{"id":2549,"date":"2023-07-03T11:11:31","date_gmt":"2023-07-03T11:11:31","guid":{"rendered":"https:\/\/www.codeastar.com\/?p=2549"},"modified":"2023-07-04T17:51:21","modified_gmt":"2023-07-04T17:51:21","slug":"learn-how-to-use-stable-diffusion-part-2-sampling-methods","status":"publish","type":"post","link":"https:\/\/www.codeastar.com\/learn-how-to-use-stable-diffusion-part-2-sampling-methods\/","title":{"rendered":"Learn how to use Stable Diffusion Part 2 &#8211; Sampling Methods"},"content":{"rendered":"\n<p>In our last post, <a href=\"https:\/\/www.codeastar.com\/stable-diffusion-quick-and-easy-guide-for-everyone-part-1\/\" target=\"_blank\" rel=\"noreferrer noopener\">Learn how to use Stable Diffusion &#8211; Part 1<\/a>, we mentioned sampling methods and stated that this is a huge topic. Right now, let&#8217;s continue our Stable Diffusion learning journey. Besides discussing the main course, Sampling Methods, we are going to have appetizers as well, like prompt and checkpoint. We hope you all can enjoy the meal this time. Let&#8217;s dig in!<\/p>\n\n\n\n<h3 class=\"wp-block-heading\">Appetizer I: Prompt<\/h3>\n\n\n\n<p>Most of the generative art starts with text inputs, the prompts, so let us start from here. We knew from our <a href=\"https:\/\/www.codeastar.com\/stable-diffusion-quick-and-easy-guide-for-everyone-part-1\/\" target=\"_blank\" rel=\"noreferrer noopener\">last post<\/a>, prompts can be inputted as a full sentence or a list of keywords. For beginners, we would suggest using full sentences as inputs first. So you may see what you have inputted in a human-readable way. When you are more familiar with Stable Diffusion, you can use keywords as the prompts for a faster process. Speaking of keywords, here are the major types we use as inputs:<\/p>\n\n\n\n<ul>\n<li>Subject &#8211; the things we are going to generate (e.g. flying birds, Joe Biden, bedroom, a bear using a computer)<\/li>\n\n\n\n<li>Position &#8211; the poses of the subject(s) (e.g. full body, half body, look at viewers, standing, sitting) <\/li>\n\n\n\n<li>Style &#8211; the mood of the artwork (e.g. sketch, oil paint, cartoonish, Picasso&#8217;s Cubism style, black and white, vintage)<\/li>\n\n\n\n<li>Resolution &#8211; how sharp and detailed the picture should be (e.g. 4K, 8K, Ultra HD, sharp focus), my personal note, more detailed output just looks more unreal<\/li>\n<\/ul>\n\n\n\n<p>Other than learning the type of prompts, we should learn the weight of prompts as well. We can use <strong>()<\/strong> with a keyword and a value to strengthen or weaken the weight of the keyword. For example, <em>(robot: 1.2)<\/em> strengthens the &#8220;robot&#8221; keyword, and vice versa, (robot: 0.9) weakens the &#8220;robot&#8221; keyword. We can also use just <strong>()<\/strong> on a keyword to emphasize the weight.<\/p>\n\n\n\n<p>When we group all the things together, we get the following prompts: <\/p>\n\n\n\n<figure class=\"wp-block-image aligncenter size-full\"><img data-attachment-id=\"2587\" data-permalink=\"https:\/\/www.codeastar.com\/learn-how-to-use-stable-diffusion-part-2-sampling-methods\/stable-diffusion-1\/\" data-orig-file=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/Stable-Diffusion-1.png?fit=626%2C111&amp;ssl=1\" data-orig-size=\"626,111\" data-comments-opened=\"1\" data-image-meta=\"{&quot;aperture&quot;:&quot;0&quot;,&quot;credit&quot;:&quot;&quot;,&quot;camera&quot;:&quot;&quot;,&quot;caption&quot;:&quot;&quot;,&quot;created_timestamp&quot;:&quot;0&quot;,&quot;copyright&quot;:&quot;&quot;,&quot;focal_length&quot;:&quot;0&quot;,&quot;iso&quot;:&quot;0&quot;,&quot;shutter_speed&quot;:&quot;0&quot;,&quot;title&quot;:&quot;&quot;,&quot;orientation&quot;:&quot;0&quot;}\" data-image-title=\"Stable-Diffusion-1\" data-image-description=\"\" data-image-caption=\"\" data-medium-file=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/Stable-Diffusion-1.png?fit=300%2C53&amp;ssl=1\" data-large-file=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/Stable-Diffusion-1.png?fit=626%2C111&amp;ssl=1\" decoding=\"async\" loading=\"lazy\" width=\"626\" height=\"111\" src=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/Stable-Diffusion-1.png?resize=626%2C111&#038;ssl=1\" alt=\"Prompts \" class=\"wp-image-2587\" srcset=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/Stable-Diffusion-1.png?w=626&amp;ssl=1 626w, https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/Stable-Diffusion-1.png?resize=300%2C53&amp;ssl=1 300w\" sizes=\"(max-width: 626px) 100vw, 626px\" data-recalc-dims=\"1\" \/><\/figure>\n\n\n\n<p>Here comes our result:<\/p>\n\n\n\n<figure class=\"wp-block-image aligncenter size-medium\"><img data-attachment-id=\"2588\" data-permalink=\"https:\/\/www.codeastar.com\/learn-how-to-use-stable-diffusion-part-2-sampling-methods\/00140-2795559298\/\" data-orig-file=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/00140-2795559298.png?fit=512%2C512&amp;ssl=1\" data-orig-size=\"512,512\" data-comments-opened=\"1\" data-image-meta=\"{&quot;aperture&quot;:&quot;0&quot;,&quot;credit&quot;:&quot;&quot;,&quot;camera&quot;:&quot;&quot;,&quot;caption&quot;:&quot;&quot;,&quot;created_timestamp&quot;:&quot;0&quot;,&quot;copyright&quot;:&quot;&quot;,&quot;focal_length&quot;:&quot;0&quot;,&quot;iso&quot;:&quot;0&quot;,&quot;shutter_speed&quot;:&quot;0&quot;,&quot;title&quot;:&quot;&quot;,&quot;orientation&quot;:&quot;0&quot;}\" data-image-title=\"00140-2795559298\" data-image-description=\"\" data-image-caption=\"\" data-medium-file=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/00140-2795559298.png?fit=300%2C300&amp;ssl=1\" data-large-file=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/00140-2795559298.png?fit=512%2C512&amp;ssl=1\" decoding=\"async\" loading=\"lazy\" width=\"300\" height=\"300\" src=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/00140-2795559298.png?resize=300%2C300&#038;ssl=1\" alt=\"Sad Keanu Reeves\" class=\"wp-image-2588\" srcset=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/00140-2795559298.png?resize=300%2C300&amp;ssl=1 300w, https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/00140-2795559298.png?resize=150%2C150&amp;ssl=1 150w, https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/00140-2795559298.png?w=512&amp;ssl=1 512w\" sizes=\"(max-width: 300px) 100vw, 300px\" data-recalc-dims=\"1\" \/><\/figure>\n\n\n\n<p>Since we have given 1.3 weight to the &#8220;taco&#8221; keyword. So we can see 2 giant tacos in our output :]] .