Get the latest tech news

LFM2-24B-A2B: Scaling Up the LFM2 Architecture


Today, we release an early checkpoint of LFM2-24B-A2B, our largest LFM2 model. This sparse Mixture of Experts (MoE) model has 24 billion total parameters with 2 billion active per token, showing that the LFM2 architecture scales effectively to larger sizes.

None

Get the Android app

Or read this on Hacker News

Read more on:

Photo of lfm2

lfm2

Photo of a2b

a2b

Photo of lfm2 architecture

lfm2 architecture