<span style="background-color:rgb(26, 27, 30);color:rgb(76, 224, 104);font-family:-apple-system, BlinkMacSystemFont, &quot;Segoe UI&quot;, Roboto, Helvetica, Arial, sans-serif, &quot;Apple Color Emoji&quot;, &quot;Segoe UI Emoji&quot;;font-size:16px">👍 Please Rate! - Your feedback, rating or a follow are greatly appreciated. - Thankyou! <span style="background-color:rgb(26, 27, 30);color:rgb(81, 237, 96);font-family:-apple-system, BlinkMacSystemFont, &quot;Segoe UI&quot;, Roboto, Helvetica, Arial, sans-serif, &quot;Apple Color Emoji&quot;, &quot;Segoe UI Emoji&quot;;font-size:16px">😊These models can use both Illustrious and XL Loras, though you may need to adjust the weights a bit.🗈 Attention: I created a fork of this model for full realism called Frisky Dingo. You can find it <a target="_blank" rel="ugc" href="https://civarchive.com/models/2408538/frisky-dingo-nsfw-il-xl-hybrid">here</a>. (I'll be updating it shortly).All v4 example images showcase the base model only, rendered in a single pass at high resolutions, with No detailers, LORAs, embeddings or upscaling, with only 2 - 3 exceptions. I'm using the <a target="_blank" rel="ugc" href="https://comfyai.run/documentation/PatchModelAddDownscale">Kohya Deep Srink</a> node to maintain stability at higher resolutions.There's a simplified workflow embedded in each of my v4 gallery images. (just verify the <a target="_blank" rel="ugc" href="https://comfyai.run/documentation/PatchModelAddDownscale">Magic Node</a> settings with the 🗈 notes section at the bottom, as I may have tweaked them slightly). - Standard SDXL VAE is baked.<hr />V4 - Cadmium - (a more fantasy / semi-realism aesthetic with a focus on high detail, high contrast &amp; saturation.)Suggested settings:<ul><li>VAE: sdxl_vae (baked in)</li><li>Clip skip: 2</li><li>Samplers / Schedulers: DPMpp_2M_SDE_GPU / SGM_Uniform is recommended, but a wide selection are supported</li><li>Resolution: up to 2048 x 1536, with the Kohya Deep Shrink node, portrait or landscape. all standard 1MP resolutions work well.</li><li>CFG: 3 - 7.0 (I typically use 3.5, - 4.8)</li><li>Steps: 30 - 36</li></ul>Prompting:<ul><li>Natural language prompting &amp; <a target="_blank" rel="ugc" href="https://danbooru.donmai.us/">Danbooru</a> tags. Generally less is more, for best results try to write clear concise prompts, (look to my sample images for examples and general formatting).</li></ul>(added new models, I'll update the credits shortly).<hr />V3 - Canny Mountain - (greater focus on retention of the illustrious knowledge base, camera film and post-processing effects)Suggested settings:<ul><li>VAE: sdxl_vae (baked in)</li><li>Clip skip: 2 was used during merge, (setting 1 or 2 should yield same results)</li><li>Samplers: DPM++ 2M SDE, DPM++ 3M SDE, Euler Ancestral Schedulers: SGM Uniform, Karras, Occasionally I use others for variance. (experimentation is recommended),</li><li>Resolution: all standard 1MP resolutions work well in portrait and landscape, (depending on context) (I often use 1024x1360 and 1120x1440 in portrait and landscape) I sometimes go as high as 1344x1728,</li><li>CFG: 3.8 - 8.0 (I typically use 3.8, -5.6 for photorealism)</li><li>Steps: 32 - 38 (I use 36 most often),</li></ul>Prompting:<ul><li><a target="_blank" rel="ugc" href="https://danbooru.donmai.us/">Danbooru</a> tags, &amp; natural language prompting. Generally less is more, for best results try to write clear concise prompts, (look to my sample images for examples and general formatting).</li></ul>positive prompts - responds well to camera related tags: photorealistic, raw photo, amiture photo, depth of field, bokeh etc.<ul><li>negative prompts - I generally recommend keeping sepia in your negatives to overcome a sepia bias. (it can be helpful to add a few things like "artificial, anime, illustration, unreal", if you're pushing for greater realism).