3 ms·
Depends what you want it for. I'm still holding out for a decent enough open model, Llama 3 is tantalisingly close, but inference speed and cost are serious bot
by specproc 2y ago
Depends what you want it for. I'm still holding out for a decent enough open model, Llama 3 is tantalisingly close, but inference speed and cost are serious bottlenecks for any corpus-based use case.