I just ran a sample that nearly out performed the state of the art in the beginning of my PhD 6 years ago, using 1/4000th the amount of sample in ~1/4th the time.

That is pretty amazing.

Comments