I0406 07:07:41.118609 5226 upgrade_proto.cpp:1082] Attempting to upgrade input file specified using deprecated 'solver_type' field (enum)': /mnt/bigdisk/DIGITS-AIN-3/digits/jobs/20210406-070739-2500/solver.prototxt I0406 07:07:41.118760 5226 upgrade_proto.cpp:1089] Successfully upgraded file specified using deprecated 'solver_type' field (enum) to 'type' field (string). W0406 07:07:41.118765 5226 upgrade_proto.cpp:1091] Note that future Caffe releases will only support 'type' field (string) for a solver's type. I0406 07:07:41.118825 5226 caffe.cpp:218] Using GPUs 0 I0406 07:07:41.143887 5226 caffe.cpp:223] GPU 0: GeForce GTX TITAN X I0406 07:07:41.395608 5226 solver.cpp:44] Initializing solver from parameters: test_iter: 51 test_interval: 102 base_lr: 0.01 display: 12 max_iter: 20400 lr_policy: "fixed" momentum: 0.9 weight_decay: 0.0001 snapshot: 102 snapshot_prefix: "snapshot" solver_mode: GPU device_id: 0 net: "train_val.prototxt" train_state { level: 0 stage: "" } type: "SGD" I0406 07:07:41.396505 5226 solver.cpp:87] Creating training net from net file: train_val.prototxt I0406 07:07:41.397136 5226 net.cpp:294] The NetState phase (0) differed from the phase (1) specified by a rule in layer val-data I0406 07:07:41.397150 5226 net.cpp:294] The NetState phase (0) differed from the phase (1) specified by a rule in layer accuracy I0406 07:07:41.397270 5226 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-AIN-3/digits/jobs/20210401-115716-aaf7/mean.binaryproto" } data_param { source: "/mnt/bigdisk/DIGITS-AIN-3/digits/jobs/20210401-115716-aaf7/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: 384 pad: 1 kernel_size: 3 group: 2 weight_filler { type: "gaussian" std: 0.01 } bias_filler { type: "constant" value: 0.1 } } } layer { name: "relu4" type: "ReLU" bottom: "conv4" top: "conv4" } layer { name: "conv5" type: "Convolution" bottom: "conv4" top: "conv5" param { lr_mult: 1 decay_mult: 1 } param { lr_mult: 2 decay_mult: 0 } convolution_param { num_output: 256 pad: 1 kernel_size: 3 group: 2 weight_filler { type: "gaussian" std: 0.01 } bias_filler { type: "constant" value: 0.1 } } } layer { name: "relu5" type: "ReLU" bottom: "conv5" top: "conv5" } layer { name: "pool5" type: "Pooling" bottom: "conv5" top: "pool5" pooling_param { pool: MAX kernel_size: 3 stride: 2 } } layer { name: "fc6" type: "InnerProduct" bottom: "pool5" top: "fc6" param { lr_mult: 1 decay_mult: 1 } param { lr_mult: 2 decay_mult: 0 } inner_product_param { num_output: 4096 weight_filler { type: "gaussian" std: 0.005 } bias_filler { type: "constant" value: 0.1 } } } layer { name: "relu6" type: "ReLU" bottom: "fc6" top: "fc6" } layer { name: "drop6" type: "Dropout" bottom: "fc6" top: "fc6" dropout_param { dropout_ratio: 0.5 } } layer { name: "fc7" type: "InnerProduct" bottom: "fc6" top: "fc7" param { lr_mult: 1 decay_mult: 1 } param { lr_mult: 2 decay_mult: 0 } inner_product_param { num_output: 4096 weight_filler { type: "gaussian" std: 0.005 } bias_filler { type: "constant" value: 0.1 } } } layer { name: "relu7" type: "ReLU" bottom: "fc7" top: "fc7" } layer { name: "drop7" type: "Dropout" bottom: "fc7" top: "fc7" dropout_param { dropout_ratio: 0.5 } } layer { name: "fc8" type: "InnerProduct" bottom: "fc7" top: "fc8" param { lr_mult: 1 decay_mult: 1 } param { lr_mult: 2 decay_mult: 0 } inner_product_param { num_output: 196 weight_filler { type: "gaussian" std: 0.01 } bias_filler { type: "constant" value: 0 } } } layer { name: "loss" type: "SoftmaxWithLoss" bottom: "fc8" bottom: "label" top: "loss" } I0406 07:07:41.397347 5226 layer_factory.hpp:77] Creating layer train-data I0406 07:07:41.421499 5226 db_lmdb.cpp:35] Opened lmdb /mnt/bigdisk/DIGITS-AIN-3/digits/jobs/20210401-115716-aaf7/train_db I0406 07:07:41.421733 5226 net.cpp:84] Creating Layer train-data I0406 07:07:41.421748 5226 net.cpp:380] train-data -> data I0406 07:07:41.421768 5226 net.cpp:380] train-data -> label I0406 07:07:41.421779 5226 data_transformer.cpp:25] Loading mean file from: /mnt/bigdisk/DIGITS-AIN-3/digits/jobs/20210401-115716-aaf7/mean.binaryproto I0406 07:07:41.462285 5226 data_layer.cpp:45] output data size: 128,3,227,227 I0406 07:07:41.609122 5226 net.cpp:122] Setting up train-data I0406 07:07:41.609144 5226 net.cpp:129] Top shape: 128 3 227 227 (19787136) I0406 07:07:41.609148 5226 net.cpp:129] Top shape: 128 (128) I0406 07:07:41.609151 5226 net.cpp:137] Memory required for data: 79149056 I0406 07:07:41.609158 5226 layer_factory.hpp:77] Creating layer conv1 I0406 07:07:41.609176 5226 net.cpp:84] Creating Layer conv1 I0406 07:07:41.609181 5226 net.cpp:406] conv1 <- data I0406 07:07:41.609192 5226 net.cpp:380] conv1 -> conv1 I0406 07:07:42.050434 5226 net.cpp:122] Setting up conv1 I0406 07:07:42.050458 5226 net.cpp:129] Top shape: 128 96 55 55 (37171200) I0406 07:07:42.050462 5226 net.cpp:137] Memory required for data: 227833856 I0406 07:07:42.050479 5226 layer_factory.hpp:77] Creating layer relu1 I0406 07:07:42.050489 5226 net.cpp:84] Creating Layer relu1 I0406 07:07:42.050493 5226 net.cpp:406] relu1 <- conv1 I0406 07:07:42.050496 5226 net.cpp:367] relu1 -> conv1 (in-place) I0406 07:07:42.050762 5226 net.cpp:122] Setting up relu1 I0406 07:07:42.050770 5226 net.cpp:129] Top shape: 128 96 55 55 (37171200) I0406 07:07:42.050772 5226 net.cpp:137] Memory required for data: 376518656 I0406 07:07:42.050776 5226 layer_factory.hpp:77] Creating layer norm1 I0406 07:07:42.050783 5226 net.cpp:84] Creating Layer norm1 I0406 07:07:42.050786 5226 net.cpp:406] norm1 <- conv1 I0406 07:07:42.050817 5226 net.cpp:380] norm1 -> norm1 I0406 07:07:42.051363 5226 net.cpp:122] Setting up norm1 I0406 07:07:42.051373 5226 net.cpp:129] Top shape: 128 96 55 55 (37171200) I0406 07:07:42.051375 5226 net.cpp:137] Memory required for data: 525203456 I0406 07:07:42.051378 5226 layer_factory.hpp:77] Creating layer pool1 I0406 07:07:42.051384 5226 net.cpp:84] Creating Layer pool1 I0406 07:07:42.051388 5226 net.cpp:406] pool1 <- norm1 I0406 07:07:42.051391 5226 net.cpp:380] pool1 -> pool1 I0406 07:07:42.051424 5226 net.cpp:122] Setting up pool1 I0406 07:07:42.051429 5226 net.cpp:129] Top shape: 128 96 27 27 (8957952) I0406 07:07:42.051431 5226 net.cpp:137] Memory required for data: 561035264 I0406 07:07:42.051434 5226 layer_factory.hpp:77] Creating layer conv2 I0406 07:07:42.051443 5226 net.cpp:84] Creating Layer conv2 I0406 07:07:42.051445 5226 net.cpp:406] conv2 <- pool1 I0406 07:07:42.051450 5226 net.cpp:380] conv2 -> conv2 I0406 07:07:42.058259 5226 net.cpp:122] Setting up conv2 I0406 07:07:42.058277 5226 net.cpp:129] Top shape: 128 256 27 27 (23887872) I0406 07:07:42.058280 5226 net.cpp:137] Memory required for data: 656586752 I0406 07:07:42.058290 5226 layer_factory.hpp:77] Creating layer relu2 I0406 07:07:42.058297 5226 net.cpp:84] Creating Layer relu2 I0406 07:07:42.058301 5226 net.cpp:406] relu2 <- conv2 I0406 07:07:42.058305 5226 net.cpp:367] relu2 -> conv2 (in-place) I0406 07:07:42.058734 5226 net.cpp:122] Setting up relu2 I0406 07:07:42.058743 5226 net.cpp:129] Top shape: 128 256 27 27 (23887872) I0406 07:07:42.058745 5226 net.cpp:137] Memory required for data: 752138240 I0406 07:07:42.058748 5226 layer_factory.hpp:77] Creating layer norm2 I0406 07:07:42.058755 5226 net.cpp:84] Creating Layer norm2 I0406 07:07:42.058758 5226 net.cpp:406] norm2 <- conv2 I0406 07:07:42.058763 5226 net.cpp:380] norm2 -> norm2 I0406 07:07:42.059024 5226 net.cpp:122] Setting up norm2 I0406 07:07:42.059032 5226 net.cpp:129] Top shape: 128 256 27 27 (23887872) I0406 07:07:42.059034 5226 net.cpp:137] Memory required for data: 847689728 I0406 07:07:42.059037 5226 layer_factory.hpp:77] Creating layer pool2 I0406 07:07:42.059044 5226 net.cpp:84] Creating Layer pool2 I0406 07:07:42.059046 5226 net.cpp:406] pool2 <- norm2 I0406 07:07:42.059051 5226 net.cpp:380] pool2 -> pool2 I0406 07:07:42.059075 5226 net.cpp:122] Setting up pool2 I0406 07:07:42.059079 5226 net.cpp:129] Top shape: 128 256 13 13 (5537792) I0406 07:07:42.059082 5226 net.cpp:137] Memory required for data: 869840896 I0406 07:07:42.059083 5226 layer_factory.hpp:77] Creating layer conv3 I0406 07:07:42.059092 5226 net.cpp:84] Creating Layer conv3 I0406 07:07:42.059094 5226 net.cpp:406] conv3 <- pool2 I0406 07:07:42.059098 5226 net.cpp:380] conv3 -> conv3 I0406 07:07:42.068939 5226 net.cpp:122] Setting up conv3 I0406 07:07:42.068958 5226 net.cpp:129] Top shape: 128 384 13 13 (8306688) I0406 07:07:42.068959 5226 net.cpp:137] Memory required for data: 903067648 I0406 07:07:42.068970 5226 layer_factory.hpp:77] Creating layer relu3 I0406 07:07:42.068979 5226 net.cpp:84] Creating Layer relu3 I0406 07:07:42.068981 5226 net.cpp:406] relu3 <- conv3 I0406 07:07:42.068986 5226 net.cpp:367] relu3 -> conv3 (in-place) I0406 07:07:42.069411 5226 net.cpp:122] Setting up relu3 I0406 07:07:42.069420 5226 net.cpp:129] Top shape: 128 384 13 13 (8306688) I0406 07:07:42.069423 5226 net.cpp:137] Memory required for data: 936294400 I0406 07:07:42.069425 5226 layer_factory.hpp:77] Creating layer conv4 I0406 07:07:42.069435 5226 net.cpp:84] Creating Layer conv4 I0406 07:07:42.069437 5226 net.cpp:406] conv4 <- conv3 I0406 07:07:42.069442 5226 net.cpp:380] conv4 -> conv4 I0406 07:07:42.078717 5226 net.cpp:122] Setting up conv4 I0406 07:07:42.078732 5226 net.cpp:129] Top shape: 128 384 13 13 (8306688) I0406 07:07:42.078735 5226 net.cpp:137] Memory required for data: 969521152 I0406 07:07:42.078742 5226 layer_factory.hpp:77] Creating layer relu4 I0406 07:07:42.078750 5226 net.cpp:84] Creating Layer relu4 I0406 07:07:42.078753 5226 net.cpp:406] relu4 <- conv4 I0406 07:07:42.078776 5226 net.cpp:367] relu4 -> conv4 (in-place) I0406 07:07:42.079092 5226 net.cpp:122] Setting up relu4 I0406 07:07:42.079100 5226 net.cpp:129] Top shape: 128 384 13 13 (8306688) I0406 07:07:42.079102 5226 net.cpp:137] Memory required for data: 1002747904 I0406 07:07:42.079105 5226 layer_factory.hpp:77] Creating layer conv5 I0406 07:07:42.079115 5226 net.cpp:84] Creating Layer conv5 I0406 07:07:42.079118 5226 net.cpp:406] conv5 <- conv4 I0406 07:07:42.079123 5226 net.cpp:380] conv5 -> conv5 I0406 07:07:42.086643 5226 net.cpp:122] Setting up conv5 I0406 07:07:42.086659 5226 net.cpp:129] Top shape: 128 256 13 13 (5537792) I0406 07:07:42.086663 5226 net.cpp:137] Memory required for data: 1024899072 I0406 07:07:42.086673 5226 layer_factory.hpp:77] Creating layer relu5 I0406 07:07:42.086683 5226 net.cpp:84] Creating Layer relu5 I0406 07:07:42.086685 5226 net.cpp:406] relu5 <- conv5 I0406 07:07:42.086691 5226 net.cpp:367] relu5 -> conv5 (in-place) I0406 07:07:42.087175 5226 net.cpp:122] Setting up relu5 I0406 07:07:42.087183 5226 net.cpp:129] Top shape: 128 256 13 13 (5537792) I0406 07:07:42.087186 5226 net.cpp:137] Memory required for data: 1047050240 I0406 07:07:42.087188 5226 layer_factory.hpp:77] Creating layer pool5 I0406 07:07:42.087194 5226 net.cpp:84] Creating Layer pool5 I0406 07:07:42.087196 5226 net.cpp:406] pool5 <- conv5 I0406 07:07:42.087203 5226 net.cpp:380] pool5 -> pool5 I0406 07:07:42.087235 5226 net.cpp:122] Setting up pool5 I0406 07:07:42.087239 5226 net.cpp:129] Top shape: 128 256 6 6 (1179648) I0406 07:07:42.087241 5226 net.cpp:137] Memory required for data: 1051768832 I0406 07:07:42.087244 5226 layer_factory.hpp:77] Creating layer fc6 I0406 07:07:42.087253 5226 net.cpp:84] Creating Layer fc6 I0406 07:07:42.087255 5226 net.cpp:406] fc6 <- pool5 I0406 07:07:42.087260 5226 net.cpp:380] fc6 -> fc6 I0406 07:07:42.420754 5226 net.cpp:122] Setting up fc6 I0406 07:07:42.420773 5226 net.cpp:129] Top shape: 128 4096 (524288) I0406 07:07:42.420775 5226 net.cpp:137] Memory required for data: 1053865984 I0406 07:07:42.420783 5226 layer_factory.hpp:77] Creating layer relu6 I0406 07:07:42.420792 5226 net.cpp:84] Creating Layer relu6 I0406 07:07:42.420795 5226 net.cpp:406] relu6 <- fc6 I0406 07:07:42.420800 5226 net.cpp:367] relu6 -> fc6 (in-place) I0406 07:07:42.421501 5226 net.cpp:122] Setting up relu6 I0406 07:07:42.421511 5226 net.cpp:129] Top shape: 128 4096 (524288) I0406 07:07:42.421514 5226 net.cpp:137] Memory required for data: 1055963136 I0406 07:07:42.421515 5226 layer_factory.hpp:77] Creating layer drop6 I0406 07:07:42.421521 5226 net.cpp:84] Creating Layer drop6 I0406 07:07:42.421525 5226 net.cpp:406] drop6 <- fc6 I0406 07:07:42.421530 5226 net.cpp:367] drop6 -> fc6 (in-place) I0406 07:07:42.421553 5226 net.cpp:122] Setting up drop6 I0406 07:07:42.421557 5226 net.cpp:129] Top shape: 128 4096 (524288) I0406 07:07:42.421559 5226 net.cpp:137] Memory required for data: 1058060288 I0406 07:07:42.421561 5226 layer_factory.hpp:77] Creating layer fc7 I0406 07:07:42.421572 5226 net.cpp:84] Creating Layer fc7 I0406 07:07:42.421574 5226 net.cpp:406] fc7 <- fc6 I0406 07:07:42.421579 5226 net.cpp:380] fc7 -> fc7 I0406 07:07:42.570488 5226 net.cpp:122] Setting up fc7 I0406 07:07:42.570508 5226 net.cpp:129] Top shape: 128 4096 (524288) I0406 07:07:42.570510 5226 net.cpp:137] Memory required for data: 1060157440 I0406 07:07:42.570518 5226 layer_factory.hpp:77] Creating layer relu7 I0406 07:07:42.570526 5226 net.cpp:84] Creating Layer relu7 I0406 07:07:42.570529 5226 net.cpp:406] relu7 <- fc7 I0406 07:07:42.570534 5226 net.cpp:367] relu7 -> fc7 (in-place) I0406 07:07:42.570909 5226 net.cpp:122] Setting up relu7 I0406 07:07:42.570919 5226 net.cpp:129] Top shape: 128 4096 (524288) I0406 07:07:42.570920 5226 net.cpp:137] Memory required for data: 1062254592 I0406 07:07:42.570922 5226 layer_factory.hpp:77] Creating layer drop7 I0406 07:07:42.570928 5226 net.cpp:84] Creating Layer drop7 I0406 07:07:42.570930 5226 net.cpp:406] drop7 <- fc7 I0406 07:07:42.570952 5226 net.cpp:367] drop7 -> fc7 (in-place) I0406 07:07:42.570974 5226 net.cpp:122] Setting up drop7 I0406 07:07:42.570978 5226 net.cpp:129] Top shape: 128 4096 (524288) I0406 07:07:42.570981 5226 net.cpp:137] Memory required for data: 1064351744 I0406 07:07:42.570982 5226 layer_factory.hpp:77] Creating layer fc8 I0406 07:07:42.570989 5226 net.cpp:84] Creating Layer fc8 I0406 07:07:42.570991 5226 net.cpp:406] fc8 <- fc7 I0406 07:07:42.570995 5226 net.cpp:380] fc8 -> fc8 I0406 07:07:42.578217 5226 net.cpp:122] Setting up fc8 I0406 07:07:42.578231 5226 net.cpp:129] Top shape: 128 196 (25088) I0406 07:07:42.578234 5226 net.cpp:137] Memory required for data: 1064452096 I0406 07:07:42.578241 5226 layer_factory.hpp:77] Creating layer loss I0406 07:07:42.578248 5226 net.cpp:84] Creating Layer loss I0406 07:07:42.578250 5226 net.cpp:406] loss <- fc8 I0406 07:07:42.578255 5226 net.cpp:406] loss <- label I0406 07:07:42.578261 5226 net.cpp:380] loss -> loss I0406 07:07:42.578270 5226 layer_factory.hpp:77] Creating layer loss I0406 07:07:42.578953 5226 net.cpp:122] Setting up loss I0406 07:07:42.578960 5226 net.cpp:129] Top shape: (1) I0406 07:07:42.578963 5226 net.cpp:132] with loss weight 1 I0406 07:07:42.578985 5226 net.cpp:137] Memory required for data: 1064452100 I0406 07:07:42.578989 5226 net.cpp:198] loss needs backward computation. I0406 07:07:42.578994 5226 net.cpp:198] fc8 needs backward computation. I0406 07:07:42.578996 5226 net.cpp:198] drop7 needs backward computation. I0406 07:07:42.578999 5226 net.cpp:198] relu7 needs backward computation. I0406 07:07:42.579000 5226 net.cpp:198] fc7 needs backward computation. I0406 07:07:42.579003 5226 net.cpp:198] drop6 needs backward computation. I0406 07:07:42.579005 5226 net.cpp:198] relu6 needs backward computation. I0406 07:07:42.579008 5226 net.cpp:198] fc6 needs backward computation. I0406 07:07:42.579011 5226 net.cpp:198] pool5 needs backward computation. I0406 07:07:42.579013 5226 net.cpp:198] relu5 needs backward computation. I0406 07:07:42.579015 5226 net.cpp:198] conv5 needs backward computation. I0406 07:07:42.579018 5226 net.cpp:198] relu4 needs backward computation. I0406 07:07:42.579020 5226 net.cpp:198] conv4 needs backward computation. I0406 07:07:42.579023 5226 net.cpp:198] relu3 needs backward computation. I0406 07:07:42.579025 5226 net.cpp:198] conv3 needs backward computation. I0406 07:07:42.579027 5226 net.cpp:198] pool2 needs backward computation. I0406 07:07:42.579030 5226 net.cpp:198] norm2 needs backward computation. I0406 07:07:42.579032 5226 net.cpp:198] relu2 needs backward computation. I0406 07:07:42.579035 5226 net.cpp:198] conv2 needs backward computation. I0406 07:07:42.579037 5226 net.cpp:198] pool1 needs backward computation. I0406 07:07:42.579039 5226 net.cpp:198] norm1 needs backward computation. I0406 07:07:42.579042 5226 net.cpp:198] relu1 needs backward computation. I0406 07:07:42.579044 5226 net.cpp:198] conv1 needs backward computation. I0406 07:07:42.579047 5226 net.cpp:200] train-data does not need backward computation. I0406 07:07:42.579049 5226 net.cpp:242] This network produces output loss I0406 07:07:42.579061 5226 net.cpp:255] Network initialization done. I0406 07:07:42.579543 5226 solver.cpp:172] Creating test net (#0) specified by net file: train_val.prototxt I0406 07:07:42.579573 5226 net.cpp:294] The NetState phase (1) differed from the phase (0) specified by a rule in layer train-data I0406 07:07:42.579708 5226 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-AIN-3/digits/jobs/20210401-115716-aaf7/mean.binaryproto" } data_param { source: "/mnt/bigdisk/DIGITS-AIN-3/digits/jobs/20210401-115716-aaf7/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: 384 pad: 1 kernel_size: 3 group: 2 weight_filler { type: "gaussian" std: 0.01 } bias_filler { type: "constant" value: 0.1 } } } layer { name: "relu4" type: "ReLU" bottom: "conv4" top: "conv4" } layer { name: "conv5" type: "Convolution" bottom: "conv4" top: "conv5" param { lr_mult: 1 decay_mult: 1 } param { lr_mult: 2 decay_mult: 0 } convolution_param { num_output: 256 pad: 1 kernel_size: 3 group: 2 weight_filler { type: "gaussian" std: 0.01 } bias_filler { type: "constant" value: 0.1 } } } layer { name: "relu5" type: "ReLU" bottom: "conv5" top: "conv5" } layer { name: "pool5" type: "Pooling" bottom: "conv5" top: "pool5" pooling_param { pool: MAX kernel_size: 3 stride: 2 } } layer { name: "fc6" type: "InnerProduct" bottom: "pool5" top: "fc6" param { lr_mult: 1 decay_mult: 1 } param { lr_mult: 2 decay_mult: 0 } inner_product_param { num_output: 4096 weight_filler { type: "gaussian" std: 0.005 } bias_filler { type: "constant" value: 0.1 } } } layer { name: "relu6" type: "ReLU" bottom: "fc6" top: "fc6" } layer { name: "drop6" type: "Dropout" bottom: "fc6" top: "fc6" dropout_param { dropout_ratio: 0.5 } } layer { name: "fc7" type: "InnerProduct" bottom: "fc6" top: "fc7" param { lr_mult: 1 decay_mult: 1 } param { lr_mult: 2 decay_mult: 0 } inner_product_param { num_output: 4096 weight_filler { type: "gaussian" std: 0.005 } bias_filler { type: "constant" value: 0.1 } } } layer { name: "relu7" type: "ReLU" bottom: "fc7" top: "fc7" } layer { name: "drop7" type: "Dropout" bottom: "fc7" top: "fc7" dropout_param { dropout_ratio: 0.5 } } layer { name: "fc8" type: "InnerProduct" bottom: "fc7" top: "fc8" param { lr_mult: 1 decay_mult: 1 } param { lr_mult: 2 decay_mult: 0 } inner_product_param { num_output: 196 weight_filler { type: "gaussian" std: 0.01 } bias_filler { type: "constant" value: 0 } } } layer { name: "accuracy" type: "Accuracy" bottom: "fc8" bottom: "label" top: "accuracy" include { phase: TEST } } layer { name: "loss" type: "SoftmaxWithLoss" bottom: "fc8" bottom: "label" top: "loss" } I0406 07:07:42.579808 5226 layer_factory.hpp:77] Creating layer val-data I0406 07:07:42.582653 5226 db_lmdb.cpp:35] Opened lmdb /mnt/bigdisk/DIGITS-AIN-3/digits/jobs/20210401-115716-aaf7/val_db I0406 07:07:42.582876 5226 net.cpp:84] Creating Layer val-data I0406 07:07:42.582885 5226 net.cpp:380] val-data -> data I0406 07:07:42.582892 5226 net.cpp:380] val-data -> label I0406 07:07:42.582899 5226 data_transformer.cpp:25] Loading mean file from: /mnt/bigdisk/DIGITS-AIN-3/digits/jobs/20210401-115716-aaf7/mean.binaryproto I0406 07:07:42.586594 5226 data_layer.cpp:45] output data size: 32,3,227,227 I0406 07:07:42.623560 5226 net.cpp:122] Setting up val-data I0406 07:07:42.623580 5226 net.cpp:129] Top shape: 32 3 227 227 (4946784) I0406 07:07:42.623584 5226 net.cpp:129] Top shape: 32 (32) I0406 07:07:42.623585 5226 net.cpp:137] Memory required for data: 19787264 I0406 07:07:42.623590 5226 layer_factory.hpp:77] Creating layer label_val-data_1_split I0406 07:07:42.623601 5226 net.cpp:84] Creating Layer label_val-data_1_split I0406 07:07:42.623605 5226 net.cpp:406] label_val-data_1_split <- label I0406 07:07:42.623610 5226 net.cpp:380] label_val-data_1_split -> label_val-data_1_split_0 I0406 07:07:42.623618 5226 net.cpp:380] label_val-data_1_split -> label_val-data_1_split_1 I0406 07:07:42.623661 5226 net.cpp:122] Setting up label_val-data_1_split I0406 07:07:42.623664 5226 net.cpp:129] Top shape: 32 (32) I0406 07:07:42.623667 5226 net.cpp:129] Top shape: 32 (32) I0406 07:07:42.623668 5226 net.cpp:137] Memory required for data: 19787520 I0406 07:07:42.623672 5226 layer_factory.hpp:77] Creating layer conv1 I0406 07:07:42.623682 5226 net.cpp:84] Creating Layer conv1 I0406 07:07:42.623683 5226 net.cpp:406] conv1 <- data I0406 07:07:42.623687 5226 net.cpp:380] conv1 -> conv1 I0406 07:07:42.632581 5226 net.cpp:122] Setting up conv1 I0406 07:07:42.632593 5226 net.cpp:129] Top shape: 32 96 55 55 (9292800) I0406 07:07:42.632596 5226 net.cpp:137] Memory required for data: 56958720 I0406 07:07:42.632604 5226 layer_factory.hpp:77] Creating layer relu1 I0406 07:07:42.632611 5226 net.cpp:84] Creating Layer relu1 I0406 07:07:42.632613 5226 net.cpp:406] relu1 <- conv1 I0406 07:07:42.632617 5226 net.cpp:367] relu1 -> conv1 (in-place) I0406 07:07:42.632874 5226 net.cpp:122] Setting up relu1 I0406 07:07:42.632889 5226 net.cpp:129] Top shape: 32 96 55 55 (9292800) I0406 07:07:42.632891 5226 net.cpp:137] Memory required for data: 94129920 I0406 07:07:42.632894 5226 layer_factory.hpp:77] Creating layer norm1 I0406 07:07:42.632901 5226 net.cpp:84] Creating Layer norm1 I0406 07:07:42.632903 5226 net.cpp:406] norm1 <- conv1 I0406 07:07:42.632908 5226 net.cpp:380] norm1 -> norm1 I0406 07:07:42.639832 5226 net.cpp:122] Setting up norm1 I0406 07:07:42.639847 5226 net.cpp:129] Top shape: 32 96 55 55 (9292800) I0406 07:07:42.639851 5226 net.cpp:137] Memory required for data: 131301120 I0406 07:07:42.639856 5226 layer_factory.hpp:77] Creating layer pool1 I0406 07:07:42.639864 5226 net.cpp:84] Creating Layer pool1 I0406 07:07:42.639869 5226 net.cpp:406] pool1 <- norm1 I0406 07:07:42.639875 5226 net.cpp:380] pool1 -> pool1 I0406 07:07:42.639914 5226 net.cpp:122] Setting up pool1 I0406 07:07:42.639920 5226 net.cpp:129] Top shape: 32 96 27 27 (2239488) I0406 07:07:42.639923 5226 net.cpp:137] Memory required for data: 140259072 I0406 07:07:42.639927 5226 layer_factory.hpp:77] Creating layer conv2 I0406 07:07:42.639938 5226 net.cpp:84] Creating Layer conv2 I0406 07:07:42.639942 5226 net.cpp:406] conv2 <- pool1 I0406 07:07:42.639977 5226 net.cpp:380] conv2 -> conv2 I0406 07:07:42.652181 5226 net.cpp:122] Setting up conv2 I0406 07:07:42.652201 5226 net.cpp:129] Top shape: 32 256 27 27 (5971968) I0406 07:07:42.652205 5226 net.cpp:137] Memory required for data: 164146944 I0406 07:07:42.652220 5226 layer_factory.hpp:77] Creating layer relu2 I0406 07:07:42.652233 5226 net.cpp:84] Creating Layer relu2 I0406 07:07:42.652238 5226 net.cpp:406] relu2 <- conv2 I0406 07:07:42.652245 5226 net.cpp:367] relu2 -> conv2 (in-place) I0406 07:07:42.652994 5226 net.cpp:122] Setting up relu2 I0406 07:07:42.653007 5226 net.cpp:129] Top shape: 32 256 27 27 (5971968) I0406 07:07:42.653010 5226 net.cpp:137] Memory required for data: 188034816 I0406 07:07:42.653014 5226 layer_factory.hpp:77] Creating layer norm2 I0406 07:07:42.653030 5226 net.cpp:84] Creating Layer norm2 I0406 07:07:42.653034 5226 net.cpp:406] norm2 <- conv2 I0406 07:07:42.653041 5226 net.cpp:380] norm2 -> norm2 I0406 07:07:42.653803 5226 net.cpp:122] Setting up norm2 I0406 07:07:42.653816 5226 net.cpp:129] Top shape: 32 256 27 27 (5971968) I0406 07:07:42.653820 5226 net.cpp:137] Memory required for data: 211922688 I0406 07:07:42.653825 5226 layer_factory.hpp:77] Creating layer pool2 I0406 07:07:42.653831 5226 net.cpp:84] Creating Layer pool2 I0406 07:07:42.653836 5226 net.cpp:406] pool2 <- norm2 I0406 07:07:42.653843 5226 net.cpp:380] pool2 -> pool2 I0406 07:07:42.653882 5226 net.cpp:122] Setting up pool2 I0406 07:07:42.653889 5226 net.cpp:129] Top shape: 32 256 13 13 (1384448) I0406 07:07:42.653892 5226 net.cpp:137] Memory required for data: 217460480 I0406 07:07:42.653896 5226 layer_factory.hpp:77] Creating layer conv3 I0406 07:07:42.653910 5226 net.cpp:84] Creating Layer conv3 I0406 07:07:42.653914 5226 net.cpp:406] conv3 <- pool2 I0406 07:07:42.653921 5226 net.cpp:380] conv3 -> conv3 I0406 07:07:42.669800 5226 net.cpp:122] Setting up conv3 I0406 07:07:42.669822 5226 net.cpp:129] Top shape: 32 384 13 13 (2076672) I0406 07:07:42.669826 5226 net.cpp:137] Memory required for data: 225767168 I0406 07:07:42.669840 5226 layer_factory.hpp:77] Creating layer relu3 I0406 07:07:42.669857 5226 net.cpp:84] Creating Layer relu3 I0406 07:07:42.669862 5226 net.cpp:406] relu3 <- conv3 I0406 07:07:42.669868 5226 net.cpp:367] relu3 -> conv3 (in-place) I0406 07:07:42.670576 5226 net.cpp:122] Setting up relu3 I0406 07:07:42.670593 5226 net.cpp:129] Top shape: 32 384 13 13 (2076672) I0406 07:07:42.670596 5226 net.cpp:137] Memory required for data: 234073856 I0406 07:07:42.670600 5226 layer_factory.hpp:77] Creating layer conv4 I0406 07:07:42.670612 5226 net.cpp:84] Creating Layer conv4 I0406 07:07:42.670617 5226 net.cpp:406] conv4 <- conv3 I0406 07:07:42.670624 5226 net.cpp:380] conv4 -> conv4 I0406 07:07:42.684892 5226 net.cpp:122] Setting up conv4 I0406 07:07:42.684912 5226 net.cpp:129] Top shape: 32 384 13 13 (2076672) I0406 07:07:42.684916 5226 net.cpp:137] Memory required for data: 242380544 I0406 07:07:42.684926 5226 layer_factory.hpp:77] Creating layer relu4 I0406 07:07:42.684937 5226 net.cpp:84] Creating Layer relu4 I0406 07:07:42.684940 5226 net.cpp:406] relu4 <- conv4 I0406 07:07:42.684949 5226 net.cpp:367] relu4 -> conv4 (in-place) I0406 07:07:42.685421 5226 net.cpp:122] Setting up relu4 I0406 07:07:42.685432 5226 net.cpp:129] Top shape: 32 384 13 13 (2076672) I0406 07:07:42.685436 5226 net.cpp:137] Memory required for data: 250687232 I0406 07:07:42.685441 5226 layer_factory.hpp:77] Creating layer conv5 I0406 07:07:42.685453 5226 net.cpp:84] Creating Layer conv5 I0406 07:07:42.685457 5226 net.cpp:406] conv5 <- conv4 I0406 07:07:42.685464 5226 net.cpp:380] conv5 -> conv5 I0406 07:07:42.697454 5226 net.cpp:122] Setting up conv5 I0406 07:07:42.697470 5226 net.cpp:129] Top shape: 32 256 13 13 (1384448) I0406 07:07:42.697474 5226 net.cpp:137] Memory required for data: 256225024 I0406 07:07:42.697485 5226 layer_factory.hpp:77] Creating layer relu5 I0406 07:07:42.697495 5226 net.cpp:84] Creating Layer relu5 I0406 07:07:42.697499 5226 net.cpp:406] relu5 <- conv5 I0406 07:07:42.697523 5226 net.cpp:367] relu5 -> conv5 (in-place) I0406 07:07:42.698045 5226 net.cpp:122] Setting up relu5 I0406 07:07:42.698055 5226 net.cpp:129] Top shape: 32 256 13 13 (1384448) I0406 07:07:42.698056 5226 net.cpp:137] Memory required for data: 261762816 I0406 07:07:42.698060 5226 layer_factory.hpp:77] Creating layer pool5 I0406 07:07:42.698069 5226 net.cpp:84] Creating Layer pool5 I0406 07:07:42.698071 5226 net.cpp:406] pool5 <- conv5 I0406 07:07:42.698076 5226 net.cpp:380] pool5 -> pool5 I0406 07:07:42.698112 5226 net.cpp:122] Setting up pool5 I0406 07:07:42.698115 5226 net.cpp:129] Top shape: 32 256 6 6 (294912) I0406 07:07:42.698117 5226 net.cpp:137] Memory required for data: 262942464 I0406 07:07:42.698119 5226 layer_factory.hpp:77] Creating layer fc6 I0406 07:07:42.698125 5226 net.cpp:84] Creating Layer fc6 I0406 07:07:42.698129 5226 net.cpp:406] fc6 <- pool5 I0406 07:07:42.698132 5226 net.cpp:380] fc6 -> fc6 I0406 07:07:43.029364 5226 net.cpp:122] Setting up fc6 I0406 07:07:43.029381 5226 net.cpp:129] Top shape: 32 4096 (131072) I0406 07:07:43.029383 5226 net.cpp:137] Memory required for data: 263466752 I0406 07:07:43.029392 5226 layer_factory.hpp:77] Creating layer relu6 I0406 07:07:43.029400 5226 net.cpp:84] Creating Layer relu6 I0406 07:07:43.029403 5226 net.cpp:406] relu6 <- fc6 I0406 07:07:43.029408 5226 net.cpp:367] relu6 -> fc6 (in-place) I0406 07:07:43.031992 5226 net.cpp:122] Setting up relu6 I0406 07:07:43.032001 5226 net.cpp:129] Top shape: 32 4096 (131072) I0406 07:07:43.032003 5226 net.cpp:137] Memory required for data: 263991040 I0406 07:07:43.032006 5226 layer_factory.hpp:77] Creating layer drop6 I0406 07:07:43.032012 5226 net.cpp:84] Creating Layer drop6 I0406 07:07:43.032014 5226 net.cpp:406] drop6 <- fc6 I0406 07:07:43.032019 5226 net.cpp:367] drop6 -> fc6 (in-place) I0406 07:07:43.032043 5226 net.cpp:122] Setting up drop6 I0406 07:07:43.032047 5226 net.cpp:129] Top shape: 32 4096 (131072) I0406 07:07:43.032050 5226 net.cpp:137] Memory required for data: 264515328 I0406 07:07:43.032052 5226 layer_factory.hpp:77] Creating layer fc7 I0406 07:07:43.032058 5226 net.cpp:84] Creating Layer fc7 I0406 07:07:43.032060 5226 net.cpp:406] fc7 <- fc6 I0406 07:07:43.032065 5226 net.cpp:380] fc7 -> fc7 I0406 07:07:43.178071 5226 net.cpp:122] Setting up fc7 I0406 07:07:43.178089 5226 net.cpp:129] Top shape: 32 4096 (131072) I0406 07:07:43.178092 5226 net.cpp:137] Memory required for data: 265039616 I0406 07:07:43.178099 5226 layer_factory.hpp:77] Creating layer relu7 I0406 07:07:43.178107 5226 net.cpp:84] Creating Layer relu7 I0406 07:07:43.178110 5226 net.cpp:406] relu7 <- fc7 I0406 07:07:43.178117 5226 net.cpp:367] relu7 -> fc7 (in-place) I0406 07:07:43.178485 5226 net.cpp:122] Setting up relu7 I0406 07:07:43.178493 5226 net.cpp:129] Top shape: 32 4096 (131072) I0406 07:07:43.178495 5226 net.cpp:137] Memory required for data: 265563904 I0406 07:07:43.178498 5226 layer_factory.hpp:77] Creating layer drop7 I0406 07:07:43.178504 5226 net.cpp:84] Creating Layer drop7 I0406 07:07:43.178506 5226 net.cpp:406] drop7 <- fc7 I0406 07:07:43.178510 5226 net.cpp:367] drop7 -> fc7 (in-place) I0406 07:07:43.178531 5226 net.cpp:122] Setting up drop7 I0406 07:07:43.178535 5226 net.cpp:129] Top shape: 32 4096 (131072) I0406 07:07:43.178537 5226 net.cpp:137] Memory required for data: 266088192 I0406 07:07:43.178539 5226 layer_factory.hpp:77] Creating layer fc8 I0406 07:07:43.178545 5226 net.cpp:84] Creating Layer fc8 I0406 07:07:43.178547 5226 net.cpp:406] fc8 <- fc7 I0406 07:07:43.178551 5226 net.cpp:380] fc8 -> fc8 I0406 07:07:43.185705 5226 net.cpp:122] Setting up fc8 I0406 07:07:43.185714 5226 net.cpp:129] Top shape: 32 196 (6272) I0406 07:07:43.185715 5226 net.cpp:137] Memory required for data: 266113280 I0406 07:07:43.185720 5226 layer_factory.hpp:77] Creating layer fc8_fc8_0_split I0406 07:07:43.185725 5226 net.cpp:84] Creating Layer fc8_fc8_0_split I0406 07:07:43.185729 5226 net.cpp:406] fc8_fc8_0_split <- fc8 I0406 07:07:43.185751 5226 net.cpp:380] fc8_fc8_0_split -> fc8_fc8_0_split_0 I0406 07:07:43.185757 5226 net.cpp:380] fc8_fc8_0_split -> fc8_fc8_0_split_1 I0406 07:07:43.185784 5226 net.cpp:122] Setting up fc8_fc8_0_split I0406 07:07:43.185788 5226 net.cpp:129] Top shape: 32 196 (6272) I0406 07:07:43.185791 5226 net.cpp:129] Top shape: 32 196 (6272) I0406 07:07:43.185792 5226 net.cpp:137] Memory required for data: 266163456 I0406 07:07:43.185794 5226 layer_factory.hpp:77] Creating layer accuracy I0406 07:07:43.185801 5226 net.cpp:84] Creating Layer accuracy I0406 07:07:43.185803 5226 net.cpp:406] accuracy <- fc8_fc8_0_split_0 I0406 07:07:43.185806 5226 net.cpp:406] accuracy <- label_val-data_1_split_0 I0406 07:07:43.185809 5226 net.cpp:380] accuracy -> accuracy I0406 07:07:43.185815 5226 net.cpp:122] Setting up accuracy I0406 07:07:43.185818 5226 net.cpp:129] Top shape: (1) I0406 07:07:43.185819 5226 net.cpp:137] Memory required for data: 266163460 I0406 07:07:43.185822 5226 layer_factory.hpp:77] Creating layer loss I0406 07:07:43.185827 5226 net.cpp:84] Creating Layer loss I0406 07:07:43.185828 5226 net.cpp:406] loss <- fc8_fc8_0_split_1 I0406 07:07:43.185832 5226 net.cpp:406] loss <- label_val-data_1_split_1 I0406 07:07:43.185834 5226 net.cpp:380] loss -> loss I0406 07:07:43.185839 5226 layer_factory.hpp:77] Creating layer loss I0406 07:07:43.186419 5226 net.cpp:122] Setting up loss I0406 07:07:43.186427 5226 net.cpp:129] Top shape: (1) I0406 07:07:43.186429 5226 net.cpp:132] with loss weight 1 I0406 07:07:43.186437 5226 net.cpp:137] Memory required for data: 266163464 I0406 07:07:43.186439 5226 net.cpp:198] loss needs backward computation. I0406 07:07:43.186444 5226 net.cpp:200] accuracy does not need backward computation. I0406 07:07:43.186446 5226 net.cpp:198] fc8_fc8_0_split needs backward computation. I0406 07:07:43.186448 5226 net.cpp:198] fc8 needs backward computation. I0406 07:07:43.186450 5226 net.cpp:198] drop7 needs backward computation. I0406 07:07:43.186452 5226 net.cpp:198] relu7 needs backward computation. I0406 07:07:43.186455 5226 net.cpp:198] fc7 needs backward computation. I0406 07:07:43.186456 5226 net.cpp:198] drop6 needs backward computation. I0406 07:07:43.186458 5226 net.cpp:198] relu6 needs backward computation. I0406 07:07:43.186460 5226 net.cpp:198] fc6 needs backward computation. I0406 07:07:43.186463 5226 net.cpp:198] pool5 needs backward computation. I0406 07:07:43.186466 5226 net.cpp:198] relu5 needs backward computation. I0406 07:07:43.186468 5226 net.cpp:198] conv5 needs backward computation. I0406 07:07:43.186470 5226 net.cpp:198] relu4 needs backward computation. I0406 07:07:43.186472 5226 net.cpp:198] conv4 needs backward computation. I0406 07:07:43.186475 5226 net.cpp:198] relu3 needs backward computation. I0406 07:07:43.186477 5226 net.cpp:198] conv3 needs backward computation. I0406 07:07:43.186480 5226 net.cpp:198] pool2 needs backward computation. I0406 07:07:43.186483 5226 net.cpp:198] norm2 needs backward computation. I0406 07:07:43.186486 5226 net.cpp:198] relu2 needs backward computation. I0406 07:07:43.186487 5226 net.cpp:198] conv2 needs backward computation. I0406 07:07:43.186489 5226 net.cpp:198] pool1 needs backward computation. I0406 07:07:43.186492 5226 net.cpp:198] norm1 needs backward computation. I0406 07:07:43.186494 5226 net.cpp:198] relu1 needs backward computation. I0406 07:07:43.186496 5226 net.cpp:198] conv1 needs backward computation. I0406 07:07:43.186498 5226 net.cpp:200] label_val-data_1_split does not need backward computation. I0406 07:07:43.186501 5226 net.cpp:200] val-data does not need backward computation. I0406 07:07:43.186503 5226 net.cpp:242] This network produces output accuracy I0406 07:07:43.186506 5226 net.cpp:242] This network produces output loss I0406 07:07:43.186520 5226 net.cpp:255] Network initialization done. I0406 07:07:43.186589 5226 solver.cpp:56] Solver scaffolding done. I0406 07:07:43.186964 5226 caffe.cpp:248] Starting Optimization I0406 07:07:43.186972 5226 solver.cpp:272] Solving I0406 07:07:43.186982 5226 solver.cpp:273] Learning Rate Policy: fixed I0406 07:07:43.188912 5226 solver.cpp:330] Iteration 0, Testing net (#0) I0406 07:07:43.188921 5226 net.cpp:676] Ignoring source layer train-data I0406 07:07:43.290103 5226 blocking_queue.cpp:49] Waiting for data I0406 07:07:47.569265 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:07:47.617841 5226 solver.cpp:397] Test net output #0: accuracy = 0.00551471 I0406 07:07:47.617877 5226 solver.cpp:397] Test net output #1: loss = 5.28124 (* 1 = 5.28124 loss) I0406 07:07:47.764514 5226 solver.cpp:218] Iteration 0 (1.16182e+36 iter/s, 4.57746s/12 iters), loss = 5.31153 I0406 07:07:47.766070 5226 solver.cpp:237] Train net output #0: loss = 5.31153 (* 1 = 5.31153 loss) I0406 07:07:47.766080 5226 sgd_solver.cpp:105] Iteration 0, lr = 0.01 I0406 07:07:51.554159 5226 solver.cpp:218] Iteration 12 (3.16785 iter/s, 3.78805s/12 iters), loss = 5.28726 I0406 07:07:51.554189 5226 solver.cpp:237] Train net output #0: loss = 5.28726 (* 1 = 5.28726 loss) I0406 07:07:51.554195 5226 sgd_solver.cpp:105] Iteration 12, lr = 0.01 I0406 07:07:56.632417 5226 solver.cpp:218] Iteration 24 (2.36305 iter/s, 5.07818s/12 iters), loss = 5.27883 I0406 07:07:56.632458 5226 solver.cpp:237] Train net output #0: loss = 5.27883 (* 1 = 5.27883 loss) I0406 07:07:56.632463 5226 sgd_solver.cpp:105] Iteration 24, lr = 0.01 I0406 07:08:01.914950 5226 solver.cpp:218] Iteration 36 (2.27167 iter/s, 5.28245s/12 iters), loss = 5.28964 I0406 07:08:01.914988 5226 solver.cpp:237] Train net output #0: loss = 5.28964 (* 1 = 5.28964 loss) I0406 07:08:01.914994 5226 sgd_solver.cpp:105] Iteration 36, lr = 0.01 I0406 07:08:07.229256 5226 solver.cpp:218] Iteration 48 (2.25809 iter/s, 5.31423s/12 iters), loss = 5.30546 I0406 07:08:07.229291 5226 solver.cpp:237] Train net output #0: loss = 5.30546 (* 1 = 5.30546 loss) I0406 07:08:07.229296 5226 sgd_solver.cpp:105] Iteration 48, lr = 0.01 I0406 07:08:12.734804 5226 solver.cpp:218] Iteration 60 (2.17965 iter/s, 5.50547s/12 iters), loss = 5.26713 I0406 07:08:12.734925 5226 solver.cpp:237] Train net output #0: loss = 5.26713 (* 1 = 5.26713 loss) I0406 07:08:12.734932 5226 sgd_solver.cpp:105] Iteration 60, lr = 0.01 I0406 07:08:17.932544 5226 solver.cpp:218] Iteration 72 (2.30877 iter/s, 5.19757s/12 iters), loss = 5.32175 I0406 07:08:17.932590 5226 solver.cpp:237] Train net output #0: loss = 5.32175 (* 1 = 5.32175 loss) I0406 07:08:17.932600 5226 sgd_solver.cpp:105] Iteration 72, lr = 0.01 I0406 07:08:23.362987 5226 solver.cpp:218] Iteration 84 (2.2098 iter/s, 5.43035s/12 iters), loss = 5.29407 I0406 07:08:23.363035 5226 solver.cpp:237] Train net output #0: loss = 5.29407 (* 1 = 5.29407 loss) I0406 07:08:23.363044 5226 sgd_solver.cpp:105] Iteration 84, lr = 0.01 I0406 07:08:28.821043 5226 solver.cpp:218] Iteration 96 (2.19863 iter/s, 5.45796s/12 iters), loss = 5.27596 I0406 07:08:28.821087 5226 solver.cpp:237] Train net output #0: loss = 5.27596 (* 1 = 5.27596 loss) I0406 07:08:28.821095 5226 sgd_solver.cpp:105] Iteration 96, lr = 0.01 I0406 07:08:30.688700 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:08:31.006948 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_102.caffemodel I0406 07:08:34.077026 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_102.solverstate I0406 07:08:36.370083 5226 solver.cpp:330] Iteration 102, Testing net (#0) I0406 07:08:36.370102 5226 net.cpp:676] Ignoring source layer train-data I0406 07:08:40.734298 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:08:40.823169 5226 solver.cpp:397] Test net output #0: accuracy = 0.00612745 I0406 07:08:40.823205 5226 solver.cpp:397] Test net output #1: loss = 5.29375 (* 1 = 5.29375 loss) I0406 07:08:42.808393 5226 solver.cpp:218] Iteration 108 (0.857927 iter/s, 13.9872s/12 iters), loss = 5.28957 I0406 07:08:42.808533 5226 solver.cpp:237] Train net output #0: loss = 5.28957 (* 1 = 5.28957 loss) I0406 07:08:42.808539 5226 sgd_solver.cpp:105] Iteration 108, lr = 0.01 I0406 07:08:47.928736 5226 solver.cpp:218] Iteration 120 (2.34368 iter/s, 5.12016s/12 iters), loss = 5.27194 I0406 07:08:47.928768 5226 solver.cpp:237] Train net output #0: loss = 5.27194 (* 1 = 5.27194 loss) I0406 07:08:47.928773 5226 sgd_solver.cpp:105] Iteration 120, lr = 0.01 I0406 07:08:53.277961 5226 solver.cpp:218] Iteration 132 (2.24335 iter/s, 5.34914s/12 iters), loss = 5.28037 I0406 07:08:53.277999 5226 solver.cpp:237] Train net output #0: loss = 5.28037 (* 1 = 5.28037 loss) I0406 07:08:53.278005 5226 sgd_solver.cpp:105] Iteration 132, lr = 0.01 I0406 07:08:58.564335 5226 solver.cpp:218] Iteration 144 (2.27003 iter/s, 5.28628s/12 iters), loss = 5.26267 I0406 07:08:58.564371 5226 solver.cpp:237] Train net output #0: loss = 5.26267 (* 1 = 5.26267 loss) I0406 07:08:58.564376 5226 sgd_solver.cpp:105] Iteration 144, lr = 0.01 I0406 07:09:03.603242 5226 solver.cpp:218] Iteration 156 (2.38151 iter/s, 5.03882s/12 iters), loss = 5.29929 I0406 07:09:03.603279 5226 solver.cpp:237] Train net output #0: loss = 5.29929 (* 1 = 5.29929 loss) I0406 07:09:03.603286 5226 sgd_solver.cpp:105] Iteration 156, lr = 0.01 I0406 07:09:08.798015 5226 solver.cpp:218] Iteration 168 (2.31006 iter/s, 5.19468s/12 iters), loss = 5.26474 I0406 07:09:08.798054 5226 solver.cpp:237] Train net output #0: loss = 5.26474 (* 1 = 5.26474 loss) I0406 07:09:08.798059 5226 sgd_solver.cpp:105] Iteration 168, lr = 0.01 I0406 07:09:14.158486 5226 solver.cpp:218] Iteration 180 (2.23865 iter/s, 5.36038s/12 iters), loss = 5.29722 I0406 07:09:14.158578 5226 solver.cpp:237] Train net output #0: loss = 5.29722 (* 1 = 5.29722 loss) I0406 07:09:14.158586 5226 sgd_solver.cpp:105] Iteration 180, lr = 0.01 I0406 07:09:19.540362 5226 solver.cpp:218] Iteration 192 (2.22976 iter/s, 5.38173s/12 iters), loss = 5.20534 I0406 07:09:19.540397 5226 solver.cpp:237] Train net output #0: loss = 5.20534 (* 1 = 5.20534 loss) I0406 07:09:19.540402 5226 sgd_solver.cpp:105] Iteration 192, lr = 0.01 I0406 07:09:23.686969 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:09:24.396731 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_204.caffemodel I0406 07:09:27.482285 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_204.solverstate I0406 07:09:29.789227 5226 solver.cpp:330] Iteration 204, Testing net (#0) I0406 07:09:29.789250 5226 net.cpp:676] Ignoring source layer train-data I0406 07:09:33.984045 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:09:34.107908 5226 solver.cpp:397] Test net output #0: accuracy = 0.0104167 I0406 07:09:34.107944 5226 solver.cpp:397] Test net output #1: loss = 5.21406 (* 1 = 5.21406 loss) I0406 07:09:34.251067 5226 solver.cpp:218] Iteration 204 (0.815741 iter/s, 14.7106s/12 iters), loss = 5.15916 I0406 07:09:34.252631 5226 solver.cpp:237] Train net output #0: loss = 5.15916 (* 1 = 5.15916 loss) I0406 07:09:34.252645 5226 sgd_solver.cpp:105] Iteration 204, lr = 0.01 I0406 07:09:38.743819 5226 solver.cpp:218] Iteration 216 (2.67192 iter/s, 4.49114s/12 iters), loss = 5.22728 I0406 07:09:38.743860 5226 solver.cpp:237] Train net output #0: loss = 5.22728 (* 1 = 5.22728 loss) I0406 07:09:38.743865 5226 sgd_solver.cpp:105] Iteration 216, lr = 0.01 I0406 07:09:44.088773 5226 solver.cpp:218] Iteration 228 (2.24515 iter/s, 5.34486s/12 iters), loss = 5.22056 I0406 07:09:44.088812 5226 solver.cpp:237] Train net output #0: loss = 5.22056 (* 1 = 5.22056 loss) I0406 07:09:44.088819 5226 sgd_solver.cpp:105] Iteration 228, lr = 0.01 I0406 07:09:49.379869 5226 solver.cpp:218] Iteration 240 (2.268 iter/s, 5.291s/12 iters), loss = 5.18359 I0406 07:09:49.380000 5226 solver.cpp:237] Train net output #0: loss = 5.18359 (* 1 = 5.18359 loss) I0406 07:09:49.380007 5226 sgd_solver.cpp:105] Iteration 240, lr = 0.01 I0406 07:09:54.555052 5226 solver.cpp:218] Iteration 252 (2.31884 iter/s, 5.175s/12 iters), loss = 5.2487 I0406 07:09:54.555094 5226 solver.cpp:237] Train net output #0: loss = 5.2487 (* 1 = 5.2487 loss) I0406 07:09:54.555099 5226 sgd_solver.cpp:105] Iteration 252, lr = 0.01 I0406 07:09:59.578614 5226 solver.cpp:218] Iteration 264 (2.38879 iter/s, 5.02346s/12 iters), loss = 5.14489 I0406 07:09:59.578672 5226 solver.cpp:237] Train net output #0: loss = 5.14489 (* 1 = 5.14489 loss) I0406 07:09:59.578681 5226 sgd_solver.cpp:105] Iteration 264, lr = 0.01 I0406 07:10:04.863327 5226 solver.cpp:218] Iteration 276 (2.27075 iter/s, 5.28461s/12 iters), loss = 5.12572 I0406 07:10:04.863373 5226 solver.cpp:237] Train net output #0: loss = 5.12572 (* 1 = 5.12572 loss) I0406 07:10:04.863379 5226 sgd_solver.cpp:105] Iteration 276, lr = 0.01 I0406 07:10:10.183627 5226 solver.cpp:218] Iteration 288 (2.25555 iter/s, 5.3202s/12 iters), loss = 5.16231 I0406 07:10:10.183665 5226 solver.cpp:237] Train net output #0: loss = 5.16231 (* 1 = 5.16231 loss) I0406 07:10:10.183670 5226 sgd_solver.cpp:105] Iteration 288, lr = 0.01 I0406 07:10:15.424337 5226 solver.cpp:218] Iteration 300 (2.2898 iter/s, 5.24062s/12 iters), loss = 5.24297 I0406 07:10:15.424374 5226 solver.cpp:237] Train net output #0: loss = 5.24297 (* 1 = 5.24297 loss) I0406 07:10:15.424379 5226 sgd_solver.cpp:105] Iteration 300, lr = 0.01 I0406 07:10:16.467451 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:10:17.586361 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_306.caffemodel I0406 07:10:20.719285 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_306.solverstate I0406 07:10:23.047116 5226 solver.cpp:330] Iteration 306, Testing net (#0) I0406 07:10:23.047137 5226 net.cpp:676] Ignoring source layer train-data I0406 07:10:27.161191 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:10:27.317126 5226 solver.cpp:397] Test net output #0: accuracy = 0.0104167 I0406 07:10:27.317157 5226 solver.cpp:397] Test net output #1: loss = 5.16469 (* 1 = 5.16469 loss) I0406 07:10:29.220289 5226 solver.cpp:218] Iteration 312 (0.869829 iter/s, 13.7958s/12 iters), loss = 5.15112 I0406 07:10:29.220326 5226 solver.cpp:237] Train net output #0: loss = 5.15112 (* 1 = 5.15112 loss) I0406 07:10:29.220332 5226 sgd_solver.cpp:105] Iteration 312, lr = 0.01 I0406 07:10:34.662358 5226 solver.cpp:218] Iteration 324 (2.20508 iter/s, 5.44198s/12 iters), loss = 5.23607 I0406 07:10:34.662395 5226 solver.cpp:237] Train net output #0: loss = 5.23607 (* 1 = 5.23607 loss) I0406 07:10:34.662400 5226 sgd_solver.cpp:105] Iteration 324, lr = 0.01 I0406 07:10:40.053938 5226 solver.cpp:218] Iteration 336 (2.22573 iter/s, 5.39149s/12 iters), loss = 5.15472 I0406 07:10:40.053977 5226 solver.cpp:237] Train net output #0: loss = 5.15472 (* 1 = 5.15472 loss) I0406 07:10:40.053982 5226 sgd_solver.cpp:105] Iteration 336, lr = 0.01 I0406 07:10:45.245923 5226 solver.cpp:218] Iteration 348 (2.31129 iter/s, 5.1919s/12 iters), loss = 5.11753 I0406 07:10:45.245961 5226 solver.cpp:237] Train net output #0: loss = 5.11753 (* 1 = 5.11753 loss) I0406 07:10:45.245967 5226 sgd_solver.cpp:105] Iteration 348, lr = 0.01 I0406 07:10:50.659412 5226 solver.cpp:218] Iteration 360 (2.21672 iter/s, 5.41341s/12 iters), loss = 5.19051 I0406 07:10:50.659442 5226 solver.cpp:237] Train net output #0: loss = 5.19051 (* 1 = 5.19051 loss) I0406 07:10:50.659448 5226 sgd_solver.cpp:105] Iteration 360, lr = 0.01 I0406 07:10:56.060766 5226 solver.cpp:218] Iteration 372 (2.2217 iter/s, 5.40127s/12 iters), loss = 5.1577 I0406 07:10:56.060873 5226 solver.cpp:237] Train net output #0: loss = 5.1577 (* 1 = 5.1577 loss) I0406 07:10:56.060889 5226 sgd_solver.cpp:105] Iteration 372, lr = 0.01 I0406 07:11:01.440582 5226 solver.cpp:218] Iteration 384 (2.23063 iter/s, 5.37965s/12 iters), loss = 5.20289 I0406 07:11:01.440636 5226 solver.cpp:237] Train net output #0: loss = 5.20289 (* 1 = 5.20289 loss) I0406 07:11:01.440649 5226 sgd_solver.cpp:105] Iteration 384, lr = 0.01 I0406 07:11:06.635514 5226 solver.cpp:218] Iteration 396 (2.30998 iter/s, 5.19484s/12 iters), loss = 5.13047 I0406 07:11:06.635550 5226 solver.cpp:237] Train net output #0: loss = 5.13047 (* 1 = 5.13047 loss) I0406 07:11:06.635555 5226 sgd_solver.cpp:105] Iteration 396, lr = 0.01 I0406 07:11:09.925401 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:11:11.458490 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_408.caffemodel I0406 07:11:14.464010 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_408.solverstate I0406 07:11:16.765609 5226 solver.cpp:330] Iteration 408, Testing net (#0) I0406 07:11:16.765626 5226 net.cpp:676] Ignoring source layer train-data I0406 07:11:20.906955 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:11:21.115682 5226 solver.cpp:397] Test net output #0: accuracy = 0.0147059 I0406 07:11:21.115717 5226 solver.cpp:397] Test net output #1: loss = 5.12747 (* 1 = 5.12747 loss) I0406 07:11:21.258468 5226 solver.cpp:218] Iteration 408 (0.820636 iter/s, 14.6228s/12 iters), loss = 5.12815 I0406 07:11:21.258514 5226 solver.cpp:237] Train net output #0: loss = 5.12815 (* 1 = 5.12815 loss) I0406 07:11:21.258522 5226 sgd_solver.cpp:105] Iteration 408, lr = 0.01 I0406 07:11:25.609107 5226 solver.cpp:218] Iteration 420 (2.75827 iter/s, 4.35055s/12 iters), loss = 5.093 I0406 07:11:25.609150 5226 solver.cpp:237] Train net output #0: loss = 5.093 (* 1 = 5.093 loss) I0406 07:11:25.609156 5226 sgd_solver.cpp:105] Iteration 420, lr = 0.01 I0406 07:11:30.603524 5226 solver.cpp:218] Iteration 432 (2.40273 iter/s, 4.99433s/12 iters), loss = 5.04658 I0406 07:11:30.603638 5226 solver.cpp:237] Train net output #0: loss = 5.04658 (* 1 = 5.04658 loss) I0406 07:11:30.603644 5226 sgd_solver.cpp:105] Iteration 432, lr = 0.01 I0406 07:11:35.849427 5226 solver.cpp:218] Iteration 444 (2.28757 iter/s, 5.24575s/12 iters), loss = 5.11776 I0406 07:11:35.849462 5226 solver.cpp:237] Train net output #0: loss = 5.11776 (* 1 = 5.11776 loss) I0406 07:11:35.849467 5226 sgd_solver.cpp:105] Iteration 444, lr = 0.01 I0406 07:11:41.185731 5226 solver.cpp:218] Iteration 456 (2.24878 iter/s, 5.33622s/12 iters), loss = 5.15248 I0406 07:11:41.185772 5226 solver.cpp:237] Train net output #0: loss = 5.15248 (* 1 = 5.15248 loss) I0406 07:11:41.185779 5226 sgd_solver.cpp:105] Iteration 456, lr = 0.01 I0406 07:11:46.489115 5226 solver.cpp:218] Iteration 468 (2.26274 iter/s, 5.3033s/12 iters), loss = 5.08704 I0406 07:11:46.489152 5226 solver.cpp:237] Train net output #0: loss = 5.08704 (* 1 = 5.08704 loss) I0406 07:11:46.489157 5226 sgd_solver.cpp:105] Iteration 468, lr = 0.01 I0406 07:11:51.581387 5226 solver.cpp:218] Iteration 480 (2.35655 iter/s, 5.09219s/12 iters), loss = 5.00496 I0406 07:11:51.581419 5226 solver.cpp:237] Train net output #0: loss = 5.00496 (* 1 = 5.00496 loss) I0406 07:11:51.581424 5226 sgd_solver.cpp:105] Iteration 480, lr = 0.01 I0406 07:11:56.632428 5226 solver.cpp:218] Iteration 492 (2.37579 iter/s, 5.05096s/12 iters), loss = 5.11756 I0406 07:11:56.632467 5226 solver.cpp:237] Train net output #0: loss = 5.11756 (* 1 = 5.11756 loss) I0406 07:11:56.632472 5226 sgd_solver.cpp:105] Iteration 492, lr = 0.01 I0406 07:12:01.945720 5226 solver.cpp:218] Iteration 504 (2.25852 iter/s, 5.3132s/12 iters), loss = 5.04789 I0406 07:12:01.945844 5226 solver.cpp:237] Train net output #0: loss = 5.04789 (* 1 = 5.04789 loss) I0406 07:12:01.945853 5226 sgd_solver.cpp:105] Iteration 504, lr = 0.01 I0406 07:12:02.175249 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:12:04.093015 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_510.caffemodel I0406 07:12:07.163301 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_510.solverstate I0406 07:12:09.464279 5226 solver.cpp:330] Iteration 510, Testing net (#0) I0406 07:12:09.464299 5226 net.cpp:676] Ignoring source layer train-data I0406 07:12:13.557585 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:12:13.801589 5226 solver.cpp:397] Test net output #0: accuracy = 0.0214461 I0406 07:12:13.801616 5226 solver.cpp:397] Test net output #1: loss = 5.05785 (* 1 = 5.05785 loss) I0406 07:12:15.738641 5226 solver.cpp:218] Iteration 516 (0.870026 iter/s, 13.7927s/12 iters), loss = 5.09813 I0406 07:12:15.738682 5226 solver.cpp:237] Train net output #0: loss = 5.09813 (* 1 = 5.09813 loss) I0406 07:12:15.738687 5226 sgd_solver.cpp:105] Iteration 516, lr = 0.01 I0406 07:12:21.254230 5226 solver.cpp:218] Iteration 528 (2.17569 iter/s, 5.5155s/12 iters), loss = 5.16702 I0406 07:12:21.254266 5226 solver.cpp:237] Train net output #0: loss = 5.16702 (* 1 = 5.16702 loss) I0406 07:12:21.254271 5226 sgd_solver.cpp:105] Iteration 528, lr = 0.01 I0406 07:12:26.438645 5226 solver.cpp:218] Iteration 540 (2.31467 iter/s, 5.18433s/12 iters), loss = 5.01441 I0406 07:12:26.438683 5226 solver.cpp:237] Train net output #0: loss = 5.01441 (* 1 = 5.01441 loss) I0406 07:12:26.438688 5226 sgd_solver.cpp:105] Iteration 540, lr = 0.01 I0406 07:12:31.549919 5226 solver.cpp:218] Iteration 552 (2.34779 iter/s, 5.11119s/12 iters), loss = 5.1222 I0406 07:12:31.549957 5226 solver.cpp:237] Train net output #0: loss = 5.1222 (* 1 = 5.1222 loss) I0406 07:12:31.549962 5226 sgd_solver.cpp:105] Iteration 552, lr = 0.01 I0406 07:12:37.008781 5226 solver.cpp:218] Iteration 564 (2.1983 iter/s, 5.45877s/12 iters), loss = 5.0354 I0406 07:12:37.008937 5226 solver.cpp:237] Train net output #0: loss = 5.0354 (* 1 = 5.0354 loss) I0406 07:12:37.008946 5226 sgd_solver.cpp:105] Iteration 564, lr = 0.01 I0406 07:12:42.297931 5226 solver.cpp:218] Iteration 576 (2.26888 iter/s, 5.28895s/12 iters), loss = 4.90123 I0406 07:12:42.297973 5226 solver.cpp:237] Train net output #0: loss = 4.90123 (* 1 = 4.90123 loss) I0406 07:12:42.297979 5226 sgd_solver.cpp:105] Iteration 576, lr = 0.01 I0406 07:12:47.422849 5226 solver.cpp:218] Iteration 588 (2.34154 iter/s, 5.12483s/12 iters), loss = 4.99606 I0406 07:12:47.422886 5226 solver.cpp:237] Train net output #0: loss = 4.99606 (* 1 = 4.99606 loss) I0406 07:12:47.422891 5226 sgd_solver.cpp:105] Iteration 588, lr = 0.01 I0406 07:12:52.957031 5226 solver.cpp:218] Iteration 600 (2.16838 iter/s, 5.5341s/12 iters), loss = 5.05287 I0406 07:12:52.957079 5226 solver.cpp:237] Train net output #0: loss = 5.05287 (* 1 = 5.05287 loss) I0406 07:12:52.957087 5226 sgd_solver.cpp:105] Iteration 600, lr = 0.01 I0406 07:12:55.310144 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:12:57.693874 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_612.caffemodel I0406 07:13:00.691838 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_612.solverstate I0406 07:13:02.985356 5226 solver.cpp:330] Iteration 612, Testing net (#0) I0406 07:13:02.985375 5226 net.cpp:676] Ignoring source layer train-data I0406 07:13:07.023017 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:13:07.305249 5226 solver.cpp:397] Test net output #0: accuracy = 0.0232843 I0406 07:13:07.305276 5226 solver.cpp:397] Test net output #1: loss = 5.0075 (* 1 = 5.0075 loss) I0406 07:13:07.446513 5226 solver.cpp:218] Iteration 612 (0.828195 iter/s, 14.4893s/12 iters), loss = 4.98406 I0406 07:13:07.446553 5226 solver.cpp:237] Train net output #0: loss = 4.98406 (* 1 = 4.98406 loss) I0406 07:13:07.446558 5226 sgd_solver.cpp:105] Iteration 612, lr = 0.01 I0406 07:13:11.916764 5226 solver.cpp:218] Iteration 624 (2.68446 iter/s, 4.47017s/12 iters), loss = 4.91334 I0406 07:13:11.916800 5226 solver.cpp:237] Train net output #0: loss = 4.91334 (* 1 = 4.91334 loss) I0406 07:13:11.916805 5226 sgd_solver.cpp:105] Iteration 624, lr = 0.01 I0406 07:13:17.145615 5226 solver.cpp:218] Iteration 636 (2.295 iter/s, 5.22876s/12 iters), loss = 5.01883 I0406 07:13:17.145668 5226 solver.cpp:237] Train net output #0: loss = 5.01883 (* 1 = 5.01883 loss) I0406 07:13:17.145679 5226 sgd_solver.cpp:105] Iteration 636, lr = 0.01 I0406 07:13:22.233357 5226 solver.cpp:218] Iteration 648 (2.35866 iter/s, 5.08764s/12 iters), loss = 4.85558 I0406 07:13:22.233395 5226 solver.cpp:237] Train net output #0: loss = 4.85558 (* 1 = 4.85558 loss) I0406 07:13:22.233402 5226 sgd_solver.cpp:105] Iteration 648, lr = 0.01 I0406 07:13:27.531522 5226 solver.cpp:218] Iteration 660 (2.26498 iter/s, 5.29807s/12 iters), loss = 4.93355 I0406 07:13:27.531569 5226 solver.cpp:237] Train net output #0: loss = 4.93355 (* 1 = 4.93355 loss) I0406 07:13:27.531586 5226 sgd_solver.cpp:105] Iteration 660, lr = 0.01 I0406 07:13:32.876922 5226 solver.cpp:218] Iteration 672 (2.24496 iter/s, 5.34531s/12 iters), loss = 5.00926 I0406 07:13:32.876958 5226 solver.cpp:237] Train net output #0: loss = 5.00926 (* 1 = 5.00926 loss) I0406 07:13:32.876965 5226 sgd_solver.cpp:105] Iteration 672, lr = 0.01 I0406 07:13:38.009531 5226 solver.cpp:218] Iteration 684 (2.33803 iter/s, 5.13253s/12 iters), loss = 4.87837 I0406 07:13:38.009645 5226 solver.cpp:237] Train net output #0: loss = 4.87837 (* 1 = 4.87837 loss) I0406 07:13:38.009651 5226 sgd_solver.cpp:105] Iteration 684, lr = 0.01 I0406 07:13:38.772302 5226 blocking_queue.cpp:49] Waiting for data I0406 07:13:43.268584 5226 solver.cpp:218] Iteration 696 (2.28185 iter/s, 5.25889s/12 iters), loss = 4.82162 I0406 07:13:43.268623 5226 solver.cpp:237] Train net output #0: loss = 4.82162 (* 1 = 4.82162 loss) I0406 07:13:43.268628 5226 sgd_solver.cpp:105] Iteration 696, lr = 0.01 I0406 07:13:48.136029 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:13:48.567907 5226 solver.cpp:218] Iteration 708 (2.26448 iter/s, 5.29924s/12 iters), loss = 4.99635 I0406 07:13:48.567945 5226 solver.cpp:237] Train net output #0: loss = 4.99635 (* 1 = 4.99635 loss) I0406 07:13:48.567950 5226 sgd_solver.cpp:105] Iteration 708, lr = 0.01 I0406 07:13:50.559027 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_714.caffemodel I0406 07:13:53.547910 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_714.solverstate I0406 07:13:55.844179 5226 solver.cpp:330] Iteration 714, Testing net (#0) I0406 07:13:55.844198 5226 net.cpp:676] Ignoring source layer train-data I0406 07:13:59.964818 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:14:00.278504 5226 solver.cpp:397] Test net output #0: accuracy = 0.0330882 I0406 07:14:00.278537 5226 solver.cpp:397] Test net output #1: loss = 4.94868 (* 1 = 4.94868 loss) I0406 07:14:02.088179 5226 solver.cpp:218] Iteration 720 (0.887565 iter/s, 13.5201s/12 iters), loss = 4.89392 I0406 07:14:02.088219 5226 solver.cpp:237] Train net output #0: loss = 4.89392 (* 1 = 4.89392 loss) I0406 07:14:02.088224 5226 sgd_solver.cpp:105] Iteration 720, lr = 0.01 I0406 07:14:07.110406 5226 solver.cpp:218] Iteration 732 (2.38942 iter/s, 5.02214s/12 iters), loss = 4.81985 I0406 07:14:07.110441 5226 solver.cpp:237] Train net output #0: loss = 4.81985 (* 1 = 4.81985 loss) I0406 07:14:07.110446 5226 sgd_solver.cpp:105] Iteration 732, lr = 0.01 I0406 07:14:12.401988 5226 solver.cpp:218] Iteration 744 (2.26779 iter/s, 5.2915s/12 iters), loss = 4.83772 I0406 07:14:12.402101 5226 solver.cpp:237] Train net output #0: loss = 4.83772 (* 1 = 4.83772 loss) I0406 07:14:12.402107 5226 sgd_solver.cpp:105] Iteration 744, lr = 0.01 I0406 07:14:17.662477 5226 solver.cpp:218] Iteration 756 (2.28123 iter/s, 5.26033s/12 iters), loss = 4.68698 I0406 07:14:17.662514 5226 solver.cpp:237] Train net output #0: loss = 4.68698 (* 1 = 4.68698 loss) I0406 07:14:17.662519 5226 sgd_solver.cpp:105] Iteration 756, lr = 0.01 I0406 07:14:22.994112 5226 solver.cpp:218] Iteration 768 (2.25075 iter/s, 5.33155s/12 iters), loss = 4.84255 I0406 07:14:22.994148 5226 solver.cpp:237] Train net output #0: loss = 4.84255 (* 1 = 4.84255 loss) I0406 07:14:22.994153 5226 sgd_solver.cpp:105] Iteration 768, lr = 0.01 I0406 07:14:28.278244 5226 solver.cpp:218] Iteration 780 (2.27099 iter/s, 5.28404s/12 iters), loss = 4.83813 I0406 07:14:28.278290 5226 solver.cpp:237] Train net output #0: loss = 4.83813 (* 1 = 4.83813 loss) I0406 07:14:28.278298 5226 sgd_solver.cpp:105] Iteration 780, lr = 0.01 I0406 07:14:33.531306 5226 solver.cpp:218] Iteration 792 (2.28442 iter/s, 5.25297s/12 iters), loss = 4.94625 I0406 07:14:33.531344 5226 solver.cpp:237] Train net output #0: loss = 4.94625 (* 1 = 4.94625 loss) I0406 07:14:33.531350 5226 sgd_solver.cpp:105] Iteration 792, lr = 0.01 I0406 07:14:38.829330 5226 solver.cpp:218] Iteration 804 (2.26503 iter/s, 5.29794s/12 iters), loss = 4.88964 I0406 07:14:38.829366 5226 solver.cpp:237] Train net output #0: loss = 4.88964 (* 1 = 4.88964 loss) I0406 07:14:38.829372 5226 sgd_solver.cpp:105] Iteration 804, lr = 0.01 I0406 07:14:40.611006 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:14:43.559803 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_816.caffemodel I0406 07:14:46.568624 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_816.solverstate I0406 07:14:48.891126 5226 solver.cpp:330] Iteration 816, Testing net (#0) I0406 07:14:48.891149 5226 net.cpp:676] Ignoring source layer train-data I0406 07:14:52.916862 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:14:53.263047 5226 solver.cpp:397] Test net output #0: accuracy = 0.0471814 I0406 07:14:53.263080 5226 solver.cpp:397] Test net output #1: loss = 4.81925 (* 1 = 4.81925 loss) I0406 07:14:53.401698 5226 solver.cpp:218] Iteration 816 (0.823484 iter/s, 14.5722s/12 iters), loss = 4.82417 I0406 07:14:53.403260 5226 solver.cpp:237] Train net output #0: loss = 4.82417 (* 1 = 4.82417 loss) I0406 07:14:53.403273 5226 sgd_solver.cpp:105] Iteration 816, lr = 0.01 I0406 07:14:57.870254 5226 solver.cpp:218] Iteration 828 (2.68639 iter/s, 4.46696s/12 iters), loss = 4.73788 I0406 07:14:57.870301 5226 solver.cpp:237] Train net output #0: loss = 4.73788 (* 1 = 4.73788 loss) I0406 07:14:57.870309 5226 sgd_solver.cpp:105] Iteration 828, lr = 0.01 I0406 07:15:03.000788 5226 solver.cpp:218] Iteration 840 (2.33898 iter/s, 5.13044s/12 iters), loss = 4.64403 I0406 07:15:03.000824 5226 solver.cpp:237] Train net output #0: loss = 4.64403 (* 1 = 4.64403 loss) I0406 07:15:03.000829 5226 sgd_solver.cpp:105] Iteration 840, lr = 0.01 I0406 07:15:08.395009 5226 solver.cpp:218] Iteration 852 (2.22464 iter/s, 5.39413s/12 iters), loss = 4.7397 I0406 07:15:08.395047 5226 solver.cpp:237] Train net output #0: loss = 4.7397 (* 1 = 4.7397 loss) I0406 07:15:08.395053 5226 sgd_solver.cpp:105] Iteration 852, lr = 0.01 I0406 07:15:13.831234 5226 solver.cpp:218] Iteration 864 (2.20745 iter/s, 5.43613s/12 iters), loss = 4.67168 I0406 07:15:13.831343 5226 solver.cpp:237] Train net output #0: loss = 4.67168 (* 1 = 4.67168 loss) I0406 07:15:13.831351 5226 sgd_solver.cpp:105] Iteration 864, lr = 0.01 I0406 07:15:19.099928 5226 solver.cpp:218] Iteration 876 (2.27767 iter/s, 5.26854s/12 iters), loss = 4.74518 I0406 07:15:19.099965 5226 solver.cpp:237] Train net output #0: loss = 4.74518 (* 1 = 4.74518 loss) I0406 07:15:19.099970 5226 sgd_solver.cpp:105] Iteration 876, lr = 0.01 I0406 07:15:24.342908 5226 solver.cpp:218] Iteration 888 (2.28881 iter/s, 5.24289s/12 iters), loss = 4.78204 I0406 07:15:24.342947 5226 solver.cpp:237] Train net output #0: loss = 4.78204 (* 1 = 4.78204 loss) I0406 07:15:24.342952 5226 sgd_solver.cpp:105] Iteration 888, lr = 0.01 I0406 07:15:29.420219 5226 solver.cpp:218] Iteration 900 (2.36349 iter/s, 5.07723s/12 iters), loss = 4.7116 I0406 07:15:29.420266 5226 solver.cpp:237] Train net output #0: loss = 4.7116 (* 1 = 4.7116 loss) I0406 07:15:29.420274 5226 sgd_solver.cpp:105] Iteration 900, lr = 0.01 I0406 07:15:33.550194 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:15:34.800318 5226 solver.cpp:218] Iteration 912 (2.23048 iter/s, 5.38001s/12 iters), loss = 4.67847 I0406 07:15:34.800351 5226 solver.cpp:237] Train net output #0: loss = 4.67847 (* 1 = 4.67847 loss) I0406 07:15:34.800356 5226 sgd_solver.cpp:105] Iteration 912, lr = 0.01 I0406 07:15:36.759296 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_918.caffemodel I0406 07:15:39.769183 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_918.solverstate I0406 07:15:42.064821 5226 solver.cpp:330] Iteration 918, Testing net (#0) I0406 07:15:42.064839 5226 net.cpp:676] Ignoring source layer train-data I0406 07:15:45.926584 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:15:46.319641 5226 solver.cpp:397] Test net output #0: accuracy = 0.0435049 I0406 07:15:46.319679 5226 solver.cpp:397] Test net output #1: loss = 4.65492 (* 1 = 4.65492 loss) I0406 07:15:48.186735 5226 solver.cpp:218] Iteration 924 (0.89644 iter/s, 13.3863s/12 iters), loss = 4.57518 I0406 07:15:48.186770 5226 solver.cpp:237] Train net output #0: loss = 4.57518 (* 1 = 4.57518 loss) I0406 07:15:48.186776 5226 sgd_solver.cpp:105] Iteration 924, lr = 0.01 I0406 07:15:53.646442 5226 solver.cpp:218] Iteration 936 (2.19796 iter/s, 5.45962s/12 iters), loss = 4.62903 I0406 07:15:53.646489 5226 solver.cpp:237] Train net output #0: loss = 4.62903 (* 1 = 4.62903 loss) I0406 07:15:53.646497 5226 sgd_solver.cpp:105] Iteration 936, lr = 0.01 I0406 07:15:58.949519 5226 solver.cpp:218] Iteration 948 (2.26288 iter/s, 5.30298s/12 iters), loss = 4.65031 I0406 07:15:58.949558 5226 solver.cpp:237] Train net output #0: loss = 4.65031 (* 1 = 4.65031 loss) I0406 07:15:58.949563 5226 sgd_solver.cpp:105] Iteration 948, lr = 0.01 I0406 07:16:04.350941 5226 solver.cpp:218] Iteration 960 (2.22167 iter/s, 5.40133s/12 iters), loss = 4.67815 I0406 07:16:04.350975 5226 solver.cpp:237] Train net output #0: loss = 4.67815 (* 1 = 4.67815 loss) I0406 07:16:04.350980 5226 sgd_solver.cpp:105] Iteration 960, lr = 0.01 I0406 07:16:09.718039 5226 solver.cpp:218] Iteration 972 (2.23588 iter/s, 5.36701s/12 iters), loss = 4.47251 I0406 07:16:09.718076 5226 solver.cpp:237] Train net output #0: loss = 4.47251 (* 1 = 4.47251 loss) I0406 07:16:09.718082 5226 sgd_solver.cpp:105] Iteration 972, lr = 0.01 I0406 07:16:15.170527 5226 solver.cpp:218] Iteration 984 (2.20087 iter/s, 5.4524s/12 iters), loss = 4.48967 I0406 07:16:15.170567 5226 solver.cpp:237] Train net output #0: loss = 4.48967 (* 1 = 4.48967 loss) I0406 07:16:15.170573 5226 sgd_solver.cpp:105] Iteration 984, lr = 0.01 I0406 07:16:20.369804 5226 solver.cpp:218] Iteration 996 (2.30805 iter/s, 5.19919s/12 iters), loss = 4.51546 I0406 07:16:20.369899 5226 solver.cpp:237] Train net output #0: loss = 4.51546 (* 1 = 4.51546 loss) I0406 07:16:20.369906 5226 sgd_solver.cpp:105] Iteration 996, lr = 0.01 I0406 07:16:25.632683 5226 solver.cpp:218] Iteration 1008 (2.28018 iter/s, 5.26273s/12 iters), loss = 4.56256 I0406 07:16:25.632732 5226 solver.cpp:237] Train net output #0: loss = 4.56256 (* 1 = 4.56256 loss) I0406 07:16:25.632741 5226 sgd_solver.cpp:105] Iteration 1008, lr = 0.01 I0406 07:16:26.735141 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:16:30.381785 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1020.caffemodel I0406 07:16:33.392508 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1020.solverstate I0406 07:16:35.701859 5226 solver.cpp:330] Iteration 1020, Testing net (#0) I0406 07:16:35.701879 5226 net.cpp:676] Ignoring source layer train-data I0406 07:16:39.570278 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:16:39.992810 5226 solver.cpp:397] Test net output #0: accuracy = 0.0606618 I0406 07:16:39.992843 5226 solver.cpp:397] Test net output #1: loss = 4.50936 (* 1 = 4.50936 loss) I0406 07:16:40.129634 5226 solver.cpp:218] Iteration 1020 (0.827769 iter/s, 14.4968s/12 iters), loss = 4.46469 I0406 07:16:40.129676 5226 solver.cpp:237] Train net output #0: loss = 4.46469 (* 1 = 4.46469 loss) I0406 07:16:40.129683 5226 sgd_solver.cpp:105] Iteration 1020, lr = 0.01 I0406 07:16:44.591004 5226 solver.cpp:218] Iteration 1032 (2.68981 iter/s, 4.46129s/12 iters), loss = 4.75816 I0406 07:16:44.591042 5226 solver.cpp:237] Train net output #0: loss = 4.75816 (* 1 = 4.75816 loss) I0406 07:16:44.591048 5226 sgd_solver.cpp:105] Iteration 1032, lr = 0.01 I0406 07:16:49.912254 5226 solver.cpp:218] Iteration 1044 (2.25515 iter/s, 5.32116s/12 iters), loss = 4.60476 I0406 07:16:49.912292 5226 solver.cpp:237] Train net output #0: loss = 4.60476 (* 1 = 4.60476 loss) I0406 07:16:49.912297 5226 sgd_solver.cpp:105] Iteration 1044, lr = 0.01 I0406 07:16:55.320588 5226 solver.cpp:218] Iteration 1056 (2.21883 iter/s, 5.40825s/12 iters), loss = 4.46903 I0406 07:16:55.320719 5226 solver.cpp:237] Train net output #0: loss = 4.46903 (* 1 = 4.46903 loss) I0406 07:16:55.320724 5226 sgd_solver.cpp:105] Iteration 1056, lr = 0.01 I0406 07:17:00.592834 5226 solver.cpp:218] Iteration 1068 (2.27615 iter/s, 5.27207s/12 iters), loss = 4.37665 I0406 07:17:00.592880 5226 solver.cpp:237] Train net output #0: loss = 4.37665 (* 1 = 4.37665 loss) I0406 07:17:00.592893 5226 sgd_solver.cpp:105] Iteration 1068, lr = 0.01 I0406 07:17:05.919016 5226 solver.cpp:218] Iteration 1080 (2.25306 iter/s, 5.32609s/12 iters), loss = 4.16893 I0406 07:17:05.919055 5226 solver.cpp:237] Train net output #0: loss = 4.16893 (* 1 = 4.16893 loss) I0406 07:17:05.919060 5226 sgd_solver.cpp:105] Iteration 1080, lr = 0.01 I0406 07:17:11.199740 5226 solver.cpp:218] Iteration 1092 (2.27245 iter/s, 5.28064s/12 iters), loss = 4.47528 I0406 07:17:11.199777 5226 solver.cpp:237] Train net output #0: loss = 4.47528 (* 1 = 4.47528 loss) I0406 07:17:11.199782 5226 sgd_solver.cpp:105] Iteration 1092, lr = 0.01 I0406 07:17:16.439673 5226 solver.cpp:218] Iteration 1104 (2.29014 iter/s, 5.23985s/12 iters), loss = 4.36203 I0406 07:17:16.439711 5226 solver.cpp:237] Train net output #0: loss = 4.36203 (* 1 = 4.36203 loss) I0406 07:17:16.439716 5226 sgd_solver.cpp:105] Iteration 1104, lr = 0.01 I0406 07:17:19.734458 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:17:21.766516 5226 solver.cpp:218] Iteration 1116 (2.25278 iter/s, 5.32676s/12 iters), loss = 4.36978 I0406 07:17:21.766553 5226 solver.cpp:237] Train net output #0: loss = 4.36978 (* 1 = 4.36978 loss) I0406 07:17:21.766558 5226 sgd_solver.cpp:105] Iteration 1116, lr = 0.01 I0406 07:17:23.851191 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1122.caffemodel I0406 07:17:26.857003 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1122.solverstate I0406 07:17:29.156157 5226 solver.cpp:330] Iteration 1122, Testing net (#0) I0406 07:17:29.156175 5226 net.cpp:676] Ignoring source layer train-data I0406 07:17:32.953393 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:17:33.422466 5226 solver.cpp:397] Test net output #0: accuracy = 0.09375 I0406 07:17:33.422498 5226 solver.cpp:397] Test net output #1: loss = 4.32848 (* 1 = 4.32848 loss) I0406 07:17:35.250597 5226 solver.cpp:218] Iteration 1128 (0.889947 iter/s, 13.4839s/12 iters), loss = 4.14385 I0406 07:17:35.250631 5226 solver.cpp:237] Train net output #0: loss = 4.14385 (* 1 = 4.14385 loss) I0406 07:17:35.250636 5226 sgd_solver.cpp:105] Iteration 1128, lr = 0.01 I0406 07:17:40.446559 5226 solver.cpp:218] Iteration 1140 (2.30952 iter/s, 5.19588s/12 iters), loss = 4.32845 I0406 07:17:40.446599 5226 solver.cpp:237] Train net output #0: loss = 4.32845 (* 1 = 4.32845 loss) I0406 07:17:40.446605 5226 sgd_solver.cpp:105] Iteration 1140, lr = 0.01 I0406 07:17:45.603284 5226 solver.cpp:218] Iteration 1152 (2.3271 iter/s, 5.15664s/12 iters), loss = 4.45601 I0406 07:17:45.603320 5226 solver.cpp:237] Train net output #0: loss = 4.45601 (* 1 = 4.45601 loss) I0406 07:17:45.603327 5226 sgd_solver.cpp:105] Iteration 1152, lr = 0.01 I0406 07:17:50.869220 5226 solver.cpp:218] Iteration 1164 (2.27883 iter/s, 5.26585s/12 iters), loss = 4.39584 I0406 07:17:50.869256 5226 solver.cpp:237] Train net output #0: loss = 4.39584 (* 1 = 4.39584 loss) I0406 07:17:50.869261 5226 sgd_solver.cpp:105] Iteration 1164, lr = 0.01 I0406 07:17:56.115993 5226 solver.cpp:218] Iteration 1176 (2.28716 iter/s, 5.24668s/12 iters), loss = 4.23961 I0406 07:17:56.116034 5226 solver.cpp:237] Train net output #0: loss = 4.23961 (* 1 = 4.23961 loss) I0406 07:17:56.116039 5226 sgd_solver.cpp:105] Iteration 1176, lr = 0.01 I0406 07:18:01.324441 5226 solver.cpp:218] Iteration 1188 (2.30399 iter/s, 5.20836s/12 iters), loss = 4.30155 I0406 07:18:01.324575 5226 solver.cpp:237] Train net output #0: loss = 4.30155 (* 1 = 4.30155 loss) I0406 07:18:01.324582 5226 sgd_solver.cpp:105] Iteration 1188, lr = 0.01 I0406 07:18:06.534363 5226 solver.cpp:218] Iteration 1200 (2.30338 iter/s, 5.20974s/12 iters), loss = 4.11412 I0406 07:18:06.534400 5226 solver.cpp:237] Train net output #0: loss = 4.11412 (* 1 = 4.11412 loss) I0406 07:18:06.534406 5226 sgd_solver.cpp:105] Iteration 1200, lr = 0.01 I0406 07:18:11.857005 5226 solver.cpp:218] Iteration 1212 (2.25456 iter/s, 5.32255s/12 iters), loss = 4.28069 I0406 07:18:11.857041 5226 solver.cpp:237] Train net output #0: loss = 4.28069 (* 1 = 4.28069 loss) I0406 07:18:11.857048 5226 sgd_solver.cpp:105] Iteration 1212, lr = 0.01 I0406 07:18:12.123777 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:18:16.567703 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1224.caffemodel I0406 07:18:20.569676 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1224.solverstate I0406 07:18:22.869602 5226 solver.cpp:330] Iteration 1224, Testing net (#0) I0406 07:18:22.869621 5226 net.cpp:676] Ignoring source layer train-data I0406 07:18:26.712373 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:18:27.216188 5226 solver.cpp:397] Test net output #0: accuracy = 0.0968137 I0406 07:18:27.216219 5226 solver.cpp:397] Test net output #1: loss = 4.26293 (* 1 = 4.26293 loss) I0406 07:18:27.353430 5226 solver.cpp:218] Iteration 1224 (0.774379 iter/s, 15.4963s/12 iters), loss = 4.17032 I0406 07:18:27.353466 5226 solver.cpp:237] Train net output #0: loss = 4.17032 (* 1 = 4.17032 loss) I0406 07:18:27.353471 5226 sgd_solver.cpp:105] Iteration 1224, lr = 0.01 I0406 07:18:31.746218 5226 solver.cpp:218] Iteration 1236 (2.7318 iter/s, 4.39271s/12 iters), loss = 4.337 I0406 07:18:31.746299 5226 solver.cpp:237] Train net output #0: loss = 4.337 (* 1 = 4.337 loss) I0406 07:18:31.746305 5226 sgd_solver.cpp:105] Iteration 1236, lr = 0.01 I0406 07:18:37.014269 5226 solver.cpp:218] Iteration 1248 (2.27794 iter/s, 5.26792s/12 iters), loss = 4.01575 I0406 07:18:37.014307 5226 solver.cpp:237] Train net output #0: loss = 4.01575 (* 1 = 4.01575 loss) I0406 07:18:37.014312 5226 sgd_solver.cpp:105] Iteration 1248, lr = 0.01 I0406 07:18:42.187361 5226 solver.cpp:218] Iteration 1260 (2.31974 iter/s, 5.173s/12 iters), loss = 4.23174 I0406 07:18:42.187414 5226 solver.cpp:237] Train net output #0: loss = 4.23174 (* 1 = 4.23174 loss) I0406 07:18:42.187423 5226 sgd_solver.cpp:105] Iteration 1260, lr = 0.01 I0406 07:18:47.607379 5226 solver.cpp:218] Iteration 1272 (2.21406 iter/s, 5.41992s/12 iters), loss = 4.21252 I0406 07:18:47.607419 5226 solver.cpp:237] Train net output #0: loss = 4.21252 (* 1 = 4.21252 loss) I0406 07:18:47.607425 5226 sgd_solver.cpp:105] Iteration 1272, lr = 0.01 I0406 07:18:52.843992 5226 solver.cpp:218] Iteration 1284 (2.2916 iter/s, 5.23652s/12 iters), loss = 4.10824 I0406 07:18:52.844030 5226 solver.cpp:237] Train net output #0: loss = 4.10824 (* 1 = 4.10824 loss) I0406 07:18:52.844036 5226 sgd_solver.cpp:105] Iteration 1284, lr = 0.01 I0406 07:18:58.077533 5226 solver.cpp:218] Iteration 1296 (2.29294 iter/s, 5.23346s/12 iters), loss = 4.13007 I0406 07:18:58.077571 5226 solver.cpp:237] Train net output #0: loss = 4.13007 (* 1 = 4.13007 loss) I0406 07:18:58.077577 5226 sgd_solver.cpp:105] Iteration 1296, lr = 0.01 I0406 07:19:03.148684 5226 solver.cpp:218] Iteration 1308 (2.36637 iter/s, 5.07107s/12 iters), loss = 4.24438 I0406 07:19:03.148825 5226 solver.cpp:237] Train net output #0: loss = 4.24438 (* 1 = 4.24438 loss) I0406 07:19:03.148831 5226 sgd_solver.cpp:105] Iteration 1308, lr = 0.01 I0406 07:19:05.525895 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:19:08.181069 5226 solver.cpp:218] Iteration 1320 (2.38464 iter/s, 5.0322s/12 iters), loss = 4.06937 I0406 07:19:08.181105 5226 solver.cpp:237] Train net output #0: loss = 4.06937 (* 1 = 4.06937 loss) I0406 07:19:08.181110 5226 sgd_solver.cpp:105] Iteration 1320, lr = 0.01 I0406 07:19:10.296854 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1326.caffemodel I0406 07:19:13.301491 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1326.solverstate I0406 07:19:15.619369 5226 solver.cpp:330] Iteration 1326, Testing net (#0) I0406 07:19:15.619387 5226 net.cpp:676] Ignoring source layer train-data I0406 07:19:19.403321 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:19:19.952392 5226 solver.cpp:397] Test net output #0: accuracy = 0.107843 I0406 07:19:19.952419 5226 solver.cpp:397] Test net output #1: loss = 4.07016 (* 1 = 4.07016 loss) I0406 07:19:21.803865 5226 solver.cpp:218] Iteration 1332 (0.880885 iter/s, 13.6227s/12 iters), loss = 3.95342 I0406 07:19:21.803905 5226 solver.cpp:237] Train net output #0: loss = 3.95342 (* 1 = 3.95342 loss) I0406 07:19:21.803910 5226 sgd_solver.cpp:105] Iteration 1332, lr = 0.01 I0406 07:19:27.465436 5226 solver.cpp:218] Iteration 1344 (2.11959 iter/s, 5.66148s/12 iters), loss = 4.07274 I0406 07:19:27.465473 5226 solver.cpp:237] Train net output #0: loss = 4.07274 (* 1 = 4.07274 loss) I0406 07:19:27.465478 5226 sgd_solver.cpp:105] Iteration 1344, lr = 0.01 I0406 07:19:32.821652 5226 solver.cpp:218] Iteration 1356 (2.24042 iter/s, 5.35613s/12 iters), loss = 3.89633 I0406 07:19:32.821687 5226 solver.cpp:237] Train net output #0: loss = 3.89633 (* 1 = 3.89633 loss) I0406 07:19:32.821692 5226 sgd_solver.cpp:105] Iteration 1356, lr = 0.01 I0406 07:19:38.148391 5226 solver.cpp:218] Iteration 1368 (2.25282 iter/s, 5.32666s/12 iters), loss = 3.79107 I0406 07:19:38.148489 5226 solver.cpp:237] Train net output #0: loss = 3.79107 (* 1 = 3.79107 loss) I0406 07:19:38.148496 5226 sgd_solver.cpp:105] Iteration 1368, lr = 0.01 I0406 07:19:39.411227 5226 blocking_queue.cpp:49] Waiting for data I0406 07:19:43.461465 5226 solver.cpp:218] Iteration 1380 (2.25864 iter/s, 5.31292s/12 iters), loss = 3.95237 I0406 07:19:43.461511 5226 solver.cpp:237] Train net output #0: loss = 3.95237 (* 1 = 3.95237 loss) I0406 07:19:43.461519 5226 sgd_solver.cpp:105] Iteration 1380, lr = 0.01 I0406 07:19:48.507176 5226 solver.cpp:218] Iteration 1392 (2.3783 iter/s, 5.04562s/12 iters), loss = 3.79814 I0406 07:19:48.507210 5226 solver.cpp:237] Train net output #0: loss = 3.79814 (* 1 = 3.79814 loss) I0406 07:19:48.507215 5226 sgd_solver.cpp:105] Iteration 1392, lr = 0.01 I0406 07:19:53.712174 5226 solver.cpp:218] Iteration 1404 (2.30551 iter/s, 5.20492s/12 iters), loss = 3.86488 I0406 07:19:53.712210 5226 solver.cpp:237] Train net output #0: loss = 3.86488 (* 1 = 3.86488 loss) I0406 07:19:53.712215 5226 sgd_solver.cpp:105] Iteration 1404, lr = 0.01 I0406 07:19:58.415917 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:19:58.799854 5226 solver.cpp:218] Iteration 1416 (2.35868 iter/s, 5.0876s/12 iters), loss = 4.04274 I0406 07:19:58.799891 5226 solver.cpp:237] Train net output #0: loss = 4.04274 (* 1 = 4.04274 loss) I0406 07:19:58.799897 5226 sgd_solver.cpp:105] Iteration 1416, lr = 0.01 I0406 07:20:03.587568 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1428.caffemodel I0406 07:20:06.554714 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1428.solverstate I0406 07:20:08.849576 5226 solver.cpp:330] Iteration 1428, Testing net (#0) I0406 07:20:08.849684 5226 net.cpp:676] Ignoring source layer train-data I0406 07:20:12.575773 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:20:13.159407 5226 solver.cpp:397] Test net output #0: accuracy = 0.136642 I0406 07:20:13.159436 5226 solver.cpp:397] Test net output #1: loss = 3.89282 (* 1 = 3.89282 loss) I0406 07:20:13.300719 5226 solver.cpp:218] Iteration 1428 (0.827545 iter/s, 14.5007s/12 iters), loss = 4.00985 I0406 07:20:13.300758 5226 solver.cpp:237] Train net output #0: loss = 4.00985 (* 1 = 4.00985 loss) I0406 07:20:13.300763 5226 sgd_solver.cpp:105] Iteration 1428, lr = 0.01 I0406 07:20:17.768146 5226 solver.cpp:218] Iteration 1440 (2.68616 iter/s, 4.46735s/12 iters), loss = 3.82675 I0406 07:20:17.768182 5226 solver.cpp:237] Train net output #0: loss = 3.82675 (* 1 = 3.82675 loss) I0406 07:20:17.768187 5226 sgd_solver.cpp:105] Iteration 1440, lr = 0.01 I0406 07:20:23.102926 5226 solver.cpp:218] Iteration 1452 (2.24943 iter/s, 5.33469s/12 iters), loss = 3.65307 I0406 07:20:23.102962 5226 solver.cpp:237] Train net output #0: loss = 3.65307 (* 1 = 3.65307 loss) I0406 07:20:23.102967 5226 sgd_solver.cpp:105] Iteration 1452, lr = 0.01 I0406 07:20:28.438781 5226 solver.cpp:218] Iteration 1464 (2.24897 iter/s, 5.33577s/12 iters), loss = 3.64242 I0406 07:20:28.438835 5226 solver.cpp:237] Train net output #0: loss = 3.64242 (* 1 = 3.64242 loss) I0406 07:20:28.438844 5226 sgd_solver.cpp:105] Iteration 1464, lr = 0.01 I0406 07:20:33.709247 5226 solver.cpp:218] Iteration 1476 (2.27688 iter/s, 5.27036s/12 iters), loss = 3.78536 I0406 07:20:33.709285 5226 solver.cpp:237] Train net output #0: loss = 3.78536 (* 1 = 3.78536 loss) I0406 07:20:33.709291 5226 sgd_solver.cpp:105] Iteration 1476, lr = 0.01 I0406 07:20:38.739935 5226 solver.cpp:218] Iteration 1488 (2.3854 iter/s, 5.03061s/12 iters), loss = 3.87757 I0406 07:20:38.739976 5226 solver.cpp:237] Train net output #0: loss = 3.87757 (* 1 = 3.87757 loss) I0406 07:20:38.739982 5226 sgd_solver.cpp:105] Iteration 1488, lr = 0.01 I0406 07:20:44.015082 5226 solver.cpp:218] Iteration 1500 (2.27486 iter/s, 5.27506s/12 iters), loss = 3.84732 I0406 07:20:44.015189 5226 solver.cpp:237] Train net output #0: loss = 3.84732 (* 1 = 3.84732 loss) I0406 07:20:44.015195 5226 sgd_solver.cpp:105] Iteration 1500, lr = 0.01 I0406 07:20:49.435204 5226 solver.cpp:218] Iteration 1512 (2.21404 iter/s, 5.41996s/12 iters), loss = 3.61355 I0406 07:20:49.435242 5226 solver.cpp:237] Train net output #0: loss = 3.61355 (* 1 = 3.61355 loss) I0406 07:20:49.435247 5226 sgd_solver.cpp:105] Iteration 1512, lr = 0.01 I0406 07:20:51.351210 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:20:54.696135 5226 solver.cpp:218] Iteration 1524 (2.281 iter/s, 5.26084s/12 iters), loss = 3.61983 I0406 07:20:54.696171 5226 solver.cpp:237] Train net output #0: loss = 3.61983 (* 1 = 3.61983 loss) I0406 07:20:54.696177 5226 sgd_solver.cpp:105] Iteration 1524, lr = 0.01 I0406 07:20:56.702572 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1530.caffemodel I0406 07:20:59.719856 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1530.solverstate I0406 07:21:02.838994 5226 solver.cpp:330] Iteration 1530, Testing net (#0) I0406 07:21:02.839013 5226 net.cpp:676] Ignoring source layer train-data I0406 07:21:06.563899 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:21:07.211084 5226 solver.cpp:397] Test net output #0: