A Breadcrumbs merge of five LFM2.5-1.2B fine-tunes on top of LiquidAI’s Instruct, blended with the Winery fusion compiler. It’s 1.17B params and 1.25 GB at Q8_0. LFM2.5’s hybrid conv/attention design makes it one of the fastest models in its class on phone CPUs. Breadcrumbs (density 0.85) over LiquidAI/LFM2.5-1.2B-Instruct, norms averaged linearly: MMLU 200 / ARC-Challenge 150 / GSM8K 60 questions, run through llama-server. Small-sample numbers from our own harness, so treat them as directional only. Works in llama.cpp (builds with LFM2 support), Jan, LM Studio and the Winery app. Licensed under the LFM Open License v1.0 inherited from LiquidAI. One donor is abliterated, so it refuses less than stock.
Модальности:
Генерация текста
Области применения:
Диалог / чат
Задача: Генерация текста
Автор: WineryLabs
Теги: gguf, merge, lfm2.5, liquid, winery, llama.cpp, breadcrumbs, endpoints_compatible
Лайков: 3 | Загрузок: 172
Описание основано на материалах HuggingFace. Перевод выполнен автоматически.