Re:Show HN: less than 650 LOC trainable GPT only using NumPy...
by PartiallyTyped on Friday November 17, 2023 @01:14PM
(#38307565)
The most interesting thing in this whole saga is that decoder only models (aka causal transformers like GPT) are as effective as they are.
[ Parent
| on HN ]
There are no comments on this comment yet.