Hey, I’m Markury. I created AndroFlux, a LoRA model for FLUX.1 dev by Black Forest Labs, because there’s a glaring lack of accurate male anatomy in AI models. Working with The Bulge community, I’m trying to fix that by making sure AI can depict male features more accurately, balancing out the overwhelming female-centric content you see everywhere.

Community and Support

Join The Bulge Discord at https://thebulge.xyz if you’re into gay AI art—it’s a great place to share and learn. If you want to support the project, whether with feedback or donations, it’s much appreciated. You can help fund AndroFlux training at https://buymeacoffee.com/y7sa4gtyv6.

Current Status

AndroFlux is still a work in progress. The model isn’t perfect yet and needs more training, so don’t expect flawless results. For now, this ComfyUI workflow works best. If you're using an older version, you can find the workflow here.

Purpose

AndroFlux is all about making sure AI models can represent male anatomy—especially penises—accurately. It’s a step towards more balanced AI-generated art.

Technical Details

AndroFlux was built and tested primarily on FLUX.1 dev, but it should also be compatible with the Schnell base models, including the quantized fp8 version. These are distilled models, making training tricky and affecting the LoRA’s performance, requiring a lot of optimization.

Training Process

• Resolution Journey: I started training at 1024px, but the model wasn’t converging, so I dropped to 512px. The quality wasn’t great, so I ended up at 768px. This resolution offers a good balance between detail and training time.

• Captioning: Captioning was crucial in training. Contributions came from the AndroAtlas dataset, part of the AndroFlow project. Ephemeron01 from The Bulge Discord assisted with captions using models like llava-hf/llava-v1.6-mistral-7b-hf and a custom-trained PaliGemma VLM. Specific prompts like "man, penis" and the activator "AndroFlux" were also used in later versions.

• Acknowledgments: Huge thanks to pseudoterminalx and Terminus Research Group for their support, including providing an A100 for some of the training runs.

Usage Tips

• For best results, use prompts that specifically mention male frontal nudity, such as "penis."

• Recommended settings:

  • Resolution: 1024x1024 or 832x1216

  • Flux Guidance: 3.5

  • LoRA Weight: Start at 1.0, reduce to 0.5 if artifacts appear

  • CFG: 2.0 (1.0 will be faster though)

Suggested Prompts

• A nude man in his {late teens | early twenties | early thirties}

• His penis is {erect | flaccid}

• His penis is {large | small | average}-sized

Version History

• v1-v4: Early tests with lots of issues.

• v5-1500: First version that started showing promise.

• v5-1800: Better results at 0.5 LoRA weight.

• v6-v11: Tweaks and changes, including a switch to 512px.

• v12: Shifted to 768px, incorporating higher resolution images and contributions from Zellian.

• v13-v14: Continued refinement, focusing on distillation quality.

• v15: Latest version, training at 768px with no CFG to preserve distillation.

Compatibility

LoRAs for Stable Diffusion 1.5 or SDXL won’t work with FLUX.1 dev due to major architectural differences. FLUX.1 dev’s transformer-based design and flow matching need specially tailored LoRAs.

Training Cost

Training Flux takes a lot of compute power, and I've already spent over $200 on these runs. If you'd like to support AndroFlux training, any tips—monetary or otherwise—are greatly appreciated! You can contribute via https://buymeacoffee.com/y7sa4gtyv6 or use my RunPod link to support me while doing your own training at no extra cost to you.

Conclusion

AndroFlux is a step towards more balanced AI models that accurately depict male anatomy. It’s an ongoing project, and your support—whether through feedback, donations, or just being part of the community—makes a big difference.