Model
Trends
.ai
Model
Trends
.ai
nejlepší open source AI modely
Spustit lokálně
multifaces_v3.safetensors
650 MB · SafeTensor
Stahování zajišťuje Civitai. Někteří autoři vyžadují nejprve přihlášení na Civitai.
AD
<multifaces3-xl>
O tomto modelu
STABLE DIFFUSION XL 1.0 · LoRA
Vytvořil
sevenof9247
na Civitai
__html__----> left with my lora <----- , , , -----> right without and other steps <-----V........... see below ..............Vtry my other face and makeup lorahttps://civitai.com/models/151594?modelVersionId=175861so carefule, it improves and needs less steps, so play with it ...i want real faces so no symmertrie and natural skin :)ATTENTION -> weight correction 0.3-0.9 results with 25 steps DPM++2M SDE Karras almost the same images as without and 70 stepsTrigger is not essential but useful for the culture ! ;)has a great impact on lower Sampling steps <30if you use DPM++3M SDE Karras with 70 steps, you meight dont need my lora-multifacesv3 -> (group, crowd, many faces), it only generates more people, distant faces are still not so goodweight (0.3-0.8)-multifacesv2 -> (trigger: "number" faces) combine with my v3, try and tell me ;)And please its for the basemodel! so you can try or merge it and finde out ;)i used also copax and juggernaut that is already not that bad at small faces so may you dont need it!-multifacesv1 -> (trigger: group, faces) combine with my v3, try and tell me ;)--> all in all i suggest "Face Editor" or refiner for small faces !-face enhancer_v3 -> more strong in close-ups and pls register my prompts ;)face enhancer_v2 -> much more smother and better ;)please merge in your checkpoint, and tell me, btw i have less trained lorasmore hints below! ->face enhancer(learning: ran on an intel-i7 with an RTX4060, 7 hours with 300 images)its a lora file, so copy in
Poznámky autora
I found that images generated with my LoRA appear bit more amateur and less like the puppy "model_flux" style. In Better Faces, I aimed to include a diverse range of cultures and ethnicities—15 if I remember correctly—with both young and old subjects, currently focusing on women. aborigines, african, arab, arctic, brazilian, chinese, egyptian, finish, german, havaiian, indian, japanese, mongolian, russian, western my FLUX-lora responds to age prompts like __yo for 20yo, 30yo, 40yo. Weight: start with 0.3 look image post for more, i was curious also negativ has an impact^^ V........... see below ..............V Try my other face expression LoRA. If you use it at a weight of only 0.4 (without trigger words), you'll achieve more realistic results. I aim for realistic faces with natural skin, avoiding perfect symmetry. :) FLUX with Euler-Sampler and SGM-Uniform-Scheduler performs well at -> 15-25 Steps Trigger words aren't essential but are useful for specifying cultural features! and "yo" ;) Model Descriptions: FLUXV3: new image set, new caption, 15 cultures, no permission to merge! FLUX: Recognizes mature (40yo) and young (20yo) subjects with reduced bokeh effect. For higher weights, it learns faces without jewelry, so consider adding jewelry and background details if desired. pony_weak: Better for close-ups; doesn't perform well with complex prompts. Not limited to "1girl" and includes backgrounds. pony_strong: works better with more complex prompts, not only 1girl, and call background face enhancer_v3: Stronger in close-ups. Please consider using my prompts. ;) face enhancer_v2: Much smoother and improved. Multi-Face Models: multifacesv_3: (Keywords: group, crowd, many faces) Generates more people, but distant faces still need improvement. Recommended weight: 0.3-0.8. multifaces_v2: (Trigger: specify the "number" of faces) i used also copax and juggernaut that is already not that bad at small faces so may you dont need it! multifaces_v1 -> (trigger: group, faces) Additional Recommendations: Overall, I suggest using Adetailer or a refiner for enhancing small faces. Face Swap with Adetailer and Reactor The new SDXL VAE (released on 2023.07.26) is significantly better than older versions for face generation. Please try my LoRA, and you'll often get more realistic faces—not those blurred, soft ones. ;) Hints: - please, almost no negative prompt is necessary! see my articles: All SFW!
Spouštěcí slova
group of people
crowd
many faces
1,436 stažení · 396 lajků na Civitai
SHA256
9114385A5475ECE18F92450B4BA5A58CB1F49784D08FDCDD28C3E6E2E21C2A52
ID modelu ModelTrends.ai
#1945
Nahlásit obsah
Ukázkové rendery
Prompty a nastavení sdílená lidmi, kteří tyto rendery vytvořili na Civitai. Vyberte render a uvidíte, jak vznikl.
Nastavení
Sampler
DPM++ 3M SDE Karras
Kroky
42
Guidance
7
Seed
4286732063
Velikost
1024x680
Prompt
photo of a large group of people in suits and ties sitting in a large concert hall, (many faces),
Negativní prompt
painting, drawing, sketch, cartoon, anime, manga, render, watermark, bw, monochrome,
Nastavení
Sampler
DPM++ 3M SDE Karras
Kroky
70
Guidance
5
Seed
498658999
Velikost
1024x680
Prompt
photo of a large crowd of young people on a demonstration , (many faces),
Negativní prompt
painting, drawing, sketch, cartoon, anime, manga, render, watermark, bw, monochrome,
Nastavení
Sampler
DPM++ 3M SDE Karras
Kroky
70
Guidance
7
Seed
4286732063
Velikost
1024x680
Prompt
photo of a large group of people sitting in a cinema, many faces,
Negativní prompt
painting, drawing, sketch, cartoon, anime, manga, render, watermark, bw, monochrome,
Nastavení
Sampler
DPM++ 3M SDE Karras
Kroky
42
Guidance
7
Seed
4286732063
Velikost
1024x680
Prompt
photo of a large group of people in suits and ties sitting in a large concert hall, (many faces), <lora:people_010_cosine_iter4_ 0000025all_jugger_adafactor_128_64_new:0.7>
Negativní prompt
painting, drawing, sketch, cartoon, anime, manga, render, watermark, bw, monochrome,
Nastavení
Sampler
DPM++ 3M SDE Karras
Kroky
70
Guidance
5
Seed
498658999
Velikost
1024x680
Prompt
photo of a large crowd of young people on a demonstration , (many faces), <lora:people_010_cosine_iter4_ 0000025all_jugger_adafactor_128_64_new:0.7>
Negativní prompt
painting, drawing, sketch, cartoon, anime, manga, render, watermark, bw, monochrome,
Nastavení
Sampler
DPM++ 3M SDE Karras
Kroky
70
Guidance
7
Seed
4286732063
Velikost
1024x680
Prompt
photo of a large group of people sitting in a cinema, many faces, <lora:people_010_cosine_iter4_ 0000025all_jugger_adafactor_128_64_new:0.3>
Negativní prompt
painting, drawing, sketch, cartoon, anime, manga, render, watermark, bw, monochrome,
AI modely pro generování obrázků, videa a textu
multifaces3 xl je LoRA pro rodinu Stable Diffusion XL 1.0 uvedený na ModelTrends.ai, katalogu open source AI modelů pro obrázky, video a text určeném jen ke čtení. Porovnejte ho s dalšími modely Stable Diffusion XL 1.0, podívejte se na jeho skóre žhavosti a vyzkoušejte ho na PirateDiffusion nebo BitVector.
Procházet podle rodiny modelů
Procházet všechny modely