accuracy = 0.137255 I0406 07:21:07.211119 5226 solver.cpp:397] Test net output #1: loss = 3.81117 (* 1 = 3.81117 loss) I0406 07:21:08.951932 5226 solver.cpp:218] Iteration 1536 (0.841771 iter/s, 14.2557s/12 iters), loss = 3.3157 I0406 07:21:08.951975 5226 solver.cpp:237] Train net output #0: loss = 3.3157 (* 1 = 3.3157 loss) I0406 07:21:08.951982 5226 sgd_solver.cpp:105] Iteration 1536, lr = 0.01 I0406 07:21:14.235147 5226 solver.cpp:218] Iteration 1548 (2.27138 iter/s, 5.28312s/12 iters), loss = 3.61828 I0406 07:21:14.235270 5226 solver.cpp:237] Train net output #0: loss = 3.61828 (* 1 = 3.61828 loss) I0406 07:21:14.235276 5226 sgd_solver.cpp:105] Iteration 1548, lr = 0.01 I0406 07:21:19.633162 5226 solver.cpp:218] Iteration 1560 (2.22311 iter/s, 5.39785s/12 iters), loss = 3.59642 I0406 07:21:19.633196 5226 solver.cpp:237] Train net output #0: loss = 3.59642 (* 1 = 3.59642 loss) I0406 07:21:19.633203 5226 sgd_solver.cpp:105] Iteration 1560, lr = 0.01 I0406 07:21:24.682137 5226 solver.cpp:218] Iteration 1572 (2.37676 iter/s, 5.04889s/12 iters), loss = 3.60166 I0406 07:21:24.682173 5226 solver.cpp:237] Train net output #0: loss = 3.60166 (* 1 = 3.60166 loss) I0406 07:21:24.682179 5226 sgd_solver.cpp:105] Iteration 1572, lr = 0.01 I0406 07:21:29.874276 5226 solver.cpp:218] Iteration 1584 (2.31122 iter/s, 5.19205s/12 iters), loss = 3.56729 I0406 07:21:29.874315 5226 solver.cpp:237] Train net output #0: loss = 3.56729 (* 1 = 3.56729 loss) I0406 07:21:29.874320 5226 sgd_solver.cpp:105] Iteration 1584, lr = 0.01 I0406 07:21:35.271520 5226 solver.cpp:218] Iteration 1596 (2.2234 iter/s, 5.39715s/12 iters), loss = 3.52012 I0406 07:21:35.271575 5226 solver.cpp:237] Train net output #0: loss = 3.52012 (* 1 = 3.52012 loss) I0406 07:21:35.271584 5226 sgd_solver.cpp:105] Iteration 1596, lr = 0.01 I0406 07:21:40.616628 5226 solver.cpp:218] Iteration 1608 (2.24509 iter/s, 5.34501s/12 iters), loss = 3.19874 I0406 07:21:40.616667 5226 solver.cpp:237] Train net output #0: loss = 3.19874 (* 1 = 3.19874 loss) I0406 07:21:40.616672 5226 sgd_solver.cpp:105] Iteration 1608, lr = 0.01 I0406 07:21:44.679390 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:21:45.899791 5226 solver.cpp:218] Iteration 1620 (2.27141 iter/s, 5.28307s/12 iters), loss = 3.49699 I0406 07:21:45.899833 5226 solver.cpp:237] Train net output #0: loss = 3.49699 (* 1 = 3.49699 loss) I0406 07:21:45.899839 5226 sgd_solver.cpp:105] Iteration 1620, lr = 0.01 I0406 07:21:50.518654 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1632.caffemodel I0406 07:21:53.557855 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1632.solverstate I0406 07:21:55.876688 5226 solver.cpp:330] Iteration 1632, Testing net (#0) I0406 07:21:55.876708 5226 net.cpp:676] Ignoring source layer train-data I0406 07:21:59.634577 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:22:00.334913 5226 solver.cpp:397] Test net output #0: accuracy = 0.159926 I0406 07:22:00.334941 5226 solver.cpp:397] Test net output #1: loss = 3.62449 (* 1 = 3.62449 loss) I0406 07:22:00.469847 5226 solver.cpp:218] Iteration 1632 (0.823615 iter/s, 14.5699s/12 iters), loss = 3.51187 I0406 07:22:00.469897 5226 solver.cpp:237] Train net output #0: loss = 3.51187 (* 1 = 3.51187 loss) I0406 07:22:00.469905 5226 sgd_solver.cpp:105] Iteration 1632, lr = 0.01 I0406 07:22:04.779532 5226 solver.cpp:218] Iteration 1644 (2.78449 iter/s, 4.30959s/12 iters), loss = 3.80759 I0406 07:22:04.779585 5226 solver.cpp:237] Train net output #0: loss = 3.80759 (* 1 = 3.80759 loss) I0406 07:22:04.779595 5226 sgd_solver.cpp:105] Iteration 1644, lr = 0.01 I0406 07:22:10.212400 5226 solver.cpp:218] Iteration 1656 (2.20882 iter/s, 5.43276s/12 iters), loss = 3.4897 I0406 07:22:10.212438 5226 solver.cpp:237] Train net output #0: loss = 3.4897 (* 1 = 3.4897 loss) I0406 07:22:10.212445 5226 sgd_solver.cpp:105] Iteration 1656, lr = 0.01 I0406 07:22:15.386605 5226 solver.cpp:218] Iteration 1668 (2.31924 iter/s, 5.17412s/12 iters), loss = 3.34191 I0406 07:22:15.386731 5226 solver.cpp:237] Train net output #0: loss = 3.34191 (* 1 = 3.34191 loss) I0406 07:22:15.386739 5226 sgd_solver.cpp:105] Iteration 1668, lr = 0.01 I0406 07:22:20.414834 5226 solver.cpp:218] Iteration 1680 (2.38661 iter/s, 5.02806s/12 iters), loss = 3.44416 I0406 07:22:20.414873 5226 solver.cpp:237] Train net output #0: loss = 3.44416 (* 1 = 3.44416 loss) I0406 07:22:20.414880 5226 sgd_solver.cpp:105] Iteration 1680, lr = 0.01 I0406 07:22:25.622228 5226 solver.cpp:218] Iteration 1692 (2.30446 iter/s, 5.2073s/12 iters), loss = 3.32771 I0406 07:22:25.622267 5226 solver.cpp:237] Train net output #0: loss = 3.32771 (* 1 = 3.32771 loss) I0406 07:22:25.622272 5226 sgd_solver.cpp:105] Iteration 1692, lr = 0.01 I0406 07:22:30.817466 5226 solver.cpp:218] Iteration 1704 (2.30985 iter/s, 5.19515s/12 iters), loss = 3.08907 I0406 07:22:30.817505 5226 solver.cpp:237] Train net output #0: loss = 3.08907 (* 1 = 3.08907 loss) I0406 07:22:30.817510 5226 sgd_solver.cpp:105] Iteration 1704, lr = 0.01 I0406 07:22:36.112320 5226 solver.cpp:218] Iteration 1716 (2.26639 iter/s, 5.29477s/12 iters), loss = 3.45159 I0406 07:22:36.112361 5226 solver.cpp:237] Train net output #0: loss = 3.45159 (* 1 = 3.45159 loss) I0406 07:22:36.112366 5226 sgd_solver.cpp:105] Iteration 1716, lr = 0.01 I0406 07:22:37.144914 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:22:41.211019 5226 solver.cpp:218] Iteration 1728 (2.35358 iter/s, 5.09861s/12 iters), loss = 3.23265 I0406 07:22:41.211058 5226 solver.cpp:237] Train net output #0: loss = 3.23265 (* 1 = 3.23265 loss) I0406 07:22:41.211064 5226 sgd_solver.cpp:105] Iteration 1728, lr = 0.01 I0406 07:22:43.223479 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1734.caffemodel I0406 07:22:46.239655 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1734.solverstate I0406 07:22:48.545070 5226 solver.cpp:330] Iteration 1734, Testing net (#0) I0406 07:22:48.545089 5226 net.cpp:676] Ignoring source layer train-data I0406 07:22:52.158030 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:22:52.851001 5226 solver.cpp:397] Test net output #0: accuracy = 0.166054 I0406 07:22:52.851029 5226 solver.cpp:397] Test net output #1: loss = 3.62371 (* 1 = 3.62371 loss) I0406 07:22:54.710803 5226 solver.cpp:218] Iteration 1740 (0.888912 iter/s, 13.4996s/12 iters), loss = 3.39027 I0406 07:22:54.710841 5226 solver.cpp:237] Train net output #0: loss = 3.39027 (* 1 = 3.39027 loss) I0406 07:22:54.710846 5226 sgd_solver.cpp:105] Iteration 1740, lr = 0.01 I0406 07:23:00.007306 5226 solver.cpp:218] Iteration 1752 (2.26568 iter/s, 5.29642s/12 iters), loss = 3.40304 I0406 07:23:00.007344 5226 solver.cpp:237] Train net output #0: loss = 3.40304 (* 1 = 3.40304 loss) I0406 07:23:00.007349 5226 sgd_solver.cpp:105] Iteration 1752, lr = 0.01 I0406 07:23:05.110262 5226 solver.cpp:218] Iteration 1764 (2.35162 iter/s, 5.10286s/12 iters), loss = 3.41792 I0406 07:23:05.110311 5226 solver.cpp:237] Train net output #0: loss = 3.41792 (* 1 = 3.41792 loss) I0406 07:23:05.110318 5226 sgd_solver.cpp:105] Iteration 1764, lr = 0.01 I0406 07:23:10.326303 5226 solver.cpp:218] Iteration 1776 (2.30064 iter/s, 5.21595s/12 iters), loss = 3.0336 I0406 07:23:10.326349 5226 solver.cpp:237] Train net output #0: loss = 3.0336 (* 1 = 3.0336 loss) I0406 07:23:10.326356 5226 sgd_solver.cpp:105] Iteration 1776, lr = 0.01 I0406 07:23:15.651890 5226 solver.cpp:218] Iteration 1788 (2.25331 iter/s, 5.32549s/12 iters), loss = 2.82969 I0406 07:23:15.651929 5226 solver.cpp:237] Train net output #0: loss = 2.82969 (* 1 = 2.82969 loss) I0406 07:23:15.651935 5226 sgd_solver.cpp:105] Iteration 1788, lr = 0.01 I0406 07:23:20.944195 5226 solver.cpp:218] Iteration 1800 (2.26748 iter/s, 5.29222s/12 iters), loss = 3.18035 I0406 07:23:20.944288 5226 solver.cpp:237] Train net output #0: loss = 3.18035 (* 1 = 3.18035 loss) I0406 07:23:20.944294 5226 sgd_solver.cpp:105] Iteration 1800, lr = 0.01 I0406 07:23:26.238842 5226 solver.cpp:218] Iteration 1812 (2.2665 iter/s, 5.29451s/12 iters), loss = 3.10267 I0406 07:23:26.238878 5226 solver.cpp:237] Train net output #0: loss = 3.10267 (* 1 = 3.10267 loss) I0406 07:23:26.238883 5226 sgd_solver.cpp:105] Iteration 1812, lr = 0.01 I0406 07:23:29.655158 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:23:31.563151 5226 solver.cpp:218] Iteration 1824 (2.25385 iter/s, 5.32422s/12 iters), loss = 3.45772 I0406 07:23:31.563189 5226 solver.cpp:237] Train net output #0: loss = 3.45772 (* 1 = 3.45772 loss) I0406 07:23:31.563194 5226 sgd_solver.cpp:105] Iteration 1824, lr = 0.01 I0406 07:23:36.043905 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1836.caffemodel I0406 07:23:39.049489 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1836.solverstate I0406 07:23:41.366464 5226 solver.cpp:330] Iteration 1836, Testing net (#0) I0406 07:23:41.366488 5226 net.cpp:676] Ignoring source layer train-data I0406 07:23:44.915056 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:23:45.670183 5226 solver.cpp:397] Test net output #0: accuracy = 0.196691 I0406 07:23:45.670212 5226 solver.cpp:397] Test net output #1: loss = 3.45419 (* 1 = 3.45419 loss) I0406 07:23:45.809367 5226 solver.cpp:218] Iteration 1836 (0.842338 iter/s, 14.2461s/12 iters), loss = 2.91034 I0406 07:23:45.809428 5226 solver.cpp:237] Train net output #0: loss = 2.91034 (* 1 = 2.91034 loss) I0406 07:23:45.809437 5226 sgd_solver.cpp:105] Iteration 1836, lr = 0.01 I0406 07:23:50.173804 5226 solver.cpp:218] Iteration 1848 (2.74956 iter/s, 4.36433s/12 iters), loss = 3.07219 I0406 07:23:50.173846 5226 solver.cpp:237] Train net output #0: loss = 3.07219 (* 1 = 3.07219 loss) I0406 07:23:50.173852 5226 sgd_solver.cpp:105] Iteration 1848, lr = 0.01 I0406 07:23:55.371817 5226 solver.cpp:218] Iteration 1860 (2.30861 iter/s, 5.19793s/12 iters), loss = 3.34904 I0406 07:23:55.371928 5226 solver.cpp:237] Train net output #0: loss = 3.34904 (* 1 = 3.34904 loss) I0406 07:23:55.371934 5226 sgd_solver.cpp:105] Iteration 1860, lr = 0.01 I0406 07:24:00.669976 5226 solver.cpp:218] Iteration 1872 (2.26501 iter/s, 5.298s/12 iters), loss = 3.02528 I0406 07:24:00.670012 5226 solver.cpp:237] Train net output #0: loss = 3.02528 (* 1 = 3.02528 loss) I0406 07:24:00.670017 5226 sgd_solver.cpp:105] Iteration 1872, lr = 0.01 I0406 07:24:05.628597 5226 solver.cpp:218] Iteration 1884 (2.42007 iter/s, 4.95853s/12 iters), loss = 3.29507 I0406 07:24:05.628648 5226 solver.cpp:237] Train net output #0: loss = 3.29507 (* 1 = 3.29507 loss) I0406 07:24:05.628659 5226 sgd_solver.cpp:105] Iteration 1884, lr = 0.01 I0406 07:24:10.858410 5226 solver.cpp:218] Iteration 1896 (2.29458 iter/s, 5.22971s/12 iters), loss = 3.04579 I0406 07:24:10.858456 5226 solver.cpp:237] Train net output #0: loss = 3.04579 (* 1 = 3.04579 loss) I0406 07:24:10.858464 5226 sgd_solver.cpp:105] Iteration 1896, lr = 0.01 I0406 07:24:16.128111 5226 solver.cpp:218] Iteration 1908 (2.27721 iter/s, 5.26961s/12 iters), loss = 2.9373 I0406 07:24:16.128149 5226 solver.cpp:237] Train net output #0: loss = 2.9373 (* 1 = 2.9373 loss) I0406 07:24:16.128154 5226 sgd_solver.cpp:105] Iteration 1908, lr = 0.01 I0406 07:24:21.435420 5226 solver.cpp:218] Iteration 1920 (2.26107 iter/s, 5.30723s/12 iters), loss = 2.82193 I0406 07:24:21.435458 5226 solver.cpp:237] Train net output #0: loss = 2.82193 (* 1 = 2.82193 loss) I0406 07:24:21.435463 5226 sgd_solver.cpp:105] Iteration 1920, lr = 0.01 I0406 07:24:21.722877 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:24:26.553185 5226 solver.cpp:218] Iteration 1932 (2.34481 iter/s, 5.11768s/12 iters), loss = 2.82071 I0406 07:24:26.553285 5226 solver.cpp:237] Train net output #0: loss = 2.82071 (* 1 = 2.82071 loss) I0406 07:24:26.553292 5226 sgd_solver.cpp:105] Iteration 1932, lr = 0.01 I0406 07:24:28.549466 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1938.caffemodel I0406 07:24:31.572197 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1938.solverstate I0406 07:24:33.917404 5226 solver.cpp:330] Iteration 1938, Testing net (#0) I0406 07:24:33.917428 5226 net.cpp:676] Ignoring source layer train-data I0406 07:24:37.581090 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:24:38.371762 5226 solver.cpp:397] Test net output #0: accuracy = 0.218137 I0406 07:24:38.371789 5226 solver.cpp:397] Test net output #1: loss = 3.36926 (* 1 = 3.36926 loss) I0406 07:24:40.250432 5226 solver.cpp:218] Iteration 1944 (0.876101 iter/s, 13.697s/12 iters), loss = 2.95375 I0406 07:24:40.250474 5226 solver.cpp:237] Train net output #0: loss = 2.95375 (* 1 = 2.95375 loss) I0406 07:24:40.250480 5226 sgd_solver.cpp:105] Iteration 1944, lr = 0.01 I0406 07:24:45.720738 5226 solver.cpp:218] Iteration 1956 (2.1937 iter/s, 5.47022s/12 iters), loss = 2.84667 I0406 07:24:45.720775 5226 solver.cpp:237] Train net output #0: loss = 2.84667 (* 1 = 2.84667 loss) I0406 07:24:45.720782 5226 sgd_solver.cpp:105] Iteration 1956, lr = 0.01 I0406 07:24:50.905449 5226 solver.cpp:218] Iteration 1968 (2.31454 iter/s, 5.18462s/12 iters), loss = 3.18985 I0406 07:24:50.905489 5226 solver.cpp:237] Train net output #0: loss = 3.18985 (* 1 = 3.18985 loss) I0406 07:24:50.905494 5226 sgd_solver.cpp:105] Iteration 1968, lr = 0.01 I0406 07:24:55.894114 5226 solver.cpp:218] Iteration 1980 (2.40549 iter/s, 4.98858s/12 iters), loss = 3.0115 I0406 07:24:55.894151 5226 solver.cpp:237] Train net output #0: loss = 3.0115 (* 1 = 3.0115 loss) I0406 07:24:55.894157 5226 sgd_solver.cpp:105] Iteration 1980, lr = 0.01 I0406 07:25:01.179332 5226 solver.cpp:218] Iteration 1992 (2.27052 iter/s, 5.28513s/12 iters), loss = 3.05164 I0406 07:25:01.179481 5226 solver.cpp:237] Train net output #0: loss = 3.05164 (* 1 = 3.05164 loss) I0406 07:25:01.179489 5226 sgd_solver.cpp:105] Iteration 1992, lr = 0.01 I0406 07:25:06.602145 5226 solver.cpp:218] Iteration 2004 (2.21295 iter/s, 5.42261s/12 iters), loss = 3.15305 I0406 07:25:06.602186 5226 solver.cpp:237] Train net output #0: loss = 3.15305 (* 1 = 3.15305 loss) I0406 07:25:06.602193 5226 sgd_solver.cpp:105] Iteration 2004, lr = 0.01 I0406 07:25:11.963425 5226 solver.cpp:218] Iteration 2016 (2.23831 iter/s, 5.3612s/12 iters), loss = 3.19823 I0406 07:25:11.963457 5226 solver.cpp:237] Train net output #0: loss = 3.19823 (* 1 = 3.19823 loss) I0406 07:25:11.963462 5226 sgd_solver.cpp:105] Iteration 2016, lr = 0.01 I0406 07:25:14.509896 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:25:17.207149 5226 solver.cpp:218] Iteration 2028 (2.28849 iter/s, 5.24364s/12 iters), loss = 3.12019 I0406 07:25:17.207185 5226 solver.cpp:237] Train net output #0: loss = 3.12019 (* 1 = 3.12019 loss) I0406 07:25:17.207190 5226 sgd_solver.cpp:105] Iteration 2028, lr = 0.01 I0406 07:25:22.115345 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2040.caffemodel I0406 07:25:25.148134 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2040.solverstate I0406 07:25:27.448951 5226 solver.cpp:330] Iteration 2040, Testing net (#0) I0406 07:25:27.448969 5226 net.cpp:676] Ignoring source layer train-data I0406 07:25:30.911226 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:25:31.728209 5226 solver.cpp:397] Test net output #0: accuracy = 0.222426 I0406 07:25:31.728292 5226 solver.cpp:397] Test net output #1: loss = 3.39818 (* 1 = 3.39818 loss) I0406 07:25:31.866468 5226 solver.cpp:218] Iteration 2040 (0.8186 iter/s, 14.6592s/12 iters), loss = 3.13997 I0406 07:25:31.866508 5226 solver.cpp:237] Train net output #0: loss = 3.13997 (* 1 = 3.13997 loss) I0406 07:25:31.866514 5226 sgd_solver.cpp:105] Iteration 2040, lr = 0.01 I0406 07:25:36.104786 5226 solver.cpp:218] Iteration 2052 (2.83137 iter/s, 4.23823s/12 iters), loss = 3.01645 I0406 07:25:36.104827 5226 solver.cpp:237] Train net output #0: loss = 3.01645 (* 1 = 3.01645 loss) I0406 07:25:36.104831 5226 sgd_solver.cpp:105] Iteration 2052, lr = 0.01 I0406 07:25:37.801060 5226 blocking_queue.cpp:49] Waiting for data I0406 07:25:41.465049 5226 solver.cpp:218] Iteration 2064 (2.23873 iter/s, 5.36017s/12 iters), loss = 2.50134 I0406 07:25:41.465085 5226 solver.cpp:237] Train net output #0: loss = 2.50134 (* 1 = 2.50134 loss) I0406 07:25:41.465090 5226 sgd_solver.cpp:105] Iteration 2064, lr = 0.01 I0406 07:25:46.810606 5226 solver.cpp:218] Iteration 2076 (2.24489 iter/s, 5.34547s/12 iters), loss = 2.76902 I0406 07:25:46.810645 5226 solver.cpp:237] Train net output #0: loss = 2.76902 (* 1 = 2.76902 loss) I0406 07:25:46.810652 5226 sgd_solver.cpp:105] Iteration 2076, lr = 0.01 I0406 07:25:52.052099 5226 solver.cpp:218] Iteration 2088 (2.28946 iter/s, 5.24141s/12 iters), loss = 3.00441 I0406 07:25:52.052136 5226 solver.cpp:237] Train net output #0: loss = 3.00441 (* 1 = 3.00441 loss) I0406 07:25:52.052141 5226 sgd_solver.cpp:105] Iteration 2088, lr = 0.01 I0406 07:25:57.265309 5226 solver.cpp:218] Iteration 2100 (2.30188 iter/s, 5.21312s/12 iters), loss = 2.76112 I0406 07:25:57.265347 5226 solver.cpp:237] Train net output #0: loss = 2.76112 (* 1 = 2.76112 loss) I0406 07:25:57.265352 5226 sgd_solver.cpp:105] Iteration 2100, lr = 0.01 I0406 07:26:02.500305 5226 solver.cpp:218] Iteration 2112 (2.2923 iter/s, 5.23491s/12 iters), loss = 2.6685 I0406 07:26:02.500427 5226 solver.cpp:237] Train net output #0: loss = 2.6685 (* 1 = 2.6685 loss) I0406 07:26:02.500434 5226 sgd_solver.cpp:105] Iteration 2112, lr = 0.01 I0406 07:26:07.365566 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:26:07.726665 5226 solver.cpp:218] Iteration 2124 (2.29613 iter/s, 5.22619s/12 iters), loss = 2.82303 I0406 07:26:07.726701 5226 solver.cpp:237] Train net output #0: loss = 2.82303 (* 1 = 2.82303 loss) I0406 07:26:07.726708 5226 sgd_solver.cpp:105] Iteration 2124, lr = 0.01 I0406 07:26:13.073390 5226 solver.cpp:218] Iteration 2136 (2.2444 iter/s, 5.34664s/12 iters), loss = 2.98934 I0406 07:26:13.073433 5226 solver.cpp:237] Train net output #0: loss = 2.98934 (* 1 = 2.98934 loss) I0406 07:26:13.073441 5226 sgd_solver.cpp:105] Iteration 2136, lr = 0.01 I0406 07:26:15.170984 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2142.caffemodel I0406 07:26:18.223866 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2142.solverstate I0406 07:26:20.526892 5226 solver.cpp:330] Iteration 2142, Testing net (#0) I0406 07:26:20.526911 5226 net.cpp:676] Ignoring source layer train-data I0406 07:26:24.024772 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:26:24.873953 5226 solver.cpp:397] Test net output #0: accuracy = 0.235907 I0406 07:26:24.873989 5226 solver.cpp:397] Test net output #1: loss = 3.27825 (* 1 = 3.27825 loss) I0406 07:26:26.736681 5226 solver.cpp:218] Iteration 2148 (0.878275 iter/s, 13.6631s/12 iters), loss = 2.80053 I0406 07:26:26.736730 5226 solver.cpp:237] Train net output #0: loss = 2.80053 (* 1 = 2.80053 loss) I0406 07:26:26.736737 5226 sgd_solver.cpp:105] Iteration 2148, lr = 0.01 I0406 07:26:32.328927 5226 solver.cpp:218] Iteration 2160 (2.14587 iter/s, 5.59214s/12 iters), loss = 2.48772 I0406 07:26:32.328969 5226 solver.cpp:237] Train net output #0: loss = 2.48772 (* 1 = 2.48772 loss) I0406 07:26:32.328975 5226 sgd_solver.cpp:105] Iteration 2160, lr = 0.01 I0406 07:26:37.600306 5226 solver.cpp:218] Iteration 2172 (2.27648 iter/s, 5.27129s/12 iters), loss = 2.69526 I0406 07:26:37.600389 5226 solver.cpp:237] Train net output #0: loss = 2.69526 (* 1 = 2.69526 loss) I0406 07:26:37.600397 5226 sgd_solver.cpp:105] Iteration 2172, lr = 0.01 I0406 07:26:42.859814 5226 solver.cpp:218] Iteration 2184 (2.28164 iter/s, 5.25937s/12 iters), loss = 2.78594 I0406 07:26:42.859869 5226 solver.cpp:237] Train net output #0: loss = 2.78594 (* 1 = 2.78594 loss) I0406 07:26:42.859879 5226 sgd_solver.cpp:105] Iteration 2184, lr = 0.01 I0406 07:26:48.213620 5226 solver.cpp:218] Iteration 2196 (2.24144 iter/s, 5.35371s/12 iters), loss = 2.66253 I0406 07:26:48.213654 5226 solver.cpp:237] Train net output #0: loss = 2.66253 (* 1 = 2.66253 loss) I0406 07:26:48.213660 5226 sgd_solver.cpp:105] Iteration 2196, lr = 0.01 I0406 07:26:53.524387 5226 solver.cpp:218] Iteration 2208 (2.2596 iter/s, 5.31068s/12 iters), loss = 2.77228 I0406 07:26:53.524435 5226 solver.cpp:237] Train net output #0: loss = 2.77228 (* 1 = 2.77228 loss) I0406 07:26:53.524441 5226 sgd_solver.cpp:105] Iteration 2208, lr = 0.01 I0406 07:26:58.809394 5226 solver.cpp:218] Iteration 2220 (2.27061 iter/s, 5.28491s/12 iters), loss = 2.52364 I0406 07:26:58.809432 5226 solver.cpp:237] Train net output #0: loss = 2.52364 (* 1 = 2.52364 loss) I0406 07:26:58.809437 5226 sgd_solver.cpp:105] Iteration 2220, lr = 0.01 I0406 07:27:00.768290 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:27:04.126536 5226 solver.cpp:218] Iteration 2232 (2.25689 iter/s, 5.31705s/12 iters), loss = 2.71301 I0406 07:27:04.126575 5226 solver.cpp:237] Train net output #0: loss = 2.71301 (* 1 = 2.71301 loss) I0406 07:27:04.126581 5226 sgd_solver.cpp:105] Iteration 2232, lr = 0.01 I0406 07:27:08.991061 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2244.caffemodel I0406 07:27:11.995447 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2244.solverstate I0406 07:27:14.296033 5226 solver.cpp:330] Iteration 2244, Testing net (#0) I0406 07:27:14.296052 5226 net.cpp:676] Ignoring source layer train-data I0406 07:27:17.663213 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:27:18.575387 5226 solver.cpp:397] Test net output #0: accuracy = 0.268382 I0406 07:27:18.575423 5226 solver.cpp:397] Test net output #1: loss = 3.14014 (* 1 = 3.14014 loss) I0406 07:27:18.720084 5226 solver.cpp:218] Iteration 2244 (0.82229 iter/s, 14.5934s/12 iters), loss = 2.32883 I0406 07:27:18.720144 5226 solver.cpp:237] Train net output #0: loss = 2.32883 (* 1 = 2.32883 loss) I0406 07:27:18.720152 5226 sgd_solver.cpp:105] Iteration 2244, lr = 0.01 I0406 07:27:23.158629 5226 solver.cpp:218] Iteration 2256 (2.70366 iter/s, 4.43844s/12 iters), loss = 2.44244 I0406 07:27:23.158671 5226 solver.cpp:237] Train net output #0: loss = 2.44244 (* 1 = 2.44244 loss) I0406 07:27:23.158677 5226 sgd_solver.cpp:105] Iteration 2256, lr = 0.01 I0406 07:27:28.498594 5226 solver.cpp:218] Iteration 2268 (2.24724 iter/s, 5.33988s/12 iters), loss = 2.4724 I0406 07:27:28.498631 5226 solver.cpp:237] Train net output #0: loss = 2.4724 (* 1 = 2.4724 loss) I0406 07:27:28.498636 5226 sgd_solver.cpp:105] Iteration 2268, lr = 0.01 I0406 07:27:33.714381 5226 solver.cpp:218] Iteration 2280 (2.30075 iter/s, 5.2157s/12 iters), loss = 2.83676 I0406 07:27:33.714419 5226 solver.cpp:237] Train net output #0: loss = 2.83676 (* 1 = 2.83676 loss) I0406 07:27:33.714426 5226 sgd_solver.cpp:105] Iteration 2280, lr = 0.01 I0406 07:27:38.969461 5226 solver.cpp:218] Iteration 2292 (2.28354 iter/s, 5.255s/12 iters), loss = 2.6109 I0406 07:27:38.969496 5226 solver.cpp:237] Train net output #0: loss = 2.6109 (* 1 = 2.6109 loss) I0406 07:27:38.969501 5226 sgd_solver.cpp:105] Iteration 2292, lr = 0.01 I0406 07:27:44.438954 5226 solver.cpp:218] Iteration 2304 (2.19402 iter/s, 5.46941s/12 iters), loss = 2.65751 I0406 07:27:44.439044 5226 solver.cpp:237] Train net output #0: loss = 2.65751 (* 1 = 2.65751 loss) I0406 07:27:44.439050 5226 sgd_solver.cpp:105] Iteration 2304, lr = 0.01 I0406 07:27:49.814252 5226 solver.cpp:218] Iteration 2316 (2.23249 iter/s, 5.37516s/12 iters), loss = 2.2846 I0406 07:27:49.814291 5226 solver.cpp:237] Train net output #0: loss = 2.2846 (* 1 = 2.2846 loss) I0406 07:27:49.814353 5226 sgd_solver.cpp:105] Iteration 2316, lr = 0.01 I0406 07:27:53.953617 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:27:55.141860 5226 solver.cpp:218] Iteration 2328 (2.25246 iter/s, 5.32752s/12 iters), loss = 2.64037 I0406 07:27:55.141896 5226 solver.cpp:237] Train net output #0: loss = 2.64037 (* 1 = 2.64037 loss) I0406 07:27:55.141902 5226 sgd_solver.cpp:105] Iteration 2328, lr = 0.01 I0406 07:28:00.500648 5226 solver.cpp:218] Iteration 2340 (2.23935 iter/s, 5.35871s/12 iters), loss = 2.31318 I0406 07:28:00.500684 5226 solver.cpp:237] Train net output #0: loss = 2.31318 (* 1 = 2.31318 loss) I0406 07:28:00.500689 5226 sgd_solver.cpp:105] Iteration 2340, lr = 0.01 I0406 07:28:02.730651 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2346.caffemodel I0406 07:28:05.641223 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2346.solverstate I0406 07:28:07.975626 5226 solver.cpp:330] Iteration 2346, Testing net (#0) I0406 07:28:07.975651 5226 net.cpp:676] Ignoring source layer train-data I0406 07:28:11.335269 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:28:12.359982 5226 solver.cpp:397] Test net output #0: accuracy = 0.254289 I0406 07:28:12.360018 5226 solver.cpp:397] Test net output #1: loss = 3.21678 (* 1 = 3.21678 loss) I0406 07:28:14.193176 5226 solver.cpp:218] Iteration 2352 (0.8764 iter/s, 13.6924s/12 iters), loss = 2.73386 I0406 07:28:14.193230 5226 solver.cpp:237] Train net output #0: loss = 2.73386 (* 1 = 2.73386 loss) I0406 07:28:14.193238 5226 sgd_solver.cpp:105] Iteration 2352, lr = 0.01 I0406 07:28:19.665506 5226 solver.cpp:218] Iteration 2364 (2.19289 iter/s, 5.47223s/12 iters), loss = 2.60615 I0406 07:28:19.665616 5226 solver.cpp:237] Train net output #0: loss = 2.60615 (* 1 = 2.60615 loss) I0406 07:28:19.665622 5226 sgd_solver.cpp:105] Iteration 2364, lr = 0.01 I0406 07:28:24.866171 5226 solver.cpp:218] Iteration 2376 (2.30747 iter/s, 5.20051s/12 iters), loss = 2.92253 I0406 07:28:24.866207 5226 solver.cpp:237] Train net output #0: loss = 2.92253 (* 1 = 2.92253 loss) I0406 07:28:24.866214 5226 sgd_solver.cpp:105] Iteration 2376, lr = 0.01 I0406 07:28:30.185734 5226 solver.cpp:218] Iteration 2388 (2.25586 iter/s, 5.31948s/12 iters), loss = 2.55916 I0406 07:28:30.185770 5226 solver.cpp:237] Train net output #0: loss = 2.55916 (* 1 = 2.55916 loss) I0406 07:28:30.185776 5226 sgd_solver.cpp:105] Iteration 2388, lr = 0.01 I0406 07:28:35.486991 5226 solver.cpp:218] Iteration 2400 (2.26365 iter/s, 5.30118s/12 iters), loss = 2.36989 I0406 07:28:35.487022 5226 solver.cpp:237] Train net output #0: loss = 2.36989 (* 1 = 2.36989 loss) I0406 07:28:35.487026 5226 sgd_solver.cpp:105] Iteration 2400, lr = 0.01 I0406 07:28:40.825139 5226 solver.cpp:218] Iteration 2412 (2.24801 iter/s, 5.33806s/12 iters), loss = 2.48392 I0406 07:28:40.825176 5226 solver.cpp:237] Train net output #0: loss = 2.48392 (* 1 = 2.48392 loss) I0406 07:28:40.825182 5226 sgd_solver.cpp:105] Iteration 2412, lr = 0.01 I0406 07:28:46.020892 5226 solver.cpp:218] Iteration 2424 (2.30962 iter/s, 5.19566s/12 iters), loss = 2.50994 I0406 07:28:46.020947 5226 solver.cpp:237] Train net output #0: loss = 2.50994 (* 1 = 2.50994 loss) I0406 07:28:46.020956 5226 sgd_solver.cpp:105] Iteration 2424, lr = 0.01 I0406 07:28:47.138969 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:28:51.407351 5226 solver.cpp:218] Iteration 2436 (2.22785 iter/s, 5.38635s/12 iters), loss = 2.52547 I0406 07:28:51.407461 5226 solver.cpp:237] Train net output #0: loss = 2.52547 (* 1 = 2.52547 loss) I0406 07:28:51.407470 5226 sgd_solver.cpp:105] Iteration 2436, lr = 0.01 I0406 07:28:56.247745 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2448.caffemodel I0406 07:28:59.258486 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2448.solverstate I0406 07:29:01.557282 5226 solver.cpp:330] Iteration 2448, Testing net (#0) I0406 07:29:01.557301 5226 net.cpp:676] Ignoring source layer train-data I0406 07:29:04.882333 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:29:05.892383 5226 solver.cpp:397] Test net output #0: accuracy = 0.27451 I0406 07:29:05.892410 5226 solver.cpp:397] Test net output #1: loss = 3.11446 (* 1 = 3.11446 loss) I0406 07:29:06.021387 5226 solver.cpp:218] Iteration 2448 (0.821141 iter/s, 14.6138s/12 iters), loss = 2.68962 I0406 07:29:06.021448 5226 solver.cpp:237] Train net output #0: loss = 2.68962 (* 1 = 2.68962 loss) I0406 07:29:06.021458 5226 sgd_solver.cpp:105] Iteration 2448, lr = 0.01 I0406 07:29:10.255609 5226 solver.cpp:218] Iteration 2460 (2.83412 iter/s, 4.23412s/12 iters), loss = 2.72334 I0406 07:29:10.255646 5226 solver.cpp:237] Train net output #0: loss = 2.72334 (* 1 = 2.72334 loss) I0406 07:29:10.255651 5226 sgd_solver.cpp:105] Iteration 2460, lr = 0.01 I0406 07:29:15.613543 5226 solver.cpp:218] Iteration 2472 (2.2397 iter/s, 5.35785s/12 iters), loss = 2.38164 I0406 07:29:15.613584 5226 solver.cpp:237] Train net output #0: loss = 2.38164 (* 1 = 2.38164 loss) I0406 07:29:15.613590 5226 sgd_solver.cpp:105] Iteration 2472, lr = 0.01 I0406 07:29:21.038807 5226 solver.cpp:218] Iteration 2484 (2.21191 iter/s, 5.42517s/12 iters), loss = 2.1673 I0406 07:29:21.038854 5226 solver.cpp:237] Train net output #0: loss = 2.1673 (* 1 = 2.1673 loss) I0406 07:29:21.038862 5226 sgd_solver.cpp:105] Iteration 2484, lr = 0.01 I0406 07:29:26.320931 5226 solver.cpp:218] Iteration 2496 (2.27185 iter/s, 5.28203s/12 iters), loss = 1.98013 I0406 07:29:26.321045 5226 solver.cpp:237] Train net output #0: loss = 1.98013 (* 1 = 1.98013 loss) I0406 07:29:26.321053 5226 sgd_solver.cpp:105] Iteration 2496, lr = 0.01 I0406 07:29:31.672318 5226 solver.cpp:218] Iteration 2508 (2.24248 iter/s, 5.35123s/12 iters), loss = 2.52243 I0406 07:29:31.672354 5226 solver.cpp:237] Train net output #0: loss = 2.52243 (* 1 = 2.52243 loss) I0406 07:29:31.672360 5226 sgd_solver.cpp:105] Iteration 2508, lr = 0.01 I0406 07:29:36.850450 5226 solver.cpp:218] Iteration 2520 (2.31748 iter/s, 5.17804s/12 iters), loss = 2.14708 I0406 07:29:36.850503 5226 solver.cpp:237] Train net output #0: loss = 2.14708 (* 1 = 2.14708 loss) I0406 07:29:36.850512 5226 sgd_solver.cpp:105] Iteration 2520, lr = 0.01 I0406 07:29:40.193704 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:29:42.157704 5226 solver.cpp:218] Iteration 2532 (2.2611 iter/s, 5.30715s/12 iters), loss = 2.4574 I0406 07:29:42.157748 5226 solver.cpp:237] Train net output #0: loss = 2.4574 (* 1 = 2.4574 loss) I0406 07:29:42.157755 5226 sgd_solver.cpp:105] Iteration 2532, lr = 0.01 I0406 07:29:47.303402 5226 solver.cpp:218] Iteration 2544 (2.33209 iter/s, 5.1456s/12 iters), loss = 2.39513 I0406 07:29:47.303438 5226 solver.cpp:237] Train net output #0: loss = 2.39513 (* 1 = 2.39513 loss) I0406 07:29:47.303445 5226 sgd_solver.cpp:105] Iteration 2544, lr = 0.01 I0406 07:29:49.364004 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2550.caffemodel I0406 07:29:52.394331 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2550.solverstate I0406 07:29:54.721765 5226 solver.cpp:330] Iteration 2550, Testing net (#0) I0406 07:29:54.721783 5226 net.cpp:676] Ignoring source layer train-data I0406 07:29:57.980566 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:29:58.983136 5226 solver.cpp:397] Test net output #0: accuracy = 0.276348 I0406 07:29:58.983170 5226 solver.cpp:397] Test net output #1: loss = 3.06047 (* 1 = 3.06047 loss) I0406 07:30:00.849547 5226 solver.cpp:218] Iteration 2556 (0.88587 iter/s, 13.546s/12 iters), loss = 2.15346 I0406 07:30:00.849598 5226 solver.cpp:237] Train net output #0: loss = 2.15346 (* 1 = 2.15346 loss) I0406 07:30:00.849606 5226 sgd_solver.cpp:105] Iteration 2556, lr = 0.01 I0406 07:30:05.706590 5226 solver.cpp:218] Iteration 2568 (2.47069 iter/s, 4.85695s/12 iters), loss = 1.99634 I0406 07:30:05.706638 5226 solver.cpp:237] Train net output #0: loss = 1.99634 (* 1 = 1.99634 loss) I0406 07:30:05.706646 5226 sgd_solver.cpp:105] Iteration 2568, lr = 0.01 I0406 07:30:11.101498 5226 solver.cpp:218] Iteration 2580 (2.22436 iter/s, 5.39481s/12 iters), loss = 2.33898 I0406 07:30:11.101536 5226 solver.cpp:237] Train net output #0: loss = 2.33898 (* 1 = 2.33898 loss) I0406 07:30:11.101541 5226 sgd_solver.cpp:105] Iteration 2580, lr = 0.01 I0406 07:30:16.395548 5226 solver.cpp:218] Iteration 2592 (2.26673 iter/s, 5.29396s/12 iters), loss = 2.23086 I0406 07:30:16.395584 5226 solver.cpp:237] Train net output #0: loss = 2.23086 (* 1 = 2.23086 loss) I0406 07:30:16.395589 5226 sgd_solver.cpp:105] Iteration 2592, lr = 0.01 I0406 07:30:21.637420 5226 solver.cpp:218] Iteration 2604 (2.2893 iter/s, 5.24178s/12 iters), loss = 2.30116 I0406 07:30:21.637461 5226 solver.cpp:237] Train net output #0: loss = 2.30116 (* 1 = 2.30116 loss) I0406 07:30:21.637467 5226 sgd_solver.cpp:105] Iteration 2604, lr = 0.01 I0406 07:30:26.795922 5226 solver.cpp:218] Iteration 2616 (2.3263 iter/s, 5.15842s/12 iters), loss = 1.97845 I0406 07:30:26.795961 5226 solver.cpp:237] Train net output #0: loss = 1.97845 (* 1 = 1.97845 loss) I0406 07:30:26.795967 5226 sgd_solver.cpp:105] Iteration 2616, lr = 0.01 I0406 07:30:31.858918 5226 solver.cpp:218] Iteration 2628 (2.37018 iter/s, 5.06291s/12 iters), loss = 2.1863 I0406 07:30:31.859036 5226 solver.cpp:237] Train net output #0: loss = 2.1863 (* 1 = 2.1863 loss) I0406 07:30:31.859042 5226 sgd_solver.cpp:105] Iteration 2628, lr = 0.01 I0406 07:30:32.334007 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:30:37.258980 5226 solver.cpp:218] Iteration 2640 (2.22226 iter/s, 5.3999s/12 iters), loss = 2.19801 I0406 07:30:37.259021 5226 solver.cpp:237] Train net output #0: loss = 2.19801 (* 1 = 2.19801 loss) I0406 07:30:37.259027 5226 sgd_solver.cpp:105] Iteration 2640, lr = 0.01 I0406 07:30:42.010056 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2652.caffemodel I0406 07:30:45.043380 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2652.solverstate I0406 07:30:47.344377 5226 solver.cpp:330] Iteration 2652, Testing net (#0) I0406 07:30:47.344396 5226 net.cpp:676] Ignoring source layer train-data I0406 07:30:50.558706 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:30:51.592180 5226 solver.cpp:397] Test net output #0: accuracy = 0.287377 I0406 07:30:51.592211 5226 solver.cpp:397] Test net output #1: loss = 3.16481 (* 1 = 3.16481 loss) I0406 07:30:51.727890 5226 solver.cpp:218] Iteration 2652 (0.829373 iter/s, 14.4688s/12 iters), loss = 2.57762 I0406 07:30:51.727926 5226 solver.cpp:237] Train net output #0: loss = 2.57762 (* 1 = 2.57762 loss) I0406 07:30:51.727931 5226 sgd_solver.cpp:105] Iteration 2652, lr = 0.01 I0406 07:30:55.859660 5226 solver.cpp:218] Iteration 2664 (2.90438 iter/s, 4.13169s/12 iters), loss = 1.93033 I0406 07:30:55.859692 5226 solver.cpp:237] Train net output #0: loss = 1.93033 (* 1 = 1.93033 loss) I0406 07:30:55.859699 5226 sgd_solver.cpp:105] Iteration 2664, lr = 0.01 I0406 07:31:01.036332 5226 solver.cpp:218] Iteration 2676 (2.31813 iter/s, 5.17659s/12 iters), loss = 2.23358 I0406 07:31:01.036371 5226 solver.cpp:237] Train net output #0: loss = 2.23358 (* 1 = 2.23358 loss) I0406 07:31:01.036378 5226 sgd_solver.cpp:105] Iteration 2676, lr = 0.01 I0406 07:31:06.379825 5226 solver.cpp:218] Iteration 2688 (2.24576 iter/s, 5.3434s/12 iters), loss = 1.98046 I0406 07:31:06.379911 5226 solver.cpp:237] Train net output #0: loss = 1.98046 (* 1 = 1.98046 loss) I0406 07:31:06.379918 5226 sgd_solver.cpp:105] Iteration 2688, lr = 0.01 I0406 07:31:11.525281 5226 solver.cpp:218] Iteration 2700 (2.33222 iter/s, 5.14532s/12 iters), loss = 2.46887 I0406 07:31:11.525326 5226 solver.cpp:237] Train net output #0: loss = 2.46887 (* 1 = 2.46887 loss) I0406 07:31:11.525333 5226 sgd_solver.cpp:105] Iteration 2700, lr = 0.01 I0406 07:31:16.678699 5226 solver.cpp:218] Iteration 2712 (2.32859 iter/s, 5.15333s/12 iters), loss = 2.32403 I0406 07:31:16.678737 5226 solver.cpp:237] Train net output #0: loss = 2.32403 (* 1 = 2.32403 loss) I0406 07:31:16.678742 5226 sgd_solver.cpp:105] Iteration 2712, lr = 0.01 I0406 07:31:21.784691 5226 solver.cpp:218] Iteration 2724 (2.35022 iter/s, 5.10591s/12 iters), loss = 2.30202 I0406 07:31:21.784734 5226 solver.cpp:237] Train net output #0: loss = 2.30202 (* 1 = 2.30202 loss) I0406 07:31:21.784739 5226 sgd_solver.cpp:105] Iteration 2724, lr = 0.01 I0406 07:31:24.576135 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:31:27.131541 5226 solver.cpp:218] Iteration 2736 (2.24435 iter/s, 5.34676s/12 iters), loss = 2.27352 I0406 07:31:27.131577 5226 solver.cpp:237] Train net output #0: loss = 2.27352 (* 1 = 2.27352 loss) I0406 07:31:27.131582 5226 sgd_solver.cpp:105] Iteration 2736, lr = 0.01 I0406 07:31:32.480998 5226 solver.cpp:218] Iteration 2748 (2.24326 iter/s, 5.34937s/12 iters), loss = 2.13631 I0406 07:31:32.481035 5226 solver.cpp:237] Train net output #0: loss = 2.13631 (* 1 = 2.13631 loss) I0406 07:31:32.481041 5226 sgd_solver.cpp:105] Iteration 2748, lr = 0.01 I0406 07:31:34.543975 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2754.caffemodel I0406 07:31:37.569041 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2754.solverstate I0406 07:31:39.866463 5226 solver.cpp:330] Iteration 2754, Testing net (#0) I0406 07:31:39.866482 5226 net.cpp:676] Ignoring source layer train-data I0406 07:31:42.831398 5226 blocking_queue.cpp:49] Waiting for data I0406 07:31:43.061197 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:31:44.152637 5226 solver.cpp:397] Test net output #0: accuracy = 0.284926 I0406 07:31:44.152662 5226 solver.cpp:397] Test net output #1: loss = 3.00838 (* 1 = 3.00838 loss) I0406 07:31:46.126389 5226 solver.cpp:218] Iteration 2760 (0.879426 iter/s, 13.6453s/12 iters), loss = 2.31702 I0406 07:31:46.126423 5226 solver.cpp:237] Train net output #0: loss = 2.31702 (* 1 = 2.31702 loss) I0406 07:31:46.126428 5226 sgd_solver.cpp:105] Iteration 2760, lr = 0.01 I0406 07:31:51.299409 5226 solver.cpp:218] Iteration 2772 (2.31977 iter/s, 5.17293s/12 iters), loss = 1.89611 I0406 07:31:51.299463 5226 solver.cpp:237] Train net output #0: loss = 1.89611 (* 1 = 1.89611 loss) I0406 07:31:51.299472 5226 sgd_solver.cpp:105] Iteration 2772, lr = 0.01 I0406 07:31:56.570516 5226 solver.cpp:218] Iteration 2784 (2.2766 iter/s, 5.27101s/12 iters), loss = 2.19765 I0406 07:31:56.570554 5226 solver.cpp:237] Train net output #0: loss = 2.19765 (* 1 = 2.19765 loss) I0406 07:31:56.570559 5226 sgd_solver.cpp:105] Iteration 2784, lr = 0.01 I0406 07:32:01.878157 5226 solver.cpp:218] Iteration 2796 (2.26093 iter/s, 5.30756s/12 iters), loss = 2.23022 I0406 07:32:01.878194 5226 solver.cpp:237] Train net output #0: loss = 2.23022 (* 1 = 2.23022 loss) I0406 07:32:01.878199 5226 sgd_solver.cpp:105] Iteration 2796, lr = 0.01 I0406 07:32:07.075815 5226 solver.cpp:218] Iteration 2808 (2.30877 iter/s, 5.19757s/12 iters), loss = 2.08422 I0406 07:32:07.075855 5226 solver.cpp:237] Train net output #0: loss = 2.08422 (* 1 = 2.08422 loss) I0406 07:32:07.075860 5226 sgd_solver.cpp:105] Iteration 2808, lr = 0.01 I0406 07:32:12.279300 5226 solver.cpp:218] Iteration 2820 (2.30618 iter/s, 5.2034s/12 iters), loss = 1.9555 I0406 07:32:12.279402 5226 solver.cpp:237] Train net output #0: loss = 1.9555 (* 1 = 1.9555 loss) I0406 07:32:12.279409 5226 sgd_solver.cpp:105] Iteration 2820, lr = 0.01 I0406 07:32:17.223429 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:32:17.552944 5226 solver.cpp:218] Iteration 2832 (2.27553 iter/s, 5.27349s/12 iters), loss = 2.25999 I0406 07:32:17.552980 5226 solver.cpp:237] Train net output #0: loss = 2.25999 (* 1 = 2.25999 loss) I0406 07:32:17.552985 5226 sgd_solver.cpp:105] Iteration 2832, lr = 0.01 I0406 07:32:22.575731 5226 solver.cpp:218] Iteration 2844 (2.38915 iter/s, 5.0227s/12 iters), loss = 2.27501 I0406 07:32:22.575768 5226 solver.cpp:237] Train net output #0: loss = 2.27501 (* 1 = 2.27501 loss) I0406 07:32:22.575773 5226 sgd_solver.cpp:105] Iteration 2844, lr = 0.01 I0406 07:32:26.959570 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2856.caffemodel I0406 07:32:30.000635 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2856.solverstate I0406 07:32:32.305446 5226 solver.cpp:330] Iteration 2856, Testing net (#0) I0406 07:32:32.305464 5226 net.cpp:676] Ignoring source layer train-data I0406 07:32:35.485757 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:32:36.598892 5226 solver.cpp:397] Test net output #0: accuracy = 0.311887 I0406 07:32:36.598923 5226 solver.cpp:397] Test net output #1: loss = 2.9559 (* 1 = 2.9559 loss) I0406 07:32:36.737751 5226 solver.cpp:218] Iteration 2856 (0.847345 iter/s, 14.1619s/12 iters), loss = 2.06675 I0406 07:32:36.737797 5226 solver.cpp:237] Train net output #0: loss = 2.06675 (* 1 = 2.06675 loss) I0406 07:32:36.737803 5226 sgd_solver.cpp:105] Iteration 2856, lr = 0.01 I0406 07:32:40.949191 5226 solver.cpp:218] Iteration 2868 (2.84944 iter/s, 4.21136s/12 iters), loss = 1.91832 I0406 07:32:40.949226 5226 solver.cpp:237] Train net output #0: loss = 1.91832 (* 1 = 1.91832 loss) I0406 07:32:40.949232 5226 sgd_solver.cpp:105] Iteration 2868, lr = 0.01 I0406 07:32:46.101409 5226 solver.cpp:218] Iteration 2880 (2.32913 iter/s, 5.15214s/12 iters), loss = 2.00836 I0406 07:32:46.101533 5226 solver.cpp:237] Train net output #0: loss = 2.00836 (* 1 = 2.00836 loss) I0406 07:32:46.101539 5226 sgd_solver.cpp:105] Iteration 2880, lr = 0.01 I0406 07:32:51.267841 5226 solver.cpp:218] Iteration 2892 (2.32276 iter/s, 5.16626s/12 iters), loss = 2.01934 I0406 07:32:51.267881 5226 solver.cpp:237] Train net output #0: loss = 2.01934 (* 1 = 2.01934 loss) I0406 07:32:51.267887 5226 sgd_solver.cpp:105] Iteration 2892, lr = 0.01 I0406 07:32:56.607414 5226 solver.cpp:218] Iteration 2904 (2.24741 iter/s, 5.33949s/12 iters), loss = 2.12261 I0406 07:32:56.607448 5226 solver.cpp:237] Train net output #0: loss = 2.12261 (* 1 = 2.12261 loss) I0406 07:32:56.607455 5226 sgd_solver.cpp:105] Iteration 2904, lr = 0.01 I0406 07:33:01.597326 5226 solver.cpp:218] Iteration 2916 (2.40489 iter/s, 4.98983s/12 iters), loss = 1.88956 I0406 07:33:01.597362 5226 solver.cpp:237] Train net output #0: loss = 1.88956 (* 1 = 1.88956 loss) I0406 07:33:01.597368 5226 sgd_solver.cpp:105] Iteration 2916, lr = 0.01 I0406 07:33:06.687352 5226 solver.cpp:218] Iteration 2928 (2.35759 iter/s, 5.08994s/12 iters), loss = 1.97696 I0406 07:33:06.687402 5226 solver.cpp:237] Train net output #0: loss = 1.97696 (* 1 = 1.97696 loss) I0406 07:33:06.687409 5226 sgd_solver.cpp:105] Iteration 2928, lr = 0.01 I0406 07:33:08.535871 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:33:11.995040 5226 solver.cpp:218] Iteration 2940 (2.26091 iter/s, 5.30759s/12 iters), loss = 1.71754 I0406 07:33:11.995079 5226 solver.cpp:237] Train net output #0: loss = 1.71754 (* 1 = 1.71754 loss) I0406 07:33:11.995083 5226 sgd_solver.cpp:105] Iteration 2940, lr = 0.01 I0406 07:33:17.293838 5226 solver.cpp:218] Iteration 2952 (2.2647 iter/s, 5.29871s/12 iters), loss = 1.85602 I0406 07:33:17.293977 5226 solver.cpp:237] Train net output #0: loss = 1.85602 (* 1 = 1.85602 loss) I0406 07:33:17.293985 5226 sgd_solver.cpp:105] Iteration 2952, lr = 0.01 I0406 07:33:19.428295 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2958.caffemodel I0406 07:33:22.450825 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2958.solverstate I0406 07:33:24.749099 5226 solver.cpp:330] Iteration 2958, Testing net (#0) I0406 07:33:24.749117 5226 net.cpp:676] Ignoring source layer train-data I0406 07:33:27.948948 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:33:29.112268 5226 solver.cpp:397] Test net output #0: accuracy = 0.308211 I0406 07:33:29.112304 5226 solver.cpp:397] Test net output #1: loss = 2.96295 (* 1 = 2.96295 loss) I0406 07:33:30.993541 5226 solver.cpp:218] Iteration 2964 (0.875947 iter/s, 13.6995s/12 iters), loss = 1.87735 I0406 07:33:30.993585 5226 solver.cpp:237] Train net output #0: loss = 1.87735 (* 1 = 1.87735 loss) I0406 07:33:30.993592 5226 sgd_solver.cpp:105] Iteration 2964, lr = 0.01 I0406 07:33:36.381143 5226 solver.cpp:218] Iteration 2976 (2.22737 iter/s, 5.38751s/12 iters), loss = 1.70259 I0406 07:33:36.381181 5226 solver.cpp:237] Train net output #0: loss = 1.70259 (* 1 = 1.70259 loss) I0406 07:33:36.381189 5226 sgd_solver.cpp:105] Iteration 2976, lr = 0.01 I0406 07:33:41.511652 5226 solver.cpp:218] Iteration 2988 (2.33899 iter/s, 5.13043s/12 iters), loss = 1.99707 I0406 07:33:41.511690 5226 solver.cpp:237] Train net output #0: loss = 1.99707 (* 1 = 1.99707 loss) I0406 07:33:41.511695 5226 sgd_solver.cpp:105] Iteration 2988, lr = 0.01 I0406 07:33:46.866042 5226 solver.cpp:218] Iteration 3000 (2.24119 iter/s, 5.3543s/12 iters), loss = 1.91498 I0406 07:33:46.866081 5226 solver.cpp:237] Train net output #0: loss = 1.91498 (* 1 = 1.91498 loss) I0406 07:33:46.866087 5226 sgd_solver.cpp:105] Iteration 3000, lr = 0.01 I0406 07:33:52.113684 5226 solver.cpp:218] Iteration 3012 (2.28678 iter/s, 5.24755s/12 iters), loss = 2.10747 I0406 07:33:52.113832 5226 solver.cpp:237] Train net output #0: loss = 2.10747 (* 1 = 2.10747 loss) I0406 07:33:52.113842 5226 sgd_solver.cpp:105] Iteration 3012, lr = 0.01 I0406 07:33:57.337056 5226 solver.cpp:218] Iteration 3024 (2.29746 iter/s, 5.22317s/12 iters), loss = 1.86945 I0406 07:33:57.337092 5226 solver.cpp:237] Train net output #0: loss = 1.86945 (* 1 = 1.86945 loss) I0406 07:33:57.337097 5226 sgd_solver.cpp:105] Iteration 3024, lr = 0.01 I0406 07:34:01.640444 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:34:02.837633 5226 solver.cpp:218] Iteration 3036 (2.18162 iter/s, 5.50049s/12 iters), loss = 1.74254 I0406 07:34:02.837669 5226 solver.cpp:237] Train net output #0: loss = 1.74254 (* 1 = 1.74254 loss) I0406 07:34:02.837675 5226 sgd_solver.cpp:105] Iteration 3036, lr = 0.01 I0406 07:34:07.857492 5226 solver.cpp:218] Iteration 3048 (2.39054 iter/s, 5.01978s/12 iters), loss = 1.80254 I0406 07:34:07.857532 5226 solver.cpp:237] Train net output #0: loss = 1.80254 (* 1 = 1.80254 loss) I0406 07:34:07.857537 5226 sgd_solver.cpp:105] Iteration 3048, lr = 0.01 I0406 07:34:12.515982 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3060.caffemodel I0406 07:34:15.508728 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3060.solverstate I0406 07:34:17.811239 5226 solver.cpp:330] Iteration 3060, Testing net (#0) I0406 07:34:17.811260 5226 net.cpp:676] Ignoring source layer train-data I0406 07:34:20.986883 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:34:22.195775 5226 solver.cpp:397] Test net output #0: accuracy = 0.308824 I0406 07:34:22.195890 5226 solver.cpp:397] Test net output #1: loss = 3.05209 (* 1 = 3.05209 loss) I0406 07:34:22.331161 5226 solver.cpp:218] Iteration 3060 (0.8291 iter/s, 14.4735s/12 iters), loss = 1.5061 I0406 07:34:22.331208 5226 solver.cpp:237] Train net output #0: loss = 1.5061 (* 1 = 1.5061 loss) I0406 07:34:22.331216 5226 sgd_solver.cpp:105] Iteration 3060, lr = 0.01 I0406 07:34:26.661587 5226 solver.cpp:218] Iteration 3072 (2.77115 iter/s, 4.33034s/12 iters), loss = 1.92765 I0406 07:34:26.661623 5226 solver.cpp:237] Train net output #0: loss = 1.92765 (* 1 = 1.92765 loss) I0406 07:34:26.661629 5226 sgd_solver.cpp:105] Iteration 3072, lr = 0.01 I0406 07:34:31.796916 5226 solver.cpp:218] Iteration 3084 (2.33679 iter/s, 5.13525s/12 iters), loss = 1.89066 I0406 07:34:31.796952 5226 solver.cpp:237] Train net output #0: loss = 1.89066 (* 1 = 1.89066 loss) I0406 07:34:31.796957 5226 sgd_solver.cpp:105] Iteration 3084, lr = 0.01 I0406 07:34:37.040153 5226 solver.cpp:218] Iteration 3096 (2.2887 iter/s, 5.24315s/12 iters), loss = 1.98083 I0406 07:34:37.040190 5226 solver.cpp:237] Train net output #0: loss = 1.98083 (* 1 = 1.98083 loss) I0406 07:34:37.040196 5226 sgd_solver.cpp:105] Iteration 3096, lr = 0.01 I0406 07:34:42.300650 5226 solver.cpp:218] Iteration 3108 (2.28119 iter/s, 5.26041s/12 iters), loss = 1.60416 I0406 07:34:42.300690 5226 solver.cpp:237] Train net output #0: loss = 1.60416 (* 1 = 1.60416 loss) I0406 07:34:42.300695 5226 sgd_solver.cpp:105] Iteration 3108, lr = 0.01 I0406 07:34:47.732422 5226 solver.cpp:218] Iteration 3120 (2.20926 iter/s, 5.43168s/12 iters), loss = 2.16043 I0406 07:34:47.732462 5226 solver.cpp:237] Train net output #0: loss = 2.16043 (* 1 = 2.16043 loss) I0406 07:34:47.732470 5226 sgd_solver.cpp:105] Iteration 3120, lr = 0.01 I0406 07:34:53.001641 5226 solver.cpp:218] Iteration 3132 (2.27742 iter/s, 5.26913s/12 iters), loss = 2.18725 I0406 07:34:53.001803 5226 solver.cpp:237] Train net output #0: loss = 2.18725 (* 1 = 2.18725 loss) I0406 07:34:53.001811 5226 sgd_solver.cpp:105] Iteration 3132, lr = 0.01 I0406 07:34:54.149668 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:34:58.375219 5226 solver.cpp:218] Iteration 3144 (2.23324 iter/s, 5.37337s/12 iters), loss = 1.74044 I0406 07:34:58.375275 5226 solver.cpp:237] Train net output #0: loss = 1.74044 (* 1 = 1.74044 loss) I0406 07:34:58.375284 5226 sgd_solver.cpp:105] Iteration 3144, lr = 0.01 I0406 07:35:03.671880 5226 solver.cpp:218] Iteration 3156 (2.26562 iter/s, 5.29655s/12 iters), loss = 2.13622 I0406 07:35:03.671918 5226 solver.cpp:237] Train net output #0: loss = 2.13622 (* 1 = 2.13622 loss) I0406 07:35:03.671924 5226 sgd_solver.cpp:105] Iteration 3156, lr = 0.01 I0406 07:35:05.815239 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3162.caffemodel I0406 07:35:08.885056 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3162.solverstate I0406 07:35:11.279619 5226 solver.cpp:330] Iteration 3162, Testing net (#0) I0406 07:35:11.279635 5226 net.cpp:676] Ignoring source layer train-data I0406 07:35:14.323442 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:35:15.565133 5226 solver.cpp:397] Test net output #0: accuracy = 0.305147 I0406 07:35:15.565165 5226 solver.cpp:397] Test net output #1: loss = 2.99407 (* 1 = 2.99407 loss) I0406 07:35:17.398216 5226 solver.cpp:218] Iteration 3168 (0.874241 iter/s, 13.7262s/12 iters), loss = 1.77368 I0406 07:35:17.398257 5226 solver.cpp:237] Train net output #0: loss = 1.77368 (* 1 = 1.77368 loss) I0406 07:35:17.398263 5226 sgd_solver.cpp:105] Iteration 3168, lr = 0.01 I0406 07:35:22.641556 5226 solver.cpp:218] Iteration 3180 (2.28866 iter/s, 5.24325s/12 iters), loss = 2.07338 I0406 07:35:22.641598 5226 solver.cpp:237] Train net output #0: loss = 2.07338 (* 1 = 2.07338 loss) I0406 07:35:22.641606 5226 sgd_solver.cpp:105] Iteration 3180, lr = 0.01 I0406 07:35:27.987327 5226 solver.cpp:218] Iteration 3192 (2.2448 iter/s, 5.34568s/12 iters), loss = 1.81821 I0406 07:35:27.987432 5226 solver.cpp:237] Train net output #0: loss = 1.81821 (* 1 = 1.81821 loss) I0406 07:35:27.987439 5226 sgd_solver.cpp:105] Iteration 3192, lr = 0.01 I0406 07:35:33.374217 5226 solver.cpp:218] Iteration 3204 (2.2277 iter/s, 5.38673s/12 iters), loss = 1.6746 I0406 07:35:33.374261 5226 solver.cpp:237] Train net output #0: loss = 1.6746 (* 1 = 1.6746 loss) I0406 07:35:33.374269 5226 sgd_solver.cpp:105] Iteration 3204, lr = 0.01 I0406 07:35:38.747400 5226 solver.cpp:218] Iteration 3216 (2.23335 iter/s, 5.37309s/12 iters), loss = 2.12646 I0406 07:35:38.747438 5226 solver.cpp:237] Train net output #0: loss = 2.12646 (* 1 = 2.12646 loss) I0406 07:35:38.747443 5226 sgd_solver.cpp:105] Iteration 3216, lr = 0.01 I0406 07:35:44.132916 5226 solver.cpp:218] Iteration 3228 (2.22824 iter/s, 5.38542s/12 iters), loss = 2.00402 I0406 07:35:44.132962 5226 solver.cpp:237] Train net output #0: loss = 2.00402 (* 1 = 2.00402 loss) I0406 07:35:44.132970 5226 sgd_solver.cpp:105] Iteration 3228, lr = 0.01 I0406 07:35:47.387502 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:35:49.317137 5226 solver.cpp:218] Iteration 3240 (2.31476 iter/s, 5.18413s/12 iters), loss = 1.80821 I0406 07:35:49.317175 5226 solver.cpp:237] Train net output #0: loss = 1.80821 (* 1 = 1.80821 loss) I0406 07:35:49.317181 5226 sgd_solver.cpp:105] Iteration 3240, lr = 0.01 I0406 07:35:54.582360 5226 solver.cpp:218] Iteration 3252 (2.27914 iter/s, 5.26514s/12 iters), loss = 2.28704 I0406 07:35:54.582397 5226 solver.cpp:237] Train net output #0: loss = 2.28704 (* 1 = 2.28704 loss) I0406 07:35:54.582402 5226 sgd_solver.cpp:105] Iteration 3252, lr = 0.01 I0406 07:35:59.110524 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3264.caffemodel I0406 07:36:02.172075 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3264.solverstate I0406 07:36:04.502154 5226 solver.cpp:330] Iteration 3264, Testing net (#0) I0406 07:36:04.502174 5226 net.cpp:676] Ignoring source layer train-data I0406 07:36:07.572098 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:36:08.926558 5226 solver.cpp:397] Test net output #0: accuracy = 0.310049 I0406 07:36:08.926586 5226 solver.cpp:397] Test net output #1: loss = 3.08834 (* 1 = 3.08834 loss) I0406 07:36:09.061693 5226 solver.cpp:218] Iteration 3264 (0.828776 iter/s, 14.4792s/12 iters), loss = 1.64652 I0406 07:36:09.061743 5226 solver.cpp:237] Train net output #0: loss = 1.64652 (* 1 = 1.64652 loss) I0406 07:36:09.061750 5226 sgd_solver.cpp:105] Iteration 3264, lr = 0.01 I0406 07:36:13.503286 5226 solver.cpp:218] Iteration 3276 (2.70179 iter/s, 4.4415s/12 iters), loss = 1.44034 I0406 07:36:13.503330 5226 solver.cpp:237] Train net output #0: loss = 1.44034 (* 1 = 1.44034 loss) I0406 07:36:13.503336 5226 sgd_solver.cpp:105] Iteration 3276, lr = 0.01 I0406 07:36:18.739557 5226 solver.cpp:218] Iteration 3288 (2.29175 iter/s, 5.23618s/12 iters), loss = 1.78925 I0406 07:36:18.739593 5226 solver.cpp:237] Train net output #0: loss = 1.78925 (* 1 = 1.78925 loss) I0406 07:36:18.739598 5226 sgd_solver.cpp:105] Iteration 3288, lr = 0.01 I0406 07:36:24.141005 5226 solver.cpp:218] Iteration 3300 (2.22166 iter/s, 5.40136s/12 iters), loss = 1.79829 I0406 07:36:24.141041 5226 solver.cpp:237] Train net output #0: loss = 1.79829 (* 1 = 1.79829 loss) I0406 07:36:24.141047 5226 sgd_solver.cpp:105] Iteration 3300, lr = 0.01 I0406 07:36:29.120118 5226 solver.cpp:218] Iteration 3312 (2.41011 iter/s, 4.97903s/12 iters), loss = 2.0169 I0406 07:36:29.120219 5226 solver.cpp:237] Train net output #0: loss = 2.0169 (* 1 = 2.0169 loss) I0406 07:36:29.120229 5226 sgd_solver.cpp:105] Iteration 3312, lr = 0.01 I0406 07:36:34.114833 5226 solver.cpp:218] Iteration 3324 (2.40261 iter/s, 4.99457s/12 iters), loss = 2.23304 I0406 07:36:34.114884 5226 solver.cpp:237] Train net output #0: loss = 2.23304 (* 1 = 2.23304 loss) I0406 07:36:34.114893 5226 sgd_solver.cpp:105] Iteration 3324, lr = 0.01 I0406 07:36:39.476953 5226 solver.cpp:218] Iteration 3336 (2.23796 iter/s, 5.36202s/12 iters), loss = 1.75603 I0406 07:36:39.477003 5226 solver.cpp:237] Train net output #0: loss = 1.75603 (* 1 = 1.75603 loss) I0406 07:36:39.477011 5226 sgd_solver.cpp:105] Iteration 3336, lr = 0.01 I0406 07:36:39.984163 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:36:44.852927 5226 solver.cpp:218] Iteration 3348 (2.23219 iter/s, 5.37587s/12 iters), loss = 1.97806 I0406 07:36:44.852977 5226 solver.cpp:237] Train net output #0: loss = 1.97806 (* 1 = 1.97806 loss) I0406 07:36:44.852986 5226 sgd_solver.cpp:105] Iteration 3348, lr = 0.01 I0406 07:36:50.265210 5226 solver.cpp:218] Iteration 3360 (2.21722 iter/s, 5.41219s/12 iters), loss = 1.63585 I0406 07:36:50.265255 5226 solver.cpp:237] Train net output #0: loss = 1.63585 (* 1 = 1.63585 loss) I0406 07:36:50.265264 5226 sgd_solver.cpp:105] Iteration 3360, lr = 0.01 I0406 07:36:52.381248 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3366.caffemodel I0406 07:36:55.435907 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3366.solverstate I0406 07:36:57.745326 5226 solver.cpp:330] Iteration 3366, Testing net (#0) I0406 07:36:57.745347 5226 net.cpp:676] Ignoring source layer train-data I0406 07:37:00.871713 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:37:02.224436 5226 solver.cpp:397] Test net output #0: accuracy = 0.300858 I0406 07:37:02.224472 5226 solver.cpp:397] Test net output #1: loss = 3.06763 (* 1 = 3.06763 loss) I0406 07:37:04.113633 5226 solver.cpp:218] Iteration 3372 (0.866534 iter/s, 13.8483s/12 iters), loss = 1.54664 I0406 07:37:04.113669 5226 solver.cpp:237] Train net output #0: loss = 1.54664 (* 1 = 1.54664 loss) I0406 07:37:04.113674 5226 sgd_solver.cpp:105] Iteration 3372, lr = 0.01 I0406 07:37:09.114207 5226 solver.cpp:218] Iteration 3384 (2.39976 iter/s, 5.00049s/12 iters), loss = 1.59694 I0406 07:37:09.114243 5226 solver.cpp:237] Train net output #0: loss = 1.59694 (* 1 = 1.59694 loss) I0406 07:37:09.114249 5226 sgd_solver.cpp:105] Iteration 3384, lr = 0.01 I0406 07:37:14.394776 5226 solver.cpp:218] Iteration 3396 (2.27252 iter/s, 5.28048s/12 iters), loss = 1.97977 I0406 07:37:14.394827 5226 solver.cpp:237] Train net output #0: loss = 1.97977 (* 1 = 1.97977 loss) I0406 07:37:14.394835 5226 sgd_solver.cpp:105] Iteration 3396, lr = 0.01 I0406 07:37:19.644085 5226 solver.cpp:218] Iteration 3408 (2.28606 iter/s, 5.24921s/12 iters), loss = 1.46625 I0406 07:37:19.644121 5226 solver.cpp:237] Train net output #0: loss = 1.46625 (* 1 = 1.46625 loss) I0406 07:37:19.644126 5226 sgd_solver.cpp:105] Iteration 3408, lr = 0.01 I0406 07:37:25.090657 5226 solver.cpp:218] Iteration 3420 (2.20326 iter/s, 5.44649s/12 iters), loss = 1.89754 I0406 07:37:25.090692 5226 solver.cpp:237] Train net output #0: loss = 1.89754 (* 1 = 1.89754 loss) I0406 07:37:25.090698 5226 sgd_solver.cpp:105] Iteration 3420, lr = 0.01 I0406 07:37:30.405474 5226 solver.cpp:218] Iteration 3432 (2.25788 iter/s, 5.31473s/12 iters), loss = 2.00795 I0406 07:37:30.405512 5226 solver.cpp:237] Train net output #0: loss = 2.00795 (* 1 = 2.00795 loss) I0406 07:37:30.405519 5226 sgd_solver.cpp:105] Iteration 3432, lr = 0.01 I0406 07:37:33.222360 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:37:35.736797 5226 solver.cpp:218] Iteration 3444 (2.25088 iter/s, 5.33124s/12 iters), loss = 1.57833 I0406 07:37:35.736832 5226 solver.cpp:237] Train net output #0: loss = 1.57833 (* 1 = 1.57833 loss) I0406 07:37:35.736837 5226 sgd_solver.cpp:105] Iteration 3444, lr = 0.01 I0406 07:37:41.088246 5226 solver.cpp:218] Iteration 3456 (2.24242 iter/s, 5.35136s/12 iters), loss = 1.89692 I0406 07:37:41.088284 5226 solver.cpp:237] Train net output #0: loss = 1.89692 (* 1 = 1.89692 loss) I0406 07:37:41.088289 5226 sgd_solver.cpp:105] Iteration 3456, lr = 0.01 I0406 07:37:45.926414 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3468.caffemodel I0406 07:37:48.977488 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3468.solverstate I0406 07:37:51.276590 5226 solver.cpp:330] Iteration 3468, Testing net (#0) I0406 07:37:51.276609 5226 net.cpp:676] Ignoring source layer train-data I0406 07:37:51.689401 5226 blocking_queue.cpp:49] Waiting for data I0406 07:37:54.217684 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:37:55.568193 5226 solver.cpp:397] Test net output #0: accuracy = 0.311274 I0406 07:37:55.568226 5226 solver.cpp:397] Test net output #1: loss = 3.06279 (* 1 = 3.06279 loss) I0406 07:37:55.708770 5226 solver.cpp:218] Iteration 3468 (0.820772 iter/s, 14.6204s/12 iters), loss = 1.84632 I0406 07:37:55.708812 5226 solver.cpp:237] Train net output #0: loss = 1.84632 (* 1 = 1.84632 loss) I0406 07:37:55.708817 5226 sgd_solver.cpp:105] Iteration 3468, lr = 0.01 I0406 07:37:59.940122 5226 solver.cpp:218] Iteration 3480 (2.83609 iter/s, 4.23118s/12 iters), loss = 1.86877 I0406 07:37:59.940165 5226 solver.cpp:237] Train net output #0: loss = 1.86877 (* 1 = 1.86877 loss) I0406 07:37:59.940171 5226 sgd_solver.cpp:105] Iteration 3480, lr = 0.01 I0406 07:38:04.944483 5226 solver.cpp:218] Iteration 3492 (2.39795 iter/s, 5.00427s/12 iters), loss = 1.77182 I0406 07:38:04.944667 5226 solver.cpp:237] Train net output #0: loss = 1.77182 (* 1 = 1.77182 loss) I0406 07:38:04.944681 5226 sgd_solver.cpp:105] Iteration 3492, lr = 0.01 I0406 07:38:10.229126 5226 solver.cpp:218] Iteration 3504 (2.27083 iter/s, 5.28442s/12 iters), loss = 1.53097 I0406 07:38:10.229166 5226 solver.cpp:237] Train net output #0: loss = 1.53097 (* 1 = 1.53097 loss) I0406 07:38:10.229172 5226 sgd_solver.cpp:105] Iteration 3504, lr = 0.01 I0406 07:38:15.509979 5226 solver.cpp:218] Iteration 3516 (2.2724 iter/s, 5.28076s/12 iters), loss = 1.72546 I0406 07:38:15.510016 5226 solver.cpp:237] Train net output #0: loss = 1.72546 (* 1 = 1.72546 loss) I0406 07:38:15.510022 5226 sgd_solver.cpp:105] Iteration 3516, lr = 0.01 I0406 07:38:20.753340 5226 solver.cpp:218] Iteration 3528 (2.28865 iter/s, 5.24328s/12 iters), loss = 1.6793 I0406 07:38:20.753377 5226 solver.cpp:237] Train net output #0: loss = 1.6793 (* 1 = 1.6793 loss) I0406 07:38:20.753382 5226 sgd_solver.cpp:105] Iteration 3528, lr = 0.01 I0406 07:38:25.631577 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:38:25.934854 5226 solver.cpp:218] Iteration 3540 (2.31596 iter/s, 5.18143s/12 iters), loss = 1.20465 I0406 07:38:25.934892 5226 solver.cpp:237] Train net output #0: loss = 1.20465 (* 1 = 1.20465 loss) I0406 07:38:25.934897 5226 sgd_solver.cpp:105] Iteration 3540, lr = 0.01 I0406 07:38:31.109880 5226 solver.cpp:218] Iteration 3552 (2.31887 iter/s, 5.17494s/12 iters), loss = 1.6886 I0406 07:38:31.109915 5226 solver.cpp:237] Train net output #0: loss = 1.6886 (* 1 = 1.6886 loss) I0406 07:38:31.109920 5226 sgd_solver.cpp:105] Iteration 3552, lr = 0.01 I0406 07:38:36.230836 5226 solver.cpp:218] Iteration 3564 (2.34335 iter/s, 5.12087s/12 iters), loss = 1.82824 I0406 07:38:36.231294 5226 solver.cpp:237] Train net output #0: loss = 1.82824 (* 1 = 1.82824 loss) I0406 07:38:36.231302 5226 sgd_solver.cpp:105] Iteration 3564, lr = 0.01 I0406 07:38:38.438297 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3570.caffemodel I0406 07:38:41.458721 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3570.solverstate I0406 07:38:43.769591 5226 solver.cpp:330] Iteration 3570, Testing net (#0) I0406 07:38:43.769611 5226 net.cpp:676] Ignoring source layer train-data I0406 07:38:46.681118 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:38:48.068444 5226 solver.cpp:397] Test net output #0: accuracy = 0.306985 I0406 07:38:48.068475 5226 solver.cpp:397] Test net output #1: loss = 3.07257 (* 1 = 3.07257 loss) I0406 07:38:49.824061 5226 solver.cpp:218] Iteration 3576 (0.882829 iter/s, 13.5927s/12 iters), loss = 1.46884 I0406 07:38:49.824111 5226 solver.cpp:237] Train net output #0: loss = 1.46884 (* 1 = 1.46884 loss) I0406 07:38:49.824120 5226 sgd_solver.cpp:105] Iteration 3576, lr = 0.01 I0406 07:38:55.115608 5226 solver.cpp:218] Iteration 3588 (2.26781 iter/s, 5.29145s/12 iters), loss = 1.8096 I0406 07:38:55.115656 5226 solver.cpp:237] Train net output #0: loss = 1.8096 (* 1 = 1.8096 loss) I0406 07:38:55.115664 5226 sgd_solver.cpp:105] Iteration 3588, lr = 0.01 I0406 07:39:00.417953 5226 solver.cpp:218] Iteration 3600 (2.26319 iter/s, 5.30225s/12 iters), loss = 1.55499 I0406 07:39:00.417995 5226 solver.cpp:237] Train net output #0: loss = 1.55499 (* 1 = 1.55499 loss) I0406 07:39:00.418002 5226 sgd_solver.cpp:105] Iteration 3600, lr = 0.01 I0406 07:39:05.712479 5226 solver.cpp:218] Iteration 3612 (2.26653 iter/s, 5.29444s/12 iters), loss = 1.75943 I0406 07:39:05.712522 5226 solver.cpp:237] Train net output #0: loss = 1.75943 (* 1 = 1.75943 loss) I0406 07:39:05.712527 5226 sgd_solver.cpp:105] Iteration 3612, lr = 0.01 I0406 07:39:11.019515 5226 solver.cpp:218] Iteration 3624 (2.26119 iter/s, 5.30695s/12 iters), loss = 1.62137 I0406 07:39:11.019640 5226 solver.cpp:237] Train net output #0: loss = 1.62137 (* 1 = 1.62137 loss) I0406 07:39:11.019647 5226 sgd_solver.cpp:105] Iteration 3624, lr = 0.01 I0406 07:39:16.083143 5226 solver.cpp:218] Iteration 3636 (2.36992 iter/s, 5.06346s/12 iters), loss = 1.40349 I0406 07:39:16.083178 5226 solver.cpp:237] Train net output #0: loss = 1.40349 (* 1 = 1.40349 loss) I0406 07:39:16.083184 5226 sgd_solver.cpp:105] Iteration 3636, lr = 0.01 I0406 07:39:18.021706 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:39:21.380954 5226 solver.cpp:218] Iteration 3648 (2.26512 iter/s, 5.29772s/12 iters), loss = 1.63095 I0406 07:39:21.381002 5226 solver.cpp:237] Train net output #0: loss = 1.63095 (* 1 = 1.63095 loss) I0406 07:39:21.381008 5226 sgd_solver.cpp:105] Iteration 3648, lr = 0.01 I0406 07:39:26.652264 5226 solver.cpp:218] Iteration 3660 (2.27651 iter/s, 5.27122s/12 iters), loss = 1.59956 I0406 07:39:26.652299 5226 solver.cpp:237] Train net output #0: loss = 1.59956 (* 1 = 1.59956 loss) I0406 07:39:26.652304 5226 sgd_solver.cpp:105] Iteration 3660, lr = 0.01 I0406 07:39:31.426250 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3672.caffemodel I0406 07:39:34.459728 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3672.solverstate I0406 07:39:36.755369 5226 solver.cpp:330] Iteration 3672, Testing net (#0) I0406 07:39:36.755388 5226 net.cpp:676] Ignoring source layer train-data I0406 07:39:39.599421 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:39:41.027675 5226 solver.cpp:397] Test net output #0: accuracy = 0.321691 I0406 07:39:41.027793 5226 solver.cpp:397] Test net output #1: loss = 3.10042 (* 1 = 3.10042 loss) I0406 07:39:41.159737 5226 solver.cpp:218] Iteration 3672 (0.827168 iter/s, 14.5073s/12 iters), loss = 1.53173 I0406 07:39:41.159788 5226 solver.cpp:237] Train net output #0: loss = 1.53173 (* 1 = 1.53173 loss) I0406 07:39:41.159795 5226 sgd_solver.cpp:105] Iteration 3672, lr = 0.01 I0406 07:39:45.523808 5226 solver.cpp:218] Iteration 3684 (2.74979 iter/s, 4.36398s/12 iters), loss = 1.54747 I0406 07:39:45.523846 5226 solver.cpp:237] Train net output #0: loss = 1.54747 (* 1 = 1.54747 loss) I0406 07:39:45.523851 5226 sgd_solver.cpp:105] Iteration 3684, lr = 0.01 I0406 07:39:50.635200 5226 solver.cpp:218] Iteration 3696 (2.34774 iter/s, 5.1113s/12 iters), loss = 1.81616 I0406 07:39:50.635237 5226 solver.cpp:237] Train net output #0: loss = 1.81616 (* 1 = 1.81616 loss) I0406 07:39:50.635242 5226 sgd_solver.cpp:105] Iteration 3696, lr = 0.01 I0406 07:39:55.909107 5226 solver.cpp:218] Iteration 3708 (2.27539 iter/s, 5.27382s/12 iters), loss = 1.42184 I0406 07:39:55.909147 5226 solver.cpp:237] Train net output #0: loss = 1.42184 (* 1 = 1.42184 loss) I0406 07:39:55.909153 5226 sgd_solver.cpp:105] Iteration 3708, lr = 0.01 I0406 07:40:01.064651 5226 solver.cpp:218] Iteration 3720 (2.32763 iter/s, 5.15546s/12 iters), loss = 1.58978 I0406 07:40:01.064697 5226 solver.cpp:237] Train net output #0: loss = 1.58978 (* 1 = 1.58978 loss) I0406 07:40:01.064704 5226 sgd_solver.cpp:105] Iteration 3720, lr = 0.01 I0406 07:40:06.113077 5226 solver.cpp:218] Iteration 3732 (2.37702 iter/s, 5.04834s/12 iters), loss = 1.78033 I0406 07:40:06.113116 5226 solver.cpp:237] Train net output #0: loss = 1.78033 (* 1 = 1.78033 loss) I0406 07:40:06.113121 5226 sgd_solver.cpp:105] Iteration 3732, lr = 0.01 I0406 07:40:10.241977 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:40:11.325973 5226 solver.cpp:218] Iteration 3744 (2.30202 iter/s, 5.21281s/12 iters), loss = 1.46165 I0406 07:40:11.326054 5226 solver.cpp:237] Train net output #0: loss = 1.46165 (* 1 = 1.46165 loss) I0406 07:40:11.326061 5226 sgd_solver.cpp:105] Iteration 3744, lr = 0.01 I0406 07:40:16.703722 5226 solver.cpp:218] Iteration 3756 (2.23147 iter/s, 5.37762s/12 iters), loss = 1.61511 I0406 07:40:16.703764 5226 solver.cpp:237] Train net output #0: loss = 1.61511 (* 1 = 1.61511 loss) I0406 07:40:16.703771 5226 sgd_solver.cpp:105] Iteration 3756, lr = 0.01 I0406 07:40:21.900269 5226 solver.cpp:218] Iteration 3768 (2.30927 iter/s, 5.19645s/12 iters), loss = 1.73678 I0406 07:40:21.900315 5226 solver.cpp:237] Train net output #0: loss = 1.73678 (* 1 = 1.73678 loss) I0406 07:40:21.900323 5226 sgd_solver.cpp:105] Iteration 3768, lr = 0.01 I0406 07:40:23.910573 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3774.caffemodel I0406 07:40:26.926476 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3774.solverstate I0406 07:40:29.257396 5226 solver.cpp:330] Iteration 3774, Testing net (#0) I0406 07:40:29.257422 5226 net.cpp:676] Ignoring source layer train-data I0406 07:40:32.103924 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:40:33.564268 5226 solver.cpp:397] Test net output #0: accuracy = 0.300858 I0406 07:40:33.564303 5226 solver.cpp:397] Test net output #1: loss = 3.14698 (* 1 = 3.14698 loss) I0406 07:40:35.401221 5226 solver.cpp:218] Iteration 3780 (0.888835 iter/s, 13.5008s/12 iters), loss = 1.7147 I0406 07:40:35.401252 5226 solver.cpp:237] Train net output #0: loss = 1.7147 (* 1 = 1.7147 loss) I0406 07:40:35.401258 5226 sgd_solver.cpp:105] Iteration 3780, lr = 0.01 I0406 07:40:40.299185 5226 solver.cpp:218] Iteration 3792 (2.45004 iter/s, 4.89789s/12 iters), loss = 1.82895 I0406 07:40:40.299217 5226 solver.cpp:237] Train net output #0: loss = 1.82895 (* 1 = 1.82895 loss) I0406 07:40:40.299222 5226 sgd_solver.cpp:105] Iteration 3792, lr = 0.01 I0406 07:40:45.502756 5226 solver.cpp:218] Iteration 3804 (2.30615 iter/s, 5.20349s/12 iters), loss = 1.64243 I0406 07:40:45.502888 5226 solver.cpp:237] Train net output #0: loss = 1.64243 (* 1 = 1.64243 loss) I0406 07:40:45.502897 5226 sgd_solver.cpp:105] Iteration 3804, lr = 0.01 I0406 07:40:50.853632 5226 solver.cpp:218] Iteration 3816 (2.2427 iter/s, 5.3507s/12 iters), loss = 1.60145 I0406 07:40:50.853683 5226 solver.cpp:237] Train net output #0: loss = 1.60145 (* 1 = 1.60145 loss) I0406 07:40:50.853691 5226 sgd_solver.cpp:105] Iteration 3816, lr = 0.01 I0406 07:40:56.027886 5226 solver.cpp:218] Iteration 3828 (2.31922 iter/s, 5.17415s/12 iters), loss = 1.59594 I0406 07:40:56.027938 5226 solver.cpp:237] Train net output #0: loss = 1.59594 (* 1 = 1.59594 loss) I0406 07:40:56.027948 5226 sgd_solver.cpp:105] Iteration 3828, lr = 0.01 I0406 07:41:01.357242 5226 solver.cpp:218] Iteration 3840 (2.25172 iter/s, 5.32926s/12 iters), loss = 1.38014 I0406 07:41:01.357282 5226 solver.cpp:237] Train net output #0: loss = 1.38014 (* 1 = 1.38014 loss) I0406 07:41:01.357288 5226 sgd_solver.cpp:105] Iteration 3840, lr = 0.01 I0406 07:41:02.562899 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:41:06.764495 5226 solver.cpp:218] Iteration 3852 (2.21928 iter/s, 5.40716s/12 iters), loss = 1.30779 I0406 07:41:06.764533 5226 solver.cpp:237] Train net output #0: loss = 1.30779 (* 1 = 1.30779 loss) I0406 07:41:06.764537 5226 sgd_solver.cpp:105] Iteration 3852, lr = 0.01 I0406 07:41:12.159624 5226 solver.cpp:218] Iteration 3864 (2.22426 iter/s, 5.39504s/12 iters), loss = 1.6621 I0406 07:41:12.159663 5226 solver.cpp:237] Train net output #0: loss = 1.6621 (* 1 = 1.6621 loss) I0406 07:41:12.159668 5226 sgd_solver.cpp:105] Iteration 3864, lr = 0.01 I0406 07:41:16.761612 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3876.caffemodel I0406 07:41:19.845855 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3876.solverstate I0406 07:41:22.159665 5226 solver.cpp:330] Iteration 3876, Testing net (#0) I0406 07:41:22.159687 5226 net.cpp:676] Ignoring source layer train-data I0406 07:41:25.036998 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:41:26.558998 5226 solver.cpp:397] Test net output #0: accuracy = 0.335172 I0406 07:41:26.559031 5226 solver.cpp:397] Test net output #1: loss = 2.88753 (* 1 = 2.88753 loss) I0406 07:41:26.694090 5226 solver.cpp:218] Iteration 3876 (0.825632 iter/s, 14.5343s/12 iters), loss = 1.29414 I0406 07:41:26.694139 5226 solver.cpp:237] Train net output #0: loss = 1.29414 (* 1 = 1.29414 loss) I0406 07:41:26.694145 5226 sgd_solver.cpp:105] Iteration 3876, lr = 0.01 I0406 07:41:30.824424 5226 solver.cpp:218] Iteration 3888 (2.9054 iter/s, 4.13024s/12 iters), loss = 1.28747 I0406 07:41:30.824476 5226 solver.cpp:237] Train net output #0: loss = 1.28747 (* 1 = 1.28747 loss) I0406 07:41:30.824483 5226 sgd_solver.cpp:105] Iteration 3888, lr = 0.01 I0406 07:41:36.169155 5226 solver.cpp:218] Iteration 3900 (2.24524 iter/s, 5.34463s/12 iters), loss = 1.49082 I0406 07:41:36.169193 5226 solver.cpp:237] Train net output #0: loss = 1.49082 (* 1 = 1.49082 loss) I0406 07:41:36.169198 5226 sgd_solver.cpp:105] Iteration 3900, lr = 0.01 I0406 07:41:41.501305 5226 solver.cpp:218] Iteration 3912 (2.25054 iter/s, 5.33206s/12 iters), loss = 1.72229 I0406 07:41:41.501343 5226 solver.cpp:237] Train net output #0: loss = 1.72229 (* 1 = 1.72229 loss) I0406 07:41:41.501348 5226 sgd_solver.cpp:105] Iteration 3912, lr = 0.01 I0406 07:41:46.855024 5226 solver.cpp:218] Iteration 3924 (2.24147 iter/s, 5.35363s/12 iters), loss = 1.9069 I0406 07:41:46.855161 5226 solver.cpp:237] Train net output #0: loss = 1.9069 (* 1 = 1.9069 loss) I0406 07:41:46.855168 5226 sgd_solver.cpp:105] Iteration 3924, lr = 0.01 I0406 07:41:51.942207 5226 solver.cpp:218] Iteration 3936 (2.35895 iter/s, 5.08701s/12 iters), loss = 1.21786 I0406 07:41:51.942242 5226 solver.cpp:237] Train net output #0: loss = 1.21786 (* 1 = 1.21786 loss) I0406 07:41:51.942248 5226 sgd_solver.cpp:105] Iteration 3936, lr = 0.01 I0406 07:41:55.467344 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:41:57.211980 5226 solver.cpp:218] Iteration 3948 (2.27718 iter/s, 5.26969s/12 iters), loss = 1.3895 I0406 07:41:57.212018 5226 solver.cpp:237] Train net output #0: loss = 1.3895 (* 1 = 1.3895 loss) I0406 07:41:57.212023 5226 sgd_solver.cpp:105] Iteration 3948, lr = 0.01 I0406 07:42:02.648705 5226 solver.cpp:218] Iteration 3960 (2.20725 iter/s, 5.43664s/12 iters), loss = 1.33313 I0406 07:42:02.648741 5226 solver.cpp:237] Train net output #0: loss = 1.33313 (* 1 = 1.33313 loss) I0406 07:42:02.648746 5226 sgd_solver.cpp:105] Iteration 3960, lr = 0.01 I0406 07:42:08.008857 5226 solver.cpp:218] Iteration 3972 (2.23878 iter/s, 5.36007s/12 iters), loss = 1.31996 I0406 07:42:08.008898 5226 solver.cpp:237] Train net output #0: loss = 1.31996 (* 1 = 1.31996 loss) I0406 07:42:08.008903 5226 sgd_solver.cpp:105] Iteration 3972, lr = 0.01 I0406 07:42:10.309165 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3978.caffemodel I0406 07:42:13.227551 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3978.solverstate I0406 07:42:15.533095 5226 solver.cpp:330] Iteration 3978, Testing net (#0) I0406 07:42:15.533119 5226 net.cpp:676] Ignoring source layer train-data I0406 07:42:18.339995 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:42:19.900418 5226 solver.cpp:397] Test net output #0: accuracy = 0.322304 I0406 07:42:19.900445 5226 solver.cpp:397] Test net output #1: loss = 3.00782 (* 1 = 3.00782 loss) I0406 07:42:21.714197 5226 solver.cpp:218] Iteration 3984 (0.875581 iter/s, 13.7052s/12 iters), loss = 1.47955 I0406 07:42:21.714257 5226 solver.cpp:237] Train net output #0: loss = 1.47955 (* 1 = 1.47955 loss) I0406 07:42:21.714263 5226 sgd_solver.cpp:105] Iteration 3984, lr = 0.01 I0406 07:42:26.971159 5226 solver.cpp:218] Iteration 3996 (2.28273 iter/s, 5.25686s/12 iters), loss = 1.57538 I0406 07:42:26.971204 5226 solver.cpp:237] Train net output #0: loss = 1.57538 (* 1 = 1.57538 loss) I0406 07:42:26.971211 5226 sgd_solver.cpp:105] Iteration 3996, lr = 0.01 I0406 07:42:32.329207 5226 solver.cpp:218] Iteration 4008 (2.23966 iter/s, 5.35795s/12 iters), loss = 1.16564 I0406 07:42:32.329242 5226 solver.cpp:237] Train net output #0: loss = 1.16564 (* 1 = 1.16564 loss) I0406 07:42:32.329248 5226 sgd_solver.cpp:105] Iteration 4008, lr = 0.01 I0406 07:42:37.623857 5226 solver.cpp:218] Iteration 4020 (2.26648 iter/s, 5.29456s/12 iters), loss = 1.41902 I0406 07:42:37.623896 5226 solver.cpp:237] Train net output #0: loss = 1.41902 (* 1 = 1.41902 loss) I0406 07:42:37.623901 5226 sgd_solver.cpp:105] Iteration 4020, lr = 0.01 I0406 07:42:42.963275 5226 solver.cpp:218] Iteration 4032 (2.24747 iter/s, 5.33933s/12 iters), loss = 1.48964 I0406 07:42:42.963333 5226 solver.cpp:237] Train net output #0: loss = 1.48964 (* 1 = 1.48964 loss) I0406 07:42:42.963346 5226 sgd_solver.cpp:105] Iteration 4032, lr = 0.01 I0406 07:42:48.142302 5226 solver.cpp:218] Iteration 4044 (2.31708 iter/s, 5.17893s/12 iters), loss = 1.55566 I0406 07:42:48.142343 5226 solver.cpp:237] Train net output #0: loss = 1.55566 (* 1 = 1.55566 loss) I0406 07:42:48.142349 5226 sgd_solver.cpp:105] Iteration 4044, lr = 0.01 I0406 07:42:48.676218 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:42:53.273089 5226 solver.cpp:218] Iteration 4056 (2.33886 iter/s, 5.1307s/12 iters), loss = 1.4053 I0406 07:42:53.273125 5226 solver.cpp:237] Train net output #0: loss = 1.4053 (* 1 = 1.4053 loss) I0406 07:42:53.273130 5226 sgd_solver.cpp:105] Iteration 4056, lr = 0.01 I0406 07:42:58.448902 5226 solver.cpp:218] Iteration 4068 (2.31852 iter/s, 5.17572s/12 iters), loss = 1.49174 I0406 07:42:58.448940 5226 solver.cpp:237] Train net output #0: loss = 1.49174 (* 1 = 1.49174 loss) I0406 07:42:58.448946 5226 sgd_solver.cpp:105] Iteration 4068, lr = 0.01 I0406 07:43:03.149427 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4080.caffemodel I0406 07:43:06.157959 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4080.solverstate I0406 07:43:08.465924 5226 solver.cpp:330] Iteration 4080, Testing net (#0) I0406 07:43:08.465942 5226 net.cpp:676] Ignoring source layer train-data I0406 07:43:11.173875 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:43:12.757417 5226 solver.cpp:397] Test net output #0: accuracy = 0.327206 I0406 07:43:12.757441 5226 solver.cpp:397] Test net output #1: loss = 3.06441 (* 1 = 3.06441 loss) I0406 07:43:12.893946 5226 solver.cpp:218] Iteration 4080 (0.830743 iter/s, 14.4449s/12 iters), loss = 1.47637 I0406 07:43:12.893985 5226 solver.cpp:237] Train net output #0: loss = 1.47637 (* 1 = 1.47637 loss) I0406 07:43:12.893990 5226 sgd_solver.cpp:105] Iteration 4080, lr = 0.01 I0406 07:43:17.437860 5226 solver.cpp:218] Iteration 4092 (2.64094 iter/s, 4.54383s/12 iters), loss = 1.70891 I0406 07:43:17.437896 5226 solver.cpp:237] Train net output #0: loss = 1.70891 (* 1 = 1.70891 loss) I0406 07:43:17.437902 5226 sgd_solver.cpp:105] Iteration 4092, lr = 0.01 I0406 07:43:23.010995 5226 solver.cpp:218] Iteration 4104 (2.15322 iter/s, 5.57305s/12 iters), loss = 1.0637 I0406 07:43:23.011080 5226 solver.cpp:237] Train net output #0: loss = 1.0637 (* 1 = 1.0637 loss) I0406 07:43:23.011085 5226 sgd_solver.cpp:105] Iteration 4104, lr = 0.01 I0406 07:43:28.396471 5226 solver.cpp:218] Iteration 4116 (2.22827 iter/s, 5.38534s/12 iters), loss = 1.2998 I0406 07:43:28.396512 5226 solver.cpp:237] Train net output #0: loss = 1.2998 (* 1 = 1.2998 loss) I0406 07:43:28.396518 5226 sgd_solver.cpp:105] Iteration 4116, lr = 0.01 I0406 07:43:33.727180 5226 solver.cpp:218] Iteration 4128 (2.25115 iter/s, 5.33062s/12 iters), loss = 1.54419 I0406 07:43:33.727221 5226 solver.cpp:237] Train net output #0: loss = 1.54419 (* 1 = 1.54419 loss) I0406 07:43:33.727227 5226 sgd_solver.cpp:105] Iteration 4128, lr = 0.01 I0406 07:43:38.791496 5226 solver.cpp:218] Iteration 4140 (2.36956 iter/s, 5.06423s/12 iters), loss = 1.35332 I0406 07:43:38.791532 5226 solver.cpp:237] Train net output #0: loss = 1.35332 (* 1 = 1.35332 loss) I0406 07:43:38.791536 5226 sgd_solver.cpp:105] Iteration 4140, lr = 0.01 I0406 07:43:41.476444 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:43:43.946408 5226 solver.cpp:218] Iteration 4152 (2.32791 iter/s, 5.15483s/12 iters), loss = 1.5944 I0406 07:43:43.946444 5226 solver.cpp:237] Train net output #0: loss = 1.5944 (* 1 = 1.5944 loss) I0406 07:43:43.946449 5226 sgd_solver.cpp:105] Iteration 4152, lr = 0.01 I0406 07:43:45.588577 5226 blocking_queue.cpp:49] Waiting for data I0406 07:43:49.212473 5226 solver.cpp:218] Iteration 4164 (2.27878 iter/s, 5.26598s/12 iters), loss = 1.41644 I0406 07:43:49.212519 5226 solver.cpp:237] Train net output #0: loss = 1.41644 (* 1 = 1.41644 loss) I0406 07:43:49.212527 5226 sgd_solver.cpp:105] Iteration 4164, lr = 0.01 I0406 07:43:54.415076 5226 solver.cpp:218] Iteration 4176 (2.30658 iter/s, 5.20251s/12 iters), loss = 1.51725 I0406 07:43:54.415203 5226 solver.cpp:237] Train net output #0: loss = 1.51725 (* 1 = 1.51725 loss) I0406 07:43:54.415211 5226 sgd_solver.cpp:105] Iteration 4176, lr = 0.01 I0406 07:43:56.462532 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4182.caffemodel I0406 07:43:59.492276 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4182.solverstate I0406 07:44:01.831262 5226 solver.cpp:330] Iteration 4182, Testing net (#0) I0406 07:44:01.831282 5226 net.cpp:676] Ignoring source layer train-data I0406 07:44:04.516916 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:44:06.203562 5226 solver.cpp:397] Test net output #0: accuracy = 0.324755 I0406 07:44:06.203596 5226 solver.cpp:397] Test net output #1: loss = 3.08177 (* 1 = 3.08177 loss) I0406 07:44:07.972090 5226 solver.cpp:218] Iteration 4188 (0.885165 iter/s, 13.5568s/12 iters), loss = 1.28694 I0406 07:44:07.972131 5226 solver.cpp:237] Train net output #0: loss = 1.28694 (* 1 = 1.28694 loss) I0406 07:44:07.972136 5226 sgd_solver.cpp:105] Iteration 4188, lr = 0.01 I0406 07:44:13.320312 5226 solver.cpp:218] Iteration 4200 (2.24378 iter/s, 5.34813s/12 iters), loss = 1.2066 I0406 07:44:13.320348 5226 solver.cpp:237] Train net output #0: loss = 1.2066 (* 1 = 1.2066 loss) I0406 07:44:13.320353 5226 sgd_solver.cpp:105] Iteration 4200, lr = 0.01 I0406 07:44:18.788738 5226 solver.cpp:218] Iteration 4212 (2.19445 iter/s, 5.46834s/12 iters), loss = 1.17914 I0406 07:44:18.788771 5226 solver.cpp:237] Train net output #0: loss = 1.17914 (* 1 = 1.17914 loss) I0406 07:44:18.788777 5226 sgd_solver.cpp:105] Iteration 4212, lr = 0.01 I0406 07:44:24.056030 5226 solver.cpp:218] Iteration 4224 (2.27825 iter/s, 5.26721s/12 iters), loss = 1.17233 I0406 07:44:24.056072 5226 solver.cpp:237] Train net output #0: loss = 1.17233 (* 1 = 1.17233 loss) I0406 07:44:24.056078 5226 sgd_solver.cpp:105] Iteration 4224, lr = 0.01 I0406 07:44:29.446471 5226 solver.cpp:218] Iteration 4236 (2.2262 iter/s, 5.39034s/12 iters), loss = 1.39547 I0406 07:44:29.446555 5226 solver.cpp:237] Train net output #0: loss = 1.39547 (* 1 = 1.39547 loss) I0406 07:44:29.446564 5226 sgd_solver.cpp:105] Iteration 4236, lr = 0.01 I0406 07:44:34.450280 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:44:34.726584 5226 solver.cpp:218] Iteration 4248 (2.27274 iter/s, 5.27998s/12 iters), loss = 1.26418 I0406 07:44:34.726629 5226 solver.cpp:237] Train net output #0: loss = 1.26418 (* 1 = 1.26418 loss) I0406 07:44:34.726635 5226 sgd_solver.cpp:105] Iteration 4248, lr = 0.01 I0406 07:44:40.007169 5226 solver.cpp:218] Iteration 4260 (2.27252 iter/s, 5.28049s/12 iters), loss = 1.43957 I0406 07:44:40.007222 5226 solver.cpp:237] Train net output #0: loss = 1.43957 (* 1 = 1.43957 loss) I0406 07:44:40.007231 5226 sgd_solver.cpp:105] Iteration 4260, lr = 0.01 I0406 07:44:45.340837 5226 solver.cpp:218] Iteration 4272 (2.2499 iter/s, 5.33357s/12 iters), loss = 1.323 I0406 07:44:45.340876 5226 solver.cpp:237] Train net output #0: loss = 1.323 (* 1 = 1.323 loss) I0406 07:44:45.340886 5226 sgd_solver.cpp:105] Iteration 4272, lr = 0.01 I0406 07:44:50.123627 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4284.caffemodel I0406 07:44:53.206562 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4284.solverstate I0406 07:44:55.546898 5226 solver.cpp:330] Iteration 4284, Testing net (#0) I0406 07:44:55.546921 5226 net.cpp:676] Ignoring source layer train-data I0406 07:44:58.198887 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:44:59.849579 5226 solver.cpp:397] Test net output #0: accuracy = 0.333946 I0406 07:44:59.849733 5226 solver.cpp:397] Test net output #1: loss = 2.985 (* 1 = 2.985 loss) I0406 07:44:59.982801 5226 solver.cpp:218] Iteration 4284 (0.819571 iter/s, 14.6418s/12 iters), loss = 1.43728 I0406 07:44:59.982842 5226 solver.cpp:237] Train net output #0: loss = 1.43728 (* 1 = 1.43728 loss) I0406 07:44:59.982848 5226 sgd_solver.cpp:105] Iteration 4284, lr = 0.01 I0406 07:45:04.194146 5226 solver.cpp:218] Iteration 4296 (2.84951 iter/s, 4.21125s/12 iters), loss = 1.46934 I0406 07:45:04.194213 5226 solver.cpp:237] Train net output #0: loss = 1.46934 (* 1 = 1.46934 loss) I0406 07:45:04.194226 5226 sgd_solver.cpp:105] Iteration 4296, lr = 0.01 I0406 07:45:09.485049 5226 solver.cpp:218] Iteration 4308 (2.26809 iter/s, 5.29079s/12 iters), loss = 1.10036 I0406 07:45:09.485085 5226 solver.cpp:237] Train net output #0: loss = 1.10036 (* 1 = 1.10036 loss) I0406 07:45:09.485091 5226 sgd_solver.cpp:105] Iteration 4308, lr = 0.01 I0406 07:45:14.804729 5226 solver.cpp:218] Iteration 4320 (2.25581 iter/s, 5.3196s/12 iters), loss = 1.11731 I0406 07:45:14.804762 5226 solver.cpp:237] Train net output #0: loss = 1.11731 (* 1 = 1.11731 loss) I0406 07:45:14.804769 5226 sgd_solver.cpp:105] Iteration 4320, lr = 0.01 I0406 07:45:19.954612 5226 solver.cpp:218] Iteration 4332 (2.33019 iter/s, 5.14979s/12 iters), loss = 0.934898 I0406 07:45:19.954659 5226 solver.cpp:237] Train net output #0: loss = 0.934898 (* 1 = 0.934898 loss) I0406 07:45:19.954668 5226 sgd_solver.cpp:105] Iteration 4332, lr = 0.01 I0406 07:45:25.265318 5226 solver.cpp:218] Iteration 4344 (2.25963 iter/s, 5.31061s/12 iters), loss = 1.36034 I0406 07:45:25.265369 5226 solver.cpp:237] Train net output #0: loss = 1.36034 (* 1 = 1.36034 loss) I0406 07:45:25.265377 5226 sgd_solver.cpp:105] Iteration 4344, lr = 0.01 I0406 07:45:27.202687 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:45:30.424093 5226 solver.cpp:218] Iteration 4356 (2.32618 iter/s, 5.15868s/12 iters), loss = 1.06496 I0406 07:45:30.424628 5226 solver.cpp:237] Train net output #0: loss = 1.06496 (* 1 = 1.06496 loss) I0406 07:45:30.424636 5226 sgd_solver.cpp:105] Iteration 4356, lr = 0.01 I0406 07:45:35.784809 5226 solver.cpp:218] Iteration 4368 (2.23875 iter/s, 5.36014s/12 iters), loss = 1.36264 I0406 07:45:35.784843 5226 solver.cpp:237] Train net output #0: loss = 1.36264 (* 1 = 1.36264 loss) I0406 07:45:35.784849 5226 sgd_solver.cpp:105] Iteration 4368, lr = 0.01 I0406 07:45:41.000325 5226 solver.cpp:218] Iteration 4380 (2.30086 iter/s, 5.21544s/12 iters), loss = 1.48893 I0406 07:45:41.000365 5226 solver.cpp:237] Train net output #0: loss = 1.48893 (* 1 = 1.48893 loss) I0406 07:45:41.000371 5226 sgd_solver.cpp:105] Iteration 4380, lr = 0.01 I0406 07:45:43.195434 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4386.caffemodel I0406 07:45:46.197104 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4386.solverstate I0406 07:45:48.496718 5226 solver.cpp:330] Iteration 4386, Testing net (#0) I0406 07:45:48.496735 5226 net.cpp:676] Ignoring source layer train-data I0406 07:45:51.056607 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:45:52.756392 5226 solver.cpp:397] Test net output #0: accuracy = 0.328431 I0406 07:45:52.756439 5226 solver.cpp:397] Test net output #1: loss = 2.9974 (* 1 = 2.9974 loss) I0406 07:45:54.592645 5226 solver.cpp:218] Iteration 4392 (0.882861 iter/s, 13.5922s/12 iters), loss = 1.34584 I0406 07:45:54.592703 5226 solver.cpp:237] Train net output #0: loss = 1.34584 (* 1 = 1.34584 loss) I0406 07:45:54.592712 5226 sgd_solver.cpp:105] Iteration 4392, lr = 0.01 I0406 07:45:59.929893 5226 solver.cpp:218] Iteration 4404 (2.2484 iter/s, 5.33714s/12 iters), loss = 1.61502 I0406 07:45:59.929946 5226 solver.cpp:237] Train net output #0: loss = 1.61502 (* 1 = 1.61502 loss) I0406 07:45:59.929955 5226 sgd_solver.cpp:105] Iteration 4404, lr = 0.01 I0406 07:46:05.188912 5226 solver.cpp:218] Iteration 4416 (2.28184 iter/s, 5.25892s/12 iters), loss = 1.2401 I0406 07:46:05.189079 5226 solver.cpp:237] Train net output #0: loss = 1.2401 (* 1 = 1.2401 loss) I0406 07:46:05.189088 5226 sgd_solver.cpp:105] Iteration 4416, lr = 0.01 I0406 07:46:10.543936 5226 solver.cpp:218] Iteration 4428 (2.24097 iter/s, 5.35481s/12 iters), loss = 1.37075 I0406 07:46:10.543972 5226 solver.cpp:237] Train net output #0: loss = 1.37075 (* 1 = 1.37075 loss) I0406 07:46:10.543977 5226 sgd_solver.cpp:105] Iteration 4428, lr = 0.01 I0406 07:46:15.717097 5226 solver.cpp:218] Iteration 4440 (2.3197 iter/s, 5.17308s/12 iters), loss = 1.19256 I0406 07:46:15.717135 5226 solver.cpp:237] Train net output #0: loss = 1.19256 (* 1 = 1.19256 loss) I0406 07:46:15.717140 5226 sgd_solver.cpp:105] Iteration 4440, lr = 0.01 I0406 07:46:20.013849 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:46:21.126646 5226 solver.cpp:218] Iteration 4452 (2.21834 iter/s, 5.40946s/12 iters), loss = 1.46465 I0406 07:46:21.126682 5226 solver.cpp:237] Train net output #0: loss = 1.46465 (* 1 = 1.46465 loss) I0406 07:46:21.126688 5226 sgd_solver.cpp:105] Iteration 4452, lr = 0.01 I0406 07:46:26.508162 5226 solver.cpp:218] Iteration 4464 (2.22989 iter/s, 5.38143s/12 iters), loss = 1.21948 I0406 07:46:26.508199 5226 solver.cpp:237] Train net output #0: loss = 1.21948 (* 1 = 1.21948 loss) I0406 07:46:26.508204 5226 sgd_solver.cpp:105] Iteration 4464, lr = 0.01 I0406 07:46:31.598654 5226 solver.cpp:218] Iteration 4476 (2.35737 iter/s, 5.09041s/12 iters), loss = 1.38957 I0406 07:46:31.598690 5226 solver.cpp:237] Train net output #0: loss = 1.38957 (* 1 = 1.38957 loss) I0406 07:46:31.598695 5226 sgd_solver.cpp:105] Iteration 4476, lr = 0.01 I0406 07:46:36.164391 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4488.caffemodel I0406 07:46:39.195258 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4488.solverstate I0406 07:46:41.496449 5226 solver.cpp:330] Iteration 4488, Testing net (#0) I0406 07:46:41.496469 5226 net.cpp:676] Ignoring source layer train-data I0406 07:46:44.070693 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:46:45.857641 5226 solver.cpp:397] Test net output #0: accuracy = 0.313113 I0406 07:46:45.857677 5226 solver.cpp:397] Test net output #1: loss = 3.21875 (* 1 = 3.21875 loss) I0406 07:46:45.994784 5226 solver.cpp:218] Iteration 4488 (0.833566 iter/s, 14.396s/12 iters), loss = 1.40557 I0406 07:46:45.994833 5226 solver.cpp:237] Train net output #0: loss = 1.40557 (* 1 = 1.40557 loss) I0406 07:46:45.994843 5226 sgd_solver.cpp:105] Iteration 4488, lr = 0.01 I0406 07:46:50.418093 5226 solver.cpp:218] Iteration 4500 (2.71296 iter/s, 4.42322s/12 iters), loss = 1.31592 I0406 07:46:50.418135 5226 solver.cpp:237] Train net output #0: loss = 1.31592 (* 1 = 1.31592 loss) I0406 07:46:50.418141 5226 sgd_solver.cpp:105] Iteration 4500, lr = 0.01 I0406 07:46:55.890537 5226 solver.cpp:218] Iteration 4512 (2.19284 iter/s, 5.47235s/12 iters), loss = 1.1564 I0406 07:46:55.890573 5226 solver.cpp:237] Train net output #0: loss = 1.1564 (* 1 = 1.1564 loss) I0406 07:46:55.890579 5226 sgd_solver.cpp:105] Iteration 4512, lr = 0.01 I0406 07:47:01.150938 5226 solver.cpp:218] Iteration 4524 (2.28123 iter/s, 5.26031s/12 iters), loss = 1.20819 I0406 07:47:01.150975 5226 solver.cpp:237] Train net output #0: loss = 1.20819 (* 1 = 1.20819 loss) I0406 07:47:01.150980 5226 sgd_solver.cpp:105] Iteration 4524, lr = 0.01 I0406 07:47:06.358862 5226 solver.cpp:218] Iteration 4536 (2.30422 iter/s, 5.20784s/12 iters), loss = 1.46608 I0406 07:47:06.359016 5226 solver.cpp:237] Train net output #0: loss = 1.46608 (* 1 = 1.46608 loss) I0406 07:47:06.359025 5226 sgd_solver.cpp:105] Iteration 4536, lr = 0.01 I0406 07:47:11.272907 5226 solver.cpp:218] Iteration 4548 (2.44208 iter/s, 4.91385s/12 iters), loss = 1.52902 I0406 07:47:11.272943 5226 solver.cpp:237] Train net output #0: loss = 1.52902 (* 1 = 1.52902 loss) I0406 07:47:11.272949 5226 sgd_solver.cpp:105] Iteration 4548, lr = 0.01 I0406 07:47:12.659642 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:47:16.570510 5226 solver.cpp:218] Iteration 4560 (2.26521 iter/s, 5.29752s/12 iters), loss = 1.1028 I0406 07:47:16.570549 5226 solver.cpp:237] Train net output #0: loss = 1.1028 (* 1 = 1.1028 loss) I0406 07:47:16.570554 5226 sgd_solver.cpp:105] Iteration 4560, lr = 0.01 I0406 07:47:21.749428 5226 solver.cpp:218] Iteration 4572 (2.31713 iter/s, 5.17883s/12 iters), loss = 1.30269 I0406 07:47:21.749469 5226 solver.cpp:237] Train net output #0: loss = 1.30269 (* 1 = 1.30269 loss) I0406 07:47:21.749473 5226 sgd_solver.cpp:105] Iteration 4572, lr = 0.01 I0406 07:47:27.030598 5226 solver.cpp:218] Iteration 4584 (2.27226 iter/s, 5.28108s/12 iters), loss = 1.34215 I0406 07:47:27.030634 5226 solver.cpp:237] Train net output #0: loss = 1.34215 (* 1 = 1.34215 loss) I0406 07:47:27.030640 5226 sgd_solver.cpp:105] Iteration 4584, lr = 0.01 I0406 07:47:29.269740 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4590.caffemodel I0406 07:47:32.235402 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4590.solverstate I0406 07:47:34.538337 5226 solver.cpp:330] Iteration 4590, Testing net (#0) I0406 07:47:34.538357 5226 net.cpp:676] Ignoring source layer train-data I0406 07:47:37.032946 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:47:38.879949 5226 solver.cpp:397] Test net output #0: accuracy = 0.328431 I0406 07:47:38.879983 5226 solver.cpp:397] Test net output #1: loss = 2.97265 (* 1 = 2.97265 loss) I0406 07:47:40.945192 5226 solver.cpp:218] Iteration 4596 (0.862413 iter/s, 13.9144s/12 iters), loss = 1.18053 I0406 07:47:40.945243 5226 solver.cpp:237] Train net output #0: loss = 1.18053 (* 1 = 1.18053 loss) I0406 07:47:40.945251 5226 sgd_solver.cpp:105] Iteration 4596, lr = 0.01 I0406 07:47:46.331529 5226 solver.cpp:218] Iteration 4608 (2.2279 iter/s, 5.38624s/12 iters), loss = 0.950801 I0406 07:47:46.331580 5226 solver.cpp:237] Train net output #0: loss = 0.950801 (* 1 = 0.950801 loss) I0406 07:47:46.331589 5226 sgd_solver.cpp:105] Iteration 4608, lr = 0.01 I0406 07:47:51.763516 5226 solver.cpp:218] Iteration 4620 (2.20918 iter/s, 5.43189s/12 iters), loss = 1.30758 I0406 07:47:51.763554 5226 solver.cpp:237] Train net output #0: loss = 1.30758 (* 1 = 1.30758 loss) I0406 07:47:51.763561 5226 sgd_solver.cpp:105] Iteration 4620, lr = 0.01 I0406 07:47:57.007606 5226 solver.cpp:218] Iteration 4632 (2.28833 iter/s, 5.244s/12 iters), loss = 1.3201 I0406 07:47:57.007660 5226 solver.cpp:237] Train net output #0: loss = 1.3201 (* 1 = 1.3201 loss) I0406 07:47:57.007668 5226 sgd_solver.cpp:105] Iteration 4632, lr = 0.01 I0406 07:48:02.299629 5226 solver.cpp:218] Iteration 4644 (2.26761 iter/s, 5.29192s/12 iters), loss = 1.49239 I0406 07:48:02.299664 5226 solver.cpp:237] Train net output #0: loss = 1.49239 (* 1 = 1.49239 loss) I0406 07:48:02.299670 5226 sgd_solver.cpp:105] Iteration 4644, lr = 0.01 I0406 07:48:05.837280 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:48:07.581893 5226 solver.cpp:218] Iteration 4656 (2.27179 iter/s, 5.28218s/12 iters), loss = 1.3102 I0406 07:48:07.582015 5226 solver.cpp:237] Train net output #0: loss = 1.3102 (* 1 = 1.3102 loss) I0406 07:48:07.582020 5226 sgd_solver.cpp:105] Iteration 4656, lr = 0.01 I0406 07:48:12.990911 5226 solver.cpp:218] Iteration 4668 (2.21859 iter/s, 5.40885s/12 iters), loss = 1.36358 I0406 07:48:12.990952 5226 solver.cpp:237] Train net output #0: loss = 1.36358 (* 1 = 1.36358 loss) I0406 07:48:12.990957 5226 sgd_solver.cpp:105] Iteration 4668, lr = 0.01 I0406 07:48:18.295114 5226 solver.cpp:218] Iteration 4680 (2.26239 iter/s, 5.30411s/12 iters), loss = 1.43451 I0406 07:48:18.295151 5226 solver.cpp:237] Train net output #0: loss = 1.43451 (* 1 = 1.43451 loss) I0406 07:48:18.295156 5226 sgd_solver.cpp:105] Iteration 4680, lr = 0.01 I0406 07:48:23.088085 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4692.caffemodel I0406 07:48:26.172093 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4692.solverstate I0406 07:48:28.474866 5226 solver.cpp:330] Iteration 4692, Testing net (#0) I0406 07:48:28.474884 5226 net.cpp:676] Ignoring source layer train-data I0406 07:48:31.003047 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:48:32.812199 5226 solver.cpp:397] Test net output #0: accuracy = 0.322304 I0406 07:48:32.812232 5226 solver.cpp:397] Test net output #1: loss = 3.10933 (* 1 = 3.10933 loss) I0406 07:48:32.953163 5226 solver.cpp:218] Iteration 4692 (0.818671 iter/s, 14.6579s/12 iters), loss = 0.981167 I0406 07:48:32.953212 5226 solver.cpp:237] Train net output #0: loss = 0.981167 (* 1 = 0.981167 loss) I0406 07:48:32.953220 5226 sgd_solver.cpp:105] Iteration 4692, lr = 0.01 I0406 07:48:37.042342 5226 solver.cpp:218] Iteration 4704 (2.93464 iter/s, 4.08908s/12 iters), loss = 1.28855 I0406 07:48:37.042387 5226 solver.cpp:237] Train net output #0: loss = 1.28855 (* 1 = 1.28855 loss) I0406 07:48:37.042392 5226 sgd_solver.cpp:105] Iteration 4704, lr = 0.01 I0406 07:48:42.288740 5226 solver.cpp:218] Iteration 4716 (2.28733 iter/s, 5.2463s/12 iters), loss = 1.11105 I0406 07:48:42.288861 5226 solver.cpp:237] Train net output #0: loss = 1.11105 (* 1 = 1.11105 loss) I0406 07:48:42.288867 5226 sgd_solver.cpp:105] Iteration 4716, lr = 0.01 I0406 07:48:47.453068 5226 solver.cpp:218] Iteration 4728 (2.32371 iter/s, 5.16416s/12 iters), loss = 1.32187 I0406 07:48:47.453100 5226 solver.cpp:237] Train net output #0: loss = 1.32187 (* 1 = 1.32187 loss) I0406 07:48:47.453106 5226 sgd_solver.cpp:105] Iteration 4728, lr = 0.01 I0406 07:48:52.675240 5226 solver.cpp:218] Iteration 4740 (2.29793 iter/s, 5.2221s/12 iters), loss = 1.43182 I0406 07:48:52.675268 5226 solver.cpp:237] Train net output #0: loss = 1.43182 (* 1 = 1.43182 loss) I0406 07:48:52.675273 5226 sgd_solver.cpp:105] Iteration 4740, lr = 0.01 I0406 07:48:57.944686 5226 solver.cpp:218] Iteration 4752 (2.27731 iter/s, 5.26937s/12 iters), loss = 1.32995 I0406 07:48:57.944734 5226 solver.cpp:237] Train net output #0: loss = 1.32995 (* 1 = 1.32995 loss) I0406 07:48:57.944741 5226 sgd_solver.cpp:105] Iteration 4752, lr = 0.01 I0406 07:48:58.504801 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:49:03.316118 5226 solver.cpp:218] Iteration 4764 (2.23408 iter/s, 5.37133s/12 iters), loss = 1.32814 I0406 07:49:03.316164 5226 solver.cpp:237] Train net output #0: loss = 1.32814 (* 1 = 1.32814 loss) I0406 07:49:03.316174 5226 sgd_solver.cpp:105] Iteration 4764, lr = 0.01 I0406 07:49:08.534395 5226 solver.cpp:218] Iteration 4776 (2.29965 iter/s, 5.21818s/12 iters), loss = 1.42889 I0406 07:49:08.534432 5226 solver.cpp:237] Train net output #0: loss = 1.42889 (* 1 = 1.42889 loss) I0406 07:49:08.534437 5226 sgd_solver.cpp:105] Iteration 4776, lr = 0.01 I0406 07:49:13.869709 5226 solver.cpp:218] Iteration 4788 (2.2492 iter/s, 5.33523s/12 iters), loss = 1.29923 I0406 07:49:13.869825 5226 solver.cpp:237] Train net output #0: loss = 1.29923 (* 1 = 1.29923 loss) I0406 07:49:13.869834 5226 sgd_solver.cpp:105] Iteration 4788, lr = 0.01 I0406 07:49:16.005098 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4794.caffemodel I0406 07:49:19.011186 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4794.solverstate I0406 07:49:21.332480 5226 solver.cpp:330] Iteration 4794, Testing net (#0) I0406 07:49:21.332502 5226 net.cpp:676] Ignoring source layer train-data I0406 07:49:23.731123 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:49:25.586869 5226 solver.cpp:397] Test net output #0: accuracy = 0.324142 I0406 07:49:25.586905 5226 solver.cpp:397] Test net output #1: loss = 3.05492 (* 1 = 3.05492 loss) I0406 07:49:27.579016 5226 solver.cpp:218] Iteration 4800 (0.875331 iter/s, 13.7091s/12 iters), loss = 1.11227 I0406 07:49:27.579053 5226 solver.cpp:237] Train net output #0: loss = 1.11227 (* 1 = 1.11227 loss) I0406 07:49:27.579058 5226 sgd_solver.cpp:105] Iteration 4800, lr = 0.01 I0406 07:49:32.750728 5226 solver.cpp:218] Iteration 4812 (2.32035 iter/s, 5.17162s/12 iters), loss = 1.2966 I0406 07:49:32.750764 5226 solver.cpp:237] Train net output #0: loss = 1.2966 (* 1 = 1.2966 loss) I0406 07:49:32.750771 5226 sgd_solver.cpp:105] Iteration 4812, lr = 0.01 I0406 07:49:37.887214 5226 solver.cpp:218] Iteration 4824 (2.33627 iter/s, 5.1364s/12 iters), loss = 1.13206 I0406 07:49:37.887262 5226 solver.cpp:237] Train net output #0: loss = 1.13206 (* 1 = 1.13206 loss) I0406 07:49:37.887270 5226 sgd_solver.cpp:105] Iteration 4824, lr = 0.01 I0406 07:49:43.078636 5226 solver.cpp:218] Iteration 4836 (2.31155 iter/s, 5.19133s/12 iters), loss = 0.997291 I0406 07:49:43.078673 5226 solver.cpp:237] Train net output #0: loss = 0.997291 (* 1 = 0.997291 loss) I0406 07:49:43.078678 5226 sgd_solver.cpp:105] Iteration 4836, lr = 0.01 I0406 07:49:45.341719 5226 blocking_queue.cpp:49] Waiting for data I0406 07:49:48.468688 5226 solver.cpp:218] Iteration 4848 (2.22636 iter/s, 5.38996s/12 iters), loss = 1.74306 I0406 07:49:48.468725 5226 solver.cpp:237] Train net output #0: loss = 1.74306 (* 1 = 1.74306 loss) I0406 07:49:48.468730 5226 sgd_solver.cpp:105] Iteration 4848, lr = 0.01 I0406 07:49:51.164045 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:49:53.547287 5226 solver.cpp:218] Iteration 4860 (2.3629 iter/s, 5.07851s/12 iters), loss = 1.10049 I0406 07:49:53.547325 5226 solver.cpp:237] Train net output #0: loss = 1.10049 (* 1 = 1.10049 loss) I0406 07:49:53.547330 5226 sgd_solver.cpp:105] Iteration 4860, lr = 0.01 I0406 07:49:58.577877 5226 solver.cpp:218] Iteration 4872 (2.38545 iter/s, 5.0305s/12 iters), loss = 1.40458 I0406 07:49:58.577913 5226 solver.cpp:237] Train net output #0: loss = 1.40458 (* 1 = 1.40458 loss) I0406 07:49:58.577919 5226 sgd_solver.cpp:105] Iteration 4872, lr = 0.01 I0406 07:50:03.783627 5226 solver.cpp:218] Iteration 4884 (2.30518 iter/s, 5.20567s/12 iters), loss = 1.23893 I0406 07:50:03.783663 5226 solver.cpp:237] Train net output #0: loss = 1.23893 (* 1 = 1.23893 loss) I0406 07:50:03.783668 5226 sgd_solver.cpp:105] Iteration 4884, lr = 0.01 I0406 07:50:08.589906 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4896.caffemodel I0406 07:50:11.617627 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4896.solverstate I0406 07:50:13.915295 5226 solver.cpp:330] Iteration 4896, Testing net (#0) I0406 07:50:13.915314 5226 net.cpp:676] Ignoring source layer train-data I0406 07:50:16.306262 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:50:18.199350 5226 solver.cpp:397] Test net output #0: accuracy = 0.318015 I0406 07:50:18.199384 5226 solver.cpp:397] Test net output #1: loss = 3.18566 (* 1 = 3.18566 loss) I0406 07:50:18.340287 5226 solver.cpp:218] Iteration 4896 (0.824373 iter/s, 14.5565s/12 iters), loss = 1.30596 I0406 07:50:18.340335 5226 solver.cpp:237] Train net output #0: loss = 1.30596 (* 1 = 1.30596 loss) I0406 07:50:18.340343 5226 sgd_solver.cpp:105] Iteration 4896, lr = 0.01 I0406 07:50:22.697813 5226 solver.cpp:218] Iteration 4908 (2.75391 iter/s, 4.35744s/12 iters), loss = 1.08516 I0406 07:50:22.697849 5226 solver.cpp:237] Train net output #0: loss = 1.08516 (* 1 = 1.08516 loss) I0406 07:50:22.697854 5226 sgd_solver.cpp:105] Iteration 4908, lr = 0.01 I0406 07:50:27.925341 5226 solver.cpp:218] Iteration 4920 (2.29558 iter/s, 5.22745s/12 iters), loss = 1.15898 I0406 07:50:27.925379 5226 solver.cpp:237] Train net output #0: loss = 1.15898 (* 1 = 1.15898 loss) I0406 07:50:27.925384 5226 sgd_solver.cpp:105] Iteration 4920, lr = 0.01 I0406 07:50:33.311020 5226 solver.cpp:218] Iteration 4932 (2.22817 iter/s, 5.38559s/12 iters), loss = 1.06388 I0406 07:50:33.311055 5226 solver.cpp:237] Train net output #0: loss = 1.06388 (* 1 = 1.06388 loss) I0406 07:50:33.311060 5226 sgd_solver.cpp:105] Iteration 4932, lr = 0.01 I0406 07:50:38.662585 5226 solver.cpp:218] Iteration 4944 (2.24237 iter/s, 5.35148s/12 iters), loss = 1.10959 I0406 07:50:38.662622 5226 solver.cpp:237] Train net output #0: loss = 1.10959 (* 1 = 1.10959 loss) I0406 07:50:38.662627 5226 sgd_solver.cpp:105] Iteration 4944, lr = 0.01 I0406 07:50:43.623112 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:50:43.871531 5226 solver.cpp:218] Iteration 4956 (2.30377 iter/s, 5.20886s/12 iters), loss = 1.37315 I0406 07:50:43.871572 5226 solver.cpp:237] Train net output #0: loss = 1.37315 (* 1 = 1.37315 loss) I0406 07:50:43.871580 5226 sgd_solver.cpp:105] Iteration 4956, lr = 0.01 I0406 07:50:49.059175 5226 solver.cpp:218] Iteration 4968 (2.31323 iter/s, 5.18755s/12 iters), loss = 1.07251 I0406 07:50:49.059303 5226 solver.cpp:237] Train net output #0: loss = 1.07251 (* 1 = 1.07251 loss) I0406 07:50:49.059312 5226 sgd_solver.cpp:105] Iteration 4968, lr = 0.01 I0406 07:50:54.432083 5226 solver.cpp:218] Iteration 4980 (2.2335 iter/s, 5.37273s/12 iters), loss = 1.22782 I0406 07:50:54.432122 5226 solver.cpp:237] Train net output #0: loss = 1.22782 (* 1 = 1.22782 loss) I0406 07:50:54.432127 5226 sgd_solver.cpp:105] Iteration 4980, lr = 0.01 I0406 07:50:59.575440 5226 solver.cpp:218] Iteration 4992 (2.33315 iter/s, 5.14327s/12 iters), loss = 1.52138 I0406 07:50:59.575486 5226 solver.cpp:237] Train net output #0: loss = 1.52138 (* 1 = 1.52138 loss) I0406 07:50:59.575495 5226 sgd_solver.cpp:105] Iteration 4992, lr = 0.01 I0406 07:51:01.859560 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4998.caffemodel I0406 07:51:04.856849 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4998.solverstate I0406 07:51:08.376473 5226 solver.cpp:330] Iteration 4998, Testing net (#0) I0406 07:51:08.376492 5226 net.cpp:676] Ignoring source layer train-data I0406 07:51:10.747936 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:51:12.676488 5226 solver.cpp:397] Test net output #0: accuracy = 0.344363 I0406 07:51:12.676518 5226 solver.cpp:397] Test net output #1: loss = 3.11989 (* 1 = 3.11989 loss) I0406 07:51:14.786027 5226 solver.cpp:218] Iteration 5004 (0.788932 iter/s, 15.2104s/12 iters), loss = 1.57642 I0406 07:51:14.786064 5226 solver.cpp:237] Train net output #0: loss = 1.57642 (* 1 = 1.57642 loss) I0406 07:51:14.786069 5226 sgd_solver.cpp:105] Iteration 5004, lr = 0.01 I0406 07:51:20.009764 5226 solver.cpp:218] Iteration 5016 (2.29724 iter/s, 5.22365s/12 iters), loss = 1.19052 I0406 07:51:20.009902 5226 solver.cpp:237] Train net output #0: loss = 1.19052 (* 1 = 1.19052 loss) I0406 07:51:20.009912 5226 sgd_solver.cpp:105] Iteration 5016, lr = 0.01 I0406 07:51:25.383754 5226 solver.cpp:218] Iteration 5028 (2.23305 iter/s, 5.37381s/12 iters), loss = 1.14189 I0406 07:51:25.383810 5226 solver.cpp:237] Train net output #0: loss = 1.14189 (* 1 = 1.14189 loss) I0406 07:51:25.383818 5226 sgd_solver.cpp:105] Iteration 5028, lr = 0.01 I0406 07:51:30.635444 5226 solver.cpp:218] Iteration 5040 (2.28502 iter/s, 5.25159s/12 iters), loss = 1.32997 I0406 07:51:30.635479 5226 solver.cpp:237] Train net output #0: loss = 1.32997 (* 1 = 1.32997 loss) I0406 07:51:30.635484 5226 sgd_solver.cpp:105] Iteration 5040, lr = 0.01 I0406 07:51:35.896574 5226 solver.cpp:218] Iteration 5052 (2.28091 iter/s, 5.26105s/12 iters), loss = 0.935736 I0406 07:51:35.896610 5226 solver.cpp:237] Train net output #0: loss = 0.935736 (* 1 = 0.935736 loss) I0406 07:51:35.896615 5226 sgd_solver.cpp:105] Iteration 5052, lr = 0.01 I0406 07:51:37.912302 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:51:40.925587 5226 solver.cpp:218] Iteration 5064 (2.38619 iter/s, 5.02893s/12 iters), loss = 1.33313 I0406 07:51:40.925623 5226 solver.cpp:237] Train net output #0: loss = 1.33313 (* 1 = 1.33313 loss) I0406 07:51:40.925629 5226 sgd_solver.cpp:105] Iteration 5064, lr = 0.01 I0406 07:51:46.135932 5226 solver.cpp:218] Iteration 5076 (2.30315 iter/s, 5.21025s/12 iters), loss = 1.14384 I0406 07:51:46.135982 5226 solver.cpp:237] Train net output #0: loss = 1.14384 (* 1 = 1.14384 loss) I0406 07:51:46.135989 5226 sgd_solver.cpp:105] Iteration 5076, lr = 0.01 I0406 07:51:51.299016 5226 solver.cpp:218] Iteration 5088 (2.32423 iter/s, 5.16299s/12 iters), loss = 1.36084 I0406 07:51:51.299127 5226 solver.cpp:237] Train net output #0: loss = 1.36084 (* 1 = 1.36084 loss) I0406 07:51:51.299134 5226 sgd_solver.cpp:105] Iteration 5088, lr = 0.01 I0406 07:51:55.968628 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_5100.caffemodel I0406 07:51:58.899629 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_5100.solverstate I0406 07:52:01.277495 5226 solver.cpp:330] Iteration 5100, Testing net (#0) I0406 07:52:01.277518 5226 net.cpp:676] Ignoring source layer train-data I0406 07:52:03.613283 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:52:05.679862 5226 solver.cpp:397] Test net output #0: accuracy = 0.346814 I0406 07:52:05.679894 5226 solver.cpp:397] Test net output #1: loss = 3.0375 (* 1 = 3.0375 loss) I0406 07:52:05.816800 5226 solver.cpp:218] Iteration 5100 (0.826585 iter/s, 14.5176s/12 iters), loss = 0.861127 I0406 07:52:05.816870 5226 solver.cpp:237] Train net output #0: loss = 0.861127 (* 1 = 0.861127 loss) I0406 07:52:05.816879 5226 sgd_solver.cpp:105] Iteration 5100, lr = 0.01 I0406 07:52:10.214721 5226 solver.cpp:218] Iteration 5112 (2.72863 iter/s, 4.39781s/12 iters), loss = 1.16427 I0406 07:52:10.214761 5226 solver.cpp:237] Train net output #0: loss = 1.16427 (* 1 = 1.16427 loss) I0406 07:52:10.214768 5226 sgd_solver.cpp:105] Iteration 5112, lr = 0.01 I0406 07:52:15.449792 5226 solver.cpp:218] Iteration 5124 (2.29227 iter/s, 5.23499s/12 iters), loss = 1.01348 I0406 07:52:15.449826 5226 solver.cpp:237] Train net output #0: loss = 1.01348 (* 1 = 1.01348 loss) I0406 07:52:15.449831 5226 sgd_solver.cpp:105] Iteration 5124, lr = 0.01 I0406 07:52:20.839573 5226 solver.cpp:218] Iteration 5136 (2.22647 iter/s, 5.3897s/12 iters), loss = 1.17138 I0406 07:52:20.839609 5226 solver.cpp:237] Train net output #0: loss = 1.17138 (* 1 = 1.17138 loss) I0406 07:52:20.839615 5226 sgd_solver.cpp:105] Iteration 5136, lr = 0.01 I0406 07:52:26.101734 5226 solver.cpp:218] Iteration 5148 (2.28047 iter/s, 5.26207s/12 iters), loss = 1.29594 I0406 07:52:26.101835 5226 solver.cpp:237] Train net output #0: loss = 1.29594 (* 1 = 1.29594 loss) I0406 07:52:26.101841 5226 sgd_solver.cpp:105] Iteration 5148, lr = 0.01 I0406 07:52:30.450764 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:52:31.408334 5226 solver.cpp:218] Iteration 5160 (2.2614 iter/s, 5.30645s/12 iters), loss = 1.16029 I0406 07:52:31.408377 5226 solver.cpp:237] Train net output #0: loss = 1.16029 (* 1 = 1.16029 loss) I0406 07:52:31.408385 5226 sgd_solver.cpp:105] Iteration 5160, lr = 0.01 I0406 07:52:36.593050 5226 solver.cpp:218] Iteration 5172 (2.31453 iter/s, 5.18463s/12 iters), loss = 1.15684 I0406 07:52:36.593087 5226 solver.cpp:237] Train net output #0: loss = 1.15684 (* 1 = 1.15684 loss) I0406 07:52:36.593092 5226 sgd_solver.cpp:105] Iteration 5172, lr = 0.01 I0406 07:52:42.041817 5226 solver.cpp:218] Iteration 5184 (2.20237 iter/s, 5.44868s/12 iters), loss = 1.26785 I0406 07:52:42.041857 5226 solver.cpp:237] Train net output #0: loss = 1.26785 (* 1 = 1.26785 loss) I0406 07:52:42.041862 5226 sgd_solver.cpp:105] Iteration 5184, lr = 0.01 I0406 07:52:47.323772 5226 solver.cpp:218] Iteration 5196 (2.27192 iter/s, 5.28187s/12 iters), loss = 1.09619 I0406 07:52:47.323805 5226 solver.cpp:237] Train net output #0: loss = 1.09619 (* 1 = 1.09619 loss) I0406 07:52:47.323812 5226 sgd_solver.cpp:105] Iteration 5196, lr = 0.01 I0406 07:52:49.493296 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_5202.caffemodel I0406 07:52:52.527582 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_5202.solverstate I0406 07:52:54.836535 5226 solver.cpp:330] Iteration 5202, Testing net (#0) I0406 07:52:54.836553 5226 net.cpp:676] Ignoring source layer train-data I0406 07:52:57.153734 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:52:59.167555 5226 solver.cpp:397] Test net output #0: accuracy = 0.35049 I0406 07:52:59.167588 5226 solver.cpp:397] Test net output #1: loss = 2.91695 (* 1 = 2.91695 loss) I0406 07:53:00.998950 5226 solver.cpp:218] Iteration 5208 (0.877511 iter/s, 13.675s/12 iters), loss = 1.74033 I0406 07:53:00.998994 5226 solver.cpp:237] Train net output #0: loss = 1.74033 (* 1 = 1.74033 loss) I0406 07:53:00.999001 5226 sgd_solver.cpp:105] Iteration 5208, lr = 0.01 I0406 07:53:06.284822 5226 solver.cpp:218] Iteration 5220 (2.27025 iter/s, 5.28577s/12 iters), loss = 1.03007 I0406 07:53:06.284878 5226 solver.cpp:237] Train net output #0: loss = 1.03007 (* 1 = 1.03007 loss) I0406 07:53:06.284893 5226 sgd_solver.cpp:105] Iteration 5220, lr = 0.01 I0406 07:53:11.703543 5226 solver.cpp:218] Iteration 5232 (2.21459 iter/s, 5.41862s/12 iters), loss = 1.47741 I0406 07:53:11.703581 5226 solver.cpp:237] Train net output #0: loss = 1.47741 (* 1 = 1.47741 loss) I0406 07:53:11.703586 5226 sgd_solver.cpp:105] Iteration 5232, lr = 0.01 I0406 07:53:17.181150 5226 solver.cpp:218] Iteration 5244 (2.19077 iter/s, 5.47752s/12 iters), loss = 1.09277 I0406 07:53:17.181188 5226 solver.cpp:237] Train net output #0: loss = 1.09277 (* 1 = 1.09277 loss) I0406 07:53:17.181195 5226 sgd_solver.cpp:105] Iteration 5244, lr = 0.01 I0406 07:53:22.675824 5226 solver.cpp:218] Iteration 5256 (2.18397 iter/s, 5.49459s/12 iters), loss = 1.53562 I0406 07:53:22.675861 5226 solver.cpp:237] Train net output #0: loss = 1.53562 (* 1 = 1.53562 loss) I0406 07:53:22.675868 5226 sgd_solver.cpp:105] Iteration 5256, lr = 0.01 I0406 07:53:24.091971 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:53:28.068123 5226 solver.cpp:218] Iteration 5268 (2.22543 iter/s, 5.39221s/12 iters), loss = 0.990537 I0406 07:53:28.068215 5226 solver.cpp:237] Train net output #0: loss = 0.990537 (* 1 = 0.990537 loss) I0406 07:53:28.068224 5226 sgd_solver.cpp:105] Iteration 5268, lr = 0.01 I0406 07:53:33.424564 5226 solver.cpp:218] Iteration 5280 (2.24035 iter/s, 5.35631s/12 iters), loss = 1.52296 I0406 07:53:33.424605 5226 solver.cpp:237] Train net output #0: loss = 1.52296 (* 1 = 1.52296 loss) I0406 07:53:33.424612 5226 sgd_solver.cpp:105] Iteration 5280, lr = 0.01 I0406 07:53:38.622931 5226 solver.cpp:218] Iteration 5292 (2.30846 iter/s, 5.19828s/12 iters), loss = 1.00433 I0406 07:53:38.622969 5226 solver.cpp:237] Train net output #0: loss = 1.00433 (* 1 = 1.00433 loss) I0406 07:53:38.622975 5226 sgd_solver.cpp:105] Iteration 5292, lr = 0.01 I0406 07:53:43.351786 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_5304.caffemodel I0406 07:53:46.348980 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_5304.solverstate I0406 07:53:48.667948 5226 solver.cpp:330] Iteration 5304, Testing net (#0) I0406 07:53:48.667969 5226 net.cpp:676] Ignoring source layer train-data I0406 07:53:50.925163 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:53:52.988348 5226 solver.cpp:397] Test net output #0: accuracy = 0.354779 I0406 07:53:52.988379 5226 solver.cpp:397] Test net output #1: loss = 3.04381 (* 1 = 3.04381 loss) I0406 07:53:53.129091 5226 solver.cpp:218] Iteration 5304 (0.827243 iter/s, 14.506s/12 iters), loss = 1.08349 I0406 07:53:53.129135 5226 solver.cpp:237] Train net output #0: loss = 1.08349 (* 1 = 1.08349 loss) I0406 07:53:53.129142 5226 sgd_solver.cpp:105] Iteration 5304, lr = 0.01 I0406 07:53:57.527644 5226 solver.cpp:218] Iteration 5316 (2.72822 iter/s, 4.39847s/12 iters), loss = 0.944097 I0406 07:53:57.527681 5226 solver.cpp:237] Train net output #0: loss = 0.944097 (* 1 = 0.944097 loss) I0406 07:53:57.527686 5226 sgd_solver.cpp:105] Iteration 5316, lr = 0.01 I0406 07:54:02.780608 5226 solver.cpp:218] Iteration 5328 (2.28446 iter/s, 5.25288s/12 iters), loss = 1.18152 I0406 07:54:02.780743 5226 solver.cpp:237] Train net output #0: loss = 1.18152 (* 1 = 1.18152 loss) I0406 07:54:02.780750 5226 sgd_solver.cpp:105] Iteration 5328, lr = 0.01 I0406 07:54:08.022302 5226 solver.cpp:218] Iteration 5340 (2.28942 iter/s, 5.24151s/12 iters), loss = 1.12153 I0406 07:54:08.022338 5226 solver.cpp:237] Train net output #0: loss = 1.12153 (* 1 = 1.12153 loss) I0406 07:54:08.022344 5226 sgd_solver.cpp:105] Iteration 5340, lr = 0.01 I0406 07:54:13.409651 5226 solver.cpp:218] Iteration 5352 (2.22748 iter/s, 5.38726s/12 iters), loss = 1.08565 I0406 07:54:13.409687 5226 solver.cpp:237] Train net output #0: loss = 1.08565 (* 1 = 1.08565 loss) I0406 07:54:13.409693 5226 sgd_solver.cpp:105] Iteration 5352, lr = 0.01 I0406 07:54:16.865134 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:54:18.572110 5226 solver.cpp:218] Iteration 5364 (2.32451 iter/s, 5.16237s/12 iters), loss = 1.32962 I0406 07:54:18.572147 5226 solver.cpp:237] Train net output #0: loss = 1.32962 (* 1 = 1.32962 loss) I0406 07:54:18.572152 5226 sgd_solver.cpp:105] Iteration 5364, lr = 0.01 I0406 07:54:23.935873 5226 solver.cpp:218] Iteration 5376 (2.23727 iter/s, 5.36367s/12 iters), loss = 1.30589 I0406 07:54:23.935910 5226 solver.cpp:237] Train net output #0: loss = 1.30589 (* 1 = 1.30589 loss) I0406 07:54:23.935915 5226 sgd_solver.cpp:105] Iteration 5376, lr = 0.01 I0406 07:54:29.298303 5226 solver.cpp:218] Iteration 5388 (2.23783 iter/s, 5.36234s/12 iters), loss = 0.812986 I0406 07:54:29.298343 5226 solver.cpp:237] Train net output #0: loss = 0.812986 (* 1 = 0.812986 loss) I0406 07:54:29.298348 5226 sgd_solver.cpp:105] Iteration 5388, lr = 0.01 I0406 07:54:34.488752 5226 solver.cpp:218] Iteration 5400 (2.31197 iter/s, 5.19037s/12 iters), loss = 1.15998 I0406 07:54:34.488857 5226 solver.cpp:237] Train net output #0: loss = 1.15998 (* 1 = 1.15998 loss) I0406 07:54:34.488863 5226 sgd_solver.cpp:105] Iteration 5400, lr = 0.01 I0406 07:54:36.517652 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_5406.caffemodel I0406 07:54:39.576598 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_5406.solverstate I0406 07:54:41.876677 5226 solver.cpp:330] Iteration 5406, Testing net (#0) I0406 07:54:41.876695 5226 net.cpp:676] Ignoring source layer train-data I0406 07:54:44.064191 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:54:46.192556 5226 solver.cpp:397] Test net output #0: accuracy = 0.334559 I0406 07:54:46.192591 5226 solver.cpp:397] Test net output #1: loss = 3.06266 (* 1 = 3.06266 loss) I0406 07:54:47.929920 5226 solver.cpp:218] Iteration 5412 (0.892793 iter/s, 13.441s/12 iters), loss = 1.11387 I0406 07:54:47.929975 5226 solver.cpp:237] Train net output #0: loss = 1.11387 (* 1 = 1.11387 loss) I0406 07:54:47.929983 5226 sgd_solver.cpp:105] Iteration 5412, lr = 0.01 I0406 07:54:53.384435 5226 solver.cpp:218] Iteration 5424 (2.20005 iter/s, 5.45441s/12 iters), loss = 1.02624 I0406 07:54:53.384474 5226 solver.cpp:237] Train net output #0: loss = 1.02624 (* 1 = 1.02624 loss) I0406 07:54:53.384480 5226 sgd_solver.cpp:105] Iteration 5424, lr = 0.01 I0406 07:54:58.695457 5226 solver.cpp:218] Iteration 5436 (2.25949 iter/s, 5.31093s/12 iters), loss = 1.27154 I0406 07:54:58.695503 5226 solver.cpp:237] Train net output #0: loss = 1.27154 (* 1 = 1.27154 loss) I0406 07:54:58.695511 5226 sgd_solver.cpp:105] Iteration 5436, lr = 0.01 I0406 07:55:03.908650 5226 solver.cpp:218] Iteration 5448 (2.30189 iter/s, 5.2131s/12 iters), loss = 1.4067 I0406 07:55:03.908695 5226 solver.cpp:237] Train net output #0: loss = 1.4067 (* 1 = 1.4067 loss) I0406 07:55:03.908704 5226 sgd_solver.cpp:105] Iteration 5448, lr = 0.01 I0406 07:55:09.126276 5226 solver.cpp:218] Iteration 5460 (2.29993 iter/s, 5.21754s/12 iters), loss = 1.03105 I0406 07:55:09.126396 5226 solver.cpp:237] Train net output #0: loss = 1.03105 (* 1 = 1.03105 loss) I0406 07:55:09.126403 5226 sgd_solver.cpp:105] Iteration 5460, lr = 0.01 I0406 07:55:09.676029 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:55:14.406342 5226 solver.cpp:218] Iteration 5472 (2.27277 iter/s, 5.27989s/12 iters), loss = 1.14117 I0406 07:55:14.406391 5226 solver.cpp:237] Train net output #0: loss = 1.14117 (* 1 = 1.14117 loss) I0406 07:55:14.406399 5226 sgd_solver.cpp:105] Iteration 5472, lr = 0.01 I0406 07:55:19.547338 5226 solver.cpp:218] Iteration 5484 (2.33422 iter/s, 5.1409s/12 iters), loss = 1.30119 I0406 07:55:19.547377 5226 solver.cpp:237] Train net output #0: loss = 1.30119 (* 1 = 1.30119 loss) I0406 07:55:19.547384 5226 sgd_solver.cpp:105] Iteration 5484, lr = 0.01 I0406 07:55:24.599706 5226 solver.cpp:218] Iteration 5496 (2.37516 iter/s, 5.05228s/12 iters), loss = 1.03323 I0406 07:55:24.599747 5226 solver.cpp:237] Train net output #0: loss = 1.03323 (* 1 = 1.03323 loss) I0406 07:55:24.599753 5226 sgd_solver.cpp:105] Iteration 5496, lr = 0.01 I0406 07:55:29.300906 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_5508.caffemodel I0406 07:55:32.312181 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_5508.solverstate I0406 07:55:34.642812 5226 solver.cpp:330] Iteration 5508, Testing net (#0) I0406 07:55:34.642838 5226 net.cpp:676] Ignoring source layer train-data I0406 07:55:36.871526 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:55:39.030405 5226 solver.cpp:397] Test net output #0: accuracy = 0.334559 I0406 07:55:39.030437 5226 solver.cpp:397] Test net output #1: loss = 3.07989 (* 1 = 3.07989 loss) I0406 07:55:39.171162 5226 solver.cpp:218] Iteration 5508 (0.823536 iter/s, 14.5713s/12 iters), loss = 1.02905 I0406 07:55:39.171241 5226 solver.cpp:237] Train net output #0: loss = 1.02905 (* 1 = 1.02905 loss) I0406 07:55:39.171249 5226 sgd_solver.cpp:105] Iteration 5508, lr = 0.01 I0406 07:55:43.412154 5226 solver.cpp:218] Iteration 5520 (2.82961 iter/s, 4.24087s/12 iters), loss = 1.17859 I0406 07:55:43.412194 5226 solver.cpp:237] Train net output #0: loss = 1.17859 (* 1 = 1.17859 loss) I0406 07:55:43.412199 5226 sgd_solver.cpp:105] Iteration 5520, lr = 0.01 I0406 07:55:45.742982 5226 blocking_queue.cpp:49] Waiting for data I0406 07:55:48.512619 5226 solver.cpp:218] Iteration 5532 (2.35277 iter/s, 5.10038s/12 iters), loss = 0.951761 I0406 07:55:48.512670 5226 solver.cpp:237] Train net output #0: loss = 0.951761 (* 1 = 0.951761 loss) I0406 07:55:48.512677 5226 sgd_solver.cpp:105] Iteration 5532, lr = 0.01 I0406 07:55:53.811420 5226 solver.cpp:218] Iteration 5544 (2.26471 iter/s, 5.2987s/12 iters), loss = 0.972357 I0406 07:55:53.811457 5226 solver.cpp:237] Train net output #0: loss = 0.972357 (* 1 = 0.972357 loss) I0406 07:55:53.811462 5226 sgd_solver.cpp:105] Iteration 5544, lr = 0.01 I0406 07:55:59.233681 5226 solver.cpp:218] Iteration 5556 (2.21313 iter/s, 5.42218s/12 iters), loss = 1.21051 I0406 07:55:59.233721 5226 solver.cpp:237] Train net output #0: loss = 1.21051 (* 1 = 1.21051 loss) I0406 07:55:59.233726 5226 sgd_solver.cpp:105] Iteration 5556, lr = 0.01 I0406 07:56:02.053340 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:56:04.480522 5226 solver.cpp:218] Iteration 5568 (2.28713 iter/s, 5.24676s/12 iters), loss = 1.04115 I0406 07:56:04.480557 5226 solver.cpp:237] Train net output #0: loss = 1.04115 (* 1 = 1.04115 loss) I0406 07:56:04.480562 5226 sgd_solver.cpp:105] Iteration 5568, lr = 0.01 I0406 07:56:09.755160 5226 solver.cpp:218] Iteration 5580 (2.27507 iter/s, 5.27455s/12 iters), loss = 1.30793 I0406 07:56:09.755280 5226 solver.cpp:237] Train net output #0: loss = 1.30793 (* 1 = 1.30793 loss) I0406 07:56:09.755288 5226 sgd_solver.cpp:105] Iteration 5580, lr = 0.01 I0406 07:56:14.976517 5226 solver.cpp:218] Iteration 5592 (2.29832 iter/s, 5.22119s/12 iters), loss = 0.838414 I0406 07:56:14.976554 5226 solver.cpp:237] Train net output #0: loss = 0.838414 (* 1 = 0.838414 loss) I0406 07:56:14.976560 5226 sgd_solver.cpp:105] Iteration 5592, lr = 0.01 I0406 07:56:20.090829 5226 solver.cpp:218] Iteration 5604 (2.34639 iter/s, 5.11423s/12 iters), loss = 1.10552 I0406 07:56:20.090863 5226 solver.cpp:237] Train net output #0: loss = 1.10552 (* 1 = 1.10552 loss) I0406 07:56:20.090868 5226 sgd_solver.cpp:105] Iteration 5604, lr = 0.01 I0406 07:56:22.290532 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_5610.caffemodel I0406 07:56:25.311790 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_5610.solverstate I0406 07:56:27.677520 5226 solver.cpp:330] Iteration 5610, Testing net (#0) I0406 07:56:27.677542 5226 net.cpp:676] Ignoring source layer train-data I0406 07:56:29.842396 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:56:32.061404 5226 solver.cpp:397] Test net output #0: accuracy = 0.334559 I0406 07:56:32.061434 5226 solver.cpp:397] Test net output #1: loss = 3.17719 (* 1 = 3.17719 loss) I0406 07:56:33.922237 5226 solver.cpp:218] Iteration 5616 (0.867599 iter/s, 13.8313s/12 iters), loss = 1.44505 I0406 07:56:33.922276 5226 solver.cpp:237] Train net output #0: loss = 1.44505 (* 1 = 1.44505 loss) I0406 07:56:33.922281 5226 sgd_solver.cpp:105] Iteration 5616, lr = 0.01 I0406 07:56:39.288908 5226 solver.cpp:218] Iteration 5628 (2.23606 iter/s, 5.36659s/12 iters), loss = 0.971849 I0406 07:56:39.288946 5226 solver.cpp:237] Train net output #0: loss = 0.971849 (* 1 = 0.971849 loss) I0406 07:56:39.288952 5226 sgd_solver.cpp:105] Iteration 5628, lr = 0.01 I0406 07:56:44.606580 5226 solver.cpp:218] Iteration 5640 (2.25666 iter/s, 5.31759s/12 iters), loss = 0.906122 I0406 07:56:44.606673 5226 solver.cpp:237] Train net output #0: loss = 0.906122 (* 1 = 0.906122 loss) I0406 07:56:44.606679 5226 sgd_solver.cpp:105] Iteration 5640, lr = 0.01 I0406 07:56:49.992328 5226 solver.cpp:218] Iteration 5652 (2.22816 iter/s, 5.38561s/12 iters), loss = 1.24248 I0406 07:56:49.992367 5226 solver.cpp:237] Train net output #0: loss = 1.24248 (* 1 = 1.24248 loss) I0406 07:56:49.992372 5226 sgd_solver.cpp:105] Iteration 5652, lr = 0.01 I0406 07:56:55.100517 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:56:55.319990 5226 solver.cpp:218] Iteration 5664 (2.25243 iter/s, 5.32757s/12 iters), loss = 1.10681 I0406 07:56:55.320027 5226 solver.cpp:237] Train net output #0: loss = 1.10681 (* 1 = 1.10681 loss) I0406 07:56:55.320034 5226 sgd_solver.cpp:105] Iteration 5664, lr = 0.01 I0406 07:57:00.502889 5226 solver.cpp:218] Iteration 5676 (2.31534 iter/s, 5.18281s/12 iters), loss = 1.09994 I0406 07:57:00.502928 5226 solver.cpp:237] Train net output #0: loss = 1.09994 (* 1 = 1.09994 loss) I0406 07:57:00.502933 5226 sgd_solver.cpp:105] Iteration 5676, lr = 0.01 I0406 07:57:05.777714 5226 solver.cpp:218] Iteration 5688 (2.275 iter/s, 5.27474s/12 iters), loss = 1.40267 I0406 07:57:05.777753 5226 solver.cpp:237] Train net output #0: loss = 1.40267 (* 1 = 1.40267 loss) I0406 07:57:05.777758 5226 sgd_solver.cpp:105] Iteration 5688, lr = 0.01 I0406 07:57:11.109906 5226 solver.cpp:218] Iteration 5700 (2.25052 iter/s, 5.33211s/12 iters), loss = 1.38615 I0406 07:57:11.109942 5226 solver.cpp:237] Train net output #0: loss = 1.38615 (* 1 = 1.38615 loss) I0406 07:57:11.109948 5226 sgd_solver.cpp:105] Iteration 5700, lr = 0.01 I0406 07:57:15.860585 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_5712.caffemodel I0406 07:57:18.809514 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_5712.solverstate I0406 07:57:21.145617 5226 solver.cpp:330] Iteration 5712, Testing net (#0) I0406 07:57:21.145648 5226 net.cpp:676] Ignoring source layer train-data I0406 07:57:23.282548 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:57:25.585129 5226 solver.cpp:397] Test net output #0: accuracy = 0.35049 I0406 07:57:25.585160 5226 solver.cpp:397] Test net output #1: loss = 3.0728 (* 1 = 3.0728 loss) I0406 07:57:25.726086 5226 solver.cpp:218] Iteration 5712 (0.821016 iter/s, 14.616s/12 iters), loss = 1.39695 I0406 07:57:25.726125 5226 solver.cpp:237] Train net output #0: loss = 1.39695 (* 1 = 1.39695 loss) I0406 07:57:25.726131 5226 sgd_solver.cpp:105] Iteration 5712, lr = 0.01 I0406 07:57:30.086277 5226 solver.cpp:218] Iteration 5724 (2.75223 iter/s, 4.3601s/12 iters), loss = 1.34678 I0406 07:57:30.086318 5226 solver.cpp:237] Train net output #0: loss = 1.34678 (* 1 = 1.34678 loss) I0406 07:57:30.086323 5226 sgd_solver.cpp:105] Iteration 5724, lr = 0.01 I0406 07:57:35.027580 5226 solver.cpp:218] Iteration 5736 (2.42855 iter/s, 4.94122s/12 iters), loss = 1.1855 I0406 07:57:35.027617 5226 solver.cpp:237] Train net output #0: loss = 1.1855 (* 1 = 1.1855 loss) I0406 07:57:35.027622 5226 sgd_solver.cpp:105] Iteration 5736, lr = 0.01 I0406 07:57:40.478971 5226 solver.cpp:218] Iteration 5748 (2.20131 iter/s, 5.4513s/12 iters), loss = 1.15293 I0406 07:57:40.479007 5226 solver.cpp:237] Train net output #0: loss = 1.15293 (* 1 = 1.15293 loss) I0406 07:57:40.479012 5226 sgd_solver.cpp:105] Iteration 5748, lr = 0.01 I0406 07:57:45.576488 5226 solver.cpp:218] Iteration 5760 (2.35413 iter/s, 5.09743s/12 iters), loss = 0.949511 I0406 07:57:45.576532 5226 solver.cpp:237] Train net output #0: loss = 0.949511 (* 1 = 0.949511 loss) I0406 07:57:45.576537 5226 sgd_solver.cpp:105] Iteration 5760, lr = 0.01 I0406 07:57:47.594240 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:57:50.879992 5226 solver.cpp:218] Iteration 5772 (2.2627 iter/s, 5.30341s/12 iters), loss = 0.99732 I0406 07:57:50.880030 5226 solver.cpp:237] Train net output #0: loss = 0.99732 (* 1 = 0.99732 loss) I0406 07:57:50.880036 5226 sgd_solver.cpp:105] Iteration 5772, lr = 0.01 I0406 07:57:56.168972 5226 solver.cpp:218] Iteration 5784 (2.2689 iter/s, 5.2889s/12 iters), loss = 1.24832 I0406 07:57:56.169008 5226 solver.cpp:237] Train net output #0: loss = 1.24832 (* 1 = 1.24832 loss) I0406 07:57:56.169013 5226 sgd_solver.cpp:105] Iteration 5784, lr = 0.01 I0406 07:58:01.455111 5226 solver.cpp:218] Iteration 5796 (2.27012 iter/s, 5.28606s/12 iters), loss = 1.11876 I0406 07:58:01.455149 5226 solver.cpp:237] Train net output #0: loss = 1.11876 (* 1 = 1.11876 loss) I0406 07:58:01.455154 5226 sgd_solver.cpp:105] Iteration 5796, lr = 0.01 I0406 07:58:06.537432 5226 solver.cpp:218] Iteration 5808 (2.36117 iter/s, 5.08224s/12 iters), loss = 1.09874 I0406 07:58:06.537472 5226 solver.cpp:237] Train net output #0: loss = 1.09874 (* 1 = 1.09874 loss) I0406 07:58:06.537478 5226 sgd_solver.cpp:105] Iteration 5808, lr = 0.01 I0406 07:58:08.592327 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_5814.caffemodel I0406 07:58:11.616477 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_5814.solverstate I0406 07:58:13.934227 5226 solver.cpp:330] Iteration 5814, Testing net (#0) I0406 07:58:13.934252 5226 net.cpp:676] Ignoring source layer train-data I0406 07:58:15.990747 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:58:18.278738 5226 solver.cpp:397] Test net output #0: accuracy = 0.318015 I0406 07:58:18.278872 5226 solver.cpp:397] Test net output #1: loss = 3.12658 (* 1 = 3.12658 loss) I0406 07:58:20.190635 5226 solver.cpp:218] Iteration 5820 (0.878924 iter/s, 13.6531s/12 iters), loss = 1.30857 I0406 07:58:20.190675 5226 solver.cpp:237] Train net output #0: loss = 1.30857 (* 1 = 1.30857 loss) I0406 07:58:20.190685 5226 sgd_solver.cpp:105] Iteration 5820, lr = 0.01 I0406 07:58:25.484483 5226 solver.cpp:218] Iteration 5832 (2.26682 iter/s, 5.29376s/12 iters), loss = 1.00455 I0406 07:58:25.484529 5226 solver.cpp:237] Train net output #0: loss = 1.00455 (* 1 = 1.00455 loss) I0406 07:58:25.484536 5226 sgd_solver.cpp:105] Iteration 5832, lr = 0.01 I0406 07:58:30.603885 5226 solver.cpp:218] Iteration 5844 (2.34407 iter/s, 5.11931s/12 iters), loss = 1.32372 I0406 07:58:30.603924 5226 solver.cpp:237] Train net output #0: loss = 1.32372 (* 1 = 1.32372 loss) I0406 07:58:30.603929 5226 sgd_solver.cpp:105] Iteration 5844, lr = 0.01 I0406 07:58:35.937963 5226 solver.cpp:218] Iteration 5856 (2.24972 iter/s, 5.33399s/12 iters), loss = 1.39658 I0406 07:58:35.938001 5226 solver.cpp:237] Train net output #0: loss = 1.39658 (* 1 = 1.39658 loss) I0406 07:58:35.938006 5226 sgd_solver.cpp:105] Iteration 5856, lr = 0.01 I0406 07:58:40.245982 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:58:41.125120 5226 solver.cpp:218] Iteration 5868 (2.31344 iter/s, 5.18707s/12 iters), loss = 1.18406 I0406 07:58:41.125157 5226 solver.cpp:237] Train net output #0: loss = 1.18406 (* 1 = 1.18406 loss) I0406 07:58:41.125164 5226 sgd_solver.cpp:105] Iteration 5868, lr = 0.01 I0406 07:58:46.382045 5226 solver.cpp:218] Iteration 5880 (2.28274 iter/s, 5.25684s/12 iters), loss = 1.24425 I0406 07:58:46.382082 5226 solver.cpp:237] Train net output #0: loss = 1.24425 (* 1 = 1.24425 loss) I0406 07:58:46.382087 5226 sgd_solver.cpp:105] Iteration 5880, lr = 0.01 I0406 07:58:51.854719 5226 solver.cpp:218] Iteration 5892 (2.19275 iter/s, 5.47259s/12 iters), loss = 1.18952 I0406 07:58:51.854820 5226 solver.cpp:237] Train net output #0: loss = 1.18952 (* 1 = 1.18952 loss) I0406 07:58:51.854825 5226 sgd_solver.cpp:105] Iteration 5892, lr = 0.01 I0406 07:58:57.209131 5226 solver.cpp:218] Iteration 5904 (2.2412 iter/s, 5.35426s/12 iters), loss = 0.900855 I0406 07:58:57.209168 5226 solver.cpp:237] Train net output #0: loss = 0.900855 (* 1 = 0.900855 loss) I0406 07:58:57.209174 5226 sgd_solver.cpp:105] Iteration 5904, lr = 0.01 I0406 07:59:02.069214 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_5916.caffemodel I0406 07:59:05.094641 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_5916.solverstate I0406 07:59:07.398937 5226 solver.cpp:330] Iteration 5916, Testing net (#0) I0406 07:59:07.398957 5226 net.cpp:676] Ignoring source layer train-data I0406 07:59:09.425787 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:59:11.757273 5226 solver.cpp:397] Test net output #0: accuracy = 0.316176 I0406 07:59:11.757319 5226 solver.cpp:397] Test net output #1: loss = 3.25887 (* 1 = 3.25887 loss) I0406 07:59:11.886507 5226 solver.cpp:218] Iteration 5916 (0.817593 iter/s, 14.6772s/12 iters), loss = 1.1649 I0406 07:59:11.886559 5226 solver.cpp:237] Train net output #0: loss = 1.1649 (* 1 = 1.1649 loss) I0406 07:59:11.886567 5226 sgd_solver.cpp:105] Iteration 5916, lr = 0.01 I0406 07:59:16.450744 5226 solver.cpp:218] Iteration 5928 (2.62919 iter/s, 4.56414s/12 iters), loss = 0.993896 I0406 07:59:16.450778 5226 solver.cpp:237] Train net output #0: loss = 0.993896 (* 1 = 0.993896 loss) I0406 07:59:16.450784 5226 sgd_solver.cpp:105] Iteration 5928, lr = 0.01 I0406 07:59:21.859767 5226 solver.cpp:218] Iteration 5940 (2.21855 iter/s, 5.40894s/12 iters), loss = 1.29924 I0406 07:59:21.859846 5226 solver.cpp:237] Train net output #0: loss = 1.29924 (* 1 = 1.29924 loss) I0406 07:59:21.859853 5226 sgd_solver.cpp:105] Iteration 5940, lr = 0.01 I0406 07:59:26.928897 5226 solver.cpp:218] Iteration 5952 (2.36733 iter/s, 5.069s/12 iters), loss = 1.43333 I0406 07:59:26.928936 5226 solver.cpp:237] Train net output #0: loss = 1.43333 (* 1 = 1.43333 loss) I0406 07:59:26.928942 5226 sgd_solver.cpp:105] Iteration 5952, lr = 0.01 I0406 07:59:32.220765 5226 solver.cpp:218] Iteration 5964 (2.26767 iter/s, 5.29178s/12 iters), loss = 1.31222 I0406 07:59:32.220800 5226 solver.cpp:237] Train net output #0: loss = 1.31222 (* 1 = 1.31222 loss) I0406 07:59:32.220805 5226 sgd_solver.cpp:105] Iteration 5964, lr = 0.01 I0406 07:59:33.680019 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 07:59:37.511986 5226 solver.cpp:218] Iteration 5976 (2.26794 iter/s, 5.29114s/12 iters), loss = 0.987962 I0406 07:59:37.512027 5226 solver.cpp:237] Train net output #0: loss = 0.987962 (* 1 = 0.987962 loss) I0406 07:59:37.512033 5226 sgd_solver.cpp:105] Iteration 5976, lr = 0.01 I0406 07:59:42.967118 5226 solver.cpp:218] Iteration 5988 (2.1998 iter/s, 5.45504s/12 iters), loss = 0.953547 I0406 07:59:42.967173 5226 solver.cpp:237] Train net output #0: loss = 0.953547 (* 1 = 0.953547 loss) I0406 07:59:42.967182 5226 sgd_solver.cpp:105] Iteration 5988, lr = 0.01 I0406 07:59:48.254253 5226 solver.cpp:218] Iteration 6000 (2.26971 iter/s, 5.28703s/12 iters), loss = 1.1414 I0406 07:59:48.254300 5226 solver.cpp:237] Train net output #0: loss = 1.1414 (* 1 = 1.1414 loss) I0406 07:59:48.254307 5226 sgd_solver.cpp:105] Iteration 6000, lr = 0.01 I0406 07:59:53.455752 5226 solver.cpp:218] Iteration 6012 (2.30707 iter/s, 5.2014s/12 iters), loss = 1.10675 I0406 07:59:53.456156 5226 solver.cpp:237] Train net output #0: loss = 1.10675 (* 1 = 1.10675 loss) I0406 07:59:53.456166 5226 sgd_solver.cpp:105] Iteration 6012, lr = 0.01 I0406 07:59:55.603401 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6018.caffemodel I0406 07:59:58.593271 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6018.solverstate I0406 08:00:00.905632 5226 solver.cpp:330] Iteration 6018, Testing net (#0) I0406 08:00:00.905651 5226 net.cpp:676] Ignoring source layer train-data I0406 08:00:02.871464 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:00:05.298772 5226 solver.cpp:397] Test net output #0: accuracy = 0.311274 I0406 08:00:05.298804 5226 solver.cpp:397] Test net output #1: loss = 3.28846 (* 1 = 3.28846 loss) I0406 08:00:07.238652 5226 solver.cpp:218] Iteration 6024 (0.870676 iter/s, 13.7824s/12 iters), loss = 0.95152 I0406 08:00:07.238687 5226 solver.cpp:237] Train net output #0: loss = 0.95152 (* 1 = 0.95152 loss) I0406 08:00:07.238693 5226 sgd_solver.cpp:105] Iteration 6024, lr = 0.01 I0406 08:00:12.680160 5226 solver.cpp:218] Iteration 6036 (2.20531 iter/s, 5.44142s/12 iters), loss = 1.35103 I0406 08:00:12.680195 5226 solver.cpp:237] Train net output #0: loss = 1.35103 (* 1 = 1.35103 loss) I0406 08:00:12.680200 5226 sgd_solver.cpp:105] Iteration 6036, lr = 0.01 I0406 08:00:18.061496 5226 solver.cpp:218] Iteration 6048 (2.22996 iter/s, 5.38125s/12 iters), loss = 1.21086 I0406 08:00:18.061534 5226 solver.cpp:237] Train net output #0: loss = 1.21086 (* 1 = 1.21086 loss) I0406 08:00:18.061540 5226 sgd_solver.cpp:105] Iteration 6048, lr = 0.01 I0406 08:00:23.017666 5226 solver.cpp:218] Iteration 6060 (2.42127 iter/s, 4.95608s/12 iters), loss = 1.25787 I0406 08:00:23.017705 5226 solver.cpp:237] Train net output #0: loss = 1.25787 (* 1 = 1.25787 loss) I0406 08:00:23.017711 5226 sgd_solver.cpp:105] Iteration 6060, lr = 0.01 I0406 08:00:26.749205 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:00:28.324869 5226 solver.cpp:218] Iteration 6072 (2.26112 iter/s, 5.30711s/12 iters), loss = 1.127 I0406 08:00:28.324923 5226 solver.cpp:237] Train net output #0: loss = 1.127 (* 1 = 1.127 loss) I0406 08:00:28.324931 5226 sgd_solver.cpp:105] Iteration 6072, lr = 0.01 I0406 08:00:33.528533 5226 solver.cpp:218] Iteration 6084 (2.30611 iter/s, 5.20356s/12 iters), loss = 1.00436 I0406 08:00:33.528582 5226 solver.cpp:237] Train net output #0: loss = 1.00436 (* 1 = 1.00436 loss) I0406 08:00:33.528591 5226 sgd_solver.cpp:105] Iteration 6084, lr = 0.01 I0406 08:00:38.765733 5226 solver.cpp:218] Iteration 6096 (2.29134 iter/s, 5.2371s/12 iters), loss = 0.84998 I0406 08:00:38.765790 5226 solver.cpp:237] Train net output #0: loss = 0.84998 (* 1 = 0.84998 loss) I0406 08:00:38.765799 5226 sgd_solver.cpp:105] Iteration 6096, lr = 0.01 I0406 08:00:44.034705 5226 solver.cpp:218] Iteration 6108 (2.27753 iter/s, 5.26887s/12 iters), loss = 1.26552 I0406 08:00:44.034760 5226 solver.cpp:237] Train net output #0: loss = 1.26552 (* 1 = 1.26552 loss) I0406 08:00:44.034768 5226 sgd_solver.cpp:105] Iteration 6108, lr = 0.01 I0406 08:00:48.858479 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6120.caffemodel I0406 08:00:51.841418 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6120.solverstate I0406 08:00:54.135193 5226 solver.cpp:330] Iteration 6120, Testing net (#0) I0406 08:00:54.135213 5226 net.cpp:676] Ignoring source layer train-data I0406 08:00:56.050230 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:00:58.405539 5226 solver.cpp:397] Test net output #0: accuracy = 0.321691 I0406 08:00:58.405665 5226 solver.cpp:397] Test net output #1: loss = 3.23157 (* 1 = 3.23157 loss) I0406 08:00:58.545665 5226 solver.cpp:218] Iteration 6120 (0.82697 iter/s, 14.5108s/12 iters), loss = 1.00214 I0406 08:00:58.545711 5226 solver.cpp:237] Train net output #0: loss = 1.00214 (* 1 = 1.00214 loss) I0406 08:00:58.545719 5226 sgd_solver.cpp:105] Iteration 6120, lr = 0.01 I0406 08:01:03.116647 5226 solver.cpp:218] Iteration 6132 (2.62531 iter/s, 4.5709s/12 iters), loss = 1.36882 I0406 08:01:03.116683 5226 solver.cpp:237] Train net output #0: loss = 1.36882 (* 1 = 1.36882 loss) I0406 08:01:03.116688 5226 sgd_solver.cpp:105] Iteration 6132, lr = 0.01 I0406 08:01:08.448292 5226 solver.cpp:218] Iteration 6144 (2.25075 iter/s, 5.33156s/12 iters), loss = 0.890568 I0406 08:01:08.448330 5226 solver.cpp:237] Train net output #0: loss = 0.890568 (* 1 = 0.890568 loss) I0406 08:01:08.448335 5226 sgd_solver.cpp:105] Iteration 6144, lr = 0.01 I0406 08:01:13.513116 5226 solver.cpp:218] Iteration 6156 (2.36932 iter/s, 5.06474s/12 iters), loss = 1.2117 I0406 08:01:13.513159 5226 solver.cpp:237] Train net output #0: loss = 1.2117 (* 1 = 1.2117 loss) I0406 08:01:13.513164 5226 sgd_solver.cpp:105] Iteration 6156, lr = 0.01 I0406 08:01:18.960605 5226 solver.cpp:218] Iteration 6168 (2.20289 iter/s, 5.44739s/12 iters), loss = 1.02632 I0406 08:01:18.960654 5226 solver.cpp:237] Train net output #0: loss = 1.02632 (* 1 = 1.02632 loss) I0406 08:01:18.960664 5226 sgd_solver.cpp:105] Iteration 6168, lr = 0.01 I0406 08:01:19.535945 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:01:24.060144 5226 solver.cpp:218] Iteration 6180 (2.3532 iter/s, 5.09945s/12 iters), loss = 1.03014 I0406 08:01:24.060179 5226 solver.cpp:237] Train net output #0: loss = 1.03014 (* 1 = 1.03014 loss) I0406 08:01:24.060184 5226 sgd_solver.cpp:105] Iteration 6180, lr = 0.01 I0406 08:01:29.351596 5226 solver.cpp:218] Iteration 6192 (2.26785 iter/s, 5.29136s/12 iters), loss = 1.10981 I0406 08:01:29.352939 5226 solver.cpp:237] Train net output #0: loss = 1.10981 (* 1 = 1.10981 loss) I0406 08:01:29.352952 5226 sgd_solver.cpp:105] Iteration 6192, lr = 0.01 I0406 08:01:34.426587 5226 solver.cpp:218] Iteration 6204 (2.36518 iter/s, 5.0736s/12 iters), loss = 1.01274 I0406 08:01:34.426626 5226 solver.cpp:237] Train net output #0: loss = 1.01274 (* 1 = 1.01274 loss) I0406 08:01:34.426632 5226 sgd_solver.cpp:105] Iteration 6204, lr = 0.01 I0406 08:01:39.719352 5226 solver.cpp:218] Iteration 6216 (2.26728 iter/s, 5.29268s/12 iters), loss = 1.49139 I0406 08:01:39.719386 5226 solver.cpp:237] Train net output #0: loss = 1.49139 (* 1 = 1.49139 loss) I0406 08:01:39.719391 5226 sgd_solver.cpp:105] Iteration 6216, lr = 0.01 I0406 08:01:41.993844 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6222.caffemodel I0406 08:01:44.933007 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6222.solverstate I0406 08:01:47.288713 5226 solver.cpp:330] Iteration 6222, Testing net (#0) I0406 08:01:47.288736 5226 net.cpp:676] Ignoring source layer train-data I0406 08:01:49.157588 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:01:50.411355 5226 blocking_queue.cpp:49] Waiting for data I0406 08:01:51.560547 5226 solver.cpp:397] Test net output #0: accuracy = 0.320466 I0406 08:01:51.560597 5226 solver.cpp:397] Test net output #1: loss = 3.17218 (* 1 = 3.17218 loss) I0406 08:01:53.476820 5226 solver.cpp:218] Iteration 6228 (0.872262 iter/s, 13.7573s/12 iters), loss = 1.21917 I0406 08:01:53.476859 5226 solver.cpp:237] Train net output #0: loss = 1.21917 (* 1 = 1.21917 loss) I0406 08:01:53.476866 5226 sgd_solver.cpp:105] Iteration 6228, lr = 0.01 I0406 08:01:58.920570 5226 solver.cpp:218] Iteration 6240 (2.2044 iter/s, 5.44366s/12 iters), loss = 1.03296 I0406 08:01:58.920608 5226 solver.cpp:237] Train net output #0: loss = 1.03296 (* 1 = 1.03296 loss) I0406 08:01:58.920614 5226 sgd_solver.cpp:105] Iteration 6240, lr = 0.01 I0406 08:02:04.259903 5226 solver.cpp:218] Iteration 6252 (2.24751 iter/s, 5.33925s/12 iters), loss = 1.21089 I0406 08:02:04.260027 5226 solver.cpp:237] Train net output #0: loss = 1.21089 (* 1 = 1.21089 loss) I0406 08:02:04.260035 5226 sgd_solver.cpp:105] Iteration 6252, lr = 0.01 I0406 08:02:09.588688 5226 solver.cpp:218] Iteration 6264 (2.252 iter/s, 5.32861s/12 iters), loss = 1.431 I0406 08:02:09.588734 5226 solver.cpp:237] Train net output #0: loss = 1.431 (* 1 = 1.431 loss) I0406 08:02:09.588742 5226 sgd_solver.cpp:105] Iteration 6264, lr = 0.01 I0406 08:02:12.515545 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:02:15.001204 5226 solver.cpp:218] Iteration 6276 (2.21712 iter/s, 5.41242s/12 iters), loss = 1.16491 I0406 08:02:15.001240 5226 solver.cpp:237] Train net output #0: loss = 1.16491 (* 1 = 1.16491 loss) I0406 08:02:15.001245 5226 sgd_solver.cpp:105] Iteration 6276, lr = 0.01 I0406 08:02:20.155663 5226 solver.cpp:218] Iteration 6288 (2.32812 iter/s, 5.15438s/12 iters), loss = 1.01929 I0406 08:02:20.155700 5226 solver.cpp:237] Train net output #0: loss = 1.01929 (* 1 = 1.01929 loss) I0406 08:02:20.155705 5226 sgd_solver.cpp:105] Iteration 6288, lr = 0.01 I0406 08:02:25.515905 5226 solver.cpp:218] Iteration 6300 (2.23874 iter/s, 5.36016s/12 iters), loss = 1.25234 I0406 08:02:25.515945 5226 solver.cpp:237] Train net output #0: loss = 1.25234 (* 1 = 1.25234 loss) I0406 08:02:25.515950 5226 sgd_solver.cpp:105] Iteration 6300, lr = 0.01 I0406 08:02:30.896199 5226 solver.cpp:218] Iteration 6312 (2.2304 iter/s, 5.38021s/12 iters), loss = 1.21727 I0406 08:02:30.896234 5226 solver.cpp:237] Train net output #0: loss = 1.21727 (* 1 = 1.21727 loss) I0406 08:02:30.896239 5226 sgd_solver.cpp:105] Iteration 6312, lr = 0.01 I0406 08:02:35.585449 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6324.caffemodel I0406 08:02:38.565816 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6324.solverstate I0406 08:02:40.891275 5226 solver.cpp:330] Iteration 6324, Testing net (#0) I0406 08:02:40.891300 5226 net.cpp:676] Ignoring source layer train-data I0406 08:02:42.803297 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:02:45.275792 5226 solver.cpp:397] Test net output #0: accuracy = 0.3125 I0406 08:02:45.275820 5226 solver.cpp:397] Test net output #1: loss = 3.16637 (* 1 = 3.16637 loss) I0406 08:02:45.416664 5226 solver.cpp:218] Iteration 6324 (0.826428 iter/s, 14.5203s/12 iters), loss = 1.13985 I0406 08:02:45.416723 5226 solver.cpp:237] Train net output #0: loss = 1.13985 (* 1 = 1.13985 loss) I0406 08:02:45.416728 5226 sgd_solver.cpp:105] Iteration 6324, lr = 0.01 I0406 08:02:49.747298 5226 solver.cpp:218] Iteration 6336 (2.77102 iter/s, 4.33053s/12 iters), loss = 1.094 I0406 08:02:49.747346 5226 solver.cpp:237] Train net output #0: loss = 1.094 (* 1 = 1.094 loss) I0406 08:02:49.747354 5226 sgd_solver.cpp:105] Iteration 6336, lr = 0.01 I0406 08:02:55.001536 5226 solver.cpp:218] Iteration 6348 (2.28391 iter/s, 5.25414s/12 iters), loss = 0.723623 I0406 08:02:55.001574 5226 solver.cpp:237] Train net output #0: loss = 0.723623 (* 1 = 0.723623 loss) I0406 08:02:55.001579 5226 sgd_solver.cpp:105] Iteration 6348, lr = 0.01 I0406 08:03:00.353610 5226 solver.cpp:218] Iteration 6360 (2.24216 iter/s, 5.35198s/12 iters), loss = 1.13863 I0406 08:03:00.353646 5226 solver.cpp:237] Train net output #0: loss = 1.13863 (* 1 = 1.13863 loss) I0406 08:03:00.353652 5226 sgd_solver.cpp:105] Iteration 6360, lr = 0.01 I0406 08:03:05.346380 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:03:05.539719 5226 solver.cpp:218] Iteration 6372 (2.31391 iter/s, 5.18602s/12 iters), loss = 1.26286 I0406 08:03:05.539757 5226 solver.cpp:237] Train net output #0: loss = 1.26286 (* 1 = 1.26286 loss) I0406 08:03:05.539762 5226 sgd_solver.cpp:105] Iteration 6372, lr = 0.01 I0406 08:03:10.735517 5226 solver.cpp:218] Iteration 6384 (2.3096 iter/s, 5.19571s/12 iters), loss = 1.30597 I0406 08:03:10.735639 5226 solver.cpp:237] Train net output #0: loss = 1.30597 (* 1 = 1.30597 loss) I0406 08:03:10.735646 5226 sgd_solver.cpp:105] Iteration 6384, lr = 0.01 I0406 08:03:15.914579 5226 solver.cpp:218] Iteration 6396 (2.3171 iter/s, 5.17889s/12 iters), loss = 0.972872 I0406 08:03:15.914623 5226 solver.cpp:237] Train net output #0: loss = 0.972872 (* 1 = 0.972872 loss) I0406 08:03:15.914631 5226 sgd_solver.cpp:105] Iteration 6396, lr = 0.01 I0406 08:03:21.281440 5226 solver.cpp:218] Iteration 6408 (2.23598 iter/s, 5.36677s/12 iters), loss = 1.09385 I0406 08:03:21.281477 5226 solver.cpp:237] Train net output #0: loss = 1.09385 (* 1 = 1.09385 loss) I0406 08:03:21.281483 5226 sgd_solver.cpp:105] Iteration 6408, lr = 0.01 I0406 08:03:26.441359 5226 solver.cpp:218] Iteration 6420 (2.32566 iter/s, 5.15983s/12 iters), loss = 1.15708 I0406 08:03:26.441396 5226 solver.cpp:237] Train net output #0: loss = 1.15708 (* 1 = 1.15708 loss) I0406 08:03:26.441402 5226 sgd_solver.cpp:105] Iteration 6420, lr = 0.01 I0406 08:03:28.528931 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6426.caffemodel I0406 08:03:31.528875 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6426.solverstate I0406 08:03:33.888156 5226 solver.cpp:330] Iteration 6426, Testing net (#0) I0406 08:03:33.888176 5226 net.cpp:676] Ignoring source layer train-data I0406 08:03:35.741004 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:03:38.222579 5226 solver.cpp:397] Test net output #0: accuracy = 0.318627 I0406 08:03:38.222615 5226 solver.cpp:397] Test net output #1: loss = 3.15036 (* 1 = 3.15036 loss) I0406 08:03:40.174453 5226 solver.cpp:218] Iteration 6432 (0.87381 iter/s, 13.733s/12 iters), loss = 1.36721 I0406 08:03:40.174492 5226 solver.cpp:237] Train net output #0: loss = 1.36721 (* 1 = 1.36721 loss) I0406 08:03:40.174499 5226 sgd_solver.cpp:105] Iteration 6432, lr = 0.01 I0406 08:03:45.636816 5226 solver.cpp:218] Iteration 6444 (2.19689 iter/s, 5.46228s/12 iters), loss = 1.13854 I0406 08:03:45.636914 5226 solver.cpp:237] Train net output #0: loss = 1.13854 (* 1 = 1.13854 loss) I0406 08:03:45.636920 5226 sgd_solver.cpp:105] Iteration 6444, lr = 0.01 I0406 08:03:50.932246 5226 solver.cpp:218] Iteration 6456 (2.26616 iter/s, 5.29529s/12 iters), loss = 0.948543 I0406 08:03:50.932288 5226 solver.cpp:237] Train net output #0: loss = 0.948543 (* 1 = 0.948543 loss) I0406 08:03:50.932297 5226 sgd_solver.cpp:105] Iteration 6456, lr = 0.01 I0406 08:03:55.976052 5226 solver.cpp:218] Iteration 6468 (2.3792 iter/s, 5.04372s/12 iters), loss = 1.34141 I0406 08:03:55.976083 5226 solver.cpp:237] Train net output #0: loss = 1.34141 (* 1 = 1.34141 loss) I0406 08:03:55.976087 5226 sgd_solver.cpp:105] Iteration 6468, lr = 0.01 I0406 08:03:57.976577 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:04:01.294920 5226 solver.cpp:218] Iteration 6480 (2.25615 iter/s, 5.31879s/12 iters), loss = 1.08114 I0406 08:04:01.294961 5226 solver.cpp:237] Train net output #0: loss = 1.08114 (* 1 = 1.08114 loss) I0406 08:04:01.294966 5226 sgd_solver.cpp:105] Iteration 6480, lr = 0.01 I0406 08:04:06.480978 5226 solver.cpp:218] Iteration 6492 (2.31394 iter/s, 5.18597s/12 iters), loss = 1.17164 I0406 08:04:06.481015 5226 solver.cpp:237] Train net output #0: loss = 1.17164 (* 1 = 1.17164 loss) I0406 08:04:06.481020 5226 sgd_solver.cpp:105] Iteration 6492, lr = 0.01 I0406 08:04:11.807209 5226 solver.cpp:218] Iteration 6504 (2.25303 iter/s, 5.32615s/12 iters), loss = 1.14135 I0406 08:04:11.807245 5226 solver.cpp:237] Train net output #0: loss = 1.14135 (* 1 = 1.14135 loss) I0406 08:04:11.807250 5226 sgd_solver.cpp:105] Iteration 6504, lr = 0.01 I0406 08:04:17.034530 5226 solver.cpp:218] Iteration 6516 (2.29567 iter/s, 5.22723s/12 iters), loss = 0.916203 I0406 08:04:17.034685 5226 solver.cpp:237] Train net output #0: loss = 0.916203 (* 1 = 0.916203 loss) I0406 08:04:17.034696 5226 sgd_solver.cpp:105] Iteration 6516, lr = 0.01 I0406 08:04:21.614195 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6528.caffemodel I0406 08:04:24.644872 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6528.solverstate I0406 08:04:26.945621 5226 solver.cpp:330] Iteration 6528, Testing net (#0) I0406 08:04:26.945639 5226 net.cpp:676] Ignoring source layer train-data I0406 08:04:28.751978 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:04:31.288625 5226 solver.cpp:397] Test net output #0: accuracy = 0.327819 I0406 08:04:31.288661 5226 solver.cpp:397] Test net output #1: loss = 3.11253 (* 1 = 3.11253 loss) I0406 08:04:31.429592 5226 solver.cpp:218] Iteration 6528 (0.833634 iter/s, 14.3948s/12 iters), loss = 0.993294 I0406 08:04:31.429632 5226 solver.cpp:237] Train net output #0: loss = 0.993294 (* 1 = 0.993294 loss) I0406 08:04:31.429637 5226 sgd_solver.cpp:105] Iteration 6528, lr = 0.01 I0406 08:04:35.823329 5226 solver.cpp:218] Iteration 6540 (2.73121 iter/s, 4.39366s/12 iters), loss = 1.01667 I0406 08:04:35.823362 5226 solver.cpp:237] Train net output #0: loss = 1.01667 (* 1 = 1.01667 loss) I0406 08:04:35.823367 5226 sgd_solver.cpp:105] Iteration 6540, lr = 0.01 I0406 08:04:40.962194 5226 solver.cpp:218] Iteration 6552 (2.33518 iter/s, 5.13878s/12 iters), loss = 1.38059 I0406 08:04:40.962230 5226 solver.cpp:237] Train net output #0: loss = 1.38059 (* 1 = 1.38059 loss) I0406 08:04:40.962239 5226 sgd_solver.cpp:105] Iteration 6552, lr = 0.01 I0406 08:04:46.304085 5226 solver.cpp:218] Iteration 6564 (2.24643 iter/s, 5.3418s/12 iters), loss = 1.35795 I0406 08:04:46.304126 5226 solver.cpp:237] Train net output #0: loss = 1.35795 (* 1 = 1.35795 loss) I0406 08:04:46.304131 5226 sgd_solver.cpp:105] Iteration 6564, lr = 0.01 I0406 08:04:50.683212 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:04:51.539520 5226 solver.cpp:218] Iteration 6576 (2.29211 iter/s, 5.23534s/12 iters), loss = 1.04343 I0406 08:04:51.539566 5226 solver.cpp:237] Train net output #0: loss = 1.04343 (* 1 = 1.04343 loss) I0406 08:04:51.539574 5226 sgd_solver.cpp:105] Iteration 6576, lr = 0.01 I0406 08:04:56.781828 5226 solver.cpp:218] Iteration 6588 (2.28911 iter/s, 5.24222s/12 iters), loss = 1.20892 I0406 08:04:56.781864 5226 solver.cpp:237] Train net output #0: loss = 1.20892 (* 1 = 1.20892 loss) I0406 08:04:56.781870 5226 sgd_solver.cpp:105] Iteration 6588, lr = 0.01 I0406 08:05:02.096091 5226 solver.cpp:218] Iteration 6600 (2.25811 iter/s, 5.31418s/12 iters), loss = 0.967889 I0406 08:05:02.096130 5226 solver.cpp:237] Train net output #0: loss = 0.967889 (* 1 = 0.967889 loss) I0406 08:05:02.096135 5226 sgd_solver.cpp:105] Iteration 6600, lr = 0.01 I0406 08:05:07.149248 5226 solver.cpp:218] Iteration 6612 (2.37479 iter/s, 5.05308s/12 iters), loss = 1.01034 I0406 08:05:07.149286 5226 solver.cpp:237] Train net output #0: loss = 1.01034 (* 1 = 1.01034 loss) I0406 08:05:07.149291 5226 sgd_solver.cpp:105] Iteration 6612, lr = 0.01 I0406 08:05:12.447765 5226 solver.cpp:218] Iteration 6624 (2.26482 iter/s, 5.29844s/12 iters), loss = 0.972174 I0406 08:05:12.447798 5226 solver.cpp:237] Train net output #0: loss = 0.972174 (* 1 = 0.972174 loss) I0406 08:05:12.447803 5226 sgd_solver.cpp:105] Iteration 6624, lr = 0.01 I0406 08:05:14.589771 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6630.caffemodel I0406 08:05:17.576299 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6630.solverstate I0406 08:05:19.872946 5226 solver.cpp:330] Iteration 6630, Testing net (#0) I0406 08:05:19.872964 5226 net.cpp:676] Ignoring source layer train-data I0406 08:05:21.664875 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:05:24.205204 5226 solver.cpp:397] Test net output #0: accuracy = 0.332721 I0406 08:05:24.205240 5226 solver.cpp:397] Test net output #1: loss = 3.27978 (* 1 = 3.27978 loss) I0406 08:05:26.305229 5226 solver.cpp:218] Iteration 6636 (0.865967 iter/s, 13.8573s/12 iters), loss = 1.41617 I0406 08:05:26.305269 5226 solver.cpp:237] Train net output #0: loss = 1.41617 (* 1 = 1.41617 loss) I0406 08:05:26.305274 5226 sgd_solver.cpp:105] Iteration 6636, lr = 0.01 I0406 08:05:31.669795 5226 solver.cpp:218] Iteration 6648 (2.23693 iter/s, 5.36448s/12 iters), loss = 1.31826 I0406 08:05:31.669833 5226 solver.cpp:237] Train net output #0: loss = 1.31826 (* 1 = 1.31826 loss) I0406 08:05:31.669838 5226 sgd_solver.cpp:105] Iteration 6648, lr = 0.01 I0406 08:05:37.038064 5226 solver.cpp:218] Iteration 6660 (2.23539 iter/s, 5.36818s/12 iters), loss = 1.51053 I0406 08:05:37.038102 5226 solver.cpp:237] Train net output #0: loss = 1.51053 (* 1 = 1.51053 loss) I0406 08:05:37.038107 5226 sgd_solver.cpp:105] Iteration 6660, lr = 0.01 I0406 08:05:42.339327 5226 solver.cpp:218] Iteration 6672 (2.26365 iter/s, 5.30118s/12 iters), loss = 1.0244 I0406 08:05:42.339363 5226 solver.cpp:237] Train net output #0: loss = 1.0244 (* 1 = 1.0244 loss) I0406 08:05:42.339370 5226 sgd_solver.cpp:105] Iteration 6672, lr = 0.01 I0406 08:05:43.782169 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:05:47.655052 5226 solver.cpp:218] Iteration 6684 (2.25749 iter/s, 5.31564s/12 iters), loss = 1.07668 I0406 08:05:47.655086 5226 solver.cpp:237] Train net output #0: loss = 1.07668 (* 1 = 1.07668 loss) I0406 08:05:47.655092 5226 sgd_solver.cpp:105] Iteration 6684, lr = 0.01 I0406 08:05:53.036336 5226 solver.cpp:218] Iteration 6696 (2.22999 iter/s, 5.3812s/12 iters), loss = 1.07263 I0406 08:05:53.036449 5226 solver.cpp:237] Train net output #0: loss = 1.07263 (* 1 = 1.07263 loss) I0406 08:05:53.036458 5226 sgd_solver.cpp:105] Iteration 6696, lr = 0.01 I0406 08:05:58.173512 5226 solver.cpp:218] Iteration 6708 (2.33598 iter/s, 5.13702s/12 iters), loss = 1.03074 I0406 08:05:58.173549 5226 solver.cpp:237] Train net output #0: loss = 1.03074 (* 1 = 1.03074 loss) I0406 08:05:58.173555 5226 sgd_solver.cpp:105] Iteration 6708, lr = 0.01 I0406 08:06:03.381147 5226 solver.cpp:218] Iteration 6720 (2.30434 iter/s, 5.20755s/12 iters), loss = 1.52513 I0406 08:06:03.381181 5226 solver.cpp:237] Train net output #0: loss = 1.52513 (* 1 = 1.52513 loss) I0406 08:06:03.381186 5226 sgd_solver.cpp:105] Iteration 6720, lr = 0.01 I0406 08:06:07.999007 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6732.caffemodel I0406 08:06:10.954542 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6732.solverstate I0406 08:06:13.259570 5226 solver.cpp:330] Iteration 6732, Testing net (#0) I0406 08:06:13.259593 5226 net.cpp:676] Ignoring source layer train-data I0406 08:06:14.961577 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:06:17.580265 5226 solver.cpp:397] Test net output #0: accuracy = 0.322304 I0406 08:06:17.580300 5226 solver.cpp:397] Test net output #1: loss = 3.12752 (* 1 = 3.12752 loss) I0406 08:06:17.717653 5226 solver.cpp:218] Iteration 6732 (0.837032 iter/s, 14.3364s/12 iters), loss = 1.19235 I0406 08:06:17.717700 5226 solver.cpp:237] Train net output #0: loss = 1.19235 (* 1 = 1.19235 loss) I0406 08:06:17.717707 5226 sgd_solver.cpp:105] Iteration 6732, lr = 0.01 I0406 08:06:21.960300 5226 solver.cpp:218] Iteration 6744 (2.82848 iter/s, 4.24256s/12 iters), loss = 1.33845 I0406 08:06:21.960342 5226 solver.cpp:237] Train net output #0: loss = 1.33845 (* 1 = 1.33845 loss) I0406 08:06:21.960348 5226 sgd_solver.cpp:105] Iteration 6744, lr = 0.01 I0406 08:06:26.987522 5226 solver.cpp:218] Iteration 6756 (2.38705 iter/s, 5.02713s/12 iters), loss = 1.57507 I0406 08:06:26.987633 5226 solver.cpp:237] Train net output #0: loss = 1.57507 (* 1 = 1.57507 loss) I0406 08:06:26.987640 5226 sgd_solver.cpp:105] Iteration 6756, lr = 0.01 I0406 08:06:32.089697 5226 solver.cpp:218] Iteration 6768 (2.35201 iter/s, 5.10202s/12 iters), loss = 0.941016 I0406 08:06:32.089735 5226 solver.cpp:237] Train net output #0: loss = 0.941016 (* 1 = 0.941016 loss) I0406 08:06:32.089740 5226 sgd_solver.cpp:105] Iteration 6768, lr = 0.01 I0406 08:06:35.689278 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:06:37.366729 5226 solver.cpp:218] Iteration 6780 (2.27404 iter/s, 5.27695s/12 iters), loss = 0.895397 I0406 08:06:37.366777 5226 solver.cpp:237] Train net output #0: loss = 0.895397 (* 1 = 0.895397 loss) I0406 08:06:37.366784 5226 sgd_solver.cpp:105] Iteration 6780, lr = 0.01 I0406 08:06:42.427404 5226 solver.cpp:218] Iteration 6792 (2.37127 iter/s, 5.06058s/12 iters), loss = 1.33512 I0406 08:06:42.427443 5226 solver.cpp:237] Train net output #0: loss = 1.33512 (* 1 = 1.33512 loss) I0406 08:06:42.427449 5226 sgd_solver.cpp:105] Iteration 6792, lr = 0.01 I0406 08:06:47.726462 5226 solver.cpp:218] Iteration 6804 (2.26459 iter/s, 5.29898s/12 iters), loss = 1.34361 I0406 08:06:47.726511 5226 solver.cpp:237] Train net output #0: loss = 1.34361 (* 1 = 1.34361 loss) I0406 08:06:47.726517 5226 sgd_solver.cpp:105] Iteration 6804, lr = 0.01 I0406 08:06:53.007879 5226 solver.cpp:218] Iteration 6816 (2.27216 iter/s, 5.28132s/12 iters), loss = 1.00271 I0406 08:06:53.007916 5226 solver.cpp:237] Train net output #0: loss = 1.00271 (* 1 = 1.00271 loss) I0406 08:06:53.007922 5226 sgd_solver.cpp:105] Iteration 6816, lr = 0.01 I0406 08:06:58.346593 5226 solver.cpp:218] Iteration 6828 (2.24777 iter/s, 5.33863s/12 iters), loss = 1.15559 I0406 08:06:58.346735 5226 solver.cpp:237] Train net output #0: loss = 1.15559 (* 1 = 1.15559 loss) I0406 08:06:58.346741 5226 sgd_solver.cpp:105] Iteration 6828, lr = 0.01 I0406 08:07:00.377619 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6834.caffemodel I0406 08:07:03.360368 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6834.solverstate I0406 08:07:05.658643 5226 solver.cpp:330] Iteration 6834, Testing net (#0) I0406 08:07:05.658660 5226 net.cpp:676] Ignoring source layer train-data I0406 08:07:07.297152 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:07:09.947367 5226 solver.cpp:397] Test net output #0: accuracy = 0.314338 I0406 08:07:09.947407 5226 solver.cpp:397] Test net output #1: loss = 3.26802 (* 1 = 3.26802 loss) I0406 08:07:11.825042 5226 solver.cpp:218] Iteration 6840 (0.890326 iter/s, 13.4782s/12 iters), loss = 1.1371 I0406 08:07:11.825080 5226 solver.cpp:237] Train net output #0: loss = 1.1371 (* 1 = 1.1371 loss) I0406 08:07:11.825085 5226 sgd_solver.cpp:105] Iteration 6840, lr = 0.01 I0406 08:07:17.130761 5226 solver.cpp:218] Iteration 6852 (2.26175 iter/s, 5.30563s/12 iters), loss = 0.999021 I0406 08:07:17.130802 5226 solver.cpp:237] Train net output #0: loss = 0.999021 (* 1 = 0.999021 loss) I0406 08:07:17.130810 5226 sgd_solver.cpp:105] Iteration 6852, lr = 0.01 I0406 08:07:22.479013 5226 solver.cpp:218] Iteration 6864 (2.24376 iter/s, 5.34816s/12 iters), loss = 0.994479 I0406 08:07:22.479053 5226 solver.cpp:237] Train net output #0: loss = 0.994479 (* 1 = 0.994479 loss) I0406 08:07:22.479058 5226 sgd_solver.cpp:105] Iteration 6864, lr = 0.01 I0406 08:07:27.873414 5226 solver.cpp:218] Iteration 6876 (2.22456 iter/s, 5.39432s/12 iters), loss = 1.32687 I0406 08:07:27.873446 5226 solver.cpp:237] Train net output #0: loss = 1.32687 (* 1 = 1.32687 loss) I0406 08:07:27.873451 5226 sgd_solver.cpp:105] Iteration 6876, lr = 0.01 I0406 08:07:28.508788 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:07:33.189221 5226 solver.cpp:218] Iteration 6888 (2.25745 iter/s, 5.31573s/12 iters), loss = 1.28745 I0406 08:07:33.189258 5226 solver.cpp:237] Train net output #0: loss = 1.28745 (* 1 = 1.28745 loss) I0406 08:07:33.189265 5226 sgd_solver.cpp:105] Iteration 6888, lr = 0.01 I0406 08:07:38.147213 5226 solver.cpp:218] Iteration 6900 (2.42038 iter/s, 4.95791s/12 iters), loss = 1.21176 I0406 08:07:38.147249 5226 solver.cpp:237] Train net output #0: loss = 1.21176 (* 1 = 1.21176 loss) I0406 08:07:38.147254 5226 sgd_solver.cpp:105] Iteration 6900, lr = 0.01 I0406 08:07:43.341907 5226 solver.cpp:218] Iteration 6912 (2.31009 iter/s, 5.19461s/12 iters), loss = 1.22255 I0406 08:07:43.341950 5226 solver.cpp:237] Train net output #0: loss = 1.22255 (* 1 = 1.22255 loss) I0406 08:07:43.341959 5226 sgd_solver.cpp:105] Iteration 6912, lr = 0.01 I0406 08:07:48.705101 5226 solver.cpp:218] Iteration 6924 (2.23751 iter/s, 5.3631s/12 iters), loss = 1.01774 I0406 08:07:48.705137 5226 solver.cpp:237] Train net output #0: loss = 1.01774 (* 1 = 1.01774 loss) I0406 08:07:48.705143 5226 sgd_solver.cpp:105] Iteration 6924, lr = 0.01 I0406 08:07:53.320048 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6936.caffemodel I0406 08:07:56.327132 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6936.solverstate I0406 08:07:58.689376 5226 solver.cpp:330] Iteration 6936, Testing net (#0) I0406 08:07:58.689456 5226 net.cpp:676] Ignoring source layer train-data I0406 08:07:59.246944 5226 blocking_queue.cpp:49] Waiting for data I0406 08:08:00.275157 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:08:02.948769 5226 solver.cpp:397] Test net output #0: accuracy = 0.341299 I0406 08:08:02.948796 5226 solver.cpp:397] Test net output #1: loss = 3.09619 (* 1 = 3.09619 loss) I0406 08:08:03.087630 5226 solver.cpp:218] Iteration 6936 (0.834353 iter/s, 14.3824s/12 iters), loss = 1.22934 I0406 08:08:03.087675 5226 solver.cpp:237] Train net output #0: loss = 1.22934 (* 1 = 1.22934 loss) I0406 08:08:03.087682 5226 