I0409 23:19:45.908432 8290 upgrade_proto.cpp:1082] Attempting to upgrade input file specified using deprecated 'solver_type' field (enum)': /mnt/bigdisk/DIGITS-MAN-2/digits/jobs/20210409-211041-f3f5/solver.prototxt I0409 23:19:45.908712 8290 upgrade_proto.cpp:1089] Successfully upgraded file specified using deprecated 'solver_type' field (enum) to 'type' field (string). W0409 23:19:45.908726 8290 upgrade_proto.cpp:1091] Note that future Caffe releases will only support 'type' field (string) for a solver's type. I0409 23:19:45.908850 8290 caffe.cpp:218] Using GPUs 2 I0409 23:19:45.933754 8290 caffe.cpp:223] GPU 2: GeForce GTX 1080 Ti I0409 23:19:46.224370 8290 solver.cpp:44] Initializing solver from parameters: test_iter: 51 test_interval: 102 base_lr: 0.01 display: 12 max_iter: 10200 lr_policy: "exp" gamma: 0.99980193 momentum: 0.9 weight_decay: 0.0001 snapshot: 102 snapshot_prefix: "snapshot" solver_mode: GPU device_id: 2 net: "train_val.prototxt" train_state { level: 0 stage: "" } type: "SGD" I0409 23:19:46.225265 8290 solver.cpp:87] Creating training net from net file: train_val.prototxt I0409 23:19:46.225858 8290 net.cpp:294] The NetState phase (0) differed from the phase (1) specified by a rule in layer val-data I0409 23:19:46.225874 8290 net.cpp:294] The NetState phase (0) differed from the phase (1) specified by a rule in layer accuracy I0409 23:19:46.226034 8290 net.cpp:51] Initializing net from parameters: state { phase: TRAIN level: 0 stage: "" } layer { name: "train-data" type: "Data" top: "data" top: "label" include { phase: TRAIN } transform_param { mirror: true crop_size: 227 mean_file: "/mnt/bigdisk/DIGITS-MAN-2/digits/jobs/20210407-214532-d396/mean.binaryproto" } data_param { source: "/mnt/bigdisk/DIGITS-MAN-2/digits/jobs/20210407-214532-d396/train_db" batch_size: 128 backend: LMDB } } layer { name: "conv1" type: "Convolution" bottom: "data" top: "conv1" param { lr_mult: 1 decay_mult: 1 } param { lr_mult: 2 decay_mult: 0 } convolution_param { num_output: 96 kernel_size: 11 stride: 4 weight_filler { type: "gaussian" std: 0.01 } bias_filler { type: "constant" value: 0 } } } layer { name: "relu1" type: "ReLU" bottom: "conv1" top: "conv1" } layer { name: "norm1" type: "LRN" bottom: "conv1" top: "norm1" lrn_param { local_size: 5 alpha: 0.0001 beta: 0.75 } } layer { name: "pool1" type: "Pooling" bottom: "norm1" top: "pool1" pooling_param { pool: MAX kernel_size: 3 stride: 2 } } layer { name: "conv2" type: "Convolution" bottom: "pool1" top: "conv2" param { lr_mult: 1 decay_mult: 1 } param { lr_mult: 2 decay_mult: 0 } convolution_param { num_output: 256 pad: 2 kernel_size: 5 group: 2 weight_filler { type: "gaussian" std: 0.01 } bias_filler { type: "constant" value: 0.1 } } } layer { name: "relu2" type: "ReLU" bottom: "conv2" top: "conv2" } layer { name: "norm2" type: "LRN" bottom: "conv2" top: "norm2" lrn_param { local_size: 5 alpha: 0.0001 beta: 0.75 } } layer { name: "pool2" type: "Pooling" bottom: "norm2" top: "pool2" pooling_param { pool: MAX kernel_size: 3 stride: 2 } } layer { name: "conv3" type: "Convolution" bottom: "pool2" top: "conv3" param { lr_mult: 1 decay_mult: 1 } param { lr_mult: 2 decay_mult: 0 } convolution_param { num_output: 384 pad: 1 kernel_size: 3 weight_filler { type: "gaussian" std: 0.01 } bias_filler { type: "constant" value: 0 } } } layer { name: "relu3" type: "ReLU" bottom: "conv3" top: "conv3" } layer { name: "conv4" type: "Convolution" bottom: "conv3" top: "conv4" param { lr_mult: 1 decay_mult: 1 } param { lr_mult: 2 decay_mult: 0 } convolution_param { num_output: 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: 1024 weight_filler { type: "gaussian" std: 0.005 } bias_filler { type: "constant" value: 0.1 } } } layer { name: "relu6" type: "ReLU" bottom: "fc6" top: "fc6" } layer { name: "drop6" type: "Dropout" bottom: "fc6" top: "fc6" dropout_param { dropout_ratio: 0.5 } } layer { name: "fc7" type: "InnerProduct" bottom: "fc6" top: "fc7" param { lr_mult: 1 decay_mult: 1 } param { lr_mult: 2 decay_mult: 0 } inner_product_param { num_output: 1024 weight_filler { type: "gaussian" std: 0.005 } bias_filler { type: "constant" value: 0.1 } } } layer { name: "relu7" type: "ReLU" bottom: "fc7" top: "fc7" } layer { name: "drop7" type: "Dropout" bottom: "fc7" top: "fc7" dropout_param { dropout_ratio: 0.5 } } layer { name: "fc8" type: "InnerProduct" bottom: "fc7" top: "fc8" param { lr_mult: 1 decay_mult: 1 } param { lr_mult: 2 decay_mult: 0 } inner_product_param { num_output: 196 weight_filler { type: "gaussian" std: 0.01 } bias_filler { type: "constant" value: 0 } } } layer { name: "loss" type: "SoftmaxWithLoss" bottom: "fc8" bottom: "label" top: "loss" } I0409 23:19:46.226128 8290 layer_factory.hpp:77] Creating layer train-data I0409 23:19:46.228001 8290 db_lmdb.cpp:35] Opened lmdb /mnt/bigdisk/DIGITS-MAN-2/digits/jobs/20210407-214532-d396/train_db I0409 23:19:46.228206 8290 net.cpp:84] Creating Layer train-data I0409 23:19:46.228216 8290 net.cpp:380] train-data -> data I0409 23:19:46.228236 8290 net.cpp:380] train-data -> label I0409 23:19:46.228247 8290 data_transformer.cpp:25] Loading mean file from: /mnt/bigdisk/DIGITS-MAN-2/digits/jobs/20210407-214532-d396/mean.binaryproto I0409 23:19:46.239208 8290 data_layer.cpp:45] output data size: 128,3,227,227 I0409 23:19:46.374977 8290 net.cpp:122] Setting up train-data I0409 23:19:46.375000 8290 net.cpp:129] Top shape: 128 3 227 227 (19787136) I0409 23:19:46.375005 8290 net.cpp:129] Top shape: 128 (128) I0409 23:19:46.375008 8290 net.cpp:137] Memory required for data: 79149056 I0409 23:19:46.375018 8290 layer_factory.hpp:77] Creating layer conv1 I0409 23:19:46.375039 8290 net.cpp:84] Creating Layer conv1 I0409 23:19:46.375044 8290 net.cpp:406] conv1 <- data I0409 23:19:46.375057 8290 net.cpp:380] conv1 -> conv1 I0409 23:19:46.988574 8290 net.cpp:122] Setting up conv1 I0409 23:19:46.988597 8290 net.cpp:129] Top shape: 128 96 55 55 (37171200) I0409 23:19:46.988601 8290 net.cpp:137] Memory required for data: 227833856 I0409 23:19:46.988621 8290 layer_factory.hpp:77] Creating layer relu1 I0409 23:19:46.988632 8290 net.cpp:84] Creating Layer relu1 I0409 23:19:46.988638 8290 net.cpp:406] relu1 <- conv1 I0409 23:19:46.988646 8290 net.cpp:367] relu1 -> conv1 (in-place) I0409 23:19:46.988961 8290 net.cpp:122] Setting up relu1 I0409 23:19:46.988970 8290 net.cpp:129] Top shape: 128 96 55 55 (37171200) I0409 23:19:46.988974 8290 net.cpp:137] Memory required for data: 376518656 I0409 23:19:46.988978 8290 layer_factory.hpp:77] Creating layer norm1 I0409 23:19:46.988987 8290 net.cpp:84] Creating Layer norm1 I0409 23:19:46.988991 8290 net.cpp:406] norm1 <- conv1 I0409 23:19:46.989017 8290 net.cpp:380] norm1 -> norm1 I0409 23:19:46.989492 8290 net.cpp:122] Setting up norm1 I0409 23:19:46.989502 8290 net.cpp:129] Top shape: 128 96 55 55 (37171200) I0409 23:19:46.989506 8290 net.cpp:137] Memory required for data: 525203456 I0409 23:19:46.989511 8290 layer_factory.hpp:77] Creating layer pool1 I0409 23:19:46.989519 8290 net.cpp:84] Creating Layer pool1 I0409 23:19:46.989523 8290 net.cpp:406] pool1 <- norm1 I0409 23:19:46.989529 8290 net.cpp:380] pool1 -> pool1 I0409 23:19:46.989567 8290 net.cpp:122] Setting up pool1 I0409 23:19:46.989573 8290 net.cpp:129] Top shape: 128 96 27 27 (8957952) I0409 23:19:46.989576 8290 net.cpp:137] Memory required for data: 561035264 I0409 23:19:46.989580 8290 layer_factory.hpp:77] Creating layer conv2 I0409 23:19:46.989591 8290 net.cpp:84] Creating Layer conv2 I0409 23:19:46.989595 8290 net.cpp:406] conv2 <- pool1 I0409 23:19:46.989600 8290 net.cpp:380] conv2 -> conv2 I0409 23:19:46.998409 8290 net.cpp:122] Setting up conv2 I0409 23:19:46.998428 8290 net.cpp:129] Top shape: 128 256 27 27 (23887872) I0409 23:19:46.998432 8290 net.cpp:137] Memory required for data: 656586752 I0409 23:19:46.998445 8290 layer_factory.hpp:77] Creating layer relu2 I0409 23:19:46.998452 8290 net.cpp:84] Creating Layer relu2 I0409 23:19:46.998456 8290 net.cpp:406] relu2 <- conv2 I0409 23:19:46.998461 8290 net.cpp:367] relu2 -> conv2 (in-place) I0409 23:19:46.998900 8290 net.cpp:122] Setting up relu2 I0409 23:19:46.998910 8290 net.cpp:129] Top shape: 128 256 27 27 (23887872) I0409 23:19:46.998914 8290 net.cpp:137] Memory required for data: 752138240 I0409 23:19:46.998917 8290 layer_factory.hpp:77] Creating layer norm2 I0409 23:19:46.998925 8290 net.cpp:84] Creating Layer norm2 I0409 23:19:46.998929 8290 net.cpp:406] norm2 <- conv2 I0409 23:19:46.998934 8290 net.cpp:380] norm2 -> norm2 I0409 23:19:46.999236 8290 net.cpp:122] Setting up norm2 I0409 23:19:46.999245 8290 net.cpp:129] Top shape: 128 256 27 27 (23887872) I0409 23:19:46.999249 8290 net.cpp:137] Memory required for data: 847689728 I0409 23:19:46.999253 8290 layer_factory.hpp:77] Creating layer pool2 I0409 23:19:46.999262 8290 net.cpp:84] Creating Layer pool2 I0409 23:19:46.999266 8290 net.cpp:406] pool2 <- norm2 I0409 23:19:46.999271 8290 net.cpp:380] pool2 -> pool2 I0409 23:19:46.999298 8290 net.cpp:122] Setting up pool2 I0409 23:19:46.999303 8290 net.cpp:129] Top shape: 128 256 13 13 (5537792) I0409 23:19:46.999306 8290 net.cpp:137] Memory required for data: 869840896 I0409 23:19:46.999310 8290 layer_factory.hpp:77] Creating layer conv3 I0409 23:19:46.999320 8290 net.cpp:84] Creating Layer conv3 I0409 23:19:46.999322 8290 net.cpp:406] conv3 <- pool2 I0409 23:19:46.999327 8290 net.cpp:380] conv3 -> conv3 I0409 23:19:47.009213 8290 net.cpp:122] Setting up conv3 I0409 23:19:47.009232 8290 net.cpp:129] Top shape: 128 384 13 13 (8306688) I0409 23:19:47.009236 8290 net.cpp:137] Memory required for data: 903067648 I0409 23:19:47.009248 8290 layer_factory.hpp:77] Creating layer relu3 I0409 23:19:47.009256 8290 net.cpp:84] Creating Layer relu3 I0409 23:19:47.009260 8290 net.cpp:406] relu3 <- conv3 I0409 23:19:47.009265 8290 net.cpp:367] relu3 -> conv3 (in-place) I0409 23:19:47.009692 8290 net.cpp:122] Setting up relu3 I0409 23:19:47.009703 8290 net.cpp:129] Top shape: 128 384 13 13 (8306688) I0409 23:19:47.009706 8290 net.cpp:137] Memory required for data: 936294400 I0409 23:19:47.009711 8290 layer_factory.hpp:77] Creating layer conv4 I0409 23:19:47.009721 8290 net.cpp:84] Creating Layer conv4 I0409 23:19:47.009725 8290 net.cpp:406] conv4 <- conv3 I0409 23:19:47.009732 8290 net.cpp:380] conv4 -> conv4 I0409 23:19:47.019748 8290 net.cpp:122] Setting up conv4 I0409 23:19:47.019769 8290 net.cpp:129] Top shape: 128 384 13 13 (8306688) I0409 23:19:47.019773 8290 net.cpp:137] Memory required for data: 969521152 I0409 23:19:47.019783 8290 layer_factory.hpp:77] Creating layer relu4 I0409 23:19:47.019791 8290 net.cpp:84] Creating Layer relu4 I0409 23:19:47.019814 8290 net.cpp:406] relu4 <- conv4 I0409 23:19:47.019821 8290 net.cpp:367] relu4 -> conv4 (in-place) I0409 23:19:47.020102 8290 net.cpp:122] Setting up relu4 I0409 23:19:47.020112 8290 net.cpp:129] Top shape: 128 384 13 13 (8306688) I0409 23:19:47.020115 8290 net.cpp:137] Memory required for data: 1002747904 I0409 23:19:47.020119 8290 layer_factory.hpp:77] Creating layer conv5 I0409 23:19:47.020129 8290 net.cpp:84] Creating Layer conv5 I0409 23:19:47.020133 8290 net.cpp:406] conv5 <- conv4 I0409 23:19:47.020139 8290 net.cpp:380] conv5 -> conv5 I0409 23:19:47.032969 8290 net.cpp:122] Setting up conv5 I0409 23:19:47.032989 8290 net.cpp:129] Top shape: 128 256 13 13 (5537792) I0409 23:19:47.032991 8290 net.cpp:137] Memory required for data: 1024899072 I0409 23:19:47.033005 8290 layer_factory.hpp:77] Creating layer relu5 I0409 23:19:47.033015 8290 net.cpp:84] Creating Layer relu5 I0409 23:19:47.033020 8290 net.cpp:406] relu5 <- conv5 I0409 23:19:47.033025 8290 net.cpp:367] relu5 -> conv5 (in-place) I0409 23:19:47.033468 8290 net.cpp:122] Setting up relu5 I0409 23:19:47.033478 8290 net.cpp:129] Top shape: 128 256 13 13 (5537792) I0409 23:19:47.033481 8290 net.cpp:137] Memory required for data: 1047050240 I0409 23:19:47.033485 8290 layer_factory.hpp:77] Creating layer pool5 I0409 23:19:47.033493 8290 net.cpp:84] Creating Layer pool5 I0409 23:19:47.033496 8290 net.cpp:406] pool5 <- conv5 I0409 23:19:47.033501 8290 net.cpp:380] pool5 -> pool5 I0409 23:19:47.033542 8290 net.cpp:122] Setting up pool5 I0409 23:19:47.033547 8290 net.cpp:129] Top shape: 128 256 6 6 (1179648) I0409 23:19:47.033550 8290 net.cpp:137] Memory required for data: 1051768832 I0409 23:19:47.033555 8290 layer_factory.hpp:77] Creating layer fc6 I0409 23:19:47.033563 8290 net.cpp:84] Creating Layer fc6 I0409 23:19:47.033566 8290 net.cpp:406] fc6 <- pool5 I0409 23:19:47.033571 8290 net.cpp:380] fc6 -> fc6 I0409 23:19:47.122498 8290 net.cpp:122] Setting up fc6 I0409 23:19:47.122520 8290 net.cpp:129] Top shape: 128 1024 (131072) I0409 23:19:47.122524 8290 net.cpp:137] Memory required for data: 1052293120 I0409 23:19:47.122534 8290 layer_factory.hpp:77] Creating layer relu6 I0409 23:19:47.122542 8290 net.cpp:84] Creating Layer relu6 I0409 23:19:47.122547 8290 net.cpp:406] relu6 <- fc6 I0409 23:19:47.122553 8290 net.cpp:367] relu6 -> fc6 (in-place) I0409 23:19:47.123090 8290 net.cpp:122] Setting up relu6 I0409 23:19:47.123100 8290 net.cpp:129] Top shape: 128 1024 (131072) I0409 23:19:47.123102 8290 net.cpp:137] Memory required for data: 1052817408 I0409 23:19:47.123106 8290 layer_factory.hpp:77] Creating layer drop6 I0409 23:19:47.123113 8290 net.cpp:84] Creating Layer drop6 I0409 23:19:47.123117 8290 net.cpp:406] drop6 <- fc6 I0409 23:19:47.123122 8290 net.cpp:367] drop6 -> fc6 (in-place) I0409 23:19:47.123149 8290 net.cpp:122] Setting up drop6 I0409 23:19:47.123154 8290 net.cpp:129] Top shape: 128 1024 (131072) I0409 23:19:47.123157 8290 net.cpp:137] Memory required for data: 1053341696 I0409 23:19:47.123162 8290 layer_factory.hpp:77] Creating layer fc7 I0409 23:19:47.123168 8290 net.cpp:84] Creating Layer fc7 I0409 23:19:47.123173 8290 net.cpp:406] fc7 <- fc6 I0409 23:19:47.123178 8290 net.cpp:380] fc7 -> fc7 I0409 23:19:47.133177 8290 net.cpp:122] Setting up fc7 I0409 23:19:47.133196 8290 net.cpp:129] Top shape: 128 1024 (131072) I0409 23:19:47.133199 8290 net.cpp:137] Memory required for data: 1053865984 I0409 23:19:47.133209 8290 layer_factory.hpp:77] Creating layer relu7 I0409 23:19:47.133217 8290 net.cpp:84] Creating Layer relu7 I0409 23:19:47.133221 8290 net.cpp:406] relu7 <- fc7 I0409 23:19:47.133229 8290 net.cpp:367] relu7 -> fc7 (in-place) I0409 23:19:47.133780 8290 net.cpp:122] Setting up relu7 I0409 23:19:47.133790 8290 net.cpp:129] Top shape: 128 1024 (131072) I0409 23:19:47.133793 8290 net.cpp:137] Memory required for data: 1054390272 I0409 23:19:47.133798 8290 layer_factory.hpp:77] Creating layer drop7 I0409 23:19:47.133806 8290 net.cpp:84] Creating Layer drop7 I0409 23:19:47.133828 8290 net.cpp:406] drop7 <- fc7 I0409 23:19:47.133836 8290 net.cpp:367] drop7 -> fc7 (in-place) I0409 23:19:47.133862 8290 net.cpp:122] Setting up drop7 I0409 23:19:47.133868 8290 net.cpp:129] Top shape: 128 1024 (131072) I0409 23:19:47.133872 8290 net.cpp:137] Memory required for data: 1054914560 I0409 23:19:47.133875 8290 layer_factory.hpp:77] Creating layer fc8 I0409 23:19:47.133883 8290 net.cpp:84] Creating Layer fc8 I0409 23:19:47.133888 8290 net.cpp:406] fc8 <- fc7 I0409 23:19:47.133893 8290 net.cpp:380] fc8 -> fc8 I0409 23:19:47.135690 8290 net.cpp:122] Setting up fc8 I0409 23:19:47.135699 8290 net.cpp:129] Top shape: 128 196 (25088) I0409 23:19:47.135701 8290 net.cpp:137] Memory required for data: 1055014912 I0409 23:19:47.135707 8290 layer_factory.hpp:77] Creating layer loss I0409 23:19:47.135713 8290 net.cpp:84] Creating Layer loss I0409 23:19:47.135717 8290 net.cpp:406] loss <- fc8 I0409 23:19:47.135721 8290 net.cpp:406] loss <- label I0409 23:19:47.135728 8290 net.cpp:380] loss -> loss I0409 23:19:47.135738 8290 layer_factory.hpp:77] Creating layer loss I0409 23:19:47.136328 8290 net.cpp:122] Setting up loss I0409 23:19:47.136338 8290 net.cpp:129] Top shape: (1) I0409 23:19:47.136343 8290 net.cpp:132] with loss weight 1 I0409 23:19:47.136359 8290 net.cpp:137] Memory required for data: 1055014916 I0409 23:19:47.136364 8290 net.cpp:198] loss needs backward computation. I0409 23:19:47.136371 8290 net.cpp:198] fc8 needs backward computation. I0409 23:19:47.136375 8290 net.cpp:198] drop7 needs backward computation. I0409 23:19:47.136379 8290 net.cpp:198] relu7 needs backward computation. I0409 23:19:47.136381 8290 net.cpp:198] fc7 needs backward computation. I0409 23:19:47.136386 8290 net.cpp:198] drop6 needs backward computation. I0409 23:19:47.136389 8290 net.cpp:198] relu6 needs backward computation. I0409 23:19:47.136394 8290 net.cpp:198] fc6 needs backward computation. I0409 23:19:47.136397 8290 net.cpp:198] pool5 needs backward computation. I0409 23:19:47.136402 8290 net.cpp:198] relu5 needs backward computation. I0409 23:19:47.136406 8290 net.cpp:198] conv5 needs backward computation. I0409 23:19:47.136411 8290 net.cpp:198] relu4 needs backward computation. I0409 23:19:47.136415 8290 net.cpp:198] conv4 needs backward computation. I0409 23:19:47.136418 8290 net.cpp:198] relu3 needs backward computation. I0409 23:19:47.136422 8290 net.cpp:198] conv3 needs backward computation. I0409 23:19:47.136428 8290 net.cpp:198] pool2 needs backward computation. I0409 23:19:47.136432 8290 net.cpp:198] norm2 needs backward computation. I0409 23:19:47.136436 8290 net.cpp:198] relu2 needs backward computation. I0409 23:19:47.136441 8290 net.cpp:198] conv2 needs backward computation. I0409 23:19:47.136445 8290 net.cpp:198] pool1 needs backward computation. I0409 23:19:47.136449 8290 net.cpp:198] norm1 needs backward computation. I0409 23:19:47.136452 8290 net.cpp:198] relu1 needs backward computation. I0409 23:19:47.136456 8290 net.cpp:198] conv1 needs backward computation. I0409 23:19:47.136461 8290 net.cpp:200] train-data does not need backward computation. I0409 23:19:47.136464 8290 net.cpp:242] This network produces output loss I0409 23:19:47.136477 8290 net.cpp:255] Network initialization done. I0409 23:19:47.136993 8290 solver.cpp:172] Creating test net (#0) specified by net file: train_val.prototxt I0409 23:19:47.137025 8290 net.cpp:294] The NetState phase (1) differed from the phase (0) specified by a rule in layer train-data I0409 23:19:47.137171 8290 net.cpp:51] Initializing net from parameters: state { phase: TEST } layer { name: "val-data" type: "Data" top: "data" top: "label" include { phase: TEST } transform_param { crop_size: 227 mean_file: "/mnt/bigdisk/DIGITS-MAN-2/digits/jobs/20210407-214532-d396/mean.binaryproto" } data_param { source: "/mnt/bigdisk/DIGITS-MAN-2/digits/jobs/20210407-214532-d396/val_db" batch_size: 32 backend: LMDB } } layer { name: "conv1" type: "Convolution" bottom: "data" top: "conv1" param { lr_mult: 1 decay_mult: 1 } param { lr_mult: 2 decay_mult: 0 } convolution_param { num_output: 96 kernel_size: 11 stride: 4 weight_filler { type: "gaussian" std: 0.01 } bias_filler { type: "constant" value: 0 } } } layer { name: "relu1" type: "ReLU" bottom: "conv1" top: "conv1" } layer { name: "norm1" type: "LRN" bottom: "conv1" top: "norm1" lrn_param { local_size: 5 alpha: 0.0001 beta: 0.75 } } layer { name: "pool1" type: "Pooling" bottom: "norm1" top: "pool1" pooling_param { pool: MAX kernel_size: 3 stride: 2 } } layer { name: "conv2" type: "Convolution" bottom: "pool1" top: "conv2" param { lr_mult: 1 decay_mult: 1 } param { lr_mult: 2 decay_mult: 0 } convolution_param { num_output: 256 pad: 2 kernel_size: 5 group: 2 weight_filler { type: "gaussian" std: 0.01 } bias_filler { type: "constant" value: 0.1 } } } layer { name: "relu2" type: "ReLU" bottom: "conv2" top: "conv2" } layer { name: "norm2" type: "LRN" bottom: "conv2" top: "norm2" lrn_param { local_size: 5 alpha: 0.0001 beta: 0.75 } } layer { name: "pool2" type: "Pooling" bottom: "norm2" top: "pool2" pooling_param { pool: MAX kernel_size: 3 stride: 2 } } layer { name: "conv3" type: "Convolution" bottom: "pool2" top: "conv3" param { lr_mult: 1 decay_mult: 1 } param { lr_mult: 2 decay_mult: 0 } convolution_param { num_output: 384 pad: 1 kernel_size: 3 weight_filler { type: "gaussian" std: 0.01 } bias_filler { type: "constant" value: 0 } } } layer { name: "relu3" type: "ReLU" bottom: "conv3" top: "conv3" } layer { name: "conv4" type: "Convolution" bottom: "conv3" top: "conv4" param { lr_mult: 1 decay_mult: 1 } param { lr_mult: 2 decay_mult: 0 } convolution_param { num_output: 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: 1024 weight_filler { type: "gaussian" std: 0.005 } bias_filler { type: "constant" value: 0.1 } } } layer { name: "relu6" type: "ReLU" bottom: "fc6" top: "fc6" } layer { name: "drop6" type: "Dropout" bottom: "fc6" top: "fc6" dropout_param { dropout_ratio: 0.5 } } layer { name: "fc7" type: "InnerProduct" bottom: "fc6" top: "fc7" param { lr_mult: 1 decay_mult: 1 } param { lr_mult: 2 decay_mult: 0 } inner_product_param { num_output: 1024 weight_filler { type: "gaussian" std: 0.005 } bias_filler { type: "constant" value: 0.1 } } } layer { name: "relu7" type: "ReLU" bottom: "fc7" top: "fc7" } layer { name: "drop7" type: "Dropout" bottom: "fc7" top: "fc7" dropout_param { dropout_ratio: 0.5 } } layer { name: "fc8" type: "InnerProduct" bottom: "fc7" top: "fc8" param { lr_mult: 1 decay_mult: 1 } param { lr_mult: 2 decay_mult: 0 } inner_product_param { num_output: 196 weight_filler { type: "gaussian" std: 0.01 } bias_filler { type: "constant" value: 0 } } } layer { name: "accuracy" type: "Accuracy" bottom: "fc8" bottom: "label" top: "accuracy" include { phase: TEST } } layer { name: "loss" type: "SoftmaxWithLoss" bottom: "fc8" bottom: "label" top: "loss" } I0409 23:19:47.137265 8290 layer_factory.hpp:77] Creating layer val-data I0409 23:19:47.138947 8290 db_lmdb.cpp:35] Opened lmdb /mnt/bigdisk/DIGITS-MAN-2/digits/jobs/20210407-214532-d396/val_db I0409 23:19:47.139151 8290 net.cpp:84] Creating Layer val-data I0409 23:19:47.139160 8290 net.cpp:380] val-data -> data I0409 23:19:47.139171 8290 net.cpp:380] val-data -> label I0409 23:19:47.139178 8290 data_transformer.cpp:25] Loading mean file from: /mnt/bigdisk/DIGITS-MAN-2/digits/jobs/20210407-214532-d396/mean.binaryproto I0409 23:19:47.143074 8290 data_layer.cpp:45] output data size: 32,3,227,227 I0409 23:19:47.177220 8290 net.cpp:122] Setting up val-data I0409 23:19:47.177242 8290 net.cpp:129] Top shape: 32 3 227 227 (4946784) I0409 23:19:47.177248 8290 net.cpp:129] Top shape: 32 (32) I0409 23:19:47.177251 8290 net.cpp:137] Memory required for data: 19787264 I0409 23:19:47.177258 8290 layer_factory.hpp:77] Creating layer label_val-data_1_split I0409 23:19:47.177270 8290 net.cpp:84] Creating Layer label_val-data_1_split I0409 23:19:47.177275 8290 net.cpp:406] label_val-data_1_split <- label I0409 23:19:47.177282 8290 net.cpp:380] label_val-data_1_split -> label_val-data_1_split_0 I0409 23:19:47.177291 8290 net.cpp:380] label_val-data_1_split -> label_val-data_1_split_1 I0409 23:19:47.177345 8290 net.cpp:122] Setting up label_val-data_1_split I0409 23:19:47.177350 8290 net.cpp:129] Top shape: 32 (32) I0409 23:19:47.177353 8290 net.cpp:129] Top shape: 32 (32) I0409 23:19:47.177356 8290 net.cpp:137] Memory required for data: 19787520 I0409 23:19:47.177361 8290 layer_factory.hpp:77] Creating layer conv1 I0409 23:19:47.177373 8290 net.cpp:84] Creating Layer conv1 I0409 23:19:47.177376 8290 net.cpp:406] conv1 <- data I0409 23:19:47.177382 8290 net.cpp:380] conv1 -> conv1 I0409 23:19:47.179538 8290 net.cpp:122] Setting up conv1 I0409 23:19:47.179550 8290 net.cpp:129] Top shape: 32 96 55 55 (9292800) I0409 23:19:47.179554 8290 net.cpp:137] Memory required for data: 56958720 I0409 23:19:47.179565 8290 layer_factory.hpp:77] Creating layer relu1 I0409 23:19:47.179572 8290 net.cpp:84] Creating Layer relu1 I0409 23:19:47.179576 8290 net.cpp:406] relu1 <- conv1 I0409 23:19:47.179581 8290 net.cpp:367] relu1 -> conv1 (in-place) I0409 23:19:47.179872 8290 net.cpp:122] Setting up relu1 I0409 23:19:47.179880 8290 net.cpp:129] Top shape: 32 96 55 55 (9292800) I0409 23:19:47.179883 8290 net.cpp:137] Memory required for data: 94129920 I0409 23:19:47.179888 8290 layer_factory.hpp:77] Creating layer norm1 I0409 23:19:47.179895 8290 net.cpp:84] Creating Layer norm1 I0409 23:19:47.179900 8290 net.cpp:406] norm1 <- conv1 I0409 23:19:47.179905 8290 net.cpp:380] norm1 -> norm1 I0409 23:19:47.180354 8290 net.cpp:122] Setting up norm1 I0409 23:19:47.180363 8290 net.cpp:129] Top shape: 32 96 55 55 (9292800) I0409 23:19:47.180367 8290 net.cpp:137] Memory required for data: 131301120 I0409 23:19:47.180371 8290 layer_factory.hpp:77] Creating layer pool1 I0409 23:19:47.180378 8290 net.cpp:84] Creating Layer pool1 I0409 23:19:47.180382 8290 net.cpp:406] pool1 <- norm1 I0409 23:19:47.180387 8290 net.cpp:380] pool1 -> pool1 I0409 23:19:47.180415 8290 net.cpp:122] Setting up pool1 I0409 23:19:47.180421 8290 net.cpp:129] Top shape: 32 96 27 27 (2239488) I0409 23:19:47.180424 8290 net.cpp:137] Memory required for data: 140259072 I0409 23:19:47.180428 8290 layer_factory.hpp:77] Creating layer conv2 I0409 23:19:47.180435 8290 net.cpp:84] Creating Layer conv2 I0409 23:19:47.180439 8290 net.cpp:406] conv2 <- pool1 I0409 23:19:47.180464 8290 net.cpp:380] conv2 -> conv2 I0409 23:19:47.189268 8290 net.cpp:122] Setting up conv2 I0409 23:19:47.189288 8290 net.cpp:129] Top shape: 32 256 27 27 (5971968) I0409 23:19:47.189292 8290 net.cpp:137] Memory required for data: 164146944 I0409 23:19:47.189303 8290 layer_factory.hpp:77] Creating layer relu2 I0409 23:19:47.189312 8290 net.cpp:84] Creating Layer relu2 I0409 23:19:47.189316 8290 net.cpp:406] relu2 <- conv2 I0409 23:19:47.189323 8290 net.cpp:367] relu2 -> conv2 (in-place) I0409 23:19:47.190894 8290 net.cpp:122] Setting up relu2 I0409 23:19:47.190907 8290 net.cpp:129] Top shape: 32 256 27 27 (5971968) I0409 23:19:47.190910 8290 net.cpp:137] Memory required for data: 188034816 I0409 23:19:47.190914 8290 layer_factory.hpp:77] Creating layer norm2 I0409 23:19:47.190925 8290 net.cpp:84] Creating Layer norm2 I0409 23:19:47.190929 8290 net.cpp:406] norm2 <- conv2 I0409 23:19:47.190937 8290 net.cpp:380] norm2 -> norm2 I0409 23:19:47.191463 8290 net.cpp:122] Setting up norm2 I0409 23:19:47.191473 8290 net.cpp:129] Top shape: 32 256 27 27 (5971968) I0409 23:19:47.191478 8290 net.cpp:137] Memory required for data: 211922688 I0409 23:19:47.191481 8290 layer_factory.hpp:77] Creating layer pool2 I0409 23:19:47.191489 8290 net.cpp:84] Creating Layer pool2 I0409 23:19:47.191493 8290 net.cpp:406] pool2 <- norm2 I0409 23:19:47.191498 8290 net.cpp:380] pool2 -> pool2 I0409 23:19:47.191529 8290 net.cpp:122] Setting up pool2 I0409 23:19:47.191536 8290 net.cpp:129] Top shape: 32 256 13 13 (1384448) I0409 23:19:47.191540 8290 net.cpp:137] Memory required for data: 217460480 I0409 23:19:47.191543 8290 layer_factory.hpp:77] Creating layer conv3 I0409 23:19:47.191552 8290 net.cpp:84] Creating Layer conv3 I0409 23:19:47.191556 8290 net.cpp:406] conv3 <- pool2 I0409 23:19:47.191563 8290 net.cpp:380] conv3 -> conv3 I0409 23:19:47.202658 8290 net.cpp:122] Setting up conv3 I0409 23:19:47.202677 8290 net.cpp:129] Top shape: 32 384 13 13 (2076672) I0409 23:19:47.202680 8290 net.cpp:137] Memory required for data: 225767168 I0409 23:19:47.202693 8290 layer_factory.hpp:77] Creating layer relu3 I0409 23:19:47.202703 8290 net.cpp:84] Creating Layer relu3 I0409 23:19:47.202708 8290 net.cpp:406] relu3 <- conv3 I0409 23:19:47.202713 8290 net.cpp:367] relu3 -> conv3 (in-place) I0409 23:19:47.203219 8290 net.cpp:122] Setting up relu3 I0409 23:19:47.203230 8290 net.cpp:129] Top shape: 32 384 13 13 (2076672) I0409 23:19:47.203235 8290 net.cpp:137] Memory required for data: 234073856 I0409 23:19:47.203238 8290 layer_factory.hpp:77] Creating layer conv4 I0409 23:19:47.203250 8290 net.cpp:84] Creating Layer conv4 I0409 23:19:47.203254 8290 net.cpp:406] conv4 <- conv3 I0409 23:19:47.203260 8290 net.cpp:380] conv4 -> conv4 I0409 23:19:47.213644 8290 net.cpp:122] Setting up conv4 I0409 23:19:47.213663 8290 net.cpp:129] Top shape: 32 384 13 13 (2076672) I0409 23:19:47.213666 8290 net.cpp:137] Memory required for data: 242380544 I0409 23:19:47.213675 8290 layer_factory.hpp:77] Creating layer relu4 I0409 23:19:47.213687 8290 net.cpp:84] Creating Layer relu4 I0409 23:19:47.213691 8290 net.cpp:406] relu4 <- conv4 I0409 23:19:47.213698 8290 net.cpp:367] relu4 -> conv4 (in-place) I0409 23:19:47.214066 8290 net.cpp:122] Setting up relu4 I0409 23:19:47.214074 8290 net.cpp:129] Top shape: 32 384 13 13 (2076672) I0409 23:19:47.214078 8290 net.cpp:137] Memory required for data: 250687232 I0409 23:19:47.214082 8290 layer_factory.hpp:77] Creating layer conv5 I0409 23:19:47.214093 8290 net.cpp:84] Creating Layer conv5 I0409 23:19:47.214097 8290 net.cpp:406] conv5 <- conv4 I0409 23:19:47.214103 8290 net.cpp:380] conv5 -> conv5 I0409 23:19:47.229457 8290 net.cpp:122] Setting up conv5 I0409 23:19:47.229476 8290 net.cpp:129] Top shape: 32 256 13 13 (1384448) I0409 23:19:47.229480 8290 net.cpp:137] Memory required for data: 256225024 I0409 23:19:47.229493 8290 layer_factory.hpp:77] Creating layer relu5 I0409 23:19:47.229504 8290 net.cpp:84] Creating Layer relu5 I0409 23:19:47.229508 8290 net.cpp:406] relu5 <- conv5 I0409 23:19:47.229533 8290 net.cpp:367] relu5 -> conv5 (in-place) I0409 23:19:47.230054 8290 net.cpp:122] Setting up relu5 I0409 23:19:47.230063 8290 net.cpp:129] Top shape: 32 256 13 13 (1384448) I0409 23:19:47.230067 8290 net.cpp:137] Memory required for data: 261762816 I0409 23:19:47.230072 8290 layer_factory.hpp:77] Creating layer pool5 I0409 23:19:47.230082 8290 net.cpp:84] Creating Layer pool5 I0409 23:19:47.230084 8290 net.cpp:406] pool5 <- conv5 I0409 23:19:47.230090 8290 net.cpp:380] pool5 -> pool5 I0409 23:19:47.230130 8290 net.cpp:122] Setting up pool5 I0409 23:19:47.230136 8290 net.cpp:129] Top shape: 32 256 6 6 (294912) I0409 23:19:47.230139 8290 net.cpp:137] Memory required for data: 262942464 I0409 23:19:47.230142 8290 layer_factory.hpp:77] Creating layer fc6 I0409 23:19:47.230149 8290 net.cpp:84] Creating Layer fc6 I0409 23:19:47.230152 8290 net.cpp:406] fc6 <- pool5 I0409 23:19:47.230159 8290 net.cpp:380] fc6 -> fc6 I0409 23:19:47.319193 8290 net.cpp:122] Setting up fc6 I0409 23:19:47.319213 8290 net.cpp:129] Top shape: 32 1024 (32768) I0409 23:19:47.319216 8290 net.cpp:137] Memory required for data: 263073536 I0409 23:19:47.319226 8290 layer_factory.hpp:77] Creating layer relu6 I0409 23:19:47.319236 8290 net.cpp:84] Creating Layer relu6 I0409 23:19:47.319239 8290 net.cpp:406] relu6 <- fc6 I0409 23:19:47.319245 8290 net.cpp:367] relu6 -> fc6 (in-place) I0409 23:19:47.320091 8290 net.cpp:122] Setting up relu6 I0409 23:19:47.320101 8290 net.cpp:129] Top shape: 32 1024 (32768) I0409 23:19:47.320103 8290 net.cpp:137] Memory required for data: 263204608 I0409 23:19:47.320108 8290 layer_factory.hpp:77] Creating layer drop6 I0409 23:19:47.320114 8290 net.cpp:84] Creating Layer drop6 I0409 23:19:47.320118 8290 net.cpp:406] drop6 <- fc6 I0409 23:19:47.320127 8290 net.cpp:367] drop6 -> fc6 (in-place) I0409 23:19:47.320152 8290 net.cpp:122] Setting up drop6 I0409 23:19:47.320158 8290 net.cpp:129] Top shape: 32 1024 (32768) I0409 23:19:47.320161 8290 net.cpp:137] Memory required for data: 263335680 I0409 23:19:47.320164 8290 layer_factory.hpp:77] Creating layer fc7 I0409 23:19:47.320171 8290 net.cpp:84] Creating Layer fc7 I0409 23:19:47.320174 8290 net.cpp:406] fc7 <- fc6 I0409 23:19:47.320180 8290 net.cpp:380] fc7 -> fc7 I0409 23:19:47.330269 8290 net.cpp:122] Setting up fc7 I0409 23:19:47.330286 8290 net.cpp:129] Top shape: 32 1024 (32768) I0409 23:19:47.330289 8290 net.cpp:137] Memory required for data: 263466752 I0409 23:19:47.330298 8290 layer_factory.hpp:77] Creating layer relu7 I0409 23:19:47.330307 8290 net.cpp:84] Creating Layer relu7 I0409 23:19:47.330312 8290 net.cpp:406] relu7 <- fc7 I0409 23:19:47.330319 8290 net.cpp:367] relu7 -> fc7 (in-place) I0409 23:19:47.330749 8290 net.cpp:122] Setting up relu7 I0409 23:19:47.330757 8290 net.cpp:129] Top shape: 32 1024 (32768) I0409 23:19:47.330760 8290 net.cpp:137] Memory required for data: 263597824 I0409 23:19:47.330765 8290 layer_factory.hpp:77] Creating layer drop7 I0409 23:19:47.330770 8290 net.cpp:84] Creating Layer drop7 I0409 23:19:47.330773 8290 net.cpp:406] drop7 <- fc7 I0409 23:19:47.330778 8290 net.cpp:367] drop7 -> fc7 (in-place) I0409 23:19:47.330803 8290 net.cpp:122] Setting up drop7 I0409 23:19:47.330807 8290 net.cpp:129] Top shape: 32 1024 (32768) I0409 23:19:47.330811 8290 net.cpp:137] Memory required for data: 263728896 I0409 23:19:47.330813 8290 layer_factory.hpp:77] Creating layer fc8 I0409 23:19:47.330821 8290 net.cpp:84] Creating Layer fc8 I0409 23:19:47.330824 8290 net.cpp:406] fc8 <- fc7 I0409 23:19:47.330829 8290 net.cpp:380] fc8 -> fc8 I0409 23:19:47.332617 8290 net.cpp:122] Setting up fc8 I0409 23:19:47.332623 8290 net.cpp:129] Top shape: 32 196 (6272) I0409 23:19:47.332626 8290 net.cpp:137] Memory required for data: 263753984 I0409 23:19:47.332633 8290 layer_factory.hpp:77] Creating layer fc8_fc8_0_split I0409 23:19:47.332638 8290 net.cpp:84] Creating Layer fc8_fc8_0_split I0409 23:19:47.332641 8290 net.cpp:406] fc8_fc8_0_split <- fc8 I0409 23:19:47.332665 8290 net.cpp:380] fc8_fc8_0_split -> fc8_fc8_0_split_0 I0409 23:19:47.332672 8290 net.cpp:380] fc8_fc8_0_split -> fc8_fc8_0_split_1 I0409 23:19:47.332706 8290 net.cpp:122] Setting up fc8_fc8_0_split I0409 23:19:47.332710 8290 net.cpp:129] Top shape: 32 196 (6272) I0409 23:19:47.332715 8290 net.cpp:129] Top shape: 32 196 (6272) I0409 23:19:47.332717 8290 net.cpp:137] Memory required for data: 263804160 I0409 23:19:47.332720 8290 layer_factory.hpp:77] Creating layer accuracy I0409 23:19:47.332726 8290 net.cpp:84] Creating Layer accuracy I0409 23:19:47.332729 8290 net.cpp:406] accuracy <- fc8_fc8_0_split_0 I0409 23:19:47.332734 8290 net.cpp:406] accuracy <- label_val-data_1_split_0 I0409 23:19:47.332741 8290 net.cpp:380] accuracy -> accuracy I0409 23:19:47.332746 8290 net.cpp:122] Setting up accuracy I0409 23:19:47.332751 8290 net.cpp:129] Top shape: (1) I0409 23:19:47.332753 8290 net.cpp:137] Memory required for data: 263804164 I0409 23:19:47.332756 8290 layer_factory.hpp:77] Creating layer loss I0409 23:19:47.332762 8290 net.cpp:84] Creating Layer loss I0409 23:19:47.332765 8290 net.cpp:406] loss <- fc8_fc8_0_split_1 I0409 23:19:47.332769 8290 net.cpp:406] loss <- label_val-data_1_split_1 I0409 23:19:47.332777 8290 net.cpp:380] loss -> loss I0409 23:19:47.332783 8290 layer_factory.hpp:77] Creating layer loss I0409 23:19:47.338613 8290 net.cpp:122] Setting up loss I0409 23:19:47.338626 8290 net.cpp:129] Top shape: (1) I0409 23:19:47.338630 8290 net.cpp:132] with loss weight 1 I0409 23:19:47.338641 8290 net.cpp:137] Memory required for data: 263804168 I0409 23:19:47.338646 8290 net.cpp:198] loss needs backward computation. I0409 23:19:47.338651 8290 net.cpp:200] accuracy does not need backward computation. I0409 23:19:47.338656 8290 net.cpp:198] fc8_fc8_0_split needs backward computation. I0409 23:19:47.338660 8290 net.cpp:198] fc8 needs backward computation. I0409 23:19:47.338663 8290 net.cpp:198] drop7 needs backward computation. I0409 23:19:47.338667 8290 net.cpp:198] relu7 needs backward computation. I0409 23:19:47.338670 8290 net.cpp:198] fc7 needs backward computation. I0409 23:19:47.338675 8290 net.cpp:198] drop6 needs backward computation. I0409 23:19:47.338678 8290 net.cpp:198] relu6 needs backward computation. I0409 23:19:47.338682 8290 net.cpp:198] fc6 needs backward computation. I0409 23:19:47.338686 8290 net.cpp:198] pool5 needs backward computation. I0409 23:19:47.338690 8290 net.cpp:198] relu5 needs backward computation. I0409 23:19:47.338693 8290 net.cpp:198] conv5 needs backward computation. I0409 23:19:47.338697 8290 net.cpp:198] relu4 needs backward computation. I0409 23:19:47.338701 8290 net.cpp:198] conv4 needs backward computation. I0409 23:19:47.338704 8290 net.cpp:198] relu3 needs backward computation. I0409 23:19:47.338708 8290 net.cpp:198] conv3 needs backward computation. I0409 23:19:47.338711 8290 net.cpp:198] pool2 needs backward computation. I0409 23:19:47.338716 8290 net.cpp:198] norm2 needs backward computation. I0409 23:19:47.338719 8290 net.cpp:198] relu2 needs backward computation. I0409 23:19:47.338722 8290 net.cpp:198] conv2 needs backward computation. I0409 23:19:47.338726 8290 net.cpp:198] pool1 needs backward computation. I0409 23:19:47.338730 8290 net.cpp:198] norm1 needs backward computation. I0409 23:19:47.338733 8290 net.cpp:198] relu1 needs backward computation. I0409 23:19:47.338737 8290 net.cpp:198] conv1 needs backward computation. I0409 23:19:47.338742 8290 net.cpp:200] label_val-data_1_split does not need backward computation. I0409 23:19:47.338745 8290 net.cpp:200] val-data does not need backward computation. I0409 23:19:47.338748 8290 net.cpp:242] This network produces output accuracy I0409 23:19:47.338752 8290 net.cpp:242] This network produces output loss I0409 23:19:47.338770 8290 net.cpp:255] Network initialization done. I0409 23:19:47.338860 8290 solver.cpp:56] Solver scaffolding done. I0409 23:19:47.339293 8290 caffe.cpp:248] Starting Optimization I0409 23:19:47.339300 8290 solver.cpp:272] Solving I0409 23:19:47.339318 8290 solver.cpp:273] Learning Rate Policy: exp I0409 23:19:47.340405 8290 solver.cpp:330] Iteration 0, Testing net (#0) I0409 23:19:47.340415 8290 net.cpp:676] Ignoring source layer train-data I0409 23:19:47.365118 8290 blocking_queue.cpp:49] Waiting for data I0409 23:19:51.924713 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:19:51.968183 8290 solver.cpp:397] Test net output #0: accuracy = 0.00306373 I0409 23:19:51.968235 8290 solver.cpp:397] Test net output #1: loss = 5.27914 (* 1 = 5.27914 loss) I0409 23:19:52.049546 8290 solver.cpp:218] Iteration 0 (5.64535e+36 iter/s, 4.71002s/12 iters), loss = 5.27669 I0409 23:19:52.049598 8290 solver.cpp:237] Train net output #0: loss = 5.27669 (* 1 = 5.27669 loss) I0409 23:19:52.049618 8290 sgd_solver.cpp:105] Iteration 0, lr = 0.01 I0409 23:19:55.973853 8290 solver.cpp:218] Iteration 12 (3.05803 iter/s, 3.9241s/12 iters), loss = 5.2816 I0409 23:19:55.973908 8290 solver.cpp:237] Train net output #0: loss = 5.2816 (* 1 = 5.2816 loss) I0409 23:19:55.973920 8290 sgd_solver.cpp:105] Iteration 12, lr = 0.00997626 I0409 23:20:00.940675 8290 solver.cpp:218] Iteration 24 (2.41615 iter/s, 4.96658s/12 iters), loss = 5.27762 I0409 23:20:00.940727 8290 solver.cpp:237] Train net output #0: loss = 5.27762 (* 1 = 5.27762 loss) I0409 23:20:00.940739 8290 sgd_solver.cpp:105] Iteration 24, lr = 0.00995257 I0409 23:20:05.773468 8290 solver.cpp:218] Iteration 36 (2.48316 iter/s, 4.83256s/12 iters), loss = 5.28161 I0409 23:20:05.773511 8290 solver.cpp:237] Train net output #0: loss = 5.28161 (* 1 = 5.28161 loss) I0409 23:20:05.773519 8290 sgd_solver.cpp:105] Iteration 36, lr = 0.00992894 I0409 23:20:10.706547 8290 solver.cpp:218] Iteration 48 (2.43267 iter/s, 4.93284s/12 iters), loss = 5.28536 I0409 23:20:10.706599 8290 solver.cpp:237] Train net output #0: loss = 5.28536 (* 1 = 5.28536 loss) I0409 23:20:10.706610 8290 sgd_solver.cpp:105] Iteration 48, lr = 0.00990537 I0409 23:20:15.732256 8290 solver.cpp:218] Iteration 60 (2.38784 iter/s, 5.02547s/12 iters), loss = 5.27684 I0409 23:20:15.732304 8290 solver.cpp:237] Train net output #0: loss = 5.27684 (* 1 = 5.27684 loss) I0409 23:20:15.732314 8290 sgd_solver.cpp:105] Iteration 60, lr = 0.00988185 I0409 23:20:20.782860 8290 solver.cpp:218] Iteration 72 (2.37607 iter/s, 5.05037s/12 iters), loss = 5.28087 I0409 23:20:20.782936 8290 solver.cpp:237] Train net output #0: loss = 5.28087 (* 1 = 5.28087 loss) I0409 23:20:20.782944 8290 sgd_solver.cpp:105] Iteration 72, lr = 0.00985839 I0409 23:20:25.646050 8290 solver.cpp:218] Iteration 84 (2.46765 iter/s, 4.86293s/12 iters), loss = 5.28276 I0409 23:20:25.646096 8290 solver.cpp:237] Train net output #0: loss = 5.28276 (* 1 = 5.28276 loss) I0409 23:20:25.646104 8290 sgd_solver.cpp:105] Iteration 84, lr = 0.00983498 I0409 23:20:30.591168 8290 solver.cpp:218] Iteration 96 (2.42675 iter/s, 4.94488s/12 iters), loss = 5.2916 I0409 23:20:30.591224 8290 solver.cpp:237] Train net output #0: loss = 5.2916 (* 1 = 5.2916 loss) I0409 23:20:30.591235 8290 sgd_solver.cpp:105] Iteration 96, lr = 0.00981163 I0409 23:20:32.276285 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:20:32.580329 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_102.caffemodel I0409 23:20:33.298744 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_102.solverstate I0409 23:20:33.819922 8290 solver.cpp:330] Iteration 102, Testing net (#0) I0409 23:20:33.819943 8290 net.cpp:676] Ignoring source layer train-data I0409 23:20:38.314425 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:20:38.392498 8290 solver.cpp:397] Test net output #0: accuracy = 0.00551471 I0409 23:20:38.392532 8290 solver.cpp:397] Test net output #1: loss = 5.28154 (* 1 = 5.28154 loss) I0409 23:20:40.231236 8290 solver.cpp:218] Iteration 108 (1.24486 iter/s, 9.63967s/12 iters), loss = 5.2848 I0409 23:20:40.231276 8290 solver.cpp:237] Train net output #0: loss = 5.2848 (* 1 = 5.2848 loss) I0409 23:20:40.231284 8290 sgd_solver.cpp:105] Iteration 108, lr = 0.00978834 I0409 23:20:45.138468 8290 solver.cpp:218] Iteration 120 (2.44549 iter/s, 4.907s/12 iters), loss = 5.27712 I0409 23:20:45.138515 8290 solver.cpp:237] Train net output #0: loss = 5.27712 (* 1 = 5.27712 loss) I0409 23:20:45.138523 8290 sgd_solver.cpp:105] Iteration 120, lr = 0.0097651 I0409 23:20:49.984782 8290 solver.cpp:218] Iteration 132 (2.47623 iter/s, 4.84608s/12 iters), loss = 5.25285 I0409 23:20:49.984834 8290 solver.cpp:237] Train net output #0: loss = 5.25285 (* 1 = 5.25285 loss) I0409 23:20:49.984844 8290 sgd_solver.cpp:105] Iteration 132, lr = 0.00974192 I0409 23:20:54.844215 8290 solver.cpp:218] Iteration 144 (2.46954 iter/s, 4.85919s/12 iters), loss = 5.29752 I0409 23:20:54.844352 8290 solver.cpp:237] Train net output #0: loss = 5.29752 (* 1 = 5.29752 loss) I0409 23:20:54.844362 8290 sgd_solver.cpp:105] Iteration 144, lr = 0.00971879 I0409 23:20:59.759099 8290 solver.cpp:218] Iteration 156 (2.44173 iter/s, 4.91456s/12 iters), loss = 5.26319 I0409 23:20:59.759140 8290 solver.cpp:237] Train net output #0: loss = 5.26319 (* 1 = 5.26319 loss) I0409 23:20:59.759150 8290 sgd_solver.cpp:105] Iteration 156, lr = 0.00969571 I0409 23:21:04.966065 8290 solver.cpp:218] Iteration 168 (2.30471 iter/s, 5.20673s/12 iters), loss = 5.27089 I0409 23:21:04.966125 8290 solver.cpp:237] Train net output #0: loss = 5.27089 (* 1 = 5.27089 loss) I0409 23:21:04.966136 8290 sgd_solver.cpp:105] Iteration 168, lr = 0.00967269 I0409 23:21:09.825464 8290 solver.cpp:218] Iteration 180 (2.46957 iter/s, 4.85915s/12 iters), loss = 5.26295 I0409 23:21:09.825523 8290 solver.cpp:237] Train net output #0: loss = 5.26295 (* 1 = 5.26295 loss) I0409 23:21:09.825536 8290 sgd_solver.cpp:105] Iteration 180, lr = 0.00964973 I0409 23:21:14.754928 8290 solver.cpp:218] Iteration 192 (2.43446 iter/s, 4.92922s/12 iters), loss = 5.27637 I0409 23:21:14.754985 8290 solver.cpp:237] Train net output #0: loss = 5.27637 (* 1 = 5.27637 loss) I0409 23:21:14.754997 8290 sgd_solver.cpp:105] Iteration 192, lr = 0.00962682 I0409 23:21:18.518112 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:21:19.177662 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_204.caffemodel I0409 23:21:19.981514 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_204.solverstate I0409 23:21:20.597496 8290 solver.cpp:330] Iteration 204, Testing net (#0) I0409 23:21:20.597517 8290 net.cpp:676] Ignoring source layer train-data I0409 23:21:25.162750 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:21:25.337328 8290 solver.cpp:397] Test net output #0: accuracy = 0.00551471 I0409 23:21:25.337381 8290 solver.cpp:397] Test net output #1: loss = 5.28376 (* 1 = 5.28376 loss) I0409 23:21:25.419421 8290 solver.cpp:218] Iteration 204 (1.12528 iter/s, 10.6641s/12 iters), loss = 5.27461 I0409 23:21:25.419473 8290 solver.cpp:237] Train net output #0: loss = 5.27461 (* 1 = 5.27461 loss) I0409 23:21:25.419484 8290 sgd_solver.cpp:105] Iteration 204, lr = 0.00960396 I0409 23:21:29.560321 8290 solver.cpp:218] Iteration 216 (2.89807 iter/s, 4.14069s/12 iters), loss = 5.28212 I0409 23:21:29.560365 8290 solver.cpp:237] Train net output #0: loss = 5.28212 (* 1 = 5.28212 loss) I0409 23:21:29.560374 8290 sgd_solver.cpp:105] Iteration 216, lr = 0.00958116 I0409 23:21:34.442471 8290 solver.cpp:218] Iteration 228 (2.45805 iter/s, 4.88192s/12 iters), loss = 5.25867 I0409 23:21:34.442519 8290 solver.cpp:237] Train net output #0: loss = 5.25867 (* 1 = 5.25867 loss) I0409 23:21:34.442528 8290 sgd_solver.cpp:105] Iteration 228, lr = 0.00955841 I0409 23:21:39.386453 8290 solver.cpp:218] Iteration 240 (2.42731 iter/s, 4.94375s/12 iters), loss = 5.28863 I0409 23:21:39.386507 8290 solver.cpp:237] Train net output #0: loss = 5.28863 (* 1 = 5.28863 loss) I0409 23:21:39.386515 8290 sgd_solver.cpp:105] Iteration 240, lr = 0.00953572 I0409 23:21:44.323591 8290 solver.cpp:218] Iteration 252 (2.43068 iter/s, 4.93689s/12 iters), loss = 5.26954 I0409 23:21:44.323652 8290 solver.cpp:237] Train net output #0: loss = 5.26954 (* 1 = 5.26954 loss) I0409 23:21:44.323663 8290 sgd_solver.cpp:105] Iteration 252, lr = 0.00951308 I0409 23:21:49.872251 8290 solver.cpp:218] Iteration 264 (2.16279 iter/s, 5.54839s/12 iters), loss = 5.2717 I0409 23:21:49.872305 8290 solver.cpp:237] Train net output #0: loss = 5.2717 (* 1 = 5.2717 loss) I0409 23:21:49.872316 8290 sgd_solver.cpp:105] Iteration 264, lr = 0.00949049 I0409 23:21:55.425114 8290 solver.cpp:218] Iteration 276 (2.16115 iter/s, 5.5526s/12 iters), loss = 5.29481 I0409 23:21:55.425262 8290 solver.cpp:237] Train net output #0: loss = 5.29481 (* 1 = 5.29481 loss) I0409 23:21:55.425274 8290 sgd_solver.cpp:105] Iteration 276, lr = 0.00946796 I0409 23:22:00.365600 8290 solver.cpp:218] Iteration 288 (2.42908 iter/s, 4.94015s/12 iters), loss = 5.28299 I0409 23:22:00.365649 8290 solver.cpp:237] Train net output #0: loss = 5.28299 (* 1 = 5.28299 loss) I0409 23:22:00.365658 8290 sgd_solver.cpp:105] Iteration 288, lr = 0.00944548 I0409 23:22:05.361449 8290 solver.cpp:218] Iteration 300 (2.40211 iter/s, 4.99561s/12 iters), loss = 5.28051 I0409 23:22:05.361495 8290 solver.cpp:237] Train net output #0: loss = 5.28051 (* 1 = 5.28051 loss) I0409 23:22:05.361501 8290 sgd_solver.cpp:105] Iteration 300, lr = 0.00942305 I0409 23:22:06.352334 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:22:07.388548 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_306.caffemodel I0409 23:22:08.087924 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_306.solverstate I0409 23:22:08.604457 8290 solver.cpp:330] Iteration 306, Testing net (#0) I0409 23:22:08.604477 8290 net.cpp:676] Ignoring source layer train-data I0409 23:22:13.147574 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:22:13.305665 8290 solver.cpp:397] Test net output #0: accuracy = 0.00551471 I0409 23:22:13.305701 8290 solver.cpp:397] Test net output #1: loss = 5.28283 (* 1 = 5.28283 loss) I0409 23:22:15.167047 8290 solver.cpp:218] Iteration 312 (1.22384 iter/s, 9.8052s/12 iters), loss = 5.28122 I0409 23:22:15.167098 8290 solver.cpp:237] Train net output #0: loss = 5.28122 (* 1 = 5.28122 loss) I0409 23:22:15.167110 8290 sgd_solver.cpp:105] Iteration 312, lr = 0.00940068 I0409 23:22:20.446130 8290 solver.cpp:218] Iteration 324 (2.27323 iter/s, 5.27882s/12 iters), loss = 5.25074 I0409 23:22:20.446189 8290 solver.cpp:237] Train net output #0: loss = 5.25074 (* 1 = 5.25074 loss) I0409 23:22:20.446199 8290 sgd_solver.cpp:105] Iteration 324, lr = 0.00937836 I0409 23:22:25.345999 8290 solver.cpp:218] Iteration 336 (2.44917 iter/s, 4.89961s/12 iters), loss = 5.25652 I0409 23:22:25.346053 8290 solver.cpp:237] Train net output #0: loss = 5.25652 (* 1 = 5.25652 loss) I0409 23:22:25.346063 8290 sgd_solver.cpp:105] Iteration 336, lr = 0.0093561 I0409 23:22:30.169062 8290 solver.cpp:218] Iteration 348 (2.48817 iter/s, 4.82282s/12 iters), loss = 5.26414 I0409 23:22:30.169179 8290 solver.cpp:237] Train net output #0: loss = 5.26414 (* 1 = 5.26414 loss) I0409 23:22:30.169191 8290 sgd_solver.cpp:105] Iteration 348, lr = 0.00933388 I0409 23:22:35.111898 8290 solver.cpp:218] Iteration 360 (2.42791 iter/s, 4.94253s/12 iters), loss = 5.27855 I0409 23:22:35.111958 8290 solver.cpp:237] Train net output #0: loss = 5.27855 (* 1 = 5.27855 loss) I0409 23:22:35.111969 8290 sgd_solver.cpp:105] Iteration 360, lr = 0.00931172 I0409 23:22:40.025184 8290 solver.cpp:218] Iteration 372 (2.44248 iter/s, 4.91304s/12 iters), loss = 5.24325 I0409 23:22:40.025246 8290 solver.cpp:237] Train net output #0: loss = 5.24325 (* 1 = 5.24325 loss) I0409 23:22:40.025259 8290 sgd_solver.cpp:105] Iteration 372, lr = 0.00928961 I0409 23:22:45.013183 8290 solver.cpp:218] Iteration 384 (2.4059 iter/s, 4.98775s/12 iters), loss = 5.24091 I0409 23:22:45.013247 8290 solver.cpp:237] Train net output #0: loss = 5.24091 (* 1 = 5.24091 loss) I0409 23:22:45.013260 8290 sgd_solver.cpp:105] Iteration 384, lr = 0.00926756 I0409 23:22:49.951985 8290 solver.cpp:218] Iteration 396 (2.42987 iter/s, 4.93854s/12 iters), loss = 5.12189 I0409 23:22:49.952057 8290 solver.cpp:237] Train net output #0: loss = 5.12189 (* 1 = 5.12189 loss) I0409 23:22:49.952071 8290 sgd_solver.cpp:105] Iteration 396, lr = 0.00924556 I0409 23:22:53.038700 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:22:54.449772 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_408.caffemodel I0409 23:22:55.171120 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_408.solverstate I0409 23:22:55.723948 8290 solver.cpp:330] Iteration 408, Testing net (#0) I0409 23:22:55.723968 8290 net.cpp:676] Ignoring source layer train-data I0409 23:23:00.043753 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:23:00.256889 8290 solver.cpp:397] Test net output #0: accuracy = 0.00735294 I0409 23:23:00.257015 8290 solver.cpp:397] Test net output #1: loss = 5.1932 (* 1 = 5.1932 loss) I0409 23:23:00.339095 8290 solver.cpp:218] Iteration 408 (1.15533 iter/s, 10.3867s/12 iters), loss = 5.27157 I0409 23:23:00.339144 8290 solver.cpp:237] Train net output #0: loss = 5.27157 (* 1 = 5.27157 loss) I0409 23:23:00.339152 8290 sgd_solver.cpp:105] Iteration 408, lr = 0.00922361 I0409 23:23:04.824095 8290 solver.cpp:218] Iteration 420 (2.67572 iter/s, 4.48478s/12 iters), loss = 5.24984 I0409 23:23:04.824137 8290 solver.cpp:237] Train net output #0: loss = 5.24984 (* 1 = 5.24984 loss) I0409 23:23:04.824146 8290 sgd_solver.cpp:105] Iteration 420, lr = 0.00920171 I0409 23:23:09.768404 8290 solver.cpp:218] Iteration 432 (2.42715 iter/s, 4.94408s/12 iters), loss = 5.16829 I0409 23:23:09.768451 8290 solver.cpp:237] Train net output #0: loss = 5.16829 (* 1 = 5.16829 loss) I0409 23:23:09.768460 8290 sgd_solver.cpp:105] Iteration 432, lr = 0.00917986 I0409 23:23:14.580699 8290 solver.cpp:218] Iteration 444 (2.49373 iter/s, 4.81207s/12 iters), loss = 5.1603 I0409 23:23:14.580744 8290 solver.cpp:237] Train net output #0: loss = 5.1603 (* 1 = 5.1603 loss) I0409 23:23:14.580754 8290 sgd_solver.cpp:105] Iteration 444, lr = 0.00915807 I0409 23:23:19.507822 8290 solver.cpp:218] Iteration 456 (2.43561 iter/s, 4.92689s/12 iters), loss = 5.1878 I0409 23:23:19.507865 8290 solver.cpp:237] Train net output #0: loss = 5.1878 (* 1 = 5.1878 loss) I0409 23:23:19.507874 8290 sgd_solver.cpp:105] Iteration 456, lr = 0.00913632 I0409 23:23:24.476239 8290 solver.cpp:218] Iteration 468 (2.41537 iter/s, 4.96818s/12 iters), loss = 5.20535 I0409 23:23:24.476295 8290 solver.cpp:237] Train net output #0: loss = 5.20535 (* 1 = 5.20535 loss) I0409 23:23:24.476306 8290 sgd_solver.cpp:105] Iteration 468, lr = 0.00911463 I0409 23:23:29.411095 8290 solver.cpp:218] Iteration 480 (2.4318 iter/s, 4.93461s/12 iters), loss = 5.10144 I0409 23:23:29.411159 8290 solver.cpp:237] Train net output #0: loss = 5.10144 (* 1 = 5.10144 loss) I0409 23:23:29.411170 8290 sgd_solver.cpp:105] Iteration 480, lr = 0.00909299 I0409 23:23:34.370455 8290 solver.cpp:218] Iteration 492 (2.41979 iter/s, 4.95911s/12 iters), loss = 5.15032 I0409 23:23:34.370551 8290 solver.cpp:237] Train net output #0: loss = 5.15032 (* 1 = 5.15032 loss) I0409 23:23:34.370561 8290 sgd_solver.cpp:105] Iteration 492, lr = 0.0090714 I0409 23:23:39.406342 8290 solver.cpp:218] Iteration 504 (2.38303 iter/s, 5.0356s/12 iters), loss = 5.15177 I0409 23:23:39.406406 8290 solver.cpp:237] Train net output #0: loss = 5.15177 (* 1 = 5.15177 loss) I0409 23:23:39.406420 8290 sgd_solver.cpp:105] Iteration 504, lr = 0.00904986 I0409 23:23:39.654234 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:23:41.406615 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_510.caffemodel I0409 23:23:42.987236 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_510.solverstate I0409 23:23:44.890985 8290 solver.cpp:330] Iteration 510, Testing net (#0) I0409 23:23:44.891008 8290 net.cpp:676] Ignoring source layer train-data I0409 23:23:49.106227 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:23:49.351322 8290 solver.cpp:397] Test net output #0: accuracy = 0.0110294 I0409 23:23:49.351366 8290 solver.cpp:397] Test net output #1: loss = 5.12622 (* 1 = 5.12622 loss) I0409 23:23:51.396620 8290 solver.cpp:218] Iteration 516 (1.00085 iter/s, 11.9898s/12 iters), loss = 5.09345 I0409 23:23:51.396672 8290 solver.cpp:237] Train net output #0: loss = 5.09345 (* 1 = 5.09345 loss) I0409 23:23:51.396683 8290 sgd_solver.cpp:105] Iteration 516, lr = 0.00902838 I0409 23:23:56.422710 8290 solver.cpp:218] Iteration 528 (2.38766 iter/s, 5.02584s/12 iters), loss = 5.15411 I0409 23:23:56.422767 8290 solver.cpp:237] Train net output #0: loss = 5.15411 (* 1 = 5.15411 loss) I0409 23:23:56.422778 8290 sgd_solver.cpp:105] Iteration 528, lr = 0.00900694 I0409 23:24:01.359074 8290 solver.cpp:218] Iteration 540 (2.43106 iter/s, 4.93612s/12 iters), loss = 5.08132 I0409 23:24:01.359131 8290 solver.cpp:237] Train net output #0: loss = 5.08132 (* 1 = 5.08132 loss) I0409 23:24:01.359146 8290 sgd_solver.cpp:105] Iteration 540, lr = 0.00898556 I0409 23:24:06.385541 8290 solver.cpp:218] Iteration 552 (2.38748 iter/s, 5.02622s/12 iters), loss = 5.12524 I0409 23:24:06.385675 8290 solver.cpp:237] Train net output #0: loss = 5.12524 (* 1 = 5.12524 loss) I0409 23:24:06.385685 8290 sgd_solver.cpp:105] Iteration 552, lr = 0.00896423 I0409 23:24:11.314667 8290 solver.cpp:218] Iteration 564 (2.43467 iter/s, 4.92881s/12 iters), loss = 5.1088 I0409 23:24:11.314721 8290 solver.cpp:237] Train net output #0: loss = 5.1088 (* 1 = 5.1088 loss) I0409 23:24:11.314733 8290 sgd_solver.cpp:105] Iteration 564, lr = 0.00894294 I0409 23:24:16.341600 8290 solver.cpp:218] Iteration 576 (2.38726 iter/s, 5.02668s/12 iters), loss = 5.0716 I0409 23:24:16.341668 8290 solver.cpp:237] Train net output #0: loss = 5.0716 (* 1 = 5.0716 loss) I0409 23:24:16.341679 8290 sgd_solver.cpp:105] Iteration 576, lr = 0.00892171 I0409 23:24:21.174381 8290 solver.cpp:218] Iteration 588 (2.48317 iter/s, 4.83253s/12 iters), loss = 5.05626 I0409 23:24:21.174435 8290 solver.cpp:237] Train net output #0: loss = 5.05626 (* 1 = 5.05626 loss) I0409 23:24:21.174448 8290 sgd_solver.cpp:105] Iteration 588, lr = 0.00890053 I0409 23:24:26.040066 8290 solver.cpp:218] Iteration 600 (2.46637 iter/s, 4.86544s/12 iters), loss = 5.12093 I0409 23:24:26.040127 8290 solver.cpp:237] Train net output #0: loss = 5.12093 (* 1 = 5.12093 loss) I0409 23:24:26.040138 8290 sgd_solver.cpp:105] Iteration 600, lr = 0.0088794 I0409 23:24:28.350975 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:24:30.426426 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_612.caffemodel I0409 23:24:31.191913 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_612.solverstate I0409 23:24:31.744830 8290 solver.cpp:330] Iteration 612, Testing net (#0) I0409 23:24:31.744853 8290 net.cpp:676] Ignoring source layer train-data I0409 23:24:36.018237 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:24:36.306394 8290 solver.cpp:397] Test net output #0: accuracy = 0.0128676 I0409 23:24:36.306439 8290 solver.cpp:397] Test net output #1: loss = 5.07849 (* 1 = 5.07849 loss) I0409 23:24:36.388813 8290 solver.cpp:218] Iteration 612 (1.15961 iter/s, 10.3483s/12 iters), loss = 5.12088 I0409 23:24:36.388979 8290 solver.cpp:237] Train net output #0: loss = 5.12088 (* 1 = 5.12088 loss) I0409 23:24:36.388990 8290 sgd_solver.cpp:105] Iteration 612, lr = 0.00885831 I0409 23:24:40.538466 8290 solver.cpp:218] Iteration 624 (2.89203 iter/s, 4.14933s/12 iters), loss = 5.09892 I0409 23:24:40.538513 8290 solver.cpp:237] Train net output #0: loss = 5.09892 (* 1 = 5.09892 loss) I0409 23:24:40.538523 8290 sgd_solver.cpp:105] Iteration 624, lr = 0.00883728 I0409 23:24:45.376135 8290 solver.cpp:218] Iteration 636 (2.48065 iter/s, 4.83744s/12 iters), loss = 4.96814 I0409 23:24:45.376188 8290 solver.cpp:237] Train net output #0: loss = 4.96814 (* 1 = 4.96814 loss) I0409 23:24:45.376199 8290 sgd_solver.cpp:105] Iteration 636, lr = 0.0088163 I0409 23:24:50.291000 8290 solver.cpp:218] Iteration 648 (2.44169 iter/s, 4.91463s/12 iters), loss = 5.13027 I0409 23:24:50.291062 8290 solver.cpp:237] Train net output #0: loss = 5.13027 (* 1 = 5.13027 loss) I0409 23:24:50.291074 8290 sgd_solver.cpp:105] Iteration 648, lr = 0.00879537 I0409 23:24:55.148797 8290 solver.cpp:218] Iteration 660 (2.47038 iter/s, 4.85755s/12 iters), loss = 5.05457 I0409 23:24:55.148857 8290 solver.cpp:237] Train net output #0: loss = 5.05457 (* 1 = 5.05457 loss) I0409 23:24:55.148869 8290 sgd_solver.cpp:105] Iteration 660, lr = 0.00877449 I0409 23:25:00.245666 8290 solver.cpp:218] Iteration 672 (2.3545 iter/s, 5.09662s/12 iters), loss = 5.01062 I0409 23:25:00.245713 8290 solver.cpp:237] Train net output #0: loss = 5.01062 (* 1 = 5.01062 loss) I0409 23:25:00.245723 8290 sgd_solver.cpp:105] Iteration 672, lr = 0.00875366 I0409 23:25:04.780802 8290 blocking_queue.cpp:49] Waiting for data I0409 23:25:05.238978 8290 solver.cpp:218] Iteration 684 (2.40333 iter/s, 4.99307s/12 iters), loss = 4.87769 I0409 23:25:05.239025 8290 solver.cpp:237] Train net output #0: loss = 4.87769 (* 1 = 4.87769 loss) I0409 23:25:05.239034 8290 sgd_solver.cpp:105] Iteration 684, lr = 0.00873287 I0409 23:25:10.179200 8290 solver.cpp:218] Iteration 696 (2.42916 iter/s, 4.93998s/12 iters), loss = 4.98702 I0409 23:25:10.179293 8290 solver.cpp:237] Train net output #0: loss = 4.98702 (* 1 = 4.98702 loss) I0409 23:25:10.179302 8290 sgd_solver.cpp:105] Iteration 696, lr = 0.00871214 I0409 23:25:14.721011 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:25:15.098070 8290 solver.cpp:218] Iteration 708 (2.43972 iter/s, 4.91859s/12 iters), loss = 5.14826 I0409 23:25:15.098115 8290 solver.cpp:237] Train net output #0: loss = 5.14826 (* 1 = 5.14826 loss) I0409 23:25:15.098124 8290 sgd_solver.cpp:105] Iteration 708, lr = 0.00869145 I0409 23:25:17.101343 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_714.caffemodel I0409 23:25:18.233521 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_714.solverstate I0409 23:25:18.874177 8290 solver.cpp:330] Iteration 714, Testing net (#0) I0409 23:25:18.874202 8290 net.cpp:676] Ignoring source layer train-data I0409 23:25:23.040326 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:25:23.392906 8290 solver.cpp:397] Test net output #0: accuracy = 0.0134804 I0409 23:25:23.392940 8290 solver.cpp:397] Test net output #1: loss = 5.04969 (* 1 = 5.04969 loss) I0409 23:25:25.204324 8290 solver.cpp:218] Iteration 720 (1.18743 iter/s, 10.1058s/12 iters), loss = 5.14904 I0409 23:25:25.204372 8290 solver.cpp:237] Train net output #0: loss = 5.14904 (* 1 = 5.14904 loss) I0409 23:25:25.204382 8290 sgd_solver.cpp:105] Iteration 720, lr = 0.00867082 I0409 23:25:30.313688 8290 solver.cpp:218] Iteration 732 (2.34874 iter/s, 5.10912s/12 iters), loss = 4.90405 I0409 23:25:30.313742 8290 solver.cpp:237] Train net output #0: loss = 4.90405 (* 1 = 4.90405 loss) I0409 23:25:30.313753 8290 sgd_solver.cpp:105] Iteration 732, lr = 0.00865023 I0409 23:25:35.197059 8290 solver.cpp:218] Iteration 744 (2.45744 iter/s, 4.88313s/12 iters), loss = 5.02938 I0409 23:25:35.197105 8290 solver.cpp:237] Train net output #0: loss = 5.02938 (* 1 = 5.02938 loss) I0409 23:25:35.197113 8290 sgd_solver.cpp:105] Iteration 744, lr = 0.0086297 I0409 23:25:40.039484 8290 solver.cpp:218] Iteration 756 (2.47822 iter/s, 4.84219s/12 iters), loss = 5.06139 I0409 23:25:40.039542 8290 solver.cpp:237] Train net output #0: loss = 5.06139 (* 1 = 5.06139 loss) I0409 23:25:40.039554 8290 sgd_solver.cpp:105] Iteration 756, lr = 0.00860921 I0409 23:25:45.002554 8290 solver.cpp:218] Iteration 768 (2.41798 iter/s, 4.96282s/12 iters), loss = 5.05668 I0409 23:25:45.002713 8290 solver.cpp:237] Train net output #0: loss = 5.05668 (* 1 = 5.05668 loss) I0409 23:25:45.002727 8290 sgd_solver.cpp:105] Iteration 768, lr = 0.00858877 I0409 23:25:49.857067 8290 solver.cpp:218] Iteration 780 (2.4721 iter/s, 4.85418s/12 iters), loss = 5.10074 I0409 23:25:49.857112 8290 solver.cpp:237] Train net output #0: loss = 5.10074 (* 1 = 5.10074 loss) I0409 23:25:49.857120 8290 sgd_solver.cpp:105] Iteration 780, lr = 0.00856838 I0409 23:25:54.788230 8290 solver.cpp:218] Iteration 792 (2.43362 iter/s, 4.93093s/12 iters), loss = 4.91015 I0409 23:25:54.788278 8290 solver.cpp:237] Train net output #0: loss = 4.91015 (* 1 = 4.91015 loss) I0409 23:25:54.788290 8290 sgd_solver.cpp:105] Iteration 792, lr = 0.00854803 I0409 23:25:59.677273 8290 solver.cpp:218] Iteration 804 (2.45459 iter/s, 4.88881s/12 iters), loss = 4.98586 I0409 23:25:59.677320 8290 solver.cpp:237] Train net output #0: loss = 4.98586 (* 1 = 4.98586 loss) I0409 23:25:59.677330 8290 sgd_solver.cpp:105] Iteration 804, lr = 0.00852774 I0409 23:26:01.409255 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:26:04.242090 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_816.caffemodel I0409 23:26:04.992431 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_816.solverstate I0409 23:26:05.534564 8290 solver.cpp:330] Iteration 816, Testing net (#0) I0409 23:26:05.534593 8290 net.cpp:676] Ignoring source layer train-data I0409 23:26:09.633242 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:26:09.993831 8290 solver.cpp:397] Test net output #0: accuracy = 0.0196078 I0409 23:26:09.993880 8290 solver.cpp:397] Test net output #1: loss = 4.99197 (* 1 = 4.99197 loss) I0409 23:26:10.075881 8290 solver.cpp:218] Iteration 816 (1.15405 iter/s, 10.3982s/12 iters), loss = 5.0223 I0409 23:26:10.075949 8290 solver.cpp:237] Train net output #0: loss = 5.0223 (* 1 = 5.0223 loss) I0409 23:26:10.075963 8290 sgd_solver.cpp:105] Iteration 816, lr = 0.00850749 I0409 23:26:14.249557 8290 solver.cpp:218] Iteration 828 (2.87532 iter/s, 4.17345s/12 iters), loss = 5.11764 I0409 23:26:14.249612 8290 solver.cpp:237] Train net output #0: loss = 5.11764 (* 1 = 5.11764 loss) I0409 23:26:14.249624 8290 sgd_solver.cpp:105] Iteration 828, lr = 0.00848729 I0409 23:26:19.198470 8290 solver.cpp:218] Iteration 840 (2.42489 iter/s, 4.94867s/12 iters), loss = 4.90733 I0409 23:26:19.198571 8290 solver.cpp:237] Train net output #0: loss = 4.90733 (* 1 = 4.90733 loss) I0409 23:26:19.198580 8290 sgd_solver.cpp:105] Iteration 840, lr = 0.00846714 I0409 23:26:24.232456 8290 solver.cpp:218] Iteration 852 (2.38394 iter/s, 5.03369s/12 iters), loss = 4.94363 I0409 23:26:24.232506 8290 solver.cpp:237] Train net output #0: loss = 4.94363 (* 1 = 4.94363 loss) I0409 23:26:24.232515 8290 sgd_solver.cpp:105] Iteration 852, lr = 0.00844704 I0409 23:26:29.170835 8290 solver.cpp:218] Iteration 864 (2.43007 iter/s, 4.93814s/12 iters), loss = 4.96307 I0409 23:26:29.170892 8290 solver.cpp:237] Train net output #0: loss = 4.96307 (* 1 = 4.96307 loss) I0409 23:26:29.170903 8290 sgd_solver.cpp:105] Iteration 864, lr = 0.00842698 I0409 23:26:34.254690 8290 solver.cpp:218] Iteration 876 (2.36053 iter/s, 5.0836s/12 iters), loss = 4.99974 I0409 23:26:34.254743 8290 solver.cpp:237] Train net output #0: loss = 4.99974 (* 1 = 4.99974 loss) I0409 23:26:34.254755 8290 sgd_solver.cpp:105] Iteration 876, lr = 0.00840698 I0409 23:26:39.224880 8290 solver.cpp:218] Iteration 888 (2.41451 iter/s, 4.96995s/12 iters), loss = 4.87933 I0409 23:26:39.224932 8290 solver.cpp:237] Train net output #0: loss = 4.87933 (* 1 = 4.87933 loss) I0409 23:26:39.224942 8290 sgd_solver.cpp:105] Iteration 888, lr = 0.00838702 I0409 23:26:44.143494 8290 solver.cpp:218] Iteration 900 (2.43983 iter/s, 4.91838s/12 iters), loss = 4.98942 I0409 23:26:44.143546 8290 solver.cpp:237] Train net output #0: loss = 4.98942 (* 1 = 4.98942 loss) I0409 23:26:44.143558 8290 sgd_solver.cpp:105] Iteration 900, lr = 0.0083671 I0409 23:26:47.993121 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:26:49.090915 8290 solver.cpp:218] Iteration 912 (2.42562 iter/s, 4.94718s/12 iters), loss = 4.76947 I0409 23:26:49.090968 8290 solver.cpp:237] Train net output #0: loss = 4.76947 (* 1 = 4.76947 loss) I0409 23:26:49.090978 8290 sgd_solver.cpp:105] Iteration 912, lr = 0.00834724 I0409 23:26:51.091018 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_918.caffemodel I0409 23:26:51.823252 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_918.solverstate I0409 23:26:52.378254 8290 solver.cpp:330] Iteration 918, Testing net (#0) I0409 23:26:52.378278 8290 net.cpp:676] Ignoring source layer train-data I0409 23:26:56.350680 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:26:56.749975 8290 solver.cpp:397] Test net output #0: accuracy = 0.0245098 I0409 23:26:56.750025 8290 solver.cpp:397] Test net output #1: loss = 4.91046 (* 1 = 4.91046 loss) I0409 23:26:58.726446 8290 solver.cpp:218] Iteration 924 (1.24544 iter/s, 9.63513s/12 iters), loss = 4.97043 I0409 23:26:58.726495 8290 solver.cpp:237] Train net output #0: loss = 4.97043 (* 1 = 4.97043 loss) I0409 23:26:58.726505 8290 sgd_solver.cpp:105] Iteration 924, lr = 0.00832742 I0409 23:27:03.876677 8290 solver.cpp:218] Iteration 936 (2.3301 iter/s, 5.14999s/12 iters), loss = 4.93574 I0409 23:27:03.876726 8290 solver.cpp:237] Train net output #0: loss = 4.93574 (* 1 = 4.93574 loss) I0409 23:27:03.876737 8290 sgd_solver.cpp:105] Iteration 936, lr = 0.00830765 I0409 23:27:08.731297 8290 solver.cpp:218] Iteration 948 (2.47199 iter/s, 4.85439s/12 iters), loss = 4.8595 I0409 23:27:08.731343 8290 solver.cpp:237] Train net output #0: loss = 4.8595 (* 1 = 4.8595 loss) I0409 23:27:08.731353 8290 sgd_solver.cpp:105] Iteration 948, lr = 0.00828793 I0409 23:27:13.796998 8290 solver.cpp:218] Iteration 960 (2.36899 iter/s, 5.06546s/12 iters), loss = 4.72768 I0409 23:27:13.797056 8290 solver.cpp:237] Train net output #0: loss = 4.72768 (* 1 = 4.72768 loss) I0409 23:27:13.797067 8290 sgd_solver.cpp:105] Iteration 960, lr = 0.00826825 I0409 23:27:18.768213 8290 solver.cpp:218] Iteration 972 (2.41401 iter/s, 4.97097s/12 iters), loss = 4.9028 I0409 23:27:18.768262 8290 solver.cpp:237] Train net output #0: loss = 4.9028 (* 1 = 4.9028 loss) I0409 23:27:18.768273 8290 sgd_solver.cpp:105] Iteration 972, lr = 0.00824862 I0409 23:27:23.718634 8290 solver.cpp:218] Iteration 984 (2.42415 iter/s, 4.95018s/12 iters), loss = 4.95181 I0409 23:27:23.718768 8290 solver.cpp:237] Train net output #0: loss = 4.95181 (* 1 = 4.95181 loss) I0409 23:27:23.718781 8290 sgd_solver.cpp:105] Iteration 984, lr = 0.00822903 I0409 23:27:28.555603 8290 solver.cpp:218] Iteration 996 (2.48105 iter/s, 4.83665s/12 iters), loss = 4.70357 I0409 23:27:28.555660 8290 solver.cpp:237] Train net output #0: loss = 4.70357 (* 1 = 4.70357 loss) I0409 23:27:28.555671 8290 sgd_solver.cpp:105] Iteration 996, lr = 0.0082095 I0409 23:27:33.686547 8290 solver.cpp:218] Iteration 1008 (2.33887 iter/s, 5.13069s/12 iters), loss = 4.91445 I0409 23:27:33.686599 8290 solver.cpp:237] Train net output #0: loss = 4.91445 (* 1 = 4.91445 loss) I0409 23:27:33.686610 8290 sgd_solver.cpp:105] Iteration 1008, lr = 0.00819001 I0409 23:27:34.694994 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:27:38.194854 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1020.caffemodel I0409 23:27:39.462452 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1020.solverstate I0409 23:27:39.978704 8290 solver.cpp:330] Iteration 1020, Testing net (#0) I0409 23:27:39.978724 8290 net.cpp:676] Ignoring source layer train-data I0409 23:27:44.040923 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:27:44.474414 8290 solver.cpp:397] Test net output #0: accuracy = 0.026348 I0409 23:27:44.474444 8290 solver.cpp:397] Test net output #1: loss = 4.84341 (* 1 = 4.84341 loss) I0409 23:27:44.556373 8290 solver.cpp:218] Iteration 1020 (1.10402 iter/s, 10.8694s/12 iters), loss = 4.75265 I0409 23:27:44.556438 8290 solver.cpp:237] Train net output #0: loss = 4.75265 (* 1 = 4.75265 loss) I0409 23:27:44.556449 8290 sgd_solver.cpp:105] Iteration 1020, lr = 0.00817056 I0409 23:27:48.648869 8290 solver.cpp:218] Iteration 1032 (2.93236 iter/s, 4.09227s/12 iters), loss = 4.73618 I0409 23:27:48.648917 8290 solver.cpp:237] Train net output #0: loss = 4.73618 (* 1 = 4.73618 loss) I0409 23:27:48.648926 8290 sgd_solver.cpp:105] Iteration 1032, lr = 0.00815116 I0409 23:27:53.536084 8290 solver.cpp:218] Iteration 1044 (2.4555 iter/s, 4.88698s/12 iters), loss = 4.77176 I0409 23:27:53.536131 8290 solver.cpp:237] Train net output #0: loss = 4.77176 (* 1 = 4.77176 loss) I0409 23:27:53.536140 8290 sgd_solver.cpp:105] Iteration 1044, lr = 0.00813181 I0409 23:27:58.400857 8290 solver.cpp:218] Iteration 1056 (2.46683 iter/s, 4.86454s/12 iters), loss = 4.84565 I0409 23:27:58.401021 8290 solver.cpp:237] Train net output #0: loss = 4.84565 (* 1 = 4.84565 loss) I0409 23:27:58.401034 8290 sgd_solver.cpp:105] Iteration 1056, lr = 0.0081125 I0409 23:28:03.348965 8290 solver.cpp:218] Iteration 1068 (2.42534 iter/s, 4.94776s/12 iters), loss = 4.84731 I0409 23:28:03.349025 8290 solver.cpp:237] Train net output #0: loss = 4.84731 (* 1 = 4.84731 loss) I0409 23:28:03.349040 8290 sgd_solver.cpp:105] Iteration 1068, lr = 0.00809324 I0409 23:28:08.245748 8290 solver.cpp:218] Iteration 1080 (2.45071 iter/s, 4.89654s/12 iters), loss = 4.73938 I0409 23:28:08.245802 8290 solver.cpp:237] Train net output #0: loss = 4.73938 (* 1 = 4.73938 loss) I0409 23:28:08.245810 8290 sgd_solver.cpp:105] Iteration 1080, lr = 0.00807403 I0409 23:28:13.249998 8290 solver.cpp:218] Iteration 1092 (2.39809 iter/s, 5.00399s/12 iters), loss = 4.67332 I0409 23:28:13.250043 8290 solver.cpp:237] Train net output #0: loss = 4.67332 (* 1 = 4.67332 loss) I0409 23:28:13.250052 8290 sgd_solver.cpp:105] Iteration 1092, lr = 0.00805486 I0409 23:28:18.150962 8290 solver.cpp:218] Iteration 1104 (2.44861 iter/s, 4.90074s/12 iters), loss = 4.67712 I0409 23:28:18.151010 8290 solver.cpp:237] Train net output #0: loss = 4.67712 (* 1 = 4.67712 loss) I0409 23:28:18.151021 8290 sgd_solver.cpp:105] Iteration 1104, lr = 0.00803573 I0409 23:28:21.255975 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:28:23.057354 8290 solver.cpp:218] Iteration 1116 (2.44591 iter/s, 4.90615s/12 iters), loss = 4.65016 I0409 23:28:23.057404 8290 solver.cpp:237] Train net output #0: loss = 4.65016 (* 1 = 4.65016 loss) I0409 23:28:23.057413 8290 sgd_solver.cpp:105] Iteration 1116, lr = 0.00801666 I0409 23:28:25.083551 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1122.caffemodel I0409 23:28:25.834357 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1122.solverstate I0409 23:28:26.360682 8290 solver.cpp:330] Iteration 1122, Testing net (#0) I0409 23:28:26.360702 8290 net.cpp:676] Ignoring source layer train-data I0409 23:28:30.308708 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:28:30.789525 8290 solver.cpp:397] Test net output #0: accuracy = 0.0410539 I0409 23:28:30.789566 8290 solver.cpp:397] Test net output #1: loss = 4.69028 (* 1 = 4.69028 loss) I0409 23:28:32.640273 8290 solver.cpp:218] Iteration 1128 (1.25228 iter/s, 9.58252s/12 iters), loss = 4.85477 I0409 23:28:32.640321 8290 solver.cpp:237] Train net output #0: loss = 4.85477 (* 1 = 4.85477 loss) I0409 23:28:32.640331 8290 sgd_solver.cpp:105] Iteration 1128, lr = 0.00799762 I0409 23:28:37.577080 8290 solver.cpp:218] Iteration 1140 (2.43084 iter/s, 4.93657s/12 iters), loss = 4.75351 I0409 23:28:37.577138 8290 solver.cpp:237] Train net output #0: loss = 4.75351 (* 1 = 4.75351 loss) I0409 23:28:37.577150 8290 sgd_solver.cpp:105] Iteration 1140, lr = 0.00797863 I0409 23:28:42.475792 8290 solver.cpp:218] Iteration 1152 (2.44974 iter/s, 4.89847s/12 iters), loss = 4.56573 I0409 23:28:42.475845 8290 solver.cpp:237] Train net output #0: loss = 4.56573 (* 1 = 4.56573 loss) I0409 23:28:42.475857 8290 sgd_solver.cpp:105] Iteration 1152, lr = 0.00795969 I0409 23:28:47.341917 8290 solver.cpp:218] Iteration 1164 (2.46615 iter/s, 4.86588s/12 iters), loss = 4.62684 I0409 23:28:47.341997 8290 solver.cpp:237] Train net output #0: loss = 4.62684 (* 1 = 4.62684 loss) I0409 23:28:47.342010 8290 sgd_solver.cpp:105] Iteration 1164, lr = 0.00794079 I0409 23:28:52.166054 8290 solver.cpp:218] Iteration 1176 (2.48763 iter/s, 4.82387s/12 iters), loss = 4.65726 I0409 23:28:52.166110 8290 solver.cpp:237] Train net output #0: loss = 4.65726 (* 1 = 4.65726 loss) I0409 23:28:52.166121 8290 sgd_solver.cpp:105] Iteration 1176, lr = 0.00792194 I0409 23:28:57.043906 8290 solver.cpp:218] Iteration 1188 (2.46022 iter/s, 4.87761s/12 iters), loss = 4.5568 I0409 23:28:57.043947 8290 solver.cpp:237] Train net output #0: loss = 4.5568 (* 1 = 4.5568 loss) I0409 23:28:57.043956 8290 sgd_solver.cpp:105] Iteration 1188, lr = 0.00790313 I0409 23:29:01.944789 8290 solver.cpp:218] Iteration 1200 (2.44865 iter/s, 4.90066s/12 iters), loss = 4.67804 I0409 23:29:01.944903 8290 solver.cpp:237] Train net output #0: loss = 4.67804 (* 1 = 4.67804 loss) I0409 23:29:01.944912 8290 sgd_solver.cpp:105] Iteration 1200, lr = 0.00788437 I0409 23:29:06.866992 8290 solver.cpp:218] Iteration 1212 (2.43808 iter/s, 4.9219s/12 iters), loss = 4.61582 I0409 23:29:06.867040 8290 solver.cpp:237] Train net output #0: loss = 4.61582 (* 1 = 4.61582 loss) I0409 23:29:06.867050 8290 sgd_solver.cpp:105] Iteration 1212, lr = 0.00786565 I0409 23:29:07.151301 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:29:11.299476 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1224.caffemodel I0409 23:29:13.169529 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1224.solverstate I0409 23:29:14.635448 8290 solver.cpp:330] Iteration 1224, Testing net (#0) I0409 23:29:14.635475 8290 net.cpp:676] Ignoring source layer train-data I0409 23:29:18.435395 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:29:18.945602 8290 solver.cpp:397] Test net output #0: accuracy = 0.0520833 I0409 23:29:18.945665 8290 solver.cpp:397] Test net output #1: loss = 4.56958 (* 1 = 4.56958 loss) I0409 23:29:19.027751 8290 solver.cpp:218] Iteration 1224 (0.98682 iter/s, 12.1603s/12 iters), loss = 4.59611 I0409 23:29:19.027813 8290 solver.cpp:237] Train net output #0: loss = 4.59611 (* 1 = 4.59611 loss) I0409 23:29:19.027824 8290 sgd_solver.cpp:105] Iteration 1224, lr = 0.00784697 I0409 23:29:23.140089 8290 solver.cpp:218] Iteration 1236 (2.9182 iter/s, 4.11212s/12 iters), loss = 4.82215 I0409 23:29:23.140134 8290 solver.cpp:237] Train net output #0: loss = 4.82215 (* 1 = 4.82215 loss) I0409 23:29:23.140146 8290 sgd_solver.cpp:105] Iteration 1236, lr = 0.00782834 I0409 23:29:27.951359 8290 solver.cpp:218] Iteration 1248 (2.49426 iter/s, 4.81105s/12 iters), loss = 4.49676 I0409 23:29:27.951401 8290 solver.cpp:237] Train net output #0: loss = 4.49676 (* 1 = 4.49676 loss) I0409 23:29:27.951409 8290 sgd_solver.cpp:105] Iteration 1248, lr = 0.00780976 I0409 23:29:32.786814 8290 solver.cpp:218] Iteration 1260 (2.48179 iter/s, 4.83523s/12 iters), loss = 4.56647 I0409 23:29:32.786942 8290 solver.cpp:237] Train net output #0: loss = 4.56647 (* 1 = 4.56647 loss) I0409 23:29:32.786953 8290 sgd_solver.cpp:105] Iteration 1260, lr = 0.00779122 I0409 23:29:37.641471 8290 solver.cpp:218] Iteration 1272 (2.47201 iter/s, 4.85435s/12 iters), loss = 4.44183 I0409 23:29:37.641521 8290 solver.cpp:237] Train net output #0: loss = 4.44183 (* 1 = 4.44183 loss) I0409 23:29:37.641533 8290 sgd_solver.cpp:105] Iteration 1272, lr = 0.00777272 I0409 23:29:42.544235 8290 solver.cpp:218] Iteration 1284 (2.44771 iter/s, 4.90253s/12 iters), loss = 4.46105 I0409 23:29:42.544282 8290 solver.cpp:237] Train net output #0: loss = 4.46105 (* 1 = 4.46105 loss) I0409 23:29:42.544292 8290 sgd_solver.cpp:105] Iteration 1284, lr = 0.00775426 I0409 23:29:47.376801 8290 solver.cpp:218] Iteration 1296 (2.48327 iter/s, 4.83233s/12 iters), loss = 4.36804 I0409 23:29:47.376866 8290 solver.cpp:237] Train net output #0: loss = 4.36804 (* 1 = 4.36804 loss) I0409 23:29:47.376878 8290 sgd_solver.cpp:105] Iteration 1296, lr = 0.00773585 I0409 23:29:52.216853 8290 solver.cpp:218] Iteration 1308 (2.47944 iter/s, 4.83981s/12 iters), loss = 4.4353 I0409 23:29:52.216907 8290 solver.cpp:237] Train net output #0: loss = 4.4353 (* 1 = 4.4353 loss) I0409 23:29:52.216918 8290 sgd_solver.cpp:105] Iteration 1308, lr = 0.00771749 I0409 23:29:54.624832 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:29:57.030966 8290 solver.cpp:218] Iteration 1320 (2.49279 iter/s, 4.81388s/12 iters), loss = 4.4104 I0409 23:29:57.031023 8290 solver.cpp:237] Train net output #0: loss = 4.4104 (* 1 = 4.4104 loss) I0409 23:29:57.031034 8290 sgd_solver.cpp:105] Iteration 1320, lr = 0.00769916 I0409 23:29:59.004508 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1326.caffemodel I0409 23:29:59.760129 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1326.solverstate I0409 23:30:00.284618 8290 solver.cpp:330] Iteration 1326, Testing net (#0) I0409 23:30:00.284644 8290 net.cpp:676] Ignoring source layer train-data I0409 23:30:04.222672 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:30:04.779300 8290 solver.cpp:397] Test net output #0: accuracy = 0.0594363 I0409 23:30:04.779350 8290 solver.cpp:397] Test net output #1: loss = 4.39001 (* 1 = 4.39001 loss) I0409 23:30:06.697471 8290 solver.cpp:218] Iteration 1332 (1.24145 iter/s, 9.6661s/12 iters), loss = 4.28586 I0409 23:30:06.697527 8290 solver.cpp:237] Train net output #0: loss = 4.28586 (* 1 = 4.28586 loss) I0409 23:30:06.697540 8290 sgd_solver.cpp:105] Iteration 1332, lr = 0.00768088 I0409 23:30:11.579267 8290 solver.cpp:218] Iteration 1344 (2.45824 iter/s, 4.88155s/12 iters), loss = 4.2958 I0409 23:30:11.579325 8290 solver.cpp:237] Train net output #0: loss = 4.2958 (* 1 = 4.2958 loss) I0409 23:30:11.579337 8290 sgd_solver.cpp:105] Iteration 1344, lr = 0.00766265 I0409 23:30:16.455214 8290 solver.cpp:218] Iteration 1356 (2.46118 iter/s, 4.87571s/12 iters), loss = 4.55699 I0409 23:30:16.455258 8290 solver.cpp:237] Train net output #0: loss = 4.55699 (* 1 = 4.55699 loss) I0409 23:30:16.455268 8290 sgd_solver.cpp:105] Iteration 1356, lr = 0.00764446 I0409 23:30:21.256958 8290 solver.cpp:218] Iteration 1368 (2.49922 iter/s, 4.80151s/12 iters), loss = 4.5433 I0409 23:30:21.257016 8290 solver.cpp:237] Train net output #0: loss = 4.5433 (* 1 = 4.5433 loss) I0409 23:30:21.257028 8290 sgd_solver.cpp:105] Iteration 1368, lr = 0.00762631 I0409 23:30:21.257306 8290 blocking_queue.cpp:49] Waiting for data I0409 23:30:26.094386 8290 solver.cpp:218] Iteration 1380 (2.48078 iter/s, 4.83719s/12 iters), loss = 4.29493 I0409 23:30:26.094434 8290 solver.cpp:237] Train net output #0: loss = 4.29493 (* 1 = 4.29493 loss) I0409 23:30:26.094442 8290 sgd_solver.cpp:105] Iteration 1380, lr = 0.0076082 I0409 23:30:30.939294 8290 solver.cpp:218] Iteration 1392 (2.47695 iter/s, 4.84468s/12 iters), loss = 4.31019 I0409 23:30:30.939340 8290 solver.cpp:237] Train net output #0: loss = 4.31019 (* 1 = 4.31019 loss) I0409 23:30:30.939349 8290 sgd_solver.cpp:105] Iteration 1392, lr = 0.00759014 I0409 23:30:35.783933 8290 solver.cpp:218] Iteration 1404 (2.47708 iter/s, 4.84441s/12 iters), loss = 4.40704 I0409 23:30:35.784130 8290 solver.cpp:237] Train net output #0: loss = 4.40704 (* 1 = 4.40704 loss) I0409 23:30:35.784145 8290 sgd_solver.cpp:105] Iteration 1404, lr = 0.00757212 I0409 23:30:40.363777 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:30:40.716806 8290 solver.cpp:218] Iteration 1416 (2.43284 iter/s, 4.9325s/12 iters), loss = 4.29338 I0409 23:30:40.716856 8290 solver.cpp:237] Train net output #0: loss = 4.29338 (* 1 = 4.29338 loss) I0409 23:30:40.716868 8290 sgd_solver.cpp:105] Iteration 1416, lr = 0.00755414 I0409 23:30:45.132454 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1428.caffemodel I0409 23:30:45.857223 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1428.solverstate I0409 23:30:46.394294 8290 solver.cpp:330] Iteration 1428, Testing net (#0) I0409 23:30:46.394320 8290 net.cpp:676] Ignoring source layer train-data I0409 23:30:50.234884 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:30:50.823966 8290 solver.cpp:397] Test net output #0: accuracy = 0.0833333 I0409 23:30:50.824012 8290 solver.cpp:397] Test net output #1: loss = 4.2649 (* 1 = 4.2649 loss) I0409 23:30:50.906538 8290 solver.cpp:218] Iteration 1428 (1.1777 iter/s, 10.1893s/12 iters), loss = 4.30635 I0409 23:30:50.906613 8290 solver.cpp:237] Train net output #0: loss = 4.30635 (* 1 = 4.30635 loss) I0409 23:30:50.906628 8290 sgd_solver.cpp:105] Iteration 1428, lr = 0.0075362 I0409 23:30:54.988200 8290 solver.cpp:218] Iteration 1440 (2.94014 iter/s, 4.08143s/12 iters), loss = 4.27452 I0409 23:30:54.988266 8290 solver.cpp:237] Train net output #0: loss = 4.27452 (* 1 = 4.27452 loss) I0409 23:30:54.988277 8290 sgd_solver.cpp:105] Iteration 1440, lr = 0.00751831 I0409 23:30:59.786989 8290 solver.cpp:218] Iteration 1452 (2.50076 iter/s, 4.79854s/12 iters), loss = 4.3055 I0409 23:30:59.787046 8290 solver.cpp:237] Train net output #0: loss = 4.3055 (* 1 = 4.3055 loss) I0409 23:30:59.787060 8290 sgd_solver.cpp:105] Iteration 1452, lr = 0.00750046 I0409 23:31:04.615373 8290 solver.cpp:218] Iteration 1464 (2.48543 iter/s, 4.82814s/12 iters), loss = 4.35901 I0409 23:31:04.615437 8290 solver.cpp:237] Train net output #0: loss = 4.35901 (* 1 = 4.35901 loss) I0409 23:31:04.615449 8290 sgd_solver.cpp:105] Iteration 1464, lr = 0.00748265 I0409 23:31:09.453071 8290 solver.cpp:218] Iteration 1476 (2.48064 iter/s, 4.83746s/12 iters), loss = 4.32005 I0409 23:31:09.453153 8290 solver.cpp:237] Train net output #0: loss = 4.32005 (* 1 = 4.32005 loss) I0409 23:31:09.453164 8290 sgd_solver.cpp:105] Iteration 1476, lr = 0.00746489 I0409 23:31:14.246240 8290 solver.cpp:218] Iteration 1488 (2.5037 iter/s, 4.7929s/12 iters), loss = 4.21909 I0409 23:31:14.246296 8290 solver.cpp:237] Train net output #0: loss = 4.21909 (* 1 = 4.21909 loss) I0409 23:31:14.246309 8290 sgd_solver.cpp:105] Iteration 1488, lr = 0.00744716 I0409 23:31:19.067498 8290 solver.cpp:218] Iteration 1500 (2.4891 iter/s, 4.82102s/12 iters), loss = 4.19458 I0409 23:31:19.067554 8290 solver.cpp:237] Train net output #0: loss = 4.19458 (* 1 = 4.19458 loss) I0409 23:31:19.067565 8290 sgd_solver.cpp:105] Iteration 1500, lr = 0.00742948 I0409 23:31:23.871696 8290 solver.cpp:218] Iteration 1512 (2.49794 iter/s, 4.80396s/12 iters), loss = 4.38337 I0409 23:31:23.871757 8290 solver.cpp:237] Train net output #0: loss = 4.38337 (* 1 = 4.38337 loss) I0409 23:31:23.871767 8290 sgd_solver.cpp:105] Iteration 1512, lr = 0.00741184 I0409 23:31:25.642429 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:31:28.821491 8290 solver.cpp:218] Iteration 1524 (2.42446 iter/s, 4.94955s/12 iters), loss = 4.24504 I0409 23:31:28.821539 8290 solver.cpp:237] Train net output #0: loss = 4.24504 (* 1 = 4.24504 loss) I0409 23:31:28.821550 8290 sgd_solver.cpp:105] Iteration 1524, lr = 0.00739425 I0409 23:31:30.830524 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1530.caffemodel I0409 23:31:31.611694 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1530.solverstate I0409 23:31:32.154698 8290 solver.cpp:330] Iteration 1530, Testing net (#0) I0409 23:31:32.154726 8290 net.cpp:676] Ignoring source layer train-data I0409 23:31:36.103706 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:31:36.789288 8290 solver.cpp:397] Test net output #0: accuracy = 0.0925245 I0409 23:31:36.789340 8290 solver.cpp:397] Test net output #1: loss = 4.14525 (* 1 = 4.14525 loss) I0409 23:31:38.512797 8290 solver.cpp:218] Iteration 1536 (1.23827 iter/s, 9.69091s/12 iters), loss = 4.31062 I0409 23:31:38.512840 8290 solver.cpp:237] Train net output #0: loss = 4.31062 (* 1 = 4.31062 loss) I0409 23:31:38.512848 8290 sgd_solver.cpp:105] Iteration 1536, lr = 0.00737669 I0409 23:31:43.371649 8290 solver.cpp:218] Iteration 1548 (2.46983 iter/s, 4.85862s/12 iters), loss = 3.86418 I0409 23:31:43.371767 8290 solver.cpp:237] Train net output #0: loss = 3.86418 (* 1 = 3.86418 loss) I0409 23:31:43.371775 8290 sgd_solver.cpp:105] Iteration 1548, lr = 0.00735918 I0409 23:31:48.266250 8290 solver.cpp:218] Iteration 1560 (2.45183 iter/s, 4.8943s/12 iters), loss = 4.20813 I0409 23:31:48.266299 8290 solver.cpp:237] Train net output #0: loss = 4.20813 (* 1 = 4.20813 loss) I0409 23:31:48.266307 8290 sgd_solver.cpp:105] Iteration 1560, lr = 0.00734171 I0409 23:31:53.236055 8290 solver.cpp:218] Iteration 1572 (2.4147 iter/s, 4.96956s/12 iters), loss = 4.05597 I0409 23:31:53.236115 8290 solver.cpp:237] Train net output #0: loss = 4.05597 (* 1 = 4.05597 loss) I0409 23:31:53.236129 8290 sgd_solver.cpp:105] Iteration 1572, lr = 0.00732427 I0409 23:31:58.109220 8290 solver.cpp:218] Iteration 1584 (2.46259 iter/s, 4.87292s/12 iters), loss = 4.16828 I0409 23:31:58.109277 8290 solver.cpp:237] Train net output #0: loss = 4.16828 (* 1 = 4.16828 loss) I0409 23:31:58.109288 8290 sgd_solver.cpp:105] Iteration 1584, lr = 0.00730688 I0409 23:32:03.032187 8290 solver.cpp:218] Iteration 1596 (2.43768 iter/s, 4.92272s/12 iters), loss = 4.0268 I0409 23:32:03.032241 8290 solver.cpp:237] Train net output #0: loss = 4.0268 (* 1 = 4.0268 loss) I0409 23:32:03.032253 8290 sgd_solver.cpp:105] Iteration 1596, lr = 0.00728954 I0409 23:32:07.969082 8290 solver.cpp:218] Iteration 1608 (2.43079 iter/s, 4.93666s/12 iters), loss = 4.32273 I0409 23:32:07.969131 8290 solver.cpp:237] Train net output #0: loss = 4.32273 (* 1 = 4.32273 loss) I0409 23:32:07.969142 8290 sgd_solver.cpp:105] Iteration 1608, lr = 0.00727223 I0409 23:32:11.770505 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:32:12.848798 8290 solver.cpp:218] Iteration 1620 (2.45928 iter/s, 4.87948s/12 iters), loss = 4.14153 I0409 23:32:12.848845 8290 solver.cpp:237] Train net output #0: loss = 4.14153 (* 1 = 4.14153 loss) I0409 23:32:12.848853 8290 sgd_solver.cpp:105] Iteration 1620, lr = 0.00725496 I0409 23:32:17.349336 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1632.caffemodel I0409 23:32:18.056066 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1632.solverstate I0409 23:32:18.592262 8290 solver.cpp:330] Iteration 1632, Testing net (#0) I0409 23:32:18.592289 8290 net.cpp:676] Ignoring source layer train-data I0409 23:32:22.471453 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:32:23.162001 8290 solver.cpp:397] Test net output #0: accuracy = 0.106618 I0409 23:32:23.162042 8290 solver.cpp:397] Test net output #1: loss = 4.07655 (* 1 = 4.07655 loss) I0409 23:32:23.244151 8290 solver.cpp:218] Iteration 1632 (1.15441 iter/s, 10.3949s/12 iters), loss = 3.96122 I0409 23:32:23.244202 8290 solver.cpp:237] Train net output #0: loss = 3.96122 (* 1 = 3.96122 loss) I0409 23:32:23.244212 8290 sgd_solver.cpp:105] Iteration 