
LoHa of Daiwa Scarlet (umamusume). Put LoHa file in your stable-diffusion-webui/models/lora folder and write LoRA notation in prompt to applying LoHa. <lora:your_loha_name:weight>
All of my works is free. If you'd like to support me, feel free to buy me a coffee. ?
Recommended options
-
LoRA weight around 0.7
-
Trigger words
-
base
-
daiwa scarlet \(umamusume\)
-
-
Racing competition costume :
-
hair bow, red bow, tiara, epaulettes, puffy long sleeves, juliet sleeves, black bowtie, white shirt, framed breasts, blue jacket, blue dress, center frills, underbust, layered skirt, garter straps, white thighhighs, white footwear
-
-
SSR Power(Swimsuit):
-
official alternate costume, official alternate hairstyle, low twintails, hair ornament, hair bobbles, navel, red bikini, frilled bikini, bikini skirt, swimsuit, frills
-
-
Santa costume :
-
official alternate costume, official alternate hairstyle, single ear cover, christmas, tiara, bare shoulders, detached sleeves, jewelry, fur trim, brooch, red gloves, fur-trimmed gloves, red dress, fur-trimmed dress, plaid dress, black skirt, black thighhighs, thigh boots, high heel boots, black footwear
-
(optional) gift box, holding gift, holding staff, snowflakes
-
-
Casual :
-
hair bow, red bow, tiara, puffy short sleeves, bare shoulders, off-shoulder shirt, collarbone, white shirt, frills, blue skirt, plaid skirt, pleated skirt
-
(optional) shoulder bag
-
-
Tracen summer uniform
-
tracen school uniform, summer uniform, serafuku, puffy short sleeves, bowtie, horseshoe ornament, tiara, sailor collar, sailor shirt, frills, white skirt, purple shirt, miniskirt, zettai ryouiki, white thighhighs, brown footwear, shoes,
-
-
Use a model derived from or mixed with animefull model.
-
Settings
-
DPM++ Series(SDE Karras, 2M Karras, etc.)
-
about 20 steps, CFG scale 3.5~6.5
-
I recommend kl-f8-anime2.vae
-
CLIP skip = 1
-
Use Hires.fix to get higher quality image
-
Upscaler
-
Latent, Latent (nearest-exact), Latent (bicubic antialiased) or other Latent series.
-
-
Denoising strength 0.50~0.65
-
LoRA training info
-
trained on sd-scripts by kohya_ss.
-
Thank you all a lot!
-
based on Animefull-pruned model
-
186 images of
-
4x low quality 56 images
-
8x medium quality 61 images
-
16x high quality 69 images
-
=> 1 epoch = 1816 images
-
-
replaced character feature tags with daiwa scarlet \(umamusume\)
-
1girl, horse tail, twintails, horse girl, red hair, animal ears, brown eyes, very long hair, tail, bangs, fang, skin fang, hair between eyes, tail through clothes, horse ears, hair intakes, long hair, solo, pink eyes, brown hair, red eyes, antenna hair
-
I think these tags are
-
the essentials of Daiwa Scarlet
-
or unnecessary for training because of duplication.
-
-
-
resolution : 768x768 with aspect ratio bucketing.
-
other training settings is included in training_data.zip or metadata in LoRA file.
All uploaded images are generated by AOM3A1
And, special thanks to my best friend. DH.
描述:
LoHa of Daiwa Scarlet (umamusume). Please install extension on WebUI.
After installing extension and put LoHa file in your sd/models/lora folder, use LoRA notation in prompt to applying LoHa. <lora:name:weight>
训练词语: daiwa scarlet \(umamusume\)
名称: daiwa_scarlet_loha-000006.safetensors
大小 (KB): 12376
类型: Model
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success
名称: img.zip
大小 (KB): 860572
类型: Training Data
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success