Join QQ group to have the opportunity to test the model in advance

You can also exchange experiences together, this is a channel where both novices and experts can communicate freely, welcome your participation, hoping to help more friends who are new to SD

Don't lose confidence because of some indifference, if you need it     Group number:226877470

Local running will provide a better experience than online generation

Suggest friends configuring SD locally to download and run the image locally

Trained based on Maigv2 Enhanced Version

Moyou 1060 repair output is very good

Trigger word is bbmb

Initial size is 512×768

Lora weight should be around 0.8

Can be adjusted as needed

High-resolution repair must be enabled

Redrawing magnitudearound 0.2 is optimal

Algorithm largely determines the character's skin color

Iteration steps:30-50

Sampling method:DPM++ 2M Karras/Restart

Others are free to try

Full body image requiresAD to be enabled

My ex-girlfriend mentioned it to me, so I casually trained one.

Welcome Hamabe Minami to join the Cyber Eternal family!

(Note: Using the model to produce images is purely personal behavior, the purpose of the image output is for personal appreciation only, and it is not allowed to use the model for any form of commercial activities, nor to disseminate or publish false information with the images generated by the model)