Sign In

入須 冬実/Irisu Fuyumi(Hyouka)

204
1.4k
269
8
Updated: Oct 5, 2024
characteranimefemalehyouka
Verified:
SafeTensor
Type
LoRA
Stats
893
269
Reviews
Published
Apr 9, 2023
Base Model
SD 1.5
Trigger Words
fuyumi
Hash
AutoV2
C7A1319295

TRY TO GET SOME KNOWLEDGE OF VERSIONS BEFORE DOWNLOADING. They are different.

2.0a(the default one) is the advanced version of 1.0 that excludes all the background. It's a hard trained one and hard to control but it can give you the best result. if you do not use block you can use a lower weight as the sample shows. Higher weight is also good if you just want to make the pictures that same as the anime.

2.0b is the lite version that FITS EVERYONE especially beginners or those who do not use block or do not like hard-trained models. It is more like my previous shizuka 3.0b that is flexible. The samples are just randomly generated pictures using hires fix and random prompt. You can see the result. Higher weight can be used.

Based on anime. If you want the best light effect try not to use block. I will try to face the background issue in next version. 0.7-0.8 seems to be good.