Did anybody of you tried Llama or Alpaca GPT models to be deployed to local computer rather than the cloud? Can this be anyhow performant in an acceptable way on desktop PC? -h1 ... Xerox Alto was the thing. Anything after we use is just a copy.
