Part of my <a target="_blank" rel="ugc" href="https://civarchive.com/tag/faces%20by%20thegipper">facial expressions</a> series. All of these LoRAs are designed to mesh well with each other and have low impact on your image framing.People on cloud nine. Smiles, laughs, etc. Did this one last because SD generally can do this fairly well, but <a target="_blank" rel="ugc" href="https://civarchive.com/models/237932/shocked-facial-expression">similar to shocked</a>, it's not great at injecting the expression into more complex compositions and mixing it with other expressions.This one sets up a full <a rel="ugc" href="https://civarchive.com/images/5580133">6 prototypical expressions that can be remixed</a> into a pretty wide palette. Definitely has a fair amount of unintended style transfer (trends towards high saturation film-grain professional style photos) I still need to train out. Probably my last fresh one for a little bit before I do some clean up of the expression models to clear out some overfitting before trying to make a combined model.Starter prompts:&lt;subject&gt;, happy&lt;subject&gt;, &lt;context&gt; BREAK happy expression

Files with SHA256: fca6aa4492b628bab3d2d8e8304174019ea267f30d772599b56ad0ee378aad2b