Introduction:
This is a LoRA that simulates the filter effects of a Polaroid camera! There are multiple filters available for selection, and it can be combined with other portrait LoRAs for generating images. When used as an auxiliary LoRA, it is recommended that the LoRA weight should not exceed 0.5.
V5.1 Z-Image-Base is here!
Compatible with both Base and Turbo models, though the Turbo model is still recommended for image generation.
V5.0 Z-Image-Base is here!
1. This update continues to expand the range of appearances to meet diverse requirements.
2. Trained on the Base model, though the Turbo model is recommended for generating images.
3. Slightly reduced the brightness contrast of the heavy colour filter to enhance scene adaptability.
4. Recommended Parameters Sampler: Euler
Scheduler: SGM_Uniform
CFG: 1
Image Size: 1024x1536
Samples: 12
Model Weight: Between 0.6 and 0.8
Prompt: OBbllai
V4.0 Z-Image-Turbo The new version is here!
1. This update introduces additional characters beyond young, attractive Asian women, including men, children, and elderly individuals.
2. Enhanced scene and skin details have been incorporated.
3. A wider variety of facial expressions has been added.
4. Recommended Parameters:
Sampler: Euler
Scheduler: SGM_Uniform
CFG: 1
Image Size: 1024x1536
Number of Samples: 9,When stacking multiple loraS, please use 22
Model Weight: Between 0.6 and 0.8
Trigger Word: OBbllai
V3.0 is coming!
1. Further enhance the skin texture! Reduces greasiness, close to real skin
2. Reduces redness of the face
3.Improved generalization, combine with other lora to create unlimited possibilities, right?
Major update for V2.0!
There are richer portraits, and 6 filters are added for selection. Of course, filters are not a must. You can just use the trigger word "OBbllai" to generate images. When only the trigger word is used, the filters will be automatically applied according to the environment and lighting conditions of the character