README.md

---
tags:
- text-to-image
- stable-diffusion
- lora
- diffusers
- template:sd-lora
widget:
- text: >-
    score_9, score_8_up, score_8_up, Anime Portrait, 1girl, dalle3 anime,
    petite, magical ambiance, best quality, pretty face, detailed art,
    <lora:dalle3anime_pony:0.8>, masterpiece, source_anime
  parameters:
    negative_prompt: >-
      score_6, score_5, score_4, kid, mature, mutated fingers, long neck, ugly,
      boring, unclear, messy, out of frame, oversaturated, bad anatomy, bad
      hands, worst quality, low quality, strokes of paint
  output:
    url: images/00315-3976938809.jpeg
- text: >-
    score_9, score_8_up, score_8_up, petite sexy 1girl, dalle3 anime, magical
    ambiance, best quality, pretty face, <lora:dalle3anime_pony:1>
  parameters:
    negative_prompt: >-
      boring, unclear, messy, out of frame, oversaturated, bad anatomy, bad
      hands, worst quality, low quality
  output:
    url: images/00296-1170526047.jpeg
base_model: LyliaEngine/Pony_Diffusion_V6_XL
instance_prompt: Anime Portrait, dalle3 anime
license: cdla-permissive-2.0
---
# dalle3anime_pony

<Gallery />

## Model description 

This model was trained on anime portraits generated by the Dall-E 3 model.

You can use it to stylize your images, but first and foremost, the model was intended as a means of creating anime portraits similar to those that can be obtained using Dall-E 3.

For portraits, add the keyword &quot;Anime Portrait&quot; in addition to &quot;dalle3 anime&quot;

Since the model was trained on Pony, it may be better to also use &quot;strokes of paint&quot; in the negative portion of the prompt to achieve the best results.

## Trigger words

You should use `Anime Portrait` to trigger the image generation.

You should use `dalle3 anime` to trigger the image generation.


## Download model

Weights for this model are available in Safetensors format.

[Download](/LyliaEngine/dalle3anime_pony/tree/main) them in the Files & versions tab.