|
--- |
|
license: openrail |
|
tags: |
|
- stable-diffusion |
|
- lora |
|
--- |
|
|
|
## Duelyst Codex Landscape LORA |
|
|
|
(Model Card will be updated later with proper example images and prompt settings, WIP) |
|
|
|
<img src="https://huggingface.co/cadaeic/duelyst-codex-lora/resolve/main/im_20230222110424_000_988913356.png"/> |
|
|
|
**Prompt:** painting of a fantasy landscape of a castle |
|
|
|
--- |
|
|
|
LORA trained on open source fantasy landscape art from the shut down game [Duelyst](https://en.wikipedia.org/wiki/Duelyst), whose assets have been uploaded to GitHub under the CC0 license. |
|
|
|
Dataset obtained from [The Public Conscious](http://library.lowframerate.games/), a collection of public domain images. |
|
|
|
### Notes |
|
|
|
- Trained over Stable Diffusion 1.5 and will work with SD 1.x models |
|
- Overtrained on fantasy landscapes |
|
- My very first Lora, and pretty much a test Lora! |
|
- Trained with [this Colab notebook](https://colab.research.google.com/github/Linaqruf/kohya-trainer/blob/main/kohya-LoRA-finetuner.ipynb) using default settings with a free Google colab account for 40 minutes |
|
|
|
--- |
|
|
|
<img src="https://huggingface.co/cadaeic/duelyst-codex-lora/resolve/main/im_20230222110707_001_141397401.png"/> |
|
|
|
**Prompt:** fantasy painting of a fat grey british shorthair cat |