 0.453428 I0412 14:00:19.936239 8032 solver.cpp:237] Train net output #0: loss = 0.453428 (* 1 = 0.453428 loss) I0412 14:00:19.936249 8032 sgd_solver.cpp:105] Iteration 7032, lr = 0.00248344 I0412 14:00:22.061321 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7038.caffemodel I0412 14:00:25.080865 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7038.solverstate I0412 14:00:27.452158 8032 solver.cpp:330] Iteration 7038, Testing net (#0) I0412 14:00:27.452183 8032 net.cpp:676] Ignoring source layer train-data I0412 14:00:29.057000 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:00:31.919149 8032 solver.cpp:397] Test net output #0: accuracy = 0.436274 I0412 14:00:31.919183 8032 solver.cpp:397] Test net output #1: loss = 2.87042 (* 1 = 2.87042 loss) I0412 14:00:33.851397 8032 solver.cpp:218] Iteration 7044 (0.862396 iter/s, 13.9147s/12 iters), loss = 0.443134 I0412 14:00:33.851436 8032 solver.cpp:237] Train net output #0: loss = 0.443134 (* 1 = 0.443134 loss) I0412 14:00:33.851446 8032 sgd_solver.cpp:105] Iteration 7044, lr = 0.00247755 I0412 14:00:39.201053 8032 solver.cpp:218] Iteration 7056 (2.24323 iter/s, 5.34944s/12 iters), loss = 0.242311 I0412 14:00:39.201287 8032 solver.cpp:237] Train net output #0: loss = 0.242311 (* 1 = 0.242311 loss) I0412 14:00:39.201308 8032 sgd_solver.cpp:105] Iteration 7056, lr = 0.00247166 I0412 14:00:44.416700 8032 solver.cpp:218] Iteration 7068 (2.30094 iter/s, 5.21525s/12 iters), loss = 0.404997 I0412 14:00:44.416746 8032 solver.cpp:237] Train net output #0: loss = 0.404997 (* 1 = 0.404997 loss) I0412 14:00:44.416755 8032 sgd_solver.cpp:105] Iteration 7068, lr = 0.0024658 I0412 14:00:49.423949 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:00:49.543160 8032 solver.cpp:218] Iteration 7080 (2.3409 iter/s, 5.12623s/12 iters), loss = 0.227999 I0412 14:00:49.543216 8032 solver.cpp:237] Train net output #0: loss = 0.227999 (* 1 = 0.227999 loss) I0412 14:00:49.543227 8032 sgd_solver.cpp:105] Iteration 7080, lr = 0.00245994 I0412 14:00:54.655086 8032 solver.cpp:218] Iteration 7092 (2.34756 iter/s, 5.1117s/12 iters), loss = 0.194628 I0412 14:00:54.655138 8032 solver.cpp:237] Train net output #0: loss = 0.194628 (* 1 = 0.194628 loss) I0412 14:00:54.655148 8032 sgd_solver.cpp:105] Iteration 7092, lr = 0.0024541 I0412 14:00:59.710202 8032 solver.cpp:218] Iteration 7104 (2.37394 iter/s, 5.05489s/12 iters), loss = 0.252814 I0412 14:00:59.710250 8032 solver.cpp:237] Train net output #0: loss = 0.252814 (* 1 = 0.252814 loss) I0412 14:00:59.710263 8032 sgd_solver.cpp:105] Iteration 7104, lr = 0.00244827 I0412 14:01:04.625763 8032 solver.cpp:218] Iteration 7116 (2.44133 iter/s, 4.91534s/12 iters), loss = 0.197578 I0412 14:01:04.625813 8032 solver.cpp:237] Train net output #0: loss = 0.197578 (* 1 = 0.197578 loss) I0412 14:01:04.625823 8032 sgd_solver.cpp:105] Iteration 7116, lr = 0.00244246 I0412 14:01:09.746362 8032 solver.cpp:218] Iteration 7128 (2.34358 iter/s, 5.12037s/12 iters), loss = 0.26795 I0412 14:01:09.746515 8032 solver.cpp:237] Train net output #0: loss = 0.26795 (* 1 = 0.26795 loss) I0412 14:01:09.746531 8032 sgd_solver.cpp:105] Iteration 7128, lr = 0.00243666 I0412 14:01:14.343806 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7140.caffemodel I0412 14:01:19.427340 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7140.solverstate I0412 14:01:24.979177 8032 solver.cpp:330] Iteration 7140, Testing net (#0) I0412 14:01:24.979203 8032 net.cpp:676] Ignoring source layer train-data I0412 14:01:26.608595 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:01:29.519480 8032 solver.cpp:397] Test net output #0: accuracy = 0.441176 I0412 14:01:29.519522 8032 solver.cpp:397] Test net output #1: loss = 2.85641 (* 1 = 2.85641 loss) I0412 14:01:29.607587 8032 solver.cpp:218] Iteration 7140 (0.604216 iter/s, 19.8605s/12 iters), loss = 0.236447 I0412 14:01:29.607633 8032 solver.cpp:237] Train net output #0: loss = 0.236447 (* 1 = 0.236447 loss) I0412 14:01:29.607642 8032 sgd_solver.cpp:105] Iteration 7140, lr = 0.00243088 I0412 14:01:33.888553 8032 solver.cpp:218] Iteration 7152 (2.80323 iter/s, 4.28077s/12 iters), loss = 0.504069 I0412 14:01:33.888602 8032 solver.cpp:237] Train net output #0: loss = 0.504069 (* 1 = 0.504069 loss) I0412 14:01:33.888613 8032 sgd_solver.cpp:105] Iteration 7152, lr = 0.00242511 I0412 14:01:38.943480 8032 solver.cpp:218] Iteration 7164 (2.37402 iter/s, 5.05471s/12 iters), loss = 0.224628 I0412 14:01:38.943516 8032 solver.cpp:237] Train net output #0: loss = 0.224628 (* 1 = 0.224628 loss) I0412 14:01:38.943526 8032 sgd_solver.cpp:105] Iteration 7164, lr = 0.00241935 I0412 14:01:43.988164 8032 solver.cpp:218] Iteration 7176 (2.37884 iter/s, 5.04448s/12 iters), loss = 0.269392 I0412 14:01:43.988284 8032 solver.cpp:237] Train net output #0: loss = 0.269392 (* 1 = 0.269392 loss) I0412 14:01:43.988294 8032 sgd_solver.cpp:105] Iteration 7176, lr = 0.0024136 I0412 14:01:46.166419 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:01:49.293481 8032 solver.cpp:218] Iteration 7188 (2.26201 iter/s, 5.30502s/12 iters), loss = 0.262146 I0412 14:01:49.293524 8032 solver.cpp:237] Train net output #0: loss = 0.262146 (* 1 = 0.262146 loss) I0412 14:01:49.293532 8032 sgd_solver.cpp:105] Iteration 7188, lr = 0.00240787 I0412 14:01:54.442689 8032 solver.cpp:218] Iteration 7200 (2.33055 iter/s, 5.149s/12 iters), loss = 0.162426 I0412 14:01:54.442729 8032 solver.cpp:237] Train net output #0: loss = 0.162426 (* 1 = 0.162426 loss) I0412 14:01:54.442737 8032 sgd_solver.cpp:105] Iteration 7200, lr = 0.00240216 I0412 14:01:59.467082 8032 solver.cpp:218] Iteration 7212 (2.38845 iter/s, 5.02419s/12 iters), loss = 0.202825 I0412 14:01:59.467125 8032 solver.cpp:237] Train net output #0: loss = 0.202825 (* 1 = 0.202825 loss) I0412 14:01:59.467133 8032 sgd_solver.cpp:105] Iteration 7212, lr = 0.00239645 I0412 14:02:04.551131 8032 solver.cpp:218] Iteration 7224 (2.36042 iter/s, 5.08384s/12 iters), loss = 0.185158 I0412 14:02:04.551187 8032 solver.cpp:237] Train net output #0: loss = 0.185158 (* 1 = 0.185158 loss) I0412 14:02:04.551196 8032 sgd_solver.cpp:105] Iteration 7224, lr = 0.00239076 I0412 14:02:09.532742 8032 solver.cpp:218] Iteration 7236 (2.40897 iter/s, 4.98139s/12 iters), loss = 0.356488 I0412 14:02:09.532796 8032 solver.cpp:237] Train net output #0: loss = 0.356488 (* 1 = 0.356488 loss) I0412 14:02:09.532809 8032 sgd_solver.cpp:105] Iteration 7236, lr = 0.00238509 I0412 14:02:11.545753 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7242.caffemodel I0412 14:02:14.569444 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7242.solverstate I0412 14:02:16.877785 8032 solver.cpp:330] Iteration 7242, Testing net (#0) I0412 14:02:16.877810 8032 net.cpp:676] Ignoring source layer train-data I0412 14:02:18.455138 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:02:21.504401 8032 solver.cpp:397] Test net output #0: accuracy = 0.449142 I0412 14:02:21.504433 8032 solver.cpp:397] Test net output #1: loss = 2.92778 (* 1 = 2.92778 loss) I0412 14:02:23.420966 8032 solver.cpp:218] Iteration 7248 (0.864072 iter/s, 13.8877s/12 iters), loss = 0.255681 I0412 14:02:23.421020 8032 solver.cpp:237] Train net output #0: loss = 0.255681 (* 1 = 0.255681 loss) I0412 14:02:23.421032 8032 sgd_solver.cpp:105] Iteration 7248, lr = 0.00237942 I0412 14:02:28.446974 8032 solver.cpp:218] Iteration 7260 (2.38769 iter/s, 5.02579s/12 iters), loss = 0.188124 I0412 14:02:28.447023 8032 solver.cpp:237] Train net output #0: loss = 0.188124 (* 1 = 0.188124 loss) I0412 14:02:28.447036 8032 sgd_solver.cpp:105] Iteration 7260, lr = 0.00237378 I0412 14:02:33.459568 8032 solver.cpp:218] Iteration 7272 (2.39408 iter/s, 5.01237s/12 iters), loss = 0.226231 I0412 14:02:33.459625 8032 solver.cpp:237] Train net output #0: loss = 0.226231 (* 1 = 0.226231 loss) I0412 14:02:33.459638 8032 sgd_solver.cpp:105] Iteration 7272, lr = 0.00236814 I0412 14:02:37.758395 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:02:38.540586 8032 solver.cpp:218] Iteration 7284 (2.36184 iter/s, 5.08079s/12 iters), loss = 0.229455 I0412 14:02:38.540634 8032 solver.cpp:237] Train net output #0: loss = 0.229455 (* 1 = 0.229455 loss) I0412 14:02:38.540643 8032 sgd_solver.cpp:105] Iteration 7284, lr = 0.00236252 I0412 14:02:43.553495 8032 solver.cpp:218] Iteration 7296 (2.39392 iter/s, 5.01269s/12 iters), loss = 0.213416 I0412 14:02:43.553547 8032 solver.cpp:237] Train net output #0: loss = 0.213416 (* 1 = 0.213416 loss) I0412 14:02:43.553560 8032 sgd_solver.cpp:105] Iteration 7296, lr = 0.00235691 I0412 14:02:48.735203 8032 solver.cpp:218] Iteration 7308 (2.31594 iter/s, 5.18148s/12 iters), loss = 0.324042 I0412 14:02:48.735373 8032 solver.cpp:237] Train net output #0: loss = 0.324042 (* 1 = 0.324042 loss) I0412 14:02:48.735391 8032 sgd_solver.cpp:105] Iteration 7308, lr = 0.00235131 I0412 14:02:53.759999 8032 solver.cpp:218] Iteration 7320 (2.38832 iter/s, 5.02446s/12 iters), loss = 0.283758 I0412 14:02:53.760047 8032 solver.cpp:237] Train net output #0: loss = 0.283758 (* 1 = 0.283758 loss) I0412 14:02:53.760056 8032 sgd_solver.cpp:105] Iteration 7320, lr = 0.00234573 I0412 14:02:58.758688 8032 solver.cpp:218] Iteration 7332 (2.40073 iter/s, 4.99847s/12 iters), loss = 0.18419 I0412 14:02:58.758744 8032 solver.cpp:237] Train net output #0: loss = 0.184189 (* 1 = 0.184189 loss) I0412 14:02:58.758754 8032 sgd_solver.cpp:105] Iteration 7332, lr = 0.00234016 I0412 14:03:03.383672 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7344.caffemodel I0412 14:03:06.409783 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7344.solverstate I0412 14:03:09.923105 8032 solver.cpp:330] Iteration 7344, Testing net (#0) I0412 14:03:09.923127 8032 net.cpp:676] Ignoring source layer train-data I0412 14:03:11.543838 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:03:14.730908 8032 solver.cpp:397] Test net output #0: accuracy = 0.461397 I0412 14:03:14.730960 8032 solver.cpp:397] Test net output #1: loss = 2.73637 (* 1 = 2.73637 loss) I0412 14:03:14.818936 8032 solver.cpp:218] Iteration 7344 (0.747213 iter/s, 16.0597s/12 iters), loss = 0.213858 I0412 14:03:14.818992 8032 solver.cpp:237] Train net output #0: loss = 0.213858 (* 1 = 0.213858 loss) I0412 14:03:14.819003 8032 sgd_solver.cpp:105] Iteration 7344, lr = 0.0023346 I0412 14:03:19.123839 8032 solver.cpp:218] Iteration 7356 (2.78766 iter/s, 4.30469s/12 iters), loss = 0.358784 I0412 14:03:19.123971 8032 solver.cpp:237] Train net output #0: loss = 0.358784 (* 1 = 0.358784 loss) I0412 14:03:19.123988 8032 sgd_solver.cpp:105] Iteration 7356, lr = 0.00232906 I0412 14:03:24.153620 8032 solver.cpp:218] Iteration 7368 (2.38593 iter/s, 5.02949s/12 iters), loss = 0.253697 I0412 14:03:24.153673 8032 solver.cpp:237] Train net output #0: loss = 0.253697 (* 1 = 0.253697 loss) I0412 14:03:24.153685 8032 sgd_solver.cpp:105] Iteration 7368, lr = 0.00232353 I0412 14:03:29.682229 8032 solver.cpp:218] Iteration 7380 (2.17062 iter/s, 5.52837s/12 iters), loss = 0.219497 I0412 14:03:29.682281 8032 solver.cpp:237] Train net output #0: loss = 0.219497 (* 1 = 0.219497 loss) I0412 14:03:29.682292 8032 sgd_solver.cpp:105] Iteration 7380, lr = 0.00231802 I0412 14:03:31.182440 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:03:34.842306 8032 solver.cpp:218] Iteration 7392 (2.32565 iter/s, 5.15985s/12 iters), loss = 0.252485 I0412 14:03:34.842360 8032 solver.cpp:237] Train net output #0: loss = 0.252485 (* 1 = 0.252485 loss) I0412 14:03:34.842371 8032 sgd_solver.cpp:105] Iteration 7392, lr = 0.00231251 I0412 14:03:39.950302 8032 solver.cpp:218] Iteration 7404 (2.34936 iter/s, 5.10777s/12 iters), loss = 0.198375 I0412 14:03:39.950353 8032 solver.cpp:237] Train net output #0: loss = 0.198375 (* 1 = 0.198375 loss) I0412 14:03:39.950366 8032 sgd_solver.cpp:105] Iteration 7404, lr = 0.00230702 I0412 14:03:44.993361 8032 solver.cpp:218] Iteration 7416 (2.37961 iter/s, 5.04284s/12 iters), loss = 0.303306 I0412 14:03:44.993414 8032 solver.cpp:237] Train net output #0: loss = 0.303306 (* 1 = 0.303306 loss) I0412 14:03:44.993427 8032 sgd_solver.cpp:105] Iteration 7416, lr = 0.00230154 I0412 14:03:50.047683 8032 solver.cpp:218] Iteration 7428 (2.37431 iter/s, 5.0541s/12 iters), loss = 0.292293 I0412 14:03:50.047852 8032 solver.cpp:237] Train net output #0: loss = 0.292293 (* 1 = 0.292293 loss) I0412 14:03:50.047866 8032 sgd_solver.cpp:105] Iteration 7428, lr = 0.00229608 I0412 14:03:55.090705 8032 solver.cpp:218] Iteration 7440 (2.37968 iter/s, 5.04269s/12 iters), loss = 0.271521 I0412 14:03:55.090754 8032 solver.cpp:237] Train net output #0: loss = 0.271521 (* 1 = 0.271521 loss) I0412 14:03:55.090762 8032 sgd_solver.cpp:105] Iteration 7440, lr = 0.00229063 I0412 14:03:57.280154 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7446.caffemodel I0412 14:04:03.114225 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7446.solverstate I0412 14:04:06.556777 8032 solver.cpp:330] Iteration 7446, Testing net (#0) I0412 14:04:06.556803 8032 net.cpp:676] Ignoring source layer train-data I0412 14:04:08.097859 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:04:11.092566 8032 solver.cpp:397] Test net output #0: accuracy = 0.469363 I0412 14:04:11.092619 8032 solver.cpp:397] Test net output #1: loss = 2.83266 (* 1 = 2.83266 loss) I0412 14:04:12.806751 8032 solver.cpp:218] Iteration 7452 (0.677375 iter/s, 17.7154s/12 iters), loss = 0.161587 I0412 14:04:12.806797 8032 solver.cpp:237] Train net output #0: loss = 0.161587 (* 1 = 0.161587 loss) I0412 14:04:12.806805 8032 sgd_solver.cpp:105] Iteration 7452, lr = 0.00228519 I0412 14:04:17.935276 8032 solver.cpp:218] Iteration 7464 (2.33995 iter/s, 5.12831s/12 iters), loss = 0.275799 I0412 14:04:17.935328 8032 solver.cpp:237] Train net output #0: loss = 0.275799 (* 1 = 0.275799 loss) I0412 14:04:17.935341 8032 sgd_solver.cpp:105] Iteration 7464, lr = 0.00227976 I0412 14:04:23.006351 8032 solver.cpp:218] Iteration 7476 (2.36647 iter/s, 5.07085s/12 iters), loss = 0.282928 I0412 14:04:23.006472 8032 solver.cpp:237] Train net output #0: loss = 0.282928 (* 1 = 0.282928 loss) I0412 14:04:23.006486 8032 sgd_solver.cpp:105] Iteration 7476, lr = 0.00227435 I0412 14:04:26.632899 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:04:28.316737 8032 solver.cpp:218] Iteration 7488 (2.25985 iter/s, 5.31008s/12 iters), loss = 0.17362 I0412 14:04:28.316797 8032 solver.cpp:237] Train net output #0: loss = 0.17362 (* 1 = 0.17362 loss) I0412 14:04:28.316812 8032 sgd_solver.cpp:105] Iteration 7488, lr = 0.00226895 I0412 14:04:33.466142 8032 solver.cpp:218] Iteration 7500 (2.33047 iter/s, 5.14918s/12 iters), loss = 0.176578 I0412 14:04:33.466197 8032 solver.cpp:237] Train net output #0: loss = 0.176577 (* 1 = 0.176577 loss) I0412 14:04:33.466210 8032 sgd_solver.cpp:105] Iteration 7500, lr = 0.00226357 I0412 14:04:38.475544 8032 solver.cpp:218] Iteration 7512 (2.3956 iter/s, 5.00919s/12 iters), loss = 0.184295 I0412 14:04:38.475585 8032 solver.cpp:237] Train net output #0: loss = 0.184295 (* 1 = 0.184295 loss) I0412 14:04:38.475594 8032 sgd_solver.cpp:105] Iteration 7512, lr = 0.00225819 I0412 14:04:43.691431 8032 solver.cpp:218] Iteration 7524 (2.30076 iter/s, 5.21567s/12 iters), loss = 0.225193 I0412 14:04:43.691483 8032 solver.cpp:237] Train net output #0: loss = 0.225193 (* 1 = 0.225193 loss) I0412 14:04:43.691494 8032 sgd_solver.cpp:105] Iteration 7524, lr = 0.00225283 I0412 14:04:49.096335 8032 solver.cpp:218] Iteration 7536 (2.2203 iter/s, 5.40468s/12 iters), loss = 0.261028 I0412 14:04:49.096381 8032 solver.cpp:237] Train net output #0: loss = 0.261028 (* 1 = 0.261028 loss) I0412 14:04:49.096391 8032 sgd_solver.cpp:105] Iteration 7536, lr = 0.00224748 I0412 14:04:53.765789 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7548.caffemodel I0412 14:04:57.426378 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7548.solverstate I0412 14:04:59.746975 8032 solver.cpp:330] Iteration 7548, Testing net (#0) I0412 14:04:59.747001 8032 net.cpp:676] Ignoring source layer train-data I0412 14:05:01.231344 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:05:04.289599 8032 solver.cpp:397] Test net output #0: accuracy = 0.454657 I0412 14:05:04.289644 8032 solver.cpp:397] Test net output #1: loss = 2.88026 (* 1 = 2.88026 loss) I0412 14:05:04.377665 8032 solver.cpp:218] Iteration 7548 (0.785299 iter/s, 15.2808s/12 iters), loss = 0.140151 I0412 14:05:04.377720 8032 solver.cpp:237] Train net output #0: loss = 0.140151 (* 1 = 0.140151 loss) I0412 14:05:04.377730 8032 sgd_solver.cpp:105] Iteration 7548, lr = 0.00224215 I0412 14:05:08.578974 8032 solver.cpp:218] Iteration 7560 (2.85639 iter/s, 4.20111s/12 iters), loss = 0.180089 I0412 14:05:08.579023 8032 solver.cpp:237] Train net output #0: loss = 0.180089 (* 1 = 0.180089 loss) I0412 14:05:08.579033 8032 sgd_solver.cpp:105] Iteration 7560, lr = 0.00223682 I0412 14:05:13.646122 8032 solver.cpp:218] Iteration 7572 (2.3683 iter/s, 5.06693s/12 iters), loss = 0.241175 I0412 14:05:13.646168 8032 solver.cpp:237] Train net output #0: loss = 0.241175 (* 1 = 0.241175 loss) I0412 14:05:13.646178 8032 sgd_solver.cpp:105] Iteration 7572, lr = 0.00223151 I0412 14:05:18.578081 8032 solver.cpp:218] Iteration 7584 (2.43322 iter/s, 4.93174s/12 iters), loss = 0.258537 I0412 14:05:18.578133 8032 solver.cpp:237] Train net output #0: loss = 0.258537 (* 1 = 0.258537 loss) I0412 14:05:18.578145 8032 sgd_solver.cpp:105] Iteration 7584, lr = 0.00222621 I0412 14:05:19.249852 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:05:23.594805 8032 solver.cpp:218] Iteration 7596 (2.3921 iter/s, 5.0165s/12 iters), loss = 0.26083 I0412 14:05:23.594856 8032 solver.cpp:237] Train net output #0: loss = 0.26083 (* 1 = 0.26083 loss) I0412 14:05:23.594866 8032 sgd_solver.cpp:105] Iteration 7596, lr = 0.00222093 I0412 14:05:28.901907 8032 solver.cpp:218] Iteration 7608 (2.26122 iter/s, 5.30687s/12 iters), loss = 0.274976 I0412 14:05:28.902009 8032 solver.cpp:237] Train net output #0: loss = 0.274976 (* 1 = 0.274976 loss) I0412 14:05:28.902019 8032 sgd_solver.cpp:105] Iteration 7608, lr = 0.00221565 I0412 14:05:34.110123 8032 solver.cpp:218] Iteration 7620 (2.30417 iter/s, 5.20794s/12 iters), loss = 0.197683 I0412 14:05:34.110177 8032 solver.cpp:237] Train net output #0: loss = 0.197683 (* 1 = 0.197683 loss) I0412 14:05:34.110189 8032 sgd_solver.cpp:105] Iteration 7620, lr = 0.00221039 I0412 14:05:36.574645 8032 blocking_queue.cpp:49] Waiting for data I0412 14:05:39.126736 8032 solver.cpp:218] Iteration 7632 (2.39216 iter/s, 5.01639s/12 iters), loss = 0.259485 I0412 14:05:39.126782 8032 solver.cpp:237] Train net output #0: loss = 0.259485 (* 1 = 0.259485 loss) I0412 14:05:39.126792 8032 sgd_solver.cpp:105] Iteration 7632, lr = 0.00220515 I0412 14:05:44.313254 8032 solver.cpp:218] Iteration 7644 (2.31379 iter/s, 5.1863s/12 iters), loss = 0.127251 I0412 14:05:44.313306 8032 solver.cpp:237] Train net output #0: loss = 0.127251 (* 1 = 0.127251 loss) I0412 14:05:44.313318 8032 sgd_solver.cpp:105] Iteration 7644, lr = 0.00219991 I0412 14:05:46.261854 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7650.caffemodel I0412 14:05:49.295347 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7650.solverstate I0412 14:05:51.686808 8032 solver.cpp:330] Iteration 7650, Testing net (#0) I0412 14:05:51.686837 8032 net.cpp:676] Ignoring source layer train-data I0412 14:05:53.210916 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:05:56.294008 8032 solver.cpp:397] Test net output #0: accuracy = 0.458946 I0412 14:05:56.294059 8032 solver.cpp:397] Test net output #1: loss = 2.93984 (* 1 = 2.93984 loss) I0412 14:05:58.096519 8032 solver.cpp:218] Iteration 7656 (0.870652 iter/s, 13.7828s/12 iters), loss = 0.225326 I0412 14:05:58.096576 8032 solver.cpp:237] Train net output #0: loss = 0.225326 (* 1 = 0.225326 loss) I0412 14:05:58.096588 8032 sgd_solver.cpp:105] Iteration 7656, lr = 0.00219469 I0412 14:06:03.244622 8032 solver.cpp:218] Iteration 7668 (2.33106 iter/s, 5.14787s/12 iters), loss = 0.170401 I0412 14:06:03.244789 8032 solver.cpp:237] Train net output #0: loss = 0.170401 (* 1 = 0.170401 loss) I0412 14:06:03.244802 8032 sgd_solver.cpp:105] Iteration 7668, lr = 0.00218948 I0412 14:06:08.483572 8032 solver.cpp:218] Iteration 7680 (2.29068 iter/s, 5.23862s/12 iters), loss = 0.238901 I0412 14:06:08.483618 8032 solver.cpp:237] Train net output #0: loss = 0.2389 (* 1 = 0.2389 loss) I0412 14:06:08.483629 8032 sgd_solver.cpp:105] Iteration 7680, lr = 0.00218428 I0412 14:06:11.407886 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:06:13.649914 8032 solver.cpp:218] Iteration 7692 (2.32283 iter/s, 5.16612s/12 iters), loss = 0.28223 I0412 14:06:13.649992 8032 solver.cpp:237] Train net output #0: loss = 0.28223 (* 1 = 0.28223 loss) I0412 14:06:13.650007 8032 sgd_solver.cpp:105] Iteration 7692, lr = 0.00217909 I0412 14:06:18.598502 8032 solver.cpp:218] Iteration 7704 (2.42505 iter/s, 4.94835s/12 iters), loss = 0.141608 I0412 14:06:18.598560 8032 solver.cpp:237] Train net output #0: loss = 0.141608 (* 1 = 0.141608 loss) I0412 14:06:18.598577 8032 sgd_solver.cpp:105] Iteration 7704, lr = 0.00217392 I0412 14:06:23.693646 8032 solver.cpp:218] Iteration 7716 (2.35529 iter/s, 5.09492s/12 iters), loss = 0.275006 I0412 14:06:23.693703 8032 solver.cpp:237] Train net output #0: loss = 0.275006 (* 1 = 0.275006 loss) I0412 14:06:23.693717 8032 sgd_solver.cpp:105] Iteration 7716, lr = 0.00216876 I0412 14:06:28.768637 8032 solver.cpp:218] Iteration 7728 (2.36464 iter/s, 5.07477s/12 iters), loss = 0.25366 I0412 14:06:28.768682 8032 solver.cpp:237] Train net output #0: loss = 0.25366 (* 1 = 0.25366 loss) I0412 14:06:28.768692 8032 sgd_solver.cpp:105] Iteration 7728, lr = 0.00216361 I0412 14:06:33.813612 8032 solver.cpp:218] Iteration 7740 (2.37871 iter/s, 5.04476s/12 iters), loss = 0.180848 I0412 14:06:33.813678 8032 solver.cpp:237] Train net output #0: loss = 0.180848 (* 1 = 0.180848 loss) I0412 14:06:33.813688 8032 sgd_solver.cpp:105] Iteration 7740, lr = 0.00215847 I0412 14:06:38.361493 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7752.caffemodel I0412 14:06:44.740015 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7752.solverstate I0412 14:06:48.456902 8032 solver.cpp:330] Iteration 7752, Testing net (#0) I0412 14:06:48.456923 8032 net.cpp:676] Ignoring source layer train-data I0412 14:06:49.786533 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:06:52.964210 8032 solver.cpp:397] Test net output #0: accuracy = 0.46875 I0412 14:06:52.964260 8032 solver.cpp:397] Test net output #1: loss = 2.86414 (* 1 = 2.86414 loss) I0412 14:06:53.052345 8032 solver.cpp:218] Iteration 7752 (0.623764 iter/s, 19.238s/12 iters), loss = 0.206609 I0412 14:06:53.052410 8032 solver.cpp:237] Train net output #0: loss = 0.206609 (* 1 = 0.206609 loss) I0412 14:06:53.052423 8032 sgd_solver.cpp:105] Iteration 7752, lr = 0.00215335 I0412 14:06:57.334020 8032 solver.cpp:218] Iteration 7764 (2.80278 iter/s, 4.28146s/12 iters), loss = 0.214502 I0412 14:06:57.334076 8032 solver.cpp:237] Train net output #0: loss = 0.214502 (* 1 = 0.214502 loss) I0412 14:06:57.334087 8032 sgd_solver.cpp:105] Iteration 7764, lr = 0.00214823 I0412 14:07:02.597443 8032 solver.cpp:218] Iteration 7776 (2.27999 iter/s, 5.26319s/12 iters), loss = 0.316924 I0412 14:07:02.597487 8032 solver.cpp:237] Train net output #0: loss = 0.316924 (* 1 = 0.316924 loss) I0412 14:07:02.597496 8032 sgd_solver.cpp:105] Iteration 7776, lr = 0.00214313 I0412 14:07:07.482023 8032 solver.cpp:218] Iteration 7788 (2.45682 iter/s, 4.88437s/12 iters), loss = 0.195496 I0412 14:07:07.482161 8032 solver.cpp:237] Train net output #0: loss = 0.195496 (* 1 = 0.195496 loss) I0412 14:07:07.482172 8032 sgd_solver.cpp:105] Iteration 7788, lr = 0.00213805 I0412 14:07:07.490131 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:07:12.476722 8032 solver.cpp:218] Iteration 7800 (2.40269 iter/s, 4.99439s/12 iters), loss = 0.305726 I0412 14:07:12.476773 8032 solver.cpp:237] Train net output #0: loss = 0.305726 (* 1 = 0.305726 loss) I0412 14:07:12.476784 8032 sgd_solver.cpp:105] Iteration 7800, lr = 0.00213297 I0412 14:07:17.594897 8032 solver.cpp:218] Iteration 7812 (2.34469 iter/s, 5.11795s/12 iters), loss = 0.174593 I0412 14:07:17.594939 8032 solver.cpp:237] Train net output #0: loss = 0.174593 (* 1 = 0.174593 loss) I0412 14:07:17.594949 8032 sgd_solver.cpp:105] Iteration 7812, lr = 0.00212791 I0412 14:07:22.653676 8032 solver.cpp:218] Iteration 7824 (2.37222 iter/s, 5.05856s/12 iters), loss = 0.167955 I0412 14:07:22.653721 8032 solver.cpp:237] Train net output #0: loss = 