This repo contains the full precision source code, in «safe tensors» format to generate GGUFs, GPTQ, EXL2, AWQ, HQQ and other formats. The source code can also be used directly. This is a Short Story fine tune (15% of the model) for 5 epochs, using a short story dataset using the following model as a base: https://huggingface.co/Goekdeniz-Guelmez/Josiefied-Qwen3-4B-Instruct-2507-gabliterated-v1 The fine tuning (using Unsloth for Win 11) alters the prose generation and general creative abilities. This model will output SHORT stories (when prompted for a «story») — that is limited output size — but it is very unique. Prose, and word choice will also be very different (see examples below). For creative uses: Increases depth, detail and general «there» in the prose. IMPORTANT: — Temps of .8 to 1.5 suggested, with rep pen of 1.05 to 1.1 — 2-5 regens will produce 2-5 very different outputs. — Model may NOT follow length requirements for a «story/scene» => regen. — You may get some refusals -> regen to get around these. — Model has an attitude too. INVALID LANGUAGE PAIR SPECIFIED. EXAMPLE: LANGPAIR=EN|IT USING 2 LETTER ISO OR RFC3066 LIKE ZH-CN. ALMOST ALL LANGUAGES SUPPORTED BUT SOME MAY HAVE NO CONTENT
Модальности:
Генерация текста
Области применения:
Генерация кода Программирование Диалог / чат Следование инструкциям
Задача: Генерация текста
Автор: DavidAU
Теги: qwen3, programming, code generation, code, coding, coder, chat, qwen
Лайков: 4 | Загрузок: 289
Описание основано на материалах HuggingFace. Перевод выполнен автоматически.
