It takes substantial time and efforts to bake models. If you appreciate my models, I would be grateful if you could support me on <a target="_blank" rel="ugc" href="https://ko-fi.com/koolchh">Ko-fi</a> ☕.<h1 id="heading-103">AnimeBoysXL</h1><h2 id="heading-104">Features</h2><ul><li>✔️ Good for inference: AnimeBoysXL is a flexible model which is good at generating images of anime boys and males-only content in a wide range of styles.</li><li>✔️ Good for training: AnimeBoysXL is suitable for further training, thanks to its neutral style and ability to recognize a great deal of concepts. Feel free to train your own anime boy model/LoRA from AnimeBoysXL.</li><li>❌ AnimeBoysXL is not optimized for creating anime girls. Please consider using other models for that purpose.</li></ul><h2 id="heading-105">Inference Guide</h2><ul><li>Prompt: Use tag-based prompts to describe your subject.<ul><li>Tag ordering matters. It is highly recommended to structure your prompt with the following templates:<ul><li><code>1boy, male focus, character name, series name, anything else you'd like to describe</code></li><li><code>2boys, male focus, multiple boys, character name(s), series name, anything else you'd like to describe</code></li></ul></li><li>Append <code>, best quality, amazing quality, best aesthetic, absurdres</code> to the prompt to improve image quality.</li><li>(Optional) Append <code>, year YYYY</code> to the prompt to shift the output toward the prevalent style of that year. <code>YYYY</code> is a 4 digit year, e.g. <code>, year 2023</code></li><li>For more detailed documentation, you can visit my <a target="_blank" rel="ugc" href="https://ko-fi.com/post/Advanced-Prompt-Guide-for-AnimeBoysXL-V3-Z8Z2WWYHS">article</a> on Ko-fi (available to supporters only).</li></ul></li><li>Negative prompt: Choose from one of the following two presets.<ol><li>Heavy (recommended): <code>lowres, (bad:1.05), text, error, missing, extra, fewer, cropped, jpeg artifacts, worst quality, bad quality, watermark, bad aesthetic, unfinished, chromatic aberration, scan, scan artifacts, 1girl, breasts</code></li><li>Light: <code>lowres, jpeg artifacts, worst quality, watermark, blurry, bad aesthetic, 1girl, breasts</code></li></ol><ul><li>(Optional) Add <code>, realistic, lips, nose</code> to the negative prompt if you need a flat anime-like style face.</li></ul></li><li>VAE: Make sure you're using <a target="_blank" rel="ugc" href="https://huggingface.co/stabilityai/sdxl-vae/tree/main">SDXL VAE</a>.</li><li>Sampling method, sampling steps and CFG scale: I find (Euler a, 28, 5) good. You are encouraged to experiment with other settings.</li><li>Width and height: 832*1216 for portrait, 1024*1024 for square, and 1216*832 for landscape.</li></ul><h2 id="heading-106">Training Details</h2>AnimeBoysXL is trained from <a target="_blank" rel="ugc" href="https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0">Stable Diffusion XL Base 1.0</a>, on ~516k images.The following tags are attached to the training data to make it easier to steer toward either more aesthetic or more flexible results.<h3 id="heading-107">Quality tags</h3><ul><li><code>best quality</code>: score &gt;= 150</li><li><code>amazing quality</code>: score in the range of [100, 150)</li><li><code>great quality</code>: score in the range of [75, 100)</li><li><code>normal quality</code>: score in the range of [0, 75)</li><li><code>bad quality</code>: score in the range of (-5, 0)</li><li><code>worst quality</code>: score &lt;= -5</li></ul><h3 id="heading-108">Aesthetic tags</h3><ul><li><code>best aesthetic</code>: score &gt;= 6.675</li><li><code>great aesthetic</code>: score in the range of [6.0, 6.675)</li><li><code>normal aesthetic</code>: score in the range of [5.0, 6.0)</li><li><code>bad aesthetic</code>: score &lt; 5.0</li></ul><h3 id="heading-109">Rating tags</h3><ul><li><code>sfw</code>: general</li><li><code>slightly nsfw</code>: sensitive</li><li><code>fairly nsfw</code>: questionable</li><li><code>very nsfw</code>: explicit</li></ul><h3 id="heading-110">Year tags</h3><code>year YYYY</code> where <code>YYYY</code> is in the range of [2005, 2023].<h3 id="heading-111">Training configurations</h3><ul><li>Hardware: 4 * Nvidia A100 80GB GPUs</li><li>Optimizer: AdaFactor</li><li>Gradient accumulation steps: 8</li><li>Batch size: 4 * 8 * 4 = 128</li><li>Learning rates:<ul><li>8e-6 for U-Net</li><li>5.2e-6 for text encoder 1 (CLIP ViT-L)</li><li>4.8e-6 for text encoder 2 (OpenCLIP ViT-bigG)</li></ul></li></ul><h3 id="heading-3">Changes from v1.0</h3><ul><li>Train with tag ordering.</li><li>Add <code>sfw</code> rating tag.</li><li>More epochs on the questionable and explicit rating subset.</li><li>FP16 mixed-precision training for final epochs.</li></ul>The above model card is for AnimeBoysXL v2.0. For the model card of v1.0, please refer to <a target="_blank" rel="ugc" href="https://huggingface.co/Koolchh/AnimeBoysXL-v1.0">here</a>.

AnimeBoysXL

<ul><li>Train with tag ordering.</li><li>Add <code>sfw</code> rating tag.</li><li>More epochs on the questionable and explicit rating subset.</li><li>FP16 mixed-precision training for final epochs.</li></ul>

Files with SHA256: 992f15b27ac280dc42ebd27e136fb6cb2f67dff0d4eddb05a6c8ffc730fe56a5