【LYCORIS】Abstract Dreamwave梦幻的动漫Dreamwave v3[690F34D2BA]

名称：Dreamwave v3.safetensors
类型：【LYCORIS】
程序版本：SD 1.5
文件大小：294.85MB
训练：STEPS: 32000,EPOCHS: 40
使用技巧：CLIP SKIP: 2
触发词：DREAMWAVE,AESTHETIC
概述：
v3: Training done as a Dreambooth, and then extracted to rank 128 locon.
v2.5: more images that are meant for purposes(tagged accordingly). Training changed.aesthetic
v2 update: 682 image training, but only 10 epochs, and on other base model. Should work quite good.
Full version update: Training on 682 images, vs 200 in Lite version. More dreamy, more abstract, better eyes, for whatever reason. But composition is less, let's say, tangible, in a way. But that's the point of this model i guess
General instruction: try some bullshit, random words, something dramatic, use wildcards. It will react differently. By default style is not very interesting(unless it is), it needs some prompting to get going.
Fourth Lazy Style Lora.
Firstly i'll release Lite version, that is based on 200 images, because training full dataset(682) will take some time, because i will use higher rank, which will make it slower, because of batch size.
Im not quite sure what that style is, but prompt on which they were generated suggests that it's Dreamwave.
Generally, this is not supposed to be used for anythin complex, because it was not tagged in a way to be, but you still can try and show me your results :3
GPTS Lazy Styles will have a lot of bleed, low editability(but better than it was in GPTS1)
If you want it to be editable and still have some style effect, try weight from 0.5 to 0.75. That should make it editable and still be looking good
Generally, this dataset processing used my custom tag filtering experiment, testing of which on GPTS1 led to increased editability, so, it might not be as hard to edit this one, but still, main purpose is just to get that bit of "wow" effect style.
Full version will be available later.
v3：作为梦幻展台进行训练，然后提取到128级机车。
v2.5：更多用于目的的图像（相应标记）。培训发生了变化。aesthetic
v2 更新：682 个图像训练，但只有 10 个 epoch，以及其他基本模型。应该工作得很好。
完整版更新：训练 682 张图片，精简版为 200 张。更梦幻，更抽象，更好的眼睛，无论出于何种原因。但在某种程度上，比方说，构图不那么有形。但我想这就是这个模型的重点
一般说明：尝试一些废话，随机单词，戏剧性的东西，使用通配符。它会有不同的反应。默认情况下，风格不是很有趣（除非是），它需要一些提示才能开始。
第四种懒惰风格劳拉。
首先，我将发布基于 200 张图像的精简版，因为训练完整的数据集 （682） 需要一些时间，因为我将使用更高的排名，这会使它变慢，因为批量大小。
我不太确定这种风格是什么，但生成它们的提示表明它是 Dreamwave。
一般来说，这不应该用于任何复杂的复合物，因为它没有以某种方式标记，但你仍然可以尝试向我展示你的结果：3
GPTS Lazy Styles 会有很多出血，可编辑性低（但比 GPTS1 更好）
如果您希望它是可编辑的并且仍然具有一些样式效果，请尝试从 0.5 到 0.75 的权重。这应该使它可编辑，并且仍然看起来不错
一般来说，这个数据集处理使用我的自定义标签过滤实验，在 GPTS1 上对其进行测试后，可编辑性提高了，因此，编辑这个可能并不难，但主要目的仍然只是为了获得一点“哇”的效果风格。

Hash parameter SHA256 690F34D2BADB6144D47CAD0655B12125DAF9C9B6E4A022CC0E0F5D9C24A0BB96 CRC32 3581BA9C BLAKE3 042F194C9A17F7B9BC91C9C9A8D8854AF76B9D27471232E7E13E6C425FCD472C AUTOV1 757B70E0 AUTOV2 690F34D2BA

作品展示：

Generation Data：

1. masterpiece, best quality, dreamwave, aesthetic, 1girl, (glass:1.1),(false:1.15),(void:1.25),(magic:1.25),
   
2. Negative prompt: (worst quality, low quality:1.4),greyscale, monochrome
   
3. Steps: 20, ENSD: 4537, Size: 512x768, Seed: 3783514673, Sampler: DPM++ 2M Karras, CFG scale: 7, Clip skip: 2, Model hash: bfa25131f5, Hires steps: 12, Hires upscale: 2, Hires upscaler: Latent (nearest-exact), Wildcard prompt: "masterpiece, best quality, dreamwave, aesthetic, 1girl, __dramatic__,__dramatic__,__dramatic__,__dramatic__,", Denoising strength: 0.55, Discard penultimate sigma: True

复制代码

Generation Data：

1. dreamwave, aesthetic, a (frozen:1.1) cosmic rose, the petals glitter with a crystalline shimmer, swirling nebulas, 8k unreal engine photorealism, ethereal lighting, purple, nighttime, darkness, surreal art
   
2. Negative prompt: (worst quality, low quality:1.4)
   
3. Steps: 20, ENSD: 4537, Size: 512x768, Seed: 859467204, Sampler: DPM++ 2M Karras, CFG scale: 7, Clip skip: 2, Model hash: bfa25131f5, Hires steps: 12, Hires upscale: 2, Hires upscaler: Latent (nearest-exact), Wildcard prompt: "dreamwave, aesthetic, __wrenchprompt__", Denoising strength: 0.55, Discard penultimate sigma: True

