Skip to first unread message
meng lin
May 22, 2016, 3:25:24 PM5/22/16
to Caffe Users
I tried to write a simple network for classifying cars, labels are simply the viewpoint.
The accuracy on training data rushes to about 0.7 then it starts to decrease. I tried to see the weights on each layer. They looks correct.
So anyone has any idea why this happens?
meng lin
May 22, 2016, 3:46:31 PM5/22/16
to Caffe Users
Also, I wonder what accuracy does caffe display, it is the accuracy of the last batch or 1/(number of batches ran in the interval)*sigma(batches ran in the interval)
meng lin
May 22, 2016, 7:16:49 PM5/22/16
to Caffe Users
trying.sh: command not found
I0522 17:33:03.098721 4784 caffe.cpp:185] Using GPUs 0
I0522 17:33:03.108016 4784 caffe.cpp:190] GPU 0: GeForce GTX 980 Ti
I0522 17:33:03.233537 4784 solver.cpp:48] Initializing solver from parameters:
test_iter: 2
test_interval: 100
base_lr: 0.001
display: 20
max_iter: 100000
lr_policy: "fixed"
momentum: 0.9
snapshot: 5000
snapshot_prefix: "/home/menglin/caffe-master/menglin_try/snapshot/"
solver_mode: GPU
device_id: 0
net: "menglin_try/cardata_resnet.prototxt"
momentum2: 0.999
type: "Adam"
I0522 17:33:03.233652 4784 solver.cpp:91] Creating training net from net file: menglin_try/cardata_resnet.prototxt
I0522 17:33:03.234083 4784 net.cpp:313] The NetState phase (0) differed from the phase (1) specified by a rule in layer data
I0522 17:33:03.234199 4784 net.cpp:49] Initializing net from parameters:
name: "mengNet"
state {
phase: TRAIN
}
layer {
name: "data"
type: "HDF5Data"
top: "data"
top: "label"
include {
phase: TRAIN
}
hdf5_data_param {
source: "/home/menglin/caffe-master/menglin_try/train.txt"
batch_size: 500
}
}
layer {
name: "conv1"
type: "Convolution"
bottom: "data"
top: "conv1"
param {
lr_mult: 1
}
param {
lr_mult: 2
}
convolution_param {
num_output: 32
kernel_size: 31
stride: 7
weight_filler {
type: "gaussian"
std: 0.0001
}
bias_filler {
type: "constant"
}
}
}
layer {
name: "res1/bn1"
type: "BatchNorm"
bottom: "conv1"
top: "res1/bn1"
param {
lr_mult: 0
}
param {
lr_mult: 0
}
param {
lr_mult: 0
}
}
layer {
name: "res1/relu1"
type: "ReLU"
bottom: "res1/bn1"
top: "res1/bn1"
}
layer {
name: "res1/conv1"
type: "Convolution"
bottom: "res1/bn1"
top: "res1/conv1"
param {
lr_mult: 1
}
param {
lr_mult: 2
}
convolution_param {
num_output: 32
pad: 2
kernel_size: 5
stride: 1
weight_filler {
type: "gaussian"
std: 0.0001
}
bias_filler {
type: "constant"
}
}
}
layer {
name: "res1/bn2"
type: "BatchNorm"
bottom: "res1/conv1"
top: "res1/bn2"
param {
lr_mult: 0
}
param {
lr_mult: 0
}
param {
lr_mult: 0
}
}
layer {
name: "res1/relu2"
type: "ReLU"
bottom: "res1/bn2"
top: "res1/bn2"
}
layer {
name: "res1/conv2"
type: "Convolution"
bottom: "res1/bn2"
top: "res1/conv2"
param {
lr_mult: 1
}
param {
lr_mult: 2
}
convolution_param {
num_output: 32
pad: 2
kernel_size: 5
stride: 1
weight_filler {
type: "gaussian"
std: 0.0001
}
bias_filler {
type: "constant"
}
}
}
layer {
name: "res1/elt"
type: "Eltwise"
bottom: "res1/conv2"
bottom: "conv1"
top: "res1/elt"
}
layer {
name: "pool1"
type: "Pooling"
bottom: "res1/elt"
top: "pool1"
pooling_param {
pool: MAX
kernel_size: 4
stride: 2
}
}
layer {
name: "res2/bn1"
type: "BatchNorm"
bottom: "pool1"
top: "res2/bn1"
param {
lr_mult: 0
}
param {
lr_mult: 0
}
param {
lr_mult: 0
}
}
layer {
name: "res2/relu1"
type: "ReLU"
bottom: "res2/bn1"
top: "res2/bn1"
}
layer {
name: "res2/conv1"
type: "Convolution"
bottom: "res2/bn1"
top: "res2/conv1"
param {
lr_mult: 1
}
param {
lr_mult: 2
}
convolution_param {
num_output: 32
pad: 3
kernel_size: 7
stride: 1
weight_filler {
type: "gaussian"
std: 0.0001
}
bias_filler {
type: "constant"
}
}
}
layer {
name: "res2/bn2"
type: "BatchNorm"
bottom: "res2/conv1"
top: "res2/bn2"
param {
lr_mult: 0
}
param {
lr_mult: 0
}
param {
lr_mult: 0
}
}
layer {
name: "res2/relu2"
type: "ReLU"
bottom: "res2/bn2"
top: "res2/bn2"
}
layer {
name: "res2/conv2"
type: "Convolution"
bottom: "res2/bn2"
top: "res2/conv2"
param {
lr_mult: 1
}
param {
lr_mult: 2
}
convolution_param {
num_output: 32
pad: 3
kernel_size: 7
stride: 1
weight_filler {
type: "gaussian"
std: 0.0001
}
bias_filler {
type: "constant"
}
}
}
layer {
name: "res2/elt"
type: "Eltwise"
bottom: "res2/conv2"
bottom: "pool1"
top: "res2/elt"
}
layer {
name: "pool2"
type: "Pooling"
bottom: "res2/elt"
top: "pool2"
pooling_param {
pool: MAX
kernel_size: 3
stride: 2
}
}
layer {
name: "ip1"
type: "InnerProduct"
bottom: "pool2"
top: "ip1"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 1
decay_mult: 0
}
inner_product_param {
num_output: 1000
weight_filler {
type: "gaussian"
std: 0.01
}
bias_filler {
type: "constant"
}
}
}
layer {
name: "bn5"
type: "BatchNorm"
bottom: "ip1"
top: "bn5"
param {
lr_mult: 0
}
param {
lr_mult: 0
}
param {
lr_mult: 0
}
}
layer {
name: "relu4"
type: "ReLU"
bottom: "bn5"
top: "bn5"
}
layer {
name: "ip2"
type: "InnerProduct"
bottom: "bn5"
top: "ip2"
inner_product_param {
num_output: 5
weight_filler {
type: "gaussian"
std: 0.01
}
bias_filler {
type: "constant"
}
}
}
layer {
name: "accuracy"
type: "Accuracy"
bottom: "ip2"
bottom: "label"
top: "accuracy"
}
layer {
name: "loss"
type: "SoftmaxWithLoss"
bottom: "ip2"
bottom: "label"
top: "loss"
}
I0522 17:33:03.234318 4784 layer_factory.hpp:77] Creating layer data
I0522 17:33:03.234339 4784 net.cpp:91] Creating Layer data
I0522 17:33:03.234344 4784 net.cpp:399] data -> data
I0522 17:33:03.234385 4784 net.cpp:399] data -> label
I0522 17:33:03.234407 4784 hdf5_data_layer.cpp:79] Loading list of HDF5 filenames from: /home/menglin/caffe-master/menglin_try/train.txt
I0522 17:33:03.234431 4784 hdf5_data_layer.cpp:93] Number of HDF5 files: 1
I0522 17:33:03.234949 4784 hdf5.cpp:32] Datatype class: H5T_FLOAT
I0522 17:33:05.712553 4784 net.cpp:141] Setting up data
I0522 17:33:05.712596 4784 net.cpp:148] Top shape: 500 3 255 255 (97537500)
I0522 17:33:05.712600 4784 net.cpp:148] Top shape: 500 1 (500)
I0522 17:33:05.712602 4784 net.cpp:156] Memory required for data: 390152000
I0522 17:33:05.712609 4784 layer_factory.hpp:77] Creating layer label_data_1_split
I0522 17:33:05.712628 4784 net.cpp:91] Creating Layer label_data_1_split
I0522 17:33:05.712633 4784 net.cpp:425] label_data_1_split <- label
I0522 17:33:05.712641 4784 net.cpp:399] label_data_1_split -> label_data_1_split_0
I0522 17:33:05.712658 4784 net.cpp:399] label_data_1_split -> label_data_1_split_1
I0522 17:33:05.712679 4784 net.cpp:141] Setting up label_data_1_split
I0522 17:33:05.712684 4784 net.cpp:148] Top shape: 500 1 (500)
I0522 17:33:05.712687 4784 net.cpp:148] Top shape: 500 1 (500)
I0522 17:33:05.712688 4784 net.cpp:156] Memory required for data: 390156000
I0522 17:33:05.712690 4784 layer_factory.hpp:77] Creating layer conv1
I0522 17:33:05.712716 4784 net.cpp:91] Creating Layer conv1
I0522 17:33:05.712719 4784 net.cpp:425] conv1 <- data
I0522 17:33:05.712733 4784 net.cpp:399] conv1 -> conv1
I0522 17:33:05.715376 4784 net.cpp:141] Setting up conv1
I0522 17:33:05.715385 4784 net.cpp:148] Top shape: 500 32 33 33 (17424000)
I0522 17:33:05.715389 4784 net.cpp:156] Memory required for data: 459852000
I0522 17:33:05.715396 4784 layer_factory.hpp:77] Creating layer conv1_conv1_0_split
I0522 17:33:05.715400 4784 net.cpp:91] Creating Layer conv1_conv1_0_split
I0522 17:33:05.715404 4784 net.cpp:425] conv1_conv1_0_split <- conv1
I0522 17:33:05.715416 4784 net.cpp:399] conv1_conv1_0_split -> conv1_conv1_0_split_0
I0522 17:33:05.715421 4784 net.cpp:399] conv1_conv1_0_split -> conv1_conv1_0_split_1
I0522 17:33:05.715440 4784 net.cpp:141] Setting up conv1_conv1_0_split
I0522 17:33:05.715445 4784 net.cpp:148] Top shape: 500 32 33 33 (17424000)
I0522 17:33:05.715456 4784 net.cpp:148] Top shape: 500 32 33 33 (17424000)
I0522 17:33:05.715458 4784 net.cpp:156] Memory required for data: 599244000
I0522 17:33:05.715461 4784 layer_factory.hpp:77] Creating layer res1/bn1
I0522 17:33:05.715478 4784 net.cpp:91] Creating Layer res1/bn1
I0522 17:33:05.715481 4784 net.cpp:425] res1/bn1 <- conv1_conv1_0_split_0
I0522 17:33:05.715492 4784 net.cpp:399] res1/bn1 -> res1/bn1
I0522 17:33:05.715849 4784 net.cpp:141] Setting up res1/bn1
I0522 17:33:05.715857 4784 net.cpp:148] Top shape: 500 32 33 33 (17424000)
I0522 17:33:05.715859 4784 net.cpp:156] Memory required for data: 668940000
