- 2 images, 2 neural nets
- Input: 20 neurons each neural net
- Hidden: 2 layers, one of 20 and another of 10
- 10.000 epochs for training for each neural net
For 3 layers (input plus 2 hidden), 20^20^10 would be the total of calculations per training epoch. The total would be (20^20^10) multiplied by 10000 (epochs) multiplied by 2 (number of neural nets).
This takes less than a second.