High-quality photorealistic SDXL LoRA trained to reproduce the same female identity across different angles, lighting conditions and compositions.Designed for realistic Instagram-style portraits and full-body photos.Trained on 25 high-resolution images with consistent identity, using minimal captions focused on pose and camera angle.Best results at LoRA weight 0.7–0.8.Works best with SDXL-based models and natural lighting prompts.Not intended for anime or heavily stylized generations.

Files with SHA256: fd8bf76c55e890d87e0b9c51e6fc4739eb20d3810fc7109b247c48b65a54eb69