I0522 17:33:05.715867 4784 layer_factory.hpp:77] Creating layer res1/relu1
I0522 17:33:05.715870 4784 net.cpp:91] Creating Layer res1/relu1
I0522 17:33:05.715873 4784 net.cpp:425] res1/relu1 <- res1/bn1
I0522 17:33:05.715886 4784 net.cpp:386] res1/relu1 -> res1/bn1 (in-place)
I0522 17:33:05.715890 4784 net.cpp:141] Setting up res1/relu1
I0522 17:33:05.715893 4784 net.cpp:148] Top shape: 500 32 33 33 (17424000)
I0522 17:33:05.715895 4784 net.cpp:156] Memory required for data: 738636000
I0522 17:33:05.715898 4784 layer_factory.hpp:77] Creating layer res1/conv1
I0522 17:33:05.715903 4784 net.cpp:91] Creating Layer res1/conv1
I0522 17:33:05.715905 4784 net.cpp:425] res1/conv1 <- res1/bn1
I0522 17:33:05.715909 4784 net.cpp:399] res1/conv1 -> res1/conv1
I0522 17:33:05.716745 4784 net.cpp:141] Setting up res1/conv1
I0522 17:33:05.716753 4784 net.cpp:148] Top shape: 500 32 33 33 (17424000)
I0522 17:33:05.716755 4784 net.cpp:156] Memory required for data: 808332000
I0522 17:33:05.716759 4784 layer_factory.hpp:77] Creating layer res1/bn2
I0522 17:33:05.716764 4784 net.cpp:91] Creating Layer res1/bn2
I0522 17:33:05.716766 4784 net.cpp:425] res1/bn2 <- res1/conv1
I0522 17:33:05.716770 4784 net.cpp:399] res1/bn2 -> res1/bn2
I0522 17:33:05.716892 4784 net.cpp:141] Setting up res1/bn2
I0522 17:33:05.716897 4784 net.cpp:148] Top shape: 500 32 33 33 (17424000)
I0522 17:33:05.716898 4784 net.cpp:156] Memory required for data: 878028000
I0522 17:33:05.716905 4784 layer_factory.hpp:77] Creating layer res1/relu2
I0522 17:33:05.716909 4784 net.cpp:91] Creating Layer res1/relu2
I0522 17:33:05.716912 4784 net.cpp:425] res1/relu2 <- res1/bn2
I0522 17:33:05.716914 4784 net.cpp:386] res1/relu2 -> res1/bn2 (in-place)
I0522 17:33:05.716928 4784 net.cpp:141] Setting up res1/relu2
I0522 17:33:05.716931 4784 net.cpp:148] Top shape: 500 32 33 33 (17424000)
I0522 17:33:05.716933 4784 net.cpp:156] Memory required for data: 947724000
I0522 17:33:05.716935 4784 layer_factory.hpp:77] Creating layer res1/conv2
I0522 17:33:05.716940 4784 net.cpp:91] Creating Layer res1/conv2
I0522 17:33:05.716943 4784 net.cpp:425] res1/conv2 <- res1/bn2
I0522 17:33:05.716946 4784 net.cpp:399] res1/conv2 -> res1/conv2
I0522 17:33:05.717557 4784 net.cpp:141] Setting up res1/conv2
I0522 17:33:05.717562 4784 net.cpp:148] Top shape: 500 32 33 33 (17424000)
I0522 17:33:05.717564 4784 net.cpp:156] Memory required for data: 1017420000
I0522 17:33:05.717567 4784 layer_factory.hpp:77] Creating layer res1/elt
I0522 17:33:05.717572 4784 net.cpp:91] Creating Layer res1/elt
I0522 17:33:05.717574 4784 net.cpp:425] res1/elt <- res1/conv2
I0522 17:33:05.717576 4784 net.cpp:425] res1/elt <- conv1_conv1_0_split_1
I0522 17:33:05.717579 4784 net.cpp:399] res1/elt -> res1/elt
I0522 17:33:05.717595 4784 net.cpp:141] Setting up res1/elt
I0522 17:33:05.717600 4784 net.cpp:148] Top shape: 500 32 33 33 (17424000)
I0522 17:33:05.717602 4784 net.cpp:156] Memory required for data: 1087116000
I0522 17:33:05.717604 4784 layer_factory.hpp:77] Creating layer pool1
I0522 17:33:05.717608 4784 net.cpp:91] Creating Layer pool1
I0522 17:33:05.717610 4784 net.cpp:425] pool1 <- res1/elt
I0522 17:33:05.717613 4784 net.cpp:399] pool1 -> pool1
I0522 17:33:05.717636 4784 net.cpp:141] Setting up pool1
I0522 17:33:05.717639 4784 net.cpp:148] Top shape: 500 32 16 16 (4096000)
I0522 17:33:05.717641 4784 net.cpp:156] Memory required for data: 1103500000
I0522 17:33:05.717643 4784 layer_factory.hpp:77] Creating layer pool1_pool1_0_split
I0522 17:33:05.717648 4784 net.cpp:91] Creating Layer pool1_pool1_0_split
I0522 17:33:05.717649 4784 net.cpp:425] pool1_pool1_0_split <- pool1
I0522 17:33:05.717651 4784 net.cpp:399] pool1_pool1_0_split -> pool1_pool1_0_split_0
I0522 17:33:05.717655 4784 net.cpp:399] pool1_pool1_0_split -> pool1_pool1_0_split_1
I0522 17:33:05.717670 4784 net.cpp:141] Setting up pool1_pool1_0_split
I0522 17:33:05.717681 4784 net.cpp:148] Top shape: 500 32 16 16 (4096000)
I0522 17:33:05.717684 4784 net.cpp:148] Top shape: 500 32 16 16 (4096000)
I0522 17:33:05.717685 4784 net.cpp:156] Memory required for data: 1136268000
I0522 17:33:05.717687 4784 layer_factory.hpp:77] Creating layer res2/bn1
I0522 17:33:05.717691 4784 net.cpp:91] Creating Layer res2/bn1
I0522 17:33:05.717694 4784 net.cpp:425] res2/bn1 <- pool1_pool1_0_split_0
I0522 17:33:05.717696 4784 net.cpp:399] res2/bn1 -> res2/bn1
I0522 17:33:05.717789 4784 net.cpp:141] Setting up res2/bn1
I0522 17:33:05.717793 4784 net.cpp:148] Top shape: 500 32 16 16 (4096000)
I0522 17:33:05.717795 4784 net.cpp:156] Memory required for data: 1152652000
I0522 17:33:05.717800 4784 layer_factory.hpp:77] Creating layer res2/relu1
I0522 17:33:05.717803 4784 net.cpp:91] Creating Layer res2/relu1
I0522 17:33:05.717805 4784 net.cpp:425] res2/relu1 <- res2/bn1
I0522 17:33:05.717808 4784 net.cpp:386] res2/relu1 -> res2/bn1 (in-place)
I0522 17:33:05.717811 4784 net.cpp:141] Setting up res2/relu1
I0522 17:33:05.717814 4784 net.cpp:148] Top shape: 500 32 16 16 (4096000)
I0522 17:33:05.717816 4784 net.cpp:156] Memory required for data: 1169036000
I0522 17:33:05.717818 4784 layer_factory.hpp:77] Creating layer res2/conv1
I0522 17:33:05.717823 4784 net.cpp:91] Creating Layer res2/conv1
I0522 17:33:05.717825 4784 net.cpp:425] res2/conv1 <- res2/bn1
I0522 17:33:05.717828 4784 net.cpp:399] res2/conv1 -> res2/conv1
I0522 17:33:05.718900 4784 net.cpp:141] Setting up res2/conv1
I0522 17:33:05.718905 4784 net.cpp:148] Top shape: 500 32 16 16 (4096000)
I0522 17:33:05.718907 4784 net.cpp:156] Memory required for data: 1185420000
I0522 17:33:05.718914 4784 layer_factory.hpp:77] Creating layer res2/bn2
I0522 17:33:05.718917 4784 net.cpp:91] Creating Layer res2/bn2
I0522 17:33:05.718920 4784 net.cpp:425] res2/bn2 <- res2/conv1
I0522 17:33:05.718924 4784 net.cpp:399] res2/bn2 -> res2/bn2
I0522 17:33:05.719013 4784 net.cpp:141] Setting up res2/bn2
I0522 17:33:05.719017 4784 net.cpp:148] Top shape: 500 32 16 16 (4096000)
I0522 17:33:05.719019 4784 net.cpp:156] Memory required for data: 1201804000
I0522 17:33:05.719023 4784 layer_factory.hpp:77] Creating layer res2/relu2
I0522 17:33:05.719027 4784 net.cpp:91] Creating Layer res2/relu2
I0522 17:33:05.719029 4784 net.cpp:425] res2/relu2 <- res2/bn2
I0522 17:33:05.719033 4784 net.cpp:386] res2/relu2 -> res2/bn2 (in-place)
I0522 17:33:05.719035 4784 net.cpp:141] Setting up res2/relu2
I0522 17:33:05.719038 4784 net.cpp:148] Top shape: 500 32 16 16 (4096000)
I0522 17:33:05.719039 4784 net.cpp:156] Memory required for data: 1218188000
I0522 17:33:05.719041 4784 layer_factory.hpp:77] Creating layer res2/conv2
I0522 17:33:05.719046 4784 net.cpp:91] Creating Layer res2/conv2
I0522 17:33:05.719048 4784 net.cpp:425] res2/conv2 <- res2/bn2
I0522 17:33:05.719053 4784 net.cpp:399] res2/conv2 -> res2/conv2
I0522 17:33:05.720123 4784 net.cpp:141] Setting up res2/conv2
I0522 17:33:05.720127 4784 net.cpp:148] Top shape: 500 32 16 16 (4096000)
I0522 17:33:05.720129 4784 net.cpp:156] Memory required for data: 1234572000
I0522 17:33:05.720134 4784 layer_factory.hpp:77] Creating layer res2/elt
I0522 17:33:05.720136 4784 net.cpp:91] Creating Layer res2/elt
I0522 17:33:05.720139 4784 net.cpp:425] res2/elt <- res2/conv2
I0522 17:33:05.720141 4784 net.cpp:425] res2/elt <- pool1_pool1_0_split_1
I0522 17:33:05.720144 4784 net.cpp:399] res2/elt -> res2/elt
I0522 17:33:05.720154 4784 net.cpp:141] Setting up res2/elt
I0522 17:33:05.720156 4784 net.cpp:148] Top shape: 500 32 16 16 (4096000)
I0522 17:33:05.720158 4784 net.cpp:156] Memory required for data: 1250956000
I0522 17:33:05.720160 4784 layer_factory.hpp:77] Creating layer pool2
I0522 17:33:05.720163 4784 net.cpp:91] Creating Layer pool2
I0522 17:33:05.720165 4784 net.cpp:425] pool2 <- res2/elt
I0522 17:33:05.720168 4784 net.cpp:399] pool2 -> pool2
I0522 17:33:05.720185 4784 net.cpp:141] Setting up pool2
I0522 17:33:05.720187 4784 net.cpp:148] Top shape: 500 32 8 8 (1024000)
I0522 17:33:05.720196 4784 net.cpp:156] Memory required for data: 1255052000
I0522 17:33:05.720197 4784 layer_factory.hpp:77] Creating layer ip1
I0522 17:33:05.720202 4784 net.cpp:91] Creating Layer ip1
I0522 17:33:05.720204 4784 net.cpp:425] ip1 <- pool2
I0522 17:33:05.720207 4784 net.cpp:399] ip1 -> ip1