<\/p>\n\n\n\n<h3 class=\"wp-block-heading\">Appetizer II: Checkpoint<\/h3>\n\n\n\n<p>We learn from our machine learning exercises, what we got is what we trained. So the model&#8217;s drawing styles and capabilities are all depended on what it has been trained. On Stable Diffusion, according to its <a href=\"https:\/\/github.com\/CompVis\/stable-diffusion\/blob\/main\/Stable_Diffusion_v1_Model_Card.md\" target=\"_blank\" rel=\"noreferrer noopener\">GitHub page<\/a>, the model is trained with a dataset of 5 billion image-text pairs from the open source project <a href=\"https:\/\/laion.ai\/\" target=\"_blank\" rel=\"noreferrer noopener\">LAION<\/a>. The large image dataset makes Stable Diffusion literally know many things \/ styles to draw. But it also brings out the general issue of &#8220;<a href=\"https:\/\/www.codeastar.com\/flask-backend-react-frontend-weather-1\/\" target=\"_blank\" rel=\"noreferrer noopener\">the bard problem<\/a>&#8221; &#8212; Jack of all trades, master of none. <\/p>\n\n\n\n<p>So people start using Stable Diffusion model as a base model, to train their own checkpoints, with their preferred image datasets. At the end, the images from the additional datasets would favor the output generation algorithm, in other to produce the preferred content and styles. There are plenty of pre-trained checkpoints with various styles and content from <a href=\"https:\/\/civitai.com\/\" target=\"_blank\" rel=\"noreferrer noopener\">CivitAI<\/a>. We can find the differences between those checkpoint models and the Stable Diffusion default model from the images below.<\/p>\n\n\n\n<pre class=\"wp-block-code\"><code>Our prompts: <em>woman, outdoor, half body, side view, busy street, (winter), 4k<\/em><\/code><\/pre>\n\n\n\n<p>Our outputs and the checkpoint models (with downloadable links):<\/p>\n\n\n\n<figure class=\"wp-block-image aligncenter size-full\"><img data-attachment-id=\"2597\" data-permalink=\"https:\/\/www.codeastar.com\/learn-how-to-use-stable-diffusion-part-2-sampling-methods\/sd_models\/\" data-orig-file=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/sd_models.png?fit=1000%2C220&amp;ssl=1\" data-orig-size=\"1000,220\" data-comments-opened=\"1\" data-image-meta=\"{&quot;aperture&quot;:&quot;0&quot;,&quot;credit&quot;:&quot;&quot;,&quot;camera&quot;:&quot;&quot;,&quot;caption&quot;:&quot;&quot;,&quot;created_timestamp&quot;:&quot;0&quot;,&quot;copyright&quot;:&quot;&quot;,&quot;focal_length&quot;:&quot;0&quot;,&quot;iso&quot;:&quot;0&quot;,&quot;shutter_speed&quot;:&quot;0&quot;,&quot;title&quot;:&quot;&quot;,&quot;orientation&quot;:&quot;0&quot;}\" data-image-title=\"sd_models\" data-image-description=\"\" data-image-caption=\"\" data-medium-file=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/sd_models.png?fit=300%2C66&amp;ssl=1\" data-large-file=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/sd_models.png?fit=1000%2C220&amp;ssl=1\" decoding=\"async\" loading=\"lazy\" width=\"1000\" height=\"220\" src=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/sd_models.png?resize=1000%2C220&#038;ssl=1\" alt=\"\" class=\"wp-image-2597\" srcset=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/sd_models.png?w=1000&amp;ssl=1 1000w, https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/sd_models.png?resize=300%2C66&amp;ssl=1 300w, https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/sd_models.png?resize=768%2C169&amp;ssl=1 768w\" sizes=\"(max-width: 1000px) 100vw, 1000px\" data-recalc-dims=\"1\" \/><\/figure>\n\n\n\n<ul>\n<li>Stable Diffusion &#8211; the original model, it does what we have inputted<\/li>\n\n\n\n<li><a href=\"https:\/\/civitai.com\/models\/4201?modelVersionId=29460\" target=\"_blank\" rel=\"noreferrer noopener\">Realistic Vision<\/a> &#8211; facial and background objects are generated with more details, i.e. more realistic :]]<\/li>\n\n\n\n<li><a href=\"https:\/\/civitai.com\/models\/43331?modelVersionId=94640\" target=\"_blank\" rel=\"noreferrer noopener\">majicMix realistic<\/a> &#8211; another realistic model, but it is trained with Asian-themed material <\/li>\n\n\n\n<li><a href=\"https:\/\/civitai.com\/models\/46049\/xenoengine-an-art-style-mega-model\" target=\"_blank\" rel=\"noreferrer noopener\">XenoEngine<\/a> &#8211; a model trained with the comic style<\/li>\n\n\n\n<li><a href=\"https:\/\/civitai.com\/models\/4468?modelVersionId=57618\" target=\"_blank\" rel=\"noreferrer noopener\">Countererfeit<\/a> &#8211; just likes the above model but with the anime style this time<\/li>\n<\/ul>\n\n\n\n<p>We have the model downable links, then how do we apply it to Stable Diffusion WebUI? Just simply copy them to  <\/p>\n\n\n\n<pre class=\"wp-block-code\"><code>&lt;stable diffusion webui path&gt;\\models\\Stable-diffusion\\<\/code><\/pre>\n\n\n\n<p>Reload your WebUI, then you can select the installed model to generate your graphics.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\">Main Course 1: Sampling Methods in Stable Diffusion<\/h3>\n\n\n\n<p>We find a list of sampling methods (samplers) available in the WebUI. Then it is always a question, which sampler should we use? Before we find out the answer, let us quickly go through those major samplers here.<\/p>\n\n\n\n<ul>\n<li><strong>Euler<\/strong> &#8211; it came from the <a href=\"https:\/\/en.wikipedia.org\/wiki\/Euler_method\" target=\"_blank\" rel=\"noreferrer noopener\">Euler method<\/a> (around the 16th centuries), which is used to approximate solutions to differential equations<\/li>\n\n\n\n<li><strong>LMS<\/strong> &#8211; <a href=\"https:\/\/en.wikipedia.org\/wiki\/Linear_multistep_method\" target=\"_blank\" rel=\"noreferrer noopener\">Linear Multistep Method<\/a>, a multi-step version of the single step&#8217;s Euler for better accuracy<\/li>\n\n\n\n<li><strong>Heun<\/strong> &#8211; an improved Euler method from <a href=\"https:\/\/en.wikipedia.org\/wiki\/Karl_Heun\" target=\"_blank\" rel=\"noreferrer noopener\">Karl Heun<\/a> <\/li>\n\n\n\n<li><strong>DPM <\/strong>&#8211; <a href=\"https:\/\/en.wikipedia.org\/wiki\/Diffusion_model\" target=\"_blank\" rel=\"noreferrer noopener\">Diffusion Probabilistic Model<\/a>, it learns how the noised image is constructed then goes to denoise it<\/li>\n\n\n\n<li><strong>DDIM<\/strong> &#8211; <a href=\"https:\/\/arxiv.org\/abs\/2010.02502\" target=\"_blank\" rel=\"noreferrer noopener\">Denoising Diffusion Implicit Model<\/a>, an improved and faster denoising model from DPM<\/li>\n\n\n\n<li><strong>PLMS<\/strong> &#8211; Pseudo-Linear Multi-Step model, actually it is a sub-process of <a href=\"https:\/\/arxiv.org\/abs\/2202.09778\" target=\"_blank\" rel=\"noreferrer noopener\">Pseudo Numerical Methods for Diffusion Models<\/a>, another improved DPM<\/li>\n\n\n\n<li><strong>UniPC<\/strong> &#8211; <a href=\"https:\/\/arxiv.org\/abs\/2302.04867\" target=\"_blank\" rel=\"noreferrer noopener\">Unified Predictor-Corrector Framework<\/a>, a new and fast model released in 2023<\/li>\n\n\n\n<li><strong>&lt;sampler&gt; a <\/strong>&#8211; a sampler with &#8220;a&#8221; means this is an ancestral sampler. While other samplers converge when they settle the error range, an ancestral would keep adding noise to sampling steps<\/li>\n<\/ul>\n\n\n\n<p>For a better comparison, we generate AI graphics using the following conditions with different samplers.