sgd_solver.cpp:105] Iteration 6936, lr = 0.01 I0406 08:08:07.308470 5226 solver.cpp:218] Iteration 6948 (2.84309 iter/s, 4.22076s/12 iters), loss = 1.4935 I0406 08:08:07.308504 5226 solver.cpp:237] Train net output #0: loss = 1.4935 (* 1 = 1.4935 loss) I0406 08:08:07.308509 5226 sgd_solver.cpp:105] Iteration 6948, lr = 0.01 I0406 08:08:12.525612 5226 solver.cpp:218] Iteration 6960 (2.30014 iter/s, 5.21706s/12 iters), loss = 1.23067 I0406 08:08:12.525645 5226 solver.cpp:237] Train net output #0: loss = 1.23067 (* 1 = 1.23067 loss) I0406 08:08:12.525650 5226 sgd_solver.cpp:105] Iteration 6960, lr = 0.01 I0406 08:08:17.686228 5226 solver.cpp:218] Iteration 6972 (2.32534 iter/s, 5.16054s/12 iters), loss = 1.23891 I0406 08:08:17.686265 5226 solver.cpp:237] Train net output #0: loss = 1.23891 (* 1 = 1.23891 loss) I0406 08:08:17.686270 5226 sgd_solver.cpp:105] Iteration 6972, lr = 0.01 I0406 08:08:20.577445 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:08:23.017037 5226 solver.cpp:218] Iteration 6984 (2.2511 iter/s, 5.33072s/12 iters), loss = 0.945253 I0406 08:08:23.017076 5226 solver.cpp:237] Train net output #0: loss = 0.945253 (* 1 = 0.945253 loss) I0406 08:08:23.017082 5226 sgd_solver.cpp:105] Iteration 6984, lr = 0.01 I0406 08:08:28.356206 5226 solver.cpp:218] Iteration 6996 (2.24758 iter/s, 5.33908s/12 iters), loss = 1.25337 I0406 08:08:28.356243 5226 solver.cpp:237] Train net output #0: loss = 1.25337 (* 1 = 1.25337 loss) I0406 08:08:28.356249 5226 sgd_solver.cpp:105] Iteration 6996, lr = 0.01 I0406 08:08:33.688304 5226 solver.cpp:218] Iteration 7008 (2.25056 iter/s, 5.33201s/12 iters), loss = 1.02682 I0406 08:08:33.688447 5226 solver.cpp:237] Train net output #0: loss = 1.02682 (* 1 = 1.02682 loss) I0406 08:08:33.688457 5226 sgd_solver.cpp:105] Iteration 7008, lr = 0.01 I0406 08:08:38.870728 5226 solver.cpp:218] Iteration 7020 (2.3156 iter/s, 5.18224s/12 iters), loss = 1.3142 I0406 08:08:38.870764 5226 solver.cpp:237] Train net output #0: loss = 1.3142 (* 1 = 1.3142 loss) I0406 08:08:38.870771 5226 sgd_solver.cpp:105] Iteration 7020, lr = 0.01 I0406 08:08:44.129168 5226 solver.cpp:218] Iteration 7032 (2.28208 iter/s, 5.25836s/12 iters), loss = 1.28904 I0406 08:08:44.129204 5226 solver.cpp:237] Train net output #0: loss = 1.28904 (* 1 = 1.28904 loss) I0406 08:08:44.129215 5226 sgd_solver.cpp:105] Iteration 7032, lr = 0.01 I0406 08:08:46.368189 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7038.caffemodel I0406 08:08:49.431262 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7038.solverstate I0406 08:08:51.748904 5226 solver.cpp:330] Iteration 7038, Testing net (#0) I0406 08:08:51.748929 5226 net.cpp:676] Ignoring source layer train-data I0406 08:08:53.341706 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:08:56.077013 5226 solver.cpp:397] Test net output #0: accuracy = 0.309436 I0406 08:08:56.077042 5226 solver.cpp:397] Test net output #1: loss = 3.14588 (* 1 = 3.14588 loss) I0406 08:08:57.839123 5226 solver.cpp:218] Iteration 7044 (0.875285 iter/s, 13.7098s/12 iters), loss = 1.21017 I0406 08:08:57.839164 5226 solver.cpp:237] Train net output #0: loss = 1.21017 (* 1 = 1.21017 loss) I0406 08:08:57.839169 5226 sgd_solver.cpp:105] Iteration 7044, lr = 0.01 I0406 08:09:02.927242 5226 solver.cpp:218] Iteration 7056 (2.35848 iter/s, 5.08803s/12 iters), loss = 1.42301 I0406 08:09:02.927281 5226 solver.cpp:237] Train net output #0: loss = 1.42301 (* 1 = 1.42301 loss) I0406 08:09:02.927286 5226 sgd_solver.cpp:105] Iteration 7056, lr = 0.01 I0406 08:09:08.095134 5226 solver.cpp:218] Iteration 7068 (2.32207 iter/s, 5.1678s/12 iters), loss = 0.962161 I0406 08:09:08.095234 5226 solver.cpp:237] Train net output #0: loss = 0.962161 (* 1 = 0.962161 loss) I0406 08:09:08.095240 5226 sgd_solver.cpp:105] Iteration 7068, lr = 0.01 I0406 08:09:13.355532 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:09:13.520596 5226 solver.cpp:218] Iteration 7080 (2.21186 iter/s, 5.42531s/12 iters), loss = 1.23596 I0406 08:09:13.520645 5226 solver.cpp:237] Train net output #0: loss = 1.23596 (* 1 = 1.23596 loss) I0406 08:09:13.520654 5226 sgd_solver.cpp:105] Iteration 7080, lr = 0.01 I0406 08:09:18.901111 5226 solver.cpp:218] Iteration 7092 (2.23031 iter/s, 5.38041s/12 iters), loss = 1.1704 I0406 08:09:18.901152 5226 solver.cpp:237] Train net output #0: loss = 1.1704 (* 1 = 1.1704 loss) I0406 08:09:18.901157 5226 sgd_solver.cpp:105] Iteration 7092, lr = 0.01 I0406 08:09:23.974205 5226 solver.cpp:218] Iteration 7104 (2.36546 iter/s, 5.07301s/12 iters), loss = 1.0578 I0406 08:09:23.974242 5226 solver.cpp:237] Train net output #0: loss = 1.0578 (* 1 = 1.0578 loss) I0406 08:09:23.974247 5226 sgd_solver.cpp:105] Iteration 7104, lr = 0.01 I0406 08:09:29.278834 5226 solver.cpp:218] Iteration 7116 (2.26221 iter/s, 5.30455s/12 iters), loss = 1.60422 I0406 08:09:29.278874 5226 solver.cpp:237] Train net output #0: loss = 1.60422 (* 1 = 1.60422 loss) I0406 08:09:29.278880 5226 sgd_solver.cpp:105] Iteration 7116, lr = 0.01 I0406 08:09:34.620187 5226 solver.cpp:218] Iteration 7128 (2.24666 iter/s, 5.34126s/12 iters), loss = 1.3641 I0406 08:09:34.620226 5226 solver.cpp:237] Train net output #0: loss = 1.3641 (* 1 = 1.3641 loss) I0406 08:09:34.620231 5226 sgd_solver.cpp:105] Iteration 7128, lr = 0.01 I0406 08:09:39.466658 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7140.caffemodel I0406 08:09:42.522153 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7140.solverstate I0406 08:09:44.835888 5226 solver.cpp:330] Iteration 7140, Testing net (#0) I0406 08:09:44.835912 5226 net.cpp:676] Ignoring source layer train-data I0406 08:09:46.393167 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:09:49.217873 5226 solver.cpp:397] Test net output #0: accuracy = 0.31924 I0406 08:09:49.217914 5226 solver.cpp:397] Test net output #1: loss = 3.29152 (* 1 = 3.29152 loss) I0406 08:09:49.353264 5226 solver.cpp:218] Iteration 7140 (0.814502 iter/s, 14.7329s/12 iters), loss = 0.985333 I0406 08:09:49.353322 5226 solver.cpp:237] Train net output #0: loss = 0.985333 (* 1 = 0.985333 loss) I0406 08:09:49.353329 5226 sgd_solver.cpp:105] Iteration 7140, lr = 0.01 I0406 08:09:53.676674 5226 solver.cpp:218] Iteration 7152 (2.77565 iter/s, 4.32331s/12 iters), loss = 1.24605 I0406 08:09:53.676712 5226 solver.cpp:237] Train net output #0: loss = 1.24605 (* 1 = 1.24605 loss) I0406 08:09:53.676717 5226 sgd_solver.cpp:105] Iteration 7152, lr = 0.01 I0406 08:09:59.005640 5226 solver.cpp:218] Iteration 7164 (2.25188 iter/s, 5.32888s/12 iters), loss = 1.33262 I0406 08:09:59.005684 5226 solver.cpp:237] Train net output #0: loss = 1.33262 (* 1 = 1.33262 loss) I0406 08:09:59.005692 5226 sgd_solver.cpp:105] Iteration 7164, lr = 0.01 I0406 08:10:03.890460 5226 solver.cpp:218] Iteration 7176 (2.45664 iter/s, 4.88473s/12 iters), loss = 1.1438 I0406 08:10:03.890501 5226 solver.cpp:237] Train net output #0: loss = 1.1438 (* 1 = 1.1438 loss) I0406 08:10:03.890506 5226 sgd_solver.cpp:105] Iteration 7176, lr = 0.01 I0406 08:10:06.190948 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:10:09.356307 5226 solver.cpp:218] Iteration 7188 (2.19549 iter/s, 5.46575s/12 iters), loss = 1.03663 I0406 08:10:09.356357 5226 solver.cpp:237] Train net output #0: loss = 1.03663 (* 1 = 1.03663 loss) I0406 08:10:09.356366 5226 sgd_solver.cpp:105] Iteration 7188, lr = 0.01 I0406 08:10:14.413187 5226 solver.cpp:218] Iteration 7200 (2.37305 iter/s, 5.05678s/12 iters), loss = 1.25818 I0406 08:10:14.414992 5226 solver.cpp:237] Train net output #0: loss = 1.25818 (* 1 = 1.25818 loss) I0406 08:10:14.415004 5226 sgd_solver.cpp:105] Iteration 7200, lr = 0.01 I0406 08:10:19.600986 5226 solver.cpp:218] Iteration 7212 (2.31394 iter/s, 5.18595s/12 iters), loss = 1.43975 I0406 08:10:19.601034 5226 solver.cpp:237] Train net output #0: loss = 1.43975 (* 1 = 1.43975 loss) I0406 08:10:19.601042 5226 sgd_solver.cpp:105] Iteration 7212, lr = 0.01 I0406 08:10:24.784196 5226 solver.cpp:218] Iteration 7224 (2.31521 iter/s, 5.18312s/12 iters), loss = 1.30981 I0406 08:10:24.784233 5226 solver.cpp:237] Train net output #0: loss = 1.30981 (* 1 = 1.30981 loss) I0406 08:10:24.784238 5226 sgd_solver.cpp:105] Iteration 7224, lr = 0.01 I0406 08:10:30.095067 5226 solver.cpp:218] Iteration 7236 (2.25956 iter/s, 5.31078s/12 iters), loss = 1.27126 I0406 08:10:30.095122 5226 solver.cpp:237] Train net output #0: loss = 1.27126 (* 1 = 1.27126 loss) I0406 08:10:30.095130 5226 sgd_solver.cpp:105] Iteration 7236, lr = 0.01 I0406 08:10:32.099313 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7242.caffemodel I0406 08:10:35.172044 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7242.solverstate I0406 08:10:37.478842 5226 solver.cpp:330] Iteration 7242, Testing net (#0) I0406 08:10:37.478864 5226 net.cpp:676] Ignoring source layer train-data I0406 08:10:39.029176 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:10:42.007102 5226 solver.cpp:397] Test net output #0: accuracy = 0.310662 I0406 08:10:42.007138 5226 solver.cpp:397] Test net output #1: loss = 3.26105 (* 1 = 3.26105 loss) I0406 08:10:43.886049 5226 solver.cpp:218] Iteration 7248 (0.870143 iter/s, 13.7908s/12 iters), loss = 1.45115 I0406 08:10:43.886091 5226 solver.cpp:237] Train net output #0: loss = 1.45115 (* 1 = 1.45115 loss) I0406 08:10:43.886097 5226 sgd_solver.cpp:105] Iteration 7248, lr = 0.01 I0406 08:10:49.166452 5226 solver.cpp:218] Iteration 7260 (2.27259 iter/s, 5.28031s/12 iters), loss = 1.67387 I0406 08:10:49.166627 5226 solver.cpp:237] Train net output #0: loss = 1.67387 (* 1 = 1.67387 loss) I0406 08:10:49.166636 5226 sgd_solver.cpp:105] Iteration 7260, lr = 0.01 I0406 08:10:54.493541 5226 solver.cpp:218] Iteration 7272 (2.25273 iter/s, 5.32687s/12 iters), loss = 1.45325 I0406 08:10:54.493592 5226 solver.cpp:237] Train net output #0: loss = 1.45325 (* 1 = 1.45325 loss) I0406 08:10:54.493599 5226 sgd_solver.cpp:105] Iteration 7272, lr = 0.01 I0406 08:10:59.118005 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:10:59.931618 5226 solver.cpp:218] Iteration 7284 (2.2067 iter/s, 5.43798s/12 iters), loss = 1.31641 I0406 08:10:59.931658 5226 solver.cpp:237] Train net output #0: loss = 1.31641 (* 1 = 1.31641 loss) I0406 08:10:59.931663 5226 sgd_solver.cpp:105] Iteration 7284, lr = 0.01 I0406 08:11:05.341847 5226 solver.cpp:218] Iteration 7296 (2.21806 iter/s, 5.41014s/12 iters), loss = 1.38925 I0406 08:11:05.341897 5226 solver.cpp:237] Train net output #0: loss = 1.38925 (* 1 = 1.38925 loss) I0406 08:11:05.341905 5226 sgd_solver.cpp:105] Iteration 7296, lr = 0.01 I0406 08:11:10.749009 5226 solver.cpp:218] Iteration 7308 (2.21932 iter/s, 5.40707s/12 iters), loss = 1.29207 I0406 08:11:10.749047 5226 solver.cpp:237] Train net output #0: loss = 1.29207 (* 1 = 1.29207 loss) I0406 08:11:10.749051 5226 sgd_solver.cpp:105] Iteration 7308, lr = 0.01 I0406 08:11:15.941305 5226 solver.cpp:218] Iteration 7320 (2.31115 iter/s, 5.19221s/12 iters), loss = 1.12175 I0406 08:11:15.941344 5226 solver.cpp:237] Train net output #0: loss = 1.12175 (* 1 = 1.12175 loss) I0406 08:11:15.941349 5226 sgd_solver.cpp:105] Iteration 7320, lr = 0.01 I0406 08:11:21.434329 5226 solver.cpp:218] Iteration 7332 (2.18463 iter/s, 5.49293s/12 iters), loss = 1.25404 I0406 08:11:21.434439 5226 solver.cpp:237] Train net output #0: loss = 1.25404 (* 1 = 1.25404 loss) I0406 08:11:21.434449 5226 sgd_solver.cpp:105] Iteration 7332, lr = 0.01 I0406 08:11:26.261636 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7344.caffemodel I0406 08:11:29.305686 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7344.solverstate I0406 08:11:31.607465 5226 solver.cpp:330] Iteration 7344, Testing net (#0) I0406 08:11:31.607483 5226 net.cpp:676] Ignoring source layer train-data I0406 08:11:33.081739 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:11:36.082605 5226 solver.cpp:397] Test net output #0: accuracy = 0.29473 I0406 08:11:36.082641 5226 solver.cpp:397] Test net output #1: loss = 3.27043 (* 1 = 3.27043 loss) I0406 08:11:36.219113 5226 solver.cpp:218] Iteration 7344 (0.811657 iter/s, 14.7846s/12 iters), loss = 1.17824 I0406 08:11:36.219177 5226 solver.cpp:237] Train net output #0: loss = 1.17824 (* 1 = 1.17824 loss) I0406 08:11:36.219187 5226 sgd_solver.cpp:105] Iteration 7344, lr = 0.01 I0406 08:11:40.517383 5226 solver.cpp:218] Iteration 7356 (2.79189 iter/s, 4.29817s/12 iters), loss = 1.36358 I0406 08:11:40.517423 5226 solver.cpp:237] Train net output #0: loss = 1.36358 (* 1 = 1.36358 loss) I0406 08:11:40.517428 5226 sgd_solver.cpp:105] Iteration 7356, lr = 0.01 I0406 08:11:45.894979 5226 solver.cpp:218] Iteration 7368 (2.23152 iter/s, 5.37751s/12 iters), loss = 1.61712 I0406 08:11:45.895026 5226 solver.cpp:237] Train net output #0: loss = 1.61712 (* 1 = 1.61712 loss) I0406 08:11:45.895033 5226 sgd_solver.cpp:105] Iteration 7368, lr = 0.01 I0406 08:11:51.193125 5226 solver.cpp:218] Iteration 7380 (2.26498 iter/s, 5.29805s/12 iters), loss = 1.38603 I0406 08:11:51.193164 5226 solver.cpp:237] Train net output #0: loss = 1.38603 (* 1 = 1.38603 loss) I0406 08:11:51.193169 5226 sgd_solver.cpp:105] Iteration 7380, lr = 0.01 I0406 08:11:52.611935 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:11:56.585515 5226 solver.cpp:218] Iteration 7392 (2.2254 iter/s, 5.3923s/12 iters), loss = 1.13427 I0406 08:11:56.585556 5226 solver.cpp:237] Train net output #0: loss = 1.13427 (* 1 = 1.13427 loss) I0406 08:11:56.585561 5226 sgd_solver.cpp:105] Iteration 7392, lr = 0.01 I0406 08:12:01.761031 5226 solver.cpp:218] Iteration 7404 (2.31865 iter/s, 5.17543s/12 iters), loss = 1.19668 I0406 08:12:01.761083 5226 solver.cpp:237] Train net output #0: loss = 1.19668 (* 1 = 1.19668 loss) I0406 08:12:01.761092 5226 sgd_solver.cpp:105] Iteration 7404, lr = 0.01 I0406 08:12:07.039425 5226 solver.cpp:218] Iteration 7416 (2.27346 iter/s, 5.27829s/12 iters), loss = 1.50726 I0406 08:12:07.039474 5226 solver.cpp:237] Train net output #0: loss = 1.50726 (* 1 = 1.50726 loss) I0406 08:12:07.039482 5226 sgd_solver.cpp:105] Iteration 7416, lr = 0.01 I0406 08:12:12.017783 5226 solver.cpp:218] Iteration 7428 (2.41048 iter/s, 4.97827s/12 iters), loss = 1.38178 I0406 08:12:12.017823 5226 solver.cpp:237] Train net output #0: loss = 1.38178 (* 1 = 1.38178 loss) I0406 08:12:12.017828 5226 sgd_solver.cpp:105] Iteration 7428, lr = 0.01 I0406 08:12:17.424818 5226 solver.cpp:218] Iteration 7440 (2.21937 iter/s, 5.40694s/12 iters), loss = 1.7034 I0406 08:12:17.424871 5226 solver.cpp:237] Train net output #0: loss = 1.7034 (* 1 = 1.7034 loss) I0406 08:12:17.424880 5226 sgd_solver.cpp:105] Iteration 7440, lr = 0.01 I0406 08:12:19.448189 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7446.caffemodel I0406 08:12:22.457324 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7446.solverstate I0406 08:12:24.791182 5226 solver.cpp:330] Iteration 7446, Testing net (#0) I0406 08:12:24.791249 5226 net.cpp:676] Ignoring source layer train-data I0406 08:12:26.255015 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:12:29.123417 5226 solver.cpp:397] Test net output #0: accuracy = 0.306985 I0406 08:12:29.123456 5226 solver.cpp:397] Test net output #1: loss = 3.25548 (* 1 = 3.25548 loss) I0406 08:12:31.111665 5226 solver.cpp:218] Iteration 7452 (0.876764 iter/s, 13.6867s/12 iters), loss = 1.26331 I0406 08:12:31.111721 5226 solver.cpp:237] Train net output #0: loss = 1.26331 (* 1 = 1.26331 loss) I0406 08:12:31.111728 5226 sgd_solver.cpp:105] Iteration 7452, lr = 0.01 I0406 08:12:36.485019 5226 solver.cpp:218] Iteration 7464 (2.23328 iter/s, 5.37325s/12 iters), loss = 1.48205 I0406 08:12:36.485072 5226 solver.cpp:237] Train net output #0: loss = 1.48205 (* 1 = 1.48205 loss) I0406 08:12:36.485081 5226 sgd_solver.cpp:105] Iteration 7464, lr = 0.01 I0406 08:12:41.842254 5226 solver.cpp:218] Iteration 7476 (2.24 iter/s, 5.35714s/12 iters), loss = 0.994821 I0406 08:12:41.842296 5226 solver.cpp:237] Train net output #0: loss = 0.994821 (* 1 = 0.994821 loss) I0406 08:12:41.842301 5226 sgd_solver.cpp:105] Iteration 7476, lr = 0.01 I0406 08:12:45.407567 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:12:46.985603 5226 solver.cpp:218] Iteration 7488 (2.33315 iter/s, 5.14326s/12 iters), loss = 1.03441 I0406 08:12:46.985646 5226 solver.cpp:237] Train net output #0: loss = 1.03441 (* 1 = 1.03441 loss) I0406 08:12:46.985651 5226 sgd_solver.cpp:105] Iteration 7488, lr = 0.01 I0406 08:12:52.286597 5226 solver.cpp:218] Iteration 7500 (2.26377 iter/s, 5.3009s/12 iters), loss = 1.299 I0406 08:12:52.286648 5226 solver.cpp:237] Train net output #0: loss = 1.299 (* 1 = 1.299 loss) I0406 08:12:52.286656 5226 sgd_solver.cpp:105] Iteration 7500, lr = 0.01 I0406 08:12:57.802179 5226 solver.cpp:218] Iteration 7512 (2.17569 iter/s, 5.51548s/12 iters), loss = 1.20659 I0406 08:12:57.802301 5226 solver.cpp:237] Train net output #0: loss = 1.20659 (* 1 = 1.20659 loss) I0406 08:12:57.802309 5226 sgd_solver.cpp:105] Iteration 7512, lr = 0.01 I0406 08:13:03.131002 5226 solver.cpp:218] Iteration 7524 (2.25198 iter/s, 5.32865s/12 iters), loss = 1.3829 I0406 08:13:03.131054 5226 solver.cpp:237] Train net output #0: loss = 1.3829 (* 1 = 1.3829 loss) I0406 08:13:03.131062 5226 sgd_solver.cpp:105] Iteration 7524, lr = 0.01 I0406 08:13:08.331003 5226 solver.cpp:218] Iteration 7536 (2.30774 iter/s, 5.1999s/12 iters), loss = 1.15164 I0406 08:13:08.331045 5226 solver.cpp:237] Train net output #0: loss = 1.15164 (* 1 = 1.15164 loss) I0406 08:13:08.331051 5226 sgd_solver.cpp:105] Iteration 7536, lr = 0.01 I0406 08:13:12.836421 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7548.caffemodel I0406 08:13:15.798626 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7548.solverstate I0406 08:13:18.113656 5226 solver.cpp:330] Iteration 7548, Testing net (#0) I0406 08:13:18.113675 5226 net.cpp:676] Ignoring source layer train-data I0406 08:13:19.476248 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:13:22.370656 5226 solver.cpp:397] Test net output #0: accuracy = 0.307598 I0406 08:13:22.370687 5226 solver.cpp:397] Test net output #1: loss = 3.36759 (* 1 = 3.36759 loss) I0406 08:13:22.511624 5226 solver.cpp:218] Iteration 7548 (0.846234 iter/s, 14.1805s/12 iters), loss = 1.37236 I0406 08:13:22.511674 5226 solver.cpp:237] Train net output #0: loss = 1.37236 (* 1 = 1.37236 loss) I0406 08:13:22.511682 5226 sgd_solver.cpp:105] Iteration 7548, lr = 0.01 I0406 08:13:27.104182 5226 solver.cpp:218] Iteration 7560 (2.61298 iter/s, 4.59246s/12 iters), loss = 1.25314 I0406 08:13:27.104228 5226 solver.cpp:237] Train net output #0: loss = 1.25314 (* 1 = 1.25314 loss) I0406 08:13:27.104238 5226 sgd_solver.cpp:105] Iteration 7560, lr = 0.01 I0406 08:13:32.443171 5226 solver.cpp:218] Iteration 7572 (2.24766 iter/s, 5.33889s/12 iters), loss = 1.41166 I0406 08:13:32.443291 5226 solver.cpp:237] Train net output #0: loss = 1.41166 (* 1 = 1.41166 loss) I0406 08:13:32.443300 5226 sgd_solver.cpp:105] Iteration 7572, lr = 0.01 I0406 08:13:37.781738 5226 solver.cpp:218] Iteration 7584 (2.24786 iter/s, 5.3384s/12 iters), loss = 1.33473 I0406 08:13:37.781798 5226 solver.cpp:237] Train net output #0: loss = 1.33473 (* 1 = 1.33473 loss) I0406 08:13:37.781806 5226 sgd_solver.cpp:105] Iteration 7584, lr = 0.01 I0406 08:13:38.446951 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:13:43.156009 5226 solver.cpp:218] Iteration 7596 (2.23291 iter/s, 5.37416s/12 iters), loss = 1.11536 I0406 08:13:43.156059 5226 solver.cpp:237] Train net output #0: loss = 1.11536 (* 1 = 1.11536 loss) I0406 08:13:43.156067 5226 sgd_solver.cpp:105] Iteration 7596, lr = 0.01 I0406 08:13:48.149021 5226 solver.cpp:218] Iteration 7608 (2.40341 iter/s, 4.99292s/12 iters), loss = 0.837079 I0406 08:13:48.149066 5226 solver.cpp:237] Train net output #0: loss = 0.837079 (* 1 = 0.837079 loss) I0406 08:13:48.149073 5226 sgd_solver.cpp:105] Iteration 7608, lr = 0.01 I0406 08:13:53.595036 5226 solver.cpp:218] Iteration 7620 (2.20349 iter/s, 5.44592s/12 iters), loss = 1.64403 I0406 08:13:53.595089 5226 solver.cpp:237] Train net output #0: loss = 1.64403 (* 1 = 1.64403 loss) I0406 08:13:53.595098 5226 sgd_solver.cpp:105] Iteration 7620, lr = 0.01 I0406 08:13:56.020975 5226 blocking_queue.cpp:49] Waiting for data I0406 08:13:58.505017 5226 solver.cpp:218] Iteration 7632 (2.44405 iter/s, 4.90988s/12 iters), loss = 1.34494 I0406 08:13:58.505072 5226 solver.cpp:237] Train net output #0: loss = 1.34494 (* 1 = 1.34494 loss) I0406 08:13:58.505080 5226 sgd_solver.cpp:105] Iteration 7632, lr = 0.01 I0406 08:14:03.942833 5226 solver.cpp:218] Iteration 7644 (2.20681 iter/s, 5.43771s/12 iters), loss = 1.4641 I0406 08:14:03.942976 5226 solver.cpp:237] Train net output #0: loss = 1.4641 (* 1 = 1.4641 loss) I0406 08:14:03.942986 5226 sgd_solver.cpp:105] Iteration 7644, lr = 0.01 I0406 08:14:06.053140 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7650.caffemodel I0406 08:14:09.180851 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7650.solverstate I0406 08:14:11.543313 5226 solver.cpp:330] Iteration 7650, Testing net (#0) I0406 08:14:11.543334 5226 net.cpp:676] Ignoring source layer train-data I0406 08:14:12.895263 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:14:15.912137 5226 solver.cpp:397] Test net output #0: accuracy = 0.310049 I0406 08:14:15.912173 5226 solver.cpp:397] Test net output #1: loss = 3.32535 (* 1 = 3.32535 loss) I0406 08:14:17.779772 5226 solver.cpp:218] Iteration 7656 (0.867259 iter/s, 13.8367s/12 iters), loss = 1.0255 I0406 08:14:17.779811 5226 solver.cpp:237] Train net output #0: loss = 1.0255 (* 1 = 1.0255 loss) I0406 08:14:17.779817 5226 sgd_solver.cpp:105] Iteration 7656, lr = 0.01 I0406 08:14:23.064656 5226 solver.cpp:218] Iteration 7668 (2.27066 iter/s, 5.2848s/12 iters), loss = 1.17211 I0406 08:14:23.064697 5226 solver.cpp:237] Train net output #0: loss = 1.17211 (* 1 = 1.17211 loss) I0406 08:14:23.064702 5226 sgd_solver.cpp:105] Iteration 7668, lr = 0.01 I0406 08:14:28.443432 5226 solver.cpp:218] Iteration 7680 (2.23103 iter/s, 5.37868s/12 iters), loss = 0.77815 I0406 08:14:28.443476 5226 solver.cpp:237] Train net output #0: loss = 0.77815 (* 1 = 0.77815 loss) I0406 08:14:28.443485 5226 sgd_solver.cpp:105] Iteration 7680, lr = 0.01 I0406 08:14:31.310232 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:14:33.717849 5226 solver.cpp:218] Iteration 7692 (2.27517 iter/s, 5.27432s/12 iters), loss = 1.04241 I0406 08:14:33.717888 5226 solver.cpp:237] Train net output #0: loss = 1.04241 (* 1 = 1.04241 loss) I0406 08:14:33.717895 5226 sgd_solver.cpp:105] Iteration 7692, lr = 0.01 I0406 08:14:39.017709 5226 solver.cpp:218] Iteration 7704 (2.26425 iter/s, 5.29977s/12 iters), loss = 1.16704 I0406 08:14:39.017829 5226 solver.cpp:237] Train net output #0: loss = 1.16704 (* 1 = 1.16704 loss) I0406 08:14:39.017838 5226 sgd_solver.cpp:105] Iteration 7704, lr = 0.01 I0406 08:14:44.279841 5226 solver.cpp:218] Iteration 7716 (2.28052 iter/s, 5.26196s/12 iters), loss = 1.18875 I0406 08:14:44.279894 5226 solver.cpp:237] Train net output #0: loss = 1.18875 (* 1 = 1.18875 loss) I0406 08:14:44.279903 5226 sgd_solver.cpp:105] Iteration 7716, lr = 0.01 I0406 08:14:49.630910 5226 solver.cpp:218] Iteration 7728 (2.24258 iter/s, 5.35097s/12 iters), loss = 1.34568 I0406 08:14:49.630949 5226 solver.cpp:237] Train net output #0: loss = 1.34568 (* 1 = 1.34568 loss) I0406 08:14:49.630954 5226 sgd_solver.cpp:105] Iteration 7728, lr = 0.01 I0406 08:14:54.879954 5226 solver.cpp:218] Iteration 7740 (2.28617 iter/s, 5.24895s/12 iters), loss = 1.15676 I0406 08:14:54.880013 5226 solver.cpp:237] Train net output #0: loss = 1.15676 (* 1 = 1.15676 loss) I0406 08:14:54.880023 5226 sgd_solver.cpp:105] Iteration 7740, lr = 0.01 I0406 08:14:59.726756 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7752.caffemodel I0406 08:15:02.754528 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7752.solverstate I0406 08:15:05.071619 5226 solver.cpp:330] Iteration 7752, Testing net (#0) I0406 08:15:05.071642 5226 net.cpp:676] Ignoring source layer train-data I0406 08:15:06.454967 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:15:09.446283 5226 solver.cpp:397] Test net output #0: accuracy = 0.300858 I0406 08:15:09.446379 5226 solver.cpp:397] Test net output #1: loss = 3.31845 (* 1 = 3.31845 loss) I0406 08:15:09.587191 5226 solver.cpp:218] Iteration 7752 (0.815934 iter/s, 14.7071s/12 iters), loss = 1.67956 I0406 08:15:09.587258 5226 solver.cpp:237] Train net output #0: loss = 1.67956 (* 1 = 1.67956 loss) I0406 08:15:09.587266 5226 sgd_solver.cpp:105] Iteration 7752, lr = 0.01 I0406 08:15:14.159837 5226 solver.cpp:218] Iteration 7764 (2.62437 iter/s, 4.57253s/12 iters), loss = 1.16685 I0406 08:15:14.159886 5226 solver.cpp:237] Train net output #0: loss = 1.16685 (* 1 = 1.16685 loss) I0406 08:15:14.159895 5226 sgd_solver.cpp:105] Iteration 7764, lr = 0.01 I0406 08:15:19.173151 5226 solver.cpp:218] Iteration 7776 (2.39367 iter/s, 5.01322s/12 iters), loss = 1.31995 I0406 08:15:19.173197 5226 solver.cpp:237] Train net output #0: loss = 1.31995 (* 1 = 1.31995 loss) I0406 08:15:19.173204 5226 sgd_solver.cpp:105] Iteration 7776, lr = 0.01 I0406 08:15:24.414840 5226 solver.cpp:218] Iteration 7788 (2.28938 iter/s, 5.24159s/12 iters), loss = 1.42039 I0406 08:15:24.414888 5226 solver.cpp:237] Train net output #0: loss = 1.42039 (* 1 = 1.42039 loss) I0406 08:15:24.414896 5226 sgd_solver.cpp:105] Iteration 7788, lr = 0.01 I0406 08:15:24.421326 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:15:29.833966 5226 solver.cpp:218] Iteration 7800 (2.21442 iter/s, 5.41903s/12 iters), loss = 0.994502 I0406 08:15:29.834007 5226 solver.cpp:237] Train net output #0: loss = 0.994502 (* 1 = 0.994502 loss) I0406 08:15:29.834014 5226 sgd_solver.cpp:105] Iteration 7800, lr = 0.01 I0406 08:15:35.034857 5226 solver.cpp:218] Iteration 7812 (2.30734 iter/s, 5.2008s/12 iters), loss = 1.02511 I0406 08:15:35.034896 5226 solver.cpp:237] Train net output #0: loss = 1.02511 (* 1 = 1.02511 loss) I0406 08:15:35.034901 5226 sgd_solver.cpp:105] Iteration 7812, lr = 0.01 I0406 08:15:40.142539 5226 solver.cpp:218] Iteration 7824 (2.34944 iter/s, 5.10759s/12 iters), loss = 1.25256 I0406 08:15:40.142664 5226 solver.cpp:237] Train net output #0: loss = 1.25256 (* 1 = 1.25256 loss) I0406 08:15:40.142670 5226 sgd_solver.cpp:105] Iteration 7824, lr = 0.01 I0406 08:15:45.343688 5226 solver.cpp:218] Iteration 7836 (2.30726 iter/s, 5.20097s/12 iters), loss = 1.59217 I0406 08:15:45.343725 5226 solver.cpp:237] Train net output #0: loss = 1.59217 (* 1 = 1.59217 loss) I0406 08:15:45.343731 5226 sgd_solver.cpp:105] Iteration 7836, lr = 0.01 I0406 08:15:50.478739 5226 solver.cpp:218] Iteration 7848 (2.33692 iter/s, 5.13496s/12 iters), loss = 1.28422 I0406 08:15:50.478787 5226 solver.cpp:237] Train net output #0: loss = 1.28422 (* 1 = 1.28422 loss) I0406 08:15:50.478796 5226 sgd_solver.cpp:105] Iteration 7848, lr = 0.01 I0406 08:15:52.602519 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7854.caffemodel I0406 08:15:55.569321 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7854.solverstate I0406 08:15:57.954066 5226 solver.cpp:330] Iteration 7854, Testing net (#0) I0406 08:15:57.954085 5226 net.cpp:676] Ignoring source layer train-data I0406 08:15:59.338569 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:16:02.464704 5226 solver.cpp:397] Test net output #0: accuracy = 0.311274 I0406 08:16:02.464740 5226 solver.cpp:397] Test net output #1: loss = 3.37465 (* 1 = 3.37465 loss) I0406 08:16:04.451552 5226 solver.cpp:218] Iteration 7860 (0.85882 iter/s, 13.9727s/12 iters), loss = 1.3159 I0406 08:16:04.451591 5226 solver.cpp:237] Train net output #0: loss = 1.3159 (* 1 = 1.3159 loss) I0406 08:16:04.451596 5226 sgd_solver.cpp:105] Iteration 7860, lr = 0.01 I0406 08:16:09.770318 5226 solver.cpp:218] Iteration 7872 (2.2562 iter/s, 5.31867s/12 iters), loss = 0.957102 I0406 08:16:09.770368 5226 solver.cpp:237] Train net output #0: loss = 0.957102 (* 1 = 0.957102 loss) I0406 08:16:09.770375 5226 sgd_solver.cpp:105] Iteration 7872, lr = 0.01 I0406 08:16:14.976897 5226 solver.cpp:218] Iteration 7884 (2.30482 iter/s, 5.20647s/12 iters), loss = 0.968111 I0406 08:16:14.977003 5226 solver.cpp:237] Train net output #0: loss = 0.968111 (* 1 = 0.968111 loss) I0406 08:16:14.977010 5226 sgd_solver.cpp:105] Iteration 7884, lr = 0.01 I0406 08:16:17.292248 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:16:20.365130 5226 solver.cpp:218] Iteration 7896 (2.22714 iter/s, 5.38808s/12 iters), loss = 1.17807 I0406 08:16:20.365170 5226 solver.cpp:237] Train net output #0: loss = 1.17807 (* 1 = 1.17807 loss) I0406 08:16:20.365176 5226 sgd_solver.cpp:105] Iteration 7896, lr = 0.01 I0406 08:16:25.706619 5226 solver.cpp:218] Iteration 7908 (2.2466 iter/s, 5.3414s/12 iters), loss = 1.23866 I0406 08:16:25.706658 5226 solver.cpp:237] Train net output #0: loss = 1.23866 (* 1 = 1.23866 loss) I0406 08:16:25.706663 5226 sgd_solver.cpp:105] Iteration 7908, lr = 0.01 I0406 08:16:30.865837 5226 solver.cpp:218] Iteration 7920 (2.32598 iter/s, 5.15913s/12 iters), loss = 1.48775 I0406 08:16:30.865890 5226 solver.cpp:237] Train net output #0: loss = 1.48775 (* 1 = 1.48775 loss) I0406 08:16:30.865898 5226 sgd_solver.cpp:105] Iteration 7920, lr = 0.01 I0406 08:16:36.146890 5226 solver.cpp:218] Iteration 7932 (2.27232 iter/s, 5.28095s/12 iters), loss = 1.09502 I0406 08:16:36.146929 5226 solver.cpp:237] Train net output #0: loss = 1.09502 (* 1 = 1.09502 loss) I0406 08:16:36.146935 5226 sgd_solver.cpp:105] Iteration 7932, lr = 0.01 I0406 08:16:41.468587 5226 solver.cpp:218] Iteration 7944 (2.25496 iter/s, 5.32161s/12 iters), loss = 1.32744 I0406 08:16:41.468643 5226 solver.cpp:237] Train net output #0: loss = 1.32744 (* 1 = 1.32744 loss) I0406 08:16:41.468652 5226 sgd_solver.cpp:105] Iteration 7944, lr = 0.01 I0406 08:16:46.070426 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7956.caffemodel I0406 08:16:49.052668 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7956.solverstate I0406 08:16:51.362301 5226 solver.cpp:330] Iteration 7956, Testing net (#0) I0406 08:16:51.362320 5226 net.cpp:676] Ignoring source layer train-data I0406 08:16:52.631026 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:16:55.792701 5226 solver.cpp:397] Test net output #0: accuracy = 0.284926 I0406 08:16:55.792737 5226 solver.cpp:397] Test net output #1: loss = 3.43636 (* 1 = 3.43636 loss) I0406 08:16:55.933223 5226 solver.cpp:218] Iteration 7956 (0.829619 iter/s, 14.4645s/12 iters), loss = 1.40492 I0406 08:16:55.933270 5226 solver.cpp:237] Train net output #0: loss = 1.40492 (* 1 = 1.40492 loss) I0406 08:16:55.933279 5226 sgd_solver.cpp:105] Iteration 7956, lr = 0.01 I0406 08:17:00.406278 5226 solver.cpp:218] Iteration 7968 (2.68278 iter/s, 4.47296s/12 iters), loss = 1.70144 I0406 08:17:00.406316 5226 solver.cpp:237] Train net output #0: loss = 1.70144 (* 1 = 1.70144 loss) I0406 08:17:00.406322 5226 sgd_solver.cpp:105] Iteration 7968, lr = 0.01 I0406 08:17:05.543663 5226 solver.cpp:218] Iteration 7980 (2.33586 iter/s, 5.1373s/12 iters), loss = 1.09255 I0406 08:17:05.543707 5226 solver.cpp:237] Train net output #0: loss = 1.09255 (* 1 = 1.09255 loss) I0406 08:17:05.543715 5226 sgd_solver.cpp:105] Iteration 7980, lr = 0.01 I0406 08:17:10.028910 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:17:10.816298 5226 solver.cpp:218] Iteration 7992 (2.27594 iter/s, 5.27254s/12 iters), loss = 0.917287 I0406 08:17:10.816349 5226 solver.cpp:237] Train net output #0: loss = 0.917287 (* 1 = 0.917287 loss) I0406 08:17:10.816359 5226 sgd_solver.cpp:105] Iteration 7992, lr = 0.01 I0406 08:17:16.058558 5226 solver.cpp:218] Iteration 8004 (2.28913 iter/s, 5.24216s/12 iters), loss = 1.26003 I0406 08:17:16.058596 5226 solver.cpp:237] Train net output #0: loss = 1.26003 (* 1 = 1.26003 loss) I0406 08:17:16.058602 5226 sgd_solver.cpp:105] Iteration 8004, lr = 0.01 I0406 08:17:21.362241 5226 solver.cpp:218] Iteration 8016 (2.26262 iter/s, 5.30359s/12 iters), loss = 1.2735 I0406 08:17:21.362334 5226 solver.cpp:237] Train net output #0: loss = 1.2735 (* 1 = 1.2735 loss) I0406 08:17:21.362341 5226 sgd_solver.cpp:105] Iteration 8016, lr = 0.01 I0406 08:17:26.656569 5226 solver.cpp:218] Iteration 8028 (2.26664 iter/s, 5.29419s/12 iters), loss = 1.02471 I0406 08:17:26.656608 5226 solver.cpp:237] Train net output #0: loss = 1.02471 (* 1 = 1.02471 loss) I0406 08:17:26.656615 5226 sgd_solver.cpp:105] Iteration 8028, lr = 0.01 I0406 08:17:32.000152 5226 solver.cpp:218] Iteration 8040 (2.24572 iter/s, 5.34349s/12 iters), loss = 1.22829 I0406 08:17:32.000207 5226 solver.cpp:237] Train net output #0: loss = 1.22829 (* 1 = 1.22829 loss) I0406 08:17:32.000216 5226 sgd_solver.cpp:105] Iteration 8040, lr = 0.01 I0406 08:17:37.124398 5226 solver.cpp:218] Iteration 8052 (2.34186 iter/s, 5.12414s/12 iters), loss = 1.16577 I0406 08:17:37.130602 5226 solver.cpp:237] Train net output #0: loss = 1.16577 (* 1 = 1.16577 loss) I0406 08:17:37.130621 5226 sgd_solver.cpp:105] Iteration 8052, lr = 0.01 I0406 08:17:39.280782 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8058.caffemodel I0406 08:17:42.320255 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8058.solverstate I0406 08:17:44.614873 5226 solver.cpp:330] Iteration 8058, Testing net (#0) I0406 08:17:44.614890 5226 net.cpp:676] Ignoring source layer train-data I0406 08:17:45.845966 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:17:48.951957 5226 solver.cpp:397] Test net output #0: accuracy = 0.321078 I0406 08:17:48.951992 5226 solver.cpp:397] Test net output #1: loss = 3.33008 (* 1 = 3.33008 loss) I0406 08:17:50.895759 5226 solver.cpp:218] Iteration 8064 (0.871772 iter/s, 13.7651s/12 iters), loss = 1.08551 I0406 08:17:50.895797 5226 solver.cpp:237] Train net output #0: loss = 1.08551 (* 1 = 1.08551 loss) I0406 08:17:50.895802 5226 sgd_solver.cpp:105] Iteration 8064, lr = 0.01 I0406 08:17:56.124384 5226 solver.cpp:218] Iteration 8076 (2.2951 iter/s, 5.22854s/12 iters), loss = 1.28771 I0406 08:17:56.124514 5226 solver.cpp:237] Train net output #0: loss = 1.28771 (* 1 = 1.28771 loss) I0406 08:17:56.124523 5226 sgd_solver.cpp:105] Iteration 8076, lr = 0.01 I0406 08:18:01.484093 5226 solver.cpp:218] Iteration 8088 (2.239 iter/s, 5.35953s/12 iters), loss = 1.1622 I0406 08:18:01.484143 5226 solver.cpp:237] Train net output #0: loss = 1.1622 (* 1 = 1.1622 loss) I0406 08:18:01.484151 5226 sgd_solver.cpp:105] Iteration 8088, lr = 0.01 I0406 08:18:02.977277 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:18:06.672235 5226 solver.cpp:218] Iteration 8100 (2.31301 iter/s, 5.18805s/12 iters), loss = 1.53249 I0406 08:18:06.672273 5226 solver.cpp:237] Train net output #0: loss = 1.53249 (* 1 = 1.53249 loss) I0406 08:18:06.672279 5226 sgd_solver.cpp:105] Iteration 8100, lr = 0.01 I0406 08:18:11.852867 5226 solver.cpp:218] Iteration 8112 (2.31636 iter/s, 5.18055s/12 iters), loss = 1.47226 I0406 08:18:11.852910 5226 solver.cpp:237] Train net output #0: loss = 1.47226 (* 1 = 1.47226 loss) I0406 08:18:11.852917 5226 sgd_solver.cpp:105] Iteration 8112, lr = 0.01 I0406 08:18:17.242426 5226 solver.cpp:218] Iteration 8124 (2.22657 iter/s, 5.38946s/12 iters), loss = 0.932258 I0406 08:18:17.242475 5226 solver.cpp:237] Train net output #0: loss = 0.932258 (* 1 = 0.932258 loss) I0406 08:18:17.242482 5226 sgd_solver.cpp:105] Iteration 8124, lr = 0.01 I0406 08:18:22.506244 5226 solver.cpp:218] Iteration 8136 (2.27975 iter/s, 5.26372s/12 iters), loss = 1.23263 I0406 08:18:22.506281 5226 solver.cpp:237] Train net output #0: loss = 1.23263 (* 1 = 1.23263 loss) I0406 08:18:22.506289 5226 sgd_solver.cpp:105] Iteration 8136, lr = 0.01 I0406 08:18:27.659448 5226 solver.cpp:218] Iteration 8148 (2.32869 iter/s, 5.15312s/12 iters), loss = 1.13365 I0406 08:18:27.659556 5226 solver.cpp:237] Train net output #0: loss = 1.13365 (* 1 = 1.13365 loss) I0406 08:18:27.659564 5226 sgd_solver.cpp:105] Iteration 8148, lr = 0.01 I0406 08:18:32.456318 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8160.caffemodel I0406 08:18:35.396615 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8160.solverstate I0406 08:18:38.212797 5226 solver.cpp:330] Iteration 8160, Testing net (#0) I0406 08:18:38.212815 5226 net.cpp:676] Ignoring source layer train-data I0406 08:18:39.357486 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:18:42.547910 5226 solver.cpp:397] Test net output #0: accuracy = 0.293505 I0406 08:18:42.547950 5226 solver.cpp:397] Test net output #1: loss = 3.32601 (* 1 = 3.32601 loss) I0406 08:18:42.690167 5226 solver.cpp:218] Iteration 8160 (0.798377 iter/s, 15.0305s/12 iters), loss = 1.36887 I0406 08:18:42.690229 5226 solver.cpp:237] Train net output #0: loss = 1.36887 (* 1 = 1.36887 loss) I0406 08:18:42.690237 5226 sgd_solver.cpp:105] Iteration 8160, lr = 0.01 I0406 08:18:47.007655 5226 solver.cpp:218] Iteration 8172 (2.77946 iter/s, 4.31738s/12 iters), loss = 1.04864 I0406 08:18:47.007694 5226 solver.cpp:237] Train net output #0: loss = 1.04864 (* 1 = 1.04864 loss) I0406 08:18:47.007699 5226 sgd_solver.cpp:105] Iteration 8172, lr = 0.01 I0406 08:18:52.347029 5226 solver.cpp:218] Iteration 8184 (2.24749 iter/s, 5.33929s/12 iters), loss = 1.419 I0406 08:18:52.347066 5226 solver.cpp:237] Train net output #0: loss = 1.419 (* 1 = 1.419 loss) I0406 08:18:52.347071 5226 sgd_solver.cpp:105] Iteration 8184, lr = 0.01 I0406 08:18:56.222564 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:18:57.685130 5226 solver.cpp:218] Iteration 8196 (2.24803 iter/s, 5.33801s/12 iters), loss = 1.50788 I0406 08:18:57.685292 5226 solver.cpp:237] Train net output #0: loss = 1.50788 (* 1 = 1.50788 loss) I0406 08:18:57.685302 5226 sgd_solver.cpp:105] Iteration 8196, lr = 0.01 I0406 08:19:02.840390 5226 solver.cpp:218] Iteration 8208 (2.32781 iter/s, 5.15505s/12 iters), loss = 1.13445 I0406 08:19:02.840437 5226 solver.cpp:237] Train net output #0: loss = 1.13445 (* 1 = 1.13445 loss) I0406 08:19:02.840445 5226 sgd_solver.cpp:105] Iteration 8208, lr = 0.01 I0406 08:19:08.220216 5226 solver.cpp:218] Iteration 8220 (2.2306 iter/s, 5.37972s/12 iters), loss = 1.22965 I0406 08:19:08.220269 5226 solver.cpp:237] Train net output #0: loss = 1.22965 (* 1 = 1.22965 loss) I0406 08:19:08.220278 5226 sgd_solver.cpp:105] Iteration 8220, lr = 0.01 I0406 08:19:13.686934 5226 solver.cpp:218] Iteration 8232 (2.19514 iter/s, 5.46661s/12 iters), loss = 1.01847 I0406 08:19:13.686986 5226 solver.cpp:237] Train net output #0: loss = 1.01847 (* 1 = 1.01847 loss) I0406 08:19:13.686995 5226 sgd_solver.cpp:105] Iteration 8232, lr = 0.01 I0406 08:19:19.153795 5226 solver.cpp:218] Iteration 8244 (2.19509 iter/s, 5.46676s/12 iters), loss = 1.13078 I0406 08:19:19.153841 5226 solver.cpp:237] Train net output #0: loss = 1.13078 (* 1 = 1.13078 loss) I0406 08:19:19.153848 5226 sgd_solver.cpp:105] Iteration 8244, lr = 0.01 I0406 08:19:24.106479 5226 solver.cpp:218] Iteration 8256 (2.42298 iter/s, 4.95259s/12 iters), loss = 1.16417 I0406 08:19:24.106530 5226 solver.cpp:237] Train net output #0: loss = 1.16417 (* 1 = 1.16417 loss) I0406 08:19:24.106537 5226 sgd_solver.cpp:105] Iteration 8256, lr = 0.01 I0406 08:19:26.105307 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8262.caffemodel I0406 08:19:29.191109 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8262.solverstate I0406 08:19:31.504225 5226 solver.cpp:330] Iteration 8262, Testing net (#0) I0406 08:19:31.504245 5226 net.cpp:676] Ignoring source layer train-data I0406 08:19:32.660403 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:19:35.859583 5226 solver.cpp:397] Test net output #0: accuracy = 0.303309 I0406 08:19:35.859609 5226 solver.cpp:397] Test net output #1: loss = 3.36301 (* 1 = 3.36301 loss) I0406 08:19:37.807512 5226 solver.cpp:218] Iteration 8268 (0.875857 iter/s, 13.7009s/12 iters), loss = 1.33566 I0406 08:19:37.807565 5226 solver.cpp:237] Train net output #0: loss = 1.33566 (* 1 = 1.33566 loss) I0406 08:19:37.807574 5226 sgd_solver.cpp:105] Iteration 8268, lr = 0.01 I0406 08:19:43.053740 5226 solver.cpp:218] Iteration 8280 (2.2874 iter/s, 5.24612s/12 iters), loss = 1.01566 I0406 08:19:43.053781 5226 solver.cpp:237] Train net output #0: loss = 1.01566 (* 1 = 1.01566 loss) I0406 08:19:43.053786 5226 sgd_solver.cpp:105] Iteration 8280, lr = 0.01 I0406 08:19:48.235564 5226 solver.cpp:218] Iteration 8292 (2.31583 iter/s, 5.18174s/12 iters), loss = 1.13512 I0406 08:19:48.235601 5226 solver.cpp:237] Train net output #0: loss = 1.13512 (* 1 = 1.13512 loss) I0406 08:19:48.235607 5226 sgd_solver.cpp:105] Iteration 8292, lr = 0.01 I0406 08:19:48.839686 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:19:53.308055 5226 solver.cpp:218] Iteration 8304 (2.36574 iter/s, 5.0724s/12 iters), loss = 1.47649 I0406 08:19:53.308095 5226 solver.cpp:237] Train net output #0: loss = 1.47649 (* 1 = 1.47649 loss) I0406 08:19:53.308100 5226 sgd_solver.cpp:105] Iteration 8304, lr = 0.01 I0406 08:19:56.178107 5226 blocking_queue.cpp:49] Waiting for data I0406 08:19:58.527791 5226 solver.cpp:218] Iteration 8316 (2.29901 iter/s, 5.21964s/12 iters), loss = 1.12538 I0406 08:19:58.527833 5226 solver.cpp:237] Train net output #0: loss = 1.12538 (* 1 = 1.12538 loss) I0406 08:19:58.527839 5226 sgd_solver.cpp:105] Iteration 8316, lr = 0.01 I0406 08:20:03.590566 5226 solver.cpp:218] Iteration 8328 (2.37028 iter/s, 5.06268s/12 iters), loss = 1.67171 I0406 08:20:03.590728 5226 solver.cpp:237] Train net output #0: loss = 1.67171 (* 1 = 1.67171 loss) I0406 08:20:03.590737 5226 sgd_solver.cpp:105] Iteration 8328, lr = 0.01 I0406 08:20:08.857568 5226 solver.cpp:218] Iteration 8340 (2.27843 iter/s, 5.26679s/12 iters), loss = 1.3362 I0406 08:20:08.857612 5226 solver.cpp:237] Train net output #0: loss = 1.3362 (* 1 = 1.3362 loss) I0406 08:20:08.857618 5226 sgd_solver.cpp:105] Iteration 8340, lr = 0.01 I0406 08:20:13.979979 5226 solver.cpp:218] Iteration 8352 (2.34269 iter/s, 5.12232s/12 iters), loss = 1.17571 I0406 08:20:13.980020 5226 solver.cpp:237] Train net output #0: loss = 1.17571 (* 1 = 1.17571 loss) I0406 08:20:13.980026 5226 sgd_solver.cpp:105] Iteration 8352, lr = 0.01 I0406 08:20:18.751482 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8364.caffemodel I0406 08:20:21.770217 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8364.solverstate I0406 08:20:24.081646 5226 solver.cpp:330] Iteration 8364, Testing net (#0) I0406 08:20:24.081665 5226 net.cpp:676] Ignoring source layer train-data I0406 08:20:25.175899 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:20:28.435474 5226 solver.cpp:397] Test net output #0: accuracy = 0.30576 I0406 08:20:28.435505 5226 solver.cpp:397] Test net output #1: loss = 3.3445 (* 1 = 3.3445 loss) I0406 08:20:28.576171 5226 solver.cpp:218] Iteration 8364 (0.822141 iter/s, 14.596s/12 iters), loss = 1.7042 I0406 08:20:28.576225 5226 solver.cpp:237] Train net output #0: loss = 1.7042 (* 1 = 1.7042 loss) I0406 08:20:28.576232 5226 sgd_solver.cpp:105] Iteration 8364, lr = 0.01 I0406 08:20:32.880053 5226 solver.cpp:218] Iteration 8376 (2.78825 iter/s, 4.30378s/12 iters), loss = 1.22799 I0406 08:20:32.880102 5226 solver.cpp:237] Train net output #0: loss = 1.22799 (* 1 = 1.22799 loss) I0406 08:20:32.880110 5226 sgd_solver.cpp:105] Iteration 8376, lr = 0.01 I0406 08:20:38.252595 5226 solver.cpp:218] Iteration 8388 (2.23362 iter/s, 5.37244s/12 iters), loss = 1.61775 I0406 08:20:38.252692 5226 solver.cpp:237] Train net output #0: loss = 1.61775 (* 1 = 1.61775 loss) I0406 08:20:38.252698 5226 sgd_solver.cpp:105] Iteration 8388, lr = 0.01 I0406 08:20:41.176440 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:20:43.463783 5226 solver.cpp:218] Iteration 8400 (2.3028 iter/s, 5.21104s/12 iters), loss = 1.59125 I0406 08:20:43.463824 5226 solver.cpp:237] Train net output #0: loss = 1.59125 (* 1 = 1.59125 loss) I0406 08:20:43.463829 5226 sgd_solver.cpp:105] Iteration 8400, lr = 0.01 I0406 08:20:48.681504 5226 solver.cpp:218] Iteration 8412 (2.29989 iter/s, 5.21763s/12 iters), loss = 1.36855 I0406 08:20:48.681541 5226 solver.cpp:237] Train net output #0: loss = 1.36855 (* 1 = 1.36855 loss) I0406 08:20:48.681546 5226 sgd_solver.cpp:105] Iteration 8412, lr = 0.01 I0406 08:20:54.009905 5226 solver.cpp:218] Iteration 8424 (2.25212 iter/s, 5.32831s/12 iters), loss = 1.62781 I0406 08:20:54.009948 5226 solver.cpp:237] Train net output #0: loss = 1.62781 (* 1 = 1.62781 loss) I0406 08:20:54.009953 5226 sgd_solver.cpp:105] Iteration 8424, lr = 0.01 I0406 08:20:59.372435 5226 solver.cpp:218] Iteration 8436 (2.23779 iter/s, 5.36244s/12 iters), loss = 1.44612 I0406 08:20:59.372488 5226 solver.cpp:237] Train net output #0: loss = 1.44612 (* 1 = 1.44612 loss) I0406 08:20:59.372496 5226 sgd_solver.cpp:105] Iteration 8436, lr = 0.01 I0406 08:21:04.682574 5226 solver.cpp:218] Iteration 8448 (2.25987 iter/s, 5.31004s/12 iters), loss = 1.05502 I0406 08:21:04.682619 5226 solver.cpp:237] Train net output #0: loss = 1.05502 (* 1 = 1.05502 loss) I0406 08:21:04.682627 5226 sgd_solver.cpp:105] Iteration 8448, lr = 0.01 I0406 08:21:09.927034 5226 solver.cpp:218] Iteration 8460 (2.28817 iter/s, 5.24436s/12 iters), loss = 1.53199 I0406 08:21:09.927155 5226 solver.cpp:237] Train net output #0: loss = 1.53199 (* 1 = 1.53199 loss) I0406 08:21:09.927162 5226 sgd_solver.cpp:105] Iteration 8460, lr = 0.01 I0406 08:21:12.165880 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8466.caffemodel I0406 08:21:15.118239 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8466.solverstate I0406 08:21:17.431269 5226 solver.cpp:330] Iteration 8466, Testing net (#0) I0406 08:21:17.431293 5226 net.cpp:676] Ignoring source layer train-data I0406 08:21:18.559489 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:21:21.874444 5226 solver.cpp:397] Test net output #0: accuracy = 0.308211 I0406 08:21:21.874486 5226 solver.cpp:397] Test net output #1: loss = 3.2345 (* 1 = 3.2345 loss) I0406 08:21:23.809402 5226 solver.cpp:218] Iteration 8472 (0.86442 iter/s, 13.8821s/12 iters), loss = 1.3732 I0406 08:21:23.809445 5226 solver.cpp:237] Train net output #0: loss = 1.3732 (* 1 = 1.3732 loss) I0406 08:21:23.809451 5226 sgd_solver.cpp:105] Iteration 8472, lr = 0.01 I0406 08:21:29.215873 5226 solver.cpp:218] Iteration 8484 (2.2196 iter/s, 5.40638s/12 iters), loss = 1.30494 I0406 08:21:29.215911 5226 solver.cpp:237] Train net output #0: loss = 1.30494 (* 1 = 1.30494 loss) I0406 08:21:29.215917 5226 sgd_solver.cpp:105] Iteration 8484, lr = 0.01 I0406 08:21:34.598186 5226 solver.cpp:218] Iteration 8496 (2.22956 iter/s, 5.38222s/12 iters), loss = 1.22598 I0406 08:21:34.598224 5226 solver.cpp:237] Train net output #0: loss = 1.22598 (* 1 = 1.22598 loss) I0406 08:21:34.598230 5226 sgd_solver.cpp:105] Iteration 8496, lr = 0.01 I0406 08:21:34.644646 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:21:39.863821 5226 solver.cpp:218] Iteration 8508 (2.27897 iter/s, 5.26554s/12 iters), loss = 1.02827 I0406 08:21:39.863857 5226 solver.cpp:237] Train net output #0: loss = 1.02827 (* 1 = 1.02827 loss) I0406 08:21:39.863862 5226 sgd_solver.cpp:105] Iteration 8508, lr = 0.01 I0406 08:21:45.076678 5226 solver.cpp:218] Iteration 8520 (2.30204 iter/s, 5.21277s/12 iters), loss = 0.969367 I0406 08:21:45.076758 5226 solver.cpp:237] Train net output #0: loss = 0.969367 (* 1 = 0.969367 loss) I0406 08:21:45.076764 5226 sgd_solver.cpp:105] Iteration 8520, lr = 0.01 I0406 08:21:50.271962 5226 solver.cpp:218] Iteration 8532 (2.30985 iter/s, 5.19515s/12 iters), loss = 0.872375 I0406 08:21:50.272001 5226 solver.cpp:237] Train net output #0: loss = 0.872375 (* 1 = 0.872375 loss) I0406 08:21:50.272006 5226 sgd_solver.cpp:105] Iteration 8532, lr = 0.01 I0406 08:21:55.517562 5226 solver.cpp:218] Iteration 8544 (2.28767 iter/s, 5.24551s/12 iters), loss = 1.03351 I0406 08:21:55.517602 5226 solver.cpp:237] Train net output #0: loss = 1.03351 (* 1 = 1.03351 loss) I0406 08:21:55.517608 5226 sgd_solver.cpp:105] Iteration 8544, lr = 0.01 I0406 08:22:00.844668 5226 solver.cpp:218] Iteration 8556 (2.25267 iter/s, 5.32701s/12 iters), loss = 1.33246 I0406 08:22:00.844705 5226 solver.cpp:237] Train net output #0: loss = 1.33246 (* 1 = 1.33246 loss) I0406 08:22:00.844712 5226 sgd_solver.cpp:105] Iteration 8556, lr = 0.01 I0406 08:22:05.432304 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8568.caffemodel I0406 08:22:08.544708 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8568.solverstate I0406 08:22:10.849048 5226 solver.cpp:330] Iteration 8568, Testing net (#0) I0406 08:22:10.849066 5226 net.cpp:676] Ignoring source layer train-data I0406 08:22:11.845108 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:22:15.296181 5226 solver.cpp:397] Test net output #0: accuracy = 0.289828 I0406 08:22:15.296315 5226 solver.cpp:397] Test net output #1: loss = 3.37105 (* 1 = 3.37105 loss) I0406 08:22:15.436966 5226 solver.cpp:218] Iteration 8568 (0.82236 iter/s, 14.5921s/12 iters), loss = 1.46079 I0406 08:22:15.438546 5226 solver.cpp:237] Train net output #0: loss = 1.46079 (* 1 = 1.46079 loss) I0406 08:22:15.438557 5226 sgd_solver.cpp:105] Iteration 8568, lr = 0.01 I0406 08:22:19.766515 5226 solver.cpp:218] Iteration 8580 (2.77269 iter/s, 4.32793s/12 iters), loss = 1.31069 I0406 08:22:19.766559 5226 solver.cpp:237] Train net output #0: loss = 1.31069 (* 1 = 1.31069 loss) I0406 08:22:19.766566 5226 sgd_solver.cpp:105] Iteration 8580, lr = 0.01 I0406 08:22:25.097669 5226 solver.cpp:218] Iteration 8592 (2.25096 iter/s, 5.33106s/12 iters), loss = 1.54103 I0406 08:22:25.097707 5226 solver.cpp:237] Train net output #0: loss = 1.54103 (* 1 = 1.54103 loss) I0406 08:22:25.097712 5226 sgd_solver.cpp:105] Iteration 8592, lr = 0.01 I0406 08:22:27.464597 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:22:30.577822 5226 solver.cpp:218] Iteration 8604 (2.18976 iter/s, 5.48006s/12 iters), loss = 1.08005 I0406 08:22:30.577862 5226 solver.cpp:237] Train net output #0: loss = 1.08005 (* 1 = 1.08005 loss) I0406 08:22:30.577867 5226 sgd_solver.cpp:105] Iteration 8604, lr = 0.01 I0406 08:22:35.865615 5226 solver.cpp:218] Iteration 8616 (2.26942 iter/s, 5.2877s/12 iters), loss = 1.39324 I0406 08:22:35.865654 5226 solver.cpp:237] Train net output #0: loss = 1.39324 (* 1 = 1.39324 loss) I0406 08:22:35.865659 5226 sgd_solver.cpp:105] Iteration 8616, lr = 0.01 I0406 08:22:41.075958 5226 solver.cpp:218] Iteration 8628 (2.30315 iter/s, 5.21025s/12 iters), loss = 1.53359 I0406 08:22:41.076004 5226 solver.cpp:237] Train net output #0: loss = 1.53359 (* 1 = 1.53359 loss) I0406 08:22:41.076012 5226 sgd_solver.cpp:105] Iteration 8628, lr = 0.01 I0406 08:22:46.318447 5226 solver.cpp:218] Iteration 8640 (2.28903 iter/s, 5.2424s/12 iters), loss = 1.19249 I0406 08:22:46.318537 5226 solver.cpp:237] Train net output #0: loss = 1.19249 (* 1 = 1.19249 loss) I0406 08:22:46.318543 5226 sgd_solver.cpp:105] Iteration 8640, lr = 0.01 I0406 08:22:51.447912 5226 solver.cpp:218] Iteration 8652 (2.33949 iter/s, 5.12932s/12 iters), loss = 1.35614 I0406 08:22:51.447960 5226 solver.cpp:237] Train net output #0: loss = 1.35614 (* 1 = 1.35614 loss) I0406 08:22:51.447968 5226 sgd_solver.cpp:105] Iteration 8652, lr = 0.01 I0406 08:22:56.760874 5226 solver.cpp:218] Iteration 8664 (2.25867 iter/s, 5.31287s/12 iters), loss = 1.21559 I0406 08:22:56.760915 5226 solver.cpp:237] Train net output #0: loss = 1.21559 (* 1 = 1.21559 loss) I0406 08:22:56.760919 5226 sgd_solver.cpp:105] Iteration 8664, lr = 0.01 I0406 08:22:59.023504 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8670.caffemodel I0406 08:23:02.107959 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8670.solverstate I0406 08:23:04.428999 5226 solver.cpp:330] Iteration 8670, Testing net (#0) I0406 08:23:04.429023 5226 net.cpp:676] Ignoring source layer train-data I0406 08:23:05.438325 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:23:08.869032 5226 solver.cpp:397] Test net output #0: accuracy = 0.296569 I0406 08:23:08.869065 5226 solver.cpp:397] Test net output #1: loss = 3.35544 (* 1 = 3.35544 loss) I0406 08:23:10.761075 5226 solver.cpp:218] Iteration 8676 (0.85714 iter/s, 14s/12 iters), loss = 1.42445 I0406 08:23:10.761126 5226 solver.cpp:237] Train net output #0: loss = 1.42445 (* 1 = 1.42445 loss) I0406 08:23:10.761134 5226 sgd_solver.cpp:105] Iteration 8676, lr = 0.01 I0406 08:23:16.192761 5226 solver.cpp:218] Iteration 8688 (2.2093 iter/s, 5.43158s/12 iters), loss = 1.24362 I0406 08:23:16.192809 5226 solver.cpp:237] Train net output #0: loss = 1.24362 (* 1 = 1.24362 loss) I0406 08:23:16.192817 5226 sgd_solver.cpp:105] Iteration 8688, lr = 0.01 I0406 08:23:20.748147 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:23:21.526031 5226 solver.cpp:218] Iteration 8700 (2.25007 iter/s, 5.33317s/12 iters), loss = 1.19184 I0406 08:23:21.526077 5226 solver.cpp:237] Train net output #0: loss = 1.19184 (* 1 = 1.19184 loss) I0406 08:23:21.526085 5226 sgd_solver.cpp:105] Iteration 8700, lr = 0.01 I0406 08:23:26.649806 5226 solver.cpp:218] Iteration 8712 (2.34207 iter/s, 5.12368s/12 iters), loss = 1.61335 I0406 08:23:26.649843 5226 solver.cpp:237] Train net output #0: loss = 1.61335 (* 1 = 1.61335 loss) I0406 08:23:26.649848 5226 sgd_solver.cpp:105] Iteration 8712, lr = 0.01 I0406 08:23:31.873006 5226 solver.cpp:218] Iteration 8724 (2.29748 iter/s, 5.22311s/12 iters), loss = 1.33385 I0406 08:23:31.873044 5226 solver.cpp:237] Train net output #0: loss = 1.33385 (* 1 = 1.33385 loss) I0406 08:23:31.873049 5226 sgd_solver.cpp:105] Iteration 8724, lr = 0.01 I0406 08:23:37.007655 5226 solver.cpp:218] Iteration 8736 (2.3371 iter/s, 5.13456s/12 iters), loss = 1.41583 I0406 08:23:37.007699 5226 solver.cpp:237] Train net output #0: loss = 1.41583 (* 1 = 1.41583 loss) I0406 08:23:37.007706 5226 sgd_solver.cpp:105] Iteration 8736, lr = 0.01 I0406 08:23:42.131088 5226 solver.cpp:218] Iteration 8748 (2.34222 iter/s, 5.12334s/12 iters), loss = 1.62163 I0406 08:23:42.131135 5226 solver.cpp:237] Train net output #0: loss = 1.62163 (* 1 = 1.62163 loss) I0406 08:23:42.131143 5226 sgd_solver.cpp:105] Iteration 8748, lr = 0.01 I0406 08:23:47.326074 5226 solver.cpp:218] Iteration 8760 (2.30996 iter/s, 5.19489s/12 iters), loss = 1.13432 I0406 08:23:47.326122 5226 solver.cpp:237] Train net output #0: loss = 1.13432 (* 1 = 1.13432 loss) I0406 08:23:47.326129 5226 sgd_solver.cpp:105] Iteration 8760, lr = 0.01 I0406 08:23:51.880928 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8772.caffemodel I0406 08:23:54.911284 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8772.solverstate I0406 08:23:57.210294 5226 solver.cpp:330] Iteration 8772, Testing net (#0) I0406 08:23:57.210312 5226 net.cpp:676] Ignoring source layer train-data I0406 08:23:58.161995 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:24:01.664041 5226 solver.cpp:397] Test net output #0: accuracy = 0.295343 I0406 08:24:01.664077 5226 solver.cpp:397] Test net output #1: loss = 3.29914 (* 1 = 3.29914 loss) I0406 08:24:01.804873 5226 solver.cpp:218] Iteration 8772 (0.828807 iter/s, 14.4786s/12 iters), loss = 1.01154 I0406 08:24:01.804920 5226 solver.cpp:237] Train net output #0: loss = 1.01154 (* 1 = 1.01154 loss) I0406 08:24:01.804926 5226 sgd_solver.cpp:105] Iteration 8772, lr = 0.01 I0406 08:24:06.319221 5226 solver.cpp:218] Iteration 8784 (2.65825 iter/s, 4.51425s/12 iters), loss = 1.20669 I0406 08:24:06.319273 5226 solver.cpp:237] Train net output #0: loss = 1.20669 (* 1 = 1.20669 loss) I0406 08:24:06.319283 5226 sgd_solver.cpp:105] Iteration 8784, lr = 0.01 I0406 08:24:11.599531 5226 solver.cpp:218] Iteration 8796 (2.27264 iter/s, 5.28021s/12 iters), loss = 1.46906 I0406 08:24:11.599570 5226 solver.cpp:237] Train net output #0: loss = 1.46906 (* 1 = 1.46906 loss) I0406 08:24:11.599575 5226 sgd_solver.cpp:105] Iteration 8796, lr = 0.01 I0406 08:24:13.049666 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:24:16.838194 5226 solver.cpp:218] Iteration 8808 (2.2907 iter/s, 5.23857s/12 iters), loss = 1.01466 I0406 08:24:16.838235 5226 solver.cpp:237] Train net output #0: loss = 1.01466 (* 1 = 1.01466 loss) I0406 08:24:16.838241 5226 sgd_solver.cpp:105] Iteration 8808, lr = 0.01 I0406 08:24:22.317852 5226 solver.cpp:218] Iteration 8820 (2.18995 iter/s, 5.47957s/12 iters), loss = 1.11431 I0406 08:24:22.317983 5226 solver.cpp:237] Train net output #0: loss = 1.11431 (* 1 = 1.11431 loss) I0406 08:24:22.317989 5226 sgd_solver.cpp:105] Iteration 8820, lr = 0.01 I0406 08:24:27.517925 5226 solver.cpp:218] Iteration 8832 (2.30774 iter/s, 5.19989s/12 iters), loss = 1.28947 I0406 08:24:27.517963 5226 solver.cpp:237] Train net output #0: loss = 1.28947 (* 1 = 1.28947 loss) I0406 08:24:27.517969 5226 sgd_solver.cpp:105] Iteration 8832, lr = 0.01 I0406 08:24:32.827482 5226 solver.cpp:218] Iteration 8844 (2.26011 iter/s, 5.30946s/12 iters), loss = 1.53667 I0406 08:24:32.827524 5226 solver.cpp:237] Train net output #0: loss = 1.53667 (* 1 = 1.53667 loss) I0406 08:24:32.827530 5226 sgd_solver.cpp:105] Iteration 8844, lr = 0.01 I0406 08:24:38.063324 5226 solver.cpp:218] Iteration 8856 (2.29193 iter/s, 5.23575s/12 iters), loss = 1.54275 I0406 08:24:38.063371 5226 solver.cpp:237] Train net output #0: loss = 1.54275 (* 1 = 1.54275 loss) I0406 08:24:38.063380 5226 sgd_solver.cpp:105] Iteration 8856, lr = 0.01 I0406 08:24:43.386808 5226 solver.cpp:218] Iteration 8868 (2.2542 iter/s, 5.32339s/12 iters), loss = 1.45895 I0406 08:24:43.386845 5226 solver.cpp:237] Train net output #0: loss = 1.45895 (* 1 = 1.45895 loss) I0406 08:24:43.386852 5226 sgd_solver.cpp:105] Iteration 8868, lr = 0.01 I0406 08:24:45.474002 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8874.caffemodel I0406 08:24:48.603260 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8874.solverstate I0406 08:24:50.954578 5226 solver.cpp:330] Iteration 8874, Testing net (#0) I0406 08:24:50.954597 5226 net.cpp:676] Ignoring source layer train-data I0406 08:24:51.817083 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:24:55.422986 5226 solver.cpp:397] Test net output #0: accuracy = 0.287377 I0406 08:24:55.423090 5226 solver.cpp:397] Test net output #1: loss = 3.28457 (* 1 = 3.28457 loss) I0406 08:24:57.304378 5226 solver.cpp:218] Iteration 8880 (0.862228 iter/s, 13.9174s/12 iters), loss = 1.31665 I0406 08:24:57.304438 5226 solver.cpp:237] Train net output #0: loss = 1.31665 (* 1 = 1.31665 loss) I0406 08:24:57.304447 5226 sgd_solver.cpp:105] Iteration 8880, lr = 0.01 I0406 08:25:02.558235 5226 solver.cpp:218] Iteration 8892 (2.2841 iter/s, 5.25372s/12 iters), loss = 1.17505 I0406 08:25:02.558275 5226 solver.cpp:237] Train net output #0: loss = 1.17505 (* 1 = 1.17505 loss) I0406 08:25:02.558280 5226 sgd_solver.cpp:105] Iteration 8892, lr = 0.01 I0406 08:25:06.364585 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:25:07.931510 5226 solver.cpp:218] Iteration 8904 (2.23331 iter/s, 5.37318s/12 iters), loss = 1.23806 I0406 08:25:07.931560 5226 solver.cpp:237] Train net output #0: loss = 1.23806 (* 1 = 1.23806 loss) I0406 08:25:07.931567 5226 sgd_solver.cpp:105] Iteration 8904, lr = 0.01 I0406 08:25:13.275257 5226 solver.cpp:218] Iteration 8916 (2.24566 iter/s, 5.34365s/12 iters), loss = 1.27448 I0406 08:25:13.275298 5226 solver.cpp:237] Train net output #0: loss = 1.27448 (* 1 = 1.27448 loss) I0406 08:25:13.275305 5226 sgd_solver.cpp:105] Iteration 8916, lr = 0.01 I0406 08:25:18.636299 5226 solver.cpp:218] Iteration 8928 (2.23841 iter/s, 5.36095s/12 iters), loss = 1.43343 I0406 08:25:18.636348 5226 solver.cpp:237] Train net output #0: loss = 1.43343 (* 1 = 1.43343 loss) I0406 08:25:18.636356 5226 sgd_solver.cpp:105] Iteration 8928, lr = 0.01 I0406 08:25:23.838407 5226 solver.cpp:218] Iteration 8940 (2.3068 iter/s, 5.202s/12 iters), loss = 1.55373 I0406 08:25:23.838454 5226 solver.cpp:237] Train net output #0: loss = 1.55373 (* 1 = 1.55373 loss) I0406 08:25:23.838461 5226 sgd_solver.cpp:105] Iteration 8940, lr = 0.01 I0406 08:25:29.048878 5226 solver.cpp:218] Iteration 8952 (2.3031 iter/s, 5.21037s/12 iters), loss = 1.40744 I0406 08:25:29.049011 5226 solver.cpp:237] Train net output #0: loss = 1.40744 (* 1 = 1.40744 loss) I0406 08:25:29.049018 5226 sgd_solver.cpp:105] Iteration 8952, lr = 0.01 I0406 08:25:34.345983 5226 solver.cpp:218] Iteration 8964 (2.26547 iter/s, 5.29692s/12 iters), loss = 0.961194 I0406 08:25:34.346020 5226 solver.cpp:237] Train net output #0: loss = 0.961194 (* 1 = 0.961194 loss) I0406 08:25:34.346026 5226 sgd_solver.cpp:105] Iteration 8964, lr = 0.01 I0406 08:25:39.175132 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8976.caffemodel I0406 08:25:42.219225 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8976.solverstate I0406 08:25:44.718933 5226 solver.cpp:330] Iteration 8976, Testing net (#0) I0406 08:25:44.718953 5226 net.cpp:676] Ignoring source layer train-data I0406 08:25:45.611171 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:25:49.078719 5226 solver.cpp:397] Test net output #0: accuracy = 0.280024 I0406 08:25:49.078747 5226 solver.cpp:397] Test net output #1: loss = 3.40016 (* 1 = 3.40016 loss) I0406 08:25:49.219588 5226 solver.cpp:218] Iteration 8976 (0.806807 iter/s, 14.8735s/12 iters), loss = 1.36654 I0406 08:25:49.219655 5226 solver.cpp:237] Train net output #0: loss = 1.36654 (* 1 = 1.36654 loss) I0406 08:25:49.219664 5226 sgd_solver.cpp:105] Iteration 8976, lr = 0.01 I0406 08:25:53.602304 5226 solver.cpp:218] Iteration 8988 (2.7381 iter/s, 4.3826s/12 iters), loss = 1.43849 I0406 08:25:53.602360 5226 solver.cpp:237] Train net output #0: loss = 1.43849 (* 1 = 1.43849 loss) I0406 08:25:53.602370 5226 sgd_solver.cpp:105] Iteration 8988, lr = 0.01 I0406 08:25:57.016037 5226 blocking_queue.cpp:49] Waiting for data I0406 08:25:58.854282 5226 solver.cpp:218] Iteration 9000 (2.2849 iter/s, 5.25188s/12 iters), loss = 1.54834 I0406 08:25:58.854321 5226 solver.cpp:237] Train net output #0: loss = 1.54834 (* 1 = 1.54834 loss) I0406 08:25:58.854326 5226 sgd_solver.cpp:105] Iteration 9000, lr = 0.01 I0406 08:25:59.499992 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:26:04.221757 5226 solver.cpp:218] Iteration 9012 (2.23572 iter/s, 5.36739s/12 iters), loss = 1.87543 I0406 08:26:04.221793 5226 solver.cpp:237] Train net output #0: loss = 1.87543 (* 1 = 1.87543 loss) I0406 08:26:04.221798 5226 sgd_solver.cpp:105] Iteration 9012, lr = 0.01 I0406 08:26:09.609632 5226 solver.cpp:218] Iteration 9024 (2.22726 iter/s, 5.38778s/12 iters), loss = 1.52809 I0406 08:26:09.609670 5226 solver.cpp:237] Train net output #0: loss = 1.52809 (* 1 = 1.52809 loss) I0406 08:26:09.609676 5226 sgd_solver.cpp:105] Iteration 9024, lr = 0.01 I0406 08:26:15.034828 5226 solver.cpp:218] Iteration 9036 (2.21194 iter/s, 5.42511s/12 iters), loss = 1.32086 I0406 08:26:15.034868 5226 solver.cpp:237] Train net output #0: loss = 1.32086 (* 1 = 1.32086 loss) I0406 08:26:15.034873 5226 sgd_solver.cpp:105] Iteration 9036, lr = 0.01 I0406 08:26:20.295110 5226 solver.cpp:218] Iteration 9048 (2.28129 iter/s, 5.26019s/12 iters), loss = 1.13095 I0406 08:26:20.295154 5226 solver.cpp:237] Train net output #0: loss = 1.13095 (* 1 = 1.13095 loss) I0406 08:26:20.295162 5226 sgd_solver.cpp:105] Iteration 9048, lr = 0.01 I0406 08:26:25.525097 5226 solver.cpp:218] Iteration 9060 (2.2945 iter/s, 5.22989s/12 iters), loss = 1.66655 I0406 08:26:25.525144 5226 solver.cpp:237] Train net output #0: loss = 1.66655 (* 1 = 1.66655 loss) I0406 08:26:25.525151 5226 sgd_solver.cpp:105] Iteration 9060, lr = 0.01 I0406 08:26:30.859648 5226 solver.cpp:218] Iteration 9072 (2.24953 iter/s, 5.33446s/12 iters), loss = 1.40013 I0406 08:26:30.859763 5226 solver.cpp:237] Train net output #0: loss = 1.40013 (* 1 = 1.40013 loss) I0406 08:26:30.859769 5226 sgd_solver.cpp:105] Iteration 9072, lr = 0.01 I0406 08:26:32.805733 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9078.caffemodel I0406 08:26:35.853482 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9078.solverstate I0406 08:26:38.156133 5226 solver.cpp:330] Iteration 9078, Testing net (#0) I0406 08:26:38.156153 5226 net.cpp:676] Ignoring source layer train-data I0406 08:26:38.973809 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:26:42.509582 5226 solver.cpp:397] Test net output #0: accuracy = 0.246936 I0406 08:26:42.509619 5226 solver.cpp:397] Test net output #1: loss = 3.52659 (* 1 = 3.52659 loss) I0406 08:26:44.306329 5226 solver.cpp:218] Iteration 9084 (0.892428 iter/s, 13.4465s/12 iters), loss = 1.38034 I0406 08:26:44.306368 5226 solver.cpp:237] Train net output #0: loss = 1.38034 (* 1 = 1.38034 loss) I0406 08:26:44.306375 5226 sgd_solver.cpp:105] Iteration 9084, lr = 0.01 I0406 08:26:49.823982 5226 solver.cpp:218] Iteration 9096 (2.17487 iter/s, 5.51756s/12 iters), loss = 1.23829 I0406 08:26:49.824038 5226 solver.cpp:237] Train net output #0: loss = 1.23829 (* 1 = 1.23829 loss) I0406 08:26:49.824046 5226 sgd_solver.cpp:105] Iteration 9096, lr = 0.01 I0406 08:26:52.848687 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:26:55.106520 5226 solver.cpp:218] Iteration 9108 (2.27168 iter/s, 5.28244s/12 iters), loss = 1.44825 I0406 08:26:55.106575 5226 solver.cpp:237] Train net output #0: loss = 1.44825 (* 1 = 1.44825 loss) I0406 08:26:55.106583 5226 sgd_solver.cpp:105] Iteration 9108, lr = 0.01 I0406 08:27:00.443120 5226 solver.cpp:218] Iteration 9120 (2.24867 iter/s, 5.3365s/12 iters), loss = 1.13138 I0406 08:27:00.443158 5226 solver.cpp:237] Train net output #0: loss = 1.13138 (* 1 = 1.13138 loss) I0406 08:27:00.443164 5226 sgd_solver.cpp:105] Iteration 9120, lr = 0.01 I0406 08:27:05.884299 5226 solver.cpp:218] Iteration 9132 (2.20544 iter/s, 5.44109s/12 iters), loss = 0.973152 I0406 08:27:05.884421 5226 solver.cpp:237] Train net output #0: loss = 0.973152 (* 1 = 0.973152 loss) I0406 08:27:05.884430 5226 sgd_solver.cpp:105] Iteration 9132, lr = 0.01 I0406 08:27:11.239425 5226 solver.cpp:218] Iteration 9144 (2.24091 iter/s, 5.35496s/12 iters), loss = 1.50677 I0406 08:27:11.239462 5226 solver.cpp:237] Train net output #0: loss = 1.50677 (* 1 = 1.50677 loss) I0406 08:27:11.239467 5226 sgd_solver.cpp:105] Iteration 9144, lr = 0.01 I0406 08:27:16.430109 5226 solver.cpp:218] Iteration 9156 (2.31187 iter/s, 5.1906s/12 iters), loss = 1.5159 I0406 08:27:16.430161 5226 solver.cpp:237] Train net output #0: loss = 1.5159 (* 1 = 1.5159 loss) I0406 08:27:16.430171 5226 sgd_solver.cpp:105] Iteration 9156, lr = 0.01 I0406 08:27:21.577827 5226 solver.cpp:218] Iteration 9168 (2.33117 iter/s, 5.14762s/12 iters), loss = 1.49927 I0406 08:27:21.577860 5226 solver.cpp:237] Train net output #0: loss = 1.49927 (* 1 = 1.49927 loss) I0406 08:27:21.577865 5226 sgd_solver.cpp:105] Iteration 9168, lr = 0.01 I0406 08:27:26.326304 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9180.caffemodel I0406 08:27:29.316359 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9180.solverstate I0406 08:27:31.619693 5226 solver.cpp:330] Iteration 9180, Testing net (#0) I0406 08:27:31.619717 5226 net.cpp:676] Ignoring source layer train-data I0406 08:27:32.450837 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:27:35.960037 5226 solver.cpp:397] Test net output #0: accuracy = 0.292892 I0406 08:27:35.960134 5226 solver.cpp:397] Test net output #1: loss = 3.30455 (* 1 = 3.30455 loss) I0406 08:27:36.100899 5226 solver.cpp:218] Iteration 9180 (0.82628 iter/s, 14.5229s/12 iters), loss = 1.15912 I0406 08:27:36.100951 5226 solver.cpp:237] Train net output #0: loss = 1.15912 (* 1 = 1.15912 loss) I0406 08:27:36.100960 5226 sgd_solver.cpp:105] Iteration 9180, lr = 0.01 I0406 08:27:40.214835 5226 solver.cpp:218] Iteration 9192 (2.91698 iter/s, 4.11384s/12 iters), loss = 1.58883 I0406 08:27:40.214876 5226 solver.cpp:237] Train net output #0: loss = 1.58883 (* 1 = 1.58883 loss) I0406 08:27:40.214881 5226 sgd_solver.cpp:105] Iteration 9192, lr = 0.01 I0406 08:27:45.708592 5226 solver.cpp:218] Iteration 9204 (2.18434 iter/s, 5.49366s/12 iters), loss = 1.2873 I0406 08:27:45.708642 5226 solver.cpp:237] Train net output #0: loss = 1.2873 (* 1 = 1.2873 loss) I0406 08:27:45.708649 5226 sgd_solver.cpp:105] Iteration 9204, lr = 0.01 I0406 08:27:45.774080 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:27:51.150384 5226 solver.cpp:218] Iteration 9216 (2.2052 iter/s, 5.44169s/12 iters), loss = 1.74755 I0406 08:27:51.150420 5226 solver.cpp:237] Train net output #0: loss = 1.74755 (* 1 = 1.74755 loss) I0406 08:27:51.150426 5226 sgd_solver.cpp:105] Iteration 9216, lr = 0.01 I0406 08:27:56.439889 5226 solver.cpp:218] Iteration 9228 (2.26868 iter/s, 5.28942s/12 iters), loss = 1.54323 I0406 08:27:56.439924 5226 solver.cpp:237] Train net output #0: loss = 1.54323 (* 1 = 1.54323 loss) I0406 08:27:56.439929 5226 sgd_solver.cpp:105] Iteration 9228, lr = 0.01 I0406 08:28:01.622162 5226 solver.cpp:218] Iteration 9240 (2.31563 iter/s, 5.18218s/12 iters), loss = 1.46492 I0406 08:28:01.622211 5226 solver.cpp:237] Train net output #0: loss = 1.46492 (* 1 = 1.46492 loss) I0406 08:28:01.622220 5226 sgd_solver.cpp:105] Iteration 9240, lr = 0.01 I0406 08:28:07.080965 5226 solver.cpp:218] Iteration 9252 (2.19833 iter/s, 5.4587s/12 iters), loss = 1.4019 I0406 08:28:07.081100 5226 solver.cpp:237] Train net output #0: loss = 1.4019 (* 1 = 1.4019 loss) I0406 08:28:07.081110 5226 sgd_solver.cpp:105] Iteration 9252, lr = 0.01 I0406 08:28:12.546842 5226 solver.cpp:218] Iteration 9264 (2.19551 iter/s, 5.46569s/12 iters), loss = 1.83513 I0406 08:28:12.546890 5226 solver.cpp:237] Train net output #0: loss = 1.83513 (* 1 = 1.83513 loss) I0406 08:28:12.546897 5226 sgd_solver.cpp:105] Iteration 9264, lr = 0.01 I0406 08:28:17.914494 5226 solver.cpp:218] Iteration 9276 (2.23566 iter/s, 5.36755s/12 iters), loss = 1.45986 I0406 08:28:17.914536 5226 solver.cpp:237] Train net output #0: loss = 1.45986 (* 1 = 1.45986 loss) I0406 08:28:17.914542 5226 sgd_solver.cpp:105] Iteration 9276, lr = 0.01 I0406 08:28:20.012820 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9282.caffemodel I0406 08:28:23.039324 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9282.solverstate I0406 08:28:25.430354 5226 solver.cpp:330] Iteration 9282, Testing net (#0) I0406 08:28:25.430374 5226 net.cpp:676] Ignoring source layer train-data I0406 08:28:26.184340 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:28:29.806226 5226 solver.cpp:397] Test net output #0: accuracy = 0.301471 I0406 08:28:29.806262 5226 solver.cpp:397] Test net output #1: loss = 3.29412 (* 1 = 3.29412 loss) I0406 08:28:31.821305 5226 solver.cpp:218] Iteration 9288 (0.862896 iter/s, 13.9067s/12 iters), loss = 1.31509 I0406 08:28:31.821344 5226 solver.cpp:237] Train net output #0: loss = 1.31509 (* 1 = 1.31509 loss) I0406 08:28:31.821349 5226 sgd_solver.cpp:105] Iteration 9288, lr = 0.01 I0406 08:28:36.970638 5226 solver.cpp:218] Iteration 9300 (2.33044 iter/s, 5.14924s/12 iters), loss = 1.36213 I0406 08:28:36.970691 5226 solver.cpp:237] Train net output #0: loss = 1.36213 (* 1 = 1.36213 loss) I0406 08:28:36.970700 5226 sgd_solver.cpp:105] Iteration 9300, lr = 0.01 I0406 08:28:39.367906 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:28:42.149413 5226 solver.cpp:218] Iteration 9312 (2.3172 iter/s, 5.17867s/12 iters), loss = 1.16864 I0406 08:28:42.149452 5226 solver.cpp:237] Train net output #0: loss = 1.16864 (* 1 = 1.16864 loss) I0406 08:28:42.149459 5226 sgd_solver.cpp:105] Iteration 9312, lr = 0.01 I0406 08:28:47.328627 5226 solver.cpp:218] Iteration 9324 (2.31699 iter/s, 5.17912s/12 iters), loss = 1.48885 I0406 08:28:47.328673 5226 solver.cpp:237] Train net output #0: loss = 1.48885 (* 1 = 1.48885 loss) I0406 08:28:47.328681 5226 sgd_solver.cpp:105] Iteration 9324, lr = 0.01 I0406 08:28:52.523437 5226 solver.cpp:218] Iteration 9336 (2.31004 iter/s, 5.19472s/12 iters), loss = 1.22772 I0406 08:28:52.523484 5226 solver.cpp:237] Train net output #0: loss = 1.22772 (* 1 = 1.22772 loss) I0406 08:28:52.523491 5226 sgd_solver.cpp:105] Iteration 9336, lr = 0.01 I0406 08:28:57.718976 5226 solver.cpp:218] Iteration 9348 (2.30972 iter/s, 5.19544s/12 iters), loss = 1.28443 I0406 08:28:57.719019 5226 solver.cpp:237] Train net output #0: loss = 1.28443 (* 1 = 1.28443 loss) I0406 08:28:57.719025 5226 sgd_solver.cpp:105] Iteration 9348, lr = 0.01 I0406 08:29:03.153055 5226 solver.cpp:218] Iteration 9360 (2.20832 iter/s, 5.43398s/12 iters), loss = 1.51051 I0406 08:29:03.153106 5226 solver.cpp:237] Train net output #0: loss = 1.51051 (* 1 = 1.51051 loss) I0406 08:29:03.153115 5226 sgd_solver.cpp:105] Iteration 9360, lr = 0.01 I0406 08:29:08.455766 5226 solver.cpp:218] Iteration 9372 (2.26304 iter/s, 5.30261s/12 iters), loss = 1.55222 I0406 08:29:08.455814 5226 solver.cpp:237] Train net output #0: loss = 1.55222 (* 1 = 1.55222 loss) I0406 08:29:08.455822 5226 sgd_solver.cpp:105] Iteration 9372, lr = 0.01 I0406 08:29:13.133136 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9384.caffemodel I0406 08:29:16.171054 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9384.solverstate I0406 08:29:18.481362 5226 solver.cpp:330] Iteration 9384, Testing net (#0) I0406 08:29:18.481381 5226 net.cpp:676] Ignoring source layer train-data I0406 08:29:19.179045 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:29:22.868168 5226 solver.cpp:397] Test net output #0: accuracy = 0.286152 I0406 08:29:22.868202 5226 solver.cpp:397] Test net output #1: loss = 3.39326 (* 1 = 3.39326 loss) I0406 08:29:23.009371 5226 solver.cpp:218] Iteration 9384 (0.824547 iter/s, 14.5535s/12 iters), loss = 1.73139 I0406 08:29:23.009418 5226 solver.cpp:237] Train net output #0: loss = 1.73139 (* 1 = 1.73139 loss) I0406 08:29:23.009424 5226 sgd_solver.cpp:105] Iteration 9384, lr = 0.01 I0406 08:29:27.324363 5226 solver.cpp:218] Iteration 9396 (2.78106 iter/s, 4.3149s/12 iters), loss = 1.73471 I0406 08:29:27.324400 5226 solver.cpp:237] Train net output #0: loss = 1.73471 (* 1 = 1.73471 loss) I0406 08:29:27.324405 5226 sgd_solver.cpp:105] Iteration 9396, lr = 0.01 I0406 08:29:31.807945 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:29:32.553489 5226 solver.cpp:218] Iteration 9408 (2.29488 iter/s, 5.22904s/12 iters), loss = 1.55081 I0406 08:29:32.553529 5226 solver.cpp:237] Train net output #0: loss = 1.55081 (* 1 = 1.55081 loss) I0406 08:29:32.553535 5226 sgd_solver.cpp:105] Iteration 9408, lr = 0.01 I0406 08:29:37.864786 5226 solver.cpp:218] Iteration 9420 (2.25938 iter/s, 5.3112s/12 iters), loss = 1.48133 I0406 08:29:37.864835 5226 solver.cpp:237] Train net output #0: loss = 1.48133 (* 1 = 1.48133 loss) I0406 08:29:37.864842 5226 sgd_solver.cpp:105] Iteration 9420, lr = 0.01 I0406 08:29:43.224578 5226 solver.cpp:218] Iteration 9432 (2.23893 iter/s, 5.35969s/12 iters), loss = 0.953196 I0406 08:29:43.224678 5226 solver.cpp:237] Train net output #0: loss = 0.953196 (* 1 = 0.953196 loss) I0406 08:29:43.224686 5226 sgd_solver.cpp:105] Iteration 9432, lr = 0.01 I0406 08:29:48.270864 5226 solver.cpp:218] Iteration 9444 (2.37805 iter/s, 5.04614s/12 iters), loss = 1.42891 I0406 08:29:48.270901 5226 solver.cpp:237] Train net output #0: loss = 1.42891 (* 1 = 1.42891 loss) I0406 08:29:48.270906 5226 sgd_solver.cpp:105] Iteration 9444, lr = 0.01 I0406 08:29:53.638362 5226 solver.cpp:218] Iteration 9456 (2.23572 iter/s, 5.3674s/12 iters), loss = 1.31815 I0406 08:29:53.638417 5226 solver.cpp:237] Train net output #0: loss = 1.31815 (* 1 = 1.31815 loss) I0406 08:29:53.638423 5226 sgd_solver.cpp:105] Iteration 9456, lr = 0.01 I0406 08:29:58.813033 5226 solver.cpp:218] Iteration 9468 (2.31903 iter/s, 5.17457s/12 iters), loss = 1.34019 I0406 08:29:58.813071 5226 solver.cpp:237] Train net output #0: loss = 1.34019 (* 1 = 1.34019 loss) I0406 08:29:58.813076 5226 sgd_solver.cpp:105] Iteration 9468, lr = 0.01 I0406 08:30:03.873073 5226 solver.cpp:218] Iteration 9480 (2.37157 iter/s, 5.05995s/12 iters), loss = 1.8316 I0406 08:30:03.873124 5226 solver.cpp:237] Train net output #0: loss = 1.8316 (* 1 = 1.8316 loss) I0406 08:30:03.873132 5226 sgd_solver.cpp:105] Iteration 9480, lr = 0.01 I0406 08:30:06.080991 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9486.caffemodel I0406 08:30:09.298688 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9486.solverstate I0406 08:30:11.599360 5226 solver.cpp:330] Iteration 9486, Testing net (#0) I0406 08:30:11.599380 5226 net.cpp:676] Ignoring source layer train-data I0406 08:30:12.215423 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:30:15.869436 5226 solver.cpp:397] Test net output #0: accuracy = 0.270221 I0406 08:30:15.869556 5226 solver.cpp:397] Test net output #1: loss = 3.41308 (* 1 = 3.41308 loss) I0406 08:30:17.733644 5226 solver.cpp:218] Iteration 9492 (0.865775 iter/s, 13.8604s/12 iters), loss = 1.60063 I0406 08:30:17.733700 5226 solver.cpp:237] Train net output #0: loss = 1.60063 (* 1 = 1.60063 loss) I0406 08:30:17.733708 5226 sgd_solver.cpp:105] Iteration 9492, lr = 0.01 I0406 08:30:23.115406 5226 solver.cpp:218] Iteration 9504 (2.22979 iter/s, 5.38166s/12 iters), loss = 1.48874 I0406 08:30:23.115442 5226 solver.cpp:237] Train net output #0: loss = 1.48874 (* 1 = 1.48874 loss) I0406 08:30:23.115447 5226 sgd_solver.cpp:105] Iteration 9504, lr = 0.01 I0406 08:30:24.630280 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:30:28.336930 5226 solver.cpp:218] Iteration 9516 (2.29822 iter/s, 5.22143s/12 iters), loss = 1.67417 I0406 08:30:28.336979 5226 solver.cpp:237] Train net output #0: loss = 1.67417 (* 1 = 1.67417 loss) I0406 08:30:28.336987 5226 sgd_solver.cpp:105] Iteration 9516, lr = 0.01 I0406 08:30:33.732908 5226 solver.cpp:218] Iteration 9528 (2.22392 iter/s, 5.39587s/12 iters), loss = 1.43106 I0406 08:30:33.732962 5226 solver.cpp:237] Train net output #0: loss = 1.43106 (* 1 = 1.43106 loss) I0406 08:30:33.732971 5226 sgd_solver.cpp:105] Iteration 9528, lr = 0.01 I0406 08:30:38.984933 5226 solver.cpp:218] Iteration 9540 (2.28488 iter/s, 5.25192s/12 iters), loss = 1.24924 I0406 08:30:38.984992 5226 solver.cpp:237] Train net output #0: loss = 1.24924 (* 1 = 1.24924 loss) I0406 08:30:38.985002 5226 sgd_solver.cpp:105] Iteration 9540, lr = 0.01 I0406 08:30:44.345671 5226 solver.cpp:218] Iteration 9552 (2.23854 iter/s, 5.36063s/12 iters), loss = 1.43932 I0406 08:30:44.345710 5226 solver.cpp:237] Train net output #0: loss = 1.43932 (* 1 = 1.43932 loss) I0406 08:30:44.345715 5226 sgd_solver.cpp:105] Iteration 9552, lr = 0.01 I0406 08:30:49.688896 5226 solver.cpp:218] Iteration 9564 (2.24587 iter/s, 5.34313s/12 iters), loss = 1.66909 I0406 08:30:49.688990 5226 solver.cpp:237] Train net output #0: loss = 1.66909 (* 1 = 1.66909 loss) I0406 08:30:49.688997 5226 sgd_solver.cpp:105] Iteration 9564, lr = 0.01 I0406 08:30:54.941321 5226 solver.cpp:218] Iteration 9576 (2.28472 iter/s, 5.25228s/12 iters), loss = 1.59738 I0406 08:30:54.941373 5226 solver.cpp:237] Train net output #0: loss = 1.59738 (* 1 = 1.59738 loss) I0406 08:30:54.941381 5226 sgd_solver.cpp:105] Iteration 9576, lr = 0.01 I0406 08:30:59.671604 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9588.caffemodel I0406 08:31:02.695439 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9588.solverstate I0406 08:31:05.054455 5226 solver.cpp:330] Iteration 9588, Testing net (#0) I0406 08:31:05.054474 5226 net.cpp:676] Ignoring source layer train-data I0406 08:31:05.720475 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:31:09.432772 5226 solver.cpp:397] Test net output #0: accuracy = 0.276348 I0406 08:31:09.432806 5226 solver.cpp:397] Test net output #1: loss = 3.42001 (* 1 = 3.42001 loss) I0406 08:31:09.573529 5226 solver.cpp:218] Iteration 9588 (0.820118 iter/s, 14.632s/12 iters), loss = 1.895 I0406 08:31:09.575094 5226 solver.cpp:237] Train net output #0: loss = 1.895 (* 1 = 1.895 loss) I0406 08:31:09.575109 5226 sgd_solver.cpp:105] Iteration 9588, lr = 0.01 I0406 08:31:13.813021 5226 solver.cpp:218] Iteration 9600 (2.8316 iter/s, 4.23789s/12 iters), loss = 1.58854 I0406 08:31:13.813076 5226 solver.cpp:237] Train net output #0: loss = 1.58854 (* 1 = 1.58854 loss) I0406 08:31:13.813084 5226 sgd_solver.cpp:105] Iteration 9600, lr = 0.01 I0406 08:31:17.642505 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:31:19.181569 5226 solver.cpp:218] Iteration 9612 (2.23528 iter/s, 5.36844s/12 iters), loss = 1.91278 I0406 08:31:19.181607 5226 solver.cpp:237] Train net output #0: loss = 1.91278 (* 1 = 1.91278 loss) I0406 08:31:19.181613 5226 sgd_solver.cpp:105] Iteration 9612, lr = 0.01 I0406 08:31:24.401324 5226 solver.cpp:218] Iteration 9624 (2.299 iter/s, 5.21967s/12 iters), loss = 1.60281 I0406 08:31:24.401437 5226 solver.cpp:237] Train net output #0: loss = 1.60281 (* 1 = 1.60281 loss) I0406 08:31:24.401444 5226 sgd_solver.cpp:105] Iteration 9624, lr = 0.01 I0406 08:31:29.839982 5226 solver.cpp:218] Iteration 9636 (2.20649 iter/s, 5.43849s/12 iters), loss = 1.48788 I0406 08:31:29.840021 5226 solver.cpp:237] Train net output #0: loss = 1.48788 (* 1 = 1.48788 loss) I0406 08:31:29.840029 5226 sgd_solver.cpp:105] Iteration 9636, lr = 0.01 I0406 08:31:35.069733 5226 solver.cpp:218] Iteration 9648 (2.2946 iter/s, 5.22966s/12 iters), loss = 1.39629 I0406 08:31:35.069773 5226 solver.cpp:237] Train net output #0: loss = 1.39629 (* 1 = 1.39629 loss) I0406 08:31:35.069778 5226 sgd_solver.cpp:105] Iteration 9648, lr = 0.01 I0406 08:31:40.405241 5226 solver.cpp:218] Iteration 9660 (2.24912 iter/s, 5.33541s/12 iters), loss = 1.48703 I0406 08:31:40.405300 5226 solver.cpp:237] Train net output #0: loss = 1.48703 (* 1 = 1.48703 