4 ms·
Not immediately, but it's not a much larger amount of work for llama than a new foundational model which typically has a tweaked compute graph.
by monocasa 8mo ago
Not immediately, but it's not a much larger amount of work for llama than a new foundational model which typically has a tweaked compute graph.