SDXL

v 1.0 Real - Maybe 😇/😈

Name: v 1.0 Real - Maybe 😇/😈
Rating: 3 (4 reviews)
Author: 🔥Mr.FiX'er

v 1.0 Real - Maybe 😇/😈 - AI Model cover image

アップロード日時: 2026/04/25 14:04
採用数: 5.2k
レビュー: 良好 (2)
Sampling Steps: 22
Sampling Method: Euler a
CFGスケール: 7.7
ネガティブ: (asinine anatomy, incompetent anatomy, anatomy disaster:1.2), A and B to C, style of B, similar to B, by B, B's signature, †, worst quality, low quality, normal quality, fused fingers, fused toes, webbed fingers, webbed toes, too many fingers, too few fingers, incorrect hand structure, clawed hand, wrong finger count, finger growing from hand, palm with no fingers, mutated hands, ugly hands, bad hands, poorly drawn hands, (hand with 6 fingers, hand with 4 fingers:1.8),
権限: 画像生成と共有を許可する
ユーザーによるモデルのダウンロードを許可する
生成画像の商用利用を許可する

詳細説明

                                ⚠️---- Instruction hard MODEL ----⚠️

                               ➡️"Key principles" for your settings;⬅️

1️⃣. Start with quality enhancers: masterpiece, best quality, amazing quality, very aesthetic, absurdres, high resolution, 8k, detailed,

(cinematic), (cinematic film:1.2), highly detailed, digital art, <- this is the basis;

2️⃣. Use CFG between 1.9 and 5.5: the image responds well within this range without overloading. The standard CFG is 5.5, but do not increase it above this value, otherwise errors and distortions will occur MODEL;

3️⃣.For STEPS: You may need to perform more STEPS for this model, but this won't affect your design;

4️⃣. Working with LoRA: not all is accurate for this MODEL. You may need to use more LoRA STRENGTH for this MODEL;

5️⃣. Be specific in your description: Instead of "beautiful picture," write "beautiful woman with long brown hair in a white dress." Instead of "epic landscape," write "epic mountain range under northern lights." Control the light and atmosphere: cinematic lighting, soft natural lighting, volumetric light, atmospheric – this greatly affects the mood MODEL.

                                                  >> OLD NEGATIVE <<

(score_6, score_5, score_4:-1.4), (worst quality, low quality, normal quality:-3.7), lowres, blurry, jpeg artifacts, watermark, text, (3d render, cgi:0.15), (semi-realistic, semi-hyperrealistic:-0.9), (color cast, wrong colors, vignette, color noise, bad color balance, too warm, too cold, blue cast, monochrome, oversaturated, dull, faded, wrong color temperature, unnatural skin tone, sickly skin:-0.2), ugly, bad anatomy, bad proportions, bad face, deformed, impossible anatomy, contorted, rubber limbs, broken anatomy, extra limbs, missing limbs, bad hands, mutated hands, broken hands, broken fingers, mutated fingers, abnormal fingers, deformed fingers, webbed fingers, claw fingers, fused fingers, missing fingers, white sclera, all white eyes, white cornea, no iris, no pupil, blank white eyes, dead white eyes, white eyeball, (realistic photo, photograph, real person, photorealistic, hyperrealistic:0.5), documentary style, candid shot, unposed, natural lighting, imperfect skin, pores, wrinkles, blemishes, stray hairs, messy hair, unkempt costume, wrinkled costume, dirty costume, worn out fabric, cheap fabric, visible seams, cosplay malfunction, costume damage, out of frame, cropped, duplicate face, cartoon, illustration, anime, painting, 3d render, CGI, sketch, abstract, (NSFW:1.3),
Sampling Steps: 11
Sampling Method: DPM++ 2M Karras
CFG Scale: 6.1

詳細説明

                                ⚠️---- Instruction hard MODEL ----⚠️

                               ➡️"Key principles" for your settings;⬅️

1️⃣. Start with quality enhancers: masterpiece, best quality, amazing quality, very aesthetic, absurdres, high resolution, 8k, detailed,

(cinematic), (cinematic film:1.2), highly detailed, digital art, <- this is the basis;

2️⃣. Use CFG between 1.9 and 5.5: the image responds well within this range without overloading. The standard CFG is 5.5, but do not increase it above this value, otherwise errors and distortions will occur MODEL;

3️⃣.For STEPS: You may need to perform more STEPS for this model, but this won't affect your design;

4️⃣. Working with LoRA: not all is accurate for this MODEL. You may need to use more LoRA STRENGTH for this MODEL;

5️⃣. Be specific in your description: Instead of "beautiful picture," write "beautiful woman with long brown hair in a white dress." Instead of "epic landscape," write "epic mountain range under northern lights." Control the light and atmosphere: cinematic lighting, soft natural lighting, volumetric light, atmospheric – this greatly affects the mood MODEL.

                                                  >> OLD NEGATIVE <<

(score_6, score_5, score_4:-1.4), (worst quality, low quality, normal quality:-3.7), lowres, blurry, jpeg artifacts, watermark, text, (3d render, cgi:0.15), (semi-realistic, semi-hyperrealistic:-0.9), (color cast, wrong colors, vignette, color noise, bad color balance, too warm, too cold, blue cast, monochrome, oversaturated, dull, faded, wrong color temperature, unnatural skin tone, sickly skin:-0.2), ugly, bad anatomy, bad proportions, bad face, deformed, impossible anatomy, contorted, rubber limbs, broken anatomy, extra limbs, missing limbs, bad hands, mutated hands, broken hands, broken fingers, mutated fingers, abnormal fingers, deformed fingers, webbed fingers, claw fingers, fused fingers, missing fingers, white sclera, all white eyes, white cornea, no iris, no pupil, blank white eyes, dead white eyes, white eyeball, (realistic photo, photograph, real person, photorealistic, hyperrealistic:0.5), documentary style, candid shot, unposed, natural lighting, imperfect skin, pores, wrinkles, blemishes, stray hairs, messy hair, unkempt costume, wrinkled costume, dirty costume, worn out fabric, cheap fabric, visible seams, cosplay malfunction, costume damage, out of frame, cropped, duplicate face, cartoon, illustration, anime, painting, 3d render, CGI, sketch, abstract, (NSFW:1.3),
Sampling Steps: 11
Sampling Method: DPM++ 2M Karras
CFG Scale: 6.1

v 1.0 Real - Maybe 😇/😈

v 1.0 Real - Maybe 😇/😈

おすすめLoRA

詳細説明

コメント

おすすめLoRA

詳細説明

コメント