Comments by "Pete Venuti" (@petevenuti7355) on "Andrej Karpathy" channel.

  1. How do you get a large model in a small ram? How do you break these up? Like training GPT in a megabyte, a bit at a time.
    1