Hugging Face Overview
Ohashi-NeMo-12B: A Roleplay-Optimized Conversational Model
Ohashi-NeMo-12B is a 12 billion parameter model developed by Delta-Vector, fine-tuned from the Mistral-Nemo-Instruct-2407 base. Its primary focus is on delivering high-quality, dialogue-heavy prose for one-on-one roleplay interactions. The model was trained for 4 epochs using Axolotl on a diverse set of conversational datasets, including NewEden/OpenCAI-ShareGPT, NewEden/vanilla-backrooms-claude-sharegpt, anthracite-org/kalo_opus_misc_240827, anthracite-org/kalo_misc_part2, and NewEden/Roleplay-Logs-V2.
Key Capabilities
- Specialized Roleplay: Optimized for generating short, sweet, and dialogue-rich responses in one-on-one roleplay.
- Character Embodiment: Designed to maintain consistent character personas, motivations, and emotional depth throughout narrative exchanges.
- Narrative Consistency: Focuses on coherent storytelling, avoiding repetition and integrating new elements naturally.
- Flexible Prompting: Tuned with Mistral formatting, supporting system prompts like Euryale's or EVA's for enhanced control over roleplay guidelines and content.
- High Context Length: Supports a context length of 32768 tokens, allowing for extended and detailed roleplay sessions.
Good For
- Interactive Storytelling: Ideal for applications requiring dynamic, character-driven narrative generation.
- Roleplay Bots: Excellent for creating AI companions or characters that engage in detailed, dialogue-focused interactions.
- Creative Writing Assistance: Can be used to generate conversational snippets or character dialogue for creative projects.