KANJI KOGANEGAWA from HAIKYU!

---

?Main Trigger Tags?

male focus, KoganegawaHU, 1boy, anime screencap

⚓Other Unique Tags⚓

• Check out Gallery (Highly Recommended)

• Experiment on your own

• Do not use all Other Unique Tags on a generation to get the character's outfit.

?Dadada's Berries?

• Use with Pony Diffusion XL Based models

• Remove "Anime screencap" for flexibility

• Remove "Other Unique Tags" or the character name within that unique tag for flexibility

• Change ___ focus if causing artefacts for ex. (siblings focus ---> Best friends focus)

• Reduce tag weight with [_tag goes here_]

• Reduce weight but hold character likeness with (__input biological tag__)

?GENDER BENDING?

• Gender bend by replacing "1boy, male focus" with "1girl, female focus" vice versa

• Use with Male centered Pony Diffusion XL Based models or LoRA styles, vice versa

• Reduce LoRA weight appropriately when prompting for male, vice versa or...

• Add tags related to the male anatomy (abs, pecs, penis, etc.), vice versa

?MODEL CONTROL?

• Better Checkpoint infusion with reduction of LoRA weight and CFG increase

• Some checkpoint models will force additional elements, add undesired elements in negative or specify in positive with "no type undesired element"

• When tagging for "full body" i recommend using Adetailer to fix facial mishaps

??‍?BAKING OWN OC!??‍?

• When MIXING Character LoRAs (mine), reduce weight to 0.6-0.8. while keeping their mother tags in.

• When MIXING Character LoRAs (not mine but with mine), reduce my LoRAs weight appropriately while keeping their mother tags in.

• When influencing own imagination with my LoRA, reduce weight appropriately

---

Koganegawa in TENSORART! @0@

https://tensor.art/models/792836029510987520

---

?NOTE:

Ask for my permission if you wish to reprint or upload my models on sites I have no knowledge of. I would love it if my LORAs can only be downloaded through CIVIT. Thank you <333

描述:

Trained on images consisting of:

• Screenshots

For more information, download the SafeTensor file and check on https://lorainfo.tools/

===

Huhu, ILXL is crazy good with character likeness, showcase images does not use Adetailer, I wanna test more checkpoint merges with it!

===

This version has the same dataset with its pony counterpart because I want to compare. Tho parameters are different based on CIVIT's default settings (Except the 1epoch ofc), O>O

===

ILXL has more room for compositions!!! XD, At first I was skeptical cause of how simple the initial images came out but once the elements are prompted---crazy. Checkpoints that look simple but have updated datasets are ?

===

Pony or ILXL,

even though text is still a mess--likeness plus a checkpoint that listens are an upgrade from "let it be a possibility" pony. Will do more chara/D. tests on ILXL! Especially with scene control Kyaaaaaaa!⸜(｡˃ ᵕ ˂ )⸝♡

训练词语: KoganegawaHU,blonde-black_KoganegawaHU_multicolored hair, brown_KoganegawaHU_eyes