【LORA】LowRA降低亮度去除噪点暗系风LowRA[348071DB54]

名称：LowRA.safetensors
类型：【LORA】
程序版本：SD 1.5
文件大小：72.1MB
触发词：dark theme
概述：
Quick tips:
Your start point (weight) is 0.6 ➜ <loraowRA:0.6>
Looking for more darkness? Just add dark theme to your prompt
The best weight range is about 0.6 ≈ 0.8
Technical information:
Dreambooth training (creation) models process is designed to turn each of your images into noise at each stage of training. This is done, so that later, during the image generation phase, Stable Diffusion could turn any random noise back into an image. The more you Train your model in Dreambooth (the more often you turn the same image into random noise), the easier it will be for Stable Diffusion to recreate the image you need from random noise (which Stable Diffusion creates itself). For example, If you turn 1 image into random noise 10 times (10 steps Dreambooth Training), then during generation, Stable Diffusion has 10 times more chance to recreate the image from noise. That's why we making thousands of training steps – to increase the probability of recreating an image.
However, by turning any image into noise we get ≈45 to ≈55 percent of Brightness (HSB). Always. You can check this on python with numpy random or in Photoshop, just take a completely white square and a completely black square and then turn both into noise. Then Filter ➜ Blur ➜ Average. You will always get ≈45 to ≈55 percent of Brightness. That's why you always fail, when you trying to generate completely black square or completely white square. You just can't.
LowRA solves this problem, when it embeds itself in the image generation process. At the beginning of the image generation process, LowRA gives you another noise, that allows you to get the image in the key you want. The higher the weight of LowRA, the more Stable Diffusion will be forced to use samples from LowRA. The samples used to train LowRA were strictly selected images that had an overall average Brightness of less than 25%. This made it possible to make the use of LowRA with less weight, so that when you use LowRA in parallel with other LoRAs, you don't get mess. So if you are using something else to get low key images in Stable Diffusion, you should forget the old techniques and just use LowRA.
The careful selection of images in LowRA training proceeded as follows: a few thousand low key images were selected to begin with, each getting one noise pattern in Photoshop and a strict color palette for each individual subject. Warm and bright tones received no more than 25% of the space in the entire image. Next, each processed image was sent to img2img to create 10 similar generations. After that, a sample of the 10 images was tested for proper color reproduction in Photoshop. If the test passed, and the bright and warm tones did not exceed 25% of the Brightness, the image that was sampled was sent to dataset. If not, the image was simply deleted and the next image was taken. This rigorous and careful selection made it possible to obtain the ideal dataset for training. And now you can see the result of my work on this page.
快速提示：
你的起点（体重）是0.6➜ <loraowRA:0.6>
寻找更多的黑暗？只需在提示中添加深色主题
最佳重量范围约为0.6≈0.8
技术信息：
Dreambooth训练（创建）模型过程旨在将您的每个图像在训练的每个阶段都变成噪音。这样做之后，在图像生成阶段，“稳定扩散”可以将任何随机噪声重新转化为图像。你在Dreambooth中训练你的模型越多（你越经常把同一张图像变成随机噪声），稳定扩散就越容易从随机噪声中重新创建你需要的图像（稳定扩散自己创建）。例如，如果将1个图像转换为随机噪声10次（10步Dreambooth Training），则在生成过程中，“稳定扩散”从噪声中重新创建图像的机会会增加10倍。这就是为什么我们要进行数千个训练步骤，以增加重建图像的概率。
然而，通过将任何图像转换为噪声，我们可以获得≈45%到≈55%的亮度（HSB）。总是你可以在python上用numpy随机或在Photoshop中检查这一点，只需取一个完全白色的正方形和一个完全黑色的正方形，然后将两者都变成噪音。然后过滤➜ 变得模糊不清➜ 平均的您将始终获得≈45%到≈55%的亮度。这就是为什么当你试图生成完全黑色的正方形或完全白色的正方形时，你总是失败。你就是做不到。
LowRA解决了这个问题，当它将自己嵌入到图像生成过程中时。在图像生成过程开始时，LowRA会为您提供另一个噪波，使您能够获得所需密钥中的图像。LowRA的权重越高，Stable Diffusion将被迫使用LowRA的样本。用于训练LowRA的样本是严格选择的整体平均亮度小于25%的图像。这使得使用LowRA的重量更小成为可能，这样当你与其他LoRA同时使用LowRA时，你就不会陷入困境。因此，如果你在稳定扩散中使用其他方法来获得低调的图像，你应该忘记旧的技术，只使用LowRA。
LowRA训练中对图像的仔细选择如下：首先选择几千张低调的图像，每个图像在Photoshop中都有一个噪声模式，每个受试者都有严格的调色板。暖色调和明亮色调占据的空间不超过整个图像的25%。接下来，每个处理后的图像都被发送到img2img，以创建10个类似的代。之后，在Photoshop中对10张图像的样本进行适当的颜色再现测试。如果测试通过，并且明亮和温暖的色调不超过亮度的25%，则将采样的图像发送到数据集。如果没有，则简单地删除该图像并拍摄下一张图像。这种严格而仔细的选择使得获得理想的训练数据集成为可能。现在你可以在这个页面上看到我的工作成果。

