模型介绍: Form for requests. I also sometimes do polls on my Patreon for the next LoRA (no payment required). Well it is done. That was honestly way more difficult then I imagined. The main problem with making this lora is quite honestly just the dataset. I have no idea why, but I couldn't find two pictures in a row where she had the same hair style. Not to mention that 99% of the available images were in her sailor uniform. I still hope you will be able to make cool pictures with it. A weight of ~1.0 is recommended. Feel free to experiment with other weights and share your results if you want to! The main prompts are "SMMars" You can try "very long hair" and "parted bangs" for the hair style (i don't know her hair style is really inconsistent) Add "1990s \(style\)" to your prompt, if you want that style. For her outfit try: "SMMars, sailor senshi uniform, red sailor collar, red skirt, elbow gloves" Let me know if you have any issues. Used embeddings for sample pictures: EasyNegative bad_prompt_version2 If you have any feedback or questions just write in the comments or hit me up on reddit or discord. ~~~~~~~~~~~~~~~~~~~~~~~o(* ̄︶ ̄*)o~~~~~~~~~~~~~~~~~~~~~~~ 示例提示词和参数: Prompt: masterpiece, best quality, absurdres, perfect antomy, 1girl, solo, SMMars, very long hair, parted bangs, sailor senshi uniform, red sailor collar, red skirt, elbow gloves, standing, cowboy shot, smile, cartoon flames in background, stylized background Negative prompt: nsfw, (worst quality, low quality:1.4), EasyNegative , bad_prompt_version2 Steps: 20, CFG scale: 5, Sampler: Euler a, Seed: 4283160086, Size: 512x768, Model: ChameleonAIRPGMix_v1.0prunedBakedVae, Model hash: ca0e3359f7, Hires upscale: 1.5, AddNet Enabled: True, AddNet Model 1: SailorMarsLORA-000026(c1cc467da016), Hires upscaler: 4x-UltraSharp, AddNet Module 1: LoRA, AddNet Weight A 1: 1, AddNet Weight B 1: 1, Denoising strength: 0.6, Clip skip: 2 |
1人表示很赞