News for hackers. Stuff that matters.
Re:Microsoft BitNet: 100B Param 1-Bit model for local CPUs by regularfry on Wednesday March 11, 2026 @09:31AM (#47335342)
You only need GPUs if you assume the training is gradient descent. GAs or anything else that can handle nonlinearities would be fine, and possibly fast enough to be interesting.

There are no comments on this comment yet.