Hash	parameter
SHA256	348071DB544B7242C5EDCB3306160D83BCDE66395153C1DAF38A575C5CEFD66E
CRC32	D8C98408
BLAKE3	420B76F142608EF6374E982578746E0795716FB610A2E7D63E5C47065A8A4CEA
AUTOV1	2DCEB6E8
AUTOV2	348071DB54

作品展示：

Generation Data：

1. <lora:LowRA:0.8> hdr, a redhead woman climbing on mountains, frightened, closeup, intricate details, hyperdetailed, cinematic, dark shot, muted colors, film grainy, soothing tones, muted colors, technicolor, (muddy:0.6)
   
2. Negative prompt: (deformed, distorted, disfigured:1.3), poorly drawn, bad anatomy, wrong anatomy, extra limb, missing limb, floating limbs, (mutated hands and fingers:1.4), disconnected limbs, mutation, mutated, ugly, disgusting, blurry, amputation
   
3. Steps: 26, ENSD: 31337, Size: 640x768, Seed: 818146396, Model: deliberate_v2, Sampler: DPM++ 2M Karras, CFG scale: 5, Model hash: 9aba26abdf, Hires steps: 8, Hires upscale: 2, Hires upscaler: 4x_NMKD-Siax_200k, Denoising strength: 0.42

复制代码

Generation Data：

1. <lora:LowRA:0.6> hdr, oldman, closeup, (crowded newspapers store:1.2), hyperdetailed, cinematic, dark shot, muted colors, film grainy, soothing tones, muted colors
   
2. Negative prompt: (deformed, distorted, disfigured:1.3), poorly drawn, bad anatomy, wrong anatomy, extra limb, missing limb, floating limbs, (mutated hands and fingers:1.4), disconnected limbs, mutation, mutated, ugly, disgusting, blurry, amputation, dof
   
3. Steps: 22, ENSD: 31337, Size: 640x768, Seed: 2478123824, Model: deliberate_v2, Sampler: DPM++ 2M Karras, CFG scale: 5, Model hash: 9aba26abdf, Hires steps: 8, Hires upscale: 2, Hires upscaler: 4x_NMKD-Siax_200k, Denoising strength: 0.36

复制代码

Generation Data：

1. <lora:LowRA:0.7> drunk, creepy santa, muddy, crowded bottles bar, intricate details, hdr, intricate details, hyperdetailed, cinematic, dark shot, muted colors, film grainy, soothing tones, muted colors, technicolor
   
2. Negative prompt: (deformed, distorted, disfigured:1.3), poorly drawn, bad anatomy, wrong anatomy, extra limb, missing limb, floating limbs, (mutated hands and fingers:1.4), disconnected limbs, mutation, mutated, ugly, disgusting, blurry, amputation
   
3. Steps: 26, ENSD: 31337, Size: 640x768, Seed: 2035073191, Model: deliberate_v2, Sampler: DPM++ 2M Karras, CFG scale: 5, Model hash: 9aba26abdf, Hires steps: 8, Hires upscale: 2, Hires upscaler: 4x_NMKD-Siax_200k, Denoising strength: 0.55

复制代码

Generation Data：

1. <lora:LowRA:0.7> a girl in the lake, (art by Jesper Ejsing:1.2)
   
2. Negative prompt: (deformed, distorted, disfigured:1.3), poorly drawn, bad anatomy, wrong anatomy, extra limb, missing limb, floating limbs, (mutated hands and fingers:1.4), disconnected limbs, mutation, mutated, ugly, disgusting, blurry, amputation
   
3. Steps: 22, ENSD: 31337, Size: 640x768, Seed: 1076673168, Model: deliberate_v2, Sampler: DPM++ 2M, CFG scale: 5.5, Model hash: 9aba26abdf, Hires steps: 8, Hires upscale: 2, Hires upscaler: 4x-AnimeSharp, Denoising strength: 0.42

复制代码

Generation Data：

1. <lora:LowRA:0.8> Маша making extreme selfie on skyscraper, bird's eye view, from above, night, smiling
   
2. Negative prompt: (deformed, distorted, disfigured), poorly drawn, bad anatomy, wrong anatomy, extra limb, missing limb, floating limbs, (mutated hands and fingers), disconnected limbs, mutation, mutated, ugly, disgusting, blurry, amputation
   
3. Steps: 22, Seed: 3135098381, Sampler: DPM++ 2M Karras, CFG scale: 6

复制代码

下载地址：

游客，如果您要查看本帖隐藏内容请回复

感谢分享，支持网站