Pro version of Babes_By_Stable_Yogi Pony is available on my Patreon
Onsite generations are permanently available on this version:
👉 Babes_By_Stable_Yogi: https://civarchive.com/models/174687?modelVersionId=2110984
Babes By Stable Yogi V6.5
V6.5 is a substantial step up from V6.0.
Anatomy, Expressions and gestures, Environments, Pose stability
A lot of these fixes came directly from feedback identified by our Patreon community — folks running the in-development checkpoints, noting what broke, and helping us prioritize what to fix for the next version.
## 🚀 Be part of the next version
We run a free image generator at discord.gg/DZEenb5wGc that ships the next-in-development checkpoint inline. Explorers generate with it, report shortfalls, and we record that feedback into the build queue for the following version.
If you've ever wanted to be part of how a checkpoint is actually made — hop in. Free to use, no install, no setup. Generate, find issues, tell us, and you'll see your fixes land in the next release.
## File lineup
Six format variants ship for general use:
(FP16) | Default for most users — best quality / speed balance |
(BF16) | RTX 3000+ with BF16 support, slight speed edge |
(FP8 Scaled) | ~8 GB VRAM cards — Forge & ComfyUI native, no extra nodes |
(NF4) | Lowest VRAM (4-6 GB) — Forge native, ComfyUI with bnb_nf4 node |
(Q6_K) | Highest quality GGUF — needs ComfyUI-GGUF or [my Forge GGUF extension for forge ui] (https://github.com/brandulateai/sd-forge-sdxl-gguf-brandulateai) |
DMD2 — 4 steps, CFG 1.2, LCM sampler. Fastest option |
## Recommended settings
Standard variants:
- Sampler: Euler a or DPM++ 2M
- Scheduler: Karras
- Steps: 25–30
- CFG: 5–7
DMD2 variant only:
- Sampler: LCM
- Steps: 4 (yes, four)
- CFG: 1.2
Resolution: native SDXL — 1024×1024, 832×1216, 1216×832 work cleanly.
## Where else to find me
Pro variants and feedback: [Patreon](https://patreon.com/brandulate)
(https://github.com/brandulateai)
Free generator with the next checkpoint waiting for your feedback at Discord Community: https://discord.gg/DZEenb5wGc
Join me on my Patreon. for exclusive perks and early access to unique resources.
To discuss custom LoRa's or models, feel free to connect on Discord.
👍 Like this model to keep me motivated and inspired to create more!
💬 Drop a comment and let me know what you'd love to see next.
🌟 Review this model to help me improve and make even better creations.
🔔 Hit that notification bell to stay updated with my latest models and updates!
Important Usage Tips
Add Stable_Yogis_PDXL_Positives at the beginning of your prompt section.
Add Stable_Yogis_PDXL_Negatives-neg at the beginning of your negative prompt section.
Description
How to Use the Inpainting Model
Recommended Settings:
Inpainting Mode: Soft inpainting
Sampler: Euler a
Scheduler Type: Uniform
Mask Blur: Set between 4 and 8 for optimal edge blending
CFG Scale: 3 to 6 for control over the prompt’s influence
Denoising Strength: 0.9 for detailed inpainting results
Soft Inpainting Adjustments:
Set "Transition Contrast Boost" to 1 (default is 4)
Adjust "Difference Threshold" to between 0.1 and 0.25 (default is 0.5)
Tips for Best Results:
Masking Quality: Use the “Only Masked” option for the best quality, ensuring that the model focuses on the masked areas.
Mask Content: Start with “Masked content: original.” If that doesn’t provide the desired results, try “latent noise” or “fill” to achieve smoother blending.
Image Size: For best results with SDXL, aim for an image size with a combined width and height between 2000 and 2500 pixels. Alternatively, scale your image around this size range using aspect ratio helpers.
Creative Masks: Use masking creatively to change poses or settings. For example, if you want to adjust the character’s posture or add specific items, mask the areas where these adjustments should occur.
Refining Results: For complex edits, use a high-quality refining model (like "realistic refinement") set between 0.8 and 0.9 to sharpen and finalize the output.
Extended Masking: When inpainting a small section, consider masking additional areas at the top or bottom so the model processes the full image context for consistency.
Minimal Prompts: Keep your prompts concise. The model responds well with minimal guidance, allowing it to draw from the mask and original image for detail.
Example Scenarios
Basic Clothing Change:
Prompt: “woman, wearing a formal gown”
Mask the existing clothing area to replace with the new outfit.
Additional prompt (if needed): "woman, standing in profile with formal gown"
Pose Adjustment:
Initial prompt: “woman, lying on her back”
To change the pose, mask specific parts (such as arms or legs) to create a new pose.
Example: “woman, sitting with crossed legs”
Context Change for Creative Effect:
Initial prompt: “woman, wearing autumn coat, standing in forest”
Mask the coat to replace it with a winter jacket, or mask the background to add fog or snowfall.
Example for new prompt: “woman, winter jacket, snowy forest”
FAQ
Comments (17)
pony v3 lightning 4 steps please
Sure, its on the way.
What are the differences bewteen DMD2 and "not DMD2" ?
In simple words its another way of Faster image generation using SDXL/PONY models, just like Hyper, LCM, Lightning. ( Requires few steps. ) More Details here : https://huggingface.co/tianweiy/DMD2
@Stable_Yogi which one is the best balance of detail and speed in your opinion? Is there a rough scoring of which on is the fastest to slowest and then does the level of detail follow that same metric? I have low VRAM so it would be great to speed things up where I can but still enjoy the crazy detail in this model!
@gatorsuf DMD2 is great with skin texture, 6 steps is good balance to cook a good image. 4 steps could work but i have seen that if you are looking for (1 shot) good images go with 6 steps. All 4 step models are faster, its your choice which suits your requirements.
Great stuff Yogi. I’ve been enjoying using your models however the Teacher/Student prompts is about to get be banned if I don’t pay attention and remove them.
Absolutely, I appreciate the heads-up! I'll make sure to review and adjust those prompts to keep everything appropriate and fun. Thanks for sticking with my models and for the valuable feedback!
Amazing model. I'll leave it at that.
Thanks a lot for the feedback. 😊
great model with good anatomy accuracy thanks
Thanks for the Feedback 👍
Decent anatomy and good resulut without extra lora's.
Amazing base models, and you will get all you need from the same creator
amazing base model, nice work
nice model, i like it, for real
Thanks for the feedback.