Article · AI translation
IT Home reported on October 28 that artificial intelligence laboratory Liquid AI released the LFM3 model family on the 27th, local time. Built on a new three-hybrid-layer architecture and pretrained on 22T tokens, it delivers stronger performance while remaining efficient.
The first LFM3 products include LFM-1.7B (text), LFM3-VL-2B (vision), d1-2B (decision-making), LFM3-Audio (audio), LFM3-Omni-2B (omni), and d1-Omni-2B (omni decision-making). All these models are based on LFM3-1.7B-Base.
Liquid AI said LFM-1.7B has a 256K context window and outperforms the previous-generation LFM2.5-2.6B while using 35% fewer parameters.
Reference
https://www.liquid.ai/blog/lfm3
Advertising disclosure: External links in this article (including but not limited to hyperlinks, QR codes, and access codes) are provided to convey more information and save time in selecting information. The results are for reference only. All IT Home articles containing external links include this disclosure.