# Multi-Style Fusion

A hybrid rap vocal model focused on improving the generation quality of Chinese rap/hip-hop music

This is a 9B-parameter language model merged using the mergekit tool, combining three models of different styles based on the TIES method

SegMoE-2x1-v0 is an untrained Segmind Diffusion Mixture of Experts model, dynamically composed of 2 expert SDXL models, offering broader knowledge, better adherence, and higher image quality.

SegMoE-4x2-v0 is an untrained Segmind Mixture of Diffusion Experts model, dynamically composed of 4 expert-level SDXL models, featuring a broader knowledge base and enhanced image generation capabilities.

A LoRA model based on Stable Diffusion XL, specifically designed for generating paintings in the impressionist style of Konstantin Korovin.

An anime-style image generation model based on Stable Diffusion, blending various popular anime styles

A text-to-image model merged from AbyssOrangeMix2, Helltaker, and Hero Academia LoRA, specializing in anime-style character generation

A hybrid model based on Stable Diffusion 1.5, focused on character design, inspired by OrangeMixs