chaRLI

詳細

ファイルをダウンロード (1)

このバージョンについて

Version 1.0 - Initial Release

This LoRA is designed to generate the character "Charli Rae Carter" with high fidelity. It is trained on a curated dataset of photorealistic images, including:

- Front, side, and back views for 3D consistency.

- "Wet" and "glistening" skin textures for intimate scenarios.

- Specific poses: mirror selfies, shower scenes, and poolside shots.

- Detailed facial features: honey-blonde hair, green eyes, and freckles.

Key Features:

- Excellent at photorealistic portraits and full-body shots.

- Captures the "Tennessee girl" aesthetic and vulnerable, intimate vibe.

- Trained to maintain consistency across different lighting and backgrounds.

How to Use:

- Trigger Word: ch@rL1 (Must be included in every prompt)

- Recommended Strength: 0.7 - 0.9 (Start at 0.8 for best results)

- Base Model: SDXL 1.0 or Wan 2.7 (Photorealistic models work best)

- Negative Prompt: "deformed, bad anatomy, extra limbs, blurry, cartoon, anime, low quality, text, watermark"

Training Notes:

- This model was trained on a mix of SFW and NSFW data to ensure versatility while maintaining character integrity.

- Best results are achieved with prompts that specify the desired lighting and mood (e.g., "soft lighting," "cinematic," "moody").

Enjoy creating with Charli!

モデル説明

This is a custom LoRA model designed to generate a photorealistic character named "Charli Rae Carter" (trigger word: ch@rL1).

What it does:

This model captures the specific facial features, body structure, and "Tennessee girl" aesthetic of the character. It is trained to reproduce her honey-blonde hair, green eyes, freckles, and slender, toned physique with high fidelity across various angles (front, side, back) and lighting conditions.

What it's good at:

- Photorealistic portraits and full-body shots.

- Specific scenarios: "Mirror Selfies," "Wet/Sweaty" textures, and intimate close-ups.

- Consistency in facial identity and body proportions.

- Reproducing specific poses like "looking away," "looking at viewer," and "side profiles."

Best use cases:

- Creating NSFW or spicy content for adult platforms.

- Generating lifestyle photography (poolside, gym, bathroom, bedroom).

- Producing consistent character sheets for video generation or animation.

How to use:

- Always include the trigger word "ch@rL1" in your prompt.

- Recommended strength: 0.7 to 0.9 (adjust based on desired likeness).

- Pair with a photorealistic base model (like SDXL or Wan 2.7) for best results.

- Use negative prompts like: "deformed, bad anatomy, extra limbs, blurry, cartoon, anime, low quality" to ensure photorealism.

Limitations:

- Like all LoRAs, it works best when paired with a strong base model.

- Extreme angles or very complex group shots may require prompt engineering to maintain consistency.

このモデルで生成された画像

並び替え

モデルタイプ	LORA
ベースモデル	SDXL 1.0
公開日	2026-07-09
トレーニングワード	ch@rL1