Type | |
Stats | 1,577 |
Reviews | (232) |
Published | May 5, 2023 |
Base Model | |
Training | Steps: 5,120 Epochs: 20 |
Trigger Words | anna tekken7 anna williams |
Hash | AutoV2 E08CA35147 |
Lora trained on Anna Williams from Tekken
models used
anime: MeinaUnreal, CoffeeMix, PrimeMix
2,5D: CamelliaMIx_2.5D_V2, SoapMix
realistic: fantasticmix_real, henmix_real, bl5
regular outfit: anna williams, 1girl, china dress, bob cut, print legwear
optional: red dress, detached sleeves
Tekken 7 outfit: anna tekken7, 1girl, hair flower, bob cut, black dress, feather trim
optional: fishnet, hair ornament
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, kl-f8-anime2, blessed2.vae (for anime models)
Sampler: DPM++ SDE Karras or DPM++ 2M Karras
Steps: 20
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
Denoise strength for img2img: 0.4 to 0.7
If you want the exact same settings, just drag and drop the picture into ComfyUI
for the pictures i used comfyui