Post @onusoz · /2026/08/08 · 04:09 PM View on I'm running some private benchmarks on @liquidai's LFM 2.5 2.6B, and if my results are correct, we might have a new champion for <10b category Scores significantly higher than gemma 4 e4b, which is 3-4x its size @vanstriendaniel · Aug 5, 2026 Can a 2.6B model earn permission to edit the @huggingface Hub? I put @liquidai's LFM2.5 on library duty. It investigates datasets with tools, then gets exactly ONE sentence to describe them. Traces are stored in a Bucket with human ratings of quality. If the community gives it 500 ratings, I'll let it start proposing dataset-card PRs! Show more Image hidden
@onusoz · /2026/08/08 · 04:09 PM View on I'm running some private benchmarks on @liquidai's LFM 2.5 2.6B, and if my results are correct, we might have a new champion for <10b category Scores significantly higher than gemma 4 e4b, which is 3-4x its size @vanstriendaniel · Aug 5, 2026 Can a 2.6B model earn permission to edit the @huggingface Hub? I put @liquidai's LFM2.5 on library duty. It investigates datasets with tools, then gets exactly ONE sentence to describe them. Traces are stored in a Bucket with human ratings of quality. If the community gives it 500 ratings, I'll let it start proposing dataset-card PRs! Show more Image hidden
I'm running some private benchmarks on @liquidai's LFM 2.5 2.6B, and if my results are correct, we might have a new champion for <10b category Scores significantly higher than gemma 4 e4b, which is 3-4x its size