2 ms·
I would like to see what happens when you go the other way. Extremely naive tokening, for instance none at all. Just a stream of bytes or nybbles. It might t
by emmelaich 3y ago
I would like to see what happens when you go the other way. Extremely naive tokening, for instance none at all. Just a stream of bytes or nybbles.
It might take far more training but also it might avoid any biases introduced by tokenisation.
[edit - see @api had the same question]
- f137 3y agoI've run across a paper "Bytes is all you need" or the like a few days ago. Probably something you'd like to read