0.167955 (* 1 = 0.167955 loss) I0412 14:07:22.653730 8032 sgd_solver.cpp:105] Iteration 7824, lr = 0.00212285 I0412 14:07:28.152202 8032 solver.cpp:218] Iteration 7836 (2.18249 iter/s, 5.4983s/12 iters), loss = 0.293593 I0412 14:07:28.152253 8032 solver.cpp:237] Train net output #0: loss = 0.293593 (* 1 = 0.293593 loss) I0412 14:07:28.152266 8032 sgd_solver.cpp:105] Iteration 7836, lr = 0.00211781 I0412 14:07:33.407768 8032 solver.cpp:218] Iteration 7848 (2.28339 iter/s, 5.25534s/12 iters), loss = 0.286235 I0412 14:07:33.407814 8032 solver.cpp:237] Train net output #0: loss = 0.286235 (* 1 = 0.286235 loss) I0412 14:07:33.407824 8032 sgd_solver.cpp:105] Iteration 7848, lr = 0.00211279 I0412 14:07:35.555192 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7854.caffemodel I0412 14:07:38.588213 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7854.solverstate I0412 14:07:40.898609 8032 solver.cpp:330] Iteration 7854, Testing net (#0) I0412 14:07:40.898633 8032 net.cpp:676] Ignoring source layer train-data I0412 14:07:42.274704 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:07:45.426096 8032 solver.cpp:397] Test net output #0: accuracy = 0.463848 I0412 14:07:45.426138 8032 solver.cpp:397] Test net output #1: loss = 2.83978 (* 1 = 2.83978 loss) I0412 14:07:47.336447 8032 solver.cpp:218] Iteration 7860 (0.861562 iter/s, 13.9282s/12 iters), loss = 0.159391 I0412 14:07:47.336494 8032 solver.cpp:237] Train net output #0: loss = 0.159391 (* 1 = 0.159391 loss) I0412 14:07:47.336504 8032 sgd_solver.cpp:105] Iteration 7860, lr = 0.00210777 I0412 14:07:52.450091 8032 solver.cpp:218] Iteration 7872 (2.34676 iter/s, 5.11343s/12 iters), loss = 0.157973 I0412 14:07:52.450136 8032 solver.cpp:237] Train net output #0: loss = 0.157973 (* 1 = 0.157973 loss) I0412 14:07:52.450145 8032 sgd_solver.cpp:105] Iteration 7872, lr = 0.00210277 I0412 14:07:57.620254 8032 solver.cpp:218] Iteration 7884 (2.32111 iter/s, 5.16995s/12 iters), loss = 0.203298 I0412 14:07:57.620298 8032 solver.cpp:237] Train net output #0: loss = 0.203298 (* 1 = 0.203298 loss) I0412 14:07:57.620309 8032 sgd_solver.cpp:105] Iteration 7884, lr = 0.00209777 I0412 14:07:59.777581 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:08:02.712586 8032 solver.cpp:218] Iteration 7896 (2.35659 iter/s, 5.09211s/12 iters), loss = 0.160718 I0412 14:08:02.712641 8032 solver.cpp:237] Train net output #0: loss = 0.160718 (* 1 = 0.160718 loss) I0412 14:08:02.712652 8032 sgd_solver.cpp:105] Iteration 7896, lr = 0.00209279 I0412 14:08:07.968820 8032 solver.cpp:218] Iteration 7908 (2.2831 iter/s, 5.25601s/12 iters), loss = 0.116061 I0412 14:08:07.968860 8032 solver.cpp:237] Train net output #0: loss = 0.116061 (* 1 = 0.116061 loss) I0412 14:08:07.968869 8032 sgd_solver.cpp:105] Iteration 7908, lr = 0.00208782 I0412 14:08:13.090914 8032 solver.cpp:218] Iteration 7920 (2.34289 iter/s, 5.12188s/12 iters), loss = 0.188641 I0412 14:08:13.091055 8032 solver.cpp:237] Train net output #0: loss = 0.188641 (* 1 = 0.188641 loss) I0412 14:08:13.091066 8032 sgd_solver.cpp:105] Iteration 7920, lr = 0.00208287 I0412 14:08:18.011970 8032 solver.cpp:218] Iteration 7932 (2.43865 iter/s, 4.92075s/12 iters), loss = 0.137276 I0412 14:08:18.012029 8032 solver.cpp:237] Train net output #0: loss = 0.137276 (* 1 = 0.137276 loss) I0412 14:08:18.012042 8032 sgd_solver.cpp:105] Iteration 7932, lr = 0.00207792 I0412 14:08:23.094835 8032 solver.cpp:218] Iteration 7944 (2.36098 iter/s, 5.08264s/12 iters), loss = 0.256588 I0412 14:08:23.094882 8032 solver.cpp:237] Train net output #0: loss = 0.256588 (* 1 = 0.256588 loss) I0412 14:08:23.094893 8032 sgd_solver.cpp:105] Iteration 7944, lr = 0.00207299 I0412 14:08:27.871206 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7956.caffemodel I0412 14:08:30.935521 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7956.solverstate I0412 14:08:33.239027 8032 solver.cpp:330] Iteration 7956, Testing net (#0) I0412 14:08:33.239051 8032 net.cpp:676] Ignoring source layer train-data I0412 14:08:34.593734 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:08:37.799301 8032 solver.cpp:397] Test net output #0: accuracy = 0.449142 I0412 14:08:37.799350 8032 solver.cpp:397] Test net output #1: loss = 2.97616 (* 1 = 2.97616 loss) I0412 14:08:37.887215 8032 solver.cpp:218] Iteration 7956 (0.811257 iter/s, 14.7919s/12 iters), loss = 0.239625 I0412 14:08:37.887269 8032 solver.cpp:237] Train net output #0: loss = 0.239625 (* 1 = 0.239625 loss) I0412 14:08:37.887280 8032 sgd_solver.cpp:105] Iteration 7956, lr = 0.00206807 I0412 14:08:42.490303 8032 solver.cpp:218] Iteration 7968 (2.60707 iter/s, 4.60288s/12 iters), loss = 0.19718 I0412 14:08:42.490355 8032 solver.cpp:237] Train net output #0: loss = 0.19718 (* 1 = 0.19718 loss) I0412 14:08:42.490366 8032 sgd_solver.cpp:105] Iteration 7968, lr = 0.00206316 I0412 14:08:47.738673 8032 solver.cpp:218] Iteration 7980 (2.28652 iter/s, 5.24814s/12 iters), loss = 0.0965604 I0412 14:08:47.738785 8032 solver.cpp:237] Train net output #0: loss = 0.0965604 (* 1 = 0.0965604 loss) I0412 14:08:47.738796 8032 sgd_solver.cpp:105] Iteration 7980, lr = 0.00205826 I0412 14:08:52.099296 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:08:52.864431 8032 solver.cpp:218] Iteration 7992 (2.34124 iter/s, 5.12548s/12 iters), loss = 0.281828 I0412 14:08:52.864480 8032 solver.cpp:237] Train net output #0: loss = 0.281828 (* 1 = 0.281828 loss) I0412 14:08:52.864490 8032 sgd_solver.cpp:105] Iteration 7992, lr = 0.00205337 I0412 14:08:57.971927 8032 solver.cpp:218] Iteration 8004 (2.34959 iter/s, 5.10727s/12 iters), loss = 0.264524 I0412 14:08:57.971978 8032 solver.cpp:237] Train net output #0: loss = 0.264524 (* 1 = 0.264524 loss) I0412 14:08:57.971990 8032 sgd_solver.cpp:105] Iteration 8004, lr = 0.0020485 I0412 14:09:03.136418 8032 solver.cpp:218] Iteration 8016 (2.32366 iter/s, 5.16427s/12 iters), loss = 0.139937 I0412 14:09:03.136473 8032 solver.cpp:237] Train net output #0: loss = 0.139937 (* 1 = 0.139937 loss) I0412 14:09:03.136487 8032 sgd_solver.cpp:105] Iteration 8016, lr = 0.00204363 I0412 14:09:08.233850 8032 solver.cpp:218] Iteration 8028 (2.35423 iter/s, 5.09721s/12 iters), loss = 0.180661 I0412 14:09:08.233902 8032 solver.cpp:237] Train net output #0: loss = 0.180661 (* 1 = 0.180661 loss) I0412 14:09:08.233914 8032 sgd_solver.cpp:105] Iteration 8028, lr = 0.00203878 I0412 14:09:13.215462 8032 solver.cpp:218] Iteration 8040 (2.40896 iter/s, 4.98139s/12 iters), loss = 0.151138 I0412 14:09:13.215518 8032 solver.cpp:237] Train net output #0: loss = 0.151138 (* 1 = 0.151138 loss) I0412 14:09:13.215531 8032 sgd_solver.cpp:105] Iteration 8040, lr = 0.00203394 I0412 14:09:18.186529 8032 solver.cpp:218] Iteration 8052 (2.41408 iter/s, 4.97084s/12 iters), loss = 0.28343 I0412 14:09:18.186694 8032 solver.cpp:237] Train net output #0: loss = 0.28343 (* 1 = 0.28343 loss) I0412 14:09:18.186707 8032 sgd_solver.cpp:105] Iteration 8052, lr = 0.00202911 I0412 14:09:20.360914 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8058.caffemodel I0412 14:09:23.331171 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8058.solverstate I0412 14:09:25.642532 8032 solver.cpp:330] Iteration 8058, Testing net (#0) I0412 14:09:25.642555 8032 net.cpp:676] Ignoring source layer train-data I0412 14:09:27.011929 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:09:30.329058 8032 solver.cpp:397] Test net output #0: accuracy = 0.466299 I0412 14:09:30.329113 8032 solver.cpp:397] Test net output #1: loss = 2.95173 (* 1 = 2.95173 loss) I0412 14:09:32.430063 8032 solver.cpp:218] Iteration 8064 (0.842524 iter/s, 14.2429s/12 iters), loss = 0.204879 I0412 14:09:32.430105 8032 solver.cpp:237] Train net output #0: loss = 0.204879 (* 1 = 0.204879 loss) I0412 14:09:32.430115 8032 sgd_solver.cpp:105] Iteration 8064, lr = 0.00202429 I0412 14:09:37.527789 8032 solver.cpp:218] Iteration 8076 (2.35409 iter/s, 5.09751s/12 iters), loss = 0.0826255 I0412 14:09:37.527837 8032 solver.cpp:237] Train net output #0: loss = 0.0826255 (* 1 = 0.0826255 loss) I0412 14:09:37.527846 8032 sgd_solver.cpp:105] Iteration 8076, lr = 0.00201949 I0412 14:09:42.683218 8032 solver.cpp:218] Iteration 8088 (2.32774 iter/s, 5.15521s/12 iters), loss = 0.135866 I0412 14:09:42.683274 8032 solver.cpp:237] Train net output #0: loss = 0.135866 (* 1 = 0.135866 loss) I0412 14:09:42.683285 8032 sgd_solver.cpp:105] Iteration 8088, lr = 0.00201469 I0412 14:09:44.136484 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:09:47.862841 8032 solver.cpp:218] Iteration 8100 (2.31687 iter/s, 5.1794s/12 iters), loss = 0.141433 I0412 14:09:47.862898 8032 solver.cpp:237] Train net output #0: loss = 0.141433 (* 1 = 0.141433 loss) I0412 14:09:47.862911 8032 sgd_solver.cpp:105] Iteration 8100, lr = 0.00200991 I0412 14:09:52.976263 8032 solver.cpp:218] Iteration 8112 (2.34687 iter/s, 5.11319s/12 iters), loss = 0.096346 I0412 14:09:52.976366 8032 solver.cpp:237] Train net output #0: loss = 0.096346 (* 1 = 0.096346 loss) I0412 14:09:52.976378 8032 sgd_solver.cpp:105] Iteration 8112, lr = 0.00200514 I0412 14:09:58.083633 8032 solver.cpp:218] Iteration 8124 (2.34967 iter/s, 5.1071s/12 iters), loss = 0.150701 I0412 14:09:58.083684 8032 solver.cpp:237] Train net output #0: loss = 0.150701 (* 1 = 0.150701 loss) I0412 14:09:58.083696 8032 sgd_solver.cpp:105] Iteration 8124, lr = 0.00200038 I0412 14:10:03.300330 8032 solver.cpp:218] Iteration 8136 (2.3004 iter/s, 5.21648s/12 iters), loss = 0.196319 I0412 14:10:03.300374 8032 solver.cpp:237] Train net output #0: loss = 0.196319 (* 1 = 0.196319 loss) I0412 14:10:03.300382 8032 sgd_solver.cpp:105] Iteration 8136, lr = 0.00199563 I0412 14:10:08.384039 8032 solver.cpp:218] Iteration 8148 (2.36058 iter/s, 5.08349s/12 iters), loss = 0.228065 I0412 14:10:08.384084 8032 solver.cpp:237] Train net output #0: loss = 0.228065 (* 1 = 0.228065 loss) I0412 14:10:08.384093 8032 sgd_solver.cpp:105] Iteration 8148, lr = 0.00199089 I0412 14:10:13.086756 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8160.caffemodel I0412 14:10:16.078676 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8160.solverstate I0412 14:10:18.428611 8032 solver.cpp:330] Iteration 8160, Testing net (#0) I0412 14:10:18.428635 8032 net.cpp:676] Ignoring source layer train-data I0412 14:10:19.580848 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:10:22.811370 8032 solver.cpp:397] Test net output #0: accuracy = 0.448529 I0412 14:10:22.811417 8032 solver.cpp:397] Test net output #1: loss = 3.00113 (* 1 = 3.00113 loss) I0412 14:10:22.899788 8032 solver.cpp:218] Iteration 8160 (0.826717 iter/s, 14.5152s/12 iters), loss = 0.270993 I0412 14:10:22.899834 8032 solver.cpp:237] Train net output #0: loss = 0.270993 (* 1 = 0.270993 loss) I0412 14:10:22.899845 8032 sgd_solver.cpp:105] Iteration 8160, lr = 0.00198616 I0412 14:10:27.230034 8032 solver.cpp:218] Iteration 8172 (2.77133 iter/s, 4.33005s/12 iters), loss = 0.302502 I0412 14:10:27.230129 8032 solver.cpp:237] Train net output #0: loss = 0.302502 (* 1 = 0.302502 loss) I0412 14:10:27.230139 8032 sgd_solver.cpp:105] Iteration 8172, lr = 0.00198145 I0412 14:10:32.749480 8032 solver.cpp:218] Iteration 8184 (2.17424 iter/s, 5.51917s/12 iters), loss = 0.118737 I0412 14:10:32.749532 8032 solver.cpp:237] Train net output #0: loss = 0.118737 (* 1 = 0.118737 loss) I0412 14:10:32.749545 8032 sgd_solver.cpp:105] Iteration 8184, lr = 0.00197674 I0412 14:10:36.741478 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:10:38.233832 8032 solver.cpp:218] Iteration 8196 (2.18814 iter/s, 5.48412s/12 iters), loss = 0.116199 I0412 14:10:38.233886 8032 solver.cpp:237] Train net output #0: loss = 0.116199 (* 1 = 0.116199 loss) I0412 14:10:38.233898 8032 sgd_solver.cpp:105] Iteration 8196, lr = 0.00197205 I0412 14:10:43.231570 8032 solver.cpp:218] Iteration 8208 (2.40119 iter/s, 4.99752s/12 iters), loss = 0.102011 I0412 14:10:43.231621 8032 solver.cpp:237] Train net output #0: loss = 0.102011 (* 1 = 0.102011 loss) I0412 14:10:43.231631 8032 sgd_solver.cpp:105] Iteration 8208, lr = 0.00196737 I0412 14:10:48.264079 8032 solver.cpp:218] Iteration 8220 (2.3846 iter/s, 5.0323s/12 iters), loss = 0.118814 I0412 14:10:48.264120 8032 solver.cpp:237] Train net output #0: loss = 0.118814 (* 1 = 0.118814 loss) I0412 14:10:48.264129 8032 