I don't think I need to describe what Lora does :DLora works without triggersThe prompting is quite simple, just take a look at my prompts from the sample photos. You can achieve good results with just a few prompts, but if you write in a little more detail, you will get better results.<hr />I have only tested Lora with the distilled model from flux2klein9b, with a strength of 1.0-1.50If you are using the base model, you should set the strength of the Lora to 1.0 or lower<hr />I created the example photos with a strength of 1.35 I use the ClownsharKsampler with eta 0.00, which works very well and produces good skin details. Exponential/res2s 4 steps at a resolution of 1280x1920. Smaller resolutions also work great, so just test it yourself to see what works best for you :)<hr />"My settings are not essential, but for me personally they provide the best detail and excellent realism."<hr />Here is a screenshot showing how I connected the sampler and the settings you can see.<img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/8253e08e-a3f4-4d8c-a62c-63ec264c0035/width=525/8253e08e-a3f4-4d8c-a62c-63ec264c0035.jpeg" />The image editing also works very well without changing the face. You just have to see which strength works best for you. Example prompt"Remove the shirt and she revealing her nude tits"If the results are not good, try a different strength.

Files with SHA256: 5bcdd89c96228823de2443da1e0e81d4c515cbdc9870ff47f4535384edaddf30