1632, lr = 0.00723774 I0409 23:32:27.439003 8290 solver.cpp:218] Iteration 1644 (2.86079 iter/s, 4.19464s/12 iters), loss = 4.08151 I0409 23:32:27.439043 8290 solver.cpp:237] Train net output #0: loss = 4.08151 (* 1 = 4.08151 loss) I0409 23:32:27.439051 8290 sgd_solver.cpp:105] Iteration 1644, lr = 0.00722056 I0409 23:32:32.327906 8290 solver.cpp:218] Iteration 1656 (2.45465 iter/s, 4.88868s/12 iters), loss = 4.11488 I0409 23:32:32.327960 8290 solver.cpp:237] Train net output #0: loss = 4.11488 (* 1 = 4.11488 loss) I0409 23:32:32.327972 8290 sgd_solver.cpp:105] Iteration 1656, lr = 0.00720341 I0409 23:32:37.232858 8290 solver.cpp:218] Iteration 1668 (2.44662 iter/s, 4.90472s/12 iters), loss = 3.87917 I0409 23:32:37.232899 8290 solver.cpp:237] Train net output #0: loss = 3.87917 (* 1 = 3.87917 loss) I0409 23:32:37.232908 8290 sgd_solver.cpp:105] Iteration 1668, lr = 0.00718631 I0409 23:32:42.140982 8290 solver.cpp:218] Iteration 1680 (2.44504 iter/s, 4.9079s/12 iters), loss = 3.99036 I0409 23:32:42.141029 8290 solver.cpp:237] Train net output #0: loss = 3.99036 (* 1 = 3.99036 loss) I0409 23:32:42.141039 8290 sgd_solver.cpp:105] Iteration 1680, lr = 0.00716925 I0409 23:32:47.341327 8290 solver.cpp:218] Iteration 1692 (2.30764 iter/s, 5.20011s/12 iters), loss = 3.95914 I0409 23:32:47.341372 8290 solver.cpp:237] Train net output #0: loss = 3.95914 (* 1 = 3.95914 loss) I0409 23:32:47.341382 8290 sgd_solver.cpp:105] Iteration 1692, lr = 0.00715223 I0409 23:32:52.276499 8290 solver.cpp:218] Iteration 1704 (2.43164 iter/s, 4.93494s/12 iters), loss = 3.90232 I0409 23:32:52.276624 8290 solver.cpp:237] Train net output #0: loss = 3.90232 (* 1 = 3.90232 loss) I0409 23:32:52.276633 8290 sgd_solver.cpp:105] Iteration 1704, lr = 0.00713525 I0409 23:32:57.151278 8290 solver.cpp:218] Iteration 1716 (2.46181 iter/s, 4.87447s/12 iters), loss = 4.00772 I0409 23:32:57.151327 8290 solver.cpp:237] Train net output #0: loss = 4.00772 (* 1 = 4.00772 loss) I0409 23:32:57.151340 8290 sgd_solver.cpp:105] Iteration 1716, lr = 0.00711831 I0409 23:32:58.185544 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:33:02.037331 8290 solver.cpp:218] Iteration 1728 (2.45609 iter/s, 4.88582s/12 iters), loss = 4.00857 I0409 23:33:02.037385 8290 solver.cpp:237] Train net output #0: loss = 4.00857 (* 1 = 4.00857 loss) I0409 23:33:02.037397 8290 sgd_solver.cpp:105] Iteration 1728, lr = 0.00710141 I0409 23:33:04.058495 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1734.caffemodel I0409 23:33:04.982288 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1734.solverstate I0409 23:33:05.679397 8290 solver.cpp:330] Iteration 1734, Testing net (#0) I0409 23:33:05.679423 8290 net.cpp:676] Ignoring source layer train-data I0409 23:33:09.271613 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:33:09.973120 8290 solver.cpp:397] Test net output #0: accuracy = 0.0931373 I0409 23:33:09.973170 8290 solver.cpp:397] Test net output #1: loss = 4.06416 (* 1 = 4.06416 loss) I0409 23:33:11.824493 8290 solver.cpp:218] Iteration 1740 (1.22615 iter/s, 9.78675s/12 iters), loss = 3.93492 I0409 23:33:11.824537 8290 solver.cpp:237] Train net output #0: loss = 3.93492 (* 1 = 3.93492 loss) I0409 23:33:11.824545 8290 sgd_solver.cpp:105] Iteration 1740, lr = 0.00708455 I0409 23:33:16.750566 8290 solver.cpp:218] Iteration 1752 (2.43613 iter/s, 4.92584s/12 iters), loss = 3.90515 I0409 23:33:16.750628 8290 solver.cpp:237] Train net output #0: loss = 3.90515 (* 1 = 3.90515 loss) I0409 23:33:16.750641 8290 sgd_solver.cpp:105] Iteration 1752, lr = 0.00706773 I0409 23:33:21.627084 8290 solver.cpp:218] Iteration 1764 (2.4609 iter/s, 4.87627s/12 iters), loss = 3.87551 I0409 23:33:21.627133 8290 solver.cpp:237] Train net output #0: loss = 3.87551 (* 1 = 3.87551 loss) I0409 23:33:21.627144 8290 sgd_solver.cpp:105] Iteration 1764, lr = 0.00705094 I0409 23:33:26.517915 8290 solver.cpp:218] Iteration 1776 (2.45369 iter/s, 4.8906s/12 iters), loss = 3.86141 I0409 23:33:26.519124 8290 solver.cpp:237] Train net output #0: loss = 3.86141 (* 1 = 3.86141 loss) I0409 23:33:26.519137 8290 sgd_solver.cpp:105] Iteration 1776, lr = 0.0070342 I0409 23:33:31.504731 8290 solver.cpp:218] Iteration 1788 (2.40702 iter/s, 4.98543s/12 iters), loss = 3.8601 I0409 23:33:31.504779 8290 solver.cpp:237] Train net output #0: loss = 3.8601 (* 1 = 3.8601 loss) I0409 23:33:31.504788 8290 sgd_solver.cpp:105] Iteration 1788, lr = 0.0070175 I0409 23:33:36.384677 8290 solver.cpp:218] Iteration 1800 (2.45916 iter/s, 4.87971s/12 iters), loss = 3.94373 I0409 23:33:36.384729 8290 solver.cpp:237] Train net output #0: loss = 3.94373 (* 1 = 3.94373 loss) I0409 23:33:36.384742 8290 sgd_solver.cpp:105] Iteration 1800, lr = 0.00700084 I0409 23:33:41.466390 8290 solver.cpp:218] Iteration 1812 (2.36152 iter/s, 5.08147s/12 iters), loss = 3.81715 I0409 23:33:41.466444 8290 solver.cpp:237] Train net output #0: loss = 3.81715 (* 1 = 3.81715 loss) I0409 23:33:41.466455 8290 sgd_solver.cpp:105] Iteration 1812, lr = 0.00698422 I0409 23:33:44.657694 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:33:46.438589 8290 solver.cpp:218] Iteration 1824 (2.41354 iter/s, 4.97196s/12 iters), loss = 3.9224 I0409 23:33:46.438645 8290 solver.cpp:237] Train net output #0: loss = 3.9224 (* 1 = 3.9224 loss) I0409 23:33:46.438657 8290 sgd_solver.cpp:105] Iteration 1824, lr = 0.00696764 I0409 23:33:50.937685 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1836.caffemodel I0409 23:33:51.701740 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1836.solverstate I0409 23:33:52.240957 8290 solver.cpp:330] Iteration 1836, Testing net (#0) I0409 23:33:52.240983 8290 net.cpp:676] Ignoring source layer train-data I0409 23:33:55.938870 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:33:56.685804 8290 solver.cpp:397] Test net output #0: accuracy = 0.112745 I0409 23:33:56.685938 8290 solver.cpp:397] Test net output #1: loss = 3.91485 (* 1 = 3.91485 loss) I0409 23:33:56.767382 8290 solver.cpp:218] Iteration 1836 (1.16185 iter/s, 10.3284s/12 iters), loss = 3.89828 I0409 23:33:56.767436 8290 solver.cpp:237] Train net output #0: loss = 3.89828 (* 1 = 3.89828 loss) I0409 23:33:56.767448 8290 sgd_solver.cpp:105] Iteration 1836, lr = 0.0069511 I0409 23:34:01.054229 8290 solver.cpp:218] Iteration 1848 (2.7994 iter/s, 4.28663s/12 iters), loss = 3.87536 I0409 23:34:01.054270 8290 solver.cpp:237] Train net output #0: loss = 3.87536 (* 1 = 3.87536 loss) I0409 23:34:01.054280 8290 sgd_solver.cpp:105] Iteration 1848, lr = 0.00693459 I0409 23:34:05.899570 8290 solver.cpp:218] Iteration 1860 (2.47672 iter/s, 4.84512s/12 iters), loss = 3.61103 I0409 23:34:05.899622 8290 solver.cpp:237] Train net output #0: loss = 3.61103 (* 1 = 3.61103 loss) I0409 23:34:05.899633 8290 sgd_solver.cpp:105] Iteration 1860, lr = 0.00691813 I0409 23:34:10.819231 8290 solver.cpp:218] Iteration 1872 (2.43931 iter/s, 4.91943s/12 iters), loss = 3.8362 I0409 23:34:10.819286 8290 solver.cpp:237] Train net output #0: loss = 3.8362 (* 1 = 3.8362 loss) I0409 23:34:10.819301 8290 sgd_solver.cpp:105] Iteration 1872, lr = 0.0069017 I0409 23:34:15.827095 8290 solver.cpp:218] Iteration 1884 (2.39635 iter/s, 5.00761s/12 iters), loss = 3.8556 I0409 23:34:15.827165 8290 solver.cpp:237] Train net output #0: loss = 3.8556 (* 1 = 3.8556 loss) I0409 23:34:15.827181 8290 sgd_solver.cpp:105] Iteration 1884, lr = 0.00688532 I0409 23:34:20.797189 8290 solver.cpp:218] Iteration 1896 (2.41456 iter/s, 4.96985s/12 iters), loss = 3.81883 I0409 23:34:20.797235 8290 solver.cpp:237] Train net output #0: loss = 3.81883 (* 1 = 3.81883 loss) I0409 23:34:20.797245 8290 sgd_solver.cpp:105] Iteration 1896, lr = 0.00686897 I0409 23:34:25.700306 8290 solver.cpp:218] Iteration 1908 (2.44754 iter/s, 4.90289s/12 iters), loss = 3.99652 I0409 23:34:25.700356 8290 solver.cpp:237] Train net output #0: loss = 3.99652 (* 1 = 3.99652 loss) I0409 23:34:25.700364 8290 sgd_solver.cpp:105] Iteration 1908, lr = 0.00685266 I0409 23:34:30.577050 8290 solver.cpp:218] Iteration 1920 (2.46078 iter/s, 4.87651s/12 iters), loss = 3.8157 I0409 23:34:30.580497 8290 solver.cpp:237] Train net output #0: loss = 3.8157 (* 1 = 3.8157 loss) I0409 23:34:30.580508 8290 sgd_solver.cpp:105] Iteration 1920, lr = 0.00683639 I0409 23:34:30.896201 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:34:35.493530 8290 solver.cpp:218] Iteration 1932 (2.44257 iter/s, 4.91285s/12 iters), loss = 3.77454 I0409 23:34:35.493588 8290 solver.cpp:237] Train net output #0: loss = 3.77454 (* 1 = 3.77454 loss) I0409 23:34:35.493602 8290 sgd_solver.cpp:105] Iteration 1932, lr = 0.00682016 I0409 23:34:37.464057 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_1938.caffemodel I0409 23:34:38.196755 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_1938.solverstate I0409 23:34:38.725883 8290 solver.cpp:330] Iteration 1938, Testing net (#0) I0409 23:34:38.725912 8290 net.cpp:676] Ignoring source layer train-data I0409 23:34:42.463874 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:34:43.244628 8290 solver.cpp:397] Test net output #0: accuracy = 0.145221 I0409 23:34:43.244668 8290 solver.cpp:397] Test net output #1: loss = 3.69411 (* 1 = 3.69411 loss) I0409 23:34:45.064960 8290 solver.cpp:218] Iteration 1944 (1.25378 iter/s, 9.57102s/12 iters), loss = 3.7683 I0409 23:34:45.065006 8290 solver.cpp:237] Train net output #0: loss = 3.7683 (* 1 = 3.7683 loss) I0409 23:34:45.065014 8290 sgd_solver.cpp:105] Iteration 1944, lr = 0.00680397 I0409 23:34:50.044875 8290 solver.cpp:218] Iteration 1956 (2.40979 iter/s, 4.97968s/12 iters), loss = 3.57099 I0409 23:34:50.044926 8290 solver.cpp:237] Train net output #0: loss = 3.57099 (* 1 = 3.57099 loss) I0409 23:34:50.044935 8290 sgd_solver.cpp:105] Iteration 1956, lr = 0.00678782 I0409 23:34:54.915809 8290 solver.cpp:218] Iteration 1968 (2.46371 iter/s, 4.8707s/12 iters), loss = 3.66782 I0409 23:34:54.915853 8290 solver.cpp:237] Train net output #0: loss = 3.66782 (* 1 = 3.66782 loss) I0409 23:34:54.915863 8290 sgd_solver.cpp:105] Iteration 1968, lr = 0.0067717 I0409 23:34:59.751766 8290 solver.cpp:218] Iteration 1980 (2.48153 iter/s, 4.83573s/12 iters), loss = 3.53628 I0409 23:34:59.751807 8290 solver.cpp:237] Train net output #0: loss = 3.53628 (* 1 = 3.53628 loss) I0409 23:34:59.751816 8290 sgd_solver.cpp:105] Iteration 1980, lr = 0.00675562 I0409 23:35:04.552168 8290 solver.cpp:218] Iteration 1992 (2.49991 iter/s, 4.80018s/12 iters), loss = 3.71887 I0409 23:35:04.552253 8290 solver.cpp:237] Train net output #0: loss = 3.71887 (* 1 = 3.71887 loss) I0409 23:35:04.552266 8290 sgd_solver.cpp:105] Iteration 1992, lr = 0.00673958 I0409 23:35:09.386368 8290 solver.cpp:218] Iteration 2004 (2.48245 iter/s, 4.83393s/12 iters), loss = 3.52654 I0409 23:35:09.386422 8290 solver.cpp:237] Train net output #0: loss = 3.52654 (* 1 = 3.52654 loss) I0409 23:35:09.386435 8290 sgd_solver.cpp:105] Iteration 2004, lr = 0.00672358 I0409 23:35:14.290952 8290 solver.cpp:218] Iteration 2016 (2.44681 iter/s, 4.90435s/12 iters), loss = 3.61567 I0409 23:35:14.290997 8290 solver.cpp:237] Train net output #0: loss = 3.61567 (* 1 = 3.61567 loss) I0409 23:35:14.291009 8290 sgd_solver.cpp:105] Iteration 2016, lr = 0.00670762 I0409 23:35:16.788079 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:35:19.172032 8290 solver.cpp:218] Iteration 2028 (2.45859 iter/s, 4.88085s/12 iters), loss = 3.44332 I0409 23:35:19.172078 8290 solver.cpp:237] Train net output #0: loss = 3.44332 (* 1 = 3.44332 loss) I0409 23:35:19.172087 8290 sgd_solver.cpp:105] Iteration 2028, lr = 0.00669169 I0409 23:35:23.688318 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2040.caffemodel I0409 23:35:24.434070 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2040.solverstate I0409 23:35:24.975539 8290 solver.cpp:330] Iteration 2040, Testing net (#0) I0409 23:35:24.975564 8290 net.cpp:676] Ignoring source layer train-data I0409 23:35:28.579335 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:35:29.403887 8290 solver.cpp:397] Test net output #0: accuracy = 0.170343 I0409 23:35:29.403923 8290 solver.cpp:397] Test net output #1: loss = 3.51427 (* 1 = 3.51427 loss) I0409 23:35:29.486008 8290 solver.cpp:218] Iteration 2040 (1.16352 iter/s, 10.3136s/12 iters), loss = 3.4685 I0409 23:35:29.486068 8290 solver.cpp:237] Train net output #0: loss = 3.4685 (* 1 = 3.4685 loss) I0409 23:35:29.486080 8290 sgd_solver.cpp:105] Iteration 2040, lr = 0.00667581 I0409 23:35:33.750592 8290 solver.cpp:218] Iteration 2052 (2.81402 iter/s, 4.26436s/12 iters), loss = 3.53668 I0409 23:35:33.750639 8290 solver.cpp:237] Train net output #0: loss = 3.53668 (* 1 = 3.53668 loss) I0409 23:35:33.750648 8290 sgd_solver.cpp:105] Iteration 2052, lr = 0.00665996 I0409 23:35:34.116518 8290 blocking_queue.cpp:49] Waiting for data I0409 23:35:38.606665 8290 solver.cpp:218] Iteration 2064 (2.47125 iter/s, 4.85584s/12 iters), loss = 3.45614 I0409 23:35:38.606786 8290 solver.cpp:237] Train net output #0: loss = 3.45614 (* 1 = 3.45614 loss) I0409 23:35:38.606796 8290 sgd_solver.cpp:105] Iteration 2064, lr = 0.00664414 I0409 23:35:43.415724 8290 solver.cpp:218] Iteration 2076 (2.49545 iter/s, 4.80876s/12 iters), loss = 3.82129 I0409 23:35:43.415768 8290 solver.cpp:237] Train net output #0: loss = 3.82129 (* 1 = 3.82129 loss) I0409 23:35:43.415778 8290 sgd_solver.cpp:105] Iteration 2076, lr = 0.00662837 I0409 23:35:48.314791 8290 solver.cpp:218] Iteration 2088 (2.44956 iter/s, 4.89883s/12 iters), loss = 3.36986 I0409 23:35:48.314846 8290 solver.cpp:237] Train net output #0: loss = 3.36986 (* 1 = 3.36986 loss) I0409 23:35:48.314860 8290 sgd_solver.cpp:105] Iteration 2088, lr = 0.00661263 I0409 23:35:53.226948 8290 solver.cpp:218] Iteration 2100 (2.44304 iter/s, 4.91192s/12 iters), loss = 3.33483 I0409 23:35:53.226992 8290 solver.cpp:237] Train net output #0: loss = 3.33483 (* 1 = 3.33483 loss) I0409 23:35:53.227001 8290 sgd_solver.cpp:105] Iteration 2100, lr = 0.00659693 I0409 23:35:58.132997 8290 solver.cpp:218] Iteration 2112 (2.44608 iter/s, 4.90582s/12 iters), loss = 3.40783 I0409 23:35:58.133051 8290 solver.cpp:237] Train net output #0: loss = 3.40783 (* 1 = 3.40783 loss) I0409 23:35:58.133064 8290 sgd_solver.cpp:105] Iteration 2112, lr = 0.00658127 I0409 23:36:02.723071 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:36:03.030594 8290 solver.cpp:218] Iteration 2124 (2.4503 iter/s, 4.89736s/12 iters), loss = 3.14603 I0409 23:36:03.030643 8290 solver.cpp:237] Train net output #0: loss = 3.14603 (* 1 = 3.14603 loss) I0409 23:36:03.030653 8290 sgd_solver.cpp:105] Iteration 2124, lr = 0.00656564 I0409 23:36:07.913058 8290 solver.cpp:218] Iteration 2136 (2.45789 iter/s, 4.88223s/12 iters), loss = 3.38008 I0409 23:36:07.913103 8290 solver.cpp:237] Train net output #0: loss = 3.38008 (* 1 = 3.38008 loss) I0409 23:36:07.913112 8290 sgd_solver.cpp:105] Iteration 2136, lr = 0.00655006 I0409 23:36:09.885239 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2142.caffemodel I0409 23:36:12.569259 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2142.solverstate I0409 23:36:13.488327 8290 solver.cpp:330] Iteration 2142, Testing net (#0) I0409 23:36:13.488346 8290 net.cpp:676] Ignoring source layer train-data I0409 23:36:17.069357 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:36:17.928733 8290 solver.cpp:397] Test net output #0: accuracy = 0.196078 I0409 23:36:17.928783 8290 solver.cpp:397] Test net output #1: loss = 3.36436 (* 1 = 3.36436 loss) I0409 23:36:19.583251 8290 solver.cpp:218] Iteration 2148 (1.0283 iter/s, 11.6697s/12 iters), loss = 3.25187 I0409 23:36:19.583303 8290 solver.cpp:237] Train net output #0: loss = 3.25187 (* 1 = 3.25187 loss) I0409 23:36:19.583315 8290 sgd_solver.cpp:105] Iteration 2148, lr = 0.00653451 I0409 23:36:24.426235 8290 solver.cpp:218] Iteration 2160 (2.47793 iter/s, 4.84275s/12 iters), loss = 3.54573 I0409 23:36:24.426276 8290 solver.cpp:237] Train net output #0: loss = 3.54573 (* 1 = 3.54573 loss) I0409 23:36:24.426285 8290 sgd_solver.cpp:105] Iteration 2160, lr = 0.00651899 I0409 23:36:29.550704 8290 solver.cpp:218] Iteration 2172 (2.34181 iter/s, 5.12423s/12 iters), loss = 3.52319 I0409 23:36:29.550760 8290 solver.cpp:237] Train net output #0: loss = 3.52319 (* 1 = 3.52319 loss) I0409 23:36:29.550770 8290 sgd_solver.cpp:105] Iteration 2172, lr = 0.00650351 I0409 23:36:34.622613 8290 solver.cpp:218] Iteration 2184 (2.36609 iter/s, 5.07166s/12 iters), loss = 3.44352 I0409 23:36:34.622673 8290 solver.cpp:237] Train net output #0: loss = 3.44352 (* 1 = 3.44352 loss) I0409 23:36:34.622686 8290 sgd_solver.cpp:105] Iteration 2184, lr = 0.00648807 I0409 23:36:39.666606 8290 solver.cpp:218] Iteration 2196 (2.37918 iter/s, 5.04375s/12 iters), loss = 3.32733 I0409 23:36:39.666661 8290 solver.cpp:237] Train net output #0: loss = 3.32733 (* 1 = 3.32733 loss) I0409 23:36:39.666672 8290 sgd_solver.cpp:105] Iteration 2196, lr = 0.00647267 I0409 23:36:45.516759 8290 solver.cpp:218] Iteration 2208 (2.05132 iter/s, 5.84988s/12 iters), loss = 3.28936 I0409 23:36:45.517407 8290 solver.cpp:237] Train net output #0: loss = 3.28936 (* 1 = 3.28936 loss) I0409 23:36:45.517416 8290 sgd_solver.cpp:105] Iteration 2208, lr = 0.0064573 I0409 23:36:50.383365 8290 solver.cpp:218] Iteration 2220 (2.46621 iter/s, 4.86577s/12 iters), loss = 3.26467 I0409 23:36:50.383424 8290 solver.cpp:237] Train net output #0: loss = 3.26467 (* 1 = 3.26467 loss) I0409 23:36:50.383435 8290 sgd_solver.cpp:105] Iteration 2220, lr = 0.00644197 I0409 23:36:52.141044 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:36:55.294721 8290 solver.cpp:218] Iteration 2232 (2.44344 iter/s, 4.91111s/12 iters), loss = 3.11914 I0409 23:36:55.294778 8290 solver.cpp:237] Train net output #0: loss = 3.11914 (* 1 = 3.11914 loss) I0409 23:36:55.294790 8290 sgd_solver.cpp:105] Iteration 2232, lr = 0.00642668 I0409 23:36:59.747815 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2244.caffemodel I0409 23:37:00.915973 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2244.solverstate I0409 23:37:01.447135 8290 solver.cpp:330] Iteration 2244, Testing net (#0) I0409 23:37:01.447163 8290 net.cpp:676] Ignoring source layer train-data I0409 23:37:05.061018 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:37:05.964597 8290 solver.cpp:397] Test net output #0: accuracy = 0.199755 I0409 23:37:05.964627 8290 solver.cpp:397] Test net output #1: loss = 3.30525 (* 1 = 3.30525 loss) I0409 23:37:06.046761 8290 solver.cpp:218] Iteration 2244 (1.11611 iter/s, 10.7516s/12 iters), loss = 3.16469 I0409 23:37:06.046810 8290 solver.cpp:237] Train net output #0: loss = 3.16469 (* 1 = 3.16469 loss) I0409 23:37:06.046818 8290 sgd_solver.cpp:105] Iteration 2244, lr = 0.00641142 I0409 23:37:10.174962 8290 solver.cpp:218] Iteration 2256 (2.90698 iter/s, 4.12799s/12 iters), loss = 2.92105 I0409 23:37:10.175022 8290 solver.cpp:237] Train net output #0: loss = 2.92105 (* 1 = 2.92105 loss) I0409 23:37:10.175033 8290 sgd_solver.cpp:105] Iteration 2256, lr = 0.0063962 I0409 23:37:15.012398 8290 solver.cpp:218] Iteration 2268 (2.48078 iter/s, 4.8372s/12 iters), loss = 3.24201 I0409 23:37:15.012454 8290 solver.cpp:237] Train net output #0: loss = 3.24201 (* 1 = 3.24201 loss) I0409 23:37:15.012465 8290 sgd_solver.cpp:105] Iteration 2268, lr = 0.00638101 I0409 23:37:19.974560 8290 solver.cpp:218] Iteration 2280 (2.41842 iter/s, 4.96193s/12 iters), loss = 3.37218 I0409 23:37:19.974663 8290 solver.cpp:237] Train net output #0: loss = 3.37218 (* 1 = 3.37218 loss) I0409 23:37:19.974678 8290 sgd_solver.cpp:105] Iteration 2280, lr = 0.00636586 I0409 23:37:24.773232 8290 solver.cpp:218] Iteration 2292 (2.50084 iter/s, 4.79839s/12 iters), loss = 3.2766 I0409 23:37:24.773286 8290 solver.cpp:237] Train net output #0: loss = 3.2766 (* 1 = 3.2766 loss) I0409 23:37:24.773298 8290 sgd_solver.cpp:105] Iteration 2292, lr = 0.00635075 I0409 23:37:29.670367 8290 solver.cpp:218] Iteration 2304 (2.45053 iter/s, 4.8969s/12 iters), loss = 3.29726 I0409 23:37:29.670418 8290 solver.cpp:237] Train net output #0: loss = 3.29726 (* 1 = 3.29726 loss) I0409 23:37:29.670429 8290 sgd_solver.cpp:105] Iteration 2304, lr = 0.00633567 I0409 23:37:34.594386 8290 solver.cpp:218] Iteration 2316 (2.43715 iter/s, 4.92378s/12 iters), loss = 3.20252 I0409 23:37:34.594434 8290 solver.cpp:237] Train net output #0: loss = 3.20252 (* 1 = 3.20252 loss) I0409 23:37:34.594444 8290 sgd_solver.cpp:105] Iteration 2316, lr = 0.00632063 I0409 23:37:38.478787 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:37:39.522573 8290 solver.cpp:218] Iteration 2328 (2.43509 iter/s, 4.92795s/12 iters), loss = 3.22207 I0409 23:37:39.522626 8290 solver.cpp:237] Train net output #0: loss = 3.22207 (* 1 = 3.22207 loss) I0409 23:37:39.522637 8290 sgd_solver.cpp:105] Iteration 2328, lr = 0.00630562 I0409 23:37:44.426620 8290 solver.cpp:218] Iteration 2340 (2.44707 iter/s, 4.90381s/12 iters), loss = 3.06521 I0409 23:37:44.426674 8290 solver.cpp:237] Train net output #0: loss = 3.06521 (* 1 = 3.06521 loss) I0409 23:37:44.426685 8290 sgd_solver.cpp:105] Iteration 2340, lr = 0.00629065 I0409 23:37:46.409425 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2346.caffemodel I0409 23:37:47.517639 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2346.solverstate I0409 23:37:48.071400 8290 solver.cpp:330] Iteration 2346, Testing net (#0) I0409 23:37:48.071424 8290 net.cpp:676] Ignoring source layer train-data I0409 23:37:51.463989 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:37:52.414507 8290 solver.cpp:397] Test net output #0: accuracy = 0.244485 I0409 23:37:52.414561 8290 solver.cpp:397] Test net output #1: loss = 3.15593 (* 1 = 3.15593 loss) I0409 23:37:54.311903 8290 solver.cpp:218] Iteration 2352 (1.21398 iter/s, 9.88487s/12 iters), loss = 3.05673 I0409 23:37:54.311960 8290 solver.cpp:237] Train net output #0: loss = 3.05673 (* 1 = 3.05673 loss) I0409 23:37:54.311972 8290 sgd_solver.cpp:105] Iteration 2352, lr = 0.00627571 I0409 23:37:59.197875 8290 solver.cpp:218] Iteration 2364 (2.45613 iter/s, 4.88574s/12 iters), loss = 2.90498 I0409 23:37:59.197921 8290 solver.cpp:237] Train net output #0: loss = 2.90498 (* 1 = 2.90498 loss) I0409 23:37:59.197932 8290 sgd_solver.cpp:105] Iteration 2364, lr = 0.00626081 I0409 23:38:04.089411 8290 solver.cpp:218] Iteration 2376 (2.45333 iter/s, 4.89131s/12 iters), loss = 3.00697 I0409 23:38:04.089462 8290 solver.cpp:237] Train net output #0: loss = 3.00697 (* 1 = 3.00697 loss) I0409 23:38:04.089474 8290 sgd_solver.cpp:105] Iteration 2376, lr = 0.00624595 I0409 23:38:09.198693 8290 solver.cpp:218] Iteration 2388 (2.34878 iter/s, 5.10904s/12 iters), loss = 2.97851 I0409 23:38:09.198745 8290 solver.cpp:237] Train net output #0: loss = 2.97851 (* 1 = 2.97851 loss) I0409 23:38:09.198756 8290 sgd_solver.cpp:105] Iteration 2388, lr = 0.00623112 I0409 23:38:14.146504 8290 solver.cpp:218] Iteration 2400 (2.42543 iter/s, 4.94758s/12 iters), loss = 3.03367 I0409 23:38:14.146550 8290 solver.cpp:237] Train net output #0: loss = 3.03367 (* 1 = 3.03367 loss) I0409 23:38:14.146560 8290 sgd_solver.cpp:105] Iteration 2400, lr = 0.00621633 I0409 23:38:19.066754 8290 solver.cpp:218] Iteration 2412 (2.43901 iter/s, 4.92002s/12 iters), loss = 2.80439 I0409 23:38:19.066805 8290 solver.cpp:237] Train net output #0: loss = 2.80439 (* 1 = 2.80439 loss) I0409 23:38:19.066817 8290 sgd_solver.cpp:105] Iteration 2412, lr = 0.00620157 I0409 23:38:23.976961 8290 solver.cpp:218] Iteration 2424 (2.44401 iter/s, 4.90997s/12 iters), loss = 3.06577 I0409 23:38:23.977120 8290 solver.cpp:237] Train net output #0: loss = 3.06577 (* 1 = 3.06577 loss) I0409 23:38:23.977134 8290 sgd_solver.cpp:105] Iteration 2424, lr = 0.00618684 I0409 23:38:25.035281 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:38:28.900079 8290 solver.cpp:218] Iteration 2436 (2.43765 iter/s, 4.92277s/12 iters), loss = 2.97528 I0409 23:38:28.900138 8290 solver.cpp:237] Train net output #0: loss = 2.97528 (* 1 = 2.97528 loss) I0409 23:38:28.900151 8290 sgd_solver.cpp:105] Iteration 2436, lr = 0.00617215 I0409 23:38:33.325450 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2448.caffemodel I0409 23:38:34.467243 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2448.solverstate I0409 23:38:35.003835 8290 solver.cpp:330] Iteration 2448, Testing net (#0) I0409 23:38:35.003865 8290 net.cpp:676] Ignoring source layer train-data I0409 23:38:38.381402 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:38:39.355664 8290 solver.cpp:397] Test net output #0: accuracy = 0.259804 I0409 23:38:39.355707 8290 solver.cpp:397] Test net output #1: loss = 3.03298 (* 1 = 3.03298 loss) I0409 23:38:39.437536 8290 solver.cpp:218] Iteration 2448 (1.13884 iter/s, 10.537s/12 iters), loss = 2.83385 I0409 23:38:39.437585 8290 solver.cpp:237] Train net output #0: loss = 2.83385 (* 1 = 2.83385 loss) I0409 23:38:39.437595 8290 sgd_solver.cpp:105] Iteration 2448, lr = 0.0061575 I0409 23:38:43.530462 8290 solver.cpp:218] Iteration 2460 (2.93204 iter/s, 4.09272s/12 iters), loss = 2.83407 I0409 23:38:43.530516 8290 solver.cpp:237] Train net output #0: loss = 2.83407 (* 1 = 2.83407 loss) I0409 23:38:43.530529 8290 sgd_solver.cpp:105] Iteration 2460, lr = 0.00614288 I0409 23:38:48.388763 8290 solver.cpp:218] Iteration 2472 (2.47012 iter/s, 4.85807s/12 iters), loss = 3.03771 I0409 23:38:48.388819 8290 solver.cpp:237] Train net output #0: loss = 3.03771 (* 1 = 3.03771 loss) I0409 23:38:48.388830 8290 sgd_solver.cpp:105] Iteration 2472, lr = 0.0061283 I0409 23:38:53.310086 8290 solver.cpp:218] Iteration 2484 (2.43849 iter/s, 4.92109s/12 iters), loss = 3.06575 I0409 23:38:53.310137 8290 solver.cpp:237] Train net output #0: loss = 3.06575 (* 1 = 3.06575 loss) I0409 23:38:53.310148 8290 sgd_solver.cpp:105] Iteration 2484, lr = 0.00611375 I0409 23:38:58.213066 8290 solver.cpp:218] Iteration 2496 (2.44761 iter/s, 4.90275s/12 iters), loss = 2.80665 I0409 23:38:58.213201 8290 solver.cpp:237] Train net output #0: loss = 2.80665 (* 1 = 2.80665 loss) I0409 23:38:58.213213 8290 sgd_solver.cpp:105] Iteration 2496, lr = 0.00609923 I0409 23:39:03.100324 8290 solver.cpp:218] Iteration 2508 (2.45552 iter/s, 4.88695s/12 iters), loss = 3.09614 I0409 23:39:03.100375 8290 solver.cpp:237] Train net output #0: loss = 3.09614 (* 1 = 3.09614 loss) I0409 23:39:03.100386 8290 sgd_solver.cpp:105] Iteration 2508, lr = 0.00608475 I0409 23:39:08.007918 8290 solver.cpp:218] Iteration 2520 (2.4453 iter/s, 4.90737s/12 iters), loss = 2.89344 I0409 23:39:08.007961 8290 solver.cpp:237] Train net output #0: loss = 2.89344 (* 1 = 2.89344 loss) I0409 23:39:08.007970 8290 sgd_solver.cpp:105] Iteration 2520, lr = 0.0060703 I0409 23:39:11.174528 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:39:12.925168 8290 solver.cpp:218] Iteration 2532 (2.4405 iter/s, 4.91703s/12 iters), loss = 2.89829 I0409 23:39:12.925205 8290 solver.cpp:237] Train net output #0: loss = 2.89829 (* 1 = 2.89829 loss) I0409 23:39:12.925215 8290 sgd_solver.cpp:105] Iteration 2532, lr = 0.00605589 I0409 23:39:17.843164 8290 solver.cpp:218] Iteration 2544 (2.44013 iter/s, 4.91777s/12 iters), loss = 2.9613 I0409 23:39:17.843209 8290 solver.cpp:237] Train net output #0: loss = 2.9613 (* 1 = 2.9613 loss) I0409 23:39:17.843219 8290 sgd_solver.cpp:105] Iteration 2544, lr = 0.00604151 I0409 23:39:19.828464 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2550.caffemodel I0409 23:39:20.809114 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2550.solverstate I0409 23:39:21.592602 8290 solver.cpp:330] Iteration 2550, Testing net (#0) I0409 23:39:21.592631 8290 net.cpp:676] Ignoring source layer train-data I0409 23:39:24.930742 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:39:25.950460 8290 solver.cpp:397] Test net output #0: accuracy = 0.264706 I0409 23:39:25.950501 8290 solver.cpp:397] Test net output #1: loss = 3.01086 (* 1 = 3.01086 loss) I0409 23:39:27.688946 8290 solver.cpp:218] Iteration 2556 (1.21885 iter/s, 9.84539s/12 iters), loss = 3.0912 I0409 23:39:27.688997 8290 solver.cpp:237] Train net output #0: loss = 3.0912 (* 1 = 3.0912 loss) I0409 23:39:27.689007 8290 sgd_solver.cpp:105] Iteration 2556, lr = 0.00602717 I0409 23:39:32.567674 8290 solver.cpp:218] Iteration 2568 (2.45978 iter/s, 4.87849s/12 iters), loss = 2.6834 I0409 23:39:32.567823 8290 solver.cpp:237] Train net output #0: loss = 2.6834 (* 1 = 2.6834 loss) I0409 23:39:32.567837 8290 sgd_solver.cpp:105] Iteration 2568, lr = 0.00601286 I0409 23:39:37.479401 8290 solver.cpp:218] Iteration 2580 (2.4433 iter/s, 4.9114s/12 iters), loss = 2.69897 I0409 23:39:37.479457 8290 solver.cpp:237] Train net output #0: loss = 2.69897 (* 1 = 2.69897 loss) I0409 23:39:37.479470 8290 sgd_solver.cpp:105] Iteration 2580, lr = 0.00599858 I0409 23:39:42.415720 8290 solver.cpp:218] Iteration 2592 (2.43108 iter/s, 4.93608s/12 iters), loss = 3.21887 I0409 23:39:42.415773 8290 solver.cpp:237] Train net output #0: loss = 3.21887 (* 1 = 3.21887 loss) I0409 23:39:42.415786 8290 sgd_solver.cpp:105] Iteration 2592, lr = 0.00598434 I0409 23:39:47.332022 8290 solver.cpp:218] Iteration 2604 (2.44097 iter/s, 4.91607s/12 iters), loss = 2.8906 I0409 23:39:47.332072 8290 solver.cpp:237] Train net output #0: loss = 2.8906 (* 1 = 2.8906 loss) I0409 23:39:47.332080 8290 sgd_solver.cpp:105] Iteration 2604, lr = 0.00597013 I0409 23:39:52.266383 8290 solver.cpp:218] Iteration 2616 (2.43204 iter/s, 4.93413s/12 iters), loss = 2.84578 I0409 23:39:52.266427 8290 solver.cpp:237] Train net output #0: loss = 2.84578 (* 1 = 2.84578 loss) I0409 23:39:52.266434 8290 sgd_solver.cpp:105] Iteration 2616, lr = 0.00595596 I0409 23:39:57.155295 8290 solver.cpp:218] Iteration 2628 (2.45465 iter/s, 4.88868s/12 iters), loss = 2.54841 I0409 23:39:57.155340 8290 solver.cpp:237] Train net output #0: loss = 2.54841 (* 1 = 2.54841 loss) I0409 23:39:57.155349 8290 sgd_solver.cpp:105] Iteration 2628, lr = 0.00594182 I0409 23:39:57.583505 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:40:02.086506 8290 solver.cpp:218] Iteration 2640 (2.43359 iter/s, 4.93098s/12 iters), loss = 2.74976 I0409 23:40:02.086565 8290 solver.cpp:237] Train net output #0: loss = 2.74976 (* 1 = 2.74976 loss) I0409 23:40:02.086577 8290 sgd_solver.cpp:105] Iteration 2640, lr = 0.00592771 I0409 23:40:06.535418 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2652.caffemodel I0409 23:40:09.203194 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2652.solverstate I0409 23:40:12.563077 8290 solver.cpp:330] Iteration 2652, Testing net (#0) I0409 23:40:12.563107 8290 net.cpp:676] Ignoring source layer train-data I0409 23:40:15.945986 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:40:16.999275 8290 solver.cpp:397] Test net output #0: accuracy = 0.294118 I0409 23:40:16.999321 8290 solver.cpp:397] Test net output #1: loss = 2.89701 (* 1 = 2.89701 loss) I0409 23:40:17.081398 8290 solver.cpp:218] Iteration 2652 (0.800304 iter/s, 14.9943s/12 iters), loss = 2.80156 I0409 23:40:17.081452 8290 solver.cpp:237] Train net output #0: loss = 2.80156 (* 1 = 2.80156 loss) I0409 23:40:17.081465 8290 sgd_solver.cpp:105] Iteration 2652, lr = 0.00591364 I0409 23:40:21.174474 8290 solver.cpp:218] Iteration 2664 (2.93193 iter/s, 4.09287s/12 iters), loss = 3.05952 I0409 23:40:21.174518 8290 solver.cpp:237] Train net output #0: loss = 3.05952 (* 1 = 3.05952 loss) I0409 23:40:21.174527 8290 sgd_solver.cpp:105] Iteration 2664, lr = 0.0058996 I0409 23:40:26.085662 8290 solver.cpp:218] Iteration 2676 (2.44351 iter/s, 4.91096s/12 iters), loss = 2.79877 I0409 23:40:26.085706 8290 solver.cpp:237] Train net output #0: loss = 2.79877 (* 1 = 2.79877 loss) I0409 23:40:26.085717 8290 sgd_solver.cpp:105] Iteration 2676, lr = 0.00588559 I0409 23:40:30.907835 8290 solver.cpp:218] Iteration 2688 (2.48862 iter/s, 4.82195s/12 iters), loss = 2.80286 I0409 23:40:30.907900 8290 solver.cpp:237] Train net output #0: loss = 2.80286 (* 1 = 2.80286 loss) I0409 23:40:30.907912 8290 sgd_solver.cpp:105] Iteration 2688, lr = 0.00587162 I0409 23:40:35.931373 8290 solver.cpp:218] Iteration 2700 (2.38887 iter/s, 5.02329s/12 iters), loss = 2.5956 I0409 23:40:35.931423 8290 solver.cpp:237] Train net output #0: loss = 2.5956 (* 1 = 2.5956 loss) I0409 23:40:35.931433 8290 sgd_solver.cpp:105] Iteration 2700, lr = 0.00585768 I0409 23:40:40.794185 8290 solver.cpp:218] Iteration 2712 (2.46782 iter/s, 4.86259s/12 iters), loss = 2.68849 I0409 23:40:40.794301 8290 solver.cpp:237] Train net output #0: loss = 2.68849 (* 1 = 2.68849 loss) I0409 23:40:40.794310 8290 sgd_solver.cpp:105] Iteration 2712, lr = 0.00584377 I0409 23:40:45.712975 8290 solver.cpp:218] Iteration 2724 (2.43977 iter/s, 4.91849s/12 iters), loss = 2.71414 I0409 23:40:45.713030 8290 solver.cpp:237] Train net output #0: loss = 2.71414 (* 1 = 2.71414 loss) I0409 23:40:45.713042 8290 sgd_solver.cpp:105] Iteration 2724, lr = 0.0058299 I0409 23:40:48.197527 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:40:50.614279 8290 solver.cpp:218] Iteration 2736 (2.44845 iter/s, 4.90107s/12 iters), loss = 2.23938 I0409 23:40:50.614334 8290 solver.cpp:237] Train net output #0: loss = 2.23938 (* 1 = 2.23938 loss) I0409 23:40:50.614346 8290 sgd_solver.cpp:105] Iteration 2736, lr = 0.00581605 I0409 23:40:55.509424 8290 solver.cpp:218] Iteration 2748 (2.45153 iter/s, 4.89491s/12 iters), loss = 2.60263 I0409 23:40:55.509471 8290 solver.cpp:237] Train net output #0: loss = 2.60263 (* 1 = 2.60263 loss) I0409 23:40:55.509480 8290 sgd_solver.cpp:105] Iteration 2748, lr = 0.00580225 I0409 23:40:57.467082 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2754.caffemodel I0409 23:40:58.158877 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2754.solverstate I0409 23:40:58.688045 8290 solver.cpp:330] Iteration 2754, Testing net (#0) I0409 23:40:58.688076 8290 net.cpp:676] Ignoring source layer train-data I0409 23:41:01.598464 8290 blocking_queue.cpp:49] Waiting for data I0409 23:41:02.155438 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:41:03.280329 8290 solver.cpp:397] Test net output #0: accuracy = 0.299632 I0409 23:41:03.280359 8290 solver.cpp:397] Test net output #1: loss = 2.81262 (* 1 = 2.81262 loss) I0409 23:41:05.119145 8290 solver.cpp:218] Iteration 2760 (1.24879 iter/s, 9.60933s/12 iters), loss = 2.53244 I0409 23:41:05.119189 8290 solver.cpp:237] Train net output #0: loss = 2.53244 (* 1 = 2.53244 loss) I0409 23:41:05.119196 8290 sgd_solver.cpp:105] Iteration 2760, lr = 0.00578847 I0409 23:41:10.132064 8290 solver.cpp:218] Iteration 2772 (2.39393 iter/s, 5.01268s/12 iters), loss = 2.48954 I0409 23:41:10.132118 8290 solver.cpp:237] Train net output #0: loss = 2.48954 (* 1 = 2.48954 loss) I0409 23:41:10.132130 8290 sgd_solver.cpp:105] Iteration 2772, lr = 0.00577473 I0409 23:41:15.007477 8290 solver.cpp:218] Iteration 2784 (2.46145 iter/s, 4.87517s/12 iters), loss = 2.66821 I0409 23:41:15.007606 8290 solver.cpp:237] Train net output #0: loss = 2.66821 (* 1 = 2.66821 loss) I0409 23:41:15.007620 8290 sgd_solver.cpp:105] Iteration 2784, lr = 0.00576102 I0409 23:41:19.903470 8290 solver.cpp:218] Iteration 2796 (2.45114 iter/s, 4.89569s/12 iters), loss = 2.59045 I0409 23:41:19.903513 8290 solver.cpp:237] Train net output #0: loss = 2.59045 (* 1 = 2.59045 loss) I0409 23:41:19.903523 8290 sgd_solver.cpp:105] Iteration 2796, lr = 0.00574734 I0409 23:41:24.817687 8290 solver.cpp:218] Iteration 2808 (2.44201 iter/s, 4.91399s/12 iters), loss = 2.52482 I0409 23:41:24.817736 8290 solver.cpp:237] Train net output #0: loss = 2.52482 (* 1 = 2.52482 loss) I0409 23:41:24.817749 8290 sgd_solver.cpp:105] Iteration 2808, lr = 0.00573369 I0409 23:41:29.740837 8290 solver.cpp:218] Iteration 2820 (2.43758 iter/s, 4.92291s/12 iters), loss = 2.81171 I0409 23:41:29.740891 8290 solver.cpp:237] Train net output #0: loss = 2.81171 (* 1 = 2.81171 loss) I0409 23:41:29.740904 8290 sgd_solver.cpp:105] Iteration 2820, lr = 0.00572008 I0409 23:41:34.350088 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:41:34.634891 8290 solver.cpp:218] Iteration 2832 (2.45207 iter/s, 4.89382s/12 iters), loss = 2.4636 I0409 23:41:34.634932 8290 solver.cpp:237] Train net output #0: loss = 2.4636 (* 1 = 2.4636 loss) I0409 23:41:34.634940 8290 sgd_solver.cpp:105] Iteration 2832, lr = 0.0057065 I0409 23:41:39.512459 8290 solver.cpp:218] Iteration 2844 (2.46036 iter/s, 4.87734s/12 iters), loss = 2.47835 I0409 23:41:39.512514 8290 solver.cpp:237] Train net output #0: loss = 2.47835 (* 1 = 2.47835 loss) I0409 23:41:39.512526 8290 sgd_solver.cpp:105] Iteration 2844, lr = 0.00569295 I0409 23:41:43.959719 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2856.caffemodel I0409 23:41:44.690907 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2856.solverstate I0409 23:41:45.263929 8290 solver.cpp:330] Iteration 2856, Testing net (#0) I0409 23:41:45.264050 8290 net.cpp:676] Ignoring source layer train-data I0409 23:41:48.619391 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:41:49.804621 8290 solver.cpp:397] Test net output #0: accuracy = 0.327819 I0409 23:41:49.804663 8290 solver.cpp:397] Test net output #1: loss = 2.68503 (* 1 = 2.68503 loss) I0409 23:41:49.886793 8290 solver.cpp:218] Iteration 2856 (1.15675 iter/s, 10.3739s/12 iters), loss = 2.38392 I0409 23:41:49.886845 8290 solver.cpp:237] Train net output #0: loss = 2.38392 (* 1 = 2.38392 loss) I0409 23:41:49.886855 8290 sgd_solver.cpp:105] Iteration 2856, lr = 0.00567944 I0409 23:41:54.026728 8290 solver.cpp:218] Iteration 2868 (2.89874 iter/s, 4.13973s/12 iters), loss = 2.49541 I0409 23:41:54.026772 8290 solver.cpp:237] Train net output #0: loss = 2.49541 (* 1 = 2.49541 loss) I0409 23:41:54.026782 8290 sgd_solver.cpp:105] Iteration 2868, lr = 0.00566595 I0409 23:41:58.934428 8290 solver.cpp:218] Iteration 2880 (2.44525 iter/s, 4.90747s/12 iters), loss = 2.78279 I0409 23:41:58.934485 8290 solver.cpp:237] Train net output #0: loss = 2.78279 (* 1 = 2.78279 loss) I0409 23:41:58.934499 8290 sgd_solver.cpp:105] Iteration 2880, lr = 0.0056525 I0409 23:42:04.012657 8290 solver.cpp:218] Iteration 2892 (2.36314 iter/s, 5.07798s/12 iters), loss = 2.44381 I0409 23:42:04.012709 8290 solver.cpp:237] Train net output #0: loss = 2.44381 (* 1 = 2.44381 loss) I0409 23:42:04.012722 8290 sgd_solver.cpp:105] Iteration 2892, lr = 0.00563908 I0409 23:42:09.032096 8290 solver.cpp:218] Iteration 2904 (2.39082 iter/s, 5.0192s/12 iters), loss = 2.58685 I0409 23:42:09.032150 8290 solver.cpp:237] Train net output #0: loss = 2.58685 (* 1 = 2.58685 loss) I0409 23:42:09.032162 8290 sgd_solver.cpp:105] Iteration 2904, lr = 0.00562569 I0409 23:42:13.934837 8290 solver.cpp:218] Iteration 2916 (2.44773 iter/s, 4.90251s/12 iters), loss = 2.54589 I0409 23:42:13.934887 8290 solver.cpp:237] Train net output #0: loss = 2.54589 (* 1 = 2.54589 loss) I0409 23:42:13.934898 8290 sgd_solver.cpp:105] Iteration 2916, lr = 0.00561233 I0409 23:42:18.891132 8290 solver.cpp:218] Iteration 2928 (2.42128 iter/s, 4.95606s/12 iters), loss = 2.49586 I0409 23:42:18.891278 8290 solver.cpp:237] Train net output #0: loss = 2.49586 (* 1 = 2.49586 loss) I0409 23:42:18.891289 8290 sgd_solver.cpp:105] Iteration 2928, lr = 0.00559901 I0409 23:42:20.704308 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:42:23.803261 8290 solver.cpp:218] Iteration 2940 (2.4431 iter/s, 4.9118s/12 iters), loss = 2.40454 I0409 23:42:23.803320 8290 solver.cpp:237] Train net output #0: loss = 2.40454 (* 1 = 2.40454 loss) I0409 23:42:23.803333 8290 sgd_solver.cpp:105] Iteration 2940, lr = 0.00558572 I0409 23:42:28.768942 8290 solver.cpp:218] Iteration 2952 (2.4167 iter/s, 4.96544s/12 iters), loss = 2.54542 I0409 23:42:28.768988 8290 solver.cpp:237] Train net output #0: loss = 2.54542 (* 1 = 2.54542 loss) I0409 23:42:28.768998 8290 sgd_solver.cpp:105] Iteration 2952, lr = 0.00557245 I0409 23:42:30.744809 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_2958.caffemodel I0409 23:42:32.435125 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_2958.solverstate I0409 23:42:33.889734 8290 solver.cpp:330] Iteration 2958, Testing net (#0) I0409 23:42:33.889755 8290 net.cpp:676] Ignoring source layer train-data I0409 23:42:37.360661 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:42:38.538379 8290 solver.cpp:397] Test net output #0: accuracy = 0.331495 I0409 23:42:38.538427 8290 solver.cpp:397] Test net output #1: loss = 2.69524 (* 1 = 2.69524 loss) I0409 23:42:40.386364 8290 solver.cpp:218] Iteration 2964 (1.03297 iter/s, 11.617s/12 iters), loss = 2.10124 I0409 23:42:40.386422 8290 solver.cpp:237] Train net output #0: loss = 2.10124 (* 1 = 2.10124 loss) I0409 23:42:40.386435 8290 sgd_solver.cpp:105] Iteration 2964, lr = 0.00555922 I0409 23:42:45.367318 8290 solver.cpp:218] Iteration 2976 (2.4093 iter/s, 4.98071s/12 iters), loss = 2.45287 I0409 23:42:45.367377 8290 solver.cpp:237] Train net output #0: loss = 2.45287 (* 1 = 2.45287 loss) I0409 23:42:45.367389 8290 sgd_solver.cpp:105] Iteration 2976, lr = 0.00554603 I0409 23:42:50.286463 8290 solver.cpp:218] Iteration 2988 (2.43957 iter/s, 4.9189s/12 iters), loss = 2.41698 I0409 23:42:50.286562 8290 solver.cpp:237] Train net output #0: loss = 2.41698 (* 1 = 2.41698 loss) I0409 23:42:50.286572 8290 sgd_solver.cpp:105] Iteration 2988, lr = 0.00553286 I0409 23:42:55.236131 8290 solver.cpp:218] Iteration 3000 (2.42455 iter/s, 4.94938s/12 iters), loss = 2.55435 I0409 23:42:55.236186 8290 solver.cpp:237] Train net output #0: loss = 2.55435 (* 1 = 2.55435 loss) I0409 23:42:55.236199 8290 sgd_solver.cpp:105] Iteration 3000, lr = 0.00551972 I0409 23:43:00.215443 8290 solver.cpp:218] Iteration 3012 (2.41009 iter/s, 4.97907s/12 iters), loss = 2.27664 I0409 23:43:00.215497 8290 solver.cpp:237] Train net output #0: loss = 2.27664 (* 1 = 2.27664 loss) I0409 23:43:00.215507 8290 sgd_solver.cpp:105] Iteration 3012, lr = 0.00550662 I0409 23:43:05.203102 8290 solver.cpp:218] Iteration 3024 (2.40606 iter/s, 4.98742s/12 iters), loss = 2.19187 I0409 23:43:05.203163 8290 solver.cpp:237] Train net output #0: loss = 2.19187 (* 1 = 2.19187 loss) I0409 23:43:05.203176 8290 sgd_solver.cpp:105] Iteration 3024, lr = 0.00549354 I0409 23:43:09.133152 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:43:10.149665 8290 solver.cpp:218] Iteration 3036 (2.42605 iter/s, 4.94632s/12 iters), loss = 2.43849 I0409 23:43:10.149724 8290 solver.cpp:237] Train net output #0: loss = 2.43849 (* 1 = 2.43849 loss) I0409 23:43:10.149736 8290 sgd_solver.cpp:105] Iteration 3036, lr = 0.0054805 I0409 23:43:15.081331 8290 solver.cpp:218] Iteration 3048 (2.43337 iter/s, 4.93143s/12 iters), loss = 2.42832 I0409 23:43:15.081372 8290 solver.cpp:237] Train net output #0: loss = 2.42832 (* 1 = 2.42832 loss) I0409 23:43:15.081379 8290 sgd_solver.cpp:105] Iteration 3048, lr = 0.00546749 I0409 23:43:19.466102 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3060.caffemodel I0409 23:43:21.203698 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3060.solverstate I0409 23:43:22.524596 8290 solver.cpp:330] Iteration 3060, Testing net (#0) I0409 23:43:22.524619 8290 net.cpp:676] Ignoring source layer train-data I0409 23:43:25.754217 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:43:26.964608 8290 solver.cpp:397] Test net output #0: accuracy = 0.349265 I0409 23:43:26.964658 8290 solver.cpp:397] Test net output #1: loss = 2.58876 (* 1 = 2.58876 loss) I0409 23:43:27.046912 8290 solver.cpp:218] Iteration 3060 (1.00292 iter/s, 11.9651s/12 iters), loss = 2.17913 I0409 23:43:27.046970 8290 solver.cpp:237] Train net output #0: loss = 2.17913 (* 1 = 2.17913 loss) I0409 23:43:27.046983 8290 sgd_solver.cpp:105] Iteration 3060, lr = 0.00545451 I0409 23:43:31.273551 8290 solver.cpp:218] Iteration 3072 (2.83928 iter/s, 4.22643s/12 iters), loss = 2.10995 I0409 23:43:31.273597 8290 solver.cpp:237] Train net output #0: loss = 2.10995 (* 1 = 2.10995 