NovitaAI
gryphe/mythomax-l2-13b (https://huggingface.co/Gryphe/MythoMax-L2-13b); nousresearch/nous-hermes-llama2-13b (https://huggingface.co/NousResearch/Nous ...
Model details
MythoMax L2 13B is a purpose-built language model centered on creative expression, designed specifically for roleplay, fiction writing, and interactive storytelling. It emerged from an experimental lineage: a perfected iteration of MythoMix, itself a merge combining MythoLogic-L2 and Huginn through a novel tensor-type merge technique. This unusual construction gives the model a distinctive aptitude for narrative coherence and character-driven dialogue that general-purpose models often struggle to maintain.
The model's development lineage includes support for fine-tuning via LoRA, enabling developers to adapt it to specific creative styles or domain needs. Its availability in GGUF format through the open-source community makes it accessible for local deployment across platforms like llama.cpp, text-generation-webui, KoboldCpp, and LM Studio, while also serving cloud API workflows. This dual accessibility—ready for creative API use or self-hosted inference—positions MythoMax L2 13B as a practical choice for writers, worldbuilders, and application developers who need consistent, engaging narrative output.
NovitaAI
gryphe/mythomax-l2-13b (https://huggingface.co/Gryphe/MythoMax-L2-13b); nousresearch/nous-hermes-llama2-13b (https://huggingface.co/NousResearch/Nous ...