---
tags:
- text-to-image
- image-to-image
- stable-diffusion
- lora
- diffusers
- template:sd-lora
widget:


- text: graphis02 style illustration of a cyberpunk asian girl with a cityscape background at sunset
  output:
    url: images/girlcyberpunk1.jpg


- text: graphis02 style illustration of a cat in a field of purple flowers
  output:
    url: images/cat1.jpg


- text: graphis02 style illustration of a woman in smoke
  output:
    url: images/womeninsmoke2.jpg


- text: graphis02 style illustration of a cityscape at sunset
  output:
    url: images/citysunset1.jpg


- text: graphis02 style illustration of the Golden Gate Bridge
  output:
    url: images/ggb.jpg


- text: graphis02 style illustration of a cup of latte
  output:
    url: images/latte.jpg


- text: graphis02 style illustration of a woman in smoke
  output:
    url: images/womeninsmoke1.jpg


- text: graphis02 style illustration of a cat in a field of purple flowers
  output:
    url: images/cat2.jpg


- text: graphis02 style illustration of a cityscape at sunset
  output:
    url: images/citysunset2.jpg


- text: graphis02 style illustration of a cyberpunk asian girl with a cityscape background at sunset
  output:
    url: images/girlcyberpunk2.jpg


base_model: black-forest-labs/FLUX.1-dev
instance_prompt: graphis02, graphis02 illustration style, in the style of graphis02
license: other
license_name: flux-dev-license
license_link: https://huggingface.co/black-forest-labs/FLUX.1-dev/blob/main/LICENSE.md

---


# Graphis-02

<Gallery />

## Model description 

Graphis-02 is a continuation of my [Graphis-01](/dicksondickson/graphis-01/tree/main) but trained on FLUX.1 Dev and with an expanded dataset. This style LoRA model generates stylized illustrations with a flat, textured look featuring smooth, grainy gradients.

You should use `graphis02`, `graphis02 illustration style`, `in the style of graphis02` to trigger the image generation. 

Recommended weightings range from 0.8 to 1.2 (when using ComfyUI). Image-to-image generations yeilds very interesting results. Have fun!

## Trigger words

You should use `graphis02`, `graphis02 illustration style`, `in the style of graphis02` to trigger the image generation.


## Download model

Weights for this model are available in Safetensors format.

[Download](/dicksondickson/graphis-02/tree/main) them in the Files & versions tab.


## Credits

Dickson Chow
[dicksonchow.com](https://dicksonchow.com)