ShoheiAI/boysfactoryXL

Ogawa Hiroshi (Boys Factory) Style

Prompt

score_9, score_8_up, score_7_up, score_6_up, source_anime, outdoors, forest, from side, 1boy, solo, black hair, brown eyes, scout uniform, brown shirt, brown shorts, neckerchief, standing, looking at viewer, salute <lora:boys_factoryXL_AMXL_v3b:1>

Negative Prompt

bara, source_pony, (hands:1.2)

Model description

This LORA is designed to recreate the style of Boys Factory by Ogawa Hiroshi. No trigger words needed, but read the description of the score tags for details. This current version tends to have difficulty with multiple characters in a scene (it tends to make 3 people when you ask for 2), so that's something I have to work on still in a future update.

Please support the original artist wherever possible so they can continue to produce their art. Without them, LORAs such as these would not be possible.

This model is also available on TensorArt and CivitAI: https://tensor.art/models/743721377674271091/Ogawa-Hiroshi-(Boys-Factory)-Style-v3b https://civitai.com/models/487524/ogawa-hiroshi-boys-factory-style?modelVersionId=542123

USAGE NOTES:

Recommended to be used at 0.9 strength. The style is mostly going to affect characters since all but 4 of the images in the dataset have simple backgrounds.

This LORA was trained specifically on Autism Mix, but I've seen good results with other checkpoints. Specifically SakuraSakuma Mix by Zaizen. Other popular mixes like Luminaverse and Incursio's Meme Diffusion have a tendency to make the characters much more feminine.

NOTE: While not intended for this use, the training dataset was filled mostly with NSFW images. If you do not want NSFW, make sure you actively prompt away from it. If you are trying to prompt for NSFW, note that many images may come out censored by default (uncensored images were tagged as such). An inpainting pass with "mosaic censoring" or "bar censor" in the negative may help with this. For this first pass, images were not tagged with rating_safe, rating_explicit, etc. Maybe in the next version.

TAGGING NOTES:

This LORA was trained specifically on Autism Mix and images were ranked in scores 6-9 as well as given the "source_anime" tag. Images with badly drawn hands and feet were tagged with "bad hands" and "bad feet" so this may be helpful in the negative. You may also need to put "white outline" in the negative if you are getting undesired outlines around characters.

Score 9 is historic works by the artist during their most prolific period (2009-ish).

Scores 7 and 8 are mostly newer works by the artist from their reactivation period (tagged with "2020s") as well as some lower quality older works.

Score 6 is entirely monochrome pencil sketches that have been tagged with "sketch"

Download model

Weights for this model are available in Safetensors format.

Download them in the Files & versions tab.