<\/p>\n\n\n\n<ul>\n<li><strong>Model<\/strong> &#8211; Stable Diffusion v1.5<\/li>\n\n\n\n<li><strong>Prompt<\/strong> &#8211; A middle century art gallery with several digital art pieces<\/li>\n\n\n\n<li><strong>Steps<\/strong> &#8211; 27<\/li>\n\n\n\n<li><strong>Image Size<\/strong> &#8211; 512 x 512<\/li>\n\n\n\n<li><strong>CFG Scale<\/strong> &#8211; 7<\/li>\n\n\n\n<li><strong>Seed<\/strong> &#8211; 1573819953<\/li>\n<\/ul>\n\n\n\n<figure class=\"wp-block-image aligncenter size-full\"><img data-attachment-id=\"2621\" data-permalink=\"https:\/\/www.codeastar.com\/learn-how-to-use-stable-diffusion-part-2-sampling-methods\/art_gallery_1a\/\" data-orig-file=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/art_gallery_1a.png?fit=1000%2C236&amp;ssl=1\" data-orig-size=\"1000,236\" data-comments-opened=\"1\" data-image-meta=\"{&quot;aperture&quot;:&quot;0&quot;,&quot;credit&quot;:&quot;&quot;,&quot;camera&quot;:&quot;&quot;,&quot;caption&quot;:&quot;&quot;,&quot;created_timestamp&quot;:&quot;0&quot;,&quot;copyright&quot;:&quot;&quot;,&quot;focal_length&quot;:&quot;0&quot;,&quot;iso&quot;:&quot;0&quot;,&quot;shutter_speed&quot;:&quot;0&quot;,&quot;title&quot;:&quot;&quot;,&quot;orientation&quot;:&quot;0&quot;}\" data-image-title=\"art_gallery_1a\" data-image-description=\"\" data-image-caption=\"\" data-medium-file=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/art_gallery_1a.png?fit=300%2C71&amp;ssl=1\" data-large-file=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/art_gallery_1a.png?fit=1000%2C236&amp;ssl=1\" decoding=\"async\" loading=\"lazy\" width=\"1000\" height=\"236\" src=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/art_gallery_1a.png?resize=1000%2C236&#038;ssl=1\" alt=\"Different sampling methods to draw an art gallery 1\" class=\"wp-image-2621\" srcset=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/art_gallery_1a.png?w=1000&amp;ssl=1 1000w, https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/art_gallery_1a.png?resize=300%2C71&amp;ssl=1 300w, https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/art_gallery_1a.png?resize=768%2C181&amp;ssl=1 768w\" sizes=\"(max-width: 1000px) 100vw, 1000px\" data-recalc-dims=\"1\" \/><\/figure>\n\n\n\n<figure class=\"wp-block-image aligncenter size-full\"><img data-attachment-id=\"2622\" data-permalink=\"https:\/\/www.codeastar.com\/learn-how-to-use-stable-diffusion-part-2-sampling-methods\/art_gallery_2a\/\" data-orig-file=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/art_gallery_2a.png?fit=1000%2C229&amp;ssl=1\" data-orig-size=\"1000,229\" data-comments-opened=\"1\" data-image-meta=\"{&quot;aperture&quot;:&quot;0&quot;,&quot;credit&quot;:&quot;&quot;,&quot;camera&quot;:&quot;&quot;,&quot;caption&quot;:&quot;&quot;,&quot;created_timestamp&quot;:&quot;0&quot;,&quot;copyright&quot;:&quot;&quot;,&quot;focal_length&quot;:&quot;0&quot;,&quot;iso&quot;:&quot;0&quot;,&quot;shutter_speed&quot;:&quot;0&quot;,&quot;title&quot;:&quot;&quot;,&quot;orientation&quot;:&quot;0&quot;}\" data-image-title=\"art_gallery_2a\" data-image-description=\"\" data-image-caption=\"\" data-medium-file=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/art_gallery_2a.png?fit=300%2C69&amp;ssl=1\" data-large-file=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/art_gallery_2a.png?fit=1000%2C229&amp;ssl=1\" decoding=\"async\" loading=\"lazy\" width=\"1000\" height=\"229\" src=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/art_gallery_2a.png?resize=1000%2C229&#038;ssl=1\" alt=\"Different sampling methods to draw an art gallery 2\" class=\"wp-image-2622\" srcset=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/art_gallery_2a.png?w=1000&amp;ssl=1 1000w, https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/art_gallery_2a.png?resize=300%2C69&amp;ssl=1 300w, https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/art_gallery_2a.png?resize=768%2C176&amp;ssl=1 768w\" sizes=\"(max-width: 1000px) 100vw, 1000px\" data-recalc-dims=\"1\" \/><\/figure>\n\n\n\n<figure class=\"wp-block-image aligncenter size-full\"><img data-attachment-id=\"2623\" data-permalink=\"https:\/\/www.codeastar.com\/learn-how-to-use-stable-diffusion-part-2-sampling-methods\/art_gallery_3a\/\" data-orig-file=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/art_gallery_3a.png?fit=1000%2C231&amp;ssl=1\" data-orig-size=\"1000,231\" data-comments-opened=\"1\" data-image-meta=\"{&quot;aperture&quot;:&quot;0&quot;,&quot;credit&quot;:&quot;&quot;,&quot;camera&quot;:&quot;&quot;,&quot;caption&quot;:&quot;&quot;,&quot;created_timestamp&quot;:&quot;0&quot;,&quot;copyright&quot;:&quot;&quot;,&quot;focal_length&quot;:&quot;0&quot;,&quot;iso&quot;:&quot;0&quot;,&quot;shutter_speed&quot;:&quot;0&quot;,&quot;title&quot;:&quot;&quot;,&quot;orientation&quot;:&quot;0&quot;}\" data-image-title=\"art_gallery_3a\" data-image-description=\"\" data-image-caption=\"\" data-medium-file=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/art_gallery_3a.png?fit=300%2C69&amp;ssl=1\" data-large-file=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/art_gallery_3a.png?fit=1000%2C231&amp;ssl=1\" decoding=\"async\" loading=\"lazy\" width=\"1000\" height=\"231\" src=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/art_gallery_3a.png?resize=1000%2C231&#038;ssl=1\" alt=\"Different sampling methods to draw an art gallery 3\" class=\"wp-image-2623\" srcset=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/art_gallery_3a.png?w=1000&amp;ssl=1 1000w, https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/art_gallery_3a.png?resize=300%2C69&amp;ssl=1 300w, https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/art_gallery_3a.png?resize=768%2C177&amp;ssl=1 768w\" sizes=\"(max-width: 1000px) 100vw, 1000px\" data-recalc-dims=\"1\" \/><\/figure>\n\n\n\n<figure class=\"wp-block-image aligncenter size-full\"><img data-attachment-id=\"2624\" data-permalink=\"https:\/\/www.codeastar.com\/learn-how-to-use-stable-diffusion-part-2-sampling-methods\/art_gallery_4a\/\" data-orig-file=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/art_gallery_4a.png?fit=1000%2C231&amp;ssl=1\" data-orig-size=\"1000,231\" data-comments-opened=\"1\" data-image-meta=\"{&quot;aperture&quot;:&quot;0&quot;,&quot;credit&quot;:&quot;&quot;,&quot;camera&quot;:&quot;&quot;,&quot;caption&quot;:&quot;&quot;,&quot;created_timestamp&quot;:&quot;0&quot;,&quot;copyright&quot;:&quot;&quot;,&quot;focal_length&quot;:&quot;0&quot;,&quot;iso&quot;:&quot;0&quot;,&quot;shutter_speed&quot;:&quot;0&quot;,&quot;title&quot;:&quot;&quot;,&quot;orientation&quot;:&quot;0&quot;}\" data-image-title=\"art_gallery_4a\" data-image-description=\"\" data-image-caption=\"\" data-medium-file=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/art_gallery_4a.png?fit=300%2C69&amp;ssl=1\" data-large-file=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/art_gallery_4a.png?fit=1000%2C231&amp;ssl=1\" decoding=\"async\" loading=\"lazy\" width=\"1000\" height=\"231\" src=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/art_gallery_4a.png?resize=1000%2C231&#038;ssl=1\" alt=\"Different