loss) I0406 08:31:40.405310 5226 sgd_solver.cpp:105] Iteration 9660, lr = 0.01 I0406 08:31:45.568725 5226 solver.cpp:218] Iteration 9672 (2.32406 iter/s, 5.16337s/12 iters), loss = 1.14586 I0406 08:31:45.568773 5226 solver.cpp:237] Train net output #0: loss = 1.14586 (* 1 = 1.14586 loss) I0406 08:31:45.568781 5226 sgd_solver.cpp:105] Iteration 9672, lr = 0.01 I0406 08:31:50.872825 5226 solver.cpp:218] Iteration 9684 (2.26244 iter/s, 5.304s/12 iters), loss = 1.49319 I0406 08:31:50.872871 5226 solver.cpp:237] Train net output #0: loss = 1.49319 (* 1 = 1.49319 loss) I0406 08:31:50.872879 5226 sgd_solver.cpp:105] Iteration 9684, lr = 0.01 I0406 08:31:52.969470 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9690.caffemodel I0406 08:31:56.136780 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9690.solverstate I0406 08:31:58.432479 5226 solver.cpp:330] Iteration 9690, Testing net (#0) I0406 08:31:58.432549 5226 net.cpp:676] Ignoring source layer train-data I0406 08:31:59.059414 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:32:01.822795 5226 blocking_queue.cpp:49] Waiting for data I0406 08:32:02.801328 5226 solver.cpp:397] Test net output #0: accuracy = 0.275735 I0406 08:32:02.801362 5226 solver.cpp:397] Test net output #1: loss = 3.38087 (* 1 = 3.38087 loss) I0406 08:32:04.627833 5226 solver.cpp:218] Iteration 9696 (0.872419 iter/s, 13.7549s/12 iters), loss = 1.349 I0406 08:32:04.627876 5226 solver.cpp:237] Train net output #0: loss = 1.349 (* 1 = 1.349 loss) I0406 08:32:04.627882 5226 sgd_solver.cpp:105] Iteration 9696, lr = 0.01 I0406 08:32:09.903945 5226 solver.cpp:218] Iteration 9708 (2.27444 iter/s, 5.27602s/12 iters), loss = 1.55017 I0406 08:32:09.903993 5226 solver.cpp:237] Train net output #0: loss = 1.55017 (* 1 = 1.55017 loss) I0406 08:32:09.904001 5226 sgd_solver.cpp:105] Iteration 9708, lr = 0.01 I0406 08:32:10.642725 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:32:15.227866 5226 solver.cpp:218] Iteration 9720 (2.25402 iter/s, 5.32382s/12 iters), loss = 1.75784 I0406 08:32:15.227916 5226 solver.cpp:237] Train net output #0: loss = 1.75784 (* 1 = 1.75784 loss) I0406 08:32:15.227924 5226 sgd_solver.cpp:105] Iteration 9720, lr = 0.01 I0406 08:32:20.313910 5226 solver.cpp:218] Iteration 9732 (2.35944 iter/s, 5.08595s/12 iters), loss = 2.01695 I0406 08:32:20.313946 5226 solver.cpp:237] Train net output #0: loss = 2.01695 (* 1 = 2.01695 loss) I0406 08:32:20.313952 5226 sgd_solver.cpp:105] Iteration 9732, lr = 0.01 I0406 08:32:25.612366 5226 solver.cpp:218] Iteration 9744 (2.26485 iter/s, 5.29837s/12 iters), loss = 1.6915 I0406 08:32:25.612427 5226 solver.cpp:237] Train net output #0: loss = 1.6915 (* 1 = 1.6915 loss) I0406 08:32:25.612435 5226 sgd_solver.cpp:105] Iteration 9744, lr = 0.01 I0406 08:32:30.965961 5226 solver.cpp:218] Iteration 9756 (2.24153 iter/s, 5.35349s/12 iters), loss = 1.7613 I0406 08:32:30.966101 5226 solver.cpp:237] Train net output #0: loss = 1.7613 (* 1 = 1.7613 loss) I0406 08:32:30.966109 5226 sgd_solver.cpp:105] Iteration 9756, lr = 0.01 I0406 08:32:36.327805 5226 solver.cpp:218] Iteration 9768 (2.23811 iter/s, 5.36166s/12 iters), loss = 1.23214 I0406 08:32:36.327841 5226 solver.cpp:237] Train net output #0: loss = 1.23214 (* 1 = 1.23214 loss) I0406 08:32:36.327847 5226 sgd_solver.cpp:105] Iteration 9768, lr = 0.01 I0406 08:32:41.593578 5226 solver.cpp:218] Iteration 9780 (2.27891 iter/s, 5.26568s/12 iters), loss = 1.73946 I0406 08:32:41.593628 5226 solver.cpp:237] Train net output #0: loss = 1.73946 (* 1 = 1.73946 loss) I0406 08:32:41.593636 5226 sgd_solver.cpp:105] Iteration 9780, lr = 0.01 I0406 08:32:46.312091 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9792.caffemodel I0406 08:32:49.400135 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9792.solverstate I0406 08:32:51.721503 5226 solver.cpp:330] Iteration 9792, Testing net (#0) I0406 08:32:51.721521 5226 net.cpp:676] Ignoring source layer train-data I0406 08:32:52.277993 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:32:56.079967 5226 solver.cpp:397] Test net output #0: accuracy = 0.264093 I0406 08:32:56.080003 5226 solver.cpp:397] Test net output #1: loss = 3.50177 (* 1 = 3.50177 loss) I0406 08:32:56.220778 5226 solver.cpp:218] Iteration 9792 (0.820398 iter/s, 14.627s/12 iters), loss = 1.34782 I0406 08:32:56.220829 5226 solver.cpp:237] Train net output #0: loss = 1.34782 (* 1 = 1.34782 loss) I0406 08:32:56.220835 5226 sgd_solver.cpp:105] Iteration 9792, lr = 0.01 I0406 08:33:00.565987 5226 solver.cpp:218] Iteration 9804 (2.76172 iter/s, 4.34511s/12 iters), loss = 1.58499 I0406 08:33:00.566027 5226 solver.cpp:237] Train net output #0: loss = 1.58499 (* 1 = 1.58499 loss) I0406 08:33:00.566032 5226 sgd_solver.cpp:105] Iteration 9804, lr = 0.01 I0406 08:33:03.602797 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:33:05.811208 5226 solver.cpp:218] Iteration 9816 (2.28784 iter/s, 5.24513s/12 iters), loss = 1.88589 I0406 08:33:05.811262 5226 solver.cpp:237] Train net output #0: loss = 1.88589 (* 1 = 1.88589 loss) I0406 08:33:05.811272 5226 sgd_solver.cpp:105] Iteration 9816, lr = 0.01 I0406 08:33:11.164916 5226 solver.cpp:218] Iteration 9828 (2.24148 iter/s, 5.3536s/12 iters), loss = 1.70971 I0406 08:33:11.164968 5226 solver.cpp:237] Train net output #0: loss = 1.70971 (* 1 = 1.70971 loss) I0406 08:33:11.164976 5226 sgd_solver.cpp:105] Iteration 9828, lr = 0.01 I0406 08:33:16.474483 5226 solver.cpp:218] Iteration 9840 (2.26011 iter/s, 5.30947s/12 iters), loss = 1.56853 I0406 08:33:16.474527 5226 solver.cpp:237] Train net output #0: loss = 1.56853 (* 1 = 1.56853 loss) I0406 08:33:16.474534 5226 sgd_solver.cpp:105] Iteration 9840, lr = 0.01 I0406 08:33:21.834518 5226 solver.cpp:218] Iteration 9852 (2.23883 iter/s, 5.35994s/12 iters), loss = 1.5757 I0406 08:33:21.834558 5226 solver.cpp:237] Train net output #0: loss = 1.5757 (* 1 = 1.5757 loss) I0406 08:33:21.834563 5226 sgd_solver.cpp:105] Iteration 9852, lr = 0.01 I0406 08:33:26.956693 5226 solver.cpp:218] Iteration 9864 (2.34279 iter/s, 5.12209s/12 iters), loss = 1.86951 I0406 08:33:26.956749 5226 solver.cpp:237] Train net output #0: loss = 1.86951 (* 1 = 1.86951 loss) I0406 08:33:26.956758 5226 sgd_solver.cpp:105] Iteration 9864, lr = 0.01 I0406 08:33:32.317757 5226 solver.cpp:218] Iteration 9876 (2.2384 iter/s, 5.36096s/12 iters), loss = 1.49468 I0406 08:33:32.317795 5226 solver.cpp:237] Train net output #0: loss = 1.49468 (* 1 = 1.49468 loss) I0406 08:33:32.317801 5226 sgd_solver.cpp:105] Iteration 9876, lr = 0.01 I0406 08:33:37.657953 5226 solver.cpp:218] Iteration 9888 (2.24715 iter/s, 5.34011s/12 iters), loss = 1.41868 I0406 08:33:37.658116 5226 solver.cpp:237] Train net output #0: loss = 1.41868 (* 1 = 1.41868 loss) I0406 08:33:37.658126 5226 sgd_solver.cpp:105] Iteration 9888, lr = 0.01 I0406 08:33:39.751976 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9894.caffemodel I0406 08:33:42.752720 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9894.solverstate I0406 08:33:45.813772 5226 solver.cpp:330] Iteration 9894, Testing net (#0) I0406 08:33:45.813793 5226 net.cpp:676] Ignoring source layer train-data I0406 08:33:46.316042 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:33:50.134459 5226 solver.cpp:397] Test net output #0: accuracy = 0.273284 I0406 08:33:50.134510 5226 solver.cpp:397] Test net output #1: loss = 3.50507 (* 1 = 3.50507 loss) I0406 08:33:52.166311 5226 solver.cpp:218] Iteration 9900 (0.827124 iter/s, 14.5081s/12 iters), loss = 1.30334 I0406 08:33:52.166349 5226 solver.cpp:237] Train net output #0: loss = 1.30334 (* 1 = 1.30334 loss) I0406 08:33:52.166354 5226 sgd_solver.cpp:105] Iteration 9900, lr = 0.01 I0406 08:33:57.361596 5226 solver.cpp:218] Iteration 9912 (2.30983 iter/s, 5.19519s/12 iters), loss = 1.7553 I0406 08:33:57.361646 5226 solver.cpp:237] Train net output #0: loss = 1.7553 (* 1 = 1.7553 loss) I0406 08:33:57.361654 5226 sgd_solver.cpp:105] Iteration 9912, lr = 0.01 I0406 08:33:57.449906 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:34:02.791034 5226 solver.cpp:218] Iteration 9924 (2.21021 iter/s, 5.42934s/12 iters), loss = 1.54744 I0406 08:34:02.791072 5226 solver.cpp:237] Train net output #0: loss = 1.54744 (* 1 = 1.54744 loss) I0406 08:34:02.791079 5226 sgd_solver.cpp:105] Iteration 9924, lr = 0.01 I0406 08:34:08.035404 5226 solver.cpp:218] Iteration 9936 (2.28821 iter/s, 5.24428s/12 iters), loss = 1.45356 I0406 08:34:08.035495 5226 solver.cpp:237] Train net output #0: loss = 1.45356 (* 1 = 1.45356 loss) I0406 08:34:08.035501 5226 sgd_solver.cpp:105] Iteration 9936, lr = 0.01 I0406 08:34:13.312279 5226 solver.cpp:218] Iteration 9948 (2.27413 iter/s, 5.27674s/12 iters), loss = 1.64733 I0406 08:34:13.312317 5226 solver.cpp:237] Train net output #0: loss = 1.64733 (* 1 = 1.64733 loss) I0406 08:34:13.312323 5226 sgd_solver.cpp:105] Iteration 9948, lr = 0.01 I0406 08:34:18.680207 5226 solver.cpp:218] Iteration 9960 (2.23554 iter/s, 5.36783s/12 iters), loss = 1.78467 I0406 08:34:18.680254 5226 solver.cpp:237] Train net output #0: loss = 1.78467 (* 1 = 1.78467 loss) I0406 08:34:18.680263 5226 sgd_solver.cpp:105] Iteration 9960, lr = 0.01 I0406 08:34:23.975453 5226 solver.cpp:218] Iteration 9972 (2.26623 iter/s, 5.29515s/12 iters), loss = 1.94637 I0406 08:34:23.975509 5226 solver.cpp:237] Train net output #0: loss = 1.94637 (* 1 = 1.94637 loss) I0406 08:34:23.975518 5226 sgd_solver.cpp:105] Iteration 9972, lr = 0.01 I0406 08:34:29.131028 5226 solver.cpp:218] Iteration 9984 (2.32762 iter/s, 5.15548s/12 iters), loss = 1.94742 I0406 08:34:29.131072 5226 solver.cpp:237] Train net output #0: loss = 1.94742 (* 1 = 1.94742 loss) I0406 08:34:29.131076 5226 sgd_solver.cpp:105] Iteration 9984, lr = 0.01 I0406 08:34:33.926730 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9996.caffemodel I0406 08:34:37.073827 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9996.solverstate I0406 08:34:39.367050 5226 solver.cpp:330] Iteration 9996, Testing net (#0) I0406 08:34:39.367133 5226 net.cpp:676] Ignoring source layer train-data I0406 08:34:39.800668 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:34:43.639911 5226 solver.cpp:397] Test net output #0: accuracy = 0.260417 I0406 08:34:43.639955 5226 solver.cpp:397] Test net output #1: loss = 3.46497 (* 1 = 3.46497 loss) I0406 08:34:43.780814 5226 solver.cpp:218] Iteration 9996 (0.819133 iter/s, 14.6496s/12 iters), loss = 1.68447 I0406 08:34:43.780856 5226 solver.cpp:237] Train net output #0: loss = 1.68447 (* 1 = 1.68447 loss) I0406 08:34:43.780861 5226 sgd_solver.cpp:105] Iteration 9996, lr = 0.01 I0406 08:34:48.034957 5226 solver.cpp:218] Iteration 10008 (2.82084 iter/s, 4.25406s/12 iters), loss = 1.74634 I0406 08:34:48.035009 5226 solver.cpp:237] Train net output #0: loss = 1.74634 (* 1 = 1.74634 loss) I0406 08:34:48.035018 5226 sgd_solver.cpp:105] Iteration 10008, lr = 0.01 I0406 08:34:50.300285 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:34:53.279476 5226 solver.cpp:218] Iteration 10020 (2.28815 iter/s, 5.24441s/12 iters), loss = 1.62686 I0406 08:34:53.279531 5226 solver.cpp:237] Train net output #0: loss = 1.62686 (* 1 = 1.62686 loss) I0406 08:34:53.279539 5226 sgd_solver.cpp:105] Iteration 10020, lr = 0.01 I0406 08:34:58.671222 5226 solver.cpp:218] Iteration 10032 (2.22567 iter/s, 5.39164s/12 iters), loss = 1.57996 I0406 08:34:58.671262 5226 solver.cpp:237] Train net output #0: loss = 1.57996 (* 1 = 1.57996 loss) I0406 08:34:58.671267 5226 sgd_solver.cpp:105] Iteration 10032, lr = 0.01 I0406 08:35:04.085789 5226 solver.cpp:218] Iteration 10044 (2.21628 iter/s, 5.41448s/12 iters), loss = 1.86189 I0406 08:35:04.085840 5226 solver.cpp:237] Train net output #0: loss = 1.86189 (* 1 = 1.86189 loss) I0406 08:35:04.085850 5226 sgd_solver.cpp:105] Iteration 10044, lr = 0.01 I0406 08:35:09.488727 5226 solver.cpp:218] Iteration 10056 (2.22106 iter/s, 5.40284s/12 iters), loss = 1.87702 I0406 08:35:09.488847 5226 solver.cpp:237] Train net output #0: loss = 1.87702 (* 1 = 1.87702 loss) I0406 08:35:09.488859 5226 sgd_solver.cpp:105] Iteration 10056, lr = 0.01 I0406 08:35:14.887974 5226 solver.cpp:218] Iteration 10068 (2.2226 iter/s, 5.39909s/12 iters), loss = 1.53069 I0406 08:35:14.888012 5226 solver.cpp:237] Train net output #0: loss = 1.53069 (* 1 = 1.53069 loss) I0406 08:35:14.888017 5226 sgd_solver.cpp:105] Iteration 10068, lr = 0.01 I0406 08:35:20.271958 5226 solver.cpp:218] Iteration 10080 (2.22887 iter/s, 5.3839s/12 iters), loss = 1.6349 I0406 08:35:20.271993 5226 solver.cpp:237] Train net output #0: loss = 1.6349 (* 1 = 1.6349 loss) I0406 08:35:20.271999 5226 sgd_solver.cpp:105] Iteration 10080, lr = 0.01 I0406 08:35:25.413131 5226 solver.cpp:218] Iteration 10092 (2.33414 iter/s, 5.14108s/12 iters), loss = 1.93234 I0406 08:35:25.413184 5226 solver.cpp:237] Train net output #0: loss = 1.93234 (* 1 = 1.93234 loss) I0406 08:35:25.413193 5226 sgd_solver.cpp:105] Iteration 10092, lr = 0.01 I0406 08:35:27.548415 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_10098.caffemodel I0406 08:35:30.562490 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_10098.solverstate I0406 08:35:32.862757 5226 solver.cpp:330] Iteration 10098, Testing net (#0) I0406 08:35:32.862776 5226 net.cpp:676] Ignoring source layer train-data I0406 08:35:33.317839 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:35:37.332799 5226 solver.cpp:397] Test net output #0: accuracy = 0.268382 I0406 08:35:37.332841 5226 solver.cpp:397] Test net output #1: loss = 3.44057 (* 1 = 3.44057 loss) I0406 08:35:39.297561 5226 solver.cpp:218] Iteration 10104 (0.864287 iter/s, 13.8843s/12 iters), loss = 1.70096 I0406 08:35:39.297621 5226 solver.cpp:237] Train net output #0: loss = 1.70096 (* 1 = 1.70096 loss) I0406 08:35:39.297631 5226 sgd_solver.cpp:105] Iteration 10104, lr = 0.01 I0406 08:35:43.851413 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:35:44.583045 5226 solver.cpp:218] Iteration 10116 (2.27041 iter/s, 5.28538s/12 iters), loss = 1.69603 I0406 08:35:44.583082 5226 solver.cpp:237] Train net output #0: loss = 1.69603 (* 1 = 1.69603 loss) I0406 08:35:44.583087 5226 sgd_solver.cpp:105] Iteration 10116, lr = 0.01 I0406 08:35:49.797902 5226 solver.cpp:218] Iteration 10128 (2.30116 iter/s, 5.21477s/12 iters), loss = 1.49449 I0406 08:35:49.797945 5226 solver.cpp:237] Train net output #0: loss = 1.49449 (* 1 = 1.49449 loss) I0406 08:35:49.797950 5226 sgd_solver.cpp:105] Iteration 10128, lr = 0.01 I0406 08:35:54.929328 5226 solver.cpp:218] Iteration 10140 (2.33857 iter/s, 5.13134s/12 iters), loss = 1.64734 I0406 08:35:54.929385 5226 solver.cpp:237] Train net output #0: loss = 1.64734 (* 1 = 1.64734 loss) I0406 08:35:54.929394 5226 sgd_solver.cpp:105] Iteration 10140, lr = 0.01 I0406 08:36:00.394415 5226 solver.cpp:218] Iteration 10152 (2.1958 iter/s, 5.46498s/12 iters), loss = 1.75535 I0406 08:36:00.394462 5226 solver.cpp:237] Train net output #0: loss = 1.75535 (* 1 = 1.75535 loss) I0406 08:36:00.394470 5226 sgd_solver.cpp:105] Iteration 10152, lr = 0.01 I0406 08:36:05.424531 5226 solver.cpp:218] Iteration 10164 (2.38568 iter/s, 5.03002s/12 iters), loss = 1.24466 I0406 08:36:05.424578 5226 solver.cpp:237] Train net output #0: loss = 1.24466 (* 1 = 1.24466 loss) I0406 08:36:05.424587 5226 sgd_solver.cpp:105] Iteration 10164, lr = 0.01 I0406 08:36:10.652312 5226 solver.cpp:218] Iteration 10176 (2.29547 iter/s, 5.22768s/12 iters), loss = 1.6235 I0406 08:36:10.652350 5226 solver.cpp:237] Train net output #0: loss = 1.6235 (* 1 = 1.6235 loss) I0406 08:36:10.652356 5226 sgd_solver.cpp:105] Iteration 10176, lr = 0.01 I0406 08:36:15.961638 5226 solver.cpp:218] Iteration 10188 (2.26021 iter/s, 5.30923s/12 iters), loss = 1.82726 I0406 08:36:15.961767 5226 solver.cpp:237] Train net output #0: loss = 1.82726 (* 1 = 1.82726 loss) I0406 08:36:15.961777 5226 sgd_solver.cpp:105] Iteration 10188, lr = 0.01 I0406 08:36:20.776013 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_10200.caffemodel I0406 08:36:23.805336 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_10200.solverstate I0406 08:36:26.123360 5226 solver.cpp:330] Iteration 10200, Testing net (#0) I0406 08:36:26.123378 5226 net.cpp:676] Ignoring source layer train-data I0406 08:36:26.476485 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:36:30.430084 5226 solver.cpp:397] Test net output #0: accuracy = 0.270221 I0406 08:36:30.430120 5226 solver.cpp:397] Test net output #1: loss = 3.58745 (* 1 = 3.58745 loss) I0406 08:36:30.569666 5226 solver.cpp:218] Iteration 10200 (0.821479 iter/s, 14.6078s/12 iters), loss = 1.93206 I0406 08:36:30.569710 5226 solver.cpp:237] Train net output #0: loss = 1.93206 (* 1 = 1.93206 loss) I0406 08:36:30.569716 5226 sgd_solver.cpp:105] Iteration 10200, lr = 0.01 I0406 08:36:34.813839 5226 solver.cpp:218] Iteration 10212 (2.82746 iter/s, 4.24409s/12 iters), loss = 2.19496 I0406 08:36:34.813879 5226 solver.cpp:237] Train net output #0: loss = 2.19496 (* 1 = 2.19496 loss) I0406 08:36:34.813885 5226 sgd_solver.cpp:105] Iteration 10212, lr = 0.01 I0406 08:36:36.356267 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:36:39.954640 5226 solver.cpp:218] Iteration 10224 (2.33431 iter/s, 5.14071s/12 iters), loss = 1.41348 I0406 08:36:39.954684 5226 solver.cpp:237] Train net output #0: loss = 1.41348 (* 1 = 1.41348 loss) I0406 08:36:39.954690 5226 sgd_solver.cpp:105] Iteration 10224, lr = 0.01 I0406 08:36:45.002214 5226 solver.cpp:218] Iteration 10236 (2.37742 iter/s, 5.04748s/12 iters), loss = 1.49571 I0406 08:36:45.002250 5226 solver.cpp:237] Train net output #0: loss = 1.49571 (* 1 = 1.49571 loss) I0406 08:36:45.002256 5226 sgd_solver.cpp:105] Iteration 10236, lr = 0.01 I0406 08:36:50.467960 5226 solver.cpp:218] Iteration 10248 (2.19553 iter/s, 5.46566s/12 iters), loss = 1.40195 I0406 08:36:50.468083 5226 solver.cpp:237] Train net output #0: loss = 1.40195 (* 1 = 1.40195 loss) I0406 08:36:50.468091 5226 sgd_solver.cpp:105] Iteration 10248, lr = 0.01 I0406 08:36:55.790956 5226 solver.cpp:218] Iteration 10260 (2.25444 iter/s, 5.32282s/12 iters), loss = 1.14686 I0406 08:36:55.790992 5226 solver.cpp:237] Train net output #0: loss = 1.14686 (* 1 = 1.14686 loss) I0406 08:36:55.790998 5226 sgd_solver.cpp:105] Iteration 10260, lr = 0.01 I0406 08:37:00.900602 5226 solver.cpp:218] Iteration 10272 (2.34854 iter/s, 5.10956s/12 iters), loss = 1.57964 I0406 08:37:00.900640 5226 solver.cpp:237] Train net output #0: loss = 1.57964 (* 1 = 1.57964 loss) I0406 08:37:00.900645 5226 sgd_solver.cpp:105] Iteration 10272, lr = 0.01 I0406 08:37:06.268011 5226 solver.cpp:218] Iteration 10284 (2.23575 iter/s, 5.36732s/12 iters), loss = 1.40208 I0406 08:37:06.268059 5226 solver.cpp:237] Train net output #0: loss = 1.40208 (* 1 = 1.40208 loss) I0406 08:37:06.268067 5226 sgd_solver.cpp:105] Iteration 10284, lr = 0.01 I0406 08:37:11.390699 5226 solver.cpp:218] Iteration 10296 (2.34256 iter/s, 5.1226s/12 iters), loss = 1.60574 I0406 08:37:11.390738 5226 solver.cpp:237] Train net output #0: loss = 1.60574 (* 1 = 1.60574 loss) I0406 08:37:11.390743 5226 sgd_solver.cpp:105] Iteration 10296, lr = 0.01 I0406 08:37:13.649708 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_10302.caffemodel I0406 08:37:16.615386 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_10302.solverstate I0406 08:37:18.906425 5226 solver.cpp:330] Iteration 10302, Testing net (#0) I0406 08:37:18.906445 5226 net.cpp:676] Ignoring source layer train-data I0406 08:37:19.243513 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:37:23.222255 5226 solver.cpp:397] Test net output #0: accuracy = 0.261642 I0406 08:37:23.222365 5226 solver.cpp:397] Test net output #1: loss = 3.59857 (* 1 = 3.59857 loss) I0406 08:37:25.137217 5226 solver.cpp:218] Iteration 10308 (0.872957 iter/s, 13.7464s/12 iters), loss = 1.85736 I0406 08:37:25.137271 5226 solver.cpp:237] Train net output #0: loss = 1.85736 (* 1 = 1.85736 loss) I0406 08:37:25.137279 5226 sgd_solver.cpp:105] Iteration 10308, lr = 0.01 I0406 08:37:28.938298 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:37:30.399622 5226 solver.cpp:218] Iteration 10320 (2.28037 iter/s, 5.2623s/12 iters), loss = 1.88133 I0406 08:37:30.399667 5226 solver.cpp:237] Train net output #0: loss = 1.88133 (* 1 = 1.88133 loss) I0406 08:37:30.399672 5226 sgd_solver.cpp:105] Iteration 10320, lr = 0.01 I0406 08:37:35.725788 5226 solver.cpp:218] Iteration 10332 (2.25307 iter/s, 5.32607s/12 iters), loss = 1.63204 I0406 08:37:35.725832 5226 solver.cpp:237] Train net output #0: loss = 1.63204 (* 1 = 1.63204 loss) I0406 08:37:35.725841 5226 sgd_solver.cpp:105] Iteration 10332, lr = 0.01 I0406 08:37:40.949409 5226 solver.cpp:218] Iteration 10344 (2.2973 iter/s, 5.22353s/12 iters), loss = 1.3453 I0406 08:37:40.949448 5226 solver.cpp:237] Train net output #0: loss = 1.3453 (* 1 = 1.3453 loss) I0406 08:37:40.949453 5226 sgd_solver.cpp:105] Iteration 10344, lr = 0.01 I0406 08:37:46.076819 5226 solver.cpp:218] Iteration 10356 (2.3404 iter/s, 5.12733s/12 iters), loss = 1.53305 I0406 08:37:46.076866 5226 solver.cpp:237] Train net output #0: loss = 1.53305 (* 1 = 1.53305 loss) I0406 08:37:46.076874 5226 sgd_solver.cpp:105] Iteration 10356, lr = 0.01 I0406 08:37:51.391896 5226 solver.cpp:218] Iteration 10368 (2.25777 iter/s, 5.31499s/12 iters), loss = 1.70293 I0406 08:37:51.391933 5226 solver.cpp:237] Train net output #0: loss = 1.70293 (* 1 = 1.70293 loss) I0406 08:37:51.391938 5226 sgd_solver.cpp:105] Iteration 10368, lr = 0.01 I0406 08:37:56.631486 5226 solver.cpp:218] Iteration 10380 (2.29029 iter/s, 5.23951s/12 iters), loss = 1.74211 I0406 08:37:56.631611 5226 solver.cpp:237] Train net output #0: loss = 1.74211 (* 1 = 1.74211 loss) I0406 08:37:56.631618 5226 sgd_solver.cpp:105] Iteration 10380, lr = 0.01 I0406 08:38:01.971520 5226 solver.cpp:218] Iteration 10392 (2.24725 iter/s, 5.33986s/12 iters), loss = 1.65588 I0406 08:38:01.971561 5226 solver.cpp:237] Train net output #0: loss = 1.65588 (* 1 = 1.65588 loss) I0406 08:38:01.971566 5226 sgd_solver.cpp:105] Iteration 10392, lr = 0.01 I0406 08:38:06.758473 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_10404.caffemodel I0406 08:38:09.749413 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_10404.solverstate I0406 08:38:12.047278 5226 solver.cpp:330] Iteration 10404, Testing net (#0) I0406 08:38:12.047297 5226 net.cpp:676] Ignoring source layer train-data I0406 08:38:12.317400 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:38:12.785883 5226 blocking_queue.cpp:49] Waiting for data I0406 08:38:16.482206 5226 solver.cpp:397] Test net output #0: accuracy = 0.255515 I0406 08:38:16.482240 5226 solver.cpp:397] Test net output #1: loss = 3.54629 (* 1 = 3.54629 loss) I0406 08:38:16.621681 5226 solver.cpp:218] Iteration 10404 (0.819112 iter/s, 14.65s/12 iters), loss = 1.90986 I0406 08:38:16.621717 5226 solver.cpp:237] Train net output #0: loss = 1.90986 (* 1 = 1.90986 loss) I0406 08:38:16.621722 5226 sgd_solver.cpp:105] Iteration 10404, lr = 0.01 I0406 08:38:20.849551 5226 solver.cpp:218] Iteration 10416 (2.83836 iter/s, 4.22779s/12 iters), loss = 1.73078 I0406 08:38:20.849591 5226 solver.cpp:237] Train net output #0: loss = 1.73078 (* 1 = 1.73078 loss) I0406 08:38:20.849596 5226 sgd_solver.cpp:105] Iteration 10416, lr = 0.01 I0406 08:38:21.710225 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:38:25.670889 5226 solver.cpp:218] Iteration 10428 (2.48898 iter/s, 4.82126s/12 iters), loss = 1.58328 I0406 08:38:25.670934 5226 solver.cpp:237] Train net output #0: loss = 1.58328 (* 1 = 1.58328 loss) I0406 08:38:25.670943 5226 sgd_solver.cpp:105] Iteration 10428, lr = 0.01 I0406 08:38:30.948463 5226 solver.cpp:218] Iteration 10440 (2.27381 iter/s, 5.27748s/12 iters), loss = 1.66798 I0406 08:38:30.948541 5226 solver.cpp:237] Train net output #0: loss = 1.66798 (* 1 = 1.66798 loss) I0406 08:38:30.948547 5226 sgd_solver.cpp:105] Iteration 10440, lr = 0.01 I0406 08:38:36.333077 5226 solver.cpp:218] Iteration 10452 (2.22863 iter/s, 5.38448s/12 iters), loss = 2.03843 I0406 08:38:36.333137 5226 solver.cpp:237] Train net output #0: loss = 2.03843 (* 1 = 2.03843 loss) I0406 08:38:36.333148 5226 sgd_solver.cpp:105] Iteration 10452, lr = 0.01 I0406 08:38:41.628131 5226 solver.cpp:218] Iteration 10464 (2.26631 iter/s, 5.29495s/12 iters), loss = 1.72326 I0406 08:38:41.628178 5226 solver.cpp:237] Train net output #0: loss = 1.72326 (* 1 = 1.72326 loss) I0406 08:38:41.628186 5226 sgd_solver.cpp:105] Iteration 10464, lr = 0.01 I0406 08:38:47.037623 5226 solver.cpp:218] Iteration 10476 (2.21836 iter/s, 5.4094s/12 iters), loss = 1.83683 I0406 08:38:47.037667 5226 solver.cpp:237] Train net output #0: loss = 1.83683 (* 1 = 1.83683 loss) I0406 08:38:47.037674 5226 sgd_solver.cpp:105] Iteration 10476, lr = 0.01 I0406 08:38:52.382683 5226 solver.cpp:218] Iteration 10488 (2.2451 iter/s, 5.34496s/12 iters), loss = 1.4834 I0406 08:38:52.382733 5226 solver.cpp:237] Train net output #0: loss = 1.4834 (* 1 = 1.4834 loss) I0406 08:38:52.382740 5226 sgd_solver.cpp:105] Iteration 10488, lr = 0.01 I0406 08:38:57.787982 5226 solver.cpp:218] Iteration 10500 (2.22008 iter/s, 5.4052s/12 iters), loss = 1.48883 I0406 08:38:57.788039 5226 solver.cpp:237] Train net output #0: loss = 1.48883 (* 1 = 1.48883 loss) I0406 08:38:57.788048 5226 sgd_solver.cpp:105] Iteration 10500, lr = 0.01 I0406 08:39:00.011535 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_10506.caffemodel I0406 08:39:03.046933 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_10506.solverstate I0406 08:39:05.357987 5226 solver.cpp:330] Iteration 10506, Testing net (#0) I0406 08:39:05.358011 5226 net.cpp:676] Ignoring source layer train-data I0406 08:39:05.625165 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:39:09.882001 5226 solver.cpp:397] Test net output #0: accuracy = 0.254289 I0406 08:39:09.882036 5226 solver.cpp:397] Test net output #1: loss = 3.62608 (* 1 = 3.62608 loss) I0406 08:39:11.853427 5226 solver.cpp:218] Iteration 10512 (0.853164 iter/s, 14.0653s/12 iters), loss = 1.74583 I0406 08:39:11.853468 5226 solver.cpp:237] Train net output #0: loss = 1.74583 (* 1 = 1.74583 loss) I0406 08:39:11.853473 5226 sgd_solver.cpp:105] Iteration 10512, lr = 0.01 I0406 08:39:14.999733 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:39:17.143844 5226 solver.cpp:218] Iteration 10524 (2.26829 iter/s, 5.29033s/12 iters), loss = 1.81615 I0406 08:39:17.143884 5226 solver.cpp:237] Train net output #0: loss = 1.81615 (* 1 = 1.81615 loss) I0406 08:39:17.143889 5226 sgd_solver.cpp:105] Iteration 10524, lr = 0.01 I0406 08:39:22.617177 5226 solver.cpp:218] Iteration 10536 (2.19248 iter/s, 5.47324s/12 iters), loss = 1.72172 I0406 08:39:22.617221 5226 solver.cpp:237] Train net output #0: loss = 1.72172 (* 1 = 1.72172 loss) I0406 08:39:22.617229 5226 sgd_solver.cpp:105] Iteration 10536, lr = 0.01 I0406 08:39:27.872583 5226 solver.cpp:218] Iteration 10548 (2.2834 iter/s, 5.25532s/12 iters), loss = 1.74302 I0406 08:39:27.872620 5226 solver.cpp:237] Train net output #0: loss = 1.74302 (* 1 = 1.74302 loss) I0406 08:39:27.872625 5226 sgd_solver.cpp:105] Iteration 10548, lr = 0.01 I0406 08:39:33.100697 5226 solver.cpp:218] Iteration 10560 (2.29532 iter/s, 5.22802s/12 iters), loss = 1.79315 I0406 08:39:33.101461 5226 solver.cpp:237] Train net output #0: loss = 1.79315 (* 1 = 1.79315 loss) I0406 08:39:33.101473 5226 sgd_solver.cpp:105] Iteration 10560, lr = 0.01 I0406 08:39:38.445492 5226 solver.cpp:218] Iteration 10572 (2.24551 iter/s, 5.34399s/12 iters), loss = 2.16118 I0406 08:39:38.445546 5226 solver.cpp:237] Train net output #0: loss = 2.16118 (* 1 = 2.16118 loss) I0406 08:39:38.445555 5226 sgd_solver.cpp:105] Iteration 10572, lr = 0.01 I0406 08:39:43.749366 5226 solver.cpp:218] Iteration 10584 (2.26254 iter/s, 5.30378s/12 iters), loss = 1.95264 I0406 08:39:43.749403 5226 solver.cpp:237] Train net output #0: loss = 1.95264 (* 1 = 1.95264 loss) I0406 08:39:43.749408 5226 sgd_solver.cpp:105] Iteration 10584, lr = 0.01 I0406 08:39:49.129091 5226 solver.cpp:218] Iteration 10596 (2.23063 iter/s, 5.37964s/12 iters), loss = 2.51427 I0406 08:39:49.129139 5226 solver.cpp:237] Train net output #0: loss = 2.51427 (* 1 = 2.51427 loss) I0406 08:39:49.129148 5226 sgd_solver.cpp:105] Iteration 10596, lr = 0.01 I0406 08:39:53.912124 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_10608.caffemodel I0406 08:39:56.931996 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_10608.solverstate I0406 08:39:59.255475 5226 solver.cpp:330] Iteration 10608, Testing net (#0) I0406 08:39:59.255494 5226 net.cpp:676] Ignoring source layer train-data I0406 08:39:59.501508 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:40:03.556175 5226 solver.cpp:397] Test net output #0: accuracy = 0.270221 I0406 08:40:03.556327 5226 solver.cpp:397] Test net output #1: loss = 3.54479 (* 1 = 3.54479 loss) I0406 08:40:03.693523 5226 solver.cpp:218] Iteration 10608 (0.823933 iter/s, 14.5643s/12 iters), loss = 1.53705 I0406 08:40:03.693574 5226 solver.cpp:237] Train net output #0: loss = 1.53705 (* 1 = 1.53705 loss) I0406 08:40:03.693581 5226 sgd_solver.cpp:105] Iteration 10608, lr = 0.01 I0406 08:40:08.023671 5226 solver.cpp:218] Iteration 10620 (2.77133 iter/s, 4.33006s/12 iters), loss = 1.8692 I0406 08:40:08.023712 5226 solver.cpp:237] Train net output #0: loss = 1.8692 (* 1 = 1.8692 loss) I0406 08:40:08.023718 5226 sgd_solver.cpp:105] Iteration 10620, lr = 0.01 I0406 08:40:08.141373 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:40:13.297097 5226 solver.cpp:218] Iteration 10632 (2.2756 iter/s, 5.27334s/12 iters), loss = 1.73755 I0406 08:40:13.297135 5226 solver.cpp:237] Train net output #0: loss = 1.73755 (* 1 = 1.73755 loss) I0406 08:40:13.297142 5226 sgd_solver.cpp:105] Iteration 10632, lr = 0.01 I0406 08:40:18.450507 5226 solver.cpp:218] Iteration 10644 (2.3286 iter/s, 5.15332s/12 iters), loss = 1.62462 I0406 08:40:18.450557 5226 solver.cpp:237] Train net output #0: loss = 1.62462 (* 1 = 1.62462 loss) I0406 08:40:18.450564 5226 sgd_solver.cpp:105] Iteration 10644, lr = 0.01 I0406 08:40:23.664012 5226 solver.cpp:218] Iteration 10656 (2.30175 iter/s, 5.21341s/12 iters), loss = 1.54483 I0406 08:40:23.664055 5226 solver.cpp:237] Train net output #0: loss = 1.54483 (* 1 = 1.54483 loss) I0406 08:40:23.664063 5226 sgd_solver.cpp:105] Iteration 10656, lr = 0.01 I0406 08:40:28.982415 5226 solver.cpp:218] Iteration 10668 (2.25636 iter/s, 5.31831s/12 iters), loss = 1.7805 I0406 08:40:28.982462 5226 solver.cpp:237] Train net output #0: loss = 1.7805 (* 1 = 1.7805 loss) I0406 08:40:28.982470 5226 sgd_solver.cpp:105] Iteration 10668, lr = 0.01 I0406 08:40:34.328495 5226 solver.cpp:218] Iteration 10680 (2.24467 iter/s, 5.34599s/12 iters), loss = 1.70867 I0406 08:40:34.328591 5226 solver.cpp:237] Train net output #0: loss = 1.70867 (* 1 = 1.70867 loss) I0406 08:40:34.328598 5226 sgd_solver.cpp:105] Iteration 10680, lr = 0.01 I0406 08:40:39.805409 5226 solver.cpp:218] Iteration 10692 (2.19107 iter/s, 5.47677s/12 iters), loss = 1.9874 I0406 08:40:39.805449 5226 solver.cpp:237] Train net output #0: loss = 1.9874 (* 1 = 1.9874 loss) I0406 08:40:39.805454 5226 sgd_solver.cpp:105] Iteration 10692, lr = 0.01 I0406 08:40:44.997350 5226 solver.cpp:218] Iteration 10704 (2.31131 iter/s, 5.19185s/12 iters), loss = 1.60257 I0406 08:40:44.997406 5226 solver.cpp:237] Train net output #0: loss = 1.60257 (* 1 = 1.60257 loss) I0406 08:40:44.997416 5226 sgd_solver.cpp:105] Iteration 10704, lr = 0.01 I0406 08:40:47.177985 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_10710.caffemodel I0406 08:40:50.182122 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_10710.solverstate I0406 08:40:53.441059 5226 solver.cpp:330] Iteration 10710, Testing net (#0) I0406 08:40:53.441082 5226 net.cpp:676] Ignoring source layer train-data I0406 08:40:53.630146 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:40:58.103081 5226 solver.cpp:397] Test net output #0: accuracy = 0.230392 I0406 08:40:58.103122 5226 solver.cpp:397] Test net output #1: loss = 3.62848 (* 1 = 3.62848 loss) I0406 08:40:59.937433 5226 solver.cpp:218] Iteration 10716 (0.803217 iter/s, 14.9399s/12 iters), loss = 1.57672 I0406 08:40:59.937482 5226 solver.cpp:237] Train net output #0: loss = 1.57672 (* 1 = 1.57672 loss) I0406 08:40:59.937490 5226 sgd_solver.cpp:105] Iteration 10716, lr = 0.01 I0406 08:41:02.272807 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:41:05.108578 5226 solver.cpp:218] Iteration 10728 (2.32061 iter/s, 5.17105s/12 iters), loss = 1.84219 I0406 08:41:05.108713 5226 solver.cpp:237] Train net output #0: loss = 1.84219 (* 1 = 1.84219 loss) I0406 08:41:05.108723 5226 sgd_solver.cpp:105] Iteration 10728, lr = 0.01 I0406 08:41:10.398701 5226 solver.cpp:218] Iteration 10740 (2.26846 iter/s, 5.28994s/12 iters), loss = 1.62112 I0406 08:41:10.398741 5226 solver.cpp:237] Train net output #0: loss = 1.62112 (* 1 = 1.62112 loss) I0406 08:41:10.398746 5226 sgd_solver.cpp:105] Iteration 10740, lr = 0.01 I0406 08:41:15.464021 5226 solver.cpp:218] Iteration 10752 (2.36909 iter/s, 5.06523s/12 iters), loss = 1.58704 I0406 08:41:15.464063 5226 solver.cpp:237] Train net output #0: loss = 1.58704 (* 1 = 1.58704 loss) I0406 08:41:15.464069 5226 sgd_solver.cpp:105] Iteration 10752, lr = 0.01 I0406 08:41:20.793922 5226 solver.cpp:218] Iteration 10764 (2.25149 iter/s, 5.32981s/12 iters), loss = 1.18151 I0406 08:41:20.793962 5226 solver.cpp:237] Train net output #0: loss = 1.18151 (* 1 = 1.18151 loss) I0406 08:41:20.793967 5226 sgd_solver.cpp:105] Iteration 10764, lr = 0.01 I0406 08:41:26.183423 5226 solver.cpp:218] Iteration 10776 (2.22659 iter/s, 5.38941s/12 iters), loss = 1.65825 I0406 08:41:26.183481 5226 solver.cpp:237] Train net output #0: loss = 1.65825 (* 1 = 1.65825 loss) I0406 08:41:26.183490 5226 sgd_solver.cpp:105] Iteration 10776, lr = 0.01 I0406 08:41:31.258934 5226 solver.cpp:218] Iteration 10788 (2.36434 iter/s, 5.07541s/12 iters), loss = 1.19596 I0406 08:41:31.258982 5226 solver.cpp:237] Train net output #0: loss = 1.19596 (* 1 = 1.19596 loss) I0406 08:41:31.258992 5226 sgd_solver.cpp:105] Iteration 10788, lr = 0.01 I0406 08:41:36.438233 5226 solver.cpp:218] Iteration 10800 (2.31696 iter/s, 5.17921s/12 iters), loss = 2.1086 I0406 08:41:36.438359 5226 solver.cpp:237] Train net output #0: loss = 2.1086 (* 1 = 2.1086 loss) I0406 08:41:36.438366 5226 sgd_solver.cpp:105] Iteration 10800, lr = 0.01 I0406 08:41:41.298038 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_10812.caffemodel I0406 08:41:45.068032 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_10812.solverstate I0406 08:41:47.371978 5226 solver.cpp:330] Iteration 10812, Testing net (#0) I0406 08:41:47.371996 5226 net.cpp:676] Ignoring source layer train-data I0406 08:41:47.500787 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:41:51.732612 5226 solver.cpp:397] Test net output #0: accuracy = 0.257966 I0406 08:41:51.732650 5226 solver.cpp:397] Test net output #1: loss = 3.56116 (* 1 = 3.56116 loss) I0406 08:41:51.873582 5226 solver.cpp:218] Iteration 10812 (0.777448 iter/s, 15.4351s/12 iters), loss = 1.68942 I0406 08:41:51.873622 5226 solver.cpp:237] Train net output #0: loss = 1.68942 (* 1 = 1.68942 loss) I0406 08:41:51.873627 5226 sgd_solver.cpp:105] Iteration 10812, lr = 0.01 I0406 08:41:55.542743 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:41:56.219074 5226 solver.cpp:218] Iteration 10824 (2.76154 iter/s, 4.34541s/12 iters), loss = 2.08939 I0406 08:41:56.219115 5226 solver.cpp:237] Train net output #0: loss = 2.08939 (* 1 = 2.08939 loss) I0406 08:41:56.219120 5226 sgd_solver.cpp:105] Iteration 10824, lr = 0.01 I0406 08:42:01.497376 5226 solver.cpp:218] Iteration 10836 (2.2735 iter/s, 5.27821s/12 iters), loss = 2.09398 I0406 08:42:01.497418 5226 solver.cpp:237] Train net output #0: loss = 2.09398 (* 1 = 2.09398 loss) I0406 08:42:01.497423 5226 sgd_solver.cpp:105] Iteration 10836, lr = 0.01 I0406 08:42:06.766347 5226 solver.cpp:218] Iteration 10848 (2.27752 iter/s, 5.26888s/12 iters), loss = 1.72878 I0406 08:42:06.766434 5226 solver.cpp:237] Train net output #0: loss = 1.72878 (* 1 = 1.72878 loss) I0406 08:42:06.766441 5226 sgd_solver.cpp:105] Iteration 10848, lr = 0.01 I0406 08:42:12.049877 5226 solver.cpp:218] Iteration 10860 (2.27127 iter/s, 5.2834s/12 iters), loss = 1.93531 I0406 08:42:12.049913 5226 solver.cpp:237] Train net output #0: loss = 1.93531 (* 1 = 1.93531 loss) I0406 08:42:12.049919 5226 sgd_solver.cpp:105] Iteration 10860, lr = 0.01 I0406 08:42:16.892480 5226 solver.cpp:218] Iteration 10872 (2.47805 iter/s, 4.84252s/12 iters), loss = 1.53606 I0406 08:42:16.892524 5226 solver.cpp:237] Train net output #0: loss = 1.53606 (* 1 = 1.53606 loss) I0406 08:42:16.892531 5226 sgd_solver.cpp:105] Iteration 10872, lr = 0.01 I0406 08:42:22.101153 5226 solver.cpp:218] Iteration 10884 (2.30389 iter/s, 5.20858s/12 iters), loss = 1.62082 I0406 08:42:22.101191 5226 solver.cpp:237] Train net output #0: loss = 1.62082 (* 1 = 1.62082 loss) I0406 08:42:22.101197 5226 sgd_solver.cpp:105] Iteration 10884, lr = 0.01 I0406 08:42:27.305347 5226 solver.cpp:218] Iteration 10896 (2.30587 iter/s, 5.20411s/12 iters), loss = 2.1832 I0406 08:42:27.305385 5226 solver.cpp:237] Train net output #0: loss = 2.1832 (* 1 = 2.1832 loss) I0406 08:42:27.305390 5226 sgd_solver.cpp:105] Iteration 10896, lr = 0.01 I0406 08:42:32.669461 5226 solver.cpp:218] Iteration 10908 (2.23712 iter/s, 5.36403s/12 iters), loss = 1.96805 I0406 08:42:32.669504 5226 solver.cpp:237] Train net output #0: loss = 1.96805 (* 1 = 1.96805 loss) I0406 08:42:32.669509 5226 sgd_solver.cpp:105] Iteration 10908, lr = 0.01 I0406 08:42:34.717959 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_10914.caffemodel I0406 08:42:37.724551 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_10914.solverstate I0406 08:42:40.040124 5226 solver.cpp:330] Iteration 10914, Testing net (#0) I0406 08:42:40.040143 5226 net.cpp:676] Ignoring source layer train-data I0406 08:42:40.158378 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:42:44.537072 5226 solver.cpp:397] Test net output #0: accuracy = 0.240809 I0406 08:42:44.537106 5226 solver.cpp:397] Test net output #1: loss = 3.59737 (* 1 = 3.59737 loss) I0406 08:42:46.492769 5226 solver.cpp:218] Iteration 10920 (0.868108 iter/s, 13.8232s/12 iters), loss = 1.98199 I0406 08:42:46.492821 5226 solver.cpp:237] Train net output #0: loss = 1.98199 (* 1 = 1.98199 loss) I0406 08:42:46.492830 5226 sgd_solver.cpp:105] Iteration 10920, lr = 0.01 I0406 08:42:47.963548 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:42:51.677886 5226 solver.cpp:218] Iteration 10932 (2.31436 iter/s, 5.18502s/12 iters), loss = 1.66489 I0406 08:42:51.677933 5226 solver.cpp:237] Train net output #0: loss = 1.66489 (* 1 = 1.66489 loss) I0406 08:42:51.677940 5226 sgd_solver.cpp:105] Iteration 10932, lr = 0.01 I0406 08:42:57.063853 5226 solver.cpp:218] Iteration 10944 (2.22805 iter/s, 5.38587s/12 iters), loss = 1.75694 I0406 08:42:57.063907 5226 solver.cpp:237] Train net output #0: loss = 1.75694 (* 1 = 1.75694 loss) I0406 08:42:57.063915 5226 sgd_solver.cpp:105] Iteration 10944, lr = 0.01 I0406 08:43:02.306483 5226 solver.cpp:218] Iteration 10956 (2.28897 iter/s, 5.24253s/12 iters), loss = 1.41838 I0406 08:43:02.306535 5226 solver.cpp:237] Train net output #0: loss = 1.41838 (* 1 = 1.41838 loss) I0406 08:43:02.306543 5226 sgd_solver.cpp:105] Iteration 10956, lr = 0.01 I0406 08:43:07.500488 5226 solver.cpp:218] Iteration 10968 (2.3104 iter/s, 5.19391s/12 iters), loss = 1.98088 I0406 08:43:07.500526 5226 solver.cpp:237] Train net output #0: loss = 1.98088 (* 1 = 1.98088 loss) I0406 08:43:07.500531 5226 sgd_solver.cpp:105] Iteration 10968, lr = 0.01 I0406 08:43:12.795881 5226 solver.cpp:218] Iteration 10980 (2.26616 iter/s, 5.29531s/12 iters), loss = 1.6785 I0406 08:43:12.795967 5226 solver.cpp:237] Train net output #0: loss = 1.6785 (* 1 = 1.6785 loss) I0406 08:43:12.795974 5226 sgd_solver.cpp:105] Iteration 10980, lr = 0.01 I0406 08:43:17.961589 5226 solver.cpp:218] Iteration 10992 (2.32307 iter/s, 5.16558s/12 iters), loss = 2.04665 I0406 08:43:17.961625 5226 solver.cpp:237] Train net output #0: loss = 2.04665 (* 1 = 2.04665 loss) I0406 08:43:17.961632 5226 sgd_solver.cpp:105] Iteration 10992, lr = 0.01 I0406 08:43:23.511693 5226 solver.cpp:218] Iteration 11004 (2.16216 iter/s, 5.55002s/12 iters), loss = 1.69658 I0406 08:43:23.511732 5226 solver.cpp:237] Train net output #0: loss = 1.69658 (* 1 = 1.69658 loss) I0406 08:43:23.511737 5226 sgd_solver.cpp:105] Iteration 11004, lr = 0.01 I0406 08:43:28.172616 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_11016.caffemodel I0406 08:43:31.239485 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_11016.solverstate I0406 08:43:33.551215 5226 solver.cpp:330] Iteration 11016, Testing net (#0) I0406 08:43:33.551239 5226 net.cpp:676] Ignoring source layer train-data I0406 08:43:33.605902 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:43:37.941057 5226 solver.cpp:397] Test net output #0: accuracy = 0.262868 I0406 08:43:37.941087 5226 solver.cpp:397] Test net output #1: loss = 3.57288 (* 1 = 3.57288 loss) I0406 08:43:38.082005 5226 solver.cpp:218] Iteration 11016 (0.8236 iter/s, 14.5702s/12 iters), loss = 1.74997 I0406 08:43:38.082054 5226 solver.cpp:237] Train net output #0: loss = 1.74997 (* 1 = 1.74997 loss) I0406 08:43:38.082062 5226 sgd_solver.cpp:105] Iteration 11016, lr = 0.01 I0406 08:43:38.635910 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:43:41.148227 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:43:42.475380 5226 solver.cpp:218] Iteration 11028 (2.73144 iter/s, 4.39328s/12 iters), loss = 2.35406 I0406 08:43:42.475436 5226 solver.cpp:237] Train net output #0: loss = 2.35406 (* 1 = 2.35406 loss) I0406 08:43:42.475445 5226 sgd_solver.cpp:105] Iteration 11028, lr = 0.01 I0406 08:43:47.879287 5226 solver.cpp:218] Iteration 11040 (2.22066 iter/s, 5.4038s/12 iters), loss = 1.90254 I0406 08:43:47.879431 5226 solver.cpp:237] Train net output #0: loss = 1.90254 (* 1 = 1.90254 loss) I0406 08:43:47.879441 5226 sgd_solver.cpp:105] Iteration 11040, lr = 0.01 I0406 08:43:53.166824 5226 solver.cpp:218] Iteration 11052 (2.26957 iter/s, 5.28734s/12 iters), loss = 2.12704 I0406 08:43:53.166877 5226 solver.cpp:237] Train net output #0: loss = 2.12704 (* 1 = 2.12704 loss) I0406 08:43:53.166887 5226 sgd_solver.cpp:105] Iteration 11052, lr = 0.01 I0406 08:43:58.449493 5226 solver.cpp:218] Iteration 11064 (2.27162 iter/s, 5.28257s/12 iters), loss = 1.83219 I0406 08:43:58.449535 5226 solver.cpp:237] Train net output #0: loss = 1.83219 (* 1 = 1.83219 loss) I0406 08:43:58.449540 5226 sgd_solver.cpp:105] Iteration 11064, lr = 0.01 I0406 08:44:03.714269 5226 solver.cpp:218] Iteration 11076 (2.27934 iter/s, 5.26468s/12 iters), loss = 1.55809 I0406 08:44:03.714316 5226 solver.cpp:237] Train net output #0: loss = 1.55809 (* 1 = 1.55809 loss) I0406 08:44:03.714324 5226 sgd_solver.cpp:105] Iteration 11076, lr = 0.01 I0406 08:44:09.132321 5226 solver.cpp:218] Iteration 11088 (2.21486 iter/s, 5.41795s/12 iters), loss = 1.88972 I0406 08:44:09.132369 5226 solver.cpp:237] Train net output #0: loss = 1.88972 (* 1 = 1.88972 loss) I0406 08:44:09.132376 5226 sgd_solver.cpp:105] Iteration 11088, lr = 0.01 I0406 08:44:12.499819 5226 blocking_queue.cpp:49] Waiting for data I0406 08:44:14.290684 5226 solver.cpp:218] Iteration 11100 (2.32636 iter/s, 5.15827s/12 iters), loss = 1.67094 I0406 08:44:14.290722 5226 solver.cpp:237] Train net output #0: loss = 1.67094 (* 1 = 1.67094 loss) I0406 08:44:14.290727 5226 sgd_solver.cpp:105] Iteration 11100, lr = 0.01 I0406 08:44:19.393898 5226 solver.cpp:218] Iteration 11112 (2.3515 iter/s, 5.10313s/12 iters), loss = 1.78406 I0406 08:44:19.394153 5226 solver.cpp:237] Train net output #0: loss = 1.78406 (* 1 = 1.78406 loss) I0406 08:44:19.394162 5226 sgd_solver.cpp:105] Iteration 11112, lr = 0.01 I0406 08:44:21.521761 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_11118.caffemodel I0406 08:44:24.534940 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_11118.solverstate I0406 08:44:26.852931 5226 solver.cpp:330] Iteration 11118, Testing net (#0) I0406 08:44:26.852949 5226 net.cpp:676] Ignoring source layer train-data I0406 08:44:31.170395 5226 solver.cpp:397] Test net output #0: accuracy = 0.246324 I0406 08:44:31.170440 5226 solver.cpp:397] Test net output #1: loss = 3.55621 (* 1 = 3.55621 loss) I0406 08:44:31.713189 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:44:33.030786 5226 solver.cpp:218] Iteration 11124 (0.879989 iter/s, 13.6365s/12 iters), loss = 1.94776 I0406 08:44:33.030838 5226 solver.cpp:237] Train net output #0: loss = 1.94776 (* 1 = 1.94776 loss) I0406 08:44:33.030845 5226 sgd_solver.cpp:105] Iteration 11124, lr = 0.01 I0406 08:44:34.016922 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:44:38.369143 5226 solver.cpp:218] Iteration 11136 (2.24792 iter/s, 5.33826s/12 iters), loss = 2.20724 I0406 08:44:38.369192 5226 solver.cpp:237] Train net output #0: loss = 2.20724 (* 1 = 2.20724 loss) I0406 08:44:38.369201 5226 sgd_solver.cpp:105] Iteration 11136, lr = 0.01 I0406 08:44:43.549877 5226 solver.cpp:218] Iteration 11148 (2.31632 iter/s, 5.18064s/12 iters), loss = 1.75799 I0406 08:44:43.549916 5226 solver.cpp:237] Train net output #0: loss = 1.75799 (* 1 = 1.75799 loss) I0406 08:44:43.549921 5226 sgd_solver.cpp:105] Iteration 11148, lr = 0.01 I0406 08:44:48.743196 5226 solver.cpp:218] Iteration 11160 (2.3107 iter/s, 5.19323s/12 iters), loss = 1.93383 I0406 08:44:48.743242 5226 solver.cpp:237] Train net output #0: loss = 1.93383 (* 1 = 1.93383 loss) I0406 08:44:48.743250 5226 sgd_solver.cpp:105] Iteration 11160, lr = 0.01 I0406 08:44:54.212308 5226 solver.cpp:218] Iteration 11172 (2.19418 iter/s, 5.46902s/12 iters), loss = 1.63516 I0406 08:44:54.212432 5226 solver.cpp:237] Train net output #0: loss = 1.63516 (* 1 = 1.63516 loss) I0406 08:44:54.212440 5226 sgd_solver.cpp:105] Iteration 11172, lr = 0.01 I0406 08:44:59.539041 5226 solver.cpp:218] Iteration 11184 (2.25286 iter/s, 5.32655s/12 iters), loss = 1.94287 I0406 08:44:59.539088 5226 solver.cpp:237] Train net output #0: loss = 1.94287 (* 1 = 1.94287 loss) I0406 08:44:59.539096 5226 sgd_solver.cpp:105] Iteration 11184, lr = 0.01 I0406 08:45:04.441113 5226 solver.cpp:218] Iteration 11196 (2.44799 iter/s, 4.90198s/12 iters), loss = 1.70559 I0406 08:45:04.441150 5226 solver.cpp:237] Train net output #0: loss = 1.70559 (* 1 = 1.70559 loss) I0406 08:45:04.441156 5226 sgd_solver.cpp:105] Iteration 11196, lr = 0.01 I0406 08:45:09.635489 5226 solver.cpp:218] Iteration 11208 (2.31023 iter/s, 5.19429s/12 iters), loss = 1.97047 I0406 08:45:09.635529 5226 solver.cpp:237] Train net output #0: loss = 1.97047 (* 1 = 1.97047 loss) I0406 08:45:09.635535 5226 sgd_solver.cpp:105] Iteration 11208, lr = 0.01 I0406 08:45:14.473090 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_11220.caffemodel I0406 08:45:17.451676 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_11220.solverstate I0406 08:45:19.750473 5226 solver.cpp:330] Iteration 11220, Testing net (#0) I0406 08:45:19.750494 5226 net.cpp:676] Ignoring source layer train-data I0406 08:45:24.186167 5226 solver.cpp:397] Test net output #0: accuracy = 0.262868 I0406 08:45:24.186194 5226 solver.cpp:397] Test net output #1: loss = 3.56059 (* 1 = 3.56059 loss) I0406 08:45:24.325489 5226 solver.cpp:218] Iteration 11220 (0.81689 iter/s, 14.6899s/12 iters), loss = 1.89074 I0406 08:45:24.325567 5226 solver.cpp:237] Train net output #0: loss = 1.89074 (* 1 = 1.89074 loss) I0406 08:45:24.325573 5226 sgd_solver.cpp:105] Iteration 11220, lr = 0.01 I0406 08:45:24.670397 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:45:26.515039 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:45:28.635885 5226 solver.cpp:218] Iteration 11232 (2.78405 iter/s, 4.31027s/12 iters), loss = 1.76823 I0406 08:45:28.635924 5226 solver.cpp:237] Train net output #0: loss = 1.76823 (* 1 = 1.76823 loss) I0406 08:45:28.635929 5226 sgd_solver.cpp:105] Iteration 11232, lr = 0.01 I0406 08:45:34.088584 5226 solver.cpp:218] Iteration 11244 (2.20078 iter/s, 5.45261s/12 iters), loss = 1.71553 I0406 08:45:34.088642 5226 solver.cpp:237] Train net output #0: loss = 1.71553 (* 1 = 1.71553 loss) I0406 08:45:34.088651 5226 sgd_solver.cpp:105] Iteration 11244, lr = 0.01 I0406 08:45:39.444531 5226 solver.cpp:218] Iteration 11256 (2.24054 iter/s, 5.35584s/12 iters), loss = 2.09982 I0406 08:45:39.444581 5226 solver.cpp:237] Train net output #0: loss = 2.09982 (* 1 = 2.09982 loss) I0406 08:45:39.444896 5226 sgd_solver.cpp:105] Iteration 11256, lr = 0.01 I0406 08:45:44.417891 5226 solver.cpp:218] Iteration 11268 (2.4129 iter/s, 4.97327s/12 iters), loss = 1.95018 I0406 08:45:44.417939 5226 solver.cpp:237] Train net output #0: loss = 1.95018 (* 1 = 1.95018 loss) I0406 08:45:44.417948 5226 sgd_solver.cpp:105] Iteration 11268, lr = 0.01 I0406 08:45:49.744841 5226 solver.cpp:218] Iteration 11280 (2.25274 iter/s, 5.32685s/12 iters), loss = 1.56122 I0406 08:45:49.744895 5226 solver.cpp:237] Train net output #0: loss = 1.56122 (* 1 = 1.56122 loss) I0406 08:45:49.744904 5226 sgd_solver.cpp:105] Iteration 11280, lr = 0.01 I0406 08:45:54.977285 5226 solver.cpp:218] Iteration 11292 (2.29342 iter/s, 5.23235s/12 iters), loss = 1.91252 I0406 08:45:54.977445 5226 solver.cpp:237] Train net output #0: loss = 1.91252 (* 1 = 1.91252 loss) I0406 08:45:54.977452 5226 sgd_solver.cpp:105] Iteration 11292, lr = 0.01 I0406 08:46:00.304841 5226 solver.cpp:218] Iteration 11304 (2.25253 iter/s, 5.32735s/12 iters), loss = 1.74319 I0406 08:46:00.304904 5226 solver.cpp:237] Train net output #0: loss = 1.74319 (* 1 = 1.74319 loss) I0406 08:46:00.304913 5226 sgd_solver.cpp:105] Iteration 11304, lr = 0.01 I0406 08:46:05.650084 5226 solver.cpp:218] Iteration 11316 (2.24503 iter/s, 5.34514s/12 iters), loss = 1.84395 I0406 08:46:05.650125 5226 solver.cpp:237] Train net output #0: loss = 1.84395 (* 1 = 1.84395 loss) I0406 08:46:05.650130 5226 sgd_solver.cpp:105] Iteration 11316, lr = 0.01 I0406 08:46:07.815052 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_11322.caffemodel I0406 08:46:10.831476 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_11322.solverstate I0406 08:46:13.159904 5226 solver.cpp:330] Iteration 11322, Testing net (#0) I0406 08:46:13.159925 5226 net.cpp:676] Ignoring source layer train-data I0406 08:46:17.661759 5226 solver.cpp:397] Test net output #0: accuracy = 0.240809 I0406 08:46:17.661789 5226 solver.cpp:397] Test net output #1: loss = 3.58143 (* 1 = 3.58143 loss) I0406 08:46:18.092340 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:46:19.601292 5226 solver.cpp:218] Iteration 11328 (0.860149 iter/s, 13.9511s/12 iters), loss = 1.7194 I0406 08:46:19.601342 5226 solver.cpp:237] Train net output #0: loss = 1.7194 (* 1 = 1.7194 loss) I0406 08:46:19.601353 5226 sgd_solver.cpp:105] Iteration 11328, lr = 0.01 I0406 08:46:19.746233 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:46:24.676730 5226 solver.cpp:218] Iteration 11340 (2.36437 iter/s, 5.07534s/12 iters), loss = 1.47063 I0406 08:46:24.676769 5226 solver.cpp:237] Train net output #0: loss = 1.47063 (* 1 = 1.47063 loss) I0406 08:46:24.676774 5226 sgd_solver.cpp:105] Iteration 11340, lr = 0.01 I0406 08:46:29.999202 5226 solver.cpp:218] Iteration 11352 (2.25463 iter/s, 5.32238s/12 iters), loss = 1.51013 I0406 08:46:29.999294 5226 solver.cpp:237] Train net output #0: loss = 1.51013 (* 1 = 1.51013 loss) I0406 08:46:29.999300 5226 sgd_solver.cpp:105] Iteration 11352, lr = 0.01 I0406 08:46:35.487938 5226 solver.cpp:218] Iteration 11364 (2.18635 iter/s, 5.4886s/12 iters), loss = 1.51933 I0406 08:46:35.487978 5226 solver.cpp:237] Train net output #0: loss = 1.51933 (* 1 = 1.51933 loss) I0406 08:46:35.487983 5226 sgd_solver.cpp:105] Iteration 11364, lr = 0.01 I0406 08:46:40.406492 5226 solver.cpp:218] Iteration 11376 (2.43978 iter/s, 4.91847s/12 iters), loss = 2.08425 I0406 08:46:40.406528 5226 solver.cpp:237] Train net output #0: loss = 2.08425 (* 1 = 2.08425 loss) I0406 08:46:40.406535 5226 sgd_solver.cpp:105] Iteration 11376, lr = 0.01 I0406 08:46:45.670711 5226 solver.cpp:218] Iteration 11388 (2.27958 iter/s, 5.26413s/12 iters), loss = 1.78707 I0406 08:46:45.670749 5226 solver.cpp:237] Train net output #0: loss = 1.78707 (* 1 = 1.78707 loss) I0406 08:46:45.670754 5226 sgd_solver.cpp:105] Iteration 11388, lr = 0.01 I0406 08:46:51.078462 5226 solver.cpp:218] Iteration 11400 (2.21907 iter/s, 5.40766s/12 iters), loss = 2.40512 I0406 08:46:51.078505 5226 solver.cpp:237] Train net output #0: loss = 2.40512 (* 1 = 2.40512 loss) I0406 08:46:51.078510 5226 sgd_solver.cpp:105] Iteration 11400, lr = 0.01 I0406 08:46:56.327033 5226 solver.cpp:218] Iteration 11412 (2.28638 iter/s, 5.24848s/12 iters), loss = 2.22505 I0406 08:46:56.327080 5226 solver.cpp:237] Train net output #0: loss = 2.22505 (* 1 = 2.22505 loss) I0406 08:46:56.327086 5226 sgd_solver.cpp:105] Iteration 11412, lr = 0.01 I0406 08:47:01.181751 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_11424.caffemodel I0406 08:47:04.233780 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_11424.solverstate I0406 08:47:06.530071 5226 solver.cpp:330] Iteration 11424, Testing net (#0) I0406 08:47:06.530094 5226 net.cpp:676] Ignoring source layer train-data I0406 08:47:10.898622 5226 solver.cpp:397] Test net output #0: accuracy = 0.230392 I0406 08:47:10.898651 5226 solver.cpp:397] Test net output #1: loss = 3.71164 (* 1 = 3.71164 loss) I0406 08:47:11.038820 5226 solver.cpp:218] Iteration 11424 (0.815681 iter/s, 14.7116s/12 iters), loss = 2.23305 I0406 08:47:11.038858 5226 solver.cpp:237] Train net output #0: loss = 2.23305 (* 1 = 2.23305 loss) I0406 08:47:11.038863 5226 sgd_solver.cpp:105] Iteration 11424, lr = 0.01 I0406 08:47:11.324560 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:47:12.549612 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:47:15.332767 5226 solver.cpp:218] Iteration 11436 (2.79469 iter/s, 4.29386s/12 iters), loss = 1.91544 I0406 08:47:15.332818 5226 solver.cpp:237] Train net output #0: loss = 1.91544 (* 1 = 1.91544 loss) I0406 08:47:15.332825 5226 sgd_solver.cpp:105] Iteration 11436, lr = 0.01 I0406 08:47:20.470675 5226 solver.cpp:218] Iteration 11448 (2.33562 iter/s, 5.13781s/12 iters), loss = 2.01688 I0406 08:47:20.470715 5226 solver.cpp:237] Train net output #0: loss = 2.01688 (* 1 = 2.01688 loss) I0406 08:47:20.470722 5226 sgd_solver.cpp:105] Iteration 11448, lr = 0.01 I0406 08:47:25.780690 5226 solver.cpp:218] Iteration 11460 (2.25992 iter/s, 5.30992s/12 iters), loss = 1.84641 I0406 08:47:25.780743 5226 solver.cpp:237] Train net output #0: loss = 1.84641 (* 1 = 1.84641 loss) I0406 08:47:25.780752 5226 sgd_solver.cpp:105] Iteration 11460, lr = 0.01 I0406 08:47:31.148954 5226 solver.cpp:218] Iteration 11472 (2.2354 iter/s, 5.36817s/12 iters), loss = 1.58496 I0406 08:47:31.148993 5226 solver.cpp:237] Train net output #0: loss = 1.58496 (* 1 = 1.58496 loss) I0406 08:47:31.148998 5226 sgd_solver.cpp:105] Iteration 11472, lr = 0.01 I0406 08:47:36.368921 5226 solver.cpp:218] Iteration 11484 (2.2989 iter/s, 5.21988s/12 iters), loss = 1.80257 I0406 08:47:36.369017 5226 solver.cpp:237] Train net output #0: loss = 1.80257 (* 1 = 1.80257 loss) I0406 08:47:36.369024 5226 sgd_solver.cpp:105] Iteration 11484, lr = 0.01 I0406 08:47:41.360534 5226 solver.cpp:218] Iteration 11496 (2.4041 iter/s, 4.99146s/12 iters), loss = 2.08389 I0406 08:47:41.360582 5226 solver.cpp:237] Train net output #0: loss = 2.08389 (* 1 = 2.08389 loss) I0406 08:47:41.360590 5226 sgd_solver.cpp:105] Iteration 11496, lr = 0.01 I0406 08:47:46.717453 5226 solver.cpp:218] Iteration 11508 (2.24013 iter/s, 5.35682s/12 iters), loss = 1.9439 I0406 08:47:46.717491 5226 solver.cpp:237] Train net output #0: loss = 1.9439 (* 1 = 1.9439 loss) I0406 08:47:46.717496 5226 sgd_solver.cpp:105] Iteration 11508, lr = 0.01 I0406 08:47:52.029862 5226 solver.cpp:218] Iteration 11520 (2.2589 iter/s, 5.31233s/12 iters), loss = 1.89792 I0406 08:47:52.029902 5226 solver.cpp:237] Train net output #0: loss = 1.89792 (* 1 = 1.89792 loss) I0406 08:47:52.029907 5226 sgd_solver.cpp:105] Iteration 11520, lr = 0.01 I0406 08:47:54.225944 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_11526.caffemodel I0406 08:47:57.290717 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_11526.solverstate I0406 08:47:59.603242 5226 solver.cpp:330] Iteration 11526, Testing net (#0) I0406 08:47:59.603261 5226 net.cpp:676] Ignoring source layer train-data I0406 08:48:03.912118 5226 solver.cpp:397] Test net output #0: accuracy = 0.23652 I0406 08:48:03.912156 5226 solver.cpp:397] Test net output #1: loss = 3.67689 (* 1 = 3.67689 loss) I0406 08:48:04.013293 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:48:05.171188 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:48:05.726559 5226 solver.cpp:218] Iteration 11532 (0.876132 iter/s, 13.6966s/12 iters), loss = 2.81239 I0406 08:48:05.726611 5226 solver.cpp:237] Train net output #0: loss = 2.81239 (* 1 = 2.81239 loss) I0406 08:48:05.726619 5226 sgd_solver.cpp:105] Iteration 11532, lr = 0.01 I0406 08:48:10.882457 5226 solver.cpp:218] Iteration 11544 (2.32748 iter/s, 5.1558s/12 iters), loss = 2.0758 I0406 08:48:10.882588 5226 solver.cpp:237] Train net output #0: loss = 2.0758 (* 1 = 2.0758 loss) I0406 08:48:10.882594 5226 sgd_solver.cpp:105] Iteration 11544, lr = 0.01 I0406 08:48:15.955760 5226 solver.cpp:218] Iteration 11556 (2.3654 iter/s, 5.07313s/12 iters), loss = 2.28721 I0406 08:48:15.955796 5226 solver.cpp:237] Train net output #0: loss = 2.28721 (* 1 = 2.28721 loss) I0406 08:48:15.955801 5226 sgd_solver.cpp:105] Iteration 11556, lr = 0.01 I0406 08:48:21.323837 5226 solver.cpp:218] Iteration 11568 (2.23547 iter/s, 5.36799s/12 iters), loss = 2.00153 I0406 08:48:21.323875 5226 solver.cpp:237] Train net output #0: loss = 2.00153 (* 1 = 2.00153 loss) I0406 08:48:21.323881 5226 sgd_solver.cpp:105] Iteration 11568, lr = 0.01 I0406 08:48:26.642590 5226 solver.cpp:218] Iteration 11580 (2.2562 iter/s, 5.31867s/12 iters), loss = 2.18757 I0406 08:48:26.642630 5226 solver.cpp:237] Train net output #0: loss = 2.18757 (* 1 = 2.18757 loss) I0406 08:48:26.642635 5226 sgd_solver.cpp:105] Iteration 11580, lr = 0.01 I0406 08:48:31.888741 5226 solver.cpp:218] Iteration 11592 (2.28743 iter/s, 5.24606s/12 iters), loss = 1.78886 I0406 08:48:31.888792 5226 solver.cpp:237] Train net output #0: loss = 1.78886 (* 1 = 1.78886 loss) I0406 08:48:31.888799 5226 sgd_solver.cpp:105] Iteration 11592, lr = 0.01 I0406 08:48:37.106726 5226 solver.cpp:218] Iteration 11604 (2.29978 iter/s, 5.21789s/12 iters), loss = 1.63513 I0406 08:48:37.106775 5226 solver.cpp:237] Train net output #0: loss = 1.63513 (* 1 = 1.63513 loss) I0406 08:48:37.106782 5226 sgd_solver.cpp:105] Iteration 11604, lr = 0.01 I0406 08:48:42.493919 5226 solver.cpp:218] Iteration 11616 (2.22754 iter/s, 5.3871s/12 iters), loss = 2.03324 I0406 08:48:42.494065 5226 solver.cpp:237] Train net output #0: loss = 2.03324 (* 1 = 2.03324 loss) I0406 08:48:42.494073 5226 sgd_solver.cpp:105] Iteration 11616, lr = 0.01 I0406 08:48:47.447613 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_11628.caffemodel I0406 08:48:50.383462 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_11628.solverstate I0406 08:48:52.709919 5226 solver.cpp:330] Iteration 11628, Testing net (#0) I0406 08:48:52.709939 5226 net.cpp:676] Ignoring source layer train-data I0406 08:48:57.166172 5226 solver.cpp:397] Test net output #0: accuracy = 0.242034 I0406 08:48:57.166209 5226 solver.cpp:397] Test net output #1: loss = 3.67389 (* 1 = 3.67389 loss) I0406 08:48:57.241222 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:48:57.301640 5226 solver.cpp:218] Iteration 11628 (0.810401 iter/s, 14.8075s/12 iters), loss = 1.84521 I0406 08:48:57.301681 5226 solver.cpp:237] Train net output #0: loss = 1.84521 (* 1 = 1.84521 loss) I0406 08:48:57.301687 5226 sgd_solver.cpp:105] Iteration 11628, lr = 0.01 I0406 08:48:58.103940 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:49:01.750751 5226 solver.cpp:218] Iteration 11640 (2.69722 iter/s, 4.44903s/12 iters), loss = 1.84684 I0406 08:49:01.750809 5226 solver.cpp:237] Train net output #0: loss = 1.84684 (* 1 = 1.84684 loss) I0406 08:49:01.750818 5226 sgd_solver.cpp:105] Iteration 11640, lr = 0.01 I0406 08:49:06.927625 5226 solver.cpp:218] Iteration 11652 (2.31805 iter/s, 5.17677s/12 iters), loss = 1.57954 I0406 08:49:06.927683 5226 solver.cpp:237] Train net output #0: loss = 1.57954 (* 1 = 1.57954 loss) I0406 08:49:06.927692 5226 sgd_solver.cpp:105] Iteration 11652, lr = 0.01 I0406 08:49:12.229154 5226 solver.cpp:218] Iteration 11664 (2.26354 iter/s, 5.30143s/12 iters), loss = 1.72684 I0406 08:49:12.229195 5226 solver.cpp:237] Train net output #0: loss = 1.72684 (* 1 = 1.72684 loss) I0406 08:49:12.229200 5226 sgd_solver.cpp:105] Iteration 11664, lr = 0.01 I0406 08:49:17.606823 5226 solver.cpp:218] Iteration 11676 (2.23149 iter/s, 5.37758s/12 iters), loss = 1.56391 I0406 08:49:17.606952 5226 solver.cpp:237] Train net output #0: loss = 1.56391 (* 1 = 1.56391 loss) I0406 08:49:17.606959 5226 sgd_solver.cpp:105] Iteration 11676, lr = 0.01 I0406 08:49:22.938372 5226 solver.cpp:218] Iteration 11688 (2.25083 iter/s, 5.33137s/12 iters), loss = 1.78094 I0406 08:49:22.938426 5226 solver.cpp:237] Train net output #0: loss = 1.78094 (* 1 = 1.78094 loss) I0406 08:49:22.938434 5226 sgd_solver.cpp:105] Iteration 11688, lr = 0.01 I0406 08:49:28.175235 5226 solver.cpp:218] Iteration 11700 (2.29149 iter/s, 5.23676s/12 iters), loss = 1.97296 I0406 08:49:28.175285 5226 solver.cpp:237] Train net output #0: loss = 1.97296 (* 1 = 1.97296 loss) I0406 08:49:28.175293 5226 sgd_solver.cpp:105] Iteration 11700, lr = 0.01 I0406 08:49:33.530195 5226 solver.cpp:218] Iteration 11712 (2.24095 iter/s, 5.35487s/12 iters), loss = 1.85167 I0406 08:49:33.530232 5226 solver.cpp:237] Train net output #0: loss = 1.85167 (* 1 = 1.85167 loss) I0406 08:49:33.530237 5226 sgd_solver.cpp:105] Iteration 11712, lr = 0.01 I0406 08:49:38.937616 5226 solver.cpp:218] Iteration 11724 (2.21921 iter/s, 5.40734s/12 iters), loss = 1.89333 I0406 08:49:38.937652 5226 solver.cpp:237] Train net output #0: loss = 1.89333 (* 1 = 1.89333 loss) I0406 08:49:38.937659 5226 sgd_solver.cpp:105] Iteration 11724, lr = 0.01 I0406 08:49:41.065572 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_11730.caffemodel I0406 08:49:44.520797 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_11730.solverstate I0406 08:49:47.031922 5226 solver.cpp:330] Iteration 11730, Testing net (#0) I0406 08:49:47.031944 5226 net.cpp:676] Ignoring source layer train-data I0406 08:49:51.289775 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:49:51.319869 5226 solver.cpp:397] Test net output #0: accuracy = 0.213848 I0406 08:49:51.319909 5226 solver.cpp:397] Test net output #1: loss = 3.78413 (* 1 = 3.78413 loss) I0406 08:49:51.936784 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:49:53.183463 5226 solver.cpp:218] Iteration 11736 (0.842359 iter/s, 14.2457s/12 iters), loss = 2.17403 I0406 08:49:53.183506 5226 solver.cpp:237] Train net output #0: loss = 2.17403 (* 1 = 2.17403 loss) I0406 08:49:53.183511 5226 sgd_solver.cpp:105] Iteration 11736, lr = 0.01 I0406 08:49:58.525866 5226 solver.cpp:218] Iteration 11748 (2.24622 iter/s, 5.34231s/12 iters), loss = 2.08566 I0406 08:49:58.525923 5226 solver.cpp:237] Train net output #0: loss = 2.08566 (* 1 = 2.08566 loss) I0406 08:49:58.525931 5226 sgd_solver.cpp:105] Iteration 11748, lr = 0.01 I0406 08:50:03.896023 5226 solver.cpp:218] Iteration 11760 (2.23462 iter/s, 5.37005s/12 iters), loss = 1.71655 I0406 08:50:03.896062 5226 solver.cpp:237] Train net output #0: loss = 1.71655 (* 1 = 1.71655 loss) I0406 08:50:03.896068 5226 sgd_solver.cpp:105] Iteration 11760, lr = 0.01 I0406 08:50:09.187372 5226 solver.cpp:218] Iteration 11772 (2.26789 iter/s, 5.29126s/12 iters), loss = 1.73329 I0406 08:50:09.187413 5226 solver.cpp:237] Train net output #0: loss = 1.73329 (* 1 = 1.73329 loss) I0406 08:50:09.187418 5226 sgd_solver.cpp:105] Iteration 11772, lr = 0.01 I0406 08:50:13.149471 5226 blocking_queue.cpp:49] Waiting for data I0406 08:50:14.576890 5226 solver.cpp:218] Iteration 11784 (2.22658 iter/s, 5.38942s/12 iters), loss = 1.95888 I0406 08:50:14.576946 5226 solver.cpp:237] Train net output #0: loss = 1.95888 (* 1 = 1.95888 loss) I0406 08:50:14.576956 5226 sgd_solver.cpp:105] Iteration 11784, lr = 0.01 I0406 08:50:19.932400 5226 solver.cpp:218] Iteration 11796 (2.24073 iter/s, 5.35541s/12 iters), loss = 1.69777 I0406 08:50:19.932440 5226 solver.cpp:237] Train net output #0: loss = 1.69777 (* 1 = 1.69777 loss) I0406 08:50:19.932446 5226 sgd_solver.cpp:105] Iteration 11796, lr = 0.01 I0406 08:50:25.074579 5226 solver.cpp:218] Iteration 11808 (2.33368 iter/s, 5.14209s/12 iters), loss = 1.52417 I0406 08:50:25.074707 5226 solver.cpp:237] Train net output #0: loss = 1.52417 (* 1 = 1.52417 loss) I0406 08:50:25.074713 5226 sgd_solver.cpp:105] Iteration 11808, lr = 0.01 I0406 08:50:30.474642 5226 solver.cpp:218] Iteration 11820 (2.22227 iter/s, 5.3999s/12 iters), loss = 1.84007 I0406 08:50:30.474678 5226 solver.cpp:237] Train net output #0: loss = 1.84007 (* 1 = 1.84007 loss) I0406 08:50:30.474682 5226 sgd_solver.cpp:105] Iteration 11820, lr = 0.01 I0406 08:50:35.161226 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_11832.caffemodel I0406 08:50:38.176852 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_11832.solverstate I0406 08:50:40.475224 5226 solver.cpp:330] Iteration 11832, Testing net (#0) I0406 08:50:40.475242 5226 net.cpp:676] Ignoring source layer train-data I0406 08:50:44.763263 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:50:44.826130 5226 solver.cpp:397] Test net output #0: accuracy = 0.259804 I0406 08:50:44.826164 5226 solver.cpp:397] Test net output #1: loss = 3.6715 (* 1 = 3.6715 loss) I0406 08:50:44.966567 5226 solver.cpp:218] Iteration 11832 (0.828056 iter/s, 14.4918s/12 iters), loss = 1.53108 I0406 08:50:44.966635 5226 solver.cpp:237] Train net output #0: loss = 1.53108 (* 1 = 1.53108 loss) I0406 08:50:44.966645 5226 sgd_solver.cpp:105] Iteration 11832, lr = 0.01 I0406 08:50:45.064098 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:50:49.490244 5226 solver.cpp:218] Iteration 11844 (2.65277 iter/s, 4.52357s/12 iters), loss = 2.03592 I0406 08:50:49.490284 5226 solver.cpp:237] Train net output #0: loss = 2.03592 (* 1 = 2.03592 loss) I0406 08:50:49.490289 5226 sgd_solver.cpp:105] Iteration 11844, lr = 0.01 I0406 08:50:54.916782 5226 solver.cpp:218] Iteration 11856 (2.21139 iter/s, 5.42645s/12 iters), loss = 1.60784 I0406 08:50:54.916819 5226 solver.cpp:237] Train net output #0: loss = 1.60784 (* 1 = 1.60784 loss) I0406 08:50:54.916824 5226 sgd_solver.cpp:105] Iteration 11856, lr = 0.01 I0406 08:51:00.330348 5226 solver.cpp:218] Iteration 11868 (2.21669 iter/s, 5.41348s/12 iters), loss = 1.80591 I0406 08:51:00.330443 5226 solver.cpp:237] Train net output #0: loss = 1.80591 (* 1 = 1.80591 loss) I0406 08:51:00.330451 5226 sgd_solver.cpp:105] Iteration 11868, lr = 0.01 I0406 08:51:05.456890 5226 solver.cpp:218] Iteration 11880 (2.34083 iter/s, 5.1264s/12 iters), loss = 1.42798 I0406 08:51:05.456930 5226 solver.cpp:237] Train net output #0: loss = 1.42798 (* 1 = 1.42798 loss) I0406 08:51:05.456936 5226 sgd_solver.cpp:105] Iteration 11880, lr = 0.01 I0406 08:51:10.642530 5226 solver.cpp:218] Iteration 11892 (2.31412 iter/s, 5.18555s/12 iters), loss = 1.89454 I0406 08:51:10.642570 5226 solver.cpp:237] Train net output #0: loss = 1.89454 (* 1 = 1.89454 loss) I0406 08:51:10.642575 5226 sgd_solver.cpp:105] Iteration 11892, lr = 0.01 I0406 08:51:16.020606 5226 solver.cpp:218] Iteration 11904 (2.23132 iter/s, 5.37799s/12 iters), loss = 1.938 I0406 08:51:16.020654 5226 solver.cpp:237] Train net output #0: loss = 1.938 (* 1 = 1.938 loss) I0406 08:51:16.020661 5226 sgd_solver.cpp:105] Iteration 11904, lr = 0.01 I0406 08:51:21.348755 5226 solver.cpp:218] Iteration 11916 (2.25223 iter/s, 5.32806s/12 iters), loss = 2.01526 I0406 08:51:21.348794 5226 solver.cpp:237] Train net output #0: loss = 2.01526 (* 1 = 2.01526 loss) I0406 08:51:21.348800 5226 sgd_solver.cpp:105] Iteration 11916, lr = 0.01 I0406 08:51:26.541290 5226 solver.cpp:218] Iteration 11928 (2.31105 iter/s, 5.19245s/12 iters), loss = 1.84794 I0406 08:51:26.541338 5226 solver.cpp:237] Train net output #0: loss = 1.84794 (* 1 = 1.84794 loss) I0406 08:51:26.541347 5226 sgd_solver.cpp:105] Iteration 11928, lr = 0.01 I0406 08:51:28.814744 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_11934.caffemodel I0406 08:51:29.604965 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:51:31.819617 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_11934.solverstate I0406 08:51:34.119446 5226 solver.cpp:330] Iteration 11934, Testing net (#0) I0406 08:51:34.119465 5226 net.cpp:676] Ignoring source layer train-data I0406 08:51:38.428683 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:51:38.534747 5226 solver.cpp:397] Test net output #0: accuracy = 0.223652 I0406 08:51:38.534783 5226 solver.cpp:397] Test net output #1: loss = 3.74559 (* 1 = 3.74559 loss) I0406 08:51:40.436434 5226 solver.cpp:218] Iteration 11940 (0.86362 iter/s, 13.895s/12 iters), loss = 2.19051 I0406 08:51:40.436487 5226 solver.cpp:237] Train net output #0: loss = 2.19051 (* 1 = 2.19051 loss) I0406 08:51:40.436496 5226 sgd_solver.cpp:105] Iteration 11940, lr = 0.01 I0406 08:51:45.733759 5226 solver.cpp:218] Iteration 11952 (2.26534 iter/s, 5.29723s/12 iters), loss = 2.09312 I0406 08:51:45.733798 5226 solver.cpp:237] Train net output #0: loss = 2.09312 (* 1 = 2.09312 loss) I0406 08:51:45.733804 5226 sgd_solver.cpp:105] Iteration 11952, lr = 0.01 I0406 08:51:51.125803 5226 solver.cpp:218] Iteration 11964 (2.22554 iter/s, 5.39195s/12 iters), loss = 1.97578 I0406 08:51:51.125855 5226 solver.cpp:237] Train net output #0: loss = 1.97578 (* 1 = 1.97578 loss) I0406 08:51:51.125864 5226 sgd_solver.cpp:105] Iteration 11964, lr = 0.01 I0406 08:51:56.505972 5226 solver.cpp:218] Iteration 11976 (2.23045 iter/s, 5.38007s/12 iters), loss = 1.82206 I0406 08:51:56.506022 5226 solver.cpp:237] Train net output #0: loss = 1.82206 (* 1 = 1.82206 loss) I0406 08:51:56.506031 5226 sgd_solver.cpp:105] Iteration 11976, lr = 0.01 I0406 08:52:01.846727 5226 solver.cpp:218] Iteration 11988 (2.24692 iter/s, 5.34065s/12 iters), loss = 2.07828 I0406 08:52:01.846848 5226 solver.cpp:237] Train net output #0: loss = 2.07828 (* 1 = 2.07828 loss) I0406 08:52:01.846858 5226 sgd_solver.cpp:105] Iteration 11988, lr = 0.01 I0406 08:52:07.132922 5226 solver.cpp:218] Iteration 12000 (2.27013 iter/s, 5.28603s/12 iters), loss = 2.26341 I0406 08:52:07.132962 5226 solver.cpp:237] Train net output #0: loss = 2.26341 (* 1 = 2.26341 loss) I0406 08:52:07.132967 5226 sgd_solver.cpp:105] Iteration 12000, lr = 0.01 I0406 08:52:12.355908 5226 solver.cpp:218] Iteration 12012 (2.29758 iter/s, 5.2229s/12 iters), loss = 2.15603 I0406 08:52:12.355957 5226 solver.cpp:237] Train net output #0: loss = 2.15603 (* 1 = 2.15603 loss) I0406 08:52:12.355965 5226 sgd_solver.cpp:105] Iteration 12012, lr = 0.01 I0406 08:52:17.347137 5226 solver.cpp:218] Iteration 12024 (2.40426 iter/s, 4.99113s/12 iters), loss = 2.04246 I0406 08:52:17.347198 5226 solver.cpp:237] Train net output #0: loss = 2.04246 (* 1 = 2.04246 loss) I0406 08:52:17.347206 5226 sgd_solver.cpp:105] Iteration 12024, lr = 0.01 I0406 08:52:22.152793 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_12036.caffemodel I0406 08:52:22.776340 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:52:25.182013 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_12036.solverstate I0406 08:52:27.483173 5226 solver.cpp:330] Iteration 12036, Testing net (#0) I0406 08:52:27.483194 5226 net.cpp:676] Ignoring source layer train-data I0406 08:52:31.621233 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:52:31.758476 5226 solver.cpp:397] Test net output #0: accuracy = 0.227941 I0406 08:52:31.758510 5226 solver.cpp:397] Test net output #1: loss = 3.71568 (* 1 = 3.71568 loss) I0406 08:52:31.897584 5226 solver.cpp:218] Iteration 12036 (0.824726 iter/s, 14.5503s/12 iters), loss = 1.87633 I0406 08:52:31.899139 5226 solver.cpp:237] Train net output #0: loss = 1.87633 (* 1 = 1.87633 loss) I0406 08:52:31.899150 5226 sgd_solver.cpp:105] Iteration 12036, lr = 0.01 I0406 08:52:36.128262 5226 solver.cpp:218] Iteration 12048 (2.83749 iter/s, 4.22909s/12 iters), loss = 1.6342 I0406 08:52:36.128304 5226 solver.cpp:237] Train net output #0: loss = 1.6342 (* 1 = 1.6342 loss) I0406 08:52:36.128312 5226 sgd_solver.cpp:105] Iteration 12048, lr = 0.01 I0406 08:52:41.390291 5226 solver.cpp:218] Iteration 12060 (2.28053 iter/s, 5.26194s/12 iters), loss = 2.14986 I0406 08:52:41.390333 5226 solver.cpp:237] Train net output #0: loss = 2.14986 (* 1 = 2.14986 loss) I0406 08:52:41.390339 5226 sgd_solver.cpp:105] Iteration 12060, lr = 0.01 I0406 08:52:46.698346 5226 solver.cpp:218] Iteration 12072 (2.26075 iter/s, 5.30797s/12 iters), loss = 1.63935 I0406 08:52:46.698393 5226 solver.cpp:237] Train net output #0: loss = 1.63935 (* 1 = 1.63935 loss) I0406 08:52:46.698401 5226 sgd_solver.cpp:105] Iteration 12072, lr = 0.01 I0406 08:52:52.049711 5226 solver.cpp:218] Iteration 12084 (2.24246 iter/s, 5.35127s/12 iters), loss = 2.53069 I0406 08:52:52.049758 5226 solver.cpp:237] Train net output #0: loss = 2.53069 (* 1 = 2.53069 loss) I0406 08:52:52.049767 5226 sgd_solver.cpp:105] Iteration 12084, lr = 0.01 I0406 08:52:57.129184 5226 solver.cpp:218] Iteration 12096 (2.3625 iter/s, 5.07937s/12 iters), loss = 1.98032 I0406 08:52:57.129236 5226 solver.cpp:237] Train net output #0: loss = 1.98032 (* 1 = 1.98032 loss) I0406 08:52:57.129245 5226 sgd_solver.cpp:105] Iteration 12096, lr = 0.01 I0406 08:53:02.560909 5226 solver.cpp:218] Iteration 12108 (2.20928 iter/s, 5.43163s/12 iters), loss = 2.2319 I0406 08:53:02.560999 5226 solver.cpp:237] Train net output #0: loss = 2.2319 (* 1 = 2.2319 loss) I0406 08:53:02.561007 5226 sgd_solver.cpp:105] Iteration 12108, lr = 0.01 I0406 08:53:07.703745 5226 solver.cpp:218] Iteration 12120 (2.3334 iter/s, 5.1427s/12 iters), loss = 2.10387 I0406 08:53:07.703783 5226 solver.cpp:237] Train net output #0: loss = 2.10387 (* 1 = 2.10387 loss) I0406 08:53:07.703788 5226 sgd_solver.cpp:105] Iteration 12120, lr = 0.01 I0406 08:53:13.020109 5226 solver.cpp:218] Iteration 12132 (2.25722 iter/s, 5.31627s/12 iters), loss = 2.05641 I0406 08:53:13.020170 5226 solver.cpp:237] Train net output #0: loss = 2.05641 (* 1 = 2.05641 loss) I0406 08:53:13.020179 5226 sgd_solver.cpp:105] Iteration 12132, lr = 0.01 I0406 08:53:15.178459 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_12138.caffemodel I0406 08:53:15.515172 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:53:18.268491 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_12138.solverstate I0406 08:53:20.576864 5226 solver.cpp:330] Iteration 12138, Testing net (#0) I0406 08:53:20.576890 5226 net.cpp:676] Ignoring source layer train-data I0406 08:53:24.799331 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:53:24.971081 5226 solver.cpp:397] Test net output #0: accuracy = 0.19424 I0406 08:53:24.971118 5226 solver.cpp:397] Test net output #1: loss = 3.99001 (* 1 = 3.99001 loss) I0406 08:53:26.742281 5226 solver.cpp:218] Iteration 12144 (0.874507 iter/s, 13.722s/12 iters), loss = 1.99688 I0406 08:53:26.742321 5226 solver.cpp:237] Train net output #0: loss = 1.99688 (* 1 = 1.99688 loss) I0406 08:53:26.742326 5226 sgd_solver.cpp:105] Iteration 12144, lr = 0.01 I0406 08:53:31.936920 5226 solver.cpp:218] Iteration 12156 (2.31011 iter/s, 5.19455s/12 iters), loss = 2.5073 I0406 08:53:31.936959 5226 solver.cpp:237] Train net output #0: loss = 2.5073 (* 1 = 2.5073 loss) I0406 08:53:31.936965 5226 sgd_solver.cpp:105] Iteration 12156, lr = 0.01 I0406 08:53:37.194880 5226 solver.cpp:218] Iteration 12168 (2.28229 iter/s, 5.25787s/12 iters), loss = 2.20415 I0406 08:53:37.195035 5226 solver.cpp:237] Train net output #0: loss = 2.20415 (* 1 = 2.20415 loss) I0406 08:53:37.195044 5226 sgd_solver.cpp:105] Iteration 12168, lr = 0.01 I0406 08:53:42.546789 5226 solver.cpp:218] Iteration 12180 (2.24227 iter/s, 5.35171s/12 iters), loss = 2.0271 I0406 08:53:42.546830 5226 solver.cpp:237] Train net output #0: loss = 2.0271 (* 1 = 2.0271 loss) I0406 08:53:42.546835 5226 sgd_solver.cpp:105] Iteration 12180, lr = 0.01 I0406 08:53:47.811519 5226 solver.cpp:218] Iteration 12192 (2.27936 iter/s, 5.26464s/12 iters), loss = 1.98259 I0406 08:53:47.811561 5226 solver.cpp:237] Train net output #0: loss = 1.98259 (* 1 = 1.98259 loss) I0406 08:53:47.811568 5226 sgd_solver.cpp:105] Iteration 12192, lr = 0.01 I0406 08:53:53.320174 5226 solver.cpp:218] Iteration 12204 (2.17843 iter/s, 5.50856s/12 iters), loss = 1.6622 I0406 08:53:53.320225 5226 solver.cpp:237] Train net output #0: loss = 1.6622 (* 1 = 1.6622 loss) I0406 08:53:53.320235 5226 sgd_solver.cpp:105] Iteration 12204, lr = 0.01 I0406 08:53:58.705425 5226 solver.cpp:218] Iteration 12216 (2.22835 iter/s, 5.38515s/12 iters), loss = 2.35338 I0406 08:53:58.705467 5226 solver.cpp:237] Train net output #0: loss = 2.35338 (* 1 = 2.35338 loss) I0406 08:53:58.705474 5226 sgd_solver.cpp:105] Iteration 12216, lr = 0.01 I0406 08:54:04.009377 5226 solver.cpp:218] Iteration 12228 (2.2625 iter/s, 5.30386s/12 iters), loss = 1.78302 I0406 08:54:04.009424 5226 solver.cpp:237] Train net output #0: loss = 1.78302 (* 1 = 1.78302 loss) I0406 08:54:04.009431 5226 sgd_solver.cpp:105] Iteration 12228, lr = 0.01 I0406 08:54:08.534565 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:54:08.671589 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_12240.caffemodel I0406 08:54:11.702525 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_12240.solverstate I0406 08:54:14.013617 5226 solver.cpp:330] Iteration 12240, Testing net (#0) I0406 08:54:14.013639 5226 net.cpp:676] Ignoring source layer train-data I0406 08:54:18.138761 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:54:18.363211 5226 solver.cpp:397] Test net output #0: accuracy = 0.242034 I0406 08:54:18.363252 5226 solver.cpp:397] Test net output #1: loss = 3.67446 (* 1 = 3.67446 loss) I0406 08:54:18.497666 5226 solver.cpp:218] Iteration 12240 (0.828264 iter/s, 14.4881s/12 iters), loss = 2.28348 I0406 08:54:18.508846 5226 solver.cpp:237] Train net output #0: loss = 2.28348 (* 1 = 2.28348 loss) I0406 08:54:18.508870 5226 sgd_solver.cpp:105] Iteration 12240, lr = 0.01 I0406 08:54:22.789860 5226 solver.cpp:218] Iteration 12252 (2.80308 iter/s, 4.281s/12 iters), loss = 1.40058 I0406 08:54:22.789913 5226 solver.cpp:237] Train net output #0: loss = 1.40058 (* 1 = 1.40058 loss) I0406 08:54:22.789919 5226 sgd_solver.cpp:105] Iteration 12252, lr = 0.01 I0406 08:54:28.061539 5226 solver.cpp:218] Iteration 12264 (2.27636 iter/s, 5.27158s/12 iters), loss = 1.86653 I0406 08:54:28.061580 5226 solver.cpp:237] Train net output #0: loss = 1.86653 (* 1 = 1.86653 loss) I0406 08:54:28.061585 5226 sgd_solver.cpp:105] Iteration 12264, lr = 0.01 I0406 08:54:33.430259 5226 solver.cpp:218] Iteration 12276 (2.23521 iter/s, 5.36863s/12 iters), loss = 2.33644 I0406 08:54:33.430307 5226 solver.cpp:237] Train net output #0: loss = 2.33644 (* 1 = 2.33644 loss) I0406 08:54:33.430316 5226 sgd_solver.cpp:105] Iteration 12276, lr = 0.01 I0406 08:54:38.601475 5226 solver.cpp:218] Iteration 12288 (2.32058 iter/s, 5.17112s/12 iters), loss = 1.94653 I0406 08:54:38.601619 5226 solver.cpp:237] Train net output #0: loss = 1.94653 (* 1 = 1.94653 loss) I0406 08:54:38.601626 5226 sgd_solver.cpp:105] Iteration 12288, lr = 0.01 I0406 08:54:43.950040 5226 solver.cpp:218] Iteration 12300 (2.24367 iter/s, 5.34837s/12 iters), loss = 1.7984 I0406 08:54:43.950094 5226 solver.cpp:237] Train net output #0: loss = 1.7984 (* 1 = 1.7984 loss) I0406 08:54:43.950103 5226 sgd_solver.cpp:105] Iteration 12300, lr = 0.01 I0406 08:54:49.299345 5226 solver.cpp:218] Iteration 12312 (2.24332 iter/s, 5.34921s/12 iters), loss = 2.4692 I0406 08:54:49.305567 5226 solver.cpp:237] Train net output #0: loss = 2.4692 (* 1 = 2.4692 loss) I0406 08:54:49.305585 5226 sgd_solver.cpp:105] Iteration 12312, lr = 0.01 I0406 08:54:54.498692 5226 solver.cpp:218] Iteration 12324 (2.31076 iter/s, 5.19309s/12 iters), loss = 2.18558 I0406 08:54:54.498742 5226 solver.cpp:237] Train net output #0: loss = 2.18558 (* 1 = 2.18558 loss) I0406 08:54:54.498751 5226 sgd_solver.cpp:105] Iteration 12324, lr = 0.01 I0406 08:54:59.726038 5226 solver.cpp:218] Iteration 12336 (2.29566 iter/s, 5.22725s/12 iters), loss = 1.80057 I0406 08:54:59.726084 5226 solver.cpp:237] Train net output #0: loss = 1.80057 (* 1 = 1.80057 loss) I0406 08:54:59.726092 5226 sgd_solver.cpp:105] Iteration 12336, lr = 0.01 I0406 08:55:01.381563 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:55:01.715597 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_12342.caffemodel I0406 08:55:04.741385 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_12342.solverstate I0406 08:55:07.049391 5226 solver.cpp:330] Iteration 12342, Testing net (#0) I0406 08:55:07.049412 5226 net.cpp:676] Ignoring source layer train-data I0406 08:55:11.206604 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:55:11.459978 