本模型基于vam建模以及相关3D图像进行训练，旨在还原3D建模风格的NSFW图像，
本lora训练目前尚未达到最佳状态，我会在之后推进lora的训练，最佳权重1.5-2.75之间，请酌情调整，使用本lora的时候最好使用2.5D系列的illustrious模型，与具有特定画风的模型搭配使用可能产生不好的画风。
This model is trained based on VAM modeling and related 3D images, aiming to reproduce NSFW images with a 3D modeling style.
The LoRa training is not yet optimal. I will continue to improve LoRa training. The optimal weights are between 1.5 and 2.75. Please adjust accordingly. When using this LoRa model, it is best to use the 2.5D series Illustrious model. Using it with models with specific art styles may produce undesirable art styles.
Description
本模型基于vam建模以及相关3D图像进行训练，旨在还原3D建模风格的NSFW图像，
本lora训练目前尚未达到最佳状态，我会在之后推进lora的训练，最佳权重1.5-2.75之间，请酌情调整，使用本lora的时候最好使用2.5D系列的illustrious模型，与具有特定画风的模型搭配使用可能产生不好的画风，请使用触发词eatyet。
This model is trained based on VAM modeling and related 3D images, aiming to reproduce NSFW images with a 3D modeling style.
The LoRa training is not yet optimal. I will continue to improve LoRa training. The optimal weights are between 1.5 and 2.75. Please adjust accordingly. When using this LoRa model, it is best to use the 2.5D series Illustrious model. Using it with models with specific art styles may produce undesirable art styles, please use "eatyet" as prompt.
Comments (4)
Can you make an anima version?
I will begin learning LoRa training with the Anima version, and will update the Anima version after the Illustrious version is updated.
好，这里一直缺vam风格的lora，能做个这个作者的风格吗
我会在收集足够高质量的相关作品之后尝试训练