This LoRA is trained based on NoobAI-XL vPred 1.0.

This LoRA was trained at a resolution of 1280*1280 using 792 in-game screenshots.

Recommended parameters:

resolution:1024*1536 or 1280*1280 (approximately 1.5 megapixels)

steps:36-40

recommended weight: 1(or lower)

Trigger Words: 3d

Negative prompt: transparent background (optional)

此lora训练分辨率为1280*1280，使用了792张游戏截图

推荐分生成参数：1024*1536或1280*1280（即150万像素左右）

步数：36-40步

推荐权重：1（或更低）

触发词：3d

负面提示词:transparent background(可选）