融合模型基于Shadowforge和我的lora，我觉得不好，但是我喜欢发,模仿nai4.5的风格 如果你喜欢，点赞，返图，这对我很重要。(｀・ω・´)—————————————————————————————————————— 注意事项（务必阅读）1.本模型微调自NoobAI-XL Eps-pred 110%2.采样推荐Euler a 迭代次数20以上 cfg5.53.推荐图片分辨率不低于512x780（最适分辨率阅读手册） ，正面或者反面tag可参考例图（包括单人和双人）。4.使用noob模型前十分建议阅读手册学习noob系模型详细参数和使用教程5[重要].noob的训练集基于danbooru，因此详细准确有效的tags最好是使用原生的danbooru的tags，获取准确的tags可以参考国内可用的安全版本Danbooru6[重要].noob系列模型可以直接生成25年以前绝大部分热门角色，同时可以使用画师tags直接使用danbooru中部分画师的画风（当然部分没有登记或训练集过少的画师效果较差）—————————————————————————————————————— 提示词推荐（推荐仅供参考）推荐质量tags： highres,absurdres,newest,very awa,masterpiece,best quality,推荐光影tags： chiaroscuro,Natural Light,Mood Lighting,cinematic lighting,volumetric lighting,ray tracing,dramatic shadows,dappled light,声明请不要使用本模型生成违反当地法律的内容，你需要对你的生成内容负责。 模型可能会无意中输出NSFW内容，确保你的提示词符合你的需求以避免在公共场所传播不当内容。 如果使用模型进行融合或者训练时候请注明使用过本模型。Model DescriptionThis fusion model is based on Shadowforge and my LoRA. I'm not entirely satisfied with it, but I enjoy sharing it anyway, mimicking the style of nai4.5.If you like it, please give it a like and share your generated images. Your support means a lot to me. (｀・ω・´)Important Notes (Must Read)This model is fine-tuned from NoobAI-XL Eps-pred 110%.Recommended sampler: Euler a, with at least 20 iterations and CFG scale 5.5.Recommended resolution: not lower than 512x780 (refer to the manual for optimal resolution). Positive and negative tags can be referenced from the sample images (including single and dual-character scenes).Before using the Noob model, it is highly recommended to read the manual to learn about the detailed parameters and usage tutorials of the Noob series models.[Important] The training dataset of the Noob model is based on Danbooru. Therefore, for detailed, accurate, and effective tags, it is best to use native Danbooru tags. To obtain accurate tags, you can refer to the domestically accessible and safe version of Danbooru.[Important] The Noob series models can directly generate most popular characters before 2025. Additionally, you can use artist tags to apply the styles of certain artists from Danbooru (though the effect may be poor for unregistered artists or those with insufficient training data). Recommended Prompt Tags (For Reference Only)Quality Tags:highres, absurdres, newest, very awa, masterpiece, best qualityLighting &amp; Shadow Tags:chiaroscuro, Natural Light, Mood Lighting, cinematic lighting, volumetric lighting, ray tracing, dramatic shadows, dappled lightDisclaimerPlease do not use this model to generate content that violates local laws. You shall be held accountable for the content you generate.The model may unintentionally output NSFW content. Ensure your prompts align with your needs to avoid spreading inappropriate content in public spaces.If you use this model for fusion or training, please acknowledge the use of this model.

NovelFabric-MIX

novelfabricMIX_v10.safetensors

v1.0

融合模型基于Shadowforge和我的lora，我觉得不好，但是我喜欢发,模仿nai4.5的风格 如果你喜欢，点赞，返图，这对我很重要。(｀・ω・´)——————————————————————————————————————注意事项（务必阅读）1.本模型微调自NoobAI-XL Eps-pred 110%2.采样推荐Euler a 迭代次数20以上 cfg5.53.推荐图片分辨率不低于512x780（最适分辨率阅读手册） ，正面或者反面tag可参考例图（包括单人和双人）。4.使用noob模型前十分建议阅读手册学习noob系模型详细参数和使用教程5[重要].noob的训练集基于danbooru，因此详细准确有效的tags最好是使用原生的danbooru的tags，获取准确的tags可以参考国内可用的安全版本Danbooru6[重要].noob系列模型可以直接生成25年以前绝大部分热门角色，同时可以使用画师tags直接使用danbooru中部分画师的画风（当然部分没有登记或训练集过少的画师效果较差）——————————————————————————————————————提示词推荐（推荐仅供参考）推荐质量tags： highres,absurdres,newest,very awa,masterpiece,best quality,推荐光影tags： chiaroscuro,Natural Light,Mood Lighting,cinematic lighting,volumetric lighting,ray tracing,dramatic shadows,dappled light,声明请不要使用本模型生成违反当地法律的内容，你需要对你的生成内容负责。 模型可能会无意中输出NSFW内容，确保你的提示词符合你的需求以避免在公共场所传播不当内容。 如果使用模型进行融合或者训练时候请注明使用过本模型。Model DescriptionThis fusion model is based on Shadowforge and my LoRA. I'm not entirely satisfied with it, but I enjoy sharing it anyway, mimicking the style of nai4.5.If you like it, please give it a like and share your generated images. Your support means a lot to me. (｀・ω・´)Important Notes (Must Read)This model is fine-tuned from NoobAI-XL Eps-pred 110%.Recommended sampler: Euler a, with at least 20 iterations and CFG scale 5.5.Recommended resolution: not lower than 512x780 (refer to the manual for optimal resolution). Positive and negative tags can be referenced from the sample images (including single and dual-character scenes).Before using the Noob model, it is highly recommended to read the manual to learn about the detailed parameters and usage tutorials of the Noob series models.[Important] The training dataset of the Noob model is based on Danbooru. Therefore, for detailed, accurate, and effective tags, it is best to use native Danbooru tags. To obtain accurate tags, you can refer to the domestically accessible and safe version of Danbooru.[Important] The Noob series models can directly generate most popular characters before 2025. Additionally, you can use artist tags to apply the styles of certain artists from Danbooru (though the effect may be poor for unregistered artists or those with insufficient training data).Recommended Prompt Tags (For Reference Only)Quality Tags:highres, absurdres, newest, very awa, masterpiece, best qualityLighting &amp; Shadow Tags:chiaroscuro, Natural Light, Mood Lighting, cinematic lighting, volumetric lighting, ray tracing, dramatic shadows, dappled lightDisclaimerPlease do not use this model to generate content that violates local laws. You shall be held accountable for the content you generate.The model may unintentionally output NSFW content. Ensure your prompts align with your needs to avoid spreading inappropriate content in public spaces.If you use this model for fusion or training, please acknowledge the use of this model.

Files with SHA256: 55ac4c446c1a6b5b77eca5f4eab278f01f40bee9edbbe74e32f2e2636b052b01