[SD 1.5][TextualInversion]Anna Kendrick Embedding V1
<p><strong>Anna Kendrick</strong> embedding created by <strong>Airathias</strong></p><p></p><p>Main V1 embedding was created in 800 steps, and a pretty high number of vectors per token. <em>Try</em> to keep your prompt short, though I've seen it work fine with longer prompts. If you find the face not defined enough, just <em>(annakendrick:1.2)</em> it.</p><p>As you'll see, it doesn't like the base model SD 1.5 much, and it leans toward better results in photorealistic renders. It <em>does</em> perform for certain fantasy semi-realistic artstyles, though. Check the X/Y grid linked below for a comparison with a few checkpoints.</p><p>This time around I decided to offer the option of more epochs, just in case. I've created a bunch of X/Y grids for you to pick and choose:</p><ul><li><p>CFG Scales (<a target="_blank" rel="ugc" href="https://airathias.com/media/public/annakendrick-cfg-scales.png">https://airathias.com/media/public/annakendrick-cfg-scales.png</a>)</p></li><li><p>Checkpoints (<a target="_blank" rel="ugc" href="https://airathias.com/media/public/annakendrick-checkpoints.png">https://airathias.com/media/public/annakendrick-checkpoints.png</a>)</p></li><li><p>Colors for checking overtraining (<a target="_blank" rel="ugc" href="https://airathias.com/media/public/annakendrick-colors-overtraincheck.png">https://airathias.com/media/public/annakendrick-colors-overtraincheck.png</a>)</p><p></p></li></ul><p>For the initial version I chose the 3 I liked best, <strong>800</strong>,<strong> 1000 </strong>and <strong>1500</strong>. If you'd like another epoch count embedding let me know.</p><p>I added prompts to the example images, should work fine. These prompts were basically blatantly copied from other examples throughout the website. If I copied yours, mea culpa. You can use this without crediting me.</p><p>The embedding was trained on stock SD 1.5, so it should work for many if not most models based off of 1.5.</p><p>Images listed are made using a handful of models to test, including:</p><ul><li><p>Stock SD 1.5</p></li><li><p>DreamShaper (<a target="_blank" rel="ugc" href="https://civitai.com/models/4384/dreamshaper">https://civitai.com/models/4384/dreamshaper</a>)</p></li><li><p>UberRealistic Dreamy PFG Porn (<a target="_blank" rel="ugc" href="https://civitai.com/models/3751/uberrealistic-dreamy-pfg-porn">https://civitai.com/models/3751/uberrealistic-dreamy-pfg-porn</a>)</p></li><li><p>Babes (<a target="_blank" rel="ugc" href="https://civitai.com/models/2220/babes">https://civitai.com/models/2220/babes</a>)</p><p></p></li></ul><p><strong>Important for the examples</strong>: In the example images I use a keyword like "<em>annakendrick-sd-v2-800</em>", this was just for testing purposes but I've elected to keep it in there so you can see what version I ended up using for that render. In your prompt you should use "<em><u>annakendrick</u></em>", "<em><u>annakendrick-1000</u></em>" or "<em><u>annakendrick-1500</u></em>".</p><p></p><p>Also, the model names in the examples are like "<em>SD_15_Photoreal_NSFW_<name></em>". This is my local model name formatting templated like "<em><basemodel>_<artstyle>_<sfw/nsfw>_<name></em>". Hashes are shown as well, but just in case you were wondering.</p>
页:
[1]