sgd_solver.cpp:105] Iteration 8220, lr = 0.0019627 I0412 14:10:53.425756 8032 solver.cpp:218] Iteration 8232 (2.32492 iter/s, 5.16146s/12 iters), loss = 0.128648 I0412 14:10:53.425808 8032 solver.cpp:237] Train net output #0: loss = 0.128648 (* 1 = 0.128648 loss) I0412 14:10:53.425820 8032 sgd_solver.cpp:105] Iteration 8232, lr = 0.00195804 I0412 14:10:58.677127 8032 solver.cpp:218] Iteration 8244 (2.28522 iter/s, 5.25115s/12 iters), loss = 0.128725 I0412 14:10:58.681403 8032 solver.cpp:237] Train net output #0: loss = 0.128725 (* 1 = 0.128725 loss) I0412 14:10:58.681414 8032 sgd_solver.cpp:105] Iteration 8244, lr = 0.00195339 I0412 14:11:04.491374 8032 solver.cpp:218] Iteration 8256 (2.06548 iter/s, 5.80978s/12 iters), loss = 0.25891 I0412 14:11:04.491427 8032 solver.cpp:237] Train net output #0: loss = 0.25891 (* 1 = 0.25891 loss) I0412 14:11:04.491441 8032 sgd_solver.cpp:105] Iteration 8256, lr = 0.00194875 I0412 14:11:06.618922 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8262.caffemodel I0412 14:11:09.602039 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8262.solverstate I0412 14:11:11.916296 8032 solver.cpp:330] Iteration 8262, Testing net (#0) I0412 14:11:11.916321 8032 net.cpp:676] Ignoring source layer train-data I0412 14:11:13.125110 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:11:16.603543 8032 solver.cpp:397] Test net output #0: accuracy = 0.466299 I0412 14:11:16.603580 8032 solver.cpp:397] Test net output #1: loss = 2.89326 (* 1 = 2.89326 loss) I0412 14:11:18.578652 8032 solver.cpp:218] Iteration 8268 (0.851862 iter/s, 14.0868s/12 iters), loss = 0.182444 I0412 14:11:18.578704 8032 solver.cpp:237] Train net output #0: loss = 0.182444 (* 1 = 0.182444 loss) I0412 14:11:18.578716 8032 sgd_solver.cpp:105] Iteration 8268, lr = 0.00194412 I0412 14:11:23.699245 8032 solver.cpp:218] Iteration 8280 (2.34358 iter/s, 5.12037s/12 iters), loss = 0.102648 I0412 14:11:23.699301 8032 solver.cpp:237] Train net output #0: loss = 0.102648 (* 1 = 0.102648 loss) I0412 14:11:23.699311 8032 sgd_solver.cpp:105] Iteration 8280, lr = 0.00193951 I0412 14:11:28.642827 8032 solver.cpp:218] Iteration 8292 (2.4275 iter/s, 4.94336s/12 iters), loss = 0.168712 I0412 14:11:28.642879 8032 solver.cpp:237] Train net output #0: loss = 0.168712 (* 1 = 0.168712 loss) I0412 14:11:28.642891 8032 sgd_solver.cpp:105] Iteration 8292, lr = 0.0019349 I0412 14:11:29.280347 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:11:33.612938 8032 solver.cpp:218] Iteration 8304 (2.41454 iter/s, 4.9699s/12 iters), loss = 0.191821 I0412 14:11:33.612980 8032 solver.cpp:237] Train net output #0: loss = 0.191821 (* 1 = 0.191821 loss) I0412 14:11:33.612989 8032 sgd_solver.cpp:105] Iteration 8304, lr = 0.00193031 I0412 14:11:36.523901 8032 blocking_queue.cpp:49] Waiting for data I0412 14:11:38.787997 8032 solver.cpp:218] Iteration 8316 (2.31891 iter/s, 5.17484s/12 iters), loss = 0.126075 I0412 14:11:38.788049 8032 solver.cpp:237] Train net output #0: loss = 0.126075 (* 1 = 0.126075 loss) I0412 14:11:38.788062 8032 sgd_solver.cpp:105] Iteration 8316, lr = 0.00192573 I0412 14:11:44.446148 8032 solver.cpp:218] Iteration 8328 (2.12092 iter/s, 5.65791s/12 iters), loss = 0.1586 I0412 14:11:44.446205 8032 solver.cpp:237] Train net output #0: loss = 0.1586 (* 1 = 0.1586 loss) I0412 14:11:44.446219 8032 sgd_solver.cpp:105] Iteration 8328, lr = 0.00192115 I0412 14:11:49.647011 8032 solver.cpp:218] Iteration 8340 (2.30741 iter/s, 5.20064s/12 iters), loss = 0.21948 I0412 14:11:49.647051 8032 solver.cpp:237] Train net output #0: loss = 0.21948 (* 1 = 0.21948 loss) I0412 14:11:49.647059 8032 sgd_solver.cpp:105] Iteration 8340, lr = 0.00191659 I0412 14:11:54.875363 8032 solver.cpp:218] Iteration 8352 (2.29527 iter/s, 5.22814s/12 iters), loss = 0.239365 I0412 14:11:54.875407 8032 solver.cpp:237] Train net output #0: loss = 0.239365 (* 1 = 0.239365 loss) I0412 14:11:54.875416 8032 sgd_solver.cpp:105] Iteration 8352, lr = 0.00191204 I0412 14:11:59.626332 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8364.caffemodel I0412 14:12:05.814157 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8364.solverstate I0412 14:12:08.381436 8032 solver.cpp:330] Iteration 8364, Testing net (#0) I0412 14:12:08.381464 8032 net.cpp:676] Ignoring source layer train-data I0412 14:12:09.636186 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:12:12.985415 8032 solver.cpp:397] Test net output #0: accuracy = 0.458333 I0412 14:12:12.985466 8032 solver.cpp:397] Test net output #1: loss = 2.92445 (* 1 = 2.92445 loss) I0412 14:12:13.073873 8032 solver.cpp:218] Iteration 8364 (0.659417 iter/s, 18.1979s/12 iters), loss = 0.0994249 I0412 14:12:13.073942 8032 solver.cpp:237] Train net output #0: loss = 0.0994248 (* 1 = 0.0994248 loss) I0412 14:12:13.073982 8032 sgd_solver.cpp:105] Iteration 8364, lr = 0.0019075 I0412 14:12:17.329398 8032 solver.cpp:218] Iteration 8376 (2.82 iter/s, 4.25531s/12 iters), loss = 0.1964 I0412 14:12:17.329454 8032 solver.cpp:237] Train net output #0: loss = 0.1964 (* 1 = 0.1964 loss) I0412 14:12:17.329468 8032 sgd_solver.cpp:105] Iteration 8376, lr = 0.00190297 I0412 14:12:22.359167 8032 solver.cpp:218] Iteration 8388 (2.38588 iter/s, 5.02959s/12 iters), loss = 0.129255 I0412 14:12:22.359210 8032 solver.cpp:237] Train net output #0: loss = 0.129255 (* 1 = 0.129255 loss) I0412 14:12:22.359220 8032 sgd_solver.cpp:105] Iteration 8388, lr = 0.00189846 I0412 14:12:25.269191 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:12:27.490798 8032 solver.cpp:218] Iteration 8400 (2.33851 iter/s, 5.13147s/12 iters), loss = 0.161738 I0412 14:12:27.490854 8032 solver.cpp:237] Train net output #0: loss = 0.161738 (* 1 = 0.161738 loss) I0412 14:12:27.490865 8032 sgd_solver.cpp:105] Iteration 8400, lr = 0.00189395 I0412 14:12:32.671485 8032 solver.cpp:218] Iteration 8412 (2.31637 iter/s, 5.18051s/12 iters), loss = 0.135826 I0412 14:12:32.682055 8032 solver.cpp:237] Train net output #0: loss = 0.135826 (* 1 = 0.135826 loss) I0412 14:12:32.682072 8032 sgd_solver.cpp:105] Iteration 8412, lr = 0.00188945 I0412 14:12:37.648283 8032 solver.cpp:218] Iteration 8424 (2.41637 iter/s, 4.96612s/12 iters), loss = 0.188408 I0412 14:12:37.648332 8032 solver.cpp:237] Train net output #0: loss = 0.188408 (* 1 = 0.188408 loss) I0412 14:12:37.648344 8032 sgd_solver.cpp:105] Iteration 8424, lr = 0.00188497 I0412 14:12:42.878099 8032 solver.cpp:218] Iteration 8436 (2.29461 iter/s, 5.22965s/12 iters), loss = 0.157586 I0412 14:12:42.878144 8032 solver.cpp:237] Train net output #0: loss = 0.157585 (* 1 = 0.157585 loss) I0412 14:12:42.878154 8032 sgd_solver.cpp:105] Iteration 8436, lr = 0.00188049 I0412 14:12:48.112253 8032 solver.cpp:218] Iteration 8448 (2.29271 iter/s, 5.23398s/12 iters), loss = 0.094879 I0412 14:12:48.112309 8032 solver.cpp:237] Train net output #0: loss = 0.094879 (* 1 = 0.094879 loss) I0412 14:12:48.112321 8032 sgd_solver.cpp:105] Iteration 8448, lr = 0.00187603 I0412 14:12:53.392211 8032 solver.cpp:218] Iteration 8460 (2.27282 iter/s, 5.27978s/12 iters), loss = 0.105675 I0412 14:12:53.392256 8032 solver.cpp:237] Train net output #0: loss = 0.105675 (* 1 = 0.105675 loss) I0412 14:12:53.392264 8032 sgd_solver.cpp:105] Iteration 8460, lr = 0.00187157 I0412 14:12:55.442150 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8466.caffemodel I0412 14:12:59.588295 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8466.solverstate I0412 14:13:02.533221 8032 solver.cpp:330] Iteration 8466, Testing net (#0) I0412 14:13:02.533246 8032 net.cpp:676] Ignoring source layer train-data I0412 14:13:03.768676 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:13:07.103238 8032 solver.cpp:397] Test net output #0: accuracy = 0.466912 I0412 14:13:07.103274 8032 solver.cpp:397] Test net output #1: loss = 2.98202 (* 1 = 2.98202 loss) I0412 14:13:09.031396 8032 solver.cpp:218] Iteration 8472 (0.767322 iter/s, 15.6388s/12 iters), loss = 0.106802 I0412 14:13:09.031459 8032 solver.cpp:237] Train net output #0: loss = 0.106802 (* 1 = 0.106802 loss) I0412 14:13:09.031476 8032 sgd_solver.cpp:105] Iteration 8472, lr = 0.00186713 I0412 14:13:14.219106 8032 solver.cpp:218] Iteration 8484 (2.31324 iter/s, 5.18752s/12 iters), loss = 0.166199 I0412 14:13:14.219163 8032 solver.cpp:237] Train net output #0: loss = 0.166199 (* 1 = 0.166199 loss) I0412 14:13:14.219175 8032 sgd_solver.cpp:105] Iteration 8484, lr = 0.0018627 I0412 14:13:19.491176 8032 solver.cpp:218] Iteration 8496 (2.27622 iter/s, 5.27189s/12 iters), loss = 0.188608 I0412 14:13:19.491219 8032 solver.cpp:237] Train net output #0: loss = 0.188607 (* 1 = 0.188607 loss) I0412 14:13:19.491230 8032 sgd_solver.cpp:105] Iteration 8496, lr = 0.00185827 I0412 14:13:19.502792 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:13:24.362501 8032 solver.cpp:218] Iteration 8508 (2.46348 iter/s, 4.87116s/12 iters), loss = 0.175845 I0412 14:13:24.362555 8032 solver.cpp:237] Train net output #0: loss = 0.175844 (* 1 = 0.175844 loss) I0412 14:13:24.362566 8032 sgd_solver.cpp:105] Iteration 8508, lr = 0.00185386 I0412 14:13:29.518333 8032 solver.cpp:218] Iteration 8520 (2.32754 iter/s, 5.15565s/12 iters), loss = 0.211023 I0412 14:13:29.518385 8032 solver.cpp:237] Train net output #0: loss = 0.211023 (* 1 = 0.211023 loss) I0412 14:13:29.518396 8032 sgd_solver.cpp:105] Iteration 8520, lr = 0.00184946 I0412 14:13:34.620967 8032 solver.cpp:218] Iteration 8532 (2.35181 iter/s, 5.10246s/12 iters), loss = 0.117492 I0412 14:13:34.621085 8032 solver.cpp:237] Train net output #0: loss = 0.117492 (* 1 = 0.117492 loss) I0412 14:13:34.621099 8032 sgd_solver.cpp:105] Iteration 8532, lr = 0.00184507 I0412 14:13:39.493652 8032 solver.cpp:218] Iteration 8544 (2.46282 iter/s, 4.87245s/12 iters), loss = 0.16915 I0412 14:13:39.493695 8032 solver.cpp:237] Train net output #0: loss = 0.16915 (* 1 = 0.16915 loss) I0412 14:13:39.493703 8032 sgd_solver.cpp:105] Iteration 8544, lr = 0.00184069 I0412 14:13:44.646688 8032 solver.cpp:218] Iteration 8556 (2.3288 iter/s, 5.15287s/12 iters), loss = 0.2243 I0412 14:13:44.646734 8032 solver.cpp:237] Train net output #0: loss = 0.2243 (* 1 = 0.2243 loss) I0412 14:13:44.646744 8032 sgd_solver.cpp:105] Iteration 8556, lr = 0.00183632 I0412 14:13:49.378983 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8568.caffemodel I0412 14:13:54.086135 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8568.solverstate I0412 14:13:58.524423 8032 solver.cpp:330] Iteration 8568, Testing net (#0) I0412 14:13:58.524451 8032 net.cpp:676] Ignoring source layer train-data I0412 14:13:59.685636 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:14:03.236188 8032 solver.cpp:397] Test net output #0: accuracy = 0.474265 I0412 14:14:03.236238 8032 solver.cpp:397] Test net output #1: loss = 2.90721 (* 1 = 2.90721 loss) I0412 14:14:03.324769 8032 solver.cpp:218] Iteration 8568 (0.64248 iter/s, 18.6776s/12 iters), loss = 0.0890709 I0412 14:14:03.324823 8032 solver.cpp:237] Train net output #0: loss = 0.0890709 (* 1 = 0.0890709 loss) I0412 14:14:03.324834 8032 sgd_solver.cpp:105] Iteration 8568, lr = 0.00183196 I0412 14:14:07.805928 8032 solver.cpp:218] Iteration 8580 (2.67798 iter/s, 4.48099s/12 iters), loss = 0.148383 I0412 14:14:07.806097 8032 solver.cpp:237] Train net output #0: loss = 0.148383 (* 1 = 0.148383 loss) I0412 14:14:07.806108 8032 sgd_solver.cpp:105] Iteration 8580, lr = 0.00182761 I0412 14:14:13.278187 8032 solver.cpp:218] Iteration 8592 (2.193 iter/s, 5.47196s/12 iters), loss = 0.188333 I0412 14:14:13.278241 8032 solver.cpp:237] Train net output #0: loss = 0.188333 (* 1 = 0.188333 loss) I0412 14:14:13.278254 8032 sgd_solver.cpp:105] Iteration 8592, lr = 0.00182327 I0412 14:14:15.696604 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:14:18.796615 8032 solver.cpp:218] Iteration 8604 (2.17461 iter/s, 5.51823s/12 iters), loss = 0.0660966 I0412 14:14:18.796679 8032 solver.cpp:237] Train net output #0: loss = 0.0660966 (* 1 = 0.0660966 loss) I0412 14:14:18.796694 8032 sgd_solver.cpp:105] Iteration 8604, lr = 0.00181894 I0412 14:14:24.251595 8032 solver.cpp:218] Iteration 8616 (2.1999 iter/s, 5.45479s/12 iters), loss = 0.0565305 I0412 14:14:24.251643 8032 solver.cpp:237] Train net output #0: loss = 0.0565305 (* 1 = 0.0565305 loss) I0412 14:14:24.251653 8032 sgd_solver.cpp:105] Iteration 