Hacker News new | ask | show | jobs
by TheCoreh 863 days ago
The networks trained were really small, with only one hidden layer, and a width of 16.