I0522 17:33:05.761791 4784 net.cpp:141] Setting up ip1
I0522 17:33:05.761811 4784 net.cpp:148] Top shape: 500 1000 (500000)
I0522 17:33:05.761814 4784 net.cpp:156] Memory required for data: 1257052000
I0522 17:33:05.761821 4784 layer_factory.hpp:77] Creating layer bn5
I0522 17:33:05.761829 4784 net.cpp:91] Creating Layer bn5
I0522 17:33:05.761833 4784 net.cpp:425] bn5 <- ip1
I0522 17:33:05.761848 4784 net.cpp:399] bn5 -> bn5
I0522 17:33:05.761983 4784 net.cpp:141] Setting up bn5
I0522 17:33:05.761987 4784 net.cpp:148] Top shape: 500 1000 (500000)
I0522 17:33:05.761989 4784 net.cpp:156] Memory required for data: 1259052000
I0522 17:33:05.761994 4784 layer_factory.hpp:77] Creating layer relu4
I0522 17:33:05.761998 4784 net.cpp:91] Creating Layer relu4
I0522 17:33:05.762001 4784 net.cpp:425] relu4 <- bn5
I0522 17:33:05.762003 4784 net.cpp:386] relu4 -> bn5 (in-place)
I0522 17:33:05.762007 4784 net.cpp:141] Setting up relu4
I0522 17:33:05.762020 4784 net.cpp:148] Top shape: 500 1000 (500000)
I0522 17:33:05.762022 4784 net.cpp:156] Memory required for data: 1261052000
I0522 17:33:05.762024 4784 layer_factory.hpp:77] Creating layer ip2
I0522 17:33:05.762029 4784 net.cpp:91] Creating Layer ip2
I0522 17:33:05.762032 4784 net.cpp:425] ip2 <- bn5
I0522 17:33:05.762034 4784 net.cpp:399] ip2 -> ip2
I0522 17:33:05.762200 4784 net.cpp:141] Setting up ip2
I0522 17:33:05.762205 4784 net.cpp:148] Top shape: 500 5 (2500)
I0522 17:33:05.762207 4784 net.cpp:156] Memory required for data: 1261062000
I0522 17:33:05.762210 4784 layer_factory.hpp:77] Creating layer ip2_ip2_0_split
I0522 17:33:05.762214 4784 net.cpp:91] Creating Layer ip2_ip2_0_split
I0522 17:33:05.762217 4784 net.cpp:425] ip2_ip2_0_split <- ip2
I0522 17:33:05.762219 4784 net.cpp:399] ip2_ip2_0_split -> ip2_ip2_0_split_0
I0522 17:33:05.762223 4784 net.cpp:399] ip2_ip2_0_split -> ip2_ip2_0_split_1
I0522 17:33:05.762251 4784 net.cpp:141] Setting up ip2_ip2_0_split
I0522 17:33:05.762255 4784 net.cpp:148] Top shape: 500 5 (2500)
I0522 17:33:05.762258 4784 net.cpp:148] Top shape: 500 5 (2500)
I0522 17:33:05.762259 4784 net.cpp:156] Memory required for data: 1261082000
I0522 17:33:05.762261 4784 layer_factory.hpp:77] Creating layer accuracy
I0522 17:33:05.762265 4784 net.cpp:91] Creating Layer accuracy
I0522 17:33:05.762267 4784 net.cpp:425] accuracy <- ip2_ip2_0_split_0
I0522 17:33:05.762270 4784 net.cpp:425] accuracy <- label_data_1_split_0
I0522 17:33:05.762274 4784 net.cpp:399] accuracy -> accuracy
I0522 17:33:05.762279 4784 net.cpp:141] Setting up accuracy
I0522 17:33:05.762280 4784 net.cpp:148] Top shape: (1)
I0522 17:33:05.762282 4784 net.cpp:156] Memory required for data: 1261082004
I0522 17:33:05.762284 4784 layer_factory.hpp:77] Creating layer loss
I0522 17:33:05.762289 4784 net.cpp:91] Creating Layer loss
I0522 17:33:05.762290 4784 net.cpp:425] loss <- ip2_ip2_0_split_1
I0522 17:33:05.762293 4784 net.cpp:425] loss <- label_data_1_split_1
I0522 17:33:05.762296 4784 net.cpp:399] loss -> loss
I0522 17:33:05.762302 4784 layer_factory.hpp:77] Creating layer loss
I0522 17:33:05.762364 4784 net.cpp:141] Setting up loss
I0522 17:33:05.762368 4784 net.cpp:148] Top shape: (1)
I0522 17:33:05.762370 4784 net.cpp:151] with loss weight 1
I0522 17:33:05.762383 4784 net.cpp:156] Memory required for data: 1261082008
I0522 17:33:05.762385 4784 net.cpp:217] loss needs backward computation.
I0522 17:33:05.762388 4784 net.cpp:219] accuracy does not need backward computation.
I0522 17:33:05.762400 4784 net.cpp:217] ip2_ip2_0_split needs backward computation.
I0522 17:33:05.762403 4784 net.cpp:217] ip2 needs backward computation.
I0522 17:33:05.762405 4784 net.cpp:217] relu4 needs backward computation.
I0522 17:33:05.762418 4784 net.cpp:217] bn5 needs backward computation.
I0522 17:33:05.762419 4784 net.cpp:217] ip1 needs backward computation.
I0522 17:33:05.762421 4784 net.cpp:217] pool2 needs backward computation.
I0522 17:33:05.762424 4784 net.cpp:217] res2/elt needs backward computation.
I0522 17:33:05.762428 4784 net.cpp:217] res2/conv2 needs backward computation.
I0522 17:33:05.762429 4784 net.cpp:217] res2/relu2 needs backward computation.
I0522 17:33:05.762431 4784 net.cpp:217] res2/bn2 needs backward computation.
I0522 17:33:05.762434 4784 net.cpp:217] res2/conv1 needs backward computation.
I0522 17:33:05.762436 4784 net.cpp:217] res2/relu1 needs backward computation.
I0522 17:33:05.762439 4784 net.cpp:217] res2/bn1 needs backward computation.
I0522 17:33:05.762441 4784 net.cpp:217] pool1_pool1_0_split needs backward computation.
I0522 17:33:05.762444 4784 net.cpp:217] pool1 needs backward computation.
I0522 17:33:05.762445 4784 net.cpp:217] res1/elt needs backward computation.
I0522 17:33:05.762449 4784 net.cpp:217] res1/conv2 needs backward computation.
I0522 17:33:05.762450 4784 net.cpp:217] res1/relu2 needs backward computation.
I0522 17:33:05.762462 4784 net.cpp:217] res1/bn2 needs backward computation.
I0522 17:33:05.762465 4784 net.cpp:217] res1/conv1 needs backward computation.
I0522 17:33:05.762466 4784 net.cpp:217] res1/relu1 needs backward computation.
I0522 17:33:05.762468 4784 net.cpp:217] res1/bn1 needs backward computation.
I0522 17:33:05.762471 4784 net.cpp:217] conv1_conv1_0_split needs backward computation.
I0522 17:33:05.762473 4784 net.cpp:217] conv1 needs backward computation.
I0522 17:33:05.762476 4784 net.cpp:219] label_data_1_split does not need backward computation.
I0522 17:33:05.762478 4784 net.cpp:219] data does not need backward computation.
I0522 17:33:05.762480 4784 net.cpp:261] This network produces output accuracy
I0522 17:33:05.762483 4784 net.cpp:261] This network produces output loss
I0522 17:33:05.762495 4784 net.cpp:274] Network initialization done.
I0522 17:33:05.762926 4784 solver.cpp:181] Creating test net (#0) specified by net file: menglin_try/cardata_resnet.prototxt
I0522 17:33:05.762989 4784 net.cpp:313] The NetState phase (1) differed from the phase (0) specified by a rule in layer data
I0522 17:33:05.763106 4784 net.cpp:49] Initializing net from parameters:
name: "mengNet"
state {
phase: TEST
}
layer {
name: "data"
type: "HDF5Data"
top: "data"
top: "label"
include {
phase: TEST
}
hdf5_data_param {
source: "/home/menglin/caffe-master/menglin_try/test.txt"
batch_size: 500
}
}
layer {
name: "conv1"
type: "Convolution"
bottom: "data"
top: "conv1"
param {
lr_mult: 1
}
param {
lr_mult: 2
}
convolution_param {
num_output: 32
kernel_size: 31
stride: 7
weight_filler {
type: "gaussian"
std: 0.0001
}
bias_filler {
type: "constant"
}
}
}
layer {
name: "res1/bn1"
type: "BatchNorm"
bottom: "conv1"
top: "res1/bn1"
param {
lr_mult: 0
}
param {
lr_mult: 0
}
param {
lr_mult: 0
}
}
layer {
name: "res1/relu1"
type: "ReLU"
bottom: "res1/bn1"
top: "res1/bn1"
}
layer {
name: "res1/conv1"
type: "Convolution"
bottom: "res1/bn1"
top: "res1/conv1"
param {
lr_mult: 1
}
param {
lr_mult: 2
}
convolution_param {
num_output: 32
pad: 2
kernel_size: 5
stride: 1
weight_filler {
type: "gaussian"
std: 0.0001
}
bias_filler {
type: "constant"
}
}
}
layer {
name: "res1/bn2"
type: "BatchNorm"
bottom: "res1/conv1"
top: "res1/bn2"
param {
lr_mult: 0
}
param {
lr_mult: 0
}
param {
lr_mult: 0
}
}
layer {
name: "res1/relu2"
type: "ReLU"
bottom: "res1/bn2"
top: "res1/bn2"
}
layer {
name: "res1/conv2"
type: "Convolution"
bottom: "res1/bn2"
top: "res1/conv2"
param {
lr_mult: 1
}
param {
lr_mult: 2
}
convolution_param {
num_output: 32
pad: 2
kernel_size: 5
stride: 1
weight_filler {
type: "gaussian"
std: 0.0001
}
bias_filler {
type: "constant"
}
}
}
layer {
name: "res1/elt"
type: "Eltwise"
bottom: "res1/conv2"
bottom: "conv1"
top: "res1/elt"
}
layer {
name: "pool1"
type: "Pooling"
bottom: "res1/elt"
top: "pool1"
pooling_param {
pool: MAX
kernel_size: 4
stride: 2
}
}
layer {
name: "res2/bn1"
type: "BatchNorm"
bottom: "pool1"
top: "res2/bn1"
param {
lr_mult: 0
}
param {
lr_mult: 0
}
param {
lr_mult: 0
}
}
layer {
name: "res2/relu1"
type: "ReLU"
bottom: "res2/bn1"
top: "res2/bn1"
}
layer {
name: "res2/conv1"
type: "Convolution"
bottom: "res2/bn1"
top: "res2/conv1"
param {
lr_mult: 1
}
param {
lr_mult: 2
}
convolution_param {
num_output: 32
pad: 3
kernel_size: 7
stride: 1
weight_filler {
type: "gaussian"
std: 0.0001
}
bias_filler {
type: "constant"
}
}
}
layer {
name: "res2/bn2"
