I decided to make this LoRA both as a test of a scraping script and an examination of a different way of training. Usage: <ul><li>As with most character LoRAs, a strength of 1 works best. </li><li>Be specific with the clothing you are looking for, otherwise, you will just get something revealing. </li><li>The keyword <code>smiling</code> tends to give the best faces.</li><li>It works best on realism models, but can even work decent on others like Dreamshaper.</li></ul>Some quick notes on the training: <ul><li>I used full BF16 training, my usual precision.</li><li>I used WD14 captioning which performed a lot better than BLIP in my opinion.</li><li>Trained at 1080x1080, but I probably could have pushed it further. <ul><li>It seems to have a preference for more vertical images because of the dataset.</li></ul></li><li>I believe it is because of the captioning method, but it took a network rank of 128 to get the face right. Otherwise, the LoRA was just getting from the neck down decent. <ul><li>Perhaps this is a clothing LoRA as well given generic tags? Curious to see what could be done with this tidbit. </li></ul></li></ul>Once again, this was created as a test. It's not the greatest character LoRA I have ever made, and didn't really take much effort to make. But that's the point of this little experiment. It does, however, produce some good results with a little experimenting.

Files with SHA256: 7da9323356e3c145628257225b2a3e32709cc67d210611e6646f78e74cb2f2e3