🖥️Welcome to try out the open-source <a target="_blank" rel="ugc" href="https://github.com/jiayev/GPT4V-Image-Captioner">GPT4V-Image-Captioner</a>, developed by my friend and me. It offers a one-click installation and comes integrated with multiple features including image pre-compression, image tagging, and tag statistics. Recently, we also launched the <a target="_blank" rel="ugc" href="https://github.com/SleeeepyZhou/sd-webui-GPT4V-Image-Captioner">webui plugin version</a> of this tool, everyone is welcome to use it!🌍欢迎加入QQ群"兔狲·AIGC梦工北厂"，群号 ：780132897 ；"兔狲·AIGC梦工南厂"，群号 ：835297318（入群答案：兔狲）。Telegram群聊“兔狲的SDXL百老汇”，链接：<a target="_blank" rel="ugc" href="https://t.me/+KkflmfLTAdwzMzI1">https://t.me/+KkflmfLTAdwzMzI1</a>This is a LoRA that can generate versatile humanoid battle weapons in an EVA-like style.The tags EVA 00, EVA 01, EVA 02, and EVA 08 are associated with Unit-00, Unit-01, Unit-02, and Unit-08, respectively. Unit-01 and Unit-02 have better results due to the larger number of training images. Unit-00 comes next, while the generation quality of Unit-08 awaits further improvement after expanding the image dataset.In addition, I have labeled all training set images with the tags "evagod" and "evangelion mecha." When only invoking these two tags without specifying EVA 00, EVA 01, EVA 02, or EVA 08, the model can generate humanoid weapons with a blended style.Please note that due to the diverse sources of training images within the model, it may not accurately reproduce the specific details of the original mecha designs, but it does exhibit a distinctive EVA style.I trained this Lora using NAI as the base model. Everyone is welcome to share and showcase the artistic effects generated on various large-scale models!这是一个可以生成类EVA风格泛用人型决战兵器的LoRA。EVA 00、EVA 01、EVA 02、EVA 08这几个tag分别关联的是零号机、初号机、二号机以及八号机。初号机和二号机因为训练图更多，所以效果更好。其次是零号机，八号机的生成效果有待图像扩充后进一步改进。另外我给所有训练集图片都打上了evagod以及evangelion mecha这两个tag。当只调用这两个tag，而不加 EVA 00、EVA 01、EVA 02、EVA 08这几个标签时，可以生成融合风格的人形兵器。请注意，因为模型中训练照片来源多样，因此并不能准确复现原作中机体的细节,但在风格上具有显著的EVA风。本Lora我是使用NAI为底模训练的。欢迎大家多多晒图分享在不同大模型上生成的画风效果!

LEOSAM's EVA『新世紀エヴァンゲリオン』新世纪福音战士 Neon Genesis EVANGELION LoRA

这个版本对训练集进行了扩充，并更仔细地调整了训练图的打标

EVAII.safetensors

V2.0

v3.0

v1.0

Leosam's EVA Neon Genesis Evangelion LoRa

This version expands the training set and adjusts the marking of training maps more carefully

Files with SHA256: fcdf7af9dbacbdd7952cbdfaba3c494a2240a6ed8315cbd68f01db0bff607e71