loss) I0409 23:43:31.273607 8290 sgd_solver.cpp:105] Iteration 3072, lr = 0.00544156 I0409 23:43:36.204576 8290 solver.cpp:218] Iteration 3084 (2.43369 iter/s, 4.93079s/12 iters), loss = 2.24212 I0409 23:43:36.204636 8290 solver.cpp:237] Train net output #0: loss = 2.24212 (* 1 = 2.24212 loss) I0409 23:43:36.204649 8290 sgd_solver.cpp:105] Iteration 3084, lr = 0.00542864 I0409 23:43:41.600889 8290 solver.cpp:218] Iteration 3096 (2.22385 iter/s, 5.39605s/12 iters), loss = 2.26911 I0409 23:43:41.600941 8290 solver.cpp:237] Train net output #0: loss = 2.26911 (* 1 = 2.26911 loss) I0409 23:43:41.600953 8290 sgd_solver.cpp:105] Iteration 3096, lr = 0.00541575 I0409 23:43:46.522462 8290 solver.cpp:218] Iteration 3108 (2.43836 iter/s, 4.92134s/12 iters), loss = 2.19154 I0409 23:43:46.522514 8290 solver.cpp:237] Train net output #0: loss = 2.19154 (* 1 = 2.19154 loss) I0409 23:43:46.522527 8290 sgd_solver.cpp:105] Iteration 3108, lr = 0.00540289 I0409 23:43:51.441433 8290 solver.cpp:218] Iteration 3120 (2.43965 iter/s, 4.91874s/12 iters), loss = 1.95425 I0409 23:43:51.441540 8290 solver.cpp:237] Train net output #0: loss = 1.95425 (* 1 = 1.95425 loss) I0409 23:43:51.441550 8290 sgd_solver.cpp:105] Iteration 3120, lr = 0.00539006 I0409 23:43:56.364061 8290 solver.cpp:218] Iteration 3132 (2.43787 iter/s, 4.92234s/12 iters), loss = 2.20819 I0409 23:43:56.364104 8290 solver.cpp:237] Train net output #0: loss = 2.20819 (* 1 = 2.20819 loss) I0409 23:43:56.364113 8290 sgd_solver.cpp:105] Iteration 3132, lr = 0.00537727 I0409 23:43:57.479707 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:44:01.295049 8290 solver.cpp:218] Iteration 3144 (2.4337 iter/s, 4.93077s/12 iters), loss = 1.97696 I0409 23:44:01.295091 8290 solver.cpp:237] Train net output #0: loss = 1.97696 (* 1 = 1.97696 loss) I0409 23:44:01.295099 8290 sgd_solver.cpp:105] Iteration 3144, lr = 0.0053645 I0409 23:44:06.289306 8290 solver.cpp:218] Iteration 3156 (2.40287 iter/s, 4.99402s/12 iters), loss = 2.174 I0409 23:44:06.289363 8290 solver.cpp:237] Train net output #0: loss = 2.174 (* 1 = 2.174 loss) I0409 23:44:06.289376 8290 sgd_solver.cpp:105] Iteration 3156, lr = 0.00535176 I0409 23:44:08.240392 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3162.caffemodel I0409 23:44:08.969074 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3162.solverstate I0409 23:44:09.497838 8290 solver.cpp:330] Iteration 3162, Testing net (#0) I0409 23:44:09.497866 8290 net.cpp:676] Ignoring source layer train-data I0409 23:44:12.629631 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:44:13.889091 8290 solver.cpp:397] Test net output #0: accuracy = 0.368873 I0409 23:44:13.889137 8290 solver.cpp:397] Test net output #1: loss = 2.57876 (* 1 = 2.57876 loss) I0409 23:44:15.651801 8290 solver.cpp:218] Iteration 3168 (1.28176 iter/s, 9.3621s/12 iters), loss = 2.00537 I0409 23:44:15.651849 8290 solver.cpp:237] Train net output #0: loss = 2.00537 (* 1 = 2.00537 loss) I0409 23:44:15.651856 8290 sgd_solver.cpp:105] Iteration 3168, lr = 0.00533906 I0409 23:44:20.568625 8290 solver.cpp:218] Iteration 3180 (2.44071 iter/s, 4.9166s/12 iters), loss = 2.22928 I0409 23:44:20.568670 8290 solver.cpp:237] Train net output #0: loss = 2.22928 (* 1 = 2.22928 loss) I0409 23:44:20.568679 8290 sgd_solver.cpp:105] Iteration 3180, lr = 0.00532638 I0409 23:44:25.484344 8290 solver.cpp:218] Iteration 3192 (2.44126 iter/s, 4.91549s/12 iters), loss = 2.18067 I0409 23:44:25.484496 8290 solver.cpp:237] Train net output #0: loss = 2.18067 (* 1 = 2.18067 loss) I0409 23:44:25.484510 8290 sgd_solver.cpp:105] Iteration 3192, lr = 0.00531374 I0409 23:44:30.363672 8290 solver.cpp:218] Iteration 3204 (2.45952 iter/s, 4.879s/12 iters), loss = 1.97483 I0409 23:44:30.363723 8290 solver.cpp:237] Train net output #0: loss = 1.97483 (* 1 = 1.97483 loss) I0409 23:44:30.363734 8290 sgd_solver.cpp:105] Iteration 3204, lr = 0.00530112 I0409 23:44:35.297948 8290 solver.cpp:218] Iteration 3216 (2.43208 iter/s, 4.93404s/12 iters), loss = 2.32981 I0409 23:44:35.298022 8290 solver.cpp:237] Train net output #0: loss = 2.32981 (* 1 = 2.32981 loss) I0409 23:44:35.298033 8290 sgd_solver.cpp:105] Iteration 3216, lr = 0.00528853 I0409 23:44:40.253381 8290 solver.cpp:218] Iteration 3228 (2.42171 iter/s, 4.95517s/12 iters), loss = 2.16532 I0409 23:44:40.253430 8290 solver.cpp:237] Train net output #0: loss = 2.16532 (* 1 = 2.16532 loss) I0409 23:44:40.253443 8290 sgd_solver.cpp:105] Iteration 3228, lr = 0.00527598 I0409 23:44:43.422979 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:44:45.141145 8290 solver.cpp:218] Iteration 3240 (2.45523 iter/s, 4.88753s/12 iters), loss = 2.07136 I0409 23:44:45.141201 8290 solver.cpp:237] Train net output #0: loss = 2.07136 (* 1 = 2.07136 loss) I0409 23:44:45.141211 8290 sgd_solver.cpp:105] Iteration 3240, lr = 0.00526345 I0409 23:44:50.085705 8290 solver.cpp:218] Iteration 3252 (2.42703 iter/s, 4.94432s/12 iters), loss = 2.1272 I0409 23:44:50.085749 8290 solver.cpp:237] Train net output #0: loss = 2.1272 (* 1 = 2.1272 loss) I0409 23:44:50.085758 8290 sgd_solver.cpp:105] Iteration 3252, lr = 0.00525095 I0409 23:44:54.501513 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3264.caffemodel I0409 23:44:55.200434 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3264.solverstate I0409 23:44:55.729660 8290 solver.cpp:330] Iteration 3264, Testing net (#0) I0409 23:44:55.729741 8290 net.cpp:676] Ignoring source layer train-data I0409 23:44:58.870083 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:45:00.166416 8290 solver.cpp:397] Test net output #0: accuracy = 0.386029 I0409 23:45:00.166468 8290 solver.cpp:397] Test net output #1: loss = 2.42004 (* 1 = 2.42004 loss) I0409 23:45:00.248785 8290 solver.cpp:218] Iteration 3264 (1.18079 iter/s, 10.1627s/12 iters), loss = 2.17537 I0409 23:45:00.248843 8290 solver.cpp:237] Train net output #0: loss = 2.17537 (* 1 = 2.17537 loss) I0409 23:45:00.248855 8290 sgd_solver.cpp:105] Iteration 3264, lr = 0.00523849 I0409 23:45:04.344063 8290 solver.cpp:218] Iteration 3276 (2.93036 iter/s, 4.09506s/12 iters), loss = 2.05759 I0409 23:45:04.344123 8290 solver.cpp:237] Train net output #0: loss = 2.05759 (* 1 = 2.05759 loss) I0409 23:45:04.344136 8290 sgd_solver.cpp:105] Iteration 3276, lr = 0.00522605 I0409 23:45:09.306692 8290 solver.cpp:218] Iteration 3288 (2.41819 iter/s, 4.96239s/12 iters), loss = 2.05552 I0409 23:45:09.306740 8290 solver.cpp:237] Train net output #0: loss = 2.05552 (* 1 = 2.05552 loss) I0409 23:45:09.306751 8290 sgd_solver.cpp:105] Iteration 3288, lr = 0.00521364 I0409 23:45:14.400722 8290 solver.cpp:218] Iteration 3300 (2.35581 iter/s, 5.09379s/12 iters), loss = 2.26359 I0409 23:45:14.400763 8290 solver.cpp:237] Train net output #0: loss = 2.26359 (* 1 = 2.26359 loss) I0409 23:45:14.400772 8290 sgd_solver.cpp:105] Iteration 3300, lr = 0.00520126 I0409 23:45:19.610350 8290 solver.cpp:218] Iteration 3312 (2.30353 iter/s, 5.20939s/12 iters), loss = 2.15509 I0409 23:45:19.610406 8290 solver.cpp:237] Train net output #0: loss = 2.15509 (* 1 = 2.15509 loss) I0409 23:45:19.610419 8290 sgd_solver.cpp:105] Iteration 3312, lr = 0.00518892 I0409 23:45:24.448082 8290 solver.cpp:218] Iteration 3324 (2.48062 iter/s, 4.8375s/12 iters), loss = 1.88312 I0409 23:45:24.448135 8290 solver.cpp:237] Train net output #0: loss = 1.88312 (* 1 = 1.88312 loss) I0409 23:45:24.448146 8290 sgd_solver.cpp:105] Iteration 3324, lr = 0.0051766 I0409 23:45:29.342063 8290 solver.cpp:218] Iteration 3336 (2.45211 iter/s, 4.89374s/12 iters), loss = 1.8928 I0409 23:45:29.342202 8290 solver.cpp:237] Train net output #0: loss = 1.8928 (* 1 = 1.8928 loss) I0409 23:45:29.342211 8290 sgd_solver.cpp:105] Iteration 3336, lr = 0.00516431 I0409 23:45:29.802016 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:45:34.286425 8290 solver.cpp:218] Iteration 3348 (2.42716 iter/s, 4.94404s/12 iters), loss = 2.09763 I0409 23:45:34.286466 8290 solver.cpp:237] Train net output #0: loss = 2.09763 (* 1 = 2.09763 loss) I0409 23:45:34.286475 8290 sgd_solver.cpp:105] Iteration 3348, lr = 0.00515204 I0409 23:45:39.156406 8290 solver.cpp:218] Iteration 3360 (2.46419 iter/s, 4.86976s/12 iters), loss = 1.90158 I0409 23:45:39.156450 8290 solver.cpp:237] Train net output #0: loss = 1.90158 (* 1 = 1.90158 loss) I0409 23:45:39.156459 8290 sgd_solver.cpp:105] Iteration 3360, lr = 0.00513981 I0409 23:45:41.160094 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3366.caffemodel I0409 23:45:41.908550 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3366.solverstate I0409 23:45:42.433007 8290 solver.cpp:330] Iteration 3366, Testing net (#0) I0409 23:45:42.433027 8290 net.cpp:676] Ignoring source layer train-data I0409 23:45:45.699524 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:45:47.374178 8290 solver.cpp:397] Test net output #0: accuracy = 0.398284 I0409 23:45:47.374217 8290 solver.cpp:397] Test net output #1: loss = 2.39045 (* 1 = 2.39045 loss) I0409 23:45:49.142846 8290 solver.cpp:218] Iteration 3372 (1.20168 iter/s, 9.98604s/12 iters), loss = 1.86262 I0409 23:45:49.142902 8290 solver.cpp:237] Train net output #0: loss = 1.86262 (* 1 = 1.86262 loss) I0409 23:45:49.142915 8290 sgd_solver.cpp:105] Iteration 3372, lr = 0.00512761 I0409 23:45:54.049203 8290 solver.cpp:218] Iteration 3384 (2.44593 iter/s, 4.90611s/12 iters), loss = 2.04228 I0409 23:45:54.049261 8290 solver.cpp:237] Train net output #0: loss = 2.04228 (* 1 = 2.04228 loss) I0409 23:45:54.049273 8290 sgd_solver.cpp:105] Iteration 3384, lr = 0.00511544 I0409 23:45:58.906373 8290 solver.cpp:218] Iteration 3396 (2.4707 iter/s, 4.85693s/12 iters), loss = 1.8072 I0409 23:45:58.906421 8290 solver.cpp:237] Train net output #0: loss = 1.8072 (* 1 = 1.8072 loss) I0409 23:45:58.906430 8290 sgd_solver.cpp:105] Iteration 3396, lr = 0.00510329 I0409 23:46:03.773345 8290 solver.cpp:218] Iteration 3408 (2.46572 iter/s, 4.86674s/12 iters), loss = 2.16614 I0409 23:46:03.773447 8290 solver.cpp:237] Train net output #0: loss = 2.16614 (* 1 = 2.16614 loss) I0409 23:46:03.773458 8290 sgd_solver.cpp:105] Iteration 3408, lr = 0.00509117 I0409 23:46:08.657644 8290 solver.cpp:218] Iteration 3420 (2.45699 iter/s, 4.88402s/12 iters), loss = 1.97075 I0409 23:46:08.657686 8290 solver.cpp:237] Train net output #0: loss = 1.97075 (* 1 = 1.97075 loss) I0409 23:46:08.657696 8290 sgd_solver.cpp:105] Iteration 3420, lr = 0.00507909 I0409 23:46:13.581550 8290 solver.cpp:218] Iteration 3432 (2.4372 iter/s, 4.92368s/12 iters), loss = 2.1129 I0409 23:46:13.581601 8290 solver.cpp:237] Train net output #0: loss = 2.1129 (* 1 = 2.1129 loss) I0409 23:46:13.581612 8290 sgd_solver.cpp:105] Iteration 3432, lr = 0.00506703 I0409 23:46:16.189293 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:46:18.609823 8290 solver.cpp:218] Iteration 3444 (2.38662 iter/s, 5.02803s/12 iters), loss = 1.72469 I0409 23:46:18.609884 8290 solver.cpp:237] Train net output #0: loss = 1.72469 (* 1 = 1.72469 loss) I0409 23:46:18.609895 8290 sgd_solver.cpp:105] Iteration 3444, lr = 0.005055 I0409 23:46:23.533217 8290 solver.cpp:218] Iteration 3456 (2.43746 iter/s, 4.92315s/12 iters), loss = 1.89476 I0409 23:46:23.533264 8290 solver.cpp:237] Train net output #0: loss = 1.89476 (* 1 = 1.89476 loss) I0409 23:46:23.533273 8290 sgd_solver.cpp:105] Iteration 3456, lr = 0.005043 I0409 23:46:27.952025 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3468.caffemodel I0409 23:46:28.722370 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3468.solverstate I0409 23:46:29.278995 8290 solver.cpp:330] Iteration 3468, Testing net (#0) I0409 23:46:29.279024 8290 net.cpp:676] Ignoring source layer train-data I0409 23:46:29.437665 8290 blocking_queue.cpp:49] Waiting for data I0409 23:46:32.440157 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:46:33.815349 8290 solver.cpp:397] Test net output #0: accuracy = 0.421569 I0409 23:46:33.815460 8290 solver.cpp:397] Test net output #1: loss = 2.2615 (* 1 = 2.2615 loss) I0409 23:46:33.897562 8290 solver.cpp:218] Iteration 3468 (1.15786 iter/s, 10.3639s/12 iters), loss = 1.78716 I0409 23:46:33.897599 8290 solver.cpp:237] Train net output #0: loss = 1.78716 (* 1 = 1.78716 loss) I0409 23:46:33.897608 8290 sgd_solver.cpp:105] Iteration 3468, lr = 0.00503102 I0409 23:46:37.945580 8290 solver.cpp:218] Iteration 3480 (2.96456 iter/s, 4.04782s/12 iters), loss = 1.73026 I0409 23:46:37.945633 8290 solver.cpp:237] Train net output #0: loss = 1.73026 (* 1 = 1.73026 loss) I0409 23:46:37.945647 8290 sgd_solver.cpp:105] Iteration 3480, lr = 0.00501908 I0409 23:46:42.825986 8290 solver.cpp:218] Iteration 3492 (2.45894 iter/s, 4.88016s/12 iters), loss = 1.87035 I0409 23:46:42.826037 8290 solver.cpp:237] Train net output #0: loss = 1.87035 (* 1 = 1.87035 loss) I0409 23:46:42.826047 8290 sgd_solver.cpp:105] Iteration 3492, lr = 0.00500716 I0409 23:46:47.653174 8290 solver.cpp:218] Iteration 3504 (2.48604 iter/s, 4.82695s/12 iters), loss = 1.94196 I0409 23:46:47.653234 8290 solver.cpp:237] Train net output #0: loss = 1.94196 (* 1 = 1.94196 loss) I0409 23:46:47.653246 8290 sgd_solver.cpp:105] Iteration 3504, lr = 0.00499527 I0409 23:46:52.564391 8290 solver.cpp:218] Iteration 3516 (2.44351 iter/s, 4.91097s/12 iters), loss = 1.73156 I0409 23:46:52.564447 8290 solver.cpp:237] Train net output #0: loss = 1.73156 (* 1 = 1.73156 loss) I0409 23:46:52.564458 8290 sgd_solver.cpp:105] Iteration 3516, lr = 0.00498341 I0409 23:46:57.508980 8290 solver.cpp:218] Iteration 3528 (2.42701 iter/s, 4.94435s/12 iters), loss = 1.75489 I0409 23:46:57.509038 8290 solver.cpp:237] Train net output #0: loss = 1.75489 (* 1 = 1.75489 loss) I0409 23:46:57.509052 8290 sgd_solver.cpp:105] Iteration 3528, lr = 0.00497158 I0409 23:47:02.195500 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:47:02.455960 8290 solver.cpp:218] Iteration 3540 (2.42584 iter/s, 4.94674s/12 iters), loss = 1.74277 I0409 23:47:02.456017 8290 solver.cpp:237] Train net output #0: loss = 1.74277 (* 1 = 1.74277 loss) I0409 23:47:02.456029 8290 sgd_solver.cpp:105] Iteration 3540, lr = 0.00495978 I0409 23:47:07.469157 8290 solver.cpp:218] Iteration 3552 (2.3938 iter/s, 5.01296s/12 iters), loss = 1.57062 I0409 23:47:07.469270 8290 solver.cpp:237] Train net output #0: loss = 1.57062 (* 1 = 1.57062 loss) I0409 23:47:07.469280 8290 sgd_solver.cpp:105] Iteration 3552, lr = 0.004948 I0409 23:47:12.381736 8290 solver.cpp:218] Iteration 3564 (2.44286 iter/s, 4.91228s/12 iters), loss = 1.7567 I0409 23:47:12.381796 8290 solver.cpp:237] Train net output #0: loss = 1.7567 (* 1 = 1.7567 loss) I0409 23:47:12.381809 8290 sgd_solver.cpp:105] Iteration 3564, lr = 0.00493626 I0409 23:47:14.389802 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3570.caffemodel I0409 23:47:15.685762 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3570.solverstate I0409 23:47:17.318701 8290 solver.cpp:330] Iteration 3570, Testing net (#0) I0409 23:47:17.318730 8290 net.cpp:676] Ignoring source layer train-data I0409 23:47:20.371796 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:47:21.794375 8290 solver.cpp:397] Test net output #0: accuracy = 0.424632 I0409 23:47:21.794420 8290 solver.cpp:397] Test net output #1: loss = 2.30742 (* 1 = 2.30742 loss) I0409 23:47:23.808280 8290 solver.cpp:218] Iteration 3576 (1.05023 iter/s, 11.4261s/12 iters), loss = 1.81491 I0409 23:47:23.808331 8290 solver.cpp:237] Train net output #0: loss = 1.81491 (* 1 = 1.81491 loss) I0409 23:47:23.808342 8290 sgd_solver.cpp:105] Iteration 3576, lr = 0.00492454 I0409 23:47:28.749543 8290 solver.cpp:218] Iteration 3588 (2.42865 iter/s, 4.94102s/12 iters), loss = 1.63647 I0409 23:47:28.749593 8290 solver.cpp:237] Train net output #0: loss = 1.63647 (* 1 = 1.63647 loss) I0409 23:47:28.749605 8290 sgd_solver.cpp:105] Iteration 3588, lr = 0.00491284 I0409 23:47:33.615478 8290 solver.cpp:218] Iteration 3600 (2.46624 iter/s, 4.8657s/12 iters), loss = 1.89781 I0409 23:47:33.615535 8290 solver.cpp:237] Train net output #0: loss = 1.89781 (* 1 = 1.89781 loss) I0409 23:47:33.615545 8290 sgd_solver.cpp:105] Iteration 3600, lr = 0.00490118 I0409 23:47:38.536669 8290 solver.cpp:218] Iteration 3612 (2.43855 iter/s, 4.92095s/12 iters), loss = 1.58957 I0409 23:47:38.536798 8290 solver.cpp:237] Train net output #0: loss = 1.58957 (* 1 = 1.58957 loss) I0409 23:47:38.536808 8290 sgd_solver.cpp:105] Iteration 3612, lr = 0.00488954 I0409 23:47:43.396014 8290 solver.cpp:218] Iteration 3624 (2.46962 iter/s, 4.85904s/12 iters), loss = 1.77983 I0409 23:47:43.396059 8290 solver.cpp:237] Train net output #0: loss = 1.77983 (* 1 = 1.77983 loss) I0409 23:47:43.396071 8290 sgd_solver.cpp:105] Iteration 3624, lr = 0.00487793 I0409 23:47:48.320739 8290 solver.cpp:218] Iteration 3636 (2.4368 iter/s, 4.92449s/12 iters), loss = 1.87013 I0409 23:47:48.320792 8290 solver.cpp:237] Train net output #0: loss = 1.87013 (* 1 = 1.87013 loss) I0409 23:47:48.320803 8290 sgd_solver.cpp:105] Iteration 3636, lr = 0.00486635 I0409 23:47:50.204185 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:47:53.279423 8290 solver.cpp:218] Iteration 3648 (2.42011 iter/s, 4.95844s/12 iters), loss = 1.71929 I0409 23:47:53.279479 8290 solver.cpp:237] Train net output #0: loss = 1.71929 (* 1 = 1.71929 loss) I0409 23:47:53.279492 8290 sgd_solver.cpp:105] Iteration 3648, lr = 0.0048548 I0409 23:47:58.237931 8290 solver.cpp:218] Iteration 3660 (2.4202 iter/s, 4.95827s/12 iters), loss = 1.64435 I0409 23:47:58.237998 8290 solver.cpp:237] Train net output #0: loss = 1.64435 (* 1 = 1.64435 loss) I0409 23:47:58.238009 8290 sgd_solver.cpp:105] Iteration 3660, lr = 0.00484327 I0409 23:48:02.689723 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3672.caffemodel I0409 23:48:03.439018 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3672.solverstate I0409 23:48:03.961390 8290 solver.cpp:330] Iteration 3672, Testing net (#0) I0409 23:48:03.961419 8290 net.cpp:676] Ignoring source layer train-data I0409 23:48:06.987442 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:48:08.466853 8290 solver.cpp:397] Test net output #0: accuracy = 0.422181 I0409 23:48:08.466902 8290 solver.cpp:397] Test net output #1: loss = 2.28559 (* 1 = 2.28559 loss) I0409 23:48:08.548833 8290 solver.cpp:218] Iteration 3672 (1.16387 iter/s, 10.3105s/12 iters), loss = 1.53923 I0409 23:48:08.548945 8290 solver.cpp:237] Train net output #0: loss = 1.53923 (* 1 = 1.53923 loss) I0409 23:48:08.548959 8290 sgd_solver.cpp:105] Iteration 3672, lr = 0.00483177 I0409 23:48:12.621711 8290 solver.cpp:218] Iteration 3684 (2.94651 iter/s, 4.07262s/12 iters), loss = 1.715 I0409 23:48:12.621749 8290 solver.cpp:237] Train net output #0: loss = 1.715 (* 1 = 1.715 loss) I0409 23:48:12.621757 8290 sgd_solver.cpp:105] Iteration 3684, lr = 0.0048203 I0409 23:48:17.500367 8290 solver.cpp:218] Iteration 3696 (2.45981 iter/s, 4.87843s/12 iters), loss = 1.4571 I0409 23:48:17.500420 8290 solver.cpp:237] Train net output #0: loss = 1.4571 (* 1 = 1.4571 loss) I0409 23:48:17.500433 8290 sgd_solver.cpp:105] Iteration 3696, lr = 0.00480886 I0409 23:48:22.345739 8290 solver.cpp:218] Iteration 3708 (2.47671 iter/s, 4.84513s/12 iters), loss = 1.78765 I0409 23:48:22.345803 8290 solver.cpp:237] Train net output #0: loss = 1.78765 (* 1 = 1.78765 loss) I0409 23:48:22.345815 8290 sgd_solver.cpp:105] Iteration 3708, lr = 0.00479744 I0409 23:48:27.176802 8290 solver.cpp:218] Iteration 3720 (2.48405 iter/s, 4.83082s/12 iters), loss = 1.75992 I0409 23:48:27.176851 8290 solver.cpp:237] Train net output #0: loss = 1.75992 (* 1 = 1.75992 loss) I0409 23:48:27.176862 8290 sgd_solver.cpp:105] Iteration 3720, lr = 0.00478605 I0409 23:48:32.039196 8290 solver.cpp:218] Iteration 3732 (2.46804 iter/s, 4.86216s/12 iters), loss = 1.54473 I0409 23:48:32.039247 8290 solver.cpp:237] Train net output #0: loss = 1.54473 (* 1 = 1.54473 loss) I0409 23:48:32.039259 8290 sgd_solver.cpp:105] Iteration 3732, lr = 0.00477469 I0409 23:48:35.975546 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:48:36.949851 8290 solver.cpp:218] Iteration 3744 (2.44378 iter/s, 4.91042s/12 iters), loss = 1.36617 I0409 23:48:36.949906 8290 solver.cpp:237] Train net output #0: loss = 1.36617 (* 1 = 1.36617 loss) I0409 23:48:36.949918 8290 sgd_solver.cpp:105] Iteration 3744, lr = 0.00476335 I0409 23:48:41.853332 8290 solver.cpp:218] Iteration 3756 (2.44736 iter/s, 4.90324s/12 iters), loss = 1.72069 I0409 23:48:41.853480 8290 solver.cpp:237] Train net output #0: loss = 1.72069 (* 1 = 1.72069 loss) I0409 23:48:41.853494 8290 sgd_solver.cpp:105] Iteration 3756, lr = 0.00475204 I0409 23:48:46.738883 8290 solver.cpp:218] Iteration 3768 (2.45639 iter/s, 4.88522s/12 iters), loss = 1.55933 I0409 23:48:46.738936 8290 solver.cpp:237] Train net output #0: loss = 1.55933 (* 1 = 1.55933 loss) I0409 23:48:46.738947 8290 sgd_solver.cpp:105] Iteration 3768, lr = 0.00474076 I0409 23:48:48.709553 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3774.caffemodel I0409 23:48:49.630421 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3774.solverstate I0409 23:48:51.620402 8290 solver.cpp:330] Iteration 3774, Testing net (#0) I0409 23:48:51.620431 8290 net.cpp:676] Ignoring source layer train-data I0409 23:48:54.565026 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:48:56.054229 8290 solver.cpp:397] Test net output #0: accuracy = 0.444853 I0409 23:48:56.054275 8290 solver.cpp:397] Test net output #1: loss = 2.20636 (* 1 = 2.20636 loss) I0409 23:48:57.856783 8290 solver.cpp:218] Iteration 3780 (1.07938 iter/s, 11.1175s/12 iters), loss = 1.61166 I0409 23:48:57.856827 8290 solver.cpp:237] Train net output #0: loss = 1.61166 (* 1 = 1.61166 loss) I0409 23:48:57.856835 8290 sgd_solver.cpp:105] Iteration 3780, lr = 0.00472951 I0409 23:49:02.756069 8290 solver.cpp:218] Iteration 3792 (2.44945 iter/s, 4.89906s/12 iters), loss = 1.75684 I0409 23:49:02.756129 8290 solver.cpp:237] Train net output #0: loss = 1.75684 (* 1 = 1.75684 loss) I0409 23:49:02.756141 8290 sgd_solver.cpp:105] Iteration 3792, lr = 0.00471828 I0409 23:49:07.678290 8290 solver.cpp:218] Iteration 3804 (2.43804 iter/s, 4.92198s/12 iters), loss = 1.59822 I0409 23:49:07.678347 8290 solver.cpp:237] Train net output #0: loss = 1.59822 (* 1 = 1.59822 loss) I0409 23:49:07.678360 8290 sgd_solver.cpp:105] Iteration 3804, lr = 0.00470707 I0409 23:49:12.589466 8290 solver.cpp:218] Iteration 3816 (2.44353 iter/s, 4.91093s/12 iters), loss = 1.68191 I0409 23:49:12.589637 8290 solver.cpp:237] Train net output #0: loss = 1.68191 (* 1 = 1.68191 loss) I0409 23:49:12.589650 8290 sgd_solver.cpp:105] Iteration 3816, lr = 0.0046959 I0409 23:49:17.516793 8290 solver.cpp:218] Iteration 3828 (2.43557 iter/s, 4.92698s/12 iters), loss = 1.4884 I0409 23:49:17.516846 8290 solver.cpp:237] Train net output #0: loss = 1.4884 (* 1 = 1.4884 loss) I0409 23:49:17.516860 8290 sgd_solver.cpp:105] Iteration 3828, lr = 0.00468475 I0409 23:49:22.450219 8290 solver.cpp:218] Iteration 3840 (2.4325 iter/s, 4.93319s/12 iters), loss = 1.65363 I0409 23:49:22.450274 8290 solver.cpp:237] Train net output #0: loss = 1.65363 (* 1 = 1.65363 loss) I0409 23:49:22.450287 8290 sgd_solver.cpp:105] Iteration 3840, lr = 0.00467363 I0409 23:49:23.569676 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:49:27.373251 8290 solver.cpp:218] Iteration 3852 (2.43764 iter/s, 4.92279s/12 iters), loss = 1.39563 I0409 23:49:27.373302 8290 solver.cpp:237] Train net output #0: loss = 1.39563 (* 1 = 1.39563 loss) I0409 23:49:27.373313 8290 sgd_solver.cpp:105] Iteration 3852, lr = 0.00466253 I0409 23:49:32.224956 8290 solver.cpp:218] Iteration 3864 (2.47347 iter/s, 4.85148s/12 iters), loss = 1.64109 I0409 23:49:32.224999 8290 solver.cpp:237] Train net output #0: loss = 1.64109 (* 1 = 1.64109 loss) I0409 23:49:32.225008 8290 sgd_solver.cpp:105] Iteration 3864, lr = 0.00465146 I0409 23:49:36.727711 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3876.caffemodel I0409 23:49:38.102047 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3876.solverstate I0409 23:49:38.698487 8290 solver.cpp:330] Iteration 3876, Testing net (#0) I0409 23:49:38.698514 8290 net.cpp:676] Ignoring source layer train-data I0409 23:49:41.749194 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:49:43.285117 8290 solver.cpp:397] Test net output #0: accuracy = 0.4375 I0409 23:49:43.285212 8290 solver.cpp:397] Test net output #1: loss = 2.23677 (* 1 = 2.23677 loss) I0409 23:49:43.367342 8290 solver.cpp:218] Iteration 3876 (1.07701 iter/s, 11.1419s/12 iters), loss = 1.28539 I0409 23:49:43.367391 8290 solver.cpp:237] Train net output #0: loss = 1.28539 (* 1 = 1.28539 loss) I0409 23:49:43.367401 8290 sgd_solver.cpp:105] Iteration 3876, lr = 0.00464042 I0409 23:49:47.519464 8290 solver.cpp:218] Iteration 3888 (2.89024 iter/s, 4.15191s/12 iters), loss = 1.31226 I0409 23:49:47.519523 8290 solver.cpp:237] Train net output #0: loss = 1.31226 (* 1 = 1.31226 loss) I0409 23:49:47.519536 8290 sgd_solver.cpp:105] Iteration 3888, lr = 0.0046294 I0409 23:49:52.439903 8290 solver.cpp:218] Iteration 3900 (2.43893 iter/s, 4.92019s/12 iters), loss = 1.64808 I0409 23:49:52.439958 8290 solver.cpp:237] Train net output #0: loss = 1.64808 (* 1 = 1.64808 loss) I0409 23:49:52.439972 8290 sgd_solver.cpp:105] Iteration 3900, lr = 0.00461841 I0409 23:49:57.465807 8290 solver.cpp:218] Iteration 3912 (2.38774 iter/s, 5.02566s/12 iters), loss = 1.5681 I0409 23:49:57.465848 8290 solver.cpp:237] Train net output #0: loss = 1.5681 (* 1 = 1.5681 loss) I0409 23:49:57.465858 8290 sgd_solver.cpp:105] Iteration 3912, lr = 0.00460744 I0409 23:50:02.369686 8290 solver.cpp:218] Iteration 3924 (2.44716 iter/s, 4.90365s/12 iters), loss = 1.87503 I0409 23:50:02.369740 8290 solver.cpp:237] Train net output #0: loss = 1.87503 (* 1 = 1.87503 loss) I0409 23:50:02.369752 8290 sgd_solver.cpp:105] Iteration 3924, lr = 0.0045965 I0409 23:50:07.314265 8290 solver.cpp:218] Iteration 3936 (2.42702 iter/s, 4.94434s/12 iters), loss = 1.37226 I0409 23:50:07.314311 8290 solver.cpp:237] Train net output #0: loss = 1.37226 (* 1 = 1.37226 loss) I0409 23:50:07.314321 8290 sgd_solver.cpp:105] Iteration 3936, lr = 0.00458559 I0409 23:50:10.603907 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:50:12.207752 8290 solver.cpp:218] Iteration 3948 (2.45235 iter/s, 4.89326s/12 iters), loss = 1.3524 I0409 23:50:12.207805 8290 solver.cpp:237] Train net output #0: loss = 1.3524 (* 1 = 1.3524 loss) I0409 23:50:12.207818 8290 sgd_solver.cpp:105] Iteration 3948, lr = 0.0045747 I0409 23:50:17.125541 8290 solver.cpp:218] Iteration 3960 (2.44024 iter/s, 4.91755s/12 iters), loss = 1.39645 I0409 23:50:17.125635 8290 solver.cpp:237] Train net output #0: loss = 1.39645 (* 1 = 1.39645 loss) I0409 23:50:17.125644 8290 sgd_solver.cpp:105] Iteration 3960, lr = 0.00456384 I0409 23:50:22.082315 8290 solver.cpp:218] Iteration 3972 (2.42106 iter/s, 4.9565s/12 iters), loss = 1.52334 I0409 23:50:22.082357 8290 solver.cpp:237] Train net output #0: loss = 1.52334 (* 1 = 1.52334 loss) I0409 23:50:22.082365 8290 sgd_solver.cpp:105] Iteration 3972, lr = 0.00455301 I0409 23:50:24.089246 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_3978.caffemodel I0409 23:50:24.861644 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_3978.solverstate I0409 23:50:25.386670 8290 solver.cpp:330] Iteration 3978, Testing net (#0) I0409 23:50:25.386698 8290 net.cpp:676] Ignoring source layer train-data I0409 23:50:28.246001 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:50:29.820086 8290 solver.cpp:397] Test net output #0: accuracy = 0.435049 I0409 23:50:29.820124 8290 solver.cpp:397] Test net output #1: loss = 2.33587 (* 1 = 2.33587 loss) I0409 23:50:31.715505 8290 solver.cpp:218] Iteration 3984 (1.24574 iter/s, 9.63279s/12 iters), loss = 1.36613 I0409 23:50:31.715555 8290 solver.cpp:237] Train net output #0: loss = 1.36613 (* 1 = 1.36613 loss) I0409 23:50:31.715564 8290 sgd_solver.cpp:105] Iteration 3984, lr = 0.0045422 I0409 23:50:36.614022 8290 solver.cpp:218] Iteration 3996 (2.44984 iter/s, 4.89828s/12 iters), loss = 1.33713 I0409 23:50:36.614075 8290 solver.cpp:237] Train net output #0: loss = 1.33713 (* 1 = 1.33713 loss) I0409 23:50:36.614086 8290 sgd_solver.cpp:105] Iteration 3996, lr = 0.00453141 I0409 23:50:41.469730 8290 solver.cpp:218] Iteration 4008 (2.47144 iter/s, 4.85547s/12 iters), loss = 1.47918 I0409 23:50:41.469781 8290 solver.cpp:237] Train net output #0: loss = 1.47918 (* 1 = 1.47918 loss) I0409 23:50:41.469792 8290 sgd_solver.cpp:105] Iteration 4008, lr = 0.00452066 I0409 23:50:46.377209 8290 solver.cpp:218] Iteration 4020 (2.44536 iter/s, 4.90725s/12 iters), loss = 1.88098 I0409 23:50:46.377251 8290 solver.cpp:237] Train net output #0: loss = 1.88098 (* 1 = 1.88098 loss) I0409 23:50:46.377260 8290 sgd_solver.cpp:105] Iteration 4020, lr = 0.00450992 I0409 23:50:51.308956 8290 solver.cpp:218] Iteration 4032 (2.43333 iter/s, 4.93151s/12 iters), loss = 1.36177 I0409 23:50:51.309039 8290 solver.cpp:237] Train net output #0: loss = 1.36177 (* 1 = 1.36177 loss) I0409 23:50:51.309048 8290 sgd_solver.cpp:105] Iteration 4032, lr = 0.00449921 I0409 23:50:56.302120 8290 solver.cpp:218] Iteration 4044 (2.40342 iter/s, 4.99289s/12 iters), loss = 1.50081 I0409 23:50:56.302168 8290 solver.cpp:237] Train net output #0: loss = 1.50081 (* 1 = 1.50081 loss) I0409 23:50:56.302177 8290 sgd_solver.cpp:105] Iteration 4044, lr = 0.00448853 I0409 23:50:56.809135 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:51:01.183562 8290 solver.cpp:218] Iteration 4056 (2.45841 iter/s, 4.88121s/12 iters), loss = 1.35092 I0409 23:51:01.183607 8290 solver.cpp:237] Train net output #0: loss = 1.35092 (* 1 = 1.35092 loss) I0409 23:51:01.183616 8290 sgd_solver.cpp:105] Iteration 4056, lr = 0.00447788 I0409 23:51:06.041770 8290 solver.cpp:218] Iteration 4068 (2.47016 iter/s, 4.85798s/12 iters), loss = 1.32497 I0409 23:51:06.041813 8290 solver.cpp:237] Train net output #0: loss = 1.32497 (* 1 = 1.32497 loss) I0409 23:51:06.041822 8290 sgd_solver.cpp:105] Iteration 4068, lr = 0.00446724 I0409 23:51:10.506534 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4080.caffemodel I0409 23:51:11.854439 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4080.solverstate I0409 23:51:14.214437 8290 solver.cpp:330] Iteration 4080, Testing net (#0) I0409 23:51:14.214462 8290 net.cpp:676] Ignoring source layer train-data I0409 23:51:17.104187 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:51:18.756247 8290 solver.cpp:397] Test net output #0: accuracy = 0.455882 I0409 23:51:18.756276 8290 solver.cpp:397] Test net output #1: loss = 2.23724 (* 1 = 2.23724 loss) I0409 23:51:18.838342 8290 solver.cpp:218] Iteration 4080 (0.937788 iter/s, 12.7961s/12 iters), loss = 1.48484 I0409 23:51:18.838392 8290 solver.cpp:237] Train net output #0: loss = 1.48484 (* 1 = 1.48484 loss) I0409 23:51:18.838402 8290 sgd_solver.cpp:105] Iteration 4080, lr = 0.00445664 I0409 23:51:23.031556 8290 solver.cpp:218] Iteration 4092 (2.86191 iter/s, 4.193s/12 iters), loss = 1.38393 I0409 23:51:23.031710 8290 solver.cpp:237] Train net output #0: loss = 1.38393 (* 1 = 1.38393 loss) I0409 23:51:23.031723 8290 sgd_solver.cpp:105] Iteration 4092, lr = 0.00444606 I0409 23:51:27.912693 8290 solver.cpp:218] Iteration 4104 (2.45861 iter/s, 4.88081s/12 iters), loss = 1.53786 I0409 23:51:27.912744 8290 solver.cpp:237] Train net output #0: loss = 1.53786 (* 1 = 1.53786 loss) I0409 23:51:27.912756 8290 sgd_solver.cpp:105] Iteration 4104, lr = 0.0044355 I0409 23:51:32.788465 8290 solver.cpp:218] Iteration 4116 (2.46127 iter/s, 4.87554s/12 iters), loss = 1.23407 I0409 23:51:32.788520 8290 solver.cpp:237] Train net output #0: loss = 1.23407 (* 1 = 1.23407 loss) I0409 23:51:32.788532 8290 sgd_solver.cpp:105] Iteration 4116, lr = 0.00442497 I0409 23:51:37.741430 8290 solver.cpp:218] Iteration 4128 (2.42291 iter/s, 4.95272s/12 iters), loss = 1.38823 I0409 23:51:37.741475 8290 solver.cpp:237] Train net output #0: loss = 1.38823 (* 1 = 1.38823 loss) I0409 23:51:37.741484 8290 sgd_solver.cpp:105] Iteration 4128, lr = 0.00441447 I0409 23:51:42.782907 8290 solver.cpp:218] Iteration 4140 (2.38037 iter/s, 5.04124s/12 iters), loss = 1.31468 I0409 23:51:42.782968 8290 solver.cpp:237] Train net output #0: loss = 1.31468 (* 1 = 1.31468 loss) I0409 23:51:42.782984 8290 sgd_solver.cpp:105] Iteration 4140, lr = 0.00440398 I0409 23:51:45.485509 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:51:47.811229 8290 solver.cpp:218] Iteration 4152 (2.3866 iter/s, 5.02807s/12 iters), loss = 1.29549 I0409 23:51:47.811285 8290 solver.cpp:237] Train net output #0: loss = 1.29549 (* 1 = 1.29549 loss) I0409 23:51:47.811297 8290 sgd_solver.cpp:105] Iteration 4152, lr = 0.00439353 I0409 23:51:48.166707 8290 blocking_queue.cpp:49] Waiting for data I0409 23:51:52.727006 8290 solver.cpp:218] Iteration 4164 (2.44123 iter/s, 4.91554s/12 iters), loss = 1.56223 I0409 23:51:52.727056 8290 solver.cpp:237] Train net output #0: loss = 1.56223 (* 1 = 1.56223 loss) I0409 23:51:52.727066 8290 sgd_solver.cpp:105] Iteration 4164, lr = 0.0043831 I0409 23:51:57.822247 8290 solver.cpp:218] Iteration 4176 (2.35525 iter/s, 5.09501s/12 iters), loss = 1.42918 I0409 23:51:57.822362 8290 solver.cpp:237] Train net output #0: loss = 1.42918 (* 1 = 1.42918 loss) I0409 23:51:57.822374 8290 sgd_solver.cpp:105] Iteration 4176, lr = 0.00437269 I0409 23:51:59.852980 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4182.caffemodel I0409 23:52:00.544700 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4182.solverstate I0409 23:52:01.078088 8290 solver.cpp:330] Iteration 4182, Testing net (#0) I0409 23:52:01.078116 8290 net.cpp:676] Ignoring source layer train-data I0409 23:52:03.958343 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:52:05.611275 8290 solver.cpp:397] Test net output #0: accuracy = 0.454044 I0409 23:52:05.611338 8290 solver.cpp:397] Test net output #1: loss = 2.19504 (* 1 = 2.19504 loss) I0409 23:52:07.512776 8290 solver.cpp:218] Iteration 4188 (1.23838 iter/s, 9.69007s/12 iters), loss = 1.57497 I0409 23:52:07.512830 8290 solver.cpp:237] Train net output #0: loss = 1.57497 (* 1 = 1.57497 loss) I0409 23:52:07.512842 8290 sgd_solver.cpp:105] Iteration 4188, lr = 0.00436231 I0409 23:52:12.411929 8290 solver.cpp:218] Iteration 4200 (2.44952 iter/s, 4.89891s/12 iters), loss = 1.38988 I0409 23:52:12.411981 8290 solver.cpp:237] Train net output #0: loss = 1.38988 (* 1 = 1.38988 loss) I0409 23:52:12.411993 8290 sgd_solver.cpp:105] Iteration 4200, lr = 0.00435195 I0409 23:52:17.352622 8290 solver.cpp:218] Iteration 4212 (2.42893 iter/s, 4.94045s/12 iters), loss = 1.452 I0409 23:52:17.352689 8290 solver.cpp:237] Train net output #0: loss = 1.452 (* 1 = 1.452 loss) I0409 23:52:17.352705 8290 sgd_solver.cpp:105] Iteration 4212, lr = 0.00434162 I0409 23:52:22.261436 8290 solver.cpp:218] Iteration 4224 (2.44471 iter/s, 4.90857s/12 iters), loss = 1.57436 I0409 23:52:22.261478 8290 solver.cpp:237] Train net output #0: loss = 1.57436 (* 1 = 1.57436 loss) I0409 23:52:22.261487 8290 sgd_solver.cpp:105] Iteration 4224, lr = 0.00433131 I0409 23:52:27.184332 8290 solver.cpp:218] Iteration 4236 (2.4377 iter/s, 4.92266s/12 iters), loss = 1.35118 I0409 23:52:27.184389 8290 solver.cpp:237] Train net output #0: loss = 1.35118 (* 1 = 1.35118 loss) I0409 23:52:27.184401 8290 sgd_solver.cpp:105] Iteration 4236, lr = 0.00432103 I0409 23:52:31.882993 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:52:32.101380 8290 solver.cpp:218] Iteration 4248 (2.44061 iter/s, 4.9168s/12 iters), loss = 1.04025 I0409 23:52:32.101461 8290 solver.cpp:237] Train net output #0: loss = 1.04025 (* 1 = 1.04025 loss) I0409 23:52:32.101480 8290 sgd_solver.cpp:105] Iteration 4248, lr = 0.00431077 I0409 23:52:36.959612 8290 solver.cpp:218] Iteration 4260 (2.47016 iter/s, 4.85798s/12 iters), loss = 1.24168 I0409 23:52:36.959672 8290 solver.cpp:237] Train net output #0: loss = 1.24168 (* 1 = 1.24168 loss) I0409 23:52:36.959686 8290 sgd_solver.cpp:105] Iteration 4260, lr = 0.00430053 I0409 23:52:41.877373 8290 solver.cpp:218] Iteration 4272 (2.44025 iter/s, 4.91752s/12 iters), loss = 1.09712 I0409 23:52:41.877416 8290 solver.cpp:237] Train net output #0: loss = 1.09712 (* 1 = 1.09712 loss) I0409 23:52:41.877425 8290 sgd_solver.cpp:105] Iteration 4272, lr = 0.00429032 I0409 23:52:46.335994 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4284.caffemodel I0409 23:52:47.084933 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4284.solverstate I0409 23:52:47.607565 8290 solver.cpp:330] Iteration 4284, Testing net (#0) I0409 23:52:47.607591 8290 net.cpp:676] Ignoring source layer train-data I0409 23:52:50.399416 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:52:52.084784 8290 solver.cpp:397] Test net output #0: accuracy = 0.446691 I0409 23:52:52.084830 8290 solver.cpp:397] Test net output #1: loss = 2.23426 (* 1 = 2.23426 loss) I0409 23:52:52.166374 8290 solver.cpp:218] Iteration 4284 (1.16634 iter/s, 10.2886s/12 iters), loss = 1.60854 I0409 23:52:52.166429 8290 solver.cpp:237] Train net output #0: loss = 1.60854 (* 1 = 1.60854 loss) I0409 23:52:52.166440 8290 sgd_solver.cpp:105] Iteration 4284, lr = 0.00428014 I0409 23:52:56.320402 8290 solver.cpp:218] Iteration 4296 (2.88891 iter/s, 4.15381s/12 iters), loss = 1.36127 I0409 23:52:56.320449 8290 solver.cpp:237] Train net output #0: loss = 1.36127 (* 1 = 1.36127 loss) I0409 23:52:56.320458 8290 sgd_solver.cpp:105] Iteration 4296, lr = 0.00426998 I0409 23:53:01.159010 8290 solver.cpp:218] Iteration 4308 (2.48017 iter/s, 4.83838s/12 iters), loss = 1.21943 I0409 23:53:01.159061 8290 solver.cpp:237] Train net output #0: loss = 1.21943 (* 1 = 1.21943 loss) I0409 23:53:01.159073 8290 sgd_solver.cpp:105] Iteration 4308, lr = 0.00425984 I0409 23:53:06.086122 8290 solver.cpp:218] Iteration 4320 (2.43562 iter/s, 4.92687s/12 iters), loss = 1.26629 I0409 23:53:06.086241 8290 solver.cpp:237] Train net output #0: loss = 1.26629 (* 1 = 1.26629 loss) I0409 23:53:06.086253 8290 sgd_solver.cpp:105] Iteration 4320, lr = 0.00424972 I0409 23:53:10.970409 8290 solver.cpp:218] Iteration 4332 (2.45701 iter/s, 4.88399s/12 iters), loss = 1.36497 I0409 23:53:10.970464 8290 solver.cpp:237] Train net output #0: loss = 1.36497 (* 1 = 1.36497 loss) I0409 23:53:10.970476 8290 sgd_solver.cpp:105] Iteration 4332, lr = 0.00423964 I0409 23:53:15.907630 8290 solver.cpp:218] Iteration 4344 (2.43063 iter/s, 4.93698s/12 iters), loss = 1.30082 I0409 23:53:15.907678 8290 solver.cpp:237] Train net output #0: loss = 1.30082 (* 1 = 1.30082 loss) I0409 23:53:15.907688 8290 sgd_solver.cpp:105] Iteration 4344, lr = 0.00422957 I0409 23:53:17.769068 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:53:20.813935 8290 solver.cpp:218] Iteration 4356 (2.44595 iter/s, 4.90607s/12 iters), loss = 1.37851 I0409 23:53:20.814004 8290 solver.cpp:237] Train net output #0: loss = 1.37851 (* 1 = 1.37851 loss) I0409 23:53:20.814014 8290 sgd_solver.cpp:105] Iteration 4356, lr = 0.00421953 I0409 23:53:25.728251 8290 solver.cpp:218] Iteration 4368 (2.44197 iter/s, 4.91407s/12 iters), loss = 1.26958 I0409 23:53:25.728304 8290 solver.cpp:237] Train net output #0: loss = 1.26958 (* 1 = 1.26958 loss) I0409 23:53:25.728315 8290 sgd_solver.cpp:105] Iteration 4368, lr = 0.00420951 I0409 23:53:31.244678 8290 solver.cpp:218] Iteration 4380 (2.17542 iter/s, 5.51617s/12 iters), loss = 1.21446 I0409 23:53:31.244719 8290 solver.cpp:237] Train net output #0: loss = 1.21446 (* 1 = 1.21446 loss) I0409 23:53:31.244729 8290 sgd_solver.cpp:105] Iteration 4380, lr = 0.00419952 I0409 23:53:33.251343 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4386.caffemodel I0409 23:53:34.027514 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4386.solverstate I0409 23:53:34.561821 8290 solver.cpp:330] Iteration 4386, Testing net (#0) I0409 23:53:34.561838 8290 net.cpp:676] Ignoring source layer train-data I0409 23:53:37.367461 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:53:39.100157 8290 solver.cpp:397] Test net output #0: accuracy = 0.468137 I0409 23:53:39.100205 8290 solver.cpp:397] Test net output #1: loss = 2.21393 (* 1 = 2.21393 loss) I0409 23:53:40.958820 8290 solver.cpp:218] Iteration 4392 (1.23536 iter/s, 9.71375s/12 iters), loss = 1.26002 I0409 23:53:40.958873 8290 solver.cpp:237] Train net output #0: loss = 1.26002 (* 1 = 1.26002 loss) I0409 23:53:40.958884 8290 sgd_solver.cpp:105] Iteration 4392, lr = 0.00418954 I0409 23:53:45.941700 8290 solver.cpp:218] Iteration 4404 (2.40836 iter/s, 4.98264s/12 iters), loss = 1.16198 I0409 23:53:45.941746 8290 solver.cpp:237] Train net output #0: loss = 1.16198 (* 1 = 1.16198 loss) I0409 23:53:45.941756 8290 sgd_solver.cpp:105] Iteration 4404, lr = 0.0041796 I0409 23:53:50.793996 8290 solver.cpp:218] Iteration 4416 (2.47318 iter/s, 4.85205s/12 iters), loss = 1.42229 I0409 23:53:50.794050 8290 solver.cpp:237] Train net output #0: loss = 1.42229 (* 1 = 1.42229 loss) I0409 23:53:50.794062 8290 sgd_solver.cpp:105] Iteration 4416, lr = 0.00416967 I0409 23:53:55.657975 8290 solver.cpp:218] Iteration 4428 (2.46724 iter/s, 4.86373s/12 iters), loss = 1.14573 I0409 23:53:55.658033 8290 solver.cpp:237] Train net output #0: loss = 1.14573 (* 1 = 1.14573 loss) I0409 23:53:55.658046 8290 sgd_solver.cpp:105] Iteration 4428, lr = 0.00415977 I0409 23:54:00.557265 8290 solver.cpp:218] Iteration 4440 (2.44945 iter/s, 4.89905s/12 iters), loss = 1.21155 I0409 23:54:00.557315 8290 solver.cpp:237] Train net output #0: loss = 1.21155 (* 1 = 1.21155 loss) I0409 23:54:00.557325 8290 sgd_solver.cpp:105] Iteration 4440, lr = 0.0041499 I0409 23:54:04.517450 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:54:05.468076 8290 solver.cpp:218] Iteration 4452 (2.44371 iter/s, 4.91057s/12 iters), loss = 1.13745 I0409 23:54:05.468133 8290 solver.cpp:237] Train net output #0: loss = 1.13745 (* 1 = 1.13745 loss) I0409 23:54:05.468145 8290 sgd_solver.cpp:105] Iteration 4452, lr = 0.00414005 I0409 23:54:10.392105 8290 solver.cpp:218] Iteration 4464 (2.43715 iter/s, 4.92378s/12 iters), loss = 1.36519 I0409 23:54:10.392246 8290 solver.cpp:237] Train net output #0: loss = 1.36519 (* 1 = 1.36519 loss) I0409 23:54:10.392259 8290 sgd_solver.cpp:105] Iteration 4464, lr = 0.00413022 I0409 23:54:15.251965 8290 solver.cpp:218] Iteration 4476 (2.46937 iter/s, 4.85954s/12 iters), loss = 1.52804 I0409 23:54:15.252023 8290 solver.cpp:237] Train net output #0: loss = 1.52804 (* 1 = 1.52804 loss) I0409 23:54:15.252036 8290 sgd_solver.cpp:105] Iteration 4476, lr = 0.00412041 I0409 23:54:19.695441 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4488.caffemodel I0409 23:54:23.417356 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4488.solverstate I0409 23:54:25.413084 8290 solver.cpp:330] Iteration 4488, Testing net (#0) I0409 23:54:25.413110 8290 net.cpp:676] Ignoring source layer train-data I0409 23:54:28.062319 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:54:29.857825 8290 solver.cpp:397] Test net output #0: accuracy = 0.468137 I0409 23:54:29.857877 8290 solver.cpp:397] Test net output #1: loss = 2.24018 (* 1 = 2.24018 loss) I0409 23:54:29.940197 8290 solver.cpp:218] Iteration 4488 (0.817013 iter/s, 14.6877s/12 iters), loss = 1.10547 I0409 23:54:29.940254 8290 solver.cpp:237] Train net output #0: loss = 1.10547 (* 1 = 1.10547 loss) I0409 23:54:29.940268 8290 sgd_solver.cpp:105] Iteration 4488, lr = 0.00411063 I0409 23:54:34.076458 8290 solver.cpp:218] Iteration 4500 (2.90132 iter/s, 4.13605s/12 iters), loss = 1.09814 I0409 23:54:34.076508 8290 solver.cpp:237] Train net output #0: loss = 1.09814 (* 1 = 1.09814 loss) I0409 23:54:34.076520 8290 sgd_solver.cpp:105] Iteration 4500, lr = 0.00410087 I0409 23:54:39.110210 8290 solver.cpp:218] Iteration 4512 (2.38402 iter/s, 5.03351s/12 iters), loss = 1.02717 I0409 23:54:39.110265 8290 solver.cpp:237] Train net