type: "BatchNorm"
bottom: "res2/conv1"
top: "res2/bn2"
param {
lr_mult: 0
}
param {
lr_mult: 0
}
param {
lr_mult: 0
}
}
layer {
name: "res2/relu2"
type: "ReLU"
bottom: "res2/bn2"
top: "res2/bn2"
}
layer {
name: "res2/conv2"
type: "Convolution"
bottom: "res2/bn2"
top: "res2/conv2"
param {
lr_mult: 1
}
param {
lr_mult: 2
}
convolution_param {
num_output: 32
pad: 3
kernel_size: 7
stride: 1
weight_filler {
type: "gaussian"
std: 0.0001
}
bias_filler {
type: "constant"
}
}
}
layer {
name: "res2/elt"
type: "Eltwise"
bottom: "res2/conv2"
bottom: "pool1"
top: "res2/elt"
}
layer {
name: "pool2"
type: "Pooling"
bottom: "res2/elt"
top: "pool2"
pooling_param {
pool: MAX
kernel_size: 3
stride: 2
}
}
layer {
name: "ip1"
type: "InnerProduct"
bottom: "pool2"
top: "ip1"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 1
decay_mult: 0
}
inner_product_param {
num_output: 1000
weight_filler {
type: "gaussian"
std: 0.01
}
bias_filler {
type: "constant"
}
}
}
layer {
name: "bn5"
type: "BatchNorm"
bottom: "ip1"
top: "bn5"
param {
lr_mult: 0
}
param {
lr_mult: 0
}
param {
lr_mult: 0
}
}
layer {
name: "relu4"
type: "ReLU"
bottom: "bn5"
top: "bn5"
}
layer {
name: "ip2"
type: "InnerProduct"
bottom: "bn5"
top: "ip2"
inner_product_param {
num_output: 5
weight_filler {
type: "gaussian"
std: 0.01
}
bias_filler {
type: "constant"
}
}
}
layer {
name: "accuracy"
type: "Accuracy"
bottom: "ip2"
bottom: "label"
top: "accuracy"
}
layer {
name: "loss"
type: "SoftmaxWithLoss"
bottom: "ip2"
bottom: "label"
top: "loss"
}
I0522 17:33:05.763185 4784 layer_factory.hpp:77] Creating layer data
I0522 17:33:05.763190 4784 net.cpp:91] Creating Layer data
I0522 17:33:05.763192 4784 net.cpp:399] data -> data
I0522 17:33:05.763197 4784 net.cpp:399] data -> label
I0522 17:33:05.763201 4784 hdf5_data_layer.cpp:79] Loading list of HDF5 filenames from: /home/menglin/caffe-master/menglin_try/test.txt
I0522 17:33:05.763216 4784 hdf5_data_layer.cpp:93] Number of HDF5 files: 1
I0522 17:33:06.088132 4784 net.cpp:141] Setting up data
I0522 17:33:06.088165 4784 net.cpp:148] Top shape: 500 3 255 255 (97537500)
I0522 17:33:06.088171 4784 net.cpp:148] Top shape: 500 1 (500)
I0522 17:33:06.088172 4784 net.cpp:156] Memory required for data: 390152000
I0522 17:33:06.088177 4784 layer_factory.hpp:77] Creating layer label_data_1_split
I0522 17:33:06.088188 4784 net.cpp:91] Creating Layer label_data_1_split
I0522 17:33:06.088191 4784 net.cpp:425] label_data_1_split <- label
I0522 17:33:06.088207 4784 net.cpp:399] label_data_1_split -> label_data_1_split_0
I0522 17:33:06.088215 4784 net.cpp:399] label_data_1_split -> label_data_1_split_1
I0522 17:33:06.088245 4784 net.cpp:141] Setting up label_data_1_split
I0522 17:33:06.088249 4784 net.cpp:148] Top shape: 500 1 (500)
I0522 17:33:06.088269 4784 net.cpp:148] Top shape: 500 1 (500)
I0522 17:33:06.088271 4784 net.cpp:156] Memory required for data: 390156000
I0522 17:33:06.088274 4784 layer_factory.hpp:77] Creating layer conv1
I0522 17:33:06.088285 4784 net.cpp:91] Creating Layer conv1
I0522 17:33:06.088287 4784 net.cpp:425] conv1 <- data
I0522 17:33:06.088300 4784 net.cpp:399] conv1 -> conv1
I0522 17:33:06.090306 4784 net.cpp:141] Setting up conv1
I0522 17:33:06.090312 4784 net.cpp:148] Top shape: 500 32 33 33 (17424000)
I0522 17:33:06.090313 4784 net.cpp:156] Memory required for data: 459852000
I0522 17:33:06.090320 4784 layer_factory.hpp:77] Creating layer conv1_conv1_0_split
I0522 17:33:06.090324 4784 net.cpp:91] Creating Layer conv1_conv1_0_split
I0522 17:33:06.090327 4784 net.cpp:425] conv1_conv1_0_split <- conv1
I0522 17:33:06.090329 4784 net.cpp:399] conv1_conv1_0_split -> conv1_conv1_0_split_0
I0522 17:33:06.090343 4784 net.cpp:399] conv1_conv1_0_split -> conv1_conv1_0_split_1
I0522 17:33:06.090360 4784 net.cpp:141] Setting up conv1_conv1_0_split
I0522 17:33:06.090365 4784 net.cpp:148] Top shape: 500 32 33 33 (17424000)
I0522 17:33:06.090368 4784 net.cpp:148] Top shape: 500 32 33 33 (17424000)
I0522 17:33:06.090369 4784 net.cpp:156] Memory required for data: 599244000
I0522 17:33:06.090371 4784 layer_factory.hpp:77] Creating layer res1/bn1
I0522 17:33:06.090376 4784 net.cpp:91] Creating Layer res1/bn1
I0522 17:33:06.090378 4784 net.cpp:425] res1/bn1 <- conv1_conv1_0_split_0
I0522 17:33:06.090381 4784 net.cpp:399] res1/bn1 -> res1/bn1
I0522 17:33:06.090497 4784 net.cpp:141] Setting up res1/bn1
I0522 17:33:06.090500 4784 net.cpp:148] Top shape: 500 32 33 33 (17424000)
I0522 17:33:06.090502 4784 net.cpp:156] Memory required for data: 668940000
I0522 17:33:06.090508 4784 layer_factory.hpp:77] Creating layer res1/relu1
I0522 17:33:06.090512 4784 net.cpp:91] Creating Layer res1/relu1
I0522 17:33:06.090514 4784 net.cpp:425] res1/relu1 <- res1/bn1
I0522 17:33:06.090517 4784 net.cpp:386] res1/relu1 -> res1/bn1 (in-place)
I0522 17:33:06.090530 4784 net.cpp:141] Setting up res1/relu1
I0522 17:33:06.090533 4784 net.cpp:148] Top shape: 500 32 33 33 (17424000)
I0522 17:33:06.090535 4784 net.cpp:156] Memory required for data: 738636000
I0522 17:33:06.090538 4784 layer_factory.hpp:77] Creating layer res1/conv1
I0522 17:33:06.090543 4784 net.cpp:91] Creating Layer res1/conv1
I0522 17:33:06.090544 4784 net.cpp:425] res1/conv1 <- res1/bn1
I0522 17:33:06.090548 4784 net.cpp:399] res1/conv1 -> res1/conv1
I0522 17:33:06.091174 4784 net.cpp:141] Setting up res1/conv1
I0522 17:33:06.091179 4784 net.cpp:148] Top shape: 500 32 33 33 (17424000)
I0522 17:33:06.091181 4784 net.cpp:156] Memory required for data: 808332000
I0522 17:33:06.091184 4784 layer_factory.hpp:77] Creating layer res1/bn2
I0522 17:33:06.091188 4784 net.cpp:91] Creating Layer res1/bn2
I0522 17:33:06.091190 4784 net.cpp:425] res1/bn2 <- res1/conv1
I0522 17:33:06.091193 4784 net.cpp:399] res1/bn2 -> res1/bn2
I0522 17:33:06.091311 4784 net.cpp:141] Setting up res1/bn2
I0522 17:33:06.091315 4784 net.cpp:148] Top shape: 500 32 33 33 (17424000)
I0522 17:33:06.091317 4784 net.cpp:156] Memory required for data: 878028000
I0522 17:33:06.091323 4784 layer_factory.hpp:77] Creating layer res1/relu2
I0522 17:33:06.091327 4784 net.cpp:91] Creating Layer res1/relu2
I0522 17:33:06.091330 4784 net.cpp:425] res1/relu2 <- res1/bn2
I0522 17:33:06.091331 4784 net.cpp:386] res1/relu2 -> res1/bn2 (in-place)
I0522 17:33:06.091346 4784 net.cpp:141] Setting up res1/relu2
I0522 17:33:06.091348 4784 net.cpp:148] Top shape: 500 32 33 33 (17424000)
I0522 17:33:06.091349 4784 net.cpp:156] Memory required for data: 947724000
I0522 17:33:06.091351 4784 layer_factory.hpp:77] Creating layer res1/conv2
I0522 17:33:06.091356 4784 net.cpp:91] Creating Layer res1/conv2
I0522 17:33:06.091358 4784 net.cpp:425] res1/conv2 <- res1/bn2
I0522 17:33:06.091361 4784 net.cpp:399] res1/conv2 -> res1/conv2
I0522 17:33:06.091980 4784 net.cpp:141] Setting up res1/conv2
I0522 17:33:06.091990 4784 net.cpp:148] Top shape: 500 32 33 33 (17424000)
I0522 17:33:06.091992 4784 net.cpp:156] Memory required for data: 1017420000
I0522 17:33:06.091996 4784 layer_factory.hpp:77] Creating layer res1/elt
I0522 17:33:06.092000 4784 net.cpp:91] Creating Layer res1/elt
I0522 17:33:06.092002 4784 net.cpp:425] res1/elt <- res1/conv2
I0522 17:33:06.092005 4784 net.cpp:425] res1/elt <- conv1_conv1_0_split_1
I0522 17:33:06.092007 4784 net.cpp:399] res1/elt -> res1/elt
I0522 17:33:06.092021 4784 net.cpp:141] Setting up res1/elt
I0522 17:33:06.092025 4784 net.cpp:148] Top shape: 500 32 33 33 (17424000)
I0522 17:33:06.092026 4784 net.cpp:156] Memory required for data: 1087116000
I0522 17:33:06.092028 4784 layer_factory.hpp:77] Creating layer pool1
I0522 17:33:06.092032 4784 net.cpp:91] Creating Layer pool1
I0522 17:33:06.092034 4784 net.cpp:425] pool1 <- res1/elt
I0522 17:33:06.092037 4784 net.cpp:399] pool1 -> pool1
I0522 17:33:06.092054 4784 net.cpp:141] Setting up pool1
I0522 17:33:06.092058 4784 net.cpp:148] Top shape: 500 32 16 16 (4096000)
I0522 17:33:06.092061 4784 net.cpp:156] Memory required for data: 1103500000
I0522 17:33:06.092061 4784 layer_factory.hpp:77] Creating layer pool1_pool1_0_split
I0522 17:33:06.092064 4784 net.cpp:91] Creating Layer pool1_pool1_0_split
I0522 17:33:06.092067 4784 net.cpp:425] pool1_pool1_0_split <- pool1
I0522 17:33:06.092069 4784 net.cpp:399] pool1_pool1_0_split -> pool1_pool1_0_split_0