8616, lr = 0.00181462 I0412 14:14:29.808848 8032 solver.cpp:218] Iteration 8628 (2.15941 iter/s, 5.55706s/12 iters), loss = 0.198804 I0412 14:14:29.808904 8032 solver.cpp:237] Train net output #0: loss = 0.198804 (* 1 = 0.198804 loss) I0412 14:14:29.808917 8032 sgd_solver.cpp:105] Iteration 8628, lr = 0.00181031 I0412 14:14:34.968487 8032 solver.cpp:218] Iteration 8640 (2.32583 iter/s, 5.15945s/12 iters), loss = 0.139953 I0412 14:14:34.968561 8032 solver.cpp:237] Train net output #0: loss = 0.139953 (* 1 = 0.139953 loss) I0412 14:14:34.968583 8032 sgd_solver.cpp:105] Iteration 8640, lr = 0.00180602 I0412 14:14:40.173409 8032 solver.cpp:218] Iteration 8652 (2.3056 iter/s, 5.20472s/12 iters), loss = 0.165303 I0412 14:14:40.173892 8032 solver.cpp:237] Train net output #0: loss = 0.165303 (* 1 = 0.165303 loss) I0412 14:14:40.173908 8032 sgd_solver.cpp:105] Iteration 8652, lr = 0.00180173 I0412 14:14:45.408201 8032 solver.cpp:218] Iteration 8664 (2.29262 iter/s, 5.23418s/12 iters), loss = 0.217175 I0412 14:14:45.408252 8032 solver.cpp:237] Train net output #0: loss = 0.217175 (* 1 = 0.217175 loss) I0412 14:14:45.408264 8032 sgd_solver.cpp:105] Iteration 8664, lr = 0.00179745 I0412 14:14:47.466298 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8670.caffemodel I0412 14:14:52.601738 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8670.solverstate I0412 14:14:57.792028 8032 solver.cpp:330] Iteration 8670, Testing net (#0) I0412 14:14:57.792055 8032 net.cpp:676] Ignoring source layer train-data I0412 14:14:58.810819 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:15:02.313160 8032 solver.cpp:397] Test net output #0: accuracy = 0.473652 I0412 14:15:02.313202 8032 solver.cpp:397] Test net output #1: loss = 2.89066 (* 1 = 2.89066 loss) I0412 14:15:04.418573 8032 solver.cpp:218] Iteration 8676 (0.631251 iter/s, 19.0099s/12 iters), loss = 0.160737 I0412 14:15:04.418635 8032 solver.cpp:237] Train net output #0: loss = 0.160737 (* 1 = 0.160737 loss) I0412 14:15:04.418648 8032 sgd_solver.cpp:105] Iteration 8676, lr = 0.00179318 I0412 14:15:09.437140 8032 solver.cpp:218] Iteration 8688 (2.39121 iter/s, 5.01838s/12 iters), loss = 0.129835 I0412 14:15:09.437196 8032 solver.cpp:237] Train net output #0: loss = 0.129835 (* 1 = 0.129835 loss) I0412 14:15:09.437207 8032 sgd_solver.cpp:105] Iteration 8688, lr = 0.00178893 I0412 14:15:13.829344 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:15:14.571853 8032 solver.cpp:218] Iteration 8700 (2.33712 iter/s, 5.13453s/12 iters), loss = 0.0952455 I0412 14:15:14.571909 8032 solver.cpp:237] Train net output #0: loss = 0.0952455 (* 1 = 0.0952455 loss) I0412 14:15:14.571920 8032 sgd_solver.cpp:105] Iteration 8700, lr = 0.00178468 I0412 14:15:19.748450 8032 solver.cpp:218] Iteration 8712 (2.31821 iter/s, 5.17641s/12 iters), loss = 0.11294 I0412 14:15:19.748497 8032 solver.cpp:237] Train net output #0: loss = 0.11294 (* 1 = 0.11294 loss) I0412 14:15:19.748509 8032 sgd_solver.cpp:105] Iteration 8712, lr = 0.00178044 I0412 14:15:24.872051 8032 solver.cpp:218] Iteration 8724 (2.34219 iter/s, 5.12342s/12 iters), loss = 0.0655398 I0412 14:15:24.872107 8032 solver.cpp:237] Train net output #0: loss = 0.0655398 (* 1 = 0.0655398 loss) I0412 14:15:24.872117 8032 sgd_solver.cpp:105] Iteration 8724, lr = 0.00177621 I0412 14:15:30.249601 8032 solver.cpp:218] Iteration 8736 (2.23158 iter/s, 5.37736s/12 iters), loss = 0.204817 I0412 14:15:30.249652 8032 solver.cpp:237] Train net output #0: loss = 0.204817 (* 1 = 0.204817 loss) I0412 14:15:30.249663 8032 sgd_solver.cpp:105] Iteration 8736, lr = 0.001772 I0412 14:15:35.755538 8032 solver.cpp:218] Iteration 8748 (2.17954 iter/s, 5.50574s/12 iters), loss = 0.106152 I0412 14:15:35.755587 8032 solver.cpp:237] Train net output #0: loss = 0.106152 (* 1 = 0.106152 loss) I0412 14:15:35.755596 8032 sgd_solver.cpp:105] Iteration 8748, lr = 0.00176779 I0412 14:15:41.161546 8032 solver.cpp:218] Iteration 8760 (2.21983 iter/s, 5.40582s/12 iters), loss = 0.110275 I0412 14:15:41.161600 8032 solver.cpp:237] Train net output #0: loss = 0.110275 (* 1 = 0.110275 loss) I0412 14:15:41.161612 8032 sgd_solver.cpp:105] Iteration 8760, lr = 0.00176359 I0412 14:15:45.958107 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8772.caffemodel I0412 14:15:49.365583 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8772.solverstate I0412 14:15:53.038798 8032 solver.cpp:330] Iteration 8772, Testing net (#0) I0412 14:15:53.038826 8032 net.cpp:676] Ignoring source layer train-data I0412 14:15:54.026794 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:15:57.479122 8032 solver.cpp:397] Test net output #0: accuracy = 0.466912 I0412 14:15:57.479164 8032 solver.cpp:397] Test net output #1: loss = 2.96563 (* 1 = 2.96563 loss) I0412 14:15:57.567765 8032 solver.cpp:218] Iteration 8772 (0.73145 iter/s, 16.4058s/12 iters), loss = 0.188505 I0412 14:15:57.567834 8032 solver.cpp:237] Train net output #0: loss = 0.188505 (* 1 = 0.188505 loss) I0412 14:15:57.567849 8032 sgd_solver.cpp:105] Iteration 8772, lr = 0.00175941 I0412 14:16:01.854912 8032 solver.cpp:218] Iteration 8784 (2.79918 iter/s, 4.28697s/12 iters), loss = 0.233927 I0412 14:16:01.854959 8032 solver.cpp:237] Train net output #0: loss = 0.233927 (* 1 = 0.233927 loss) I0412 14:16:01.854967 8032 sgd_solver.cpp:105] Iteration 8784, lr = 0.00175523 I0412 14:16:07.158852 8032 solver.cpp:218] Iteration 8796 (2.26255 iter/s, 5.30376s/12 iters), loss = 0.170884 I0412 14:16:07.158900 8032 solver.cpp:237] Train net output #0: loss = 0.170884 (* 1 = 0.170884 loss) I0412 14:16:07.158908 8032 sgd_solver.cpp:105] Iteration 8796, lr = 0.00175106 I0412 14:16:08.647789 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:16:12.541723 8032 solver.cpp:218] Iteration 8808 (2.22937 iter/s, 5.38268s/12 iters), loss = 0.109316 I0412 14:16:12.541765 8032 solver.cpp:237] Train net output #0: loss = 0.109316 (* 1 = 0.109316 loss) I0412 14:16:12.541774 8032 sgd_solver.cpp:105] Iteration 8808, lr = 0.0017469 I0412 14:16:17.640730 8032 solver.cpp:218] Iteration 8820 (2.35348 iter/s, 5.09883s/12 iters), loss = 0.13025 I0412 14:16:17.640902 8032 solver.cpp:237] Train net output #0: loss = 0.13025 (* 1 = 0.13025 loss) I0412 14:16:17.640916 8032 sgd_solver.cpp:105] Iteration 8820, lr = 0.00174276 I0412 14:16:22.586339 8032 solver.cpp:218] Iteration 8832 (2.42654 iter/s, 4.94531s/12 iters), loss = 0.0707032 I0412 14:16:22.586387 8032 solver.cpp:237] Train net output #0: loss = 0.0707032 (* 1 = 0.0707032 loss) I0412 14:16:22.586400 8032 sgd_solver.cpp:105] Iteration 8832, lr = 0.00173862 I0412 14:16:27.442807 8032 solver.cpp:218] Iteration 8844 (2.47102 iter/s, 4.85629s/12 iters), loss = 0.138285 I0412 14:16:27.442864 8032 solver.cpp:237] Train net output #0: loss = 0.138285 (* 1 = 0.138285 loss) I0412 14:16:27.442876 8032 sgd_solver.cpp:105] Iteration 8844, lr = 0.00173449 I0412 14:16:32.718210 8032 solver.cpp:218] Iteration 8856 (2.27479 iter/s, 5.27521s/12 iters), loss = 0.0911897 I0412 14:16:32.718266 8032 solver.cpp:237] Train net output #0: loss = 0.0911897 (* 1 = 0.0911897 loss) I0412 14:16:32.718279 8032 sgd_solver.cpp:105] Iteration 8856, lr = 0.00173037 I0412 14:16:37.729157 8032 solver.cpp:218] Iteration 8868 (2.39485 iter/s, 5.01076s/12 iters), loss = 0.0682498 I0412 14:16:37.729218 8032 solver.cpp:237] Train net output #0: loss = 0.0682498 (* 1 = 0.0682498 loss) I0412 14:16:37.729230 8032 sgd_solver.cpp:105] Iteration 8868, lr = 0.00172626 I0412 14:16:39.757081 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8874.caffemodel I0412 14:16:42.856169 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8874.solverstate I0412 14:16:45.175671 8032 solver.cpp:330] Iteration 8874, Testing net (#0) I0412 14:16:45.175698 8032 net.cpp:676] Ignoring source layer train-data I0412 14:16:46.114815 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:16:49.664372 8032 solver.cpp:397] Test net output #0: accuracy = 0.479167 I0412 14:16:49.664489 8032 solver.cpp:397] Test net output #1: loss = 2.9533 (* 1 = 2.9533 loss) I0412 14:16:51.710521 8032 solver.cpp:218] Iteration 8880 (0.85831 iter/s, 13.981s/12 iters), loss = 0.180451 I0412 14:16:51.710575 8032 solver.cpp:237] Train net output #0: loss = 0.180451 (* 1 = 0.180451 loss) I0412 14:16:51.710590 8032 sgd_solver.cpp:105] Iteration 8880, lr = 0.00172217 I0412 14:16:57.039144 8032 solver.cpp:218] Iteration 8892 (2.25207 iter/s, 5.32842s/12 iters), loss = 0.133444 I0412 14:16:57.039192 8032 solver.cpp:237] Train net output #0: loss = 0.133444 (* 1 = 0.133444 loss) I0412 14:16:57.039202 8032 sgd_solver.cpp:105] Iteration 8892, lr = 0.00171808 I0412 14:17:00.632334 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:17:02.152292 8032 solver.cpp:218] Iteration 8904 (2.34697 iter/s, 5.11297s/12 iters), loss = 0.0248051 I0412 14:17:02.152335 8032 solver.cpp:237] Train net output #0: loss = 0.0248051 (* 1 = 0.0248051 loss) I0412 14:17:02.152348 8032 sgd_solver.cpp:105] Iteration 8904, lr = 0.001714 I0412 14:17:07.560132 8032 solver.cpp:218] Iteration 8916 (2.21908 iter/s, 5.40765s/12 iters), loss = 0.128694 I0412 14:17:07.560195 8032 solver.cpp:237] Train net output #0: loss = 0.128694 (* 1 = 0.128694 loss) I0412 14:17:07.560211 8032 sgd_solver.cpp:105] Iteration 8916, lr = 0.00170993 I0412 14:17:12.529254 8032 solver.cpp:218] Iteration 8928 (2.41501 iter/s, 4.96893s/12 iters), loss = 0.190176 I0412 14:17:12.529297 8032 solver.cpp:237] Train net output #0: loss = 0.190176 (* 1 = 0.190176 loss) I0412 14:17:12.529305 8032 sgd_solver.cpp:105] Iteration 8928, lr = 0.00170587 I0412 14:17:17.616299 8032 solver.cpp:218] Iteration 8940 (2.35902 iter/s, 5.08686s/12 iters), loss = 0.142237 I0412 14:17:17.616358 8032 solver.cpp:237] Train net output #0: loss = 0.142237 (* 1 = 0.142237 loss) I0412 14:17:17.616376 8032 sgd_solver.cpp:105] Iteration 8940, lr = 0.00170182 I0412 14:17:23.001731 8032 solver.cpp:218] Iteration 8952 (2.22832 iter/s, 5.38523s/12 iters), loss = 0.0964781 I0412 14:17:23.001884 8032 solver.cpp:237] Train net output #0: loss = 0.0964781 (* 1 = 0.0964781 loss) I0412 14:17:23.001899 8032 sgd_solver.cpp:105] Iteration 8952, lr = 0.00169778 I0412 14:17:28.231853 8032 solver.cpp:218] Iteration 8964 (2.29453 iter/s, 5.22984s/12 iters), loss = 0.113249 I0412 14:17:28.231899 8032 solver.cpp:237] Train net output #0: loss = 0.113249 (* 1 = 0.113249 loss) I0412 14:17:28.231907 8032 sgd_solver.cpp:105] Iteration 8964, lr = 0.00169375 I0412 14:17:32.844199 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8976.caffemodel I0412 14:17:36.054044 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8976.solverstate I0412 14:17:38.419034 8032 solver.cpp:330] Iteration 8976, Testing net (#0) I0412 14:17:38.419060 8032 net.cpp:676] Ignoring source layer train-data I0412 14:17:39.370043 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:17:42.901741 8032 solver.cpp:397] Test net output #0: accuracy = 0.476103 I0412 14:17:42.901784 8032 solver.cpp:397] Test net output #1: loss = 2.91971 (* 1 = 2.91971 loss) I0412 14:17:42.990197 8032 solver.cpp:218] Iteration 8976 (0.813122 iter/s, 14.7579s/12 iters), loss = 0.0253515 I0412 14:17:42.990254 8032 solver.cpp:237] Train net output #0: loss = 0.0253515 (* 1 = 0.0253515 loss) I0412 14:17:42.990267 8032 sgd_solver.cpp:105] Iteration 8976, lr = 0.00168973 I0412 14:17:47.696908 8032 solver.cpp:218] Iteration 8988 (2.54966 iter/s, 4.70652s/12 iters), loss = 0.0518003 I0412 14:17:47.696964 8032 solver.cpp:237] Train net output #0: loss = 0.0518003 (* 1 = 0.0518003 loss) I0412 14:17:47.696976 8032 sgd_solver.cpp:105] Iteration 8988, lr = 0.00168571 I0412 14:17:51.030730 8032 blocking_queue.cpp:49] Waiting for data I0412 14:17:52.811408 8032 solver.cpp:218] Iteration 9000 (2.34636 iter/s, 5.11431s/12 iters), loss = 0.269431 I0412 14:17:52.811457 8032 solver.cpp:237] Train net output #0: loss = 0.269431 (* 1 = 0.269431 loss) I0412 14:17:52.811468 8032 sgd_solver.cpp:105] Iteration 9000, lr = 0.00168171 I0412 14:17:53.473449 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:17:57.613404 8032 solver.cpp:218] Iteration 9012 (2.49905 iter/s, 4.80182s/12 iters), loss = 0.0645164 I0412 14:17:57.613451 8032 solver.cpp:237] Train net output #0: loss = 0.0645165 (* 1 = 0.0645165 loss) I0412 14:17:57.613462 