<h1 id="v1.1-update:">v1.1 Update:</h1><p>This update uses the previous model v1 as the base and was mixed with a couple of comic/toon checkpoints until I achieved a more "Toon" look than previously. I like how it turned out and think it may be more flexible than previously.</p><h1 id="v1.2-update:">v1.2 Update:</h1><p>This update uses the previous model v1.1 as the base and was mixed with a couple of checkpoints in an attempt to improve the generation of hands and feet.  </p><h2 id="prompting:">Prompting:</h2><p>I've started omitting the boilerplate quality tags as they result in a different look than I really was looking to achieve. With this model, I do like to use the following in the positive prompt:</p><pre><code>western comics \(style\), pinup \(style\), animification, </code></pre><p>The negative prompt I like:</p><pre><code>child, kid, childish, childlike, asian, muted color, flat color, anime screenshot, anime, patreon username, watermark, 3d,</code></pre><p>An example prompt:</p><pre><code>1girl, sexy, beautiful, upturned nose, adorable eyes, dutch angle, looking at viewer, twi'lek, from side,</code></pre><p><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/22195867-68be-480a-ab1e-d6bed54375b9/original=true/22195867-68be-480a-ab1e-d6bed54375b9.jpeg" /></p><p>I also use DPM++ 2S a when upscaling.</p><hr /><h2 id="end-update">END UPDATE</h2><hr /><h2 id="description">Description</h2><p>This model borrows from my own Illustrious model <a target="_blank" rel="ugc" href="https://civarchive.com/models/1309559"><strong>GDP Illustrious</strong></a> and several Loras, some I created and some I downloaded, as well as the absolutely fantastic <a target="_blank" rel="ugc" href="https://civarchive.com/models/1570391?modelVersionId=2329740"><strong>Nova Cartoon XL</strong></a> model to create more cartoonish/comic style images, but retains a lot of the flexibility of vanilla Illustrious models.</p><h2 id="ingredients">Ingredients</h2><p>This is a merge of several amazing Illustrious models that created<strong> </strong><a target="_blank" rel="ugc" href="https://civarchive.com/models/1309559"><strong>GDP Illustrious</strong></a> :</p><p>The primary model is an excellent realistic model, <a target="_blank" rel="ugc" href="https://civarchive.com/models/1025051?modelVersionId=1418758"><strong>NovaRealityXL</strong></a>. This model produces fantastic images, please check it out!</p><p>Nova was then merged with the awesome <a target="_blank" rel="ugc" href="https://civarchive.com/models/827184?modelVersionId=1410435"><strong>WAI-NSFW_Illustrious</strong></a>. I love this checkpoint!</p><p>The resultant checkpoint was then merged with another fantastic model that is more 2.5D, <a target="_blank" rel="ugc" href="https://civarchive.com/models/1025051?modelVersionId=1418758"><strong>Illustrij.</strong></a></p><p>Then, since I really love the results I was getting with this model: <a target="_blank" rel="ugc" href="https://civarchive.com/models/1570391?modelVersionId=2329740"><strong>Nova Cartoon XL</strong></a><strong> </strong>was merged back in along with <a target="_blank" rel="ugc" href="https://civarchive.com/models/681120?modelVersionId=1683072">DisneyLikeIllust </a>Lora, and a few merged other Loras.</p><h3 id="i-highly-recommend-consulting-the-descriptions-on-the-four-above-models-to-get-the-best-results."><strong>I highly recommend consulting the descriptions on the four above models to get the best results.</strong></h3><p></p><h2 id="parameters">Parameters</h2><p>Sampler: Euler A</p><p>Steps: 30 - 35</p><p>CFG: 5 - 7</p><p>VAE: sdxl_vae</p><p>Clip Skip: 2</p><p></p><h2 id="prompting">Prompting</h2><p>In the positive prompt, I almost always include some quality tags that I borrowed from this excellent <a target="_blank" rel="ugc" href="https://civarchive.com/articles/8380/tips-for-illustrious-xl-prompting-updates">article </a>by <span class="MentionNode_mention__q78sK" data-type="mention" data-id="mention:2156482" data-label="Cicatriz">@Cicatriz</span>. If you haven't checked it out, you should!</p><pre><code>masterpiece, best quality, absurdres, amazing quality, very aesthetic, newest, perfect quality, absolutely eye-catching, </code></pre><p>I've noticed Illustrious is more sensitive to negative prompting, so this might be overkill. I usually use the following in the negative prompt:</p><pre><code>child, kid, childish, childlike, loli, asian, </code></pre><p>An example prompt:</p><pre><code>masterpiece, best quality, absurdres, amazing quality, very aesthetic, newest, perfect quality, absolutely eye-catching, subject focus, 1girl, depth of field, realistic, photorealistic, large breasts, teasing, flirting, squatting, leaning back, arm support, cameltoe, spread legs, from above, highleg swimsuit, eyeliner, looking at viewer, tiger print, disney, jungle, blurry background, blurry foreground, {gray|brown|blonde|white|red|auburn|black|pink|blue} hair, {hair up|short twintails|updo|half updo|short hair|long hair|bob cut|pixie cut|medium hair|high ponytail|high twintails}, __devilkkw/body-1/hair_styles_front__, from {side|behind|front|above}, {green|brown|blue|grey} eyes,

Negative prompt:
child, kid, childish, childlike, loli, asian,  </code></pre><p></p>

GDP Illustrious Toon

<p>Updated model with different checkpoint merges</p>

Files with SHA256: 1f93cd800ff83a0c4a20b896ae97178a7b5461baafa0dad2655e8f1bcbb33f44