I0522 17:33:06.092072 4784 net.cpp:399] pool1_pool1_0_split -> pool1_pool1_0_split_1
I0522 17:33:06.092087 4784 net.cpp:141] Setting up pool1_pool1_0_split
I0522 17:33:06.092092 4784 net.cpp:148] Top shape: 500 32 16 16 (4096000)
I0522 17:33:06.092093 4784 net.cpp:148] Top shape: 500 32 16 16 (4096000)
I0522 17:33:06.092095 4784 net.cpp:156] Memory required for data: 1136268000
I0522 17:33:06.092097 4784 layer_factory.hpp:77] Creating layer res2/bn1
I0522 17:33:06.092100 4784 net.cpp:91] Creating Layer res2/bn1
I0522 17:33:06.092103 4784 net.cpp:425] res2/bn1 <- pool1_pool1_0_split_0
I0522 17:33:06.092105 4784 net.cpp:399] res2/bn1 -> res2/bn1
I0522 17:33:06.092200 4784 net.cpp:141] Setting up res2/bn1
I0522 17:33:06.092205 4784 net.cpp:148] Top shape: 500 32 16 16 (4096000)
I0522 17:33:06.092206 4784 net.cpp:156] Memory required for data: 1152652000
I0522 17:33:06.092211 4784 layer_factory.hpp:77] Creating layer res2/relu1
I0522 17:33:06.092213 4784 net.cpp:91] Creating Layer res2/relu1
I0522 17:33:06.092216 4784 net.cpp:425] res2/relu1 <- res2/bn1
I0522 17:33:06.092218 4784 net.cpp:386] res2/relu1 -> res2/bn1 (in-place)
I0522 17:33:06.092221 4784 net.cpp:141] Setting up res2/relu1
I0522 17:33:06.092224 4784 net.cpp:148] Top shape: 500 32 16 16 (4096000)
I0522 17:33:06.092226 4784 net.cpp:156] Memory required for data: 1169036000
I0522 17:33:06.092227 4784 layer_factory.hpp:77] Creating layer res2/conv1
I0522 17:33:06.092233 4784 net.cpp:91] Creating Layer res2/conv1
I0522 17:33:06.092236 4784 net.cpp:425] res2/conv1 <- res2/bn1
I0522 17:33:06.092238 4784 net.cpp:399] res2/conv1 -> res2/conv1
I0522 17:33:06.093336 4784 net.cpp:141] Setting up res2/conv1
I0522 17:33:06.093343 4784 net.cpp:148] Top shape: 500 32 16 16 (4096000)
I0522 17:33:06.093344 4784 net.cpp:156] Memory required for data: 1185420000
I0522 17:33:06.093349 4784 layer_factory.hpp:77] Creating layer res2/bn2
I0522 17:33:06.093354 4784 net.cpp:91] Creating Layer res2/bn2
I0522 17:33:06.093356 4784 net.cpp:425] res2/bn2 <- res2/conv1
I0522 17:33:06.093359 4784 net.cpp:399] res2/bn2 -> res2/bn2
I0522 17:33:06.093454 4784 net.cpp:141] Setting up res2/bn2
I0522 17:33:06.093458 4784 net.cpp:148] Top shape: 500 32 16 16 (4096000)
I0522 17:33:06.093461 4784 net.cpp:156] Memory required for data: 1201804000
I0522 17:33:06.093464 4784 layer_factory.hpp:77] Creating layer res2/relu2
I0522 17:33:06.093468 4784 net.cpp:91] Creating Layer res2/relu2
I0522 17:33:06.093471 4784 net.cpp:425] res2/relu2 <- res2/bn2
I0522 17:33:06.093473 4784 net.cpp:386] res2/relu2 -> res2/bn2 (in-place)
I0522 17:33:06.093482 4784 net.cpp:141] Setting up res2/relu2
I0522 17:33:06.093484 4784 net.cpp:148] Top shape: 500 32 16 16 (4096000)
I0522 17:33:06.093487 4784 net.cpp:156] Memory required for data: 1218188000
I0522 17:33:06.093488 4784 layer_factory.hpp:77] Creating layer res2/conv2
I0522 17:33:06.093493 4784 net.cpp:91] Creating Layer res2/conv2
I0522 17:33:06.093495 4784 net.cpp:425] res2/conv2 <- res2/bn2
I0522 17:33:06.093498 4784 net.cpp:399] res2/conv2 -> res2/conv2
I0522 17:33:06.094894 4784 net.cpp:141] Setting up res2/conv2
I0522 17:33:06.094902 4784 net.cpp:148] Top shape: 500 32 16 16 (4096000)
I0522 17:33:06.094904 4784 net.cpp:156] Memory required for data: 1234572000
I0522 17:33:06.094908 4784 layer_factory.hpp:77] Creating layer res2/elt
I0522 17:33:06.094913 4784 net.cpp:91] Creating Layer res2/elt
I0522 17:33:06.094914 4784 net.cpp:425] res2/elt <- res2/conv2
I0522 17:33:06.094918 4784 net.cpp:425] res2/elt <- pool1_pool1_0_split_1
I0522 17:33:06.094920 4784 net.cpp:399] res2/elt -> res2/elt
I0522 17:33:06.094930 4784 net.cpp:141] Setting up res2/elt
I0522 17:33:06.094933 4784 net.cpp:148] Top shape: 500 32 16 16 (4096000)
I0522 17:33:06.094935 4784 net.cpp:156] Memory required for data: 1250956000
I0522 17:33:06.094938 4784 layer_factory.hpp:77] Creating layer pool2
I0522 17:33:06.094941 4784 net.cpp:91] Creating Layer pool2
I0522 17:33:06.094944 4784 net.cpp:425] pool2 <- res2/elt
I0522 17:33:06.094945 4784 net.cpp:399] pool2 -> pool2
I0522 17:33:06.094964 4784 net.cpp:141] Setting up pool2
I0522 17:33:06.094966 4784 net.cpp:148] Top shape: 500 32 8 8 (1024000)
I0522 17:33:06.094969 4784 net.cpp:156] Memory required for data: 1255052000
I0522 17:33:06.094970 4784 layer_factory.hpp:77] Creating layer ip1
I0522 17:33:06.094974 4784 net.cpp:91] Creating Layer ip1
I0522 17:33:06.094976 4784 net.cpp:425] ip1 <- pool2
I0522 17:33:06.094980 4784 net.cpp:399] ip1 -> ip1
I0522 17:33:06.137181 4784 net.cpp:141] Setting up ip1
I0522 17:33:06.137202 4784 net.cpp:148] Top shape: 500 1000 (500000)
I0522 17:33:06.137204 4784 net.cpp:156] Memory required for data: 1257052000
I0522 17:33:06.137212 4784 layer_factory.hpp:77] Creating layer bn5
I0522 17:33:06.137222 4784 net.cpp:91] Creating Layer bn5
I0522 17:33:06.137235 4784 net.cpp:425] bn5 <- ip1
I0522 17:33:06.137241 4784 net.cpp:399] bn5 -> bn5
I0522 17:33:06.137384 4784 net.cpp:141] Setting up bn5
I0522 17:33:06.137388 4784 net.cpp:148] Top shape: 500 1000 (500000)
I0522 17:33:06.137390 4784 net.cpp:156] Memory required for data: 1259052000
I0522 17:33:06.137395 4784 layer_factory.hpp:77] Creating layer relu4
I0522 17:33:06.137399 4784 net.cpp:91] Creating Layer relu4
I0522 17:33:06.137403 4784 net.cpp:425] relu4 <- bn5
I0522 17:33:06.137404 4784 net.cpp:386] relu4 -> bn5 (in-place)
I0522 17:33:06.137418 4784 net.cpp:141] Setting up relu4
I0522 17:33:06.137421 4784 net.cpp:148] Top shape: 500 1000 (500000)
I0522 17:33:06.137423 4784 net.cpp:156] Memory required for data: 1261052000
I0522 17:33:06.137425 4784 layer_factory.hpp:77] Creating layer ip2
I0522 17:33:06.137431 4784 net.cpp:91] Creating Layer ip2
I0522 17:33:06.137434 4784 net.cpp:425] ip2 <- bn5
I0522 17:33:06.137436 4784 net.cpp:399] ip2 -> ip2
I0522 17:33:06.137605 4784 net.cpp:141] Setting up ip2
I0522 17:33:06.137610 4784 net.cpp:148] Top shape: 500 5 (2500)
I0522 17:33:06.137611 4784 net.cpp:156] Memory required for data: 1261062000
I0522 17:33:06.137615 4784 layer_factory.hpp:77] Creating layer ip2_ip2_0_split
I0522 17:33:06.137619 4784 net.cpp:91] Creating Layer ip2_ip2_0_split
I0522 17:33:06.137621 4784 net.cpp:425] ip2_ip2_0_split <- ip2
I0522 17:33:06.137624 4784 net.cpp:399] ip2_ip2_0_split -> ip2_ip2_0_split_0
I0522 17:33:06.137627 4784 net.cpp:399] ip2_ip2_0_split -> ip2_ip2_0_split_1
I0522 17:33:06.137655 4784 net.cpp:141] Setting up ip2_ip2_0_split
I0522 17:33:06.137660 4784 net.cpp:148] Top shape: 500 5 (2500)
I0522 17:33:06.137662 4784 net.cpp:148] Top shape: 500 5 (2500)
I0522 17:33:06.137665 4784 net.cpp:156] Memory required for data: 1261082000
I0522 17:33:06.137676 4784 layer_factory.hpp:77] Creating layer accuracy
I0522 17:33:06.137681 4784 net.cpp:91] Creating Layer accuracy
I0522 17:33:06.137682 4784 net.cpp:425] accuracy <- ip2_ip2_0_split_0
I0522 17:33:06.137686 4784 net.cpp:425] accuracy <- label_data_1_split_0
I0522 17:33:06.137688 4784 net.cpp:399] accuracy -> accuracy
I0522 17:33:06.137694 4784 net.cpp:141] Setting up accuracy
I0522 17:33:06.137698 4784 net.cpp:148] Top shape: (1)
I0522 17:33:06.137701 4784 net.cpp:156] Memory required for data: 1261082004
I0522 17:33:06.137701 4784 layer_factory.hpp:77] Creating layer loss
I0522 17:33:06.137706 4784 net.cpp:91] Creating Layer loss
I0522 17:33:06.137707 4784 net.cpp:425] loss <- ip2_ip2_0_split_1
I0522 17:33:06.137709 4784 net.cpp:425] loss <- label_data_1_split_1
I0522 17:33:06.137712 4784 net.cpp:399] loss -> loss
I0522 17:33:06.137718 4784 layer_factory.hpp:77] Creating layer loss
I0522 17:33:06.137783 4784 net.cpp:141] Setting up loss
I0522 17:33:06.137786 4784 net.cpp:148] Top shape: (1)
I0522 17:33:06.137789 4784 net.cpp:151] with loss weight 1
I0522 17:33:06.137800 4784 net.cpp:156] Memory required for data: 1261082008
I0522 17:33:06.137802 4784 net.cpp:217] loss needs backward computation.
I0522 17:33:06.137805 4784 net.cpp:219] accuracy does not need backward computation.
I0522 17:33:06.137817 4784 net.cpp:217] ip2_ip2_0_split needs backward computation.
I0522 17:33:06.137820 4784 net.cpp:217] ip2 needs backward computation.
I0522 17:33:06.137821 4784 net.cpp:217] relu4 needs backward computation.
I0522 17:33:06.137823 4784 net.cpp:217] bn5 needs backward computation.
