I would prefer if you guys don't use this model for Commercial Use.You can check out my pixiv where I use my lora to create images.<a target="_blank" rel="ugc" href="https://www.pixiv.net/en/users/34413372">https://www.pixiv.net/en/users/34413372</a>support me by following me in pixiv <a target="_blank" rel="ugc" href="https://www.pixiv.net/en/users/34413372">https://www.pixiv.net/en/users/34413372</a>I made this lora for consistent 1 room, I trained this lora with 3d images that I rendered and used Photoshop to add some character in image. It is not perfect because it is first time I made a lora for non human stuff.v2The lora is not able to create humans and you need to use other lora model. I made the caption better and added 4 extra new original images in dataset. it is way harder than I thought when it comes to making a lora for a consistent building that can also be flexible with other lora. I would suggest the strength of this lora should be from 0.3-0.8. The lower you go, the more it won't be able to create the room.This lora requires good prompting to get decent result.I would say the lora have failed, but I am still uploading to see if other people able to use it.Go with 648x976, 768x1152, 768x1024, 768x768, 976x648, 1024x768these are the keywords that I trained the lora with mix with auto tagger: wooden panel, pillows, bed, light, blanket, no humans, gym bag, carpet, flower, lamp, cup, drawer, light, sun light, vase, rose, clock, book, closet, tree branch stick, 2 drawers, glass vase, curtains, ceiling light, ceiling, stick, from above, from side,v1the lora only give good result if you upscale the image so going for example 768x1152 will not give good result. the low resolution does give very bad images. try to use low weight (around 0.5-0.7) on lora. The bedroom lora is somewhat compatible with other lora.The lora make the character way to big and you can add giant, tall woman/man etc. in negative,to trigger the lora: bedroom_1, bedroom, picture frame,there is only one perspective I trained the lora on that is from above, but it should still be able to do other perspective, the only one the lora is bad at is from side perspective.in the room there where many item and I added keyword for them that are: picture frame, ceiling light, pilow (I wrote it wrong but I was too late lol), bag, lamp, flower and rose (yes they are separate words I used to train the ai to trigger specific plant), cup, vase, book, sunlight, table, bed, carpet, curtains, branch stick, blanket,

Files with SHA256: b9452f1a255cdf257f60af1c19295c1b2ae8eb897155983a010d92b47288fafa