8032 sgd_solver.cpp:105] Iteration 9012, lr = 0.00167772 I0412 14:18:02.960711 8032 solver.cpp:218] Iteration 9024 (2.2442 iter/s, 5.34712s/12 iters), loss = 0.0714488 I0412 14:18:02.960757 8032 solver.cpp:237] Train net output #0: loss = 0.0714489 (* 1 = 0.0714489 loss) I0412 14:18:02.960767 8032 sgd_solver.cpp:105] Iteration 9024, lr = 0.00167374 I0412 14:18:08.678411 8032 solver.cpp:218] Iteration 9036 (2.09882 iter/s, 5.7175s/12 iters), loss = 0.0341797 I0412 14:18:08.678465 8032 solver.cpp:237] Train net output #0: loss = 0.0341798 (* 1 = 0.0341798 loss) I0412 14:18:08.678476 8032 sgd_solver.cpp:105] Iteration 9036, lr = 0.00166976 I0412 14:18:13.675083 8032 solver.cpp:218] Iteration 9048 (2.40169 iter/s, 4.99649s/12 iters), loss = 0.154288 I0412 14:18:13.675130 8032 solver.cpp:237] Train net output #0: loss = 0.154288 (* 1 = 0.154288 loss) I0412 14:18:13.675140 8032 sgd_solver.cpp:105] Iteration 9048, lr = 0.0016658 I0412 14:18:18.863066 8032 solver.cpp:218] Iteration 9060 (2.31312 iter/s, 5.18779s/12 iters), loss = 0.0404 I0412 14:18:18.863128 8032 solver.cpp:237] Train net output #0: loss = 0.0404001 (* 1 = 0.0404001 loss) I0412 14:18:18.863142 8032 sgd_solver.cpp:105] Iteration 9060, lr = 0.00166184 I0412 14:18:24.363513 8032 solver.cpp:218] Iteration 9072 (2.18172 iter/s, 5.50024s/12 iters), loss = 0.163352 I0412 14:18:24.363668 8032 solver.cpp:237] Train net output #0: loss = 0.163352 (* 1 = 0.163352 loss) I0412 14:18:24.363682 8032 sgd_solver.cpp:105] Iteration 9072, lr = 0.0016579 I0412 14:18:26.459237 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9078.caffemodel I0412 14:18:29.923486 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9078.solverstate I0412 14:18:32.199476 8032 solver.cpp:330] Iteration 9078, Testing net (#0) I0412 14:18:32.199501 8032 net.cpp:676] Ignoring source layer train-data I0412 14:18:33.111455 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:18:36.758924 8032 solver.cpp:397] Test net output #0: accuracy = 0.483456 I0412 14:18:36.758961 8032 solver.cpp:397] Test net output #1: loss = 2.97421 (* 1 = 2.97421 loss) I0412 14:18:38.765723 8032 solver.cpp:218] Iteration 9084 (0.833235 iter/s, 14.4017s/12 iters), loss = 0.164739 I0412 14:18:38.765777 8032 solver.cpp:237] Train net output #0: loss = 0.164739 (* 1 = 0.164739 loss) I0412 14:18:38.765789 8032 sgd_solver.cpp:105] Iteration 9084, lr = 0.00165396 I0412 14:18:43.900225 8032 solver.cpp:218] Iteration 9096 (2.33722 iter/s, 5.13431s/12 iters), loss = 0.131306 I0412 14:18:43.900271 8032 solver.cpp:237] Train net output #0: loss = 0.131306 (* 1 = 0.131306 loss) I0412 14:18:43.900280 8032 sgd_solver.cpp:105] Iteration 9096, lr = 0.00165003 I0412 14:18:46.970414 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:18:49.068632 8032 solver.cpp:218] Iteration 9108 (2.32188 iter/s, 5.16822s/12 iters), loss = 0.0836777 I0412 14:18:49.068684 8032 solver.cpp:237] Train net output #0: loss = 0.0836777 (* 1 = 0.0836777 loss) I0412 14:18:49.068696 8032 sgd_solver.cpp:105] Iteration 9108, lr = 0.00164612 I0412 14:18:54.196276 8032 solver.cpp:218] Iteration 9120 (2.34034 iter/s, 5.12745s/12 iters), loss = 0.135569 I0412 14:18:54.196328 8032 solver.cpp:237] Train net output #0: loss = 0.135569 (* 1 = 0.135569 loss) I0412 14:18:54.196339 8032 sgd_solver.cpp:105] Iteration 9120, lr = 0.00164221 I0412 14:18:59.378994 8032 solver.cpp:218] Iteration 9132 (2.31547 iter/s, 5.18252s/12 iters), loss = 0.112213 I0412 14:18:59.379112 8032 solver.cpp:237] Train net output #0: loss = 0.112213 (* 1 = 0.112213 loss) I0412 14:18:59.379124 8032 sgd_solver.cpp:105] Iteration 9132, lr = 0.00163831 I0412 14:19:04.671983 8032 solver.cpp:218] Iteration 9144 (2.26726 iter/s, 5.29273s/12 iters), loss = 0.0985844 I0412 14:19:04.672034 8032 solver.cpp:237] Train net output #0: loss = 0.0985844 (* 1 = 0.0985844 loss) I0412 14:19:04.672044 8032 sgd_solver.cpp:105] Iteration 9144, lr = 0.00163442 I0412 14:19:10.294775 8032 solver.cpp:218] Iteration 9156 (2.13425 iter/s, 5.6226s/12 iters), loss = 0.0960109 I0412 14:19:10.294816 8032 solver.cpp:237] Train net output #0: loss = 0.0960109 (* 1 = 0.0960109 loss) I0412 14:19:10.294824 8032 sgd_solver.cpp:105] Iteration 9156, lr = 0.00163054 I0412 14:19:15.464419 8032 solver.cpp:218] Iteration 9168 (2.32133 iter/s, 5.16946s/12 iters), loss = 0.0445728 I0412 14:19:15.464463 8032 solver.cpp:237] Train net output #0: loss = 0.0445728 (* 1 = 0.0445728 loss) I0412 14:19:15.464473 8032 sgd_solver.cpp:105] Iteration 9168, lr = 0.00162667 I0412 14:19:20.419634 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9180.caffemodel I0412 14:19:23.405599 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9180.solverstate I0412 14:19:25.778075 8032 solver.cpp:330] Iteration 9180, Testing net (#0) I0412 14:19:25.778098 8032 net.cpp:676] Ignoring source layer train-data I0412 14:19:26.576834 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:19:30.439774 8032 solver.cpp:397] Test net output #0: accuracy = 0.473652 I0412 14:19:30.439934 8032 solver.cpp:397] Test net output #1: loss = 3.01549 (* 1 = 3.01549 loss) I0412 14:19:30.528579 8032 solver.cpp:218] Iteration 9180 (0.796615 iter/s, 15.0637s/12 iters), loss = 0.0607797 I0412 14:19:30.528630 8032 solver.cpp:237] Train net output #0: loss = 0.0607797 (* 1 = 0.0607797 loss) I0412 14:19:30.528641 8032 sgd_solver.cpp:105] Iteration 9180, lr = 0.00162281 I0412 14:19:34.915321 8032 solver.cpp:218] Iteration 9192 (2.73563 iter/s, 4.38656s/12 iters), loss = 0.148058 I0412 14:19:34.915380 8032 solver.cpp:237] Train net output #0: loss = 0.148058 (* 1 = 0.148058 loss) I0412 14:19:34.915392 8032 sgd_solver.cpp:105] Iteration 9192, lr = 0.00161895 I0412 14:19:40.008697 8032 solver.cpp:218] Iteration 9204 (2.35609 iter/s, 5.09318s/12 iters), loss = 0.174324 I0412 14:19:40.008752 8032 solver.cpp:237] Train net output #0: loss = 0.174324 (* 1 = 0.174324 loss) I0412 14:19:40.008765 8032 sgd_solver.cpp:105] Iteration 9204, lr = 0.00161511 I0412 14:19:40.091717 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:19:45.058917 8032 solver.cpp:218] Iteration 9216 (2.37623 iter/s, 5.05002s/12 iters), loss = 0.121579 I0412 14:19:45.058972 8032 solver.cpp:237] Train net output #0: loss = 0.121579 (* 1 = 0.121579 loss) I0412 14:19:45.058984 8032 sgd_solver.cpp:105] Iteration 9216, lr = 0.00161128 I0412 14:19:50.134557 8032 solver.cpp:218] Iteration 9228 (2.36432 iter/s, 5.07545s/12 iters), loss = 0.061577 I0412 14:19:50.134598 8032 solver.cpp:237] Train net output #0: loss = 0.061577 (* 1 = 0.061577 loss) I0412 14:19:50.134605 8032 sgd_solver.cpp:105] Iteration 9228, lr = 0.00160745 I0412 14:19:55.155894 8032 solver.cpp:218] Iteration 9240 (2.38989 iter/s, 5.02116s/12 iters), loss = 0.119659 I0412 14:19:55.155934 8032 solver.cpp:237] Train net output #0: loss = 0.119659 (* 1 = 0.119659 loss) I0412 14:19:55.155943 8032 sgd_solver.cpp:105] Iteration 9240, lr = 0.00160363 I0412 14:20:00.424996 8032 solver.cpp:218] Iteration 9252 (2.27751 iter/s, 5.26891s/12 iters), loss = 0.135695 I0412 14:20:00.425050 8032 solver.cpp:237] Train net output #0: loss = 0.135695 (* 1 = 0.135695 loss) I0412 14:20:00.425063 8032 sgd_solver.cpp:105] Iteration 9252, lr = 0.00159983 I0412 14:20:05.776206 8032 solver.cpp:218] Iteration 9264 (2.24257 iter/s, 5.351s/12 iters), loss = 0.0724511 I0412 14:20:05.776322 8032 solver.cpp:237] Train net output #0: loss = 0.0724511 (* 1 = 0.0724511 loss) I0412 14:20:05.776335 8032 sgd_solver.cpp:105] Iteration 9264, lr = 0.00159603 I0412 14:20:10.873718 8032 solver.cpp:218] Iteration 9276 (2.35421 iter/s, 5.09725s/12 iters), loss = 0.0976801 I0412 14:20:10.873770 8032 solver.cpp:237] Train net output #0: loss = 0.0976801 (* 1 = 0.0976801 loss) I0412 14:20:10.873780 8032 sgd_solver.cpp:105] Iteration 9276, lr = 0.00159224 I0412 14:20:12.990300 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9282.caffemodel I0412 14:20:16.028350 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9282.solverstate I0412 14:20:18.351660 8032 solver.cpp:330] Iteration 9282, Testing net (#0) I0412 14:20:18.351686 8032 net.cpp:676] Ignoring source layer train-data I0412 14:20:19.149487 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:20:22.932770 8032 solver.cpp:397] Test net output #0: accuracy = 0.466299 I0412 14:20:22.932811 8032 solver.cpp:397] Test net output #1: loss = 3.00211 (* 1 = 3.00211 loss) I0412 14:20:24.913323 8032 solver.cpp:218] Iteration 9288 (0.85475 iter/s, 14.0392s/12 iters), loss = 0.0847554 I0412 14:20:24.913370 8032 solver.cpp:237] Train net output #0: loss = 0.0847554 (* 1 = 0.0847554 loss) I0412 14:20:24.913380 8032 sgd_solver.cpp:105] Iteration 9288, lr = 0.00158846 I0412 14:20:30.033509 8032 solver.cpp:218] Iteration 9300 (2.34375 iter/s, 5.11999s/12 iters), loss = 0.0293582 I0412 14:20:30.033569 8032 solver.cpp:237] Train net output #0: loss = 0.0293582 (* 1 = 0.0293582 loss) I0412 14:20:30.033583 8032 sgd_solver.cpp:105] Iteration 9300, lr = 0.00158469 I0412 14:20:32.345096 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:20:35.191722 8032 solver.cpp:218] Iteration 9312 (2.32648 iter/s, 5.15801s/12 iters), loss = 0.0791316 I0412 14:20:35.191771 8032 solver.cpp:237] Train net output #0: loss = 0.0791316 (* 1 = 0.0791316 loss) I0412 14:20:35.191781 8032 sgd_solver.cpp:105] Iteration 9312, lr = 0.00158092 I0412 14:20:40.641611 8032 solver.cpp:218] Iteration 9324 (2.20196 iter/s, 5.44969s/12 iters), loss = 0.0548814 I0412 14:20:40.641727 8032 solver.cpp:237] Train net output #0: loss = 0.0548814 (* 1 = 0.0548814 loss) I0412 14:20:40.641736 8032 sgd_solver.cpp:105] Iteration 9324, lr = 0.00157717 I0412 14:20:45.855505 8032 solver.cpp:218] Iteration 9336 (2.30166 iter/s, 5.21363s/12 iters), loss = 0.0464497 I0412 14:20:45.855549 8032 solver.cpp:237] Train net output #0: loss = 0.0464497 (* 1 = 0.0464497 loss) I0412 14:20:45.855558 8032 sgd_solver.cpp:105] Iteration 9336, lr = 0.00157343 I0412 14:20:51.446015 8032 solver.cpp:218] Iteration 9348 (2.14657 iter/s, 5.59032s/12 iters), loss = 0.114463 I0412 14:20:51.446058 8032 solver.cpp:237] Train net output #0: loss = 0.114463 (* 1 = 0.114463 loss) I0412 14:20:51.446069 8032 sgd_solver.cpp:105] Iteration 9348, lr = 0.00156969 I0412 14:20:56.558940 8032 solver.cpp:218] Iteration 9360 (2.34708 iter/s, 5.11274s/12 iters), loss = 0.0697213 I0412 14:20:56.558982 8032 solver.cpp:237] Train net output #0: loss = 0.0697213 (* 1 = 0.0697213 loss) I0412 14:20:56.558991 8032 sgd_solver.cpp:105] Iteration 9360, lr = 0.00156596 I0412 14:21:01.622268 8032 solver.cpp:218] Iteration 9372 (2.37007 iter/s, 5.06314s/12 iters), loss = 0.0667925 I0412 14:21:01.622314 8032 solver.cpp:237] Train net output #0: loss = 0.0667925 (* 1 = 0.0667925 loss) I0412 14:21:01.622323 8032 sgd_solver.cpp:105] Iteration 9372, lr = 0.00156225 I0412 14:21:06.173506 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9384.caffemodel I0412 14:21:09.249609 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9384.solverstate I0412 14:21:11.617535 8032 solver.cpp:330] Iteration 9384, Testing net (#0) I0412 14:21:11.617609 8032 net.cpp:676] Ignoring source layer train-data I0412 14:21:12.403970 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:21:16.074429 8032 solver.cpp:397] Test net output #0: accuracy = 0.478554 I0412 14:21:16.074478 8032 solver.cpp:397] Test net output #1: loss = 3.01758 (* 1 = 3.01758 loss) I0412 14:21:16.163761 8032 solver.cpp:218] Iteration 9384 (0.825249 iter/s, 14.5411s/12 iters), loss = 0.0796176 I0412 14:21:16.163825 8032 solver.cpp:237] Train net output #0: loss = 0.0796176 (* 1 = 0.0796176 loss) I0412 14:21:16.163836 8032 sgd_solver.cpp:105] Iteration 9384, lr = 0.00155854 I0412 14:21:20.737800 8032 solver.cpp:218] Iteration 9396 (2.62361 iter/s, 4.57384s/12 iters), loss = 0.0961561 I0412 14:21:20.737867 8032 solver.cpp:237] Train net output #0: loss = 0.0961561 (* 1 = 0.0961561 loss) I0412 14:21:20.737886 8032 sgd_solver.cpp:105] Iteration 9396, lr = 0.00155484 I0412 14:21:25.227699 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:21:25.893719 8032 solver.cpp:218] Iteration 9408 (2.32752 iter/s, 5.15571s/12 iters), loss = 0.134061 I0412 14:21:25.893796 8032 solver.cpp:237] Train net output #0: loss = 0.134061 (* 1 = 0.134061 loss) I0412 14:21:25.893810 8032 