I0522 17:33:06.137825 4784 net.cpp:217] ip1 needs backward computation.
I0522 17:33:06.137827 4784 net.cpp:217] pool2 needs backward computation.
I0522 17:33:06.137830 4784 net.cpp:217] res2/elt needs backward computation.
I0522 17:33:06.137832 4784 net.cpp:217] res2/conv2 needs backward computation.
I0522 17:33:06.137835 4784 net.cpp:217] res2/relu2 needs backward computation.
I0522 17:33:06.137836 4784 net.cpp:217] res2/bn2 needs backward computation.
I0522 17:33:06.137838 4784 net.cpp:217] res2/conv1 needs backward computation.
I0522 17:33:06.137841 4784 net.cpp:217] res2/relu1 needs backward computation.
I0522 17:33:06.137843 4784 net.cpp:217] res2/bn1 needs backward computation.
I0522 17:33:06.137845 4784 net.cpp:217] pool1_pool1_0_split needs backward computation.
I0522 17:33:06.137847 4784 net.cpp:217] pool1 needs backward computation.
I0522 17:33:06.137850 4784 net.cpp:217] res1/elt needs backward computation.
I0522 17:33:06.137852 4784 net.cpp:217] res1/conv2 needs backward computation.
I0522 17:33:06.137855 4784 net.cpp:217] res1/relu2 needs backward computation.
I0522 17:33:06.137857 4784 net.cpp:217] res1/bn2 needs backward computation.
I0522 17:33:06.137859 4784 net.cpp:217] res1/conv1 needs backward computation.
I0522 17:33:06.137861 4784 net.cpp:217] res1/relu1 needs backward computation.
I0522 17:33:06.137872 4784 net.cpp:217] res1/bn1 needs backward computation.
I0522 17:33:06.137874 4784 net.cpp:217] conv1_conv1_0_split needs backward computation.
I0522 17:33:06.137876 4784 net.cpp:217] conv1 needs backward computation.
I0522 17:33:06.137878 4784 net.cpp:219] label_data_1_split does not need backward computation.
I0522 17:33:06.137881 4784 net.cpp:219] data does not need backward computation.
I0522 17:33:06.137883 4784 net.cpp:261] This network produces output accuracy
I0522 17:33:06.137886 4784 net.cpp:261] This network produces output loss
I0522 17:33:06.137897 4784 net.cpp:274] Network initialization done.
I0522 17:33:06.137982 4784 solver.cpp:60] Solver scaffolding done.
I0522 17:33:06.138679 4784 caffe.cpp:219] Starting Optimization
I0522 17:33:06.138684 4784 solver.cpp:279] Solving mengNet
I0522 17:33:06.138685 4784 solver.cpp:280] Learning Rate Policy: fixed
I0522 17:33:06.139313 4784 solver.cpp:337] Iteration 0, Testing net (#0)
I0522 17:33:07.191285 4784 solver.cpp:404] Test net output #0: accuracy = 0.111
I0522 17:33:07.191323 4784 solver.cpp:404] Test net output #1: loss = 31.5449 (* 1 = 31.5449 loss)
I0522 17:33:08.166841 4784 solver.cpp:228] Iteration 0, loss = 1.62908
I0522 17:33:08.166858 4784 solver.cpp:244] Train net output #0: accuracy = 0.18
I0522 17:33:08.166865 4784 solver.cpp:244] Train net output #1: loss = 1.62908 (* 1 = 1.62908 loss)
I0522 17:33:08.166869 4784 sgd_solver.cpp:106] Iteration 0, lr = 0.001
I0522 17:33:33.155853 4784 solver.cpp:228] Iteration 20, loss = 1.01674
I0522 17:33:33.155987 4784 solver.cpp:244] Train net output #0: accuracy = 0.61
I0522 17:33:33.155997 4784 solver.cpp:244] Train net output #1: loss = 1.01674 (* 1 = 1.01674 loss)
I0522 17:33:33.156000 4784 sgd_solver.cpp:106] Iteration 20, lr = 0.001
I0522 17:33:58.819301 4784 solver.cpp:228] Iteration 40, loss = 1.07871
I0522 17:33:58.819347 4784 solver.cpp:244] Train net output #0: accuracy = 0.65
I0522 17:33:58.819355 4784 solver.cpp:244] Train net output #1: loss = 1.07871 (* 1 = 1.07871 loss)
I0522 17:33:58.819360 4784 sgd_solver.cpp:106] Iteration 40, lr = 0.001
I0522 17:34:24.790990 4784 solver.cpp:228] Iteration 60, loss = 0.957721
I0522 17:34:24.791084 4784 solver.cpp:244] Train net output #0: accuracy = 0.584
I0522 17:34:24.791102 4784 solver.cpp:244] Train net output #1: loss = 0.957721 (* 1 = 0.957721 loss)
I0522 17:34:24.791106 4784 sgd_solver.cpp:106] Iteration 60, lr = 0.001
I0522 17:34:51.278635 4784 solver.cpp:228] Iteration 80, loss = 0.765386
I0522 17:34:51.278659 4784 solver.cpp:244] Train net output #0: accuracy = 0.67
I0522 17:34:51.278666 4784 solver.cpp:244] Train net output #1: loss = 0.765386 (* 1 = 0.765386 loss)
I0522 17:34:51.278669 4784 sgd_solver.cpp:106] Iteration 80, lr = 0.001
I0522 17:35:17.271759 4784 solver.cpp:337] Iteration 100, Testing net (#0)
I0522 17:35:18.389349 4784 solver.cpp:404] Test net output #0: accuracy = 0.427
I0522 17:35:18.389376 4784 solver.cpp:404] Test net output #1: loss = 1.18771 (* 1 = 1.18771 loss)
I0522 17:35:19.464973 4784 solver.cpp:228] Iteration 100, loss = 0.760981
I0522 17:35:19.464998 4784 solver.cpp:244] Train net output #0: accuracy = 0.702
I0522 17:35:19.465005 4784 solver.cpp:244] Train net output #1: loss = 0.760981 (* 1 = 0.760981 loss)
I0522 17:35:19.465008 4784 sgd_solver.cpp:106] Iteration 100, lr = 0.001
I0522 17:35:46.555075 4784 solver.cpp:228] Iteration 120, loss = 0.959433
I0522 17:35:46.555099 4784 solver.cpp:244] Train net output #0: accuracy = 0.678
I0522 17:35:46.555105 4784 solver.cpp:244] Train net output #1: loss = 0.959433 (* 1 = 0.959433 loss)
I0522 17:35:46.555109 4784 sgd_solver.cpp:106] Iteration 120, lr = 0.001
I0522 17:36:12.798771 4784 solver.cpp:228] Iteration 140, loss = 0.807995
I0522 17:36:12.798897 4784 solver.cpp:244] Train net output #0: accuracy = 0.662
I0522 17:36:12.798907 4784 solver.cpp:244] Train net output #1: loss = 0.807995 (* 1 = 0.807995 loss)
I0522 17:36:12.798912 4784 sgd_solver.cpp:106] Iteration 140, lr = 0.001
I0522 17:36:39.076234 4784 solver.cpp:228] Iteration 160, loss = 0.755289
I0522 17:36:39.076258 4784 solver.cpp:244] Train net output #0: accuracy = 0.676
I0522 17:36:39.076264 4784 solver.cpp:244] Train net output #1: loss = 0.755289 (* 1 = 0.755289 loss)
I0522 17:36:39.076268 4784 sgd_solver.cpp:106] Iteration 160, lr = 0.001
I0522 17:37:05.253271 4784 solver.cpp:228] Iteration 180, loss = 0.748647
I0522 17:37:05.253399 4784 solver.cpp:244] Train net output #0: accuracy = 0.716
I0522 17:37:05.253408 4784 solver.cpp:244] Train net output #1: loss = 0.748647 (* 1 = 0.748647 loss)
I0522 17:37:05.253412 4784 sgd_solver.cpp:106] Iteration 180, lr = 0.001
I0522 17:37:30.447172 4784 solver.cpp:337] Iteration 200, Testing net (#0)
I0522 17:37:31.528594 4784 solver.cpp:404] Test net output #0: accuracy = 0.425
I0522 17:37:31.528619 4784 solver.cpp:404] Test net output #1: loss = 1.76517 (* 1 = 1.76517 loss)
I0522 17:37:32.550200 4784 solver.cpp:228] Iteration 200, loss = 1.04188
I0522 17:37:32.550221 4784 solver.cpp:244] Train net output #0: accuracy = 0.674
I0522 17:37:32.550228 4784 solver.cpp:244] Train net output #1: loss = 1.04188 (* 1 = 1.04188 loss)
I0522 17:37:32.550231 4784 sgd_solver.cpp:106] Iteration 200, lr = 0.001
I0522 17:37:58.518641 4784 solver.cpp:228] Iteration 220, loss = 0.860798
I0522 17:37:58.518751 4784 solver.cpp:244] Train net output #0: accuracy = 0.638
I0522 17:37:58.518769 4784 solver.cpp:244] Train net output #1: loss = 0.860798 (* 1 = 0.860798 loss)
I0522 17:37:58.518772 4784 sgd_solver.cpp:106] Iteration 220, lr = 0.001
I0522 17:38:24.644518 4784 solver.cpp:228] Iteration 240, loss = 0.831552
I0522 17:38:24.644539 4784 solver.cpp:244] Train net output #0: accuracy = 0.656
I0522 17:38:24.644546 4784 solver.cpp:244] Train net output #1: loss = 0.831552 (* 1 = 0.831552 loss)
I0522 17:38:24.644548 4784 sgd_solver.cpp:106] Iteration 240, lr = 0.001
I0522 17:38:50.661337 4784 solver.cpp:228] Iteration 260, loss = 0.917817
I0522 17:38:50.661470 4784 solver.cpp:244] Train net output #0: accuracy = 0.656
I0522 17:38:50.661480 4784 solver.cpp:244] Train net output #1: loss = 0.917817 (* 1 = 0.917817 loss)
I0522 17:38:50.661484 4784 sgd_solver.cpp:106] Iteration 260, lr = 0.001
I0522 17:39:16.566133 4784 solver.cpp:228] Iteration 280, loss = 1.14495
I0522 17:39:16.566154 4784 solver.cpp:244] Train net output #0: accuracy = 0.628
I0522 17:39:16.566160 4784 solver.cpp:244] Train net output #1: loss = 1.14495 (* 1 = 1.14495 loss)
I0522 17:39:16.566164 4784 sgd_solver.cpp:106] Iteration 280, lr = 0.001
I0522 17:39:41.512784 4784 solver.cpp:337] Iteration 300, Testing net (#0)
I0522 17:39:42.592100 4784 solver.cpp:404] Test net output #0: accuracy = 0.471
I0522 17:39:42.592125 4784 solver.cpp:404] Test net output #1: loss = 1.59105 (* 1 = 1.59105 loss)
I0522 17:39:43.611470 4784 solver.cpp:228] Iteration 300, loss = 1.03568
I0522 17:39:43.611492 4784 solver.cpp:244] Train net output #0: accuracy = 0.564
