This is my first model trained for LoRa. It turned out pretty good (especially highres), considering the time spent on making the dataset.I recommend using 0.6-0.8 for weights, although 1 still looks decent. No trigger words are required, however you can also include <code>bartolomeobari</code> in your prompt. Trained on <a target="_blank" rel="ugc" href="https://civarchive.com/models/14305/heavenorangemix">HeavenOrangeMix</a>, but other anime models should also work. It plays well with other LoRas, but weights need to be lowered otherwise the image gets fried.NotesThe model tends to produce close-ups, so I recommend adding <code>close-up</code> to negatives. Additionally, the default face expression isn't neutral and tends towards tokens like <code>:3</code> or <code>wavy mouth</code>. Adding face expressions to your prompts or adding the above to your negative prompt should improve your generations. After I get some time to clean up the dataset and retrain the model, these problems should go away.

Files with SHA256: 1200a3f17c2ae720e5e0aa745e83ecf86ed57fae73172af55962801fb3337c7c