This lora can do both the scenery and characters in style of BLAME! but for best results use the dedicated character loras for Killy, Cibo and Sanakan
https://civarchive.com/models/204863/killy-from-blame
https://civarchive.com/models/204845/cibo-from-blame
https://civarchive.com/models/204885/sanakan-from-blame
Master trigger: tsutomu nihei
Style triggers: nothing for default blame manga style, movie for the blame movie and knights of sidonia style, ova for the blame OVA style, knights of sidonia for the knights of sidonia style
Tags: tsutomu nihei, by tsutomu nihei, movie, ova, original, fanart, blame, knights of sidonia, biomega, killy, cibo, sanakan, dhomochevsky, tzuru, safeguard, black knight, silicon life, scenery, GBE, exterminator, forehead mark, arm cannon, GBEAC
Use the tag fanart as a negative tag to filter out non-original style influences
Description
FAQ
Comments (15)
So many of these images with no prompts. :(
yeah i save as JPG cause PNG takes up a lot of hard drive space so metadata gets lost
@pillowaifu jpeg has all metadata, I only save jpeg too
@omenizer mine doesnt :C
With no prompts and model to use to reproduce results, LoRAs are almost useless :/
My client doesnt save metadata to jpg for some reason, models used are abyss orange mix 2, nijidiffuse20 and a merge of aom2-easter9 and a merge of that with niji20.
As for reproducing, you could always just make something new instead of trying to do whats already been done
@pillowaifu It's not about reproducing what you've done. It's more about getting an idea of what words/triggers are used to get the desired effects.
@pillowaifu exactly, as mentioned it's not about doing 1:1 the same image, it's that without the information you cannot get anything similar, so the LoRA is useless. JPGs do not store the information, they need to be PNGs or you can write the prompts and model information in the description
@stanleypain007313 @victorc25744 i wrote all the triggering tags in the description, just write tsutomu nihei and press generate and you will get all kinds stuff, like the image of a girl in black armor against a wall was tsutomu nihei, default cube along with the obligatory best quality tags
@stanleypain007313 @victorc25744 heres how you use it https://civitai.com/images/3762342 https://civitai.com/images/3762345 https://civitai.com/images/3762344 https://civitai.com/images/3762396
this is literally all there is to it. just experiment with different styling and quality tags and find what works best for you and the model you are using, use your imagination.
Looks great! what checkpoint did you use to make the example pictures?
thanks, used nijidiffuse20, abyss orange mix 2 and 2 merges of my own of AOM2 with easter9 and a merge of that with niji20
This is the best LoRA of Nihei Tsutomu style!
glad you like it <3
This is worst LoRA of Nihei Tsutomu style!