<blockquote>⚠️ Compliance: This LoRA was trained exclusively on synthetic (AI-generated) images. No real persons — living or deceased — appear in the training data. Any resemblance to actual individuals is coincidental. "J-Rogue" (formerly Project J-Kinetics) is a fictional character concept intended for conceptual rendering and animation.OverviewJ-Rogue is a high-fidelity identity LoRA built for strict character consistency across still-image pipelines. It locks a specific performer's face, body proportions, signature attire, and a rigid six-point tattoo map that survives pose changes where weaker identity models bleed or drift. The intended look is photoreal, club/venue-lit, and athletic-performer energy — calm confidence rather than exaggerated expression.Locked Identity Specs<ul><li>Face: Soft, slightly rounded cheeks; warm brown eyes; full lips; natural soft smile</li><li>Hair: Soft blonde bob, chin-to-shoulder length, natural wave. Not pixie, not long</li><li>Build: Voluptuous hourglass; full hips, soft athletic curves; fair skin</li><li>Jewelry: Small gold hoop earrings when visible</li></ul>Default Attire<ul><li>Tight ribbed white tank top</li><li>Black thong</li><li>Footwear: barefoot, heels, or platform heels</li></ul>Tattoo Map (Strict Placement)Tattoo Exact Location Visible When Anchor Center upper chest, sternum Front / ¾ front Camera Right shoulder / upper deltoid Front / ¾ / right-side Celtic Knot Left upper back, over scapula Rear / left ¾ rear Sun Circle Right upper back, over scapula Rear / right ¾ rear Alien Head Left forearm, mid-forearm Left arm forward / raised Mandalas Right forearm, mid-forearm Right arm forward / raisedNever relocate, resize dramatically, or duplicate these tattoos. Only render tattoos that would realistically face the camera from the current angle.Expression Baseline<ul><li>Default: quiet confident smile, private amusement, calm ownership of the room</li><li>Secondary: focused athletic intensity (stretches, movement)</li><li>Avoid: blank, severe, overly soft, or aggressive expressions unless requested</li></ul>Recommended Prompt StructureJ-Rogue, 1woman, soft blonde bob, brown eyes, full lips, soft smile, voluptuous hourglass, fair skin, anchor tattoo on chest, camera tattoo on shoulder, [attire tags], [pose tags], [lighting tags]Recommended Settings<ul><li>LoRA Weight: 0.75–0.90 (above 0.95 may warp tattoo mapping)</li><li>Sampling Steps: 25–35</li><li>CFG Scale: 5.5–7.0</li><li>Base Model: Pony Diffusion XL / SDXL</li><li>Trigger Words: <code>jamie</code> <code>J-Rogue</code>, <code>projectJKinetics</code></li></ul><hr />Category: Character Tags: <code>J-Rogue</code>, <code>projectJKinetics</code>, <code>character</code>, <code>identity</code>, <code>lora</code>, <code>pony</code>, <code>sdxl</code>, <code>blonde</code>, <code>freckles</code>, <code>brown eyes</code>, <code>tattoo</code>, <code>anchor tattoo</code>, <code>camera tattoo</code>, <code>hourglass</code>, <code>soft curves</code>, <code>tank top</code>, <code>thong</code>, <code>photorealistic</code>, <code>confident</code>, <code>performer</code>, <code>consistent character</code>, <code>tattoo mapping</code></blockquote>

J-Rogue: Identity LoRA (SDXL / Pony)

3183120_training_data.zip

motion_V1 (1).safetensors

NsfwPovAllInOneLoraSdxl-000009.safetensors

V1 (from epoch 14) (from epoch 14) (from epoch 14)

V1 (from epoch 14) (from epoch 14)

V1 (from epoch 14)

⚠️ COMPLIANCE NOTICE: This motion LoRA and any associated outputs are purely fictional concepts. To prevent any confusion: this model does not utilize scraped footage of real individuals, nor is it based on a real person's likeness. The foundational logic, movement data, and physics frameworks were trained entirely on AI-generated synthetic content, localized ML dependencies, and open-source anatomical rigs.Overview: Let’s cut the bullshit. You aren’t downloading this to animate a sunset. This is a purpose-built motion LoRA engineered for top-tier, consenting adult animations. If you want stiff, lifeless, robotic movements, look elsewhere. If you want raw, fluid motion designed to turn heads, raise pulses, and finish the job, this asset is exactly where you need to be.This model doesn’t pull punches or play it safe. It understands weight, friction, and rhythm. The J-Series framework was trained specifically to handle high-arousal scenes with the mechanical precision and raw intensity required for premium NSFW content. It bridges the gap between static imagery and visceral, responsive motion.Technical Specifications &amp; Deployment:<ul><li>Base Framework Compatibility: AnimateDiff / SD 1.5</li><li>Optimal LoRA Weight: <code>0.75 - 0.85</code>. Pushing past <code>0.9</code> will induce skeletal artifacting and break the anatomical constraints.</li><li>Trigger Protocol: <code>j-series kinetics</code>, <code>visceral dynamic motion</code></li><li>Generation Parameters: 30-40 Steps | CFG Scale: 5.5 - 7.0</li><li>Training Integrity: Compiled locally via Kohya_ss over highly optimized step counts to ensure clean latent caching and zero dependency conflicts.</li></ul>Prompting Strategy:<ul><li>Keep negative prompts aggressive against structural mutations (e.g., <code>extra limbs, disjointed anatomy, bad physics</code>).</li><li>Use specific environmental friction tags to ground the motion (e.g., <code>heavy breathing, skin contact, weight shifting</code>).</li></ul>

Files with SHA256: 425c2834125df0b637caf336051b5f0934a6c26b42c01afa6f0cf9bc756e3583