> [!NOTE] > WARNING: This model has been trained on instructions but has not undergone safety or value alignment. > > Work In Progress New versions will be available during the coming weeks/months. > > Sampling Parameters: For optimal performance, we recommend using temperatures close to zero (0 — 0.2). Additionally, we advise against using any type of repetition penalty, as from our experience, it negatively impacts instructed model’s responses. ALIA-40b-instruct-2605 is the latest release in the ALIA model family. While development is ongoing and further updates are expected, this version already incorporates several notable improvements over previous releases. — Exact Instruction Following: Enhanced instruction-tuning, leading to more reliable adherence to user constrains. — Formatting: Better output formatting capabilities, such as JSON or Markdown. — Long-Context: Improved long-context and multi-turn capabilities compared to previous versions of ALIA-40b-instruct. — System prompt: Compared to previous versions, this one substancially improves system-following capabilities. — Summarization: Improved ability to perform concise and accurate summarization of texts of variable…
Модальности:
Генерация текста
Области применения:
Диалог / чат Следование инструкциям
Задача: Генерация текста
Автор: BSC-LT
Теги: llama, conversational, ca, en, es, eu, gl, text-generation-inference
Лайков: 4 | Загрузок: 326
Описание основано на материалах HuggingFace. Перевод выполнен автоматически.