sampling methods to draw an art gallery 4\" class=\"wp-image-2624\" srcset=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/art_gallery_4a.png?w=1000&amp;ssl=1 1000w, https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/art_gallery_4a.png?resize=300%2C69&amp;ssl=1 300w, https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/art_gallery_4a.png?resize=768%2C177&amp;ssl=1 768w\" sizes=\"(max-width: 1000px) 100vw, 1000px\" data-recalc-dims=\"1\" \/><\/figure>\n\n\n\n<p>All samplers can generate decent outputs in general, but I would prefer a mid-century art gallery built with arches. And it looks like the steps are not enough to generate art pieces and visitors. If I have to select the best 3 from the above samplers, I should pick  <em>Heun<\/em>, <em>DPM2 Karras<\/em> and <em>PLMS<\/em>.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\">Main Course 2: Sampling Methods in Stable Diffusion (cont.)<\/h3>\n\n\n\n<p>Now, let us test another topic, here is our input:<\/p>\n\n\n\n<p>A portrait of a woman set against a dark background. The subject is positioned in a three-quarter view, facing slightly toward the viewer. The woman is portrayed from the chest up, with her upper body and face prominently displayed. She has a serene expression, characterized by a slight smile that seems to hold a sense of mystery. Her eyes are captivating, with a gaze that follows the viewer from various angles. The woman is portrayed with remarkable realism, with delicate brushwork capturing subtle details in her face and skin tone. Her brown hair is gently layered and frames her face. She is adorned in clothing typical of the era, wearing a dark-colored garment with a veil covering her hair. The painting&#8217;s composition is relatively simple, focusing primarily on the subject and her engaging presence.<\/p>\n\n\n\n<p>The origin of the above input is the textual description of the <em>Mona Lisa<\/em>. Let&#8217; see what can Stable Diffusion give us:<\/p>\n\n\n\n<figure class=\"wp-block-image aligncenter size-full\"><img data-attachment-id=\"2625\" data-permalink=\"https:\/\/www.codeastar.com\/learn-how-to-use-stable-diffusion-part-2-sampling-methods\/mona_lisa_1a\/\" data-orig-file=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/mona_lisa_1a.png?fit=1000%2C228&amp;ssl=1\" data-orig-size=\"1000,228\" data-comments-opened=\"1\" data-image-meta=\"{&quot;aperture&quot;:&quot;0&quot;,&quot;credit&quot;:&quot;&quot;,&quot;camera&quot;:&quot;&quot;,&quot;caption&quot;:&quot;&quot;,&quot;created_timestamp&quot;:&quot;0&quot;,&quot;copyright&quot;:&quot;&quot;,&quot;focal_length&quot;:&quot;0&quot;,&quot;iso&quot;:&quot;0&quot;,&quot;shutter_speed&quot;:&quot;0&quot;,&quot;title&quot;:&quot;&quot;,&quot;orientation&quot;:&quot;0&quot;}\" data-image-title=\"mona_lisa_1a\" data-image-description=\"\" data-image-caption=\"\" data-medium-file=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/mona_lisa_1a.png?fit=300%2C68&amp;ssl=1\" data-large-file=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/mona_lisa_1a.png?fit=1000%2C228&amp;ssl=1\" decoding=\"async\" loading=\"lazy\" width=\"1000\" height=\"228\" src=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/mona_lisa_1a.png?resize=1000%2C228&#038;ssl=1\" alt=\"Different sampling methods to draw The Mona Liza 1\" class=\"wp-image-2625\" srcset=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/mona_lisa_1a.png?w=1000&amp;ssl=1 1000w, https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/mona_lisa_1a.png?resize=300%2C68&amp;ssl=1 300w, https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/mona_lisa_1a.png?resize=768%2C175&amp;ssl=1 768w\" sizes=\"(max-width: 1000px) 100vw, 1000px\" data-recalc-dims=\"1\" \/><\/figure>\n\n\n\n<figure class=\"wp-block-image aligncenter size-full\"><img data-attachment-id=\"2627\" data-permalink=\"https:\/\/www.codeastar.com\/learn-how-to-use-stable-diffusion-part-2-sampling-methods\/mona_lisa_2a\/\" data-orig-file=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/mona_lisa_2a.png?fit=1000%2C231&amp;ssl=1\" data-orig-size=\"1000,231\" data-comments-opened=\"1\" data-image-meta=\"{&quot;aperture&quot;:&quot;0&quot;,&quot;credit&quot;:&quot;&quot;,&quot;camera&quot;:&quot;&quot;,&quot;caption&quot;:&quot;&quot;,&quot;created_timestamp&quot;:&quot;0&quot;,&quot;copyright&quot;:&quot;&quot;,&quot;focal_length&quot;:&quot;0&quot;,&quot;iso&quot;:&quot;0&quot;,&quot;shutter_speed&quot;:&quot;0&quot;,&quot;title&quot;:&quot;&quot;,&quot;orientation&quot;:&quot;0&quot;}\" data-image-title=\"mona_lisa_2a\" data-image-description=\"\" data-image-caption=\"\" data-medium-file=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/mona_lisa_2a.png?fit=300%2C69&amp;ssl=1\" data-large-file=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/mona_lisa_2a.png?fit=1000%2C231&amp;ssl=1\" decoding=\"async\" loading=\"lazy\" width=\"1000\" height=\"231\" src=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/mona_lisa_2a.png?resize=1000%2C231&#038;ssl=1\" alt=\"Different sampling methods to draw The Mona Liza 2\" class=\"wp-image-2627\" srcset=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/mona_lisa_2a.png?w=1000&amp;ssl=1 1000w, https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/mona_lisa_2a.png?resize=300%2C69&amp;ssl=1 300w, https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/mona_lisa_2a.png?resize=768%2C177&amp;ssl=1 768w\" sizes=\"(max-width: 1000px) 100vw, 1000px\" data-recalc-dims=\"1\" \/><\/figure>\n\n\n\n<figure class=\"wp-block-image aligncenter size-full\"><img data-attachment-id=\"2628\" data-permalink=\"https:\/\/www.codeastar.com\/learn-how-to-use-stable-diffusion-part-2-sampling-methods\/mona_lisa_3a\/\" data-orig-file=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/mona_lisa_3a.png?fit=1000%2C231&amp;ssl=1\" data-orig-size=\"1000,231\" data-comments-opened=\"1\" data-image-meta=\"{&quot;aperture&quot;:&quot;0&quot;,&quot;credit&quot;:&quot;&quot;,&quot;camera&quot;:&quot;&quot;,&quot;caption&quot;:&quot;&quot;,&quot;created_timestamp&quot;:&quot;0&quot;,&quot;copyright&quot;:&quot;&quot;,&quot;focal_length&quot;:&quot;0&quot;,&quot;iso&quot;:&quot;0&quot;,&quot;shutter_speed&quot;:&quot;0&quot;,&quot;title&quot;:&quot;&quot;,&quot;orientation&quot;:&quot;0&quot;}\" data-image-title=\"mona_lisa_3a\" data-image-description=\"\" data-image-caption=\"\" data-medium-file=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/mona_lisa_3a.png?fit=300%2C69&amp;ssl=1\" data-large-file=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/mona_lisa_3a.png?fit=1000%2C231&amp;ssl=1\" decoding=\"async\" loading=\"lazy\" width=\"1000\" height=\"231\" src=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/mona_lisa_3a.png?resize=1000%2C231&#038;ssl=1\" alt=\"Different