<h1 id="heading-1135">El Andrew</h1>Hi there! Back again with another LoRa. This one was a bit harder to cook. It's based on the amazing <a target="_blank" rel="ugc" href="https://www.tiktok.com/@andrewtwk1">Andrew TWK</a>, and was trained on 84 different hand-captioned images from his <a target="_blank" rel="ugc" href="https://www.instagram.com/andrewtwk1/">Instagram</a>, <a target="_blank" rel="ugc" href="https://www.tiktok.com/@andrewtwk1">TikTok</a>, and Twitter. Should work fine at default strength, but ofc you can adjust it however you want.If you liked my model, feel free to drop a 👍😊!<h2 id="heading-106">Tips:</h2><ol><li>Don't forget the trigger word!</li><li>If you want his other longer hair hairstyle, use messy hair in the prompt.</li><li>Add closed eyes to the negative prompt if you're getting too many closed eyes images.</li><li>For images that aren't close-ups/portraits, I recommend using high-res fix so that the face doesn't look bad from far away.</li><li>Recommended Models<ul><li>Any decent realistic model (don't use the base checkpoint, it sucks)</li><li>Cartoon/anime checkpoints work as well</li></ul></li></ol>

Files with SHA256: d48846835ddbe1e2d0e858c6b6ef31a340939a48242ba7f625ed872a16b7a661