Total score: 3/7 OK.
Basically can be restored correctly
The overall style is simple, even in the nightsky without adding much decoration, Abyss Orange has the simplest lines. But it is difficult to output 3D and life-life style photos. is a pure 2D stylized lora.
There are a small number of limb errors, and it is difficult to generate feet with the resolution of the test
Another lora with multiple outfits, straight out mostly regular clothing. The test of the specified costume was not carried out because of interest. Guess is suitable for adding quality words to get a beautiful 2D picture, but the versatility is weak.
This test is designed to indicate the availability of the role LoRa. 40 images were generated using 4 seeds and 10 models. It is evaluated from 3 dimensions and 1 subjective dimension of character setting, style and failure. Objective dimension score: 0 Not Recommended, 1 Available, 2 Recommended. Subjective dimensions: 0 Not recommended, 1 Recommended. The total score is 7.
该测试旨在指示角色 LoRa 的可用性.使用4个种子和10个模型生成了40个图像。从人物设定、风格与失败3个维度和1个主观维度进行评估。客观维度得分：0 不推荐，1 可用，2 推荐。主观维度：0 不推荐，1 推荐。总分是 7 。