niji-3d icon colorful wind SDXL

0
0
0
0
Cartoon styleLogo designStyle BoostLogo & Icon
Cập nhật gần đây: Đăng tải lần đầu:
Cartoon style,Logo design,Style Boost,LoRA,SDXLImage info
Cartoon style,Logo design,Style Boost,LoRA,SDXLImage info
Cartoon style,Logo design,Style Boost,LoRA,SDXLImage info
Cartoon style,Logo design,Style Boost,LoRA,SDXLImage info
Cartoon style,Logo design,Style Boost,LoRA,SDXLImage info
Cartoon style,Logo design,Style Boost,LoRA,SDXLImage info
Cartoon style,Logo design,Style Boost,LoRA,SDXLImage info
Cartoon style,Logo design,Style Boost,LoRA,SDXLImage info
Cartoon style,Logo design,Style Boost,LoRA,SDXLImage info
Cartoon style,Logo design,Style Boost,LoRA,SDXLImage info

This is a colorful lora created in a 3D mode similar to Midjourney's Niji, with a texture resembling highly glossy and iridescent semi-transparent plastic. Of course, it can also be glass or frosted glass through keying and weight of words.

The training image dataset includes some daily objects, excelling in generating structurally simple objects such as daily necessities, footwear, hats, clothing, transportation, landmarks, etc. To achieve the effect of a cover image, start by writing a (object name, like camera) icon, 3d icon, ui ux; for a clean background write: white background, studio lighting; leave the negative terms blank, if you want to avoid logos, signatures, watermarks, write text, signature, watermark, username.

Since SDXL on LibLib does not yet support the control net, please download and use it locally. If you generate good 3D icons, please come back and share them with everyone to showcase your results!

**The next version of this lora plans to incorporate character training. If you happen to need it, feel free to like and bookmark it.**


Items: earphone, chair, children's shoes, basketball shoes, alarm clock, castle, hourglass, guitar, vehicle, airship, etc.

Design aesthetic: Bauhaus, industrial design

Texture words: fine lustre, many details, octane render, colored plastic, gradient color, frosted glass


Here are the recommended parameters:

Base film: Use the official basic algorithm XL (SDXL 1.0 official version). If you need the SD 1.5 version of lora, leave a message below.

Lora weight: 0.6-1

Sampling algorithm: DPM++ 2S a Karras or Euler a

Iteration steps: 30-50

Output image size 1024x1024

CFG scale: 7-9

Note: Be sure to choose XL models for the base film; otherwise, they won't be found in the lora options.

Thảo luận

Phổ biến nhất
|
Mới nhất
Gửi
Đến sớm
Tải xuống
(0.00KB)
Chi tiết
Loại
Số lần tạo hình ảnh trực tuyến
0
Tải xuống
0
Tham Số Đề Xuất
Kết Hợp Đề XuấtCheckpoint
Không
Cân nặng được khuyến nghị.
0.8
CFG
7
VAE
Không

Bộ sưu tập hình ảnh

Phổ biến nhất
|
Mới nhất