11 ms·
It's a combination of tokens consumed, dataset + model storage cost, and inference + training compute cost.
by marcellodb 11mo ago
It's a combination of tokens consumed, dataset + model storage cost, and inference + training compute cost.
11 ms·