----------ITERATION 0:

Recognition accuracy: 422136 right out of 1647232, 25.63% correct.
CV accuracy: 42520 right out of 167227, 25.43% correct.
1: Training accuracy:  799068 right out of 1480005, 53.99% correct.
   CV accuracy: 67779 right out of 167227, 40.53% correct.
2: Training accuracy:  845189 right out of 1480005, 57.11% correct.
   CV accuracy: 69405 right out of 167227, 41.50% correct.
3: Training accuracy:  856353 right out of 1480005, 57.86% correct.
   CV accuracy: 70234 right out of 167227, 42.00% correct.
4: Training accuracy:  831431 right out of 1480005, 56.18% correct.
   CV accuracy: 75748 right out of 167227, 45.30% correct.
5: Training accuracy:  817753 right out of 1480005, 55.25% correct.
   CV accuracy: 77479 right out of 167227, 46.33% correct.
6: Training accuracy:  799419 right out of 1480005, 54.01% correct.
   CV accuracy: 78906 right out of 167227, 47.18% correct.
7: Training accuracy:  774531 right out of 1480005, 52.33% correct.
   CV accuracy: 78645 right out of 167227, 47.03% correct.
Best CV accuracy: 47.18% correct in epoch 6.
Recognition accuracy: 821517 right out of 1647232, 49.87% correct.

----------ITERATION 2:
Recognition accuracy: 821517 right out of 1647232, 49.87% correct.
CV accuracy: 39696 right out of 167227, 23.74% correct.
1: Training accuracy:  847703 right out of 1480005, 57.28% correct.
   CV accuracy: 81057 right out of 167227, 48.47% correct.
2: Training accuracy:  907437 right out of 1480005, 61.31% correct.
   CV accuracy: 82758 right out of 167227, 49.49% correct.
3: Training accuracy:  922754 right out of 1480005, 62.35% correct.
   CV accuracy: 84398 right out of 167227, 50.47% correct.
4: Training accuracy:  929557 right out of 1480005, 62.81% correct.
   CV accuracy: 84296 right out of 167227, 50.41% correct.
5: Training accuracy:  922000 right out of 1480005, 62.30% correct.
   CV accuracy: 89648 right out of 167227, 53.61% correct.
6: Training accuracy:  922067 right out of 1480005, 62.30% correct.
   CV accuracy: 92713 right out of 167227, 55.44% correct.
7: Training accuracy:  913508 right out of 1480005, 61.72% correct.
   CV accuracy: 93367 right out of 167227, 55.83% correct.
Best CV accuracy: 55.83% correct in epoch 7.
Recognition accuracy: 977138 right out of 1647232, 59.32% correct.

----------ITERATION 4:
Recognition accuracy: 977138 right out of 1647232, 59.32% correct.
CV accuracy: 38899 right out of 167227, 23.26% correct.
1: Training accuracy:  858108 right out of 1480005, 57.98% correct.
   CV accuracy: 82873 right out of 167227, 49.56% correct.
2: Training accuracy:  920915 right out of 1480005, 62.22% correct.
   CV accuracy: 83917 right out of 167227, 50.18% correct.
3: Training accuracy:  938036 right out of 1480005, 63.38% correct.
   CV accuracy: 86119 right out of 167227, 51.50% correct.
4: Training accuracy:  946132 right out of 1480005, 63.93% correct.
   CV accuracy: 86377 right out of 167227, 51.65% correct.
5: Training accuracy:  943577 right out of 1480005, 63.75% correct.
   CV accuracy: 90967 right out of 167227, 54.40% correct.
6: Training accuracy:  943302 right out of 1480005, 63.74% correct.
   CV accuracy: 94167 right out of 167227, 56.31% correct.
7: Training accuracy:  936619 right out of 1480005, 63.28% correct.
   CV accuracy: 94897 right out of 167227, 56.75% correct.
Best CV accuracy: 56.75% correct in epoch 7.
Recognition accuracy: 1002165 right out of 1647232, 60.84% correct.