⭐🎠 Hello, this is Ether PDXL 🎠⭐

Ether PDXL is a pony based model focused on flexibility and producing more interesting, anime style illustrations.

Please note that this model has biases in generating females.

✨ Please Share Your Cool Creations Below! ✨

* Ether PDXL - Version A2B

Figured out why the model was half the size after saving thanks to foxmlg!

My AUTO1111 was using SDXL LORAs at fp8 which resulted in the model also being saved at that precision. I disabled that setting, then re-tuned the weights to have a similar but (imo) nicer overall aesthetic.

Seems to be more consistent, stable and less 'fried' than A2 (?). Not entirely sure though as I wanted to push this release as a 'quick fix'. Please let me know about your experience with version A2B!

Thank you all so much for trying out my models.

Still in its early stages. I'm trying to understand SDXL merging as well as navigate resources that align with my goals.

VAE is baked in. It's XL VAE C9.1 as it's more vibrant than standard SDXL which better fits the aesthetic I'm looking for.

Looking forward to your feedback!

Please refer to Pony V6 and PD for Anime's model pages for more detailed prompting guidelines.

* Huggingface Archive

☄️ Generation Recommendations

* All preview images were generated with no LORA's, embeddings, extensions or img to img/inpainting. Just pure text to image with hi-res fix upscaling.

Most sample images were generated with:

Sampler: Eular A
Schedule Type: SGM Uniform
Steps: 28 - 38
CFG: 4 - 9
Clip Skip: 2
ENSD: 31337

Negative Prompt Example:

"score_4, score_5, score_6, source_pony, source_furry, fewer digits, extra digits, polydactyly"

I recommend this negative prompt if you wish to achieve a similar style to the preview images. You can further alter positive and negative prompts based on the type of style you'd like to achieve.

Sample images were further upscaled using Hi-res fix. Please utilize it if you wish to achieve similar results.

Latent (nearest-exact)
Upscale by: 1.75
Denoising strength: 0.55

Are your results changing too much after Hi-res fix? Please try 4x-foolhardy-Remarci.

(Also recommended for simple style illustrations, anime, cartoon, etc.)

4x-foolhardy-Remarci
Upscale by: 1.75
Denoising strength: 0.45 - 0.6

📘 Model Merging Info - vA2B

I have no training or real knowledge of machine learning. I merged the model using the information provided on WarriorMama777's AOM model page as well as the ones in these Rentrys. I very arbitrarily mixed the UBM weights of each model, modifying the values with trial and error until I was visually satisfied with it.

I utilized the SuperMerger extension for AUTO1111 by hako-mikan.

Please refer to the 'About This Version' for more detailed information.

💙 Credits

Thank you to WarriorMama777 for providing AOM and other various mixes as well as detailing your workflow which inspired me to try mixing my own models.

Thank you to hako-mikan for creating the SuperMerger extension allowing a quicker workflow for merging.

Thank you to those that have contributed to these MBW Rentrys and this very detailed MBW Figma.

Thank you to PurpleSmartAI for Pony Diffusion V6.

Thank you to gsdf, rqdwdw for PD for Anime.

Thank you to CagliostroLab for Animagine XL V3.1.

Thank you to LEOSAM for HelloWorld XL.

Thank you to advokat for Concept B.

Thank you to ZyloO for RELSM - Semi Realism.

Thank you to ComprehensivePrize65844 for nai style for pony v3.1.

Thank you to Fish788 for Worldly XL and MeMaXL.

Thank you to KAtogh for AIイラストおじさん.

Thank you to AUTO1111 for creating the web UI.

Thank you to StabilityAI for starting everything with Stable Diffusion.

Thank you to CivitAI and Huggingface for providing places where open source AI can flourish.

Thank you to the entire SD community for continuing to openly share and create