</li></ul><hr />V2 - Fully REALized - (greater photorealism while maintaining much of the illustrious knowledge base)Suggested settings:<ul><li>VAE: sdxl_vae (baked in)</li><li>Clip skip: 2 was used during merge, (setting 1 or 2 should yield same results)</li><li>Sampler: DPM++ 2M SDE - SGM Uniform, (good option for photorealism) or Euler Ancestral - SGM Uniform (experimentation is recommended),</li><li>Resolution: all standard 1MP resolutions work well in portrait and landscape, (depending on context) (I often use 1024x1360 and 1120x1440 in portrait and landscape)</li><li>CFG: 2.8 - 9.0 (I commonly use 3.8, 5 &amp; 7)</li><li>Steps: 24 - 38 (I use 36 most often, though I'm starting to use lower values with solutions like CFG rescale &amp; Zero Star).</li></ul><hr />V1 - Beyond the ValleySuggested settings:<ul><li>VAE: sdxl_vae (baked in)</li><li>Clip skip: 2 was used during merge, (setting 1 or 2 should yield same results)</li><li>Sampler: Euler Ancestral - SGM Uniform (most consistent good results), DPM++ 2M SDE - SGM Uniform, (good option for photorealism)</li><li>Resolution: all standard 1MP resolutions work well in portrait and landscape, (depending on context) (I often use 1024x1360 and 1120x1440 in portrait and landscape)</li><li>CFG: 2.8 - 8.0 (lower for more photorealism - I commonly use 2.8, 3.8, 5 &amp; 7)</li><li>Steps: 24 - 38 (I use 36 most often)</li></ul>Prompting:<ul><li>Primarily <a target="_blank" rel="ugc" href="https://danbooru.donmai.us/">Danbooru</a> tags, mixed with a bit of natural language prompting. Generally less is more, for best results try to write clear concise prompts, (look to my sample images for examples and general formatting).</li></ul>positive prompts - Hype4realistic can be added to push realism a bit further.<ul><li>negative prompts - I generally recommend starting with none and adding tags as needed. (it can be helpful to add a few things like "toon, illustration, unreal", if you're pushing for greater realism).</li></ul><hr />I started this project to in an attempt to recreated a specific aesthetic created by <a target="_blank" rel="ugc" href="https://civarchive.com/user/blinkdotleh">blinkdotleh</a> using his <a target="_blank" rel="ugc" href="https://civarchive.com/models/1417633/ultra-photo-realistic-workflow-or-refiner-upscaler-adetailer-index-elements-inpainting-controlnet-civitai-metadata-noobai-xlillustrioussdxl-comfyui">workflow</a> where 2 models split the steps during image generation. He created a series of images using <a target="_blank" rel="ugc" href="https://civarchive.com/models/507472?modelVersionId=1413921">Uncanny valley</a> for the initial steps &amp; my <a target="_blank" rel="ugc" href="https://civarchive.com/models/885893?modelVersionId=1989438">fabled Illusion</a> model as a refiner. I created a style Lora from those outputs and included it in this merge. This is an attempt to recreate that look in a singe model while preserving as much of the Illustrious knowledge base as possible.on the image generation side, this is roughly 50% Illustrious &amp; 50% XL (mostly <a target="_blank" rel="ugc" href="https://civarchive.com/models/502468?modelVersionId=991916">bigASP</a>), while the CLIP more heavily favors Illustrious at about 65%.<h3 id="notes-and-tips:">🗈 Notes &amp; Tips:</h3>(to be expanded over time),<pre><code>Preferred Sampler / Scheduler combos: in order preference: 
DPM++ 2M_SDE_GPU, / SGM_uniform, Simple, beta, 
DPM++ 2M_SDE, / SGM_uniform, Simple, beta, 
DPM++ 3M_SDE_GPU, / Simple, SGM_uniform, beta,
Euler_A, / Simple, beta, SGM_uniform, exponential,
DPM++ 3M_SDE, / Simple, SGM_uniform, beta,
heun, / Simple, beta, (Sometimes garbage, sometimes pure gold)
dpmpp_2M, / beta, (same as the last one, hit or miss depending on scenario, but when it behaves, it's a clear winner, after some trial &amp; error it becomes more predictable).
