This Lora generates images of a fully artificially generated character of a young woman. The Lora has been created using the consistent character workflow described in my tutorial article <a target="_blank" rel="ugc" href="https://civarchive.com/articles/14863">Simple SDXL Consistent Character Generation using just Forge and the Civitai Trainer</a>. Ashley can do NSFW stuff, but she hasn't specifically been trained to do so, so don't expect a very high quality.Training data for Ashley and all SDXL version sample images have been generated with Stoiqo Afrodite, hence she will probably work best on this base model. Sample Images for FLUX version have been generated using the FLUX.1 [dev] base checkpoint.<h1 id="usage-and-prompting">Usage and Prompting</h1>FLUX: Drop the lora in (the default strength of 1.0 is fine for the FLUX base checkpoint, you will need a higher strength when using finetunes) and ask for a picture of AshleyAI. Please be aware that this LoRA works best with the original FLUX.1 D checkpoint, expect inferior results when using with finetuned checkpoints. Ashley is also fully NSFW capable.SDXL: There are two versions of this model: The "Face Only" version has only been trained on facial and upper body shots, the "Full Body" version has been trained on facial, upper and full body shots (which have been generated using the Face Only version. I recommend you use the Full Body version to generate your images and the Face version for inpainting (e.g. in conjunction with ADetailer). Prompt for "AshleyAI" with a Lora weight of around 0.7.<h1 id="training-data">Training Data</h1>The training data for both the face model and the full body model can be found here:<ul><li><a target="_blank" rel="ugc" href="https://civarchive.com/posts/17062335">Face model</a></li><li><a target="_blank" rel="ugc" href="https://civarchive.com/posts/17103858">Full Body model</a></li></ul>Note: The full body model has been trained on both the face and full body model training datasets.

Artificial Character - Ashley

Files with SHA256: 503fdfb436f8c2ddb60e0618aa5e9e102be844e2462942027b31c7ae12463af5