<h1 id="20241027-8wpdbdpxh">20241027</h1>The training of this model and the images it generates are only for learning purposes.Hinoa from Monster Hunter RiseTrigger Words: mhhinoaI think she’s the most beautiful 受付孃/看板娘 🤩Right now, there are still many details the model haven't learned well, such as the hair ornaments on the back view, the standing pose in full-body shot and the outdoor backgroundOn the other hand, since all the front view images of this character have hair ornaments, it’s almost impossible to remove them through prompts. I’ve tried manually removing the hair ornaments from a few images and retraining the dataset, but the effect is that I can occasionally remove the headpiece by adding 'hair ornament' to the negative prompt.Adding tags like 'realistic' and 'photorealistic' can enhance the character's features. Among them, 'photorealistic' will give a smoother effect<h2 id="note-coh3nj37l">Note</h2><ul><li>The model's output may heavily rely on the checkpoint.</li><li>You could find the example prompts from my post.</li><li>Adding tag such as 'realistic', 'realism' could enhance the features of the character.</li><li>My prompts are basically composed in the order of [character traits] + [style] + [expression] + [clothing] + [camera and action] + [background], and you can delete or modify them as needed.</li><li>Recommended weight: 1~0.6, adjust as needed until the character's appearance meets your requirements.</li><li>Upscale value recommendation is around 1.3~2.0, denoising strength is 0.2</li><li>Facial distortion may easily occur in situations such as full-body shots. If there is facial distortion, consider using ADetailer for repair</li></ul>

Files with SHA256: 6c31a94ea5686b7ad439030f972b1c2ecdb01d310e8bad51fdf2269935371d65