News for hackers. Stuff that matters.
Re:DeepSeek-R1 with Dynamic 1.58-bit Quantization by sitkack on Tuesday January 28, 2025 @08:58AM (#42852342)
How are you doing your evals?

Being able to do semantic diffs of the output of the two models should tell you what you need to do.

There are no comments on this comment yet.