









Hello,
I made this version from the real images of chernobyl disaster. Some pictures look blurry due to the low quality of data set, but I love the details of this version.
Trigger word for v3 is "chern4byl enviroment"
Recommended settings:
Lora Weight: 0.6-0.8
Sampler: DPM++ SDE Karras
Resolution (W,H): 1280x1024
Steps: 40-50
CFG scale: 4-6
You can use this lora with SDXL to make post apocalyptic style background scene.
Trigger word "SZ_4poXL enviroment"
I recommend using the wildcard here: https://civitai.com/models/70264
In case you don't want to use wildcard, play around with these prompts:
inside a post-apocalyptic library, dust-shrouded books, sagging shelves, strewn broken furniture, ceiling cobwebs.
on a post-apocalyptic kitchen, scattered rusted pots, cracked dining table, moldy open refrigerator.
inside a post-apocalyptic collapsed bedroom, dust-laden bed, peeling wallpaper, shattered mirror, moth-ravaged wardrobe.
on a post-apocalyptic living room, torn sofas, shattered tv, faded family photographs, worn carpet.
inside a post-apocalyptic nursery, broken crib, strewn dust-covered stuffed animals, peeling wallpaper.
in a post-apocalyptic abandoned workshop, scattered rusted tools, lingering oil smell, yellowed blueprints.
in a post-apocalyptic empty classroom, overturned desks, dusty chalkboard, yellowed textbooks.
inside a post-apocalyptic silent hospital ward, stained bed sheets, dusty medical equipment, lingering antiseptic odor.
on a post-apocalyptic deserted train station, rusting benches, cracked information board, graffiti-covered abandoned train.
in a post-apocalyptic crumbling cinema, torn posters, dusty projector, broken seats, lopsided screen.
on a post-apocalyptic supermarket, empty shelves, scattered trolleys, broken glass, decaying food packaging.
in a post-apocalyptic factory, rusting machinery, broken windows, discarded materials, smashed control room.
in a post-apocalyptic neglected greenhouse, shattered glass, overgrown plants, rusted tools, dried fountain.
***V2.0***
I improved the data set and training parameters, and optimized the lora size. Have fun!
Recommended settings:
Lora Weight: 0.6-0.8
Sampler: DPM++ SDE Karras
Resolution (W,H): 1280x1024
Steps: 40-50
CFG scale: 6-7
***V1.0***
Recommended settings:
Lora Weight: 0.6-0.8
Sampler: DPM++ SDE Karras
Resolution (W,H): 1280x1024
Steps: 30-40
Denoising strenght : 0.3-0.5
CFG scale: 5-6
This version is huge since it’s my 1st experiment, I will optimize it better soon.
Welcome any feedback so I can make better version in the future!
1. Quyền đối với các mô hình được đăng lại thuộc về người sáng tạo ban đầu.
2. Người sáng tạo gốc muốn xác nhận mô hình vui lòng liên hệ nhân viên SeaArt AI qua kênh chính thức. Nhấp để xác nhận
