CAT-Paws is an agentic LLM that thinks in Japanese (e.g., reasoning trace is in Japanese). The model is based on Qwen3-Swallow-v0.2 which is a continual pretraining model based on Qwen3 to read and write fluently in Japanese. CAT-Paws is trained for multi-turn interactions involving coding, interacting with terminal, and tool usages. For non-agentic single-turn tasks, we recommend using its sibling model CAT-Thinking-8B. CAT-Paws is designed to reason in Japanese even if the input text is in English. The model is trained with the maximum output token length of 4096. We recommend setting maxnewtokens to at least 4096, and larger for difficult problems. Although the model is trained to respond within 4096 tokens, it tends to generate longer responses, especially for difficult and/or confusing instructions. It often gets stuck in repetition, especially when the instruction is confusing (e.g., two contradicting instructions are given). We recommend using a harness with minimum system prompts for CAT-Paws. Because the context length is quite limited (40k), it doesn’t work well with long system prompts used for harnessing the frontier models. Instead, it will function well with simple…
Модальности:
Генерация текста
Области применения:
Диалог / чат
Задача: Генерация текста
Автор: cyberagent
Теги: qwen3, japanese, english, conversational, ja, en, text-generation-inference, endpoints_compatible
Лайков: 3 | Загрузок: 186
Описание основано на материалах HuggingFace. Перевод выполнен автоматически.
