This model is fine-tuned for generating summaries of diverse text types in German. It produces concise, variable-length summaries, with the ability to reference key sections of the original text. Designed to handle long-context documents, it can efficiently summarize even extensive texts—such as 100-page documents—within seconds, thanks to its large-scale architecture. Attention: The model won’t follow the output text requirements when used with structured output via VLLM/Outlines. It is advised to sanitize/parse the JSON manually. Please report back to us if you find a way to use structured output without quality degradation. The model was trained with these prompts. It is advised to follow them verbatim and only vary the prompt variables.
Модальности:
Генерация текста
Области применения:
Диалог / чат
Задача: Генерация текста
Автор: ellamind
Теги: mistral, conversational, de, text-generation-inference, endpoints_compatible
Лайков: 3 | Загрузок: 0
Описание основано на материалах HuggingFace. Перевод выполнен автоматически.
