Sign In

wanxiangXL

61
416
5
Verified:
SafeTensor
Type
Checkpoint Trained
Stats
416
Reviews
Published
Sep 16, 2023
Base Model
SDXL 1.0
Hash
AutoV2
C64DA72D7A
default creator card background decoration
Holiday 2023: Red Victory
SHiJ's Avatar
SHiJ

Model play recommendation

There are more pictures in the Gallery for reference

  1. Recommended Sampling Method:(DPM 2M Karras)(DPM SDE Karras)(DPM2)(Euler a)

  2. Iteration steps recommended: 30-60 according to the actual situation of the individual adjustment

  3. HD repair redrawing amplitude recommendation: (default 7) can also be adjusted according to individual mold

  4. Size recommendation: 1024 starting effect (good)

  5. Recommendation of prompt words: 7-16 coefficient

It is recommended to use it together with the LORA model for better results. (万享XL_LORA - v1.0 | Stable Diffusion LoRA | Civitai)

Lora helps to generate different styles of pictures, adding more details, gloss, color, texture, light and shadow, better rendering effect.

Please return more pictures when it is convenient for family members. Some pictures are convenient for everyone to have more inspiration and reference in the drawing.