---
license: creativeml-openrail-m
---

Prompt with **"rizaDB anime girl"**

**Training details (as far as I remember):**

- Trained with [JoePenna Dreambooth repository](https://github.com/JoePenna/Dreambooth-Stable-Diffusion)
- data set: 54 concept images + a number of custom reg images
- default learning rate for 4500 steps

**Example generations:**

![00037-2764761054-rizaDB](https://huggingface.co/alea31415/ryza-atelier-ryza/resolve/main/00037-2764761054-rizaDB.png)
![00061-3019598551-rizaDB](https://huggingface.co/alea31415/ryza-atelier-ryza/resolve/main/00061-3019598551-rizaDB.png)
![00226-1649286225-HQ](https://huggingface.co/alea31415/ryza-atelier-ryza/resolve/main/00226-1649286225-HQ.png)
![00231-1843099238-HQ](https://huggingface.co/alea31415/ryza-atelier-ryza/resolve/main/00231-1843099238-HQ.png)
![00244-3595448462-HQ](https://huggingface.co/alea31415/ryza-atelier-ryza/resolve/main/00244-3595448462-HQ.png)