LoRA trained on Akshan from League of Legends
requested by: @LadyNuggets
tags: akshan, 1boy, male focus, cape, pants
models used
anime: Sweet-mix - v1.8 - flat, MeinaPastel - V5
2,5D: PastelDiffusedMix - V2.1, LusterMix - v1.5, ReV Animated - v1.2.2, ZavyMix - a2
realistic: Paragon V1.0
negative embeddings: bad_prompt, negative_hand, ng_deepnegative_v1_75t, bad-artist, 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(euler a for img2img upscaling) from time to time im also using DPM++ 2M alt Karras
Steps: 20 to 35
Clipskip: 2
Upscaler : GitHub - pkuliyi2015/multidiffusion-upscaler-for-automatic1111 4x-UltraSharp(for 2,5D & realistic), 4xAnimesharp(for anime)
CFG Scale : 5 to 9 (mostly 7)
LORA weight for txt2img: 0.5-1 (for the examples i mostly used 0.8 and 1)
Denoise strength upscaling: 0.3 to 0.55
If you want to autocomplete the tags & trigger im using, I recommend using Booru style tag autocompletion for AUTOMATIC1111's Stable Diffusion web UI
For full body shots i recommend using: ComfyUI Impact Pack
or the a1111 alternative GitHub - Bing-su/adetailer
Description
FAQ
Comments (6)
this looks genuinely perfect even without using any sort of fine tuning methods used on my end. I'd love to see how you manage to get all of your loras trained so accurately
Nothing special, honestly, besides maybe the fact that I train on 768 resolution. Other than that, it's pretty basic. I sometimes need multiple tries though to get it to work well, luckily not on this one. Im Glad you enjoy it! :)