Two weeks ago I was running the Opt2.3B (I think) language model, which is not very capable and ran like an absolute dog on my machine. Last week, I downloaded Alpaca, which was better, twice the size, and ran super fast. Four days later I downloaded GPT4All, which is even better than that, and now I'm eyeing Vicuna, which does better on many tasks than Bard, thinking nothing but "gimmee" (so far that one isn't available for download, but man is the online demo impressive).

I was actually sort of surprised that Vicuna didn't become available for easy download overnight. This snail-pace has got to stop! \s\s\s\s\s