Type | |
Stats | 1,172 1,090 |
Reviews | (178) |
Published | May 12, 2023 |
Base Model | |
Training | Steps: 4,950 Epochs: 25 |
Trigger Words | juliaclassic |
Hash | AutoV2 DC87278DE4 |
Lora trained on Julia Chang from Tekken
tags: juliaclassic, 1girl
optional: twin braids, headband, crop top, denim skirt, feather hair ornament
models used:
2,5D: MeinaUnreal - V3, majicMIX fantasy - v2.0, PsyGon Animerge - v1.1, CamelliaMIx_2.5D_V2
realistic: fantasticmix_real, henmix_real
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: 0.6-1 (0.8 used in the pictures)
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