sgd_solver.cpp:105] Iteration 9408, lr = 0.00155114 I0412 14:21:31.030999 8032 solver.cpp:218] Iteration 9420 (2.33597 iter/s, 5.13706s/12 iters), loss = 0.109167 I0412 14:21:31.031050 8032 solver.cpp:237] Train net output #0: loss = 0.109167 (* 1 = 0.109167 loss) I0412 14:21:31.031061 8032 sgd_solver.cpp:105] Iteration 9420, lr = 0.00154746 I0412 14:21:36.200948 8032 solver.cpp:218] Iteration 9432 (2.3212 iter/s, 5.16975s/12 iters), loss = 0.174548 I0412 14:21:36.201000 8032 solver.cpp:237] Train net output #0: loss = 0.174548 (* 1 = 0.174548 loss) I0412 14:21:36.201012 8032 sgd_solver.cpp:105] Iteration 9432, lr = 0.00154379 I0412 14:21:41.351135 8032 solver.cpp:218] Iteration 9444 (2.3301 iter/s, 5.14999s/12 iters), loss = 0.0597359 I0412 14:21:41.351188 8032 solver.cpp:237] Train net output #0: loss = 0.0597359 (* 1 = 0.0597359 loss) I0412 14:21:41.351200 8032 sgd_solver.cpp:105] Iteration 9444, lr = 0.00154012 I0412 14:21:46.621006 8032 solver.cpp:218] Iteration 9456 (2.27718 iter/s, 5.26967s/12 iters), loss = 0.104595 I0412 14:21:46.621124 8032 solver.cpp:237] Train net output #0: loss = 0.104595 (* 1 = 0.104595 loss) I0412 14:21:46.621134 8032 sgd_solver.cpp:105] Iteration 9456, lr = 0.00153647 I0412 14:21:51.645800 8032 solver.cpp:218] Iteration 9468 (2.38828 iter/s, 5.02454s/12 iters), loss = 0.0721537 I0412 14:21:51.645840 8032 solver.cpp:237] Train net output #0: loss = 0.0721536 (* 1 = 0.0721536 loss) I0412 14:21:51.645849 8032 sgd_solver.cpp:105] Iteration 9468, lr = 0.00153282 I0412 14:21:56.877791 8032 solver.cpp:218] Iteration 9480 (2.29367 iter/s, 5.2318s/12 iters), loss = 0.0597184 I0412 14:21:56.877848 8032 solver.cpp:237] Train net output #0: loss = 0.0597184 (* 1 = 0.0597184 loss) I0412 14:21:56.877861 8032 sgd_solver.cpp:105] Iteration 9480, lr = 0.00152918 I0412 14:21:59.029103 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9486.caffemodel I0412 14:22:01.954408 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9486.solverstate I0412 14:22:04.249675 8032 solver.cpp:330] Iteration 9486, Testing net (#0) I0412 14:22:04.249701 8032 net.cpp:676] Ignoring source layer train-data I0412 14:22:05.007411 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:22:08.865906 8032 solver.cpp:397] Test net output #0: accuracy = 0.477328 I0412 14:22:08.865949 8032 solver.cpp:397] Test net output #1: loss = 2.98022 (* 1 = 2.98022 loss) I0412 14:22:10.922524 8032 solver.cpp:218] Iteration 9492 (0.854439 iter/s, 14.0443s/12 iters), loss = 0.123621 I0412 14:22:10.922577 8032 solver.cpp:237] Train net output #0: loss = 0.123621 (* 1 = 0.123621 loss) I0412 14:22:10.922590 8032 sgd_solver.cpp:105] Iteration 9492, lr = 0.00152555 I0412 14:22:16.303386 8032 solver.cpp:218] Iteration 9504 (2.23021 iter/s, 5.38066s/12 iters), loss = 0.0422153 I0412 14:22:16.303431 8032 solver.cpp:237] Train net output #0: loss = 0.0422153 (* 1 = 0.0422153 loss) I0412 14:22:16.303440 8032 sgd_solver.cpp:105] Iteration 9504, lr = 0.00152193 I0412 14:22:17.881726 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:22:21.678236 8032 solver.cpp:218] Iteration 9516 (2.2327 iter/s, 5.37465s/12 iters), loss = 0.13533 I0412 14:22:21.678273 8032 solver.cpp:237] Train net output #0: loss = 0.13533 (* 1 = 0.13533 loss) I0412 14:22:21.678282 8032 sgd_solver.cpp:105] Iteration 9516, lr = 0.00151831 I0412 14:22:26.774379 8032 solver.cpp:218] Iteration 9528 (2.35481 iter/s, 5.09596s/12 iters), loss = 0.0500655 I0412 14:22:26.774425 8032 solver.cpp:237] Train net output #0: loss = 0.0500655 (* 1 = 0.0500655 loss) I0412 14:22:26.774433 8032 sgd_solver.cpp:105] Iteration 9528, lr = 0.00151471 I0412 14:22:31.992512 8032 solver.cpp:218] Iteration 9540 (2.29976 iter/s, 5.21794s/12 iters), loss = 0.0946619 I0412 14:22:31.992569 8032 solver.cpp:237] Train net output #0: loss = 0.0946619 (* 1 = 0.0946619 loss) I0412 14:22:31.992584 8032 sgd_solver.cpp:105] Iteration 9540, lr = 0.00151111 I0412 14:22:37.056393 8032 solver.cpp:218] Iteration 9552 (2.36982 iter/s, 5.06367s/12 iters), loss = 0.0438274 I0412 14:22:37.056447 8032 solver.cpp:237] Train net output #0: loss = 0.0438274 (* 1 = 0.0438274 loss) I0412 14:22:37.056458 8032 sgd_solver.cpp:105] Iteration 9552, lr = 0.00150752 I0412 14:22:42.067278 8032 solver.cpp:218] Iteration 9564 (2.39488 iter/s, 5.01068s/12 iters), loss = 0.0484788 I0412 14:22:42.067332 8032 solver.cpp:237] Train net output #0: loss = 0.0484788 (* 1 = 0.0484788 loss) I0412 14:22:42.067343 8032 sgd_solver.cpp:105] Iteration 9564, lr = 0.00150395 I0412 14:22:47.175320 8032 solver.cpp:218] Iteration 9576 (2.34933 iter/s, 5.10784s/12 iters), loss = 0.0581066 I0412 14:22:47.175372 8032 solver.cpp:237] Train net output #0: loss = 0.0581065 (* 1 = 0.0581065 loss) I0412 14:22:47.175382 8032 sgd_solver.cpp:105] Iteration 9576, lr = 0.00150037 I0412 14:22:51.750952 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9588.caffemodel I0412 14:22:54.788075 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9588.solverstate I0412 14:22:57.101752 8032 solver.cpp:330] Iteration 9588, Testing net (#0) I0412 14:22:57.101779 8032 net.cpp:676] Ignoring source layer train-data I0412 14:22:57.848184 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:23:01.694289 8032 solver.cpp:397] Test net output #0: accuracy = 0.48652 I0412 14:23:01.694337 8032 solver.cpp:397] Test net output #1: loss = 2.98046 (* 1 = 2.98046 loss) I0412 14:23:01.782825 8032 solver.cpp:218] Iteration 9588 (0.821521 iter/s, 14.6071s/12 iters), loss = 0.115572 I0412 14:23:01.782872 8032 solver.cpp:237] Train net output #0: loss = 0.115572 (* 1 = 0.115572 loss) I0412 14:23:01.782882 8032 sgd_solver.cpp:105] Iteration 9588, lr = 0.00149681 I0412 14:23:06.782405 8032 solver.cpp:218] Iteration 9600 (2.40029 iter/s, 4.99939s/12 iters), loss = 0.120474 I0412 14:23:06.782449 8032 solver.cpp:237] Train net output #0: loss = 0.120474 (* 1 = 0.120474 loss) I0412 14:23:06.782457 8032 sgd_solver.cpp:105] Iteration 9600, lr = 0.00149326 I0412 14:23:10.641324 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:23:12.177336 8032 solver.cpp:218] Iteration 9612 (2.22439 iter/s, 5.39473s/12 iters), loss = 0.0878812 I0412 14:23:12.177387 8032 solver.cpp:237] Train net output #0: loss = 0.0878812 (* 1 = 0.0878812 loss) I0412 14:23:12.177398 8032 sgd_solver.cpp:105] Iteration 9612, lr = 0.00148971 I0412 14:23:17.622433 8032 solver.cpp:218] Iteration 9624 (2.2039 iter/s, 5.44489s/12 iters), loss = 0.0611142 I0412 14:23:17.622484 8032 solver.cpp:237] Train net output #0: loss = 0.0611142 (* 1 = 0.0611142 loss) I0412 14:23:17.622495 8032 sgd_solver.cpp:105] Iteration 9624, lr = 0.00148618 I0412 14:23:22.667330 8032 solver.cpp:218] Iteration 9636 (2.37873 iter/s, 5.0447s/12 iters), loss = 0.0564873 I0412 14:23:22.667409 8032 solver.cpp:237] Train net output #0: loss = 0.0564873 (* 1 = 0.0564873 loss) I0412 14:23:22.667423 8032 sgd_solver.cpp:105] Iteration 9636, lr = 0.00148265 I0412 14:23:27.751112 8032 solver.cpp:218] Iteration 9648 (2.36055 iter/s, 5.08355s/12 iters), loss = 0.102642 I0412 14:23:27.751159 8032 solver.cpp:237] Train net output #0: loss = 0.102642 (* 1 = 0.102642 loss) I0412 14:23:27.751170 8032 sgd_solver.cpp:105] Iteration 9648, lr = 0.00147913 I0412 14:23:32.868710 8032 solver.cpp:218] Iteration 9660 (2.34494 iter/s, 5.1174s/12 iters), loss = 0.0993326 I0412 14:23:32.868769 8032 solver.cpp:237] Train net output #0: loss = 0.0993326 (* 1 = 0.0993326 loss) I0412 14:23:32.868786 8032 sgd_solver.cpp:105] Iteration 9660, lr = 0.00147562 I0412 14:23:37.985744 8032 solver.cpp:218] Iteration 9672 (2.3452 iter/s, 5.11683s/12 iters), loss = 0.0485323 I0412 14:23:37.985792 8032 solver.cpp:237] Train net output #0: loss = 0.0485322 (* 1 = 0.0485322 loss) I0412 14:23:37.985801 8032 sgd_solver.cpp:105] Iteration 9672, lr = 0.00147211 I0412 14:23:43.220562 8032 solver.cpp:218] Iteration 9684 (2.29243 iter/s, 5.23462s/12 iters), loss = 0.102614 I0412 14:23:43.220605 8032 solver.cpp:237] Train net output #0: loss = 0.102614 (* 1 = 0.102614 loss) I0412 14:23:43.220613 8032 sgd_solver.cpp:105] Iteration 9684, lr = 0.00146862 I0412 14:23:45.253438 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9690.caffemodel I0412 14:23:48.876560 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9690.solverstate I0412 14:23:51.246268 8032 solver.cpp:330] Iteration 9690, Testing net (#0) I0412 14:23:51.246295 8032 net.cpp:676] Ignoring source layer train-data I0412 14:23:51.862074 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:23:54.958139 8032 blocking_queue.cpp:49] Waiting for data I0412 14:23:55.952543 8032 solver.cpp:397] Test net output #0: accuracy = 0.493873 I0412 14:23:55.952581 8032 solver.cpp:397] Test net output #1: loss = 2.95954 (* 1 = 2.95954 loss) I0412 14:23:57.875437 8032 solver.cpp:218] Iteration 9696 (0.818865 iter/s, 14.6544s/12 iters), loss = 0.158326 I0412 14:23:57.875483 8032 solver.cpp:237] Train net output #0: loss = 0.158326 (* 1 = 0.158326 loss) I0412 14:23:57.875491 8032 sgd_solver.cpp:105] Iteration 9696, lr = 0.00146513 I0412 14:24:02.948668 8032 solver.cpp:218] Iteration 9708 (2.36545 iter/s, 5.07303s/12 iters), loss = 0.0827165 I0412 14:24:02.948734 8032 solver.cpp:237] Train net output #0: loss = 0.0827165 (* 1 = 0.0827165 loss) I0412 14:24:02.948746 8032 sgd_solver.cpp:105] Iteration 9708, lr = 0.00146165 I0412 14:24:03.677026 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:24:07.899219 8032 solver.cpp:218] Iteration 9720 (2.42408 iter/s, 4.95034s/12 iters), loss = 0.0958916 I0412 14:24:07.899276 8032 solver.cpp:237] Train net output #0: loss = 0.0958916 (* 1 = 0.0958916 loss) I0412 14:24:07.899288 8032 sgd_solver.cpp:105] Iteration 9720, lr = 0.00145818 I0412 14:24:12.921582 8032 solver.cpp:218] Iteration 9732 (2.38941 iter/s, 5.02216s/12 iters), loss = 0.036704 I0412 14:24:12.921643 8032 solver.cpp:237] Train net output #0: loss = 0.036704 (* 1 = 0.036704 loss) I0412 14:24:12.921654 8032 sgd_solver.cpp:105] Iteration 9732, lr = 0.00145472 I0412 14:24:17.901343 8032 solver.cpp:218] Iteration 9744 (2.40986 iter/s, 4.97955s/12 iters), loss = 0.14867 I0412 14:24:17.901408 8032 solver.cpp:237] Train net output #0: loss = 0.14867 (* 1 = 0.14867 loss) I0412 14:24:17.901420 8032 sgd_solver.cpp:105] Iteration 9744, lr = 0.00145127 I0412 14:24:22.822016 8032 solver.cpp:218] Iteration 9756 (2.43879 iter/s, 4.92046s/12 iters), loss = 0.250834 I0412 14:24:22.822070 8032 solver.cpp:237] Train net output #0: loss = 0.250834 (* 1 = 0.250834 loss) I0412 14:24:22.822083 8032 sgd_solver.cpp:105] Iteration 9756, lr = 0.00144782 I0412 14:24:27.757845 8032 solver.cpp:218] Iteration 9768 (2.4313 iter/s, 4.93563s/12 iters), loss = 0.0531197 I0412 14:24:27.757933 8032 solver.cpp:237] Train net output #0: loss = 0.0531197 (* 1 = 0.0531197 loss) I0412 14:24:27.757946 8032 sgd_solver.cpp:105] Iteration 9768, lr = 0.00144438 I0412 14:24:32.767982 8032 solver.cpp:218] Iteration 9780 (2.39526 iter/s, 5.0099s/12 iters), loss = 0.0831798 I0412 14:24:32.768035 8032 solver.cpp:237] Train net output #0: loss = 0.0831798 (* 1 = 0.0831798 loss) I0412 14:24:32.768047 8032 sgd_solver.cpp:105] Iteration 9780, lr = 0.00144095 I0412 14:24:37.308048 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9792.caffemodel I0412 14:24:41.428753 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9792.solverstate I0412 14:24:45.067683 8032 solver.cpp:330] Iteration 9792, Testing net (#0) I0412 14:24:45.067711 8032 net.cpp:676] Ignoring source layer train-data I0412 14:24:45.620959 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:24:49.525378 8032 solver.cpp:397] Test net output #0: accuracy = 0.485294 I0412 14:24:49.525432 8032 solver.cpp:397] Test net output #1: loss = 2.99133 (* 1 = 2.99133 loss) I0412 14:24:49.613996 8032 solver.cpp:218] Iteration 9792 (0.712357 iter/s, 16.8455s/12 iters), loss = 0.126416 I0412 14:24:49.614049 8032 solver.cpp:237] Train net output #0: loss = 0.126416 (* 1 = 0.126416 loss) I0412 14:24:49.614060 8032 sgd_solver.cpp:105] Iteration 9792, lr = 0.00143753 I0412 14:24:53.779126 8032 solver.cpp:218] Iteration 9804 (2.88119 iter/s, 4.16495s/12 iters), loss = 0.0271288 I0412 14:24:53.779191 8032 solver.cpp:237] Train net output #0: loss = 0.0271287 (* 1 = 0.0271287 loss) I0412 14:24:53.779206 8032 sgd_solver.cpp:105] Iteration 