5226 solver.cpp:397] Test net output #0: accuracy = 0.210784 I0406 08:55:11.460011 5226 solver.cpp:397] Test net output #1: loss = 3.72294 (* 1 = 3.72294 loss) I0406 08:55:13.343698 5226 solver.cpp:218] Iteration 12348 (0.881218 iter/s, 13.6175s/12 iters), loss = 2.02862 I0406 08:55:13.343737 5226 solver.cpp:237] Train net output #0: loss = 2.02862 (* 1 = 2.02862 loss) I0406 08:55:13.343744 5226 sgd_solver.cpp:105] Iteration 12348, lr = 0.01 I0406 08:55:18.647181 5226 solver.cpp:218] Iteration 12360 (2.2627 iter/s, 5.3034s/12 iters), loss = 1.78042 I0406 08:55:18.647222 5226 solver.cpp:237] Train net output #0: loss = 1.78042 (* 1 = 1.78042 loss) I0406 08:55:18.647228 5226 sgd_solver.cpp:105] Iteration 12360, lr = 0.01 I0406 08:55:23.925997 5226 solver.cpp:218] Iteration 12372 (2.27327 iter/s, 5.27873s/12 iters), loss = 1.92886 I0406 08:55:23.926033 5226 solver.cpp:237] Train net output #0: loss = 1.92886 (* 1 = 1.92886 loss) I0406 08:55:23.926038 5226 sgd_solver.cpp:105] Iteration 12372, lr = 0.01 I0406 08:55:29.264027 5226 solver.cpp:218] Iteration 12384 (2.24806 iter/s, 5.33795s/12 iters), loss = 1.66089 I0406 08:55:29.264065 5226 solver.cpp:237] Train net output #0: loss = 1.66089 (* 1 = 1.66089 loss) I0406 08:55:29.264070 5226 sgd_solver.cpp:105] Iteration 12384, lr = 0.01 I0406 08:55:34.621311 5226 solver.cpp:218] Iteration 12396 (2.23998 iter/s, 5.3572s/12 iters), loss = 1.8297 I0406 08:55:34.621346 5226 solver.cpp:237] Train net output #0: loss = 1.8297 (* 1 = 1.8297 loss) I0406 08:55:34.621352 5226 sgd_solver.cpp:105] Iteration 12396, lr = 0.01 I0406 08:55:39.919286 5226 solver.cpp:218] Iteration 12408 (2.26505 iter/s, 5.29789s/12 iters), loss = 2.10458 I0406 08:55:39.919325 5226 solver.cpp:237] Train net output #0: loss = 2.10458 (* 1 = 2.10458 loss) I0406 08:55:39.919330 5226 sgd_solver.cpp:105] Iteration 12408, lr = 0.01 I0406 08:55:45.282624 5226 solver.cpp:218] Iteration 12420 (2.23745 iter/s, 5.36325s/12 iters), loss = 1.9067 I0406 08:55:45.282791 5226 solver.cpp:237] Train net output #0: loss = 1.9067 (* 1 = 1.9067 loss) I0406 08:55:45.282800 5226 sgd_solver.cpp:105] Iteration 12420, lr = 0.01 I0406 08:55:50.699417 5226 solver.cpp:218] Iteration 12432 (2.21542 iter/s, 5.41658s/12 iters), loss = 2.26223 I0406 08:55:50.699455 5226 solver.cpp:237] Train net output #0: loss = 2.26223 (* 1 = 2.26223 loss) I0406 08:55:50.699461 5226 sgd_solver.cpp:105] Iteration 12432, lr = 0.01 I0406 08:55:54.801519 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:55:55.537760 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_12444.caffemodel I0406 08:55:58.541608 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_12444.solverstate I0406 08:56:00.839605 5226 solver.cpp:330] Iteration 12444, Testing net (#0) I0406 08:56:00.839622 5226 net.cpp:676] Ignoring source layer train-data I0406 08:56:04.903287 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:56:05.200853 5226 solver.cpp:397] Test net output #0: accuracy = 0.240196 I0406 08:56:05.200899 5226 solver.cpp:397] Test net output #1: loss = 3.71219 (* 1 = 3.71219 loss) I0406 08:56:05.337018 5226 solver.cpp:218] Iteration 12444 (0.819815 iter/s, 14.6375s/12 iters), loss = 2.07233 I0406 08:56:05.338582 5226 solver.cpp:237] Train net output #0: loss = 2.07233 (* 1 = 2.07233 loss) I0406 08:56:05.338595 5226 sgd_solver.cpp:105] Iteration 12444, lr = 0.01 I0406 08:56:09.628427 5226 solver.cpp:218] Iteration 12456 (2.79733 iter/s, 4.28981s/12 iters), loss = 2.00891 I0406 08:56:09.628466 5226 solver.cpp:237] Train net output #0: loss = 2.00891 (* 1 = 2.00891 loss) I0406 08:56:09.628473 5226 sgd_solver.cpp:105] Iteration 12456, lr = 0.01 I0406 08:56:13.858142 5226 blocking_queue.cpp:49] Waiting for data I0406 08:56:14.926502 5226 solver.cpp:218] Iteration 12468 (2.26501 iter/s, 5.29798s/12 iters), loss = 1.96054 I0406 08:56:14.926553 5226 solver.cpp:237] Train net output #0: loss = 1.96054 (* 1 = 1.96054 loss) I0406 08:56:14.926561 5226 sgd_solver.cpp:105] Iteration 12468, lr = 0.01 I0406 08:56:20.228616 5226 solver.cpp:218] Iteration 12480 (2.26329 iter/s, 5.30202s/12 iters), loss = 2.04537 I0406 08:56:20.228696 5226 solver.cpp:237] Train net output #0: loss = 2.04537 (* 1 = 2.04537 loss) I0406 08:56:20.228703 5226 sgd_solver.cpp:105] Iteration 12480, lr = 0.01 I0406 08:56:25.330332 5226 solver.cpp:218] Iteration 12492 (2.35221 iter/s, 5.10159s/12 iters), loss = 1.93161 I0406 08:56:25.330382 5226 solver.cpp:237] Train net output #0: loss = 1.93161 (* 1 = 1.93161 loss) I0406 08:56:25.330391 5226 sgd_solver.cpp:105] Iteration 12492, lr = 0.01 I0406 08:56:30.625582 5226 solver.cpp:218] Iteration 12504 (2.26622 iter/s, 5.29515s/12 iters), loss = 2.0567 I0406 08:56:30.625633 5226 solver.cpp:237] Train net output #0: loss = 2.0567 (* 1 = 2.0567 loss) I0406 08:56:30.625640 5226 sgd_solver.cpp:105] Iteration 12504, lr = 0.01 I0406 08:56:35.653348 5226 solver.cpp:218] Iteration 12516 (2.38679 iter/s, 5.02767s/12 iters), loss = 1.86373 I0406 08:56:35.653388 5226 solver.cpp:237] Train net output #0: loss = 1.86373 (* 1 = 1.86373 loss) I0406 08:56:35.653393 5226 sgd_solver.cpp:105] Iteration 12516, lr = 0.01 I0406 08:56:40.862766 5226 solver.cpp:218] Iteration 12528 (2.30356 iter/s, 5.20933s/12 iters), loss = 1.92142 I0406 08:56:40.862807 5226 solver.cpp:237] Train net output #0: loss = 1.92142 (* 1 = 1.92142 loss) I0406 08:56:40.862812 5226 sgd_solver.cpp:105] Iteration 12528, lr = 0.01 I0406 08:56:46.062629 5226 solver.cpp:218] Iteration 12540 (2.30779 iter/s, 5.19977s/12 iters), loss = 2.30835 I0406 08:56:46.062669 5226 solver.cpp:237] Train net output #0: loss = 2.30835 (* 1 = 2.30835 loss) I0406 08:56:46.062674 5226 sgd_solver.cpp:105] Iteration 12540, lr = 0.01 I0406 08:56:46.984927 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:56:48.059630 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_12546.caffemodel I0406 08:56:51.104948 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_12546.solverstate I0406 08:56:53.420801 5226 solver.cpp:330] Iteration 12546, Testing net (#0) I0406 08:56:53.420821 5226 net.cpp:676] Ignoring source layer train-data I0406 08:56:57.460188 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:56:57.812516 5226 solver.cpp:397] Test net output #0: accuracy = 0.199755 I0406 08:56:57.812552 5226 solver.cpp:397] Test net output #1: loss = 3.89406 (* 1 = 3.89406 loss) I0406 08:56:59.699388 5226 solver.cpp:218] Iteration 12552 (0.879984 iter/s, 13.6366s/12 iters), loss = 2.40158 I0406 08:56:59.699425 5226 solver.cpp:237] Train net output #0: loss = 2.40158 (* 1 = 2.40158 loss) I0406 08:56:59.699431 5226 sgd_solver.cpp:105] Iteration 12552, lr = 0.01 I0406 08:57:04.692265 5226 solver.cpp:218] Iteration 12564 (2.40346 iter/s, 4.9928s/12 iters), loss = 1.93384 I0406 08:57:04.692302 5226 solver.cpp:237] Train net output #0: loss = 1.93384 (* 1 = 1.93384 loss) I0406 08:57:04.692307 5226 sgd_solver.cpp:105] Iteration 12564, lr = 0.01 I0406 08:57:09.812453 5226 solver.cpp:218] Iteration 12576 (2.3437 iter/s, 5.1201s/12 iters), loss = 2.54624 I0406 08:57:09.812507 5226 solver.cpp:237] Train net output #0: loss = 2.54624 (* 1 = 2.54624 loss) I0406 08:57:09.812516 5226 sgd_solver.cpp:105] Iteration 12576, lr = 0.01 I0406 08:57:15.190047 5226 solver.cpp:218] Iteration 12588 (2.23152 iter/s, 5.37749s/12 iters), loss = 2.34897 I0406 08:57:15.190084 5226 solver.cpp:237] Train net output #0: loss = 2.34897 (* 1 = 2.34897 loss) I0406 08:57:15.190090 5226 sgd_solver.cpp:105] Iteration 12588, lr = 0.01 I0406 08:57:20.326215 5226 solver.cpp:218] Iteration 12600 (2.33641 iter/s, 5.13608s/12 iters), loss = 1.67807 I0406 08:57:20.326261 5226 solver.cpp:237] Train net output #0: loss = 1.67807 (* 1 = 1.67807 loss) I0406 08:57:20.326269 5226 sgd_solver.cpp:105] Iteration 12600, lr = 0.01 I0406 08:57:25.650372 5226 solver.cpp:218] Iteration 12612 (2.25392 iter/s, 5.32406s/12 iters), loss = 1.93017 I0406 08:57:25.650496 5226 solver.cpp:237] Train net output #0: loss = 1.93017 (* 1 = 1.93017 loss) I0406 08:57:25.650507 5226 sgd_solver.cpp:105] Iteration 12612, lr = 0.01 I0406 08:57:30.959216 5226 solver.cpp:218] Iteration 12624 (2.26045 iter/s, 5.30867s/12 iters), loss = 2.17282 I0406 08:57:30.959260 5226 solver.cpp:237] Train net output #0: loss = 2.17282 (* 1 = 2.17282 loss) I0406 08:57:30.959267 5226 sgd_solver.cpp:105] Iteration 12624, lr = 0.01 I0406 08:57:36.210465 5226 solver.cpp:218] Iteration 12636 (2.28521 iter/s, 5.25115s/12 iters), loss = 1.70949 I0406 08:57:36.210505 5226 solver.cpp:237] Train net output #0: loss = 1.70949 (* 1 = 1.70949 loss) I0406 08:57:36.210510 5226 sgd_solver.cpp:105] Iteration 12636, lr = 0.01 I0406 08:57:39.295540 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:57:40.936054 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_12648.caffemodel I0406 08:57:44.759793 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_12648.solverstate I0406 08:57:47.094141 5226 solver.cpp:330] Iteration 12648, Testing net (#0) I0406 08:57:47.094159 5226 net.cpp:676] Ignoring source layer train-data I0406 08:57:51.291172 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:57:51.681748 5226 solver.cpp:397] Test net output #0: accuracy = 0.210172 I0406 08:57:51.681774 5226 solver.cpp:397] Test net output #1: loss = 3.81345 (* 1 = 3.81345 loss) I0406 08:57:51.822319 5226 solver.cpp:218] Iteration 12648 (0.768654 iter/s, 15.6117s/12 iters), loss = 1.82351 I0406 08:57:51.822367 5226 solver.cpp:237] Train net output #0: loss = 1.82351 (* 1 = 1.82351 loss) I0406 08:57:51.822374 5226 sgd_solver.cpp:105] Iteration 12648, lr = 0.01 I0406 08:57:55.986485 5226 solver.cpp:218] Iteration 12660 (2.88179 iter/s, 4.16408s/12 iters), loss = 1.79833 I0406 08:57:55.986598 5226 solver.cpp:237] Train net output #0: loss = 1.79833 (* 1 = 1.79833 loss) I0406 08:57:55.986604 5226 sgd_solver.cpp:105] Iteration 12660, lr = 0.01 I0406 08:58:01.163935 5226 solver.cpp:218] Iteration 12672 (2.31781 iter/s, 5.17729s/12 iters), loss = 2.13792 I0406 08:58:01.163982 5226 solver.cpp:237] Train net output #0: loss = 2.13792 (* 1 = 2.13792 loss) I0406 08:58:01.163990 5226 sgd_solver.cpp:105] Iteration 12672, lr = 0.01 I0406 08:58:06.449580 5226 solver.cpp:218] Iteration 12684 (2.27034 iter/s, 5.28555s/12 iters), loss = 2.03929 I0406 08:58:06.449621 5226 solver.cpp:237] Train net output #0: loss = 2.03929 (* 1 = 2.03929 loss) I0406 08:58:06.449626 5226 sgd_solver.cpp:105] Iteration 12684, lr = 0.01 I0406 08:58:11.833459 5226 solver.cpp:218] Iteration 12696 (2.22891 iter/s, 5.38379s/12 iters), loss = 2.30356 I0406 08:58:11.833506 5226 solver.cpp:237] Train net output #0: loss = 2.30356 (* 1 = 2.30356 loss) I0406 08:58:11.833514 5226 sgd_solver.cpp:105] Iteration 12696, lr = 0.01 I0406 08:58:17.075721 5226 solver.cpp:218] Iteration 12708 (2.28913 iter/s, 5.24216s/12 iters), loss = 2.05397 I0406 08:58:17.075762 5226 solver.cpp:237] Train net output #0: loss = 2.05397 (* 1 = 2.05397 loss) I0406 08:58:17.075769 5226 sgd_solver.cpp:105] Iteration 12708, lr = 0.01 I0406 08:58:22.359730 5226 solver.cpp:218] Iteration 12720 (2.27104 iter/s, 5.28392s/12 iters), loss = 2.47962 I0406 08:58:22.359769 5226 solver.cpp:237] Train net output #0: loss = 2.47962 (* 1 = 2.47962 loss) I0406 08:58:22.359776 5226 sgd_solver.cpp:105] Iteration 12720, lr = 0.01 I0406 08:58:27.598197 5226 solver.cpp:218] Iteration 12732 (2.29078 iter/s, 5.23838s/12 iters), loss = 2.39226 I0406 08:58:27.598467 5226 solver.cpp:237] Train net output #0: loss = 2.39226 (* 1 = 2.39226 loss) I0406 08:58:27.598474 5226 sgd_solver.cpp:105] Iteration 12732, lr = 0.01 I0406 08:58:32.732507 5226 solver.cpp:218] Iteration 12744 (2.33736 iter/s, 5.13399s/12 iters), loss = 2.38341 I0406 08:58:32.732549 5226 solver.cpp:237] Train net output #0: loss = 2.38341 (* 1 = 2.38341 loss) I0406 08:58:32.732556 5226 sgd_solver.cpp:105] Iteration 12744, lr = 0.01 I0406 08:58:32.933468 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:58:34.807253 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_12750.caffemodel I0406 08:58:37.836956 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_12750.solverstate I0406 08:58:40.152251 5226 solver.cpp:330] Iteration 12750, Testing net (#0) I0406 08:58:40.152271 5226 net.cpp:676] Ignoring source layer train-data I0406 08:58:44.336971 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:58:44.754827 5226 solver.cpp:397] Test net output #0: accuracy = 0.206495 I0406 08:58:44.754853 5226 solver.cpp:397] Test net output #1: loss = 3.89849 (* 1 = 3.89849 loss) I0406 08:58:46.635798 5226 solver.cpp:218] Iteration 12756 (0.863114 iter/s, 13.9032s/12 iters), loss = 2.03739 I0406 08:58:46.635843 5226 solver.cpp:237] Train net output #0: loss = 2.03739 (* 1 = 2.03739 loss) I0406 08:58:46.635849 5226 sgd_solver.cpp:105] Iteration 12756, lr = 0.01 I0406 08:58:52.024415 5226 solver.cpp:218] Iteration 12768 (2.22696 iter/s, 5.38852s/12 iters), loss = 1.84004 I0406 08:58:52.024462 5226 solver.cpp:237] Train net output #0: loss = 1.84004 (* 1 = 1.84004 loss) I0406 08:58:52.024471 5226 sgd_solver.cpp:105] Iteration 12768, lr = 0.01 I0406 08:58:57.289966 5226 solver.cpp:218] Iteration 12780 (2.27901 iter/s, 5.26545s/12 iters), loss = 1.90052 I0406 08:58:57.290017 5226 solver.cpp:237] Train net output #0: loss = 1.90052 (* 1 = 1.90052 loss) I0406 08:58:57.290026 5226 sgd_solver.cpp:105] Iteration 12780, lr = 0.01 I0406 08:59:02.556270 5226 solver.cpp:218] Iteration 12792 (2.27868 iter/s, 5.26621s/12 iters), loss = 2.2727 I0406 08:59:02.556380 5226 solver.cpp:237] Train net output #0: loss = 2.2727 (* 1 = 2.2727 loss) I0406 08:59:02.556387 5226 sgd_solver.cpp:105] Iteration 12792, lr = 0.01 I0406 08:59:08.030645 5226 solver.cpp:218] Iteration 12804 (2.1921 iter/s, 5.47421s/12 iters), loss = 1.71102 I0406 08:59:08.030687 5226 solver.cpp:237] Train net output #0: loss = 1.71102 (* 1 = 1.71102 loss) I0406 08:59:08.030692 5226 sgd_solver.cpp:105] Iteration 12804, lr = 0.01 I0406 08:59:13.457456 5226 solver.cpp:218] Iteration 12816 (2.21128 iter/s, 5.42672s/12 iters), loss = 1.67122 I0406 08:59:13.457513 5226 solver.cpp:237] Train net output #0: loss = 1.67122 (* 1 = 1.67122 loss) I0406 08:59:13.457521 5226 sgd_solver.cpp:105] Iteration 12816, lr = 0.01 I0406 08:59:18.853266 5226 solver.cpp:218] Iteration 12828 (2.22399 iter/s, 5.39571s/12 iters), loss = 2.01368 I0406 08:59:18.853302 5226 solver.cpp:237] Train net output #0: loss = 2.01368 (* 1 = 2.01368 loss) I0406 08:59:18.853307 5226 sgd_solver.cpp:105] Iteration 12828, lr = 0.01 I0406 08:59:24.041209 5226 solver.cpp:218] Iteration 12840 (2.31309 iter/s, 5.18786s/12 iters), loss = 2.36339 I0406 08:59:24.041255 5226 solver.cpp:237] Train net output #0: loss = 2.36339 (* 1 = 2.36339 loss) I0406 08:59:24.041261 5226 sgd_solver.cpp:105] Iteration 12840, lr = 0.01 I0406 08:59:26.539247 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:59:28.799561 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_12852.caffemodel I0406 08:59:31.975042 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_12852.solverstate I0406 08:59:34.300062 5226 solver.cpp:330] Iteration 12852, Testing net (#0) I0406 08:59:34.300978 5226 net.cpp:676] Ignoring source layer train-data I0406 08:59:38.344899 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 08:59:38.850325 5226 solver.cpp:397] Test net output #0: accuracy = 0.228554 I0406 08:59:38.850356 5226 solver.cpp:397] Test net output #1: loss = 3.8433 (* 1 = 3.8433 loss) I0406 08:59:38.991447 5226 solver.cpp:218] Iteration 12852 (0.802671 iter/s, 14.9501s/12 iters), loss = 2.05897 I0406 08:59:38.991494 5226 solver.cpp:237] Train net output #0: loss = 2.05897 (* 1 = 2.05897 loss) I0406 08:59:38.991503 5226 sgd_solver.cpp:105] Iteration 12852, lr = 0.01 I0406 08:59:43.207427 5226 solver.cpp:218] Iteration 12864 (2.84637 iter/s, 4.21589s/12 iters), loss = 2.38214 I0406 08:59:43.207466 5226 solver.cpp:237] Train net output #0: loss = 2.38214 (* 1 = 2.38214 loss) I0406 08:59:43.207473 5226 sgd_solver.cpp:105] Iteration 12864, lr = 0.01 I0406 08:59:48.704298 5226 solver.cpp:218] Iteration 12876 (2.18309 iter/s, 5.49678s/12 iters), loss = 2.61828 I0406 08:59:48.704345 5226 solver.cpp:237] Train net output #0: loss = 2.61828 (* 1 = 2.61828 loss) I0406 08:59:48.704353 5226 sgd_solver.cpp:105] Iteration 12876, lr = 0.01 I0406 08:59:54.035287 5226 solver.cpp:218] Iteration 12888 (2.25103 iter/s, 5.33089s/12 iters), loss = 2.34082 I0406 08:59:54.035337 5226 solver.cpp:237] Train net output #0: loss = 2.34082 (* 1 = 2.34082 loss) I0406 08:59:54.035346 5226 sgd_solver.cpp:105] Iteration 12888, lr = 0.01 I0406 08:59:59.464380 5226 solver.cpp:218] Iteration 12900 (2.21035 iter/s, 5.42899s/12 iters), loss = 2.30015 I0406 08:59:59.464437 5226 solver.cpp:237] Train net output #0: loss = 2.30015 (* 1 = 2.30015 loss) I0406 08:59:59.464444 5226 sgd_solver.cpp:105] Iteration 12900, lr = 0.01 I0406 09:00:04.838977 5226 solver.cpp:218] Iteration 12912 (2.23277 iter/s, 5.3745s/12 iters), loss = 2.0681 I0406 09:00:04.839087 5226 solver.cpp:237] Train net output #0: loss = 2.0681 (* 1 = 2.0681 loss) I0406 09:00:04.839094 5226 sgd_solver.cpp:105] Iteration 12912, lr = 0.01 I0406 09:00:10.167570 5226 solver.cpp:218] Iteration 12924 (2.25207 iter/s, 5.32843s/12 iters), loss = 1.97818 I0406 09:00:10.167621 5226 solver.cpp:237] Train net output #0: loss = 1.97818 (* 1 = 1.97818 loss) I0406 09:00:10.167630 5226 sgd_solver.cpp:105] Iteration 12924, lr = 0.01 I0406 09:00:15.536777 5226 solver.cpp:218] Iteration 12936 (2.23501 iter/s, 5.36911s/12 iters), loss = 1.9848 I0406 09:00:15.536829 5226 solver.cpp:237] Train net output #0: loss = 1.9848 (* 1 = 1.9848 loss) I0406 09:00:15.536837 5226 sgd_solver.cpp:105] Iteration 12936, lr = 0.01 I0406 09:00:20.264463 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:00:20.856742 5226 solver.cpp:218] Iteration 12948 (2.2557 iter/s, 5.31987s/12 iters), loss = 2.13956 I0406 09:00:20.856799 5226 solver.cpp:237] Train net output #0: loss = 2.13956 (* 1 = 2.13956 loss) I0406 09:00:20.856808 5226 sgd_solver.cpp:105] Iteration 12948, lr = 0.01 I0406 09:00:23.122495 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_12954.caffemodel I0406 09:00:26.181989 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_12954.solverstate I0406 09:00:28.496124 5226 solver.cpp:330] Iteration 12954, Testing net (#0) I0406 09:00:28.496143 5226 net.cpp:676] Ignoring source layer train-data I0406 09:00:32.423219 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:00:32.975376 5226 solver.cpp:397] Test net output #0: accuracy = 0.226716 I0406 09:00:32.975414 5226 solver.cpp:397] Test net output #1: loss = 3.83009 (* 1 = 3.83009 loss) I0406 09:00:34.929587 5226 solver.cpp:218] Iteration 12960 (0.852716 iter/s, 14.0727s/12 iters), loss = 2.02595 I0406 09:00:34.929744 5226 solver.cpp:237] Train net output #0: loss = 2.02595 (* 1 = 2.02595 loss) I0406 09:00:34.929754 5226 sgd_solver.cpp:105] Iteration 12960, lr = 0.01 I0406 09:00:40.274217 5226 solver.cpp:218] Iteration 12972 (2.24533 iter/s, 5.34443s/12 iters), loss = 1.85272 I0406 09:00:40.274266 5226 solver.cpp:237] Train net output #0: loss = 1.85272 (* 1 = 1.85272 loss) I0406 09:00:40.274276 5226 sgd_solver.cpp:105] Iteration 12972, lr = 0.01 I0406 09:00:45.675988 5226 solver.cpp:218] Iteration 12984 (2.22154 iter/s, 5.40167s/12 iters), loss = 1.9954 I0406 09:00:45.676039 5226 solver.cpp:237] Train net output #0: loss = 1.9954 (* 1 = 1.9954 loss) I0406 09:00:45.676048 5226 sgd_solver.cpp:105] Iteration 12984, lr = 0.01 I0406 09:00:50.995613 5226 solver.cpp:218] Iteration 12996 (2.25584 iter/s, 5.31953s/12 iters), loss = 1.96171 I0406 09:00:50.995654 5226 solver.cpp:237] Train net output #0: loss = 1.96171 (* 1 = 1.96171 loss) I0406 09:00:50.995659 5226 sgd_solver.cpp:105] Iteration 12996, lr = 0.01 I0406 09:00:56.387122 5226 solver.cpp:218] Iteration 13008 (2.22576 iter/s, 5.39142s/12 iters), loss = 2.28275 I0406 09:00:56.387176 5226 solver.cpp:237] Train net output #0: loss = 2.28275 (* 1 = 2.28275 loss) I0406 09:00:56.387184 5226 sgd_solver.cpp:105] Iteration 13008, lr = 0.01 I0406 09:01:01.810196 5226 solver.cpp:218] Iteration 13020 (2.21281 iter/s, 5.42297s/12 iters), loss = 2.59514 I0406 09:01:01.810236 5226 solver.cpp:237] Train net output #0: loss = 2.59514 (* 1 = 2.59514 loss) I0406 09:01:01.810242 5226 sgd_solver.cpp:105] Iteration 13020, lr = 0.01 I0406 09:01:06.956879 5226 solver.cpp:218] Iteration 13032 (2.33164 iter/s, 5.1466s/12 iters), loss = 2.2502 I0406 09:01:06.956995 5226 solver.cpp:237] Train net output #0: loss = 2.2502 (* 1 = 2.2502 loss) I0406 09:01:06.957001 5226 sgd_solver.cpp:105] Iteration 13032, lr = 0.01 I0406 09:01:12.347457 5226 solver.cpp:218] Iteration 13044 (2.22617 iter/s, 5.39042s/12 iters), loss = 2.58961 I0406 09:01:12.347492 5226 solver.cpp:237] Train net output #0: loss = 2.58961 (* 1 = 2.58961 loss) I0406 09:01:12.347497 5226 sgd_solver.cpp:105] Iteration 13044, lr = 0.01 I0406 09:01:14.220393 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:01:17.242931 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_13056.caffemodel I0406 09:01:20.306138 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_13056.solverstate I0406 09:01:22.603654 5226 solver.cpp:330] Iteration 13056, Testing net (#0) I0406 09:01:22.603675 5226 net.cpp:676] Ignoring source layer train-data I0406 09:01:26.578012 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:01:27.131258 5226 solver.cpp:397] Test net output #0: accuracy = 0.219975 I0406 09:01:27.131283 5226 solver.cpp:397] Test net output #1: loss = 3.81154 (* 1 = 3.81154 loss) I0406 09:01:27.271960 5226 solver.cpp:218] Iteration 13056 (0.804055 iter/s, 14.9244s/12 iters), loss = 2.36838 I0406 09:01:27.271998 5226 solver.cpp:237] Train net output #0: loss = 2.36838 (* 1 = 2.36838 loss) I0406 09:01:27.272003 5226 sgd_solver.cpp:105] Iteration 13056, lr = 0.01 I0406 09:01:31.507488 5226 solver.cpp:218] Iteration 13068 (2.83323 iter/s, 4.23545s/12 iters), loss = 1.82797 I0406 09:01:31.507526 5226 solver.cpp:237] Train net output #0: loss = 1.82797 (* 1 = 1.82797 loss) I0406 09:01:31.507531 5226 sgd_solver.cpp:105] Iteration 13068, lr = 0.01 I0406 09:01:37.022451 5226 solver.cpp:218] Iteration 13080 (2.17593 iter/s, 5.51487s/12 iters), loss = 1.78876 I0406 09:01:37.022584 5226 solver.cpp:237] Train net output #0: loss = 1.78876 (* 1 = 1.78876 loss) I0406 09:01:37.022593 5226 sgd_solver.cpp:105] Iteration 13080, lr = 0.01 I0406 09:01:42.455754 5226 solver.cpp:218] Iteration 13092 (2.20867 iter/s, 5.43313s/12 iters), loss = 1.90038 I0406 09:01:42.455793 5226 solver.cpp:237] Train net output #0: loss = 1.90038 (* 1 = 1.90038 loss) I0406 09:01:42.455799 5226 sgd_solver.cpp:105] Iteration 13092, lr = 0.01 I0406 09:01:47.490206 5226 solver.cpp:218] Iteration 13104 (2.38362 iter/s, 5.03436s/12 iters), loss = 2.18204 I0406 09:01:47.490247 5226 solver.cpp:237] Train net output #0: loss = 2.18204 (* 1 = 2.18204 loss) I0406 09:01:47.490254 5226 sgd_solver.cpp:105] Iteration 13104, lr = 0.01 I0406 09:01:52.895884 5226 solver.cpp:218] Iteration 13116 (2.21992 iter/s, 5.40559s/12 iters), loss = 1.95494 I0406 09:01:52.895920 5226 solver.cpp:237] Train net output #0: loss = 1.95494 (* 1 = 1.95494 loss) I0406 09:01:52.895926 5226 sgd_solver.cpp:105] Iteration 13116, lr = 0.01 I0406 09:01:58.123517 5226 solver.cpp:218] Iteration 13128 (2.29553 iter/s, 5.22755s/12 iters), loss = 2.51104 I0406 09:01:58.123569 5226 solver.cpp:237] Train net output #0: loss = 2.51104 (* 1 = 2.51104 loss) I0406 09:01:58.123577 5226 sgd_solver.cpp:105] Iteration 13128, lr = 0.01 I0406 09:02:03.396598 5226 solver.cpp:218] Iteration 13140 (2.27575 iter/s, 5.27298s/12 iters), loss = 2.20705 I0406 09:02:03.396648 5226 solver.cpp:237] Train net output #0: loss = 2.20705 (* 1 = 2.20705 loss) I0406 09:02:03.396656 5226 sgd_solver.cpp:105] Iteration 13140, lr = 0.01 I0406 09:02:07.565629 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:02:08.811480 5226 solver.cpp:218] Iteration 13152 (2.21616 iter/s, 5.41478s/12 iters), loss = 2.71286 I0406 09:02:08.811535 5226 solver.cpp:237] Train net output #0: loss = 2.71286 (* 1 = 2.71286 loss) I0406 09:02:08.811545 5226 sgd_solver.cpp:105] Iteration 13152, lr = 0.01 I0406 09:02:10.970021 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_13158.caffemodel I0406 09:02:13.991351 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_13158.solverstate I0406 09:02:16.299675 5226 solver.cpp:330] Iteration 13158, Testing net (#0) I0406 09:02:16.299695 5226 net.cpp:676] Ignoring source layer train-data I0406 09:02:19.897347 5226 blocking_queue.cpp:49] Waiting for data I0406 09:02:20.122828 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:02:20.685557 5226 solver.cpp:397] Test net output #0: accuracy = 0.215686 I0406 09:02:20.685595 5226 solver.cpp:397] Test net output #1: loss = 3.76541 (* 1 = 3.76541 loss) I0406 09:02:22.625780 5226 solver.cpp:218] Iteration 13164 (0.868675 iter/s, 13.8141s/12 iters), loss = 2.09108 I0406 09:02:22.625819 5226 solver.cpp:237] Train net output #0: loss = 2.09108 (* 1 = 2.09108 loss) I0406 09:02:22.625825 5226 sgd_solver.cpp:105] Iteration 13164, lr = 0.01 I0406 09:02:27.822975 5226 solver.cpp:218] Iteration 13176 (2.30898 iter/s, 5.1971s/12 iters), loss = 1.87697 I0406 09:02:27.823022 5226 solver.cpp:237] Train net output #0: loss = 1.87697 (* 1 = 1.87697 loss) I0406 09:02:27.823030 5226 sgd_solver.cpp:105] Iteration 13176, lr = 0.01 I0406 09:02:33.051116 5226 solver.cpp:218] Iteration 13188 (2.29531 iter/s, 5.22804s/12 iters), loss = 1.73069 I0406 09:02:33.051175 5226 solver.cpp:237] Train net output #0: loss = 1.73069 (* 1 = 1.73069 loss) I0406 09:02:33.051184 5226 sgd_solver.cpp:105] Iteration 13188, lr = 0.01 I0406 09:02:38.401268 5226 solver.cpp:218] Iteration 13200 (2.24297 iter/s, 5.35005s/12 iters), loss = 2.105 I0406 09:02:38.403254 5226 solver.cpp:237] Train net output #0: loss = 2.105 (* 1 = 2.105 loss) I0406 09:02:38.403265 5226 sgd_solver.cpp:105] Iteration 13200, lr = 0.01 I0406 09:02:43.525346 5226 solver.cpp:218] Iteration 13212 (2.34281 iter/s, 5.12205s/12 iters), loss = 1.99243 I0406 09:02:43.525393 5226 solver.cpp:237] Train net output #0: loss = 1.99243 (* 1 = 1.99243 loss) I0406 09:02:43.525400 5226 sgd_solver.cpp:105] Iteration 13212, lr = 0.01 I0406 09:02:48.898905 5226 solver.cpp:218] Iteration 13224 (2.2332 iter/s, 5.37347s/12 iters), loss = 2.5856 I0406 09:02:48.898943 5226 solver.cpp:237] Train net output #0: loss = 2.5856 (* 1 = 2.5856 loss) I0406 09:02:48.898948 5226 sgd_solver.cpp:105] Iteration 13224, lr = 0.01 I0406 09:02:54.243583 5226 solver.cpp:218] Iteration 13236 (2.24526 iter/s, 5.34459s/12 iters), loss = 2.48104 I0406 09:02:54.243635 5226 solver.cpp:237] Train net output #0: loss = 2.48104 (* 1 = 2.48104 loss) I0406 09:02:54.243643 5226 sgd_solver.cpp:105] Iteration 13236, lr = 0.01 I0406 09:02:59.618660 5226 solver.cpp:218] Iteration 13248 (2.23257 iter/s, 5.37498s/12 iters), loss = 2.18527 I0406 09:02:59.618697 5226 solver.cpp:237] Train net output #0: loss = 2.18527 (* 1 = 2.18527 loss) I0406 09:02:59.618702 5226 sgd_solver.cpp:105] Iteration 13248, lr = 0.01 I0406 09:03:00.657335 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:03:04.438789 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_13260.caffemodel I0406 09:03:08.596582 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_13260.solverstate I0406 09:03:11.041599 5226 solver.cpp:330] Iteration 13260, Testing net (#0) I0406 09:03:11.041617 5226 net.cpp:676] Ignoring source layer train-data I0406 09:03:14.718336 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:03:15.336138 5226 solver.cpp:397] Test net output #0: accuracy = 0.193627 I0406 09:03:15.336166 5226 solver.cpp:397] Test net output #1: loss = 3.99448 (* 1 = 3.99448 loss) I0406 09:03:15.476671 5226 solver.cpp:218] Iteration 13260 (0.756722 iter/s, 15.8579s/12 iters), loss = 2.17935 I0406 09:03:15.476711 5226 solver.cpp:237] Train net output #0: loss = 2.17935 (* 1 = 2.17935 loss) I0406 09:03:15.476716 5226 sgd_solver.cpp:105] Iteration 13260, lr = 0.01 I0406 09:03:19.759227 5226 solver.cpp:218] Iteration 13272 (2.80212 iter/s, 4.28247s/12 iters), loss = 2.30286 I0406 09:03:19.759276 5226 solver.cpp:237] Train net output #0: loss = 2.30286 (* 1 = 2.30286 loss) I0406 09:03:19.759284 5226 sgd_solver.cpp:105] Iteration 13272, lr = 0.01 I0406 09:03:25.018007 5226 solver.cpp:218] Iteration 13284 (2.28194 iter/s, 5.25868s/12 iters), loss = 2.85717 I0406 09:03:25.018060 5226 solver.cpp:237] Train net output #0: loss = 2.85717 (* 1 = 2.85717 loss) I0406 09:03:25.018069 5226 sgd_solver.cpp:105] Iteration 13284, lr = 0.01 I0406 09:03:30.364535 5226 solver.cpp:218] Iteration 13296 (2.24449 iter/s, 5.34643s/12 iters), loss = 2.16949 I0406 09:03:30.364570 5226 solver.cpp:237] Train net output #0: loss = 2.16949 (* 1 = 2.16949 loss) I0406 09:03:30.364575 5226 sgd_solver.cpp:105] Iteration 13296, lr = 0.01 I0406 09:03:35.632984 5226 solver.cpp:218] Iteration 13308 (2.27775 iter/s, 5.26837s/12 iters), loss = 1.94612 I0406 09:03:35.633021 5226 solver.cpp:237] Train net output #0: loss = 1.94612 (* 1 = 1.94612 loss) I0406 09:03:35.633026 5226 sgd_solver.cpp:105] Iteration 13308, lr = 0.01 I0406 09:03:40.697268 5226 solver.cpp:218] Iteration 13320 (2.36957 iter/s, 5.0642s/12 iters), loss = 2.41577 I0406 09:03:40.697402 5226 solver.cpp:237] Train net output #0: loss = 2.41577 (* 1 = 2.41577 loss) I0406 09:03:40.697408 5226 sgd_solver.cpp:105] Iteration 13320, lr = 0.01 I0406 09:03:46.089785 5226 solver.cpp:218] Iteration 13332 (2.22538 iter/s, 5.39233s/12 iters), loss = 2.12187 I0406 09:03:46.089835 5226 solver.cpp:237] Train net output #0: loss = 2.12187 (* 1 = 2.12187 loss) I0406 09:03:46.089843 5226 sgd_solver.cpp:105] Iteration 13332, lr = 0.01 I0406 09:03:51.420637 5226 solver.cpp:218] Iteration 13344 (2.25109 iter/s, 5.33074s/12 iters), loss = 2.26784 I0406 09:03:51.420689 5226 solver.cpp:237] Train net output #0: loss = 2.26784 (* 1 = 2.26784 loss) I0406 09:03:51.420699 5226 sgd_solver.cpp:105] Iteration 13344, lr = 0.01 I0406 09:03:54.765344 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:03:56.772083 5226 solver.cpp:218] Iteration 13356 (2.24243 iter/s, 5.35134s/12 iters), loss = 1.95577 I0406 09:03:56.772119 5226 solver.cpp:237] Train net output #0: loss = 1.95577 (* 1 = 1.95577 loss) I0406 09:03:56.772125 5226 sgd_solver.cpp:105] Iteration 13356, lr = 0.01 I0406 09:03:58.874276 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_13362.caffemodel I0406 09:04:01.904891 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_13362.solverstate I0406 09:04:04.220733 5226 solver.cpp:330] Iteration 13362, Testing net (#0) I0406 09:04:04.220752 5226 net.cpp:676] Ignoring source layer train-data I0406 09:04:08.101148 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:04:08.783672 5226 solver.cpp:397] Test net output #0: accuracy = 0.197917 I0406 09:04:08.783699 5226 solver.cpp:397] Test net output #1: loss = 3.89871 (* 1 = 3.89871 loss) I0406 09:04:10.765869 5226 solver.cpp:218] Iteration 13368 (0.857532 iter/s, 13.9936s/12 iters), loss = 2.30211 I0406 09:04:10.765964 5226 solver.cpp:237] Train net output #0: loss = 2.30211 (* 1 = 2.30211 loss) I0406 09:04:10.765969 5226 sgd_solver.cpp:105] Iteration 13368, lr = 0.01 I0406 09:04:16.125847 5226 solver.cpp:218] Iteration 13380 (2.23888 iter/s, 5.35983s/12 iters), loss = 2.62597 I0406 09:04:16.125897 5226 solver.cpp:237] Train net output #0: loss = 2.62597 (* 1 = 2.62597 loss) I0406 09:04:16.125905 5226 sgd_solver.cpp:105] Iteration 13380, lr = 0.01 I0406 09:04:21.463968 5226 solver.cpp:218] Iteration 13392 (2.24802 iter/s, 5.33802s/12 iters), loss = 2.61026 I0406 09:04:21.464023 5226 solver.cpp:237] Train net output #0: loss = 2.61026 (* 1 = 2.61026 loss) I0406 09:04:21.464032 5226 sgd_solver.cpp:105] Iteration 13392, lr = 0.01 I0406 09:04:26.816376 5226 solver.cpp:218] Iteration 13404 (2.24203 iter/s, 5.3523s/12 iters), loss = 2.43207 I0406 09:04:26.816421 5226 solver.cpp:237] Train net output #0: loss = 2.43207 (* 1 = 2.43207 loss) I0406 09:04:26.816427 5226 sgd_solver.cpp:105] Iteration 13404, lr = 0.01 I0406 09:04:32.042165 5226 solver.cpp:218] Iteration 13416 (2.29635 iter/s, 5.22569s/12 iters), loss = 2.22829 I0406 09:04:32.042222 5226 solver.cpp:237] Train net output #0: loss = 2.22829 (* 1 = 2.22829 loss) I0406 09:04:32.042230 5226 sgd_solver.cpp:105] Iteration 13416, lr = 0.01 I0406 09:04:37.348088 5226 solver.cpp:218] Iteration 13428 (2.26167 iter/s, 5.30582s/12 iters), loss = 2.20421 I0406 09:04:37.348135 5226 solver.cpp:237] Train net output #0: loss = 2.20421 (* 1 = 2.20421 loss) I0406 09:04:37.348145 5226 sgd_solver.cpp:105] Iteration 13428, lr = 0.01 I0406 09:04:42.691124 5226 solver.cpp:218] Iteration 13440 (2.24595 iter/s, 5.34294s/12 iters), loss = 2.17978 I0406 09:04:42.691224 5226 solver.cpp:237] Train net output #0: loss = 2.17978 (* 1 = 2.17978 loss) I0406 09:04:42.691231 5226 sgd_solver.cpp:105] Iteration 13440, lr = 0.01 I0406 09:04:47.783449 5226 solver.cpp:218] Iteration 13452 (2.35655 iter/s, 5.09218s/12 iters), loss = 2.74603 I0406 09:04:47.783488 5226 solver.cpp:237] Train net output #0: loss = 2.74603 (* 1 = 2.74603 loss) I0406 09:04:47.783493 5226 sgd_solver.cpp:105] Iteration 13452, lr = 0.01 I0406 09:04:47.994081 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:04:52.495338 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_13464.caffemodel I0406 09:04:55.508199 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_13464.solverstate I0406 09:04:57.821684 5226 solver.cpp:330] Iteration 13464, Testing net (#0) I0406 09:04:57.821707 5226 net.cpp:676] Ignoring source layer train-data I0406 09:05:01.583039 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:05:02.279616 5226 solver.cpp:397] Test net output #0: accuracy = 0.213848 I0406 09:05:02.279667 5226 solver.cpp:397] Test net output #1: loss = 3.81124 (* 1 = 3.81124 loss) I0406 09:05:02.417029 5226 solver.cpp:218] Iteration 13464 (0.82004 iter/s, 14.6334s/12 iters), loss = 2.2499 I0406 09:05:02.417075 5226 solver.cpp:237] Train net output #0: loss = 2.2499 (* 1 = 2.2499 loss) I0406 09:05:02.417081 5226 sgd_solver.cpp:105] Iteration 13464, lr = 0.01 I0406 09:05:06.669373 5226 solver.cpp:218] Iteration 13476 (2.82203 iter/s, 4.25226s/12 iters), loss = 1.69158 I0406 09:05:06.669414 5226 solver.cpp:237] Train net output #0: loss = 1.69158 (* 1 = 1.69158 loss) I0406 09:05:06.669418 5226 sgd_solver.cpp:105] Iteration 13476, lr = 0.01 I0406 09:05:11.885988 5226 solver.cpp:218] Iteration 13488 (2.30038 iter/s, 5.21652s/12 iters), loss = 2.443 I0406 09:05:11.886026 5226 solver.cpp:237] Train net output #0: loss = 2.443 (* 1 = 2.443 loss) I0406 09:05:11.886031 5226 sgd_solver.cpp:105] Iteration 13488, lr = 0.01 I0406 09:05:16.765650 5226 solver.cpp:218] Iteration 13500 (2.45923 iter/s, 4.87958s/12 iters), loss = 2.49768 I0406 09:05:16.765754 5226 solver.cpp:237] Train net output #0: loss = 2.49768 (* 1 = 2.49768 loss) I0406 09:05:16.765761 5226 sgd_solver.cpp:105] Iteration 13500, lr = 0.01 I0406 09:05:22.051703 5226 solver.cpp:218] Iteration 13512 (2.27019 iter/s, 5.2859s/12 iters), loss = 2.15436 I0406 09:05:22.051751 5226 solver.cpp:237] Train net output #0: loss = 2.15436 (* 1 = 2.15436 loss) I0406 09:05:22.051759 5226 sgd_solver.cpp:105] Iteration 13512, lr = 0.01 I0406 09:05:27.413342 5226 solver.cpp:218] Iteration 13524 (2.23816 iter/s, 5.36155s/12 iters), loss = 1.98547 I0406 09:05:27.413379 5226 solver.cpp:237] Train net output #0: loss = 1.98547 (* 1 = 1.98547 loss) I0406 09:05:27.413388 5226 sgd_solver.cpp:105] Iteration 13524, lr = 0.01 I0406 09:05:32.665179 5226 solver.cpp:218] Iteration 13536 (2.28495 iter/s, 5.25175s/12 iters), loss = 2.20417 I0406 09:05:32.665231 5226 solver.cpp:237] Train net output #0: loss = 2.20417 (* 1 = 2.20417 loss) I0406 09:05:32.665236 5226 sgd_solver.cpp:105] Iteration 13536, lr = 0.01 I0406 09:05:38.094023 5226 solver.cpp:218] Iteration 13548 (2.21046 iter/s, 5.42874s/12 iters), loss = 2.3796 I0406 09:05:38.094071 5226 solver.cpp:237] Train net output #0: loss = 2.3796 (* 1 = 2.3796 loss) I0406 09:05:38.094080 5226 sgd_solver.cpp:105] Iteration 13548, lr = 0.01 I0406 09:05:40.548434 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:05:43.338945 5226 solver.cpp:218] Iteration 13560 (2.28797 iter/s, 5.24483s/12 iters), loss = 1.73444 I0406 09:05:43.338985 5226 solver.cpp:237] Train net output #0: loss = 1.73444 (* 1 = 1.73444 loss) I0406 09:05:43.338991 5226 sgd_solver.cpp:105] Iteration 13560, lr = 0.01 I0406 09:05:45.381743 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_13566.caffemodel I0406 09:05:48.490562 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_13566.solverstate I0406 09:05:50.800822 5226 solver.cpp:330] Iteration 13566, Testing net (#0) I0406 09:05:50.800843 5226 net.cpp:676] Ignoring source layer train-data I0406 09:05:54.394147 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:05:55.115319 5226 solver.cpp:397] Test net output #0: accuracy = 0.221814 I0406 09:05:55.115351 5226 solver.cpp:397] Test net output #1: loss = 3.82109 (* 1 = 3.82109 loss) I0406 09:05:56.926829 5226 solver.cpp:218] Iteration 13572 (0.883149 iter/s, 13.5877s/12 iters), loss = 2.14733 I0406 09:05:56.926867 5226 solver.cpp:237] Train net output #0: loss = 2.14733 (* 1 = 2.14733 loss) I0406 09:05:56.926872 5226 sgd_solver.cpp:105] Iteration 13572, lr = 0.01 I0406 09:06:02.222796 5226 solver.cpp:218] Iteration 13584 (2.26591 iter/s, 5.29588s/12 iters), loss = 2.42594 I0406 09:06:02.222851 5226 solver.cpp:237] Train net output #0: loss = 2.42594 (* 1 = 2.42594 loss) I0406 09:06:02.222860 5226 sgd_solver.cpp:105] Iteration 13584, lr = 0.01 I0406 09:06:07.441560 5226 solver.cpp:218] Iteration 13596 (2.29944 iter/s, 5.21866s/12 iters), loss = 2.05399 I0406 09:06:07.441614 5226 solver.cpp:237] Train net output #0: loss = 2.05399 (* 1 = 2.05399 loss) I0406 09:06:07.441622 5226 sgd_solver.cpp:105] Iteration 13596, lr = 0.01 I0406 09:06:12.749780 5226 solver.cpp:218] Iteration 13608 (2.26069 iter/s, 5.30812s/12 iters), loss = 2.33492 I0406 09:06:12.749816 5226 solver.cpp:237] Train net output #0: loss = 2.33492 (* 1 = 2.33492 loss) I0406 09:06:12.749821 5226 sgd_solver.cpp:105] Iteration 13608, lr = 0.01 I0406 09:06:17.960744 5226 solver.cpp:218] Iteration 13620 (2.30287 iter/s, 5.21088s/12 iters), loss = 2.19593 I0406 09:06:17.960783 5226 solver.cpp:237] Train net output #0: loss = 2.19593 (* 1 = 2.19593 loss) I0406 09:06:17.960788 5226 sgd_solver.cpp:105] Iteration 13620, lr = 0.01 I0406 09:06:23.237072 5226 solver.cpp:218] Iteration 13632 (2.27435 iter/s, 5.27624s/12 iters), loss = 1.59323 I0406 09:06:23.237193 5226 solver.cpp:237] Train net output #0: loss = 1.59323 (* 1 = 1.59323 loss) I0406 09:06:23.237200 5226 sgd_solver.cpp:105] Iteration 13632, lr = 0.01 I0406 09:06:28.348304 5226 solver.cpp:218] Iteration 13644 (2.34785 iter/s, 5.11107s/12 iters), loss = 1.97531 I0406 09:06:28.348345 5226 solver.cpp:237] Train net output #0: loss = 1.97531 (* 1 = 1.97531 loss) I0406 09:06:28.348349 5226 sgd_solver.cpp:105] Iteration 13644, lr = 0.01 I0406 09:06:33.132251 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:06:33.557981 5226 solver.cpp:218] Iteration 13656 (2.30345 iter/s, 5.20959s/12 iters), loss = 2.3836 I0406 09:06:33.558020 5226 solver.cpp:237] Train net output #0: loss = 2.3836 (* 1 = 2.3836 loss) I0406 09:06:33.558027 5226 sgd_solver.cpp:105] Iteration 13656, lr = 0.01 I0406 09:06:38.320755 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_13668.caffemodel I0406 09:06:41.643426 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_13668.solverstate I0406 09:06:44.027274 5226 solver.cpp:330] Iteration 13668, Testing net (#0) I0406 09:06:44.027292 5226 net.cpp:676] Ignoring source layer train-data I0406 09:06:47.896504 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:06:48.651468 5226 solver.cpp:397] Test net output #0: accuracy = 0.205882 I0406 09:06:48.651504 5226 solver.cpp:397] Test net output #1: loss = 3.93181 (* 1 = 3.93181 loss) I0406 09:06:48.789579 5226 solver.cpp:218] Iteration 13668 (0.787844 iter/s, 15.2314s/12 iters), loss = 2.39534 I0406 09:06:48.789636 5226 solver.cpp:237] Train net output #0: loss = 2.39534 (* 1 = 2.39534 loss) I0406 09:06:48.789642 5226 sgd_solver.cpp:105] Iteration 13668, lr = 0.01 I0406 09:06:53.043021 5226 solver.cpp:218] Iteration 13680 (2.82131 iter/s, 4.25334s/12 iters), loss = 2.39708 I0406 09:06:53.043062 5226 solver.cpp:237] Train net output #0: loss = 2.39708 (* 1 = 2.39708 loss) I0406 09:06:53.043068 5226 sgd_solver.cpp:105] Iteration 13680, lr = 0.01 I0406 09:06:58.556622 5226 solver.cpp:218] Iteration 13692 (2.17647 iter/s, 5.5135s/12 iters), loss = 2.40788 I0406 09:06:58.560961 5226 solver.cpp:237] Train net output #0: loss = 2.40788 (* 1 = 2.40788 loss) I0406 09:06:58.560977 5226 sgd_solver.cpp:105] Iteration 13692, lr = 0.01 I0406 09:07:03.825335 5226 solver.cpp:218] Iteration 13704 (2.27949 iter/s, 5.26435s/12 iters), loss = 2.02432 I0406 09:07:03.825373 5226 solver.cpp:237] Train net output #0: loss = 2.02432 (* 1 = 2.02432 loss) I0406 09:07:03.825378 5226 sgd_solver.cpp:105] Iteration 13704, lr = 0.01 I0406 09:07:09.164023 5226 solver.cpp:218] Iteration 13716 (2.24778 iter/s, 5.3386s/12 iters), loss = 2.33532 I0406 09:07:09.164063 5226 solver.cpp:237] Train net output #0: loss = 2.33532 (* 1 = 2.33532 loss) I0406 09:07:09.164068 5226 sgd_solver.cpp:105] Iteration 13716, lr = 0.01 I0406 09:07:14.518608 5226 solver.cpp:218] Iteration 13728 (2.24111 iter/s, 5.3545s/12 iters), loss = 2.23236 I0406 09:07:14.518647 5226 solver.cpp:237] Train net output #0: loss = 2.23236 (* 1 = 2.23236 loss) I0406 09:07:14.518652 5226 sgd_solver.cpp:105] Iteration 13728, lr = 0.01 I0406 09:07:19.845876 5226 solver.cpp:218] Iteration 13740 (2.2526 iter/s, 5.32718s/12 iters), loss = 2.40155 I0406 09:07:19.845927 5226 solver.cpp:237] Train net output #0: loss = 2.40155 (* 1 = 2.40155 loss) I0406 09:07:19.845935 5226 sgd_solver.cpp:105] Iteration 13740, lr = 0.01 I0406 09:07:25.187639 5226 solver.cpp:218] Iteration 13752 (2.24649 iter/s, 5.34166s/12 iters), loss = 2.35743 I0406 09:07:25.187695 5226 solver.cpp:237] Train net output #0: loss = 2.35743 (* 1 = 2.35743 loss) I0406 09:07:25.187705 5226 sgd_solver.cpp:105] Iteration 13752, lr = 0.01 I0406 09:07:27.066740 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:07:30.522723 5226 solver.cpp:218] Iteration 13764 (2.24931 iter/s, 5.33497s/12 iters), loss = 1.79331 I0406 09:07:30.522908 5226 solver.cpp:237] Train net output #0: loss = 1.79331 (* 1 = 1.79331 loss) I0406 09:07:30.522917 5226 sgd_solver.cpp:105] Iteration 13764, lr = 0.01 I0406 09:07:32.513229 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_13770.caffemodel I0406 09:07:35.577848 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_13770.solverstate I0406 09:07:37.903879 5226 solver.cpp:330] Iteration 13770, Testing net (#0) I0406 09:07:37.903898 5226 net.cpp:676] Ignoring source layer train-data I0406 09:07:41.456990 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:07:42.288944 5226 solver.cpp:397] Test net output #0: accuracy = 0.199142 I0406 09:07:42.288972 5226 solver.cpp:397] Test net output #1: loss = 3.95267 (* 1 = 3.95267 loss) I0406 09:07:44.128840 5226 solver.cpp:218] Iteration 13776 (0.881974 iter/s, 13.6058s/12 iters), loss = 2.28799 I0406 09:07:44.128880 5226 solver.cpp:237] Train net output #0: loss = 2.28799 (* 1 = 2.28799 loss) I0406 09:07:44.128893 5226 sgd_solver.cpp:105] Iteration 13776, lr = 0.01 I0406 09:07:49.357278 5226 solver.cpp:218] Iteration 13788 (2.29518 iter/s, 5.22835s/12 iters), loss = 2.19909 I0406 09:07:49.357331 5226 solver.cpp:237] Train net output #0: loss = 2.19909 (* 1 = 2.19909 loss) I0406 09:07:49.357342 5226 sgd_solver.cpp:105] Iteration 13788, lr = 0.01 I0406 09:07:54.596021 5226 solver.cpp:218] Iteration 13800 (2.29067 iter/s, 5.23864s/12 iters), loss = 2.34205 I0406 09:07:54.596071 5226 solver.cpp:237] Train net output #0: loss = 2.34205 (* 1 = 2.34205 loss) I0406 09:07:54.596078 5226 sgd_solver.cpp:105] Iteration 13800, lr = 0.01 I0406 09:07:59.938014 5226 solver.cpp:218] Iteration 13812 (2.24639 iter/s, 5.3419s/12 iters), loss = 2.15847 I0406 09:07:59.938055 5226 solver.cpp:237] Train net output #0: loss = 2.15847 (* 1 = 2.15847 loss) I0406 09:07:59.938060 5226 sgd_solver.cpp:105] Iteration 13812, lr = 0.01 I0406 09:08:05.393174 5226 solver.cpp:218] Iteration 13824 (2.19979 iter/s, 5.45506s/12 iters), loss = 2.41394 I0406 09:08:05.393296 5226 solver.cpp:237] Train net output #0: loss = 2.41394 (* 1 = 2.41394 loss) I0406 09:08:05.393309 5226 sgd_solver.cpp:105] Iteration 13824, lr = 0.01 I0406 09:08:10.583961 5226 solver.cpp:218] Iteration 13836 (2.31186 iter/s, 5.19062s/12 iters), loss = 2.4472 I0406 09:08:10.584013 5226 solver.cpp:237] Train net output #0: loss = 2.4472 (* 1 = 2.4472 loss) I0406 09:08:10.584020 5226 sgd_solver.cpp:105] Iteration 13836, lr = 0.01 I0406 09:08:15.764874 5226 solver.cpp:218] Iteration 13848 (2.31624 iter/s, 5.18082s/12 iters), loss = 2.31541 I0406 09:08:15.764940 5226 solver.cpp:237] Train net output #0: loss = 2.31541 (* 1 = 2.31541 loss) I0406 09:08:15.764953 5226 sgd_solver.cpp:105] Iteration 13848, lr = 0.01 I0406 09:08:19.718096 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:08:20.926285 5226 solver.cpp:218] Iteration 13860 (2.325 iter/s, 5.1613s/12 iters), loss = 2.08953 I0406 09:08:20.926326 5226 solver.cpp:237] Train net output #0: loss = 2.08953 (* 1 = 2.08953 loss) I0406 09:08:20.926331 5226 sgd_solver.cpp:105] Iteration 13860, lr = 0.01 I0406 09:08:25.692596 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_13872.caffemodel I0406 09:08:28.748028 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_13872.solverstate I0406 09:08:31.056689 5226 solver.cpp:330] Iteration 13872, Testing net (#0) I0406 09:08:31.056710 5226 net.cpp:676] Ignoring source layer train-data I0406 09:08:31.971725 5226 blocking_queue.cpp:49] Waiting for data I0406 09:08:34.499881 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:08:35.333228 5226 solver.cpp:397] Test net output #0: accuracy = 0.22549 I0406 09:08:35.333256 5226 solver.cpp:397] Test net output #1: loss = 3.83533 (* 1 = 3.83533 loss) I0406 09:08:35.474154 5226 solver.cpp:218] Iteration 13872 (0.824871 iter/s, 14.5477s/12 iters), loss = 2.11235 I0406 09:08:35.474295 5226 solver.cpp:237] Train net output #0: loss = 2.11235 (* 1 = 2.11235 loss) I0406 09:08:35.474304 5226 sgd_solver.cpp:105] Iteration 13872, lr = 0.01 I0406 09:08:39.734441 5226 solver.cpp:218] Iteration 13884 (2.81683 iter/s, 4.26011s/12 iters), loss = 1.84818 I0406 09:08:39.734483 5226 solver.cpp:237] Train net output #0: loss = 1.84818 (* 1 = 1.84818 loss) I0406 09:08:39.734488 5226 sgd_solver.cpp:105] Iteration 13884, lr = 0.01 I0406 09:08:44.962476 5226 solver.cpp:218] Iteration 13896 (2.29536 iter/s, 5.22794s/12 iters), loss = 1.96291 I0406 09:08:44.962527 5226 solver.cpp:237] Train net output #0: loss = 1.96291 (* 1 = 1.96291 loss) I0406 09:08:44.962536 5226 sgd_solver.cpp:105] Iteration 13896, lr = 0.01 I0406 09:08:50.325089 5226 solver.cpp:218] Iteration 13908 (2.23776 iter/s, 5.36251s/12 iters), loss = 1.94858 I0406 09:08:50.325130 5226 solver.cpp:237] Train net output #0: loss = 1.94858 (* 1 = 1.94858 loss) I0406 09:08:50.325135 5226 sgd_solver.cpp:105] Iteration 13908, lr = 0.01 I0406 09:08:55.592846 5226 solver.cpp:218] Iteration 13920 (2.27805 iter/s, 5.26767s/12 iters), loss = 1.95682 I0406 09:08:55.592890 5226 solver.cpp:237] Train net output #0: loss = 1.95682 (* 1 = 1.95682 loss) I0406 09:08:55.592896 5226 sgd_solver.cpp:105] Iteration 13920, lr = 0.01 I0406 09:09:00.912190 5226 solver.cpp:218] Iteration 13932 (2.25595 iter/s, 5.31926s/12 iters), loss = 2.67654 I0406 09:09:00.912236 5226 solver.cpp:237] Train net output #0: loss = 2.67654 (* 1 = 2.67654 loss) I0406 09:09:00.912243 5226 sgd_solver.cpp:105] Iteration 13932, lr = 0.01 I0406 09:09:06.248520 5226 solver.cpp:218] Iteration 13944 (2.24878 iter/s, 5.33624s/12 iters), loss = 2.03481 I0406 09:09:06.248631 5226 solver.cpp:237] Train net output #0: loss = 2.03481 (* 1 = 2.03481 loss) I0406 09:09:06.248636 5226 sgd_solver.cpp:105] Iteration 13944, lr = 0.01 I0406 09:09:11.415364 5226 solver.cpp:218] Iteration 13956 (2.32257 iter/s, 5.16669s/12 iters), loss = 2.15872 I0406 09:09:11.415412 5226 solver.cpp:237] Train net output #0: loss = 2.15872 (* 1 = 2.15872 loss) I0406 09:09:11.415421 5226 sgd_solver.cpp:105] Iteration 13956, lr = 0.01 I0406 09:09:12.509493 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:09:16.900771 5226 solver.cpp:218] Iteration 13968 (2.18766 iter/s, 5.48531s/12 iters), loss = 2.22588 I0406 09:09:16.900818 5226 solver.cpp:237] Train net output #0: loss = 2.22588 (* 1 = 2.22588 loss) I0406 09:09:16.900825 5226 sgd_solver.cpp:105] Iteration 13968, lr = 0.01 I0406 09:09:18.990226 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_13974.caffemodel I0406 09:09:21.985267 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_13974.solverstate I0406 09:09:24.302176 5226 solver.cpp:330] Iteration 13974, Testing net (#0) I0406 09:09:24.302197 5226 net.cpp:676] Ignoring source layer train-data I0406 09:09:27.825876 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:09:28.753934 5226 solver.cpp:397] Test net output #0: accuracy = 0.227941 I0406 09:09:28.753971 5226 solver.cpp:397] Test net output #1: loss = 3.92318 (* 1 = 3.92318 loss) I0406 09:09:30.565485 5226 solver.cpp:218] Iteration 13980 (0.878184 iter/s, 13.6646s/12 iters), loss = 2.24545 I0406 09:09:30.565527 5226 solver.cpp:237] Train net output #0: loss = 2.24545 (* 1 = 2.24545 loss) I0406 09:09:30.565533 5226 sgd_solver.cpp:105] Iteration 13980, lr = 0.01 I0406 09:09:35.854341 5226 solver.cpp:218] Iteration 13992 (2.26896 iter/s, 5.28876s/12 iters), loss = 2.71619 I0406 09:09:35.854389 5226 solver.cpp:237] Train net output #0: loss = 2.71619 (* 1 = 2.71619 loss) I0406 09:09:35.854396 5226 sgd_solver.cpp:105] Iteration 13992, lr = 0.01 I0406 09:09:41.050032 5226 solver.cpp:218] Iteration 14004 (2.30965 iter/s, 5.1956s/12 iters), loss = 2.34296 I0406 09:09:41.050182 5226 solver.cpp:237] Train net output #0: loss = 2.34296 (* 1 = 2.34296 loss) I0406 09:09:41.050191 5226 sgd_solver.cpp:105] Iteration 14004, lr = 0.01 I0406 09:09:46.450433 5226 solver.cpp:218] Iteration 14016 (2.22214 iter/s, 5.4002s/12 iters), loss = 2.04413 I0406 09:09:46.450479 5226 solver.cpp:237] Train net output #0: loss = 2.04413 (* 1 = 2.04413 loss) I0406 09:09:46.450486 5226 sgd_solver.cpp:105] Iteration 14016, lr = 0.01 I0406 09:09:51.604075 5226 solver.cpp:218] Iteration 14028 (2.32849 iter/s, 5.15355s/12 iters), loss = 2.12931 I0406 09:09:51.604115 5226 solver.cpp:237] Train net output #0: loss = 2.12931 (* 1 = 2.12931 loss) I0406 09:09:51.604120 5226 sgd_solver.cpp:105] Iteration 14028, lr = 0.01 I0406 09:09:56.836527 5226 solver.cpp:218] Iteration 14040 (2.29342 iter/s, 5.23236s/12 iters), loss = 2.28087 I0406 09:09:56.836572 5226 solver.cpp:237] Train net output #0: loss = 2.28087 (* 1 = 2.28087 loss) I0406 09:09:56.836578 5226 sgd_solver.cpp:105] Iteration 14040, lr = 0.01 I0406 09:10:02.146047 5226 solver.cpp:218] Iteration 14052 (2.26013 iter/s, 5.30942s/12 iters), loss = 2.83542 I0406 09:10:02.146100 5226 solver.cpp:237] Train net output #0: loss = 2.83542 (* 1 = 2.83542 loss) I0406 09:10:02.146108 5226 sgd_solver.cpp:105] Iteration 14052, lr = 0.01 I0406 09:10:05.410022 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:10:07.317672 5226 solver.cpp:218] Iteration 14064 (2.3204 iter/s, 5.17153s/12 iters), loss = 2.37138 I0406 09:10:07.317711 5226 solver.cpp:237] Train net output #0: loss = 2.37138 (* 1 = 2.37138 loss) I0406 09:10:07.317716 5226 sgd_solver.cpp:105] Iteration 14064, lr = 0.01 I0406 09:10:12.190904 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_14076.caffemodel I0406 09:10:15.142534 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_14076.solverstate I0406 09:10:17.439829 5226 solver.cpp:330] Iteration 14076, Testing net (#0) I0406 09:10:17.439859 5226 net.cpp:676] Ignoring source layer train-data I0406 09:10:21.125290 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:10:22.044435 5226 solver.cpp:397] Test net output #0: accuracy = 0.208946 I0406 09:10:22.044471 5226 solver.cpp:397] Test net output #1: loss = 3.99557 (* 1 = 3.99557 loss) I0406 09:10:22.186200 5226 solver.cpp:218] Iteration 14076 (0.807082 iter/s, 14.8684s/12 iters), loss = 2.35432 I0406 09:10:22.186252 5226 solver.cpp:237] Train net output #0: loss = 2.35432 (* 1 = 2.35432 loss) I0406 09:10:22.186259 5226 sgd_solver.cpp:105] Iteration 14076, lr = 0.01 I0406 09:10:26.680339 5226 solver.cpp:218] Iteration 14088 (2.6702 iter/s, 4.49404s/12 iters), loss = 2.1479 I0406 09:10:26.680382 5226 solver.cpp:237] Train net output #0: loss = 2.1479 (* 1 = 2.1479 loss) I0406 09:10:26.680387 5226 sgd_solver.cpp:105] Iteration 14088, lr = 0.01 I0406 09:10:32.234671 5226 solver.cpp:218] Iteration 14100 (2.16051 iter/s, 5.55424s/12 iters), loss = 2.17677 I0406 09:10:32.234707 5226 solver.cpp:237] Train net output #0: loss = 2.17677 (* 1 = 2.17677 loss) I0406 09:10:32.234714 5226 sgd_solver.cpp:105] Iteration 14100, lr = 0.01 I0406 09:10:37.562297 5226 solver.cpp:218] Iteration 14112 (2.25245 iter/s, 5.32754s/12 iters), loss = 2.77335 I0406 09:10:37.562351 5226 solver.cpp:237] Train net output #0: loss = 2.77335 (* 1 = 2.77335 loss) I0406 09:10:37.562361 5226 sgd_solver.cpp:105] Iteration 14112, lr = 0.01 I0406 09:10:42.872081 5226 solver.cpp:218] Iteration 14124 (2.26002 iter/s, 5.30969s/12 iters), loss = 2.2758 I0406 09:10:42.872303 5226 solver.cpp:237] Train net output #0: loss = 2.2758 (* 1 = 2.2758 loss) I0406 09:10:42.872309 5226 sgd_solver.cpp:105] Iteration 14124, lr = 0.01 I0406 09:10:47.989812 5226 solver.cpp:218] Iteration 14136 (2.34491 iter/s, 5.11746s/12 iters), loss = 2.23622 I0406 09:10:47.989864 5226 solver.cpp:237] Train net output #0: loss = 2.23622 (* 1 = 2.23622 loss) I0406 09:10:47.989873 5226 sgd_solver.cpp:105] Iteration 14136, lr = 0.01 I0406 09:10:53.362319 5226 solver.cpp:218] Iteration 14148 (2.23364 iter/s, 5.37241s/12 iters), loss = 2.61299 I0406 09:10:53.362360 5226 solver.cpp:237] Train net output #0: loss = 2.61299 (* 1 = 2.61299 loss) I0406 09:10:53.362365 5226 sgd_solver.cpp:105] Iteration 14148, lr = 0.01 I0406 09:10:58.696686 5226 solver.cpp:218] Iteration 14160 (2.2496 iter/s, 5.33428s/12 iters), loss = 2.44028 I0406 09:10:58.696727 5226 solver.cpp:237] Train net output #0: loss = 2.44028 (* 1 = 2.44028 loss) I0406 09:10:58.696732 5226 sgd_solver.cpp:105] Iteration 14160, lr = 0.01 I0406 09:10:58.955461 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:11:03.876199 5226 solver.cpp:218] Iteration 14172 (2.31686 iter/s, 5.17942s/12 iters), loss = 2.32228 I0406 09:11:03.876245 5226 solver.cpp:237] Train net output #0: loss = 2.32228 (* 1 = 2.32228 loss) I0406 09:11:03.876252 5226 sgd_solver.cpp:105] Iteration 14172, lr = 0.01 I0406 09:11:05.972968 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_14178.caffemodel I0406 09:11:09.039983 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_14178.solverstate I0406 09:11:11.338547 5226 solver.cpp:330] Iteration 14178, Testing net (#0) I0406 09:11:11.338567 5226 net.cpp:676] Ignoring source layer train-data I0406 09:11:14.804132 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:11:15.748437 5226 solver.cpp:397] Test net output #0: accuracy = 0.183824 I0406 09:11:15.748471 5226 solver.cpp:397] Test net output #1: loss = 4.07589 (* 1 = 4.07589 loss) I0406 09:11:17.532730 5226 solver.cpp:218] Iteration 14184 (0.87871 iter/s, 13.6564s/12 iters), loss = 1.95331 I0406 09:11:17.532781 5226 solver.cpp:237] Train net output #0: loss = 1.95331 (* 1 = 1.95331 loss) I0406 09:11:17.532788 5226 sgd_solver.cpp:105] Iteration 14184, lr = 0.01 I0406 09:11:22.770311 5226 solver.cpp:218] Iteration 14196 (2.29118 iter/s, 5.23749s/12 iters), loss = 2.34385 I0406 09:11:22.770347 5226 solver.cpp:237] Train net output #0: loss = 2.34385 (* 1 = 2.34385 loss) I0406 09:11:22.770352 5226 sgd_solver.cpp:105] Iteration 14196, lr = 0.01 I0406 09:11:27.762193 5226 solver.cpp:218] Iteration 14208 (2.40394 iter/s, 4.9918s/12 iters), loss = 2.73436 I0406 09:11:27.762228 5226 solver.cpp:237] Train net output #0: loss = 2.73436 (* 1 = 2.73436 loss) I0406 09:11:27.762233 5226 sgd_solver.cpp:105] Iteration 14208, lr = 0.01 I0406 09:11:32.953739 5226 solver.cpp:218] Iteration 14220 (2.31149 iter/s, 5.19146s/12 iters), loss = 2.25962 I0406 09:11:32.953791 5226 solver.cpp:237] Train net output #0: loss = 2.25962 (* 1 = 2.25962 loss) I0406 09:11:32.953796 5226 sgd_solver.cpp:105] Iteration 14220, lr = 0.01 I0406 09:11:38.132721 5226 solver.cpp:218] Iteration 14232 (2.31711 iter/s, 5.17887s/12 iters), loss = 2.4327 I0406 09:11:38.132777 5226 solver.cpp:237] Train net output #0: loss = 2.4327 (* 1 = 2.4327 loss) I0406 09:11:38.132786 5226 sgd_solver.cpp:105] Iteration 14232, lr = 0.01 I0406 09:11:43.345237 5226 solver.cpp:218] Iteration 14244 (2.3022 iter/s, 5.21241s/12 iters), loss = 2.42727 I0406 09:11:43.345273 5226 solver.cpp:237] Train net output #0: loss = 2.42727 (* 1 = 2.42727 loss) I0406 09:11:43.345278 5226 sgd_solver.cpp:105] Iteration 14244, lr = 0.01 I0406 09:11:48.614328 5226 solver.cpp:218] Iteration 14256 (2.27747 iter/s, 5.26901s/12 iters), loss = 2.05933 I0406 09:11:48.614446 5226 solver.cpp:237] Train net output #0: loss = 2.05933 (* 1 = 2.05933 loss) I0406 09:11:48.614452 5226 sgd_solver.cpp:105] Iteration 14256, lr = 0.01 I0406 09:11:51.097167 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:11:53.835638 5226 solver.cpp:218] Iteration 14268 (2.29835 iter/s, 5.22114s/12 iters), loss = 2.7042 I0406 09:11:53.835687 5226 solver.cpp:237] Train net output #0: loss = 2.7042 (* 1 = 2.7042 loss) I0406 09:11:53.835695 5226 sgd_solver.cpp:105] Iteration 14268, lr = 0.01 I0406 09:11:58.653678 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_14280.caffemodel I0406 09:12:01.705651 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_14280.solverstate I0406 09:12:04.021034 5226 solver.cpp:330] Iteration 14280, Testing net (#0) I0406 09:12:04.021054 5226 net.cpp:676] Ignoring source layer train-data I0406 09:12:07.416474 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:12:08.399963 5226 solver.cpp:397] Test net output #0: accuracy = 0.202819 I0406 09:12:08.399998 5226 solver.cpp:397] Test net output #1: loss = 3.98414 (* 1 = 3.98414 loss) I0406 09:12:08.540485 5226 solver.cpp:218] Iteration 14280 (0.816066 iter/s, 14.7047s/12 iters), loss = 2.41422 I0406 09:12:08.540524 5226 solver.cpp:237] Train net output #0: loss = 2.41422 (* 1 = 2.41422 loss) I0406 09:12:08.540529 5226 sgd_solver.cpp:105] Iteration 14280, lr = 0.01 I0406 09:12:12.849885 5226 solver.cpp:218] Iteration 14292 (2.78467 iter/s, 4.30931s/12 iters), loss = 2.07011 I0406 09:12:12.849941 5226 solver.cpp:237] Train net output #0: loss = 2.07011 (* 1 = 2.07011 loss) I0406 09:12:12.849951 5226 sgd_solver.cpp:105] Iteration 14292, lr = 0.01 I0406 09:12:18.045759 5226 solver.cpp:218] Iteration 14304 (2.30957 iter/s, 5.19577s/12 iters), loss = 2.31967 I0406 09:12:18.045817 5226 solver.cpp:237] Train net output #0: loss = 2.31967 (* 1 = 2.31967 loss) I0406 09:12:18.045826 5226 sgd_solver.cpp:105] Iteration 14304, lr = 0.01 I0406 09:12:23.363026 5226 solver.cpp:218] Iteration 14316 (2.25684 iter/s, 5.31716s/12 iters), loss = 2.21341 I0406 09:12:23.363126 5226 solver.cpp:237] Train net output #0: loss = 2.21341 (* 1 = 2.21341 loss) I0406 09:12:23.363133 5226 sgd_solver.cpp:105] Iteration 14316, lr = 0.01 I0406 09:12:28.760185 5226 solver.cpp:218] Iteration 14328 (2.22345 iter/s, 5.39701s/12 iters), loss = 2.02363 I0406 09:12:28.760222 5226 solver.cpp:237] Train net output #0: loss = 2.02363 (* 1 = 2.02363 loss) I0406 09:12:28.760227 5226 sgd_solver.cpp:105] Iteration 14328, lr = 0.01 I0406 09:12:33.790174 5226 solver.cpp:218] Iteration 14340 (2.38573 iter/s, 5.0299s/12 iters), loss = 2.54744 I0406 09:12:33.790230 5226 solver.cpp:237] Train net output #0: loss = 2.54744 (* 1 = 2.54744 loss) I0406 09:12:33.790237 5226 sgd_solver.cpp:105] Iteration 14340, lr = 0.01 I0406 09:12:39.134004 5226 solver.cpp:218] Iteration 14352 (2.24562 iter/s, 5.34373s/12 iters), loss = 1.9294 I0406 09:12:39.134039 5226 solver.cpp:237] Train net output #0: loss = 1.9294 (* 1 = 1.9294 loss) I0406 09:12:39.134045 5226 sgd_solver.cpp:105] Iteration 14352, lr = 0.01 I0406 09:12:43.991971 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:12:44.376108 5226 solver.cpp:218] Iteration 14364 (2.28919 iter/s, 5.24202s/12 iters), loss = 2.35754 I0406 09:12:44.376143 5226 solver.cpp:237] Train net output #0: loss = 2.35754 (* 1 = 2.35754 loss) I0406 09:12:44.376148 5226 sgd_solver.cpp:105] Iteration 14364, lr = 0.01 I0406 09:12:49.595145 5226 solver.cpp:218] Iteration 14376 (2.29931 iter/s, 5.21895s/12 iters), loss = 2.16566 I0406 09:12:49.595201 5226 solver.cpp:237] Train net output #0: loss = 2.16566 (* 1 = 2.16566 loss) I0406 09:12:49.595211 5226 sgd_solver.cpp:105] Iteration 14376, lr = 0.01 I0406 09:12:51.574071 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_14382.caffemodel I0406 09:12:54.612736 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_14382.solverstate I0406 09:12:56.929548 5226 solver.cpp:330] Iteration 14382, Testing net (#0) I0406 09:12:56.929567 5226 net.cpp:676] Ignoring source layer train-data I0406 09:13:00.257303 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:13:01.287714 5226 solver.cpp:397] Test net output #0: accuracy = 0.186275 I0406 09:13:01.287750 5226 solver.cpp:397] Test net output #1: loss = 3.96429 (* 1 = 3.96429 loss) I0406 09:13:03.142288 5226 solver.cpp:218] Iteration 14388 (0.885805 iter/s, 13.547s/12 iters), loss = 2.94084 I0406 09:13:03.142347 5226 solver.cpp:237] Train net output #0: loss = 2.94084 (* 1 = 2.94084 loss) I0406 09:13:03.142356 5226 sgd_solver.cpp:105] Iteration 14388, lr = 0.01 I0406 09:13:08.341923 5226 solver.cpp:218] Iteration 14400 (2.3079 iter/s, 5.19953s/12 iters), loss = 2.35891 I0406 09:13:08.341962 5226 solver.cpp:237] Train net output #0: loss = 2.35891 (* 1 = 2.35891 loss) I0406 09:13:08.341969 5226 sgd_solver.cpp:105] Iteration 14400, lr = 0.01 I0406 09:13:13.585629 5226 solver.cpp:218] Iteration 14412 (2.2885 iter/s, 5.24362s/12 iters), loss = 2.00138 I0406 09:13:13.585667 5226 solver.cpp:237] Train net output #0: loss = 2.00138 (* 1 = 2.00138 loss) I0406 09:13:13.585672 5226 sgd_solver.cpp:105] Iteration 14412, lr = 0.01 I0406 09:13:19.070232 5226 solver.cpp:218] Iteration 14424 (2.18798 iter/s, 5.48451s/12 iters), loss = 2.44084 I0406 09:13:19.070281 5226 solver.cpp:237] Train net output #0: loss = 2.44084 (* 1 = 2.44084 loss) I0406 09:13:19.070288 5226 sgd_solver.cpp:105] Iteration 14424, lr = 0.01 I0406 09:13:23.964534 5226 solver.cpp:218] Iteration 14436 (2.45188 iter/s, 4.89421s/12 iters), loss = 2.31559 I0406 09:13:23.964574 5226 solver.cpp:237] Train net output #0: loss = 2.31559 (* 1 = 2.31559 loss) I0406 09:13:23.964581 5226 sgd_solver.cpp:105] Iteration 14436, lr = 0.01 I0406 09:13:29.241680 5226 solver.cpp:218] Iteration 14448 (2.27399 iter/s, 5.27706s/12 iters), loss = 2.0162 I0406 09:13:29.241775 5226 solver.cpp:237] Train net output #0: loss = 2.0162 (* 1 = 2.0162 loss) I0406 09:13:29.241782 5226 sgd_solver.cpp:105] Iteration 14448, lr = 0.01 I0406 09:13:34.563563 5226 solver.cpp:218] Iteration 14460 (2.2549 iter/s, 5.32174s/12 iters), loss = 2.97605 I0406 09:13:34.563612 5226 solver.cpp:237] Train net output #0: loss = 2.97605 (* 1 = 2.97605 loss) I0406 09:13:34.563621 5226 sgd_solver.cpp:105] Iteration 14460, lr = 0.01 I0406 09:13:36.455925 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:13:39.933153 5226 solver.cpp:218] Iteration 14472 (2.23485 iter/s, 5.3695s/12 iters), loss = 2.18069 I0406 09:13:39.933193 5226 solver.cpp:237] Train net output #0: loss = 2.18069 (* 1 = 2.18069 loss) I0406 09:13:39.933198 5226 sgd_solver.cpp:105] Iteration 14472, lr = 0.01 I0406 09:13:44.687784 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_14484.caffemodel I0406 09:13:47.770009 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_14484.solverstate I0406 09:13:50.092334 5226 solver.cpp:330] Iteration 14484, Testing net (#0) I0406 09:13:50.092353 5226 net.cpp:676] Ignoring source layer train-data I0406 09:13:53.543886 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:13:54.630252 5226 solver.cpp:397] Test net output #0: accuracy = 0.21201 I0406 09:13:54.630288 5226 solver.cpp:397] Test net output #1: loss = 3.96919 (* 1 = 3.96919 loss) I0406 09:13:54.771375 5226 solver.cpp:218] Iteration 14484 (0.80873 iter/s, 14.8381s/12 iters), loss = 1.81004 I0406 09:13:54.772941 5226 solver.cpp:237] Train net output #0: loss = 1.81004 (* 1 = 1.81004 loss) I0406 09:13:54.772955 5226 sgd_solver.cpp:105] Iteration 14484, lr = 0.01 I0406 09:13:59.139078 5226 solver.cpp:218] Iteration 14496 (2.74844 iter/s, 4.36611s/12 iters), loss = 2.10455 I0406 09:13:59.139120 5226 solver.cpp:237] Train net output #0: loss = 2.10455 (* 1 = 2.10455 loss) I0406 09:13:59.139127 5226 sgd_solver.cpp:105] Iteration 14496, lr = 0.01 I0406 09:14:04.435921 5226 solver.cpp:218] Iteration 14508 (2.26554 iter/s, 5.29675s/12 iters), loss = 1.72575 I0406 09:14:04.436141 5226 solver.cpp:237] Train net output #0: loss = 1.72575 (* 1 = 1.72575 loss) I0406 09:14:04.436147 5226 sgd_solver.cpp:105] Iteration 14508, lr = 0.01 I0406 09:14:09.510411 5226 solver.cpp:218] Iteration 14520 (2.36489 iter/s, 5.07422s/12 iters), loss = 2.38361 I0406 09:14:09.510458 5226 solver.cpp:237] Train net output #0: loss = 2.38361 (* 1 = 2.38361 loss) I0406 09:14:09.510466 5226 sgd_solver.cpp:105] Iteration 14520, lr = 0.01 I0406 09:14:14.804949 5226 solver.cpp:218] Iteration 14532 (2.26653 iter/s, 5.29444s/12 iters), loss = 2.61797 I0406 09:14:14.805265 5226 solver.cpp:237] Train net output #0: loss = 2.61797 (* 1 = 2.61797 loss) I0406 09:14:14.805277 5226 sgd_solver.cpp:105] Iteration 14532, lr = 0.01 I0406 09:14:20.027874 5226 solver.cpp:218] Iteration 14544 (2.29772 iter/s, 5.22257s/12 iters), loss = 2.36757 I0406 09:14:20.027911 5226 solver.cpp:237] Train net output #0: loss = 2.36757 (* 1 = 2.36757 loss) I0406 09:14:20.027917 5226 sgd_solver.cpp:105] Iteration 14544, lr = 0.01 I0406 09:14:25.457240 5226 solver.cpp:218] Iteration 14556 (2.21024 iter/s, 5.42928s/12 iters), loss = 2.39384 I0406 09:14:25.457293 5226 solver.cpp:237] Train net output #0: loss = 2.39384 (* 1 = 2.39384 loss) I0406 09:14:25.457302 5226 sgd_solver.cpp:105] Iteration 14556, lr = 0.01 I0406 09:14:29.470706 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:14:29.715241 5226 blocking_queue.cpp:49] Waiting for data I0406 09:14:30.603055 5226 solver.cpp:218] Iteration 14568 (2.33204 iter/s, 5.14572s/12 iters), loss = 2.43081 I0406 09:14:30.603096 5226 solver.cpp:237] Train net output #0: loss = 2.43081 (* 1 = 2.43081 loss) I0406 09:14:30.603101 5226 sgd_solver.cpp:105] Iteration 14568, lr = 0.01 I0406 09:14:35.699571 5226 solver.cpp:218] Iteration 14580 (2.35459 iter/s, 5.09643s/12 iters), loss = 2.08512 I0406 09:14:35.699673 5226 solver.cpp:237] Train net output #0: loss = 2.08512 (* 1 = 2.08512 loss) I0406 09:14:35.699679 5226 sgd_solver.cpp:105] Iteration 14580, lr = 0.01 I0406 09:14:37.686326 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_14586.caffemodel I0406 09:14:40.951568 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_14586.solverstate I0406 09:14:43.353724 5226 solver.cpp:330] Iteration 14586, Testing net (#0) I0406 09:14:43.353744 5226 net.cpp:676] Ignoring source layer train-data I0406 09:14:46.618400 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:14:47.730461 5226 solver.cpp:397] Test net output #0: accuracy = 0.20527 I0406 09:14:47.730495 5226 solver.cpp:397] Test net output #1: loss = 3.96151 (* 1 = 3.96151 loss) I0406 09:14:49.622854 5226 solver.cpp:218] Iteration 14592 (0.861878 iter/s, 13.9231s/12 iters), loss = 1.99538 I0406 09:14:49.622896 5226 solver.cpp:237] Train net output #0: loss = 1.99538 (* 1 = 1.99538 loss) I0406 09:14:49.622902 5226 sgd_solver.cpp:105] Iteration 14592, lr = 0.01 I0406 09:14:54.920375 5226 solver.cpp:218] Iteration 14604 (2.26525 iter/s, 5.29743s/12 iters), loss = 2.3153 I0406 09:14:54.920415 5226 solver.cpp:237] Train net output #0: loss = 2.3153 (* 1 = 2.3153 loss) I0406 09:14:54.920420 5226 sgd_solver.cpp:105] Iteration 14604, lr = 0.01 I0406 09:15:00.224246 5226 solver.cpp:218] Iteration 14616 (2.26253 iter/s, 5.30379s/12 iters), loss = 3.94873 I0406 09:15:00.224283 5226 solver.cpp:237] Train net output #0: loss = 3.94873 (* 1 = 3.94873 loss) I0406 09:15:00.224289 5226 sgd_solver.cpp:105] Iteration 14616, lr = 0.01 I0406 09:15:05.285670 5226 solver.cpp:218] Iteration 14628 (2.37091 iter/s, 5.06134s/12 iters), loss = 2.2769 I0406 09:15:05.285714 5226 solver.cpp:237] Train net output #0: loss = 2.2769 (* 1 = 2.2769 loss) I0406 09:15:05.285722 5226 sgd_solver.cpp:105] Iteration 14628, lr = 0.01 I0406 09:15:10.546106 5226 solver.cpp:218] Iteration 14640 (2.28122 iter/s, 5.26034s/12 iters), loss = 1.92349 I0406 09:15:10.546252 5226 solver.cpp:237] Train net output #0: loss = 1.92349 (* 1 = 1.92349 loss) I0406 09:15:10.546259 5226 sgd_solver.cpp:105] Iteration 14640, lr = 0.01 I0406 09:15:15.747860 5226 solver.cpp:218] Iteration 14652 (2.307 iter/s, 5.20156s/12 iters), loss = 2.65605 I0406 09:15:15.747902 5226 solver.cpp:237] Train net output #0: loss = 2.65605 (* 1 = 2.65605 loss) I0406 09:15:15.747910 5226 sgd_solver.cpp:105] Iteration 14652, lr = 0.01 I0406 09:15:21.098929 5226 solver.cpp:218] Iteration 14664 (2.24258 iter/s, 5.35098s/12 iters), loss = 2.44919 I0406 09:15:21.098969 5226 solver.cpp:237] Train net output #0: loss = 2.44919 (* 1 = 2.44919 loss) I0406 09:15:21.098975 5226 sgd_solver.cpp:105] Iteration 14664, lr = 0.01 I0406 09:15:22.196290 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:15:26.346282 5226 solver.cpp:218] Iteration 14676 (2.2869 iter/s, 5.24727s/12 iters), loss = 2.45059 I0406 09:15:26.346321 5226 solver.cpp:237] Train net output #0: loss = 2.45059 (* 1 = 2.45059 loss) I0406 09:15:26.346326 5226 sgd_solver.cpp:105] Iteration 14676, lr = 0.01 I0406 09:15:30.952576 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_14688.caffemodel I0406 09:15:34.012209 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_14688.solverstate I0406 09:15:36.319867 5226 solver.cpp:330] Iteration 14688, Testing net (#0) I0406 09:15:36.319886 5226 net.cpp:676] Ignoring source layer train-data I0406 09:15:39.660727 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:15:40.849191 5226 solver.cpp:397] Test net output #0: accuracy = 0.186275 I0406 09:15:40.849269 5226 solver.cpp:397] Test net output #1: loss = 3.9224 (* 1 = 3.9224 loss) I0406 09:15:40.989961 5226 solver.cpp:218] Iteration 14688 (0.819474 iter/s, 14.6435s/12 iters), loss = 2.23342 I0406 09:15:40.989997 5226 solver.cpp:237] Train net output #0: loss = 2.23342 (* 1 = 2.23342 loss) I0406 09:15:40.990002 5226 sgd_solver.cpp:105] Iteration 14688, lr = 0.01 I0406 09:15:45.192642 5226 solver.cpp:218] Iteration 14700 (2.85538 iter/s, 4.2026s/12 iters), loss = 2.71329 I0406 09:15:45.192680 5226 solver.cpp:237] Train net output #0: loss = 2.71329 (* 1 = 2.71329 loss) I0406 09:15:45.192685 5226 sgd_solver.cpp:105] Iteration 14700, lr = 0.01 I0406 09:15:50.747629 5226 solver.cpp:218] Iteration 14712 (2.16026 iter/s, 5.5549s/12 iters), loss = 3.03435 I0406 09:15:50.747668 5226 solver.cpp:237] Train net output #0: loss = 3.03435 (* 1 = 3.03435 loss) I0406 09:15:50.747674 5226 sgd_solver.cpp:105] Iteration 14712, lr = 0.01 I0406 09:15:55.900198 5226 solver.cpp:218] Iteration 14724 (2.32897 iter/s, 5.15248s/12 iters), loss = 2.28689 I0406 09:15:55.900234 5226 solver.cpp:237] Train net output #0: loss = 2.28689 (* 1 = 2.28689 loss) I0406 09:15:55.900240 5226 sgd_solver.cpp:105] Iteration 14724, lr = 0.01 I0406 09:16:01.137030 5226 solver.cpp:218] Iteration 14736 (2.2915 iter/s, 5.23674s/12 iters), loss = 2.14654 I0406 09:16:01.137079 5226 solver.cpp:237] Train net output #0: loss = 2.14654 (* 1 = 2.14654 loss) I0406 09:16:01.137085 5226 sgd_solver.cpp:105] Iteration 14736, lr = 0.01 I0406 09:16:06.478664 5226 solver.cpp:218] Iteration 14748 (2.24654 iter/s, 5.34154s/12 iters), loss = 2.59674 I0406 09:16:06.478706 5226 solver.cpp:237] Train net output #0: loss = 2.59674 (* 1 = 2.59674 loss) I0406 09:16:06.478713 5226 sgd_solver.cpp:105] Iteration 14748, lr = 0.01 I0406 09:16:11.556572 5226 solver.cpp:218] Iteration 14760 (2.36322 iter/s, 5.07782s/12 iters), loss = 2.39623 I0406 09:16:11.556695 5226 solver.cpp:237] Train net output #0: loss = 2.39623 (* 1 = 2.39623 loss) I0406 09:16:11.556702 5226 sgd_solver.cpp:105] Iteration 14760, lr = 0.01 I0406 09:16:14.906227 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:16:16.736292 5226 solver.cpp:218] Iteration 14772 (2.3168 iter/s, 5.17955s/12 iters), loss = 2.43255 I0406 09:16:16.736330 5226 solver.cpp:237] Train net output #0: loss = 2.43255 (* 1 = 2.43255 loss) I0406 09:16:16.736335 5226 sgd_solver.cpp:105] Iteration 14772, lr = 0.01 I0406 09:16:21.715226 5226 solver.cpp:218] Iteration 14784 (2.4102 iter/s, 4.97884s/12 iters), loss = 1.70921 I0406 09:16:21.715283 5226 solver.cpp:237] Train net output #0: loss = 1.70921 (* 1 = 1.70921 loss) I0406 09:16:21.715291 5226 sgd_solver.cpp:105] Iteration 14784, lr = 0.01 I0406 09:16:23.955961 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_14790.caffemodel I0406 09:16:26.951361 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_14790.solverstate I0406 09:16:29.268870 5226 solver.cpp:330] Iteration 14790, Testing net (#0) I0406 09:16:29.268900 5226 net.cpp:676] Ignoring source layer train-data I0406 09:16:32.498005 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:16:33.700686 5226 solver.cpp:397] Test net output #0: accuracy = 0.207108 I0406 09:16:33.700718 5226 solver.cpp:397] Test net output #1: loss = 3.95197 (* 1 = 3.95197 loss) I0406 09:16:35.555115 5226 solver.cpp:218] Iteration 14796 (0.867068 iter/s, 13.8397s/12 iters), loss = 2.19746 I0406 09:16:35.555148 5226 solver.cpp:237] Train net output #0: loss = 2.19746 (* 1 = 2.19746 loss) I0406 09:16:35.555153 5226 sgd_solver.cpp:105] Iteration 14796, lr = 0.01 I0406 09:16:40.647352 5226 solver.cpp:218] Iteration 14808 (2.35657 iter/s, 5.09216s/12 iters), loss = 2.64486 I0406 09:16:40.647392 5226 solver.cpp:237] Train net output #0: loss = 2.64486 (* 1 = 2.64486 loss) I0406 09:16:40.647397 5226 sgd_solver.cpp:105] Iteration 14808, lr = 0.01 I0406 09:16:45.768039 5226 solver.cpp:218] Iteration 14820 (2.34348 iter/s, 5.12059s/12 iters), loss = 2.31862 I0406 09:16:45.768158 5226 solver.cpp:237] Train net output #0: loss = 2.31862 (* 1 = 2.31862 loss) I0406 09:16:45.768167 5226 sgd_solver.cpp:105] Iteration 14820, lr = 0.01 I0406 09:16:50.991766 5226 solver.cpp:218] Iteration 14832 (2.29728 iter/s, 5.22356s/12 iters), loss = 1.9846 I0406 09:16:50.991802 5226 solver.cpp:237] Train net output #0: loss = 1.9846 (* 1 = 1.9846 loss) I0406 09:16:50.991808 5226 sgd_solver.cpp:105] Iteration 14832, lr = 0.01 I0406 09:16:56.397054 5226 solver.cpp:218] Iteration 14844 (2.22008 iter/s, 5.4052s/12 iters), loss = 2.37669 I0406 09:16:56.397090 5226 solver.cpp:237] Train net output #0: loss = 2.37669 (* 1 = 2.37669 loss) I0406 09:16:56.397095 5226 sgd_solver.cpp:105] Iteration 14844, lr = 0.01 I0406 09:17:01.716208 5226 solver.cpp:218] Iteration 14856 (2.25603 iter/s, 5.31907s/12 iters), loss = 2.46445 I0406 09:17:01.716255 5226 solver.cpp:237] Train net output #0: loss = 2.46445 (* 1 = 2.46445 loss) I0406 09:17:01.716260 5226 sgd_solver.cpp:105] Iteration 14856, lr = 0.01 I0406 09:17:06.970147 5226 solver.cpp:218] Iteration 14868 (2.28404 iter/s, 5.25385s/12 iters), loss = 2.61024 I0406 09:17:06.970188 5226 solver.cpp:237] Train net output #0: loss = 2.61024 (* 1 = 2.61024 loss) I0406 09:17:06.970193 5226 sgd_solver.cpp:105] Iteration 14868, lr = 0.01 I0406 09:17:07.268007 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:17:12.335629 5226 solver.cpp:218] Iteration 14880 (2.23656 iter/s, 5.36539s/12 iters), loss = 2.44326 I0406 09:17:12.335671 5226 solver.cpp:237] Train net output #0: loss = 2.44326 (* 1 = 2.44326 loss) I0406 09:17:12.335680 5226 sgd_solver.cpp:105] Iteration 14880, lr = 0.01 I0406 09:17:17.105007 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_14892.caffemodel I0406 09:17:20.123386 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_14892.solverstate I0406 09:17:22.431665 5226 solver.cpp:330] Iteration 14892, Testing net (#0) I0406 09:17:22.431684 5226 net.cpp:676] Ignoring source layer train-data I0406 09:17:25.541198 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:17:26.833526 5226 solver.cpp:397] Test net output #0: accuracy = 0.196691 I0406 09:17:26.833560 5226 solver.cpp:397] Test net output #1: loss = 3.94578 (* 1 = 3.94578 loss) I0406 09:17:26.968325 5226 solver.cpp:218] Iteration 14892 (0.820089 iter/s, 14.6326s/12 iters), loss = 2.20364 I0406 09:17:26.968370 5226 solver.cpp:237] Train net output #0: loss = 2.20364 (* 1 = 2.20364 loss) I0406 09:17:26.968377 5226 sgd_solver.cpp:105] Iteration 14892, lr = 0.01 I0406 09:17:31.358001 5226 solver.cpp:218] Iteration 14904 (2.73374 iter/s, 4.38959s/12 iters), loss = 2.41026 I0406 09:17:31.358039 5226 solver.cpp:237] Train net output #0: loss = 2.41026 (* 1 = 2.41026 loss) I0406 09:17:31.358045 5226 sgd_solver.cpp:105] Iteration 14904, lr = 0.01 I0406 09:17:36.484997 5226 solver.cpp:218] Iteration 14916 (2.34059 iter/s, 5.12691s/12 iters), loss = 2.39415 I0406 09:17:36.485051 5226 solver.cpp:237] Train net output #0: loss = 2.39415 (* 1 = 2.39415 loss) I0406 09:17:36.485059 5226 sgd_solver.cpp:105] Iteration 14916, lr = 0.01 I0406 09:17:41.760682 5226 solver.cpp:218] Iteration 14928 (2.27463 iter/s, 5.27558s/12 iters), loss = 2.1506 I0406 09:17:41.760735 5226 solver.cpp:237] Train net output #0: loss = 2.1506 (* 1 = 2.1506 loss) I0406 09:17:41.760746 5226 sgd_solver.cpp:105] Iteration 14928, lr = 0.01 I0406 09:17:46.966197 5226 solver.cpp:218] Iteration 14940 (2.30529 iter/s, 5.20541s/12 iters), loss = 2.83912 I0406 09:17:46.966245 5226 solver.cpp:237] Train net output #0: loss = 2.83912 (* 1 = 2.83912 loss) I0406 09:17:46.966253 5226 sgd_solver.cpp:105] Iteration 14940, lr = 0.01 I0406 09:17:52.316517 5226 solver.cpp:218] Iteration 14952 (2.24289 iter/s, 5.35023s/12 iters), loss = 2.47864 I0406 09:17:52.316603 5226 solver.cpp:237] Train net output #0: loss = 2.47864 (* 1 = 2.47864 loss) I0406 09:17:52.316612 5226 sgd_solver.cpp:105] Iteration 14952, lr = 0.01 I0406 09:17:57.693423 5226 solver.cpp:218] Iteration 14964 (2.23182 iter/s, 5.37677s/12 iters), loss = 2.31652 I0406 09:17:57.693459 5226 solver.cpp:237] Train net output #0: loss = 2.31652 (* 1 = 2.31652 loss) I0406 09:17:57.693465 5226 sgd_solver.cpp:105] Iteration 14964, lr = 0.01 I0406 09:18:00.402683 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:18:03.090602 5226 solver.cpp:218] Iteration 14976 (2.22342 iter/s, 5.39709s/12 iters), loss = 2.91963 I0406 09:18:03.090662 5226 solver.cpp:237] Train net output #0: loss = 2.91963 (* 1 = 2.91963 loss) I0406 09:18:03.090670 5226 sgd_solver.cpp:105] Iteration 14976, lr = 0.01 I0406 09:18:08.442405 5226 solver.cpp:218] Iteration 14988 (2.24228 iter/s, 5.3517s/12 iters), loss = 2.49171 I0406 09:18:08.442445 5226 solver.cpp:237] Train net output #0: loss = 2.49171 (* 1 = 2.49171 loss) I0406 09:18:08.442451 5226 sgd_solver.cpp:105] Iteration 14988, lr = 0.01 I0406 09:18:10.659570 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_14994.caffemodel I0406 09:18:13.696950 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_14994.solverstate I0406 09:18:16.008486 5226 solver.cpp:330] Iteration 14994, Testing net (#0) I0406 09:18:16.008504 5226 net.cpp:676] Ignoring source layer train-data I0406 09:18:19.016944 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:18:20.351589 5226 solver.cpp:397] Test net output #0: accuracy = 0.186887 I0406 09:18:20.351624 5226 solver.cpp:397] Test net output #1: loss = 4.0286 (* 1 = 4.0286 loss) I0406 09:18:22.124931 5226 solver.cpp:218] Iteration 15000 (0.877039 iter/s, 13.6824s/12 iters), loss = 2.57134 I0406 09:18:22.124965 5226 solver.cpp:237] Train net output #0: loss = 2.57134 (* 1 = 2.57134 loss) I0406 09:18:22.124971 5226 sgd_solver.cpp:105] Iteration 15000, lr = 0.01 I0406 09:18:27.247741 5226 solver.cpp:218] Iteration 15012 (2.3425 iter/s, 5.12273s/12 iters), loss = 2.50652 I0406 09:18:27.247875 5226 solver.cpp:237] Train net output #0: loss = 2.50652 (* 1 = 2.50652 loss) I0406 09:18:27.247885 5226 sgd_solver.cpp:105] Iteration 15012, lr = 0.01 I0406 09:18:32.524755 5226 solver.cpp:218] Iteration 15024 (2.27409 iter/s, 5.27683s/12 iters), loss = 2.14104 I0406 09:18:32.524796 5226 solver.cpp:237] Train net output #0: loss = 2.14104 (* 1 = 2.14104 loss) I0406 09:18:32.524802 5226 sgd_solver.cpp:105] Iteration 15024, lr = 0.01 I0406 09:18:37.882902 5226 solver.cpp:218] Iteration 15036 (2.23961 iter/s, 5.35806s/12 iters), loss = 2.29173 