Overview
sh0ck0r/Strawberrylemonade-L3-70B-v1.2-heretic3 is a 70 billion parameter language model based on the Llama 3 architecture. It is a decensored version of the original sophosympatheia/Strawberrylemonade-L3-70B-v1.2 model, created using the Heretic v1.2.0 tool. This model is a merge of several Llama 3-based models, including zerofata/L3.3-GeneticLemonade-Final-v2-70B and zerofata/L3.3-GeneticLemonade-Unleashed-v3-70B, using the NuSLERP merge method.
Key Capabilities
- Decensored Output: Engineered to produce responses with significantly fewer refusals (18/100) compared to the original model (99/100), allowing for broader content generation.
- Enhanced Creativity and Expressiveness: Specifically designed and improved for creative writing, roleplaying, and storytelling tasks.
- Customizable Output: Highly responsive to sampling parameters like
temperature and min-p, enabling users to fine-tune for either reliable, coherent responses or more creative, diverse outputs. - Llama 3 Foundation: Benefits from the underlying capabilities of the Llama 3 architecture, providing a strong base for language generation.
Good For
- Roleplaying and Storytelling: Its primary strength lies in generating engaging and immersive narratives, making it ideal for interactive fiction, character development, and creative writing projects.
- Applications Requiring Unrestricted Content: Suitable for use cases where content moderation or refusal to generate certain topics is undesirable, such as specific creative or research applications.
- Experimentation with Sampling Settings: Developers and users who wish to explore the impact of various sampling parameters on model output will find it highly adaptable.