Model
Trends
.ai
Model
Trends
.ai
the best open source ai models
Run Locally
rouwei_v080Vpred.safetensors
6.46 GB · SafeTensor · fp16 · pruned
rouwei_v080Vpred_trainingData.zip
1 MB · Other
Download
Downloads are served by Civitai. Some creators ask for a Civitai login first.
AD
Install / Run on Graydient
Requires membership
<rouwei-epsilon06-xl>
About This Model
STABLE DIFFUSION XL 1.0 · Base Models
Created by
Minthybasis
on Civitai
by Minthybasis - Advanced illustrious model, somewhere between an anime and furry style. Hard to work with but gives incredible range of lightning and poses for those that are up for the challenge. This is a large scale fine-tune with 4.5 million pictures using danbooru tags, so e621 isn't understood. Use guidance 5-7 at 20-30 steps. To get a sharp photo, render three steps: upscale by 1.5x at strength 0.6, then again at 0.3x at strength 0.55. General styles: 2.5d, anime screencap, bold line, sketch, cgi, digital painting, flat colors, smooth shading, minimalistic, ink style, oil style, pastel style Booru tags styles: 1950s (style), 1960s (style), 1970s (style), 1980s (style), 1990s (style), 1990s (style), animification, art nouveau, pinup (style), toon (style), western comics (style), nihonga, shikishi, minimalism, fine art parody and everything from this group. Characters: Use full name booru tag and proper formatting, like "karin_(blue_archive)" -> "karin \(blue_archive\)", use skin tags for better reproducing, like "karin \(bunny \(blue_archive\)". Lots of Tail/Ears-related concepts: tail censor, holding own tail, hugging own tail, holding another's tail, tail grab, tail raised, tail down, ears down, hand on own ear, tail around own leg, tail around penis, tail through clothes, tail under clothes, lifted by tail, tail biting, tail insertion, tail masturbation, holding with tail, ...and many others with natural text. The majority works perfectly, some requires rolling. Bri
Creator notes
In depth retraining of Illustrious to achieve best prompt adherence, knowledge and state of the art performance. Big dreams come true The version number is just an index of current final release, not a fraction of the planned training. Large scale finetune using gpu cluster with a dataset of ~13M pictures (~4M with natural text captions) - Fresh and wast knowledge about characters, concepts, styles, cultural and related things - The best prompt adherence among SDXL anime models at the moment of release - Solved main problems with tags bleeding and biases, common for Illustrious, NoobAi and other checkpoints - Excellent aesthetics and knowledge across a wide range of styles ( (), including hundreds of unique cherry-picked datasets from private galleries, including those received from the artists themselves) - High flexibility and variety without stability tradeoff - No more annoying watermarks for popular styles thanks to clean dataset - Vibrant colors and smooth gradients without trace of burning, full range even with epsilon - Pure training from Illustrious v0.1 without involving third-party checkpoints, Loras, tweakers, etc. There are also some issues and changes compared to the previous version, please RTFM. Dataset cut-off - end of April 2025. Features and prompting: Important change: When you are prompting artist styles, especially mixing several, their tags MUST BE in a separate CLIP chunk. Just add BREAK after it (for A1111 and derivatives), use conditioning concat node (for Comfy) or at least put them in the very end. Otherwise, significant degradation of results is likely. Basic: The checkpoint works both with short-simple and long-complex prompts. However, if there are contradictory or weird things - unlike with others they won't be ignored affecting the output. No guide-rails, no safeguards, no lobotomy. Just prompt what you want to see and don't prompt what shouldn't be on the picture. If you want to have a view from above - don't put ceiling into positive, if you want to have crop view with head out of frame - don't make detailed description of character facial features, and so on. Pretty simple but sometimes people are missing it. Version 0.8 comes with advanced understanding of natural text prompts. It doesn't mean that you are obligated to use it, tags only - completely fine, especially because understanding of tags combinations is also improved. Do not expect it to perform like Flux or other models based on T5 or LLM text