9804, lr = 0.00143412 I0412 14:24:56.777307 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:24:58.801709 8032 solver.cpp:218] Iteration 9816 (2.38931 iter/s, 5.02237s/12 iters), loss = 0.0441181 I0412 14:24:58.801862 8032 solver.cpp:237] Train net output #0: loss = 0.044118 (* 1 = 0.044118 loss) I0412 14:24:58.801873 8032 sgd_solver.cpp:105] Iteration 9816, lr = 0.00143072 I0412 14:25:03.821761 8032 solver.cpp:218] Iteration 9828 (2.39056 iter/s, 5.01975s/12 iters), loss = 0.031117 I0412 14:25:03.821818 8032 solver.cpp:237] Train net output #0: loss = 0.0311169 (* 1 = 0.0311169 loss) I0412 14:25:03.821830 8032 sgd_solver.cpp:105] Iteration 9828, lr = 0.00142732 I0412 14:25:09.134682 8032 solver.cpp:218] Iteration 9840 (2.25874 iter/s, 5.3127s/12 iters), loss = 0.0692747 I0412 14:25:09.134732 8032 solver.cpp:237] Train net output #0: loss = 0.0692747 (* 1 = 0.0692747 loss) I0412 14:25:09.134743 8032 sgd_solver.cpp:105] Iteration 9840, lr = 0.00142393 I0412 14:25:14.257208 8032 solver.cpp:218] Iteration 9852 (2.34269 iter/s, 5.12232s/12 iters), loss = 0.187117 I0412 14:25:14.257264 8032 solver.cpp:237] Train net output #0: loss = 0.187117 (* 1 = 0.187117 loss) I0412 14:25:14.257277 8032 sgd_solver.cpp:105] Iteration 9852, lr = 0.00142055 I0412 14:25:19.341421 8032 solver.cpp:218] Iteration 9864 (2.36034 iter/s, 5.08401s/12 iters), loss = 0.0485055 I0412 14:25:19.341468 8032 solver.cpp:237] Train net output #0: loss = 0.0485054 (* 1 = 0.0485054 loss) I0412 14:25:19.341480 8032 sgd_solver.cpp:105] Iteration 9864, lr = 0.00141718 I0412 14:25:24.276262 8032 solver.cpp:218] Iteration 9876 (2.43179 iter/s, 4.93464s/12 iters), loss = 0.116283 I0412 14:25:24.276324 8032 solver.cpp:237] Train net output #0: loss = 0.116283 (* 1 = 0.116283 loss) I0412 14:25:24.276340 8032 sgd_solver.cpp:105] Iteration 9876, lr = 0.00141381 I0412 14:25:29.376780 8032 solver.cpp:218] Iteration 9888 (2.3528 iter/s, 5.10031s/12 iters), loss = 0.0913596 I0412 14:25:29.376864 8032 solver.cpp:237] Train net output #0: loss = 0.0913595 (* 1 = 0.0913595 loss) I0412 14:25:29.376873 8032 sgd_solver.cpp:105] Iteration 9888, lr = 0.00141045 I0412 14:25:31.456961 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9894.caffemodel I0412 14:25:34.452499 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9894.solverstate I0412 14:25:36.719099 8032 solver.cpp:330] Iteration 9894, Testing net (#0) I0412 14:25:36.719121 8032 net.cpp:676] Ignoring source layer train-data I0412 14:25:37.347957 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:25:41.472829 8032 solver.cpp:397] Test net output #0: accuracy = 0.484681 I0412 14:25:41.472880 8032 solver.cpp:397] Test net output #1: loss = 2.94409 (* 1 = 2.94409 loss) I0412 14:25:43.223009 8032 solver.cpp:218] Iteration 9900 (0.866692 iter/s, 13.8458s/12 iters), loss = 0.21525 I0412 14:25:43.223064 8032 solver.cpp:237] Train net output #0: loss = 0.21525 (* 1 = 0.21525 loss) I0412 14:25:43.223076 8032 sgd_solver.cpp:105] Iteration 9900, lr = 0.00140711 I0412 14:25:48.256497 8032 solver.cpp:218] Iteration 9912 (2.38413 iter/s, 5.03328s/12 iters), loss = 0.127196 I0412 14:25:48.256541 8032 solver.cpp:237] Train net output #0: loss = 0.127196 (* 1 = 0.127196 loss) I0412 14:25:48.256549 8032 sgd_solver.cpp:105] Iteration 9912, lr = 0.00140377 I0412 14:25:48.354678 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:25:53.324400 8032 solver.cpp:218] Iteration 9924 (2.36794 iter/s, 5.0677s/12 iters), loss = 0.0472139 I0412 14:25:53.324458 8032 solver.cpp:237] Train net output #0: loss = 0.0472139 (* 1 = 0.0472139 loss) I0412 14:25:53.324471 8032 sgd_solver.cpp:105] Iteration 9924, lr = 0.00140043 I0412 14:25:58.269395 8032 solver.cpp:218] Iteration 9936 (2.42679 iter/s, 4.94479s/12 iters), loss = 0.0585075 I0412 14:25:58.269439 8032 solver.cpp:237] Train net output #0: loss = 0.0585074 (* 1 = 0.0585074 loss) I0412 14:25:58.269450 8032 sgd_solver.cpp:105] Iteration 9936, lr = 0.00139711 I0412 14:26:03.260586 8032 solver.cpp:218] Iteration 9948 (2.40433 iter/s, 4.991s/12 iters), loss = 0.136667 I0412 14:26:03.260741 8032 solver.cpp:237] Train net output #0: loss = 0.136667 (* 1 = 0.136667 loss) I0412 14:26:03.260754 8032 sgd_solver.cpp:105] Iteration 9948, lr = 0.00139379 I0412 14:26:08.544381 8032 solver.cpp:218] Iteration 9960 (2.27123 iter/s, 5.28348s/12 iters), loss = 0.114557 I0412 14:26:08.544433 8032 solver.cpp:237] Train net output #0: loss = 0.114557 (* 1 = 0.114557 loss) I0412 14:26:08.544445 8032 sgd_solver.cpp:105] Iteration 9960, lr = 0.00139048 I0412 14:26:13.818668 8032 solver.cpp:218] Iteration 9972 (2.27528 iter/s, 5.27408s/12 iters), loss = 0.0365471 I0412 14:26:13.818722 8032 solver.cpp:237] Train net output #0: loss = 0.0365471 (* 1 = 0.0365471 loss) I0412 14:26:13.818732 8032 sgd_solver.cpp:105] Iteration 9972, lr = 0.00138718 I0412 14:26:19.103731 8032 solver.cpp:218] Iteration 9984 (2.27064 iter/s, 5.28485s/12 iters), loss = 0.086662 I0412 14:26:19.103775 8032 solver.cpp:237] Train net output #0: loss = 0.086662 (* 1 = 0.086662 loss) I0412 14:26:19.103785 8032 sgd_solver.cpp:105] Iteration 9984, lr = 0.00138389 I0412 14:26:23.912631 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9996.caffemodel I0412 14:26:26.937786 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9996.solverstate I0412 14:26:29.323062 8032 solver.cpp:330] Iteration 9996, Testing net (#0) I0412 14:26:29.323089 8032 net.cpp:676] Ignoring source layer train-data I0412 14:26:29.835678 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:26:33.959442 8032 solver.cpp:397] Test net output #0: accuracy = 0.5 I0412 14:26:33.959506 8032 solver.cpp:397] Test net output #1: loss = 3.00734 (* 1 = 3.00734 loss) I0412 14:26:34.047878 8032 solver.cpp:218] Iteration 9996 (0.803015 iter/s, 14.9437s/12 iters), loss = 0.136005 I0412 14:26:34.047928 8032 solver.cpp:237] Train net output #0: loss = 0.136004 (* 1 = 0.136004 loss) I0412 14:26:34.047940 8032 sgd_solver.cpp:105] Iteration 9996, lr = 0.0013806 I0412 14:26:38.523937 8032 solver.cpp:218] Iteration 10008 (2.68104 iter/s, 4.47587s/12 iters), loss = 0.0475579 I0412 14:26:38.523989 8032 solver.cpp:237] Train net output #0: loss = 0.0475579 (* 1 = 0.0475579 loss) I0412 14:26:38.524000 8032 sgd_solver.cpp:105] Iteration 10008, lr = 0.00137732 I0412 14:26:40.944492 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:26:43.705631 8032 solver.cpp:218] Iteration 10020 (2.31594 iter/s, 5.18148s/12 iters), loss = 0.0423334 I0412 14:26:43.705686 8032 solver.cpp:237] Train net output #0: loss = 0.0423334 (* 1 = 0.0423334 loss) I0412 14:26:43.705698 8032 sgd_solver.cpp:105] Iteration 10020, lr = 0.00137405 I0412 14:26:48.873785 8032 solver.cpp:218] Iteration 10032 (2.32201 iter/s, 5.16794s/12 iters), loss = 0.0707297 I0412 14:26:48.873836 8032 solver.cpp:237] Train net output #0: loss = 0.0707297 (* 1 = 0.0707297 loss) I0412 14:26:48.873847 8032 sgd_solver.cpp:105] Iteration 10032, lr = 0.00137079 I0412 14:26:54.506557 8032 solver.cpp:218] Iteration 10044 (2.13047 iter/s, 5.63255s/12 iters), loss = 0.0377946 I0412 14:26:54.506605 8032 solver.cpp:237] Train net output #0: loss = 0.0377946 (* 1 = 0.0377946 loss) I0412 14:26:54.506615 8032 sgd_solver.cpp:105] Iteration 10044, lr = 0.00136754 I0412 14:26:59.569088 8032 solver.cpp:218] Iteration 10056 (2.37045 iter/s, 5.06232s/12 iters), loss = 0.0826398 I0412 14:26:59.569141 8032 solver.cpp:237] Train net output #0: loss = 0.0826398 (* 1 = 0.0826398 loss) I0412 14:26:59.569154 8032 sgd_solver.cpp:105] Iteration 10056, lr = 0.00136429 I0412 14:27:04.685366 8032 solver.cpp:218] Iteration 10068 (2.34555 iter/s, 5.11607s/12 iters), loss = 0.0869745 I0412 14:27:04.685483 8032 solver.cpp:237] Train net output #0: loss = 0.0869745 (* 1 = 0.0869745 loss) I0412 14:27:04.685493 8032 sgd_solver.cpp:105] Iteration 10068, lr = 0.00136105 I0412 14:27:10.108362 8032 solver.cpp:218] Iteration 10080 (2.21292 iter/s, 5.42271s/12 iters), loss = 0.049313 I0412 14:27:10.108417 8032 solver.cpp:237] Train net output #0: loss = 0.049313 (* 1 = 0.049313 loss) I0412 14:27:10.108428 8032 sgd_solver.cpp:105] Iteration 10080, lr = 0.00135782 I0412 14:27:15.573760 8032 solver.cpp:218] Iteration 10092 (2.19572 iter/s, 5.46518s/12 iters), loss = 0.0792774 I0412 14:27:15.573813 8032 solver.cpp:237] Train net output #0: loss = 0.0792774 (* 1 = 0.0792774 loss) I0412 14:27:15.573825 8032 sgd_solver.cpp:105] Iteration 10092, lr = 0.0013546 I0412 14:27:17.851361 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_10098.caffemodel I0412 14:27:23.722730 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_10098.solverstate I0412 14:27:27.610383 8032 solver.cpp:330] Iteration 10098, Testing net (#0) I0412 14:27:27.610414 8032 net.cpp:676] Ignoring source layer train-data I0412 14:27:28.093248 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:27:32.231264 8032 solver.cpp:397] Test net output #0: accuracy = 0.49326 I0412 14:27:32.231298 8032 solver.cpp:397] Test net output #1: loss = 2.93962 (* 1 = 2.93962 loss) I0412 14:27:34.268633 8032 solver.cpp:218] Iteration 10104 (0.641907 iter/s, 18.6943s/12 iters), loss = 0.0912263 I0412 14:27:34.268688 8032 solver.cpp:237] Train net output #0: loss = 0.0912263 (* 1 = 0.0912263 loss) I0412 14:27:34.268699 8032 sgd_solver.cpp:105] Iteration 10104, lr = 0.00135138 I0412 14:27:38.636579 8036 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:27:39.279255 8032 solver.cpp:218] Iteration 10116 (2.39501 iter/s, 5.01041s/12 iters), loss = 0.211717 I0412 14:27:39.279297 8032 solver.cpp:237] Train net output #0: loss = 0.211717 (* 1 = 0.211717 loss) I0412 14:27:39.279306 8032 sgd_solver.cpp:105] Iteration 10116, lr = 0.00134817 I0412 14:27:44.434705 8032 solver.cpp:218] Iteration 10128 (2.32773 iter/s, 5.15524s/12 iters), loss = 0.186586 I0412 14:27:44.434756 8032 solver.cpp:237] Train net output #0: loss = 0.186586 (* 1 = 0.186586 loss) I0412 14:27:44.434767 8032 sgd_solver.cpp:105] Iteration 10128, lr = 0.00134497 I0412 14:27:49.662498 8032 solver.cpp:218] Iteration 10140 (2.29551 iter/s, 5.22759s/12 iters), loss = 0.0665451 I0412 14:27:49.662537 8032 solver.cpp:237] Train net output #0: loss = 0.0665451 (* 1 = 0.0665451 loss) I0412 14:27:49.662546 8032 sgd_solver.cpp:105] Iteration 10140, lr = 0.00134178 I0412 14:27:54.853602 8032 solver.cpp:218] Iteration 10152 (2.31174 iter/s, 5.1909s/12 iters), loss = 0.0725212 I0412 14:27:54.853644 8032 solver.cpp:237] Train net output #0: loss = 0.0725212 (* 1 = 0.0725212 loss) I0412 14:27:54.853652 8032 sgd_solver.cpp:105] Iteration 10152, lr = 0.00133859 I0412 14:28:00.181365 8032 solver.cpp:218] Iteration 10164 (2.25244 iter/s, 5.32755s/12 iters), loss = 0.0217107 I0412 14:28:00.181419 8032 solver.cpp:237] Train net output #0: loss = 0.0217107 (* 1 = 0.0217107 loss) I0412 14:28:00.181434 8032 sgd_solver.cpp:105] Iteration 10164, lr = 0.00133541 I0412 14:28:05.566229 8032 solver.cpp:218] Iteration 10176 (2.22856 iter/s, 5.38464s/12 iters), loss = 0.0435951 I0412 14:28:05.566274 8032 solver.cpp:237] Train net output #0: loss = 0.0435951 (* 1 = 0.0435951 loss) I0412 14:28:05.566282 8032 sgd_solver.cpp:105] Iteration 10176, lr = 0.00133224 I0412 14:28:10.735625 8032 solver.cpp:218] Iteration 10188 (2.32145 iter/s, 5.16919s/12 iters), loss = 0.0166095 I0412 14:28:10.735787 8032 solver.cpp:237] Train net output #0: loss = 0.0166095 (* 1 = 0.0166095 loss) I0412 14:28:10.735800 8032 sgd_solver.cpp:105] Iteration 10188, lr = 0.00132908 I0412 14:28:15.495311 8032 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_10200.caffemodel I0412 14:28:18.498514 8032 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_10200.solverstate I0412 14:28:20.846794 8032 solver.cpp:310] Iteration 10200, loss = 0.0776016 I0412 14:28:20.846822 8032 solver.cpp:330] Iteration 10200, Testing net (#0) I0412 14:28:20.846827 8032 net.cpp:676] Ignoring source layer train-data I0412 14:28:21.330977 8037 data_layer.cpp:73] Restarting data prefetching from start. I0412 14:28:25.585819 8032 solver.cpp:397] Test net output #0: accuracy = 0.497549 I0412 14:28:25.585871 8032 solver.cpp:397] Test net output #1: loss = 3.06837 (* 1 = 3.06837 loss) I0412 14:28:25.585884 8032 solver.cpp:315] Optimization Done. I0412 14:28:25.585891 8032 caffe.cpp:259] Optimization Done.