---
tags:
- text-to-image
- stable-diffusion
- lora
- diffusers
- template:sd-lora
widget:
- text: a peaceful fox curled up sleeping peacefully, in ATKSN style
  output:
    url: images/A_fox_curled_up_sleeping_peacefully_in_ATKSN_style.png
- text: a lighthouse standing tall against the ocean waves in ATKSN style
  output:
    url: >-
      images/A_lighthouse_standing_tall_against_the_ocean_waves_in_ATKSN_style.png
- text: a single rose blooming minimalistic in ATKSN style
  output:
    url: images/A_single_rose_blooming_minimalistic._in_ATKSN_style.png
- text: >-
    a young woman with long hair looking over her shoulder portrait in ATKSN
    style
  output:
    url: >-
      images/A_young_woman_with_long_hair_looking_over_her_shoulder_portrait_in_ATKSN_style.png
- text: >-
    cute illustration, a pair of birds building a nest on a branch surrounded by
    cherry blossoms in ATKSN style
  output:
    url: >-
      images/cute_illustration._A_pair_of_birds_building_a_nest_on_a_branch_surrounded_by_blooming_cherry_blossoms_in_ATKSN_style.png
- text: illustration, a large black cat sitting on a windowsill in ATKSN style
  output:
    url: >-
      images/illustration_A_large_black_cat_sitting_on_a_windowsill_in_ATKSN_style.png
base_model: black-forest-labs/FLUX.1-dev
instance_prompt: in ATKSN style
license: cc0-1.0
---
# Dithering

<Gallery />

## Model description 

**Creator:** [Yerko Lucic](https:&#x2F;&#x2F;glif.app&#x2F;@Yerko)

**Style:** Dithering is a Flux LoRa that applies a black-and-white 1-bit Atkinson dithering effect, trained on 45 synthetic AI-generated images. The dataset includes realistic images to ensure proper shading, but works well for illustrations too.

### Key Features

*Dithering Effect:*  Applies 1-bit Atkinson dithering, great for detailed black-and-white shading.
*Color Versatility:* Prompts for drawings, illustrations, or paintings will often produce color outputs due to the realistic dataset, but the dithering remains effective.
*Trigger Phrase:* Use &quot;in ATKSN style&quot; at the end of prompts to activate the effect.
*Flexible Use:* Works at a strength of 1.0, and outputs can be used as masks or as a post-processing effect.

### Use Cases

*As-Is:* Strong black-and-white dithering at full strength.
*Color Outputs:* When prompting for drawings or paintings, expect color results with dithering still visible.
*Masking &amp; Post-Processing:* Use the outputs for further effects or compositing.

**Click here to try it on Glif:**

[![Try it on Glif!](https:&#x2F;&#x2F;cdn-uploads.huggingface.co&#x2F;production&#x2F;uploads&#x2F;635dd6cd4fabde0df74aeae6&#x2F;2aHGiskVFk6BniLGuhRzf.png)](https:&#x2F;&#x2F;glif.app&#x2F;@Yerko&#x2F;glifs&#x2F;cm15t5km50007uu4rtf2orynu)

*This LoRA is a part of the Glif Loradex Trainer project. The goal of the project is to explore different training techniques together, create and release opensource LoRAs, and fill the world with more art. This project is sponsored by [Glif](https:&amp;#x2F;&amp;#x2F;glif.app&amp;#x2F;), offering free early access to Flux Dev training on the Glif Discord.*

*All models in the Loradex Trainer program are released under the BFL Flux Dev license. Attribution for the original LoRA creator is always appreciated!*


## Trigger words

You should use `in ATKSN style` to trigger the image generation.


## Download model

Weights for this model are available in Safetensors format.

[Download](/glif/dithering/tree/main) them in the Files & versions tab.