I am admittedly kind of a noob at this, but if you enjoy my humble work, any donations are greatly appreciated: <a target="_blank" rel="ugc" href="https://ko-fi.com/insult_to_ninjary">https://ko-fi.com/insult_to_ninjary</a>Your generosity helps me stay motivated to keep improving and to keep trying new things!At a Glance:<ul><li>Style: Realistic</li><li>Works with LoRAs?: It should (I don't use a ton of realistic LoRAs, but it seemed ok!)</li><li>VAE: vae-ft-mse-840000-ema-pruned is baked (but you can override if desired)</li><li>Clip skip: 1 or 2 (experiment with both!)</li><li>Hires. fix: use non-default upscaler; denoising strength 0.5 - 0.4</li></ul>Description:This is my first true attempt at a realistic checkpoint merge. My goal for this model was to make something that produced people that were not always conventionally attractive with AI sameface. I was also hoping to get the model as stable as possible for SD1.5, so that hopefully one could get decent results more quickly, without having to gen a million times.In the end, I'd call this a partial success. It is possible to get some just normal everyday looking people, but you might need to prompt for specific features that reduce conventional attractiveness (like ugly, natural skin texture, freckles, wrinkles, etc.). I also feel if you throw in some prompts for an amateur photography style (like slight motion blur, grainy, low-resolution, digital artifacts, etc.), it can help to mask some of the glaring AI image hallmarks while still looking kind of natural for a 1980s-2000s photograph or careless camera phone snap. Without more specific prompting however, you may often get more generic appearances.And as for being a more stable model, it's not bad for SD1.5, but it's still SD1.5. So this means you'll still get janky hands/fingers, weird limbs, and strange compositions a little more often than you'd want. That's just part of livin' the SD1.5 lifestyle, lol. But I have some basic tips below!Troubleshooting:<ul><li>In general, the biggest issue you're going to come across is problems with anatomy - hands, fingers, etc. Sometimes just some general weirdness. For an SD1.5 model, I'd say it's not the most stable one I've used, but it's also not as bad as many others. If you have issues, just try your best with all the usual fixes - different prompting, use a different VAE, try different clip skip values, use adetailer, etc.</li></ul>Outro:If you make anything awesome, please consider posting them to the gallery! That goes for all of my models! I really like seeing what kind of stuff you all are able to make!If you have any questions, feel free to ask! I will try my best to help! Also, if you have any ideas for future models, please don't hesitate to share them with me!

ItN's Realistic Merge

itnsRealisticMerge_v10.safetensors

v1.0

I am admittedly kind of a noob at this, but if you enjoy my humble work, any donations are greatly appreciated: <a target="_blank" rel="noopener noreferrer nofollow" href="https://ko-fi.com/insult_to_ninjary">https://ko-fi.com/insult_to_ninjary</a>Your generosity helps me stay motivated to keep improving and to keep trying new things!At a Glance:<ul><li>Style: Realistic</li><li>Works with LoRAs?: It should (I don't use a ton of realistic LoRAs, but it seemed ok!)</li><li>VAE: vae-ft-mse-840000-ema-pruned is baked (but you can override if desired)</li><li>Clip skip: 1 or 2 (experiment with both!)</li><li>Hires. fix: use non-default upscaler; denoising strength 0.5 - 0.4</li></ul>Description:This is my first true attempt at a realistic checkpoint merge. My goal for this model was to make something that produced people that were not always conventionally attractive with AI sameface. I was also hoping to get the model as stable as possible for SD1.5, so that hopefully one could get decent results more quickly, without having to gen a million times.In the end, I'd call this a partial success. It is possible to get some just normal everyday looking people, but you might need to prompt for specific features that reduce conventional attractiveness (like ugly, natural skin texture, freckles, wrinkles, etc.). I also feel if you throw in some prompts for an ******* photography style (like slight motion blur, grainy, low-resolution, digital artifacts, etc.), it can help to mask some of the glaring AI image hallmarks while still looking kind of natural for a 1980s-2000s photograph or careless camera phone snap. Without more specific prompting however, you may often get more generic appearances.And as for being a more stable model, it's not bad for SD1.5, but it's still SD1.5. So this means you'll still get janky hands/fingers, weird limbs, and strange compositions a little more often than you'd want. That's just part of livin' the SD1.5 lifestyle, lol. But I have some basic tips below!Troubleshooting:<ul><li>In general, the biggest issue you're going to come across is problems with anatomy - hands, fingers, etc. Sometimes just some general weirdness. For an SD1.5 model, I'd say it's not the most stable one I've used, but it's also not as bad as many others. If you have issues, just try your best with all the usual fixes - different prompting, use a different VAE, try different clip skip values, use adetailer, etc.</li></ul>Outro:If you make anything awesome, please consider posting them to the gallery! That goes for all of my models! I really like seeing what kind of stuff you all are able to make!If you have any questions, feel free to ask! I will try my best to help! Also, if you have any ideas for future models, please don't hesitate to share them with me!

Files with SHA256: c9fed6b1f85fa33aa33d5c57c86612d0d6d9b413368d44d5acfd9e375821e520