I0406 09:18:37.882938 5226 solver.cpp:237] Train net output #0: loss = 2.29173 (* 1 = 2.29173 loss) I0406 09:18:37.882943 5226 sgd_solver.cpp:105] Iteration 15036, lr = 0.01 I0406 09:18:43.310381 5226 solver.cpp:218] Iteration 15048 (2.21101 iter/s, 5.42739s/12 iters), loss = 2.10765 I0406 09:18:43.310436 5226 solver.cpp:237] Train net output #0: loss = 2.10765 (* 1 = 2.10765 loss) I0406 09:18:43.310444 5226 sgd_solver.cpp:105] Iteration 15048, lr = 0.01 I0406 09:18:48.697759 5226 solver.cpp:218] Iteration 15060 (2.22747 iter/s, 5.38727s/12 iters), loss = 2.27025 I0406 09:18:48.697804 5226 solver.cpp:237] Train net output #0: loss = 2.27025 (* 1 = 2.27025 loss) I0406 09:18:48.697813 5226 sgd_solver.cpp:105] Iteration 15060, lr = 0.01 I0406 09:18:53.620030 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:18:53.987581 5226 solver.cpp:218] Iteration 15072 (2.26855 iter/s, 5.28973s/12 iters), loss = 2.44537 I0406 09:18:53.987619 5226 solver.cpp:237] Train net output #0: loss = 2.44537 (* 1 = 2.44537 loss) I0406 09:18:53.987625 5226 sgd_solver.cpp:105] Iteration 15072, lr = 0.01 I0406 09:18:59.247442 5226 solver.cpp:218] Iteration 15084 (2.28147 iter/s, 5.25977s/12 iters), loss = 2.55182 I0406 09:18:59.247560 5226 solver.cpp:237] Train net output #0: loss = 2.55182 (* 1 = 2.55182 loss) I0406 09:18:59.247568 5226 sgd_solver.cpp:105] Iteration 15084, lr = 0.01 I0406 09:19:04.022390 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_15096.caffemodel I0406 09:19:07.022927 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_15096.solverstate I0406 09:19:09.320590 5226 solver.cpp:330] Iteration 15096, Testing net (#0) I0406 09:19:09.320609 5226 net.cpp:676] Ignoring source layer train-data I0406 09:19:12.397886 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:19:13.749411 5226 solver.cpp:397] Test net output #0: accuracy = 0.172794 I0406 09:19:13.749449 5226 solver.cpp:397] Test net output #1: loss = 4.14015 (* 1 = 4.14015 loss) I0406 09:19:13.889973 5226 solver.cpp:218] Iteration 15096 (0.819542 iter/s, 14.6423s/12 iters), loss = 2.67633 I0406 09:19:13.890013 5226 solver.cpp:237] Train net output #0: loss = 2.67633 (* 1 = 2.67633 loss) I0406 09:19:13.890018 5226 sgd_solver.cpp:105] Iteration 15096, lr = 0.01 I0406 09:19:18.247953 5226 solver.cpp:218] Iteration 15108 (2.75362 iter/s, 4.35789s/12 iters), loss = 2.38707 I0406 09:19:18.248005 5226 solver.cpp:237] Train net output #0: loss = 2.38707 (* 1 = 2.38707 loss) I0406 09:19:18.248013 5226 sgd_solver.cpp:105] Iteration 15108, lr = 0.01 I0406 09:19:23.507370 5226 solver.cpp:218] Iteration 15120 (2.28167 iter/s, 5.25931s/12 iters), loss = 2.49154 I0406 09:19:23.507416 5226 solver.cpp:237] Train net output #0: loss = 2.49154 (* 1 = 2.49154 loss) I0406 09:19:23.507422 5226 sgd_solver.cpp:105] Iteration 15120, lr = 0.01 I0406 09:19:28.733196 5226 solver.cpp:218] Iteration 15132 (2.29633 iter/s, 5.22573s/12 iters), loss = 2.67081 I0406 09:19:28.733233 5226 solver.cpp:237] Train net output #0: loss = 2.67081 (* 1 = 2.67081 loss) I0406 09:19:28.733239 5226 sgd_solver.cpp:105] Iteration 15132, lr = 0.01 I0406 09:19:33.988365 5226 solver.cpp:218] Iteration 15144 (2.2835 iter/s, 5.25508s/12 iters), loss = 2.34375 I0406 09:19:33.988521 5226 solver.cpp:237] Train net output #0: loss = 2.34375 (* 1 = 2.34375 loss) I0406 09:19:33.988531 5226 sgd_solver.cpp:105] Iteration 15144, lr = 0.01 I0406 09:19:39.321030 5226 solver.cpp:218] Iteration 15156 (2.25037 iter/s, 5.33247s/12 iters), loss = 2.33162 I0406 09:19:39.321069 5226 solver.cpp:237] Train net output #0: loss = 2.33162 (* 1 = 2.33162 loss) I0406 09:19:39.321074 5226 sgd_solver.cpp:105] Iteration 15156, lr = 0.01 I0406 09:19:44.450824 5226 solver.cpp:218] Iteration 15168 (2.33932 iter/s, 5.1297s/12 iters), loss = 2.21661 I0406 09:19:44.450866 5226 solver.cpp:237] Train net output #0: loss = 2.21661 (* 1 = 2.21661 loss) I0406 09:19:44.450875 5226 sgd_solver.cpp:105] Iteration 15168, lr = 0.01 I0406 09:19:46.396755 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:19:49.702390 5226 solver.cpp:218] Iteration 15180 (2.28507 iter/s, 5.25147s/12 iters), loss = 2.24728 I0406 09:19:49.702442 5226 solver.cpp:237] Train net output #0: loss = 2.24728 (* 1 = 2.24728 loss) I0406 09:19:49.702450 5226 sgd_solver.cpp:105] Iteration 15180, lr = 0.01 I0406 09:19:55.229676 5226 solver.cpp:218] Iteration 15192 (2.17109 iter/s, 5.52718s/12 iters), loss = 2.48743 I0406 09:19:55.229720 5226 solver.cpp:237] Train net output #0: loss = 2.48743 (* 1 = 2.48743 loss) I0406 09:19:55.229728 5226 sgd_solver.cpp:105] Iteration 15192, lr = 0.01 I0406 09:19:57.417174 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_15198.caffemodel I0406 09:20:00.501956 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_15198.solverstate I0406 09:20:02.801457 5226 solver.cpp:330] Iteration 15198, Testing net (#0) I0406 09:20:02.801477 5226 net.cpp:676] Ignoring source layer train-data I0406 09:20:05.848068 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:20:07.345039 5226 solver.cpp:397] Test net output #0: accuracy = 0.220588 I0406 09:20:07.345068 5226 solver.cpp:397] Test net output #1: loss = 3.85039 (* 1 = 3.85039 loss) I0406 09:20:09.301525 5226 solver.cpp:218] Iteration 15204 (0.852775 iter/s, 14.0717s/12 iters), loss = 2.9643 I0406 09:20:09.301564 5226 solver.cpp:237] Train net output #0: loss = 2.9643 (* 1 = 2.9643 loss) I0406 09:20:09.301570 5226 sgd_solver.cpp:105] Iteration 15204, lr = 0.01 I0406 09:20:14.589720 5226 solver.cpp:218] Iteration 15216 (2.26924 iter/s, 5.2881s/12 iters), loss = 2.35505 I0406 09:20:14.589766 5226 solver.cpp:237] Train net output #0: loss = 2.35505 (* 1 = 2.35505 loss) I0406 09:20:14.589774 5226 sgd_solver.cpp:105] Iteration 15216, lr = 0.01 I0406 09:20:19.901767 5226 solver.cpp:218] Iteration 15228 (2.25906 iter/s, 5.31195s/12 iters), loss = 2.2205 I0406 09:20:19.901806 5226 solver.cpp:237] Train net output #0: loss = 2.2205 (* 1 = 2.2205 loss) I0406 09:20:19.901811 5226 sgd_solver.cpp:105] Iteration 15228, lr = 0.01 I0406 09:20:25.255931 5226 solver.cpp:218] Iteration 15240 (2.24128 iter/s, 5.35408s/12 iters), loss = 2.34903 I0406 09:20:25.255972 5226 solver.cpp:237] Train net output #0: loss = 2.34903 (* 1 = 2.34903 loss) I0406 09:20:25.255977 5226 sgd_solver.cpp:105] Iteration 15240, lr = 0.01 I0406 09:20:30.022627 5226 blocking_queue.cpp:49] Waiting for data I0406 09:20:30.556344 5226 solver.cpp:218] Iteration 15252 (2.26401 iter/s, 5.30032s/12 iters), loss = 2.69788 I0406 09:20:30.556396 5226 solver.cpp:237] Train net output #0: loss = 2.69788 (* 1 = 2.69788 loss) I0406 09:20:30.556403 5226 sgd_solver.cpp:105] Iteration 15252, lr = 0.01 I0406 09:20:35.737082 5226 solver.cpp:218] Iteration 15264 (2.31631 iter/s, 5.18065s/12 iters), loss = 2.11559 I0406 09:20:35.737118 5226 solver.cpp:237] Train net output #0: loss = 2.11559 (* 1 = 2.11559 loss) I0406 09:20:35.737123 5226 sgd_solver.cpp:105] Iteration 15264, lr = 0.01 I0406 09:20:39.957659 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:20:41.120762 5226 solver.cpp:218] Iteration 15276 (2.229 iter/s, 5.38359s/12 iters), loss = 2.74078 I0406 09:20:41.120810 5226 solver.cpp:237] Train net output #0: loss = 2.74078 (* 1 = 2.74078 loss) I0406 09:20:41.120816 5226 sgd_solver.cpp:105] Iteration 15276, lr = 0.01 I0406 09:20:46.350329 5226 solver.cpp:218] Iteration 15288 (2.29469 iter/s, 5.22947s/12 iters), loss = 2.18095 I0406 09:20:46.350394 5226 solver.cpp:237] Train net output #0: loss = 2.18095 (* 1 = 2.18095 loss) I0406 09:20:46.350404 5226 sgd_solver.cpp:105] Iteration 15288, lr = 0.01 I0406 09:20:51.213483 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_15300.caffemodel I0406 09:20:54.249163 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_15300.solverstate I0406 09:20:56.560293 5226 solver.cpp:330] Iteration 15300, Testing net (#0) I0406 09:20:56.560314 5226 net.cpp:676] Ignoring source layer train-data I0406 09:20:59.546308 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:21:00.911293 5226 solver.cpp:397] Test net output #0: accuracy = 0.186887 I0406 09:21:00.911329 5226 solver.cpp:397] Test net output #1: loss = 4.08283 (* 1 = 4.08283 loss) I0406 09:21:01.047930 5226 solver.cpp:218] Iteration 15300 (0.816469 iter/s, 14.6974s/12 iters), loss = 2.78283 I0406 09:21:01.047977 5226 solver.cpp:237] Train net output #0: loss = 2.78283 (* 1 = 2.78283 loss) I0406 09:21:01.047984 5226 sgd_solver.cpp:105] Iteration 15300, lr = 0.01 I0406 09:21:05.214291 5226 solver.cpp:218] Iteration 15312 (2.88027 iter/s, 4.16628s/12 iters), loss = 2.17308 I0406 09:21:05.214331 5226 solver.cpp:237] Train net output #0: loss = 2.17308 (* 1 = 2.17308 loss) I0406 09:21:05.214337 5226 sgd_solver.cpp:105] Iteration 15312, lr = 0.01 I0406 09:21:10.382927 5226 solver.cpp:218] Iteration 15324 (2.32173 iter/s, 5.16855s/12 iters), loss = 2.40783 I0406 09:21:10.383041 5226 solver.cpp:237] Train net output #0: loss = 2.40783 (* 1 = 2.40783 loss) I0406 09:21:10.383051 5226 sgd_solver.cpp:105] Iteration 15324, lr = 0.01 I0406 09:21:15.723915 5226 solver.cpp:218] Iteration 15336 (2.24685 iter/s, 5.34082s/12 iters), loss = 2.73631 I0406 09:21:15.723965 5226 solver.cpp:237] Train net output #0: loss = 2.73631 (* 1 = 2.73631 loss) I0406 09:21:15.723973 5226 sgd_solver.cpp:105] Iteration 15336, lr = 0.01 I0406 09:21:21.042649 5226 solver.cpp:218] Iteration 15348 (2.25622 iter/s, 5.31864s/12 iters), loss = 2.58036 I0406 09:21:21.042690 5226 solver.cpp:237] Train net output #0: loss = 2.58036 (* 1 = 2.58036 loss) I0406 09:21:21.042695 5226 sgd_solver.cpp:105] Iteration 15348, lr = 0.01 I0406 09:21:26.416411 5226 solver.cpp:218] Iteration 15360 (2.23311 iter/s, 5.37367s/12 iters), loss = 2.95746 I0406 09:21:26.416460 5226 solver.cpp:237] Train net output #0: loss = 2.95746 (* 1 = 2.95746 loss) I0406 09:21:26.416465 5226 sgd_solver.cpp:105] Iteration 15360, lr = 0.01 I0406 09:21:31.639911 5226 solver.cpp:218] Iteration 15372 (2.29735 iter/s, 5.22341s/12 iters), loss = 2.75997 I0406 09:21:31.639948 5226 solver.cpp:237] Train net output #0: loss = 2.75997 (* 1 = 2.75997 loss) I0406 09:21:31.639953 5226 sgd_solver.cpp:105] Iteration 15372, lr = 0.01 I0406 09:21:32.775804 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:21:37.087393 5226 solver.cpp:218] Iteration 15384 (2.20289 iter/s, 5.44739s/12 iters), loss = 2.73533 I0406 09:21:37.087451 5226 solver.cpp:237] Train net output #0: loss = 2.73533 (* 1 = 2.73533 loss) I0406 09:21:37.087460 5226 sgd_solver.cpp:105] Iteration 15384, lr = 0.01 I0406 09:21:42.339174 5226 solver.cpp:218] Iteration 15396 (2.28499 iter/s, 5.25167s/12 iters), loss = 2.5165 I0406 09:21:42.339334 5226 solver.cpp:237] Train net output #0: loss = 2.5165 (* 1 = 2.5165 loss) I0406 09:21:42.339344 5226 sgd_solver.cpp:105] Iteration 15396, lr = 0.01 I0406 09:21:44.460127 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_15402.caffemodel I0406 09:21:47.484427 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_15402.solverstate I0406 09:21:49.812829 5226 solver.cpp:330] Iteration 15402, Testing net (#0) I0406 09:21:49.812846 5226 net.cpp:676] Ignoring source layer train-data I0406 09:21:52.724072 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:21:54.157788 5226 solver.cpp:397] Test net output #0: accuracy = 0.179534 I0406 09:21:54.157826 5226 solver.cpp:397] Test net output #1: loss = 4.16302 (* 1 = 4.16302 loss) I0406 09:21:55.918220 5226 solver.cpp:218] Iteration 15408 (0.883731 iter/s, 13.5788s/12 iters), loss = 2.98151 I0406 09:21:55.918257 5226 solver.cpp:237] Train net output #0: loss = 2.98151 (* 1 = 2.98151 loss) I0406 09:21:55.918263 5226 sgd_solver.cpp:105] Iteration 15408, lr = 0.01 I0406 09:22:01.181650 5226 solver.cpp:218] Iteration 15420 (2.27992 iter/s, 5.26334s/12 iters), loss = 2.73224 I0406 09:22:01.181697 5226 solver.cpp:237] Train net output #0: loss = 2.73224 (* 1 = 2.73224 loss) I0406 09:22:01.181705 5226 sgd_solver.cpp:105] Iteration 15420, lr = 0.01 I0406 09:22:06.605823 5226 solver.cpp:218] Iteration 15432 (2.21236 iter/s, 5.42408s/12 iters), loss = 2.54501 I0406 09:22:06.605860 5226 solver.cpp:237] Train net output #0: loss = 2.54501 (* 1 = 2.54501 loss) I0406 09:22:06.605866 5226 sgd_solver.cpp:105] Iteration 15432, lr = 0.01 I0406 09:22:11.951035 5226 solver.cpp:218] Iteration 15444 (2.24504 iter/s, 5.34513s/12 iters), loss = 2.64866 I0406 09:22:11.951073 5226 solver.cpp:237] Train net output #0: loss = 2.64866 (* 1 = 2.64866 loss) I0406 09:22:11.951078 5226 sgd_solver.cpp:105] Iteration 15444, lr = 0.01 I0406 09:22:17.229717 5226 solver.cpp:218] Iteration 15456 (2.27333 iter/s, 5.27859s/12 iters), loss = 2.65273 I0406 09:22:17.229821 5226 solver.cpp:237] Train net output #0: loss = 2.65273 (* 1 = 2.65273 loss) I0406 09:22:17.229830 5226 sgd_solver.cpp:105] Iteration 15456, lr = 0.01 I0406 09:22:22.430706 5226 solver.cpp:218] Iteration 15468 (2.30732 iter/s, 5.20084s/12 iters), loss = 2.24809 I0406 09:22:22.430760 5226 solver.cpp:237] Train net output #0: loss = 2.24809 (* 1 = 2.24809 loss) I0406 09:22:22.430768 5226 sgd_solver.cpp:105] Iteration 15468, lr = 0.01 I0406 09:22:25.821961 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:22:27.789016 5226 solver.cpp:218] Iteration 15480 (2.23955 iter/s, 5.35821s/12 iters), loss = 3.06516 I0406 09:22:27.789068 5226 solver.cpp:237] Train net output #0: loss = 3.06516 (* 1 = 3.06516 loss) I0406 09:22:27.789077 5226 sgd_solver.cpp:105] Iteration 15480, lr = 0.01 I0406 09:22:33.179409 5226 solver.cpp:218] Iteration 15492 (2.22622 iter/s, 5.39029s/12 iters), loss = 2.2507 I0406 09:22:33.179453 5226 solver.cpp:237] Train net output #0: loss = 2.2507 (* 1 = 2.2507 loss) I0406 09:22:33.179459 5226 sgd_solver.cpp:105] Iteration 15492, lr = 0.01 I0406 09:22:38.117730 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_15504.caffemodel I0406 09:22:41.886471 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_15504.solverstate I0406 09:22:44.186674 5226 solver.cpp:330] Iteration 15504, Testing net (#0) I0406 09:22:44.186693 5226 net.cpp:676] Ignoring source layer train-data I0406 09:22:47.243474 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:22:48.715207 5226 solver.cpp:397] Test net output #0: accuracy = 0.175245 I0406 09:22:48.715241 5226 solver.cpp:397] Test net output #1: loss = 4.11399 (* 1 = 4.11399 loss) I0406 09:22:48.855852 5226 solver.cpp:218] Iteration 15504 (0.765487 iter/s, 15.6763s/12 iters), loss = 2.26472 I0406 09:22:48.855919 5226 solver.cpp:237] Train net output #0: loss = 2.26472 (* 1 = 2.26472 loss) I0406 09:22:48.855927 5226 sgd_solver.cpp:105] Iteration 15504, lr = 0.01 I0406 09:22:53.148175 5226 solver.cpp:218] Iteration 15516 (2.79576 iter/s, 4.29222s/12 iters), loss = 2.68273 I0406 09:22:53.148226 5226 solver.cpp:237] Train net output #0: loss = 2.68273 (* 1 = 2.68273 loss) I0406 09:22:53.148233 5226 sgd_solver.cpp:105] Iteration 15516, lr = 0.01 I0406 09:22:58.472012 5226 solver.cpp:218] Iteration 15528 (2.25406 iter/s, 5.32374s/12 iters), loss = 3.27756 I0406 09:22:58.472060 5226 solver.cpp:237] Train net output #0: loss = 3.27756 (* 1 = 3.27756 loss) I0406 09:22:58.472066 5226 sgd_solver.cpp:105] Iteration 15528, lr = 0.01 I0406 09:23:03.689270 5226 solver.cpp:218] Iteration 15540 (2.3001 iter/s, 5.21716s/12 iters), loss = 2.70905 I0406 09:23:03.689306 5226 solver.cpp:237] Train net output #0: loss = 2.70905 (* 1 = 2.70905 loss) I0406 09:23:03.689311 5226 sgd_solver.cpp:105] Iteration 15540, lr = 0.01 I0406 09:23:08.780079 5226 solver.cpp:218] Iteration 15552 (2.35723 iter/s, 5.09072s/12 iters), loss = 3.03579 I0406 09:23:08.780122 5226 solver.cpp:237] Train net output #0: loss = 3.03579 (* 1 = 3.03579 loss) I0406 09:23:08.780128 5226 sgd_solver.cpp:105] Iteration 15552, lr = 0.01 I0406 09:23:14.082496 5226 solver.cpp:218] Iteration 15564 (2.26316 iter/s, 5.30233s/12 iters), loss = 2.88521 I0406 09:23:14.082535 5226 solver.cpp:237] Train net output #0: loss = 2.88521 (* 1 = 2.88521 loss) I0406 09:23:14.082540 5226 sgd_solver.cpp:105] Iteration 15564, lr = 0.01 I0406 09:23:19.063846 5226 solver.cpp:218] Iteration 15576 (2.40903 iter/s, 4.98127s/12 iters), loss = 2.58066 I0406 09:23:19.063937 5226 solver.cpp:237] Train net output #0: loss = 2.58066 (* 1 = 2.58066 loss) I0406 09:23:19.063943 5226 sgd_solver.cpp:105] Iteration 15576, lr = 0.01 I0406 09:23:19.517916 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:23:24.326666 5226 solver.cpp:218] Iteration 15588 (2.28021 iter/s, 5.26268s/12 iters), loss = 2.96647 I0406 09:23:24.326706 5226 solver.cpp:237] Train net output #0: loss = 2.96647 (* 1 = 2.96647 loss) I0406 09:23:24.326711 5226 sgd_solver.cpp:105] Iteration 15588, lr = 0.01 I0406 09:23:29.554989 5226 solver.cpp:218] Iteration 15600 (2.29523 iter/s, 5.22823s/12 iters), loss = 2.60254 I0406 09:23:29.555037 5226 solver.cpp:237] Train net output #0: loss = 2.60254 (* 1 = 2.60254 loss) I0406 09:23:29.555045 5226 sgd_solver.cpp:105] Iteration 15600, lr = 0.01 I0406 09:23:31.653286 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_15606.caffemodel I0406 09:23:34.696998 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_15606.solverstate I0406 09:23:37.014838 5226 solver.cpp:330] Iteration 15606, Testing net (#0) I0406 09:23:37.014863 5226 net.cpp:676] Ignoring source layer train-data I0406 09:23:40.060835 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:23:41.554417 5226 solver.cpp:397] Test net output #0: accuracy = 0.175245 I0406 09:23:41.554456 5226 solver.cpp:397] Test net output #1: loss = 4.13645 (* 1 = 4.13645 loss) I0406 09:23:43.566253 5226 solver.cpp:218] Iteration 15612 (0.856463 iter/s, 14.0111s/12 iters), loss = 2.80721 I0406 09:23:43.566291 5226 solver.cpp:237] Train net output #0: loss = 2.80721 (* 1 = 2.80721 loss) I0406 09:23:43.566296 5226 sgd_solver.cpp:105] Iteration 15612, lr = 0.01 I0406 09:23:48.929411 5226 solver.cpp:218] Iteration 15624 (2.23753 iter/s, 5.36306s/12 iters), loss = 2.46183 I0406 09:23:48.929459 5226 solver.cpp:237] Train net output #0: loss = 2.46183 (* 1 = 2.46183 loss) I0406 09:23:48.929467 5226 sgd_solver.cpp:105] Iteration 15624, lr = 0.01 I0406 09:23:54.320904 5226 solver.cpp:218] Iteration 15636 (2.22577 iter/s, 5.39139s/12 iters), loss = 2.79031 I0406 09:23:54.321048 5226 solver.cpp:237] Train net output #0: loss = 2.79031 (* 1 = 2.79031 loss) I0406 09:23:54.321058 5226 sgd_solver.cpp:105] Iteration 15636, lr = 0.01 I0406 09:23:59.483031 5226 solver.cpp:218] Iteration 15648 (2.32471 iter/s, 5.16194s/12 iters), loss = 2.364 I0406 09:23:59.483067 5226 solver.cpp:237] Train net output #0: loss = 2.364 (* 1 = 2.364 loss) I0406 09:23:59.483072 5226 sgd_solver.cpp:105] Iteration 15648, lr = 0.01 I0406 09:24:04.723836 5226 solver.cpp:218] Iteration 15660 (2.28976 iter/s, 5.24072s/12 iters), loss = 2.516 I0406 09:24:04.723882 5226 solver.cpp:237] Train net output #0: loss = 2.516 (* 1 = 2.516 loss) I0406 09:24:04.723891 5226 sgd_solver.cpp:105] Iteration 15660, lr = 0.01 I0406 09:24:10.012141 5226 solver.cpp:218] Iteration 15672 (2.2692 iter/s, 5.28821s/12 iters), loss = 2.47083 I0406 09:24:10.012179 5226 solver.cpp:237] Train net output #0: loss = 2.47083 (* 1 = 2.47083 loss) I0406 09:24:10.012184 5226 sgd_solver.cpp:105] Iteration 15672, lr = 0.01 I0406 09:24:12.794190 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:24:15.433780 5226 solver.cpp:218] Iteration 15684 (2.21339 iter/s, 5.42155s/12 iters), loss = 2.50643 I0406 09:24:15.433818 5226 solver.cpp:237] Train net output #0: loss = 2.50643 (* 1 = 2.50643 loss) I0406 09:24:15.433823 5226 sgd_solver.cpp:105] Iteration 15684, lr = 0.01 I0406 09:24:20.868903 5226 solver.cpp:218] Iteration 15696 (2.2079 iter/s, 5.43503s/12 iters), loss = 2.96342 I0406 09:24:20.868942 5226 solver.cpp:237] Train net output #0: loss = 2.96342 (* 1 = 2.96342 loss) I0406 09:24:20.868949 5226 sgd_solver.cpp:105] Iteration 15696, lr = 0.01 I0406 09:24:25.595543 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_15708.caffemodel I0406 09:24:28.685792 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_15708.solverstate I0406 09:24:31.021780 5226 solver.cpp:330] Iteration 15708, Testing net (#0) I0406 09:24:31.021806 5226 net.cpp:676] Ignoring source layer train-data I0406 09:24:33.884191 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:24:35.513459 5226 solver.cpp:397] Test net output #0: accuracy = 0.179534 I0406 09:24:35.513496 5226 solver.cpp:397] Test net output #1: loss = 4.13492 (* 1 = 4.13492 loss) I0406 09:24:35.653733 5226 solver.cpp:218] Iteration 15708 (0.811651 iter/s, 14.7847s/12 iters), loss = 2.78042 I0406 09:24:35.653789 5226 solver.cpp:237] Train net output #0: loss = 2.78042 (* 1 = 2.78042 loss) I0406 09:24:35.653798 5226 sgd_solver.cpp:105] Iteration 15708, lr = 0.01 I0406 09:24:39.828651 5226 solver.cpp:218] Iteration 15720 (2.87437 iter/s, 4.17483s/12 iters), loss = 2.57444 I0406 09:24:39.828707 5226 solver.cpp:237] Train net output #0: loss = 2.57444 (* 1 = 2.57444 loss) I0406 09:24:39.828717 5226 sgd_solver.cpp:105] Iteration 15720, lr = 0.01 I0406 09:24:45.107062 5226 solver.cpp:218] Iteration 15732 (2.27346 iter/s, 5.27831s/12 iters), loss = 2.91971 I0406 09:24:45.107103 5226 solver.cpp:237] Train net output #0: loss = 2.91971 (* 1 = 2.91971 loss) I0406 09:24:45.107108 5226 sgd_solver.cpp:105] Iteration 15732, lr = 0.01 I0406 09:24:50.298753 5226 solver.cpp:218] Iteration 15744 (2.31143 iter/s, 5.1916s/12 iters), loss = 2.88435 I0406 09:24:50.298816 5226 solver.cpp:237] Train net output #0: loss = 2.88435 (* 1 = 2.88435 loss) I0406 09:24:50.298827 5226 sgd_solver.cpp:105] Iteration 15744, lr = 0.01 I0406 09:24:55.691529 5226 solver.cpp:218] Iteration 15756 (2.22524 iter/s, 5.39266s/12 iters), loss = 2.47781 I0406 09:24:55.691637 5226 solver.cpp:237] Train net output #0: loss = 2.47781 (* 1 = 2.47781 loss) I0406 09:24:55.691648 5226 sgd_solver.cpp:105] Iteration 15756, lr = 0.01 I0406 09:25:00.901211 5226 solver.cpp:218] Iteration 15768 (2.30347 iter/s, 5.20953s/12 iters), loss = 2.61691 I0406 09:25:00.901263 5226 solver.cpp:237] Train net output #0: loss = 2.61691 (* 1 = 2.61691 loss) I0406 09:25:00.901271 5226 sgd_solver.cpp:105] Iteration 15768, lr = 0.01 I0406 09:25:05.842293 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:25:06.171766 5226 solver.cpp:218] Iteration 15780 (2.27685 iter/s, 5.27045s/12 iters), loss = 2.44525 I0406 09:25:06.171813 5226 solver.cpp:237] Train net output #0: loss = 2.44525 (* 1 = 2.44525 loss) I0406 09:25:06.171819 5226 sgd_solver.cpp:105] Iteration 15780, lr = 0.01 I0406 09:25:11.286185 5226 solver.cpp:218] Iteration 15792 (2.34635 iter/s, 5.11433s/12 iters), loss = 2.42336 I0406 09:25:11.286224 5226 solver.cpp:237] Train net output #0: loss = 2.42336 (* 1 = 2.42336 loss) I0406 09:25:11.286229 5226 sgd_solver.cpp:105] Iteration 15792, lr = 0.01 I0406 09:25:16.488040 5226 solver.cpp:218] Iteration 15804 (2.30691 iter/s, 5.20177s/12 iters), loss = 2.65934 I0406 09:25:16.488080 5226 solver.cpp:237] Train net output #0: loss = 2.65934 (* 1 = 2.65934 loss) I0406 09:25:16.488086 5226 sgd_solver.cpp:105] Iteration 15804, lr = 0.01 I0406 09:25:18.625020 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_15810.caffemodel I0406 09:25:22.178592 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_15810.solverstate I0406 09:25:25.489333 5226 solver.cpp:330] Iteration 15810, Testing net (#0) I0406 09:25:25.489358 5226 net.cpp:676] Ignoring source layer train-data I0406 09:25:28.430613 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:25:30.061458 5226 solver.cpp:397] Test net output #0: accuracy = 0.189951 I0406 09:25:30.061494 5226 solver.cpp:397] Test net output #1: loss = 4.02824 (* 1 = 4.02824 loss) I0406 09:25:32.001039 5226 solver.cpp:218] Iteration 15816 (0.773552 iter/s, 15.5128s/12 iters), loss = 2.58526 I0406 09:25:32.001080 5226 solver.cpp:237] Train net output #0: loss = 2.58526 (* 1 = 2.58526 loss) I0406 09:25:32.001085 5226 sgd_solver.cpp:105] Iteration 15816, lr = 0.01 I0406 09:25:37.270018 5226 solver.cpp:218] Iteration 15828 (2.27752 iter/s, 5.26889s/12 iters), loss = 2.4471 I0406 09:25:37.270069 5226 solver.cpp:237] Train net output #0: loss = 2.4471 (* 1 = 2.4471 loss) I0406 09:25:37.270077 5226 sgd_solver.cpp:105] Iteration 15828, lr = 0.01 I0406 09:25:42.790802 5226 solver.cpp:218] Iteration 15840 (2.17364 iter/s, 5.52069s/12 iters), loss = 2.22162 I0406 09:25:42.790840 5226 solver.cpp:237] Train net output #0: loss = 2.22162 (* 1 = 2.22162 loss) I0406 09:25:42.790846 5226 sgd_solver.cpp:105] Iteration 15840, lr = 0.01 I0406 09:25:48.082165 5226 solver.cpp:218] Iteration 15852 (2.26789 iter/s, 5.29127s/12 iters), loss = 2.47819 I0406 09:25:48.082213 5226 solver.cpp:237] Train net output #0: loss = 2.47819 (* 1 = 2.47819 loss) I0406 09:25:48.082223 5226 sgd_solver.cpp:105] Iteration 15852, lr = 0.01 I0406 09:25:53.189289 5226 solver.cpp:218] Iteration 15864 (2.3497 iter/s, 5.10703s/12 iters), loss = 2.40515 I0406 09:25:53.189330 5226 solver.cpp:237] Train net output #0: loss = 2.40515 (* 1 = 2.40515 loss) I0406 09:25:53.189335 5226 sgd_solver.cpp:105] Iteration 15864, lr = 0.01 I0406 09:25:58.372988 5226 solver.cpp:218] Iteration 15876 (2.31499 iter/s, 5.18361s/12 iters), loss = 2.54443 I0406 09:25:58.373042 5226 solver.cpp:237] Train net output #0: loss = 2.54443 (* 1 = 2.54443 loss) I0406 09:25:58.373051 5226 sgd_solver.cpp:105] Iteration 15876, lr = 0.01 I0406 09:26:00.325109 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:26:03.781400 5226 solver.cpp:218] Iteration 15888 (2.21881 iter/s, 5.40831s/12 iters), loss = 2.13872 I0406 09:26:03.781440 5226 solver.cpp:237] Train net output #0: loss = 2.13872 (* 1 = 2.13872 loss) I0406 09:26:03.781445 5226 sgd_solver.cpp:105] Iteration 15888, lr = 0.01 I0406 09:26:08.996037 5226 solver.cpp:218] Iteration 15900 (2.30125 iter/s, 5.21455s/12 iters), loss = 2.4455 I0406 09:26:08.996074 5226 solver.cpp:237] Train net output #0: loss = 2.4455 (* 1 = 2.4455 loss) I0406 09:26:08.996080 5226 sgd_solver.cpp:105] Iteration 15900, lr = 0.01 I0406 09:26:13.553409 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_15912.caffemodel I0406 09:26:16.570116 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_15912.solverstate I0406 09:26:18.882848 5226 solver.cpp:330] Iteration 15912, Testing net (#0) I0406 09:26:18.882871 5226 net.cpp:676] Ignoring source layer train-data I0406 09:26:21.599714 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:26:23.344414 5226 solver.cpp:397] Test net output #0: accuracy = 0.183211 I0406 09:26:23.344462 5226 solver.cpp:397] Test net output #1: loss = 4.08466 (* 1 = 4.08466 loss) I0406 09:26:23.485219 5226 solver.cpp:218] Iteration 15912 (0.828212 iter/s, 14.489s/12 iters), loss = 2.62416 I0406 09:26:23.485285 5226 solver.cpp:237] Train net output #0: loss = 2.62416 (* 1 = 2.62416 loss) I0406 09:26:23.485291 5226 sgd_solver.cpp:105] Iteration 15912, lr = 0.01 I0406 09:26:27.983083 5226 solver.cpp:218] Iteration 15924 (2.668 iter/s, 4.49776s/12 iters), loss = 2.52435 I0406 09:26:27.983119 5226 solver.cpp:237] Train net output #0: loss = 2.52435 (* 1 = 2.52435 loss) I0406 09:26:27.983124 5226 sgd_solver.cpp:105] Iteration 15924, lr = 0.01 I0406 09:26:33.121044 5226 solver.cpp:218] Iteration 15936 (2.3356 iter/s, 5.13787s/12 iters), loss = 2.92511 I0406 09:26:33.121196 5226 solver.cpp:237] Train net output #0: loss = 2.92511 (* 1 = 2.92511 loss) I0406 09:26:33.121207 5226 sgd_solver.cpp:105] Iteration 15936, lr = 0.01 I0406 09:26:33.121475 5226 blocking_queue.cpp:49] Waiting for data I0406 09:26:38.272338 5226 solver.cpp:218] Iteration 15948 (2.3296 iter/s, 5.1511s/12 iters), loss = 2.54054 I0406 09:26:38.272387 5226 solver.cpp:237] Train net output #0: loss = 2.54054 (* 1 = 2.54054 loss) I0406 09:26:38.272395 5226 sgd_solver.cpp:105] Iteration 15948, lr = 0.01 I0406 09:26:43.181797 5226 solver.cpp:218] Iteration 15960 (2.44431 iter/s, 4.90937s/12 iters), loss = 2.45221 I0406 09:26:43.181831 5226 solver.cpp:237] Train net output #0: loss = 2.45221 (* 1 = 2.45221 loss) I0406 09:26:43.181836 5226 sgd_solver.cpp:105] Iteration 15960, lr = 0.01 I0406 09:26:48.453789 5226 solver.cpp:218] Iteration 15972 (2.27622 iter/s, 5.27191s/12 iters), loss = 2.25819 I0406 09:26:48.453842 5226 solver.cpp:237] Train net output #0: loss = 2.25819 (* 1 = 2.25819 loss) I0406 09:26:48.453850 5226 sgd_solver.cpp:105] Iteration 15972, lr = 0.01 I0406 09:26:52.694708 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:26:53.893669 5226 solver.cpp:218] Iteration 15984 (2.20597 iter/s, 5.43978s/12 iters), loss = 2.6535 I0406 09:26:53.893718 5226 solver.cpp:237] Train net output #0: loss = 2.6535 (* 1 = 2.6535 loss) I0406 09:26:53.893725 5226 sgd_solver.cpp:105] Iteration 15984, lr = 0.01 I0406 09:26:59.242192 5226 solver.cpp:218] Iteration 15996 (2.24365 iter/s, 5.34843s/12 iters), loss = 2.31555 I0406 09:26:59.242229 5226 solver.cpp:237] Train net output #0: loss = 2.31555 (* 1 = 2.31555 loss) I0406 09:26:59.242235 5226 sgd_solver.cpp:105] Iteration 15996, lr = 0.01 I0406 09:27:04.434032 5226 solver.cpp:218] Iteration 16008 (2.31136 iter/s, 5.19175s/12 iters), loss = 2.09383 I0406 09:27:04.434135 5226 solver.cpp:237] Train net output #0: loss = 2.09383 (* 1 = 2.09383 loss) I0406 09:27:04.434141 5226 sgd_solver.cpp:105] Iteration 16008, lr = 0.01 I0406 09:27:06.611964 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_16014.caffemodel I0406 09:27:09.559796 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_16014.solverstate I0406 09:27:11.905735 5226 solver.cpp:330] Iteration 16014, Testing net (#0) I0406 09:27:11.905755 5226 net.cpp:676] Ignoring source layer train-data I0406 09:27:14.549826 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:27:16.242141 5226 solver.cpp:397] Test net output #0: accuracy = 0.175858 I0406 09:27:16.242172 5226 solver.cpp:397] Test net output #1: loss = 4.11342 (* 1 = 4.11342 loss) I0406 09:27:18.265439 5226 solver.cpp:218] Iteration 16020 (0.867603 iter/s, 13.8312s/12 iters), loss = 2.28053 I0406 09:27:18.265480 5226 solver.cpp:237] Train net output #0: loss = 2.28053 (* 1 = 2.28053 loss) I0406 09:27:18.265484 5226 sgd_solver.cpp:105] Iteration 16020, lr = 0.01 I0406 09:27:23.520648 5226 solver.cpp:218] Iteration 16032 (2.28349 iter/s, 5.25512s/12 iters), loss = 2.53497 I0406 09:27:23.520699 5226 solver.cpp:237] Train net output #0: loss = 2.53497 (* 1 = 2.53497 loss) I0406 09:27:23.520706 5226 sgd_solver.cpp:105] Iteration 16032, lr = 0.01 I0406 09:27:28.923040 5226 solver.cpp:218] Iteration 16044 (2.22128 iter/s, 5.4023s/12 iters), loss = 2.16649 I0406 09:27:28.923077 5226 solver.cpp:237] Train net output #0: loss = 2.16649 (* 1 = 2.16649 loss) I0406 09:27:28.923084 5226 sgd_solver.cpp:105] Iteration 16044, lr = 0.01 I0406 09:27:34.145128 5226 solver.cpp:218] Iteration 16056 (2.29797 iter/s, 5.222s/12 iters), loss = 2.12509 I0406 09:27:34.145176 5226 solver.cpp:237] Train net output #0: loss = 2.12509 (* 1 = 2.12509 loss) I0406 09:27:34.145184 5226 sgd_solver.cpp:105] Iteration 16056, lr = 0.01 I0406 09:27:39.536896 5226 solver.cpp:218] Iteration 16068 (2.22566 iter/s, 5.39167s/12 iters), loss = 2.24504 I0406 09:27:39.537031 5226 solver.cpp:237] Train net output #0: loss = 2.24504 (* 1 = 2.24504 loss) I0406 09:27:39.537039 5226 sgd_solver.cpp:105] Iteration 16068, lr = 0.01 I0406 09:27:44.818182 5226 solver.cpp:218] Iteration 16080 (2.27225 iter/s, 5.2811s/12 iters), loss = 2.47746 I0406 09:27:44.818235 5226 solver.cpp:237] Train net output #0: loss = 2.47746 (* 1 = 2.47746 loss) I0406 09:27:44.818244 5226 sgd_solver.cpp:105] Iteration 16080, lr = 0.01 I0406 09:27:45.955237 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:27:50.237565 5226 solver.cpp:218] Iteration 16092 (2.21432 iter/s, 5.41928s/12 iters), loss = 2.80119 I0406 09:27:50.237604 5226 solver.cpp:237] Train net output #0: loss = 2.80119 (* 1 = 2.80119 loss) I0406 09:27:50.237610 5226 sgd_solver.cpp:105] Iteration 16092, lr = 0.01 I0406 09:27:55.725623 5226 solver.cpp:218] Iteration 16104 (2.1866 iter/s, 5.48797s/12 iters), loss = 2.4888 I0406 09:27:55.725662 5226 solver.cpp:237] Train net output #0: loss = 2.4888 (* 1 = 2.4888 loss) I0406 09:27:55.725667 5226 sgd_solver.cpp:105] Iteration 16104, lr = 0.01 I0406 09:28:00.595273 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_16116.caffemodel I0406 09:28:03.664355 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_16116.solverstate I0406 09:28:05.961359 5226 solver.cpp:330] Iteration 16116, Testing net (#0) I0406 09:28:05.961377 5226 net.cpp:676] Ignoring source layer train-data I0406 09:28:08.829941 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:28:10.649439 5226 solver.cpp:397] Test net output #0: accuracy = 0.192402 I0406 09:28:10.649514 5226 solver.cpp:397] Test net output #1: loss = 3.99475 (* 1 = 3.99475 loss) I0406 09:28:10.784075 5226 solver.cpp:218] Iteration 16116 (0.796903 iter/s, 15.0583s/12 iters), loss = 2.7622 I0406 09:28:10.785636 5226 solver.cpp:237] Train net output #0: loss = 2.7622 (* 1 = 2.7622 loss) I0406 09:28:10.785651 5226 sgd_solver.cpp:105] Iteration 16116, lr = 0.01 I0406 09:28:15.193295 5226 solver.cpp:218] Iteration 16128 (2.72256 iter/s, 4.40762s/12 iters), loss = 2.56036 I0406 09:28:15.193333 5226 solver.cpp:237] Train net output #0: loss = 2.56036 (* 1 = 2.56036 loss) I0406 09:28:15.193339 5226 sgd_solver.cpp:105] Iteration 16128, lr = 0.01 I0406 09:28:20.784226 5226 solver.cpp:218] Iteration 16140 (2.14637 iter/s, 5.59084s/12 iters), loss = 2.95733 I0406 09:28:20.784261 5226 solver.cpp:237] Train net output #0: loss = 2.95733 (* 1 = 2.95733 loss) I0406 09:28:20.784266 5226 sgd_solver.cpp:105] Iteration 16140, lr = 0.01 I0406 09:28:26.447095 5226 solver.cpp:218] Iteration 16152 (2.1191 iter/s, 5.66278s/12 iters), loss = 2.48393 I0406 09:28:26.447144 5226 solver.cpp:237] Train net output #0: loss = 2.48393 (* 1 = 2.48393 loss) I0406 09:28:26.447151 5226 sgd_solver.cpp:105] Iteration 16152, lr = 0.01 I0406 09:28:32.172912 5226 solver.cpp:218] Iteration 16164 (2.09581 iter/s, 5.72571s/12 iters), loss = 2.54062 I0406 09:28:32.172950 5226 solver.cpp:237] Train net output #0: loss = 2.54062 (* 1 = 2.54062 loss) I0406 09:28:32.172956 5226 sgd_solver.cpp:105] Iteration 16164, lr = 0.01 I0406 09:28:37.882836 5226 solver.cpp:218] Iteration 16176 (2.10164 iter/s, 5.70983s/12 iters), loss = 2.63387 I0406 09:28:37.882884 5226 solver.cpp:237] Train net output #0: loss = 2.63387 (* 1 = 2.63387 loss) I0406 09:28:37.882891 5226 sgd_solver.cpp:105] Iteration 16176, lr = 0.01 I0406 09:28:41.187204 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:28:43.320613 5226 solver.cpp:218] Iteration 16188 (2.20682 iter/s, 5.43768s/12 iters), loss = 3.08065 I0406 09:28:43.320663 5226 solver.cpp:237] Train net output #0: loss = 3.08065 (* 1 = 3.08065 loss) I0406 09:28:43.320672 5226 sgd_solver.cpp:105] Iteration 16188, lr = 0.01 I0406 09:28:48.868044 5226 solver.cpp:218] Iteration 16200 (2.1632 iter/s, 5.54733s/12 iters), loss = 2.41528 I0406 09:28:48.868093 5226 solver.cpp:237] Train net output #0: loss = 2.41528 (* 1 = 2.41528 loss) I0406 09:28:48.868100 5226 sgd_solver.cpp:105] Iteration 16200, lr = 0.01 I0406 09:28:54.500289 5226 solver.cpp:218] Iteration 16212 (2.13063 iter/s, 5.63214s/12 iters), loss = 2.35922 I0406 09:28:54.506508 5226 solver.cpp:237] Train net output #0: loss = 2.35922 (* 1 = 2.35922 loss) I0406 09:28:54.506528 5226 sgd_solver.cpp:105] Iteration 16212, lr = 0.01 I0406 09:28:56.807871 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_16218.caffemodel I0406 09:29:00.080796 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_16218.solverstate I0406 09:29:02.375846 5226 solver.cpp:330] Iteration 16218, Testing net (#0) I0406 09:29:02.375869 5226 net.cpp:676] Ignoring source layer train-data I0406 09:29:05.283919 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:29:07.255527 5226 solver.cpp:397] Test net output #0: accuracy = 0.165441 I0406 09:29:07.255573 5226 solver.cpp:397] Test net output #1: loss = 4.11181 (* 1 = 4.11181 loss) I0406 09:29:09.287992 5226 solver.cpp:218] Iteration 16224 (0.811831 iter/s, 14.7814s/12 iters), loss = 2.15621 I0406 09:29:09.288041 5226 solver.cpp:237] Train net output #0: loss = 2.15621 (* 1 = 2.15621 loss) I0406 09:29:09.288049 5226 sgd_solver.cpp:105] Iteration 16224, lr = 0.01 I0406 09:29:14.942277 5226 solver.cpp:218] Iteration 16236 (2.12232 iter/s, 5.65419s/12 iters), loss = 2.31498 I0406 09:29:14.942384 5226 solver.cpp:237] Train net output #0: loss = 2.31498 (* 1 = 2.31498 loss) I0406 09:29:14.942394 5226 sgd_solver.cpp:105] Iteration 16236, lr = 0.01 I0406 09:29:20.515019 5226 solver.cpp:218] Iteration 16248 (2.1534 iter/s, 5.57259s/12 iters), loss = 2.68945 I0406 09:29:20.515064 5226 solver.cpp:237] Train net output #0: loss = 2.68945 (* 1 = 2.68945 loss) I0406 09:29:20.515072 5226 sgd_solver.cpp:105] Iteration 16248, lr = 0.01 I0406 09:29:26.049319 5226 solver.cpp:218] Iteration 16260 (2.16841 iter/s, 5.534s/12 iters), loss = 2.55721 I0406 09:29:26.049367 5226 solver.cpp:237] Train net output #0: loss = 2.55721 (* 1 = 2.55721 loss) I0406 09:29:26.049374 5226 sgd_solver.cpp:105] Iteration 16260, lr = 0.01 I0406 09:29:31.664438 5226 solver.cpp:218] Iteration 16272 (2.13712 iter/s, 5.61502s/12 iters), loss = 2.33488 I0406 09:29:31.664479 5226 solver.cpp:237] Train net output #0: loss = 2.33488 (* 1 = 2.33488 loss) I0406 09:29:31.664484 5226 sgd_solver.cpp:105] Iteration 16272, lr = 0.01 I0406 09:29:37.314996 5226 solver.cpp:218] Iteration 16284 (2.12372 iter/s, 5.65046s/12 iters), loss = 2.37757 I0406 09:29:37.315050 5226 solver.cpp:237] Train net output #0: loss = 2.37757 (* 1 = 2.37757 loss) I0406 09:29:37.315059 5226 sgd_solver.cpp:105] Iteration 16284, lr = 0.01 I0406 09:29:37.840929 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:29:43.282966 5226 solver.cpp:218] Iteration 16296 (2.01077 iter/s, 5.96787s/12 iters), loss = 2.91481 I0406 09:29:43.283004 5226 solver.cpp:237] Train net output #0: loss = 2.91481 (* 1 = 2.91481 loss) I0406 09:29:43.283010 5226 sgd_solver.cpp:105] Iteration 16296, lr = 0.01 I0406 09:29:48.870033 5226 solver.cpp:218] Iteration 16308 (2.14787 iter/s, 5.58693s/12 iters), loss = 2.85057 I0406 09:29:48.870216 5226 solver.cpp:237] Train net output #0: loss = 2.85057 (* 1 = 2.85057 loss) I0406 09:29:48.870226 5226 sgd_solver.cpp:105] Iteration 16308, lr = 0.01 I0406 09:29:54.265199 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_16320.caffemodel I0406 09:29:57.426184 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_16320.solverstate I0406 09:29:59.833500 5226 solver.cpp:330] Iteration 16320, Testing net (#0) I0406 09:29:59.833529 5226 net.cpp:676] Ignoring source layer train-data I0406 09:30:02.847280 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:30:04.648279 5226 solver.cpp:397] Test net output #0: accuracy = 0.153186 I0406 09:30:04.648316 5226 solver.cpp:397] Test net output #1: loss = 4.10669 (* 1 = 4.10669 loss) I0406 09:30:04.788470 5226 solver.cpp:218] Iteration 16320 (0.753857 iter/s, 15.9181s/12 iters), loss = 2.56875 I0406 09:30:04.788522 5226 solver.cpp:237] Train net output #0: loss = 2.56875 (* 1 = 2.56875 loss) I0406 09:30:04.788529 5226 sgd_solver.cpp:105] Iteration 16320, lr = 0.01 I0406 09:30:09.277542 5226 solver.cpp:218] Iteration 16332 (2.67322 iter/s, 4.48898s/12 iters), loss = 2.33193 I0406 09:30:09.277601 5226 solver.cpp:237] Train net output #0: loss = 2.33193 (* 1 = 2.33193 loss) I0406 09:30:09.277611 5226 sgd_solver.cpp:105] Iteration 16332, lr = 0.01 I0406 09:30:14.888790 5226 solver.cpp:218] Iteration 16344 (2.1386 iter/s, 5.61114s/12 iters), loss = 2.56185 I0406 09:30:14.888829 5226 solver.cpp:237] Train net output #0: loss = 2.56185 (* 1 = 2.56185 loss) I0406 09:30:14.888834 5226 sgd_solver.cpp:105] Iteration 16344, lr = 0.01 I0406 09:30:20.303822 5226 solver.cpp:218] Iteration 16356 (2.21609 iter/s, 5.41494s/12 iters), loss = 2.94448 I0406 09:30:20.303939 5226 solver.cpp:237] Train net output #0: loss = 2.94448 (* 1 = 2.94448 loss) I0406 09:30:20.303951 5226 sgd_solver.cpp:105] Iteration 16356, lr = 0.01 I0406 09:30:26.114989 5226 solver.cpp:218] Iteration 16368 (2.06505 iter/s, 5.811s/12 iters), loss = 2.92951 I0406 09:30:26.115039 5226 solver.cpp:237] Train net output #0: loss = 2.92951 (* 1 = 2.92951 loss) I0406 09:30:26.115047 5226 sgd_solver.cpp:105] Iteration 16368, lr = 0.01 I0406 09:30:31.776924 5226 solver.cpp:218] Iteration 16380 (2.11945 iter/s, 5.66184s/12 iters), loss = 3.13383 I0406 09:30:31.776962 5226 solver.cpp:237] Train net output #0: loss = 3.13383 (* 1 = 3.13383 loss) I0406 09:30:31.776968 5226 sgd_solver.cpp:105] Iteration 16380, lr = 0.01 I0406 09:30:34.733310 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:30:37.509531 5226 solver.cpp:218] Iteration 16392 (2.09332 iter/s, 5.73252s/12 iters), loss = 2.46985 I0406 09:30:37.509572 5226 solver.cpp:237] Train net output #0: loss = 2.46985 (* 1 = 2.46985 loss) I0406 09:30:37.509577 5226 sgd_solver.cpp:105] Iteration 16392, lr = 0.01 I0406 09:30:43.187196 5226 solver.cpp:218] Iteration 16404 (2.11358 iter/s, 5.67757s/12 iters), loss = 2.23007 I0406 09:30:43.187242 5226 solver.cpp:237] Train net output #0: loss = 2.23007 (* 1 = 2.23007 loss) I0406 09:30:43.187252 5226 sgd_solver.cpp:105] Iteration 16404, lr = 0.01 I0406 09:30:48.800408 5226 solver.cpp:218] Iteration 16416 (2.13785 iter/s, 5.61312s/12 iters), loss = 2.45317 I0406 09:30:48.800459 5226 solver.cpp:237] Train net output #0: loss = 2.45317 (* 1 = 2.45317 loss) I0406 09:30:48.800469 5226 sgd_solver.cpp:105] Iteration 16416, lr = 0.01 I0406 09:30:51.342778 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_16422.caffemodel I0406 09:30:54.561311 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_16422.solverstate I0406 09:30:56.969089 5226 solver.cpp:330] Iteration 16422, Testing net (#0) I0406 09:30:56.969106 5226 net.cpp:676] Ignoring source layer train-data I0406 09:30:59.958966 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:31:01.891489 5226 solver.cpp:397] Test net output #0: accuracy = 0.172181 I0406 09:31:01.891527 5226 solver.cpp:397] Test net output #1: loss = 4.19572 (* 1 = 4.19572 loss) I0406 09:31:03.790828 5226 solver.cpp:218] Iteration 16428 (0.80052 iter/s, 14.9903s/12 iters), loss = 3.18151 I0406 09:31:03.790881 5226 solver.cpp:237] Train net output #0: loss = 3.18151 (* 1 = 3.18151 loss) I0406 09:31:03.790891 5226 sgd_solver.cpp:105] Iteration 16428, lr = 0.01 I0406 09:31:09.088055 5226 solver.cpp:218] Iteration 16440 (2.26538 iter/s, 5.29712s/12 iters), loss = 2.36082 I0406 09:31:09.088104 5226 solver.cpp:237] Train net output #0: loss = 2.36082 (* 1 = 2.36082 loss) I0406 09:31:09.088111 5226 sgd_solver.cpp:105] Iteration 16440, lr = 0.01 I0406 09:31:14.630079 5226 solver.cpp:218] Iteration 16452 (2.16538 iter/s, 5.54175s/12 iters), loss = 2.34356 I0406 09:31:14.630134 5226 solver.cpp:237] Train net output #0: loss = 2.34356 (* 1 = 2.34356 loss) I0406 09:31:14.630146 5226 sgd_solver.cpp:105] Iteration 16452, lr = 0.01 I0406 09:31:20.147900 5226 solver.cpp:218] Iteration 16464 (2.17481 iter/s, 5.51771s/12 iters), loss = 1.83815 I0406 09:31:20.147950 5226 solver.cpp:237] Train net output #0: loss = 1.83815 (* 1 = 1.83815 loss) I0406 09:31:20.147958 5226 sgd_solver.cpp:105] Iteration 16464, lr = 0.01 I0406 09:31:25.491257 5226 solver.cpp:218] Iteration 16476 (2.24582 iter/s, 5.34325s/12 iters), loss = 2.64603 I0406 09:31:25.491374 5226 solver.cpp:237] Train net output #0: loss = 2.64603 (* 1 = 2.64603 loss) I0406 09:31:25.491385 5226 sgd_solver.cpp:105] Iteration 16476, lr = 0.01 I0406 09:31:30.937216 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:31:31.239321 5226 solver.cpp:218] Iteration 16488 (2.08772 iter/s, 5.7479s/12 iters), loss = 2.93763 I0406 09:31:31.239357 5226 solver.cpp:237] Train net output #0: loss = 2.93763 (* 1 = 2.93763 loss) I0406 09:31:31.239362 5226 sgd_solver.cpp:105] Iteration 16488, lr = 0.01 I0406 09:31:36.937832 5226 solver.cpp:218] Iteration 16500 (2.10585 iter/s, 5.69842s/12 iters), loss = 2.73476 I0406 09:31:36.937886 5226 solver.cpp:237] Train net output #0: loss = 2.73476 (* 1 = 2.73476 loss) I0406 09:31:36.937894 5226 sgd_solver.cpp:105] Iteration 16500, lr = 0.01 I0406 09:31:42.717741 5226 solver.cpp:218] Iteration 16512 (2.0762 iter/s, 5.7798s/12 iters), loss = 2.10951 I0406 09:31:42.717789 5226 solver.cpp:237] Train net output #0: loss = 2.10951 (* 1 = 2.10951 loss) I0406 09:31:42.717798 5226 sgd_solver.cpp:105] Iteration 16512, lr = 0.01 I0406 09:31:47.854550 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_16524.caffemodel I0406 09:31:50.970706 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_16524.solverstate I0406 09:31:53.339844 5226 solver.cpp:330] Iteration 16524, Testing net (#0) I0406 09:31:53.339865 5226 net.cpp:676] Ignoring source layer train-data I0406 09:31:56.025452 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:31:58.217926 5226 solver.cpp:397] Test net output #0: accuracy = 0.169118 I0406 09:31:58.217952 5226 solver.cpp:397] Test net output #1: loss = 4.11092 (* 1 = 4.11092 loss) I0406 09:31:58.361522 5226 solver.cpp:218] Iteration 16524 (0.767086 iter/s, 15.6436s/12 iters), loss = 2.64053 I0406 09:31:58.361572 5226 solver.cpp:237] Train net output #0: loss = 2.64053 (* 1 = 2.64053 loss) I0406 09:31:58.361579 5226 sgd_solver.cpp:105] Iteration 16524, lr = 0.01 I0406 09:32:02.910775 5226 solver.cpp:218] Iteration 16536 (2.63785 iter/s, 4.54916s/12 iters), loss = 2.09279 I0406 09:32:02.910821 5226 solver.cpp:237] Train net output #0: loss = 2.09279 (* 1 = 2.09279 loss) I0406 09:32:02.910830 5226 sgd_solver.cpp:105] Iteration 16536, lr = 0.01 I0406 09:32:08.656677 5226 solver.cpp:218] Iteration 16548 (2.08851 iter/s, 5.74572s/12 iters), loss = 2.44516 I0406 09:32:08.656728 5226 solver.cpp:237] Train net output #0: loss = 2.44516 (* 1 = 2.44516 loss) I0406 09:32:08.656734 5226 sgd_solver.cpp:105] Iteration 16548, lr = 0.01 I0406 09:32:14.274456 5226 solver.cpp:218] Iteration 16560 (2.13611 iter/s, 5.61768s/12 iters), loss = 2.54443 I0406 09:32:14.274497 5226 solver.cpp:237] Train net output #0: loss = 2.54443 (* 1 = 2.54443 loss) I0406 09:32:14.274502 5226 sgd_solver.cpp:105] Iteration 16560, lr = 0.01 I0406 09:32:19.852234 5226 solver.cpp:218] Iteration 16572 (2.15143 iter/s, 5.57769s/12 iters), loss = 2.59858 I0406 09:32:19.852274 5226 solver.cpp:237] Train net output #0: loss = 2.59858 (* 1 = 2.59858 loss) I0406 09:32:19.852279 5226 sgd_solver.cpp:105] Iteration 16572, lr = 0.01 I0406 09:32:25.674960 5226 solver.cpp:218] Iteration 16584 (2.06093 iter/s, 5.82263s/12 iters), loss = 2.65429 I0406 09:32:25.675009 5226 solver.cpp:237] Train net output #0: loss = 2.65429 (* 1 = 2.65429 loss) I0406 09:32:25.675017 5226 sgd_solver.cpp:105] Iteration 16584, lr = 0.01 I0406 09:32:27.852551 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:32:31.481380 5226 solver.cpp:218] Iteration 16596 (2.06671 iter/s, 5.80632s/12 iters), loss = 2.42607 I0406 09:32:31.481438 5226 solver.cpp:237] Train net output #0: loss = 2.42607 (* 1 = 2.42607 loss) I0406 09:32:31.481448 5226 sgd_solver.cpp:105] Iteration 16596, lr = 0.01 I0406 09:32:37.138252 5226 solver.cpp:218] Iteration 16608 (2.12135 iter/s, 5.65677s/12 iters), loss = 2.49398 I0406 09:32:37.138291 5226 solver.cpp:237] Train net output #0: loss = 2.49398 (* 1 = 2.49398 loss) I0406 09:32:37.138296 5226 sgd_solver.cpp:105] Iteration 16608, lr = 0.01 I0406 09:32:42.953439 5226 solver.cpp:218] Iteration 16620 (2.0636 iter/s, 5.81509s/12 iters), loss = 2.7301 I0406 09:32:42.953492 5226 solver.cpp:237] Train net output #0: loss = 2.7301 (* 1 = 2.7301 loss) I0406 09:32:42.953500 5226 sgd_solver.cpp:105] Iteration 16620, lr = 0.01 I0406 09:32:45.127641 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_16626.caffemodel I0406 09:32:49.070082 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_16626.solverstate I0406 09:32:51.479532 5226 solver.cpp:330] Iteration 16626, Testing net (#0) I0406 09:32:51.479550 5226 net.cpp:676] Ignoring source layer train-data I0406 09:32:54.278700 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:32:55.733291 5226 blocking_queue.cpp:49] Waiting for data I0406 09:32:56.470927 5226 solver.cpp:397] Test net output #0: accuracy = 0.167279 I0406 09:32:56.470960 5226 solver.cpp:397] Test net output #1: loss = 4.1523 (* 1 = 4.1523 loss) I0406 09:32:58.396672 5226 solver.cpp:218] Iteration 16632 (0.777048 iter/s, 15.4431s/12 iters), loss = 2.21863 I0406 09:32:58.398195 5226 solver.cpp:237] Train net output #0: loss = 2.21863 (* 1 = 2.21863 loss) I0406 09:32:58.398211 5226 sgd_solver.cpp:105] Iteration 16632, lr = 0.01 I0406 09:33:04.069341 5226 solver.cpp:218] Iteration 16644 (2.11599 iter/s, 5.67111s/12 iters), loss = 2.2941 I0406 09:33:04.069380 5226 solver.cpp:237] Train net output #0: loss = 2.2941 (* 1 = 2.2941 loss) I0406 09:33:04.069386 5226 sgd_solver.cpp:105] Iteration 16644, lr = 0.01 I0406 09:33:09.934734 5226 solver.cpp:218] Iteration 16656 (2.04593 iter/s, 5.8653s/12 iters), loss = 1.89636 I0406 09:33:09.934779 5226 solver.cpp:237] Train net output #0: loss = 1.89636 (* 1 = 1.89636 loss) I0406 09:33:09.934787 5226 sgd_solver.cpp:105] Iteration 16656, lr = 0.01 I0406 09:33:15.585810 5226 solver.cpp:218] Iteration 16668 (2.12353 iter/s, 5.65098s/12 iters), loss = 2.40289 I0406 09:33:15.585858 5226 solver.cpp:237] Train net output #0: loss = 2.40289 (* 1 = 2.40289 loss) I0406 09:33:15.585866 5226 sgd_solver.cpp:105] Iteration 16668, lr = 0.01 I0406 09:33:20.888341 5226 solver.cpp:218] Iteration 16680 (2.26311 iter/s, 5.30244s/12 iters), loss = 2.3542 I0406 09:33:20.888377 5226 solver.cpp:237] Train net output #0: loss = 2.3542 (* 1 = 2.3542 loss) I0406 09:33:20.888382 5226 sgd_solver.cpp:105] Iteration 16680, lr = 0.01 I0406 09:33:25.343715 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:33:26.509707 5226 solver.cpp:218] Iteration 16692 (2.13475 iter/s, 5.62128s/12 iters), loss = 2.99205 I0406 09:33:26.515930 5226 solver.cpp:237] Train net output #0: loss = 2.99205 (* 1 = 2.99205 loss) I0406 09:33:26.515947 5226 sgd_solver.cpp:105] Iteration 16692, lr = 0.01 I0406 09:33:32.121196 5226 solver.cpp:218] Iteration 16704 (2.14086 iter/s, 5.60523s/12 iters), loss = 2.66269 I0406 09:33:32.121381 5226 solver.cpp:237] Train net output #0: loss = 2.66269 (* 1 = 2.66269 loss) I0406 09:33:32.121393 5226 sgd_solver.cpp:105] Iteration 16704, lr = 0.01 I0406 09:33:37.883632 5226 solver.cpp:218] Iteration 16716 (2.08255 iter/s, 5.76217s/12 iters), loss = 2.70606 I0406 09:33:37.883687 5226 solver.cpp:237] Train net output #0: loss = 2.70606 (* 1 = 2.70606 loss) I0406 09:33:37.883695 5226 sgd_solver.cpp:105] Iteration 16716, lr = 0.01 I0406 09:33:42.890591 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_16728.caffemodel I0406 09:33:46.102897 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_16728.solverstate I0406 09:33:48.429646 5226 solver.cpp:330] Iteration 16728, Testing net (#0) I0406 09:33:48.429666 5226 net.cpp:676] Ignoring source layer train-data I0406 09:33:51.269460 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:33:53.218272 5226 solver.cpp:397] Test net output #0: accuracy = 0.202819 I0406 09:33:53.218307 5226 solver.cpp:397] Test net output #1: loss = 4.02315 (* 1 = 4.02315 loss) I0406 09:33:53.348867 5226 solver.cpp:218] Iteration 16728 (0.775942 iter/s, 15.4651s/12 iters), loss = 2.55238 I0406 09:33:53.348917 5226 solver.cpp:237] Train net output #0: loss = 2.55238 (* 1 = 2.55238 loss) I0406 09:33:53.348925 5226 sgd_solver.cpp:105] Iteration 16728, lr = 0.01 I0406 09:33:58.225090 5226 solver.cpp:218] Iteration 16740 (2.46097 iter/s, 4.87613s/12 iters), loss = 2.19731 I0406 09:33:58.225147 5226 solver.cpp:237] Train net output #0: loss = 2.19731 (* 1 = 2.19731 loss) I0406 09:33:58.225157 5226 sgd_solver.cpp:105] Iteration 16740, lr = 0.01 I0406 09:34:04.144661 5226 solver.cpp:218] Iteration 16752 (2.02721 iter/s, 5.91947s/12 iters), loss = 2.38449 I0406 09:34:04.144774 5226 solver.cpp:237] Train net output #0: loss = 2.38449 (* 1 = 2.38449 loss) I0406 09:34:04.144784 5226 sgd_solver.cpp:105] Iteration 16752, lr = 0.01 I0406 09:34:09.579020 5226 solver.cpp:218] Iteration 16764 (2.20824 iter/s, 5.4342s/12 iters), loss = 2.6434 I0406 09:34:09.579064 5226 solver.cpp:237] Train net output #0: loss = 2.6434 (* 1 = 2.6434 loss) I0406 09:34:09.579071 5226 sgd_solver.cpp:105] Iteration 16764, lr = 0.01 I0406 09:34:15.014813 5226 solver.cpp:218] Iteration 16776 (2.20763 iter/s, 5.4357s/12 iters), loss = 2.96203 I0406 09:34:15.014851 5226 solver.cpp:237] Train net output #0: loss = 2.96203 (* 1 = 2.96203 loss) I0406 09:34:15.014856 5226 sgd_solver.cpp:105] Iteration 16776, lr = 0.01 I0406 09:34:20.651443 5226 solver.cpp:218] Iteration 16788 (2.12897 iter/s, 5.63654s/12 iters), loss = 2.26781 I0406 09:34:20.651497 5226 solver.cpp:237] Train net output #0: loss = 2.26781 (* 1 = 2.26781 loss) I0406 09:34:20.651506 5226 sgd_solver.cpp:105] Iteration 16788, lr = 0.01 I0406 09:34:21.892722 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:34:26.436740 5226 solver.cpp:218] Iteration 16800 (2.07426 iter/s, 5.78519s/12 iters), loss = 2.64273 I0406 09:34:26.442966 5226 solver.cpp:237] Train net output #0: loss = 2.64273 (* 1 = 2.64273 loss) I0406 09:34:26.442988 5226 sgd_solver.cpp:105] Iteration 16800, lr = 0.01 I0406 09:34:32.254184 5226 solver.cpp:218] Iteration 16812 (2.06498 iter/s, 5.81118s/12 iters), loss = 2.42553 I0406 09:34:32.254233 5226 solver.cpp:237] Train net output #0: loss = 2.42553 (* 1 = 2.42553 loss) I0406 09:34:32.254243 5226 sgd_solver.cpp:105] Iteration 16812, lr = 0.01 I0406 09:34:37.731284 5226 solver.cpp:218] Iteration 16824 (2.191 iter/s, 5.47696s/12 iters), loss = 2.73933 I0406 09:34:37.731446 5226 solver.cpp:237] Train net output #0: loss = 2.73933 (* 1 = 2.73933 loss) I0406 09:34:37.731454 5226 sgd_solver.cpp:105] Iteration 16824, lr = 0.01 I0406 09:34:39.879523 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_16830.caffemodel I0406 09:34:42.924937 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_16830.solverstate I0406 09:34:45.260212 5226 solver.cpp:330] Iteration 16830, Testing net (#0) I0406 09:34:45.260237 5226 net.cpp:676] Ignoring source layer train-data I0406 09:34:47.935811 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:34:50.216485 5226 solver.cpp:397] Test net output #0: accuracy = 0.161765 I0406 09:34:50.216511 5226 solver.cpp:397] Test net output #1: loss = 4.21224 (* 1 = 4.21224 loss) I0406 09:34:52.167304 5226 solver.cpp:218] Iteration 16836 (0.831269 iter/s, 14.4358s/12 iters), loss = 2.99011 I0406 09:34:52.167353 5226 solver.cpp:237] Train net output #0: loss = 2.99011 (* 1 = 2.99011 loss) I0406 09:34:52.167358 5226 sgd_solver.cpp:105] Iteration 16836, lr = 0.01 I0406 09:34:57.731110 5226 solver.cpp:218] Iteration 16848 (2.15684 iter/s, 5.5637s/12 iters), loss = 2.67494 I0406 09:34:57.731156 5226 solver.cpp:237] Train net output #0: loss = 2.67494 (* 1 = 2.67494 loss) I0406 09:34:57.731164 5226 sgd_solver.cpp:105] Iteration 16848, lr = 0.01 I0406 09:35:03.158469 5226 solver.cpp:218] Iteration 16860 (2.21106 iter/s, 5.42726s/12 iters), loss = 2.42591 I0406 09:35:03.158517 5226 solver.cpp:237] Train net output #0: loss = 2.42591 (* 1 = 2.42591 loss) I0406 09:35:03.158526 5226 sgd_solver.cpp:105] Iteration 16860, lr = 0.01 I0406 09:35:08.689682 5226 solver.cpp:218] Iteration 16872 (2.16954 iter/s, 5.53113s/12 iters), loss = 2.9865 I0406 09:35:08.689777 5226 solver.cpp:237] Train net output #0: loss = 2.9865 (* 1 = 2.9865 loss) I0406 09:35:08.689785 5226 sgd_solver.cpp:105] Iteration 16872, lr = 0.01 I0406 09:35:14.223196 5226 solver.cpp:218] Iteration 16884 (2.16866 iter/s, 5.53337s/12 iters), loss = 2.26302 I0406 09:35:14.223246 5226 solver.cpp:237] Train net output #0: loss = 2.26302 (* 1 = 2.26302 loss) I0406 09:35:14.223255 5226 sgd_solver.cpp:105] Iteration 16884, lr = 0.01 I0406 09:35:18.036418 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:35:19.994814 5226 solver.cpp:218] Iteration 16896 (2.07918 iter/s, 5.77151s/12 iters), loss = 2.89087 I0406 09:35:19.994863 5226 solver.cpp:237] Train net output #0: loss = 2.89087 (* 1 = 2.89087 loss) I0406 09:35:19.994872 5226 sgd_solver.cpp:105] Iteration 16896, lr = 0.01 I0406 09:35:25.494658 5226 solver.cpp:218] Iteration 16908 (2.18192 iter/s, 5.49973s/12 iters), loss = 2.29341 I0406 09:35:25.494705 5226 solver.cpp:237] Train net output #0: loss = 2.29341 (* 1 = 2.29341 loss) I0406 09:35:25.494715 5226 sgd_solver.cpp:105] Iteration 16908, lr = 0.01 I0406 09:35:31.193148 5226 solver.cpp:218] Iteration 16920 (2.10586 iter/s, 5.69839s/12 iters), loss = 2.6235 I0406 09:35:31.193195 5226 solver.cpp:237] Train net output #0: loss = 2.6235 (* 1 = 2.6235 loss) I0406 09:35:31.193203 5226 sgd_solver.cpp:105] Iteration 16920, lr = 0.01 I0406 09:35:36.222177 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_16932.caffemodel I0406 09:35:39.301908 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_16932.solverstate I0406 09:35:41.610008 5226 solver.cpp:330] Iteration 16932, Testing net (#0) I0406 09:35:41.610025 5226 net.cpp:676] Ignoring source layer train-data I0406 09:35:44.274611 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:35:46.581073 5226 solver.cpp:397] Test net output #0: accuracy = 0.179534 I0406 09:35:46.581110 5226 solver.cpp:397] Test net output #1: loss = 4.11929 (* 1 = 4.11929 loss) I0406 09:35:46.717133 5226 solver.cpp:218] Iteration 16932 (0.773006 iter/s, 15.5238s/12 iters), loss = 2.70538 I0406 09:35:46.717183 5226 solver.cpp:237] Train net output #0: loss = 2.70538 (* 1 = 2.70538 loss) I0406 09:35:46.717191 5226 sgd_solver.cpp:105] Iteration 16932, lr = 0.01 I0406 09:35:51.336233 5226 solver.cpp:218] Iteration 16944 (2.59796 iter/s, 4.619s/12 iters), loss = 2.58273 I0406 09:35:51.336282 5226 solver.cpp:237] Train net output #0: loss = 2.58273 (* 1 = 2.58273 loss) I0406 09:35:51.336293 5226 sgd_solver.cpp:105] Iteration 16944, lr = 0.01 I0406 09:35:56.939513 5226 solver.cpp:218] Iteration 16956 (2.14164 iter/s, 5.60318s/12 iters), loss = 2.86786 I0406 09:35:56.939565 5226 solver.cpp:237] Train net output #0: loss = 2.86786 (* 1 = 2.86786 loss) I0406 09:35:56.939574 5226 sgd_solver.cpp:105] Iteration 16956, lr = 0.01 I0406 09:36:02.742823 5226 solver.cpp:218] Iteration 16968 (2.06782 iter/s, 5.8032s/12 iters), loss = 2.36557 I0406 09:36:02.742877 5226 solver.cpp:237] Train net output #0: loss = 2.36557 (* 1 = 2.36557 loss) I0406 09:36:02.742884 5226 sgd_solver.cpp:105] Iteration 16968, lr = 0.01 I0406 09:36:08.606348 5226 solver.cpp:218] Iteration 16980 (2.04659 iter/s, 5.86342s/12 iters), loss = 2.72038 I0406 09:36:08.606396 5226 solver.cpp:237] Train net output #0: loss = 2.72038 (* 1 = 2.72038 loss) I0406 09:36:08.606403 5226 sgd_solver.cpp:105] Iteration 16980, lr = 0.01 I0406 09:36:14.171015 5226 solver.cpp:218] Iteration 16992 (2.15656 iter/s, 5.56441s/12 iters), loss = 3.18683 I0406 09:36:14.171133 5226 solver.cpp:237] Train net output #0: loss = 3.18683 (* 1 = 3.18683 loss) I0406 09:36:14.171142 5226 sgd_solver.cpp:105] Iteration 16992, lr = 0.01 I0406 09:36:14.638798 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:36:19.679415 5226 solver.cpp:218] Iteration 17004 (2.17856 iter/s, 5.50823s/12 iters), loss = 2.47616 I0406 09:36:19.679463 5226 solver.cpp:237] Train net output #0: loss = 2.47616 (* 1 = 2.47616 loss) I0406 09:36:19.679471 5226 sgd_solver.cpp:105] Iteration 17004, lr = 0.01 I0406 09:36:25.376852 5226 solver.cpp:218] Iteration 17016 (2.10625 iter/s, 5.69733s/12 iters), loss = 2.57479 I0406 09:36:25.376921 5226 solver.cpp:237] Train net output #0: loss = 2.57479 (* 1 = 2.57479 loss) I0406 09:36:25.376929 5226 sgd_solver.cpp:105] Iteration 17016, lr = 0.01 I0406 09:36:31.233106 5226 solver.cpp:218] Iteration 17028 (2.04913 iter/s, 5.85614s/12 iters), loss = 2.88804 I0406 09:36:31.233141 5226 solver.cpp:237] Train net output #0: loss = 2.88804 (* 1 = 2.88804 loss) I0406 09:36:31.233147 5226 sgd_solver.cpp:105] Iteration 17028, lr = 0.01 I0406 09:36:33.404342 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_17034.caffemodel I0406 09:36:36.647012 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_17034.solverstate I0406 09:36:39.246140 5226 solver.cpp:330] Iteration 17034, Testing net (#0) I0406 09:36:39.246160 5226 net.cpp:676] Ignoring source layer train-data I0406 09:36:41.766379 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:36:44.265537 5226 solver.cpp:397] Test net output #0: accuracy = 0.16299 I0406 09:36:44.265625 5226 solver.cpp:397] Test net output #1: loss = 4.21995 (* 1 = 4.21995 loss) I0406 09:36:46.347651 5226 solver.cpp:218] Iteration 17040 (0.793945 iter/s, 15.1144s/12 iters), loss = 2.37067 I0406 09:36:46.352939 5226 solver.cpp:237] Train net output #0: loss = 2.37067 (* 1 = 2.37067 loss) I0406 09:36:46.352959 5226 sgd_solver.cpp:105] Iteration 17040, lr = 0.01 I0406 09:36:52.245440 5226 solver.cpp:218] Iteration 17052 (2.0365 iter/s, 5.89247s/12 iters), loss = 2.20378 I0406 09:36:52.245486 5226 solver.cpp:237] Train net output #0: loss = 2.20378 (* 1 = 2.20378 loss) I0406 09:36:52.245494 5226 sgd_solver.cpp:105] Iteration 17052, lr = 0.01 I0406 09:36:57.807592 5226 solver.cpp:218] Iteration 17064 (2.15748 iter/s, 5.56206s/12 iters), loss = 2.47126 I0406 09:36:57.807641 5226 solver.cpp:237] Train net output #0: loss = 2.47126 (* 1 = 2.47126 loss) I0406 09:36:57.807648 5226 sgd_solver.cpp:105] Iteration 17064, lr = 0.01 I0406 09:37:03.242210 5226 solver.cpp:218] Iteration 17076 (2.20811 iter/s, 5.43452s/12 iters), loss = 2.51492 I0406 09:37:03.242252 5226 solver.cpp:237] Train net output #0: loss = 2.51492 (* 1 = 2.51492 loss) I0406 09:37:03.242259 5226 sgd_solver.cpp:105] Iteration 17076, lr = 0.01 I0406 09:37:08.610327 5226 solver.cpp:218] Iteration 17088 (2.23546 iter/s, 5.36803s/12 iters), loss = 2.83734 I0406 09:37:08.610363 5226 solver.cpp:237] Train net output #0: loss = 2.83734 (* 1 = 2.83734 loss) I0406 09:37:08.610369 5226 sgd_solver.cpp:105] Iteration 17088, lr = 0.01 I0406 09:37:11.652858 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:37:14.167340 5226 solver.cpp:218] Iteration 17100 (2.15947 iter/s, 5.55692s/12 iters), loss = 2.34522 I0406 09:37:14.167379 5226 solver.cpp:237] Train net output #0: loss = 2.34522 (* 1 = 2.34522 loss) I0406 09:37:14.167385 5226 sgd_solver.cpp:105] Iteration 17100, lr = 0.01 I0406 09:37:19.485165 5226 solver.cpp:218] Iteration 17112 (2.2566 iter/s, 5.31774s/12 iters), loss = 2.60643 I0406 09:37:19.485324 5226 solver.cpp:237] Train net output #0: loss = 2.60643 (* 1 = 2.60643 loss) I0406 09:37:19.485333 5226 sgd_solver.cpp:105] Iteration 17112, lr = 0.01 I0406 09:37:25.069727 5226 solver.cpp:218] Iteration 17124 (2.14888 iter/s, 5.58431s/12 iters), loss = 2.55089 I0406 09:37:25.069777 5226 solver.cpp:237] Train net output #0: loss = 2.55089 (* 1 = 2.55089 loss) I0406 09:37:25.069784 5226 sgd_solver.cpp:105] Iteration 17124, lr = 0.01 I0406 09:37:30.427904 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_17136.caffemodel I0406 09:37:33.598701 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_17136.solverstate I0406 09:37:35.929870 5226 solver.cpp:330] Iteration 17136, Testing net (#0) I0406 09:37:35.929891 5226 net.cpp:676] Ignoring source layer train-data I0406 09:37:38.582072 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:37:40.948179 5226 solver.cpp:397] Test net output #0: accuracy = 0.191176 I0406 09:37:40.948215 5226 solver.cpp:397] Test net output #1: loss = 4.05446 (* 1 = 4.05446 loss) I0406 09:37:41.080511 5226 solver.cpp:218] Iteration 17136 (0.749503 iter/s, 16.0106s/12 iters), loss = 2.55505 I0406 09:37:41.080564 5226 solver.cpp:237] Train net output #0: loss = 2.55505 (* 1 = 2.55505 loss) I0406 09:37:41.080571 5226 sgd_solver.cpp:105] Iteration 17136, lr = 0.01 I0406 09:37:45.866030 5226 solver.cpp:218] Iteration 17148 (2.50762 iter/s, 4.78542s/12 iters), loss = 2.87978 I0406 09:37:45.866082 5226 solver.cpp:237] Train net output #0: loss = 2.87978 (* 1 = 2.87978 loss) I0406 09:37:45.866091 5226 sgd_solver.cpp:105] Iteration 17148, lr = 0.01 I0406 09:37:51.538857 5226 solver.cpp:218] Iteration 17160 (2.11539 iter/s, 5.67272s/12 iters), loss = 2.34037 I0406 09:37:51.539177 5226 solver.cpp:237] Train net output #0: loss = 2.34037 (* 1 = 2.34037 loss) I0406 09:37:51.539187 5226 sgd_solver.cpp:105] Iteration 17160, lr = 0.01 I0406 09:37:57.113538 5226 solver.cpp:218] Iteration 17172 (2.15273 iter/s, 5.57431s/12 iters), loss = 1.82207 I0406 09:37:57.113579 5226 solver.cpp:237] Train net output #0: loss = 1.82207 (* 1 = 1.82207 loss) I0406 09:37:57.113584 5226 sgd_solver.cpp:105] Iteration 17172, lr = 0.01 I0406 09:38:03.035051 5226 solver.cpp:218] Iteration 17184 (2.02654 iter/s, 5.92141s/12 iters), loss = 2.83974 I0406 09:38:03.035102 5226 solver.cpp:237] Train net output #0: loss = 2.83974 (* 1 = 2.83974 loss) I0406 09:38:03.035113 5226 sgd_solver.cpp:105] Iteration 17184, lr = 0.01 I0406 09:38:08.331537 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:38:08.619920 5226 solver.cpp:218] Iteration 17196 (2.1487 iter/s, 5.58477s/12 iters), loss = 3.0153 I0406 09:38:08.619969 5226 solver.cpp:237] Train net output #0: loss = 3.0153 (* 1 = 3.0153 loss) I0406 09:38:08.619980 5226 sgd_solver.cpp:105] Iteration 17196, lr = 0.01 I0406 09:38:14.214838 5226 solver.cpp:218] Iteration 17208 (2.14484 iter/s, 5.59482s/12 iters), loss = 2.6266 I0406 09:38:14.214887 5226 solver.cpp:237] Train net output #0: loss = 2.6266 (* 1 = 2.6266 loss) I0406 09:38:14.214895 5226 sgd_solver.cpp:105] Iteration 17208, lr = 0.01 I0406 09:38:19.997965 5226 solver.cpp:218] Iteration 17220 (2.07504 iter/s, 5.78303s/12 iters), loss = 2.96225 I0406 09:38:19.998005 5226 solver.cpp:237] Train net output #0: loss = 2.96225 (* 1 = 2.96225 loss) I0406 09:38:19.998010 5226 sgd_solver.cpp:105] Iteration 17220, lr = 0.01 I0406 09:38:25.590109 5226 solver.cpp:218] Iteration 17232 (2.1459 iter/s, 5.59205s/12 iters), loss = 2.98649 I0406 09:38:25.590284 5226 solver.cpp:237] Train net output #0: loss = 2.98649 (* 1 = 2.98649 loss) I0406 09:38:25.590296 5226 sgd_solver.cpp:105] Iteration 17232, lr = 0.01 I0406 09:38:27.937952 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_17238.caffemodel I0406 09:38:31.036271 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_17238.solverstate I0406 09:38:33.343325 5226 solver.cpp:330] Iteration 17238, Testing net (#0) I0406 09:38:33.343348 5226 net.cpp:676] Ignoring source layer train-data I0406 09:38:35.849823 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:38:38.255728 5226 solver.cpp:397] Test net output #0: accuracy = 0.147059 I0406 09:38:38.255766 5226 solver.cpp:397] Test net output #1: loss = 4.31391 (* 1 = 4.31391 loss) I0406 09:38:40.310954 5226 solver.cpp:218] Iteration 17244 (0.815188 iter/s, 14.7205s/12 iters), loss = 2.32494 I0406 09:38:40.310992 5226 solver.cpp:237] Train net output #0: loss = 2.32494 (* 1 = 2.32494 loss) I0406 09:38:40.310998 5226 sgd_solver.cpp:105] Iteration 17244, lr = 0.01 I0406 09:38:46.042059 5226 solver.cpp:218] Iteration 17256 (2.09387 iter/s, 5.73101s/12 iters), loss = 2.72251 I0406 09:38:46.042109 5226 solver.cpp:237] Train net output #0: loss = 2.72251 (* 1 = 2.72251 loss) I0406 09:38:46.042116 5226 sgd_solver.cpp:105] Iteration 17256, lr = 0.01 I0406 09:38:51.863628 5226 solver.cpp:218] Iteration 17268 (2.06134 iter/s, 5.82147s/12 iters), loss = 2.9911 I0406 09:38:51.863667 5226 solver.cpp:237] Train net output #0: loss = 2.9911 (* 1 = 2.9911 loss) I0406 09:38:51.863672 5226 sgd_solver.cpp:105] Iteration 17268, lr = 0.01 I0406 09:38:57.512382 5226 solver.cpp:218] Iteration 17280 (2.1244 iter/s, 5.64867s/12 iters), loss = 2.57827 I0406 09:38:57.512476 5226 solver.cpp:237] Train net output #0: loss = 2.57827 (* 1 = 2.57827 loss) I0406 09:38:57.512485 5226 sgd_solver.cpp:105] Iteration 17280, lr = 0.01 I0406 09:39:03.168375 5226 solver.cpp:218] Iteration 17292 (2.12171 iter/s, 5.6558s/12 iters), loss = 2.88501 I0406 09:39:03.168431 5226 solver.cpp:237] Train net output #0: loss = 2.88501 (* 1 = 2.88501 loss) I0406 09:39:03.168440 5226 sgd_solver.cpp:105] Iteration 17292, lr = 0.01 I0406 09:39:05.353889 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:39:09.009541 5226 solver.cpp:218] Iteration 17304 (2.05442 iter/s, 5.84106s/12 iters), loss = 2.35144 I0406 09:39:09.009588 5226 solver.cpp:237] Train net output #0: loss = 2.35144 (* 1 = 2.35144 loss) I0406 09:39:09.009594 5226 sgd_solver.cpp:105] Iteration 17304, lr = 0.01 I0406 09:39:14.392957 5226 solver.cpp:218] Iteration 17316 (2.22911 iter/s, 5.38332s/12 iters), loss = 2.58688 I0406 09:39:14.393007 5226 solver.cpp:237] Train net output #0: loss = 2.58688 (* 1 = 2.58688 loss) I0406 09:39:14.393015 5226 sgd_solver.cpp:105] Iteration 17316, lr = 0.01 I0406 09:39:20.129498 5226 solver.cpp:218] Iteration 17328 (2.09189 iter/s, 5.73644s/12 iters), loss = 2.35575 I0406 09:39:20.129536 5226 solver.cpp:237] Train net output #0: loss = 2.35575 (* 1 = 2.35575 loss) I0406 09:39:20.129541 5226 sgd_solver.cpp:105] Iteration 17328, lr = 0.01 I0406 09:39:25.106586 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_17340.caffemodel I0406 09:39:28.809131 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_17340.solverstate I0406 09:39:32.086956 5226 solver.cpp:330] Iteration 17340, Testing net (#0) I0406 09:39:32.086977 5226 net.cpp:676] Ignoring source layer train-data I0406 09:39:33.291972 5226 blocking_queue.cpp:49] Waiting for data I0406 09:39:34.595803 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:39:37.039052 5226 solver.cpp:397] Test net output #0: accuracy = 0.177696 I0406 09:39:37.039090 5226 solver.cpp:397] Test net output #1: loss = 4.11346 (* 1 = 4.11346 loss) I0406 09:39:37.180872 5226 solver.cpp:218] Iteration 17340 (0.703762 iter/s, 17.0512s/12 iters), loss = 2.49296 I0406 09:39:37.180927 5226 solver.cpp:237] Train net output #0: loss = 2.49296 (* 1 = 2.49296 loss) I0406 09:39:37.180934 5226 sgd_solver.cpp:105] Iteration 17340, lr = 0.01 I0406 09:39:41.820314 5226 solver.cpp:218] Iteration 17352 (2.58657 iter/s, 4.63934s/12 iters), loss = 2.5964 I0406 09:39:41.820354 5226 solver.cpp:237] Train net output #0: loss = 2.5964 (* 1 = 2.5964 loss) I0406 09:39:41.820360 5226 sgd_solver.cpp:105] Iteration 17352, lr = 0.01 I0406 09:39:47.570683 5226 solver.cpp:218] Iteration 17364 (2.08686 iter/s, 5.75028s/12 iters), loss = 3.09626 I0406 09:39:47.570719 5226 solver.cpp:237] Train net output #0: loss = 3.09626 (* 1 = 3.09626 loss) I0406 09:39:47.570724 5226 sgd_solver.cpp:105] Iteration 17364, lr = 0.01 I0406 09:39:53.394436 5226 solver.cpp:218] Iteration 17376 (2.06056 iter/s, 5.82366s/12 iters), loss = 3.01261 I0406 09:39:53.394482 5226 solver.cpp:237] Train net output #0: loss = 3.01261 (* 1 = 3.01261 loss) I0406 09:39:53.394490 5226 sgd_solver.cpp:105] Iteration 17376, lr = 0.01 I0406 09:39:58.968819 5226 solver.cpp:218] Iteration 17388 (2.15274 iter/s, 5.57428s/12 iters), loss = 2.8225 I0406 09:39:58.968945 5226 solver.cpp:237] Train net output #0: loss = 2.8225 (* 1 = 2.8225 loss) I0406 09:39:58.968953 5226 sgd_solver.cpp:105] Iteration 17388, lr = 0.01 I0406 09:40:03.167459 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:40:04.281426 5226 solver.cpp:218] Iteration 17400 (2.25889 iter/s, 5.31236s/12 iters), loss = 2.58022 I0406 09:40:04.281474 5226 solver.cpp:237] Train net output #0: loss = 2.58022 (* 1 = 2.58022 loss) I0406 09:40:04.281482 5226 sgd_solver.cpp:105] Iteration 17400, lr = 0.01 I0406 09:40:10.085455 5226 solver.cpp:218] Iteration 17412 (2.06756 iter/s, 5.80393s/12 iters), loss = 2.47987 I0406 09:40:10.085494 5226 solver.cpp:237] Train net output #0: loss = 2.47987 (* 1 = 2.47987 loss) I0406 09:40:10.085500 5226 sgd_solver.cpp:105] Iteration 17412, lr = 0.01 I0406 09:40:15.592054 5226 solver.cpp:218] Iteration 17424 (2.17924 iter/s, 5.50651s/12 iters), loss = 2.34384 I0406 09:40:15.592103 5226 solver.cpp:237] Train net output #0: loss = 2.34384 (* 1 = 2.34384 loss) I0406 09:40:15.592111 5226 sgd_solver.cpp:105] Iteration 17424, lr = 0.01 I0406 09:40:21.275457 5226 solver.cpp:218] Iteration 17436 (2.11145 iter/s, 5.6833s/12 iters), loss = 2.36691 I0406 09:40:21.275506 5226 solver.cpp:237] Train net output #0: loss = 2.36691 (* 1 = 2.36691 loss) I0406 09:40:21.275513 5226 sgd_solver.cpp:105] Iteration 17436, lr = 0.01 I0406 09:40:23.631533 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_17442.caffemodel I0406 09:40:26.797827 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_17442.solverstate I0406 09:40:29.124644 5226 solver.cpp:330] Iteration 17442, Testing net (#0) I0406 09:40:29.124733 5226 net.cpp:676] Ignoring source layer train-data I0406 09:40:31.525674 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:40:33.914196 5226 solver.cpp:397] Test net output #0: accuracy = 0.18076 I0406 09:40:33.914232 5226 solver.cpp:397] Test net output #1: loss = 4.08108 (* 1 = 4.08108 loss) I0406 09:40:35.803243 5226 solver.cpp:218] Iteration 17448 (0.826012 iter/s, 14.5276s/12 iters), loss = 2.74822 I0406 09:40:35.803297 5226 solver.cpp:237] Train net output #0: loss = 2.74822 (* 1 = 2.74822 loss) I0406 09:40:35.803306 5226 sgd_solver.cpp:105] Iteration 17448, lr = 0.01 I0406 09:40:41.419297 5226 solver.cpp:218] Iteration 17460 (2.13677 iter/s, 5.61595s/12 iters), loss = 2.78272 I0406 09:40:41.419353 5226 solver.cpp:237] Train net output #0: loss = 2.78272 (* 1 = 2.78272 loss) I0406 09:40:41.419361 5226 sgd_solver.cpp:105] Iteration 17460, lr = 0.01 I0406 09:40:47.009218 5226 solver.cpp:218] Iteration 17472 (2.14676 iter/s, 5.58982s/12 iters), loss = 2.55214 I0406 09:40:47.009258 5226 solver.cpp:237] Train net output #0: loss = 2.55214 (* 1 = 2.55214 loss) I0406 09:40:47.009263 5226 sgd_solver.cpp:105] Iteration 17472, lr = 0.01 I0406 09:40:52.788568 5226 solver.cpp:218] Iteration 17484 (2.07639 iter/s, 5.77926s/12 iters), loss = 2.87774 I0406 09:40:52.788615 5226 solver.cpp:237] Train net output #0: loss = 2.87774 (* 1 = 2.87774 loss) I0406 09:40:52.788625 5226 sgd_solver.cpp:105] Iteration 17484, lr = 0.01 I0406 09:40:58.089170 5226 solver.cpp:218] Iteration 17496 (2.26393 iter/s, 5.30051s/12 iters), loss = 2.46048 I0406 09:40:58.089208 5226 solver.cpp:237] Train net output #0: loss = 2.46048 (* 1 = 2.46048 loss) I0406 09:40:58.089213 5226 sgd_solver.cpp:105] Iteration 17496, lr = 0.01 I0406 09:40:59.556021 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:41:03.845177 5226 solver.cpp:218] Iteration 17508 (2.08481 iter/s, 5.75591s/12 iters), loss = 2.50711 I0406 09:41:03.845230 5226 solver.cpp:237] Train net output #0: loss = 2.50711 (* 1 = 2.50711 loss) I0406 09:41:03.845240 5226 sgd_solver.cpp:105] Iteration 17508, lr = 0.01 I0406 09:41:09.743705 5226 solver.cpp:218] Iteration 17520 (2.03444 iter/s, 5.89843s/12 iters), loss = 3.28702 I0406 09:41:09.743748 5226 solver.cpp:237] Train net output #0: loss = 3.28702 (* 1 = 3.28702 loss) I0406 09:41:09.743754 5226 sgd_solver.cpp:105] Iteration 17520, lr = 0.01 I0406 09:41:15.334728 5226 solver.cpp:218] Iteration 17532 (2.14633 iter/s, 5.59093s/12 iters), loss = 2.5581 I0406 09:41:15.334772 5226 solver.cpp:237] Train net output #0: loss = 2.5581 (* 1 = 2.5581 loss) I0406 09:41:15.334780 5226 sgd_solver.cpp:105] Iteration 17532, lr = 0.01 I0406 09:41:20.443552 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_17544.caffemodel I0406 09:41:23.565848 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_17544.solverstate I0406 09:41:25.899669 5226 solver.cpp:330] Iteration 17544, Testing net (#0) I0406 09:41:25.899690 5226 net.cpp:676] Ignoring source layer train-data I0406 09:41:28.103119 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:41:30.590183 5226 solver.cpp:397] Test net output #0: accuracy = 0.16973 I0406 09:41:30.590278 5226 solver.cpp:397] Test net output #1: loss = 4.19862 (* 1 = 4.19862 loss) I0406 09:41:30.732620 5226 solver.cpp:218] Iteration 17544 (0.779335 iter/s, 15.3977s/12 iters), loss = 2.6258 I0406 09:41:30.732661 5226 solver.cpp:237] Train net output #0: loss = 2.6258 (* 1 = 2.6258 loss) I0406 09:41:30.732667 5226 sgd_solver.cpp:105] Iteration 17544, lr = 0.01 I0406 09:41:35.509629 5226 solver.cpp:218] Iteration 17556 (2.51208 iter/s, 4.77691s/12 iters), loss = 2.36219 I0406 09:41:35.509680 5226 solver.cpp:237] Train net output #0: loss = 2.36219 (* 1 = 2.36219 loss) I0406 09:41:35.509688 5226 sgd_solver.cpp:105] Iteration 17556, lr = 0.01 I0406 09:41:41.038118 5226 solver.cpp:218] Iteration 17568 (2.17062 iter/s, 5.52838s/12 iters), loss = 2.63459 I0406 09:41:41.038172 5226 solver.cpp:237] Train net output #0: loss = 2.63459 (* 1 = 2.63459 loss) I0406 09:41:41.038179 5226 sgd_solver.cpp:105] Iteration 17568, lr = 0.01 I0406 09:41:46.357728 5226 solver.cpp:218] Iteration 17580 (2.25585 iter/s, 5.31951s/12 iters), loss = 2.49699 I0406 09:41:46.357781 5226 solver.cpp:237] Train net output #0: loss = 2.49699 (* 1 = 2.49699 loss) I0406 09:41:46.357791 5226 sgd_solver.cpp:105] Iteration 17580, lr = 0.01 I0406 09:41:51.956640 5226 solver.cpp:218] Iteration 17592 (2.14331 iter/s, 5.59881s/12 iters), loss = 2.92978 I0406 09:41:51.956688 5226 solver.cpp:237] Train net output #0: loss = 2.92978 (* 1 = 2.92978 loss) I0406 09:41:51.956696 5226 sgd_solver.cpp:105] Iteration 17592, lr = 0.01 I0406 09:41:55.978201 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:41:57.912783 5226 solver.cpp:218] Iteration 17604 (2.01476 iter/s, 5.95604s/12 iters), loss = 2.82676 I0406 09:41:57.912840 5226 solver.cpp:237] Train net output #0: loss = 2.82676 (* 1 = 2.82676 loss) I0406 09:41:57.912850 5226 sgd_solver.cpp:105] Iteration 17604, lr = 0.01 I0406 09:42:03.658324 5226 solver.cpp:218] Iteration 17616 (2.08861 iter/s, 5.74544s/12 iters), loss = 2.90723 I0406 09:42:03.658469 5226 solver.cpp:237] Train net output #0: loss = 2.90723 (* 1 = 2.90723 loss) I0406 09:42:03.658480 5226 sgd_solver.cpp:105] Iteration 17616, lr = 0.01 I0406 09:42:09.091034 5226 solver.cpp:218] Iteration 17628 (2.20892 iter/s, 5.43251s/12 iters), loss = 2.54831 I0406 09:42:09.091084 5226 solver.cpp:237] Train net output #0: loss = 2.54831 (* 1 = 2.54831 loss) I0406 09:42:09.091091 5226 sgd_solver.cpp:105] Iteration 17628, lr = 0.01 I0406 09:42:14.881716 5226 solver.cpp:218] Iteration 17640 (2.07233 iter/s, 5.79058s/12 iters), loss = 2.67157 I0406 09:42:14.881767 5226 solver.cpp:237] Train net output #0: loss = 2.67157 (* 1 = 2.67157 loss) I0406 09:42:14.881774 5226 sgd_solver.cpp:105] Iteration 17640, lr = 0.01 I0406 09:42:17.205209 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_17646.caffemodel I0406 09:42:21.559206 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_17646.solverstate I0406 09:42:25.024138 5226 solver.cpp:330] Iteration 17646, Testing net (#0) I0406 09:42:25.024155 5226 net.cpp:676] Ignoring source layer train-data I0406 09:42:27.448318 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:42:29.877810 5226 solver.cpp:397] Test net output #0: accuracy = 0.164828 I0406 09:42:29.877846 5226 solver.cpp:397] Test net output #1: loss = 4.17957 (* 1 = 4.17957 loss) I0406 09:42:31.970463 5226 solver.cpp:218] Iteration 17652 (0.702224 iter/s, 17.0886s/12 iters), loss = 2.66953 I0406 09:42:31.970516 5226 solver.cpp:237] Train net output #0: loss = 2.66953 (* 1 = 2.66953 loss) I0406 09:42:31.970525 5226 sgd_solver.cpp:105] Iteration 17652, lr = 0.01 I0406 09:42:37.527065 5226 solver.cpp:218] Iteration 17664 (2.15963 iter/s, 5.5565s/12 iters), loss = 2.8272 I0406 09:42:37.527209 5226 solver.cpp:237] Train net output #0: loss = 2.8272 (* 1 = 2.8272 loss) I0406 09:42:37.527217 5226 sgd_solver.cpp:105] Iteration 17664, lr = 0.01 I0406 09:42:43.192404 5226 solver.cpp:218] Iteration 17676 (2.11822 iter/s, 5.66515s/12 iters), loss = 2.84847 I0406 09:42:43.192459 5226 solver.cpp:237] Train net output #0: loss = 2.84847 (* 1 = 2.84847 loss) I0406 09:42:43.192468 5226 sgd_solver.cpp:105] Iteration 17676, lr = 0.01 I0406 09:42:48.638018 5226 solver.cpp:218] Iteration 17688 (2.20365 iter/s, 5.44551s/12 iters), loss = 2.73282 I0406 09:42:48.638067 5226 solver.cpp:237] Train net output #0: loss = 2.73282 (* 1 = 2.73282 loss) I0406 09:42:48.638075 5226 sgd_solver.cpp:105] Iteration 17688, lr = 0.01 I0406 09:42:54.344620 5226 solver.cpp:218] Iteration 17700 (2.10286 iter/s, 5.7065s/12 iters), loss = 2.72733 I0406 09:42:54.344671 5226 solver.cpp:237] Train net output #0: loss = 2.72733 (* 1 = 2.72733 loss) I0406 09:42:54.344679 5226 sgd_solver.cpp:105] Iteration 17700, lr = 0.01 I0406 09:42:54.978006 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:43:00.075264 5226 solver.cpp:218] Iteration 17712 (2.09404 iter/s, 5.73054s/12 iters), loss = 2.45168 I0406 09:43:00.075321 5226 solver.cpp:237] Train net output #0: loss = 2.45168 (* 1 = 2.45168 loss) I0406 09:43:00.075330 5226 sgd_solver.cpp:105] Iteration 17712, lr = 0.01 I0406 09:43:05.775935 5226 solver.cpp:218] Iteration 17724 (2.10505 iter/s, 5.70057s/12 iters), loss = 2.6602 I0406 09:43:05.775977 5226 solver.cpp:237] Train net output #0: loss = 2.6602 (* 1 = 2.6602 loss) I0406 09:43:05.775982 5226 sgd_solver.cpp:105] Iteration 17724, lr = 0.01 I0406 09:43:11.458418 5226 