P prvpns v6
Training corpus 18+
Live training exhibit

Krea2 LoRA
training corpus

Source videos and reference frames used to fine-tune the prvpns / prvpene adapter — identity + contact geometry. Face-excluded · creator-owned · model-card documentation.

93 frames 2 source videos Base Krea2 Rank 16 · 3200 steps
Start here

Source training videos

Silent web encodes of the clips used to harvest frames. Play either — full provenance of the corpus.

Corpus

Frame catalog

Click any frame for full resolution and the training caption used at fine-tune time.

Filter
Docs

Methodology

How to read this page as training documentation.

Purpose of this corpus

Adapter training for anatomical consistency under viewpoint and contact deformation. Multi-angle identity stills + junction / occlusion examples so the model is not locked to a single pose.

Captions use trainer tokens (prvpns, optional prvpene for contact scenes).

How to read this page

  • Videos first — source clips for frame harvest.
  • Thumbs — full frame for audit (no decorative crop).
  • Click — full resolution + train-time caption.
  • Filters — split by role (video / still / solo).

Category roles

  • Video frames — stills extracted from source clips.
  • Still / contact — junction geometry.
  • Identity stills — multi-view lock.

Access & use

Restricted technical documentation for the published LoRA card. Do not redistribute frames. AI showcase images on model pages are generated samples; this catalog is training material only.

Train

Specifications

Productprvpns_penis_krea2_v6_final.safetensors
BaseKrea2 · Qwen-Image / flowmatch
NetworkLoRA rank 16 / alpha 16
Steps3200 · ai-toolkit · bf16 · adafactor · lr 1e-4
Resolution768 × 1024
Triggersprvpns (identity) · prvpene (contact)
WeightFree Smell turbo couples 0.42–0.44 · Gonzalomo ~0.55 · solo close 0.85–0.90
Mix~96% contact-weighted slots · stills + curated video frames

Adult technical material (18+). Private consensual source photography / video stills. Intended for model card transparency and reproducible fine-tuning documentation.