train_output.txt

There is 1 NVIDIA GPU available in this machine.
MemoryDataLayer dataTrain loading data: 
  179.443 MB
  name:mnist train images dim[4]={60000,1,28,28}
  234.375 KB
  name:mnist train labels dim[4]={60000,1,1,1}
MemoryDataLayer dataTest loading data: 
  29.9072 MB
  name:mnist test images dim[4]={10000,1,28,28}
  39.0625 KB
  name:mnist test labels dim[4]={10000,1,1,1}
=====================================================================================================================================
  Layers:                                                                        Responses:                                          
=====================================================================================================================================
  dataTrain
                                                                                 data[4]={100,1,28,28}
                                                                                 label[4]={100,1,1,1}
  dataTest
* conv1 weight[4]={20,1,5,5} bias[4]={1,20,1,1}
                                                                               * conv1[4]={100,20,24,24}
  pool1
                                                                               * pool1[4]={100,20,12,12}
* ip1 weight[2]={500,2880} bias[1]={500}
                                                                               * ip1[4]={100,500,1,1}
  relu1
* ip2 weight[2]={10,500} bias[1]={10}
                                                                               * prob[4]={100,10,1,1}
  prob
  loss
=====================================================================================================================================
[Net] GPU 0: Total GPU memory: 17.1944 MB
=====================================================================================================================================
[Solver] GPU 0: Total GPU memory: 28.2267 MB
All GPUs: Total GPU memory: 28.2267 MB
===========================================================================================================================================
  Training:                                                                            Testing:                                            
===========================================================================================================================================
                                                                                        Epoch 0  Iteration 0 loss = 2.31218 * 1  eval = 0.11   
Epoch 0  Iteration 0  learning_rate = 0.01 loss = 2.33974 * 1  eval = 0.12   
Epoch 0  Iteration 100  learning_rate = 0.00992565 loss = 0.782117 * 1  eval = 0.79   
                                                                                        Epoch 0  Iteration 200 loss = 0.34111 * 1  eval = 0.92   
Epoch 0  Iteration 200  learning_rate = 0.00985258 loss = 0.296047 * 1  eval = 0.91   
Epoch 0  Iteration 300  learning_rate = 0.00978075 loss = 0.373389 * 1  eval = 0.91   
                                                                                        Epoch 0  Iteration 400 loss = 0.317869 * 1  eval = 0.91   
Epoch 0  Iteration 400  learning_rate = 0.00971013 loss = 0.19207 * 1  eval = 0.98   
Epoch 0  Iteration 500  learning_rate = 0.00964069 loss = 0.293197 * 1  eval = 0.93   
                                                                                        Epoch 1  Iteration 600 loss = 0.309794 * 1  eval = 0.88   
Epoch 1  Iteration 600  learning_rate = 0.0095724 loss = 0.204257 * 1  eval = 0.94   
Epoch 1  Iteration 700  learning_rate = 0.00950522 loss = 0.223026 * 1  eval = 0.94   
                                                                                        Epoch 1  Iteration 800 loss = 0.28211 * 1  eval = 0.94   
Epoch 1  Iteration 800  learning_rate = 0.00943913 loss = 0.236288 * 1  eval = 0.92   
Epoch 1  Iteration 900  learning_rate = 0.00937411 loss = 0.419412 * 1  eval = 0.88   
                                                                                        Epoch 1  Iteration 1000 loss = 0.287435 * 1  eval = 0.9   
Epoch 1  Iteration 1000  learning_rate = 0.00931012 loss = 0.223635 * 1  eval = 0.95   
After Train 1000 Iterations Needs Time: Time passes 22.1424 seconds