I0522 17:39:43.611498 4784 solver.cpp:244] Train net output #1: loss = 1.03568 (* 1 = 1.03568 loss)
I0522 17:39:43.611502 4784 sgd_solver.cpp:106] Iteration 300, lr = 0.001
I0522 17:40:09.526494 4784 solver.cpp:228] Iteration 320, loss = 0.987076
I0522 17:40:09.526515 4784 solver.cpp:244] Train net output #0: accuracy = 0.572
I0522 17:40:09.526521 4784 solver.cpp:244] Train net output #1: loss = 0.987076 (* 1 = 0.987076 loss)
I0522 17:40:09.526525 4784 sgd_solver.cpp:106] Iteration 320, lr = 0.001
I0522 17:40:35.481556 4784 solver.cpp:228] Iteration 340, loss = 1.01115
I0522 17:40:35.481685 4784 solver.cpp:244] Train net output #0: accuracy = 0.644
I0522 17:40:35.481695 4784 solver.cpp:244] Train net output #1: loss = 1.01115 (* 1 = 1.01115 loss)
I0522 17:40:35.481698 4784 sgd_solver.cpp:106] Iteration 340, lr = 0.001
I0522 17:41:01.751333 4784 solver.cpp:228] Iteration 360, loss = 1.13727
I0522 17:41:01.751355 4784 solver.cpp:244] Train net output #0: accuracy = 0.584
I0522 17:41:01.751363 4784 solver.cpp:244] Train net output #1: loss = 1.13727 (* 1 = 1.13727 loss)
I0522 17:41:01.751366 4784 sgd_solver.cpp:106] Iteration 360, lr = 0.001
I0522 17:41:27.686775 4784 solver.cpp:228] Iteration 380, loss = 0.974676
I0522 17:41:27.686899 4784 solver.cpp:244] Train net output #0: accuracy = 0.552
I0522 17:41:27.686909 4784 solver.cpp:244] Train net output #1: loss = 0.974676 (* 1 = 0.974676 loss)
I0522 17:41:27.686913 4784 sgd_solver.cpp:106] Iteration 380, lr = 0.001
I0522 17:41:52.615531 4784 solver.cpp:337] Iteration 400, Testing net (#0)
I0522 17:41:53.691292 4784 solver.cpp:404] Test net output #0: accuracy = 0.447
I0522 17:41:53.691315 4784 solver.cpp:404] Test net output #1: loss = 1.1416 (* 1 = 1.1416 loss)
I0522 17:41:54.701591 4784 solver.cpp:228] Iteration 400, loss = 0.990989
I0522 17:41:54.701613 4784 solver.cpp:244] Train net output #0: accuracy = 0.564
I0522 17:41:54.701620 4784 solver.cpp:244] Train net output #1: loss = 0.990989 (* 1 = 0.990989 loss)
I0522 17:41:54.701623 4784 sgd_solver.cpp:106] Iteration 400, lr = 0.001
I0522 17:42:20.633812 4784 solver.cpp:228] Iteration 420, loss = 0.881144
I0522 17:42:20.634812 4784 solver.cpp:244] Train net output #0: accuracy = 0.632
I0522 17:42:20.634822 4784 solver.cpp:244] Train net output #1: loss = 0.881144 (* 1 = 0.881144 loss)
I0522 17:42:20.634826 4784 sgd_solver.cpp:106] Iteration 420, lr = 0.001
I0522 17:42:46.588696 4784 solver.cpp:228] Iteration 440, loss = 1.19581
I0522 17:42:46.588719 4784 solver.cpp:244] Train net output #0: accuracy = 0.56
I0522 17:42:46.588726 4784 solver.cpp:244] Train net output #1: loss = 1.19581 (* 1 = 1.19581 loss)
I0522 17:42:46.588729 4784 sgd_solver.cpp:106] Iteration 440, lr = 0.001
I0522 17:43:12.478281 4784 solver.cpp:228] Iteration 460, loss = 1.02179
I0522 17:43:12.478423 4784 solver.cpp:244] Train net output #0: accuracy = 0.524
I0522 17:43:12.478433 4784 solver.cpp:244] Train net output #1: loss = 1.02179 (* 1 = 1.02179 loss)
I0522 17:43:12.478437 4784 sgd_solver.cpp:106] Iteration 460, lr = 0.001
I0522 17:43:38.443531 4784 solver.cpp:228] Iteration 480, loss = 1.01613
I0522 17:43:38.443555 4784 solver.cpp:244] Train net output #0: accuracy = 0.55
I0522 17:43:38.443562 4784 solver.cpp:244] Train net output #1: loss = 1.01613 (* 1 = 1.01613 loss)
I0522 17:43:38.443564 4784 sgd_solver.cpp:106] Iteration 480, lr = 0.001
I0522 17:44:03.331502 4784 solver.cpp:337] Iteration 500, Testing net (#0)
I0522 17:44:04.405691 4784 solver.cpp:404] Test net output #0: accuracy = 0.369
I0522 17:44:04.405716 4784 solver.cpp:404] Test net output #1: loss = 1.77825 (* 1 = 1.77825 loss)
I0522 17:44:05.420948 4784 solver.cpp:228] Iteration 500, loss = 1.02574
I0522 17:44:05.420971 4784 solver.cpp:244] Train net output #0: accuracy = 0.598
I0522 17:44:05.420979 4784 solver.cpp:244] Train net output #1: loss = 1.02574 (* 1 = 1.02574 loss)
I0522 17:44:05.420981 4784 sgd_solver.cpp:106] Iteration 500, lr = 0.001
I0522 17:44:31.456296 4784 solver.cpp:228] Iteration 520, loss = 1.13687
I0522 17:44:31.456320 4784 solver.cpp:244] Train net output #0: accuracy = 0.578
I0522 17:44:31.456326 4784 solver.cpp:244] Train net output #1: loss = 1.13687 (* 1 = 1.13687 loss)
I0522 17:44:31.456329 4784 sgd_solver.cpp:106] Iteration 520, lr = 0.001
I0522 17:44:57.392630 4784 solver.cpp:228] Iteration 540, loss = 1.02987
I0522 17:44:57.392760 4784 solver.cpp:244] Train net output #0: accuracy = 0.492
I0522 17:44:57.392771 4784 solver.cpp:244] Train net output #1: loss = 1.02987 (* 1 = 1.02987 loss)
I0522 17:44:57.392774 4784 sgd_solver.cpp:106] Iteration 540, lr = 0.001
I0522 17:45:23.431924 4784 solver.cpp:228] Iteration 560, loss = 0.93721
I0522 17:45:23.431948 4784 solver.cpp:244] Train net output #0: accuracy = 0.532
I0522 17:45:23.431954 4784 solver.cpp:244] Train net output #1: loss = 0.93721 (* 1 = 0.93721 loss)
I0522 17:45:23.431957 4784 sgd_solver.cpp:106] Iteration 560, lr = 0.001
I0522 17:45:49.399096 4784 solver.cpp:228] Iteration 580, loss = 0.982253
I0522 17:45:49.399209 4784 solver.cpp:244] Train net output #0: accuracy = 0.616
I0522 17:45:49.399219 4784 solver.cpp:244] Train net output #1: loss = 0.982253 (* 1 = 0.982253 loss)
I0522 17:45:49.399224 4784 sgd_solver.cpp:106] Iteration 580, lr = 0.001
I0522 17:46:14.355550 4784 solver.cpp:337] Iteration 600, Testing net (#0)
I0522 17:46:15.440210 4784 solver.cpp:404] Test net output #0: accuracy = 0.256
I0522 17:46:15.440234 4784 solver.cpp:404] Test net output #1: loss = 1.60479 (* 1 = 1.60479 loss)
I0522 17:46:16.464331 4784 solver.cpp:228] Iteration 600, loss = 1.0996
I0522 17:46:16.464354 4784 solver.cpp:244] Train net output #0: accuracy = 0.554
I0522 17:46:16.464360 4784 solver.cpp:244] Train net output #1: loss = 1.0996 (* 1 = 1.0996 loss)
I0522 17:46:16.464365 4784 sgd_solver.cpp:106] Iteration 600, lr = 0.001
I0522 17:46:42.423319 4784 solver.cpp:228] Iteration 620, loss = 1.01845
I0522 17:46:42.423445 4784 solver.cpp:244] Train net output #0: accuracy = 0.536
I0522 17:46:42.423454 4784 solver.cpp:244] Train net output #1: loss = 1.01845 (* 1 = 1.01845 loss)
I0522 17:46:42.423458 4784 sgd_solver.cpp:106] Iteration 620, lr = 0.001
I0522 17:47:08.422520 4784 solver.cpp:228] Iteration 640, loss = 1.01987
I0522 17:47:08.422544 4784 solver.cpp:244] Train net output #0: accuracy = 0.54
I0522 17:47:08.422551 4784 solver.cpp:244] Train net output #1: loss = 1.01987 (* 1 = 1.01987 loss)
I0522 17:47:08.422555 4784 sgd_solver.cpp:106] Iteration 640, lr = 0.001
I0522 17:47:34.423115 4784 solver.cpp:228] Iteration 660, loss = 1.05065
I0522 17:47:34.423248 4784 solver.cpp:244] Train net output #0: accuracy = 0.582
I0522 17:47:34.423257 4784 solver.cpp:244] Train net output #1: loss = 1.05065 (* 1 = 1.05065 loss)
I0522 17:47:34.423261 4784 sgd_solver.cpp:106] Iteration 660, lr = 0.001
I0522 17:48:00.368378 4784 solver.cpp:228] Iteration 680, loss = 1.31215
I0522 17:48:00.368403 4784 solver.cpp:244] Train net output #0: accuracy = 0.56
I0522 17:48:00.368410 4784 solver.cpp:244] Train net output #1: loss = 1.31215 (* 1 = 1.31215 loss)
I0522 17:48:00.368413 4784 sgd_solver.cpp:106] Iteration 680, lr = 0.001
I0522 17:48:25.286612 4784 solver.cpp:337] Iteration 700, Testing net (#0)
I0522 17:48:26.367760 4784 solver.cpp:404] Test net output #0: accuracy = 0.318
I0522 17:48:26.367784 4784 solver.cpp:404] Test net output #1: loss = 2.49873 (* 1 = 2.49873 loss)
I0522 17:48:27.389021 4784 solver.cpp:228] Iteration 700, loss = 1.1296
I0522 17:48:27.389045 4784 solver.cpp:244] Train net output #0: accuracy = 0.514
I0522 17:48:27.389051 4784 solver.cpp:244] Train net output #1: loss = 1.1296 (* 1 = 1.1296 loss)
I0522 17:48:27.389055 4784 sgd_solver.cpp:106] Iteration 700, lr = 0.001
I0522 17:48:53.397667 4784 solver.cpp:228] Iteration 720, loss = 1.0282
I0522 17:48:53.397691 4784 solver.cpp:244] Train net output #0: accuracy = 0.546
I0522 17:48:53.397698 4784 solver.cpp:244] Train net output #1: loss = 1.0282 (* 1 = 1.0282 loss)
I0522 17:48:53.397702 4784 sgd_solver.cpp:106] Iteration 720, lr = 0.001
I0522 17:49:19.439896 4784 solver.cpp:228] Iteration 740, loss = 1.02349
I0522 17:49:19.440012 4784 solver.cpp:244] Train net output #0: accuracy = 0.57
I0522 17:49:19.440033 4784 solver.cpp:244] Train net output #1: loss = 1.02349 (* 1 = 1.02349 loss)
I0522 17:49:19.440037 4784 sgd_solver.cpp:106] Iteration 740, lr = 0.001
