4 ms·
Good language models are NOT easy to run, so I’d imagine local compute would take many orders of magnitude more power
by circuit10 3y ago
Good language models are NOT easy to run, so I’d imagine local compute would take many orders of magnitude more power