4 ms·
> Llammas are showing the world that it doesn't take monopoly-level hardware to run those things. LLaMA was not necessarily the model that did that. A fairer a
by smoldesu 4y ago
> Llammas are showing the world that it doesn't take monopoly-level hardware to run those things.
LLaMA was not necessarily the model that did that. A fairer attribution might be BERT or GPT-Neo.