sampling methods to draw The Mona Liza 3\" class=\"wp-image-2628\" srcset=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/mona_lisa_3a.png?w=1000&amp;ssl=1 1000w, https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/mona_lisa_3a.png?resize=300%2C69&amp;ssl=1 300w, https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/mona_lisa_3a.png?resize=768%2C177&amp;ssl=1 768w\" sizes=\"(max-width: 1000px) 100vw, 1000px\" data-recalc-dims=\"1\" \/><\/figure>\n\n\n\n<figure class=\"wp-block-image aligncenter size-full\"><img data-attachment-id=\"2626\" data-permalink=\"https:\/\/www.codeastar.com\/learn-how-to-use-stable-diffusion-part-2-sampling-methods\/mona_lisa_4a\/\" data-orig-file=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/mona_lisa_4a.png?fit=1000%2C231&amp;ssl=1\" data-orig-size=\"1000,231\" data-comments-opened=\"1\" data-image-meta=\"{&quot;aperture&quot;:&quot;0&quot;,&quot;credit&quot;:&quot;&quot;,&quot;camera&quot;:&quot;&quot;,&quot;caption&quot;:&quot;&quot;,&quot;created_timestamp&quot;:&quot;0&quot;,&quot;copyright&quot;:&quot;&quot;,&quot;focal_length&quot;:&quot;0&quot;,&quot;iso&quot;:&quot;0&quot;,&quot;shutter_speed&quot;:&quot;0&quot;,&quot;title&quot;:&quot;&quot;,&quot;orientation&quot;:&quot;0&quot;}\" data-image-title=\"mona_lisa_4a\" data-image-description=\"\" data-image-caption=\"\" data-medium-file=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/mona_lisa_4a.png?fit=300%2C69&amp;ssl=1\" data-large-file=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/mona_lisa_4a.png?fit=1000%2C231&amp;ssl=1\" decoding=\"async\" loading=\"lazy\" width=\"1000\" height=\"231\" src=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/mona_lisa_4a.png?resize=1000%2C231&#038;ssl=1\" alt=\"Different sampling methods to draw The Mona Liza 4\" class=\"wp-image-2626\" srcset=\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/mona_lisa_4a.png?w=1000&amp;ssl=1 1000w, https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/mona_lisa_4a.png?resize=300%2C69&amp;ssl=1 300w, https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/06\/mona_lisa_4a.png?resize=768%2C177&amp;ssl=1 768w\" sizes=\"(max-width: 1000px) 100vw, 1000px\" data-recalc-dims=\"1\" \/><\/figure>\n\n\n\n<p>When it comes to generating a human subject other than a scene from AI, we will become more picky. Guess we are more familiar with our facial features and we can spot the differences more easily. And I think I see Mr. Bean from the above outputs&#8230; This time, my best 3 picks are  <em>DPM++ 2M<\/em>, <em>DPM2 Karras<\/em> and <em>DPM2 a Karras<\/em>.<\/p>\n\n\n\n<p>From the 2 inputs, we can see certain samplers are aligned with a similar style. Such as <em>Euler<\/em>, <em>LMS<\/em>, <em>DPM++ 2M<\/em>, <em>DPM2 Karras<\/em> and others are in one group. <em>DPM2<\/em> <em>a<\/em>, <em>DPM++2S a<\/em> and <em>DPM2 a Karras<\/em> are in another group. And <em>DPM fast<\/em> is on its group. Even if samplers are in the same group, they may have a wide range of  variance, just look at <em>LMS<\/em> and <em>PLMS<\/em>.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\">Dessert: Summary of Our Findings<\/h3>\n\n\n\n<p>Let us summarize what we have learned so far:<\/p>\n\n\n\n<ul>\n<li>types of prompts<\/li>\n\n\n\n<li>usage of models<\/li>\n\n\n\n<li>differences on samplers<\/li>\n<\/ul>\n\n\n\n<p>We may use more specific keywords for constructing prompts. And the selection of a model, it is depending on the purpose of your generative AI graphics. Rule of Thumb, go to a model download site, pick your genre then pick the highest-rated model there. Then on the sampler side, separate samplers into different groups according to their styles and generate a testing image from each group. Once you find your favorite style, try to fine tune with each sampler within the group, to get your desired result.<\/p>\n\n\n\n<p>We can take a look at the following chart to find a faster sampler in different groups. <\/p>\n\n\n\n<center><iframe loading=\"lazy\" width=\"900\" height=\"400\" frameborder=\"0\" scrolling=\"no\" src=\"\/\/plotly.com\/~Ravenhon\/3.embed\"><\/iframe><\/center>\n\n\n\n<p>Our Mona Liza prompt is longer and more complex than our art gallery prompt. But in some sampling methods, the Mona Liza prompt just uses a shorter time to generate. The longer prompt does not mean it needs a longer time to generate, but the number of objects to draw does. Our Mona Liza prompt focuses on describing a portrait. While on our art gallery prompt, it needs to generate the venue, art pieces and visitors.<\/p>\n\n\n\n<p>We hope you all can utilize the use of prompts, models and samplers to make great generative graphics. Let&#8217;s keep on drawing!<\/p>\n","protected":false},"excerpt":{"rendered":"<p>In our last post, Learn how to use Stable Diffusion &#8211; Part 1, we mentioned sampling methods and stated that this is a huge topic. Right now, let&#8217;s continue our Stable Diffusion learning journey. Besides discussing the main course, Sampling Methods, we are going to have appetizers as well, like prompt and checkpoint. We hope [&hellip;]<\/p>\n","protected":false},"author":1,"featured_media":2639,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"om_disable_all_campaigns":false,"_monsterinsights_skip_tracking":false,"_monsterinsights_sitenote_active":false,"_monsterinsights_sitenote_note":"","_monsterinsights_sitenote_category":0,"site-sidebar-layout":"default","site-content-layout":"","ast-site-content-layout":"","site-content-style":"default","site-sidebar-style":"default","ast-global-header-display":"","ast-banner-title-visibility":"","ast-main-header-display":"","ast-hfb-above-header-display":"","ast-hfb-below-header-display":"","ast-hfb-mobile-header-display":"","site-post-title":"","ast-breadcrumbs-content":"","ast-featured-img":"","footer-sml-layout":"","theme-transparent-header-meta":"","adv-header-id-meta":"","stick-header-meta":"","header-above-stick-meta":"","header-main-stick-meta":"","header-below-stick-meta":"","astra-migrate-meta-layouts":"default","ast-page-background-enabled":"default","ast-page-background-meta":{"desktop":{"background-color":"","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"tablet":{"background-color":"","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"mobile":{"background-color":"","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""}},"ast-content-background-meta":{"desktop":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"tablet":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"mobile":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""}},"jetpack_post_was_ever_published":false,"_jetpack_newsletter_access":"","_jetpack_newsletter_tier_id":0,"jetpack_publicize_message":"","jetpack_is_tweetstorm":false,"jetpack_publicize_feature_enabled":true,"jetpack_social_post_already_shared":true,"jetpack_social_options":{"image_generator_settings":{"template":"highway","enabled":false}}},"categories":[185],"tags":[187,188,186,180],"jetpack_publicize_connections":[],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v21.8.1 - https:\/\/yoast.com\/wordpress\/plugins\/seo\/ -->\n<title>Sampling Methods in Stable Diffusion &#8902; Code A Star<\/title>\n<meta name=\"description\" content=\"In this session, we will discuss the use of Sampling Methods, Prompts and Checkpoints in Stable Diffusion with examples.