The BFS (Best Face Swap) LoRA series was developed for Qwen Image Edit 2509, specialized in high-fidelity face and head replacement tasks with natural tone blending and consistent lighting.
Each version builds upon the previous one:
🧠 V1 – Focus Faces: precise face swaps, keeping the original head shape and hair while transferring facial identity and expression.
🧩 V2 – Focus Head: stronger head swaps, replacing the full head (including hair and pose orientation).
The 2 versions complement each other, one is focused on face swapping and the other is focused on head swapping.
IMPORTANT NOTE: In version 2, I trained up to 12,000 steps, and this version you'll see now has 10,000 steps. I had some issues with the head proportions, so this version 2 may be temporary at this time, as I'm trying to resolve the head issue. If anyone can resolve this through their workflow (prompts, settings), please let me know. Additionally, one possible temporary solution to the issue of face toning compared to the body is to try a second pass asking for color correction between the body and face, or you can try passing a grayscale image of the face so the model can't assume the original color but you need to specify the eye color in the prompt. I'll attach an image below of what this would look like. Anyway, I'm trying to work on resolving these issues as quickly as possible.
Second pass:
prompt: Match the head’s skin tone to the body and correct the head size for natural proportions.
Grayscale Image:
⚙️ BFS V1 — “Focus Faces”
Trained on 240 image triplets (face, body, and result),
with a LoRA rank of 16 → later increased to 32,
and gradient accumulation = 2, running for 5500 steps on an NVIDIA L40S GPU.
This version produces stable and detailed face swaps, preserving expression, lighting, and gaze direction while maintaining the body’s natural look.
🔧 Model Notes
Quantization: not guaranteed to work below FP8 (avoid GGUF Q4).
Face mask: optional — remove if MediaPipe or Planar Overlay cause issues.
Pose conditioning: use MediaPipe Face Mesh if you need more alignment control.
Lightning LoRA: may produce plastic-like skin, especially when mixed with other Qwen-based LoRAs.
⚙️ Recommended Settings
Samplers:
er_sde + beta57 / kl_optimal / ddim_uniform(best results)ddim + ddim_uniform (sometimes most realistic)res_2s + beta57
Don't get attached to one setting, sometimes if it doesn't work well with one, switch to another.
Precision:
🧠 Best:
fp16⚙️ Recommended:
gguf q8orfp8⚠️ Below fp8: noticeable degradation
Inference Tips:
With LightX2V LoRA → use 4 steps for fast generation.
Without it → use 20–25 steps, CFG 1.5–2.5 for realism.
If eyes glitch, try CFG 3.5 or toggle Lightning LoRA.
🧬 BFS V2 — “Focus Head”
The V2 “Focus Head” version was trained as a continuation of V1, extending the dataset and shifting focus toward full head swaps.
It was trained on a NVIDIA RTX 6000 PRO, rank 32, for 12,000 steps, using 628 image pairs (face, body, target, and sometimes pose maps generated via MediaPipe).
🔹 Training Phases
Standard Face Swap – same as v1, focusing on facial identity.
Pose-Conditioned Face Swap – added pose maps to align gaze and head angle.
Full Head Swap – replaced the entire head (including hair) for stronger identity control.
After ~2000 steps, the focus moved toward head swap refinement.
At ~4000 steps, the dataset was narrowed to perfect skin-tone matches, and by the end of training,
the dataset evolved from 628 → 138 → 76 high-quality samples for final fine-tuning.
⚠️ Note:
While V2 can still perform standard face swaps, it’s more naturally inclined toward full head swaps due to its data balance.
This was intentional in part, but also a side-effect of dataset distribution and mixed conditioning.
Future versions aim to restore the fine facial control of V1 while keeping the strong tone and pose consistency of V2.
⚠️ Important Notice
Do not use or share results involving real people, celebrities, or public figures.
Civitai’s moderation may disable posts that violate likeness or consent rules.
This model is intended only for artistic and fictional characters, educational use, and AI experimentation.
I take no responsibility for any misuse of this model. Please use it responsibly and respect all likeness rights.