Danbooru Target SDXL: Finetuned with a focus on Danbooru's content.
Best with Danbooru Tags: Works optimally when prompted with Danbooru-style tags.
Lora Compatibility: Compatible with most Loras available on the hub.

Usage Instructions:

It is recommended to separate tags with commas; tags can be used in any order.
Use "year 20XX" to generate art in the style of that specific year (e.g., year 2023).
High step size could be beneficial.

Disclaimer: This model may produce unexpected results. Use at your own risk.

Community: Join our Discord community and share your creations!

If you enjoy using Momo XL, feel free to drop by and say hi.

Check out the official release page on Hugging Face: Momo XL

Detailed information about the training will be released later on the repository.

Thank you! 😊

Momo XL - Training Details (Oct 15, 2024)

Dataset:
Momo XL was trained on a dataset of 400,000+ images sourced from Danbooru.

Base Model:
Momo XL is built on top of SDXL and includes elements from two additional finetuned models:

Formula: SDXL_base + (Animagine 3.0 base - SDXL_base) * 1.0 + (Pony V6 - SDXL_base) * 0.5

For more information about the source models:

For more detailed information about the training process, visit the official Hugging Face repository here.