Comment by kalu | Hacker News Reader

Top New Best Ask Show Jobs

Comment by kalu | Hacker News Reader

So this author trained a neural network billions of times using different hyper parameters ? How much dod that cost ?

"overnight on an A100" per https://twitter.com/jaschasd/status/1757056127991439832

The networks trained were really small, with only one hidden layer, and a width of 16.

Very small networks are cheap and easy (even 20 years ago).

Reply on news.ycombinator.com