2 ms·The networks trained were really small, with only one hidden layer, and a width of 16.by TheCoreh 3y agoThe networks trained were really small, with only one hidden layer, and a width of 16.