In this version I used a slightly different method of training and as a result the quality of the eyes and their colour correctness are improved, also when comparing the 1st and 2nd version the 2nd one shows itself better in sharpness and flexibility
模型描述
loRA of Washington from Azur Lane
I use it on weights from 0.5 to 0.8 and I like the results, you are welcom to experiment
