Alin Krea is a photorealistic character LoRA trained on Krea 2.
This LoRA is designed to reproduce the Alin character consistently across a wide range of facial expressions, camera angles, outfits, lighting conditions, and backgrounds.
The training dataset contains multiple expression variations and portrait compositions intended to improve facial identity consistency while maintaining prompt flexibility.
Trigger word:
Alin
Recommended uses:
Photorealistic portraits
Character-focused photography
Facial expression generation
Close-up and head-and-shoulders portraits
Lifestyle and cinematic photography
Different camera angles, clothing, lighting, and environments
Example prompt:
Alin, woman, photorealistic portrait, natural skin texture, soft natural lighting, close-up portraitExpression examples:
neutral, happy, playful, bored, angry, terrified, determinedThis is the Krea 2 version of the Alin character LoRA.
Description
Continued training from the epoch 9 checkpoint of v1.0.
This revision extends the previous Krea 2 training to further improve:
Facial identity consistency
Expression accuracy
Facial detail across different camera angles
Generalization across lighting, clothing, and backgrounds
Trigger word: Alin
This version is a continuation of the previous training run, not a new training from scratch.



