I have created some articles comparing v0.01 to v0.02 so it is easier to see the improvements and changes I have made (one for PG and one for everything else):<ul><li><a target="_blank" rel="ugc" href="https://civarchive.com/articles/29602/pg-13-comparing-hyphoria-v001-vs-v002">https://civarchive.com/articles/29602/pg-comparing-hyphoria-v001-vs-v002</a></li><li><a target="_blank" rel="ugc" href="https://civarchive.com/articles/29604/xxx-comparing-hyphoria-v001-vs-v002">https://civarchive.com/articles/29604/xxx-comparing-hyphoria-v001-vs-v002</a></li></ul><hr />The long awaited new version of Hyphoria is here, after almost a year of waiting, I present <s>v1</s> v0.02!When I made the first version I was unsure of its quality or if it would even be worth using, but you all really have seemed to enjoy using it and doing things well beyond what I would have imagined with it. I'm glad everyone has loved it but that also made it hard to come up with something to improve, there were always things to improve but with merging and even training it isn't easy to make something that is a categorical improvement at least on my limited budget and available time.So after experimenting with other models (my real version on Illustrious and my Qwen Image model) I learned a ton and over the last half year I experimented and tried to find something that I liked and felt like an improvement while not being a step down in most ways. I wanted it to feel like the same model but better and that is where v0.02 came to be.It has much better contrast compared to v0.01, which was washed out especially when you did img2img or hires. The background details often were non-sense or didn't resolve to anything useful, v0.02 in a lot of ways improves on that but can still be nonsense but that is SDXL/Illu for you.The other major area of improvement was around higher resolution stability being able to reliably generate at 1.5 megapixels with little to no anatomy issues (at least compared to v0.01) and that translates to better stability when doing hires passes on images. In my testing I rarely see anatomy issues like extra body parts even after multiple hires passes (but the denoise and resolution settings you choose will affect that obviously). It should be noted that ultra wide resolutions or high resolution wide aspect ratio generations can have issues with duplication or anatomy warping (e.g. subject laying horizontally) for both versions, portrait aspect ratios are more stable and support higher resolution without issues in v0.02.<h3 id="v0.02-recommended-settings">v0.02 Recommended Settings</h3><ul><li>Steps: 20-35</li><li>CFG: 3 (2.5-6)</li><li>Sampler: Euler a OR DPM++ 2M (Should be stable on most)</li><li>Scheduler: Should be stable on most</li><li>Resolution: 1mp to 1.5mp</li><li>Prompting: Less is more, I often don't even use quality tags outside of <code>worst quality</code> in the negatives. Quality tags can lead to a glossier and less flexible result, but sometimes more attractive poses and lighting, experiment.</li></ul><h3 id="v0.01-recommended-settings">v0.01 Recommended Settings</h3><ul><li>Steps: 20-35</li><li>CFG: 3 (2.5-6)</li><li>Sampler: DPM++ 2M SDE (or DPM++ 2M on CIVITAI Gen)</li><li>Scheduler: Karras</li><li>Resolution: I recommend 1024x1024 - 1280x1280 for normal generation. Supports up to 1536x1536 without breaking, but can suffer some body extension. Any aspect ratio that equals the same number of pixels should work.</li><li>Prompting: This model listens to your prompts, like really listens, so avoid the fluff and keep your prompts focused.<ul><li>Recommended Positive: <code>masterpiece, best quality, absurdres</code></li><li>Recommended Negative: <code>worst quality, low quality</code></li><li>You can often omit pos or neg quality tags depending on what you're trying to make, grungy images for instance will want to pull out some of those quality tags. Some of the other quality tags do work if you want those types of images.</li></ul></li></ul><h3 id="license-and-use">License &amp; Use</h3>SDXL - <a target="_blank" rel="ugc" href="https://github.com/Stability-AI/generative-models/blob/main/model_licenses/LICENSE-SDXL1.0">CreativeML Open RAIL++-M</a> Illustrious - <a target="_blank" rel="ugc" href="https://freedevproject.org/faipl-1.0-sd/￼NoobAI-XL">https://freedevproject.org/faipl-1.0-sd/ </a>NoobAI-XL - <a target="_blank" rel="ugc" href="https://civarchive.com/models/license/1140829">https://civarchive.com/models/license/1140829</a>Copied the above license list from <a target="_blank" rel="ugc" href="https://civarchive.com/models/1162518/plant-milk-model-suite">Plant Milk's</a> page.

Hyphoria

Experimental merge, no real logic behind the merge outside of picking models I liked and some numbers I thought made sense for weighting them. Future versions will have structured and less subjective evaluations to determine more optimal configurations.

hyphoria_v001.safetensors

Hyphoria [Illu & NAI]

v0.01

v0.02

thanks to the help of hyphoria and plant milk created this artstyle of paint,oil style,have fun

tazercubes networth

tazercubesNetworth_v10.safetensors

v1.0

beastercube

An experimental merge of various models I like, using an algorithm that ideally takes the strengths of each.NOTE: This model is very much WIP, but results are promising by all the people graciously testing.Models Used<ul><li>IllustriousXL V2.0 Stable - Used as base merge target</li><li>Rouwei v0.7 eps</li><li>IllumiYume v3.1</li><li>Hassaku v1.3 Style A</li><li>ionsyx v3.0</li><li>Wicked Illustrious Mix v1.1</li><li>mdntIllus Syn v1</li><li>Kokio v2.0</li><li>Diving Illustrious Anime v11</li><li>Bismuth Illustrious Mix v2.0</li><li>NoobAI v1.1 eps</li><li>Unreleased Merge</li><li>Plant Milk Hemp II</li><li>Plant Milk Coconut</li></ul>The above may not make the most sense, but using the algorithm, each model's UNET and CLIP were compared and the best was chosen tensor by tensor, using the above weights as my subjective adjustment to influence how likely they would be chosen. I can't tell you exactly what part of what model is used where, my merge script just chose what it considered the best.For future versions I will be experimenting with different models, along with more granular adjustments (at the block level, or even more granular than that).License &amp; UseSDXL - <a target="_blank" rel="ugc noopener noreferrer nofollow" href="https://github.com/Stability-AI/generative-models/blob/main/model_licenses/LICENSE-SDXL1.0">CreativeML Open RAIL++-M</a>Illustrious - <a target="_blank" rel="ugc noopener noreferrer nofollow" href="https://freedevproject.org/faipl-1.0-sd/">https://freedevproject.org/faipl-1.0-sd/</a>NoobAI-XL - <a target="_blank" rel="ugc noopener noreferrer nofollow" href="https://civitai.com/models/license/1140829">https://civitai.com/models/license/1140829</a>

Hyphoria [Illustrious 2.0 & NAI]

Experimental merge, no real logic behind the merge outside of picking models I liked and some numbers I thought made sense for weighting them. Future versions will have structured and less subjective evaluations to determine more optimal configurations.

Files with SHA256: 416daecfbef167dc59ffce19bc79987ba0e8b0de8e79f79ed215c196d893479d