Creator-owned reference stills used to fine-tune the prvpns anatomical LoRA on Krea2.
Below: how the set is organized, real example frames per class, and the captions used at train time.
End-to-end path from curated stills to the shipped adapter — illustrated with real training images (not AI mockups).
01 Curate → 02 Caption → 03 Fine-tune → 04 Export
Image on the left, explanation on the right — one row per pipeline stage.
Mid-train samples for the dedicated contact prompt (sample index 3). Left = sample image, right = step metrics + what improves.
No separate mid-train sample prompt was set for fluid — the class is learned from dataset stills mixed into the 2200-step loop. Cards show real train stills + what they teach.
Compact catalog for audit — full-res + caption on click.
Run metrics from the completed Krea2 LoRA job (2200 steps).
| Product | prvpns_penis_krea2_v9.safetensors |
|---|---|
| Base | Krea2 · Qwen-Image / flowmatch |
| Network | LoRA rank 16 / alpha 16 |
| Dataset | 106 captioned stills · 768×1024 |
| Batch | 1 · gradient checkpointing |
| Trigger | prvpns |
| Caption dropout | 0.05 |
| Checkpoints | every 250 steps (kept: 250…2000 + final) |
Each row is a training class: count, share bar, and real stills from that class. Click a thumb to open full-res.
Window averages — start (1–100): 0.08765 ·
mid (1000–1100): 0.05919 ·
end (2100–2199): 0.0766
Mid-train samples (same prompt seed family) next to step / loss / progress bars — layout from the older exhibit style.
One real training still per class with share bars and extra samples beside the metrics.
How to read this page as model-card documentation.
LoRA adapter for stable anatomical identity under viewpoint and contact. Multi-state stills (erect / soft / contact / fluid) prevent pose lock-in.
Main token: prvpns.
Private technical documentation for the published model card. Do not redistribute stills. Creator-owned reference material only.
| Product | prvpns_penis_krea2_v9_training.safetensors |
|---|---|
| Base | Krea2 · flowmatch · bf16 |
| Network | LoRA rank 16 / alpha 16 |
| Steps | 2200 · ai-toolkit · adafactor · lr 1e-4 |
| Resolution | 768 × 1024 |
| Trigger | prvpns (identité) |
| Source | Creator-owned stills · 106 curated JPG |
| Typical weight | couples 0.45–0.55 · solo 0.55–0.65 |
| Mix | solo identity + contact stills + flaccid · 0% IA showcase |
Adult technical material (18+). Private consensual source photography. Reproducible fine-tune documentation — real corpus only.