output #0: loss = 1.02717 (* 1 = 1.02717 loss) I0409 23:54:39.110275 8290 sgd_solver.cpp:105] Iteration 4512, lr = 0.00409113 I0409 23:54:43.955699 8290 solver.cpp:218] Iteration 4524 (2.47665 iter/s, 4.84525s/12 iters), loss = 1.11081 I0409 23:54:43.955785 8290 solver.cpp:237] Train net output #0: loss = 1.11081 (* 1 = 1.11081 loss) I0409 23:54:43.955796 8290 sgd_solver.cpp:105] Iteration 4524, lr = 0.00408142 I0409 23:54:48.821203 8290 solver.cpp:218] Iteration 4536 (2.46648 iter/s, 4.86523s/12 iters), loss = 1.36165 I0409 23:54:48.821262 8290 solver.cpp:237] Train net output #0: loss = 1.36165 (* 1 = 1.36165 loss) I0409 23:54:48.821274 8290 sgd_solver.cpp:105] Iteration 4536, lr = 0.00407173 I0409 23:54:53.878635 8290 solver.cpp:218] Iteration 4548 (2.37286 iter/s, 5.05718s/12 iters), loss = 1.22499 I0409 23:54:53.878692 8290 solver.cpp:237] Train net output #0: loss = 1.22499 (* 1 = 1.22499 loss) I0409 23:54:53.878703 8290 sgd_solver.cpp:105] Iteration 4548, lr = 0.00406206 I0409 23:54:55.085114 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:54:58.710942 8290 solver.cpp:218] Iteration 4560 (2.48341 iter/s, 4.83206s/12 iters), loss = 1.02395 I0409 23:54:58.710999 8290 solver.cpp:237] Train net output #0: loss = 1.02395 (* 1 = 1.02395 loss) I0409 23:54:58.711010 8290 sgd_solver.cpp:105] Iteration 4560, lr = 0.00405242 I0409 23:55:03.567121 8290 solver.cpp:218] Iteration 4572 (2.4712 iter/s, 4.85594s/12 iters), loss = 1.26879 I0409 23:55:03.567173 8290 solver.cpp:237] Train net output #0: loss = 1.26879 (* 1 = 1.26879 loss) I0409 23:55:03.567183 8290 sgd_solver.cpp:105] Iteration 4572, lr = 0.0040428 I0409 23:55:08.544296 8290 solver.cpp:218] Iteration 4584 (2.41112 iter/s, 4.97693s/12 iters), loss = 0.887832 I0409 23:55:08.544355 8290 solver.cpp:237] Train net output #0: loss = 0.887832 (* 1 = 0.887832 loss) I0409 23:55:08.544366 8290 sgd_solver.cpp:105] Iteration 4584, lr = 0.0040332 I0409 23:55:10.566144 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4590.caffemodel I0409 23:55:11.369118 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4590.solverstate I0409 23:55:11.908617 8290 solver.cpp:330] Iteration 4590, Testing net (#0) I0409 23:55:11.908648 8290 net.cpp:676] Ignoring source layer train-data I0409 23:55:14.781257 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:55:16.592936 8290 solver.cpp:397] Test net output #0: accuracy = 0.483456 I0409 23:55:16.592983 8290 solver.cpp:397] Test net output #1: loss = 2.17275 (* 1 = 2.17275 loss) I0409 23:55:18.391978 8290 solver.cpp:218] Iteration 4596 (1.21861 iter/s, 9.84726s/12 iters), loss = 0.970405 I0409 23:55:18.392042 8290 solver.cpp:237] Train net output #0: loss = 0.970405 (* 1 = 0.970405 loss) I0409 23:55:18.392055 8290 sgd_solver.cpp:105] Iteration 4596, lr = 0.00402362 I0409 23:55:23.222213 8290 solver.cpp:218] Iteration 4608 (2.48448 iter/s, 4.82999s/12 iters), loss = 1.03565 I0409 23:55:23.222268 8290 solver.cpp:237] Train net output #0: loss = 1.03565 (* 1 = 1.03565 loss) I0409 23:55:23.222277 8290 sgd_solver.cpp:105] Iteration 4608, lr = 0.00401407 I0409 23:55:28.126272 8290 solver.cpp:218] Iteration 4620 (2.44707 iter/s, 4.90382s/12 iters), loss = 1.16419 I0409 23:55:28.126314 8290 solver.cpp:237] Train net output #0: loss = 1.16419 (* 1 = 1.16419 loss) I0409 23:55:28.126323 8290 sgd_solver.cpp:105] Iteration 4620, lr = 0.00400454 I0409 23:55:33.185132 8290 solver.cpp:218] Iteration 4632 (2.37219 iter/s, 5.05862s/12 iters), loss = 1.13456 I0409 23:55:33.185181 8290 solver.cpp:237] Train net output #0: loss = 1.13456 (* 1 = 1.13456 loss) I0409 23:55:33.185190 8290 sgd_solver.cpp:105] Iteration 4632, lr = 0.00399503 I0409 23:55:38.274279 8290 solver.cpp:218] Iteration 4644 (2.35807 iter/s, 5.08891s/12 iters), loss = 1.25694 I0409 23:55:38.274331 8290 solver.cpp:237] Train net output #0: loss = 1.25694 (* 1 = 1.25694 loss) I0409 23:55:38.274343 8290 sgd_solver.cpp:105] Iteration 4644, lr = 0.00398555 I0409 23:55:41.643256 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:55:43.192955 8290 solver.cpp:218] Iteration 4656 (2.4398 iter/s, 4.91844s/12 iters), loss = 1.23005 I0409 23:55:43.192999 8290 solver.cpp:237] Train net output #0: loss = 1.23005 (* 1 = 1.23005 loss) I0409 23:55:43.193008 8290 sgd_solver.cpp:105] Iteration 4656, lr = 0.00397608 I0409 23:55:48.064340 8290 solver.cpp:218] Iteration 4668 (2.46348 iter/s, 4.87115s/12 iters), loss = 1.20181 I0409 23:55:48.064463 8290 solver.cpp:237] Train net output #0: loss = 1.20181 (* 1 = 1.20181 loss) I0409 23:55:48.064477 8290 sgd_solver.cpp:105] Iteration 4668, lr = 0.00396664 I0409 23:55:52.949193 8290 solver.cpp:218] Iteration 4680 (2.45673 iter/s, 4.88455s/12 iters), loss = 1.20693 I0409 23:55:52.949249 8290 solver.cpp:237] Train net output #0: loss = 1.20693 (* 1 = 1.20693 loss) I0409 23:55:52.949261 8290 sgd_solver.cpp:105] Iteration 4680, lr = 0.00395723 I0409 23:55:57.420397 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4692.caffemodel I0409 23:55:58.178609 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4692.solverstate I0409 23:55:58.719936 8290 solver.cpp:330] Iteration 4692, Testing net (#0) I0409 23:55:58.719962 8290 net.cpp:676] Ignoring source layer train-data I0409 23:56:01.282091 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:56:03.127385 8290 solver.cpp:397] Test net output #0: accuracy = 0.468137 I0409 23:56:03.127415 8290 solver.cpp:397] Test net output #1: loss = 2.16981 (* 1 = 2.16981 loss) I0409 23:56:03.209278 8290 solver.cpp:218] Iteration 4692 (1.16963 iter/s, 10.2597s/12 iters), loss = 1.0826 I0409 23:56:03.209321 8290 solver.cpp:237] Train net output #0: loss = 1.0826 (* 1 = 1.0826 loss) I0409 23:56:03.209331 8290 sgd_solver.cpp:105] Iteration 4692, lr = 0.00394783 I0409 23:56:07.343286 8290 solver.cpp:218] Iteration 4704 (2.9029 iter/s, 4.1338s/12 iters), loss = 1.01998 I0409 23:56:07.343348 8290 solver.cpp:237] Train net output #0: loss = 1.01998 (* 1 = 1.01998 loss) I0409 23:56:07.343361 8290 sgd_solver.cpp:105] Iteration 4704, lr = 0.00393846 I0409 23:56:12.263005 8290 solver.cpp:218] Iteration 4716 (2.43929 iter/s, 4.91947s/12 iters), loss = 0.993161 I0409 23:56:12.263058 8290 solver.cpp:237] Train net output #0: loss = 0.993161 (* 1 = 0.993161 loss) I0409 23:56:12.263069 8290 sgd_solver.cpp:105] Iteration 4716, lr = 0.00392911 I0409 23:56:17.374104 8290 solver.cpp:218] Iteration 4728 (2.34794 iter/s, 5.11085s/12 iters), loss = 1.09407 I0409 23:56:17.374152 8290 solver.cpp:237] Train net output #0: loss = 1.09407 (* 1 = 1.09407 loss) I0409 23:56:17.374164 8290 sgd_solver.cpp:105] Iteration 4728, lr = 0.00391978 I0409 23:56:22.278601 8290 solver.cpp:218] Iteration 4740 (2.44685 iter/s, 4.90426s/12 iters), loss = 0.878493 I0409 23:56:22.278759 8290 solver.cpp:237] Train net output #0: loss = 0.878493 (* 1 = 0.878493 loss) I0409 23:56:22.278774 8290 sgd_solver.cpp:105] Iteration 4740, lr = 0.00391047 I0409 23:56:27.192579 8290 solver.cpp:218] Iteration 4752 (2.44218 iter/s, 4.91364s/12 iters), loss = 1.03671 I0409 23:56:27.192625 8290 solver.cpp:237] Train net output #0: loss = 1.03671 (* 1 = 1.03671 loss) I0409 23:56:27.192634 8290 sgd_solver.cpp:105] Iteration 4752, lr = 0.00390119 I0409 23:56:27.713654 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:56:32.122736 8290 solver.cpp:218] Iteration 4764 (2.43411 iter/s, 4.92993s/12 iters), loss = 0.920601 I0409 23:56:32.122788 8290 solver.cpp:237] Train net output #0: loss = 0.920601 (* 1 = 0.920601 loss) I0409 23:56:32.122799 8290 sgd_solver.cpp:105] Iteration 4764, lr = 0.00389193 I0409 23:56:37.028599 8290 solver.cpp:218] Iteration 4776 (2.44617 iter/s, 4.90562s/12 iters), loss = 1.10183 I0409 23:56:37.028656 8290 solver.cpp:237] Train net output #0: loss = 1.10183 (* 1 = 1.10183 loss) I0409 23:56:37.028668 8290 sgd_solver.cpp:105] Iteration 4776, lr = 0.00388269 I0409 23:56:41.937158 8290 solver.cpp:218] Iteration 4788 (2.44483 iter/s, 4.90832s/12 iters), loss = 1.11958 I0409 23:56:41.937203 8290 solver.cpp:237] Train net output #0: loss = 1.11958 (* 1 = 1.11958 loss) I0409 23:56:41.937212 8290 sgd_solver.cpp:105] Iteration 4788, lr = 0.00387347 I0409 23:56:43.918778 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4794.caffemodel I0409 23:56:44.693334 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4794.solverstate I0409 23:56:45.569761 8290 solver.cpp:330] Iteration 4794, Testing net (#0) I0409 23:56:45.569790 8290 net.cpp:676] Ignoring source layer train-data I0409 23:56:48.150605 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:56:50.042027 8290 solver.cpp:397] Test net output #0: accuracy = 0.501838 I0409 23:56:50.042062 8290 solver.cpp:397] Test net output #1: loss = 2.08856 (* 1 = 2.08856 loss) I0409 23:56:51.847458 8290 solver.cpp:218] Iteration 4800 (1.21091 iter/s, 9.90989s/12 iters), loss = 1.10591 I0409 23:56:51.847515 8290 solver.cpp:237] Train net output #0: loss = 1.10591 (* 1 = 1.10591 loss) I0409 23:56:51.847527 8290 sgd_solver.cpp:105] Iteration 4800, lr = 0.00386427 I0409 23:56:56.692636 8290 solver.cpp:218] Iteration 4812 (2.47681 iter/s, 4.84494s/12 iters), loss = 1.15976 I0409 23:56:56.692744 8290 solver.cpp:237] Train net output #0: loss = 1.15976 (* 1 = 1.15976 loss) I0409 23:56:56.692756 8290 sgd_solver.cpp:105] Iteration 4812, lr = 0.0038551 I0409 23:57:01.552069 8290 solver.cpp:218] Iteration 4824 (2.46957 iter/s, 4.85915s/12 iters), loss = 1.18468 I0409 23:57:01.552116 8290 solver.cpp:237] Train net output #0: loss = 1.18468 (* 1 = 1.18468 loss) I0409 23:57:01.552125 8290 sgd_solver.cpp:105] Iteration 4824, lr = 0.00384594 I0409 23:57:06.441391 8290 solver.cpp:218] Iteration 4836 (2.45445 iter/s, 4.88909s/12 iters), loss = 0.981182 I0409 23:57:06.441442 8290 solver.cpp:237] Train net output #0: loss = 0.981182 (* 1 = 0.981182 loss) I0409 23:57:06.441452 8290 sgd_solver.cpp:105] Iteration 4836, lr = 0.00383681 I0409 23:57:07.221670 8290 blocking_queue.cpp:49] Waiting for data I0409 23:57:11.393298 8290 solver.cpp:218] Iteration 4848 (2.42343 iter/s, 4.95166s/12 iters), loss = 1.55507 I0409 23:57:11.393354 8290 solver.cpp:237] Train net output #0: loss = 1.55507 (* 1 = 1.55507 loss) I0409 23:57:11.393368 8290 sgd_solver.cpp:105] Iteration 4848, lr = 0.0038277 I0409 23:57:14.035589 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:57:16.362376 8290 solver.cpp:218] Iteration 4860 (2.41506 iter/s, 4.96883s/12 iters), loss = 0.749391 I0409 23:57:16.362443 8290 solver.cpp:237] Train net output #0: loss = 0.749391 (* 1 = 0.749391 loss) I0409 23:57:16.362457 8290 sgd_solver.cpp:105] Iteration 4860, lr = 0.00381862 I0409 23:57:21.294312 8290 solver.cpp:218] Iteration 4872 (2.43325 iter/s, 4.93168s/12 iters), loss = 1.01604 I0409 23:57:21.294369 8290 solver.cpp:237] Train net output #0: loss = 1.01604 (* 1 = 1.01604 loss) I0409 23:57:21.294381 8290 sgd_solver.cpp:105] Iteration 4872, lr = 0.00380955 I0409 23:57:26.175042 8290 solver.cpp:218] Iteration 4884 (2.45877 iter/s, 4.88049s/12 iters), loss = 1.03738 I0409 23:57:26.175102 8290 solver.cpp:237] Train net output #0: loss = 1.03738 (* 1 = 1.03738 loss) I0409 23:57:26.175115 8290 sgd_solver.cpp:105] Iteration 4884, lr = 0.0038005 I0409 23:57:30.611259 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4896.caffemodel I0409 23:57:31.708482 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4896.solverstate I0409 23:57:32.883111 8290 solver.cpp:330] Iteration 4896, Testing net (#0) I0409 23:57:32.883139 8290 net.cpp:676] Ignoring source layer train-data I0409 23:57:35.437299 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:57:37.441174 8290 solver.cpp:397] Test net output #0: accuracy = 0.495098 I0409 23:57:37.441222 8290 solver.cpp:397] Test net output #1: loss = 2.13307 (* 1 = 2.13307 loss) I0409 23:57:37.523397 8290 solver.cpp:218] Iteration 4896 (1.05746 iter/s, 11.3479s/12 iters), loss = 1.0057 I0409 23:57:37.523440 8290 solver.cpp:237] Train net output #0: loss = 1.0057 (* 1 = 1.0057 loss) I0409 23:57:37.523449 8290 sgd_solver.cpp:105] Iteration 4896, lr = 0.00379148 I0409 23:57:41.700645 8290 solver.cpp:218] Iteration 4908 (2.87285 iter/s, 4.17704s/12 iters), loss = 0.875682 I0409 23:57:41.700707 8290 solver.cpp:237] Train net output #0: loss = 0.875682 (* 1 = 0.875682 loss) I0409 23:57:41.700721 8290 sgd_solver.cpp:105] Iteration 4908, lr = 0.00378248 I0409 23:57:46.509399 8290 solver.cpp:218] Iteration 4920 (2.49558 iter/s, 4.80851s/12 iters), loss = 0.883767 I0409 23:57:46.509461 8290 solver.cpp:237] Train net output #0: loss = 0.883767 (* 1 = 0.883767 loss) I0409 23:57:46.509474 8290 sgd_solver.cpp:105] Iteration 4920, lr = 0.0037735 I0409 23:57:51.313369 8290 solver.cpp:218] Iteration 4932 (2.49806 iter/s, 4.80373s/12 iters), loss = 0.791983 I0409 23:57:51.313429 8290 solver.cpp:237] Train net output #0: loss = 0.791983 (* 1 = 0.791983 loss) I0409 23:57:51.313442 8290 sgd_solver.cpp:105] Iteration 4932, lr = 0.00376454 I0409 23:57:56.142195 8290 solver.cpp:218] Iteration 4944 (2.4852 iter/s, 4.82858s/12 iters), loss = 1.15994 I0409 23:57:56.142256 8290 solver.cpp:237] Train net output #0: loss = 1.15994 (* 1 = 1.15994 loss) I0409 23:57:56.142268 8290 sgd_solver.cpp:105] Iteration 4944, lr = 0.0037556 I0409 23:58:00.765650 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:58:00.958300 8290 solver.cpp:218] Iteration 4956 (2.49177 iter/s, 4.81586s/12 iters), loss = 0.879959 I0409 23:58:00.958359 8290 solver.cpp:237] Train net output #0: loss = 0.879959 (* 1 = 0.879959 loss) I0409 23:58:00.958376 8290 sgd_solver.cpp:105] Iteration 4956, lr = 0.00374669 I0409 23:58:05.914618 8290 solver.cpp:218] Iteration 4968 (2.42127 iter/s, 4.95607s/12 iters), loss = 1.14483 I0409 23:58:05.914667 8290 solver.cpp:237] Train net output #0: loss = 1.14483 (* 1 = 1.14483 loss) I0409 23:58:05.914680 8290 sgd_solver.cpp:105] Iteration 4968, lr = 0.00373779 I0409 23:58:10.811332 8290 solver.cpp:218] Iteration 4980 (2.45074 iter/s, 4.89648s/12 iters), loss = 0.814194 I0409 23:58:10.811388 8290 solver.cpp:237] Train net output #0: loss = 0.814194 (* 1 = 0.814194 loss) I0409 23:58:10.811400 8290 sgd_solver.cpp:105] Iteration 4980, lr = 0.00372892 I0409 23:58:15.709762 8290 solver.cpp:218] Iteration 4992 (2.44988 iter/s, 4.89819s/12 iters), loss = 1.07069 I0409 23:58:15.709816 8290 solver.cpp:237] Train net output #0: loss = 1.07069 (* 1 = 1.07069 loss) I0409 23:58:15.709830 8290 sgd_solver.cpp:105] Iteration 4992, lr = 0.00372006 I0409 23:58:17.676952 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_4998.caffemodel I0409 23:58:19.287154 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_4998.solverstate I0409 23:58:20.037914 8290 solver.cpp:330] Iteration 4998, Testing net (#0) I0409 23:58:20.037933 8290 net.cpp:676] Ignoring source layer train-data I0409 23:58:22.525213 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:58:24.526917 8290 solver.cpp:397] Test net output #0: accuracy = 0.496324 I0409 23:58:24.526968 8290 solver.cpp:397] Test net output #1: loss = 2.10013 (* 1 = 2.10013 loss) I0409 23:58:26.380173 8290 solver.cpp:218] Iteration 5004 (1.12465 iter/s, 10.67s/12 iters), loss = 0.864698 I0409 23:58:26.380234 8290 solver.cpp:237] Train net output #0: loss = 0.864698 (* 1 = 0.864698 loss) I0409 23:58:26.380246 8290 sgd_solver.cpp:105] Iteration 5004, lr = 0.00371123 I0409 23:58:31.335803 8290 solver.cpp:218] Iteration 5016 (2.42161 iter/s, 4.95538s/12 iters), loss = 0.9993 I0409 23:58:31.338155 8290 solver.cpp:237] Train net output #0: loss = 0.9993 (* 1 = 0.9993 loss) I0409 23:58:31.338171 8290 sgd_solver.cpp:105] Iteration 5016, lr = 0.00370242 I0409 23:58:36.262377 8290 solver.cpp:218] Iteration 5028 (2.43702 iter/s, 4.92404s/12 iters), loss = 1.14541 I0409 23:58:36.262428 8290 solver.cpp:237] Train net output #0: loss = 1.14541 (* 1 = 1.14541 loss) I0409 23:58:36.262440 8290 sgd_solver.cpp:105] Iteration 5028, lr = 0.00369363 I0409 23:58:41.235661 8290 solver.cpp:218] Iteration 5040 (2.41301 iter/s, 4.97305s/12 iters), loss = 0.948971 I0409 23:58:41.235721 8290 solver.cpp:237] Train net output #0: loss = 0.948971 (* 1 = 0.948971 loss) I0409 23:58:41.235734 8290 sgd_solver.cpp:105] Iteration 5040, lr = 0.00368486 I0409 23:58:46.132496 8290 solver.cpp:218] Iteration 5052 (2.45069 iter/s, 4.89659s/12 iters), loss = 1.04798 I0409 23:58:46.132551 8290 solver.cpp:237] Train net output #0: loss = 1.04798 (* 1 = 1.04798 loss) I0409 23:58:46.132563 8290 sgd_solver.cpp:105] Iteration 5052, lr = 0.00367611 I0409 23:58:48.032507 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:58:51.033787 8290 solver.cpp:218] Iteration 5064 (2.44846 iter/s, 4.90105s/12 iters), loss = 1.03006 I0409 23:58:51.033833 8290 solver.cpp:237] Train net output #0: loss = 1.03006 (* 1 = 1.03006 loss) I0409 23:58:51.033841 8290 sgd_solver.cpp:105] Iteration 5064, lr = 0.00366738 I0409 23:58:55.949652 8290 solver.cpp:218] Iteration 5076 (2.44119 iter/s, 4.91563s/12 iters), loss = 1.01222 I0409 23:58:55.949715 8290 solver.cpp:237] Train net output #0: loss = 1.01222 (* 1 = 1.01222 loss) I0409 23:58:55.949728 8290 sgd_solver.cpp:105] Iteration 5076, lr = 0.00365868 I0409 23:59:00.905095 8290 solver.cpp:218] Iteration 5088 (2.4217 iter/s, 4.9552s/12 iters), loss = 0.738719 I0409 23:59:00.905144 8290 solver.cpp:237] Train net output #0: loss = 0.738719 (* 1 = 0.738719 loss) I0409 23:59:00.905153 8290 sgd_solver.cpp:105] Iteration 5088, lr = 0.00364999 I0409 23:59:05.470147 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_5100.caffemodel I0409 23:59:06.466974 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_5100.solverstate I0409 23:59:08.034795 8290 solver.cpp:330] Iteration 5100, Testing net (#0) I0409 23:59:08.034819 8290 net.cpp:676] Ignoring source layer train-data I0409 23:59:10.357369 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:59:12.363190 8290 solver.cpp:397] Test net output #0: accuracy = 0.508578 I0409 23:59:12.363241 8290 solver.cpp:397] Test net output #1: loss = 2.07228 (* 1 = 2.07228 loss) I0409 23:59:12.443960 8290 solver.cpp:218] Iteration 5100 (1.04001 iter/s, 11.5384s/12 iters), loss = 1.01295 I0409 23:59:12.444016 8290 solver.cpp:237] Train net output #0: loss = 1.01295 (* 1 = 1.01295 loss) I0409 23:59:12.444028 8290 sgd_solver.cpp:105] Iteration 5100, lr = 0.00364132 I0409 23:59:16.713414 8290 solver.cpp:218] Iteration 5112 (2.81081 iter/s, 4.26923s/12 iters), loss = 0.899726 I0409 23:59:16.713471 8290 solver.cpp:237] Train net output #0: loss = 0.899726 (* 1 = 0.899726 loss) I0409 23:59:16.713482 8290 sgd_solver.cpp:105] Iteration 5112, lr = 0.00363268 I0409 23:59:21.698756 8290 solver.cpp:218] Iteration 5124 (2.40718 iter/s, 4.98509s/12 iters), loss = 0.969474 I0409 23:59:21.698818 8290 solver.cpp:237] Train net output #0: loss = 0.969474 (* 1 = 0.969474 loss) I0409 23:59:21.698832 8290 sgd_solver.cpp:105] Iteration 5124, lr = 0.00362405 I0409 23:59:26.606659 8290 solver.cpp:218] Iteration 5136 (2.44516 iter/s, 4.90765s/12 iters), loss = 0.890349 I0409 23:59:26.606717 8290 solver.cpp:237] Train net output #0: loss = 0.890349 (* 1 = 0.890349 loss) I0409 23:59:26.606729 8290 sgd_solver.cpp:105] Iteration 5136, lr = 0.00361545 I0409 23:59:31.526281 8290 solver.cpp:218] Iteration 5148 (2.43933 iter/s, 4.91938s/12 iters), loss = 0.794216 I0409 23:59:31.526330 8290 solver.cpp:237] Train net output #0: loss = 0.794216 (* 1 = 0.794216 loss) I0409 23:59:31.526343 8290 sgd_solver.cpp:105] Iteration 5148, lr = 0.00360687 I0409 23:59:35.497975 8294 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:59:36.407846 8290 solver.cpp:218] Iteration 5160 (2.45835 iter/s, 4.88133s/12 iters), loss = 0.65668 I0409 23:59:36.407900 8290 solver.cpp:237] Train net output #0: loss = 0.65668 (* 1 = 0.65668 loss) I0409 23:59:36.407914 8290 sgd_solver.cpp:105] Iteration 5160, lr = 0.0035983 I0409 23:59:41.324265 8290 solver.cpp:218] Iteration 5172 (2.44092 iter/s, 4.91619s/12 iters), loss = 0.955993 I0409 23:59:41.324303 8290 solver.cpp:237] Train net output #0: loss = 0.955993 (* 1 = 0.955993 loss) I0409 23:59:41.324311 8290 sgd_solver.cpp:105] Iteration 5172, lr = 0.00358976 I0409 23:59:46.271700 8290 solver.cpp:218] Iteration 5184 (2.42561 iter/s, 4.94722s/12 iters), loss = 1.09478 I0409 23:59:46.271739 8290 solver.cpp:237] Train net output #0: loss = 1.09478 (* 1 = 1.09478 loss) I0409 23:59:46.271746 8290 sgd_solver.cpp:105] Iteration 5184, lr = 0.00358124 I0409 23:59:51.109177 8290 solver.cpp:218] Iteration 5196 (2.48075 iter/s, 4.83725s/12 iters), loss = 0.875414 I0409 23:59:51.109231 8290 solver.cpp:237] Train net output #0: loss = 0.875414 (* 1 = 0.875414 loss) I0409 23:59:51.109243 8290 sgd_solver.cpp:105] Iteration 5196, lr = 0.00357273 I0409 23:59:53.103684 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_5202.caffemodel I0409 23:59:53.836236 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_5202.solverstate I0409 23:59:54.365973 8290 solver.cpp:330] Iteration 5202, Testing net (#0) I0409 23:59:54.365994 8290 net.cpp:676] Ignoring source layer train-data I0409 23:59:56.775156 8295 data_layer.cpp:73] Restarting data prefetching from start. I0409 23:59:58.911933 8290 solver.cpp:397] Test net output #0: accuracy = 0.515319 I0409 23:59:58.911986 8290 solver.cpp:397] Test net output #1: loss = 2.04432 (* 1 = 2.04432 loss) I0410 00:00:00.876673 8290 solver.cpp:218] Iteration 5208 (1.22862 iter/s, 9.76709s/12 iters), loss = 0.709369 I0410 00:00:00.876730 8290 solver.cpp:237] Train net output #0: loss = 0.709369 (* 1 = 0.709369 loss) I0410 00:00:00.876744 8290 sgd_solver.cpp:105] Iteration 5208, lr = 0.00356425 I0410 00:00:05.909173 8290 solver.cpp:218] Iteration 5220 (2.38462 iter/s, 5.03226s/12 iters), loss = 0.925779 I0410 00:00:05.910245 8290 solver.cpp:237] Train net output #0: loss = 0.925779 (* 1 = 0.925779 loss) I0410 00:00:05.910256 8290 sgd_solver.cpp:105] Iteration 5220, lr = 0.00355579 I0410 00:00:10.843024 8290 solver.cpp:218] Iteration 5232 (2.4328 iter/s, 4.93259s/12 iters), loss = 0.840775 I0410 00:00:10.843087 8290 solver.cpp:237] Train net output #0: loss = 0.840775 (* 1 = 0.840775 loss) I0410 00:00:10.843101 8290 sgd_solver.cpp:105] Iteration 5232, lr = 0.00354735 I0410 00:00:15.736297 8290 solver.cpp:218] Iteration 5244 (2.45247 iter/s, 4.89303s/12 iters), loss = 0.836296 I0410 00:00:15.736344 8290 solver.cpp:237] Train net output #0: loss = 0.836296 (* 1 = 0.836296 loss) I0410 00:00:15.736354 8290 sgd_solver.cpp:105] Iteration 5244, lr = 0.00353892 I0410 00:00:20.616495 8290 solver.cpp:218] Iteration 5256 (2.45904 iter/s, 4.87996s/12 iters), loss = 0.809875 I0410 00:00:20.616567 8290 solver.cpp:237] Train net output #0: loss = 0.809875 (* 1 = 0.809875 loss) I0410 00:00:20.616583 8290 sgd_solver.cpp:105] Iteration 5256, lr = 0.00353052 I0410 00:00:21.914547 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:00:25.549396 8290 solver.cpp:218] Iteration 5268 (2.43277 iter/s, 4.93265s/12 iters), loss = 0.984964 I0410 00:00:25.549438 8290 solver.cpp:237] Train net output #0: loss = 0.984964 (* 1 = 0.984964 loss) I0410 00:00:25.549449 8290 sgd_solver.cpp:105] Iteration 5268, lr = 0.00352214 I0410 00:00:30.337952 8290 solver.cpp:218] Iteration 5280 (2.50609 iter/s, 4.78833s/12 iters), loss = 1.08922 I0410 00:00:30.338029 8290 solver.cpp:237] Train net output #0: loss = 1.08922 (* 1 = 1.08922 loss) I0410 00:00:30.338043 8290 sgd_solver.cpp:105] Iteration 5280, lr = 0.00351378 I0410 00:00:35.200825 8290 solver.cpp:218] Iteration 5292 (2.46781 iter/s, 4.86261s/12 iters), loss = 0.910194 I0410 00:00:35.200891 8290 solver.cpp:237] Train net output #0: loss = 0.910194 (* 1 = 0.910194 loss) I0410 00:00:35.200904 8290 sgd_solver.cpp:105] Iteration 5292, lr = 0.00350544 I0410 00:00:39.579545 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_5304.caffemodel I0410 00:00:40.297986 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_5304.solverstate I0410 00:00:40.846813 8290 solver.cpp:330] Iteration 5304, Testing net (#0) I0410 00:00:40.846837 8290 net.cpp:676] Ignoring source layer train-data I0410 00:00:43.295428 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:00:45.383836 8290 solver.cpp:397] Test net output #0: accuracy = 0.525123 I0410 00:00:45.383886 8290 solver.cpp:397] Test net output #1: loss = 2.07691 (* 1 = 2.07691 loss) I0410 00:00:45.465824 8290 solver.cpp:218] Iteration 5304 (1.16907 iter/s, 10.2646s/12 iters), loss = 0.674758 I0410 00:00:45.465893 8290 solver.cpp:237] Train net output #0: loss = 0.674758 (* 1 = 0.674758 loss) I0410 00:00:45.465911 8290 sgd_solver.cpp:105] Iteration 5304, lr = 0.00349711 I0410 00:00:49.690873 8290 solver.cpp:218] Iteration 5316 (2.84036 iter/s, 4.22482s/12 iters), loss = 0.721853 I0410 00:00:49.690924 8290 solver.cpp:237] Train net output #0: loss = 0.721853 (* 1 = 0.721853 loss) I0410 00:00:49.690934 8290 sgd_solver.cpp:105] Iteration 5316, lr = 0.00348881 I0410 00:00:54.610493 8290 solver.cpp:218] Iteration 5328 (2.43933 iter/s, 4.91939s/12 iters), loss = 0.829293 I0410 00:00:54.610533 8290 solver.cpp:237] Train net output #0: loss = 0.829293 (* 1 = 0.829293 loss) I0410 00:00:54.610543 8290 sgd_solver.cpp:105] Iteration 5328, lr = 0.00348053 I0410 00:00:59.488955 8290 solver.cpp:218] Iteration 5340 (2.4599 iter/s, 4.87824s/12 iters), loss = 0.759039 I0410 00:00:59.489006 8290 solver.cpp:237] Train net output #0: loss = 0.759039 (* 1 = 0.759039 loss) I0410 00:00:59.489017 8290 sgd_solver.cpp:105] Iteration 5340, lr = 0.00347226 I0410 00:01:04.322804 8290 solver.cpp:218] Iteration 5352 (2.48261 iter/s, 4.83361s/12 iters), loss = 0.768784 I0410 00:01:04.322860 8290 solver.cpp:237] Train net output #0: loss = 0.768784 (* 1 = 0.768784 loss) I0410 00:01:04.322871 8290 sgd_solver.cpp:105] Iteration 5352, lr = 0.00346402 I0410 00:01:07.663305 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:01:09.212760 8290 solver.cpp:218] Iteration 5364 (2.45413 iter/s, 4.88972s/12 iters), loss = 0.796877 I0410 00:01:09.212807 8290 solver.cpp:237] Train net output #0: loss = 0.796877 (* 1 = 0.796877 loss) I0410 00:01:09.212816 8290 sgd_solver.cpp:105] Iteration 5364, lr = 0.0034558 I0410 00:01:14.127781 8290 solver.cpp:218] Iteration 5376 (2.44161 iter/s, 4.91479s/12 iters), loss = 0.904363 I0410 00:01:14.127936 8290 solver.cpp:237] Train net output #0: loss = 0.904363 (* 1 = 0.904363 loss) I0410 00:01:14.127949 8290 sgd_solver.cpp:105] Iteration 5376, lr = 0.00344759 I0410 00:01:19.029549 8290 solver.cpp:218] Iteration 5388 (2.44827 iter/s, 4.90143s/12 iters), loss = 0.794196 I0410 00:01:19.029603 8290 solver.cpp:237] Train net output #0: loss = 0.794196 (* 1 = 0.794196 loss) I0410 00:01:19.029613 8290 sgd_solver.cpp:105] Iteration 5388, lr = 0.00343941 I0410 00:01:23.898237 8290 solver.cpp:218] Iteration 5400 (2.46485 iter/s, 4.86845s/12 iters), loss = 0.743833 I0410 00:01:23.898288 8290 solver.cpp:237] Train net output #0: loss = 0.743833 (* 1 = 0.743833 loss) I0410 00:01:23.898299 8290 sgd_solver.cpp:105] Iteration 5400, lr = 0.00343124 I0410 00:01:25.890710 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_5406.caffemodel I0410 00:01:27.031088 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_5406.solverstate I0410 00:01:29.059310 8290 solver.cpp:330] Iteration 5406, Testing net (#0) I0410 00:01:29.059340 8290 net.cpp:676] Ignoring source layer train-data I0410 00:01:31.334719 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:01:33.456110 8290 solver.cpp:397] Test net output #0: accuracy = 0.518995 I0410 00:01:33.456161 8290 solver.cpp:397] Test net output #1: loss = 2.0315 (* 1 = 2.0315 loss) I0410 00:01:35.368889 8290 solver.cpp:218] Iteration 5412 (1.04619 iter/s, 11.4702s/12 iters), loss = 0.588786 I0410 00:01:35.368949 8290 solver.cpp:237] Train net output #0: loss = 0.588786 (* 1 = 0.588786 loss) I0410 00:01:35.368963 8290 sgd_solver.cpp:105] Iteration 5412, lr = 0.00342309 I0410 00:01:40.358317 8290 solver.cpp:218] Iteration 5424 (2.40521 iter/s, 4.98918s/12 iters), loss = 0.666277 I0410 00:01:40.358372 8290 solver.cpp:237] Train net output #0: loss = 0.666277 (* 1 = 0.666277 loss) I0410 00:01:40.358387 8290 sgd_solver.cpp:105] Iteration 5424, lr = 0.00341497 I0410 00:01:45.230075 8290 solver.cpp:218] Iteration 5436 (2.4633 iter/s, 4.87152s/12 iters), loss = 0.842297 I0410 00:01:45.230180 8290 solver.cpp:237] Train net output #0: loss = 0.842297 (* 1 = 0.842297 loss) I0410 00:01:45.230190 8290 sgd_solver.cpp:105] Iteration 5436, lr = 0.00340686 I0410 00:01:50.101874 8290 solver.cpp:218] Iteration 5448 (2.4633 iter/s, 4.87151s/12 iters), loss = 0.70531 I0410 00:01:50.101938 8290 solver.cpp:237] Train net output #0: loss = 0.70531 (* 1 = 0.70531 loss) I0410 00:01:50.101951 8290 sgd_solver.cpp:105] Iteration 5448, lr = 0.00339877 I0410 00:01:54.996147 8290 solver.cpp:218] Iteration 5460 (2.45197 iter/s, 4.89403s/12 iters), loss = 0.949301 I0410 00:01:54.996199 8290 solver.cpp:237] Train net output #0: loss = 0.949301 (* 1 = 0.949301 loss) I0410 00:01:54.996210 8290 sgd_solver.cpp:105] Iteration 5460, lr = 0.0033907 I0410 00:01:55.552121 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:01:59.948421 8290 solver.cpp:218] Iteration 5472 (2.42325 iter/s, 4.95204s/12 iters), loss = 0.589799 I0410 00:01:59.948475 8290 solver.cpp:237] Train net output #0: loss = 0.589799 (* 1 = 0.589799 loss) I0410 00:01:59.948490 8290 sgd_solver.cpp:105] Iteration 5472, lr = 0.00338265 I0410 00:02:04.899266 8290 solver.cpp:218] Iteration 5484 (2.42395 iter/s, 4.9506s/12 iters), loss = 0.634544 I0410 00:02:04.899318 8290 solver.cpp:237] Train net output #0: loss = 0.634544 (* 1 = 0.634544 loss) I0410 00:02:04.899329 8290 sgd_solver.cpp:105] Iteration 5484, lr = 0.00337462 I0410 00:02:09.867267 8290 solver.cpp:218] Iteration 5496 (2.41557 iter/s, 4.96776s/12 iters), loss = 0.740435 I0410 00:02:09.867309 8290 solver.cpp:237] Train net output #0: loss = 0.740435 (* 1 = 0.740435 loss) I0410 00:02:09.867318 8290 sgd_solver.cpp:105] Iteration 5496, lr = 0.00336661 I0410 00:02:14.281339 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_5508.caffemodel I0410 00:02:16.389374 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_5508.solverstate I0410 00:02:17.577189 8290 solver.cpp:330] Iteration 5508, Testing net (#0) I0410 00:02:17.577212 8290 net.cpp:676] Ignoring source layer train-data I0410 00:02:19.859067 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:02:22.023528 8290 solver.cpp:397] Test net output #0: accuracy = 0.51348 I0410 00:02:22.023567 8290 solver.cpp:397] Test net output #1: loss = 2.10371 (* 1 = 2.10371 loss) I0410 00:02:22.105638 8290 solver.cpp:218] Iteration 5508 (0.980562 iter/s, 12.2379s/12 iters), loss = 0.91706 I0410 00:02:22.105695 8290 solver.cpp:237] Train net output #0: loss = 0.91706 (* 1 = 0.91706 loss) I0410 00:02:22.105708 8290 sgd_solver.cpp:105] Iteration 5508, lr = 0.00335861 I0410 00:02:26.332438 8290 solver.cpp:218] Iteration 5520 (2.83917 iter/s, 4.22658s/12 iters), loss = 0.655388 I0410 00:02:26.332481 8290 solver.cpp:237] Train net output #0: loss = 0.655388 (* 1 = 0.655388 loss) I0410 00:02:26.332490 8290 sgd_solver.cpp:105] Iteration 5520, lr = 0.00335064 I0410 00:02:27.489584 8290 blocking_queue.cpp:49] Waiting for data I0410 00:02:31.182519 8290 solver.cpp:218] Iteration 5532 (2.4743 iter/s, 4.84986s/12 iters), loss = 0.7302 I0410 00:02:31.182564 8290 solver.cpp:237] Train net output #0: loss = 0.7302 (* 1 = 0.7302 loss) I0410 00:02:31.182576 8290 sgd_solver.cpp:105] Iteration 5532, lr = 0.00334268 I0410 00:02:36.040218 8290 solver.cpp:218] Iteration 5544 (2.47042 iter/s, 4.85747s/12 iters), loss = 0.735182 I0410 00:02:36.040277 8290 solver.cpp:237] Train net output #0: loss = 0.735182 (* 1 = 0.735182 loss) I0410 00:02:36.040289 8290 sgd_solver.cpp:105] Iteration 5544, lr = 0.00333475 I0410 00:02:40.974306 8290 solver.cpp:218] Iteration 5556 (2.43218 iter/s, 4.93385s/12 iters), loss = 0.690315 I0410 00:02:40.974346 8290 solver.cpp:237] Train net output #0: loss = 0.690315 (* 1 = 0.690315 loss) I0410 00:02:40.974355 8290 sgd_solver.cpp:105] Iteration 5556, lr = 0.00332683 I0410 00:02:43.631821 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:02:45.917627 8290 solver.cpp:218] Iteration 5568 (2.42763 iter/s, 4.94309s/12 iters), loss = 0.605914 I0410 00:02:45.917686 8290 solver.cpp:237] Train net output #0: loss = 0.605914 (* 1 = 0.605914 loss) I0410 00:02:45.917701 8290 sgd_solver.cpp:105] Iteration 5568, lr = 0.00331893 I0410 00:02:50.832407 8290 solver.cpp:218] Iteration 5580 (2.44173 iter/s, 4.91454s/12 iters), loss = 0.73981 I0410 00:02:50.832502 8290 solver.cpp:237] Train net output #0: loss = 0.73981 (* 1 = 0.73981 loss) I0410 00:02:50.832512 8290 sgd_solver.cpp:105] Iteration 5580, lr = 0.00331105 I0410 00:02:55.877570 8290 solver.cpp:218] Iteration 5592 (2.37865 iter/s, 5.04488s/12 iters), loss = 0.817539 I0410 00:02:55.877631 8290 solver.cpp:237] Train net output #0: loss = 0.817539 (* 1 = 0.817539 loss) I0410 00:02:55.877643 8290 sgd_solver.cpp:105] Iteration 5592, lr = 0.00330319 I0410 00:03:00.822158 8290 solver.cpp:218] Iteration 5604 (2.42702 iter/s, 4.94434s/12 iters), loss = 0.754433 I0410 00:03:00.822213 8290 solver.cpp:237] Train net output #0: loss = 0.754433 (* 1 = 0.754433 loss) I0410 00:03:00.822227 8290 sgd_solver.cpp:105] Iteration 5604, lr = 0.00329535 I0410 00:03:02.840323 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_5610.caffemodel I0410 00:03:05.227392 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_5610.solverstate I0410 00:03:05.989826 8290 solver.cpp:330] Iteration 5610, Testing net (#0) I0410 00:03:05.989861 8290 net.cpp:676] Ignoring source layer train-data I0410 00:03:08.231685 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:03:10.431178 8290 solver.cpp:397] Test net output #0: accuracy = 0.516544 I0410 00:03:10.431226 8290 solver.cpp:397] Test net output #1: loss = 2.05436 (* 1 = 2.05436 loss) I0410 00:03:12.412919 8290 solver.cpp:218] Iteration 5616 (1.03535 iter/s, 11.5903s/12 iters), loss = 0.474643 I0410 00:03:12.412971 8290 solver.cpp:237] Train net output #0: loss = 0.474643 (* 1 = 0.474643 loss) I0410 00:03:12.412981 8290 sgd_solver.cpp:105] Iteration 5616, lr = 0.00328752 I0410 00:03:17.265465 8290 solver.cpp:218] Iteration 5628 (2.47305 iter/s, 4.85231s/12 iters), loss = 0.929968 I0410 00:03:17.265523 8290 solver.cpp:237] Train net output #0: loss = 0.929968 (* 1 = 0.929968 loss) I0410 00:03:17.265535 8290 sgd_solver.cpp:105] Iteration 5628, lr = 0.00327972 I0410 00:03:22.131011 8290 solver.cpp:218] Iteration 5640 (2.46644 iter/s, 4.86531s/12 iters), loss = 0.70494 I0410 00:03:22.131171 8290 solver.cpp:237] Train net output #0: loss = 0.70494 (* 1 = 0.70494 loss) I0410 00:03:22.131186 8290 sgd_solver.cpp:105] Iteration 5640, lr = 0.00327193 I0410 00:03:27.049278 8290 solver.cpp:218] Iteration 5652 (2.44005 iter/s, 4.91793s/12 iters), loss = 0.595452 I0410 00:03:27.049326 8290 solver.cpp:237] Train net output #0: loss = 0.595452 (* 1 = 0.595452 loss) I0410 00:03:27.049340 8290 sgd_solver.cpp:105] Iteration 5652, lr = 0.00326416 I0410 00:03:31.908968 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:03:32.074568 8290 solver.cpp:218] Iteration 5664 (2.38804 iter/s, 5.02505s/12 iters), loss = 0.548925 I0410 00:03:32.074623 8290 solver.cpp:237] Train net output #0: loss = 0.548925 (* 1 = 0.548925 loss) I0410 00:03:32.074636 8290 sgd_solver.cpp:105] Iteration 5664, lr = 0.00325641 I0410 00:03:36.882417 8290 solver.cpp:218] Iteration 5676 (2.49604 iter/s, 4.80762s/12 iters), loss = 0.712179 I0410 00:03:36.882468 8290 solver.cpp:237] Train net output #0: loss = 0.712179 (* 1 = 0.712179 loss) I0410 00:03:36.882479 8290 sgd_solver.cpp:105] Iteration 5676, lr = 0.00324868 I0410 00:03:41.769889 8290 solver.cpp:218] Iteration 5688 (2.45538 iter/s, 4.88724s/12 iters), loss = 0.59573 I0410 00:03:41.769946 8290 solver.cpp:237] Train net output #0: loss = 0.59573 (* 1 = 0.59573 loss) I0410 00:03:41.769977 8290 sgd_solver.cpp:105] Iteration 5688, lr = 0.00324097 I0410 00:03:46.698716 8290 solver.cpp:218] Iteration 5700 (2.43478 iter/s, 4.92859s/12 iters), loss = 0.913676 I0410 00:03:46.698773 8290 solver.cpp:237] Train net output #0: loss = 0.913676 (* 1 = 0.913676 loss) I0410 00:03:46.698786 8290 sgd_solver.cpp:105] Iteration 5700, lr = 0.00323328 I0410 00:03:51.143872 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_5712.caffemodel I0410 00:03:52.272445 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_5712.solverstate I0410 00:03:53.265769 8290 solver.cpp:330] Iteration 5712, Testing net (#0) I0410 00:03:53.265799 8290 net.cpp:676] Ignoring source layer train-data I0410 00:03:55.475025 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:03:57.866689 8290 solver.cpp:397] Test net output #0: accuracy = 0.512255 I0410 00:03:57.866729 8290 solver.cpp:397] Test net output #1: loss = 2.10088 (* 1 = 2.10088 loss) I0410 00:03:57.948678 8290 solver.cpp:218] Iteration 5712 (1.06671 iter/s, 11.2495s/12 iters), loss = 0.788975 I0410 00:03:57.948731 8290 solver.cpp:237] Train net output #0: loss = 0.788975 (* 1 = 0.788975 loss) I0410 00:03:57.948742 8290 sgd_solver.cpp:105] Iteration 5712, lr = 0.0032256 I0410 00:04:02.030028 8290 solver.cpp:218] Iteration 5724 (2.94035 iter/s, 4.08115s/12 iters), loss = 0.626184 I0410 00:04:02.030069 8290 solver.cpp:237] Train net output #0: loss = 0.626184 (* 1 = 0.626184 loss) I0410 00:04:02.030078 8290 sgd_solver.cpp:105] Iteration 5724, lr = 0.00321794 I0410 00:04:06.940503 8290 solver.cpp:218] Iteration 5736 (2.44387 iter/s, 4.91025s/12 iters), loss = 0.946896 I0410 00:04:06.940546 8290 solver.cpp:237] Train net output #0: loss = 0.946896 (* 1 = 0.946896 loss) I0410 00:04:06.940554 8290 sgd_solver.cpp:105] Iteration 5736, lr = 0.0032103 I0410 00:04:11.798530 8290 solver.cpp:218] Iteration 5748 (2.47026 iter/s, 4.8578s/12 iters), loss = 0.715861 I0410 00:04:11.798591 8290 solver.cpp:237] Train net output #0: loss = 0.715861 (* 1 = 0.715861 loss) I0410 00:04:11.798604 8290 sgd_solver.cpp:105] Iteration 5748, lr = 0.00320268 I0410 00:04:16.605235 8290 solver.cpp:218] Iteration 5760 (2.49664 iter/s, 4.80645s/12 iters), loss = 0.77179 I0410 00:04:16.605309 8290 solver.cpp:237] Train net output #0: loss = 0.77179 (* 1 = 0.77179 loss) I0410 00:04:16.605327 8290 sgd_solver.cpp:105] Iteration 5760, lr = 0.00319508 I0410 00:04:18.416002 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:04:21.394546 8290 solver.cpp:218] Iteration 5772 (2.50571 iter/s, 4.78906s/12 iters), loss = 0.801741 I0410 00:04:21.394605 8290 solver.cpp:237] Train net output #0: loss = 0.801741 (* 1 = 0.801741 loss) I0410 00:04:21.394619 8290 sgd_solver.cpp:105] Iteration 5772, lr = 0.00318749 I0410 00:04:26.241533 8290 solver.cpp:218] Iteration 5784 (2.47589 iter/s, 4.84675s/12 iters), loss = 0.789568 I0410 00:04:26.241684 8290 solver.cpp:237] Train net output #0: loss = 0.789568 (* 1 = 0.789568 loss) I0410 00:04:26.241698 8290 sgd_solver.cpp:105] Iteration 5784, lr = 0.00317992 I0410 00:04:31.154621 8290 solver.cpp:218] Iteration 5796 (2.44262 iter/s, 4.91276s/12 iters), loss = 0.603219 I0410 00:04:31.154665 8290 solver.cpp:237] Train net output #0: loss = 0.603219 (* 1 = 0.603219 loss) I0410 00:04:31.154675 8290 sgd_solver.cpp:105] Iteration 5796, lr = 0.00317237 I0410 00:04:36.021534 8290 solver.cpp:218] Iteration 5808 (2.46574 iter/s, 4.86669s/12 iters), loss = 0.699942 I0410 00:04:36.021579 8290 solver.cpp:237] Train net output #0: loss = 0.699942 (* 1 = 0.699942 loss) I0410 00:04:36.021586 8290 sgd_solver.cpp:105] Iteration 5808, lr = 0.00316484 I0410 00:04:38.006338 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_5814.caffemodel I0410 00:04:38.775503 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_5814.solverstate I0410 00:04:39.789558 8290 solver.cpp:330] Iteration 5814, Testing net (#0) I0410 00:04:39.789589 8290 net.cpp:676] Ignoring source layer train-data I0410 00:04:41.983492 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:04:44.263751 8290 solver.cpp:397] Test net output #0: accuracy = 0.549632 I0410 00:04:44.263789 8290 solver.cpp:397] Test net output #1: loss = 1.94675 (* 1 = 1.94675 loss) I0410 00:04:46.106922 8290 solver.cpp:218] Iteration 5820 (1.18989 iter/s, 10.085s/12 iters), loss = 0.620817 I0410 00:04:46.106969 8290 solver.cpp:237] Train net output #0: loss = 0.620817 (* 1 = 0.620817 loss) I0410 00:04:46.106979 8290 sgd_solver.cpp:105] Iteration 5820, lr = 0.00315733 I0410 00:04:50.966054 8290 solver.cpp:218] Iteration 5832 (2.4697 iter/s, 4.85889s/12 iters), loss = 0.652825 I0410 00:04:50.966116 8290 solver.cpp:237] Train net output #0: loss = 0.652825 (* 1 = 0.652825 loss) I0410 00:04:50.966130 8290 sgd_solver.cpp:105] Iteration 5832, lr = 0.00314983 I0410 00:04:55.790983 8290 solver.cpp:218] Iteration 5844 (2.48721 iter/s, 4.82469s/12 iters), loss = 0.573823 I0410 00:04:55.791039 8290 solver.cpp:237] Train net output #0: loss = 0.573823 (* 1 = 0.573823 loss) I0410 00:04:55.791051 8290 sgd_solver.cpp:105] Iteration 5844, lr = 0.00314235 I0410 00:05:00.628418 8290 solver.cpp:218] Iteration 5856 (2.48078 iter/s, 4.8372s/12 iters), loss = 0.576656 I0410 00:05:00.628571 8290 solver.cpp:237] Train net output #0: loss = 0.576656 (* 1 = 0.576656 loss) I0410 00:05:00.628584 8290 sgd_solver.cpp:105] Iteration 5856, lr = 0.00313489 I0410 00:05:04.682401 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:05:05.527017 8290 solver.cpp:218] Iteration 5868 (2.44985 iter/s, 4.89827s/12 iters), loss = 0.521781 I0410 00:05:05.527058 8290 solver.cpp:237] Train net output #0: loss = 0.521781 (* 1 = 0.521781 loss) I0410 00:05:05.527068 8290 sgd_solver.cpp:105] Iteration 5868, lr = 0.00312745 I0410 00:05:10.583498 8290 solver.cpp:218] Iteration 5880 (2.3733 iter/s, 5.05625s/12 iters), loss = 0.722767 I0410 00:05:10.583544 8290 solver.cpp:237] Train net output #0: loss = 0.722767 (* 1 = 0.722767 loss) I0410 00:05:10.583554 8290 sgd_solver.cpp:105] Iteration 5880, lr = 0.00312002 I0410 00:05:15.512598 8290 solver.cpp:218] Iteration 5892 (2.43464 iter/s, 4.92886s/12 iters), loss = 0.578501 I0410 00:05:15.512660 8290 solver.cpp:237] Train net output #0: loss = 0.578501 (* 1 = 0.578501 loss) I0410 00:05:15.512674 8290 sgd_solver.cpp:105] Iteration 5892, lr = 0.00311262 I0410 00:05:20.471253 8290 solver.cpp:218] Iteration 5904 (2.42013 iter/s, 4.95841s/12 iters), loss = 0.580986 I0410 00:05:20.471311 8290 solver.cpp:237] Train net output #0: loss = 0.580986 (* 1 = 0.580986 loss) I0410 00:05:20.471325 8290 sgd_solver.cpp:105] Iteration 5904, lr = 0.00310523 I0410 00:05:24.979913 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_5916.caffemodel I0410 00:05:25.836776 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_5916.solverstate I0410 00:05:27.184896 8290 solver.cpp:330] Iteration 5916, Testing net (#0) I0410 00:05:27.184931 8290 net.cpp:676] Ignoring source layer train-data I0410 00:05:29.309374 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:05:31.629616 8290 solver.cpp:397] Test net output #0: accuracy = 0.536765 I0410 00:05:31.629726 8290 solver.cpp:397] Test net output #1: loss = 2.00299 (* 1 = 2.00299 loss) I0410 00:05:31.711707 8290 solver.cpp:218] Iteration 5916 (1.06762 iter/s, 11.24s/12 iters), loss = 0.483691 I0410 00:05:31.711760 8290 solver.cpp:237] Train net output #0: loss = 0.483691 (* 1 = 0.483691 loss) I0410 00:05:31.711771 8290 sgd_solver.cpp:105] Iteration 5916, lr = 0.00309785 I0410 00:05:35.892721 8290 solver.cpp:218] Iteration 5928 (2.87026 iter/s, 4.18081s/12 iters), loss = 0.690266 I0410 00:05:35.892760 8290 solver.cpp:237] Train net output #0: loss = 0.690266 (* 1 = 0.690266 loss) I0410 00:05:35.892768 8290 sgd_solver.cpp:105] Iteration 5928, lr = 0.0030905 I0410 00:05:40.839993 8290 solver.cpp:218] Iteration 5940 (2.42569 iter/s, 4.94704s/12 iters), loss = 0.587467 