<h1 id="wingardium-leviosa!"><em><u>Wingardium Leviosa!</u></em></h1><h1></h1><h2 id="hello-fellow-citizens!">Hello fellow Citizens!</h2><p></p><p>So, its been a minute. A lot happened. Anyhow, I am still saving for a new GPU. Meanwhile, I recently found myself interested in Hagrid from the Harry Potter series. Everyone talks about dumbledore being gay (meme) but why not a <em>manly</em> version of Hagrid!<br /><br />Now, this is my first ever realistic lora. It is <em>not</em> perfect. A lot of the images were from Movie 1 and Movie two and were a bitch to find :/ So, I will eventually revisit this in future. If any of you know how to train a realistic lora, please inform me. I want to learn ;-;<br /><br />So, I have done some small testing. Not a lot as I use a timed service. However, there is some promising works. I would recommend keeping things to a realistic checkpoint, and keep things (no pun intended) above belt as they are the best images. I also recommend using ADetailer / Face Detailer for best results. However, there were some issues with clothing - <strong>if you do nude, add "Long Coat, red shirt, brown vest" to the negative prompts.</strong> Otherwise, it will mix it in with other clothing.</p><p></p><p>So far, other lora's seem to be okay with it, however not the best. As I said, this is only an early version - I will with your help and knowledge, improve this in future ^^</p><p></p>

Files with SHA256: fd7560c491732377ec47c665f6600131ddcff4cd7a3bdd4f1afaa2ac9353b949