solver.cpp:218] Iteration 17736 (2.11179 iter/s, 5.68239s/12 iters), loss = 2.49153 I0406 09:43:11.458571 5226 solver.cpp:237] Train net output #0: loss = 2.49153 (* 1 = 2.49153 loss) I0406 09:43:11.458580 5226 sgd_solver.cpp:105] Iteration 17736, lr = 0.01 I0406 09:43:16.625604 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_17748.caffemodel I0406 09:43:19.796114 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_17748.solverstate I0406 09:43:23.001834 5226 solver.cpp:330] Iteration 17748, Testing net (#0) I0406 09:43:23.001857 5226 net.cpp:676] Ignoring source layer train-data I0406 09:43:25.183641 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:43:27.665089 5226 solver.cpp:397] Test net output #0: accuracy = 0.18076 I0406 09:43:27.665128 5226 solver.cpp:397] Test net output #1: loss = 4.09905 (* 1 = 4.09905 loss) I0406 09:43:27.799749 5226 solver.cpp:218] Iteration 17748 (0.734346 iter/s, 16.3411s/12 iters), loss = 2.44148 I0406 09:43:27.804927 5226 solver.cpp:237] Train net output #0: loss = 2.44148 (* 1 = 2.44148 loss) I0406 09:43:27.804949 5226 sgd_solver.cpp:105] Iteration 17748, lr = 0.01 I0406 09:43:32.584304 5226 solver.cpp:218] Iteration 17760 (2.5108 iter/s, 4.77935s/12 iters), loss = 2.87457 I0406 09:43:32.584342 5226 solver.cpp:237] Train net output #0: loss = 2.87457 (* 1 = 2.87457 loss) I0406 09:43:32.584347 5226 sgd_solver.cpp:105] Iteration 17760, lr = 0.01 I0406 09:43:38.103519 5226 solver.cpp:218] Iteration 17772 (2.17426 iter/s, 5.51913s/12 iters), loss = 2.29188 I0406 09:43:38.103559 5226 solver.cpp:237] Train net output #0: loss = 2.29188 (* 1 = 2.29188 loss) I0406 09:43:38.103564 5226 sgd_solver.cpp:105] Iteration 17772, lr = 0.01 I0406 09:43:43.797216 5226 solver.cpp:218] Iteration 17784 (2.10763 iter/s, 5.69361s/12 iters), loss = 2.96446 I0406 09:43:43.797338 5226 solver.cpp:237] Train net output #0: loss = 2.96446 (* 1 = 2.96446 loss) I0406 09:43:43.797343 5226 sgd_solver.cpp:105] Iteration 17784, lr = 0.01 I0406 09:43:49.245136 5226 solver.cpp:218] Iteration 17796 (2.20274 iter/s, 5.44775s/12 iters), loss = 2.7726 I0406 09:43:49.245188 5226 solver.cpp:237] Train net output #0: loss = 2.7726 (* 1 = 2.7726 loss) I0406 09:43:49.245199 5226 sgd_solver.cpp:105] Iteration 17796, lr = 0.01 I0406 09:43:52.279206 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:43:54.981252 5226 solver.cpp:218] Iteration 17808 (2.09205 iter/s, 5.73601s/12 iters), loss = 2.39674 I0406 09:43:54.981302 5226 solver.cpp:237] Train net output #0: loss = 2.39674 (* 1 = 2.39674 loss) I0406 09:43:54.981309 5226 sgd_solver.cpp:105] Iteration 17808, lr = 0.01 I0406 09:44:00.669909 5226 solver.cpp:218] Iteration 17820 (2.1095 iter/s, 5.68856s/12 iters), loss = 2.37164 I0406 09:44:00.669946 5226 solver.cpp:237] Train net output #0: loss = 2.37164 (* 1 = 2.37164 loss) I0406 09:44:00.669951 5226 sgd_solver.cpp:105] Iteration 17820, lr = 0.01 I0406 09:44:06.489933 5226 solver.cpp:218] Iteration 17832 (2.06188 iter/s, 5.81992s/12 iters), loss = 2.66357 I0406 09:44:06.489987 5226 solver.cpp:237] Train net output #0: loss = 2.66357 (* 1 = 2.66357 loss) I0406 09:44:06.489996 5226 sgd_solver.cpp:105] Iteration 17832, lr = 0.01 I0406 09:44:12.069243 5226 solver.cpp:218] Iteration 17844 (2.15084 iter/s, 5.57921s/12 iters), loss = 2.5385 I0406 09:44:12.069300 5226 solver.cpp:237] Train net output #0: loss = 2.5385 (* 1 = 2.5385 loss) I0406 09:44:12.069308 5226 sgd_solver.cpp:105] Iteration 17844, lr = 0.01 I0406 09:44:14.220479 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_17850.caffemodel I0406 09:44:17.476472 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_17850.solverstate I0406 09:44:20.665552 5226 solver.cpp:330] Iteration 17850, Testing net (#0) I0406 09:44:20.665571 5226 net.cpp:676] Ignoring source layer train-data I0406 09:44:22.897233 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:44:25.493468 5226 solver.cpp:397] Test net output #0: accuracy = 0.150123 I0406 09:44:25.493507 5226 solver.cpp:397] Test net output #1: loss = 4.28018 (* 1 = 4.28018 loss) I0406 09:44:27.499778 5226 solver.cpp:218] Iteration 17856 (0.777687 iter/s, 15.4304s/12 iters), loss = 2.42105 I0406 09:44:27.499830 5226 solver.cpp:237] Train net output #0: loss = 2.42105 (* 1 = 2.42105 loss) I0406 09:44:27.499840 5226 sgd_solver.cpp:105] Iteration 17856, lr = 0.01 I0406 09:44:33.018020 5226 solver.cpp:218] Iteration 17868 (2.17464 iter/s, 5.51815s/12 iters), loss = 2.36433 I0406 09:44:33.018066 5226 solver.cpp:237] Train net output #0: loss = 2.36433 (* 1 = 2.36433 loss) I0406 09:44:33.018074 5226 sgd_solver.cpp:105] Iteration 17868, lr = 0.01 I0406 09:44:38.647756 5226 solver.cpp:218] Iteration 17880 (2.13158 iter/s, 5.62963s/12 iters), loss = 3.07001 I0406 09:44:38.647810 5226 solver.cpp:237] Train net output #0: loss = 3.07001 (* 1 = 3.07001 loss) I0406 09:44:38.647819 5226 sgd_solver.cpp:105] Iteration 17880, lr = 0.01 I0406 09:44:44.034554 5226 solver.cpp:218] Iteration 17892 (2.22771 iter/s, 5.3867s/12 iters), loss = 2.85333 I0406 09:44:44.034605 5226 solver.cpp:237] Train net output #0: loss = 2.85333 (* 1 = 2.85333 loss) I0406 09:44:44.034616 5226 sgd_solver.cpp:105] Iteration 17892, lr = 0.01 I0406 09:44:49.361501 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:44:49.670512 5226 solver.cpp:218] Iteration 17904 (2.12922 iter/s, 5.63586s/12 iters), loss = 2.99441 I0406 09:44:49.670562 5226 solver.cpp:237] Train net output #0: loss = 2.99441 (* 1 = 2.99441 loss) I0406 09:44:49.670569 5226 sgd_solver.cpp:105] Iteration 17904, lr = 0.01 I0406 09:44:55.390885 5226 solver.cpp:218] Iteration 17916 (2.09782 iter/s, 5.72022s/12 iters), loss = 2.85584 I0406 09:44:55.390933 5226 solver.cpp:237] Train net output #0: loss = 2.85584 (* 1 = 2.85584 loss) I0406 09:44:55.390941 5226 sgd_solver.cpp:105] Iteration 17916, lr = 0.01 I0406 09:45:01.098598 5226 solver.cpp:218] Iteration 17928 (2.10245 iter/s, 5.70762s/12 iters), loss = 2.64041 I0406 09:45:01.098647 5226 solver.cpp:237] Train net output #0: loss = 2.64041 (* 1 = 2.64041 loss) I0406 09:45:01.098655 5226 sgd_solver.cpp:105] Iteration 17928, lr = 0.01 I0406 09:45:06.719770 5226 solver.cpp:218] Iteration 17940 (2.13482 iter/s, 5.62107s/12 iters), loss = 2.74439 I0406 09:45:06.719822 5226 solver.cpp:237] Train net output #0: loss = 2.74439 (* 1 = 2.74439 loss) I0406 09:45:06.719830 5226 sgd_solver.cpp:105] Iteration 17940, lr = 0.01 I0406 09:45:11.672817 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_17952.caffemodel I0406 09:45:14.797783 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_17952.solverstate I0406 09:45:17.644163 5226 solver.cpp:330] Iteration 17952, Testing net (#0) I0406 09:45:17.644187 5226 net.cpp:676] Ignoring source layer train-data I0406 09:45:19.912022 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:45:22.637588 5226 solver.cpp:397] Test net output #0: accuracy = 0.173407 I0406 09:45:22.637622 5226 solver.cpp:397] Test net output #1: loss = 4.1446 (* 1 = 4.1446 loss) I0406 09:45:22.810892 5226 solver.cpp:218] Iteration 17952 (0.74576 iter/s, 16.091s/12 iters), loss = 3.14317 I0406 09:45:22.812471 5226 solver.cpp:237] Train net output #0: loss = 3.14317 (* 1 = 3.14317 loss) I0406 09:45:22.812486 5226 sgd_solver.cpp:105] Iteration 17952, lr = 0.01 I0406 09:45:27.159741 5226 solver.cpp:218] Iteration 17964 (2.76037 iter/s, 4.34724s/12 iters), loss = 2.29988 I0406 09:45:27.159785 5226 solver.cpp:237] Train net output #0: loss = 2.29988 (* 1 = 2.29988 loss) I0406 09:45:27.159792 5226 sgd_solver.cpp:105] Iteration 17964, lr = 0.01 I0406 09:45:32.843408 5226 solver.cpp:218] Iteration 17976 (2.11135 iter/s, 5.68357s/12 iters), loss = 2.84172 I0406 09:45:32.843456 5226 solver.cpp:237] Train net output #0: loss = 2.84172 (* 1 = 2.84172 loss) I0406 09:45:32.843464 5226 sgd_solver.cpp:105] Iteration 17976, lr = 0.01 I0406 09:45:38.705754 5226 solver.cpp:218] Iteration 17988 (2.047 iter/s, 5.86224s/12 iters), loss = 2.48493 I0406 09:45:38.705799 5226 solver.cpp:237] Train net output #0: loss = 2.48493 (* 1 = 2.48493 loss) I0406 09:45:38.705807 5226 sgd_solver.cpp:105] Iteration 17988, lr = 0.01 I0406 09:45:44.316476 5226 solver.cpp:218] Iteration 18000 (2.1388 iter/s, 5.61063s/12 iters), loss = 2.77124 I0406 09:45:44.316524 5226 solver.cpp:237] Train net output #0: loss = 2.77124 (* 1 = 2.77124 loss) I0406 09:45:44.316530 5226 sgd_solver.cpp:105] Iteration 18000, lr = 0.01 I0406 09:45:46.298668 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:45:49.905369 5226 solver.cpp:218] Iteration 18012 (2.14715 iter/s, 5.5888s/12 iters), loss = 2.40614 I0406 09:45:49.905408 5226 solver.cpp:237] Train net output #0: loss = 2.40614 (* 1 = 2.40614 loss) I0406 09:45:49.905413 5226 sgd_solver.cpp:105] Iteration 18012, lr = 0.01 I0406 09:45:55.460817 5226 solver.cpp:218] Iteration 18024 (2.16007 iter/s, 5.55536s/12 iters), loss = 3.00819 I0406 09:45:55.460912 5226 solver.cpp:237] Train net output #0: loss = 3.00819 (* 1 = 3.00819 loss) I0406 09:45:55.460918 5226 sgd_solver.cpp:105] Iteration 18024, lr = 0.01 I0406 09:46:01.024089 5226 solver.cpp:218] Iteration 18036 (2.15706 iter/s, 5.56313s/12 iters), loss = 2.86129 I0406 09:46:01.024142 5226 solver.cpp:237] Train net output #0: loss = 2.86129 (* 1 = 2.86129 loss) I0406 09:46:01.024148 5226 sgd_solver.cpp:105] Iteration 18036, lr = 0.01 I0406 09:46:01.024410 5226 blocking_queue.cpp:49] Waiting for data I0406 09:46:06.534756 5226 solver.cpp:218] Iteration 18048 (2.17763 iter/s, 5.51057s/12 iters), loss = 2.72902 I0406 09:46:06.534806 5226 solver.cpp:237] Train net output #0: loss = 2.72902 (* 1 = 2.72902 loss) I0406 09:46:06.534814 5226 sgd_solver.cpp:105] Iteration 18048, lr = 0.01 I0406 09:46:08.666837 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_18054.caffemodel I0406 09:46:11.846040 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_18054.solverstate I0406 09:46:14.187741 5226 solver.cpp:330] Iteration 18054, Testing net (#0) I0406 09:46:14.187764 5226 net.cpp:676] Ignoring source layer train-data I0406 09:46:16.292692 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:46:19.177889 5226 solver.cpp:397] Test net output #0: accuracy = 0.17402 I0406 09:46:19.177927 5226 solver.cpp:397] Test net output #1: loss = 4.12615 (* 1 = 4.12615 loss) I0406 09:46:21.263212 5226 solver.cpp:218] Iteration 18060 (0.814758 iter/s, 14.7283s/12 iters), loss = 3.00604 I0406 09:46:21.263268 5226 solver.cpp:237] Train net output #0: loss = 3.00604 (* 1 = 3.00604 loss) I0406 09:46:21.263278 5226 sgd_solver.cpp:105] Iteration 18060, lr = 0.01 I0406 09:46:26.970561 5226 solver.cpp:218] Iteration 18072 (2.10259 iter/s, 5.70724s/12 iters), loss = 2.908 I0406 09:46:26.970734 5226 solver.cpp:237] Train net output #0: loss = 2.908 (* 1 = 2.908 loss) I0406 09:46:26.970746 5226 sgd_solver.cpp:105] Iteration 18072, lr = 0.01 I0406 09:46:32.481107 5226 solver.cpp:218] Iteration 18084 (2.17773 iter/s, 5.51033s/12 iters), loss = 3.11263 I0406 09:46:32.481148 5226 solver.cpp:237] Train net output #0: loss = 3.11263 (* 1 = 3.11263 loss) I0406 09:46:32.481153 5226 sgd_solver.cpp:105] Iteration 18084, lr = 0.01 I0406 09:46:37.988409 5226 solver.cpp:218] Iteration 18096 (2.17896 iter/s, 5.50721s/12 iters), loss = 2.71296 I0406 09:46:37.988445 5226 solver.cpp:237] Train net output #0: loss = 2.71296 (* 1 = 2.71296 loss) I0406 09:46:37.988451 5226 sgd_solver.cpp:105] Iteration 18096, lr = 0.01 I0406 09:46:42.274039 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:46:43.411275 5226 solver.cpp:218] Iteration 18108 (2.21289 iter/s, 5.42278s/12 iters), loss = 3.14407 I0406 09:46:43.411327 5226 solver.cpp:237] Train net output #0: loss = 3.14407 (* 1 = 3.14407 loss) I0406 09:46:43.411336 5226 sgd_solver.cpp:105] Iteration 18108, lr = 0.01 I0406 09:46:49.133911 5226 solver.cpp:218] Iteration 18120 (2.09697 iter/s, 5.72253s/12 iters), loss = 2.93491 I0406 09:46:49.133958 5226 solver.cpp:237] Train net output #0: loss = 2.93491 (* 1 = 2.93491 loss) I0406 09:46:49.133966 5226 sgd_solver.cpp:105] Iteration 18120, lr = 0.01 I0406 09:46:54.562464 5226 solver.cpp:218] Iteration 18132 (2.21059 iter/s, 5.42841s/12 iters), loss = 3.30812 I0406 09:46:54.562516 5226 solver.cpp:237] Train net output #0: loss = 3.30812 (* 1 = 3.30812 loss) I0406 09:46:54.562526 5226 sgd_solver.cpp:105] Iteration 18132, lr = 0.01 I0406 09:47:00.502377 5226 solver.cpp:218] Iteration 18144 (2.02027 iter/s, 5.93981s/12 iters), loss = 2.33789 I0406 09:47:00.502491 5226 solver.cpp:237] Train net output #0: loss = 2.33789 (* 1 = 2.33789 loss) I0406 09:47:00.502498 5226 sgd_solver.cpp:105] Iteration 18144, lr = 0.01 I0406 09:47:05.533103 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_18156.caffemodel I0406 09:47:08.666015 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_18156.solverstate I0406 09:47:12.214016 5226 solver.cpp:330] Iteration 18156, Testing net (#0) I0406 09:47:12.214040 5226 net.cpp:676] Ignoring source layer train-data I0406 09:47:14.216341 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:47:16.972262 5226 solver.cpp:397] Test net output #0: accuracy = 0.139093 I0406 09:47:16.972298 5226 solver.cpp:397] Test net output #1: loss = 4.27292 (* 1 = 4.27292 loss) I0406 09:47:17.118527 5226 solver.cpp:218] Iteration 18156 (0.722199 iter/s, 16.6159s/12 iters), loss = 2.86034 I0406 09:47:17.118567 5226 solver.cpp:237] Train net output #0: loss = 2.86034 (* 1 = 2.86034 loss) I0406 09:47:17.118573 5226 sgd_solver.cpp:105] Iteration 18156, lr = 0.01 I0406 09:47:21.930510 5226 solver.cpp:218] Iteration 18168 (2.49385 iter/s, 4.81184s/12 iters), loss = 2.69061 I0406 09:47:21.930557 5226 solver.cpp:237] Train net output #0: loss = 2.69061 (* 1 = 2.69061 loss) I0406 09:47:21.930565 5226 sgd_solver.cpp:105] Iteration 18168, lr = 0.01 I0406 09:47:27.801985 5226 solver.cpp:218] Iteration 18180 (2.04381 iter/s, 5.87137s/12 iters), loss = 2.9345 I0406 09:47:27.802031 5226 solver.cpp:237] Train net output #0: loss = 2.9345 (* 1 = 2.9345 loss) I0406 09:47:27.802039 5226 sgd_solver.cpp:105] Iteration 18180, lr = 0.01 I0406 09:47:33.652521 5226 solver.cpp:218] Iteration 18192 (2.05113 iter/s, 5.85044s/12 iters), loss = 2.59265 I0406 09:47:33.652631 5226 solver.cpp:237] Train net output #0: loss = 2.59265 (* 1 = 2.59265 loss) I0406 09:47:33.652638 5226 sgd_solver.cpp:105] Iteration 18192, lr = 0.01 I0406 09:47:39.501034 5226 solver.cpp:218] Iteration 18204 (2.05186 iter/s, 5.84835s/12 iters), loss = 2.48892 I0406 09:47:39.501085 5226 solver.cpp:237] Train net output #0: loss = 2.48892 (* 1 = 2.48892 loss) I0406 09:47:39.501094 5226 sgd_solver.cpp:105] Iteration 18204, lr = 0.01 I0406 09:47:40.780498 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:47:44.769239 5226 solver.cpp:218] Iteration 18216 (2.27786 iter/s, 5.26811s/12 iters), loss = 2.83364 I0406 09:47:44.769287 5226 solver.cpp:237] Train net output #0: loss = 2.83364 (* 1 = 2.83364 loss) I0406 09:47:44.769295 5226 sgd_solver.cpp:105] Iteration 18216, lr = 0.01 I0406 09:47:50.581367 5226 solver.cpp:218] Iteration 18228 (2.06468 iter/s, 5.81203s/12 iters), loss = 3.19854 I0406 09:47:50.581427 5226 solver.cpp:237] Train net output #0: loss = 3.19854 (* 1 = 3.19854 loss) I0406 09:47:50.581436 5226 sgd_solver.cpp:105] Iteration 18228, lr = 0.01 I0406 09:47:56.336108 5226 solver.cpp:218] Iteration 18240 (2.08528 iter/s, 5.75463s/12 iters), loss = 2.73519 I0406 09:47:56.336158 5226 solver.cpp:237] Train net output #0: loss = 2.73519 (* 1 = 2.73519 loss) I0406 09:47:56.336164 5226 sgd_solver.cpp:105] Iteration 18240, lr = 0.01 I0406 09:48:02.105165 5226 solver.cpp:218] Iteration 18252 (2.0801 iter/s, 5.76896s/12 iters), loss = 2.92738 I0406 09:48:02.105206 5226 solver.cpp:237] Train net output #0: loss = 2.92738 (* 1 = 2.92738 loss) I0406 09:48:02.105211 5226 sgd_solver.cpp:105] Iteration 18252, lr = 0.01 I0406 09:48:04.437423 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_18258.caffemodel I0406 09:48:07.635370 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_18258.solverstate I0406 09:48:10.733316 5226 solver.cpp:330] Iteration 18258, Testing net (#0) I0406 09:48:10.733337 5226 net.cpp:676] Ignoring source layer train-data I0406 09:48:12.767627 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:48:15.403087 5226 solver.cpp:397] Test net output #0: accuracy = 0.173407 I0406 09:48:15.403117 5226 solver.cpp:397] Test net output #1: loss = 4.19531 (* 1 = 4.19531 loss) I0406 09:48:17.299959 5226 solver.cpp:218] Iteration 18264 (0.789752 iter/s, 15.1946s/12 iters), loss = 2.40677 I0406 09:48:17.300009 5226 solver.cpp:237] Train net output #0: loss = 2.40677 (* 1 = 2.40677 loss) I0406 09:48:17.300017 5226 sgd_solver.cpp:105] Iteration 18264, lr = 0.01 I0406 09:48:23.098910 5226 solver.cpp:218] Iteration 18276 (2.06938 iter/s, 5.79885s/12 iters), loss = 2.47545 I0406 09:48:23.098953 5226 solver.cpp:237] Train net output #0: loss = 2.47545 (* 1 = 2.47545 loss) I0406 09:48:23.098958 5226 sgd_solver.cpp:105] Iteration 18276, lr = 0.01 I0406 09:48:28.724840 5226 solver.cpp:218] Iteration 18288 (2.13302 iter/s, 5.62583s/12 iters), loss = 2.91271 I0406 09:48:28.724900 5226 solver.cpp:237] Train net output #0: loss = 2.91271 (* 1 = 2.91271 loss) I0406 09:48:28.724908 5226 sgd_solver.cpp:105] Iteration 18288, lr = 0.01 I0406 09:48:34.601153 5226 solver.cpp:218] Iteration 18300 (2.04213 iter/s, 5.87621s/12 iters), loss = 2.48929 I0406 09:48:34.601267 5226 solver.cpp:237] Train net output #0: loss = 2.48929 (* 1 = 2.48929 loss) I0406 09:48:34.601276 5226 sgd_solver.cpp:105] Iteration 18300, lr = 0.01 I0406 09:48:38.165952 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:48:39.846613 5226 solver.cpp:218] Iteration 18312 (2.28776 iter/s, 5.2453s/12 iters), loss = 2.75617 I0406 09:48:39.846655 5226 solver.cpp:237] Train net output #0: loss = 2.75617 (* 1 = 2.75617 loss) I0406 09:48:39.846662 5226 sgd_solver.cpp:105] Iteration 18312, lr = 0.01 I0406 09:48:45.696408 5226 solver.cpp:218] Iteration 18324 (2.05139 iter/s, 5.8497s/12 iters), loss = 2.58896 I0406 09:48:45.696458 5226 solver.cpp:237] Train net output #0: loss = 2.58896 (* 1 = 2.58896 loss) I0406 09:48:45.696468 5226 sgd_solver.cpp:105] Iteration 18324, lr = 0.01 I0406 09:48:51.285014 5226 solver.cpp:218] Iteration 18336 (2.14726 iter/s, 5.58851s/12 iters), loss = 2.33807 I0406 09:48:51.285054 5226 solver.cpp:237] Train net output #0: loss = 2.33807 (* 1 = 2.33807 loss) I0406 09:48:51.285059 5226 sgd_solver.cpp:105] Iteration 18336, lr = 0.01 I0406 09:48:56.978015 5226 solver.cpp:218] Iteration 18348 (2.10789 iter/s, 5.69291s/12 iters), loss = 2.7505 I0406 09:48:56.978067 5226 solver.cpp:237] Train net output #0: loss = 2.7505 (* 1 = 2.7505 loss) I0406 09:48:56.978075 5226 sgd_solver.cpp:105] Iteration 18348, lr = 0.01 I0406 09:49:01.800734 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_18360.caffemodel I0406 09:49:04.967018 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_18360.solverstate I0406 09:49:07.329947 5226 solver.cpp:330] Iteration 18360, Testing net (#0) I0406 09:49:07.329973 5226 net.cpp:676] Ignoring source layer train-data I0406 09:49:09.247212 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:49:12.172071 5226 solver.cpp:397] Test net output #0: accuracy = 0.147059 I0406 09:49:12.172102 5226 solver.cpp:397] Test net output #1: loss = 4.25444 (* 1 = 4.25444 loss) I0406 09:49:12.318850 5226 solver.cpp:218] Iteration 18360 (0.782234 iter/s, 15.3407s/12 iters), loss = 3.18243 I0406 09:49:12.318897 5226 solver.cpp:237] Train net output #0: loss = 3.18243 (* 1 = 3.18243 loss) I0406 09:49:12.318905 5226 sgd_solver.cpp:105] Iteration 18360, lr = 0.01 I0406 09:49:16.935120 5226 solver.cpp:218] Iteration 18372 (2.59955 iter/s, 4.61618s/12 iters), loss = 2.61853 I0406 09:49:16.935166 5226 solver.cpp:237] Train net output #0: loss = 2.61853 (* 1 = 2.61853 loss) I0406 09:49:16.935174 5226 sgd_solver.cpp:105] Iteration 18372, lr = 0.01 I0406 09:49:22.504688 5226 solver.cpp:218] Iteration 18384 (2.1546 iter/s, 5.56947s/12 iters), loss = 2.93539 I0406 09:49:22.504736 5226 solver.cpp:237] Train net output #0: loss = 2.93539 (* 1 = 2.93539 loss) I0406 09:49:22.504745 5226 sgd_solver.cpp:105] Iteration 18384, lr = 0.01 I0406 09:49:28.151232 5226 solver.cpp:218] Iteration 18396 (2.12523 iter/s, 5.64644s/12 iters), loss = 2.97436 I0406 09:49:28.151281 5226 solver.cpp:237] Train net output #0: loss = 2.97436 (* 1 = 2.97436 loss) I0406 09:49:28.151289 5226 sgd_solver.cpp:105] Iteration 18396, lr = 0.01 I0406 09:49:33.820475 5226 solver.cpp:218] Iteration 18408 (2.11674 iter/s, 5.6691s/12 iters), loss = 2.53136 I0406 09:49:33.820524 5226 solver.cpp:237] Train net output #0: loss = 2.53136 (* 1 = 2.53136 loss) I0406 09:49:33.820533 5226 sgd_solver.cpp:105] Iteration 18408, lr = 0.01 I0406 09:49:34.504357 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:49:39.332718 5226 solver.cpp:218] Iteration 18420 (2.17701 iter/s, 5.51215s/12 iters), loss = 3.38627 I0406 09:49:39.332814 5226 solver.cpp:237] Train net output #0: loss = 3.38627 (* 1 = 3.38627 loss) I0406 09:49:39.332821 5226 sgd_solver.cpp:105] Iteration 18420, lr = 0.01 I0406 09:49:45.152784 5226 solver.cpp:218] Iteration 18432 (2.06189 iter/s, 5.81991s/12 iters), loss = 2.90654 I0406 09:49:45.152834 5226 solver.cpp:237] Train net output #0: loss = 2.90654 (* 1 = 2.90654 loss) I0406 09:49:45.152842 5226 sgd_solver.cpp:105] Iteration 18432, lr = 0.01 I0406 09:49:51.120024 5226 solver.cpp:218] Iteration 18444 (2.01101 iter/s, 5.96714s/12 iters), loss = 2.96142 I0406 09:49:51.120061 5226 solver.cpp:237] Train net output #0: loss = 2.96142 (* 1 = 2.96142 loss) I0406 09:49:51.120067 5226 sgd_solver.cpp:105] Iteration 18444, lr = 0.01 I0406 09:49:56.531935 5226 solver.cpp:218] Iteration 18456 (2.21737 iter/s, 5.41182s/12 iters), loss = 2.9578 I0406 09:49:56.531991 5226 solver.cpp:237] Train net output #0: loss = 2.9578 (* 1 = 2.9578 loss) I0406 09:49:56.532001 5226 sgd_solver.cpp:105] Iteration 18456, lr = 0.01 I0406 09:49:58.817632 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_18462.caffemodel I0406 09:50:01.912143 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_18462.solverstate I0406 09:50:04.320299 5226 solver.cpp:330] Iteration 18462, Testing net (#0) I0406 09:50:04.320322 5226 net.cpp:676] Ignoring source layer train-data I0406 09:50:06.446846 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:50:09.369204 5226 solver.cpp:397] Test net output #0: accuracy = 0.14277 I0406 09:50:09.369320 5226 solver.cpp:397] Test net output #1: loss = 4.3483 (* 1 = 4.3483 loss) I0406 09:50:11.520042 5226 solver.cpp:218] Iteration 18468 (0.800643 iter/s, 14.988s/12 iters), loss = 2.92021 I0406 09:50:11.520078 5226 solver.cpp:237] Train net output #0: loss = 2.92021 (* 1 = 2.92021 loss) I0406 09:50:11.520084 5226 sgd_solver.cpp:105] Iteration 18468, lr = 0.01 I0406 09:50:17.148154 5226 solver.cpp:218] Iteration 18480 (2.1322 iter/s, 5.62799s/12 iters), loss = 2.09946 I0406 09:50:17.148192 5226 solver.cpp:237] Train net output #0: loss = 2.09946 (* 1 = 2.09946 loss) I0406 09:50:17.148198 5226 sgd_solver.cpp:105] Iteration 18480, lr = 0.01 I0406 09:50:22.918964 5226 solver.cpp:218] Iteration 18492 (2.07946 iter/s, 5.77072s/12 iters), loss = 2.91972 I0406 09:50:22.919000 5226 solver.cpp:237] Train net output #0: loss = 2.91972 (* 1 = 2.91972 loss) I0406 09:50:22.919006 5226 sgd_solver.cpp:105] Iteration 18492, lr = 0.01 I0406 09:50:28.322741 5226 solver.cpp:218] Iteration 18504 (2.22071 iter/s, 5.40369s/12 iters), loss = 2.6155 I0406 09:50:28.322782 5226 solver.cpp:237] Train net output #0: loss = 2.6155 (* 1 = 2.6155 loss) I0406 09:50:28.322788 5226 sgd_solver.cpp:105] Iteration 18504, lr = 0.01 I0406 09:50:31.406450 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:50:33.979977 5226 solver.cpp:218] Iteration 18516 (2.12121 iter/s, 5.65715s/12 iters), loss = 2.52292 I0406 09:50:33.980015 5226 solver.cpp:237] Train net output #0: loss = 2.52292 (* 1 = 2.52292 loss) I0406 09:50:33.980019 5226 sgd_solver.cpp:105] Iteration 18516, lr = 0.01 I0406 09:50:39.558295 5226 solver.cpp:218] Iteration 18528 (2.15122 iter/s, 5.57823s/12 iters), loss = 2.75023 I0406 09:50:39.558429 5226 solver.cpp:237] Train net output #0: loss = 2.75023 (* 1 = 2.75023 loss) I0406 09:50:39.558440 5226 sgd_solver.cpp:105] Iteration 18528, lr = 0.01 I0406 09:50:45.211774 5226 solver.cpp:218] Iteration 18540 (2.12266 iter/s, 5.65329s/12 iters), loss = 2.36854 I0406 09:50:45.211825 5226 solver.cpp:237] Train net output #0: loss = 2.36854 (* 1 = 2.36854 loss) I0406 09:50:45.211833 5226 sgd_solver.cpp:105] Iteration 18540, lr = 0.01 I0406 09:50:50.958618 5226 solver.cpp:218] Iteration 18552 (2.08814 iter/s, 5.74674s/12 iters), loss = 2.65339 I0406 09:50:50.958657 5226 solver.cpp:237] Train net output #0: loss = 2.65339 (* 1 = 2.65339 loss) I0406 09:50:50.958662 5226 sgd_solver.cpp:105] Iteration 18552, lr = 0.01 I0406 09:50:56.086372 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_18564.caffemodel I0406 09:50:59.246240 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_18564.solverstate I0406 09:51:01.598513 5226 solver.cpp:330] Iteration 18564, Testing net (#0) I0406 09:51:01.598533 5226 net.cpp:676] Ignoring source layer train-data I0406 09:51:03.717177 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:51:06.516773 5226 solver.cpp:397] Test net output #0: accuracy = 0.150735 I0406 09:51:06.516808 5226 solver.cpp:397] Test net output #1: loss = 4.2536 (* 1 = 4.2536 loss) I0406 09:51:06.660565 5226 solver.cpp:218] Iteration 18564 (0.764244 iter/s, 15.7018s/12 iters), loss = 2.60679 I0406 09:51:06.660615 5226 solver.cpp:237] Train net output #0: loss = 2.60679 (* 1 = 2.60679 loss) I0406 09:51:06.660622 5226 sgd_solver.cpp:105] Iteration 18564, lr = 0.01 I0406 09:51:11.341235 5226 solver.cpp:218] Iteration 18576 (2.56379 iter/s, 4.68058s/12 iters), loss = 2.42891 I0406 09:51:11.341364 5226 solver.cpp:237] Train net output #0: loss = 2.42891 (* 1 = 2.42891 loss) I0406 09:51:11.341373 5226 sgd_solver.cpp:105] Iteration 18576, lr = 0.01 I0406 09:51:16.923838 5226 solver.cpp:218] Iteration 18588 (2.1496 iter/s, 5.58243s/12 iters), loss = 2.71232 I0406 09:51:16.923894 5226 solver.cpp:237] Train net output #0: loss = 2.71232 (* 1 = 2.71232 loss) I0406 09:51:16.923902 5226 sgd_solver.cpp:105] Iteration 18588, lr = 0.01 I0406 09:51:22.183441 5226 solver.cpp:218] Iteration 18600 (2.28159 iter/s, 5.2595s/12 iters), loss = 2.8243 I0406 09:51:22.183481 5226 solver.cpp:237] Train net output #0: loss = 2.8243 (* 1 = 2.8243 loss) I0406 09:51:22.183487 5226 sgd_solver.cpp:105] Iteration 18600, lr = 0.01 I0406 09:51:27.247663 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:51:27.467072 5226 solver.cpp:218] Iteration 18612 (2.2712 iter/s, 5.28354s/12 iters), loss = 2.77226 I0406 09:51:27.467131 5226 solver.cpp:237] Train net output #0: loss = 2.77226 (* 1 = 2.77226 loss) I0406 09:51:27.467140 5226 sgd_solver.cpp:105] Iteration 18612, lr = 0.01 I0406 09:51:32.820556 5226 solver.cpp:218] Iteration 18624 (2.24157 iter/s, 5.35338s/12 iters), loss = 2.97261 I0406 09:51:32.820595 5226 solver.cpp:237] Train net output #0: loss = 2.97261 (* 1 = 2.97261 loss) I0406 09:51:32.820601 5226 sgd_solver.cpp:105] Iteration 18624, lr = 0.01 I0406 09:51:38.232952 5226 solver.cpp:218] Iteration 18636 (2.21717 iter/s, 5.4123s/12 iters), loss = 2.11176 I0406 09:51:38.233001 5226 solver.cpp:237] Train net output #0: loss = 2.11176 (* 1 = 2.11176 loss) I0406 09:51:38.233011 5226 sgd_solver.cpp:105] Iteration 18636, lr = 0.01 I0406 09:51:43.320188 5226 solver.cpp:218] Iteration 18648 (2.35889 iter/s, 5.08714s/12 iters), loss = 2.73249 I0406 09:51:43.320327 5226 solver.cpp:237] Train net output #0: loss = 2.73249 (* 1 = 2.73249 loss) I0406 09:51:43.320333 5226 sgd_solver.cpp:105] Iteration 18648, lr = 0.01 I0406 09:51:48.694944 5226 solver.cpp:218] Iteration 18660 (2.23274 iter/s, 5.37457s/12 iters), loss = 2.76798 I0406 09:51:48.694980 5226 solver.cpp:237] Train net output #0: loss = 2.76798 (* 1 = 2.76798 loss) I0406 09:51:48.694986 5226 sgd_solver.cpp:105] Iteration 18660, lr = 0.01 I0406 09:51:50.936535 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_18666.caffemodel I0406 09:51:53.880911 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_18666.solverstate I0406 09:51:56.192927 5226 solver.cpp:330] Iteration 18666, Testing net (#0) I0406 09:51:56.192951 5226 net.cpp:676] Ignoring source layer train-data I0406 09:51:57.929293 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:52:00.614500 5226 solver.cpp:397] Test net output #0: accuracy = 0.168505 I0406 09:52:00.614533 5226 solver.cpp:397] Test net output #1: loss = 4.29495 (* 1 = 4.29495 loss) I0406 09:52:02.608079 5226 solver.cpp:218] Iteration 18672 (0.862503 iter/s, 13.913s/12 iters), loss = 2.53058 I0406 09:52:02.608134 5226 solver.cpp:237] Train net output #0: loss = 2.53058 (* 1 = 2.53058 loss) I0406 09:52:02.608144 5226 sgd_solver.cpp:105] Iteration 18672, lr = 0.01 I0406 09:52:07.690367 5226 solver.cpp:218] Iteration 18684 (2.36119 iter/s, 5.08219s/12 iters), loss = 2.62814 I0406 09:52:07.690407 5226 solver.cpp:237] Train net output #0: loss = 2.62814 (* 1 = 2.62814 loss) I0406 09:52:07.690412 5226 sgd_solver.cpp:105] Iteration 18684, lr = 0.01 I0406 09:52:12.886394 5226 solver.cpp:218] Iteration 18696 (2.3095 iter/s, 5.19594s/12 iters), loss = 2.3242 I0406 09:52:12.886431 5226 solver.cpp:237] Train net output #0: loss = 2.3242 (* 1 = 2.3242 loss) I0406 09:52:12.886436 5226 sgd_solver.cpp:105] Iteration 18696, lr = 0.01 I0406 09:52:18.100653 5226 solver.cpp:218] Iteration 18708 (2.30142 iter/s, 5.21417s/12 iters), loss = 2.65449 I0406 09:52:18.100817 5226 solver.cpp:237] Train net output #0: loss = 2.65449 (* 1 = 2.65449 loss) I0406 09:52:18.100826 5226 sgd_solver.cpp:105] Iteration 18708, lr = 0.01 I0406 09:52:20.133059 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:52:23.469902 5226 solver.cpp:218] Iteration 18720 (2.23503 iter/s, 5.36905s/12 iters), loss = 2.86272 I0406 09:52:23.469935 5226 solver.cpp:237] Train net output #0: loss = 2.86272 (* 1 = 2.86272 loss) I0406 09:52:23.469941 5226 sgd_solver.cpp:105] Iteration 18720, lr = 0.01 I0406 09:52:23.835044 5226 blocking_queue.cpp:49] Waiting for data I0406 09:52:28.891965 5226 solver.cpp:218] Iteration 18732 (2.21321 iter/s, 5.42198s/12 iters), loss = 3.0482 I0406 09:52:28.892014 5226 solver.cpp:237] Train net output #0: loss = 3.0482 (* 1 = 3.0482 loss) I0406 09:52:28.892024 5226 sgd_solver.cpp:105] Iteration 18732, lr = 0.01 I0406 09:52:34.206876 5226 solver.cpp:218] Iteration 18744 (2.25784 iter/s, 5.31482s/12 iters), loss = 2.91395 I0406 09:52:34.206933 5226 solver.cpp:237] Train net output #0: loss = 2.91395 (* 1 = 2.91395 loss) I0406 09:52:34.206941 5226 sgd_solver.cpp:105] Iteration 18744, lr = 0.01 I0406 09:52:39.583050 5226 solver.cpp:218] Iteration 18756 (2.23211 iter/s, 5.37607s/12 iters), loss = 2.73775 I0406 09:52:39.583086 5226 solver.cpp:237] Train net output #0: loss = 2.73775 (* 1 = 2.73775 loss) I0406 09:52:39.583091 5226 sgd_solver.cpp:105] Iteration 18756, lr = 0.01 I0406 09:52:44.217043 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_18768.caffemodel I0406 09:52:47.200744 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_18768.solverstate I0406 09:52:49.504376 5226 solver.cpp:330] Iteration 18768, Testing net (#0) I0406 09:52:49.504472 5226 net.cpp:676] Ignoring source layer train-data I0406 09:52:51.314322 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:52:54.057606 5226 solver.cpp:397] Test net output #0: accuracy = 0.160539 I0406 09:52:54.057642 5226 solver.cpp:397] Test net output #1: loss = 4.18962 (* 1 = 4.18962 loss) I0406 09:52:54.198197 5226 solver.cpp:218] Iteration 18768 (0.821074 iter/s, 14.615s/12 iters), loss = 2.57012 I0406 09:52:54.198249 5226 solver.cpp:237] Train net output #0: loss = 2.57012 (* 1 = 2.57012 loss) I0406 09:52:54.198257 5226 sgd_solver.cpp:105] Iteration 18768, lr = 0.01 I0406 09:52:58.684337 5226 solver.cpp:218] Iteration 18780 (2.67496 iter/s, 4.48604s/12 iters), loss = 3.12243 I0406 09:52:58.684391 5226 solver.cpp:237] Train net output #0: loss = 3.12243 (* 1 = 3.12243 loss) I0406 09:52:58.684401 5226 sgd_solver.cpp:105] Iteration 18780, lr = 0.01 I0406 09:53:03.669354 5226 solver.cpp:218] Iteration 18792 (2.40726 iter/s, 4.98492s/12 iters), loss = 3.04408 I0406 09:53:03.669395 5226 solver.cpp:237] Train net output #0: loss = 3.04408 (* 1 = 3.04408 loss) I0406 09:53:03.669400 5226 sgd_solver.cpp:105] Iteration 18792, lr = 0.01 I0406 09:53:08.985697 5226 solver.cpp:218] Iteration 18804 (2.25723 iter/s, 5.31625s/12 iters), loss = 3.1562 I0406 09:53:08.985735 5226 solver.cpp:237] Train net output #0: loss = 3.1562 (* 1 = 3.1562 loss) I0406 09:53:08.985740 5226 sgd_solver.cpp:105] Iteration 18804, lr = 0.01 I0406 09:53:13.379719 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:53:14.256330 5226 solver.cpp:218] Iteration 18816 (2.2768 iter/s, 5.27055s/12 iters), loss = 2.61566 I0406 09:53:14.256367 5226 solver.cpp:237] Train net output #0: loss = 2.61566 (* 1 = 2.61566 loss) I0406 09:53:14.256372 5226 sgd_solver.cpp:105] Iteration 18816, lr = 0.01 I0406 09:53:19.445952 5226 solver.cpp:218] Iteration 18828 (2.31235 iter/s, 5.18954s/12 iters), loss = 3.05859 I0406 09:53:19.445993 5226 solver.cpp:237] Train net output #0: loss = 3.05859 (* 1 = 3.05859 loss) I0406 09:53:19.445999 5226 sgd_solver.cpp:105] Iteration 18828, lr = 0.01 I0406 09:53:24.810252 5226 solver.cpp:218] Iteration 18840 (2.23705 iter/s, 5.36421s/12 iters), loss = 2.46742 I0406 09:53:24.810375 5226 solver.cpp:237] Train net output #0: loss = 2.46742 (* 1 = 2.46742 loss) I0406 09:53:24.810384 5226 sgd_solver.cpp:105] Iteration 18840, lr = 0.01 I0406 09:53:30.164876 5226 solver.cpp:218] Iteration 18852 (2.24112 iter/s, 5.35446s/12 iters), loss = 2.69689 I0406 09:53:30.164934 5226 solver.cpp:237] Train net output #0: loss = 2.69689 (* 1 = 2.69689 loss) I0406 09:53:30.164942 5226 sgd_solver.cpp:105] Iteration 18852, lr = 0.01 I0406 09:53:35.764940 5226 solver.cpp:218] Iteration 18864 (2.14287 iter/s, 5.59996s/12 iters), loss = 2.58727 I0406 09:53:35.764981 5226 solver.cpp:237] Train net output #0: loss = 2.58727 (* 1 = 2.58727 loss) I0406 09:53:35.764986 5226 sgd_solver.cpp:105] Iteration 18864, lr = 0.01 I0406 09:53:37.993623 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_18870.caffemodel I0406 09:53:41.036381 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_18870.solverstate I0406 09:53:43.337386 5226 solver.cpp:330] Iteration 18870, Testing net (#0) I0406 09:53:43.337405 5226 net.cpp:676] Ignoring source layer train-data I0406 09:53:44.917209 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:53:47.726856 5226 solver.cpp:397] Test net output #0: accuracy = 0.160539 I0406 09:53:47.726883 5226 solver.cpp:397] Test net output #1: loss = 4.28674 (* 1 = 4.28674 loss) I0406 09:53:49.493155 5226 solver.cpp:218] Iteration 18876 (0.874121 iter/s, 13.7281s/12 iters), loss = 2.61103 I0406 09:53:49.493208 5226 solver.cpp:237] Train net output #0: loss = 2.61103 (* 1 = 2.61103 loss) I0406 09:53:49.493216 5226 sgd_solver.cpp:105] Iteration 18876, lr = 0.01 I0406 09:53:54.832093 5226 solver.cpp:218] Iteration 18888 (2.24768 iter/s, 5.33884s/12 iters), loss = 2.96768 I0406 09:53:54.832211 5226 solver.cpp:237] Train net output #0: loss = 2.96768 (* 1 = 2.96768 loss) I0406 09:53:54.832217 5226 sgd_solver.cpp:105] Iteration 18888, lr = 0.01 I0406 09:53:59.930821 5226 solver.cpp:218] Iteration 18900 (2.3536 iter/s, 5.09856s/12 iters), loss = 3.17591 I0406 09:53:59.930861 5226 solver.cpp:237] Train net output #0: loss = 3.17591 (* 1 = 3.17591 loss) I0406 09:53:59.930866 5226 sgd_solver.cpp:105] Iteration 18900, lr = 0.01 I0406 09:54:05.211365 5226 solver.cpp:218] Iteration 18912 (2.27253 iter/s, 5.28046s/12 iters), loss = 2.8022 I0406 09:54:05.211405 5226 solver.cpp:237] Train net output #0: loss = 2.8022 (* 1 = 2.8022 loss) I0406 09:54:05.211410 5226 sgd_solver.cpp:105] Iteration 18912, lr = 0.01 I0406 09:54:06.656607 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:54:10.640969 5226 solver.cpp:218] Iteration 18924 (2.21014 iter/s, 5.42951s/12 iters), loss = 2.35531 I0406 09:54:10.641024 5226 solver.cpp:237] Train net output #0: loss = 2.35531 (* 1 = 2.35531 loss) I0406 09:54:10.641033 5226 sgd_solver.cpp:105] Iteration 18924, lr = 0.01 I0406 09:54:15.817670 5226 solver.cpp:218] Iteration 18936 (2.31812 iter/s, 5.17661s/12 iters), loss = 2.43513 I0406 09:54:15.817715 5226 solver.cpp:237] Train net output #0: loss = 2.43513 (* 1 = 2.43513 loss) I0406 09:54:15.817724 5226 sgd_solver.cpp:105] Iteration 18936, lr = 0.01 I0406 09:54:21.219298 5226 solver.cpp:218] Iteration 18948 (2.22159 iter/s, 5.40154s/12 iters), loss = 2.18946 I0406 09:54:21.219345 5226 solver.cpp:237] Train net output #0: loss = 2.18946 (* 1 = 2.18946 loss) I0406 09:54:21.219352 5226 sgd_solver.cpp:105] Iteration 18948, lr = 0.01 I0406 09:54:26.653434 5226 solver.cpp:218] Iteration 18960 (2.2083 iter/s, 5.43404s/12 iters), loss = 2.9931 I0406 09:54:26.653564 5226 solver.cpp:237] Train net output #0: loss = 2.9931 (* 1 = 2.9931 loss) I0406 09:54:26.653575 5226 sgd_solver.cpp:105] Iteration 18960, lr = 0.01 I0406 09:54:31.512817 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_18972.caffemodel I0406 09:54:34.557161 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_18972.solverstate I0406 09:54:36.871708 5226 solver.cpp:330] Iteration 18972, Testing net (#0) I0406 09:54:36.871728 5226 net.cpp:676] Ignoring source layer train-data I0406 09:54:38.538091 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:54:41.308542 5226 solver.cpp:397] Test net output #0: accuracy = 0.140319 I0406 09:54:41.308574 5226 solver.cpp:397] Test net output #1: loss = 4.34298 (* 1 = 4.34298 loss) I0406 09:54:41.449098 5226 solver.cpp:218] Iteration 18972 (0.811061 iter/s, 14.7954s/12 iters), loss = 2.47181 I0406 09:54:41.449138 5226 solver.cpp:237] Train net output #0: loss = 2.47181 (* 1 = 2.47181 loss) I0406 09:54:41.449144 5226 sgd_solver.cpp:105] Iteration 18972, lr = 0.01 I0406 09:54:45.870874 5226 solver.cpp:218] Iteration 18984 (2.7139 iter/s, 4.42169s/12 iters), loss = 2.4347 I0406 09:54:45.870921 5226 solver.cpp:237] Train net output #0: loss = 2.4347 (* 1 = 2.4347 loss) I0406 09:54:45.870929 5226 sgd_solver.cpp:105] Iteration 18984, lr = 0.01 I0406 09:54:51.296736 5226 solver.cpp:218] Iteration 18996 (2.21167 iter/s, 5.42576s/12 iters), loss = 2.43389 I0406 09:54:51.296795 5226 solver.cpp:237] Train net output #0: loss = 2.43389 (* 1 = 2.43389 loss) I0406 09:54:51.296805 5226 sgd_solver.cpp:105] Iteration 18996, lr = 0.01 I0406 09:54:56.524983 5226 solver.cpp:218] Iteration 19008 (2.29527 iter/s, 5.22815s/12 iters), loss = 2.51239 I0406 09:54:56.525022 5226 solver.cpp:237] Train net output #0: loss = 2.51239 (* 1 = 2.51239 loss) I0406 09:54:56.525027 5226 sgd_solver.cpp:105] Iteration 19008, lr = 0.01 I0406 09:55:00.243628 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:55:01.752319 5226 solver.cpp:218] Iteration 19020 (2.29566 iter/s, 5.22725s/12 iters), loss = 2.84622 I0406 09:55:01.752362 5226 solver.cpp:237] Train net output #0: loss = 2.84622 (* 1 = 2.84622 loss) I0406 09:55:01.752368 5226 sgd_solver.cpp:105] Iteration 19020, lr = 0.01 I0406 09:55:07.149793 5226 solver.cpp:218] Iteration 19032 (2.2233 iter/s, 5.39738s/12 iters), loss = 2.81705 I0406 09:55:07.149832 5226 solver.cpp:237] Train net output #0: loss = 2.81705 (* 1 = 2.81705 loss) I0406 09:55:07.149837 5226 sgd_solver.cpp:105] Iteration 19032, lr = 0.01 I0406 09:55:12.421520 5226 solver.cpp:218] Iteration 19044 (2.27633 iter/s, 5.27164s/12 iters), loss = 2.51426 I0406 09:55:12.421567 5226 solver.cpp:237] Train net output #0: loss = 2.51426 (* 1 = 2.51426 loss) I0406 09:55:12.421576 5226 sgd_solver.cpp:105] Iteration 19044, lr = 0.01 I0406 09:55:17.797022 5226 solver.cpp:218] Iteration 19056 (2.2324 iter/s, 5.37538s/12 iters), loss = 2.23711 I0406 09:55:17.797060 5226 solver.cpp:237] Train net output #0: loss = 2.23711 (* 1 = 2.23711 loss) I0406 09:55:17.797065 5226 sgd_solver.cpp:105] Iteration 19056, lr = 0.01 I0406 09:55:23.108865 5226 solver.cpp:218] Iteration 19068 (2.25914 iter/s, 5.31175s/12 iters), loss = 2.54198 I0406 09:55:23.108924 5226 solver.cpp:237] Train net output #0: loss = 2.54198 (* 1 = 2.54198 loss) I0406 09:55:23.108932 5226 sgd_solver.cpp:105] Iteration 19068, lr = 0.01 I0406 09:55:25.266986 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_19074.caffemodel I0406 09:55:28.334154 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_19074.solverstate I0406 09:55:30.646607 5226 solver.cpp:330] Iteration 19074, Testing net (#0) I0406 09:55:30.646693 5226 net.cpp:676] Ignoring source layer train-data I0406 09:55:32.183472 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:55:35.186436 5226 solver.cpp:397] Test net output #0: accuracy = 0.181985 I0406 09:55:35.186471 5226 solver.cpp:397] Test net output #1: loss = 4.14378 (* 1 = 4.14378 loss) I0406 09:55:36.995689 5226 solver.cpp:218] Iteration 19080 (0.864138 iter/s, 13.8867s/12 iters), loss = 2.66738 I0406 09:55:36.995740 5226 solver.cpp:237] Train net output #0: loss = 2.66738 (* 1 = 2.66738 loss) I0406 09:55:36.995749 5226 sgd_solver.cpp:105] Iteration 19080, lr = 0.01 I0406 09:55:46.395162 5226 solver.cpp:218] Iteration 19092 (1.27669 iter/s, 9.39934s/12 iters), loss = 2.26242 I0406 09:55:46.395226 5226 solver.cpp:237] Train net output #0: loss = 2.26242 (* 1 = 2.26242 loss) I0406 09:55:46.395234 5226 sgd_solver.cpp:105] Iteration 19092, lr = 0.01 I0406 09:55:54.763990 5226 solver.cpp:218] Iteration 19104 (1.43391 iter/s, 8.3687s/12 iters), loss = 2.51026 I0406 09:55:54.764039 5226 solver.cpp:237] Train net output #0: loss = 2.51026 (* 1 = 2.51026 loss) I0406 09:55:54.764047 5226 sgd_solver.cpp:105] Iteration 19104, lr = 0.01 I0406 09:56:03.312409 5226 solver.cpp:218] Iteration 19116 (1.40379 iter/s, 8.54829s/12 iters), loss = 2.756 I0406 09:56:03.324959 5226 solver.cpp:237] Train net output #0: loss = 2.756 (* 1 = 2.756 loss) I0406 09:56:03.324971 5226 sgd_solver.cpp:105] Iteration 19116, lr = 0.01 I0406 09:56:04.214323 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:56:10.854852 5226 solver.cpp:218] Iteration 19128 (1.59366 iter/s, 7.52984s/12 iters), loss = 2.74721 I0406 09:56:10.854902 5226 solver.cpp:237] Train net output #0: loss = 2.74721 (* 1 = 2.74721 loss) I0406 09:56:10.854909 5226 sgd_solver.cpp:105] Iteration 19128, lr = 0.01 I0406 09:56:16.023772 5226 solver.cpp:218] Iteration 19140 (2.32161 iter/s, 5.16882s/12 iters), loss = 3.16032 I0406 09:56:16.023808 5226 solver.cpp:237] Train net output #0: loss = 3.16032 (* 1 = 3.16032 loss) I0406 09:56:16.023813 5226 sgd_solver.cpp:105] Iteration 19140, lr = 0.01 I0406 09:56:21.082613 5226 solver.cpp:218] Iteration 19152 (2.37213 iter/s, 5.05875s/12 iters), loss = 2.77754 I0406 09:56:21.082662 5226 solver.cpp:237] Train net output #0: loss = 2.77754 (* 1 = 2.77754 loss) I0406 09:56:21.082670 5226 sgd_solver.cpp:105] Iteration 19152, lr = 0.01 I0406 09:56:26.236485 5226 solver.cpp:218] Iteration 19164 (2.32839 iter/s, 5.15378s/12 iters), loss = 2.4267 I0406 09:56:26.236526 5226 solver.cpp:237] Train net output #0: loss = 2.4267 (* 1 = 2.4267 loss) I0406 09:56:26.236531 5226 sgd_solver.cpp:105] Iteration 19164, lr = 0.01 I0406 09:56:30.834393 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_19176.caffemodel I0406 09:56:33.869146 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_19176.solverstate I0406 09:56:36.187420 5226 solver.cpp:330] Iteration 19176, Testing net (#0) I0406 09:56:36.187439 5226 net.cpp:676] Ignoring source layer train-data I0406 09:56:37.823941 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:56:40.828379 5226 solver.cpp:397] Test net output #0: accuracy = 0.154412 I0406 09:56:40.828408 5226 solver.cpp:397] Test net output #1: loss = 4.27568 (* 1 = 4.27568 loss) I0406 09:56:40.970129 5226 solver.cpp:218] Iteration 19176 (0.81447 iter/s, 14.7335s/12 iters), loss = 2.65843 I0406 09:56:40.971714 5226 solver.cpp:237] Train net output #0: loss = 2.65843 (* 1 = 2.65843 loss) I0406 09:56:40.971724 5226 sgd_solver.cpp:105] Iteration 19176, lr = 0.01 I0406 09:56:45.339912 5226 solver.cpp:218] Iteration 19188 (2.74715 iter/s, 4.36816s/12 iters), loss = 2.61225 I0406 09:56:45.339965 5226 solver.cpp:237] Train net output #0: loss = 2.61225 (* 1 = 2.61225 loss) I0406 09:56:45.339973 5226 sgd_solver.cpp:105] Iteration 19188, lr = 0.01 I0406 09:56:50.495851 5226 solver.cpp:218] Iteration 19200 (2.32746 iter/s, 5.15584s/12 iters), loss = 2.87117 I0406 09:56:50.495898 5226 solver.cpp:237] Train net output #0: loss = 2.87117 (* 1 = 2.87117 loss) I0406 09:56:50.495906 5226 sgd_solver.cpp:105] Iteration 19200, lr = 0.01 I0406 09:56:55.760635 5226 solver.cpp:218] Iteration 19212 (2.27934 iter/s, 5.26469s/12 iters), loss = 2.69491 I0406 09:56:55.760676 5226 solver.cpp:237] Train net output #0: loss = 2.69491 (* 1 = 2.69491 loss) I0406 09:56:55.760682 5226 sgd_solver.cpp:105] Iteration 19212, lr = 0.01 I0406 09:56:58.665115 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:57:01.137848 5226 solver.cpp:218] Iteration 19224 (2.23168 iter/s, 5.37712s/12 iters), loss = 2.80032 I0406 09:57:01.137903 5226 solver.cpp:237] Train net output #0: loss = 2.80032 (* 1 = 2.80032 loss) I0406 09:57:01.137912 5226 sgd_solver.cpp:105] Iteration 19224, lr = 0.01 I0406 09:57:06.427059 5226 solver.cpp:218] Iteration 19236 (2.26881 iter/s, 5.28911s/12 iters), loss = 2.46941 I0406 09:57:06.427202 5226 solver.cpp:237] Train net output #0: loss = 2.46941 (* 1 = 2.46941 loss) I0406 09:57:06.427212 5226 sgd_solver.cpp:105] Iteration 19236, lr = 0.01 I0406 09:57:11.696939 5226 solver.cpp:218] Iteration 19248 (2.27717 iter/s, 5.26969s/12 iters), loss = 2.76649 I0406 09:57:11.696977 5226 solver.cpp:237] Train net output #0: loss = 2.76649 (* 1 = 2.76649 loss) I0406 09:57:11.696983 5226 sgd_solver.cpp:105] Iteration 19248, lr = 0.01 I0406 09:57:17.008790 5226 solver.cpp:218] Iteration 19260 (2.25914 iter/s, 5.31176s/12 iters), loss = 2.79233 I0406 09:57:17.008836 5226 solver.cpp:237] Train net output #0: loss = 2.79233 (* 1 = 2.79233 loss) I0406 09:57:17.008844 5226 sgd_solver.cpp:105] Iteration 19260, lr = 0.01 I0406 09:57:22.378641 5226 solver.cpp:218] Iteration 19272 (2.23474 iter/s, 5.36976s/12 iters), loss = 2.12638 I0406 09:57:22.378679 5226 solver.cpp:237] Train net output #0: loss = 2.12638 (* 1 = 2.12638 loss) I0406 09:57:22.378684 5226 sgd_solver.cpp:105] Iteration 19272, lr = 0.01 I0406 09:57:24.549031 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_19278.caffemodel I0406 09:57:27.734087 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_19278.solverstate I0406 09:57:30.550173 5226 solver.cpp:330] Iteration 19278, Testing net (#0) I0406 09:57:30.550197 5226 net.cpp:676] Ignoring source layer train-data I0406 09:57:32.041424 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:57:35.055517 5226 solver.cpp:397] Test net output #0: accuracy = 0.147672 I0406 09:57:35.055554 5226 solver.cpp:397] Test net output #1: loss = 4.29293 (* 1 = 4.29293 loss) I0406 09:57:36.900943 5226 solver.cpp:218] Iteration 19284 (0.826323 iter/s, 14.5222s/12 iters), loss = 2.55369 I0406 09:57:36.901058 5226 solver.cpp:237] Train net output #0: loss = 2.55369 (* 1 = 2.55369 loss) I0406 09:57:36.901064 5226 sgd_solver.cpp:105] Iteration 19284, lr = 0.01 I0406 09:57:42.110467 5226 solver.cpp:218] Iteration 19296 (2.30354 iter/s, 5.20936s/12 iters), loss = 2.61398 I0406 09:57:42.110505 5226 solver.cpp:237] Train net output #0: loss = 2.61398 (* 1 = 2.61398 loss) I0406 09:57:42.110510 5226 sgd_solver.cpp:105] Iteration 19296, lr = 0.01 I0406 09:57:47.332265 5226 solver.cpp:218] Iteration 19308 (2.2981 iter/s, 5.22171s/12 iters), loss = 3.00053 I0406 09:57:47.332305 5226 solver.cpp:237] Train net output #0: loss = 3.00053 (* 1 = 3.00053 loss) I0406 09:57:47.332310 5226 sgd_solver.cpp:105] Iteration 19308, lr = 0.01 I0406 09:57:52.428169 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:57:52.621233 5226 solver.cpp:218] Iteration 19320 (2.26891 iter/s, 5.28888s/12 iters), loss = 2.60861 I0406 09:57:52.621273 5226 solver.cpp:237] Train net output #0: loss = 2.60861 (* 1 = 2.60861 loss) I0406 09:57:52.621277 5226 sgd_solver.cpp:105] Iteration 19320, lr = 0.01 I0406 09:57:58.161604 5226 solver.cpp:218] Iteration 19332 (2.16596 iter/s, 5.54028s/12 iters), loss = 2.86653 I0406 09:57:58.161660 5226 solver.cpp:237] Train net output #0: loss = 2.86653 (* 1 = 2.86653 loss) I0406 09:57:58.161669 5226 sgd_solver.cpp:105] Iteration 19332, lr = 0.01 I0406 09:58:03.448232 5226 solver.cpp:218] Iteration 19344 (2.26992 iter/s, 5.28653s/12 iters), loss = 3.0863 I0406 09:58:03.448288 5226 solver.cpp:237] Train net output #0: loss = 3.0863 (* 1 = 3.0863 loss) I0406 09:58:03.448297 5226 sgd_solver.cpp:105] Iteration 19344, lr = 0.01 I0406 09:58:08.817152 5226 solver.cpp:218] Iteration 19356 (2.23513 iter/s, 5.36882s/12 iters), loss = 3.54207 I0406 09:58:08.817243 5226 solver.cpp:237] Train net output #0: loss = 3.54207 (* 1 = 3.54207 loss) I0406 09:58:08.817250 5226 sgd_solver.cpp:105] Iteration 19356, lr = 0.01 I0406 09:58:14.051836 5226 solver.cpp:218] Iteration 19368 (2.29248 iter/s, 5.2345s/12 iters), loss = 3.14982 I0406 09:58:14.051882 5226 solver.cpp:237] Train net output #0: loss = 3.14982 (* 1 = 3.14982 loss) I0406 09:58:14.051889 5226 sgd_solver.cpp:105] Iteration 19368, lr = 0.01 I0406 09:58:18.897032 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_19380.caffemodel I0406 09:58:21.959190 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_19380.solverstate I0406 09:58:24.272936 5226 solver.cpp:330] Iteration 19380, Testing net (#0) I0406 09:58:24.272958 5226 net.cpp:676] Ignoring source layer train-data I0406 09:58:25.640415 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:58:28.567831 5226 solver.cpp:397] Test net output #0: accuracy = 0.162377 I0406 09:58:28.567864 5226 solver.cpp:397] Test net output #1: loss = 4.26514 (* 1 = 4.26514 loss) I0406 09:58:28.708443 5226 solver.cpp:218] Iteration 19380 (0.818751 iter/s, 14.6565s/12 iters), loss = 2.70815 I0406 09:58:28.708480 5226 solver.cpp:237] Train net output #0: loss = 2.70815 (* 1 = 2.70815 loss) I0406 09:58:28.708487 5226 sgd_solver.cpp:105] Iteration 19380, lr = 0.01 I0406 09:58:33.107223 5226 solver.cpp:218] Iteration 19392 (2.72808 iter/s, 4.3987s/12 iters), loss = 2.92216 I0406 09:58:33.107264 5226 solver.cpp:237] Train net output #0: loss = 2.92216 (* 1 = 2.92216 loss) I0406 09:58:33.107270 5226 sgd_solver.cpp:105] Iteration 19392, lr = 0.01 I0406 09:58:38.612675 5226 solver.cpp:218] Iteration 19404 (2.17969 iter/s, 5.50536s/12 iters), loss = 2.81593 I0406 09:58:38.612725 5226 solver.cpp:237] Train net output #0: loss = 2.81593 (* 1 = 2.81593 loss) I0406 09:58:38.612733 5226 sgd_solver.cpp:105] Iteration 19404, lr = 0.01 I0406 09:58:39.407426 5226 blocking_queue.cpp:49] Waiting for data I0406 09:58:43.780370 5226 solver.cpp:218] Iteration 19416 (2.32216 iter/s, 5.1676s/12 iters), loss = 2.86767 I0406 09:58:43.780407 5226 solver.cpp:237] Train net output #0: loss = 2.86767 (* 1 = 2.86767 loss) I0406 09:58:43.780412 5226 sgd_solver.cpp:105] Iteration 19416, lr = 0.01 I0406 09:58:45.892205 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:58:49.135965 5226 solver.cpp:218] Iteration 19428 (2.24068 iter/s, 5.35551s/12 iters), loss = 2.91555 I0406 09:58:49.136013 5226 solver.cpp:237] Train net output #0: loss = 2.91555 (* 1 = 2.91555 loss) I0406 09:58:49.136020 5226 sgd_solver.cpp:105] Iteration 19428, lr = 0.01 I0406 09:58:54.453687 5226 solver.cpp:218] Iteration 19440 (2.25665 iter/s, 5.31763s/12 iters), loss = 2.48498 I0406 09:58:54.453729 5226 solver.cpp:237] Train net output #0: loss = 2.48498 (* 1 = 2.48498 loss) I0406 09:58:54.453735 5226 sgd_solver.cpp:105] Iteration 19440, lr = 0.01 I0406 09:58:59.676683 5226 solver.cpp:218] Iteration 19452 (2.29757 iter/s, 5.22291s/12 iters), loss = 2.7795 I0406 09:58:59.676719 5226 solver.cpp:237] Train net output #0: loss = 2.7795 (* 1 = 2.7795 loss) I0406 09:58:59.676725 5226 sgd_solver.cpp:105] Iteration 19452, lr = 0.01 I0406 09:59:05.138038 5226 solver.cpp:218] Iteration 19464 (2.19729 iter/s, 5.46127s/12 iters), loss = 3.36903 I0406 09:59:05.138077 5226 solver.cpp:237] Train net output #0: loss = 3.36903 (* 1 = 3.36903 loss) I0406 09:59:05.138082 5226 sgd_solver.cpp:105] Iteration 19464, lr = 0.01 I0406 09:59:10.583138 5226 solver.cpp:218] Iteration 19476 (2.20385 iter/s, 5.44501s/12 iters), loss = 2.70699 I0406 09:59:10.583237 5226 solver.cpp:237] Train net output #0: loss = 2.70699 (* 1 = 2.70699 loss) I0406 09:59:10.583243 5226 sgd_solver.cpp:105] Iteration 19476, lr = 0.01 I0406 09:59:12.816386 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_19482.caffemodel I0406 09:59:17.359834 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_19482.solverstate I0406 09:59:21.091992 5226 solver.cpp:330] Iteration 19482, Testing net (#0) I0406 09:59:21.092016 5226 net.cpp:676] Ignoring source layer train-data I0406 09:59:22.522255 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:59:25.505995 5226 solver.cpp:397] Test net output #0: accuracy = 0.145833 I0406 09:59:25.506029 5226 solver.cpp:397] Test net output #1: loss = 4.29541 (* 1 = 4.29541 loss) I0406 09:59:27.425757 5226 solver.cpp:218] Iteration 19488 (0.712487 iter/s, 16.8424s/12 iters), loss = 2.67586 I0406 09:59:27.425809 5226 solver.cpp:237] Train net output #0: loss = 2.67586 (* 1 = 2.67586 loss) I0406 09:59:27.425819 5226 sgd_solver.cpp:105] Iteration 19488, lr = 0.01 I0406 09:59:32.769673 5226 solver.cpp:218] Iteration 19500 (2.24559 iter/s, 5.34382s/12 iters), loss = 2.838 I0406 09:59:32.769714 5226 solver.cpp:237] Train net output #0: loss = 2.838 (* 1 = 2.838 loss) I0406 09:59:32.769721 5226 sgd_solver.cpp:105] Iteration 19500, lr = 0.01 I0406 09:59:38.277453 5226 solver.cpp:218] Iteration 19512 (2.17877 iter/s, 5.50769s/12 iters), loss = 2.52217 I0406 09:59:38.277494 5226 solver.cpp:237] Train net output #0: loss = 2.52217 (* 1 = 2.52217 loss) I0406 09:59:38.277500 5226 sgd_solver.cpp:105] Iteration 19512, lr = 0.01 I0406 09:59:42.738787 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 09:59:43.664301 5226 solver.cpp:218] Iteration 19524 (2.22769 iter/s, 5.38676s/12 iters), loss = 2.15964 I0406 09:59:43.664353 5226 solver.cpp:237] Train net output #0: loss = 2.15964 (* 1 = 2.15964 loss) I0406 09:59:43.664363 5226 sgd_solver.cpp:105] Iteration 19524, lr = 0.01 I0406 09:59:49.126029 5226 solver.cpp:218] Iteration 19536 (2.19715 iter/s, 5.46162s/12 iters), loss = 2.33063 I0406 09:59:49.126098 5226 solver.cpp:237] Train net output #0: loss = 2.33063 (* 1 = 2.33063 loss) I0406 09:59:49.126109 5226 sgd_solver.cpp:105] Iteration 19536, lr = 0.01 I0406 09:59:54.579237 5226 solver.cpp:218] Iteration 19548 (2.20059 iter/s, 5.45309s/12 iters), loss = 2.25702 I0406 09:59:54.579290 5226 solver.cpp:237] Train net output #0: loss = 2.25702 (* 1 = 2.25702 loss) I0406 09:59:54.579298 5226 sgd_solver.cpp:105] Iteration 19548, lr = 0.01 I0406 09:59:59.938196 5226 solver.cpp:218] Iteration 19560 (2.23928 iter/s, 5.35886s/12 iters), loss = 2.56183 I0406 09:59:59.938241 5226 solver.cpp:237] Train net output #0: loss = 2.56183 (* 1 = 2.56183 loss) I0406 09:59:59.938249 5226 sgd_solver.cpp:105] Iteration 19560, lr = 0.01 I0406 10:00:05.521911 5226 solver.cpp:218] Iteration 19572 (2.14914 iter/s, 5.58362s/12 iters), loss = 2.52675 I0406 10:00:05.521948 5226 solver.cpp:237] Train net output #0: loss = 2.52675 (* 1 = 2.52675 loss) I0406 10:00:05.521955 5226 sgd_solver.cpp:105] Iteration 19572, lr = 0.01 I0406 10:00:10.219933 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_19584.caffemodel I0406 10:00:13.370455 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_19584.solverstate I0406 10:00:16.436864 5226 solver.cpp:330] Iteration 19584, Testing net (#0) I0406 10:00:16.436890 5226 net.cpp:676] Ignoring source layer train-data I0406 10:00:17.859496 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 10:00:21.042773 5226 solver.cpp:397] Test net output #0: accuracy = 0.180147 I0406 10:00:21.042809 5226 solver.cpp:397] Test net output #1: loss = 4.1592 (* 1 = 4.1592 loss) I0406 10:00:21.187947 5226 solver.cpp:218] Iteration 19584 (0.765995 iter/s, 15.6659s/12 iters), loss = 2.55748 I0406 10:00:21.187988 5226 solver.cpp:237] Train net output #0: loss = 2.55748 (* 1 = 2.55748 loss) I0406 10:00:21.187992 5226 sgd_solver.cpp:105] Iteration 19584, lr = 0.01 I0406 10:00:25.739646 5226 solver.cpp:218] Iteration 19596 (2.63643 iter/s, 4.55161s/12 iters), loss = 2.67315 I0406 10:00:25.739693 5226 solver.cpp:237] Train net output #0: loss = 2.67315 (* 1 = 2.67315 loss) I0406 10:00:25.739701 5226 sgd_solver.cpp:105] Iteration 19596, lr = 0.01 I0406 10:00:31.497824 5226 solver.cpp:218] Iteration 19608 (2.08403 iter/s, 5.75808s/12 iters), loss = 2.80414 I0406 10:00:31.497865 5226 solver.cpp:237] Train net output #0: loss = 2.80414 (* 1 = 2.80414 loss) I0406 10:00:31.497870 5226 sgd_solver.cpp:105] Iteration 19608, lr = 0.01 I0406 10:00:37.089128 5226 solver.cpp:218] Iteration 19620 (2.14622 iter/s, 5.59121s/12 iters), loss = 2.99022 I0406 10:00:37.089169 5226 solver.cpp:237] Train net output #0: loss = 2.99022 (* 1 = 2.99022 loss) I0406 10:00:37.089174 5226 sgd_solver.cpp:105] Iteration 19620, lr = 0.01 I0406 10:00:38.773588 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 10:00:42.826838 5226 solver.cpp:218] Iteration 19632 (2.09146 iter/s, 5.73762s/12 iters), loss = 2.47468 I0406 10:00:42.826886 5226 solver.cpp:237] Train net output #0: loss = 2.47468 (* 1 = 2.47468 loss) I0406 10:00:42.826894 5226 sgd_solver.cpp:105] Iteration 19632, lr = 0.01 I0406 10:00:48.278543 5226 solver.cpp:218] Iteration 19644 (2.20119 iter/s, 5.45161s/12 iters), loss = 2.7661 I0406 10:00:48.278720 5226 solver.cpp:237] Train net output #0: loss = 2.7661 (* 1 = 2.7661 loss) I0406 10:00:48.278729 5226 sgd_solver.cpp:105] Iteration 19644, lr = 0.01 I0406 10:00:53.589474 5226 solver.cpp:218] Iteration 19656 (2.25959 iter/s, 5.31071s/12 iters), loss = 2.61602 I0406 10:00:53.589524 5226 solver.cpp:237] Train net output #0: loss = 2.61602 (* 1 = 2.61602 loss) I0406 10:00:53.589531 5226 sgd_solver.cpp:105] Iteration 19656, lr = 0.01 I0406 10:00:58.988624 5226 solver.cpp:218] Iteration 19668 (2.22261 iter/s, 5.39905s/12 iters), loss = 2.82244 I0406 10:00:58.988672 5226 solver.cpp:237] Train net output #0: loss = 2.82244 (* 1 = 2.82244 loss) I0406 10:00:58.988679 5226 sgd_solver.cpp:105] Iteration 19668, lr = 0.01 I0406 10:01:04.245191 5226 solver.cpp:218] Iteration 19680 (2.2829 iter/s, 5.25647s/12 iters), loss = 2.45953 I0406 10:01:04.245242 5226 solver.cpp:237] Train net output #0: loss = 2.45953 (* 1 = 2.45953 loss) I0406 10:01:04.245250 5226 sgd_solver.cpp:105] Iteration 19680, lr = 0.01 I0406 10:01:06.639740 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_19686.caffemodel I0406 10:01:09.652125 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_19686.solverstate I0406 10:01:13.198583 5226 solver.cpp:330] Iteration 19686, Testing net (#0) I0406 10:01:13.198604 5226 net.cpp:676] Ignoring source layer train-data I0406 10:01:14.521813 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 10:01:17.851245 5226 solver.cpp:397] Test net output #0: accuracy = 0.174632 I0406 10:01:17.851282 5226 solver.cpp:397] Test net output #1: loss = 4.19452 (* 1 = 4.19452 loss) I0406 10:01:19.787679 5226 solver.cpp:218] Iteration 19692 (0.772085 iter/s, 15.5423s/12 iters), loss = 2.93832 I0406 10:01:19.787804 5226 solver.cpp:237] Train net output #0: loss = 2.93832 (* 1 = 2.93832 loss) I0406 10:01:19.787817 5226 sgd_solver.cpp:105] Iteration 19692, lr = 0.01 I0406 10:01:25.043987 5226 solver.cpp:218] Iteration 19704 (2.28304 iter/s, 5.25614s/12 iters), loss = 2.85992 I0406 10:01:25.044035 5226 solver.cpp:237] Train net output #0: loss = 2.85992 (* 1 = 2.85992 loss) I0406 10:01:25.044042 5226 sgd_solver.cpp:105] Iteration 19704, lr = 0.01 I0406 10:01:30.476754 5226 solver.cpp:218] Iteration 19716 (2.20886 iter/s, 5.43267s/12 iters), loss = 2.76936 I0406 10:01:30.476807 5226 solver.cpp:237] Train net output #0: loss = 2.76936 (* 1 = 2.76936 loss) I0406 10:01:30.476815 5226 sgd_solver.cpp:105] Iteration 19716, lr = 0.01 I0406 10:01:34.316648 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 10:01:36.029168 5226 solver.cpp:218] Iteration 19728 (2.16126 iter/s, 5.55232s/12 iters), loss = 2.64841 I0406 10:01:36.029207 5226 solver.cpp:237] Train net output #0: loss = 2.64841 (* 1 = 2.64841 loss) I0406 10:01:36.029218 5226 sgd_solver.cpp:105] Iteration 19728, lr = 0.01 I0406 10:01:41.592639 5226 solver.cpp:218] Iteration 19740 (2.15696 iter/s, 5.56337s/12 iters), loss = 2.73363 I0406 10:01:41.592691 5226 solver.cpp:237] Train net output #0: loss = 2.73363 (* 1 = 2.73363 loss) I0406 10:01:41.592700 5226 sgd_solver.cpp:105] Iteration 19740, lr = 0.01 I0406 10:01:47.095300 5226 solver.cpp:218] Iteration 19752 (2.1808 iter/s, 5.50256s/12 iters), loss = 2.18588 I0406 10:01:47.095358 5226 solver.cpp:237] Train net output #0: loss = 2.18588 (* 1 = 2.18588 loss) I0406 10:01:47.095368 5226 sgd_solver.cpp:105] Iteration 19752, lr = 0.01 I0406 10:01:52.475334 5226 solver.cpp:218] Iteration 19764 (2.23051 iter/s, 5.37993s/12 iters), loss = 2.68636 I0406 10:01:52.475489 5226 solver.cpp:237] Train net output #0: loss = 2.68636 (* 1 = 2.68636 loss) I0406 10:01:52.475498 5226 sgd_solver.cpp:105] Iteration 19764, lr = 0.01 I0406 10:01:58.304586 5226 solver.cpp:218] Iteration 19776 (2.05866 iter/s, 5.82905s/12 iters), loss = 2.37485 I0406 10:01:58.304636 5226 solver.cpp:237] Train net output #0: loss = 2.37485 (* 1 = 2.37485 loss) I0406 10:01:58.304643 5226 sgd_solver.cpp:105] Iteration 19776, lr = 0.01 I0406 10:02:03.341866 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_19788.caffemodel I0406 10:02:06.477465 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_19788.solverstate I0406 10:02:10.218739 5226 solver.cpp:330] Iteration 19788, Testing net (#0) I0406 10:02:10.218761 5226 net.cpp:676] Ignoring source layer train-data I0406 10:02:11.487207 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 10:02:14.822784 5226 solver.cpp:397] Test net output #0: accuracy = 0.164216 I0406 10:02:14.822819 5226 solver.cpp:397] Test net output #1: loss = 4.23622 (* 1 = 4.23622 loss) I0406 10:02:14.960054 5226 solver.cpp:218] Iteration 19788 (0.720491 iter/s, 16.6553s/12 iters), loss = 2.22507 I0406 10:02:14.960114 5226 solver.cpp:237] Train net output #0: loss = 2.22507 (* 1 = 2.22507 loss) I0406 10:02:14.960122 5226 sgd_solver.cpp:105] Iteration 19788, lr = 0.01 I0406 10:02:19.360641 5226 solver.cpp:218] Iteration 19800 (2.72697 iter/s, 4.40048s/12 iters), loss = 2.69649 I0406 10:02:19.360697 5226 solver.cpp:237] Train net output #0: loss = 2.69649 (* 1 = 2.69649 loss) I0406 10:02:19.360707 5226 sgd_solver.cpp:105] Iteration 19800, lr = 0.01 I0406 10:02:24.752535 5226 solver.cpp:218] Iteration 19812 (2.22561 iter/s, 5.39179s/12 iters), loss = 2.933 I0406 10:02:24.752658 5226 solver.cpp:237] Train net output #0: loss = 2.933 (* 1 = 2.933 loss) I0406 10:02:24.752671 5226 sgd_solver.cpp:105] Iteration 19812, lr = 0.01 I0406 10:02:30.157891 5226 solver.cpp:218] Iteration 19824 (2.22009 iter/s, 5.40519s/12 iters), loss = 2.82492 I0406 10:02:30.157929 5226 solver.cpp:237] Train net output #0: loss = 2.82492 (* 1 = 2.82492 loss) I0406 10:02:30.157934 5226 sgd_solver.cpp:105] Iteration 19824, lr = 0.01 I0406 10:02:30.838217 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 10:02:35.788676 5226 solver.cpp:218] Iteration 19836 (2.13117 iter/s, 5.6307s/12 iters), loss = 2.92864 I0406 10:02:35.788715 5226 solver.cpp:237] Train net output #0: loss = 2.92864 (* 1 = 2.92864 loss) I0406 10:02:35.788720 5226 sgd_solver.cpp:105] Iteration 19836, lr = 0.01 I0406 10:02:41.293627 5226 solver.cpp:218] Iteration 19848 (2.17989 iter/s, 5.50486s/12 iters), loss = 2.98564 I0406 10:02:41.293673 5226 solver.cpp:237] Train net output #0: loss = 2.98564 (* 1 = 2.98564 loss) I0406 10:02:41.293681 5226 sgd_solver.cpp:105] Iteration 19848, lr = 0.01 I0406 10:02:46.821861 5226 solver.cpp:218] Iteration 19860 (2.17071 iter/s, 5.52814s/12 iters), loss = 2.59119 I0406 10:02:46.821916 5226 solver.cpp:237] Train net output #0: loss = 2.59119 (* 1 = 2.59119 loss) I0406 10:02:46.821925 5226 sgd_solver.cpp:105] Iteration 19860, lr = 0.01 I0406 10:02:52.340847 5226 solver.cpp:218] Iteration 19872 (2.17435 iter/s, 5.51888s/12 iters), loss = 2.7598 I0406 10:02:52.340909 5226 solver.cpp:237] Train net output #0: loss = 2.7598 (* 1 = 2.7598 loss) I0406 10:02:52.340919 5226 sgd_solver.cpp:105] Iteration 19872, lr = 0.01 I0406 10:02:57.715147 5226 solver.cpp:218] Iteration 19884 (2.2329 iter/s, 5.37419s/12 iters), loss = 2.81648 I0406 10:02:57.715274 5226 solver.cpp:237] Train net output #0: loss = 2.81648 (* 1 = 2.81648 loss) I0406 10:02:57.715283 5226 sgd_solver.cpp:105] Iteration 19884, lr = 0.01 I0406 10:03:00.006028 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_19890.caffemodel I0406 10:03:03.048367 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_19890.solverstate I0406 10:03:06.431736 5226 solver.cpp:330] Iteration 19890, Testing net (#0) I0406 10:03:06.431761 5226 net.cpp:676] Ignoring source layer train-data I0406 10:03:07.632391 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 10:03:11.090910 5226 solver.cpp:397] Test net output #0: accuracy = 0.163603 I0406 10:03:11.090945 5226 solver.cpp:397] Test net output #1: loss = 4.29033 (* 1 = 4.29033 loss) I0406 10:03:12.970165 5226 solver.cpp:218] Iteration 19896 (0.786638 iter/s, 15.2548s/12 iters), loss = 2.81393 I0406 10:03:12.970223 5226 solver.cpp:237] Train net output #0: loss = 2.81393 (* 1 = 2.81393 loss) I0406 10:03:12.970232 5226 sgd_solver.cpp:105] Iteration 19896, lr = 0.01 I0406 10:03:18.350425 5226 solver.cpp:218] Iteration 19908 (2.23042 iter/s, 5.38015s/12 iters), loss = 2.29048 I0406 10:03:18.350466 5226 solver.cpp:237] Train net output #0: loss = 2.29048 (* 1 = 2.29048 loss) I0406 10:03:18.350471 5226 sgd_solver.cpp:105] Iteration 19908, lr = 0.01 I0406 10:03:23.800796 5226 solver.cpp:218] Iteration 19920 (2.20172 iter/s, 5.45028s/12 iters), loss = 2.59675 I0406 10:03:23.800833 5226 solver.cpp:237] Train net output #0: loss = 2.59675 (* 1 = 2.59675 loss) I0406 10:03:23.800838 5226 sgd_solver.cpp:105] Iteration 19920, lr = 0.01 I0406 10:03:26.722972 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 10:03:29.197059 5226 solver.cpp:218] Iteration 19932 (2.2238 iter/s, 5.39618s/12 iters), loss = 2.73143 I0406 10:03:29.197175 5226 solver.cpp:237] Train net output #0: loss = 2.73143 (* 1 = 2.73143 loss) I0406 10:03:29.197181 5226 sgd_solver.cpp:105] Iteration 19932, lr = 0.01 I0406 10:03:34.904337 5226 solver.cpp:218] Iteration 19944 (2.10264 iter/s, 5.70711s/12 iters), loss = 3.0777 I0406 10:03:34.904386 5226 solver.cpp:237] Train net output #0: loss = 3.0777 (* 1 = 3.0777 loss) I0406 10:03:34.904392 5226 sgd_solver.cpp:105] Iteration 19944, lr = 0.01 I0406 10:03:40.273808 5226 solver.cpp:218] Iteration 19956 (2.2349 iter/s, 5.36938s/12 iters), loss = 2.64861 I0406 10:03:40.273856 5226 solver.cpp:237] Train net output #0: loss = 2.64861 (* 1 = 2.64861 loss) I0406 10:03:40.273864 5226 sgd_solver.cpp:105] Iteration 19956, lr = 0.01 I0406 10:03:45.456115 5226 solver.cpp:218] Iteration 19968 (2.31562 iter/s, 5.1822s/12 iters), loss = 2.64935 I0406 10:03:45.456171 5226 solver.cpp:237] Train net output #0: loss = 2.64935 (* 1 = 2.64935 loss) I0406 10:03:45.456182 5226 sgd_solver.cpp:105] Iteration 19968, lr = 0.01 I0406 10:03:50.800523 5226 solver.cpp:218] Iteration 19980 (2.24538 iter/s, 5.34431s/12 iters), loss = 2.70214 I0406 10:03:50.800575 5226 solver.cpp:237] Train net output #0: loss = 2.70214 (* 1 = 2.70214 loss) I0406 10:03:50.800583 5226 sgd_solver.cpp:105] Iteration 19980, lr = 0.01 I0406 10:03:55.829727 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_19992.caffemodel I0406 10:03:58.849689 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_19992.solverstate I0406 10:04:01.469600 5226 solver.cpp:330] Iteration 19992, Testing net (#0) I0406 10:04:01.469662 5226 net.cpp:676] Ignoring source layer train-data I0406 10:04:02.646924 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 10:04:05.841603 5226 solver.cpp:397] Test net output #0: accuracy = 0.167892 I0406 10:04:05.841639 5226 solver.cpp:397] Test net output #1: loss = 4.1912 (* 1 = 4.1912 loss) I0406 10:04:05.979552 5226 solver.cpp:218] Iteration 19992 (0.790572 iter/s, 15.1789s/12 iters), loss = 3.25358 I0406 10:04:05.979607 5226 solver.cpp:237] Train net output #0: loss = 3.25358 (* 1 = 3.25358 loss) I0406 10:04:05.979615 5226 sgd_solver.cpp:105] Iteration 19992, lr = 0.01 I0406 10:04:10.333920 5226 solver.cpp:218] Iteration 20004 (2.75591 iter/s, 4.35427s/12 iters), loss = 2.75259 I0406 10:04:10.333961 5226 solver.cpp:237] Train net output #0: loss = 2.75259 (* 1 = 2.75259 loss) I0406 10:04:10.333966 5226 sgd_solver.cpp:105] Iteration 20004, lr = 0.01 I0406 10:04:15.942981 5226 solver.cpp:218] Iteration 20016 (2.13943 iter/s, 5.60896s/12 iters), loss = 2.79368 I0406 10:04:15.943027 5226 solver.cpp:237] Train net output #0: loss = 2.79368 (* 1 = 2.79368 loss) I0406 10:04:15.943035 5226 sgd_solver.cpp:105] Iteration 20016, lr = 0.01 I0406 10:04:21.229457 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 10:04:21.394726 5226 solver.cpp:218] Iteration 20028 (2.20117 iter/s, 5.45165s/12 iters), loss = 2.57669 I0406 10:04:21.394781 5226 solver.cpp:237] Train net output #0: loss = 2.57669 (* 1 = 2.57669 loss) I0406 10:04:21.394790 5226 sgd_solver.cpp:105] Iteration 20028, lr = 0.01 I0406 10:04:26.973359 5226 solver.cpp:218] Iteration 20040 (2.1511 iter/s, 5.57853s/12 iters), loss = 2.99062 I0406 10:04:26.973420 5226 solver.cpp:237] Train net output #0: loss = 2.99062 (* 1 = 2.99062 loss) I0406 10:04:26.973428 5226 sgd_solver.cpp:105] Iteration 20040, lr = 0.01 I0406 10:04:32.324486 5226 solver.cpp:218] Iteration 20052 (2.24256 iter/s, 5.35103s/12 iters), loss = 2.56087 I0406 10:04:32.324627 5226 solver.cpp:237] Train net output #0: loss = 2.56087 (* 1 = 2.56087 loss) I0406 10:04:32.324636 5226 sgd_solver.cpp:105] Iteration 20052, lr = 0.01 I0406 10:04:40.948962 5226 solver.cpp:218] Iteration 20064 (1.39179 iter/s, 8.62201s/12 iters), loss = 2.90471 I0406 10:04:40.949019 5226 solver.cpp:237] Train net output #0: loss = 2.90471 (* 1 = 2.90471 loss) I0406 10:04:40.949028 5226 sgd_solver.cpp:105] Iteration 20064, lr = 0.01 I0406 10:04:50.913915 5226 solver.cpp:218] Iteration 20076 (1.20457 iter/s, 9.96204s/12 iters), loss = 2.82843 I0406 10:04:50.913975 5226 solver.cpp:237] Train net output #0: loss = 2.82843 (* 1 = 2.82843 loss) I0406 10:04:50.913983 5226 sgd_solver.cpp:105] Iteration 20076, lr = 0.01 I0406 10:04:58.682845 5226 solver.cpp:218] Iteration 20088 (1.54464 iter/s, 7.76881s/12 iters), loss = 2.53284 I0406 10:04:58.684938 5226 solver.cpp:237] Train net output #0: loss = 2.53284 (* 1 = 2.53284 loss) I0406 10:04:58.684958 5226 sgd_solver.cpp:105] Iteration 20088, lr = 0.01 I0406 10:05:01.583570 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_20094.caffemodel I0406 10:05:06.201537 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_20094.solverstate I0406 10:05:09.618099 5226 solver.cpp:330] Iteration 20094, Testing net (#0) I0406 10:05:09.618127 5226 net.cpp:676] Ignoring source layer train-data I0406 10:05:12.187541 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 10:05:17.569209 5226 blocking_queue.cpp:49] Waiting for data I0406 10:05:18.524080 5226 solver.cpp:397] Test net output #0: accuracy = 0.14277 I0406 10:05:18.524119 5226 solver.cpp:397] Test net output #1: loss = 4.42466 (* 1 = 4.42466 loss) I0406 10:05:21.724939 5226 solver.cpp:218] Iteration 20100 (0.520925 iter/s, 23.036s/12 iters), loss = 2.85385 I0406 10:05:21.725000 5226 solver.cpp:237] Train net output #0: loss = 2.85385 (* 1 = 2.85385 loss) I0406 10:05:21.725008 5226 sgd_solver.cpp:105] Iteration 20100, lr = 0.01 I0406 10:05:30.065737 5226 solver.cpp:218] Iteration 20112 (1.43873 iter/s, 8.34066s/12 iters), loss = 2.83841 I0406 10:05:30.065793 5226 solver.cpp:237] Train net output #0: loss = 2.83841 (* 1 = 2.83841 loss) I0406 10:05:30.065802 5226 sgd_solver.cpp:105] Iteration 20112, lr = 0.01 I0406 10:05:37.080931 5226 solver.cpp:218] Iteration 20124 (1.71301 iter/s, 7.00522s/12 iters), loss = 2.47944 I0406 10:05:37.081384 5226 solver.cpp:237] Train net output #0: loss = 2.47944 (* 1 = 2.47944 loss) I0406 10:05:37.081394 5226 sgd_solver.cpp:105] Iteration 20124, lr = 0.01 I0406 10:05:40.017272 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 10:05:43.877558 5226 solver.cpp:218] Iteration 20136 (1.76571 iter/s, 6.79612s/12 iters), loss = 2.62401 I0406 10:05:43.877617 5226 solver.cpp:237] Train net output #0: loss = 2.62401 (* 1 = 2.62401 loss) I0406 10:05:43.877626 5226 sgd_solver.cpp:105] Iteration 20136, lr = 0.01 I0406 10:05:50.228932 5226 solver.cpp:218] Iteration 20148 (1.89371 iter/s, 6.33678s/12 iters), loss = 2.69378 I0406 10:05:50.228986 5226 solver.cpp:237] Train net output #0: loss = 2.69378 (* 1 = 2.69378 loss) I0406 10:05:50.228996 5226 sgd_solver.cpp:105] Iteration 20148, lr = 0.01 I0406 10:05:56.992928 5226 solver.cpp:218] Iteration 20160 (1.77583 iter/s, 6.7574s/12 iters), loss = 2.29788 I0406 10:05:56.992980 5226 solver.cpp:237] Train net output #0: loss = 2.29788 (* 1 = 2.29788 loss) I0406 10:05:56.992988 5226 sgd_solver.cpp:105] Iteration 20160, lr = 0.01 I0406 10:06:04.076555 5226 solver.cpp:218] Iteration 20172 (1.69407 iter/s, 7.08351s/12 iters), loss = 2.44717 I0406 10:06:04.076604 5226 solver.cpp:237] Train net output #0: loss = 2.44717 (* 1 = 2.44717 loss) I0406 10:06:04.076612 5226 sgd_solver.cpp:105] Iteration 20172, lr = 0.01 I0406 10:06:10.535975 5226 solver.cpp:218] Iteration 20184 (1.85778 iter/s, 6.45931s/12 iters), loss = 2.75424 I0406 10:06:10.536149 5226 solver.cpp:237] Train net output #0: loss = 2.75424 (* 1 = 2.75424 loss) I0406 10:06:10.536160 5226 sgd_solver.cpp:105] Iteration 20184, lr = 0.01 I0406 10:06:16.711539 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_20196.caffemodel I0406 10:06:20.339686 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_20196.solverstate I0406 10:06:23.778192 5226 solver.cpp:330] Iteration 20196, Testing net (#0) I0406 10:06:23.778215 5226 net.cpp:676] Ignoring source layer train-data I0406 10:06:25.327039 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 10:06:28.911279 5226 solver.cpp:397] Test net output #0: accuracy = 0.168505 I0406 10:06:28.911314 5226 solver.cpp:397] Test net output #1: loss = 4.23951 (* 1 = 4.23951 loss) I0406 10:06:29.051805 5226 solver.cpp:218] Iteration 20196 (0.648104 iter/s, 18.5155s/12 iters), loss = 2.20535 I0406 10:06:29.051856 5226 solver.cpp:237] Train net output #0: loss = 2.20535 (* 1 = 2.20535 loss) I0406 10:06:29.051864 5226 sgd_solver.cpp:105] Iteration 20196, lr = 0.01 I0406 10:06:33.950533 5226 solver.cpp:218] Iteration 20208 (2.44966 iter/s, 4.89863s/12 iters), loss = 2.79625 I0406 10:06:33.950587 5226 solver.cpp:237] Train net output #0: loss = 2.79625 (* 1 = 2.79625 loss) I0406 10:06:33.950596 5226 sgd_solver.cpp:105] Iteration 20208, lr = 0.01 I0406 10:06:39.542714 5226 solver.cpp:218] Iteration 20220 (2.14589 iter/s, 5.59208s/12 iters), loss = 2.83598 I0406 10:06:39.542754 5226 solver.cpp:237] Train net output #0: loss = 2.83598 (* 1 = 2.83598 loss) I0406 10:06:39.542760 5226 sgd_solver.cpp:105] Iteration 20220, lr = 0.01 I0406 10:06:45.237886 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 10:06:47.852932 5226 solver.cpp:218] Iteration 20232 (1.44406 iter/s, 8.30992s/12 iters), loss = 2.70664 I0406 10:06:47.852983 5226 solver.cpp:237] Train net output #0: loss = 2.70664 (* 1 = 2.70664 loss) I0406 10:06:47.852993 5226 sgd_solver.cpp:105] Iteration 20232, lr = 0.01 I0406 10:06:58.563769 5226 solver.cpp:218] Iteration 20244 (1.12043 iter/s, 10.7101s/12 iters), loss = 2.64755 I0406 10:06:58.563823 5226 solver.cpp:237] Train net output #0: loss = 2.64755 (* 1 = 2.64755 loss) I0406 10:06:58.563830 5226 sgd_solver.cpp:105] Iteration 20244, lr = 0.01 I0406 10:07:11.468931 5226 solver.cpp:218] Iteration 20256 (0.930074 iter/s, 12.9022s/12 iters), loss = 2.33334 I0406 10:07:11.468987 5226 solver.cpp:237] Train net output #0: loss = 2.33334 (* 1 = 2.33334 loss) I0406 10:07:11.468994 5226 sgd_solver.cpp:105] Iteration 20256, lr = 0.01 I0406 10:07:23.288929 5226 solver.cpp:218] Iteration 20268 (1.01529 iter/s, 11.8193s/12 iters), loss = 2.78814 I0406 10:07:23.289147 5226 solver.cpp:237] Train net output #0: loss = 2.78814 (* 1 = 2.78814 loss) I0406 10:07:23.289156 5226 sgd_solver.cpp:105] Iteration 20268, lr = 0.01 I0406 10:07:30.058640 5226 solver.cpp:218] Iteration 20280 (1.77267 iter/s, 6.76944s/12 iters), loss = 2.64901 I0406 10:07:30.058686 5226 solver.cpp:237] Train net output #0: loss = 2.64901 (* 1 = 2.64901 loss) I0406 10:07:30.058692 5226 sgd_solver.cpp:105] Iteration 20280, lr = 0.01 I0406 10:07:36.943645 5226 solver.cpp:218] Iteration 20292 (1.74296 iter/s, 6.88485s/12 iters), loss = 2.75297 I0406 10:07:36.943696 5226 solver.cpp:237] Train net output #0: loss = 2.75297 (* 1 = 2.75297 loss) I0406 10:07:36.943704 5226 sgd_solver.cpp:105] Iteration 20292, lr = 0.01 I0406 10:07:39.623667 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_20298.caffemodel I0406 10:07:43.239082 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_20298.solverstate I0406 10:07:48.326705 5226 solver.cpp:330] Iteration 20298, Testing net (#0) I0406 10:07:48.326730 5226 net.cpp:676] Ignoring source layer train-data I0406 10:07:49.678532 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 10:07:54.613147 5226 solver.cpp:397] Test net output #0: accuracy = 0.151348 I0406 10:07:54.613283 5226 solver.cpp:397] Test net output #1: loss = 4.2782 (* 1 = 4.2782 loss) I0406 10:07:56.934345 5226 solver.cpp:218] Iteration 20304 (0.600285 iter/s, 19.9905s/12 iters), loss = 2.91154 I0406 10:07:56.934396 5226 solver.cpp:237] Train net output #0: loss = 2.91154 (* 1 = 2.91154 loss) I0406 10:07:56.934403 5226 sgd_solver.cpp:105] Iteration 20304, lr = 0.01 I0406 10:08:03.890859 5226 solver.cpp:218] Iteration 20316 (1.72503 iter/s, 6.9564s/12 iters), loss = 2.29866 I0406 10:08:03.890918 5226 solver.cpp:237] Train net output #0: loss = 2.29866 (* 1 = 2.29866 loss) I0406 10:08:03.890926 5226 sgd_solver.cpp:105] Iteration 20316, lr = 0.01 I0406 10:08:10.638542 5226 solver.cpp:218] Iteration 20328 (1.77842 iter/s, 6.74756s/12 iters), loss = 2.68673 I0406 10:08:10.638592 5226 solver.cpp:237] Train net output #0: loss = 2.68673 (* 1 = 2.68673 loss) I0406 10:08:10.638600 5226 sgd_solver.cpp:105] Iteration 20328, lr = 0.01 I0406 10:08:12.460681 5252 data_layer.cpp:73] Restarting data prefetching from start. I0406 10:08:17.238793 5226 solver.cpp:218] Iteration 20340 (1.81814 iter/s, 6.60014s/12 iters), loss = 2.39676 I0406 10:08:17.258757 5226 solver.cpp:237] Train net output #0: loss = 2.39676 (* 1 = 2.39676 loss) I0406 10:08:17.258780 5226 sgd_solver.cpp:105] Iteration 20340, lr = 0.01 I0406 10:08:23.880574 5226 solver.cpp:218] Iteration 20352 (1.8122 iter/s, 6.62178s/12 iters), loss = 2.60019 I0406 10:08:23.880630 5226 solver.cpp:237] Train net output #0: loss = 2.60019 (* 1 = 2.60019 loss) I0406 10:08:23.880637 5226 sgd_solver.cpp:105] Iteration 20352, lr = 0.01 I0406 10:08:30.384938 5226 solver.cpp:218] Iteration 20364 (1.84796 iter/s, 6.49363s/12 iters), loss = 3.03568 I0406 10:08:30.385072 5226 solver.cpp:237] Train net output #0: loss = 3.03568 (* 1 = 3.03568 loss) I0406 10:08:30.385082 5226 sgd_solver.cpp:105] Iteration 20364, lr = 0.01 I0406 10:08:36.180506 5226 solver.cpp:218] Iteration 20376 (2.07061 iter/s, 5.79538s/12 iters), loss = 2.34857 I0406 10:08:36.180553 5226 solver.cpp:237] Train net output #0: loss = 2.34857 (* 1 = 2.34857 loss) I0406 10:08:36.180559 5226 sgd_solver.cpp:105] Iteration 20376, lr = 0.01 I0406 10:08:41.693197 5226 solver.cpp:218] Iteration 20388 (2.17683 iter/s, 5.51259s/12 iters), loss = 2.17682 I0406 10:08:41.693245 5226 solver.cpp:237] Train net output #0: loss = 2.17682 (* 1 = 2.17682 loss) I0406 10:08:41.693253 5226 sgd_solver.cpp:105] Iteration 20388, lr = 0.01 I0406 10:08:46.666807 5226 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_20400.caffemodel I0406 10:08:49.926069 5226 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_20400.solverstate I0406 10:08:54.345849 5226 solver.cpp:310] Iteration 20400, loss = 2.60572 I0406 10:08:54.345885 5226 solver.cpp:330] Iteration 20400, Testing net (#0) I0406 10:08:54.345890 5226 net.cpp:676] Ignoring source layer train-data I0406 10:08:55.404270 5283 data_layer.cpp:73] Restarting data prefetching from start. I0406 10:08:58.876803 5226 solver.cpp:397] Test net output #0: accuracy = 0.174632 I0406 10:08:58.876838 5226 solver.cpp:397] Test net output #1: loss = 4.15245 (* 1 = 4.15245 loss) I0406 10:08:58.876844 5226 solver.cpp:315] Optimization Done. I0406 10:08:58.876847 5226 caffe.cpp:259] Optimization Done.