I came across an Esports player who had shared pictures of creating a LoRa model, so I thought I would share it with you.
❤️ - Ko-fi ❤️ Where creators get support from fans through donations, memberships, shop sales, and more! The original 'Buy Me a Coffee Page.;https://ko-fi.com/steverogers080863

DISCLAIMER:

I'm doing this for fun and not for commercial motives, My primary purpose is to compose a fictional character using AI with the beauty of all persons above, please support them if you love their works. I do not support any indecent images generated of all persons in this model, I do not support selling generated images of persons in this model. please refrain from what they have been posting online.

All persons in this model are 21+ and not a minor, please DO NOT post explicit nudity, suggestive, or pornographic content online using this model. Refer to CivitAi TOS for more details.

===

Request or Suggestion Rule:

1. No request for a minor!

2. The requested should have public data (not private)

3. Should have more than 30 good-quality images

4. Since all requests or suggestions are free, I will still need to like your request, otherwise its a work not a hobby for me.

5. Be a decent human being and leave a review if your request is up It takes you only 5 mins while it takes me hours, electricity, and effort to do your requests.

6. If you don't get what you want, do not review the bomb, there could be other creators' models or Lora that can get better results than mine, use theirs instead, and leave a good review on them too, support any other creators for their hard works!

NOTE: I will keep updating more actresses in versions update

===

Usage Guide

--

A. Lora Weight

• 0.1 ~ 0.4 -- is to combine multiple faces or features, or if you experience overbaking

• 0.5 ~ 0.7 -- is a balanced weight to get still accurate features but with more flexibility

• 0.8 ~ 1.0 -- is to create an accurate representation of the features.

B. CFG Scale

• A lower CFG Scale means AI gets more creative and less overbaking problems.

• A higher CFG Scale means results get very accurate with training sets and are more prone to overbaking problems.

• Dynamic CFG Threshold, using this method allows you to bump up quality and achieve photorealistic easily, but with lycoris, I found out recently that this method actually lowers the success rate, so only use it when you are composing with a Lora / model that works fine with this method.

C. Sampling Steps, High Res Fix & Upscaling

• I like to use sampling steps at 50, with high res steps at 20, or 1/2 of the sampling steps.

• I find that 2M Karras works better and faster than SDE Karras. with ERSGAN 4x+ method + ERSGAN 4x for Upscaling. This method actually gives quite good results without being too sharp or too plasticky

• Upscaling - GPFGAN @0.25, Codeformer @0.3 & Codeformer Weight @0.15 seems to yield superior results.

Enjoy.