encoders. The whole size ot SDXL checkpoint is less then only that text encoder, in addition illustrious-v0.1 which is used as the base completely forgot a lot of general things from vanilla sdxl-base. However, even in current state it works much better, allows to do new things usually impossible without external guidance, as well making manual editing, inpainting, etc more convenient. To achieve best performance you should keep track of CLIP chunks. In SDXL the prompt is separated into a chunks of 75 (77 including BOS and EOS) tokens, that are processing by CLIP separately, and only then are concatinating and comes as conditions to unet. If you want to specify some features for character/object and separate them from other prompt parts - make sure they are in the same chunk and optionally separate it with BREAK. It will not solve problem of traits mixing completely, but can reduce it improving overall understanding, since text encoders on RouWei are able to process the whole sequence, not individual concepts better then others. Dataset contains only booru-style tags and natural text expressions. Despite having a share of furries, real life photos, western media, etc. all captions have been converted to classic booru style to avoid a number of problems from mixing of different systems. So e621 tags won't be understanded properly. Sampling parameters: - ~1 megapixel for txt2img, any AR with resolution multiple of 32 (1024x1024, 1056x, 1152x, 1216x832,...). Euler_a, 20..28steps. - CFG: for epsilon version 4..9 (7 is best), for vpred version, 3..5 - Sigmas multiply may improve results a bit, CFG++ samplers work fine. LCM/PCM/DMD/... and exotic samplers untested. - Some schedulers doesn't work well. - Highresfix - x1.5 latent + denoise 0.6 or any gan + denoise 0.3..0.55. - For vpred version lower CFG 3..5 is needed! For vpred version lower CFG 3..5 is needed! Quality classification: Only 4 quality tags: masterpiece, best quality for positive and low quality, worst quality for negative. Nothing else. Actually you can even omit positive and reduce negative to low quality only, since they can affect basic style and composition. Meta tags like lowres have been removed and don't work, better not to use them. Low resolution images have been either removed or upscaled and cleaned with DAT depending on their importance. Negative prompt: worst quality, low quality, watermark That's all, no need of "rusty trombone", "farting on prey" and others. Do not put tags like greyscale, monochrome in negative unless you understand what are you doing. Extra tags for brightness/colors/contrast section below can be used Artist styles: , (also can be found in "training data"). Used with "by " it's mandatory. It will not work properly without it. "by " is a meta-token for styles to avoid mixing/misinterpret with tags/characters of similar or close name. This allows to have a better results for styles and at the same time avoid random style fluctuation that you may observe in other checkpoints. Multiple give very interesting results, can be controlled with prompt weights and spells. YOU MUST ADD BREAK after artists/style tags (for A1111) or concat conditioning (for Comfy) or put them in the very end of your prompt. For example: by kantoku, by wlop, best quality, masterpiece BREAK 1girl, ... General styles: 2.5d, anime screencap, bold line, sketch, cgi, digital painting, flat colors, smooth shading, minimalistic, ink style, oil st
6,515 downloads · 576 likes on Civitai
SHA256
1A40B1BABCA6A143DE9E98BCC4EF1277FF688B33892DF5DB683EE25753C0FD3A
ModelTrends.ai Model ID
#14762
Flag content
Example renders
Prompts and settings shared by the people who made these renders on Civitai. Pick a render to see how it was made.
Settings
Guidance
3
Size
1660x1956
Prompt
wide shot, wallpaper, fantasy, epic, 1girl, solo, adventurer, standing, mountains, (explosion:1.1), (nuclear explosion:1.1), (mushroom cloud:1.1), river, forest, (shockwave:1.1), outdoor, rocks, hut
Negative prompt
pixel art, pixelated, furry, cloud, eruption, volcano, by bkub, by matsunaga kouyou, loli, bad quality, worst quality, sketch, watermark, symmetry, plain simple background, 3d, realistic, photo
Settings
Guidance
4
Size
1510x2085
Prompt
1girl, solo, (ghost:1.1), grey hair, very long hair, blue eyes, pale skin, expressionless, hair over one eye, ethereal, wedding dress, veil, tiara, gloves, ribbon, torn clothes, chain, floating, holding bouquet, looking at viewer, arms at sides, subtle pose, translucent, castle, indoors, candle, cobweb, ruins, night, moonlight, stone wall, eerie lighting, foggy