I0410 00:05:40.840051 8290 solver.cpp:237] Train net output #0: loss = 0.587467 (* 1 = 0.587467 loss) I0410 00:05:40.840063 8290 sgd_solver.cpp:105] Iteration 5940, lr = 0.00308316 I0410 00:05:45.659807 8290 solver.cpp:218] Iteration 5952 (2.48985 iter/s, 4.81958s/12 iters), loss = 0.569671 I0410 00:05:45.659862 8290 solver.cpp:237] Train net output #0: loss = 0.569671 (* 1 = 0.569671 loss) I0410 00:05:45.659873 8290 sgd_solver.cpp:105] Iteration 5952, lr = 0.00307584 I0410 00:05:50.610430 8290 solver.cpp:218] Iteration 5964 (2.42406 iter/s, 4.95038s/12 iters), loss = 0.709331 I0410 00:05:50.610476 8290 solver.cpp:237] Train net output #0: loss = 0.709331 (* 1 = 0.709331 loss) I0410 00:05:50.610484 8290 sgd_solver.cpp:105] Iteration 5964, lr = 0.00306854 I0410 00:05:51.925163 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:05:55.520113 8290 solver.cpp:218] Iteration 5976 (2.44427 iter/s, 4.90944s/12 iters), loss = 0.60588 I0410 00:05:55.520184 8290 solver.cpp:237] Train net output #0: loss = 0.60588 (* 1 = 0.60588 loss) I0410 00:05:55.520201 8290 sgd_solver.cpp:105] Iteration 5976, lr = 0.00306125 I0410 00:06:00.440176 8290 solver.cpp:218] Iteration 5988 (2.43912 iter/s, 4.91982s/12 iters), loss = 0.62738 I0410 00:06:00.440224 8290 solver.cpp:237] Train net output #0: loss = 0.62738 (* 1 = 0.62738 loss) I0410 00:06:00.440235 8290 sgd_solver.cpp:105] Iteration 5988, lr = 0.00305398 I0410 00:06:05.333123 8290 solver.cpp:218] Iteration 6000 (2.45263 iter/s, 4.89271s/12 iters), loss = 0.463792 I0410 00:06:05.333276 8290 solver.cpp:237] Train net output #0: loss = 0.463792 (* 1 = 0.463792 loss) I0410 00:06:05.333288 8290 sgd_solver.cpp:105] Iteration 6000, lr = 0.00304673 I0410 00:06:10.169826 8290 solver.cpp:218] Iteration 6012 (2.4812 iter/s, 4.83637s/12 iters), loss = 0.58877 I0410 00:06:10.169879 8290 solver.cpp:237] Train net output #0: loss = 0.58877 (* 1 = 0.58877 loss) I0410 00:06:10.169890 8290 sgd_solver.cpp:105] Iteration 6012, lr = 0.0030395 I0410 00:06:12.130745 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6018.caffemodel I0410 00:06:13.078645 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6018.solverstate I0410 00:06:14.175768 8290 solver.cpp:330] Iteration 6018, Testing net (#0) I0410 00:06:14.175797 8290 net.cpp:676] Ignoring source layer train-data I0410 00:06:16.160061 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:06:18.540982 8290 solver.cpp:397] Test net output #0: accuracy = 0.530637 I0410 00:06:18.541030 8290 solver.cpp:397] Test net output #1: loss = 2.01316 (* 1 = 2.01316 loss) I0410 00:06:20.380129 8290 solver.cpp:218] Iteration 6024 (1.17533 iter/s, 10.2099s/12 iters), loss = 0.60988 I0410 00:06:20.380187 8290 solver.cpp:237] Train net output #0: loss = 0.60988 (* 1 = 0.60988 loss) I0410 00:06:20.380199 8290 sgd_solver.cpp:105] Iteration 6024, lr = 0.00303228 I0410 00:06:25.199316 8290 solver.cpp:218] Iteration 6036 (2.49017 iter/s, 4.81895s/12 iters), loss = 0.572384 I0410 00:06:25.199368 8290 solver.cpp:237] Train net output #0: loss = 0.572384 (* 1 = 0.572384 loss) I0410 00:06:25.199379 8290 sgd_solver.cpp:105] Iteration 6036, lr = 0.00302508 I0410 00:06:30.128013 8290 solver.cpp:218] Iteration 6048 (2.43484 iter/s, 4.92846s/12 iters), loss = 0.539664 I0410 00:06:30.128065 8290 solver.cpp:237] Train net output #0: loss = 0.539664 (* 1 = 0.539664 loss) I0410 00:06:30.128077 8290 sgd_solver.cpp:105] Iteration 6048, lr = 0.0030179 I0410 00:06:35.060801 8290 solver.cpp:218] Iteration 6060 (2.43282 iter/s, 4.93255s/12 iters), loss = 0.452945 I0410 00:06:35.060856 8290 solver.cpp:237] Train net output #0: loss = 0.452945 (* 1 = 0.452945 loss) I0410 00:06:35.060868 8290 sgd_solver.cpp:105] Iteration 6060, lr = 0.00301074 I0410 00:06:38.417596 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:06:39.945755 8290 solver.cpp:218] Iteration 6072 (2.45664 iter/s, 4.88471s/12 iters), loss = 0.730669 I0410 00:06:39.945814 8290 solver.cpp:237] Train net output #0: loss = 0.730669 (* 1 = 0.730669 loss) I0410 00:06:39.945827 8290 sgd_solver.cpp:105] Iteration 6072, lr = 0.00300359 I0410 00:06:44.894287 8290 solver.cpp:218] Iteration 6084 (2.42508 iter/s, 4.94829s/12 iters), loss = 0.48598 I0410 00:06:44.894336 8290 solver.cpp:237] Train net output #0: loss = 0.48598 (* 1 = 0.48598 loss) I0410 00:06:44.894347 8290 sgd_solver.cpp:105] Iteration 6084, lr = 0.00299646 I0410 00:06:49.787099 8290 solver.cpp:218] Iteration 6096 (2.45269 iter/s, 4.89258s/12 iters), loss = 0.504175 I0410 00:06:49.787147 8290 solver.cpp:237] Train net output #0: loss = 0.504175 (* 1 = 0.504175 loss) I0410 00:06:49.787158 8290 sgd_solver.cpp:105] Iteration 6096, lr = 0.00298934 I0410 00:06:54.639062 8290 solver.cpp:218] Iteration 6108 (2.47334 iter/s, 4.85173s/12 iters), loss = 0.609597 I0410 00:06:54.639112 8290 solver.cpp:237] Train net output #0: loss = 0.609597 (* 1 = 0.609597 loss) I0410 00:06:54.639122 8290 sgd_solver.cpp:105] Iteration 6108, lr = 0.00298225 I0410 00:06:59.127777 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6120.caffemodel I0410 00:07:00.735793 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6120.solverstate I0410 00:07:01.452494 8290 solver.cpp:330] Iteration 6120, Testing net (#0) I0410 00:07:01.452525 8290 net.cpp:676] Ignoring source layer train-data I0410 00:07:03.488490 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:07:05.887006 8290 solver.cpp:397] Test net output #0: accuracy = 0.540441 I0410 00:07:05.887044 8290 solver.cpp:397] Test net output #1: loss = 2.07781 (* 1 = 2.07781 loss) I0410 00:07:05.968050 8290 solver.cpp:218] Iteration 6120 (1.05927 iter/s, 11.3285s/12 iters), loss = 0.561702 I0410 00:07:05.968112 8290 solver.cpp:237] Train net output #0: loss = 0.561702 (* 1 = 0.561702 loss) I0410 00:07:05.968123 8290 sgd_solver.cpp:105] Iteration 6120, lr = 0.00297517 I0410 00:07:10.081701 8290 solver.cpp:218] Iteration 6132 (2.91727 iter/s, 4.11343s/12 iters), loss = 0.505237 I0410 00:07:10.081826 8290 solver.cpp:237] Train net output #0: loss = 0.505237 (* 1 = 0.505237 loss) I0410 00:07:10.081840 8290 sgd_solver.cpp:105] Iteration 6132, lr = 0.0029681 I0410 00:07:14.959841 8290 solver.cpp:218] Iteration 6144 (2.46011 iter/s, 4.87784s/12 iters), loss = 0.579004 I0410 00:07:14.959887 8290 solver.cpp:237] Train net output #0: loss = 0.579004 (* 1 = 0.579004 loss) I0410 00:07:14.959897 8290 sgd_solver.cpp:105] Iteration 6144, lr = 0.00296105 I0410 00:07:19.913527 8290 solver.cpp:218] Iteration 6156 (2.42255 iter/s, 4.95345s/12 iters), loss = 0.486451 I0410 00:07:19.913581 8290 solver.cpp:237] Train net output #0: loss = 0.486451 (* 1 = 0.486451 loss) I0410 00:07:19.913592 8290 sgd_solver.cpp:105] Iteration 6156, lr = 0.00295402 I0410 00:07:25.060628 8290 solver.cpp:218] Iteration 6168 (2.33152 iter/s, 5.14685s/12 iters), loss = 0.586908 I0410 00:07:25.060693 8290 solver.cpp:237] Train net output #0: loss = 0.586908 (* 1 = 0.586908 loss) I0410 00:07:25.060708 8290 sgd_solver.cpp:105] Iteration 6168, lr = 0.00294701 I0410 00:07:25.643815 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:07:29.969993 8290 solver.cpp:218] Iteration 6180 (2.44444 iter/s, 4.90911s/12 iters), loss = 0.449448 I0410 00:07:29.970053 8290 solver.cpp:237] Train net output #0: loss = 0.449448 (* 1 = 0.449448 loss) I0410 00:07:29.970067 8290 sgd_solver.cpp:105] Iteration 6180, lr = 0.00294001 I0410 00:07:34.908761 8290 solver.cpp:218] Iteration 6192 (2.42988 iter/s, 4.93852s/12 iters), loss = 0.480697 I0410 00:07:34.908802 8290 solver.cpp:237] Train net output #0: loss = 0.480697 (* 1 = 0.480697 loss) I0410 00:07:34.908811 8290 sgd_solver.cpp:105] Iteration 6192, lr = 0.00293303 I0410 00:07:39.865079 8290 solver.cpp:218] Iteration 6204 (2.42126 iter/s, 4.95609s/12 iters), loss = 0.577426 I0410 00:07:39.865123 8290 solver.cpp:237] Train net output #0: loss = 0.577426 (* 1 = 0.577426 loss) I0410 00:07:39.865132 8290 sgd_solver.cpp:105] Iteration 6204, lr = 0.00292607 I0410 00:07:44.858558 8290 solver.cpp:218] Iteration 6216 (2.40325 iter/s, 4.99325s/12 iters), loss = 0.584556 I0410 00:07:44.858665 8290 solver.cpp:237] Train net output #0: loss = 0.584556 (* 1 = 0.584556 loss) I0410 00:07:44.858675 8290 sgd_solver.cpp:105] Iteration 6216, lr = 0.00291912 I0410 00:07:46.877518 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6222.caffemodel I0410 00:07:48.155856 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6222.solverstate I0410 00:07:48.692081 8290 solver.cpp:330] Iteration 6222, Testing net (#0) I0410 00:07:48.692107 8290 net.cpp:676] Ignoring source layer train-data I0410 00:07:50.690016 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:07:51.809947 8290 blocking_queue.cpp:49] Waiting for data I0410 00:07:53.250202 8290 solver.cpp:397] Test net output #0: accuracy = 0.515931 I0410 00:07:53.250239 8290 solver.cpp:397] Test net output #1: loss = 2.25744 (* 1 = 2.25744 loss) I0410 00:07:55.194708 8290 solver.cpp:218] Iteration 6228 (1.16103 iter/s, 10.3357s/12 iters), loss = 0.63461 I0410 00:07:55.194758 8290 solver.cpp:237] Train net output #0: loss = 0.63461 (* 1 = 0.63461 loss) I0410 00:07:55.194769 8290 sgd_solver.cpp:105] Iteration 6228, lr = 0.00291219 I0410 00:08:00.038091 8290 solver.cpp:218] Iteration 6240 (2.47773 iter/s, 4.84315s/12 iters), loss = 0.580539 I0410 00:08:00.038143 8290 solver.cpp:237] Train net output #0: loss = 0.580539 (* 1 = 0.580539 loss) I0410 00:08:00.038156 8290 sgd_solver.cpp:105] Iteration 6240, lr = 0.00290528 I0410 00:08:04.888216 8290 solver.cpp:218] Iteration 6252 (2.47428 iter/s, 4.84989s/12 iters), loss = 0.406778 I0410 00:08:04.888276 8290 solver.cpp:237] Train net output #0: loss = 0.406778 (* 1 = 0.406778 loss) I0410 00:08:04.888289 8290 sgd_solver.cpp:105] Iteration 6252, lr = 0.00289838 I0410 00:08:09.780768 8290 solver.cpp:218] Iteration 6264 (2.45283 iter/s, 4.89231s/12 iters), loss = 0.574151 I0410 00:08:09.780822 8290 solver.cpp:237] Train net output #0: loss = 0.574151 (* 1 = 0.574151 loss) I0410 00:08:09.780835 8290 sgd_solver.cpp:105] Iteration 6264, lr = 0.0028915 I0410 00:08:12.412778 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:08:14.610579 8290 solver.cpp:218] Iteration 6276 (2.48469 iter/s, 4.82958s/12 iters), loss = 0.614238 I0410 00:08:14.610635 8290 solver.cpp:237] Train net output #0: loss = 0.614238 (* 1 = 0.614238 loss) I0410 00:08:14.610646 8290 sgd_solver.cpp:105] Iteration 6276, lr = 0.00288463 I0410 00:08:19.505864 8290 solver.cpp:218] Iteration 6288 (2.45146 iter/s, 4.89504s/12 iters), loss = 0.575991 I0410 00:08:19.506048 8290 solver.cpp:237] Train net output #0: loss = 0.575991 (* 1 = 0.575991 loss) I0410 00:08:19.506063 8290 sgd_solver.cpp:105] Iteration 6288, lr = 0.00287779 I0410 00:08:24.394554 8290 solver.cpp:218] Iteration 6300 (2.45483 iter/s, 4.88833s/12 iters), loss = 0.473831 I0410 00:08:24.394608 8290 solver.cpp:237] Train net output #0: loss = 0.473831 (* 1 = 0.473831 loss) I0410 00:08:24.394621 8290 sgd_solver.cpp:105] Iteration 6300, lr = 0.00287095 I0410 00:08:29.317675 8290 solver.cpp:218] Iteration 6312 (2.4376 iter/s, 4.92288s/12 iters), loss = 0.503258 I0410 00:08:29.317720 8290 solver.cpp:237] Train net output #0: loss = 0.503258 (* 1 = 0.503258 loss) I0410 00:08:29.317730 8290 sgd_solver.cpp:105] Iteration 6312, lr = 0.00286414 I0410 00:08:33.773023 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6324.caffemodel I0410 00:08:34.519413 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6324.solverstate I0410 00:08:35.091976 8290 solver.cpp:330] Iteration 6324, Testing net (#0) I0410 00:08:35.091995 8290 net.cpp:676] Ignoring source layer train-data I0410 00:08:36.966316 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:08:39.480388 8290 solver.cpp:397] Test net output #0: accuracy = 0.538603 I0410 00:08:39.480440 8290 solver.cpp:397] Test net output #1: loss = 2.11558 (* 1 = 2.11558 loss) I0410 00:08:39.562875 8290 solver.cpp:218] Iteration 6324 (1.17133 iter/s, 10.2448s/12 iters), loss = 0.371492 I0410 00:08:39.562927 8290 solver.cpp:237] Train net output #0: loss = 0.371492 (* 1 = 0.371492 loss) I0410 00:08:39.562937 8290 sgd_solver.cpp:105] Iteration 6324, lr = 0.00285734 I0410 00:08:43.633905 8290 solver.cpp:218] Iteration 6336 (2.94781 iter/s, 4.07082s/12 iters), loss = 0.535114 I0410 00:08:43.633975 8290 solver.cpp:237] Train net output #0: loss = 0.535114 (* 1 = 0.535114 loss) I0410 00:08:43.633989 8290 sgd_solver.cpp:105] Iteration 6336, lr = 0.00285055 I0410 00:08:48.705104 8290 solver.cpp:218] Iteration 6348 (2.36642 iter/s, 5.07095s/12 iters), loss = 0.488435 I0410 00:08:48.705159 8290 solver.cpp:237] Train net output #0: loss = 0.488435 (* 1 = 0.488435 loss) I0410 00:08:48.705171 8290 sgd_solver.cpp:105] Iteration 6348, lr = 0.00284379 I0410 00:08:53.529034 8290 solver.cpp:218] Iteration 6360 (2.48772 iter/s, 4.8237s/12 iters), loss = 0.798207 I0410 00:08:53.529165 8290 solver.cpp:237] Train net output #0: loss = 0.798207 (* 1 = 0.798207 loss) I0410 00:08:53.529175 8290 sgd_solver.cpp:105] Iteration 6360, lr = 0.00283703 I0410 00:08:58.279906 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:08:58.424419 8290 solver.cpp:218] Iteration 6372 (2.45144 iter/s, 4.89507s/12 iters), loss = 0.405343 I0410 00:08:58.424464 8290 solver.cpp:237] Train net output #0: loss = 0.405343 (* 1 = 0.405343 loss) I0410 00:08:58.424474 8290 sgd_solver.cpp:105] Iteration 6372, lr = 0.0028303 I0410 00:09:03.331527 8290 solver.cpp:218] Iteration 6384 (2.44555 iter/s, 4.90688s/12 iters), loss = 0.534917 I0410 00:09:03.331583 8290 solver.cpp:237] Train net output #0: loss = 0.534917 (* 1 = 0.534917 loss) I0410 00:09:03.331594 8290 sgd_solver.cpp:105] Iteration 6384, lr = 0.00282358 I0410 00:09:08.244582 8290 solver.cpp:218] Iteration 6396 (2.44259 iter/s, 4.91282s/12 iters), loss = 0.520506 I0410 00:09:08.244639 8290 solver.cpp:237] Train net output #0: loss = 0.520506 (* 1 = 0.520506 loss) I0410 00:09:08.244652 8290 sgd_solver.cpp:105] Iteration 6396, lr = 0.00281687 I0410 00:09:13.203960 8290 solver.cpp:218] Iteration 6408 (2.41978 iter/s, 4.95913s/12 iters), loss = 0.4304 I0410 00:09:13.204020 8290 solver.cpp:237] Train net output #0: loss = 0.4304 (* 1 = 0.4304 loss) I0410 00:09:13.204032 8290 sgd_solver.cpp:105] Iteration 6408, lr = 0.00281019 I0410 00:09:18.103283 8290 solver.cpp:218] Iteration 6420 (2.44944 iter/s, 4.89908s/12 iters), loss = 0.560743 I0410 00:09:18.103340 8290 solver.cpp:237] Train net output #0: loss = 0.560743 (* 1 = 0.560743 loss) I0410 00:09:18.103353 8290 sgd_solver.cpp:105] Iteration 6420, lr = 0.00280351 I0410 00:09:20.211769 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6426.caffemodel I0410 00:09:21.220504 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6426.solverstate I0410 00:09:21.765604 8290 solver.cpp:330] Iteration 6426, Testing net (#0) I0410 00:09:21.765632 8290 net.cpp:676] Ignoring source layer train-data I0410 00:09:23.659780 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:09:26.183552 8290 solver.cpp:397] Test net output #0: accuracy = 0.552696 I0410 00:09:26.183610 8290 solver.cpp:397] Test net output #1: loss = 2.06168 (* 1 = 2.06168 loss) I0410 00:09:28.070467 8290 solver.cpp:218] Iteration 6432 (1.204 iter/s, 9.96677s/12 iters), loss = 0.447168 I0410 00:09:28.070520 8290 solver.cpp:237] Train net output #0: loss = 0.447168 (* 1 = 0.447168 loss) I0410 00:09:28.070529 8290 sgd_solver.cpp:105] Iteration 6432, lr = 0.00279686 I0410 00:09:32.945250 8290 solver.cpp:218] Iteration 6444 (2.46177 iter/s, 4.87455s/12 iters), loss = 0.437946 I0410 00:09:32.945302 8290 solver.cpp:237] Train net output #0: loss = 0.437946 (* 1 = 0.437946 loss) I0410 00:09:32.945314 8290 sgd_solver.cpp:105] Iteration 6444, lr = 0.00279022 I0410 00:09:37.841274 8290 solver.cpp:218] Iteration 6456 (2.45109 iter/s, 4.89579s/12 iters), loss = 0.458514 I0410 00:09:37.841329 8290 solver.cpp:237] Train net output #0: loss = 0.458514 (* 1 = 0.458514 loss) I0410 00:09:37.841341 8290 sgd_solver.cpp:105] Iteration 6456, lr = 0.00278359 I0410 00:09:42.664781 8290 solver.cpp:218] Iteration 6468 (2.48794 iter/s, 4.82327s/12 iters), loss = 0.607888 I0410 00:09:42.664849 8290 solver.cpp:237] Train net output #0: loss = 0.607888 (* 1 = 0.607888 loss) I0410 00:09:42.664861 8290 sgd_solver.cpp:105] Iteration 6468, lr = 0.00277698 I0410 00:09:44.592664 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:09:47.508502 8290 solver.cpp:218] Iteration 6480 (2.47756 iter/s, 4.84347s/12 iters), loss = 0.606123 I0410 00:09:47.508558 8290 solver.cpp:237] Train net output #0: loss = 0.606123 (* 1 = 0.606123 loss) I0410 00:09:47.508571 8290 sgd_solver.cpp:105] Iteration 6480, lr = 0.00277039 I0410 00:09:52.310124 8290 solver.cpp:218] Iteration 6492 (2.49928 iter/s, 4.80138s/12 iters), loss = 0.534001 I0410 00:09:52.310181 8290 solver.cpp:237] Train net output #0: loss = 0.534001 (* 1 = 0.534001 loss) I0410 00:09:52.310194 8290 sgd_solver.cpp:105] Iteration 6492, lr = 0.00276381 I0410 00:09:57.146239 8290 solver.cpp:218] Iteration 6504 (2.48145 iter/s, 4.83587s/12 iters), loss = 0.47894 I0410 00:09:57.146394 8290 solver.cpp:237] Train net output #0: loss = 0.47894 (* 1 = 0.47894 loss) I0410 00:09:57.146407 8290 sgd_solver.cpp:105] Iteration 6504, lr = 0.00275725 I0410 00:10:01.946913 8290 solver.cpp:218] Iteration 6516 (2.49982 iter/s, 4.80034s/12 iters), loss = 0.483504 I0410 00:10:01.946966 8290 solver.cpp:237] Train net output #0: loss = 0.483504 (* 1 = 0.483504 loss) I0410 00:10:01.946979 8290 sgd_solver.cpp:105] Iteration 6516, lr = 0.00275071 I0410 00:10:06.335464 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6528.caffemodel I0410 00:10:07.100059 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6528.solverstate I0410 00:10:07.656551 8290 solver.cpp:330] Iteration 6528, Testing net (#0) I0410 00:10:07.656574 8290 net.cpp:676] Ignoring source layer train-data I0410 00:10:09.437356 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:10:12.134176 8290 solver.cpp:397] Test net output #0: accuracy = 0.54473 I0410 00:10:12.134225 8290 solver.cpp:397] Test net output #1: loss = 2.10635 (* 1 = 2.10635 loss) I0410 00:10:12.216429 8290 solver.cpp:218] Iteration 6528 (1.16856 iter/s, 10.2691s/12 iters), loss = 0.660729 I0410 00:10:12.216482 8290 solver.cpp:237] Train net output #0: loss = 0.660729 (* 1 = 0.660729 loss) I0410 00:10:12.216495 8290 sgd_solver.cpp:105] Iteration 6528, lr = 0.00274418 I0410 00:10:16.382246 8290 solver.cpp:218] Iteration 6540 (2.88073 iter/s, 4.16561s/12 iters), loss = 0.644949 I0410 00:10:16.382293 8290 solver.cpp:237] Train net output #0: loss = 0.644949 (* 1 = 0.644949 loss) I0410 00:10:16.382302 8290 sgd_solver.cpp:105] Iteration 6540, lr = 0.00273766 I0410 00:10:21.283988 8290 solver.cpp:218] Iteration 6552 (2.44822 iter/s, 4.90151s/12 iters), loss = 0.482584 I0410 00:10:21.284040 8290 solver.cpp:237] Train net output #0: loss = 0.482584 (* 1 = 0.482584 loss) I0410 00:10:21.284054 8290 sgd_solver.cpp:105] Iteration 6552, lr = 0.00273116 I0410 00:10:26.139734 8290 solver.cpp:218] Iteration 6564 (2.47142 iter/s, 4.85551s/12 iters), loss = 0.613881 I0410 00:10:26.139786 8290 solver.cpp:237] Train net output #0: loss = 0.613881 (* 1 = 0.613881 loss) I0410 00:10:26.139797 8290 sgd_solver.cpp:105] Iteration 6564, lr = 0.00272468 I0410 00:10:30.297065 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:10:31.053843 8290 solver.cpp:218] Iteration 6576 (2.44206 iter/s, 4.91388s/12 iters), loss = 0.465202 I0410 00:10:31.053895 8290 solver.cpp:237] Train net output #0: loss = 0.465202 (* 1 = 0.465202 loss) I0410 00:10:31.053906 8290 sgd_solver.cpp:105] Iteration 6576, lr = 0.00271821 I0410 00:10:35.918792 8290 solver.cpp:218] Iteration 6588 (2.46674 iter/s, 4.86472s/12 iters), loss = 0.331781 I0410 00:10:35.918833 8290 solver.cpp:237] Train net output #0: loss = 0.331781 (* 1 = 0.331781 loss) I0410 00:10:35.918843 8290 sgd_solver.cpp:105] Iteration 6588, lr = 0.00271175 I0410 00:10:40.812351 8290 solver.cpp:218] Iteration 6600 (2.45232 iter/s, 4.89333s/12 iters), loss = 0.44854 I0410 00:10:40.812412 8290 solver.cpp:237] Train net output #0: loss = 0.44854 (* 1 = 0.44854 loss) I0410 00:10:40.812425 8290 sgd_solver.cpp:105] Iteration 6600, lr = 0.00270532 I0410 00:10:45.692826 8290 solver.cpp:218] Iteration 6612 (2.4589 iter/s, 4.88023s/12 iters), loss = 0.310617 I0410 00:10:45.692881 8290 solver.cpp:237] Train net output #0: loss = 0.310617 (* 1 = 0.310617 loss) I0410 00:10:45.692893 8290 sgd_solver.cpp:105] Iteration 6612, lr = 0.00269889 I0410 00:10:50.575812 8290 solver.cpp:218] Iteration 6624 (2.45763 iter/s, 4.88275s/12 iters), loss = 0.59819 I0410 00:10:50.575865 8290 solver.cpp:237] Train net output #0: loss = 0.59819 (* 1 = 0.59819 loss) I0410 00:10:50.575878 8290 sgd_solver.cpp:105] Iteration 6624, lr = 0.00269248 I0410 00:10:52.580266 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6630.caffemodel I0410 00:10:53.322365 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6630.solverstate I0410 00:10:53.856657 8290 solver.cpp:330] Iteration 6630, Testing net (#0) I0410 00:10:53.856688 8290 net.cpp:676] Ignoring source layer train-data I0410 00:10:55.612336 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:10:58.355756 8290 solver.cpp:397] Test net output #0: accuracy = 0.544118 I0410 00:10:58.355808 8290 solver.cpp:397] Test net output #1: loss = 2.09781 (* 1 = 2.09781 loss) I0410 00:11:00.240206 8290 solver.cpp:218] Iteration 6636 (1.24172 iter/s, 9.66399s/12 iters), loss = 0.315081 I0410 00:11:00.240262 8290 solver.cpp:237] Train net output #0: loss = 0.315081 (* 1 = 0.315081 loss) I0410 00:11:00.240274 8290 sgd_solver.cpp:105] Iteration 6636, lr = 0.00268609 I0410 00:11:05.217854 8290 solver.cpp:218] Iteration 6648 (2.4109 iter/s, 4.9774s/12 iters), loss = 0.513665 I0410 00:11:05.218032 8290 solver.cpp:237] Train net output #0: loss = 0.513665 (* 1 = 0.513665 loss) I0410 00:11:05.218046 8290 sgd_solver.cpp:105] Iteration 6648, lr = 0.00267971 I0410 00:11:10.061324 8290 solver.cpp:218] Iteration 6660 (2.47774 iter/s, 4.84311s/12 iters), loss = 0.36853 I0410 00:11:10.061378 8290 solver.cpp:237] Train net output #0: loss = 0.36853 (* 1 = 0.36853 loss) I0410 00:11:10.061393 8290 sgd_solver.cpp:105] Iteration 6660, lr = 0.00267335 I0410 00:11:15.067926 8290 solver.cpp:218] Iteration 6672 (2.39695 iter/s, 5.00636s/12 iters), loss = 0.358161 I0410 00:11:15.067986 8290 solver.cpp:237] Train net output #0: loss = 0.358161 (* 1 = 0.358161 loss) I0410 00:11:15.067999 8290 sgd_solver.cpp:105] Iteration 6672, lr = 0.00266701 I0410 00:11:16.388207 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:11:19.979506 8290 solver.cpp:218] Iteration 6684 (2.44333 iter/s, 4.91133s/12 iters), loss = 0.449289 I0410 00:11:19.979563 8290 solver.cpp:237] Train net output #0: loss = 0.449289 (* 1 = 0.449289 loss) I0410 00:11:19.979578 8290 sgd_solver.cpp:105] Iteration 6684, lr = 0.00266067 I0410 00:11:24.903851 8290 solver.cpp:218] Iteration 6696 (2.43699 iter/s, 4.9241s/12 iters), loss = 0.581136 I0410 00:11:24.903906 8290 solver.cpp:237] Train net output #0: loss = 0.581136 (* 1 = 0.581136 loss) I0410 00:11:24.903916 8290 sgd_solver.cpp:105] Iteration 6696, lr = 0.00265436 I0410 00:11:29.780802 8290 solver.cpp:218] Iteration 6708 (2.46067 iter/s, 4.87672s/12 iters), loss = 0.492784 I0410 00:11:29.780856 8290 solver.cpp:237] Train net output #0: loss = 0.492784 (* 1 = 0.492784 loss) I0410 00:11:29.780869 8290 sgd_solver.cpp:105] Iteration 6708, lr = 0.00264805 I0410 00:11:34.647174 8290 solver.cpp:218] Iteration 6720 (2.46602 iter/s, 4.86613s/12 iters), loss = 0.488919 I0410 00:11:34.647230 8290 solver.cpp:237] Train net output #0: loss = 0.488919 (* 1 = 0.488919 loss) I0410 00:11:34.647243 8290 sgd_solver.cpp:105] Iteration 6720, lr = 0.00264177 I0410 00:11:39.116151 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6732.caffemodel I0410 00:11:40.570005 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6732.solverstate I0410 00:11:43.522338 8290 solver.cpp:330] Iteration 6732, Testing net (#0) I0410 00:11:43.522369 8290 net.cpp:676] Ignoring source layer train-data I0410 00:11:45.327801 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:11:47.955147 8290 solver.cpp:397] Test net output #0: accuracy = 0.550858 I0410 00:11:47.955193 8290 solver.cpp:397] Test net output #1: loss = 2.05955 (* 1 = 2.05955 loss) I0410 00:11:48.036324 8290 solver.cpp:218] Iteration 6732 (0.896284 iter/s, 13.3886s/12 iters), loss = 0.335625 I0410 00:11:48.036381 8290 solver.cpp:237] Train net output #0: loss = 0.335625 (* 1 = 0.335625 loss) I0410 00:11:48.036393 8290 sgd_solver.cpp:105] Iteration 6732, lr = 0.0026355 I0410 00:11:52.268337 8290 solver.cpp:218] Iteration 6744 (2.83567 iter/s, 4.2318s/12 iters), loss = 0.356864 I0410 00:11:52.268383 8290 solver.cpp:237] Train net output #0: loss = 0.356864 (* 1 = 0.356864 loss) I0410 00:11:52.268394 8290 sgd_solver.cpp:105] Iteration 6744, lr = 0.00262924 I0410 00:11:57.153434 8290 solver.cpp:218] Iteration 6756 (2.45657 iter/s, 4.88487s/12 iters), loss = 0.411146 I0410 00:11:57.153475 8290 solver.cpp:237] Train net output #0: loss = 0.411146 (* 1 = 0.411146 loss) I0410 00:11:57.153486 8290 sgd_solver.cpp:105] Iteration 6756, lr = 0.002623 I0410 00:12:02.010110 8290 solver.cpp:218] Iteration 6768 (2.47094 iter/s, 4.85645s/12 iters), loss = 0.440689 I0410 00:12:02.010152 8290 solver.cpp:237] Train net output #0: loss = 0.440689 (* 1 = 0.440689 loss) I0410 00:12:02.010160 8290 sgd_solver.cpp:105] Iteration 6768, lr = 0.00261677 I0410 00:12:05.354015 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:12:06.821746 8290 solver.cpp:218] Iteration 6780 (2.49407 iter/s, 4.8114s/12 iters), loss = 0.53905 I0410 00:12:06.821807 8290 solver.cpp:237] Train net output #0: loss = 0.53905 (* 1 = 0.53905 loss) I0410 00:12:06.821820 8290 sgd_solver.cpp:105] Iteration 6780, lr = 0.00261056 I0410 00:12:11.740805 8290 solver.cpp:218] Iteration 6792 (2.43961 iter/s, 4.91882s/12 iters), loss = 0.522296 I0410 00:12:11.740962 8290 solver.cpp:237] Train net output #0: loss = 0.522296 (* 1 = 0.522296 loss) I0410 00:12:11.740978 8290 sgd_solver.cpp:105] Iteration 6792, lr = 0.00260436 I0410 00:12:16.718518 8290 solver.cpp:218] Iteration 6804 (2.41091 iter/s, 4.97738s/12 iters), loss = 0.400508 I0410 00:12:16.718576 8290 solver.cpp:237] Train net output #0: loss = 0.400508 (* 1 = 0.400508 loss) I0410 00:12:16.718590 8290 sgd_solver.cpp:105] Iteration 6804, lr = 0.00259817 I0410 00:12:21.595844 8290 solver.cpp:218] Iteration 6816 (2.46049 iter/s, 4.87708s/12 iters), loss = 0.393078 I0410 00:12:21.595901 8290 solver.cpp:237] Train net output #0: loss = 0.393078 (* 1 = 0.393078 loss) I0410 00:12:21.595913 8290 sgd_solver.cpp:105] Iteration 6816, lr = 0.00259201 I0410 00:12:26.512301 8290 solver.cpp:218] Iteration 6828 (2.4409 iter/s, 4.91622s/12 iters), loss = 0.344447 I0410 00:12:26.512357 8290 solver.cpp:237] Train net output #0: loss = 0.344447 (* 1 = 0.344447 loss) I0410 00:12:26.512369 8290 sgd_solver.cpp:105] Iteration 6828, lr = 0.00258585 I0410 00:12:28.528702 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6834.caffemodel I0410 00:12:30.118472 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6834.solverstate I0410 00:12:30.648898 8290 solver.cpp:330] Iteration 6834, Testing net (#0) I0410 00:12:30.648926 8290 net.cpp:676] Ignoring source layer train-data I0410 00:12:32.513252 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:12:35.177022 8290 solver.cpp:397] Test net output #0: accuracy = 0.554534 I0410 00:12:35.177054 8290 solver.cpp:397] Test net output #1: loss = 2.13941 (* 1 = 2.13941 loss) I0410 00:12:36.897424 8290 solver.cpp:218] Iteration 6840 (1.15555 iter/s, 10.3847s/12 iters), loss = 0.384576 I0410 00:12:36.897473 8290 solver.cpp:237] Train net output #0: loss = 0.384576 (* 1 = 0.384576 loss) I0410 00:12:36.897485 8290 sgd_solver.cpp:105] Iteration 6840, lr = 0.00257971 I0410 00:12:41.809419 8290 solver.cpp:218] Iteration 6852 (2.44312 iter/s, 4.91176s/12 iters), loss = 0.415139 I0410 00:12:41.809506 8290 solver.cpp:237] Train net output #0: loss = 0.415139 (* 1 = 0.415139 loss) I0410 00:12:41.809520 8290 sgd_solver.cpp:105] Iteration 6852, lr = 0.00257359 I0410 00:12:46.713253 8290 solver.cpp:218] Iteration 6864 (2.4472 iter/s, 4.90357s/12 iters), loss = 0.444716 I0410 00:12:46.713308 8290 solver.cpp:237] Train net output #0: loss = 0.444716 (* 1 = 0.444716 loss) I0410 00:12:46.713322 8290 sgd_solver.cpp:105] Iteration 6864, lr = 0.00256748 I0410 00:12:51.882786 8290 solver.cpp:218] Iteration 6876 (2.3214 iter/s, 5.16929s/12 iters), loss = 0.378115 I0410 00:12:51.882830 8290 solver.cpp:237] Train net output #0: loss = 0.378115 (* 1 = 0.378115 loss) I0410 00:12:51.882838 8290 sgd_solver.cpp:105] Iteration 6876, lr = 0.00256138 I0410 00:12:52.495540 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:12:56.720636 8290 solver.cpp:218] Iteration 6888 (2.48055 iter/s, 4.83763s/12 iters), loss = 0.45459 I0410 00:12:56.720679 8290 solver.cpp:237] Train net output #0: loss = 0.45459 (* 1 = 0.45459 loss) I0410 00:12:56.720688 8290 sgd_solver.cpp:105] Iteration 6888, lr = 0.0025553 I0410 00:13:01.530130 8290 solver.cpp:218] Iteration 6900 (2.49518 iter/s, 4.80926s/12 iters), loss = 0.319153 I0410 00:13:01.530191 8290 solver.cpp:237] Train net output #0: loss = 0.319153 (* 1 = 0.319153 loss) I0410 00:13:01.530202 8290 sgd_solver.cpp:105] Iteration 6900, lr = 0.00254923 I0410 00:13:06.415678 8290 solver.cpp:218] Iteration 6912 (2.45635 iter/s, 4.88531s/12 iters), loss = 0.291447 I0410 00:13:06.415725 8290 solver.cpp:237] Train net output #0: loss = 0.291447 (* 1 = 0.291447 loss) I0410 00:13:06.415735 8290 sgd_solver.cpp:105] Iteration 6912, lr = 0.00254318 I0410 00:13:11.321732 8290 solver.cpp:218] Iteration 6924 (2.44607 iter/s, 4.90582s/12 iters), loss = 0.5254 I0410 00:13:11.321789 8290 solver.cpp:237] Train net output #0: loss = 0.5254 (* 1 = 0.5254 loss) I0410 00:13:11.321802 8290 sgd_solver.cpp:105] Iteration 6924, lr = 0.00253714 I0410 00:13:15.798259 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_6936.caffemodel I0410 00:13:17.803653 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_6936.solverstate I0410 00:13:18.414649 8290 solver.cpp:330] Iteration 6936, Testing net (#0) I0410 00:13:18.414677 8290 net.cpp:676] Ignoring source layer train-data I0410 00:13:18.771971 8290 blocking_queue.cpp:49] Waiting for data I0410 00:13:20.253600 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:13:22.963068 8290 solver.cpp:397] Test net output #0: accuracy = 0.558824 I0410 00:13:22.963117 8290 solver.cpp:397] Test net output #1: loss = 2.12237 (* 1 = 2.12237 loss) I0410 00:13:23.045222 8290 solver.cpp:218] Iteration 6936 (1.02363 iter/s, 11.723s/12 iters), loss = 0.33607 I0410 00:13:23.045284 8290 solver.cpp:237] Train net output #0: loss = 0.33607 (* 1 = 0.33607 loss) I0410 00:13:23.045298 8290 sgd_solver.cpp:105] Iteration 6936, lr = 0.00253112 I0410 00:13:27.181895 8290 solver.cpp:218] Iteration 6948 (2.90103 iter/s, 4.13646s/12 iters), loss = 0.267889 I0410 00:13:27.181943 8290 solver.cpp:237] Train net output #0: loss = 0.267889 (* 1 = 0.267889 loss) I0410 00:13:27.181952 8290 sgd_solver.cpp:105] Iteration 6948, lr = 0.00252511 I0410 00:13:32.019124 8290 solver.cpp:218] Iteration 6960 (2.48087 iter/s, 4.837s/12 iters), loss = 0.375408 I0410 00:13:32.019166 8290 solver.cpp:237] Train net output #0: loss = 0.375408 (* 1 = 0.375408 loss) I0410 00:13:32.019174 8290 sgd_solver.cpp:105] Iteration 6960, lr = 0.00251911 I0410 00:13:36.896560 8290 solver.cpp:218] Iteration 6972 (2.46042 iter/s, 4.87721s/12 iters), loss = 0.37758 I0410 00:13:36.896620 8290 solver.cpp:237] Train net output #0: loss = 0.37758 (* 1 = 0.37758 loss) I0410 00:13:36.896631 8290 sgd_solver.cpp:105] Iteration 6972, lr = 0.00251313 I0410 00:13:39.548703 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:13:41.697029 8290 solver.cpp:218] Iteration 6984 (2.49988 iter/s, 4.80023s/12 iters), loss = 0.396917 I0410 00:13:41.697089 8290 solver.cpp:237] Train net output #0: loss = 0.396917 (* 1 = 0.396917 loss) I0410 00:13:41.697101 8290 sgd_solver.cpp:105] Iteration 6984, lr = 0.00250717 I0410 00:13:46.521322 8290 solver.cpp:218] Iteration 6996 (2.48753 iter/s, 4.82406s/12 iters), loss = 0.370897 I0410 00:13:46.521481 8290 solver.cpp:237] Train net output #0: loss = 0.370897 (* 1 = 0.370897 loss) I0410 00:13:46.521493 8290 sgd_solver.cpp:105] Iteration 6996, lr = 0.00250121 I0410 00:13:51.304394 8290 solver.cpp:218] Iteration 7008 (2.50902 iter/s, 4.78274s/12 iters), loss = 0.39137 I0410 00:13:51.304438 8290 solver.cpp:237] Train net output #0: loss = 0.39137 (* 1 = 0.39137 loss) I0410 00:13:51.304450 8290 sgd_solver.cpp:105] Iteration 7008, lr = 0.00249528 I0410 00:13:56.113278 8290 solver.cpp:218] Iteration 7020 (2.4955 iter/s, 4.80865s/12 iters), loss = 0.360271 I0410 00:13:56.113343 8290 solver.cpp:237] Train net output #0: loss = 0.360271 (* 1 = 0.360271 loss) I0410 00:13:56.113354 8290 sgd_solver.cpp:105] Iteration 7020, lr = 0.00248935 I0410 00:14:00.920302 8290 solver.cpp:218] Iteration 7032 (2.49647 iter/s, 4.80678s/12 iters), loss = 0.215464 I0410 00:14:00.920361 8290 solver.cpp:237] Train net output #0: loss = 0.215464 (* 1 = 0.215464 loss) I0410 00:14:00.920372 8290 sgd_solver.cpp:105] Iteration 7032, lr = 0.00248344 I0410 00:14:02.898269 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7038.caffemodel I0410 00:14:03.715867 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7038.solverstate I0410 00:14:04.247495 8290 solver.cpp:330] Iteration 7038, Testing net (#0) I0410 00:14:04.247515 8290 net.cpp:676] Ignoring source layer train-data I0410 00:14:05.965710 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:14:08.881884 8290 solver.cpp:397] Test net output #0: accuracy = 0.548407 I0410 00:14:08.881923 8290 solver.cpp:397] Test net output #1: loss = 2.1936 (* 1 = 2.1936 loss) I0410 00:14:10.711333 8290 solver.cpp:218] Iteration 7044 (1.22566 iter/s, 9.79062s/12 iters), loss = 0.395046 I0410 00:14:10.711392 8290 solver.cpp:237] Train net output #0: loss = 0.395046 (* 1 = 0.395046 loss) I0410 00:14:10.711405 8290 sgd_solver.cpp:105] Iteration 7044, lr = 0.00247755 I0410 00:14:15.656577 8290 solver.cpp:218] Iteration 7056 (2.42669 iter/s, 4.945s/12 iters), loss = 0.213187 I0410 00:14:15.656632 8290 solver.cpp:237] Train net output #0: loss = 0.213187 (* 1 = 0.213187 loss) I0410 00:14:15.656646 8290 sgd_solver.cpp:105] Iteration 7056, lr = 0.00247166 I0410 00:14:20.613742 8290 solver.cpp:218] Iteration 7068 (2.42086 iter/s, 4.95692s/12 iters), loss = 0.441787 I0410 00:14:20.613878 8290 solver.cpp:237] Train net output #0: loss = 0.441787 (* 1 = 0.441787 loss) I0410 00:14:20.613893 8290 sgd_solver.cpp:105] Iteration 7068, lr = 0.0024658 I0410 00:14:25.413007 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:14:25.513123 8290 solver.cpp:218] Iteration 7080 (2.44945 iter/s, 4.89907s/12 iters), loss = 0.436123 I0410 00:14:25.513171 8290 solver.cpp:237] Train net output #0: loss = 0.436123 (* 1 = 0.436123 loss) I0410 00:14:25.513181 8290 sgd_solver.cpp:105] Iteration 7080, lr = 0.00245994 I0410 00:14:30.408571 8290 solver.cpp:218] Iteration 7092 (2.45137 iter/s, 4.89522s/12 iters), loss = 0.251691 I0410 00:14:30.408624 8290 solver.cpp:237] Train net output #0: loss = 0.251691 (* 1 = 0.251691 loss) I0410 00:14:30.408638 8290 sgd_solver.cpp:105] Iteration 7092, lr = 0.0024541 I0410 00:14:35.335680 8290 solver.cpp:218] Iteration 7104 (2.43562 iter/s, 4.92688s/12 iters), loss = 0.409799 I0410 00:14:35.335716 8290 solver.cpp:237] Train net output #0: loss = 0.409799 (* 1 = 0.409799 loss) I0410 00:14:35.335726 8290 sgd_solver.cpp:105] Iteration 7104, lr = 0.00244827 I0410 00:14:40.254062 8290 solver.cpp:218] Iteration 7116 (2.43993 iter/s, 4.91817s/12 iters), loss = 0.434978 I0410 00:14:40.254107 8290 solver.cpp:237] Train net output #0: loss = 0.434978 (* 1 = 0.434978 loss) I0410 00:14:40.254117 8290 sgd_solver.cpp:105] Iteration 7116, lr = 0.00244246 I0410 00:14:45.205792 8290 solver.cpp:218] Iteration 7128 (2.42351 iter/s, 4.9515s/12 iters), loss = 0.212282 I0410 00:14:45.205847 8290 solver.cpp:237] Train net output #0: loss = 0.212282 (* 1 = 0.212282 loss) I0410 00:14:45.205858 8290 sgd_solver.cpp:105] Iteration 7128, lr = 0.00243666 I0410 00:14:49.663784 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7140.caffemodel I0410 00:14:51.019032 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7140.solverstate I0410 00:14:51.626639 8290 solver.cpp:330] Iteration 7140, Testing net (#0) I0410 00:14:51.626662 8290 net.cpp:676] Ignoring source layer train-data I0410 00:14:53.261564 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:14:56.073537 8290 solver.cpp:397] Test net output #0: accuracy = 0.559436 I0410 00:14:56.073581 8290 solver.cpp:397] Test net output #1: loss = 2.08959 (* 1 = 2.08959 loss) I0410 00:14:56.155583 8290 solver.cpp:218] Iteration 7140 (1.09596 iter/s, 10.9494s/12 iters), loss = 0.377493 I0410 00:14:56.155629 8290 solver.cpp:237] Train net output #0: loss = 0.377493 (* 1 = 0.377493 loss) I0410 00:14:56.155640 8290 sgd_solver.cpp:105] Iteration 7140, lr = 0.00243088 I0410 00:15:00.215585 8290 solver.cpp:218] Iteration 7152 (2.95581 iter/s, 4.0598s/12 iters), loss = 0.45525 I0410 00:15:00.215641 8290 solver.cpp:237] Train net output #0: loss = 0.45525 (* 1 = 0.45525 loss) I0410 00:15:00.215654 8290 sgd_solver.cpp:105] Iteration 7152, lr = 0.00242511 I0410 00:15:05.112862 8290 solver.cpp:218] Iteration 7164 (2.45046 iter/s, 4.89704s/12 iters), loss = 0.403663 I0410 00:15:05.112916 8290 solver.cpp:237] Train net output #0: loss = 0.403663 (* 1 = 0.403663 loss) I0410 00:15:05.112928 8290 sgd_solver.cpp:105] Iteration 7164, lr = 0.00241935 I0410 00:15:10.089776 8290 solver.cpp:218] Iteration 7176 (2.41125 iter/s, 4.97668s/12 iters), loss = 0.320715 I0410 00:15:10.089818 8290 solver.cpp:237] Train net output #0: loss = 0.320715 (* 1 = 0.320715 loss) I0410 00:15:10.089828 8290 sgd_solver.cpp:105] Iteration 7176, lr = 0.0024136 I0410 00:15:12.182691 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:15:15.030546 8290 solver.cpp:218] Iteration 7188 (2.42888 iter/s, 4.94054s/12 iters), loss = 0.40741 I0410 00:15:15.030593 8290 solver.cpp:237] Train net output #0: loss = 0.40741 (* 1 = 0.40741 loss) I0410 00:15:15.030602 8290 sgd_solver.cpp:105] Iteration 7188, lr = 0.00240787 I0410 00:15:20.007007 8290 solver.cpp:218] Iteration 7200 (2.41147 iter/s, 4.97622s/12 iters), loss = 0.292301 I0410 00:15:20.007068 8290 solver.cpp:237] Train net output #0: loss = 0.292301 (* 1 = 0.292301 loss) I0410 00:15:20.007082 8290 sgd_solver.cpp:105] Iteration 7200, lr = 0.00240216 I0410 00:15:24.914036 8290 solver.cpp:218] Iteration 7212 (2.44559 iter/s, 4.90679s/12 iters), loss = 0.307401 I0410 00:15:24.914155 8290 solver.cpp:237] Train net output #0: loss = 0.307401 (* 1 = 0.307401 loss) I0410 00:15:24.914170 8290 sgd_solver.cpp:105] Iteration 7212, lr = 0.00239645 I0410 00:15:29.815397 8290 solver.cpp:218] Iteration 7224 (2.44845 iter/s, 4.90106s/12 iters), loss = 0.334204 I0410 00:15:29.815454 8290 solver.cpp:237] Train net output #0: loss = 0.334204 (* 1 = 0.334204 loss) I0410 00:15:29.815465 8290 sgd_solver.cpp:105] Iteration 7224, lr = 0.00239076 I0410 00:15:34.719729 8290 solver.cpp:218] Iteration 7236 (2.44693 iter/s, 4.9041s/12 iters), loss = 0.284888 I0410 00:15:34.719784 8290 solver.cpp:237] Train net output #0: loss = 0.284888 (* 1 = 0.284888 loss) I0410 00:15:34.719796 8290 sgd_solver.cpp:105] Iteration 7236, lr = 0.00238509 I0410 00:15:36.717363 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7242.caffemodel I0410 00:15:37.445669 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7242.solverstate I0410 00:15:37.967248 8290 solver.cpp:330] Iteration 7242, Testing net (#0) I0410 00:15:37.967275 8290 net.cpp:676] Ignoring source layer train-data I0410 00:15:39.588606 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:15:42.718309 8290 solver.cpp:397] Test net output #0: accuracy = 0.553309 I0410 00:15:42.718358 8290 solver.cpp:397] Test net output #1: loss = 2.05818 (* 1 = 2.05818 loss) I0410 00:15:44.495118 8290 solver.cpp:218] Iteration 7248 (1.22762 iter/s, 9.77498s/12 iters), loss = 0.35937 I0410 00:15:44.495187 8290 solver.cpp:237] Train net output #0: loss = 0.35937 (* 1 = 0.35937 loss) I0410 00:15:44.495198 8290 sgd_solver.cpp:105] Iteration 7248, lr = 0.00237942 I0410 00:15:49.294003 8290 solver.cpp:218] Iteration 7260 (2.50071 iter/s, 4.79864s/12 iters), loss = 0.444432 I0410 00:15:49.294064 8290 solver.cpp:237] Train net output #0: loss = 0.444432 (* 1 = 0.444432 loss) I0410 00:15:49.294077 8290 sgd_solver.cpp:105] Iteration 7260, lr = 0.00237378 I0410 00:15:54.189496 8290 solver.cpp:218] Iteration 7272 (2.45136 iter/s, 4.89525s/12 iters), loss = 0.32969 I0410 00:15:54.189541 8290 solver.cpp:237] Train net output #0: loss = 0.32969 (* 1 = 0.32969 loss) I0410 00:15:54.189549 8290 sgd_solver.cpp:105] Iteration 7272, lr = 0.00236814 I0410 00:15:58.357163 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:15:59.095175 8290 solver.cpp:218] Iteration 7284 (2.44626 iter/s, 4.90544s/12 iters), loss = 0.301174 I0410 00:15:59.095233 8290 solver.cpp:237] Train net output #0: loss = 0.301174 (* 1 = 0.301174 loss) I0410 00:15:59.095245 8290 sgd_solver.cpp:105] Iteration 7284, lr = 0.00236252 I0410 00:16:04.019037 8290 solver.cpp:218] Iteration 7296 (2.43723 iter/s, 4.92362s/12 iters), loss = 0.359445 I0410 00:16:04.019083 8290 solver.cpp:237] Train net output #0: loss = 0.359445 (* 1 = 0.359445 loss) I0410 00:16:04.019093 8290 sgd_solver.cpp:105] Iteration 7296, lr = 0.00235691 I0410 00:16:09.046658 8290 solver.cpp:218] Iteration 7308 (2.38693 iter/s, 5.02739s/12 iters), loss = 0.257474 I0410 00:16:09.046706 8290 solver.cpp:237] Train net output #0: loss = 0.257474 (* 1 = 0.257474 loss) I0410 00:16:09.046716 8290 sgd_solver.cpp:105] Iteration 7308, lr = 0.00235131 I0410 00:16:14.087198 8290 solver.cpp:218] Iteration 7320 (2.38081 iter/s, 5.0403s/12 iters), loss = 0.23265 I0410 00:16:14.087251 8290 solver.cpp:237] Train net output #0: loss = 0.23265 (* 1 = 0.23265 loss) I0410 00:16:14.087263 8290 sgd_solver.cpp:105] Iteration 7320, lr = 0.00234573 I0410 00:16:18.978297 8290 solver.cpp:218] Iteration 7332 (2.45356 iter/s, 4.89086s/12 iters), loss = 0.318418 I0410 00:16:18.978363 8290 solver.cpp:237] Train net output #0: loss = 0.318418 (* 1 = 0.318418 loss) I0410 00:16:18.978376 8290 sgd_solver.cpp:105] Iteration 7332, lr = 0.00234016 I0410 00:16:23.890508 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7344.caffemodel I0410 00:16:24.656288 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7344.solverstate I0410 00:16:25.192394 8290 solver.cpp:330] Iteration 7344, Testing net (#0) I0410 00:16:25.192423 8290 net.cpp:676] Ignoring source layer train-data I0410 00:16:26.699359 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:16:29.600721 8290 solver.cpp:397] Test net output #0: accuracy = 0.565564 I0410 00:16:29.600843 8290 solver.cpp:397] Test net output #1: loss = 2.11968 (* 1 = 2.11968 loss) I0410 00:16:29.682945 8290 solver.cpp:218] Iteration 7344 (1.12106 iter/s, 10.7042s/12 iters), loss = 0.302619 I0410 00:16:29.682996 8290 solver.cpp:237] Train net output #0: loss = 0.302619 (* 1 = 0.302619 loss) I0410 00:16:29.683009 8290 sgd_solver.cpp:105] Iteration 7344, lr = 0.0023346 I0410 00:16:33.847020 8290 solver.cpp:218] Iteration 7356 (2.88194 iter/s, 4.16387s/12 iters), loss = 0.435911 I0410 00:16:33.847064 8290 solver.cpp:237] Train net output #0: loss = 0.435911 (* 1 = 0.435911 loss) I0410 00:16:33.847074 8290 sgd_solver.cpp:105] Iteration 7356, lr = 0.00232906 I0410 00:16:38.706588 8290 