\" \/>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/www.codeastar.com\/learn-how-to-use-stable-diffusion-part-2-sampling-methods\/\" \/>\n<meta property=\"og:locale\" content=\"en_US\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Sampling Methods in Stable Diffusion &#8902; Code A Star\" \/>\n<meta property=\"og:description\" content=\"In this session, we will discuss the use of Sampling Methods, Prompts and Checkpoints in Stable Diffusion with examples.\" \/>\n<meta property=\"og:url\" content=\"https:\/\/www.codeastar.com\/learn-how-to-use-stable-diffusion-part-2-sampling-methods\/\" \/>\n<meta property=\"og:site_name\" content=\"Code A Star\" \/>\n<meta property=\"article:publisher\" content=\"codeastar\" \/>\n<meta property=\"article:author\" content=\"codeastar\" \/>\n<meta property=\"article:published_time\" content=\"2023-07-03T11:11:31+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2023-07-04T17:51:21+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/www.codeastar.com\/wp-content\/uploads\/2023\/07\/l_supper.png\" \/>\n\t<meta property=\"og:image:width\" content=\"1023\" \/>\n\t<meta property=\"og:image:height\" content=\"433\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/png\" \/>\n<meta name=\"author\" content=\"Raven Hon\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:creator\" content=\"@codeastar\" \/>\n<meta name=\"twitter:site\" content=\"@codeastar\" \/>\n<meta name=\"twitter:label1\" content=\"Written by\" \/>\n\t<meta name=\"twitter:data1\" content=\"Raven Hon\" \/>\n\t<meta name=\"twitter:label2\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data2\" content=\"9 minutes\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"Article\",\"@id\":\"https:\/\/www.codeastar.com\/learn-how-to-use-stable-diffusion-part-2-sampling-methods\/#article\",\"isPartOf\":{\"@id\":\"https:\/\/www.codeastar.com\/learn-how-to-use-stable-diffusion-part-2-sampling-methods\/\"},\"author\":{\"name\":\"Raven Hon\",\"@id\":\"https:\/\/www.codeastar.com\/#\/schema\/person\/832d202eb92a3d430097e88c6d0550bd\"},\"headline\":\"Learn how to use Stable Diffusion Part 2 &#8211; Sampling Methods\",\"datePublished\":\"2023-07-03T11:11:31+00:00\",\"dateModified\":\"2023-07-04T17:51:21+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\/\/www.codeastar.com\/learn-how-to-use-stable-diffusion-part-2-sampling-methods\/\"},\"wordCount\":1443,\"commentCount\":1,\"publisher\":{\"@id\":\"https:\/\/www.codeastar.com\/#\/schema\/person\/832d202eb92a3d430097e88c6d0550bd\"},\"keywords\":[\"checkpoint\",\"CivitAI\",\"Sampling\",\"Stable Diffusion\"],\"articleSection\":[\"Stable Diffusion Artist\"],\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"CommentAction\",\"name\":\"Comment\",\"target\":[\"https:\/\/www.codeastar.com\/learn-how-to-use-stable-diffusion-part-2-sampling-methods\/#respond\"]}]},{\"@type\":\"WebPage\",\"@id\":\"https:\/\/www.codeastar.com\/learn-how-to-use-stable-diffusion-part-2-sampling-methods\/\",\"url\":\"https:\/\/www.codeastar.com\/learn-how-to-use-stable-diffusion-part-2-sampling-methods\/\",\"name\":\"Sampling Methods in Stable Diffusion &#8902; Code A Star\",\"isPartOf\":{\"@id\":\"https:\/\/www.codeastar.com\/#website\"},\"datePublished\":\"2023-07-03T11:11:31+00:00\",\"dateModified\":\"2023-07-04T17:51:21+00:00\",\"description\":\"In this session, we will discuss the use of Sampling Methods, Prompts and Checkpoints in Stable Diffusion with examples.\",\"breadcrumb\":{\"@id\":\"https:\/\/www.codeastar.com\/learn-how-to-use-stable-diffusion-part-2-sampling-methods\/#breadcrumb\"},\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/www.codeastar.com\/learn-how-to-use-stable-diffusion-part-2-sampling-methods\/\"]}]},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\/\/www.codeastar.com\/learn-how-to-use-stable-diffusion-part-2-sampling-methods\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\/\/www.codeastar.com\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"Learn how to use Stable Diffusion Part 2 &#8211; Sampling Methods\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\/\/www.codeastar.com\/#website\",\"url\":\"https:\/\/www.codeastar.com\/\",\"name\":\"Code A Star\",\"description\":\"We don&#039;t wish upon a star, we code a star\",\"publisher\":{\"@id\":\"https:\/\/www.codeastar.com\/#\/schema\/person\/832d202eb92a3d430097e88c6d0550bd\"},\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\/\/www.codeastar.com\/?s={search_term_string}\"},\"query-input\":\"required name=search_term_string\"}],\"inLanguage\":\"en-US\"},{\"@type\":[\"Person\",\"Organization\"],\"@id\":\"https:\/\/www.codeastar.com\/#\/schema\/person\/832d202eb92a3d430097e88c6d0550bd\",\"name\":\"Raven Hon\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\/\/www.codeastar.com\/#\/schema\/person\/image\/\",\"url\":\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2018\/08\/logo70.png?fit=70%2C70&ssl=1\",\"contentUrl\":\"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2018\/08\/logo70.png?fit=70%2C70&ssl=1\",\"width\":70,\"height\":70,\"caption\":\"Raven Hon\"},\"logo\":{\"@id\":\"https:\/\/www.codeastar.com\/#\/schema\/person\/image\/\"},\"description\":\"Raven Hon is\u00a0a 20 years+ veteran in information technology industry who has worked on various projects from console, web, game, banking and mobile applications in different sized companies.