Negative prompt
loli, bad quality, worst quality, sketch, watermark, symmetry, plain simple background, (blank eyes, empty eyes, simple eyes:0.8), symmetry
Settings
Guidance
5
Size
2013x2530
Prompt
1girl, solo, angel, angel wings, white wings, blonde hair, long hair, blue eyes, large breasts, determined, closed eyes, hair over one eye, white dress, armor, red cape, sword, holding sword, standing, legs apart, wind, stormy sky, ocean, cliff, lightning, glowing eyes, halo, from side
Negative prompt
loli, bad quality, worst quality, watermark, censor, censored, symmetry, (blank eyes, empty eyes, simple eyes:0.9), (sparks, dust, noise:1.0)
Settings
Guidance
4
Size
1617x1972
Prompt
1girl, solo, blonde hair, blue eyes, light blush, dirty face, sweat, determined expression, hair over one eye, hair between eyes, military uniform, historical, 19th century, epaulettes, aiguillette, torn clothes, black boots, fur hat, belt, white gloves, sash, gold trim, standing, leaning on weapon, holding musket, gun, looking at viewer, wind lift, hair spread out, tired expression, battlefield, smoke, ruins, sunset, cloudy sky, grass, outdoors
Negative prompt
pixel art, pixelated, abstract background, depth of field, blurry, loli, bad quality, worst quality, sketch, watermark, symmetry, plain simple background, (blank eyes, empty eyes, simple eyes:0.8), (sparks, dust, noise:1.0)
Settings
Size
1510x2203
Prompt
1girl, solo, (monster girl:1.3), curly hair, grey hair, very long hair, magenta eyes, licking lips, thick lips, small breasts, flaming eyes, fiery wings, red and orange flame, flying, midair, latex bodysuit, skin tight, zipper, (unzipped:1.1), (navel:1.1), (groin:1.1), midriff, nude, choker, eye mask, latex gloves, shiny latex, shiny clothes, latex, covered nipples, magic, spell, (swirl:1.1), (ice magic:1.1), crystal cave, ice crystals, crystal, snow, fantasy, detailed snowflakes, black trench coat, fur trim
Negative prompt
loli, bad quality, worst quality, sketch, watermark, symmetry, plain simple background, (blank eyes, empty eyes, simple eyes:0.8), skinny, 3d, realistic, photo
Settings
Guidance
5
Size
1574x2090
Prompt
1girl, solo, elf, (frieren:1.1), (white hair:0.7), no earrings, alternate hairstyle, (motoyui:1.3), high ponytail, black lips, alternate costume, dark persona, (goth fashion:1.1), hexagon print, honeycomb \(pattern\), cross-laced corset, half-turned pose, sitting on chest, leaning back, finger to mouth, (holding book:1.1), looking at object, (will-o'-the-wisp \(mythology\) light:0.9), treasure chest, mimic chest, horror \(theme\), dungeon, full body
Negative prompt
loli, bad quality, worst quality, sketch, watermark, censor, censored, symmetry, (blank eyes, empty eyes, simple eyes:0.9), (sparks, dust, noise:1.0), (earrings:1.2), limited palette, sidelocks, floral print, twintails, cross, day, coin, pixel art, book
Settings
Sampler
DPM++ 2M SDE
Steps
28
Guidance
6
Seed
2406612522
Size
1024x1536
Prompt
by ranko no ane, masterpiece, best quality, BREAK, checkered sand, 1girl, long black hair, faceless, black robe, pale skin, leaning on object, curled fingers, wide smile, evil smile, collage, photo bashing, photoshop \(medium\), BREAK, black wolf, giant piano, black streaks, dithering, zoom layer, halo,
Negative prompt
worst quality,
Settings
Size
1585x2358
Prompt
1girl, demon girl, solo, thick eyelashes, long hair, hair over one eye, large breasts, demon horns, multiple horns, mixed horns, curled horns, smoking, cigarette, holding cigarette, grin, (scar on mouth:1.1), looking at viewer, upper body, earrings, indoors, smoke, coat, showing middle finger with other hand, black nails, hands up, hallway, (table:0.9), (cup:0.9)
Negative prompt
loli, bad quality, worst quality, sketch, watermark, censor, censored, symmetry, (blank eyes, empty eyes, simple eyes:0.9), (sparks, dust, noise:1.0), cigarette in mouth
AI Models for Image, Video & Text Generation
rouwei epsilon06 xl is a base model for the Stable Diffusion XL 1.0 family listed on ModelTrends.ai, a read-only catalog of open source AI image, video and text models. Compare it with other Stable Diffusion XL 1.0 models, check its heat score to see how it is trending, and open it on PirateDiffusion or BitVector to try it.
Browse by Model Family
Browse all models