Lora for Sakura from series: "The Reprise of the Spear Hero". Trained on both adult and kid version. Dataset is 27 images, I barely scrapped from all sources. Character is rarely appear in color.
Description
Retrained with better dataset and slightly bigger amount of data. Trained on 27 random images.
FAQ
Comments (1)
Actually I think I can improve and expand dataset since I missed some images for dataset, and retrain Lora for better result. Which model choose for training? Anima 1.0, Anima 2.9B or maybe Anima 3.9B?