
This LoRA creates a realistic deepfake version of Youtuber and marketing influencer John Crestani.
Combine this LoRA with any face swap controlNET such as REACTOR. In the A1111 img2img section, choose the image you want to use. I recommend selecting the face with the brush in the inpaint section. Once you've selected the face and are using the LoRA as a prompt network, choose a picture of your face and place it in REACTOR, hit generate, and you will have the image with your face.
If you want to generate images in txt2img, you must generate the images with SD1.5 at 512x512 pixels. After generating them at 512x512, go to the img2img section and use the image as a reference but this time generate at 1024x1024 (using one of the LoRAs as the prompt network), finally add REACTOR (face swap).
Trigger word: john crestani
###
This was commissioned by John Crestani.
描述:
Trained on both public and private images of John Crestani.
训练词语: john crestani
名称: john crestani.safetensors
大小 (KB): 147570
类型: Model
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success