Dev Radar
Support
LiveUpdated 2026-11-03 19:56 UTC

LFM3 1.7B 1 GB GGUF + EAR98 GGUF

LFM3 1.7B 1 GB GGUF + EAR98 GGUF I used RCO and GSQ to produce two improved GGUF checkpoints. RCO directly optimized…

This is a dev post classified by AGK Intelligence as Other (a tool drop), kept by the Dev Radar because it carries real work, not commentary.

LFM3 1.7B 1 GB GGUF + EAR98 GGUF I used RCO and GSQ to produce two improved GGUF checkpoints. RCO directly optimized EAR as its objective. All reported EAR scores were computed on a held-out dataset. Both GGUFs significantly outperform their respective donor GGUFs of similar size from LiquidAI and AtomicChat. I also computed EAR in French, Japanese, and Hindi, where the impact of quantization is more pronounced. Good reminder that quantization typically causes greater performance degradation on non-English tasks. Checkpoints and results: https://huggingface.co/kaitchup/LFM3-1.7B-GSQ-GGUF

Posted by Benjamin Marie (7.6k followers) 2 h ago · 12 likes · 807 views · view the original post on X. Kept by the Dev Radar as Other. Tools mentioned: lfm3-1.7b-gsq-gguf.

More dev work like this

Every post is read and classified by AGK Intelligence: what it is, which market it belongs to, and whether the link is a real tool. 31.2k posts from 5k X accounts over the last 21 days, 9.2k tools, 12 markets. New posts are collected several times an hour and the full ranking is recomputed every hour — last update 2026-11-03 19:56 UTC. Full method.