复制代码

Generation Data：

1. 1girl,dreamwave,aesthetic,
   
2. Negative prompt: badpic,EasyNegativeV2,
   
3. Steps: 20, VAE: klF8Anime2.safetensors, Size: 1024x1536, Seed: 1738112196, Model: Dreamshaper_8, "badpic: cc64d0aca8fd, Version: v1.6.0, Sampler: DPM++ 2M Karras, VAE hash: 551eac7037, CFG scale: 7, Clip skip: 2, Model hash: 879db523c3, EasyNegativeV2: 339cc9210f70", Denoising strength: 0.35, Ultimate SD upscale padding: 32, Ultimate SD upscale upscaler: R-ESRGAN 4x+ Anime6B, Ultimate SD upscale mask_blur: 8, Ultimate SD upscale tile_width: 512, Ultimate SD upscale tile_height: 512

复制代码

Generation Data：

1. (intricate detail, beautiful:1.2), <lora:smol02:0.4>, <lora:Dreamwave v3:0.5>,
   
2. 1 girl, adult italian woman, freckles, ruby eyes, black wavy hair,  
   
3. Style-GravityMagic, solo, from front, front view, half shot,  detailed background, detailed face, (lovecraftian theme:1.1), haunted expression, investigator of lovecraftian  horrors,   adventuring clothes, cloak, dark color scheme, desperate search for answers, , existential dread, , lovecraftian story, , inexplicable phenomena, dreaded texts,  mysterious mist, sinister shadows,  mysterious lovecraftian atmosphere,
   
4. Negative prompt: EasyNegative, negative_hand-neg, (multiple views:1.3), BadDream, UnrealisticDream,(male, man, boy, masculine:1.2), (undead, monster, human-monster, tentacle limbs:1.3),
   
5. Steps: 25, VAE: vae-ft-mse-840000-ema-pruned.safetensors, Size: 512x768, Seed: 3392830685, Model: antlersmix_11, smol02: 0.4>, "smol02: dcd18f6e1a3a, Sampler: DPM++ 2M Karras, VAE hash: df3c506e51, CFG scale: 20, Clip skip: 2, beautiful: 1.2), Model hash: 7f499cd636, Hires steps: 10, Dreamwave v3: 0.5>,\n__character-alt__\n__full-prompt-fantasy__", Hires upscale: 2, Hires upscaler: 4x-UltraSharp, ADetailer model: face_yolov8n.pt, Wildcard prompt: "(intricate detail, ADetailer version: 23.11.0, Denoising strength: 0.5, ADetailer mask blur: 4, ADetailer confidence: 0.3, ADetailer dilate erode: 32, ADetailer inpaint padding: 32, ADetailer denoising strength: 0.4, ADetailer inpaint only masked: True

复制代码

Generation Data：

1. dreamwave, aesthetic, 1girl, a (frozen:1.1) cosmic rose, the petals glitter with a crystalline shimmer, swirling nebulas, 8k unreal engine photorealism, ethereal lighting, purple, nighttime, darkness, surreal art, zentangle patterns
   
2. Negative prompt: (worst quality, low quality:1.4)
   
3. Steps: 40, VAE: kl-f8-anime2.ckpt, Size: 512x768, Seed: 3711767605, Model: anreal_v10, "Module: inpaint_global_harmonious, Sampler: DPM++ 2M Karras, CFG mode: Half Cosine Up, VAE hash: 63aeecb90f, CFG scale: 20, Clip skip: 2, Mimic mode: Half Cosine Up, Model hash: 59b31c16c1, Hires steps: 20, Mimic scale: 7, Hires upscale: 2.25, Hires upscaler: ESRGAN_4x2, ADetailer model: mediapipe_face_full, Interpolate Phi: 1, ADetailer version: 23.10.1, CFG scale minimum: 6.5, Denoising strength: 0.5, Scaling Startpoint: MEAN, ADetailer mask blur: 4, ADetailer model 2nd: hand_yolov8n.pt, Mimic scale minimum: 6.5, Variability Measure: AD, ADetailer confidence: 0.3, ADetailer prompt 2nd: (beautiful hand), Threshold percentile: 100, ADetailer dilate/erode: 32, ADetailer mask blur 2nd: 4, ADetailer confidence 2nd: 0.3, ADetailer inpaint padding: 32, Separate Feature Channels: True, ADetailer ControlNet model: controlnet11Models_inpaint [be8bc0ed], ADetailer dilate/erode 2nd: 32, ADetailer ControlNet module: inpaint_global_harmonious, ADetailer denoising strength: 0.4, Dynamic thresholding enabled: True, ADetailer inpaint only masked: True, ADetailer inpaint padding 2nd: 32, ADetailer negative prompt 2nd: "negative_hand-neg, fused hands", ADetailer ControlNet model 2nd: controlnet11Models_inpaint [be8bc0ed], ADetailer ControlNet module 2nd: inpaint_global_harmonious, ADetailer denoising strength 2nd: 0.4, ADetailer inpaint only masked 2nd: True

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

我测试了它以获得赏金，我不得不说我对你能从中获得的结果感到非常满意。
使用 MeinaMix、简短的提示和 ：0.7-1 的强度，您可以生成相当不错的图像，看起来不会太通用。

优秀到大家集体隐藏图片不发表！ps：up加油，作品无敌，GPT整个系列都形式感极强，你迷恋的粉橙色系统使很多经典底模再次闪闪发光！