艾希 魔女 Ashe

https://s2.xiaonaofu.top/material/1ef93ca1-8e03-4229-91a1-da39aa01751f.jpeghttps://s2.xiaonaofu.top/material/6b74e280-3013-4a45-8c06-1c4dc7b7ccf9.jpeghttps://s2.xiaonaofu.top/material/445d9f41-e121-405b-aa44-532138340eb8.jpeghttps://s2.xiaonaofu.top/material/331faa93-4c79-419a-a6d3-c72e076db3a5.jpeghttps://s2.xiaonaofu.top/material/65d5133c-87da-4189-a019-69c53635620d.jpeghttps://s2.xiaonaofu.top/material/b4f645c0-3ab0-435f-a536-e6b0f5992372.jpeghttps://s2.xiaonaofu.top/material/73b411cd-7e2a-4bd2-bc5c-3a2041a40071.jpeghttps://s2.xiaonaofu.top/material/3f19504d-1384-45a9-b0d1-d904549d74e9.jpegcoven ashe

LoRA trained on Coven Ashe from League of Legends

requested by @BangBangFukanawa

trained on 12 images lol

models used

anime: YabaL_Mix V3,

2,5D: CamelliaMIx_2.5D, HELLmix, DarkSun, MeinaUnreal

tags: coven ashe, 1girl,

optional tags: hood, covered eyes, cape, dress,

negative embeddings: embedding:bad-hands-5, embedding:ng_deepnegative_v1_75t, embedding:bad-artist, embedding:EasyNegative

Recommended settings:

Make sure you are aware on the usage instructions of LORA

VAE: vae-ft-mse-840000-ema-pruned (for realistic models & 2,5D), kl-f8-anime2, blessed2.vae (for anime models)

Sampler: DPM++ SDE Karras or DPM++ 2M Karras

Steps: 20 to 35

Clipskip: 2

Upscaler : img2img in comfyui using 4x-UltraMix_Smooth

CFG Scale : 5 to 9 (mostly 7 and 8)

LORA weight for txt2img: 0.5-1 (for the examples i mostly used 0.8)

Denoise strength for img2img: 0.4 to 0.7

for the pictures i used comfyui

Hide

发表回复

后才能评论