batch 0
  minibatch loss: 2.11096167564
  sample 1:
    input     > [5 2 8 4 6 6 3 2]
    predicted > [7 0 0 0 0 0 0 0 0]
  sample 2:
    input     > [9 5 4 2 6 4 5 6]
    predicted > [2 2 0 0 0 0 0 0 0]
  sample 3:
    input     > [4 8 8 9 2 9 3 0]
    predicted > [0 0 0 0 0 0 0 0 0]
batch 1000
  minibatch loss: 0.0359240211546
  sample 1:
    input     > [6 5 6 4 8 6 0 0]
    predicted > [6 5 6 4 8 6 1 0 0]
  sample 2:
    input     > [5 4 2 9 3 0 0 0]
    predicted > [5 4 2 9 3 1 0 0 0]
  sample 3:
    input     > [4 2 9 0 0 0 0 0]
    predicted > [4 2 9 1 0 0 0 0 0]
batch 2000
  minibatch loss: 0.0119506847113
  sample 1:
    input     > [9 4 7 4 2 0 0 0]
    predicted > [9 4 7 4 2 1 0 0 0]
  sample 2:
    input     > [7 2 6 6 6 9 4 0]
    predicted > [7 2 6 6 6 9 4 1 0]
  sample 3:
    input     > [4 5 4 6 2 0 0 0]
    predicted > [4 5 4 6 2 1 0 0 0]
batch 3000
  minibatch loss: 0.00151053641457
  sample 1:
    input     > [8 8 3 3 6 7 7 0]
    predicted > [8 8 3 3 6 7 7 1 0]
  sample 2:
    input     > [2 8 5 7 0 0 0 0]
    predicted > [2 8 5 7 1 0 0 0 0]
  sample 3:
    input     > [4 8 9 9 6 9 8 3]
    predicted > [4 8 9 9 6 9 8 3 1]