\",\"sameAs\":[\"https:\/\/www.codeastar.com\",\"codeastar\",\"https:\/\/twitter.com\/codeastar\"]}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"Sampling Methods in Stable Diffusion &#8902; Code A Star","description":"In this session, we will discuss the use of Sampling Methods, Prompts and Checkpoints in Stable Diffusion with examples.","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/www.codeastar.com\/learn-how-to-use-stable-diffusion-part-2-sampling-methods\/","og_locale":"en_US","og_type":"article","og_title":"Sampling Methods in Stable Diffusion &#8902; Code A Star","og_description":"In this session, we will discuss the use of Sampling Methods, Prompts and Checkpoints in Stable Diffusion with examples.","og_url":"https:\/\/www.codeastar.com\/learn-how-to-use-stable-diffusion-part-2-sampling-methods\/","og_site_name":"Code A Star","article_publisher":"codeastar","article_author":"codeastar","article_published_time":"2023-07-03T11:11:31+00:00","article_modified_time":"2023-07-04T17:51:21+00:00","og_image":[{"width":1023,"height":433,"url":"https:\/\/www.codeastar.com\/wp-content\/uploads\/2023\/07\/l_supper.png","type":"image\/png"}],"author":"Raven Hon","twitter_card":"summary_large_image","twitter_creator":"@codeastar","twitter_site":"@codeastar","twitter_misc":{"Written by":"Raven Hon","Est. reading time":"9 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"Article","@id":"https:\/\/www.codeastar.com\/learn-how-to-use-stable-diffusion-part-2-sampling-methods\/#article","isPartOf":{"@id":"https:\/\/www.codeastar.com\/learn-how-to-use-stable-diffusion-part-2-sampling-methods\/"},"author":{"name":"Raven Hon","@id":"https:\/\/www.codeastar.com\/#\/schema\/person\/832d202eb92a3d430097e88c6d0550bd"},"headline":"Learn how to use Stable Diffusion Part 2 &#8211; Sampling Methods","datePublished":"2023-07-03T11:11:31+00:00","dateModified":"2023-07-04T17:51:21+00:00","mainEntityOfPage":{"@id":"https:\/\/www.codeastar.com\/learn-how-to-use-stable-diffusion-part-2-sampling-methods\/"},"wordCount":1443,"commentCount":1,"publisher":{"@id":"https:\/\/www.codeastar.com\/#\/schema\/person\/832d202eb92a3d430097e88c6d0550bd"},"keywords":["checkpoint","CivitAI","Sampling","Stable Diffusion"],"articleSection":["Stable Diffusion Artist"],"inLanguage":"en-US","potentialAction":[{"@type":"CommentAction","name":"Comment","target":["https:\/\/www.codeastar.com\/learn-how-to-use-stable-diffusion-part-2-sampling-methods\/#respond"]}]},{"@type":"WebPage","@id":"https:\/\/www.codeastar.com\/learn-how-to-use-stable-diffusion-part-2-sampling-methods\/","url":"https:\/\/www.codeastar.com\/learn-how-to-use-stable-diffusion-part-2-sampling-methods\/","name":"Sampling Methods in Stable Diffusion &#8902; Code A Star","isPartOf":{"@id":"https:\/\/www.codeastar.com\/#website"},"datePublished":"2023-07-03T11:11:31+00:00","dateModified":"2023-07-04T17:51:21+00:00","description":"In this session, we will discuss the use of Sampling Methods, Prompts and Checkpoints in Stable Diffusion with examples.","breadcrumb":{"@id":"https:\/\/www.codeastar.com\/learn-how-to-use-stable-diffusion-part-2-sampling-methods\/#breadcrumb"},"inLanguage":"en-US","potentialAction":[{"@type":"ReadAction","target":["https:\/\/www.codeastar.com\/learn-how-to-use-stable-diffusion-part-2-sampling-methods\/"]}]},{"@type":"BreadcrumbList","@id":"https:\/\/www.codeastar.com\/learn-how-to-use-stable-diffusion-part-2-sampling-methods\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/www.codeastar.com\/"},{"@type":"ListItem","position":2,"name":"Learn how to use Stable Diffusion Part 2 &#8211; Sampling Methods"}]},{"@type":"WebSite","@id":"https:\/\/www.codeastar.com\/#website","url":"https:\/\/www.codeastar.com\/","name":"Code A Star","description":"We don&#039;t wish upon a star, we code a star","publisher":{"@id":"https:\/\/www.codeastar.com\/#\/schema\/person\/832d202eb92a3d430097e88c6d0550bd"},"potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/www.codeastar.com\/?s={search_term_string}"},"query-input":"required name=search_term_string"}],"inLanguage":"en-US"},{"@type":["Person","Organization"],"@id":"https:\/\/www.codeastar.com\/#\/schema\/person\/832d202eb92a3d430097e88c6d0550bd","name":"Raven Hon","image":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/www.codeastar.com\/#\/schema\/person\/image\/","url":"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2018\/08\/logo70.png?fit=70%2C70&ssl=1","contentUrl":"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2018\/08\/logo70.png?fit=70%2C70&ssl=1","width":70,"height":70,"caption":"Raven Hon"},"logo":{"@id":"https:\/\/www.codeastar.com\/#\/schema\/person\/image\/"},"description":"Raven Hon is\u00a0a 20 years+ veteran in information technology industry who has worked on various projects from console, web, game, banking and mobile applications in different sized companies.","sameAs":["https:\/\/www.codeastar.com","codeastar","https:\/\/twitter.com\/codeastar"]}]}},"jetpack_featured_media_url":"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/07\/l_supper.png?fit=1023%2C433&ssl=1","jetpack_sharing_enabled":true,"jetpack_shortlink":"https:\/\/wp.me\/p8PcRO-F7","jetpack-related-posts":[{"id":2529,"url":"https:\/\/www.codeastar.com\/stable-diffusion-quick-and-easy-guide-for-everyone-part-1\/","url_meta":{"origin":2549,"position":0},"title":"Easy Guide for Beginner: Learn how to use Stable Diffusion WebUI &#8211; Part 1","author":"Raven Hon","date":"May 29, 2023","format":false,"excerpt":"We have been learning programming and AI since the beginning of this website. Now, we are excited to explore the world of generative AI art using the Stable Diffusion WebUI. This innovative web interface provides us with a simple and efficient way to generate AI art. In one of our\u2026","rel":"","context":"In &quot;Learn Machine Learning&quot;","block_context":{"text":"Learn Machine Learning","link":"https:\/\/www.codeastar.com\/category\/machine-learning\/"},"img":{"alt_text":"Stable Diffusion","src":"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/05\/Stable_Diffusion.png?fit=1024%2C512&ssl=1&resize=350%2C200","width":350,"height":200,"srcset":"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/05\/Stable_Diffusion.png?fit=1024%2C512&ssl=1&resize=350%2C200 1x, https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/05\/Stable_Diffusion.png?fit=1024%2C512&ssl=1&resize=525%2C300 1.5x, https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2023\/05\/Stable_Diffusion.png?fit=1024%2C512&ssl=1&resize=700%2C400 2x"},"classes":[]},{"id":1336,"url":"https:\/\/www.codeastar.com\/u-net-object-detection-iou\/","url_meta":{"origin":2549,"position":1},"title":"U-Net and IoU for Object Detection in Image Processing","author":"Raven Hon","date":"September 26, 2018","format":false,"excerpt":"Other than our last\u00a0\u00a0hand writing challenge, there is another\u00a0Kaggle\u00a0challenge featuring image recognition --\u00a0TGS Salt Identification Challenge. But this time, we are going for an \"upgrade\". As we are dealing with object detection. It is all about salt. In this challenge, our mission is finding geophysical\u00a0images that contain salt. Oh wait,\u2026","rel":"","context":"In &quot;Learn Machine Learning&quot;","block_context":{"text":"Learn Machine Learning","link":"https:\/\/www.codeastar.com\/category\/machine-learning\/"},"img":{"alt_text":"Object Detection","src":"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2018\/09\/unet-1.png?fit=1004%2C661&ssl=1&resize=350%2C200","width":350,"height":200,"srcset":"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2018\/09\/unet-1.png?fit=1004%2C661&ssl=1&resize=350%2C200 1x, https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2018\/09\/unet-1.png?fit=1004%2C661&ssl=1&resize=525%2C300 1.5x, https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2018\/09\/unet-1.png?fit=1004%2C661&ssl=1&resize=700%2C400 2x"},"classes":[]},{"id":46,"url":"https:\/\/www.codeastar.com\/become-a-coder-in-10-seconds\/","url_meta":{"origin":2549,"position":2},"title":"[Hands-on] Become a coder in 10 seconds!","author":"Raven Hon","date":"June 1, 2017","format":false,"excerpt":"Yes, it is a catchy title, but it still sounds legit. Last time, we talked the talk on the code we pick to code\u00a0--- the Python language.