A 0.4B parameter causal language model built for edge deployment. Fast, small, and honest about what it can do. — Improved instruction adherence on structured output formats — More stable multi-sentence generation (fewer mid-sequence repetitions) — Reduced hallucination rate on short factual queries (internal held-out eval) A 0.4B model cannot compete with 7B+ on reasoning — Geode doesn’t pretend to. Q4KM, Q6K, and Q80 are the recommended daily drivers. TQ10 and TQ20 are included for constrained edge hardware but will lose measurable reasoning and factual accuracy. For deterministic output, use —temp 0 and keep prompts short. The GGUF metadata includes the chat template. If your runtime doesn’t apply it automatically:
Модальности:
Генерация текста
Области применения:
Диалог / чат
Задача: Генерация текста
Автор: North-ML1
Теги: gguf, 0.4b, causal-lm, edge, north-ml, small-language-model, wind, wind-edge
Лайков: 3 | Загрузок: 266
Описание основано на материалах HuggingFace. Перевод выполнен автоматически.
