Hugging Face Blog·· 6 h ago
Build specialized multimodal models with LFM3
Build specialized multimodal models with LFM3
AI summary
LFM3 is our new generation of Liquid Foundation Models (LFMs), built to be the starting point for specialized intelligence. It brings text, vision, audio, omni, and decision models onto one shared backbone, all in the 2B parameter range.
Selection record
Threshold 60Official, first-handFirst 74Second 78
AdmittedSum of both 152 ≥ twice the threshold 120
- Source tier
- Official, first-hand; this tier's threshold is 60
- Pre-filter
- passed：LFM3多模态模型发布，含架构与用法
- Why it was chosen
- Covers how one 2B-class backbone spans text, vision, audio, omni and decision models, plus single-GPU fine-tuning and on-device deployment paths, which helps in judging the specialised small-model route.
A model scores each item twice, independently, against one written standard, out of 100. An item is admitted only when the two scores add up to twice the threshold. The threshold is set per source tier.
Source: Hugging Face Blog · huggingface.co