I0522 17:49:45.490361 4784 solver.cpp:228] Iteration 760, loss = 1.24412
I0522 17:49:45.490386 4784 solver.cpp:244] Train net output #0: accuracy = 0.55
I0522 17:49:45.490392 4784 solver.cpp:244] Train net output #1: loss = 1.24412 (* 1 = 1.24412 loss)
I0522 17:49:45.490396 4784 sgd_solver.cpp:106] Iteration 760, lr = 0.001
I0522 17:50:11.561460 4784 solver.cpp:228] Iteration 780, loss = 1.03187
I0522 17:50:11.561590 4784 solver.cpp:244] Train net output #0: accuracy = 0.512
I0522 17:50:11.561600 4784 solver.cpp:244] Train net output #1: loss = 1.03187 (* 1 = 1.03187 loss)
I0522 17:50:11.561604 4784 sgd_solver.cpp:106] Iteration 780, lr = 0.001
I0522 17:50:36.536679 4784 solver.cpp:337] Iteration 800, Testing net (#0)
I0522 17:50:37.614876 4784 solver.cpp:404] Test net output #0: accuracy = 0.351
I0522 17:50:37.614902 4784 solver.cpp:404] Test net output #1: loss = 1.48911 (* 1 = 1.48911 loss)
I0522 17:50:38.635115 4784 solver.cpp:228] Iteration 800, loss = 1.0097
I0522 17:50:38.635138 4784 solver.cpp:244] Train net output #0: accuracy = 0.532
I0522 17:50:38.635144 4784 solver.cpp:244] Train net output #1: loss = 1.0097 (* 1 = 1.0097 loss)
I0522 17:50:38.635149 4784 sgd_solver.cpp:106] Iteration 800, lr = 0.001
I0522 17:51:04.791569 4784 solver.cpp:228] Iteration 820, loss = 0.97889
I0522 17:51:04.791723 4784 solver.cpp:244] Train net output #0: accuracy = 0.598
I0522 17:51:04.791733 4784 solver.cpp:244] Train net output #1: loss = 0.97889 (* 1 = 0.97889 loss)
I0522 17:51:04.791736 4784 sgd_solver.cpp:106] Iteration 820, lr = 0.001
I0522 17:51:30.898741 4784 solver.cpp:228] Iteration 840, loss = 1.25591
I0522 17:51:30.898775 4784 solver.cpp:244] Train net output #0: accuracy = 0.54
I0522 17:51:30.898782 4784 solver.cpp:244] Train net output #1: loss = 1.25591 (* 1 = 1.25591 loss)
I0522 17:51:30.898795 4784 sgd_solver.cpp:106] Iteration 840, lr = 0.001
I0522 17:51:57.034945 4784 solver.cpp:228] Iteration 860, loss = 1.11175
I0522 17:51:57.035073 4784 solver.cpp:244] Train net output #0: accuracy = 0.502
I0522 17:51:57.035082 4784 solver.cpp:244] Train net output #1: loss = 1.11175 (* 1 = 1.11175 loss)
I0522 17:51:57.035086 4784 sgd_solver.cpp:106] Iteration 860, lr = 0.001
I0522 17:52:23.191498 4784 solver.cpp:228] Iteration 880, loss = 1.03711
I0522 17:52:23.191524 4784 solver.cpp:244] Train net output #0: accuracy = 0.548
I0522 17:52:23.191529 4784 solver.cpp:244] Train net output #1: loss = 1.03711 (* 1 = 1.03711 loss)
I0522 17:52:23.191534 4784 sgd_solver.cpp:106] Iteration 880, lr = 0.001
I0522 17:52:48.250437 4784 solver.cpp:337] Iteration 900, Testing net (#0)
I0522 17:52:49.334789 4784 solver.cpp:404] Test net output #0: accuracy = 0.427
I0522 17:52:49.334813 4784 solver.cpp:404] Test net output #1: loss = 1.27754 (* 1 = 1.27754 loss)
I0522 17:52:50.358955 4784 solver.cpp:228] Iteration 900, loss = 1.09171
I0522 17:52:50.358978 4784 solver.cpp:244] Train net output #0: accuracy = 0.572
I0522 17:52:50.358985 4784 solver.cpp:244] Train net output #1: loss = 1.09171 (* 1 = 1.09171 loss)
I0522 17:52:50.358989 4784 sgd_solver.cpp:106] Iteration 900, lr = 0.001
I0522 17:53:16.496212 4784 solver.cpp:228] Iteration 920, loss = 1.24933
I0522 17:53:16.496235 4784 solver.cpp:244] Train net output #0: accuracy = 0.548
I0522 17:53:16.496242 4784 solver.cpp:244] Train net output #1: loss = 1.24933 (* 1 = 1.24933 loss)
I0522 17:53:16.496245 4784 sgd_solver.cpp:106] Iteration 920, lr = 0.001
I0522 17:53:42.699348 4784 solver.cpp:228] Iteration 940, loss = 1.06739
I0522 17:53:42.699460 4784 solver.cpp:244] Train net output #0: accuracy = 0.504
I0522 17:53:42.699467 4784 solver.cpp:244] Train net output #1: loss = 1.06739 (* 1 = 1.06739 loss)
I0522 17:53:42.699471 4784 sgd_solver.cpp:106] Iteration 940, lr = 0.001
I0522 17:54:08.810178 4784 solver.cpp:228] Iteration 960, loss = 1.13492
I0522 17:54:08.810201 4784 solver.cpp:244] Train net output #0: accuracy = 0.52
I0522 17:54:08.810209 4784 solver.cpp:244] Train net output #1: loss = 1.13492 (* 1 = 1.13492 loss)
I0522 17:54:08.810212 4784 sgd_solver.cpp:106] Iteration 960, lr = 0.001
I0522 17:54:34.998816 4784 solver.cpp:228] Iteration 980, loss = 1.14369
I0522 17:54:34.998940 4784 solver.cpp:244] Train net output #0: accuracy = 0.568
I0522 17:54:34.998958 4784 solver.cpp:244] Train net output #1: loss = 1.14369 (* 1 = 1.14369 loss)
I0522 17:54:34.998961 4784 sgd_solver.cpp:106] Iteration 980, lr = 0.001
I0522 17:55:00.170233 4784 solver.cpp:337] Iteration 1000, Testing net (#0)
I0522 17:55:01.256238 4784 solver.cpp:404] Test net output #0: accuracy = 0.394
I0522 17:55:01.256263 4784 solver.cpp:404] Test net output #1: loss = 1.34368 (* 1 = 1.34368 loss)
I0522 17:55:02.280261 4784 solver.cpp:228] Iteration 1000, loss = 1.29931
I0522 17:55:02.280283 4784 solver.cpp:244] Train net output #0: accuracy = 0.516
I0522 17:55:02.280289 4784 solver.cpp:244] Train net output #1: loss = 1.29931 (* 1 = 1.29931 loss)
I0522 17:55:02.280293 4784 sgd_solver.cpp:106] Iteration 1000, lr = 0.001
I0522 17:55:28.382843 4784 solver.cpp:228] Iteration 1020, loss = 1.09446
I0522 17:55:28.382931 4784 solver.cpp:244] Train net output #0: accuracy = 0.516
I0522 17:55:28.382939 4784 solver.cpp:244] Train net output #1: loss = 1.09446 (* 1 = 1.09446 loss)
I0522 17:55:28.382943 4784 sgd_solver.cpp:106] Iteration 1020, lr = 0.001
I0522 17:55:54.615752 4784 solver.cpp:228] Iteration 1040, loss = 1.08742
I0522 17:55:54.615777 4784 solver.cpp:244] Train net output #0: accuracy = 0.508
I0522 17:55:54.615782 4784 solver.cpp:244] Train net output #1: loss = 1.08742 (* 1 = 1.08742 loss)
I0522 17:55:54.615785 4784 sgd_solver.cpp:106] Iteration 1040, lr = 0.001
I0522 17:56:20.853593 4784 solver.cpp:228] Iteration 1060, loss = 1.11572
I0522 17:56:20.853746 4784 solver.cpp:244] Train net output #0: accuracy = 0.552
I0522 17:56:20.853756 4784 solver.cpp:244] Train net output #1: loss = 1.11572 (* 1 = 1.11572 loss)
I0522 17:56:20.853760 4784 sgd_solver.cpp:106] Iteration 1060, lr = 0.001
I0522 17:56:47.106473 4784 solver.cpp:228] Iteration 1080, loss = 1.40598
I0522 17:56:47.106498 4784 solver.cpp:244] Train net output #0: accuracy = 0.484
I0522 17:56:47.106504 4784 solver.cpp:244] Train net output #1: loss = 1.40598 (* 1 = 1.40598 loss)
I0522 17:56:47.106509 4784 sgd_solver.cpp:106] Iteration 1080, lr = 0.001
I0522 17:57:12.172627 4784 solver.cpp:337] Iteration 1100, Testing net (#0)
I0522 17:57:13.263082 4784 solver.cpp:404] Test net output #0: accuracy = 0.458
I0522 17:57:13.263106 4784 solver.cpp:404] Test net output #1: loss = 1.55694 (* 1 = 1.55694 loss)
I0522 17:57:14.289500 4784 solver.cpp:228] Iteration 1100, loss = 1.17419
I0522 17:57:14.289522 4784 solver.cpp:244] Train net output #0: accuracy = 0.478
I0522 17:57:14.289530 4784 solver.cpp:244] Train net output #1: loss = 1.17419 (* 1 = 1.17419 loss)
I0522 17:57:14.289532 4784 sgd_solver.cpp:106] Iteration 1100, lr = 0.001
I0522 17:57:40.568991 4784 solver.cpp:228] Iteration 1120, loss = 1.07978
I0522 17:57:40.569015 4784 solver.cpp:244] Train net output #0: accuracy = 0.492
I0522 17:57:40.569022 4784 solver.cpp:244] Train net output #1: loss = 1.07978 (* 1 = 1.07978 loss)
I0522 17:57:40.569025 4784 sgd_solver.cpp:106] Iteration 1120, lr = 0.001
I0522 17:58:06.746948 4784 solver.cpp:228] Iteration 1140, loss = 1.07904
I0522 17:58:06.747062 4784 solver.cpp:244] Train net output #0: accuracy = 0.534
I0522 17:58:06.747071 4784 solver.cpp:244] Train net output #1: loss = 1.07904 (* 1 = 1.07904 loss)
I0522 17:58:06.747076 4784 sgd_solver.cpp:106] Iteration 1140, lr = 0.001
I0522 17:58:32.890458 4784 solver.cpp:228] Iteration 1160, loss = 1.36037
I0522 17:58:32.890481 4784 solver.cpp:244] Train net output #0: accuracy = 0.504
I0522 17:58:32.890487 4784 solver.cpp:244] Train net output #1: loss = 1.36037 (* 1 = 1.36037 loss)
I0522 17:58:32.890491 4784 sgd_solver.cpp:106] Iteration 1160, lr = 0.001
I0522 17:58:59.054553 4784 solver.cpp:228] Iteration 1180, loss = 1.1291
I0522 17:58:59.054685 4784 solver.cpp:244] Train net output #0: accuracy = 0.482
I0522 17:58:59.054694 4784 solver.cpp:244] Train net output #1: loss = 1.1291 (* 1 = 1.1291 loss)
I0522 17:58:59.054698 4784 sgd_solver.cpp:106] Iteration 1180, lr = 0.001
I post the log file here..
Reply all
Reply to author
Forward
0 new messages