<pre><code>**Trigger:** `lovina` (required only for multi-character LoRAs)

Krea2 LoRA of a female character.

**Training:** Trained with OneTrainer on the Krea2-RAW model; tested with the Krea2-TURBO model at 8 steps, CFG 0, using the `exp_heun_2_x0` sampler and `ddim_uniform` scheduler.

**Disclaimer:** No images of real people were used in training. Any resemblance to a real person is purely coincidental. All source images were generated with SDXL Juggernaut or Grok Imagine. Qwen Image Edit 2511 and Grok Imagine were used to generate additional camera angles. The dataset is available upon request.
</code></pre><p></p>

Files with SHA256: dbc3c9753f721dbabd453fbcc2847a41cc3ce6be1c15fed47c9e32bdfc75caad