solver.cpp:218] Iteration 7368 (2.46947 iter/s, 4.85934s/12 iters), loss = 0.244273 I0410 00:16:38.706645 8290 solver.cpp:237] Train net output #0: loss = 0.244273 (* 1 = 0.244273 loss) I0410 00:16:38.706656 8290 sgd_solver.cpp:105] Iteration 7368, lr = 0.00232353 I0410 00:16:43.727018 8290 solver.cpp:218] Iteration 7380 (2.39035 iter/s, 5.02018s/12 iters), loss = 0.262518 I0410 00:16:43.727077 8290 solver.cpp:237] Train net output #0: loss = 0.262518 (* 1 = 0.262518 loss) I0410 00:16:43.727090 8290 sgd_solver.cpp:105] Iteration 7380, lr = 0.00231802 I0410 00:16:45.098707 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:16:48.627827 8290 solver.cpp:218] Iteration 7392 (2.44869 iter/s, 4.90058s/12 iters), loss = 0.442333 I0410 00:16:48.627866 8290 solver.cpp:237] Train net output #0: loss = 0.442333 (* 1 = 0.442333 loss) I0410 00:16:48.627876 8290 sgd_solver.cpp:105] Iteration 7392, lr = 0.00231251 I0410 00:16:53.562304 8290 solver.cpp:218] Iteration 7404 (2.43198 iter/s, 4.93424s/12 iters), loss = 0.429372 I0410 00:16:53.562352 8290 solver.cpp:237] Train net output #0: loss = 0.429372 (* 1 = 0.429372 loss) I0410 00:16:53.562362 8290 sgd_solver.cpp:105] Iteration 7404, lr = 0.00230702 I0410 00:16:58.468744 8290 solver.cpp:218] Iteration 7416 (2.44588 iter/s, 4.90621s/12 iters), loss = 0.282718 I0410 00:16:58.468792 8290 solver.cpp:237] Train net output #0: loss = 0.282718 (* 1 = 0.282718 loss) I0410 00:16:58.468801 8290 sgd_solver.cpp:105] Iteration 7416, lr = 0.00230154 I0410 00:17:03.343253 8290 solver.cpp:218] Iteration 7428 (2.46191 iter/s, 4.87427s/12 iters), loss = 0.257598 I0410 00:17:03.343406 8290 solver.cpp:237] Train net output #0: loss = 0.257598 (* 1 = 0.257598 loss) I0410 00:17:03.343420 8290 sgd_solver.cpp:105] Iteration 7428, lr = 0.00229608 I0410 00:17:08.261379 8290 solver.cpp:218] Iteration 7440 (2.44012 iter/s, 4.9178s/12 iters), loss = 0.242623 I0410 00:17:08.261425 8290 solver.cpp:237] Train net output #0: loss = 0.242623 (* 1 = 0.242623 loss) I0410 00:17:08.261435 8290 sgd_solver.cpp:105] Iteration 7440, lr = 0.00229063 I0410 00:17:10.230624 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7446.caffemodel I0410 00:17:10.943652 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7446.solverstate I0410 00:17:11.469255 8290 solver.cpp:330] Iteration 7446, Testing net (#0) I0410 00:17:11.469274 8290 net.cpp:676] Ignoring source layer train-data I0410 00:17:12.894330 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:17:15.785437 8290 solver.cpp:397] Test net output #0: accuracy = 0.558824 I0410 00:17:15.785488 8290 solver.cpp:397] Test net output #1: loss = 2.10302 (* 1 = 2.10302 loss) I0410 00:17:17.558974 8290 solver.cpp:218] Iteration 7452 (1.29071 iter/s, 9.29721s/12 iters), loss = 0.24377 I0410 00:17:17.559031 8290 solver.cpp:237] Train net output #0: loss = 0.24377 (* 1 = 0.24377 loss) I0410 00:17:17.559044 8290 sgd_solver.cpp:105] Iteration 7452, lr = 0.00228519 I0410 00:17:22.378674 8290 solver.cpp:218] Iteration 7464 (2.48991 iter/s, 4.81946s/12 iters), loss = 0.259899 I0410 00:17:22.378737 8290 solver.cpp:237] Train net output #0: loss = 0.259899 (* 1 = 0.259899 loss) I0410 00:17:22.378751 8290 sgd_solver.cpp:105] Iteration 7464, lr = 0.00227976 I0410 00:17:27.204377 8290 solver.cpp:218] Iteration 7476 (2.48681 iter/s, 4.82546s/12 iters), loss = 0.375804 I0410 00:17:27.204438 8290 solver.cpp:237] Train net output #0: loss = 0.375804 (* 1 = 0.375804 loss) I0410 00:17:27.204449 8290 sgd_solver.cpp:105] Iteration 7476, lr = 0.00227435 I0410 00:17:30.584125 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:17:32.120705 8290 solver.cpp:218] Iteration 7488 (2.44096 iter/s, 4.91609s/12 iters), loss = 0.358901 I0410 00:17:32.120749 8290 solver.cpp:237] Train net output #0: loss = 0.358901 (* 1 = 0.358901 loss) I0410 00:17:32.120757 8290 sgd_solver.cpp:105] Iteration 7488, lr = 0.00226895 I0410 00:17:37.006624 8290 solver.cpp:218] Iteration 7500 (2.45615 iter/s, 4.88569s/12 iters), loss = 0.244359 I0410 00:17:37.006748 8290 solver.cpp:237] Train net output #0: loss = 0.244359 (* 1 = 0.244359 loss) I0410 00:17:37.006762 8290 sgd_solver.cpp:105] Iteration 7500, lr = 0.00226357 I0410 00:17:41.884639 8290 solver.cpp:218] Iteration 7512 (2.46017 iter/s, 4.87771s/12 iters), loss = 0.330834 I0410 00:17:41.884699 8290 solver.cpp:237] Train net output #0: loss = 0.330834 (* 1 = 0.330834 loss) I0410 00:17:41.884711 8290 sgd_solver.cpp:105] Iteration 7512, lr = 0.00225819 I0410 00:17:46.810220 8290 solver.cpp:218] Iteration 7524 (2.43638 iter/s, 4.92533s/12 iters), loss = 0.354861 I0410 00:17:46.810283 8290 solver.cpp:237] Train net output #0: loss = 0.354861 (* 1 = 0.354861 loss) I0410 00:17:46.810297 8290 sgd_solver.cpp:105] Iteration 7524, lr = 0.00225283 I0410 00:17:51.727787 8290 solver.cpp:218] Iteration 7536 (2.44035 iter/s, 4.91732s/12 iters), loss = 0.157747 I0410 00:17:51.727838 8290 solver.cpp:237] Train net output #0: loss = 0.157747 (* 1 = 0.157747 loss) I0410 00:17:51.727849 8290 sgd_solver.cpp:105] Iteration 7536, lr = 0.00224748 I0410 00:17:56.122534 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7548.caffemodel I0410 00:17:56.937644 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7548.solverstate I0410 00:17:57.481019 8290 solver.cpp:330] Iteration 7548, Testing net (#0) I0410 00:17:57.481047 8290 net.cpp:676] Ignoring source layer train-data I0410 00:17:58.868945 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:18:02.006114 8290 solver.cpp:397] Test net output #0: accuracy = 0.563725 I0410 00:18:02.006166 8290 solver.cpp:397] Test net output #1: loss = 2.13985 (* 1 = 2.13985 loss) I0410 00:18:02.088582 8290 solver.cpp:218] Iteration 7548 (1.15826 iter/s, 10.3604s/12 iters), loss = 0.370903 I0410 00:18:02.088639 8290 solver.cpp:237] Train net output #0: loss = 0.370903 (* 1 = 0.370903 loss) I0410 00:18:02.088651 8290 sgd_solver.cpp:105] Iteration 7548, lr = 0.00224215 I0410 00:18:06.331981 8290 solver.cpp:218] Iteration 7560 (2.82806 iter/s, 4.24318s/12 iters), loss = 0.267183 I0410 00:18:06.332024 8290 solver.cpp:237] Train net output #0: loss = 0.267183 (* 1 = 0.267183 loss) I0410 00:18:06.332033 8290 sgd_solver.cpp:105] Iteration 7560, lr = 0.00223682 I0410 00:18:11.183434 8290 solver.cpp:218] Iteration 7572 (2.4736 iter/s, 4.85123s/12 iters), loss = 0.306771 I0410 00:18:11.183571 8290 solver.cpp:237] Train net output #0: loss = 0.306771 (* 1 = 0.306771 loss) I0410 00:18:11.183581 8290 sgd_solver.cpp:105] Iteration 7572, lr = 0.00223151 I0410 00:18:16.073840 8290 solver.cpp:218] Iteration 7584 (2.45394 iter/s, 4.89009s/12 iters), loss = 0.320192 I0410 00:18:16.073892 8290 solver.cpp:237] Train net output #0: loss = 0.320192 (* 1 = 0.320192 loss) I0410 00:18:16.073904 8290 sgd_solver.cpp:105] Iteration 7584, lr = 0.00222621 I0410 00:18:16.702793 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:18:20.929947 8290 solver.cpp:218] Iteration 7596 (2.47124 iter/s, 4.85587s/12 iters), loss = 0.356399 I0410 00:18:20.930024 8290 solver.cpp:237] Train net output #0: loss = 0.356399 (* 1 = 0.356399 loss) I0410 00:18:20.930037 8290 sgd_solver.cpp:105] Iteration 7596, lr = 0.00222093 I0410 00:18:25.769974 8290 solver.cpp:218] Iteration 7608 (2.47947 iter/s, 4.83975s/12 iters), loss = 0.307699 I0410 00:18:25.770032 8290 solver.cpp:237] Train net output #0: loss = 0.307699 (* 1 = 0.307699 loss) I0410 00:18:25.770043 8290 sgd_solver.cpp:105] Iteration 7608, lr = 0.00221565 I0410 00:18:30.695399 8290 solver.cpp:218] Iteration 7620 (2.43646 iter/s, 4.92519s/12 iters), loss = 0.292419 I0410 00:18:30.695444 8290 solver.cpp:237] Train net output #0: loss = 0.292419 (* 1 = 0.292419 loss) I0410 00:18:30.695453 8290 sgd_solver.cpp:105] Iteration 7620, lr = 0.00221039 I0410 00:18:31.835811 8290 blocking_queue.cpp:49] Waiting for data I0410 00:18:35.561364 8290 solver.cpp:218] Iteration 7632 (2.46623 iter/s, 4.86574s/12 iters), loss = 0.373504 I0410 00:18:35.561414 8290 solver.cpp:237] Train net output #0: loss = 0.373504 (* 1 = 0.373504 loss) I0410 00:18:35.561425 8290 sgd_solver.cpp:105] Iteration 7632, lr = 0.00220515 I0410 00:18:40.506567 8290 solver.cpp:218] Iteration 7644 (2.42671 iter/s, 4.94496s/12 iters), loss = 0.305735 I0410 00:18:40.506623 8290 solver.cpp:237] Train net output #0: loss = 0.305735 (* 1 = 0.305735 loss) I0410 00:18:40.506635 8290 sgd_solver.cpp:105] Iteration 7644, lr = 0.00219991 I0410 00:18:42.472445 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7650.caffemodel I0410 00:18:46.394789 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7650.solverstate I0410 00:18:50.092567 8290 solver.cpp:330] Iteration 7650, Testing net (#0) I0410 00:18:50.092599 8290 net.cpp:676] Ignoring source layer train-data I0410 00:18:51.550907 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:18:54.536432 8290 solver.cpp:397] Test net output #0: accuracy = 0.566176 I0410 00:18:54.536497 8290 solver.cpp:397] Test net output #1: loss = 2.11352 (* 1 = 2.11352 loss) I0410 00:18:56.308341 8290 solver.cpp:218] Iteration 7656 (0.759438 iter/s, 15.8012s/12 iters), loss = 0.386883 I0410 00:18:56.308391 8290 solver.cpp:237] Train net output #0: loss = 0.386883 (* 1 = 0.386883 loss) I0410 00:18:56.308403 8290 sgd_solver.cpp:105] Iteration 7656, lr = 0.00219469 I0410 00:19:01.144536 8290 solver.cpp:218] Iteration 7668 (2.48141 iter/s, 4.83596s/12 iters), loss = 0.313453 I0410 00:19:01.144593 8290 solver.cpp:237] Train net output #0: loss = 0.313453 (* 1 = 0.313453 loss) I0410 00:19:01.144606 8290 sgd_solver.cpp:105] Iteration 7668, lr = 0.00218948 I0410 00:19:06.223179 8290 solver.cpp:218] Iteration 7680 (2.36295 iter/s, 5.0784s/12 iters), loss = 0.178307 I0410 00:19:06.223224 8290 solver.cpp:237] Train net output #0: loss = 0.178307 (* 1 = 0.178307 loss) I0410 00:19:06.223234 8290 sgd_solver.cpp:105] Iteration 7680, lr = 0.00218428 I0410 00:19:08.907179 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:19:11.017411 8290 solver.cpp:218] Iteration 7692 (2.50313 iter/s, 4.79401s/12 iters), loss = 0.392266 I0410 00:19:11.017460 8290 solver.cpp:237] Train net output #0: loss = 0.392266 (* 1 = 0.392266 loss) I0410 00:19:11.017470 8290 sgd_solver.cpp:105] Iteration 7692, lr = 0.00217909 I0410 00:19:15.868811 8290 solver.cpp:218] Iteration 7704 (2.47363 iter/s, 4.85117s/12 iters), loss = 0.312185 I0410 00:19:15.868932 8290 solver.cpp:237] Train net output #0: loss = 0.312185 (* 1 = 0.312185 loss) I0410 00:19:15.868947 8290 sgd_solver.cpp:105] Iteration 7704, lr = 0.00217392 I0410 00:19:20.688429 8290 solver.cpp:218] Iteration 7716 (2.48998 iter/s, 4.81932s/12 iters), loss = 0.202147 I0410 00:19:20.688482 8290 solver.cpp:237] Train net output #0: loss = 0.202147 (* 1 = 0.202147 loss) I0410 00:19:20.688495 8290 sgd_solver.cpp:105] Iteration 7716, lr = 0.00216876 I0410 00:19:25.565441 8290 solver.cpp:218] Iteration 7728 (2.46064 iter/s, 4.87678s/12 iters), loss = 0.158035 I0410 00:19:25.565501 8290 solver.cpp:237] Train net output #0: loss = 0.158035 (* 1 = 0.158035 loss) I0410 00:19:25.565515 8290 sgd_solver.cpp:105] Iteration 7728, lr = 0.00216361 I0410 00:19:30.440733 8290 solver.cpp:218] Iteration 7740 (2.46151 iter/s, 4.87505s/12 iters), loss = 0.341519 I0410 00:19:30.440790 8290 solver.cpp:237] Train net output #0: loss = 0.341519 (* 1 = 0.341519 loss) I0410 00:19:30.440804 8290 sgd_solver.cpp:105] Iteration 7740, lr = 0.00215847 I0410 00:19:34.900364 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7752.caffemodel I0410 00:19:36.022127 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7752.solverstate I0410 00:19:36.999070 8290 solver.cpp:330] Iteration 7752, Testing net (#0) I0410 00:19:36.999089 8290 net.cpp:676] Ignoring source layer train-data I0410 00:19:38.405129 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:19:41.485286 8290 solver.cpp:397] Test net output #0: accuracy = 0.57598 I0410 00:19:41.485334 8290 solver.cpp:397] Test net output #1: loss = 2.12999 (* 1 = 2.12999 loss) I0410 00:19:41.568823 8290 solver.cpp:218] Iteration 7752 (1.0784 iter/s, 11.1276s/12 iters), loss = 0.237162 I0410 00:19:41.568878 8290 solver.cpp:237] Train net output #0: loss = 0.237162 (* 1 = 0.237162 loss) I0410 00:19:41.568890 8290 sgd_solver.cpp:105] Iteration 7752, lr = 0.00215335 I0410 00:19:45.660111 8290 solver.cpp:218] Iteration 7764 (2.93322 iter/s, 4.09107s/12 iters), loss = 0.339464 I0410 00:19:45.660163 8290 solver.cpp:237] Train net output #0: loss = 0.339464 (* 1 = 0.339464 loss) I0410 00:19:45.660173 8290 sgd_solver.cpp:105] Iteration 7764, lr = 0.00214823 I0410 00:19:50.758641 8290 solver.cpp:218] Iteration 7776 (2.35373 iter/s, 5.09829s/12 iters), loss = 0.155812 I0410 00:19:50.758786 8290 solver.cpp:237] Train net output #0: loss = 0.155812 (* 1 = 0.155812 loss) I0410 00:19:50.758800 8290 sgd_solver.cpp:105] Iteration 7776, lr = 0.00214313 I0410 00:19:55.640420 8290 solver.cpp:218] Iteration 7788 (2.45828 iter/s, 4.88146s/12 iters), loss = 0.2173 I0410 00:19:55.640470 8290 solver.cpp:237] Train net output #0: loss = 0.2173 (* 1 = 0.2173 loss) I0410 00:19:55.640480 8290 sgd_solver.cpp:105] Iteration 7788, lr = 0.00213805 I0410 00:19:55.648499 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:20:00.465906 8290 solver.cpp:218] Iteration 7800 (2.48692 iter/s, 4.82525s/12 iters), loss = 0.237393 I0410 00:20:00.465982 8290 solver.cpp:237] Train net output #0: loss = 0.237393 (* 1 = 0.237393 loss) I0410 00:20:00.465996 8290 sgd_solver.cpp:105] Iteration 7800, lr = 0.00213297 I0410 00:20:05.263590 8290 solver.cpp:218] Iteration 7812 (2.50134 iter/s, 4.79743s/12 iters), loss = 0.211319 I0410 00:20:05.263638 8290 solver.cpp:237] Train net output #0: loss = 0.211319 (* 1 = 0.211319 loss) I0410 00:20:05.263648 8290 sgd_solver.cpp:105] Iteration 7812, lr = 0.00212791 I0410 00:20:10.148105 8290 solver.cpp:218] Iteration 7824 (2.45686 iter/s, 4.88428s/12 iters), loss = 0.284689 I0410 00:20:10.148151 8290 solver.cpp:237] Train net output #0: loss = 0.284689 (* 1 = 0.284689 loss) I0410 00:20:10.148159 8290 sgd_solver.cpp:105] Iteration 7824, lr = 0.00212285 I0410 00:20:15.119771 8290 solver.cpp:218] Iteration 7836 (2.41379 iter/s, 4.97143s/12 iters), loss = 0.306284 I0410 00:20:15.119827 8290 solver.cpp:237] Train net output #0: loss = 0.306284 (* 1 = 0.306284 loss) I0410 00:20:15.119840 8290 sgd_solver.cpp:105] Iteration 7836, lr = 0.00211781 I0410 00:20:20.033447 8290 solver.cpp:218] Iteration 7848 (2.44228 iter/s, 4.91343s/12 iters), loss = 0.238862 I0410 00:20:20.033501 8290 solver.cpp:237] Train net output #0: loss = 0.238862 (* 1 = 0.238862 loss) I0410 00:20:20.033514 8290 sgd_solver.cpp:105] Iteration 7848, lr = 0.00211279 I0410 00:20:22.027971 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7854.caffemodel I0410 00:20:23.723453 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7854.solverstate I0410 00:20:26.684101 8290 solver.cpp:330] Iteration 7854, Testing net (#0) I0410 00:20:26.684130 8290 net.cpp:676] Ignoring source layer train-data I0410 00:20:28.059895 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:20:31.121490 8290 solver.cpp:397] Test net output #0: accuracy = 0.571078 I0410 00:20:31.121536 8290 solver.cpp:397] Test net output #1: loss = 2.13229 (* 1 = 2.13229 loss) I0410 00:20:32.955960 8290 solver.cpp:218] Iteration 7860 (0.928649 iter/s, 12.922s/12 iters), loss = 0.421062 I0410 00:20:32.956015 8290 solver.cpp:237] Train net output #0: loss = 0.421062 (* 1 = 0.421062 loss) I0410 00:20:32.956027 8290 sgd_solver.cpp:105] Iteration 7860, lr = 0.00210777 I0410 00:20:37.876070 8290 solver.cpp:218] Iteration 7872 (2.43909 iter/s, 4.91987s/12 iters), loss = 0.222452 I0410 00:20:37.876117 8290 solver.cpp:237] Train net output #0: loss = 0.222452 (* 1 = 0.222452 loss) I0410 00:20:37.876129 8290 sgd_solver.cpp:105] Iteration 7872, lr = 0.00210277 I0410 00:20:42.784296 8290 solver.cpp:218] Iteration 7884 (2.44499 iter/s, 4.908s/12 iters), loss = 0.132319 I0410 00:20:42.784339 8290 solver.cpp:237] Train net output #0: loss = 0.132319 (* 1 = 0.132319 loss) I0410 00:20:42.784348 8290 sgd_solver.cpp:105] Iteration 7884, lr = 0.00209777 I0410 00:20:44.874143 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:20:47.670291 8290 solver.cpp:218] Iteration 7896 (2.45611 iter/s, 4.88577s/12 iters), loss = 0.390463 I0410 00:20:47.670356 8290 solver.cpp:237] Train net output #0: loss = 0.390463 (* 1 = 0.390463 loss) I0410 00:20:47.670370 8290 sgd_solver.cpp:105] Iteration 7896, lr = 0.00209279 I0410 00:20:52.582386 8290 solver.cpp:218] Iteration 7908 (2.44307 iter/s, 4.91185s/12 iters), loss = 0.280547 I0410 00:20:52.582501 8290 solver.cpp:237] Train net output #0: loss = 0.280547 (* 1 = 0.280547 loss) I0410 00:20:52.582513 8290 sgd_solver.cpp:105] Iteration 7908, lr = 0.00208782 I0410 00:20:57.531527 8290 solver.cpp:218] Iteration 7920 (2.42481 iter/s, 4.94884s/12 iters), loss = 0.431512 I0410 00:20:57.531575 8290 solver.cpp:237] Train net output #0: loss = 0.431512 (* 1 = 0.431512 loss) I0410 00:20:57.531586 8290 sgd_solver.cpp:105] Iteration 7920, lr = 0.00208287 I0410 00:21:02.562114 8290 solver.cpp:218] Iteration 7932 (2.38552 iter/s, 5.03034s/12 iters), loss = 0.265816 I0410 00:21:02.562167 8290 solver.cpp:237] Train net output #0: loss = 0.265816 (* 1 = 0.265816 loss) I0410 00:21:02.562180 8290 sgd_solver.cpp:105] Iteration 7932, lr = 0.00207792 I0410 00:21:07.747567 8290 solver.cpp:218] Iteration 7944 (2.31428 iter/s, 5.18521s/12 iters), loss = 0.246522 I0410 00:21:07.747622 8290 solver.cpp:237] Train net output #0: loss = 0.246522 (* 1 = 0.246522 loss) I0410 00:21:07.747635 8290 sgd_solver.cpp:105] Iteration 7944, lr = 0.00207299 I0410 00:21:12.144661 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_7956.caffemodel I0410 00:21:12.870913 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_7956.solverstate I0410 00:21:13.401293 8290 solver.cpp:330] Iteration 7956, Testing net (#0) I0410 00:21:13.401319 8290 net.cpp:676] Ignoring source layer train-data I0410 00:21:14.756047 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:21:17.852914 8290 solver.cpp:397] Test net output #0: accuracy = 0.564951 I0410 00:21:17.852964 8290 solver.cpp:397] Test net output #1: loss = 2.19199 (* 1 = 2.19199 loss) I0410 00:21:17.934376 8290 solver.cpp:218] Iteration 7956 (1.17804 iter/s, 10.1864s/12 iters), loss = 0.187908 I0410 00:21:17.934425 8290 solver.cpp:237] Train net output #0: loss = 0.187908 (* 1 = 0.187908 loss) I0410 00:21:17.934438 8290 sgd_solver.cpp:105] Iteration 7956, lr = 0.00206807 I0410 00:21:22.093056 8290 solver.cpp:218] Iteration 7968 (2.88567 iter/s, 4.15847s/12 iters), loss = 0.222756 I0410 00:21:22.093106 8290 solver.cpp:237] Train net output #0: loss = 0.222756 (* 1 = 0.222756 loss) I0410 00:21:22.093117 8290 sgd_solver.cpp:105] Iteration 7968, lr = 0.00206316 I0410 00:21:26.966959 8290 solver.cpp:218] Iteration 7980 (2.46221 iter/s, 4.87367s/12 iters), loss = 0.368423 I0410 00:21:26.972957 8290 solver.cpp:237] Train net output #0: loss = 0.368423 (* 1 = 0.368423 loss) I0410 00:21:26.972971 8290 sgd_solver.cpp:105] Iteration 7980, lr = 0.00205826 I0410 00:21:31.198251 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:21:31.908491 8290 solver.cpp:218] Iteration 7992 (2.43144 iter/s, 4.93536s/12 iters), loss = 0.202405 I0410 00:21:31.908548 8290 solver.cpp:237] Train net output #0: loss = 0.202405 (* 1 = 0.202405 loss) I0410 00:21:31.908561 8290 sgd_solver.cpp:105] Iteration 7992, lr = 0.00205337 I0410 00:21:36.791064 8290 solver.cpp:218] Iteration 8004 (2.45784 iter/s, 4.88233s/12 iters), loss = 0.315508 I0410 00:21:36.791115 8290 solver.cpp:237] Train net output #0: loss = 0.315508 (* 1 = 0.315508 loss) I0410 00:21:36.791126 8290 sgd_solver.cpp:105] Iteration 8004, lr = 0.0020485 I0410 00:21:41.999848 8290 solver.cpp:218] Iteration 8016 (2.30391 iter/s, 5.20854s/12 iters), loss = 0.262591 I0410 00:21:41.999893 8290 solver.cpp:237] Train net output #0: loss = 0.262591 (* 1 = 0.262591 loss) I0410 00:21:41.999903 8290 sgd_solver.cpp:105] Iteration 8016, lr = 0.00204363 I0410 00:21:46.901751 8290 solver.cpp:218] Iteration 8028 (2.44815 iter/s, 4.90167s/12 iters), loss = 0.213203 I0410 00:21:46.901808 8290 solver.cpp:237] Train net output #0: loss = 0.213203 (* 1 = 0.213203 loss) I0410 00:21:46.901821 8290 sgd_solver.cpp:105] Iteration 8028, lr = 0.00203878 I0410 00:21:51.784634 8290 solver.cpp:218] Iteration 8040 (2.45768 iter/s, 4.88265s/12 iters), loss = 0.302689 I0410 00:21:51.784682 8290 solver.cpp:237] Train net output #0: loss = 0.302689 (* 1 = 0.302689 loss) I0410 00:21:51.784693 8290 sgd_solver.cpp:105] Iteration 8040, lr = 0.00203394 I0410 00:21:56.659981 8290 solver.cpp:218] Iteration 8052 (2.46148 iter/s, 4.87511s/12 iters), loss = 0.272849 I0410 00:21:56.660037 8290 solver.cpp:237] Train net output #0: loss = 0.272849 (* 1 = 0.272849 loss) I0410 00:21:56.660049 8290 sgd_solver.cpp:105] Iteration 8052, lr = 0.00202911 I0410 00:21:58.672499 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8058.caffemodel I0410 00:22:01.041780 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8058.solverstate I0410 00:22:01.803067 8290 solver.cpp:330] Iteration 8058, Testing net (#0) I0410 00:22:01.803090 8290 net.cpp:676] Ignoring source layer train-data I0410 00:22:03.233884 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:22:06.460572 8290 solver.cpp:397] Test net output #0: accuracy = 0.571691 I0410 00:22:06.460613 8290 solver.cpp:397] Test net output #1: loss = 2.17252 (* 1 = 2.17252 loss) I0410 00:22:08.211179 8290 solver.cpp:218] Iteration 8064 (1.0389 iter/s, 11.5507s/12 iters), loss = 0.264658 I0410 00:22:08.211241 8290 solver.cpp:237] Train net output #0: loss = 0.264658 (* 1 = 0.264658 loss) I0410 00:22:08.211252 8290 sgd_solver.cpp:105] Iteration 8064, lr = 0.00202429 I0410 00:22:13.012990 8290 solver.cpp:218] Iteration 8076 (2.49918 iter/s, 4.80157s/12 iters), loss = 0.143362 I0410 00:22:13.013052 8290 solver.cpp:237] Train net output #0: loss = 0.143362 (* 1 = 0.143362 loss) I0410 00:22:13.013064 8290 sgd_solver.cpp:105] Iteration 8076, lr = 0.00201949 I0410 00:22:17.828493 8290 solver.cpp:218] Iteration 8088 (2.49208 iter/s, 4.81526s/12 iters), loss = 0.115795 I0410 00:22:17.828537 8290 solver.cpp:237] Train net output #0: loss = 0.115795 (* 1 = 0.115795 loss) I0410 00:22:17.828547 8290 sgd_solver.cpp:105] Iteration 8088, lr = 0.00201469 I0410 00:22:19.265069 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:22:22.720925 8290 solver.cpp:218] Iteration 8100 (2.45289 iter/s, 4.8922s/12 iters), loss = 0.150658 I0410 00:22:22.720983 8290 solver.cpp:237] Train net output #0: loss = 0.150658 (* 1 = 0.150658 loss) I0410 00:22:22.720996 8290 sgd_solver.cpp:105] Iteration 8100, lr = 0.00200991 I0410 00:22:27.602357 8290 solver.cpp:218] Iteration 8112 (2.45842 iter/s, 4.88119s/12 iters), loss = 0.159067 I0410 00:22:27.602416 8290 solver.cpp:237] Train net output #0: loss = 0.159067 (* 1 = 0.159067 loss) I0410 00:22:27.602428 8290 sgd_solver.cpp:105] Iteration 8112, lr = 0.00200514 I0410 00:22:32.587342 8290 solver.cpp:218] Iteration 8124 (2.40735 iter/s, 4.98474s/12 iters), loss = 0.193695 I0410 00:22:32.587440 8290 solver.cpp:237] Train net output #0: loss = 0.193695 (* 1 = 0.193695 loss) I0410 00:22:32.587451 8290 sgd_solver.cpp:105] Iteration 8124, lr = 0.00200038 I0410 00:22:37.461678 8290 solver.cpp:218] Iteration 8136 (2.46202 iter/s, 4.87405s/12 iters), loss = 0.168223 I0410 00:22:37.461730 8290 solver.cpp:237] Train net output #0: loss = 0.168223 (* 1 = 0.168223 loss) I0410 00:22:37.461740 8290 sgd_solver.cpp:105] Iteration 8136, lr = 0.00199563 I0410 00:22:42.396459 8290 solver.cpp:218] Iteration 8148 (2.43183 iter/s, 4.93455s/12 iters), loss = 0.2309 I0410 00:22:42.396507 8290 solver.cpp:237] Train net output #0: loss = 0.2309 (* 1 = 0.2309 loss) I0410 00:22:42.396514 8290 sgd_solver.cpp:105] Iteration 8148, lr = 0.00199089 I0410 00:22:46.853055 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8160.caffemodel I0410 00:22:48.134265 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8160.solverstate I0410 00:22:49.223165 8290 solver.cpp:330] Iteration 8160, Testing net (#0) I0410 00:22:49.223191 8290 net.cpp:676] Ignoring source layer train-data I0410 00:22:50.479188 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:22:53.694967 8290 solver.cpp:397] Test net output #0: accuracy = 0.596814 I0410 00:22:53.694996 8290 solver.cpp:397] Test net output #1: loss = 2.13111 (* 1 = 2.13111 loss) I0410 00:22:53.777209 8290 solver.cpp:218] Iteration 8160 (1.05445 iter/s, 11.3803s/12 iters), loss = 0.254937 I0410 00:22:53.777276 8290 solver.cpp:237] Train net output #0: loss = 0.254937 (* 1 = 0.254937 loss) I0410 00:22:53.777288 8290 sgd_solver.cpp:105] Iteration 8160, lr = 0.00198616 I0410 00:22:57.938599 8290 solver.cpp:218] Iteration 8172 (2.88381 iter/s, 4.16116s/12 iters), loss = 0.274137 I0410 00:22:57.938655 8290 solver.cpp:237] Train net output #0: loss = 0.274137 (* 1 = 0.274137 loss) I0410 00:22:57.938666 8290 sgd_solver.cpp:105] Iteration 8172, lr = 0.00198145 I0410 00:23:02.835796 8290 solver.cpp:218] Iteration 8184 (2.4505 iter/s, 4.89696s/12 iters), loss = 0.316743 I0410 00:23:02.835903 8290 solver.cpp:237] Train net output #0: loss = 0.316743 (* 1 = 0.316743 loss) I0410 00:23:02.835912 8290 sgd_solver.cpp:105] Iteration 8184, lr = 0.00197674 I0410 00:23:06.350543 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:23:07.781509 8290 solver.cpp:218] Iteration 8196 (2.42649 iter/s, 4.94542s/12 iters), loss = 0.218738 I0410 00:23:07.781558 8290 solver.cpp:237] Train net output #0: loss = 0.218738 (* 1 = 0.218738 loss) I0410 00:23:07.781569 8290 sgd_solver.cpp:105] Iteration 8196, lr = 0.00197205 I0410 00:23:12.667675 8290 solver.cpp:218] Iteration 8208 (2.45603 iter/s, 4.88593s/12 iters), loss = 0.309058 I0410 00:23:12.667723 8290 solver.cpp:237] Train net output #0: loss = 0.309058 (* 1 = 0.309058 loss) I0410 00:23:12.667733 8290 sgd_solver.cpp:105] Iteration 8208, lr = 0.00196737 I0410 00:23:17.597995 8290 solver.cpp:218] Iteration 8220 (2.43404 iter/s, 4.93008s/12 iters), loss = 0.286609 I0410 00:23:17.598053 8290 solver.cpp:237] Train net output #0: loss = 0.286609 (* 1 = 0.286609 loss) I0410 00:23:17.598067 8290 sgd_solver.cpp:105] Iteration 8220, lr = 0.0019627 I0410 00:23:22.435389 8290 solver.cpp:218] Iteration 8232 (2.4808 iter/s, 4.83715s/12 iters), loss = 0.157205 I0410 00:23:22.435442 8290 solver.cpp:237] Train net output #0: loss = 0.157205 (* 1 = 0.157205 loss) I0410 00:23:22.435454 8290 sgd_solver.cpp:105] Iteration 8232, lr = 0.00195804 I0410 00:23:27.275046 8290 solver.cpp:218] Iteration 8244 (2.47964 iter/s, 4.83942s/12 iters), loss = 0.348536 I0410 00:23:27.275104 8290 solver.cpp:237] Train net output #0: loss = 0.348536 (* 1 = 0.348536 loss) I0410 00:23:27.275116 8290 sgd_solver.cpp:105] Iteration 8244, lr = 0.00195339 I0410 00:23:32.108201 8290 solver.cpp:218] Iteration 8256 (2.48297 iter/s, 4.83292s/12 iters), loss = 0.184274 I0410 00:23:32.108244 8290 solver.cpp:237] Train net output #0: loss = 0.184274 (* 1 = 0.184274 loss) I0410 00:23:32.108254 8290 sgd_solver.cpp:105] Iteration 8256, lr = 0.00194875 I0410 00:23:34.116806 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8262.caffemodel I0410 00:23:34.846777 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8262.solverstate I0410 00:23:35.364845 8290 solver.cpp:330] Iteration 8262, Testing net (#0) I0410 00:23:35.364868 8290 net.cpp:676] Ignoring source layer train-data I0410 00:23:36.578696 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:23:39.872459 8290 solver.cpp:397] Test net output #0: accuracy = 0.581495 I0410 00:23:39.872510 8290 solver.cpp:397] Test net output #1: loss = 2.10737 (* 1 = 2.10737 loss) I0410 00:23:41.796787 8290 solver.cpp:218] Iteration 8268 (1.23862 iter/s, 9.68819s/12 iters), loss = 0.259163 I0410 00:23:41.796829 8290 solver.cpp:237] Train net output #0: loss = 0.259163 (* 1 = 0.259163 loss) I0410 00:23:41.796838 8290 sgd_solver.cpp:105] Iteration 8268, lr = 0.00194412 I0410 00:23:46.663822 8290 solver.cpp:218] Iteration 8280 (2.46568 iter/s, 4.86681s/12 iters), loss = 0.163577 I0410 00:23:46.663877 8290 solver.cpp:237] Train net output #0: loss = 0.163577 (* 1 = 0.163577 loss) I0410 00:23:46.663888 8290 sgd_solver.cpp:105] Iteration 8280, lr = 0.00193951 I0410 00:23:51.531723 8290 solver.cpp:218] Iteration 8292 (2.46525 iter/s, 4.86766s/12 iters), loss = 0.152952 I0410 00:23:51.531776 8290 solver.cpp:237] Train net output #0: loss = 0.152952 (* 1 = 0.152952 loss) I0410 00:23:51.531788 8290 sgd_solver.cpp:105] Iteration 8292, lr = 0.0019349 I0410 00:23:52.218973 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:23:56.427489 8290 solver.cpp:218] Iteration 8304 (2.45122 iter/s, 4.89552s/12 iters), loss = 0.194439 I0410 00:23:56.427542 8290 solver.cpp:237] Train net output #0: loss = 0.194439 (* 1 = 0.194439 loss) I0410 00:23:56.427552 8290 sgd_solver.cpp:105] Iteration 8304, lr = 0.00193031 I0410 00:23:57.997330 8290 blocking_queue.cpp:49] Waiting for data I0410 00:24:01.274127 8290 solver.cpp:218] Iteration 8316 (2.47606 iter/s, 4.8464s/12 iters), loss = 0.314609 I0410 00:24:01.274173 8290 solver.cpp:237] Train net output #0: loss = 0.314609 (* 1 = 0.314609 loss) I0410 00:24:01.274183 8290 sgd_solver.cpp:105] Iteration 8316, lr = 0.00192573 I0410 00:24:06.123714 8290 solver.cpp:218] Iteration 8328 (2.47456 iter/s, 4.84935s/12 iters), loss = 0.16997 I0410 00:24:06.123859 8290 solver.cpp:237] Train net output #0: loss = 0.16997 (* 1 = 0.16997 loss) I0410 00:24:06.123872 8290 sgd_solver.cpp:105] Iteration 8328, lr = 0.00192115 I0410 00:24:11.015341 8290 solver.cpp:218] Iteration 8340 (2.45333 iter/s, 4.89131s/12 iters), loss = 0.300923 I0410 00:24:11.015380 8290 solver.cpp:237] Train net output #0: loss = 0.300923 (* 1 = 0.300923 loss) I0410 00:24:11.015388 8290 sgd_solver.cpp:105] Iteration 8340, lr = 0.00191659 I0410 00:24:15.902874 8290 solver.cpp:218] Iteration 8352 (2.45534 iter/s, 4.8873s/12 iters), loss = 0.266017 I0410 00:24:15.902936 8290 solver.cpp:237] Train net output #0: loss = 0.266017 (* 1 = 0.266017 loss) I0410 00:24:15.902948 8290 sgd_solver.cpp:105] Iteration 8352, lr = 0.00191204 I0410 00:24:20.435041 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8364.caffemodel I0410 00:24:23.751924 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8364.solverstate I0410 00:24:24.307135 8290 solver.cpp:330] Iteration 8364, Testing net (#0) I0410 00:24:24.307161 8290 net.cpp:676] Ignoring source layer train-data I0410 00:24:25.471869 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:24:28.807790 8290 solver.cpp:397] Test net output #0: accuracy = 0.583946 I0410 00:24:28.807840 8290 solver.cpp:397] Test net output #1: loss = 2.11828 (* 1 = 2.11828 loss) I0410 00:24:28.890017 8290 solver.cpp:218] Iteration 8364 (0.924028 iter/s, 12.9866s/12 iters), loss = 0.100368 I0410 00:24:28.890064 8290 solver.cpp:237] Train net output #0: loss = 0.100368 (* 1 = 0.100368 loss) I0410 00:24:28.890076 8290 sgd_solver.cpp:105] Iteration 8364, lr = 0.0019075 I0410 00:24:32.929672 8290 solver.cpp:218] Iteration 8376 (2.9707 iter/s, 4.03945s/12 iters), loss = 0.175626 I0410 00:24:32.929731 8290 solver.cpp:237] Train net output #0: loss = 0.175626 (* 1 = 0.175626 loss) I0410 00:24:32.929744 8290 sgd_solver.cpp:105] Iteration 8376, lr = 0.00190297 I0410 00:24:37.843675 8290 solver.cpp:218] Iteration 8388 (2.44212 iter/s, 4.91376s/12 iters), loss = 0.185903 I0410 00:24:37.843838 8290 solver.cpp:237] Train net output #0: loss = 0.185903 (* 1 = 0.185903 loss) I0410 00:24:37.843853 8290 sgd_solver.cpp:105] Iteration 8388, lr = 0.00189846 I0410 00:24:40.596304 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:24:42.718523 8290 solver.cpp:218] Iteration 8400 (2.46179 iter/s, 4.8745s/12 iters), loss = 0.20911 I0410 00:24:42.718571 8290 solver.cpp:237] Train net output #0: loss = 0.20911 (* 1 = 0.20911 loss) I0410 00:24:42.718580 8290 sgd_solver.cpp:105] Iteration 8400, lr = 0.00189395 I0410 00:24:47.679744 8290 solver.cpp:218] Iteration 8412 (2.41887 iter/s, 4.96099s/12 iters), loss = 0.245319 I0410 00:24:47.679792 8290 solver.cpp:237] Train net output #0: loss = 0.245319 (* 1 = 0.245319 loss) I0410 00:24:47.679802 8290 sgd_solver.cpp:105] Iteration 8412, lr = 0.00188945 I0410 00:24:52.579453 8290 solver.cpp:218] Iteration 8424 (2.44925 iter/s, 4.89947s/12 iters), loss = 0.130997 I0410 00:24:52.579515 8290 solver.cpp:237] Train net output #0: loss = 0.130997 (* 1 = 0.130997 loss) I0410 00:24:52.579528 8290 sgd_solver.cpp:105] Iteration 8424, lr = 0.00188497 I0410 00:24:57.426086 8290 solver.cpp:218] Iteration 8436 (2.47607 iter/s, 4.84639s/12 iters), loss = 0.179863 I0410 00:24:57.426128 8290 solver.cpp:237] Train net output #0: loss = 0.179863 (* 1 = 0.179863 loss) I0410 00:24:57.426136 8290 sgd_solver.cpp:105] Iteration 8436, lr = 0.00188049 I0410 00:25:02.242312 8290 solver.cpp:218] Iteration 8448 (2.49169 iter/s, 4.816s/12 iters), loss = 0.230896 I0410 00:25:02.242367 8290 solver.cpp:237] Train net output #0: loss = 0.230896 (* 1 = 0.230896 loss) I0410 00:25:02.242380 8290 sgd_solver.cpp:105] Iteration 8448, lr = 0.00187603 I0410 00:25:07.144034 8290 solver.cpp:218] Iteration 8460 (2.44824 iter/s, 4.90148s/12 iters), loss = 0.304391 I0410 00:25:07.144093 8290 solver.cpp:237] Train net output #0: loss = 0.304391 (* 1 = 0.304391 loss) I0410 00:25:07.144105 8290 sgd_solver.cpp:105] Iteration 8460, lr = 0.00187157 I0410 00:25:09.138064 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8466.caffemodel I0410 00:25:11.049363 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8466.solverstate I0410 00:25:12.191504 8290 solver.cpp:330] Iteration 8466, Testing net (#0) I0410 00:25:12.191536 8290 net.cpp:676] Ignoring source layer train-data I0410 00:25:13.339028 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:25:16.856205 8290 solver.cpp:397] Test net output #0: accuracy = 0.573529 I0410 00:25:16.856256 8290 solver.cpp:397] Test net output #1: loss = 2.11242 (* 1 = 2.11242 loss) I0410 00:25:18.618288 8290 solver.cpp:218] Iteration 8472 (1.04586 iter/s, 11.4738s/12 iters), loss = 0.294388 I0410 00:25:18.618345 8290 solver.cpp:237] Train net output #0: loss = 0.294388 (* 1 = 0.294388 loss) I0410 00:25:18.618355 8290 sgd_solver.cpp:105] Iteration 8472, lr = 0.00186713 I0410 00:25:23.516219 8290 solver.cpp:218] Iteration 8484 (2.45013 iter/s, 4.89769s/12 iters), loss = 0.292131 I0410 00:25:23.516264 8290 solver.cpp:237] Train net output #0: loss = 0.292131 (* 1 = 0.292131 loss) I0410 00:25:23.516273 8290 sgd_solver.cpp:105] Iteration 8484, lr = 0.0018627 I0410 00:25:28.409649 8290 solver.cpp:218] Iteration 8496 (2.45238 iter/s, 4.8932s/12 iters), loss = 0.219757 I0410 00:25:28.409698 8290 solver.cpp:237] Train net output #0: loss = 0.219757 (* 1 = 0.219757 loss) I0410 00:25:28.409708 8290 sgd_solver.cpp:105] Iteration 8496, lr = 0.00185827 I0410 00:25:28.456806 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:25:33.285233 8290 solver.cpp:218] Iteration 8508 (2.46137 iter/s, 4.87534s/12 iters), loss = 0.221145 I0410 00:25:33.285290 8290 solver.cpp:237] Train net output #0: loss = 0.221145 (* 1 = 0.221145 loss) I0410 00:25:33.285300 8290 sgd_solver.cpp:105] Iteration 8508, lr = 0.00185386 I0410 00:25:38.076993 8290 solver.cpp:218] Iteration 8520 (2.50442 iter/s, 4.79152s/12 iters), loss = 0.184015 I0410 00:25:38.077042 8290 solver.cpp:237] Train net output #0: loss = 0.184015 (* 1 = 0.184015 loss) I0410 00:25:38.077051 8290 sgd_solver.cpp:105] Iteration 8520, lr = 0.00184946 I0410 00:25:42.970315 8290 solver.cpp:218] Iteration 8532 (2.45244 iter/s, 4.89308s/12 iters), loss = 0.156644 I0410 00:25:42.970486 8290 solver.cpp:237] Train net output #0: loss = 0.156644 (* 1 = 0.156644 loss) I0410 00:25:42.970499 8290 sgd_solver.cpp:105] Iteration 8532, lr = 0.00184507 I0410 00:25:47.900785 8290 solver.cpp:218] Iteration 8544 (2.43402 iter/s, 4.93012s/12 iters), loss = 0.238158 I0410 00:25:47.900837 8290 solver.cpp:237] Train net output #0: loss = 0.238158 (* 1 = 0.238158 loss) I0410 00:25:47.900849 8290 sgd_solver.cpp:105] Iteration 8544, lr = 0.00184069 I0410 00:25:52.803990 8290 solver.cpp:218] Iteration 8556 (2.4475 iter/s, 4.90297s/12 iters), loss = 0.194764 I0410 00:25:52.804040 8290 solver.cpp:237] Train net output #0: loss = 0.194764 (* 1 = 0.194764 loss) I0410 00:25:52.804051 8290 sgd_solver.cpp:105] Iteration 8556, lr = 0.00183632 I0410 00:25:57.242966 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8568.caffemodel I0410 00:25:58.003305 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8568.solverstate I0410 00:25:58.544335 8290 solver.cpp:330] Iteration 8568, Testing net (#0) I0410 00:25:58.544363 8290 net.cpp:676] Ignoring source layer train-data I0410 00:25:59.649576 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:26:03.136148 8290 solver.cpp:397] Test net output #0: accuracy = 0.582721 I0410 00:26:03.136198 8290 solver.cpp:397] Test net output #1: loss = 2.1201 (* 1 = 2.1201 loss) I0410 00:26:03.218354 8290 solver.cpp:218] Iteration 8568 (1.1523 iter/s, 10.4139s/12 iters), loss = 0.167068 I0410 00:26:03.218402 8290 solver.cpp:237] Train net output #0: loss = 0.167068 (* 1 = 0.167068 loss) I0410 00:26:03.218413 8290 sgd_solver.cpp:105] Iteration 8568, lr = 0.00183196 I0410 00:26:07.367878 8290 solver.cpp:218] Iteration 8580 (2.89204 iter/s, 4.14932s/12 iters), loss = 0.178444 I0410 00:26:07.367930 8290 solver.cpp:237] Train net output #0: loss = 0.178444 (* 1 = 0.178444 loss) I0410 00:26:07.367942 8290 sgd_solver.cpp:105] Iteration 8580, lr = 0.00182761 I0410 00:26:12.257040 8290 solver.cpp:218] Iteration 8592 (2.45452 iter/s, 4.88893s/12 iters), loss = 0.14112 I0410 00:26:12.257086 8290 solver.cpp:237] Train net output #0: loss = 0.14112 (* 1 = 0.14112 loss) I0410 00:26:12.257098 8290 sgd_solver.cpp:105] Iteration 8592, lr = 0.00182327 I0410 00:26:14.381743 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:26:17.164995 8290 solver.cpp:218] Iteration 8604 (2.44512 iter/s, 4.90773s/12 iters), loss = 0.184819 I0410 00:26:17.165048 8290 solver.cpp:237] Train net output #0: loss = 0.184819 (* 1 = 0.184819 loss) I0410 00:26:17.165060 8290 sgd_solver.cpp:105] Iteration 8604, lr = 0.00181894 I0410 00:26:22.057543 8290 solver.cpp:218] Iteration 8616 (2.45283 iter/s, 4.89231s/12 iters), loss = 0.103497 I0410 00:26:22.057597 8290 solver.cpp:237] Train net output #0: loss = 0.103497 (* 1 = 0.103497 loss) I0410 00:26:22.057610 8290 sgd_solver.cpp:105] Iteration 8616, lr = 0.00181462 I0410 00:26:26.935436 8290 solver.cpp:218] Iteration 8628 (2.4602 iter/s, 4.87766s/12 iters), loss = 0.223018 I0410 00:26:26.935489 8290 solver.cpp:237] Train net output #0: loss = 0.223018 (* 1 = 0.223018 loss) I0410 00:26:26.935503 8290 sgd_solver.cpp:105] Iteration 8628, lr = 0.00181031 I0410 00:26:31.848541 8290 solver.cpp:218] Iteration 8640 (2.44257 iter/s, 4.91287s/12 iters), loss = 0.189599 I0410 00:26:31.848592 8290 solver.cpp:237] Train net output #0: loss = 0.189599 (* 1 = 0.189599 loss) I0410 00:26:31.848603 8290 sgd_solver.cpp:105] Iteration 8640, lr = 0.00180602 I0410 00:26:36.673072 8290 solver.cpp:218] Iteration 8652 (2.48741 iter/s, 4.8243s/12 iters), loss = 0.232721 I0410 00:26:36.673128 8290 solver.cpp:237] Train net output #0: loss = 0.232721 (* 1 = 0.232721 loss) I0410 00:26:36.673139 8290 sgd_solver.cpp:105] Iteration 8652, lr = 0.00180173 I0410 00:26:41.556663 8290 solver.cpp:218] Iteration 8664 (2.45733 iter/s, 4.88336s/12 iters), loss = 0.230148 I0410 00:26:41.556717 8290 solver.cpp:237] Train net output #0: loss = 0.230148 (* 1 = 0.230148 loss) I0410 00:26:41.556728 8290 sgd_solver.cpp:105] Iteration 8664, lr = 0.00179745 I0410 00:26:43.550441 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8670.caffemodel I0410 00:26:44.318194 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8670.solverstate I0410 00:26:44.968897 8290 solver.cpp:330] Iteration 8670, Testing net (#0) I0410 00:26:44.969041 8290 net.cpp:676] Ignoring source layer train-data I0410 00:26:45.935660 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:26:49.428129 8290 solver.cpp:397] Test net output #0: accuracy = 0.58701 I0410 00:26:49.428179 8290 solver.cpp:397] Test net output #1: loss = 2.10335 (* 1 = 2.10335 loss) I0410 00:26:51.254488 8290 solver.cpp:218] Iteration 8676 (1.23744 iter/s, 9.69742s/12 iters), loss = 0.201568 I0410 00:26:51.254545 8290 solver.cpp:237] Train net output #0: loss = 0.201568 (* 1 = 0.201568 loss) I0410 00:26:51.254559 8290 sgd_solver.cpp:105] Iteration 8676, lr = 0.00179318 I0410 00:26:56.131996 8290 solver.cpp:218] Iteration 8688 (2.46039 iter/s, 4.87727s/12 iters), loss = 0.135361 I0410 00:26:56.132051 8290 solver.cpp:237] Train net output #0: loss = 0.135361 (* 1 = 0.135361 loss) I0410 00:26:56.132063 8290 sgd_solver.cpp:105] Iteration 8688, lr = 0.00178893 I0410 00:27:00.360118 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:27:01.039868 8290 solver.cpp:218] Iteration 8700 (2.44517 iter/s, 4.90764s/12 iters), loss = 0.143948 I0410 00:27:01.039916 8290 solver.cpp:237] Train net output #0: loss = 0.143948 (* 1 = 0.143948 loss) I0410 00:27:01.039927 8290 sgd_solver.cpp:105] Iteration 8700, lr = 0.00178468 I0410 00:27:06.219358 8290 solver.cpp:218] Iteration 8712 (2.31694 iter/s, 5.17925s/12 iters), loss = 0.197376 I0410 00:27:06.219406 8290 solver.cpp:237] Train net output #0: loss = 0.197376 (* 1 = 0.197376 loss) I0410 00:27:06.219416 8290 sgd_solver.cpp:105] Iteration 8712, lr = 0.00178044 I0410 00:27:11.084776 8290 solver.cpp:218] Iteration 8724 (2.46651 iter/s, 4.86518s/12 iters), loss = 0.199831 I0410 00:27:11.084841 8290 solver.cpp:237] Train net output #0: loss = 0.199831 (* 1 = 0.199831 loss) I0410 00:27:11.084853 8290 sgd_solver.cpp:105] Iteration 8724, lr = 0.00177621 I0410 00:27:16.008332 8290 solver.cpp:218] Iteration 8736 (2.43738 iter/s, 4.92331s/12 iters), loss = 0.213477 I0410 00:27:16.008425 8290 solver.cpp:237] Train net output #0: loss = 0.213477 (* 1 = 0.213477 loss) I0410 00:27:16.008437 8290 sgd_solver.cpp:105] Iteration 8736, lr = 0.001772 I0410 00:27:20.955869 8290 solver.cpp:218] Iteration 8748 (2.42559 iter/s, 4.94726s/12 iters), loss = 0.274857 I0410 00:27:20.955925 8290 solver.cpp:237] Train net output #0: loss = 0.274857 (* 1 = 0.274857 loss) I0410 00:27:20.955938 8290 sgd_solver.cpp:105] Iteration 8748, lr = 0.00176779 I0410 00:27:25.833679 8290 solver.cpp:218] Iteration 8760 (2.46024 iter/s, 4.87757s/12 iters), loss = 0.116077 I0410 00:27:25.833730 8290 solver.cpp:237] Train net output #0: loss = 0.116077 (* 1 = 0.116077 loss) I0410 00:27:25.833743 8290 sgd_solver.cpp:105] Iteration 8760, lr = 0.00176359 I0410 00:27:30.341877 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8772.caffemodel I0410 00:27:31.100474 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8772.solverstate I0410 00:27:31.637431 8290 solver.cpp:330] Iteration 8772, Testing net (#0) I0410 00:27:31.637454 8290 net.cpp:676] Ignoring source layer train-data I0410 00:27:32.616006 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:27:36.036389 8290 solver.cpp:397] Test net output #0: accuracy = 0.582108 I0410 00:27:36.036440 8290 solver.cpp:397] Test net output #1: loss = 2.10336 (* 1 = 2.10336 loss) I0410 00:27:36.118360 8290 solver.cpp:218] Iteration 8772 (1.16683 iter/s, 10.2842s/12 iters), loss = 0.261093 I0410 00:27:36.118418 8290 solver.cpp:237] Train net output #0: loss = 0.261093 (* 1 = 0.261093 loss) I0410 00:27:36.118432 8290 sgd_solver.cpp:105] Iteration 8772, lr = 0.00175941 I0410 00:27:40.341454 8290 solver.cpp:218] Iteration 8784 (2.84166 iter/s, 4.22288s/12 iters), loss = 0.107797 I0410 00:27:40.341504 8290 solver.cpp:237] Train net output #0: loss = 