Comradeship LU (Lumina)

战友之情必须延续。

@ = Karcher-merge (仅限 DiT)

(a @ b @ c) = Karcher_merge.py --models a.safetensors b.safetensors c.safetensors

v2 许可证为 fair-ai-public-license-1.0-sd

comradeshipLU-v2T28

牛肝菌的反击

Comradeship LU v2T25 + (LoRA 2A × 2.2) = Comradeship LU v2T27

NetaYume Lumina v3.0 + (LoRA 2A × 2.2) = NetaYume Lumina Hentai MOD

Comradeship LU v2T27 @ NetaYume Lumina Hentai MOD = Comradeship LU v2T28

comradeshipLU-v2T25

我是牛肝菌

(infallssssuumnnnn/checkpoint-e6_s138000 @ infallssssuumnnnna5n/checkpoint-e3_s95000) = Comradeship LU v2T23

(infallssssuumnnnn/checkpoint-e6_s138000 & infallssssuumnnnna5n/checkpoint-e3_s95000) = Comradeship LU v2PH1

Comradeship LU v2T23 + ((Comradeship LU v2T23 - Comradeship LU v2PH1) × -3.5) = Comradeship LU v2T24

Comradeship LU v2T23 @ Comradeship LU v2T24 = Comradeship LU v2T25

叠层刺猬菌

comradeshipLU-v2T22

neta-lumina-v1.0 + LoRA A + LoRA C + ... = Comradeship LU v2T22

https://huggingface.co/hanzogak/comradeshipLUv2

辣味肉丸

comradeshipLU-v2T18

neta-lumina-alpha-aes-full-a5n-ep2-s58714.pth + LoRA 1A + (LoRA 1B × -1) + LoRA 1C + LoRA A + LoRA C + (LoRA D × -0.35) + LoRA E + LoRA 1C + (LoRA 1D × 1.3) = Comradeship LU v2T18

DiT 过载

comradeshipLU-v2T14

简单美学 LoRA 合并 × 4.5 = 极致美学

Comradeship LU v2T9 + A LoRA + (B LoRA × -1) + C LoRA + (D LoRA × -1) + E LoRA + (F LoRA × -0.55) + G LoRA + (H LoRA × -0.25) + (I LoRA × 0.5) = Comradeship LU v2T14

DiT 梦想者

comradeshipLU-v2T12

简单美学 LoRA 合并 × 3 = 极致美学

Comradeship LU v2T9 + LoRA A + (LoRA B × -1) + LoRA C + (LoRA D × -1) + LoRA E + (LoRA F × -0.55) = Comradeship LU v2T12

comradeshipLU-v2T10

简单美学 LoRA 合并

(Neta Lumina Alpha full_round5_ep5_s182000 @ Neta Lumina Alpha NetaAniLumina_Alpha_full_round7_ep4_s82000 @ Neta Lumina Alpha full_roundnnnn_ep1_s38000 @ Neta Lumina Alpha full_roundnnnn_ep3_s63858 @ Neta Lumina Alpha full_roundnnnn_ep4_s85144 @ Neta Lumina Alpha full_roundnnnn_ep5_s106430 @ Neta Lumina Alpha full_roundnnnn_ep6_s127716 @ Neta Lumina Alpha full_roundnnnn_ep7_s149002 @ Neta Lumina Alpha full_roundnnnn_ep7_s160000) = km1

(beta-lumina-beta-0624-aes-experimental - km1) × 0.5 + (km1) = Comradeship LU v2T9

Comradeship LU v2T9 + LoRA A + (LoRA B × -1) = Comradeship LU v2T10

comradeshipLU-v2T6

((neta-lumina-beta-0624 @ Neta Lumina Alpha(full_round7_ep4_s68303)) + ((LeX-Lumina - Lumina-Image-2.0) × 0.2) × 0.5) + ((neta-lumina-beta-0624 @ Neta Lumina Alpha(full_roundnnnn_ep7_s149002)) + ((LeX-Lumina - Lumina-Image-2.0) × 0.2) × 0.5) = Comradeship LU v2T6

comradeshipLU-v2T3

这是一个测试用的 Karcher-merge 模型。（https://github.com/win10ogod/Karcher-merge）

(neta-lumina-beta-0624 @ Neta Lumina Alpha(full_round7_ep4_s68303)) + ((LeX-Lumina - Lumina-Image-2.0) × 0.2) = Comradeship LU v2T3

comradeshipLU-v1T2

这是一个基于 Illustrious-Lumina-v0.03 的合成人像模型，为测试合并模型。

Illustrious-Lumina-v0.03 + ((LeX-Lumina - Lumina-Image-2.0) × 0.6) = Comradeship LU v1T2