Elizezen/Antler-7B
Elizezen/Antler-7B is a 7-billion parameter decoder-only Japanese language model, fine-tuned on novel datasets. Built upon the Japanese Stable LM Base Gamma 7B architecture, it specializes in generating Japanese novels. This model leverages a 4096-token context length and is primarily optimized for creative text generation rather than instruction-based responses.
Loading preview...
Antler-7B Overview
Elizezen/Antler-7B is a 7-billion parameter, decoder-only Japanese language model. It is built on the foundation of the Japanese Stable LM Base Gamma 7B and has been specifically fine-tuned using novel Japanese datasets. The model's training involved a combination of web novels, including less than 1GB of non-PG content and 70GB of PG content.
Key Capabilities
- Japanese Novel Generation: The model is primarily designed and optimized for generating creative Japanese narratives and stories.
- Base Architecture: Leverages the robust architecture of Japanese Stable LM Base Gamma 7B.
- Context Length: Supports a context window of 4096 tokens, suitable for generating coherent longer-form text.
Intended Use
Antler-7B is best suited for applications requiring the generation of Japanese novels and creative writing. Developers should note that while it excels in narrative generation, it may not perform as effectively with instruction-based prompts or tasks requiring precise instruction following.