DDPM / ddim_uniform, (High accuracy with clean details, a bit of a washout - overexposed effect, but with higher CFG's &amp; / or charged terms in your prompting, this can become the best option.)
Karras, tends to overcook things and glitch eyes, I almost never use it, (I know people swear by it, it's great for many scenarios / models but generally not with mine. 
Exponential, pairs well with LCM to mitigate washout
AYS, works well, but I prefer LCM with this model.
------
Patch Model Add Downscale (Kohya Deep Shrink), 
is a default node in Comfy that greatly increases image stability by compressing the Unet model during the 1st compositional stage of image generation then re-expanding it for fine details.
Parameters of concern include: 
Block number: (3 is default but 4 can also be useful in some instances).
Downscale Factor: (I use 1.5 as a default but will go as low as 1.25 for resolutions close to standard, &amp; sometimes as high as 2 if pushing the upper limits.
End Percent: (this is the percentage of total steps, the point at witch the unet will decompress. You can generally leave this at 0.35, but I sometimes make slight adjustments if I'm trying to eliminate a pesky error.
-------</code></pre><h3 id="resources-used-creator-thanks">Resources used / Creator thanks</h3>Checkpoints:<a target="_blank" rel="ugc" href="https://civarchive.com/models/1110783/ilustmix">iLustMix</a> by <a target="_blank" rel="ugc" href="https://civarchive.com/user/GZees">GZees</a><a target="_blank" rel="ugc" href="https://civarchive.com/models/502468/bigasp">BigASP v2.0</a> by <a target="_blank" rel="ugc" href="https://civarchive.com/user/nutbutter">nutbutter</a><a target="_blank" rel="ugc" href="https://civarchive.com/models/507472?modelVersionId=1413921">Uncanny valley</a> by <a target="_blank" rel="ugc" href="https://civarchive.com/user/meden">meden</a> - (clip only)<a target="_blank" rel="ugc" href="https://civarchive.com/models/840857/itercomp">IterComp</a> by <a target="_blank" rel="ugc" href="https://civarchive.com/user/advokat">Advokat</a><a target="_blank" rel="ugc" href="https://civarchive.com/models/885893?modelVersionId=1989438">Fabled Illusion</a> by <a target="_blank" rel="ugc" href="https://civarchive.com/user/prodajie">self</a><a target="_blank" rel="ugc" href="https://civarchive.com/models/139562?modelVersionId=789646">RealVisXL V5.0</a> by <a target="_blank" rel="ugc" href="https://civarchive.com/user/SG_161222">SG_161222</a>Loras:<a target="_blank" rel="ugc" href="https://civarchive.com/models/626603/hyperrealistic-pony-or-illustrious">Hyperrealistic [Pony | Illustrious]</a> by <a target="_blank" rel="ugc" href="https://civarchive.com/user/Zoropaton">Zoropaton</a><a target="_blank" rel="ugc" href="https://civarchive.com/models/510261/spo-sdxl4k-p10eplorawebui">SPO-SDXL_4k-p_10ep_LoRA_webui</a> by <a target="_blank" rel="ugc" href="https://civarchive.com/user/rockeycoss">rockeycoss</a>custom style Lora - (not yet publicly available)(Additionally, thanks to everyone whos prompts I've pilfered for testing).

Fabled Infusion (N/SFW) 🔮

Note: keep "sepia" in Neg Prompt to correct color bias, (v3 only).v3 focuses on greater retention of the Illustrious knowledge base. During iterative testing, I rendered a handful of characters and found the point where pushing further realism was causing the loss of key details, then dialed it back a bit.There's a (situationally dependent) sepia bias. I recommended keeping sepia in your neg prompt.CFG: 3ish - 9ish, (4.8 - 5.8 sweet spot)32 - 38 steps, DPM++2M_SDE, Euler_A, DPM++3M_SDE, (SGM_Uniform, Karras).Standard SDXL VAE is recommended &amp; included.

Files with SHA256: d48301ca99007bdf21c01b62c48e09b5dad3478fe6d1422571842d629f871ded