\u00a0Now, let's walk the walk and make no joke --- the hands-on workshop. First things first, install the Python into your machine. No matter\u2026","rel":"","context":"In &quot;We code therefore we are&quot;","block_context":{"text":"We code therefore we are","link":"https:\/\/www.codeastar.com\/category\/we-code-therefore-we-are\/"},"img":{"alt_text":"","src":"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2017\/06\/10s_coder.png?resize=350%2C200&ssl=1","width":350,"height":200,"srcset":"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2017\/06\/10s_coder.png?resize=350%2C200&ssl=1 1x, https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2017\/06\/10s_coder.png?resize=525%2C300&ssl=1 1.5x, https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2017\/06\/10s_coder.png?resize=700%2C400&ssl=1 2x"},"classes":[]},{"id":1769,"url":"https:\/\/www.codeastar.com\/react-frontend-weather-forecast-2\/","url_meta":{"origin":2549,"position":3},"title":"React Frontend and Flask Backend for Weather Forecast \u2013 Part 2","author":"Raven Hon","date":"March 4, 2019","format":false,"excerpt":"From our last post, \"Flask Backend and React Frontend for Weather Forecast \u2013 Part 1\", we built our magic user, RESTful API as the backend module with Flask. This time, we are going to build our sword swinger, who always deals with challenges at the front line. Before we go\u2026","rel":"","context":"In &quot;We code therefore we are&quot;","block_context":{"text":"We code therefore we are","link":"https:\/\/www.codeastar.com\/category\/we-code-therefore-we-are\/"},"img":{"alt_text":"React Frontend","src":"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2019\/03\/wq_happy.png?fit=1000%2C500&ssl=1&resize=350%2C200","width":350,"height":200,"srcset":"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2019\/03\/wq_happy.png?fit=1000%2C500&ssl=1&resize=350%2C200 1x, https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2019\/03\/wq_happy.png?fit=1000%2C500&ssl=1&resize=525%2C300 1.5x, https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2019\/03\/wq_happy.png?fit=1000%2C500&ssl=1&resize=700%2C400 2x"},"classes":[]},{"id":930,"url":"https:\/\/www.codeastar.com\/install-docker-raspberry-pi\/","url_meta":{"origin":2549,"position":4},"title":"How to install Docker on Raspberry Pi","author":"Raven Hon","date":"April 3, 2018","format":false,"excerpt":"In this article, we will talk about Docker and Raspberry Pi.\u00a0Docker is a platform you may probably hear it often recently. It is hot because it helps people simplify the DevOps process. It lets developers build light-weight applications. Then they can ship and run their applications \"anywhere\" (we will discuss\u2026","rel":"","context":"In &quot;Have a pie, Raspberry Pi(e)&quot;","block_context":{"text":"Have a pie, Raspberry Pi(e)","link":"https:\/\/www.codeastar.com\/category\/have-a-raspberry-pie\/"},"img":{"alt_text":"Docker and Raspberry Pi","src":"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2018\/04\/raspberry.png?fit=1000%2C542&ssl=1&resize=350%2C200","width":350,"height":200,"srcset":"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2018\/04\/raspberry.png?fit=1000%2C542&ssl=1&resize=350%2C200 1x, https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2018\/04\/raspberry.png?fit=1000%2C542&ssl=1&resize=525%2C300 1.5x, https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2018\/04\/raspberry.png?fit=1000%2C542&ssl=1&resize=700%2C400 2x"},"classes":[]},{"id":1941,"url":"https:\/\/www.codeastar.com\/recurrent-neural-network-rnn-in-nlp-and-python-part-2\/","url_meta":{"origin":2549,"position":5},"title":"RNN (Recurrent Neural Network) in NLP and Python &#8211; Part 2","author":"Raven Hon","date":"May 15, 2019","format":false,"excerpt":"From our Part 1 of NLP and Python topic, we talked about word pre-processing for a machine to handle words. This time, we are going to talk about building a model for a machine to classify words. We learned to use CNN to classify images in past. Then we use\u2026","rel":"","context":"In &quot;Learn Machine Learning&quot;","block_context":{"text":"Learn Machine Learning","link":"https:\/\/www.codeastar.com\/category\/machine-learning\/"},"img":{"alt_text":"Recurrent Neural Network","src":"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2019\/05\/rnn.png?fit=1000%2C723&ssl=1&resize=350%2C200","width":350,"height":200,"srcset":"https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2019\/05\/rnn.png?fit=1000%2C723&ssl=1&resize=350%2C200 1x, https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2019\/05\/rnn.png?fit=1000%2C723&ssl=1&resize=525%2C300 1.5x, https:\/\/i0.wp.com\/www.codeastar.com\/wp-content\/uploads\/2019\/05\/rnn.png?fit=1000%2C723&ssl=1&resize=700%2C400 2x"},"classes":[]}],"_links":{"self":[{"href":"https:\/\/www.codeastar.com\/wp-json\/wp\/v2\/posts\/2549"}],"collection":[{"href":"https:\/\/www.codeastar.com\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.codeastar.com\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.codeastar.com\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.codeastar.com\/wp-json\/wp\/v2\/comments?post=2549"}],"version-history":[{"count":39,"href":"https:\/\/www.codeastar.com\/wp-json\/wp\/v2\/posts\/2549\/revisions"}],"predecessor-version":[{"id":2640,"href":"https:\/\/www.codeastar.com\/wp-json\/wp\/v2\/posts\/2549\/revisions\/2640"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.codeastar.com\/wp-json\/wp\/v2\/media\/2639"}],"wp:attachment":[{"href":"https:\/\/www.codeastar.com\/wp-json\/wp\/v2\/media?parent=2549"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.codeastar.com\/wp-json\/wp\/v2\/categories?post=2549"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.codeastar.com\/wp-json\/wp\/v2\/tags?post=2549"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}