0.107797 (* 1 = 0.107797 loss) I0410 00:27:40.341514 8290 sgd_solver.cpp:105] Iteration 8784, lr = 0.00175523 I0410 00:27:45.263326 8290 solver.cpp:218] Iteration 8796 (2.43821 iter/s, 4.92164s/12 iters), loss = 0.307385 I0410 00:27:45.263373 8290 solver.cpp:237] Train net output #0: loss = 0.307385 (* 1 = 0.307385 loss) I0410 00:27:45.263382 8290 sgd_solver.cpp:105] Iteration 8796, lr = 0.00175106 I0410 00:27:46.690704 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:27:50.307494 8290 solver.cpp:218] Iteration 8808 (2.3791 iter/s, 5.04393s/12 iters), loss = 0.265095 I0410 00:27:50.307545 8290 solver.cpp:237] Train net output #0: loss = 0.265095 (* 1 = 0.265095 loss) I0410 00:27:50.307557 8290 sgd_solver.cpp:105] Iteration 8808, lr = 0.0017469 I0410 00:27:55.187048 8290 solver.cpp:218] Iteration 8820 (2.45936 iter/s, 4.87931s/12 iters), loss = 0.280195 I0410 00:27:55.187108 8290 solver.cpp:237] Train net output #0: loss = 0.280195 (* 1 = 0.280195 loss) I0410 00:27:55.187119 8290 sgd_solver.cpp:105] Iteration 8820, lr = 0.00174276 I0410 00:28:00.010776 8290 solver.cpp:218] Iteration 8832 (2.48783 iter/s, 4.82349s/12 iters), loss = 0.213436 I0410 00:28:00.010829 8290 solver.cpp:237] Train net output #0: loss = 0.213436 (* 1 = 0.213436 loss) I0410 00:28:00.010840 8290 sgd_solver.cpp:105] Iteration 8832, lr = 0.00173862 I0410 00:28:04.929276 8290 solver.cpp:218] Iteration 8844 (2.43989 iter/s, 4.91826s/12 iters), loss = 0.2066 I0410 00:28:04.929335 8290 solver.cpp:237] Train net output #0: loss = 0.2066 (* 1 = 0.2066 loss) I0410 00:28:04.929348 8290 sgd_solver.cpp:105] Iteration 8844, lr = 0.00173449 I0410 00:28:09.858299 8290 solver.cpp:218] Iteration 8856 (2.43468 iter/s, 4.92877s/12 iters), loss = 0.0969012 I0410 00:28:09.858355 8290 solver.cpp:237] Train net output #0: loss = 0.0969012 (* 1 = 0.0969012 loss) I0410 00:28:09.858368 8290 sgd_solver.cpp:105] Iteration 8856, lr = 0.00173037 I0410 00:28:14.789523 8290 solver.cpp:218] Iteration 8868 (2.43359 iter/s, 4.93099s/12 iters), loss = 0.155928 I0410 00:28:14.789567 8290 solver.cpp:237] Train net output #0: loss = 0.155928 (* 1 = 0.155928 loss) I0410 00:28:14.789577 8290 sgd_solver.cpp:105] Iteration 8868, lr = 0.00172626 I0410 00:28:16.801411 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8874.caffemodel I0410 00:28:18.113499 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8874.solverstate I0410 00:28:19.305073 8290 solver.cpp:330] Iteration 8874, Testing net (#0) I0410 00:28:19.305104 8290 net.cpp:676] Ignoring source layer train-data I0410 00:28:20.297330 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:28:23.801442 8290 solver.cpp:397] Test net output #0: accuracy = 0.596814 I0410 00:28:23.801488 8290 solver.cpp:397] Test net output #1: loss = 2.10785 (* 1 = 2.10785 loss) I0410 00:28:25.646703 8290 solver.cpp:218] Iteration 8880 (1.1053 iter/s, 10.8567s/12 iters), loss = 0.265009 I0410 00:28:25.646760 8290 solver.cpp:237] Train net output #0: loss = 0.265009 (* 1 = 0.265009 loss) I0410 00:28:25.646771 8290 sgd_solver.cpp:105] Iteration 8880, lr = 0.00172217 I0410 00:28:30.789779 8290 solver.cpp:218] Iteration 8892 (2.33335 iter/s, 5.14282s/12 iters), loss = 0.23544 I0410 00:28:30.789836 8290 solver.cpp:237] Train net output #0: loss = 0.23544 (* 1 = 0.23544 loss) I0410 00:28:30.789850 8290 sgd_solver.cpp:105] Iteration 8892, lr = 0.00171808 I0410 00:28:34.225215 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:28:35.622805 8290 solver.cpp:218] Iteration 8904 (2.48304 iter/s, 4.83279s/12 iters), loss = 0.190199 I0410 00:28:35.622859 8290 solver.cpp:237] Train net output #0: loss = 0.190199 (* 1 = 0.190199 loss) I0410 00:28:35.622871 8290 sgd_solver.cpp:105] Iteration 8904, lr = 0.001714 I0410 00:28:40.723546 8290 solver.cpp:218] Iteration 8916 (2.35271 iter/s, 5.1005s/12 iters), loss = 0.228277 I0410 00:28:40.723594 8290 solver.cpp:237] Train net output #0: loss = 0.228277 (* 1 = 0.228277 loss) I0410 00:28:40.723605 8290 sgd_solver.cpp:105] Iteration 8916, lr = 0.00170993 I0410 00:28:45.712471 8290 solver.cpp:218] Iteration 8928 (2.40544 iter/s, 4.98869s/12 iters), loss = 0.170864 I0410 00:28:45.712522 8290 solver.cpp:237] Train net output #0: loss = 0.170864 (* 1 = 0.170864 loss) I0410 00:28:45.712535 8290 sgd_solver.cpp:105] Iteration 8928, lr = 0.00170587 I0410 00:28:50.581276 8290 solver.cpp:218] Iteration 8940 (2.46479 iter/s, 4.86857s/12 iters), loss = 0.194494 I0410 00:28:50.581403 8290 solver.cpp:237] Train net output #0: loss = 0.194494 (* 1 = 0.194494 loss) I0410 00:28:50.581418 8290 sgd_solver.cpp:105] Iteration 8940, lr = 0.00170182 I0410 00:28:55.509181 8290 solver.cpp:218] Iteration 8952 (2.43527 iter/s, 4.92759s/12 iters), loss = 0.248457 I0410 00:28:55.509243 8290 solver.cpp:237] Train net output #0: loss = 0.248457 (* 1 = 0.248457 loss) I0410 00:28:55.509258 8290 sgd_solver.cpp:105] Iteration 8952, lr = 0.00169778 I0410 00:29:00.404501 8290 solver.cpp:218] Iteration 8964 (2.45144 iter/s, 4.89508s/12 iters), loss = 0.191258 I0410 00:29:00.404551 8290 solver.cpp:237] Train net output #0: loss = 0.191258 (* 1 = 0.191258 loss) I0410 00:29:00.404561 8290 sgd_solver.cpp:105] Iteration 8964, lr = 0.00169375 I0410 00:29:04.857890 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_8976.caffemodel I0410 00:29:07.009693 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_8976.solverstate I0410 00:29:07.985846 8290 solver.cpp:330] Iteration 8976, Testing net (#0) I0410 00:29:07.985875 8290 net.cpp:676] Ignoring source layer train-data I0410 00:29:08.988958 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:29:12.496011 8290 solver.cpp:397] Test net output #0: accuracy = 0.590074 I0410 00:29:12.496064 8290 solver.cpp:397] Test net output #1: loss = 2.04515 (* 1 = 2.04515 loss) I0410 00:29:12.578445 8290 solver.cpp:218] Iteration 8976 (0.985751 iter/s, 12.1735s/12 iters), loss = 0.11501 I0410 00:29:12.578529 8290 solver.cpp:237] Train net output #0: loss = 0.11501 (* 1 = 0.11501 loss) I0410 00:29:12.578545 8290 sgd_solver.cpp:105] Iteration 8976, lr = 0.00168973 I0410 00:29:16.723217 8290 solver.cpp:218] Iteration 8988 (2.89538 iter/s, 4.14454s/12 iters), loss = 0.131813 I0410 00:29:16.723256 8290 solver.cpp:237] Train net output #0: loss = 0.131813 (* 1 = 0.131813 loss) I0410 00:29:16.723265 8290 sgd_solver.cpp:105] Iteration 8988, lr = 0.00168571 I0410 00:29:18.756350 8290 blocking_queue.cpp:49] Waiting for data I0410 00:29:21.646869 8290 solver.cpp:218] Iteration 9000 (2.43733 iter/s, 4.92343s/12 iters), loss = 0.159584 I0410 00:29:21.646952 8290 solver.cpp:237] Train net output #0: loss = 0.159584 (* 1 = 0.159584 loss) I0410 00:29:21.646963 8290 sgd_solver.cpp:105] Iteration 9000, lr = 0.00168171 I0410 00:29:22.329551 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:29:26.399751 8290 solver.cpp:218] Iteration 9012 (2.52492 iter/s, 4.75262s/12 iters), loss = 0.251617 I0410 00:29:26.399803 8290 solver.cpp:237] Train net output #0: loss = 0.251617 (* 1 = 0.251617 loss) I0410 00:29:26.399816 8290 sgd_solver.cpp:105] Iteration 9012, lr = 0.00167772 I0410 00:29:31.362857 8290 solver.cpp:218] Iteration 9024 (2.41796 iter/s, 4.96286s/12 iters), loss = 0.259443 I0410 00:29:31.362916 8290 solver.cpp:237] Train net output #0: loss = 0.259443 (* 1 = 0.259443 loss) I0410 00:29:31.362931 8290 sgd_solver.cpp:105] Iteration 9024, lr = 0.00167374 I0410 00:29:36.295392 8290 solver.cpp:218] Iteration 9036 (2.43295 iter/s, 4.93229s/12 iters), loss = 0.18201 I0410 00:29:36.295442 8290 solver.cpp:237] Train net output #0: loss = 0.18201 (* 1 = 0.18201 loss) I0410 00:29:36.295455 8290 sgd_solver.cpp:105] Iteration 9036, lr = 0.00166976 I0410 00:29:41.230516 8290 solver.cpp:218] Iteration 9048 (2.43167 iter/s, 4.93489s/12 iters), loss = 0.215586 I0410 00:29:41.230567 8290 solver.cpp:237] Train net output #0: loss = 0.215586 (* 1 = 0.215586 loss) I0410 00:29:41.230578 8290 sgd_solver.cpp:105] Iteration 9048, lr = 0.0016658 I0410 00:29:46.139770 8290 solver.cpp:218] Iteration 9060 (2.44448 iter/s, 4.90902s/12 iters), loss = 0.266254 I0410 00:29:46.139822 8290 solver.cpp:237] Train net output #0: loss = 0.266254 (* 1 = 0.266254 loss) I0410 00:29:46.139833 8290 sgd_solver.cpp:105] Iteration 9060, lr = 0.00166184 I0410 00:29:51.066745 8290 solver.cpp:218] Iteration 9072 (2.43569 iter/s, 4.92674s/12 iters), loss = 0.170245 I0410 00:29:51.066790 8290 solver.cpp:237] Train net output #0: loss = 0.170245 (* 1 = 0.170245 loss) I0410 00:29:51.066802 8290 sgd_solver.cpp:105] Iteration 9072, lr = 0.0016579 I0410 00:29:53.074246 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9078.caffemodel I0410 00:29:53.817332 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9078.solverstate I0410 00:29:54.349016 8290 solver.cpp:330] Iteration 9078, Testing net (#0) I0410 00:29:54.349045 8290 net.cpp:676] Ignoring source layer train-data I0410 00:29:55.240712 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:29:58.780174 8290 solver.cpp:397] Test net output #0: accuracy = 0.604167 I0410 00:29:58.780222 8290 solver.cpp:397] Test net output #1: loss = 2.09165 (* 1 = 2.09165 loss) I0410 00:30:00.547986 8290 solver.cpp:218] Iteration 9084 (1.26571 iter/s, 9.48085s/12 iters), loss = 0.125276 I0410 00:30:00.548051 8290 solver.cpp:237] Train net output #0: loss = 0.125276 (* 1 = 0.125276 loss) I0410 00:30:00.548063 8290 sgd_solver.cpp:105] Iteration 9084, lr = 0.00165396 I0410 00:30:05.348831 8290 solver.cpp:218] Iteration 9096 (2.49969 iter/s, 4.8006s/12 iters), loss = 0.0673945 I0410 00:30:05.348892 8290 solver.cpp:237] Train net output #0: loss = 0.0673945 (* 1 = 0.0673945 loss) I0410 00:30:05.348903 8290 sgd_solver.cpp:105] Iteration 9096, lr = 0.00165003 I0410 00:30:08.194993 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:30:10.203805 8290 solver.cpp:218] Iteration 9108 (2.47181 iter/s, 4.85473s/12 iters), loss = 0.215655 I0410 00:30:10.203847 8290 solver.cpp:237] Train net output #0: loss = 0.215655 (* 1 = 0.215655 loss) I0410 00:30:10.203856 8290 sgd_solver.cpp:105] Iteration 9108, lr = 0.00164612 I0410 00:30:15.119817 8290 solver.cpp:218] Iteration 9120 (2.44112 iter/s, 4.91578s/12 iters), loss = 0.124072 I0410 00:30:15.119865 8290 solver.cpp:237] Train net output #0: loss = 0.124072 (* 1 = 0.124072 loss) I0410 00:30:15.119874 8290 sgd_solver.cpp:105] Iteration 9120, lr = 0.00164221 I0410 00:30:19.951079 8290 solver.cpp:218] Iteration 9132 (2.48394 iter/s, 4.83103s/12 iters), loss = 0.1231 I0410 00:30:19.951122 8290 solver.cpp:237] Train net output #0: loss = 0.1231 (* 1 = 0.1231 loss) I0410 00:30:19.951131 8290 sgd_solver.cpp:105] Iteration 9132, lr = 0.00163831 I0410 00:30:24.819855 8290 solver.cpp:218] Iteration 9144 (2.4648 iter/s, 4.86854s/12 iters), loss = 0.16798 I0410 00:30:24.819983 8290 solver.cpp:237] Train net output #0: loss = 0.16798 (* 1 = 0.16798 loss) I0410 00:30:24.819994 8290 sgd_solver.cpp:105] Iteration 9144, lr = 0.00163442 I0410 00:30:29.725283 8290 solver.cpp:218] Iteration 9156 (2.44642 iter/s, 4.90512s/12 iters), loss = 0.159549 I0410 00:30:29.725342 8290 solver.cpp:237] Train net output #0: loss = 0.159549 (* 1 = 0.159549 loss) I0410 00:30:29.725353 8290 sgd_solver.cpp:105] Iteration 9156, lr = 0.00163054 I0410 00:30:34.648039 8290 solver.cpp:218] Iteration 9168 (2.43778 iter/s, 4.92252s/12 iters), loss = 0.189383 I0410 00:30:34.648147 8290 solver.cpp:237] Train net output #0: loss = 0.189383 (* 1 = 0.189383 loss) I0410 00:30:34.648160 8290 sgd_solver.cpp:105] Iteration 9168, lr = 0.00162667 I0410 00:30:39.260759 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9180.caffemodel I0410 00:30:40.958412 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9180.solverstate I0410 00:30:42.071524 8290 solver.cpp:330] Iteration 9180, Testing net (#0) I0410 00:30:42.071552 8290 net.cpp:676] Ignoring source layer train-data I0410 00:30:42.964655 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:30:46.805498 8290 solver.cpp:397] Test net output #0: accuracy = 0.602941 I0410 00:30:46.805541 8290 solver.cpp:397] Test net output #1: loss = 2.15935 (* 1 = 2.15935 loss) I0410 00:30:46.887812 8290 solver.cpp:218] Iteration 9180 (0.980451 iter/s, 12.2393s/12 iters), loss = 0.115548 I0410 00:30:46.887876 8290 solver.cpp:237] Train net output #0: loss = 0.115548 (* 1 = 0.115548 loss) I0410 00:30:46.887889 8290 sgd_solver.cpp:105] Iteration 9180, lr = 0.00162281 I0410 00:30:51.091737 8290 solver.cpp:218] Iteration 9192 (2.85463 iter/s, 4.2037s/12 iters), loss = 0.149557 I0410 00:30:51.091796 8290 solver.cpp:237] Train net output #0: loss = 0.149557 (* 1 = 0.149557 loss) I0410 00:30:51.091809 8290 sgd_solver.cpp:105] Iteration 9192, lr = 0.00161895 I0410 00:30:55.942015 8290 solver.cpp:218] Iteration 9204 (2.47421 iter/s, 4.85004s/12 iters), loss = 0.164562 I0410 00:30:55.942124 8290 solver.cpp:237] Train net output #0: loss = 0.164562 (* 1 = 0.164562 loss) I0410 00:30:55.942135 8290 sgd_solver.cpp:105] Iteration 9204, lr = 0.00161511 I0410 00:30:56.021745 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:31:00.862694 8290 solver.cpp:218] Iteration 9216 (2.43883 iter/s, 4.92039s/12 iters), loss = 0.195578 I0410 00:31:00.862744 8290 solver.cpp:237] Train net output #0: loss = 0.195578 (* 1 = 0.195578 loss) I0410 00:31:00.862754 8290 sgd_solver.cpp:105] Iteration 9216, lr = 0.00161128 I0410 00:31:05.691622 8290 solver.cpp:218] Iteration 9228 (2.48514 iter/s, 4.82869s/12 iters), loss = 0.234862 I0410 00:31:05.691679 8290 solver.cpp:237] Train net output #0: loss = 0.234862 (* 1 = 0.234862 loss) I0410 00:31:05.691692 8290 sgd_solver.cpp:105] Iteration 9228, lr = 0.00160745 I0410 00:31:10.593757 8290 solver.cpp:218] Iteration 9240 (2.44803 iter/s, 4.90189s/12 iters), loss = 0.189641 I0410 00:31:10.593812 8290 solver.cpp:237] Train net output #0: loss = 0.189641 (* 1 = 0.189641 loss) I0410 00:31:10.593824 8290 sgd_solver.cpp:105] Iteration 9240, lr = 0.00160363 I0410 00:31:15.413465 8290 solver.cpp:218] Iteration 9252 (2.4899 iter/s, 4.81948s/12 iters), loss = 0.115376 I0410 00:31:15.413504 8290 solver.cpp:237] Train net output #0: loss = 0.115376 (* 1 = 0.115376 loss) I0410 00:31:15.413512 8290 sgd_solver.cpp:105] Iteration 9252, lr = 0.00159983 I0410 00:31:20.346664 8290 solver.cpp:218] Iteration 9264 (2.43261 iter/s, 4.93297s/12 iters), loss = 0.118274 I0410 00:31:20.346721 8290 solver.cpp:237] Train net output #0: loss = 0.118274 (* 1 = 0.118274 loss) I0410 00:31:20.346732 8290 sgd_solver.cpp:105] Iteration 9264, lr = 0.00159603 I0410 00:31:25.461244 8290 solver.cpp:218] Iteration 9276 (2.34635 iter/s, 5.11433s/12 iters), loss = 0.0920208 I0410 00:31:25.461305 8290 solver.cpp:237] Train net output #0: loss = 0.0920208 (* 1 = 0.0920208 loss) I0410 00:31:25.461318 8290 sgd_solver.cpp:105] Iteration 9276, lr = 0.00159224 I0410 00:31:27.604758 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9282.caffemodel I0410 00:31:28.379467 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9282.solverstate I0410 00:31:28.921085 8290 solver.cpp:330] Iteration 9282, Testing net (#0) I0410 00:31:28.921114 8290 net.cpp:676] Ignoring source layer train-data I0410 00:31:29.645447 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:31:33.256567 8290 solver.cpp:397] Test net output #0: accuracy = 0.594975 I0410 00:31:33.256613 8290 solver.cpp:397] Test net output #1: loss = 2.15205 (* 1 = 2.15205 loss) I0410 00:31:35.147245 8290 solver.cpp:218] Iteration 9288 (1.23895 iter/s, 9.68559s/12 iters), loss = 0.126785 I0410 00:31:35.147290 8290 solver.cpp:237] Train net output #0: loss = 0.126785 (* 1 = 0.126785 loss) I0410 00:31:35.147300 8290 sgd_solver.cpp:105] Iteration 9288, lr = 0.00158846 I0410 00:31:39.938809 8290 solver.cpp:218] Iteration 9300 (2.50452 iter/s, 4.79133s/12 iters), loss = 0.132317 I0410 00:31:39.938868 8290 solver.cpp:237] Train net output #0: loss = 0.132317 (* 1 = 0.132317 loss) I0410 00:31:39.938879 8290 sgd_solver.cpp:105] Iteration 9300, lr = 0.00158469 I0410 00:31:42.115159 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:31:44.856489 8290 solver.cpp:218] Iteration 9312 (2.4403 iter/s, 4.91744s/12 iters), loss = 0.142326 I0410 00:31:44.856534 8290 solver.cpp:237] Train net output #0: loss = 0.142326 (* 1 = 0.142326 loss) I0410 00:31:44.856544 8290 sgd_solver.cpp:105] Iteration 9312, lr = 0.00158092 I0410 00:31:50.038736 8290 solver.cpp:218] Iteration 9324 (2.31571 iter/s, 5.18201s/12 iters), loss = 0.184327 I0410 00:31:50.038792 8290 solver.cpp:237] Train net output #0: loss = 0.184327 (* 1 = 0.184327 loss) I0410 00:31:50.038805 8290 sgd_solver.cpp:105] Iteration 9324, lr = 0.00157717 I0410 00:31:54.916597 8290 solver.cpp:218] Iteration 9336 (2.46021 iter/s, 4.87763s/12 iters), loss = 0.139177 I0410 00:31:54.916633 8290 solver.cpp:237] Train net output #0: loss = 0.139177 (* 1 = 0.139177 loss) I0410 00:31:54.916640 8290 sgd_solver.cpp:105] Iteration 9336, lr = 0.00157343 I0410 00:31:59.907281 8290 solver.cpp:218] Iteration 9348 (2.40459 iter/s, 4.99046s/12 iters), loss = 0.142394 I0410 00:31:59.907366 8290 solver.cpp:237] Train net output #0: loss = 0.142394 (* 1 = 0.142394 loss) I0410 00:31:59.907377 8290 sgd_solver.cpp:105] Iteration 9348, lr = 0.00156969 I0410 00:32:04.762895 8290 solver.cpp:218] Iteration 9360 (2.4715 iter/s, 4.85535s/12 iters), loss = 0.0649428 I0410 00:32:04.762941 8290 solver.cpp:237] Train net output #0: loss = 0.0649428 (* 1 = 0.0649428 loss) I0410 00:32:04.762951 8290 sgd_solver.cpp:105] Iteration 9360, lr = 0.00156596 I0410 00:32:09.692785 8290 solver.cpp:218] Iteration 9372 (2.43425 iter/s, 4.92966s/12 iters), loss = 0.212618 I0410 00:32:09.692842 8290 solver.cpp:237] Train net output #0: loss = 0.212618 (* 1 = 0.212618 loss) I0410 00:32:09.692854 8290 sgd_solver.cpp:105] Iteration 9372, lr = 0.00156225 I0410 00:32:14.104759 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9384.caffemodel I0410 00:32:14.827345 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9384.solverstate I0410 00:32:15.363853 8290 solver.cpp:330] Iteration 9384, Testing net (#0) I0410 00:32:15.363874 8290 net.cpp:676] Ignoring source layer train-data I0410 00:32:16.108899 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:32:19.764768 8290 solver.cpp:397] Test net output #0: accuracy = 0.598652 I0410 00:32:19.764819 8290 solver.cpp:397] Test net output #1: loss = 2.13416 (* 1 = 2.13416 loss) I0410 00:32:19.846912 8290 solver.cpp:218] Iteration 9384 (1.18183 iter/s, 10.1537s/12 iters), loss = 0.134881 I0410 00:32:19.846971 8290 solver.cpp:237] Train net output #0: loss = 0.134881 (* 1 = 0.134881 loss) I0410 00:32:19.846983 8290 sgd_solver.cpp:105] Iteration 9384, lr = 0.00155854 I0410 00:32:23.930310 8290 solver.cpp:218] Iteration 9396 (2.93888 iter/s, 4.08318s/12 iters), loss = 0.136849 I0410 00:32:23.930369 8290 solver.cpp:237] Train net output #0: loss = 0.136849 (* 1 = 0.136849 loss) I0410 00:32:23.930382 8290 sgd_solver.cpp:105] Iteration 9396, lr = 0.00155484 I0410 00:32:28.117674 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:32:28.757885 8290 solver.cpp:218] Iteration 9408 (2.48584 iter/s, 4.82734s/12 iters), loss = 0.206749 I0410 00:32:28.757941 8290 solver.cpp:237] Train net output #0: loss = 0.206749 (* 1 = 0.206749 loss) I0410 00:32:28.757951 8290 sgd_solver.cpp:105] Iteration 9408, lr = 0.00155114 I0410 00:32:33.762195 8290 solver.cpp:218] Iteration 9420 (2.39805 iter/s, 5.00407s/12 iters), loss = 0.100561 I0410 00:32:33.762358 8290 solver.cpp:237] Train net output #0: loss = 0.100561 (* 1 = 0.100561 loss) I0410 00:32:33.762372 8290 sgd_solver.cpp:105] Iteration 9420, lr = 0.00154746 I0410 00:32:38.728732 8290 solver.cpp:218] Iteration 9432 (2.41634 iter/s, 4.96619s/12 iters), loss = 0.167086 I0410 00:32:38.728782 8290 solver.cpp:237] Train net output #0: loss = 0.167086 (* 1 = 0.167086 loss) I0410 00:32:38.728793 8290 sgd_solver.cpp:105] Iteration 9432, lr = 0.00154379 I0410 00:32:43.641361 8290 solver.cpp:218] Iteration 9444 (2.4428 iter/s, 4.9124s/12 iters), loss = 0.17093 I0410 00:32:43.641419 8290 solver.cpp:237] Train net output #0: loss = 0.17093 (* 1 = 0.17093 loss) I0410 00:32:43.641433 8290 sgd_solver.cpp:105] Iteration 9444, lr = 0.00154012 I0410 00:32:48.571614 8290 solver.cpp:218] Iteration 9456 (2.43407 iter/s, 4.93001s/12 iters), loss = 0.114931 I0410 00:32:48.571663 8290 solver.cpp:237] Train net output #0: loss = 0.114931 (* 1 = 0.114931 loss) I0410 00:32:48.571676 8290 sgd_solver.cpp:105] Iteration 9456, lr = 0.00153647 I0410 00:32:53.421175 8290 solver.cpp:218] Iteration 9468 (2.47457 iter/s, 4.84932s/12 iters), loss = 0.118632 I0410 00:32:53.421233 8290 solver.cpp:237] Train net output #0: loss = 0.118632 (* 1 = 0.118632 loss) I0410 00:32:53.421244 8290 sgd_solver.cpp:105] Iteration 9468, lr = 0.00153282 I0410 00:32:58.256907 8290 solver.cpp:218] Iteration 9480 (2.48165 iter/s, 4.83549s/12 iters), loss = 0.179933 I0410 00:32:58.256951 8290 solver.cpp:237] Train net output #0: loss = 0.179933 (* 1 = 0.179933 loss) I0410 00:32:58.256959 8290 sgd_solver.cpp:105] Iteration 9480, lr = 0.00152918 I0410 00:33:00.203931 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9486.caffemodel I0410 00:33:00.908666 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9486.solverstate I0410 00:33:01.437599 8290 solver.cpp:330] Iteration 9486, Testing net (#0) I0410 00:33:01.437623 8290 net.cpp:676] Ignoring source layer train-data I0410 00:33:02.109122 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:33:05.802000 8290 solver.cpp:397] Test net output #0: accuracy = 0.608456 I0410 00:33:05.802088 8290 solver.cpp:397] Test net output #1: loss = 2.14265 (* 1 = 2.14265 loss) I0410 00:33:07.607544 8290 solver.cpp:218] Iteration 9492 (1.28339 iter/s, 9.35025s/12 iters), loss = 0.215604 I0410 00:33:07.607595 8290 solver.cpp:237] Train net output #0: loss = 0.215604 (* 1 = 0.215604 loss) I0410 00:33:07.607605 8290 sgd_solver.cpp:105] Iteration 9492, lr = 0.00152555 I0410 00:33:12.517712 8290 solver.cpp:218] Iteration 9504 (2.44403 iter/s, 4.90993s/12 iters), loss = 0.147493 I0410 00:33:12.517760 8290 solver.cpp:237] Train net output #0: loss = 0.147493 (* 1 = 0.147493 loss) I0410 00:33:12.517768 8290 sgd_solver.cpp:105] Iteration 9504, lr = 0.00152193 I0410 00:33:13.948803 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:33:17.429267 8290 solver.cpp:218] Iteration 9516 (2.44334 iter/s, 4.91132s/12 iters), loss = 0.171664 I0410 00:33:17.429329 8290 solver.cpp:237] Train net output #0: loss = 0.171664 (* 1 = 0.171664 loss) I0410 00:33:17.429342 8290 sgd_solver.cpp:105] Iteration 9516, lr = 0.00151831 I0410 00:33:22.360376 8290 solver.cpp:218] Iteration 9528 (2.43365 iter/s, 4.93086s/12 iters), loss = 0.0868357 I0410 00:33:22.360435 8290 solver.cpp:237] Train net output #0: loss = 0.0868357 (* 1 = 0.0868357 loss) I0410 00:33:22.360448 8290 sgd_solver.cpp:105] Iteration 9528, lr = 0.00151471 I0410 00:33:27.311655 8290 solver.cpp:218] Iteration 9540 (2.42374 iter/s, 4.95103s/12 iters), loss = 0.226056 I0410 00:33:27.311708 8290 solver.cpp:237] Train net output #0: loss = 0.226055 (* 1 = 0.226055 loss) I0410 00:33:27.311719 8290 sgd_solver.cpp:105] Iteration 9540, lr = 0.00151111 I0410 00:33:32.231431 8290 solver.cpp:218] Iteration 9552 (2.43925 iter/s, 4.91954s/12 iters), loss = 0.140518 I0410 00:33:32.231482 8290 solver.cpp:237] Train net output #0: loss = 0.140518 (* 1 = 0.140518 loss) I0410 00:33:32.231493 8290 sgd_solver.cpp:105] Iteration 9552, lr = 0.00150752 I0410 00:33:37.162298 8290 solver.cpp:218] Iteration 9564 (2.43376 iter/s, 4.93063s/12 iters), loss = 0.103133 I0410 00:33:37.162418 8290 solver.cpp:237] Train net output #0: loss = 0.103133 (* 1 = 0.103133 loss) I0410 00:33:37.162428 8290 sgd_solver.cpp:105] Iteration 9564, lr = 0.00150395 I0410 00:33:42.063607 8290 solver.cpp:218] Iteration 9576 (2.44847 iter/s, 4.90101s/12 iters), loss = 0.0799665 I0410 00:33:42.063647 8290 solver.cpp:237] Train net output #0: loss = 0.0799664 (* 1 = 0.0799664 loss) I0410 00:33:42.063657 8290 sgd_solver.cpp:105] Iteration 9576, lr = 0.00150037 I0410 00:33:46.545234 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9588.caffemodel I0410 00:33:47.939637 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9588.solverstate I0410 00:33:48.871769 8290 solver.cpp:330] Iteration 9588, Testing net (#0) I0410 00:33:48.871799 8290 net.cpp:676] Ignoring source layer train-data I0410 00:33:49.557360 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:33:53.557411 8290 solver.cpp:397] Test net output #0: accuracy = 0.598652 I0410 00:33:53.557461 8290 solver.cpp:397] Test net output #1: loss = 2.16206 (* 1 = 2.16206 loss) I0410 00:33:53.639933 8290 solver.cpp:218] Iteration 9588 (1.03664 iter/s, 11.5759s/12 iters), loss = 0.11849 I0410 00:33:53.639981 8290 solver.cpp:237] Train net output #0: loss = 0.11849 (* 1 = 0.11849 loss) I0410 00:33:53.639994 8290 sgd_solver.cpp:105] Iteration 9588, lr = 0.00149681 I0410 00:33:57.837108 8290 solver.cpp:218] Iteration 9600 (2.85921 iter/s, 4.19697s/12 iters), loss = 0.179366 I0410 00:33:57.837157 8290 solver.cpp:237] Train net output #0: loss = 0.179365 (* 1 = 0.179365 loss) I0410 00:33:57.837167 8290 sgd_solver.cpp:105] Iteration 9600, lr = 0.00149326 I0410 00:34:01.385905 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:34:02.748914 8290 solver.cpp:218] Iteration 9612 (2.44321 iter/s, 4.91157s/12 iters), loss = 0.191174 I0410 00:34:02.748966 8290 solver.cpp:237] Train net output #0: loss = 0.191174 (* 1 = 0.191174 loss) I0410 00:34:02.748977 8290 sgd_solver.cpp:105] Iteration 9612, lr = 0.00148971 I0410 00:34:07.719799 8290 solver.cpp:218] Iteration 9624 (2.41417 iter/s, 4.97065s/12 iters), loss = 0.222863 I0410 00:34:07.719877 8290 solver.cpp:237] Train net output #0: loss = 0.222863 (* 1 = 0.222863 loss) I0410 00:34:07.719887 8290 sgd_solver.cpp:105] Iteration 9624, lr = 0.00148618 I0410 00:34:12.702438 8290 solver.cpp:218] Iteration 9636 (2.40849 iter/s, 4.98238s/12 iters), loss = 0.152528 I0410 00:34:12.702488 8290 solver.cpp:237] Train net output #0: loss = 0.152528 (* 1 = 0.152528 loss) I0410 00:34:12.702497 8290 sgd_solver.cpp:105] Iteration 9636, lr = 0.00148265 I0410 00:34:17.631826 8290 solver.cpp:218] Iteration 9648 (2.43449 iter/s, 4.92916s/12 iters), loss = 0.118202 I0410 00:34:17.631878 8290 solver.cpp:237] Train net output #0: loss = 0.118202 (* 1 = 0.118202 loss) I0410 00:34:17.631888 8290 sgd_solver.cpp:105] Iteration 9648, lr = 0.00147913 I0410 00:34:22.616988 8290 solver.cpp:218] Iteration 9660 (2.40726 iter/s, 4.98492s/12 iters), loss = 0.185248 I0410 00:34:22.617031 8290 solver.cpp:237] Train net output #0: loss = 0.185248 (* 1 = 0.185248 loss) I0410 00:34:22.617039 8290 sgd_solver.cpp:105] Iteration 9660, lr = 0.00147562 I0410 00:34:27.522323 8290 solver.cpp:218] Iteration 9672 (2.44643 iter/s, 4.9051s/12 iters), loss = 0.17702 I0410 00:34:27.522385 8290 solver.cpp:237] Train net output #0: loss = 0.17702 (* 1 = 0.17702 loss) I0410 00:34:27.522397 8290 sgd_solver.cpp:105] Iteration 9672, lr = 0.00147211 I0410 00:34:32.530542 8290 solver.cpp:218] Iteration 9684 (2.39618 iter/s, 5.00797s/12 iters), loss = 0.155898 I0410 00:34:32.530596 8290 solver.cpp:237] Train net output #0: loss = 0.155898 (* 1 = 0.155898 loss) I0410 00:34:32.530608 8290 sgd_solver.cpp:105] Iteration 9684, lr = 0.00146862 I0410 00:34:34.530082 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9690.caffemodel I0410 00:34:36.131292 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9690.solverstate I0410 00:34:38.692986 8290 solver.cpp:330] Iteration 9690, Testing net (#0) I0410 00:34:38.693109 8290 net.cpp:676] Ignoring source layer train-data I0410 00:34:39.316804 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:34:41.926131 8290 blocking_queue.cpp:49] Waiting for data I0410 00:34:43.221021 8290 solver.cpp:397] Test net output #0: accuracy = 0.597426 I0410 00:34:43.221065 8290 solver.cpp:397] Test net output #1: loss = 2.23705 (* 1 = 2.23705 loss) I0410 00:34:45.085345 8290 solver.cpp:218] Iteration 9696 (0.955848 iter/s, 12.5543s/12 iters), loss = 0.126858 I0410 00:34:45.085393 8290 solver.cpp:237] Train net output #0: loss = 0.126858 (* 1 = 0.126858 loss) I0410 00:34:45.085403 8290 sgd_solver.cpp:105] Iteration 9696, lr = 0.00146513 I0410 00:34:49.933279 8290 solver.cpp:218] Iteration 9708 (2.4754 iter/s, 4.84769s/12 iters), loss = 0.123293 I0410 00:34:49.933341 8290 solver.cpp:237] Train net output #0: loss = 0.123293 (* 1 = 0.123293 loss) I0410 00:34:49.933353 8290 sgd_solver.cpp:105] Iteration 9708, lr = 0.00146165 I0410 00:34:50.667102 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:34:54.817234 8290 solver.cpp:218] Iteration 9720 (2.45715 iter/s, 4.88371s/12 iters), loss = 0.210054 I0410 00:34:54.817284 8290 solver.cpp:237] Train net output #0: loss = 0.210054 (* 1 = 0.210054 loss) I0410 00:34:54.817296 8290 sgd_solver.cpp:105] Iteration 9720, lr = 0.00145818 I0410 00:34:59.644315 8290 solver.cpp:218] Iteration 9732 (2.48609 iter/s, 4.82685s/12 iters), loss = 0.163827 I0410 00:34:59.644372 8290 solver.cpp:237] Train net output #0: loss = 0.163827 (* 1 = 0.163827 loss) I0410 00:34:59.644385 8290 sgd_solver.cpp:105] Iteration 9732, lr = 0.00145472 I0410 00:35:04.554832 8290 solver.cpp:218] Iteration 9744 (2.44385 iter/s, 4.91028s/12 iters), loss = 0.0851454 I0410 00:35:04.554874 8290 solver.cpp:237] Train net output #0: loss = 0.0851453 (* 1 = 0.0851453 loss) I0410 00:35:04.554883 8290 sgd_solver.cpp:105] Iteration 9744, lr = 0.00145127 I0410 00:35:09.408694 8290 solver.cpp:218] Iteration 9756 (2.47238 iter/s, 4.85363s/12 iters), loss = 0.190312 I0410 00:35:09.408792 8290 solver.cpp:237] Train net output #0: loss = 0.190312 (* 1 = 0.190312 loss) I0410 00:35:09.408805 8290 sgd_solver.cpp:105] Iteration 9756, lr = 0.00144782 I0410 00:35:14.348487 8290 solver.cpp:218] Iteration 9768 (2.42939 iter/s, 4.93951s/12 iters), loss = 0.216784 I0410 00:35:14.348536 8290 solver.cpp:237] Train net output #0: loss = 0.216784 (* 1 = 0.216784 loss) I0410 00:35:14.348546 8290 sgd_solver.cpp:105] Iteration 9768, lr = 0.00144438 I0410 00:35:19.372787 8290 solver.cpp:218] Iteration 9780 (2.38851 iter/s, 5.02406s/12 iters), loss = 0.157838 I0410 00:35:19.372839 8290 solver.cpp:237] Train net output #0: loss = 0.157838 (* 1 = 0.157838 loss) I0410 00:35:19.372849 8290 sgd_solver.cpp:105] Iteration 9780, lr = 0.00144095 I0410 00:35:23.760664 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9792.caffemodel I0410 00:35:24.574888 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9792.solverstate I0410 00:35:25.939250 8290 solver.cpp:330] Iteration 9792, Testing net (#0) I0410 00:35:25.939282 8290 net.cpp:676] Ignoring source layer train-data I0410 00:35:26.503574 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:35:30.317384 8290 solver.cpp:397] Test net output #0: accuracy = 0.599877 I0410 00:35:30.317432 8290 solver.cpp:397] Test net output #1: loss = 2.1071 (* 1 = 2.1071 loss) I0410 00:35:30.399585 8290 solver.cpp:218] Iteration 9792 (1.0883 iter/s, 11.0263s/12 iters), loss = 0.0671187 I0410 00:35:30.399643 8290 solver.cpp:237] Train net output #0: loss = 0.0671187 (* 1 = 0.0671187 loss) I0410 00:35:30.399655 8290 sgd_solver.cpp:105] Iteration 9792, lr = 0.00143753 I0410 00:35:34.721369 8290 solver.cpp:218] Iteration 9804 (2.77677 iter/s, 4.32156s/12 iters), loss = 0.225088 I0410 00:35:34.721418 8290 solver.cpp:237] Train net output #0: loss = 0.225088 (* 1 = 0.225088 loss) I0410 00:35:34.721427 8290 sgd_solver.cpp:105] Iteration 9804, lr = 0.00143412 I0410 00:35:37.590451 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:35:39.568425 8290 solver.cpp:218] Iteration 9816 (2.47585 iter/s, 4.84682s/12 iters), loss = 0.169179 I0410 00:35:39.568533 8290 solver.cpp:237] Train net output #0: loss = 0.169179 (* 1 = 0.169179 loss) I0410 00:35:39.568544 8290 sgd_solver.cpp:105] Iteration 9816, lr = 0.00143072 I0410 00:35:44.491994 8290 solver.cpp:218] Iteration 9828 (2.4374 iter/s, 4.92328s/12 iters), loss = 0.249713 I0410 00:35:44.492048 8290 solver.cpp:237] Train net output #0: loss = 0.249713 (* 1 = 0.249713 loss) I0410 00:35:44.492061 8290 sgd_solver.cpp:105] Iteration 9828, lr = 0.00142732 I0410 00:35:49.356606 8290 solver.cpp:218] Iteration 9840 (2.46691 iter/s, 4.86438s/12 iters), loss = 0.106891 I0410 00:35:49.356655 8290 solver.cpp:237] Train net output #0: loss = 0.106891 (* 1 = 0.106891 loss) I0410 00:35:49.356665 8290 sgd_solver.cpp:105] Iteration 9840, lr = 0.00142393 I0410 00:35:54.209165 8290 solver.cpp:218] Iteration 9852 (2.47304 iter/s, 4.85233s/12 iters), loss = 0.147359 I0410 00:35:54.209224 8290 solver.cpp:237] Train net output #0: loss = 0.147359 (* 1 = 0.147359 loss) I0410 00:35:54.209236 8290 sgd_solver.cpp:105] Iteration 9852, lr = 0.00142055 I0410 00:35:59.078310 8290 solver.cpp:218] Iteration 9864 (2.46462 iter/s, 4.86891s/12 iters), loss = 0.142445 I0410 00:35:59.078364 8290 solver.cpp:237] Train net output #0: loss = 0.142445 (* 1 = 0.142445 loss) I0410 00:35:59.078377 8290 sgd_solver.cpp:105] Iteration 9864, lr = 0.00141718 I0410 00:36:03.945679 8290 solver.cpp:218] Iteration 9876 (2.46552 iter/s, 4.86713s/12 iters), loss = 0.219695 I0410 00:36:03.945734 8290 solver.cpp:237] Train net output #0: loss = 0.219695 (* 1 = 0.219695 loss) I0410 00:36:03.945747 8290 sgd_solver.cpp:105] Iteration 9876, lr = 0.00141381 I0410 00:36:08.859158 8290 solver.cpp:218] Iteration 9888 (2.44238 iter/s, 4.91324s/12 iters), loss = 0.151694 I0410 00:36:08.859211 8290 solver.cpp:237] Train net output #0: loss = 0.151694 (* 1 = 0.151694 loss) I0410 00:36:08.859225 8290 sgd_solver.cpp:105] Iteration 9888, lr = 0.00141045 I0410 00:36:10.863267 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9894.caffemodel I0410 00:36:11.615602 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9894.solverstate I0410 00:36:12.153570 8290 solver.cpp:330] Iteration 9894, Testing net (#0) I0410 00:36:12.153597 8290 net.cpp:676] Ignoring source layer train-data I0410 00:36:12.701270 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:36:16.681671 8290 solver.cpp:397] Test net output #0: accuracy = 0.609069 I0410 00:36:16.681721 8290 solver.cpp:397] Test net output #1: loss = 2.08726 (* 1 = 2.08726 loss) I0410 00:36:18.430235 8290 solver.cpp:218] Iteration 9900 (1.25383 iter/s, 9.57068s/12 iters), loss = 0.132175 I0410 00:36:18.430290 8290 solver.cpp:237] Train net output #0: loss = 0.132175 (* 1 = 0.132175 loss) I0410 00:36:18.430302 8290 sgd_solver.cpp:105] Iteration 9900, lr = 0.00140711 I0410 00:36:23.257194 8290 solver.cpp:218] Iteration 9912 (2.48616 iter/s, 4.82673s/12 iters), loss = 0.145986 I0410 00:36:23.257230 8290 solver.cpp:237] Train net output #0: loss = 0.145986 (* 1 = 0.145986 loss) I0410 00:36:23.257238 8290 sgd_solver.cpp:105] Iteration 9912, lr = 0.00140377 I0410 00:36:23.355392 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:36:28.105561 8290 solver.cpp:218] Iteration 9924 (2.47517 iter/s, 4.84815s/12 iters), loss = 0.218555 I0410 00:36:28.105604 8290 solver.cpp:237] Train net output #0: loss = 0.218555 (* 1 = 0.218555 loss) I0410 00:36:28.105613 8290 sgd_solver.cpp:105] Iteration 9924, lr = 0.00140043 I0410 00:36:33.091598 8290 solver.cpp:218] Iteration 9936 (2.40683 iter/s, 4.98581s/12 iters), loss = 0.130835 I0410 00:36:33.091640 8290 solver.cpp:237] Train net output #0: loss = 0.130835 (* 1 = 0.130835 loss) I0410 00:36:33.091648 8290 sgd_solver.cpp:105] Iteration 9936, lr = 0.00139711 I0410 00:36:37.965807 8290 solver.cpp:218] Iteration 9948 (2.46205 iter/s, 4.87398s/12 iters), loss = 0.145394 I0410 00:36:37.965864 8290 solver.cpp:237] Train net output #0: loss = 0.145394 (* 1 = 0.145394 loss) I0410 00:36:37.965879 8290 sgd_solver.cpp:105] Iteration 9948, lr = 0.00139379 I0410 00:36:42.804741 8290 solver.cpp:218] Iteration 9960 (2.48001 iter/s, 4.8387s/12 iters), loss = 0.124511 I0410 00:36:42.804877 8290 solver.cpp:237] Train net output #0: loss = 0.124511 (* 1 = 0.124511 loss) I0410 00:36:42.804889 8290 sgd_solver.cpp:105] Iteration 9960, lr = 0.00139048 I0410 00:36:47.691112 8290 solver.cpp:218] Iteration 9972 (2.45597 iter/s, 4.88606s/12 iters), loss = 0.133182 I0410 00:36:47.691161 8290 solver.cpp:237] Train net output #0: loss = 0.133182 (* 1 = 0.133182 loss) I0410 00:36:47.691170 8290 sgd_solver.cpp:105] Iteration 9972, lr = 0.00138718 I0410 00:36:52.588011 8290 solver.cpp:218] Iteration 9984 (2.45065 iter/s, 4.89667s/12 iters), loss = 0.111334 I0410 00:36:52.588057 8290 solver.cpp:237] Train net output #0: loss = 0.111334 (* 1 = 0.111334 loss) I0410 00:36:52.588068 8290 sgd_solver.cpp:105] Iteration 9984, lr = 0.00138389 I0410 00:36:57.070251 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_9996.caffemodel I0410 00:36:58.823534 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_9996.solverstate I0410 00:37:00.408417 8290 solver.cpp:330] Iteration 9996, Testing net (#0) I0410 00:37:00.408447 8290 net.cpp:676] Ignoring source layer train-data I0410 00:37:00.941998 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:37:05.043601 8290 solver.cpp:397] Test net output #0: accuracy = 0.615196 I0410 00:37:05.043649 8290 solver.cpp:397] Test net output #1: loss = 2.08788 (* 1 = 2.08788 loss) I0410 00:37:05.126503 8290 solver.cpp:218] Iteration 9996 (0.957091 iter/s, 12.538s/12 iters), loss = 0.0860792 I0410 00:37:05.126585 8290 solver.cpp:237] Train net output #0: loss = 0.0860792 (* 1 = 0.0860792 loss) I0410 00:37:05.126605 8290 sgd_solver.cpp:105] Iteration 9996, lr = 0.0013806 I0410 00:37:09.444545 8290 solver.cpp:218] Iteration 10008 (2.77919 iter/s, 4.3178s/12 iters), loss = 0.142106 I0410 00:37:09.444594 8290 solver.cpp:237] Train net output #0: loss = 0.142106 (* 1 = 0.142106 loss) I0410 00:37:09.444604 8290 sgd_solver.cpp:105] Iteration 10008, lr = 0.00137732 I0410 00:37:11.613652 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:37:14.267819 8290 solver.cpp:218] Iteration 10020 (2.48806 iter/s, 4.82304s/12 iters), loss = 0.208689 I0410 00:37:14.267904 8290 solver.cpp:237] Train net output #0: loss = 0.208689 (* 1 = 0.208689 loss) I0410 00:37:14.267915 8290 sgd_solver.cpp:105] Iteration 10020, lr = 0.00137405 I0410 00:37:19.170140 8290 solver.cpp:218] Iteration 10032 (2.44795 iter/s, 4.90205s/12 iters), loss = 0.0948505 I0410 00:37:19.170183 8290 solver.cpp:237] Train net output #0: loss = 0.0948505 (* 1 = 0.0948505 loss) I0410 00:37:19.170192 8290 sgd_solver.cpp:105] Iteration 10032, lr = 0.00137079 I0410 00:37:24.075242 8290 solver.cpp:218] Iteration 10044 (2.44655 iter/s, 4.90487s/12 iters), loss = 0.0727096 I0410 00:37:24.075301 8290 solver.cpp:237] Train net output #0: loss = 0.0727096 (* 1 = 0.0727096 loss) I0410 00:37:24.075314 8290 sgd_solver.cpp:105] Iteration 10044, lr = 0.00136754 I0410 00:37:28.957461 8290 solver.cpp:218] Iteration 10056 (2.45802 iter/s, 4.88198s/12 iters), loss = 0.165142 I0410 00:37:28.957509 8290 solver.cpp:237] Train net output #0: loss = 0.165142 (* 1 = 0.165142 loss) I0410 00:37:28.957517 8290 sgd_solver.cpp:105] Iteration 10056, lr = 0.00136429 I0410 00:37:33.900094 8290 solver.cpp:218] Iteration 10068 (2.42797 iter/s, 4.9424s/12 iters), loss = 0.166083 I0410 00:37:33.900143 8290 solver.cpp:237] Train net output #0: loss = 0.166083 (* 1 = 0.166083 loss) I0410 00:37:33.900154 8290 sgd_solver.cpp:105] Iteration 10068, lr = 0.00136105 I0410 00:37:38.965620 8290 solver.cpp:218] Iteration 10080 (2.36907 iter/s, 5.06529s/12 iters), loss = 0.212723 I0410 00:37:38.965674 8290 solver.cpp:237] Train net output #0: loss = 0.212723 (* 1 = 0.212723 loss) I0410 00:37:38.965685 8290 sgd_solver.cpp:105] Iteration 10080, lr = 0.00135782 I0410 00:37:43.933156 8290 solver.cpp:218] Iteration 10092 (2.4158 iter/s, 4.9673s/12 iters), loss = 0.16935 I0410 00:37:43.933202 8290 solver.cpp:237] Train net output #0: loss = 0.16935 (* 1 = 0.16935 loss) I0410 00:37:43.933213 8290 sgd_solver.cpp:105] Iteration 10092, lr = 0.0013546 I0410 00:37:46.055474 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_10098.caffemodel I0410 00:37:47.328294 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_10098.solverstate I0410 00:37:48.062430 8290 solver.cpp:330] Iteration 10098, Testing net (#0) I0410 00:37:48.062450 8290 net.cpp:676] Ignoring source layer train-data I0410 00:37:48.538372 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:37:52.501708 8290 solver.cpp:397] Test net output #0: accuracy = 0.609069 I0410 00:37:52.501739 8290 solver.cpp:397] Test net output #1: loss = 2.11572 (* 1 = 2.11572 loss) I0410 00:37:54.352721 8290 solver.cpp:218] Iteration 10104 (1.15173 iter/s, 10.4191s/12 iters), loss = 0.0977715 I0410 00:37:54.352777 8290 solver.cpp:237] Train net output #0: loss = 0.0977715 (* 1 = 0.0977715 loss) I0410 00:37:54.352787 8290 sgd_solver.cpp:105] Iteration 10104, lr = 0.00135138 I0410 00:37:58.552284 8294 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:37:59.147158 8290 solver.cpp:218] Iteration 10116 (2.50302 iter/s, 4.7942s/12 iters), loss = 0.0882249 I0410 00:37:59.147215 8290 solver.cpp:237] Train net output #0: loss = 0.0882249 (* 1 = 0.0882249 loss) I0410 00:37:59.147228 8290 sgd_solver.cpp:105] Iteration 10116, lr = 0.00134817 I0410 00:38:03.966773 8290 solver.cpp:218] Iteration 10128 (2.48995 iter/s, 4.81938s/12 iters), loss = 0.249112 I0410 00:38:03.966827 8290 solver.cpp:237] Train net output #0: loss = 0.249112 (* 1 = 0.249112 loss) I0410 00:38:03.966840 8290 sgd_solver.cpp:105] Iteration 10128, lr = 0.00134497 I0410 00:38:08.920841 8290 solver.cpp:218] Iteration 10140 (2.42237 iter/s, 4.95383s/12 iters), loss = 0.0939745 I0410 00:38:08.920890 8290 solver.cpp:237] Train net output #0: loss = 0.0939745 (* 1 = 0.0939745 loss) I0410 00:38:08.920899 8290 sgd_solver.cpp:105] Iteration 10140, lr = 0.00134178 I0410 00:38:13.870360 8290 solver.cpp:218] Iteration 10152 (2.42459 iter/s, 4.94929s/12 iters), loss = 0.0703192 I0410 00:38:13.870406 8290 solver.cpp:237] Train net output #0: loss = 0.0703191 (* 1 = 0.0703191 loss) I0410 00:38:13.870416 8290 sgd_solver.cpp:105] Iteration 10152, lr = 0.00133859 I0410 00:38:18.714112 8290 solver.cpp:218] Iteration 10164 (2.47754 iter/s, 4.84352s/12 iters), loss = 0.124716 I0410 00:38:18.714257 8290 solver.cpp:237] Train net output #0: loss = 0.124716 (* 1 = 0.124716 loss) I0410 00:38:18.714267 8290 sgd_solver.cpp:105] Iteration 10164, lr = 0.00133541 I0410 00:38:23.588946 8290 solver.cpp:218] Iteration 10176 (2.46179 iter/s, 4.87451s/12 iters), loss = 0.107524 I0410 00:38:23.589005 8290 solver.cpp:237] Train net output #0: loss = 0.107524 (* 1 = 0.107524 loss) I0410 00:38:23.589017 8290 sgd_solver.cpp:105] Iteration 10176, lr = 0.00133224 I0410 00:38:28.485512 8290 solver.cpp:218] Iteration 10188 (2.45082 iter/s, 4.89632s/12 iters), loss = 0.0592905 I0410 00:38:28.485558 8290 solver.cpp:237] Train net output #0: loss = 0.0592904 (* 1 = 0.0592904 loss) I0410 00:38:28.485567 8290 sgd_solver.cpp:105] Iteration 10188, lr = 0.00132908 I0410 00:38:32.955538 8290 solver.cpp:447] Snapshotting to binary proto file snapshot_iter_10200.caffemodel I0410 00:38:34.173661 8290 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshot_iter_10200.solverstate I0410 00:38:35.317695 8290 solver.cpp:310] Iteration 10200, loss = 0.115279 I0410 00:38:35.317731 8290 solver.cpp:330] Iteration 10200, Testing net (#0) I0410 00:38:35.317740 8290 net.cpp:676] Ignoring source layer train-data I0410 00:38:35.728943 8295 data_layer.cpp:73] Restarting data prefetching from start. I0410 00:38:39.791018 8290 solver.cpp:397] Test net output #0: accuracy = 0.609069 I0410 00:38:39.791056 8290 solver.cpp:397] Test net output #1: loss = 2.08302 (* 1 = 2.08302 loss) I0410 00:38:39.791065 8290 solver.cpp:315] Optimization Done. I0410 00:38:39.791071 8290 caffe.cpp:259] Optimization Done.