News for hackers. Stuff that matters.
Re:You can now train a 70B language model at home by chessgecko on Friday March 08, 2024 @02:52PM (#39645514)
also just to add, I think the 1.58 bit is mostly faster for inference because training still had to multiply a lot of floating point gradients by integer activations, hold floating point weights/gradients for round, and deal with norms and stuff. could be wrong about that though

There are no comments on this comment yet.